The American Journal of Cardiology Copyright В© 2005 Elsevier Inc. All rights reserved Volume 95, Issue 2, Pages 161-310 (15 January 2005) 1. g c d e f Editorial board • EDITORIAL BOARD Page A3 Coronary Artery Disease 2. g c d e f 3. g c d e f 4. g c d e f 5. g c d e f Impact of sirolimus-eluting stent on the outcome of patients with chronic total occlusions • ARTICLE Pages 161-166 Sunao Nakamura, Tamil Selvan Muthusamy, Jang-Ho Bae, Yeo Hans Cahyadi, Wasan Udayachalerm and Damras Tresukosol Impact of baseline renal function on mortality after percutaneous coronary intervention with sirolimus-eluting stents or bare metal stents • ARTICLE Pages 167-172 Pedro A. Lemos, Chourmouzios A. Arampatzis, Angela Hoye, Joost Daemen, Andrew T.L. Ong, Francesco Saia, Willem J. van der Giessen, Eugene P. McFadden, Georgios Sianos, Pieter C. Smits et al. Determinants of 30-day adverse events following saphenous vein graft intervention with and without a distal occlusion embolic protection device • ARTICLE Pages 173-177 Gregory R. Giugliano, Richard E. Kuntz, Jeffrey J. Popma, Donald E. Cutlip, Donald S. Baim and Saphenous Vein Graft Angioplasty Free of Emboli Randomized (SAFER) Trial Investigators Artificial neural network modeling of stress single-photon emission computed tomographic imaging for detecting extensive coronary artery disease • ARTICLE Pages 178-181 J. Scott Allison, Jaekyeong Heo and Ami E. Iskandrian 6. g c d e f Incremental prognostic power of single-photon emission computed tomographic myocardial perfusion imaging in patients with known or suspected coronary artery disease • ARTICLE Pages 182-188 Salvador Borges-Neto, Linda K. Shaw, Robert H. Tuttle, John H. Alexander, William T. Smith IV, Marianna Chambless, R. Edward Coleman, Robert A. Harrington and Robert M. Califf Preventive Cardiology 7. g c d e f 8. g c d e f Comparative effects of rosuvastatin and gemfibrozil on glucose, insulin, and lipid metabolism in insulin-resistant, nondiabetic patients with combined dyslipidemia • ARTICLE Pages 189-193 Cindy Lamendola, Fahim Abbasi, James W. Chu, Howard Hutchinson, Valerie Cain, Elizabeth Leary, Tracey McLaughlin, Evan Stein and Gerald Reaven Relation between atherogenic dyslipidemia and the Adult Treatment Program-III definition of metabolic syndrome (Genetic Epidemiology of Metabolic Syndrome Project) • ARTICLE Pages 194-198 Diego F. Wyszynski, Dawn M. Waterworth, Philip J. Barter, Jonathan Cohen, Y. Antero KesГ¤niemi, Robert W. Mahley, Ruth McPherson, GГ©rard Waeber, Thomas P. Bersot, Sanjay S. Sharma et al. Miscellaneous 9. g c d e f New predictors of outcome in idiopathic pulmonary arterial hypertension • ARTICLE Pages 199-203 Steven M. Kawut, Evelyn M. Horn, Ketevan K. Berekashvili, Robert P. Garofano, Rochelle L. Goldsmith, Allison C. Widlitz, Erika B. Rosenzweig, Diane Kerstein and Robyn J. Barst Interview 10. g c d e f Anthony Nicholas Demaria, MD: A conversation with the editor • DISCUSSION Pages 204-223 Brief Reports 11. g c d e f 12. g c d e f 13. g c d e f 14. g c d e f 15. g c d e f 16. g c d e f Relation of impaired Thrombolysis In Myocardial Infarction myocardial perfusion grades to residual thrombus following the restoration of epicardial patency in ST-elevation myocardial infarction • SHORT COMMUNICATION Pages 224-227 Ajay J. Kirtane, Aaron Weisbord, Dimitrios Karmpaliotis, Sabina A. Murphy, Robert P. Giugliano, Christopher P. Cannon, Elliott M. Antman, E. Magnus Ohman, Matthew T. Roe, Eugene Braunwald et al. Angiography and revascularization in patients with heart failure following fibrinolytic therapy for ST-elevation acute myocardial infarction • SHORT COMMUNICATION Pages 228-233 Amir Kashani, C. Michael Gibson, Sabina A. Murphy, Marc S. Sabatine, David A. Morrow, Elliott M. Antman and Robert P. Giugliano Impact of ST-segment depression resolution on mortality after successful mechanical reperfusion in patients with ST-segment elevation acute myocardial infarction • SHORT COMMUNICATION Pages 234-236 Giuseppe De Luca, Arthur C. Maas, Arnoud W.J. van't Hof, Jan Paul Ottervanger, Jan C.A. Hoorntje, A.T. Marcel Gosselink, Jan-Henk E. Dambrink, Menko-Jan de Boer and Harry Suryapranata Effects of tirofiban and statins on high-sensitivity C-reactive protein, interleukin-6, and soluble CD40 ligand following percutaneous coronary interventions in patients with stable coronary artery disease • SHORT COMMUNICATION Pages 236-240 Rabih R. Azar, Georges Badaoui, Antoine Sarkis, Roland Kassab, Elie SalamГ©, Samira KlaymГ©, Roger Naman and Mirna Germanos Value of preprocedure multislice computed tomographic coronary angiography to predict the outcome of percutaneous recanalization of chronic total occlusions • SHORT COMMUNICATION Pages 240-243 Nico R. Mollet, Angela Hoye, Pedro A. Lemos, Filippo Cademartiri, Georgios Sianos, Eugene P. McFadden, Gabriel P. Krestin, Patrick W. Serruys and Pim J. de Feyter Angiographic and clinical outcomes of polytetrafluoroethylenecovered stent use in significant coronary perforations • SHORT COMMUNICATION Pages 244-246 Hung Ly, Jean-Pierre S. Awaida, Jacques LespГ©rance and Luc Bilodeau 17. g c d e f Role of postoperative vasopressor use in occurrence of atrial fibrillation after coronary artery bypass grafting • SHORT COMMUNICATION Pages 247-249 Vikrant Salaria, Nirav J. Mehta, Syed Abdul-Aziz, Syed M. Mohiuddin and Ijaz A. Khan 18. g c d e f 19. g c d e f Reaching recommended lipid and blood pressure targets with amlodipine/atorvastatin combination in patients with coronary heart disease • SHORT COMMUNICATION Pages 249-253 Jean-Francois Dorval, Todd Anderson, Jean Buithieu, Sammy Chan, Stuart Hutchison, Thao Huynh, Jean Jobin, Eva Lonn, Paul Poirier, Lawrence Title et al. Benefits of niacin by glycemic status in patients with healed myocardial infarction (from the Coronary Drug Project) • SHORT COMMUNICATION Pages 254-257 Paul L. Canner, Curt D. Furberg, Michael L. Terrin and Mark E. McGovern 20. g c d e f Occurrence of acute myocardial infarction in Worcester, Massachusetts, before, during, and after the terrorists attacks in New York City and Washington, DC, on 11 September 2001 • SHORT COMMUNICATION Pages 258-260 Robert J. Goldberg, Frederick Spencer, Darleen Lessard, Jorge Yarzebski, Craig Lareau and Joel M. Gore 21. g c d e f Characteristics and outcomes of patients with acute myocardial infarction and angiographically normal coronary arteries • SHORT COMMUNICATION Pages 261-263 Alf Inge Larsen, P. Diane Galbraith, William A. Ghali, Colleen M. Norris, Michelle M. Graham, Merril L. Knudtson and APPROACH Investigators 22. g c d e f Long-term safety and efficacy of high-dose atorvastatin treatment in patients with familial hypercholesterolemia • SHORT COMMUNICATION Pages 264-266 Sanne van Wissen, Tineke J. Smilde, Mieke D. Trip, Anton F.H. Stalenhoef and John J.P. Kastelein 23. g c d e f 24. g c d e f 25. g c d e f 26. g c d e f Effect of medical and surgical weight loss on endothelial vasomotor function in obese patients • SHORT COMMUNICATION Pages 266-268 Noyan Gokce, Joseph A. Vita, Marie McDonnell, Armour R. Forse, Nawfal Istfan, Maria Stoeckl, Izabella Lipinska, John F. Keaney, Jr and Caroline M. Apovian Prevalence of adequate control of increased serum low-density lipoprotein cholesterol in self-pay or Medicare patients versus Medicaid or private insurance patients followed in a University General Medicine Clinic • SHORT COMMUNICATION Pages 269-270 Raja Varma, Wilbert S. Aronow, Glenn Gandelman and Christopher Zammit Clinical variables predicting inappropriate use of implantable cardioverter-defibrillator in patients with coronary heart disease or nonischemic dilated cardiomyopathy • SHORT COMMUNICATION Pages 271-274 Dominic A.M.J. Theuns, A. Peter J. Klootwijk, Maarten L. Simoons and Luc J. Jordaens Design of the SHock Inhibition Evaluation with Azimilide (SHIELD) study: A novel method to assess antiarrhythmic drug effect in patients with an implantable cardioverter-defibrillator • SHORT COMMUNICATION Pages 274-276 Craig M. Pratt, Paul Dorian, Hussein R. Al-Khalidi, Jose M. Brum, Martin Borggrefe, Daljit S. Tatla, Johannes Brachmann, Robert J. Myerburg, David S. Cannom, Michael J. Holroyde et al. 27. g c d e f 28. g c d e f Refeeding normalizes the QT rate dependence of female anorexic patients • SHORT COMMUNICATION Pages 277-280 FrГ©dГ©ric Roche, Jean-Claude BarthГ©lГ©my, Norbert Mayaud, Vincent Pichot, David Duverney, Natacha Germain, FranГ§ois Lang and Bruno Estour Left ventricular structure and function in sedentary and physically active subjects with left ventricular hypertrophy (the LIFE Study) • SHORT COMMUNICATION Pages 280-283 Kurt Boman, Mona Olofsson, BjГ¶rn DahlГ¶f, Eva Gerdts, Markku S. Nieminen, Vasilios Papademetriou, Kristian Wachtell and Richard B. Devereux 29. g c d e f 30. g c d e f Comparison of angiotensin-converting enzyme inhibitors in the treatment of congestive heart failure • SHORT COMMUNICATION Pages 283-286 Karen Tu, Muhammad Mamdani, Alex Kopp and Douglas Lee Simplified peak power reserve in patients with an implantable cardioverter-defibrillator and advanced heart failure • SHORT COMMUNICATION Pages 286-288 William T. Katsiyiannis, Alan D. Waggoner, Benico Barzilai, Brian F. Gage, Jose M. Sanchez, Joseph G. Rogers, Bruce D. Lindsay and Marye J. Gleva 31. g c d e f 32. g c d e f 33. g c d e f 34. g c d e f 35. g c d e f Effect of hormone therapy on mortality rates among women with heart failure and coronary artery disease • SHORT COMMUNICATION Pages 289-291 Kirsten Bibbins-Domingo, Feng Lin, Eric Vittinghoff, Elizabeth BarrettConnor, Stephen B. Hulley, Deborah Grady and Michael G. Shlipak Aortic dimensions in patients with bicuspid aortic valve without significant valve dysfunction • SHORT COMMUNICATION Pages 292-294 Moreno Cecconi, Marcello Manfrin, Alessandra Moraca, Raffaele Zanoli, Pier Luigi Colonna, Maria Grazia Bettuzzi, Stefano Moretti, Domenico Gabrielli and Gian Piero Perna Is transesophageal echocardiography overused in the diagnosis of infective endocarditis? • SHORT COMMUNICATION Pages 295-297 Molly Thangaroopan and Jonathan B. Choy Symptomatic patients have similar outcomes compared with asymptomatic patients after carotid artery stenting with emboli protection • SHORT COMMUNICATION Pages 297-300 Michael H. Yen, David S. Lee, Samir Kapadia, Ravish Sachar, Deepak L. Bhatt, Christopher T. Bajzer and Jay S. Yadav Prognostic importance of isolated T-wave abnormalities • SHORT COMMUNICATION Pages 300-304 Takuya Yamazaki, Jonathan Myers and Victor F. Froelicher 36. g c d e f 37. g c d e f Relation between effects of adenosine on brachial artery reactivity and perfusion pattern in patients with known or suspected coronary artery disease • SHORT COMMUNICATION Pages 304-307 Deval Mehta, Gurpreet Baweja, Rajesh Venkataraman, Gilbert J. Zoghbi, Thein Htay, Jaekyeong Heo, Navin C. Nanda and Ami E. Iskandrian Usefulness of real-time three-dimensional echocardiography for reliable measurement of cardiac output in patients with ischemic or idiopathic dilated cardiomyopathy • SHORT COMMUNICATION Pages 308-310 Sean M. Fleming, Barry Cumberledge, Christoph Kiesewetter, Gareth Parry and Antoinette Kenny EDITOR IN CHIEF William C. Roberts, ASSOCIATE EDITORS Baylor Heart & Vascular Hospital Baylor University Medical Center Wadley Tower No. 457 3600 Gaston Avenue Dallas, Texas 75246 (214)826-8252 Fax: (214)826-2855 CARDIOVASCULAR MEDICINE In Adults Jonathan Abrams Robert J. Adolph Joseph S. Alpert Martin A. Alpert Ezra A. Amsterdam Jeffrey Anderson William F. Armstrong Richard W. Asinger Donald S. Baim Gary John Balady Thomas M. Bashore Eric Bates George A. Beller David G. Benditt Peter C. Block William E. Boden Monty M. Bodenheimer Robert O. Bonow Jeffrey S. Borer Harisios Boudoulas Martial G. Bourassa Eugene Braunwald Jeffrey A. Brinker Bruce R. Brodie Alfred E. Buxton Michael E. Cain Richard O. Cannon III Samuel Ward Casscells Agustin Castellanos Bernard R. Chaitman Kanu Chatterjee John S. Child Robert J. Cody Lawrence S. Cohen Marc Cohen C. Richard Conti Michael H. Crawford Robert F. DeBusk Prakash C. Deedwania Gregory J. Dehmer Efthymios N. Deliargyris James A. de Lemos Anthony N. DeMaria Pablo Denes Nicholas L. DePace Richard B. Devereux George A. Diamond John P. DiMarco Michael J. Domanski Gerald Dorros John S. Douglas, Jr. Pamela S. Douglas Eric J. Eichhorn Mark J. Eisenberg Uri Elkayam Thomas C. Andrews Paul A. Grayburn MD ASSISTANT EDITORS Calixto A. Romero Robert L. Rosenthal Carlos E. Velasco EDITORIAL BOARD Kenneth A. Ellenbogen Myrvin H. Ellestad Stephen G. Ellis Toby R. Engel Andrew E. Epstein N. A. Mark Estes, III Michael Ezekowitz Rodney H. Falk John A. Farmer David P. Faxon Robert L. Feldman Ted Feldman Jack Ferlinz Jerome L. Fleg Gerald F. Fletcher Nancy C. Flowers James S. Forrester Joseph A. Franciosa Gary S. Francis Victor F. Froelicher, Jr. W. Bruce Fye William H. Gaasch William Ganz Julius M. Gardin Bernard J. Gersh Mihai Gheorghiade Raymond Gibbons D. Luke Glancy Stephen P. Glasser Michael R. Gold Samuel Z. Goldhaber Robert E. Goldstein Sidney Goldstein Steven A. Goldstein J. Anthony Gomes Antonio M. Gotto, Jr. K. Lance Gould Donald C. Harrison Richard H. Helfant Gary V. Heller Philip D. Henry L. David Hillis David R. Holmes, Jr. William G. Hundley Abdulmassih S. Iskandrian Allan S. Jaffe William B. Kannel Norman M. Kaplan Joel S. Karliner John A. Kastor Sanjiv Kaul Harold L. Kennedy Kenneth M. Kent Richard E. Kerber Dean J. Kereiakes Morton J. Kern Spencer B. King III Robert E. Kleiger George J. Klein Lloyd W. Klein Paul Kligfield Robert A. Kloner John B. Kostis Morris N. Kotler Charles Landau Richard L. Lange Carl J. Lavie Carl V. Leier Joseph Lindsay, Jr. Gregory Y.H. Lip Joseph Loscalzo G.B. John Mancini Francis E. Marchlinski Frank I. Marcus Barry J. Maron Randolph P. Martin Attilo Maseri Dean T. Mason Michael D. McGoon Raymond G. McKay Jawahar L. Mehta Richard S. Meltzer Franz H. Messerli Eric L. Michelson Alan B. Miller Gary S. Mintz Fred Morady Arthur J. Moss James E. Muller Robert J. Myerburg Gerald B. Naccarelli Navin C. Nanda Christopher O’Connor Robert A. O’Rourke Erik Magnus Ohman Antonio Pacifico Richard L. Page Eugene R. Passamani Richard C. Pasternak Alan S. Pearlman Carl J. Pepine Joseph K. Perloff Ileana Pina Bertram Pitt Philip J. Podrid Arshed A. Quyyumi Charles E. Rackley C. Venkata Ram Nathaniel Reichek Robert Roberts William J. Rogers Melvin M. Scheinman Nelson B. Schiller David J. Schneider John S. Schroeder Pravin M. Shah Prediman K. Shah Jamshid Shirani Robert J. Siegel Marc A. Silver Mark E. Silverman Ross J. Simpson, Jr. Steven N. Singh Sidney C. Smith, Jr. Burton E. Sobel John C. Somberg Peter C. Spittell David H. Spodick Lynne W. Stevenson John R. Stratton Jonathan M. Tobis Eric J. Topol Byron F. Vandenberg Hector O. Ventura George W. Vetrovec Ronald G. Victor Robert A. Vogel Frans J. Wackers David D. Waters Nanette K. Wenger William B. White Robert Wilensky James T. Willerson John R. Wilson Miguel Zabalgoitia Barry L. Zaret Douglas P. Zipes In Infants and Children Hugh D. Allen Bruce S. Alpert Arthur Garson, Jr. Stanley J. Goldberg Thomas P. Graham, Jr. Warren G. Guntheroth Howard P. Gutgesell John D. Kugler James E. Lock John W. Moore Lowell W. Perry David J. Sahn Richard M. Schieken A. Rebecca Snider CARDIOVASCULAR SURGERY Eugene H. Blackstone Lawrence I. Bonchek Lawrence H. Cohn John A. Elefteriades Thomas L. Spray RELATED SPECIALISTS L. Maximilian Buja Barry A. Franklin Charles B. Higgins Jeffrey E. Saffitz Renu Virmani Redford B. Williams A3 Impact of Sirolimus-Eluting Stent on the Outcome of Patients With Chronic Total Occlusions Sunao Nakamura, MD, PhD, Tamil Selvan Muthusamy, MD, Jang-Ho Bae, MD, Yeo Hans Cahyadi, MD, Wasan Udayachalerm, MD, and Damras Tresukosol, MD Several randomized trials have demonstrated that stent implantation after successful recanalization of long-term total occlusions decreases restenosis and reocclusion rates. The sirolimus-eluting stent (SES) has recently proved its efficacy to decrease restenosis in selected patients. However, the efficacy of SES implantation in patients who have chronic total occlusions is currently unknown. Therefore, we investigated procedural and 6and 12-month angiographic outcomes (analyzed by quantitative coronary angiography) and left ventricular function in 60 patients who received SESs and 120 patients who received bare metal stents (BMSs). Minimum luminal diameter did not differ immediately after recanalization (SES group 3.04 ШЋ 0.50 mm vs BMS group 3.12 ШЋ 0.48 mm). After 6 months, the SES group still had significantly larger luminal diameters (3.04 ШЋ 0.44 mm vs 1.94 ШЋ 0.98 mm) and significantly lower restenosis and reocclusion rates (2% and 0%, respectively) than did the BMS group (32% and 6%, respectively). Late loss was significantly smaller in the SES group than in the BMS group. At follow-up, the SES group had fewer cardiac events, including target lesion revascularization (p <0.001), than did the BMS group. In conclusion, SES implantation after recanalization of chronic total occlusion provides a better clinical outcome with less restenosis and target lesion revascularization after 6 months than does BMSs. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:161–166) he long-term success after percutaneous coronary intervention for chronic total occlusions (CTOs) is T hampered by high rates of subacute reocclusion and tation in patients who have CTO is currently unknown. Therefore, we performed a prospective, multicenter, observational study in patients who had CTO and underwent elective SES implantation or conventional bare metal stent (BMS) implantation to evaluate success rates, in-hospital and clinical outcomes, and their correlation with predefined clinical and angiographic variables. late restenosis. After successful recanalization of total occlusion (provided the vessel remains patent), patients report relief from symptoms and develop improvement of left ventricular function, fewer cardiac events, and fewer requirements for bypass surgery.1,2 In addition, several randomized trials have demonstrated that stent implantation decreases restenosis and reocclusion rates3–9 and confers a long-term survival advantage.10 However, restenosis remains the major complication limiting late outcome after percutaneous coronary intervention. The efficacy of the sirolimuseluting stent (SES; Cypher, Cordis Corp., Johnson & Johnson, Miami, Florida) has recently been proved to decrease restenosis in selected populations.11–13 Importantly, by maintaining all mechanical properties, the late benefit observed with the SES has been accomplished without compromising the excellent procedural and acute results obtained with conventional metallic stents. However, the efficacy of SES implanFrom the Department of Cardiology, New Tokyo Hospital, Matsudo, Japan; the Department of Cardiology, Damansara Specialist Hospital, Kuala Lumpur, Malaysia; the Department of Cardiology, Konyang University Hospital, Daejeon, Korea; the Department of Cardiology, Husada Hospital, Jakarta, Indonesia; the Department of Cardiology, King Chulalongkorn Memorial Hospital, Bangkok, Thailand; and the Department of Cardiology, Siriaj Hospital, Bangkok, Thailand. Manuscript received May 7, 2004; revised manuscript received and accepted August 31, 2004. Address for reprints: Sunao Nakamura, MD, PhD, Department of Cardiology, New Tokyo Hospital, 473-1 Nemoto, Matsudo-shi, Chiba 271-0077, Japan. E-mail: [email protected]. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 METHODS Study design: From March 2002 to March 2003, a multicenter, prospective, observational clinical study was conducted in 6 Asian heart centers. All 180 consecutive patients who had CTO successfully recanalized with percutaneous coronary intervention and who underwent coronary stenting were enrolled in this study. Among them, 60 patients (60 CTOs in total) underwent SES implantation and were enrolled in the SES group, and 120 patients (120 CTOs in total) underwent BMS implantation (Medtronic S670, Medtronic Vascular, Santa Rosa, California; MultiLink Tristar, Guidant, Temecula, California; BX Velocity, Cordis Corp.) and were enrolled in the BMS group. Patients who had a previous revascularization procedure (percutaneous coronary intervention or coronary artery bypass graft) or myocardial infarction Х…3 months of the procedure and patients who had an estimated CTO duration ПЅ90 days were excluded. Lesion-related criteria for exclusion were a reference diameter ПЅ2.5 mm by visual estimation, an angiographically visible thrombus adjacent to the occlusion site, significant left main coronary artery disease, and totally occlusive diffuse in-stent restenosis. CTO was defined as complete interruption of the vessel with 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.088 161 TABLE 1 Baseline Characteristics of Patients Who Received Bare Metal Stent or Sirolimus-eluting Stent Age (yrs) Men Hypertension Hyperlipidemia Diabetes mellitus Current smoker Previous myocardial infarction Previous coronary bypass Ejection fraction (%) Estimated duration of occlusion (mo) No. of narrowed coronary arteries 1 2 3 Occluded coronary arteries Left anterior descending Left circumflex Right BMS (n П 120) SES (n П 60) 68.6 П® 7.8 75 (63%) 72 (60%) 40 (33%) 39 (33%) 59 (49%) 42 (35%) 10 (8%) 55.8 П® 6.8 7.8 П® 6.6 69.5 П® 8.8 42 (70%) 35 (58%) 21 (35%) 20 (33%) 30 (50%) 22 (37%) 7 (11%) 51.8 П® 7.7 8.8 П® 7.8 20 (17%) 62 (52%) 38 (32%) 13 (22%) 33 (55%) 14 (23%) 54% 27% 33% 52% 28% 28% Values are mean П® SD or numbers of patients (percentages). TABLE 2 In-hospital Clinical Outcome Angiographic success Procedural success Major adverse cardiac event Death Q-wave myocardial infarction Urgent coronary bypass Urgent percutaneous coronary intervention Minor complications Vessel perforation Cardiac tamponade Vascular hematoma BMS (n П 120) SES (n П 60) 120 (100%) 119 (99%) 60 (100%) 60 (100%) 0 1 0 1 0 0 0 0 0 0 2 (2%) 0 0 1 (2%) Thrombolysis In Myocardial Infarction flow grade 0 that was Х†3 months old. Occlusion age was estimated from the date of myocardial infarction on the distribution of the occluded vessel, abrupt worsening of angina pectoris, or information provided by sequential angiograms when they were available. All patients gave written, informed consent to the procedure. In addition, at the time of initial stent implantation, patients were informed about the need for 6- and 12month angiographic and late clinical follow-up. Stent implantation procedure: Premedication consisted of 200 mg/day of aspirin orally and 150 mg of clopidogrel starting Х†24 hours before percutaneous coronary intervention. A bolus of 10,000 U of heparin was administered after sheath insertion, with repeat bolus given as needed to maintain an activated clotting time of Пѕ250 seconds. All patients underwent percutaneous coronary intervention by a standard over-the-wire technique using an exchange catheter (Excelsior, Scimed, Boston Scientific Corporation, Natick, Massachusetts) in which the femoral approach was used. The technique has been 162 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 TABLE 3 Angiographic Characteristics (Baseline) Reference diameter (proximal) (mm) Minimum luminal diameter after stenting (mm) Angiographic calcification Side branch: present Lesion length (mm) Chronic total occlusion length (mm) Stent length (mm) Maximum inflation pressure (atm) Stent overlap BMS (n П 120) SES (n П 60) 3.32 П® 0.50 3.12 П® 0.60 3.12 П® 0.48 3.04 П® 0.50 57% 32% 32.5 П® 7.8 23.6 П® 6.6 68% 30% 35.0 П® 8.8 24.8 П® 5.5 30.8 П® 8.8 15.9 П® 4.4 36.5 П® 2.2 16.8 П® 3.0 43% 47% Values are mean П® SD or percentages. modified over time using second-generation guidewires, including Cross-It (Guidant, Santa Clara, California) and Conquest Pro (Asahi Intecc Co., Ltd., Nagoya, Japan), or use of sophisticated techniques for crossing the occluded artery.14 After the lesion was crossed by the guidewire and successfully recanalized with standard angioplasty that resulted in a reference diameter Х†2.5 mm on on-line quantitative coronary angiography, BMSs and SESs were implanted in patients on the BMS and SES registries, respectively. Stent deployment was performed using an initial inflation of 16 atm in the 2 groups. Stent expansion was completed by additional inflation (maximum 20 atm) using other short balloons until good angiographic results were obtained. Definitions: CTO was defined as a lesion that exhibited Thrombolysis In Myocardial Infarction flow grade 0 in a native coronary artery. Angiographic technical success was defined as the ability to cross the occluded segment with wire and balloon and successfully open the artery with ПЅ40% residual stenosis from all views. Procedural success was defined as angiographic success with no in-hospital major adverse cardiac events. Major adverse cardiac events were defined as the occurrence of death, Q-wave myocardial infarction, or urgent revascularization. Urgent revascularization was classified by operators who cared for patients and required repeat percutaneous coronary intervention of the target vessel during the same admission or coronary artery bypass grafting, including bypass of the target vessel. Follow-up: Coronary events during and after the procedure were recorded. Clinical follow-up was performed in the cardiology clinic 1, 3, 6, 9, and 12 months after the procedure and included physical examination, stress testing, and documenting a history of cardiac complaints for all patients. Six and 12 months after recanalization, patients were hospitalized for repeat angiography as much as possible (follow-up angiographic rates of 67% at 6 months and 75% at 12 months in the SES group and 97% at 6 months in the BMS group). Angiographic analysis: Angiographic analysis was performed at a central core laboratory by 2 experiJANUARY 15, 2005 period, 120 of 130 screened patients who had CTO (92%) were successSES fully recanalized, implanted with BMS at BMS (BX Velocity 59%, Multi-Link 6 months 6 months 12 months Tristar 22%, Medtronic AVE 19%), (n П 120) (n П 60) (n П 60) p Value and enrolled in the multicenter BMS Clinical follow-up 120/120 60/60 60/60 NS registry. Angiographic follow-up 116/120 40/60 45/60 0.05 Baseline clinical characteristics: BaseReference diameter (mm) 3.24 П® 0.48 3.14 П® 0.55 3.16 П® 0.50 NS Minimum luminal diameter (mm) 1.94 П® 0.98 3.04 П® 0.44 3.00 П® 0.50 0.001 line characteristics of the 2 groups Late loss (mm) 1.36 П® 0.88 0.08 П® 0.10 0.12 П® 0.10 0.001 are listed in Table 1. There was no Loss index 0.41 П® 0.48 0.03 П® 0.02 0.04 П® 0.03 0.001 significant difference in baseline Restenosis rate 32% 2% 2% 0.001 clinical and angiographic characterReocclusion rate 6% 0% 0% 0.001 Target vessel revascularization 30% 3% 3% 0.001 istics between groups. Most patients Target lesion revascularization 23% 2% 2% 0.001 presented with multivessel disease (BMS 83%, SES 78%). About 33% Values are numbers of patients, mean П® SD, or percentages. of patients had a previous myocardial infarction (in the region of the myocardium supplied by the target enced operators who were not involved in the stenting vessel; BMS 35%, SES 37%). Viability of the myoprocedure. Angiograms were quantitatively analyzed cardium supplied by the target vessel was not invesby a computer-assisted cardiovascular measurement tigated by protocol, but in 70% of patients, the pressystem (MEDIS, Nuenen, The Netherlands) that used ence of normal or only decreased regional systolic semiautomated edge detection. Minimum luminal di- wall motion of that area evaluated by angiography or ameter, reference diameter, and percent diameter ste- echocardiography suggested the presence of a viable nosis were measured before and after the procedure myocardium. and at 6- and 12-month follow-up; results from the In-hospital outcomes: Table 2 lists in-hospital single “worst” view were recorded. events. Angiographic success was obtained in 100% Study end points: The primary end point in this of the 2 groups. Procedural success was obtained in study was the occurrence of major adverse cardiac 100% of the SES group but was 99% of the BMS events, defined as cardiac death, any myocardial in- group because of 1 patient who developed subacute farction, or target lesion revascularization with repeat thrombosis. This patient was treated with repeat perpercutaneous coronary intervention or coronary artery cutaneous coronary intervention and eventually develbypass graft after SES implantation versus BMS im- oped Q-wave myocardial infarction. Bleeding compliplantation. Death was considered cardiac unless oth- cations in the 2 groups were also not significantly erwise demonstrated. Myocardial infarction was doc- different. umented by the presence of a new pathologic Q wave Angiographic results: Baseline and follow-up anon an electrocardiogram, according to the Minnesota giographic results are presented in Tables 3 and 4. Code (F-11), or an increase in creatinine kinase to Пѕ2 There were no significant differences between the times the upper limit of normal. Anginal symptoms at groups in terms of reference diameter, minimum lufollow-up and angiographic restenosis after SES im- minal diameter after stenting, lesion length, CTO plantation versus BMS implantation were secondary length, and stent length. Angiographic follow-up at 6 end points. These symptoms were graded according to months was performed in 67% of the SES group and the classification of the Canadian Cardiovascular in 97% of the BMS group. In the SES group, 75% of Society. patients consented to 12-month follow-up angiograStatistical analysis: Categorical values are ex- phy. At follow-up after 6 and 12 months, the SES pressed as absolute numbers and percent values. Con- group had significantly larger minimum luminal ditinuous variables are expressed as mean П® SD. Dif- ameters because of a significantly smaller late loss. ferences between groups were assessed with Target lesion revascularization and target vessel recontinuity-adjusted chi-square test or Fisher’s exact vascularization were also significantly lower in the test for categorical variables and with a 2-tailed un- SES group. The restenosis rates at follow-up, defined paired Student’s t test or Mann-Whitney U statistical as Пѕ50% stenosis, were 32% in the BMS group and test for continuous variables. Event-free survival rates 2% in the SES group (p ПЅ0.001). Reocclusion rates were estimated with the Kaplan-Meier method, with were 6% in the BMS group and 0% in the SES group the difference between groups assessed with a log- (p ПЅ0.001). Figure 1 shows the cumulative distriburank test of significance. A p value ПЅ0.05 was con- tion curve of changing minimum luminal diameter of sidered statistically significant. the BMS and SES groups. The cumulative distribution curve of the SES group did not differ from baseline (after stenting) to follow-up. RESULTS Patients: From March 2002 to March 2003, 60 of Events and clinical outcome: Long-term clinical out66 screened patients who had CTO (91%) were suc- come is presented in Table 5. Clinical follow-up data cessfully recanalized, implanted with SES, and en- were available from 100% of patients in the 2 groups. rolled in the multicenter SES registry. In the same Three nonfatal myocardial infarctions were observed, TABLE 4 Angiographic Characteristics and Clinical Outcome (Follow-up) CORONARY ARTERY DISEASE/SIROLIMUS-ELUTING STENT IN CTO 163 FIGURE 2. Elective implantation of the Cypher SES after successful recanalization improved left ventricular (LV) ejection fraction significantly. FIGURE 1. Cumulative distribution curves of mininum luminal diameter (MLD) assigned to (A) BMS and (B) SES implantations. Binary restenosis rates of the BMS and SES groups were 27.5% and 2.2%, respectively. mo f/u ‫ ؍‬months follow-up. TABLE 5 Long-term Clinical Outcome (12 months) Major adverse cardiac event Death Myocardial infarction Coronary bypass Repeat percutaneous coronary intervention Any event BMS (n П 120) SES (n П 60) p Value 0 3 (3%) 7 (6%) 44 (37%) 0 0 0 2 (3%) NS NS 0.01 0.001 50 (42%) 2 (3%) 0.001 and 7 patients in the BMS group underwent coronary artery bypass grafting. Repeat percutaneous coronary intervention was performed in 37% of patients in the BMS group versus only 3% of patients in the SES group. Ten patients in the BMS group and 1 patient in the SES group showed no involvement of the study vessel and were treated by progression of the disease in other vessels; the other 28% of patients in the BMS group and 1 patient in the SES group underwent target lesion revascularization procedures. Overall, surgical or percutaneous target lesion revascularization after 6 months was performed in 42% of patients in the BMS group and in 3% of patients in the SES group. Stress 164 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 testing and left ventricular function data analysis (Figure 2) showed a strong tendency toward a better outcome in patients who received SESs. Patients in the SES group showed a significantly lower percentage of positive results from stress testing at follow-up than did those in the BMS group (3% vs 37%). This difference was statistically significant. Estimated probabilities of freedom from major adverse cardiac events at 6 months were 92% in the BMS group and 100% in the SES group, and probabilities of freedom from major adverse cardiac events and repeat percutaneous coronary intervention were 58% in the BMS group and 97% in the SES group (p ПЅ0.001). This difference was essentially determined by a higher estimated 6-month survival rate without target lesion revascularization in the SES group. Kaplan-Meier analyses of freedom from major adverse cardiac events and freedom from major adverse cardiac events and repeat percutaneous coronary intervention are shown in Figure 3. DISCUSSION In the present study, we report the first clinical experience with the Cypher SES for the management of CTO. The major findings of this study were as follows. First, implantation of SES was feasible and safe for the management of CTO; no cases of subacute stent thrombosis were noted. Second, late loss 6 months after SES implantation seemingly decreased compared with late loss in the BMS group. Third, the angiographic restenosis rate was 2% in the SES group compared with 32% in the BMS group. Fourth, the target lesion revascularization rate of 2% in the SES group was surprisingly lower than the rate of 23% in the BMS group. Several clinical studies have documented the challenge of controlling CTO because of the lower success rate and high incidence of restenosis (30% to 60%).15,16 The recent availability of new specialized guidewires or more sophisticated techniques for crossing occluded arteries has contributed to the higher success rate of CTO management.17–21 In addition, JANUARY 15, 2005 enotic effect after management of CTO without the attendant risk of stent thrombosis observed with other effective treatment modalities. Long-term follow-up will be required to determine whether the inhibition of restenosis observed in this study is sustained. Randomized controlled trials will be necessary. Acknowledgment: We are indebted to Shigeru Saito, MD, Shonan Kamakura General Hospital, Japan for valuable comments. We also thank Ritsuko Otaka for diligent work in organizing and typing the manuscript. 1. Heyder M, Engel HJ, Hormann E. Angioplasty of chronic coronary artery FIGURE 3. Kaplan-Meier analysis of survival rates free of (A) major adverse cardiac events (MACE) and (B) MACE Ш‰ target lesion revascularization (TLR) in patients assigned to BMS implantation (solid line) or SES implantation (dotted line). randomized studies have demonstrated that stent implantation decreases restenosis and reocclusion rates.3–9 However, the remaining restenosis is mainly attributed to reoccurrence of major adverse cardiac events. Therefore, solving the problem of restenosis is apparently the key to improving the long-term outcome in these patients. In this study, notwithstanding the challenging population treated, we found strikingly similar results in terms of suppression of neointimal proliferation to that reported previously in low-risk patient populations.11,22,23 Acute procedural, in-hospital, and 12month outcomes were favorable. At 6- and 12-month angiographic follow-up, only 1 patient required target lesion revascularization due to having a gap between 2 stents. In the remaining patient, late luminal loss averaged 0.08 mm. This was extremely low compared with the BMS group and provides an extremely low incidence of restenosis. This was a small observational study. However, the lack of stent thrombosis, myocardial infarction, or death and only 1 case of repeat percutaneous coronary intervention indicate that antiproliferative and anti-inflammatory actions of SES may provide a long-lasting antirest- occlusions: impact on functional status and left ventricular performance. Z Kardiol 1996;85:125–132. 2. Bell MR, Berger PB, Bresnahan JF, Reeder GS, Bailey KR, Holmes DR Jr. Initial and long-term outcome of 354 patients after coronary balloon angioplasty of total coronary occlusions. Circulation 1992;85:1003–1011. 3. Sirnes PA, Golf S, Myreng Y, Molstad P, Emanuelsson H, Albertsson P, Brekke M, Mangschau A, Endresen K, Kjekshus J. Stenting In Chronic Coronary Occlusion (SICCO): a randomized, controlled trial of adding stent implantation after successful angioplasty. J Am Coll Cardiol 1996;28:1444 –1451. 4. Hancock J, Thomas MR, Holmberg S, Wainwright RJ, Jewitt DE. Randomized trial of elective stenting after successful percutaneous transluminal coronary angioplasty of occluded coronary arteries. Heart 1998;79:18 –23. 5. Rubartelli P, Niccoli L, Verna E, Giachero C, Zimarino M, Fontanelli A, Vassanelli C, Campolo L, Martuscelli E, Tommasini G. Stent implantation versus balloon angioplasty in chronic coronary occlusions: results from the GISSOC trial. Gruppo Italiano di Studio sullo Stent nelle Occlusioni Coronariche. J Am Coll Cardiol 1998;32:90 –96. 6. Buller CE, Dzavik V, Carere RG, Mancini GBJ, Barbeau G, Lazzam C, Anderson TJ, Knudtson ML, Marquis JF, Suzuki T, et al. Primary stenting versus balloon angioplasty in occluded coronary arteries: the Total Occlusion Study of Canada (TOSCA). Circulation 1999;100:236 –242. 7. Sievert H, Rohde S, Utech A, Schulze R, Scherer D, Merle H, Ensslen R, Schrader R, Spies H, Fach A. Stent or angioplasty after recanalization of chronic coronary occlusions? (The SARECCO Trial). Am J Cardiol 1999;84:386 –390. 8. Hoher M, Wohrle J, Grebe OC, Kochs M, Osterhues HH, Hombach V, Buchwald AB. A randomized trial of elective stenting after balloon recanalization of chronic total occlusions. J Am Coll Cardiol 1999;34:722–729. 9. Lotan C, Rozenman Y, Hendler A, Turgeman Y, Ayzenberg O, Beyar R, Krakover R, Rosenfeld T, Gotsman MS. Stents in total occlusion for restenosis prevention. The multicentre randomized STOP study. The Israeli Working Group for Interventional Cardiology. Eur Heart J Med 2000;21:1960 –1966. 10. Olivari Z, Rubartelli P, Piscione F, Ettori F, Fontanelli A, Salemme L, Giachero C, Di Mario C, Gabrielli G, Spedicato L, Bedogni F. Immediate results and one-year clinical outcome after percutaneous coronary interventions in chronic total occlusions. J Am Coll Cardiol 2003;41:1672–1678. 11. Morice MC, Serruys PW, Sousa JE, Fajadet J, Hayashi EB, Perin M, Colombo A, Schuler G, Barragan P, Guagliumi G, et al. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med 2002;346:1773–1780. 12. Degertekin M, Serruys PW, Foley DP, Tanabe K, Regar E, Vos J, Smits PC, van der Giessen WJ, van den Brand M, de Feyter P, Popma JJ. Persistent inhibition of neointimal hyperplasia after Sirolimus-eluting stent implantation: long-term (up to 2 years) clinical, angiographic, and intravascular ultrasound follow-up. Circulation 2002;106:1610 –1613. 13. Rensing BJ, Vos J, Smits PC, Foley DP, van den Brand MJ, van der Giessen WJ, de Feijter PJ, Serruys PW. Coronary restenosis elimination with a sirolimus eluting stent: first European human experience with 6-month angiographic and intravascular ultrasonic follow-up. Eur Heart J 2001;22:2125–2130. 14. Lefevre T, Louvard Y, Loubeyre C, Dumas P, Piechaud JF, Krol M, Benslimane A, Premchand RK, Morice MC. A randomized study comparing two guidewire strategies for angioplasty of chronic total coronary occlusion. Am J Cardiol 2000;85:1144 –1147. 15. Violaris AG, Melkert R, Serruys PW. Long-term luminal renarrowing after successful elective coronary angioplasty of total occlusions. A quantitative angiographic analysis. Circulation 1995;91:2140 –2150. 16. Berger PB, Holmes DR Jr, Ohman EM, O’Hanesian MA, Murphy JG, Schwartz RS, Serruys PW, Faxon DP. Restenosis, reocclusion and adverse cardiovascular events after successful balloon angioplasty of occluded versus nonoccluded coronary arteries. Results from the Multicenter American Research Trial With Cilazapril After Angioplasty to Prevent Transluminal Coronary Obstruction and Restenosis (MARCATOR). J Am Coll Cardiol 1996;27:1–7. CORONARY ARTERY DISEASE/SIROLIMUS-ELUTING STENT IN CTO 165 17. Allemann Y, Kaufmann UP, Meyer BJ, Gunnes P, Urban P, Mulhauser B, Dorsaz PA, Meier B. Magnum wire for percutaneous coronary balloon angioplasty in 800 total chronic occlusions. Am J Cardiol 1997;80:634 – 637. 18. Reimers B, Camassa N, Di Mario C, Akiyama T, Di Francesco L, Finci L, Colombo A. Mechanical recanalization of total coronary occlusions with the use of a new guide wire. Am Heart J 1998;135:726 –731. 19. Kahler J, Koster R, Brockhoff C, Reimers J, Baldus S, Terres W, Meinertz T, Hamm CW. Initial experience with a hydrophilic-coated guidewire for recanalization of chronic coronary occlusions. Catheter Cardiovasc Interv 2000;49:45–50. 20. Serruys PW, Hamburger JN, Koolen JJ, Fajadet J, Haude M, Klues H, SeabraGomes R, Corcos T, Hamm C, Pizzuli L, et al. Total occlusion trial with angioplasty by using laser guidewire. The TOTAL trial. Eur Heart J 2000;21:1797–1805. 166 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 21. Piscione F, Galasso G, Maione AG, Pisani A, Golino P, Leosco D, Indolfi C, Chiariello M. Immediate and long term outcome of recanalization of chronic total coronary occlusions. J Interv Cardiol 2002;15:173–179. 22. Sousa JE, Costa MA, Abizaid A, Abizaid AS, Feres F, Pinto IMF, Seixas AC, Staico R, Mattos LA, Sousa AGMR, et al. Lack of neointimal proliferation after implantation of sirolimus-coated stents in human coronary arteries: a quantitative coronary angiography and three-dimensional intravascular ultrasound study. Circulation 2001;103:192–195. 23. Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, O’Shaughnessy C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, et al. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2002;349:1315–1323. JANUARY 15, 2005 Impact of Baseline Renal Function on Mortality After Percutaneous Coronary Intervention With Sirolimus-Eluting Stents or Bare Metal Stents Pedro A. Lemos, MD, PhD, Chourmouzios A. Arampatzis, MD, PhD, Angela Hoye, MB, ChB, MRCP, Joost Daemen, Andrew T.L., Ong, MBBS, Francesco Saia, MD, Willem J. van der Giessen, MD, PhD, Eugene P. McFadden, MB, ChB, MD, Georgios Sianos, MD, PhD, Pieter C. Smits, MD, Pim de Feyter, MD, PhD, Sjoerd H. Hofma, MD, Ron T. van Domburg, PhD, and Patrick W. Serruys, MD, PhD PhD, Renal impairment is an important predictor of mortality after percutaneous coronary intervention and may increase the restenosis rate. However, the relation between restenosis and the risk of death in patients who have renal impairment remains unclear. We evaluated the incidences of repeat revascularization and mortality in patients who had renal impairment and those who did not and who received sirolimus-eluting stents or bare stents. A total of 1,080 consecutive patients treated for 1 year had available data to calculate baseline creatinine clearance. Patients received bare stents (first 6 months, n ‫ ؍‬543) or sirolimus-eluting stents (last 6 months, n ‫ ؍‬537) and were grouped according to the presence or absence of renal impairment (creatinine clearance <60 ml/min). Patients who had renal impairment had a higher mortality rate at 1 year (7.6% vs 2.5%, hazard ratio 3.14, 95% confidence interval 1.68 to 5.88, p <0.01), with no differences in mortality be- tween patients who received bare stents and those who received sirolimus-eluting stents (hazard ratio 0.91, 95% confidence interval 0.49 to 1.68, p ‫ ؍‬0.8). The incidence of target vessel revascularization decreased significantly in patients who were treated with sirolimus-eluting stents and did not have renal impairment (hazard ratio 0.59, 95% confidence interval 0.39 to 0.90, p ‫ ؍‬0.01) and in those who had decreased renal function (hazard ratio 0.37, 95% confidence interval 0.15 to 0.90, p ‫؍‬ 0.03). Thus, sirolimus-eluting stents compared with conventional stents decreased clinical restenosis in patients who had renal impairment. However, this benefit was not paralleled by a decrease in the risk of death in this population. It seems unlikely that restenosis could be a contributing factor that influenced the increased mortality of patients who had impaired renal function. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:167–172) hronic kidney disease has been shown to strongly increase the risk of short- and long-term adverse C events in patients who have atherosclerotic dis- clinical trials conducted to date have excluded this subset of patients who had impaired renal function and, as a consequence, the effect of SES implantation on the outcomes of this subset of patients is currently unknown. The present study evaluated the effect of baseline renal function on 1-year mortality of patients who had been treated with conventional bare stents or SESs. ease,1–11 and patients who have renal failure have been reported to have higher mortality rates after successful percutaneous coronary intervention.10 –12 Neither surgical nor percutaneous revascularization has been shown to eliminate the increased risk of patients who have renal impairment.7–12 Sirolimus-eluting stents (SESs) have proved to markedly decrease neointimal growth and in-stent restenosis compared with conventional stents, with an impressive decrease in the risk of subsequent repeat revascularization.13–15 However, all From the Heart Institute (InCor), University of SГЈo Paulo Medical School, SГЈo Paulo, Brazil; and the Erasmus Medical Center, Thoraxcenter, Rotterdam, The Netherlands. This study was supported by the Erasmus Medical Center, Erasmus University, Rotterdam, The Netherlands, and by an unrestricted institutional grant from Cordis, a Johnson & Johnson Company, Miami Lakes, Florida. Manuscript received June 23, 2004; revised manuscript received and accepted August 31, 2004. Address for reprints: Patrick W. Serruys, MD, PhD, Thoraxcenter, Bd-406, Dr. Molewaterplein 40, 3015-GD Rotterdam, The Netherlands. E-mail: [email protected]. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 METHODS Patient population: Since April 2002, SES implantation (Cypher; Cordis, Johnson & Johnson, Cordis Europa NV, Roden, The Netherlands) has been adopted as the default interventional strategy for all patients who are treated in our institution, as described elsewhere.16,17 For comparison, a control group was composed of all consecutive patients who had been treated with conventional bare stents in the period before the introduction of SESs.16,17 From October 2001 to October 2002 (6-month enrollment for presirolimus and sirolimus phases), 1,262 consecutive patients who were not on dialysis were treated with bare stents or SESs in the 2 study periods. Among 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.089 167 these patients, 1,080 (86%) had preprocedural serum creatinine measured in our institution and comprised the present study population (bare stent group, n П 543 patients; SES group, n П 537 patients). This protocol was approved by the local ethics committee and was conducted in accordance with the Declaration of Helsinki. Written, informed consent was obtained from every patient. Interventional procedures: All interventions were performed according to standard techniques, and the final strategy was left to the discretion of the operators. Angiographic success was defined as residual stenosis ПЅ30% by visual analysis with Thrombolysis In Myocardial Infarction grade 3 anterograde flow. Periprocedural use of a glycoprotein IIb/IIIa inhibitor was left to the discretion of the operator. All patients were advised to maintain lifelong use of aspirin. Clopidogrel was prescribed for Х†1 month in the bare stent group. For patients who were treated with SESs, clopidogrel was recommended for 3 months, unless patients had Х†1 of the following characteristics (in which case clopidogrel was maintained for Х†6 months): multiple SES implantations (Пѕ3 stents), total stented length Пѕ36 mm, chronic total occlusion, bifurcations, and in-stent restenosis. Clinical follow-up and end points: In-hospital clinical information was retrieved from an electronic database of patients maintained in our hospital and by review of hospital records for those who had been discharged to referring hospitals. Postdischarge survival status was obtained from the municipal civil registries. Repeat revascularization procedures (surgical or percutaneous) and rehospitalizations were prospectively collected during follow-up. Patients were directly approached and/or the referring physicians and institutions contacted whenever necessary for additional information. The primary end point of the present study was all-cause mortality at 1 year. Incidence of target vessel revascularization was assessed to evaluate the antirestenotic effect of SESs versus bare stents. Target vessel revascularization was defined as a reintervention (surgical or percutaneous) to treat any lesion located in the same epicardial vessel treated at the index procedure. Renal function evaluation: The closest creatinine values before the procedure were used to calculate baseline creatinine clearance according to the formula proposed by Cockcroft and Gault18: creatinine clearance (milliliters/minute) П (140 ПЄ age) П« weight (kilograms) П¬ 72 П« serum creatinine (milligrams/ deciliter) (П« 0.85 for women). Renal impairment was defined as a calculated creatinine clearance ПЅ60 ml/min, a cut-off value previously proposed by the National Kidney Foundation’s Kidney Disease Outcome Quality Initiative Advisory Board to identify patients who have moderate renal impairment5 and the American Heart Association’s Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention.6 Statistical analysis: Continuous variables are presented as mean П® SD and were compared with Student’s t test. Categorical variables are presented as 168 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 counts and percentages and were compared with Fisher’s exact test. An unadjusted cumulative incidence of death and target vessel revascularization was evaluated by the Kaplan-Meier method. Cox’s proportional hazards models were used to examine the effects of renal impairment, SES implantation, and the interaction between stent type versus renal function on clinical end points. To adjust for baseline differences between study groups, all variables associated with the clinical end points at univariate analyses (p Х…0.2 for selection) were tested in multivariate analyses to identify independent predictors of 1-year mortality (tested variables were gender, acute myocardial infarction at admission, 3-vessel disease, hypercholesterolemia, current smoking status, diabetes, angiographic success, statin prescription, left anterior descending artery stenting, left main coronary artery stenting, bypass graft stenting, and renal function impairment) and target vessel revascularization (tested variables were previous bypass surgery, acute myocardial infarction at admission, 3-vessel disease, hypercholesterolemia, current smoking status, diabetes, left main coronary artery stenting, bypass graft stenting, number of stents implanted, treatment of in-stent restenosis, and SES use). The final models were built by backward stepwise variable selection, with a p value ПЅ0.05 used as a criterion for entry and removal of variables. All reported p values are 2-tailed, and a p value ПЅ0.05 was considered statistically significant. RESULTS Baseline and procedural characteristics: Among 543 patients who had been treated with bare stents, 92 (17%) had renal dysfunction at baseline; among 537 patients who had been treated with SESs, renal dysfunction was present in 94 (18%). Table 1 presents baseline and procedural characteristics of patients who had normal renal function or renal impairment, according to the type of stent used. Pooled over the stent type used, patients who had renal impairment were older, more frequently women, and more often had hypertension, previous coronary surgery, 3-vessel disease, bypass graft stenting, and more stents implanted per procedure. In addition, patients who had a lower rate of clearance weighed less, were shorter, and were less likely to smoke and to have received stenting in the right coronary artery. Overall, in patients who did or did not have renal impairment, average serum creatinine level was 0.9 П® 0.2 versus 1.3 П® 0.4 mg/dl, respectively (p ПЅ0.01), and average creatinine clearance was 99 П® 27 versus 49 П® 9 mg/dl, respectively (p ПЅ0.01). Preprocedural baseline characteristics of patients who had been treated with SESs and those who had been treated with bare stents were similar, except for a lower frequency of current smokers and a higher rate of previous percutaneous intervention and treatment of restenotic lesions in patients who received SESs (Table 1). The average creatinine clearance was similar between patients who had been treated with SESs or with bare stents. In the SES group, use of glycoprotein IIb/IIIa inhibitors was lower and the number of JANUARY 15, 2005 TABLE 1 Baseline and Procedural Characteristics of Patients With or Without Renal Impairment Treated With Bare Stents or Sirolimus-eluting Stents Normal Renal Function Variable Men Age (yrs) Height (cm) Weight (kg)†Hypercholesterolemia or use of lipid-lowering drug Hypertension Current smoker†Diabetes mellitus Insulin dependent Non–insulin dependent Previous myocardial infarction Previous bypass surgery Previous percutaneous coronary intervention†Clinical presentation Stable angina pectoris Unstable angina pectoris Acute myocardial infarction Coronary vessel disease 1-vessel 2-vessel 3-vessel Coronary vessel treated Right coronary artery Left anterior descending Left circumflex artery Left main coronary Bypass graft No. of in-stent restenosis (Х†1 lesion)†No. of stents implanted per patient‡ Angiographic success for all lesions Periprocedural use of glycoprotein IIb/IIIa inhibitor†Statin use at discharge ACE inhibitor use at discharge Clopidogrel prescription (months)‡ Serum creatinine (mg/dl) Creatinine clearance (ml/min) Renal Impairment Bare Stent (n П 451) SES (n П 443) Bare Stent (n П 92) SES (n П 94) 79% 59 П® 11 174 П® 8 84 П® 13 56% 36% 37% 14% 4% 10% 39% 7% 21% 73% 59 П® 10 173 П® 9 82 П® 14 58% 39% 31% 18% 5% 13% 33% 9% 27% 48% 72 П® 8 167 П® 10 71 П® 11 51% 49% 20% 20% 9% 11% 39% 25% 25% 50% 72 П® 9 167 П® 9 70 П® 11 62% 50% 17% 20% 6% 14% 31% 21% 31% 47% 34% 19% 50% 35% 16% 55% 35% 10% 38% 47% 15% 49% 36% 16% 48% 34% 19% 36% 29% 35% 33% 31% 36% 39% 57% 34% 4% 3% 4% 1.9 П® 1.1 97% 39% 66% 30% 3.0 П® 2.1 0.9 П® 0.8 101 П® 29 39% 57% 31% 2% 3% 10% 2.1 П® 1.4 97% 17% 65% 25% 4.2 П® 2.0 0.9 П® 0.2 98 П® 25 32% 52% 33% 5% 12% 6% 2.1 П® 1.4 98% 29% 59% 26% 3.0 П® 1.8 1.3 П® 0.4 49 П® 9 28% 63% 33% 5% 11% 13% 2.4 П® 1.6 98% 21% 62% 28% 4.1 П® 2.0 1.3 П® 0.4 50 П® 9 p Value* ПЅ0.01 ПЅ0.01 ПЅ0.01 ПЅ0.01 0.9 ПЅ0.01 ПЅ0.01 0.2 0.1 0.7 0.8 ПЅ0.01 0.2 0.1 ПЅ0.01 0.02 0.9 1.0 0.2 ПЅ0.01 0.2 ПЅ0.01 0.8 0.5 0.2 1.0 0.9 ПЅ0.01 ПЅ0.01 *Patients who had normal renal function versus patients who had renal impairment and were pooled over stent type group. †p ПЅ0.05 for bare stents versus SESs pooled over renal function group. ‡ p ПЅ0.01 for bare stents versus SESs pooled over renal function group. Values are mean П® SD or percentages. ACE П angiotensin-converting enzyme. stents implanted per procedure was greater (Table 1). Clopidogrel use was longer in the SES group, according to the predefined management protocol. There were no differences with regard to postprocedural prescription of statins between the SES and bare stent groups. One-year mortality: Clinical follow-up data were available for 1,074 patients (99.4%; median follow-up period 421 days, interquartile range 391 to 459). When all patients were pooled together, regardless of SES or bare stent use, the unadjusted risk of death at 1 year was significantly higher in patients who had renal impairment than in patients who had normal renal function (7.6% vs 2.5%, respectively, hazard ratio 3.14, 95% confidence interval [CI] 1.68 to 5.88, p ПЅ0.01; Figure 1). Similarly, when analyzed separately, baseline renal impairment significantly increased the risk of death in patients who had been treated with bare stents or SESs (Figure 1). When evaluated irrespective of renal function, patients who had been treated with bare stents or SESs had similar 1-year mortality rates (3.6% vs 3.2%, respectively, hazard ratio 0.91, 95% CI 0.49 to 1.68, p П 0.8; Figure 2). The interaction factor for the relation of the effects of renal impairment and stent type on risk of death was not significant (p П 0.7 for interaction). Independent predictors of mortality at 1 year by multivariate analysis are presented in Table 2. Use of SESs had no influence in risk of death at 1 year (adjusted hazard ratio 1.10, 95% CI 0.59 to 2.07, p П 0.8). Repeat revascularization: Overall, SES implantation significantly decreased the incidence of target vessel revascularization at 1 year compared with bare stent implantation (6.9% vs 13.1%, unadjusted hazard ratio 0.54, 95% CI 0.37 to 0.79, p ПЅ0.01). SES implantation was effective in decreasing the risk of target vessel revascularization in patients who did not have CORONARY ARTERY DISEASE/RENAL FUNCTION AFFECTS OUTCOMES AFTER SIROLIMUS STENTS 169 FIGURE 2. Unadjusted 1-year incidence of all-cause death for patients who had been treated with bare stents versus those who had been treated with SESs pooled over renal function. TABLE 2 Independent Multivariate Predictors of One-year Mortality and Target Vessel Revascularization Rates Variable 1-yr mortality rate Women Renal impairment Acute myocardial infarction 3-Vessel disease SES use 1-yr target vessel revascularization rate Renal impairment SES use Current smoker Treatment of in-stent restenosis No. of stents implanted Hazard Ratio 95% CI p Value 2.00 2.15 3.00 2.75 1.10 1.04–3.85 1.10–4.28 1.54–5.86 1.44–5.22 0.59–2.07 0.04 0.03 ПЅ0.01 ПЅ0.01 0.8 1.22 0.43 0.56 3.29 1.23 0.79–1.88 0.29–0.64 0.36–0.88 2.05–5.28 1.09–1.38 0.4 ПЅ0.01 0.01 ПЅ0.01 ПЅ0.01 0.37, 95% CI 0.15 to 0.90, p П 0.03). At multivariate analysis, SES use remained an important factor that decreased the risk of repeat revascularization (adjusted hazard ratio 0.43, 95% CI 0.29 to 0.64, p ПЅ0.01). Importantly, the presence of renal impairment did not significantly influence the risk of target vessel revascularization (adjusted hazard ratio 1.22, 95% CI 0.79 to 1.88, p П 0.4). Other independent predictors of 1-year target vessel revascularization are presented in Table 2. FIGURE 1. Unadjusted 1-year incidence of all-cause death after percutaneous coronary intervention for patients who had renal impairment and those who did not. Total study population pooled over stent type (top) and patients who had been treated with bare stents (middle) or SESs (bottom). renal impairment (7.2% vs 11.8%, unadjusted hazard ratio 0.59, 95% CI 0.39 to 0.90, p П 0.01) and in those who did (5.6% vs 19.6%, unadjusted hazard ratio 170 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 DISCUSSION The main finding of the present study was that impaired renal function significantly increases 1-year mortality after percutaneous coronary revascularization, regardless of stent type. Despite the clear antirestenotic effect of SESs, which markedly decreased the incidence of target vessel revascularization compared with bare stents, mortality rates in patients who had renal dysfunction and those who did not were similar for the 2 treatment strategies. JANUARY 15, 2005 Impaired renal function has been shown to negatively influence survival rates after percutaneous intervention.7–12 Although patients who have renal impairment are well known to have an increased prevalence of associated risk factors, it has been identified in our series and in previous reports as an important independent predictor of mortality. Several inflammatory, procoagulant, and atherogenic markers have been described in patients who have renal impairment,19 –26 which might accelerate disease progression and account for a higher tendency to acute events. Some of these factors have been associated with an increased risk of late restenosis.27–29 Accordingly, patients who have end-stage renal failure have been shown to present with increased levels of fibrinogen and higher rates of restenosis than patients not on dialysis.29 It has been hypothesized that the high incidence of restenosis may contribute to the increased mortality rate seen in patients who have renal impairment.10,12 Our results challenge this concept by demonstrating that the striking decrease in clinical restenosis after SES implantation was not paralleled by any decrease in mortality rate among patients who had renal impairment. Although drug-eluting stents did not decrease mortality risk after coronary intervention, the decrease in restenosis represents an important therapeutic achievement for the clinical management of patients who have renal impairment. In our series, SES implantation decreased the risk of repeat revascularization by Пѕ50% of that observed with bare stents in patients who had renal impairment. The marked decrease in the risk of repeat revascularization may shift the focus of clinical attention after percutaneous intervention from restenosis prevention toward the institution of more aggressive disease-modifying strategies. Although the present study has limitations related to its nonrandomized nature, the bare stent and SES groups had comparable baseline characteristics. Creatinine clearance was calculated because it has been shown to correlate well with actual values18 and provide a better estimate of renal function than serum creatinine alone.30 Complete data for preprocedural creatinine clearance calculation were available for 86% of patients, which may have introduced a selection bias in the analysis. Nevertheless, patients who had been excluded from this study due to missing creatinine clearance data had a 1-year mortality rate of Пі5.5%, which was intermediate between patients who had renal impairment and those who did not, indicating that those who had been excluded likely included similar proportions of patients who had renal impairment and those who did not. 1. Al Suwaidi J, Reddan DN, Williams K, Pieper KS, Harrington RA, Califf RM, Granger CB, Ohman EM, Holmes DR Jr. Prognostic implications of abnormalities in renal function in patients with acute coronary syndromes. Circulation 2002;106:974 –980. 2. Shlipak MG, Simon JA, Grady D, Lin F, Wenger NK, Furberg CD. Renal insufficiency and cardiovascular events in postmenopausal women with coronary heart disease. J Am Coll Cardiol 2001;38:705–711. 3. Mann JF, Gerstein HC, Pogue J, Bosch J, Yusuf S. Renal insufficiency as a predictor of cardiovascular outcomes and the impact of ramipril: the HOPE randomized trial. Ann Intern Med 2001;134:629 – 636. 4. Gibson CM, Pinto DS, Murphy SA, Morrow DA, Hobbach HP, Wiviott SD, Giugliano RP, Cannon CP, Antman EM, Braunwald E. Association of creatinine and creatinine clearance on presentation in acute myocardial infarction with subsequent mortality. J Am Coll Cardiol 2003;42:1535–1543. 5. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39:S1–S246. 6. Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL, McCullough PA, Kasiske BL, Kelepouris E, Klag MJ, et al. Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention. Circulation 2003;108:2154 –2169. 7. Gruberg L, Dangas G, Mehran R, Mintz GS, Kent KM, Pichard AD, Satler LF, Lansky AJ, Stone GW, Leon MB. Clinical outcome following percutaneous coronary interventions in patients with chronic renal failure. Catheter Cardiovasc Interv 2002;55:66 –72. 8. Szczech LA, Best PJ, Crowley E, Brooks MM, Berger PB, Bittner V, Gersh BJ, Jones R, Califf RM, Ting HH, et al. Outcomes of patients with chronic renal insufficiency in the bypass angioplasty revascularization investigation. Circulation 2002;105:2253–2258. 9. Dixon SR, O’Neill WW, Sadeghi HM, Stone GW, Brodie B, Cox DA, Garcia E, Mattos L, Grines LL, Boura JA, et al. Usefulness of creatinine clearance in predicting early and late death after primary angioplasty for acute myocardial infarction. Am J Cardiol 2003;91:1454 –1457. 10. Best PJ, Lennon R, Ting HH, Bell MR, Rihal CS, Holmes DR, Berger PB. The impact of renal insufficiency on clinical outcomes in patients undergoing percutaneous coronary interventions. J Am Coll Cardiol 2002;39:1113–1119. 11. Rubenstein MH, Harrell LC, Sheynberg BV, Schunkert H, Bazari H, Palacios IF. Are patients with renal failure good candidates for percutaneous coronary revascularization in the new device era? Circulation 2000;102:2966 –2972. 12. Herzog CA, Ma JZ, Collins AJ. Comparative survival of dialysis patients in the United States after coronary angioplasty, coronary artery stenting, and coronary artery bypass surgery and impact of diabetes. Circulation 2002;106:2207–2211. 13. Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, O’Shaughnessy C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, et al. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315–1323. 14. Morice MC, Serruys PW, Sousa JE, Fajadet J, Ban Hayashi E, Perin M, Colombo A, Schuler G, Barragan P, Guagliumi G, et al. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med 2002;346:1773–1780. 15. Lemos PA, Serruys PW, van Domburg RT, Saia F, Arampatzis CA, Hoye A, Degertekin M, Tanabe K, Daemen J, Liu TK, et al. Unrestricted utilization of sirolimus-eluting stents compared with conventional bare stent implantation in the “real world”: the Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital (RESEARCH) registry. Circulation 2004;109:190 –195. 16. Lemos PA, Lee C, Degertekin M, Saia F, Tanabe K, Arampatzis CA, Hoye A, van Duuren M, Sianos G, Smits PC, et al. Early outcome after sirolimuseluting stent implantation in patients with acute coronary syndromes. Insights from the Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital (RESEARCH) registry. J Am Coll Cardiol 2003;41:2093–2099. 17. Lemos PA, Serruys PW, van Domburg RT, Saia F, Arampatzis CA, Hoye A, Degertekin M, Tanabe K, Daemen J, Liu TKK, et al. Unrestricted utilization of sirolimus-eluting stents compared to conventional bare stent implantation in the “real world”: The Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital (RESEARCH) registry. Circulation 2004;109:190 –195. 18. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31– 41. 19. Shlipak MG, Fried LF, Crump C, Bleyer AJ, Manolio TA, Tracy RP, Furberg CD, Psaty BM. Elevations of inflammatory and procoagulant biomarkers in elderly persons with renal insufficiency. Circulation 2003;107:87–92. 20. Henry RM, Kostense PJ, Bos G, Dekker JM, Nijpels G, Heine RJ, Bouter LM, Stehouwer CD. Mild renal insufficiency is associated with increased cardiovascular mortality: the Hoorn Study. Kidney Int 2002;62:1402–1407. 21. Dzurik R, Spustova V, Janekova K. The prevalence of insulin resistance in kidney disease patients before the development of renal failure. Nephron 1995; 69:281–285. 22. Sechi LA, Zingaro L, De Carli S, Sechi G, Catena C, Falleti E, Dell’Anna E, Bartoli E. Increased serum lipoprotein(a) levels in patients with early renal failure. Ann Intern Med 1998;129:457– 461. 23. Stuveling EM, Hillege HL, Bakker SJ, Gans RO, De Jong PE, De Zeeuw D. C-reactive protein is associated with renal function abnormalities in a nondiabetic population. Kidney Int 2003;63:654 – 661. 24. Holvoet P, Donck J, Landeloos M, Brouwers E, Luijtens K, Arnout J, Lesaffre E, Vanrenterghem Y, Collen D. Correlation between oxidized low density lipoproteins and von Willebrand factor in chronic renal failure. Thromb Haemost 1996;76:663– 669. CORONARY ARTERY DISEASE/RENAL FUNCTION AFFECTS OUTCOMES AFTER SIROLIMUS STENTS 171 25. Thambyrajah J, Landray MJ, McGlynn FJ, Jones HJ, Wheeler DC, Townend JN. Abnormalities of endothelial function in patients with predialysis renal failure. Heart 2000;83:205–209. 26. Parsons DS, Reaveley DA, Pavitt DV, Brown EA. Relationship of renal function to homocysteine and lipoprotein(a) levels: the frequency of the combination of both risk factors in chronic renal impairment. Am J Kidney Dis 2002;40:916 –923. 27. Otsuka M, Hayashi Y, Ueda H, Imazu M, Kohno N. Predictive value of preprocedural fibrinogen concerning coronary stenting. Atherosclerosis 2002; 164:371–378. 172 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 28. Benchimol D, Bonnet J, Benchimol H, Drouillet F, Duplaa C, Couffinhal T, Desgranges C, Bricaud H. Biological risk factors for restenosis after percutaneous transluminal coronary angioplasty. Int J Cardiol 1993;38:7–18. 29. Schoebel FC, Gradaus F, Ivens K, Heering P, Jax TW, Grabensee B, Strauer BE, Leschke M. Restenosis after elective coronary balloon angioplasty in patients with end stage renal disease: a case-control study using quantitative coronary angiography. Heart 1997;78:337–342. 30. Bostom AG, Kronenberg F, Ritz E. Predictive performance of renal function equations for patients with chronic kidney disease and normal serum creatinine levels. J Am Soc Nephrol 2002;13:2140 –2144. JANUARY 15, 2005 Determinants of 30-Day Adverse Events Following Saphenous Vein Graft Intervention With and Without a Distal Occlusion Embolic Protection Device Gregory R. Giugliano, MD, SM, Richard E. Kuntz, MD, MSc, Jeffrey J. Popma, MD, Donald E. Cutlip, MD, and Donald S. Baim, MD, on behalf of the Saphenous Vein Graft Angioplasty Free of Emboli Randomized (SAFER) Trial Investigators Distal balloon occlusion was approved as a means of embolic protection during saphenous vein graft intervention based on its ability to decrease major adverse clinical events (MACEs) by 42% in the 801-patient Saphenous Vein Graft Angioplasty Free of Emboli Randomized (SAFER) trial. However, the cost and technical complexity of this device have limited its widespread use and prompted some to avoid its use in cases that appear at “low risk” for complications. If predictors of MACEs and their potential decrease by distal balloon occlusion could be identified, this would have important clinical implications in this challenging population. We therefore used standard demographic and angiographic variables and 2 new angiographic markers (extent of graft degeneration and estimated volume of plaque in the target lesion) to construct multivariable logistic regres- sion models of 30-day of MACEs in the SAFER trial. Independent correlates of increased 30-day MACEs were more extensive vein graft degeneration (p ‫؍‬ 0.0001) and bulkier lesions (larger estimated plaque volume, p ‫ ؍‬0.0005). Use of a distal balloon occlusion device was independently predictive of lower 30-day rates of MACE (p ‫ ؍‬0.01), with uniform benefit across risk strata (no significant interaction between device use and independent angiographic risk factors). Thus, the risk of 30-day MACEs after percutaneous intervention in aortocoronary saphenous vein grafts is increased in more diffusely diseased grafts and in bulkier lesions, but a significant benefit of the GuardWire was seen across all levels of MACE risk rather than just those perceived to be at highest risk. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:173–177) lthough distal atheroemboli have been implicated as the cause of major adverse clinical events A (MACEs) during stenting of saphenous vein graft vessel diameters of 3 to 6 mm, ejection fraction Пѕ25%, and serum creatinine levels ПЅ2.5 mg/dl. The primary end point was 30-day MACEs, defined as the composite of death, myocardial infarction (creatine kinase-MB level Пѕ3 times the upper limit of normal), emergency bypass surgery, or target vessel revascularization Х…30 days of the index procedure. Detailed descriptions of the SAFER, including patient and lesion characteristics, participating hospitals, data gathering, analysis process, and end point definitions, have been previously published.8 For the present analysis, standard clinical and angiographic variables were supplemented by 2 new ad hoc angiographic variables derived from the core laboratory: SVG percent degeneration score, which was defined in quartiles as the portion (0% to 25%, 26% to 50%, 51% to 75%, Пѕ75%) of the treated graft with evident (Пѕ20%) disease, and estimated lesion plaque volume, which was defined as the volume of a cylinder whose diameter was equal to the reference vessel diameter and whose length was the shoulder-to-shoulder lesion length, as measured by the core laboratory, minus the volume of a cylinder of the same length and a diameter equal to the measured minimal luminal diameter (Appendix A). The former captures the state of graft disease outside the target lesion, and the latter captures the amount of plaque at risk for embolization when the lumen of the target lesion is normalized by stent placement. (SVG) lesions,1– 8 only 1 study9 has reported correlates of MACEs after SVG percutaneous coronary intervention that used distal embolic protection. We therefore conducted this analysis of a selected set of angiographic and demographic characteristics in the Saphenous Vein Graft Angioplasty Free of Emboli Randomized (SAFER) trial to identify the potential determinants of MACEs at 30 days and their interaction with a randomized balloon occlusion device. METHODS The SAFER trial8 enrolled 801 patients between June 1999 and August 2000 in a multicenter study in which patients who underwent SVG graft intervention were randomized to undergo stenting over a conventional guidewire or over the GuardWire distal protection device (Medtronic, Minneapolis, Minnesota). All patients had coronary ischemia, culprit lesions of Пѕ50% diameter stenosis by visual estimate, reference From the Divisions of Cardiology and Clinical Biometrics, Brigham and Women’s Hospital, Boston, Massachusetts; and the Division of Cardiology, Beth Israel Deaconess Medical Center, Boston, Massachusetts. Manuscript received August 13, 2004; revised manuscript received and accepted August 31, 2004. Address for reprints: Donald S. Baim, MD, Brigham and Women’s Hospital, One Brigham Circle, Boston, Massachusetts 02115. E-mail: [email protected]. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.090 173 blocker, longer lesion, angiographic evidence of thrombus, SVG angulaUnivariable Predictors of 30-day MACE Odds Ratio SE p Value tion Пѕ45В°, larger reference vessel diameter, postprocedural in-stent perAge (yrs) 1.00 0.017 0.99 cent diameter stenosis, and treatment Male gender 1.02 0.439 0.96 History of diabetes mellitus 0.68 0.383 0.31 assignment (no distal protection). History of myocardial infarction 1.28 0.359 0.49 The 2 new angiographic variables History of hypertension 0.86 0.376 0.68 derived from the core laboratory, History of hyperlipidemia 1.79 0.460 0.20 SVG degeneration score and estiSmoking Х…1 year 1.41 0.512 0.50 Conventional stenting with predilatation 3.32 0.382 0.002 mated plaque volume, also proved to Vein graft age 1.03 0.033 0.33 be significant univariable associates Use of glycoprotein IIb/IIIa inhibitors 1.67 0.230 0.03 of 30-day MACEs (Table 1). Lesion length 1.06 0.015 0.0001 SVG percent degeneration was Thrombus 1.89 0.343 0.06 categorized in quartiles based on the Angulation Пѕ45В° 3.11 0.501 0.02 Reference vessel diameter 1.72 0.237 0.02 axial proportion (0% to 25%, 26% to Postprocedural in-stent percent diameter stenosis 1.03 0.010 0.002 50%, 51% to 75%, 76% to 100%), SVG degeneration score* 1.90 0.168 0.0001 measured by the core laboratory, of †Plaque volume 1.01 0.0006 0.0001 the graft that showed angiographic *Four categories represent percent graft with angiographic evidence of atherosclerotic plaque (0% to evidence of atherosclerotic disease. 25%, 26% to 50%, 51% to 75%, 76% to 100%). The number of lesions in each group †2 2 Determined by the equation: вђІ(lesion length) [(RVD/2) ПЄ (MLD/2) ], where вђІ equals 3.14, RVD is and the corresponding incidence of reference vessel diameter in millimeters, and MLD is minimal luminal diameter of the lesion in millimeters. MACEa are presented in Table 2, which associates higher degeneration scores with a higher incidence of Statistical analysis: Continuous variables were com- MACEs in the control group (p П 0.0002 for trend by pared with Student’s t test when variables were nor- chi-square test) and GuardWire group (p П 0.0001) mally distributed and with Wilcoxon’s rank-sum test and associates the use of a GuardWire with decreased when variables were not normally distributed. Binary MACE at all levels of graft degeneration score. variables were compared with chi-square test with Estimated lesion plaque volume was calculated for normal approximation or Fisher’s exact test when each target lesion, and the resulting values were diappropriate. A 2-tailed p value Х…0.05 was considered vided into equal quartiles (minimum to maximum statistically significant. All data were analyzed based plaque volumes within each quartile: 8 to 66, 67 to on the intention-to-treat principle. Logistic regression 108, 109 to 178, and 179 to 1,109 mm3, respectively). was used to identify predictors of 30-day MACEs. Number and percentage of MACEs in each quartile of Selected demographic and angiographic characteris- plaque volume are listed in Table 3. Lesions with tics recorded in the SAFER trial in addition to graft larger estimated plaque volume were associated with a degeneration score and lesion plaque volume were higher incidence of MACEs in the control group (p П considered in univariable analyses as potential predic- 0.0012 for trend by chi-square test) and GuardWire tors. A candidate list of significant variables (p ПЅ0.15) group (p П 0.0003). As presented in Table 3, use of from the logistic univariable models was added to the GuardWire was associated with decreased MACEs those variables of particular clinical interest into a at each level of estimated plaque volume. stepwise final multivariable model in which a 2-sided Multivariable determinants of MACEs: Multivariable p value Х…0.05 was considered statistically significant. modeling identified the more conventional variable, leAn interaction was sought between randomized treat- sion length, as collinear with SVG degeneration score ment assignment (distal occlusion embolic protection and not as strong a predictor; thus, lesion length and device) and final multivariable main effects. All sta- reference vessel diameter were excluded from the multistical analyses were performed with SAS 6.12 (SAS tivariable model. Therefore, only greater SVG degenerInstitute, Cary, North Carolina). ation score, larger plaque volume, and lack of GuardWire management were found to be significant independent determinants of 30-day MACEs (Table 4). RESULTS SAFER trial: Between June 1999 and August 2000, Interaction terms between the 2 vein graft independent 801 patients who had 875 lesions were enrolled in the determinants and treatment assignment were then examSAFER trial. The trial showed a 42% relative decrease ined to explore whether the risk conferred by degenerain 30-day MACEs (16.5% for control patients vs 9.6% tion and plaque volume was modified by assignment to for patients who received the GuardWire, p П 0.004), the GuardWire distal protection device. Those 2 interacindicating superior safety of stent implantation over tion terms were rejected, suggesting that the main effect the GuardWire compared with a standard guidewire. of the GuardWire distal protection device is uniformly The 2 arms of this trial were well matched for all additive across the other 2 risk prediction variables. That is, MACEs increase with more diffuse SVG degenerabaseline clinical and angiographic variables. Univariable correlates of MACEs: Factors linked to tion and larger estimated lesion plaque volume in the an increased risk of MACEs included use of balloon actual analysis groups (Figures 1 and 2) and the model predilatation, use of a platelet glycoprotein IIb/IIIa predictions (Figure 3), with significant relative decreased TABLE 1 List of Candidate Variables 174 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 TABLE 2 Adverse Event Rates and Treatment Effect by Saphenous Vein Graft Degeneration Score Rate of MACE SVG Degeneration n Guidewire GuardWire 0–25% 26–50% 51–75% 76–100% 435 208 78 45 9.5% 22.9% 32.5% 22.7% 4.7% 12.5% 21.1% 21.7% OR (95% CI) 0.46 0.48 0.55 0.94 (0.21–1.0) (0.23–1.0) (0.20–1.54) (0.23–3.85) Relative Decrease in MACE p Value* 51% 45% 35% 4% 0.06 0.07 0.31 0.99 Relative Decrease in MACE p Value* 49% 77% 21% 32% 0.37 0.01 0.54 0.21 *For trend, p П 0.0002 and p П 0.0001. CI П confidence interval; OR П odds ratio. TABLE 3 Adverse Event Rates and Treatment Effect by Plaque Volume Quartile 30-Day Rate of MACE Plaque Volume Quartile n Guidewire GuardWire 1 2 3 4 189 190 189 188 8.4% 13.0% 16.7% 25.3% 4.3% 3.1% 13.1% 17.2% OR (95% CI) 0.48 0.21 0.76 0.61 (0.14–1.7) (0.06–0.77) (0.34–1.69) (0.30–1.25) *For trend, p П 0.0012 and p П 0.0003. Abbreviations as in Table 2. TABLE 4 Multivariable Regression Analysis of Predictors of 30-Day Rates of Main Adverse Clinical Events Predictors Odds Ratio Treatment assignment (1 П GuardWire, 0 П Guidewire) SVG degeneration Score (1–4) Plaque volume (continuous) SE p Value 0.55 0.23 0.01 1.56 1.002 0.12 0.0007 0.0001 0.0005 FIGURE 1. Percent SVG degeneration was categorized based on the axial proportion determined by the core laboratory (0% to 25%, 26% to 50%, 51% to 75%, 76% to 100%) of each treated graft that showed angiographic evidence of atherosclerotic disease. The decrease in relative risk of MACES using the GuardWire versus conventional guidewire is shown for each level of degeneration. More diffusely diseased SVGs convey a higher overall risk of MACES, with somewhat less relative benefit of GuardWire protection. rates of MACE by the GuardWire, are seen across levels of degeneration or plaque volume. DISCUSSION The SAFER trial8 showed that the use of a distal occlusion/aspiration embolic protection device (MedtronicPercuSurge GuardWire and associated Export catheter) during SVG intervention decreased incidences of 30-day MACEs (by 42%, from 19% to 9%, p ПЅ0.001) and no-reflow (by 60%, from 8.3% to 3.3%, p ПЅ0.001). This finding implicates distal embolic debris as the proximate cause of these 2 complications during SVG intervention. Nonetheless, the clinical community has been slow to adopt this device due in part to its high cost (Пѕ$1,000) and the moderate technical complexity that requires careful coordination of 2 operators to minimize the obligate 4- to 6-minute ischemic occlusion time. One rationalization of that limited adoption might be the assertion that certain techniques for performing the procedure (e.g., use of direct stenting, use of platelet glycoprotein IIb/IIIa receptor blockers) or ways of selecting patients at low enough risk for embolic complications might obviate distal protection. The first variable we tested was percent graft length that showed evident disease (angiographic SVG degeneration score; Figure 1). In the control group, MACEs increased when Пѕ25% of the graft was diseased (25%, 40 of 158) compared with when ПЅ25% of the graft was diseased (9.5%, 21 of 220). However, the relative decrease in MACEs using the GuardWire was greatest (51%, from 9.5% to 4.7%) in the group with the lowest SVG score. The observation of least benefit in the group with the highest SVG score (4%, from 22.7% to 21.7%) may suggest an CORONARY ARTERY DISEASE/PREDICTORS OF MACE AFTER SVG PERCUTANEOUS CORONARY INTERVENTION 175 FIGURE 2. Relative risk decrease of MACES when using the GuardWire versus conventional guidewire for each quartile of plaque volume. The relative benefit of GuardWire protection is greatest in the lower quartiles. Larger plaque volumes indicate a higher overall risk of MACES, which is decreased significantly by distal protection. with a diameter equal to the measured minimal luminal diameter. This indirect measurement of plaque burden proved to be a very potent predictor of complications, with monotonic increases in MACEs in the conventional guidewire group (from 8.4% to 25.3%, p П 0.0012 for trend) from lowest to highest quartile of estimated lesion plaque volume. Moreover, there was a substantial decrease in MACEs when using the GuardWire in even the 2 lowest quartiles of estimated plaque volume (49%, from 8.4% to 4.3%, and 77%, from 13.0% to 3.0%, in those quartiles, respectively, p П 0.01; Figure 2). The somewhat lower treatment effect in the 2 highest quartiles of estimated lesion plaque volume (21% and 32%, respectively) may suggest overwhelming embolic load, inefficiency of the export catheter with bulkier loads, or sufficient plaque to protrude through the stent struts and embolize when normal flow is restored after the distal protection device has been removed. This additional analysis of the 801-patient SAFER trial thus shows that even the lowest risk grafts have nontrivial (4% to 11%) event rates without distal protection and that those risks are decreased across the board using distal embolic protection. Over and above these findings, the present study has identified important new angiographic parameters that can be used to compare expected baseline rates of MACEs of patient cohorts across various device trials, including distal filters and proximal occlusion aspiration devices. APPENDIX A FIGURE 3. Predicted rates of MACES between control and treatment (GuardWire) groups based on 2 predictor variables (plaque volume and percent SVG degeneration) in this analysis calculated with the equation presented in Appendix B. Use of distal embolic protection provides treatment benefit across all levels of risk. overwhelming embolic load or offsetting device-produced emboli when there is no relatively disease-free landing zone for the GuardWire. Beyond the state of degeneration of the graft as a whole, we hypothesized that the amount of debris available for embolization from the lesion site itself might also be a predictor of potential embolic load and consequent MACEs. Without intravascular ultrasound, no actual measurement of lesion volume could be ascertained. However, we chose to calculate an estimated plaque volume as the volume of a cylinder whose diameter was equal to the vessel reference diameter and whose length was equal to the lesion length minus the volume of a cylinder of equal length 176 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 Angiographic variable definitions: All procedural angiograms were reviewed according to standard morphologic criteria. Lesion length was defined as the axial extent of the lesion that contained a shoulder-toshoulder luminal decrease by Х†20%. The degree of graft degeneration (Пѕ20% luminal irregularity) throughout the SVG was estimated as percent total SVG length and reported in quartiles as 0% to 25%, 26% to 50%, 51% to 75%, and 75% to 100%. Estimated lesion plaque volume was defined as the volume of a cylinder whose diameter was equal to the reference vessel diameter and whose length was the shoulder-to-shoulder lesion length as measured by the core laboratory. Plaque volume was determined by this equation: вђІ(lesion length)[(RVD/2)2 ПЄ (MLD/2)2], where вђІ equals 3.14, RVD is the reference vessel diameter in millimeters, and MLD is the minimal luminal diameter of the lesion in millimeters. The resultant patient lesion volumes were divided into 4 quartiles that contained equal numbers of patients. APPENDIX B Model-predicted rates of MACEs: The probability of MACEs based on the final multivariable model was calculated for each quartile of SVG degeneration score and plaque volume. For SVG degeneration score, a number from 1 to 4 was entered based on percent degeneration as previously defined. Plaque volume was derived as a continuous variable and used as such for multivariable modeling. For purposes of establishing predicted rates of MACE, plaque volume was categorized by quartiles, and the mean value within each quartile was used in the following formula: Probability of MACE e(ПЄ2.44ПЄ0.59Н“treatment assignmentН”П©0.45Н“Degeneration ScoreН”П©0.0024Н“plaque volumeН”) . П 1 П© e(ПЄ2.44ПЄ0.59Н“treatment assignmentН”П©0.45Н“Degeneration ScoreН”П©0.0024Н“plaque volumeН”) JANUARY 15, 2005 1. Grube E, Gerckens U, Yeung AC, Rowold S, Kirchhof N, Sedgewick J, Yadav JS, Stertzer S. Prevention of distal embolization during coronary angioplasty in saphenous vein grafts and native vessels using porous filter protection. Circulation 2001;104:2436 –2441. 2. Webb JG, Carere RG, Virmani R, Baim D, Teirstein PS, Whitlow P, McQueen C, Kolodgie FD, Buller E, Dodek A, Mancini GB, Oesterle S. Retrieval and analysis of particulate debris after saphenous vein graft intervention. J Am Coll Cardiol 1999;34:468 – 475. 3. Lefkovits J, Holmes DR, Califf RM, Safian RD, Pieper K, Keeler G, Topol EJ. Predictors and sequelae of distal embolization during saphenous vein graft intervention from the CAVEAT-II trial. Coronary Angioplasty Versus Excisional Atherectomy Trial. Circulation 1995;92:734 –740. 4. Waksman R, Douglas JS Jr, Scott NA, Ghazzal ZM, Yee-Peterson J, King SB III. Distal embolization is common after directional atherectomy in coronary arteries and saphenous vein grafts. Am Heart J 1995;129:430 – 435. 5. Waksman R, Weintraub WS, Ghazzal Z, Scott NA, Shen Y, King SB III, Douglas JS Jr. Short- and long-term outcome of narrowed saphenous vein bypass graft: a comparison of Palmaz-Schatz stent, directional coronary atherectomy, and balloon angioplasty. Am Heart J 1997;134:274 –281. 6. Hong MK, Mehran R, Dangas G, Mintz GS, Lansky AJ, Pichard AD, Kent KM, Satler LF, Stone GW, Leon MB. Creatine kinase-MB enzyme elevation following successful saphenous vein graft intervention is associated with late mortality. Circulation 1999;100:2400 –2405. 7. Waksman R, Ghazzal ZM, Baim DS, Steenkiste AR, Yeh W, Detre KM, King SB III. Myocardial infarction as a complication of new interventional devices. Am J Cardiol 1996;78:751–756. 8. Baim DS, Wahr D, George B, Leon MB, Greenberg J, Cutlip DE, Kaya U, Popma JJ, Ho KK, Kuntz RE. Randomized trial of a distal embolic protection device during percutaneous intervention of saphenous vein aorto-coronary bypass grafts. Circulation 2002;105:1285–1290. 9. Stone GW, Rogers C, Hermiller J, Feldman R, Hall P, Haber R, Masud A, Cambier P, Caputo RP, Turco M, et al. Randomized comparison of distal protection with a filter-based catheter and a balloon occlusion and aspiration system during percutaneous intervention of diseased saphenous vein aorto-coronary bypass grafts. Circulation 2003;108:548 –553. CORONARY ARTERY DISEASE/PREDICTORS OF MACE AFTER SVG PERCUTANEOUS CORONARY INTERVENTION 177 Artificial Neural Network Modeling of Stress Single-Photon Emission Computed Tomographic Imaging for Detecting Extensive Coronary Artery Disease J. Scott Allison, MD, Jaekyeong Heo, MD, and Ami E. Iskandrian, MD Single-photon emission computed tomographic imaging is a useful, noninvasive method to detect coronary artery disease (CAD). We tested the hypothesis that artificial neural network modeling could predict CAD extent better than visual interpretation; 109 patients who underwent stress single-photon emission computed tomography and coronary angiography were selected. Twenty patients who had a <5% probability of CAD were also selected for calculation of normalcy rate. A model was trained for each vessel. Stress images were decreased to 25 points by pixel averaging the polar map. The model output was 1 for vessel stenosis >60% and 0 otherwise. Model sensitivities were 92% (55 of 60) for left anterior descending artery versus 62% (37 of 60) for visual interpretation (p ‫ ؍‬0.0002), 69% (20 of 29) for left circumflex artery versus 55% for visual interpretation (p ‫ ؍‬0.30), and 94% (45 of 48) for right coronary artery versus 78% for visual interpretation (p ‫ ؍‬0.024). Model specificities and normalcy rates were 78% and 85% for the left anterior descending artery, 93% and 100% for the left circumflex artery, and 85% and 90% for the right circumflex artery, respectively. Single-vessel CAD was predicted in 27 of 28 patients (96%) by modeling versus 23 of 28 patients (82%) by visual interpretation (p ‫ ؍‬0.11). Multivessel CAD was correctly predicted in 30 of 46 patients (65%) by modeling versus 16 of 46 patients (35%) by visual interpretation (p ‫ ؍‬0.004). Thus, artificial neural network models can predict CAD from stress single-photon emission computed tomographic images when using separate models for the 3 major epicardial vessels. Because of their high sensitivity and specificity in detecting extensive CAD, these models have great promise as an aid to correctly identify patients at high risk for CAD. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:178 –181) ecause of the powerful ability to model complex input/output relations, nonlinear artificial neural netB work models can be used to map single-photon emission raphy separated by Х…3 months between January 2001 and April 2003 were selected. Patients who had previous coronary revascularization or any event between studies were excluded. Based on Х†60% diameter stenosis by coronary angiography, 17 patients had 3-vessel CAD, 29 had 2-vessel CAD, 28 had 1-vessel CAD, and 35 had no significant CAD. Patients who had left main disease (n П 8) were considered to have 2-vessel disease (LAD and left circumflex). Pertinent data on the study group are presented in Table 1. Twenty patients who had a ПЅ5% probability of CAD and normal stress SPECT studies by visual interpretation were selected to calculate a normalcy rate. These 20 patients did not undergo coronary angiography. Stress and imaging procedure: Patients underwent a stress treadmill exercise test according to Bruce’s protocol (n П 37) or received adenosine (n П 67) or dobutamine (n П 5). Patients were instructed to not use вђ¤ blockers and long-acting nitrates for 24 to 48 hours before the study. Methods of stress testing, image acquisition and processing, and image interpretation have previously been described.6 All images were acquired in a gated mode 60 minutes after injection of technetium 99m tetrofosmin with a dual-head detector. Although attenuation correction was used when available for visual interpretation, only data from the uncorrected images were used for training and validating the neural networks. Same-day or separate-day protocols during stress and at rest were used as appropriate based on patients’ weight and body computed tomographic (SPECT) perfusion images to coronary narrowings detected by coronary angiography. Previous studies have indicated that these models detect coronary artery disease (CAD) at least as effectively as human experts.1–5 However, these studies combined right and left circumflex coronary arteries into a single vascular territory, produced models from images during stress and at rest with only a small number of inputs from the stress data, or used models with a large number of inputs and a small number of training cases. This study tested the hypothesis that neural network models could predict extensive CAD by using separate models for the right, left circumflex, and left anterior descending (LAD) arteries better than visual interpretation of SPECT images. METHODS Study patients: One hundred nine patients who had stress SPECT perfusion studies and coronary angiog- From the Division of Cardiovascular Diseases, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama. Manuscript received June 24, 2004; revised manuscript received and accepted September 3, 2004. Address for reprints: Ami E. Iskandrian, MD, Division of Cardiovascular Diseases, University of Alabama at Birmingham, 318 LHRB, 1900 University Boulevard, Birmingham, Alabama 35294-0006. E-mail: [email protected]. 178 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.003 averaging the polar map with 25 equal-area sectors, as shown in Figure 1. Before averaging, pixels were scaled from 0 to 1, with 1 indicating the most perfuAge (yrs) 62 П® 11 sion and 0 indicating the least. These 25 data points Men 64 (59%) were used as inputs to the model. Because this study did Systemic hypertension 83 (76%) not attempt to identify ischemia, perfusion data at rest Diabetes mellitus 38 (35%) Hyperlipidemia (LDL Пѕ100 or on therapy) 40 (37%) were not included in the model inputs. The model output Previous myocardial Infarction 13 (12%) was 1 for vessel stenosis Пѕ60% and 0 otherwise. Ejection fraction (%) 52 П® 14 Model creation: Separate models were created and Chest pain during stress 18 (17%) trained for each of the 3 vessels. The model architecPositive electrocardiogram during stress 30 (28%) ture (Figure 2) used the logistic function, 1.0/[1.0 П© LDL П low-density lipoprotein. e(ПЄx)], as the transfer function, f. The same inputs were used for all models as previously described, with the output based on the existence of a significant lesion in the vessel modeled. The model’s input layer contained 25 neurons that corresponded to the 25 inputs and 1 output layer neuron. The number of hidden layer neurons was optimized for each model, with the LAD model containing 8, the left circumflex model containing 4, and the right model containing 5. EasyNN Plus 3.0 (Neural Planner Software, Cheadle, United Kingdom) was used for model creation, training, and validation. Each model was trained with back-propagation. Data from the 109 patients were split randomly into 66% training (72 of FIGURE 1. Stress polar map divided into 25 equal-area sectors. 109) and 33% validation (37 of 109) datasets for each model. More patients were assigned to the training data set than to the validation data set because of the relatively small number of patients.7 Statistical analysis: Data are presentedby model (vessel) and by extent of disease (single vessel or multivessel). Model results were compared with visual interpretation. Model predictions by vessel and extent of disease were compared with visual interFIGURE 2. Multilayer perception neural network. f ‫ ؍‬transfer function; w ‫ ؍‬synaptic pretation by Fisher’s exact test. A p weight. value ПЅ0.05 was considered statistically significant. TABLE 1 Pertinent Demographic Data on Study Population (n П 109) habitus. Perfusion defects in the anterior wall or septum were considered to represent LAD disease, those in the inferior wall were considered to represent right coronary artery disease, and those in the lateral wall were considered to represent left circumflex disease. Polar maps were generated with AutoQUANT 4.2.1 (ADAC, Sunnyvale, California). Coronary angiography was performed according to standard techniques. Significant stenosis was defined as Х†60% diameter stenosis in the LAD, left circumflex, or right coronary arteries. Angiograms and the SPECT images were read without knowledge of the other results. Model inputs and output: Stress SPECT perfusion images were decreased to 25 data points by pixel RESULTS There were 60 diseased LADs, 48 diseased right, and 29 diseased left circumflex coronary arteries. Table 2 lists the sensitivity of the model for detecting CAD compared with visual interpretation. These results are presented graphically in Figures 3 and 4. In patients who had no significant CAD, all 3 models had correct predictions in 21 of 35 patients, with a specificity of 60% (visual interpretation 29%, p П 0.0095). The low specificity most likely reflected post-test referral bias, and most patients had CAD but ПЅ60% stenosis, which could be physiologically significant. Results in the training and validation groups were similar (data not shown). CORONARY ARTERY DISEASE/ARTIFICIAL NEURAL NETWORK MODELING 179 The most common method of SPECT interpretation is visual, which often is supplemented by semiquantiModel Visual p Value tative algorithms. These algorithms include the use of a summed stress score, Location of disease LAD 92% (55/60) 62% (37/60) 0.0002 polar maps, or circumferential proLeft circumflex 69% (20/29) 55% (16/29) 0.2964 files.8 –13 The use of gated imaging and Right 94% (45/48) 78% (37/48) 0.0237 attenuation correction have improved No. of coronary artery narrowings diagnostic accuracy.14 –20 It is for these 1 96% (27/28) 82% (23/28) 0.1080 2 59% (17/29) 31% (9/29) 0.0402 reasons that SPECT imaging has been 3 76% (13/17) 41% (7/17) 0.0466 firmly established in detection of CAD 2/3 65% (30/46) 35% (16/46) 0.0040 and risk assessment. There are 2 issues that need further refinement, the ability to correctly identify patients at high risk (those with extensive CAD) and to improve the interpretation of less experienced readers. Multiple methods of automated interpretation of SPECT imaging have previously been studied. Most of these studies have used the polar map as input. Perfusion quantification applied to assigned vessel regions has been used to localize fixed and reversible defects to a vessel.8 –13 Expert systems have been developed that use a knowledge base of rules to interpret images.21 Case-based systems match a study patient to the best fit in a library of patients.22 These methods have the advantage that some justification can be made for the results produced. Garcia et al21 FIGURE 3. Sensitivity of the model (gray bars) versus that of vifound that interpretations of an expert system agree sual interpretation (black bars) for detecting CAD. well with expert visual interpretation and are as accurate when coronary angiography is used as the gold standard. Similarly, Khorsand et al22 showed that a case-based system has a diagnostic accuracy similar to visual interpretation and the use of normal limits. Neural networks are true “black box” models that contain complex interconnections of neurons with no interpretable intermediate data. However, because of their ability to perform complex pattern-recognition tasks, neural networks have the potential to outperform the previously described methods. Multiple previous studies have examined neural network models of myocardial perfusion images. Fujita et al1 showed that a neural network with 256 inputs could be trained to interpret SPECT polar map images better than a radiology FIGURE 4. Sensitivity of the model (gray bars) versus that of visual interpretation (black bars) for detecting extent of CAD. resident but worse than an expert radiologist. However, the study used a very large number of inputs, 256, with Twenty patients who had a ПЅ5% likelihood of CAD a small number of test cases, 71. The study also did not were used to determine normalcy rates. The LAD model contain any cases in which the angiographic results did correctly predicted 17 of 20 patients (85%), the left not agree with the visual interpretation of the SPECT circumflex model correctly predicted 20 of 20 patients images. Porenta et al2 reported that a neural network with (100%), and the right model correctly predicted 18 of 20 45 inputs from thallium-201 dipyridamole stress-redispatients (90%). The combined normalcy rate was 75% tribution planar images predicted significant CAD with a (15 of 20) for all 3 models. sensitivity of 51% and a specificity of 90% compared with angiography. The study examined 159 patients, 81 of whom underwent angiography, and considered 2 vasDISCUSSION The present results show that fully automated neural cular territories (LAD and combined right/left circumflex network models can be used to detect any CAD and arteries). The neural network was not significantly better extensive CAD with a high sensitivity and specificity than expert visual interpretation compared with angiogthat are as good as or better than visual interpretation raphy. A neural network trained to predict expert visual alone. Sensitivities were 96% for single-vessel disease interpretation performed better than the neural network and 65% for correctly identifying multivessel disease. trained to predict angiographic lesions. Hamilton et al3 These values were higher than those by visual interpre- examined 410 men and found that a neural network tation (82% and 35%, respectively). could be trained to detect hypoperfused segments in a TABLE 2 Model Sensitivity Versus Visual Interpretation in Detecting Coronary Artery Disease 180 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 polar map divided into 8 or 24 segments. Hypoperfused segments were defined as ПЅ2.5 SD of the expected mean count from a normal database. Coronary angiography was not used in this study. Lindahl et al4 reported that a neural network could be trained to predict angiographic LAD lesions significantly better than visual interpretation (77.2% vs 73% sensitivity, p П 0.038). The study examined 135 patients and considered 2 vascular territories (LAD and combined right/left circumflex arteries). The neural network trained for the right/left circumflex territory was not significantly better than visual interpretation. Inputs to the 2 neural networks were 30 Fourier components, which were obtained by applying a 2-dimensional Fourier transform to the polar maps during stress and at rest, which has a smaller root-mean-square error than pixel averaging. However, neural networks were not trained with both methods. Lindahl et al,5 in a separate study, found that physicians who classify polar maps benefit from an artificial neural network as measured by an increased area under the receiver-operating characteristic curve.5 The areas under the receiver-operating characteristic curve increased from 0.65 to 0.70 (p П 0.18) and from 0.79 to 0.82 (p П 0.006) for the 2 vascular territories in the same 135 patients assessed in the previous study. Based on the results of this and other studies, neural network models appear to be a promising method to improve diagnostic accuracy of SPECT imaging. This study has a few limitations that must be considered. First, coronary angiography was used as the gold standard for CAD, which may not be valid, especially in patients who have intermediate severity of stenoses.23–25 Further, no independent method to validate the physiologic significance of coronary stenosis such as flow reserve ratio, pressure gradients, or resistance was used. These methods have shown better correlation with visual interpretation of SPECT images than with coronary angiography.24 –27 We elected to use 60% diameter stenosis as a cutoff but even that is open to criticism. Second, all data obtained were retrospective with no prospective testing. Third, the sample size of 109 patients was relatively small for creation of a neural network, but the cases were diverse. Diverse cases are more desirable than number of patients because a neural network trained on data from many patients who do not have CAD will not perform well in patients who do have CAD. Further work needed in this area includes modeling left main disease and prospective testing. A model to improve the detection of significant left main disease would be valuable. Because of results from these 3 models, a left main model with good performance should be possible with enough training data. Prospective testing that measures improvement in visual interpretations with model advice is needed. 1. Fujita H, Katafuchi T, Uehara T, Nishimura T. Application of artificial neural network to computer-aided diagnosis of coronary artery disease in myocardial SPECT bull’s eye images. J Nucl Med 1992;33:272–276. 2. Porenta G, Dorffner G, Kundrat S, Petta P, Duit-Schedlmayer J, Sochor H. Automated interpretation of planar thallium-201-dipyridamole stress-redistribution scintigrams using artificial neural networks. J Nucl Med 1994;35:2041–2047. 3. Hamilton D, Riley PJ, Miola UJ, Amro AA. A feed forward neural network for classification of bull’s-eye myocardial perfusion images. Eur J Nucl Med 1995; 22:108 –115. 4. Lindahl D, Palmer J, Ohlsson M, Peterson C, Lundin A, Edenbrandt L. Automated interpretation of myocardial SPECT perfusion images using artificial neural networks. J Nucl Med 1997;38:1870 –1875. 5. Lindahl D, Lanke J, Lundin A, Palmer J, Edenbrandt L. Improved classifications of myocardial bull’s eye scintigrams with computer-based decision support system. J Nucl Med 1999;40:96 –101. 6. Iskandrian AE, Verani MS. Nuclear Cardiac Imaging: Principles and Applications, 3rd Ed. New York: Oxford University Press, 2003:137–162. 7. Patterson DW. Artificial Neural Networks: Theory and Applications. Singapore: Prentice-Hall, 1996:213. 8. Garcia EV, Van Train K, Maddahi J, Prigent F, Friedman J, Areeda J, Waxman A, Berman DS. Quantification of rotational thallium-201 myocardial tomography. J Nucl Med 1985;26:17–26. 9. DePasquale EE, Nody AC, DePuey EG, Garcia EV, Pilcher G, Bredlau C, Roubin G, Gober A, Gruentzig A, D’Amato P, Berger HJ. Quantitative rotational thallium-201 tomography for identifying and localizing coronary artery disease. Circulation 1988;77:316 –327. 10. Maddahi J, Van Train K, Prigent F, Garcia EV, Friedman J, Ostrzega E, Berman D. Quantitative single photon emission computed thallium-201 tomography for detection and localization of coronary artery disease: optimization and prospective validation of a new technique. J Am Coll Cardiol 1989;14:1689 –1699. 11. Klein JL, Garcia EV, DePuey EG, Campbell J, Taylor AT, Pettigrew RI, D’Amato P, Folks R, Alazraki N. Reversibility bull’s-eye: a new polar bull’s eye map to quantify reversibility of stress-induced SPECT thallium-201 myocardial perfusion defects. J Nucl Med 1990;31:1240 –1246. 12. Garcia EV, DePuey EG, Sonnemaker RE, Neely HR, DePasquale EE, Robbins WL, Moore WH, Heo J, Iskandrian AS, Campbell J. Quantification of the reversibility of stress-induced thallium-201 myocardial perfusion defects: a multicenter trial using bull’s-eye polar maps and standard normal limits. J Nucl Med 1990;31:1761–1765. 13. Van Train KF, Areeda J, Garcia EV, Cooke CD, Maddahi J, Kiat H, Germano G, Silagan G, Folks R, Berman DS. Quantitative same-day rest-stress technetium99m-sestamibi SPECT: definition and validation of stress normal limits and criteria for abnormality. J Nucl Med 1993;34:1494 –1502. 14. Chua T, Kiat H, Germano G, Maurer G, Van Train K, Friedman J, Berman D. Gated technetium 99m sestamibi for simultaneous assessment of stress myocardial perfusion, post-exercise regional ventricular function and myocardial viability. J Am Coll Cardiol 1994;23:1107–1114. 15. DePuey EG, Rozanski A. Using gated technetium-99m-sestamibi SPECT to characterize fixed myocardial defects as infarct or artifact. J Nucl Med 1995;36: 952–955. 16. Smanio PE, Watson DD, Segalla DL, Vinson EL, Smith WH, Beller GA. Value of gating of technetium-99m sestamibi single-photon emission computed tomographic imaging. J Am Coll Cardiol 1997;30:1687–1692. 17. Taillefer R, DePuey EG, Udelson JE, Beller GA, Latour Y, Reeves F. Comparative diagnostic accuracy of Tl-201 and Tc-99m sestamibi SPECT imaging (perfusion and ECG-gated SPECT) in detecting coronary artery disease in women. J Am Coll Cardiol 1997;29:69 –77. 18. Sharir T, Germano G, Kavanagh PB, Lai S, Cohen I, Lewin HC, Friedman JD, Zellweger MJ, Berman DS. Incremental prognostic value of post-stress left ventricular ejection fraction and volume by gated myocardial perfusion single photon emission computed tomography. Circulation 1999;100:1035–1042. 19. DePuey EG, Garcia EV. Optimal specificity of thallium-201 SPECT through recognition of imaging artifacts. J Nucl Med 1989;30:441– 449. 20. Links J, DePuey EG, Taillefer R, Becker LC. Attenuation correction and gating synergistically improve the diagnostic accuracy of myocardial perfusion SPECT. J Nucl Cardiol 2002;9:183–187. 21. Garcia EV, Cooke CD, Folks RD, Santana CA, Krawczynska EG, De Braal L, Ezquerra NF. Diagnostic performance of an expert system for the interpretation of myocardial perfusion SPECT studies. J Nucl Med 2001;42:1185–1191. 22. Khorsand A, Haddad M, Graf S, Moertl D, Sochor H, Porenta G. Automated assessment of dipyridamole 201Tl myocardial SPECT perfusion scintigraphy by case-based reasoning. J Nucl Med 2001;42:189 –193. 23. White CW, Wright CB, Doty DB, Hiratza LF, Eastham CL, Harrison DG, Marcus ML. Does visual interpretation of the coronary arteriogram predict the physiologic importance of a coronary stenosis? N Engl J Med 1984;310:819 – 824. 24. Miller DD, Donohue TJ, Younis LT, Bach RG, Aguirre FV, Wittry MD, Goodgold HM, Chaitman BR, Kern MJ. Correlation of pharmacologic 99mTcsestamibi myocardial perfusion imaging with poststenotic coronary flow reserve in patients with angiographically intermediate coronary artery stenoses. Circulation 1994;89:2150 –2160. 25. Pijls NH, De Bruyne B, Peels K, Van Der Voort PH, Bonnier HJ, Bartunek J, Koolen JJ. Measurement of myocardial fractional flow reserve to assess the functional severity of coronary artery stenosis. N Engl J Med 1996;334:1703–1708. 26. Donohue TJ, Kern MJ, Aguirre FV, Bach RG, Wolford T, Bell CA, Segal J. Assessing the hemodynamic significance of coronary artery stenoses: analysis of translesional pressure-flow velocity relationships in patients. J Am Coll Cardiol 1993;22:449 – 458. 27. Heller LI, Cates C, Popma J, Deckelbaum LI, Joye JD, Dahlberg ST, Villegas BJ, Arnold A, Kipperman R, Grinstead WC, et al. Intracoronary Doppler assessment of moderate coronary artery disease: comparison with 201Tl imaging and coronary angiography. FACTS Study Group. Circulation 1997;96:484 – 490. CORONARY ARTERY DISEASE/ARTIFICIAL NEURAL NETWORK MODELING 181 Incremental Prognostic Power of Single-Photon Emission Computed Tomographic Myocardial Perfusion Imaging in Patients With Known or Suspected Coronary Artery Disease Salvador Borges-Neto, MD, Linda K. Shaw, MS, Robert H. Tuttle, MSPH, John H. Alexander, MD, William T. Smith IV, MD, Marianna Chambless, BS, R. Edward Coleman, MD, Robert A. Harrington, MD, and Robert M. Califf, MD Noninvasive stress testing provides prognostic information in patients who have suspected coronary artery disease, but limited data are available on the incremental value of myocardial perfusion testing in high-risk patients. We studied 3,275 patients who underwent cardiac catheterization and single-photon emission computed tomographic (SPECT) perfusion imaging. Median follow-up was 3.1 years for death, cardiovascular death, and a composite of cardiovascular death or nonfatal myocardial infarction. Using Cox’s proportional hazards regression models, we examined the relation of SPECT summed stress score (SSS) to each outcome. A 1-unit change in SSS was associated with increased risks of 4%, 7%, and 5% for death, cardiovascular death, and death or nonfatal myocardial infarction, respectively (all p <0.0001). To examine the prognostic utility of SPECT, after baseline adjustments, SSS and angiographic results provided incremental prognostic information for each outcome. Thus, SPECT SSS provides information beyond clinical and angiographic data in patients who have known or suspected coronary artery disease. This information may be useful for stratifying patients into multiple risk categories for future cardiovascular events and potentially guiding therapy. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:182–188) e undertook this study (1) to evaluate the value of single-photon emission computed tomoW graphic (SPECT) myocardial perfusion imaging for Center (Durham, North Carolina) specifically for percutaneous coronary intervention often have incomplete angiographic data in the Duke Databank for Cardiovascular Disease. Ten percent of otherwise eligible patients had incomplete angiograms. The study population consisted of 3,275 patients who had complete angiographic and SPECT myocardial perfusion imaging studies. This research was performed with the approval of the institutional review board of the Duke University Medical Center. Clinical information: Clinical characteristics were prospectively collected for each patient at the time of cardiac catheterization and stored in the Duke Databank for Cardiovascular Disease.1,2 Clinical variables included age; history of congestive heart failure, peripheral vascular disease, cerebrovascular disease, hypertension, diabetes, or myocardial infarction; New York Heart Association congestive heart failure classification; presence of carotid bruits or ventricular gallop; chest pain (pain type, course, and frequency); gender; race; and a modified Charlson’s co-morbidity index.3 Histories of congestive heart failure and myocardial infarction were removed from Charlson’s index and examined independently because of their predictive value in high-risk patients who have cardiac problems. Stress testing: Patients who were capable of exercising underwent exercise treadmill stress testing according to a standard Bruce’s protocol, unless another protocol was specifically requested by the attending predicting adverse events in a high-risk patient population for coronary artery disease (CAD) and (2) to determine the incremental benefit of SPECT data versus those of clinical and angiographic data. With the increasing number of options available for medical and interventional management of CAD, risk stratification of patients who have CAD may help to guide and monitor responses to therapy. METHODS Study population: Between September 1993 and July 2000, 17,048 patients underwent SPECT procedures during stress (exercise or pharmacologic) or at rest at our laboratory. Of these patients, 4,829 underwent coronary angiography Х…180 days before or after the nuclear procedure. Patients were excluded from the study if they underwent subsequent revascularization Х…60 days after SPECT perfusion imaging and if the angiographic results were incomplete. Patients who are referred to the Duke University Medical From the Departments of Medicine (Cardiology) and Radiology (Nuclear Medicine) and the Duke Clinical Research Institute, Duke University Medical Center, Durham, North Carolina. Manuscript received March 11, 2004; revised manuscript received and accepted September 8, 2004. Address for reprints: Salvador Borges-Neto, MD, Duke University Medical Center, Box 3949, Durham, North Carolina 27710. E-mail: [email protected]. 182 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.004 FIGURE 1. Twelve-segment system used to generate the perfusion score for nuclear imaging at rest and during stress: (1) high anterior, (2) low anterior, (3) high anterolateral, (4) low anterolateral, (5) anteroseptal, (6) inferoseptal, (7) high posterolateral, (8) low posterolateral, (9) inferior, (10) posterobasal, (11) inferoapical, and (12) anteroapical. physician.4 Patients who were unable to exercise underwent pharmacologic stress testing. SPECT myocardial perfusion imaging protocol: The protocol for performing SPECT myocardial perfusion imaging studies has been previously described.5,6 In summary, SPECT data were obtained with multihead detectors using a step-and-shoot protocol. Images at rest were obtained for 30 seconds/projection and those during stress were obtained for 20 seconds/projection. A dual-isotope protocol was used for most patients, with thallium-201 being the agent of choice for images at rest and technetium-99m perfusion agents being used for images during stress. Patients who weighed Пѕ280 lb (Пѕ127 kg) had technetium-99m agents used for portions of the examination at rest and during stress. Image interpretation and candidate nuclear variables: Studies were independently reviewed by physi- cian specialists in nuclear medicine or nuclear cardiology. A 12-segment reporting system (Figure 1) was used to quantify perfusion in various segments. The relative perfusion grade to each segment was quantified by 4 gradations, with 0 representing no defect, 1 representing a mild defect, 2 representing a moderate defect, and 3 representing a severe defect. A cumulative summed stress score (SSS) was obtained by adding the scores of the 12 segments. Therefore, SSS would equal 0 in a normal study, and the maximum score would be 36 (severe perfusion defect in all 12 segments). The score variable has been shown to be highly predictive of cardiovascular outcomes in a 20segment model.7 The summed rest score, which is the perfusion score at rest, and the summed difference score, the score difference between stress and rest, were also constructed by applying the same scoring system. Coronary angiography: Coronary angiography was performed in multiple left and right anterior oblique projections. Angiograms were interpreted by consensus of Х†2 experienced angiographers who applied the following ordinal scale: 0, ПЅ25%, 25%, 50%, 75%, 95%, or 100% occlusion. Extent of CAD was determined by the traditional 1-, 2-, or 3-vessel disease characterization. Significant disease was defined as Х†75% occlusion of a major epicardial coronary artery. The number of major epicardial vessels with significant stenosis (Х†75% narrowing of luminal diameter) was used to evaluate the prognostic value of angiographic disease severity with SPECT. Follow-up: During follow-up, information was collected on death, cardiovascular death, nonfatal myocardial infarction, and date of last known status. These data were prospectively obtained by mailed questionnaires or telephone interviews at 6 months, 1 year, and annually thereafter. Information from clinic visits and rehospitalizations was also evaluated to ascertain follow-up and end point determination. Considering all time intervals and all patients, follow-up was 93% complete. An independent clinical events committee reviewed and classified all deaths and nonfatal myocardial infarctions without knowledge of the patient’s clinical, cardiac catheterization, or SPECT myocardial perfusion results. Criteria used to diagnose nonfatal myocardial infarction and cardiovascular death have been described and validated.8,9 Statistical analysis: Clinical characteristics of subjects were described in terms of percentage for categorical variables and by the median (25th and 75th percentiles) for continuous variables. Significant differences at the 0.05 level across risk strata were determined with Pearson’s chi-square tests for categorical variables and the Kruskal-Wallis test for continuous variables. Cumulative survival rates as a function of time after SPECT myocardial perfusion imaging were calculated using the Kaplan-Meier method.10 These rates are reported in tabular form at specific time points stratified by SSS score category and by number of diseased vessels. Cox’s proportional hazards regression modeling techniques were used to assess unadjusted and adjusted relations between SSS and patient outcomes.11,12 After examining the results of a flexible Cox’s model-fitting approach involving cubic polynomial spline functions,13–17 the linearity of the unadjusted relation between SSS and each outcome was assessed. Further application of Cox’s regression techniques resulted in multivariable clinical models that were developed to determine the most important predictors of the outcomes based on patients’ histories and physical examinations. Candidate variables were selected on the basis of clinical importance and from previous models.1,18 –20 Multivariable models for each end point that consisted of variables that were statistically significant (p ПЅ0.05) were derived from a stepwise selection process of the candidate variables: age, gender, race, history of hypertension, vascular disease, Charlson’s co-morbidity index, history and severity of CORONARY ARTERY DISEASE/PROGNOSTIC POWER OF SPECT 183 TABLE 1 Clinical Characteristics of Study Population Clinical Characteristics Men White Age (yrs) Systemic hypertension Previous myocardial infarction Previous congestive heart failure Diabetes mellitus Peripheral vascular disease Previous revascularization SRS SDS Angiographic results No. of diseased vessels 0 1 2 3 SSS 0 – 4 (n П 1,669) SSS 5–14 (n П 1,297) SSS Х†15 (n П 309) p Value 54.2% 77.7% 61 (52, 70) 65.6% 26.8% 28.1% 26.1% 14.9% 42.5% 0 (0, 0) 0 (0, 2) 71.3% 77.2% 61 (52, 69) 64.6% 58.9% 32.1% 31.2% 17.7% 58.4% 5 (2, 8) 2 (0, 5) 78.6% 78.3% 62 (54, 71) 65.1% 78.3% 50.0% 37.9% 18.8% 66.0% 15 (11, 18) 2 (0, 6) 0.001 0.910 0.208 0.851 0.001 0.001 0.001 0.054 0.001 0.001 0.001 42.9% 26.2% 14.7% 16.2% 16.7% 23.3% 24.8% 35.2% 7.4% 17.2% 21.0% 54.4% 0.001 Values are percentages or medians (25th, 75th percentiles). SDS П summed difference score; SRS П summed rest score. congestive heart failure, ventricular gallop, diabetes, previous myocardial infarction, and previous revascularization with percutaneous coronary intervention or coronary artery bypass grafting. After determining the most robust model for each end point, angiographic results (defined by number of diseased vessels) and perfusion SSS were added to the model in a stepwise manner to assess their incremental benefit. An interaction term consisting of SSS by angiographic results was tested in each model for each end point to determine whether the relation between outcome and SSS varied significantly over the extent of CAD. In addition, the measurements summed rest score and summed difference score were evaluated separately in a similar incremental fashion, with each score replacing SSS in the modeling scenario. Thus, we used prognostic risk-adjusted models that contained significant clinical variables to provide a base for evaluating the incremental prognostic value of angiography and perfusion imaging versus that of clinical data. RESULTS Clinical characteristics: Clinical characteristics, perfusion results, and angiographic results for the 3,275 patients stratified by levels of SSS are presented in Table 1. Higher-risk SSS levels were associated with male gender, previous revascularization procedures, history of myocardial infarction, congestive heart failure, diabetes, and increased severity of coronary disease on angiography. Differences in age, race, and anginal symptoms were not noted across perfusion levels. Follow-up, end points, and outcome events: Of 3,275 subjects, 571 died and 231 developed nonfatal myocardial infarctions. Of the deaths, 377 (66%) were classified as cardiovascular. The composite end point of cardiovascular death or myocardial infarction in- 184 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 cluded 558 patients, 50 of whom had a nonfatal myocardial infarction and later died. Median follow-up was 3.1 years (25th and 75th percentiles 1.6 and 5.1), and maximum follow-up was 8.1 years. Unadjusted end point relations: The unadjusted relation between perfusion and angiography with study end points was statistically significant in all patients. Chi-square values associated with SSS for death, cardiovascular death, and the composite of cardiovascular death or nonfatal myocardial infarction were 36.6, 57.6, and 51.7, respectively. Each unit change in SSS was associated with an absolute 4% increase in risk for death (95% confidence interval [CI] 3 to 6), a 7% increase in risk for cardiovascular death (95% CI 5 to 8), and a 5% increase in risk for death or nonfatal myocardial infarction (95% CI 4 to 7, p ПЅ0.001 for each). Unadjusted chi-square values associated with angiography for these end points were 49.5, 65.7, and 74.8, respectively. Each increase in the number of vessels with significant stenosis was associated with a 31% increased risk for death (95% CI 21 to 41), a 47% increased risk for cardiovascular death (95% CI 33 to 61), and a 39% increased risk for death or nonfatal myocardial infarction (95% CI 29 to 50, p ПЅ0.001 for each). Figure 2 shows the relation between SSS and cardiovascular death that was produced with a Cox’s model-fitting technique that involved cubic spline functions. A continuous linear plot of the spline between SSS and the logarithm of the hazard ratio for cardiovascular death exists, thus satisfying the linearity assumption of Cox’s proportional hazards model. Similar linear trends were observed for SSS and the other 2 end points. To illustrate risk stratification by SSS, we grouped SSS values into the following 3 categories of 0 to 4, 5 to 14, and Х†15. Survival rate by SSS category is shown in Figure 3. Similar trends were observed for the all-cause mortality rate and the composite of nonJANUARY 15, 2005 fatal myocardial infarction and cardiovascular mortality (Figure 4). Unadjusted 5-year event rates for each risk group stratified by angiographic results are presented in Table 2. Unadjusted 5-year event rates for patients who had an SSS of 0 stratified by clinical characteristics are presented in the Table 3. Multivariable Analyses: CLINICAL Significant risk factors from the clinical multivariable models for all 3 outcomes included age, New York Heart Association congestive heart failure severity class, Charlson’s co-morbidity index, diabetes, vascular disease, history of myocardial infarction, and previous percutaneous coronary intervention. Each variable was assoFIGURE 2. Polynomial spline plot of a linear relation between cardiovascular morciated with an increased risk, extality rate and results of SPECT imaging (SSS). LL ‫ ؍‬lower limit of confidence cept for previous percutaneous band; UL ‫ ؍‬upper limit of confidence band. coronary intervention, which was associated with decreased risk. In addition to this list, the all-cause mortality end point model included ventricular gallop. The covariate list for the cardiovascular mortality end point model also included race, with minority race being associated with a lower cardiovascular survival rate. The cardiovascular mortality or nonfatal myocardial infarction end point model also included hypertension and minority race as significant covariates. ANGIOGRAPHIC AND SPECT MODELING RESULTS: Clinical information was a powerful predictor of outcomes; however, coronary angiography and SSS provided additional and incremental prognostic information. Figure 5 shows chi-square values for the 3 models after angiographic data and SSS were added to the clinical models. The 8 clinical variables provided 91% of the prognostic information for estimating the risk of death, 83% for cardiovascular death, and 85% for cardiovascular death or myocardial infarction. However, SSS and angiographic data provided further prognostic information beyond these clinical variables and beyond each other (p ПЅ0.001 for all incremental chi-square values; Figure 5). For each outcome (cardiovascular death or cardiovascular death/myocardial infarction), summed rest score (chi-square 33.4, p П 0.0001 and chi-square 17.5, p П 0.0001, respectively) and summed difference score (chi-square 4.1, p П 0.0418 and chi-square 5.4, p П 0.0197, respectively) contributed no more than a similar amount of additional prognostic information to the clinical variables compared with SSS results. None of the interaction terms, SSS by angiographic results, tested in the multivariable models were significant (all interaction terms p Пѕ0.2). Thus, the relation between SSS and outcome remained constant FIGURE 3. Kaplan-Meier cardiovascular event-free survival across the extent of CAD (defined by the number of curves demonstrating risk represented by SSS values 0 to 4, 5 to diseased vessels) for each end point. 14, and >15. VARIABLES: CORONARY ARTERY DISEASE/PROGNOSTIC POWER OF SPECT 185 graphic results or did not use clinical models that adjusted for noncardiac (Charlson’s co-morbidity index)3 and cardiac co-morbidities. In our series, in which all patients underwent cardiac catheterization and SPECT imaging, we observed 5.1% and 6.3% 1-year rates for death and combined cardiovascular death or myocardial infarction, respectively. Despite a high prevalence of previous myocardial infarction in patients who had an SSS Х†15, in the overall patient population, the summed difference score and summed rest score also provided independent prognostic information. When we observed a higher prevalence of patients who had previFIGURE 4. Event rates at follow-up intervals for all end points for SSS values 0 to 4, 5 ous myocardial infarction, the summed to 14, and >15. CV ‫ ؍‬cardiovascular; MI ‫ ؍‬myocardial infarction. rest score and, hence, SSS were higher, indicating worsening outcomes. Although we excluded patients DISCUSSION This study involved a high-risk cohort of patients who underwent revascularization Х…60 days, we exwho had a high rate of revascularization and under- amined and determined that these patients had higher went cardiac catheterization Х…6 months of their index summed difference scores compared with the overall SPECT study. We have collected a comprehensive list study population (p П 0.0001). Further, the summed of clinical, angiographic, and nuclear imaging vari- rest score was not different (p П 0.3976). Including ables and have developed a robust clinical model to these patients did not change the interpretation of understand their prognostic value. We found a strong SPECT results. No interactions between SSS and number of disassociation between SSS and all-cause mortality rate and between SSS and cardiovascular mortality rate. eased vessels were found; however, even after adjustClinical variables provided Пѕ80% of the prognostic ment for clinical and angiographic data, SSS was information available from the combination of clinical prognostically important. Thus, these data support the data, SPECT imaging results, and angiographic re- use of SPECT imaging to provide important informasults. After adjustment for these clinical and angio- tion to physicians, even when angiographic results are graphic variables, SPECT imaging provided addi- known. Our research used a 12-segment system model to tional important prognostic information. Many diagnostic tests and procedures, including describe the location of cardiac perfusion defects becardiac catheterization, are available to risk stratify cause our database supported only 12 segments at the patients who have CAD into high- and low-risk sub- outset of the data collection period. Although use of groups to guide further invasive or noninvasive treat- the recommended segmental model27 would provide a ment. The Duke Databank for Cardiovascular Disease greater ability to define regions of defect within the has been used to validate the prognostic importance of myocardium and possibly offer a higher resolution of exercise treadmill testing and of clinical variables and the SSS, the use of a 12-segment model does not to assess the prognostic utility of exercise ejection detract from the validity of our results. Despite multiple previous studies that have demfraction calculated by radionuclide angiography. More recently, it was used to demonstrate the incremental onstrated the usefulness of the poststress ejection fracprognostic value of ejection fraction by first-pass tion to predict outcome,28 we did not incorporate a gated SPECT ejection fraction into this study. At the RNA during pharmacologic stress testing. Although previous studies have evaluated the time of data collection for our database, gated SPECT relative importance of thallium 201 imaging in pa- imaging was not a standard component of the diagtients who undergo cardiac catheterization21,22 and nostic study in our laboratory. Based on previous have made efforts to quantify the relative contribu- results, it is likely that a gated SPECT ejection fraction of various radionuclide, clinical, and catheter- tion would provide further incremental prognostic ization variables,23 they have not addressed the value beyond the clinical variables and perhaps beprognostic contribution when the extent of diseased yond the perfusion score, particularly for the prediccoronary anatomy was known.24 –26 The use of tion of cardiovascular mortality rate. Despite these limitations but because of the strong SPECT as a prognostic aid has been validated in numerous studies,7,24 –26 but these populations were association between SSS and clinical outcomes, the generally at low risk for disease or had low ob- potential exists to use SPECT perfusion imaging as an served event rates, or their respective methodolo- intermediate biomarker to draw inferences about those gies did not assess the effect of coronary angio- therapies that may ultimately produce clinical benefit. 186 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 TABLE 2 Unadjusted Five-year Event Rates Categorized by Summed Stress Score Risk Strata and Number of Diseased Vessels SSS Group No. of Diseased Vessels No. of Patients/Group Death (%) CV Death (%) CV Death/MI (%) 0 1 2 3 0 1 2 3 0 1 2 3 717 437 245 270 217 302 322 456 23* 53 65 168 14.8 21.1 22.5 28.9 14.3 21.8 24.6 33.9 — 21.3 28.9 42.9 8.2 10.2 14.9 22.2 10.2 13.3 18.1 26.2 — 14.0 27.8 35.3 14.0 18.1 26.3 29.8 14.6 17.6 27.3 36.1 — 17.3 47.7 42.0 0–4 5–14 Х†15 *Insufficient number of patients to assess event rates. CV П cardiovascular; MI П myocardial infarction. TABLE 3 Unadjusted Five-year Event Rates at Zero Summed Stress Score Clinical Characteristic No. of diseased vessels Stress type Age group Charlson’s co-morbidity categories Levels No. of Patients/Group Death Rate (%) CV Death Rate (%) CV Death/MI Rate (%) 0 1 Multiple Exercise Pharmacologic Young (ПЅ70 yrs) Old (Х†70 yrs) Пѕ1 0 or 1 364 247 238 344 505 613 236 204 645 17.7 20.0 21.9 10.1 36.0 16.7 28.8 27.3 17.0 9.5 9.4 12.7 5.8 19.2 14.8 23.3 14.7 9.0 14.4 16.6 20.9 10.9 27.7 7.9 18.1 22.9 15.0 Abbreviations as in Table 2. approach, if successful, would further support the usefulness of SSS to cardiology and clinical decision making. 1. Pryor DB, Shaw L, McCants CB, Lee KL, Mark DB, Harrell FE Jr, Muhlbaier LH, Califf RM. Value of the history and physical in identifying patients at increased risk for coronary artery disease. Ann Intern Med 1993; 118:81–90. 2. Borges-Neto S, Shaw LJ, Kesler K, Sell T, Peterson ED, Coleman RE, Jones RH. Usefulness of serial radionuclide angiography in predicting cardiac death after coronary artery bypass grafting and comparison with clinical and cardiac catheterization data. Am J Cardiol 1997;79:851– 855. 3. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40:373–383. 4. Gibbons RJ, Balady GJ, Beasley JW, Bricker JT, Duvernoy WF, Froelicher VF, Mark DB, Marwick TH, FIGURE 5. Model chi-square values demonstrating incremental prognostic information McCallister BD, Thompson PD, et al. ACC/AHA for clinical variables (black bars), clinical Ш‰ angiographic variables (bars with horiguidelines for exercise testing: executive summary. A zontal lines), clinical variables Ш‰ SSS (white bars), and clinical variables Ш‰ angioreport of the American College of Cardiology/Amerigraphic variables Ш‰ SSS (bars with vertical lines). Abbreviations as in Figure 4. can Heart Association Task Force on Practice Guidelines (Committee on Exercise Testing). Circulation 1997;96:345–354. Novel anti-ischemic regimens or new revasculariza- 5. Mast ST, Shaw LK, Ravizzini GC, Chambless M, Joski P, Coleman RE, S. Incremental prognostic value of RNA ejection fraction measuretion techniques could potentially be screened initially Borges-Neto ments during pharmacologic stress testing: a comparison with clinical and perfor an effect on SSS. Interventions that decrease SSS fusion variables. J Nucl Med 2001;42:871– 877. would be promising to test in adequately powered 6. Borges-Neto S. Perfusion and function assessment by nuclear cardiology techniques. Curr Opin Cardiol 1997;12:581–586. clinical trials with long-term follow-up to determine 7. Berman DS, Kang X, Van Train KF, Lewin HC, Cohen I, Areeda J, Freidman whether they also improve clinical outcomes. This JD, Germano G, Shaw LJ, Hachamovitch R. Comparative prognostic value of CORONARY ARTERY DISEASE/PROGNOSTIC POWER OF SPECT 187 automatic quantitative analysis versus semiquantitative visual analysis of exercise myocardial perfusion single-photon emission computed tomography. J Am Coll Cardiol 1998;32:1987–1995. 8. Harris PJ, Harrell FE Jr, Lee KL, Rosati RA, Nonfatal myocardial infarction in medically treated patients with coronary artery disease. Am J Cardiol 1980; 46:937–942. 9. Shaw LJ, Hachamovitch R, Heller G, Marwick TH, Travin MI, Iskandrian AE, Kesler K, Lauer MS, Hendel R, Borges-Neto S, et al. Noninvasive strategies for the estimation of cardiac risk in stable chest pain patients. Am J Cardiol 2000; 86:1–7. 10. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 481. 11. Cox DR. Regression models and life tables (with discussion). J R Stat Soc B 1972;34:187–220. 12. Breslow NE. Covariance analysis of censored survival data. Biometrics 1974;30:89 –99. 13. Harrell FE Jr, Pollock BG, Lee KL. Graphical methods for the analysis of survival data. In: Proceedings of the 12th Annual SAS Users Group International Conference. Cary, NC: SAS Institute, 1987:1107–1115. 14. Smith PL. Splines as a useful and convenient statistical tool. Am Stat 1979;33:57– 62. 15. Stone CJ, Koo C. Additive splines in statistics. In: Proceedings of the Statistical Computing Section. Alexandria, VA: American Statistical Association, 1985:45– 48. 16. Devlin TF, Weeks BJ. Spline functions for logistic regression modeling. In: Proceedings of the 11th Annual SAS Users Group International Conference. Cary, NC: SAS Institute, 1986:646 – 651. 17. Harrell FE Jr, Lee KL, Pollock BG. Regression models in clinical studies: determining relationships between predictors and response. J Natl Cancer Inst 1988;80:1198 –1202. 18. Rosati RA, McNeer JF, Starmer CF, Mittler BS, Morris JJ Jr, Wallace AG. A new information system for medical practice. Arch Intern Med 1985;135: 1017–1024. 19. Califf RM, Mark DB, Harrell FE Jr, Hlatky MA, Lee KL, Rosati RA, Pryor DB. Importance of clinical measures of ischemia in the prognosis of patients with documented coronary artery disease. J Am Coll Cardiol 1988;11:20 –26. 188 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 20. Harris PJ, Lee KL, Harrell FE Jr, Behar VS, Rosati RA. Outcome in medically treated coronary artery disease: ischemic events: nonfatal infarction and death. Circulation 1980;62:718 –726. 21. Kaul S, Finkelstein DM, Homma S, Leavitt M, Okada RD, Boucher CA. Superiority of quantitative exercise thallium-201 variables in determining longterm prognosis in ambulatory patients with chest pain: a comparison with cardiac catheterization. J Am Coll Cardiol 1988;12:25–34. 22. Kaul S, Lilly DR, Gascho JA, Watson DD, Gibson RS, Oliner CA, Ryan JM. Prognostic utility of the exercise thallium-201 tests in ambulatory patients with chest pain: comparison with cardiac catheterization. J Am Coll Cardiol 1988; 12:25–34. 23. Lee KL, Pryor DB, Pieper KS, Harrell FE Jr, Califf RM, Mark DB, Hlatky MA, Coleman RE, Cobb FR, Jones RH. Prognostic value of radionuclide angiography in medically treated patients with coronary artery disease. Circulation 1990;82:1705–1717. 24. Hachamovitch R, Berman DS, Shaw LJ, Kiat H, Cohen I, Cabico JA, Friedman J, Diamond GA. Incremental prognostic value of myocardial perfusion single photon emission computed tomography for the prediction of cardiac death. Differential stratification for risk of cardiac death and myocardial infarction. Circulation 1998;97:535–543. 25. Hachamovitch R, Berman DS, Kiat H, Cohen I, Cabico JA, Friedman J, Diamond GA. Exercise myocardial perfusion SPECT in patients without known coronary artery disease: incremental prognostic value and impact on subsequent patient management. Circulation 1996;93:905–914. 26. Iskandrian S, Iskandrian AE. Risk assessment using single-photon emission computed tomographic technetium-99m sestamibi imaging. J Am Coll Cardiol 1998;32:57– 62. 27. Berman D, Port SC, Garcia E, Sinusas A, Wackers F. Imaging guidelines for nuclear cardiology procedures. J Nucl Cardiol 1996;6:G49 –G84. 28. Sharir T, Germano G, Kang X, Lewin HC, Miranda R, Cohen I, Agafitei RD, Friedman JD, Berman DS. Prediction of myocardial infarction versus cardiac death by gated myocardial perfusion SPECT: risk stratification by the amount of stress-induced ischemia and the poststress ejection fraction. J Nucl Med 2001; 42:831– 837. JANUARY 15, 2005 Comparative Effects of Rosuvastatin and Gemfibrozil on Glucose, Insulin, and Lipid Metabolism in Insulin-Resistant, Nondiabetic Patients With Combined Dyslipidemia Cindy Lamendola, MSN, Fahim Abbasi, MD, James W. Chu, MD, Howard Hutchinson, MD, Valerie Cain, MS, Elizabeth Leary, PhD, Tracey McLaughlin, MD, Evan Stein, MD, PhD, and Gerald Reaven, MD To evaluate the pharmacologic intervention most likely to decrease cardiovascular disease risk in insulin-resistant patients with combined dyslipidemia, 39 patients with this abnormality were assessed before and after 3 months of treatment with gemfibrozil (1,200 mg/day) or rosuvastatin (40 mg/day) with regard to: (1) steady-state plasma glucose concentration at the end of a 180-minute infusion of octreotide, insulin, and glucose; (2) fasting lipid, lipoprotein, and apolipoprotein concentrations; and (3) daylong glucose, insulin, triglyceride, and remnant lipoprotein cholesterol concentrations in response to breakfast and lunch. The 2 groups were similar at baseline in age, gender, body mass index and in measurements of carbohydrate and lipoprotein metabolism. Neither gemfibrozil nor rosuvastatin enhanced insulin sensitivity or lowered daylong glucose and insulin concentrations in insulin-resistant patients with combined dyslipidemia, but both drugs significantly decreased fasting triglyceride concentrations. However, only rosuvastatin treatment significantly (p <0.05 to <0.001) reduced fasting low-density lipoprotein cholesterol, apolipoprotein B-100, apolipoprotein C-III, apolipoprotein C-III:B particles, the apolipoprotein B-100:apolipoprotein A-I ratio, and increased apolipoprotein A-I (p <0.05). The degree of improvement in fasting and postprandial remnant lipoprotein cholesterol concentrations was significantly greater (p <0.05) in rosuvastatin-treated patients, and this difference in the relative effectiveness of the drugs was also true of the decrease in non– highdensity lipoprotein cholesterol concentrations. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:189 –193) pidemiologic studies have demonstrated that cardiovascular disease (CVD) risk is greater in paE tients with combined dyslipidemia, defined by eleva- in insulin-resistant, nondiabetic patients with combined dyslipidemia. tions in both plasma triglyceride and low-density lipoprotein (LDL) cholesterol concentrations, than in those with either abnormality alone.1,2 In addition to their abnormal lipoprotein metabolism, patients with combined dyslipidemia are also insulin resistant and hyperinsulinemic.3 Because the clinical utility of statin versus fibric acid drug treatment has not been evaluated in patients with this clinical syndrome, we compared the relative effects of gemfibrozil with rosuvastatin treatment on multiple aspects of fasting and postprandial carbohydrate and lipoprotein metabolism From the Department of Medicine, Stanford University School of Medicine, Stanford, California; AstraZeneca, Wilmington, Delaware; Pacific Biometrics Inc., Seattle, Washington; and Medical Research Laboratories International, Highland Heights, Kentucky. This study was supported by Research Grant RR-00070 from the National Institutes of Health, Bethesda, Maryland, and a grant from AstraZeneca, Wilmington, Delaware. Drs. Reaven and Stein received funding from AstraZeneca for this study. Manuscript received June 8, 2004; revised manuscript received and accepted September 8, 2004 Address for reprints: Gerald M. Reaven, MD, Falk Cardiovascular Research Center, Stanford University School of Medicine, 300 Pasteur Drive, Stanford, California 94305. E-mail: greaven@ cvmed.stanford.edu. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 METHODS The Human Subjects Committee approved this study, and all subjects provided written, informed consent at the time of their initial visit to the General Clinical Research Center. The experimental group consisted of 39 apparently healthy patients who volunteered for this study in response to advertisements in local newspapers. They were in good general health, with a body mass index ПЅ33.0 kg/m2, and had elevations (Пѕ200 mg/dl) of both total plasma cholesterol and triglyceride concentrations while not taking any lipid-lowering drugs for Х†6 weeks. In addition, they had no known disease other than dyslipidemia, were nondiabetic, with normal blood count, urinalysis, and liver, kidney, and thyroid function tests. Volunteers meeting these inclusion criteria were instructed on a weight-maintaining, low saturated fat diet and were seen at intervals over the next 6 weeks. Those who continued to have a total cholesterol concentration Пѕ200 mg/dl and triglyceride concentrations between 200 and 600 mg/dl (with values differing by ПЅ30%) at the end of the 6-week dietary run-in were then randomized to receive treatment with either gemfibrozil or rosuvastatin. The age, gender distribution, 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.005 189 ide concentrations were measured as previously described, and remnant lipoproteins were isolated by an imp Value munoseparation method based on the use of monoclonal antibodies to hu0.45 0.62 man apolipoproteins B-100 and A-I 0.28 to remove most of the apolipoprotein B-100 – and apolipoprotein A-I– containing lipoproteins. The unbound fraction is designated as remnant lipoproteins and quantified by determining the cholesterol concentration with a highly sensitive enzymatic assay.7–9 On another morning after an overnight fast, insulin-mediated glucose disposal was quantified by a modification10 of the insulin suppression test.11,12 Subjects were infused for 180 minutes with octreotide acetate (0.27 вђ®g/m2/min), insulin (32 mU/m2/min), and glucose (267 mg/m2/min), blood drawn at 10minute intervals from 150 to 180 minutes, plasma glucose and insulin concentrations measured, and the mean of these values used as the steady-state plasma insulin and glucose concentrations for each patient. The steady-state plasma glucose concentration provides a direct measure of the ability of insulin to mediate disposal of an infused glucose load; the higher the steady-state plasma glucose concentration, the more insulin-resistant the patient. After completion of these baseline measurements, treatment was initiated with gemfibrozil (600 mg twice daily) or rosuvastatin (40 mg/day). Subjects were seen at 2-week intervals to evaluate their wellbeing, drug and dietary compliance, and liver function; creatine kinase was measured at monthly intervals. All baseline measurements were repeated after 3 months of drug treatment. Results are expressed as mean П® SEM unless otherwise stated. Statistical significance of the absolute changes in the fasting concentrations of the experimental variables resulting from each treatment were compared by Student’s paired t test, with comparisons of the differences between the effects of each treatment by Student’s unpaired t test. Absolute changes in postprandial values were evaluated by 2-way analysis of variance. In all cases, p values ПЅ0.05 were considered statistically significant. TABLE 1 Baseline Clinical Characteristics of the Two Experimental Groups Variable Gemfibrozil (n П 19) Rosuvastatin (n П 20) Age (yrs) Men/women Body mass index (kg/m2) 51 П® 10 11/8 29.5 П® 2.9 54 П® 9 10/10 28.1 П® 4.1 Values are expressed as mean П® SD. and body mass index of the 2 groups are compared in Table 1, and it is apparent that these variables were similarly distributed. Before starting drug treatment, blood samples were obtained after an overnight fast, and plasma separated for determination of fasting lipid, lipoprotein, and apolipoprotein concentrations in the following manner. Total cholesterol and triglyceride concentrations were determined by enzymatic methods on a Hitachi 747 analyzer (Boehringer Mannheim Diagnostics, Indianapolis, Indiana) as previously described.4 Highdensity lipoprotein (HDL) was isolated using heparin-2 mol/L manganese chloride,5 and the cholesterol and triglyceride content of the isolated HDL measured. Very-low-density lipoprotein (VLDL) and LDL cholesterol and triglyceride concentrations were determined after the ultracentrifugal separation of LDL and HDL from VLDL at a density Пѕ1.006.6 Cholesterol and triglyceride concentrations were measured at a density Пѕ1.006 (bottom) fraction, and the previously measured HDL cholesterol and HDL triglyceride values subtracted to obtain LDL cholesterol and LDL triglyceride levels. VLDL cholesterol and VLDL triglyceride levels were calculated by subtracting the values obtained on the bottom fraction from those measured in whole plasma. Apolipoproteins A-I, B, and E were analyzed by rate immunonephelometry (Dade-Behring, Marburg, Germany). Apolipoprotein C-III was measured by electroimmunoassay, and the apolipoprotein C-III content of nonapolipoprotein Bcontaining lipoproteins (apolipoprotein C-III:non B) similarly measured after precipitation of all apolipoprotein B-containing particles with antisera to apolipoprotein B. The amount of apolipoprotein B associated C-III (apolipoprotien C-III:B) was obtained by subtraction. Plasma glucose and insulin concentrations were measured every hour in response to 2 meals, given at 8 A.M. and at noon. Both meals contained (as percentages of total calories) 15% protein, 43% carbohydrate, and 42% fat (ПЅ10% saturated fat), with breakfast comprising 20% and lunch comprising 40% of estimated daily caloric requirements. Insulin was measured in a stepwise sandwich enzyme-linked immunosorbent assay procedure on an ES 300 (Boehringer Mannheim Diagnostics). Glucose was measured using the hexokinase method on the Hitachi 747 (Boehringer Mannheim Diagnostics). During the daylong meal tolerance test, blood was also obtained at 2-hour intervals for determination of fasting and postprandial plasma triglyceride and remnant lipoprotein cholesterol concentrations. Triglycer190 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 RESULTS Fasting plasma lipid and lipoprotein concentrations, before and after treatment, are listed in Table 2; importantly, at baseline both groups had the adverse CVD risk profile of patients with combined dyslipidemia (i.e., elevations in total cholesterol and triglyceride concentrations) associated with low HDL cholesterol concentrations. Total cholesterol concentrations decreased in response to both drugs (p ПЅ0.001), but the reduction was greater in response to rosuvastatin (ПЄ104 to ПЄ17 mg/dl). Not surprisingly, LDL cholesterol concentrations decreased significantly (p ПЅ0.001) after administration of rosuvastatin (ПЄ76 mg/dl), but did not change in response to gemfibrozil (П©5 mg/dl). VLDL and non-HDL cholesterol concentrations decreased significantly (p ПЅ0.001) JANUARY 15, 2005 TABLE 2 Fasting Lipid and Lipoprotein Concentrations Before and After Treatment Gemfibrozil Variable Total cholesterol VLDL cholesterol LDL cholesterol HDL cholesterol Non–HDL cholesterol Triglycerides VLDL triglycerides LDL triglycerides HDL triglycerides RLP cholesterol Total cholesterol/HDL cholesterol Non–HDL cholesterol/ HDL cholesterol Before 223 52 126 41 181 284 199 26 43 15 5.6 П® П® П® П® П® П® П® П® П® П® П® 4 4 5 2 4 17 15 4 3 2 0.2 4.6 П® 0.2 Rosuvastatin After p Value Before vs After П® П® П® П® П® П® П® П® П® П® П® 5 4 5 3 4 23 20 4 4 1 0.2 ПЅ0.001 ПЅ0.001 NS ПЅ0.001 ПЅ0.001 ПЅ0.001 ПЅ0.001 NS ПЅ0.01 ПЅ0.005 ПЅ0.001 3.7 П® 0.2 ПЅ0.001 206 31 131 46 160 166 106 26 34 10 4.7 Before 242 66 138 40 202 324 235 33 45 23 6.6 П® П® П® П® П® П® П® П® П® П® П® 5 7 7 3 6 19 27 6 2 4 0.4 5.6 П® 0.4 After p Value Before vs After p Value* П® П® П® П® П® П® П® П® П® П® П® 6 4 4 3 6 18 19 2 2 1 0.2 ПЅ0.001 ПЅ0.001 ПЅ0.001 NS ПЅ0.001 ПЅ0.001 ПЅ0.05 ПЅ0.05 ПЅ0.05 ПЅ0.002 ПЅ0.001 ПЅ0.001 NS ПЅ0.001 NS ПЅ0.001 NS NS ПЅ0.05 NS ПЅ0.05 ПЅ0.001 2.4 П® 0.2 ПЅ0.001 ПЅ0.001 138 35 62 42 96 211 163 14 39 9 3.4 Values are expressed as mean П® SEM. *Changes in gemfibrozil treatment compared with changes in rosuvastatin treatment. RLP П remnant lipoprotein. TABLE 3 Fasting Apolipoprotein Concentrations Before and After Treatment in Patients With Combined Dyslipidemia Gemfibrozil Variable Apolipoprotein A-I Apolipoprotein B-100 Apolipoprotein C-III Apolipoprotein C-III:B Apolipoprotein C-III:non-B Apolipoprotein E Apolipoprotein B-100/ apolipoprotein A-I Before 141 137 4.5 2.5 2.1 4.8 1.0 П® П® П® П® П® П® П® 6 4 0.3 0.2 0.2 0.3 0.1 After 144 129 3.8 2.1 1.8 4.2 0.9 П® П® П® П® П® П® П® 8 4 0.4 0.2 0.2 0.3 0.0 Rosuvastatin p Value Before vs After NS NS NS NS NS ПЅ0.05 NS Before 132 147 5.2 3.1 2.1 6.0 1.2 П® П® П® П® П® П® П® 7 6 0.3 0.3 0.2 0.6 0.1 After p Value Before vs After p Value* П® П® П® П® П® П® П® ПЅ0.05 ПЅ0.001 ПЅ0.01 ПЅ0.05 NS ПЅ0.001 ПЅ0.001 NS ПЅ0.001 NS NS NS ПЅ0.001 ПЅ0.001 146 80 4.3 2.5 2.0 3.4 0.6 8 4 0.4 0.3 0.1 0.3 0.0 Values are expressed as mean П® SEM. *p values П changes in gemfibrozil treatment compared with changes in rosuvastatin treatment. in response to either treatment (ПЄ21 and ПЄ21 mg/dl, respectively, for gemfibrozil vs ПЄ31 and ПЄ106 mg/dl for rosuvastatin), but the decrease in non-HDL cholesterol concentration was significantly greater in rosuvastatintreated patients (p ПЅ0.001). The only variable that improved in response to gemfibrozil, but not rosuvastatin, was an increase in HDL cholesterol concentrations (5 vs 2 mg/dl), but there was no significant difference when the change associated with the 2 treatments was directly compared. In contrast to the disparate effects of the 2 drugs on LDL cholesterol concentrations, both drugs lowered total plasma triglyceride concentrations (p ПЅ0.001), and to a comparable degree (ПЄ118 vs ПЄ113 mg/dl). In addition, VLDL triglyceride and HDL triglyceride concentrations were significantly lower after treatment with either drugs, and the magnitude of the decrease was not significantly different as a function of the drug used. However, a decrease in LDL triglycerides was only seen in rosuvastatin-treated patients (ПЄ19 mg/dl), and these patients also had a significantly greater decrease in fasting remnant lipoprotein cholesterol concentrations (ПЄ14 mg/dl for rosuvastatin vs ПЄ5 mg/dl for gemfibrozil). Finally, although the ratios of total cholesterol/HDL cholesterol and non-HDL cholesterol/HDL cholesterol were significantly lower after treatment with either drug, they decreased to a significantly greater degree in response to rosuvastatin (ПЄ3.2 and ПЄ3.2 for rosuvastatin vs ПЄ0.9 and ПЄ0.9 for gemfibrozil, respectively). Treatment-associated changes in apolipoprotein concentrations are listed in Table 3. The only change after gemfibrozil administration was a significant decrease in plasma apolipoprotein E concentrations (ПЄ0.6 mg/dl). Apolipoprotein E concentrations also decreased in rosuvastatin-treated patients (ПЄ2.6 mg/dl), and the magnitude of the decline was significantly greater (p ПЅ0.001). In addition, rosuvastatin treatment was associated with an increase in apolipoprotein A-I concentrations (p ПЅ0.05) of 14 mg/dl, and decreases of ПЄ67 mg/dl in the concentrations of apolipoprotein B-100 (p ПЅ0.001), ПЄ0.9 mg/dl in apolipoprotein C-III (p ПЅ0.01), and ПЄ0.7 mg/dl in apolipoprotein C-III:B (p ПЅ0.05), as well as ПЄ0.6 in the plasma concen- PREVENTIVE CARDIOLOGY/ROSUVASTATIN AND GEMFIBROZIL IN COMBINED DYSLIPIDEMIA 191 FIGURE 1. Comparison of daylong plasma triglycerides and remnant lipoprotein cholesterol (RLP) concentration in response to test meals before and after treatment with gemfibrozil or rosuvastatin. The test meals contained (as a percentage of total calories) 15% protein, 42% fat (<10% saturated fat), and 43% carbohydrates. Subjects ate breakfast at 8 A.M. (20% of daily caloric requirements) and lunch at noon (40% of daily caloric requirements) with blood samples obtained after fasting and then every 2 hours for 8 hours. tration ratio of apolipoprotein B-100/apolipoprotein A-I (p ПЅ0.001). Steady-state plasma glucose concentrations did not change in response to treatment with either gemfibrozil (192 П® 15 to 201 П® 14 mg/dl) or rosuvastatin (192 П® 13 to 207 П® 16 mg/dl). Because steady-state plasma insulin concentrations were also similar before and after treatment with either drug (Пі80 mU/L), insulin-mediated glucose disposal did not change with treatment. Consequently, it was not surprising that daylong plasma glucose and insulin levels were comparable before and after in both treatment groups. It is worth emphasizing that the steady-state plasma glucose concentrations were quite high in both groups, similar to previous data indicating that insulin resistance characterizes patients with combined dyslipidemia.3 Daylong concentrations of plasma triglycerides, and remnant lipoprotein cholesterol, before and after treatment with the 2 drugs are shown in Figure 1. It is apparent from these data that treatment with either drug led to significant (p ПЅ0.001) decreases in daylong plasma triglyceride and remnant lipoprotein cholesterol concentrations. Although no difference was observed between them with regard to effects on daylong triglyceride concentrations, the reductions in remnant lipoprotein cholesterol were significantly 192 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 greater (p ПЅ0.01) than with gemfibrozil after rosuvastatin treatment. DISCUSSION The results presented indicate that rosuvastatin treatment of individuals with combined dyslipidemia not only decreases LDL cholesterol concentration to a value that could no longer be considered to contribute to CVD risk but also lowers triglyceride, VLDL triglyceride, and VLDL cholesterol concentrations to approximately the same degree as that in gemfibrozil-treated patients. The finding that all of the lipoprotein abnormalities that characterize subjects with combined dyslipidemia improved significantly in rosuvastatin-treated patients raises the possibility that effective treatment of these patients frequently may not require combination therapy with a statin and a fibric acid. Although treatment with either drug was associated with significant decreases in plasma non-HDL cholesterol concentrations, as well in the plasma concentration ratios of total cholesterol/HDL cholesterol, and non-HDL cholesterol/HDL cholesterol, in all 3 cases the magnitude of the change was significantly greater in rosuvastatin-treated patients. In addition, rosuvasJANUARY 15, 2005 tatin-treated patients had several changes in plasma apolipoprotein concentrations that would suggest decreased CVD risk, including a statistically significant increase in apolipoprotein A-I and decreases in apolipoprotein B-100, apolipoprotein C-III, apolipoprotein C-III:B, apolipoprotein E, and the ratio of apolipoprotein B-100/apolipoprotein A-I. In contrast, the only statistically significant changes in plasma apoprotein concentrations seen in gemfibrozil-treated patients were decreases in apolipoprotein E levels; even in this case, the decrease in rosuvastatin-treated patients was significantly greater. Finally, the observation that the decrease in the postprandial accumulation of remnant lipoproteins was significantly greater after treatment with rosuvastatin is worthy of comment for 2 reasons. Clinically, the atherogenic potential of increased concentrations of triglyceride-rich remnant lipoproteins, initially emphasized by Zilversmit,13 has been repeatedly confirmed.14 –16 Physiologically, the daylong decreases in postprandial triglyceride concentrations were similar after both drugs, and the decrements in remnant lipoprotein cholesterol concentrations in gemfibroziltreated patents were similar to changes in postprandial triglyceride levels. In contrast, the decrease in daylong remnant lipoprotein cholesterol concentrations in rosuvastatin-treated subjects were accentuated compared with the decreases in postprandial triglycerides, resulting in much lower postprandial remnant lipoprotein levels than levels in gemfibrozil-treated patients. These data (1) are similar to our findings in patients with combined dyslipidemia who also had type 2 diabetes,17 (2) demonstrate that the ability of satins to upregulate the hepatic LDL receptor will also increase the removal rate from plasma of remnant lipoproteins, and (3) offer another example of how statin treatment can decrease CVD risk above and beyond the ability to lower LDL cholesterol levels. 1. Manninen V, Tenkanen L, Koskinen P, Huttunen JK, Manttari M, Heinonen OP, Frick MH. Joint effects of serum triglyceride and LDL cholesterol and HDL cholesterol concentrations on coronary heart disease risk in the Helsinki Heart Study. Implications for treatment. Circulation 1992;85:37– 45. 2. Assmann G, Schulte H, Funke H, Von Eckardstein A. The emergence of triglycerides as a significant independent risk factor in coronary artery disease. Eur Heart J 1998;19(suppl):M8 –M14. 3. Sheu WH-H, Shieh SM, Fuh MM, Shen DD, Jeng CY, Chen Y-D, Reaven GM. Insulin resistance, glucose intolerance, and hyperinsulinemia. Hypertriglyceridemia versus hypercholesterolemia. Arterioiscler Thromb Vasc Biol 1993;13:367– 370. 4. Steiner P, Freidel J, Bremner W, Stein E. Standardization of micromethods for plasma cholesterol, triglyceride and HDL-cholesterol with the lipid clinics’ methodology. J Clin Chem 1981;19:850. 5. Warnick G, Albers J. A comprehensive evaluation of the heparin manganese precipitation procedure for estimating high-density lipoprotein cholesterol. J Lipid Res 1978;19:65–76. 6. Lipid Research Clinics Program. Manual of laboratory operations: Lipid and lipoprotein analysis. Washington DC: US Department of Health Education and Welfare. Publication NIH 75-628, 1982. 7. Campos E, Nakajima K, Tanaka A, Havel RJ. Properties of an apolipoprotein. E-enriched fraction of triglyceride-rich lipoproteins isolated from human blood plasma with a monoclonal antibody to apolipoprotein B-100. J Lipid Res 1992; 33:369 –380. 8. Nakajima K, Saito T, Tamura A, Suzuki M, Nakano T, Adachi M, Tanaka A, Tada N, Nakamura H, Campos E, Havel R. Cholesterol in remnant-like lipoproteins in human serum using monoclonal anti apo B-100 and anti apo A-I immunoaffinity mixed gels. Clin Chim Acta 1993;223:53–71. 9. Leary ET, Wang T, Baker DJ, Cilla DD, Zhong J, Warnick GR, Nakajima K, Havel RJ. Evaluation of an immunoseparation method for quantitative measurement of remnant-like particle-cholesterol in serum and plasma. Clin Chem 1998; 44:2490 –2498. 10. Pei D, Jones CN, Bhargava R, Chen YD, Reaven GM. Evaluation of octreotide to assess insulin-mediated glucose disposal by the insulin suppression test. Diabetolgia 1994;37:843– 845. 11. Shen S-W, Reaven GM, Farquhar J. Comparison of impedance to insulinmediated glucose uptake in normal subjects and in subjects with latent diabetes. J Clin Invest 1980;49:2151–2160. 12. Greenfield MS, Doberne L, Kraemer FB, Tobey TA, Reaven GM. Assessment of insulin resistance with the insulin suppression test and the euglycemic clamp. Diabetes 1981;30:387–92. 13. Zilversmit DB. Atherogenesis: a postprandial phenomenon. Circulation 1979;60:473– 485. 14. Groot PH, van Stiphout WA, Krauss XH, Jansen H, van Tol A, van Ramshorst E, Chin-On S, Hofman A, Cresswell SR, Havekes L. Postprandial lipoprotein metabolism in normolipidemic men with and without coronary artery disease. Arterioiscler Thromb Vasc Biol 1991;11:653– 662. 15. Patsch JR, Miesenbock G, Hopferwieser T, Muhlberger V, Knapp E, Dunn JK, Gotto AM Jr, Patsch W. Relation of triglyceride metabolism and coronary artery disease. Studies in the postprandial state. Arterioiscler Thromb Vasc Biol 1992;12:1336 –1345. 16. Karpe F, Bard JM, Steiner G, Carlson LA, Fruchart JC, Hamsten A. HDLs and alimentary lipemia. Studies in men with previous myocardial infarction at a young age. Arterioiscler Thromb Vasc Biol 1993;13:11–22. 17. McLaughlin T, Abbasi F, Lamendola C, Leary E, Reaven GM. Comparison in patients with type 2 diabetes of fibric acid versus hepatic hydroxymethyl glutaryl-coenzyme A reductase inhibitor treatment of combined dyslipidemia. Metabolism 2002;51:1355–1359. PREVENTIVE CARDIOLOGY/ROSUVASTATIN AND GEMFIBROZIL IN COMBINED DYSLIPIDEMIA 193 Relation Between Atherogenic Dyslipidemia and the Adult Treatment Program-III Definition of Metabolic Syndrome (Genetic Epidemiology of Metabolic Syndrome Project) Diego F. Wyszynski, PhD, Dawn M. Waterworth, PhD, Philip J. Barter, MD, Jonathan Cohen, PhD, Y. Antero KesГ¤niemi, MD, Robert W. Mahley, MD, PhD, Ruth McPherson, MD, GГ©rard Waeber, MD, Thomas P. Bersot, MD, Sanjay S. Sharma, BSc, Vikki Nolan, MPH, Lefkos T. Middleton, MD, PhD, Scott S. Sundseth, PhD, Lindsay A. Farrer, MD, Vincent Mooser, MD, and Scott M. Grundy, MD, PhD Genetic Epidemiology of Metabolic Syndrome is a multinational, family-based study to explore the genetic basis of the metabolic syndrome. Atherogenic dyslipidemia (defined as low plasma high-density lipoprotein cholesterol with elevated triglycerides (<25th and >75th percentile for age, gender, and country, respectively) identified affected subjects for the metabolic syndrome. This report examines the frequency at which atherogenic dyslipidemia predicts the metabolic syndrome of the National Cholesterol Education Program Adult Treatment Panel III (ATP-III). One thousand four hundred thirty-six (854 men/582 women) affected patients by our criteria were compared with 1,672 (737 men/935 women) unaffected persons. Affected patients had more hypertension, obesity, and hyperglycemia, and they met a higher number of ATP-III criteria (3.2 ШЋ 1.1 SD vs 1.3 ШЋ 1.1 SD, p <0.001). Overall, 76% of affected persons also qualified for the ATP-III definition (Cohen’s вђ¬ 0.61, 95% confidence interval 0.59 to 0.64), similar to a separate group of 464 sporadic, unrelated cases (75%). Concordance increased from 41% to 82% and 88% for ages <35, 36 to 55, and >55 years, respectively. Affected status was also independently associated with waist circumference (p <0.001) and fasting glucose (p <0.001) but not systolic blood pressure (p ‫ ؍‬0.43). Thus, the lipid-based criteria used to define affection status in this study substantially parallels the ATP-III definition of metabolic syndrome in subjects aged >35 years. In subjects aged <35 years, atherogenic dyslipidemia frequently occurs in the absence of other metabolic syndrome risk factors. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:194 –198) he present report is derived from the Genetic Epidemiology of Metabolic Syndrome project. This T is a large, multinational, family-based study exploring affection status: elevated triglycerides and concomitant low high-density lipoprotein (HDL) cholesterol. These criteria were used because they are (1) primary features of atherogenic dyslipidemia, (2) associated with insulin resistance, (3) detectable early in the development of metabolic syndrome, (4) individually highly heritable, and (5) easy to measure. The essential question addressed was how strongly familial atherogenic dyslipidemia is associated with other components of the metabolic syndrome as defined by the ATP-III.1 The analysis thus was designed to determine whether most instances of atherogenic dyslipidemia occur together with the metabolic syndrome or as an isolated dyslipidemia. the genetic basis of the metabolic syndrome. Two simple lipid-based criteria were used to define the From the Department of Medicine (Genetics Program), Boston University School of Medicine, Boston, Massachusetts; Medical Genetics, Genetic Research, GlaxoSmithKline, Collegeville, Pennsylvania; Heart Research Institute, New South Wales, and Hanson Institute, University of Adelaide, Adelaide, Australia; Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, Texas; Department of Internal Medicine and Biocenter Oulu, University of Oulu, Oulu, Finland; Gladstone Institute of Cardiovascular Disease and the University of California, San Francisco, California, and American Hospital, Istanbul, Turkey; University of Ottawa Heart Institute, Ottawa, Canada; and Department of Medicine, Lausanne University Hospital, Lausanne, Switzerland. This research project was funded by GlaxoSmithKline, Inc., Collegeville, Pennsylvania. Manuscript received June 22, 2004; revised manuscript received and accepted August 31, 2004. Address for reprints: Scott M. Grundy, MD, PhD, Center for Human Nutrition, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Y3.206, Dallas, Texas 753909052. E-mail: [email protected]. 194 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 METHODS Accrual centers in Australia, Canada, Finland, Switzerland, Turkey, and the United States participated in this study. Analysis was performed in the epidemiology center in Boston University. Caucasian participants (aged 18 to 70 years) were considered affected if they had both plasma triglycerides Х†75th 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.091 TABLE 1 Plasma Triglycerides and High-density Lipoprotein (HDL) Cholesterol Values Corresponding to Triglycerides Х†75th Percentile and HDL in Х…25th Percentile for Specific Ages by Sex and Site* Australia TG HDL Canada TG Finland HDL TG Switzerland Turkey United States HDL TG HDL TG HDL TG HDL 46 41 41 40 40 40 124 124 133 151 151 151 41 41 42 43 43 39 109 155 191 189 168 120 35 35 35 35 35 35 138 159 198 194 184 174 39 40 36 37 37 37 52 51 50 49 49 49 106 106 106 115 124 124 54 53 52 54 51 51 83 110 136 165 146 145 40 40 40 40 40 40 116 129 145 183 195 171 45 44 44 45 46 46 Male Subjects Age (yrs) 20 30 40 50 60 70 133 151 195 204 204 195 43 39 39 39 39 39 130 154 193 202 201 198 39 39 39 39 39 39 98 148 178 198 190 190 Female Subjects Age (yrs) 20 30 40 50 60 70 124 115 124 168 186 195 46 46 46 46 46 46 121 118 120 170 182 197 46 46 46 46 46 46 94 110 112 133 161 161 *The youngest participants in both Switzerland and Australia were 18 years of age. Despite both Australia and the United States using NHANES data to calculate percentiles, the cutpoints were originally calculated using different age groupings, resulting in slightly different cutpoints by country. Values are expressed as milligrams per deciliter. percentile and serum HDL cholesterol Х…25th percentile, both adjusted for age and gender. Those meeting the criteria will be referred to throughout as affected by current criteria. Lipid values were either currently obtained (80%) or obtained up to 3 years before entry (20%). Population values vary in different countries; in particular, HDL levels in Turkey are lower.2 Thus, cut points were defined by national databases. The following sources were used for different sites: Australia (National Health and Nutrition Examination Survey [NHANES III] survey website [http://www. cdc.gov/nchs/about/major/nhanes/nh3data.htm]); Canada (for subjects Х†18 years, Canadian Heart Health Surveys 1986 to 1992 data,3 and for younger subjects, Lipid Research Clinics prevalence study),4 Finland, cardiovascular risk factor data5; Switzerland, Geneva database on the epidemiology of cardiovascular risk factors in public servants, Turkey [Turkish Heart Study],6 and United States (NHANES III survey as mentioned). Age and gender cut points for the 75th percentiles for triglyceride and HDL cholesterol levels for each study site are listed in Table 1. Subjects were excluded if they (1) had a fasting blood glucose levels Х†126 mg/dl (probands only), (2) had a body mass index Х†35 kg/m2, (3) were positive for human immunodeficiency virus, (4) were recipients of an organ transplant, (5) had familial hypercholesterolemia, or (6) were heavy alcohol users (Пѕ8 U/day). The minimal unit for collection of a family was an affected sib-pair, but larger nuclear or extended families were collected if possible. In a small number of instances, some patients originally believed to be affected, on closer examination turned out not to be affected, leading to a small proportion of families with only 1 affected. An additional 464 singly-ascer- tained patients were recruited from the same sites with the same lipid phenotype as that used for the family collection. The institutional review board at each participating site approved the protocol and informed consent forms. A standardized questionnaire was administered to every family member. In addition to demographic data, it captured co-morbid conditions, medications, tobacco, and alcohol. Height, weight, and waist circumference, as well as 3 blood pressure measurements, were obtained for each subject. The average of the second and third systolic and diastolic blood pressure readings was used in the analyses. Blood was collected after a 12-hour fast. Laboratory tests for the lipids and glucose levels were performed according to standard procedures. Statistical analyses were performed using SAS version 8 (SAS Institute, Cary, North Carolina) and Stata version 8. Descriptive statistics are expressed as number (percent) or mean П® SD, except for triglyceride levels and blood pressure measurements, which are expressed as medians. Statistical significance was determined using univariate logistic regression for continuous variables or a chi-square test for categorical variables. For the linear regression analysis, PROC GLM in SAS (SAS Institute) was used; waist circumference, fasting glucose, and systolic blood pressure were treated as continuous variables and currently affected status was a categorical variable. RESULTS In all, 3,273 patients from 504 families were recruited in the study. The 6 accrual centers contributed 33 to 136 families, and the size of the families ranged from 2 to 179. Of 3,273 participants from the family PREVENTIVE CARDIOLOGY/ATHEROGENIC DYSLIPIDEMIA AND METABOLIC SYNDROME 195 TABLE 2 Characteristics of Participants Based on the Genetic Epidemiology of Metabolic Syndrome Criteria* Men Variable Age (yrs) Ever smoked Current smoker No. of cigarettes/d No. of alcoholic drinks/wk Waist circumference (cm) High waist circumference Пѕ102 cm (M), 88 cm (F) Waist-to-hip ratio Body mass index (kg/m2) Body mass index Х†30 kg/m2 Diagnosed diabetes mellitus Diagnosed systemic hypertension Taking blood pressure medication Blood pressure (median, mm Hg) Elevated blood pressure†Taking lipid-altering medication Total cholesterol (mg/dl) HDL cholesterol (mg/dl) HDL cholesterol ПЅ40 mg/dl (M), ПЅ50 mg/dl (F) LDL cholesterol (mg/dL) Triglycerides (median, mg/dl) Triglycerides (Х†150 mg/dl) Fasting glucose (mg/dl) Glucose Х†110 mg/dl‡ Women Affected (n П 854) Unaffected (n П 737) Affected (n П 582) Unaffected (n П 935) 45.9 П® 14.0 541 (63%) 229 (26%) 21.2 П® 14.3 5.5 П® 6.7 100.2 П® 12.0 333 (41%) 0.94 П® 0.06 28.2 П® 3.9 291 (34%) 75 (9%) 275 (32%) 208 (24%) 133/84 583 (69%) 282 (33%) 211.9 П® 56.1 32.4 П® 5.4 829 (97%) 118.2 П® 40.5 265.5 796 (93%) 99.6 П® 23.5 143 (17%) 40.4 П® 18.4В§ 392 (53%)В§ 210 (29%)В§ 19.3 П® 13.1 5.3 П® 6.4 90.0 П® 14.4В§ 121 (17%)В§ 0.89 П® 0.08В§ 25.1 П® 4.4В§ 96 (13%)В§ 25 (3%)В§ 110 (15%)В§ 74 (10%) 128/80В§ 378 (53%)В§ 64 (9%)В§ 188.2 П® 46.0В§ 46.0 П® 11.4В§ 221 (30%)В§ 118.1 П® 39.9 109.0В§ 190 (26%)В§ 91.9 П® 17.9В§ 52 (7%)В§ 45.3 П® 16.4 244 (42%) 105 (18%) 15.5 П® 11.3 3.6 П® 3.8 91.5 П® 14.0 332 (60%) 0.85 П® 0.08 28.1 П® 4.7 211 (36%) 60 (11%) 198 (34%) 165 (28%) 131/83 380 (67%) 140 (24%) 210.0 П® 49.6 37.3 П® 6.5 573 (98%) 121.8 П® 42.8 215.0 476 (82%) 99.0 П® 31.3 84 (15%) 42.1 П® 18.0В§ 326 (35%)К€ 151 (16%) 12.5 П® 9.4К€ 4.1 П® 5.5 81.7 П® 12.9В§ 270 (30%)В§ 0.82 П® 0.09В§ 24.7 П® 4.5В§ 128 (14%)В§ 19 (2%)В§ 148 (16%)В§ 108 (12%)В§ 129/79В§ 422 (46%)В§ 60 (7%)В§ 192.6 П® 43.9В§ 57.0 П® 14.0В§ 290 (31%)В§ 114.8 П® 38.1К€ 88.7В§ 129 (14%)В§ 88.3 П® 16.8В§ 32 (3%)В§ *The Genetic Epidemiology of Metabolic Syndrome plasma lipid criteria: triglycerides Х†75th percentile, HDL cholesterol Х…25th percentile (age-adjusted, country and gender-specific). †Systolic Х†130 mm Hg and/or diastolic Х†85 mm Hg. ‡ Does not include treated diabetics В§ p Х…0.05; К€p Х…0.01; В§p Х…0.001 versus affected subjects. Values are expressed as number (%) or mean П® SD. ATP-III criteria to define metabolic syndrome are depicted in italics. TABLE 3 Characteristics of Singly-ascertained Cases Variable Men Age (yrs) Ever smoked Current smoker No. of cigarettes/d No. of alcoholic drinks/wk Waist circumference (cm) High waist circumference Пѕ102 cm (M), Пѕ88 cm (W) Waist-to-hip ratio Body mass index (kg/m2) Body mass index Х†30 kg/m2 Diagnosed hypertension Blood pressure (median, mm Hg) Elevated blood pressure Taking lipid medication Total cholesterol (mg/dl) HDL cholesterol (mg/dl) HDL cholesterol ПЅ40 mg/dl (M), ПЅ50 mg/dl (W) LDL cholesterol (mg/dl) Triglycerides (median, mg/dl) Triglycerides Х†150 mg/dl Fasting glucose (mg/dl) Glucose Х†110 mg/dl Cases (n П 464) 340 (73%) 50.6 П® 8.6 295 (64%) 109 (24%) 21.0 П® 13.3 NA 97.6 П® 15.3 230 (50%) 0.92 П® 0.07 28.9 П® 3.3 179 (39%) 160 (34%) 132/81 279 (60%) 206 (45%) 221.9 П® 45.3 35.7 П® 6.2 429 (92%) 132.1 П® 45.3 310.7 443 (95%) 96.1 П® 11.2 43 (11%) Values are presented as number percentage or mean П® SD. LDL П low-density lipoprotein. 196 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 dataset, 1,436 (44%) were considered as affected using the current definition, 1,672 (51%) were unaffected, and the remaining 165 patients (5%) were considered unknown due to absence of reliable measurements of either HDL cholesterol or triglycerides and thus were excluded from analysis. The clinical characteristics of the 1,436 (854 men/582 women) affected and the 1,672 (737 men/935 women) unaffected family members, as well as the proportion of these patients who met each patient ATP-III criterion, are listed in Table 2. Age, waist circumference, waistto-hip ratio, body mass index, the proportion of obese patients, blood pressure, and the proportion of hypertensive subjects, blood glucose levels, and proportion of persons with blood glucose levels Пѕ110 mg/dl were all significantly higher among affected versus unaffected men and women (p ПЅ0.001). These differences were not due to differences in alcohol consumption or current cigarette smoking. In addition, a separate set of 464 unrelated patients was ascertained based on the same criteria as the familial subjects, except that they were not required to have an affected relative. The characteristics of this singly-ascertained group are listed in Table 3. To explore the concordance between the current and the ATP-III definitions for metabolic syndrome, JANUARY 15, 2005 FIGURE 1. Frequency distribution (percent) of genetic epidemiology of metabolic syndrome-affected (closed bars) and -unaffected (open bars) patients according to the number of ATP-III criteria and age. (ATP-III criteria to define metabolic syndrome are listed in Table 1.) The diagnosis of metabolic syndrome is made if >3 of these criteria are met (corresponding to right side of dotted line). we first partitioned 3,030 subjects from the family dataset who had sufficient information for ATP-III classification according to these 2 definitions. Among those affected by current criteria, 76% met the ATPIII definition for metabolic syndrome, whereas 85% of those unaffected by current criteria were also negative for the ATP-III definition, with a Cohen’s вђ¬ of 0.61 (95% confidence interval [CI] 0.59 to 0.64), indicating overall substantial agreement. To establish how much of the discordance between the 2 definitions was due to our different criteria for dyslipidemia, and how much was due to coexistence of dyslipidemia with other metabolic syndrome factors, we repeated this analysis after replacing the absolute thresholds for triglyceride and HDL cholesterol in ATP-III with the current thresholds to create a pseudodefinition and again looked for concordance with the current definition. Among those positive for both definitions, 81% also met the ATP-III definition for metabolic syndrome, whereas 91% of those unaffected by current criteria were also unaffected by ATP-III, with a Cohen’s вђ¬ of 0.72 (95% CI 0.69 to 0.74). Thus, overall concordance was increased, and only a small proportion of subjects (4.7%) affected by current criteria did not achieve ATP-III status due to differences in the lipid thresholds. We next stratified the participants in this study according to the number of ATP-III criteria they met and by age categories (Figure 1). In this figure, solid bars located left of the dotted line correspond to “false positives,” whereas solid bars on the right correspond to “true positives,” if the ATP-III definition is used as the “gold standard” to define metabolic syndrome. This analysis revealed that in the age group Х…35 years old, Пѕ50% of the participants (58%) who were affected by current criteria were not affected by ATPIII. In contrast, the rate of false positives decreased to 14% in those aged 36 to 55 years and to 12% in older subjects. The overall agreement between the 2 criteria as described by the Cohen’s вђ¬ showed a similar pattern, with only moderate agreement in those ПЅ35 years (вђ¬ 0.42, 95% CI 0.36 to 0.48), but substantial agreement in both those 35 to 55 years (вђ¬ 0.66, 95% CI 0.60 to 0.71) and in those Пѕ55 years (вђ¬ 0.58, 95% CI 0.51 to 0.65). Because currently defined status was fairly well correlated with ATP-III status overall, it was interesting to explore which nondyslipidemic components of ATP-III (waist, fasting glucose, and systolic blood pressure) were most closely associated with the current definition of dyslipidemia. A linear regression analysis performed on waist circumference, fasting glucose, and systolic blood pressure is presented in Table 4. The parameter estimates correspond to the mean increase in the trait under evaluation in affected subjects by current criteria compared with unaffected subjects. Univariate analysis indicates unadjusted association, whereas model 1 adjusts for standard confounders (age, gender, center, smoking, and low-density lipoprotein cholesterol), and model 2 further adjusts for metabolic syndrome-associated factors. Affection status by current data was closely associated with each of the 3 traits in the univariate analysis, and the association remained significant after adjustment for confounders, although with a reduced magnitude of effect for both waist circumference and fasting glucose levels. In contrast, after adjustment for factors related to metabolic syndrome, systolic blood pressure was no longer associated with status by current criteria, suggesting that the observed association with current data was a secondary effect of Х†1 of the syndromic factors. Further analysis showed that waist was the specific variable that ablated this association (data not shown). To assess if the concordance between the 2 definitions was influenced by the familial nature of dyslipidemia in the present study, the same comparison was performed in the additional set of 464 patients affected by current criteria who were singly ascertained. Overall, 75% of these sporadic cases also were affected for ATP-III metabolic syndrome, a figure similar to that observed in the family sample. Thus, the level of concordance between the 2 criteria does not appear to have been substantially affected by the familial nature of affected family members. PREVENTIVE CARDIOLOGY/ATHEROGENIC DYSLIPIDEMIA AND METABOLIC SYNDROME 197 TABLE 4 Association of the Genetic Epidemiology of Metabolic Syndrome Criteria With Systolic Blood Pressure, Waist Circumference, and Fasting Glucose Systolic Blood Pressure (mm Hg) Univariate Model 1* Model 2†Parameter Estimate 95% CI 7.1 3.8 0.6 5.6–8.6 2.4–5.1 ПЄ0.9–2.2 p Value ПЅ0.0001 ПЅ0.0001 0.43 Waist Circumference (cm) Parameter Estimate 95% CI 11.3 8.2 7.0 10.3–12.3 7.3–9.1 6.0–7.9 p Value ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 Fasting Glucose (mg/dl) Parameter Estimate 95% CI 9.5 7.0 3.6 7.9–11.1 5.4–8.6 1.8–5.3 p Value ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 *Adjusted for age, gender, center, smoking, and LDL cholesterol. †Systolic blood pressure further adjusted for waist, fasting glucose, and hypertension medications; waist circumference further adjusted for systolic blood pressure, fasting glucose, and hypertension medications; fasting glucose further adjusted for waist, systolic blood pressure, and hypertension medications. DISCUSSION In our family collection, 86% of Caucasian patients aged Пѕ35 years affected with atherogenic dyslipidemia met the ATP-III definition for metabolic syndrome. A similar high association was present in sporadic cases of atherogenic dyslipidemia. Thus, most middle-aged subjects who have atherogenic dyslipidemia seemingly will have it as a component of the metabolic syndrome. This supports our contention that atherogenic dyslipidemia, whether familial or sporadic, is a robust identifier of predisposition for the syndrome. The strongest association was with abdominal obesity. The weaker concordance between atherogenic dyslipidemia and ATP-III metabolic syndrome before age 35 years may signify a particularly strong predisposition to an isolated lipid disorder, but in many cases, it likely foreshadows the development of the metabolic syndrome later in life. The relation between dyslipidemia and systolic blood pressure was less striking, suggesting a weaker causal linkage. This corroborates well with published findings, where a factor analyses investigating metabolic syndrome shows that blood pressure often falls out as a distinct factor,7 although it often correlates with obesity.8 Elevated plasma glucose was also not strongly associated, but here it is known that compensatory hyperinsulinemia can prevent the development of overt hyperglycemia for many years in persons with the metabolic syndrome. Finally, it is known that the metabolic syndrome manifests differently in different populations, and the essential findings obtained in the current Caucasian population will not necessarily be applicable to other populations, notably blacks, Hispanics, and Asians. Acknowledgment: We thank all the families whose participation made this project possible. We gratefully acknowledge the contributions of the study personnel at each of the collaborating sites: John Farrell, 198 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 Nicholas Nikolopoulos and Maureen Sutton (Boston); Judy Walshe, Monica Prentice, Anne Whitehouse, Julie Butters, and Tori Nicholls (Australia); Heather Doelle, Lynn Lewis, and Anna Toma (Canada); Kari Kervinen, Seppo Poykko, Liisa Mannermaa, and Sari Paavola (Finland); Claire Hurrel, Diane Morin, Alice Mermod, Myriam Genoud, and Roger Darioli (Switzerland); Guy Pepin, Sibel Tanir, Erhan Palaoglu, Kerem Ozer, Linda Mahley, and Aysen Agacdiken (Turkey); and Deborah A. Widmer, Rhonda Harris, and Selena Dixon (United States). We also would like to thank Allen Roses, Shyama Brewster, and Julia Perry for their support, and Dennis Sprecher, Annie McNeill, and Sara Ephross for their useful discussions. 1. National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment III). Third report of the National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) final report. Circulation 2002;106:3143–3421. 2. Bersot TP, Vega GL, Grundy SM, Palaoglu KE, Atagunduz P, Ozbayrakci S, Gokdemir O, Mahley RW. Elevated hepatic lipase activity and low levels of high density lipoprotein in a normotriglyceridemic, nonobese Turkish population. J Lipid Res 1999;40:432– 438. 3. Langille DB, Joffres MR, MacPherson KM, Andreou P, Kirkland SA, MacLean DR. Prevalence of risk factors for cardiovascular disease in Canadians 55 to 74 years of age: results from the Canadian Heart Health Surveys, 1986 –1992. Can Med Assoc J 1999;161(suppl 8):S3–S9. 4. The Lipid Research Clinics Program Epidemiology Committee. Plasma lipid distributions in selected North American populations: the Lipid Research Clinics Program Prevalence Study. Circulation 1979;60:427– 439. 5. Vartiainen E, Jousilahti P, Alfthan G, Sundvall J, Pietinen P, Puska P. Cardiovascular risk factor changes in Finland, 1972–1997. Int J Epidemiol 2000;29:49 –56. 6. Mahley RW, Palaoglu KE, Atak Z, Dawson-Pepin J, Langlois AM, Cheung V, Onat H, Fulks P, Mahley LL, Vakar F. Turkish Heart Study: lipids, lipoproteins, and apolipoproteins. J Lipid Res 1995;36:839 – 859. 7. Hanley AJ, Karter AJ, Festa A, D’Agostino R Jr, Wagenknecht LE, Savage P, Tracy RP, Saad MF, Haffner S. Factor analysis of metabolic syndrome using directly measured insulin sensitivity: The Insulin Resistance Atherosclerosis Study. Diabetes 2002;51:2642–2647. 8. Meigs JB, D’Agostino RB Sr, Wilson PW, Cupples LA, Nathan DM, Singer DE. Risk variable clustering in the insulin resistance syndrome. The Framingham Offspring Study. Diabetes 1997;46:1594 –1600. JANUARY 15, 2005 New Predictors of Outcome in Idiopathic Pulmonary Arterial Hypertension Steven M. Kawut, MD, Evelyn M. Horn, MD, Ketevan K. Berekashvili, MD, Robert P. Garofano, EdD, Rochelle L. Goldsmith, EdD, Allison C. Widlitz, PA, MS, Erika B. Rosenzweig, MD, Diane Kerstein, MD, and Robyn J. Barst, MD Idiopathic pulmonary arterial hypertension (PAH) is a rare disease with a poor prognosis. New therapies have improved the outcome of this condition; accordingly, the factors that determine outcome may have changed. We aimed to identify determinants of survival in a cohort of consecutive patients with PAH: which was idiopathic, familial, or associated with anorexigen use. We performed a retrospective cohort study of 84 consecutive patients with PAH who underwent initial evaluation at our center from January 1994 to June 2002. The primary outcome was death or lung transplantation. Survival at 1, 2, and 3 years was 87%, 75%, and 61%, respectively. Multivariate analysis showed that being of African-American or Asian descent was associated with an increased risk of death. Warfarin use was associated with a reduced risk of death. Higher serum albumin and cardiac index and acute vasoreactivity were independently associated with improved survival. These data suggest that the determinants of outcome have changed. Race is identified as a new risk factor, which may be attributable to biologic or socioeconomic differences. Cardiac function and acute reactivity of the pulmonary vascular bed remain strong independent predictors of outcome. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:199 –203) he National Institutes of Health (NIH) Patient Registry for the Characterization of Primary Pulmonary T Hypertension found that right-sided heart function (as- PAH who underwent initial evaluation from January 1994 to June 2002 with follow-up through June 2003 at our center. The study was approved by the institutional review board. Study subjects: To assemble the cohort, we queried the Data Warehouse at our institution for all patients with an International Classification of Diseases-Ninth Revision code for primary or secondary pulmonary hypertension who were assessed by clinicians at our center. Second, we hand searched all outpatient records from our center. Last, we reviewed a comprehensive list of all patients from our center who underwent cardiac catheterization since 1994. The following criteria were required for inclusion in the study: (1) PAH that was idiopathic, familial, or associated with anorexigen use5; (2) age Пѕ16 years; and (3) initial evaluation performed at our center between January 1994 and July 2002. The exclusion criteria were (1) previous cardiac catheterization with acute vasodilator study and initiation of PAH therapy and (2) other forms of PAH. Usual clinical and laboratory criteria were used to rule out other diseases associated with PAH (e.g., systemic sclerosis and systemic lupus erythematosus). Measurements: Patients routinely underwent laboratory testing, pulmonary function testing, radionuclide angiography,6 echocardiography, cardiopulmonary exercise testing, and cardiac catheterization at initial evaluation. During cardiac catheterization, short-acting vasodilators, such as inhaled nitric oxide or intravenous epoprostenol, were administered to test acute vasoreactivity. Acute vasoreactivity was defined as a reduction in mean pulmonary artery pressure of Пѕ20% without a decrease in cardiac index. Exercise testing, cardiac catheterization, or vasodilator testing were not performed if deemed unsafe (e.g., severe symptoms at rest). sessed by right atrial pressure and cardiac index) and severity of pulmonary arterial hypertension (PAH) (assessed by pulmonary artery pressure) were associated with survival.1 Since the NIH Registry, clinical trials have demonstrated the effectiveness of several therapies in idiopathic PAH: continuous intravenous epoprostenol, continuous subcutaneous treprostinil, and oral bosentan have been approved for treatment of PAH.2– 4 In light of these new therapies, it is possible that the prognosis of PAH has been altered. Accordingly, the determinants of survival may have also changed. If so, the NIH Registry prediction rule may no longer provide accurate prognostic information. Our first aim was to identify the determinants of survival in idiopathic PAH in the recent therapeutic era. Our second aim was to reevaluate the accuracy and generalizability of the NIH prediction rule for patients with idiopathic PAH diagnosed during the last decade. METHODS Study design: We performed a retrospective cohort study of all consecutive adult patients with idiopathic From the Departments of Internal Medicine and Pediatrics, College of Physicians and Surgeons, and Department of Epidemiology, Joseph L. Mailman School of Public Health, Columbia University, New York, New York. Dr. Kawut was supported by Grant HL67771 from the National Institutes of Health, Bethesda, Maryland, and the Florence and Herbert Irving Clinical Research Career Award, New York, New York. Manuscript received July 21, 2004; revised manuscript received and accepted September 8, 2004. Address for reprints: Steven Kawut, MD, Division of Pulmonary, Allergy, and Critical Care Medicine, PH 8E, Room 101, 622 West 168th Street, New York, New York 10032. E-mail: sk2097@ columbia.edu. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.006 199 TABLE 1 Baseline Laboratory (n П 84), Pulmonary Function Tests (n П 84), and Cardiopulmonary Exercise Tests (n П 72) Data Characteristic Blood urea nitrogen (mg/dl) Creatinine (mg/dl) White blood cell count (10 В· 9/l) Hemoglobin (g/dl) Hematocrit (%) Mean corpuscular volume (fl) Platelets (10 В· 9/l) Total bilirubin (mg/dl)* Direct bilirubin (mg/dl)* Alkaline phosphatase (U/L)* Alanine aminotransferase (U/L)* Aspartate aminotransferase (U/L)* Albumin (g/dl)* Prothrombin time (s)†Positive antinuclear antibody‡ Forced vital capacity (% predicted) Forced vital capacity (L) Forced expiratory volume in 1 s (% predicted) Forced expiratory volume in 1 s (L) Total lung capacity (% predicted)В§ Total lung capacity (L)В§ Diffusion capacity for carbon monoxide adjusted for hemoglobin (ml/min/ mm Hg)‡ Diffusion capacity for carbon monoxide adjusted for hemoglobin (% predicted)‡ Peak heart rate (beats/min) Peak systolic blood pressure (mm Hg)К€ Peak diastolic blood pressure (mm Hg)К€ Peak oxygen consumption (ml/kg/min) Peak oxygen pulse (ml/beat) Peak respiratory exchange ratio Peak ventilatory equivalent for carbon dioxide Peak minute ventilation (L/min) TABLE 2 Hemodynamic Parameters at Baseline and With Acute Vasodilator Testing Value 17 (14–20) 0.9 (0.8–1.1) 7.8 (6.8–9.6) 14.8 (14–16) 43.9 (41.6–47.6) 88.6 (86.5–92.4) 212 (178–262) 0.9 (0.7–1.3) 0.2 (0.1–0.3) 84 (62–111) 23 (17–32) 24.5 (20–32) 4.5 (4.2–4.9) 13.4 (12.7–14.3) 16 (20%) 90 П® 20 3.1 П® 1.0 86 П® 18 2.4 93 4.8 16 П® П® П® П® 0.7 16 1.1 6 63 П® 22 137 133 77 11.4 6.1 1.1 56 П® П® П® П® П® П® П® 21 26 13 4 2.5 0.1 15 47 П® 17 Values are expressed as mean П® SD, median (interquartile range) or number (%). *n П 82, †n П 81, ‡n П 80, В§n П 70, К€n П 71. The choice of therapy adhered to usual algorithms.7 Patients with acute vasoreactivity were treated with calcium channel blockers. Nonreactive patients were treated with bosentan (when available) for New York Heart Association (NYHA) class III and IV symptoms, treprostinil (when available) for NYHA III or IV symptoms, and intravenous epoprostenol for NYHA III symptoms (before the availability of bosentan and treprostinil) and NYHA IV symptoms. Although certain therapies were available at different times during the study period, we adjusted for these differences in our analysis. The primary combined end point was death or lung transplantation. We assessed outcomes through chart review and query of the Social Security Death Index. A patient was censored as alive at the last medical contact recorded in the computerized clinical database through June 2003. Statistical analysis: Continuous variables were summarized by mean П® SD or median (interquartile range). Categorical variables were summarized by frequencies with 95% confidence intervals (CIs). Sur200 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 Variable Heart rate (beats/min) Mean pulmonary artery pressure (mm Hg) Mean right atrial pressure (mm Hg) Cardiac index (L/min/m2) Pulmonary vascular resistance index (U В· m2) Mixed venous oxygen saturation (%) Baseline (n П 80)* With Acute Vasodilator (n П 74) 82 (74–93) 55 (48–61) 81 (67–92) 52 (42–58) 8 (5–12) 8 (5–12)†1.8 (1.5–2.4) 27 (16–33) 2.1 (1.7–2.5) 21 (13–29) 60 (50–66) 64 (57–69) Values are expressed as median (interquartile range). *Baseline hemodynamics were similar for patients with acute vasodilator studies. †n П 60. vival analyses were performed using Cox proportional-hazards models.8 We constructed univariate models for each variable. Therapies were modeled using timevarying covariates, which allow patient-time to be accurately classified according to treatment status. For example, we included patient-time while receiving various therapies (e.g., warfarin, calcium channel blockers, epoprostenol, bosentan, and/or treprostinil), while also including patient-time before initiation and after discontinuation of these therapies. This statistical method allowed us to control for the availability and use of different therapies in patients at different times over the course of the study. Multivariate models were constructed using covariates with a p value ПЅ0.20 on univariate analysis or that were believed to be clinically important. Likelihood ratio testing was used to identify the most parsimonious model. The variables in the final multivariate model met the assumption of proportional hazards.9 We also assessed the accuracy of the NIH Registry prediction rule.10 “Discrimination” describes the ability of the rule to distinguish patients who die from patients who live and is expressed by the area under the receiveroperating characteristic curve (AUC). The AUC ranges from 0.5 (for chance) to 1.0 (for perfect prediction). “Calibration” is the ability of the rule to accurately predict the probability of death for a patient. To assess calibration, we divided the cohort into quintiles based on the predicted probability of death or transplant and calculated the Hosmer-Lemeshow goodness-of-fit statistic. A p value Пѕ0.05 indicated adequate prediction. Univariate analyses were performed with available data. We performed simple imputation for missing data points in the multivariable analysis. We performed sensitivity analyses using mean values and the “worst” value in the data set in place of the missing values for the multivariate analysis. A p value ПЅ0.05 was considered statistically significant. Stata, version 7.0 (College Station, Texas), was used for all analyses. JANUARY 15, 2005 RESULTS Our cohort was composed of 84 newly diagnosed patients with PAH (mean age 42 П® 14 years). Sixty-eight (81%) were women. The cohort included 69 whites (82%, of whom 9 were Hispanic and 60 were non-Hispanic) and 15 non-whites (18%, 9 Asians and 6 African-Americans). Sixtysix patients (78%) had idiopathic PAH, 14 (17%) had familial PAH, and 4 (5%) had PAH related to anorexigen use. The median time between the performance of echocardiography and/or cardiac catheterization consistent with a diagnosis of pulmonary hypertension and evaluation at our center was 181 days (interquartile range 91 to 374). The median duration of symptoms before our evaluation was 584 days (interquartile range 251 to 1,392; n П 81). Seventy-nine patients (94%) were treated with warfarin and 72 (86%) with digoxin at baseline or after evaluation. Thirty-three patients (39%) received oral calcium channel blockers and 61 (73%) received spironolactone. Thirty-eight patients (45%) were treated with intravenous epoprostenol, 31 within 60 days of the baseline catheterization. Twelve patients (14%) were treated with subcutaneous treprostinil and 23 (27%) with oral bosentan. Eleven patients (13%) received combination therapy with bosentan and either intravenous epoprostenol or subcutaneous treprostinil. Laboratory, pulmonary function, cardiopulmonary exercise test, and hemodynamic data are listed in Tables 1 and 2. Four patients did not undergo rightsided cardiac catheterization due to severity of illness and safety concerns; 10 patients did not undergo vasodilator testing. Echocardiography showed a pericardial effusion in 28 of 83 patients (34%). The peak tricuspid regurgitant gradient was 83 П® 24 mm Hg (n П 74). Right ventricular ejection fraction on radionuclide angiography was 30 П® 9% (n П 65, normal Х†45%).11 Survival: The median follow-up was 764 days (interquartile range 505 to 1,573). Twenty-four patients died and 1 patient underwent lung transplantation. Transplant-free survival at 1 year was 87% (95% CI 77% to 93%), at 3 years 75% (95% CI 63% to 84%), and at 5 years 61% (95% CI 45% to 73%) (Figure 1). Univariate survival analyses: We found an increased risk of death for non-whites (Asians and AfricanAmericans) in our cohort (Table 3). Hepatic and renal dysfunction were associated with an increased risk of death. The presence of antinuclear antibodies (titer Х†1:40) was not associated with outcome (hazard ratio [HR] 1.6, 95% CI 0.63 to 4.1, p П 0.32, n П 80). The time of patient evaluation was not associated with survival (data not shown). Lower systolic blood pressure and higher ventilatory equivalent for carbon dioxide at peak exercise were associated with worse outcome (p ПЅ0.05); however, other parameters from cardiopulmonary exercise testing were not (data not shown). A pericardial effusion on echocardiography more than doubled the risk of death (HR 2.6, 95% CI 1.2 to 5.9, p П 0.017), whereas the peak tricuspid regurgitant gradient was FIGURE 1. Kaplan-Meier estimate of transplant-free survival in patients with idiopathic pulmonary arterial hypertension. not associated with outcome (HR 1.0, 95% CI 0.98 to 1.02, p П 0.83). A 5% increase in right ventricular ejection fraction on radionuclide angiography was associated with a reduced risk of death (HR 0.62, 95% CI 0.43 to 0.89, p П 0.009). Multivariate survival analysis: Asian or AfricanAmerican race, low serum albumin, lack of warfarin use, low cardiac index, and absence of acute vasoreactivity were independently associated with an increased risk of death (Table 4). Replacement of cardiac index with right ventricular ejection fraction in the model showed similar results. Repeat analysis of the cohort without potentially influential and/or outlying subjects did not change the conclusions, supporting the robustness of the final multivariate model. The effect estimates from the sensitivity analyses for missing data were not substantially different from those using simple imputation (data not shown). Assessment of the NIH prediction rule: The NIH prediction rule was poorly calibrated to predict the probability of 1-, 2-, and 3-year survival in our cohort (p ПЅ0.005). Reclassifying the 1 patient who underwent lung transplantation as alive did not change the results. Although better than chance alone, the NIH prediction rule was not sufficiently discriminating between patients who died and those who were alive at each time period (1-year AUC 0.76 [95% CI 0.59 to 0.93], 2-year AUC 0.70 [95% CI 0.55 to 0.86], and 3-year AUC 0.72 [95% CI 0.57 to 0.88]). DISCUSSION This is the largest study of outcomes in patients with idiopathic PAH treated with currently approved PAH therapies. We have shown that cardiac index, acute vasoreactivity, race, and serum albumin independently predict transplant-free survival in patients with idiopathic PAH. Although warfarin use was associated with increased survival, the use of other PAH medications was not. These results differ from those of recently published studies.12–15 Our data also demMISCELLANEOUS/NEW PREDICTORS OF OUTCOME IN PAH 201 relation to outcome. Although rightsided cardiac function continues to Variable HR 95% CI p Value be an important factor in determining survival, the currently approved theraAge 0.99 0.96–1.0 0.44 peutic agents may have “uncoupled” Gender: men 1.0 0.40–2.6 0.98 Body mass index 0.96 0.90–1.0 0.29 the former association between seRace verity of PAH and outcome. Asian or African-American 3.2 1.3–7.6 0.008 The purpose of acute vasoreactivWhite 1.0 — — ity testing in PAH is to predict the Diagnosis Idiopathic 1.0 — — response to chronic oral calcium Familial 1.6 0.58–4.2 0.37 channel blockade.20,21 Recently, inRelated to anorexigens 2.3 0.53–10 0.27 vestigators have questioned the vaWarfarin 0.33 0.12–0.90 0.03 lidity of using only a 20% decrease Digoxin 0.57 0.23–1.4 0.24 in mean pulmonary artery pressure in Spironolactone 2.1 0.87–5.1 0.10 Calcium channel blockers 0.34 0.08–1.4 0.14 defining acute vasoreactivity, beEpoprostenol 1.4 0.62–3.1 0.42 cause near normalization of pulmoBosentan 0.64 0.15–2.7 0.54 nary artery pressures is a better preTreprostinil 0.64 0.15–2.7 0.54 dictor of the long-term response to Combination therapy 0.75 0.10–5.6 0.78 Blood urea nitrogen 1.05 1.0–1.1 0.04 chronic calcium channel blockade.22 Creatinine 6.1 1.7–22 0.006 However, our data suggest that paHematocrit 1.0 0.95–1.1 0.41 tients with an acute response to vaPlatelet count 1.0 0.99–1.0 0.28 sodilator testing have an improved Prothrombin time 1.1 1.0–1.1 0.025 survival compared with patients who Total bilirubin 1.9 1.2–2.8 0.003 Direct bilirubin 15.3 3.7–63 0.001 are nonresponders, irrespective of Albumin 0.46 0.19–1.11 0.09 treatment. This suggests that acute Alanine aminotransaminase 1.01 1.0–1.02 0.13 vasoreactivity has implications reAspartate transaminase 1.02 1.0–1.03 0.055 garding survival beyond the theraAlkaline phosphatase 1.01 1.0–1.02 0.006 Forced vital capacity % predicted 0.97 0.95–0.99 0.015 peutic response to calcium channel Forced expiratory volume in 1-s % 0.97 0.95–0.99 0.035 blockers. predicted Although we did not find signifiTotal lung capacity % predicted 0.96 0.93–0.99 0.025 cant associations between the variDiffusion capacity for carbon dioxide 1.0 0.99–1.03 0.54 ous PAH therapies and survival, adjusted for hemoglobin % predicted these results should be interpreted Heart rate 1.06 1.02–1.1 0.005 with caution. Confounding by disMean pulmonary artery pressure 1.02 0.98–1.05 0.29 ease severity may bias the associaMean right atrial pressure 1.05 0.99–1.1 0.09 tions between therapies and outCardiac index 0.36 0.17–0.76 0.008 Pulmonary vascular resistance index 1.03 1.01–1.06 0.005 comes. For example, sicker patients Mixed venous oxygen saturation 0.94 0.90–0.98 0.003 may receive intravenous epoprosteAcute vasoreactivity 0.11 0.01–0.81 0.03 nol earlier than healthier patients. Therefore, epoprostenol may appear to be associated with worse outcomes in a cohort study, despite definitive evidence TABLE 4 Multivariate Analysis of Transplant-free Survival that epoprostenol improves survival.4 Accordingly, a Variable HR 95% CI p Value cohort study cannot estimate the clinical effects of therapies. However, we can conclude that adjustment Asian or African-American 4.3 1.7–11 0.002 for differing therapeutic regimens did not affect the race Warfarin use 0.35 0.12–0.99 0.05 main determinants of survival in the final multivariate Albumin 0.37 0.16–0.84 0.019 model. Cardiac index 0.41 0.19–0.90 0.026 We found an association between warfarin use and Acute vasoreactivity 0.13 0.02–0.96 0.046 improved survival, consistent with other nonrandomized studies.20,23–25 Anticoagulation with warfarin may prevent in situ thrombosis, a frequent finding in onstrate that right ventricular ejection fraction on ra- the pulmonary vasculature in PAH, and venous thromdionuclide angiography is associated with survival in boembolism. However, anticoagulation is withheld in PAH. This could be a useful clinical parameter for patients with previous bleeding episodes and/or severe evaluating disease severity and a potential surrogate thrombocytopenia, so that confounding by disease severity may explain these findings. A randomized conend point for PAH clinical trials.16 –19 The degree of PAH and right-sided cardiac dys- trolled trial of anticoagulation in PAH may therefore function have been considered the most significant be warranted. Traditionally, liver function test abnormalities in factors in the disease course.1 We found that cardiac index and right ventricular ejection fraction predicted PAH have been believed to merely reflect increased right survival, whereas pulmonary artery pressure had no atrial pressure. However, we identified low albumin as a TABLE 3 Univariate Analyses of Transplant-free Survival 202 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 risk factor for mortality, independent of right atrial pressure or cardiac index. Although a low albumin may result from malnutrition or chronic losses, these findings highlight the need for further investigation of hepatic dysfunction in patients with PAH. To our knowledge, our study is the first to suggest a difference in survival in PAH based on race.1,12,14,26,27 Although 1 study found an increased age-specific mortality for African-American women with PAH, the diagnosis was based on International Classification of Diseases-Ninth Revision coding without clinical data, and actual survival was not assessed.28 Minority patients may be identified later in their disease course; however, we did not find associations between race and duration of symptoms, referral time, or hemodynamics. Our findings are unfortunately consistent with studies of health disparities in other cardiovascular diseases and should spur outreach to “at-risk” patients and the physicians who serve them.29 Despite similar demographics and possibly lower mean cardiac index in our cohort compared with the NIH Registry, the NIH prediction rule underestimated the probability of survival in each quintile of risk in our cohort (data not shown), likely attributable to advances in treatment. Because adequate calibration of a predictive index is most important for counseling patients regarding their prognosis,10 the NIH rule may no longer be sufficient for that task. These findings require confirmation in other centers. Although there were missing data for some variables, sensitivity analyses based on different imputation methods did not change our conclusions. Finally, the small number of events limited the precision of certain analyses; however, we did not find significant model “overfitting.” In conclusion, cardiac function and acute reactivity of the pulmonary vascular bed at diagnosis remain strong predictors of outcome in PAH regardless of which therapy is initiated. Asian or African-American race and low albumin are new independent risk factors. Right ventricular ejection fraction from radionuclide angiography predicts long-term outcome in PAH. Future studies should derive and validate new prediction tools for patients with PAH diagnosed and treated in the 21st century. 1. D’Alonzo GE, Barst RJ, Ayres SM, Bergofsky EH, Brundage BH, Detre KM, Fishman AP, Goldring RM, Groves BM, Kernis JT, et al. Survival in patients with primary pulmonary hypertension. Results from a national prospective registry. Ann Intern Med 1991;115:343–349. 2. Rubin LJ, Badesch DB, Barst RJ, Galie N, Black CM, Keogh A, Pulido T, Frost A, Roux S, Leconte I, Landzberg M, Simonneau G. Bosentan therapy for pulmonary arterial hypertension. N Engl J Med 2002;346:896 –903. 3. Simonneau G, Barst RJ, Galie N, Naeije R, Rich S, Bourge RC, Keogh A, Oudiz R, Frost A, Blackburn SD, Crow JW, Rubin LJ. Continuous subcutaneous infusion of treprostinil, a prostacyclin analogue, in patients with pulmonary arterial hypertension: a double-blind, randomized, placebo-controlled trial. Am J Respir Crit Care Med 2002;165:800 – 804. 4. Barst RJ, Rubin LJ, Long WA, McGoon MD, Rich S, Badesch DB, Groves BM, Tapson VF, Bourge RC, et al. A comparison of continuous intravenous epoprostenol with conventional therapy for primary pulmonary hypertension. N Engl J Med 1996;334:296 –302. 5. Simonneau G, Galie N, Rubin LJ, Langleben D, Seeger W, Domenighetti G, Gibbs S, Lebrec D, Speich R, Beghetti M, Rich S, Fishman A. Clinical classification of pulmonary hypertension. J Am Coll Cardiol 2004;43:5S–12S. 6. Nichols K, Saouaf R, Ababneh AA, Barst RJ, Rosenbaum MS, Groch MW, Shoyeb AH, Bergmann SR. Validation of SPECT equilibrium radionuclide angiographic right ventricular parameters by cardiac magnetic resonance imaging. J Nucl Cardiol 2002;9:153–160. 7. Galie N, Seeger W, Naeije R, Simonneau G, Rubin LJ. Comparative analysis of clinical trials and evidence-based treatment algorithm in pulmonary arterial hypertension. J Am Coll Cardiol 2004;(suppl)43:81S– 88S. 8. Cox DR. Regression models and life tables (with discussion). J Royal Stat Soc 1972;34:187–220. 9. Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 1994;81:515–526. 10. Justice AC, Covinsky KE, Berlin JA. Assessing the generalizability of prognostic information. Ann Intern Med 1999;130:515–524. 11. Berger HJ, Matthay RA, Loke J, Marshall RC, Gottschalk A, Zaret BL. Assessment of cardiac performance with quantitative radionuclide angiocardiography: right ventricular ejection fraction with reference to findings in chronic obstructive pulmonary disease. Am J Cardiol 1978;41:897–905. 12. McLaughlin VV, Shillington A, Rich S. Survival in primary pulmonary hypertension: the impact of epoprostenol therapy. Circulation 2002;106:1477–1482. 13. Sitbon O, Humbert M, Nunes H, Parent F, Garcia G, Herve P, Rainisio M, Simonneau G. Long-term intravenous epoprostenol infusion in primary pulmonary hypertension: prognostic factors and survival. J Am Coll Cardiol 2002;40: 780 –788. 14. Kuhn KP, Byrne DW, Arbogast PG, Doyle TP, Loyd JE, Robbins IM. Outcome in 91 consecutive patients with pulmonary arterial hypertension receiving epoprostenol. Am J Respir Crit Care Med 2003;167:580 –586. 15. Wensel R, Opitz CF, Anker SD, Winkler J, Hoffken G, Kleber FX, Sharma R, Hummel M, Hetzer R, Ewert R. Assessment of survival in patients with primary pulmonary hypertension: importance of cardiopulmonary exercise testing. Circulation 2002;106:319 –324. 16. Kawut SM, Palevsky HI. Surrogate endpoints in pulmonary arterial hypertension. Am Heart J 2004;148:559 –565. 17. Kawut SM, Palevsky HI. New answers raise new questions in pulmonary arterial hypertension. Eur Respir J 2004;23:799 – 801. 18. Peacock A, Naeije R, Galie N, Reeves JT. End points in pulmonary arterial hypertension: the way forward. Eur Respir J 2004;23:947–953. 19. Hoeper MM, Oudiz RJ, Peacock A, Tapson VF, Haworth SG, Frost AE, Torbicki A. End points and clinical trial designs in pulmonary arterial hypertension: clinical and regulatory perspectives. J Am Coll Cardiol 2004;43(suppl): 48S–55S. 20. Rich S, Kaufmann E, Levy PS. The effect of high doses of calcium-channel blockers on survival in primary pulmonary hypertension. N Engl J Med 1992; 327:76 – 81. 21. Weir EK, Rubin LJ, Ayres SM, Bergofsky EH, Brundage BH, Detre KM, Elliott CG, Fishman AP, Goldring RM, Groves BM, et al. The acute administration of vasodilators in primary pulmonary hypertension. Experience from the National Institutes of Health Registry on Primary Pulmonary Hypertension. Am Rev Respir Dis 1989;140:1623–1630. 22. Sitbon O, Humbert M, Ioos V, Jais X, Parent F, Garcia G, Herve P, Simonneau G. Who benefits from long-term calcium-channel blocker therapy in primary pulmonary hypertension? (abstr). Am J Respir Crit Care Med 2004;167: A440. 23. Fuster V, Steele PM, Edwards WD, Gersh BJ, McGoon MD, Frye RL. Primary pulmonary hypertension: natural history and the importance of thrombosis. Circulation 1984;70:580 –587. 24. Cohen M, Fuster V, Edwards WD. Anticoagulation in the treatment of pulmonary hypertension. In: Fishman AP, ed. The Pulmonary Circulation: Normal and Abnormal. Philadelphia, PA: University of Pennsylvania Press, 1990: 501–510. 25. Frank H, Mlczoch J, Huber K, Schuster E, Gurtner HP, Kneussl M. The effect of anticoagulant therapy in primary and anorectic drug-induced pulmonary hypertension. Chest 1997;112:714 –721. 26. Kawut SM, Taichman DB, Archer-Chicko CL, Palevsky HI, Kimmel SE. Hemodynamics and survival in patients with pulmonary arterial hypertension related to systemic sclerosis. Chest 2003;123:344 –350. 27. Bossone E, Paciocco G, Iarussi D, Agretto A, Iacono A, Gillespie BW, Rubenfire M. The prognostic role of the ECG in primary pulmonary hypertension. Chest 2002;121:513–518. 28. Lilienfeld DE, Rubin LJ. Mortality from primary pulmonary hypertension in the United States, 1979 –1996. Chest 2000;117:796 – 800. 29. Cooper R, Cutler J, Desvigne-Nickens P, Fortmann SP, Friedman L, Havlik R, Hogelin G, Marler J, McGovern P, Morosco G, et al. Trends and disparities in coronary heart disease, stroke, and other cardiovascular diseases in the United States: findings of the national conference on cardiovascular disease prevention. Circulation 2000;102:3137–3147. MISCELLANEOUS/NEW PREDICTORS OF OUTCOME IN PAH 203 ANTHONY NICHOLAS DeMARIA, MD: A Conversation With the Editor* ony DeMaria (Figure 1) was born in Elizabeth, New Jersey on January 12, 1943, and grew up in T Bayonne, New Jersey. After public and parochial schools, he went to the College of the Holy Cross, graduating in 1964, and then to the New Jersey College of Medicine, graduating in 1968. His internship in internal medicine was at the St. Vincent Hospital in Worcester, Massachusetts, and his junior and senior medical residency was at the US Public Health Service Hospital in Staten Island, New York. His fellowship in cardiovascular medicine was at the University of California at Davis. Following completion of his fellowship in 1973, he remained on the faculty until June 1981, when he moved to Lexington, Kentucky, to be Professor of Medicine and Chief of the cardiovascular division of the University of Kentucky College of Medicine. In May 1992 he returned to California, this time to San Diego, as Professor of Medicine and Chief of the cardiovascular division of the University of California at San Diego. In October 2001, he was made the Judith and Jack White Chair in Cardiology. Dr. DeMaria has been, for nearly 3 decades, an international authority in cardiology, particularly in the area of echocardiography. His investigations have led to the publication of Пѕ450 articles, almost all in peer-reviewed medical journals. He is also the author or editor of 3 books. For his investigative work and teaching abilities, he has received many honors, including several Internal Medicine Distinguished Faculty Teaching Awards, Teaching Scholar of the American Heart Association, Distinguished Alumnus Award of the University of Medicine & Dentistry of New Jersey, and honorary degrees from the Kagawa Medical University and the University of Bordeaux, among others. He was president of the American College of Cardiology in 1988 and president of the American Society of Echocardiography from 1985 to 1987. He became editor in chief of the Journal of the American College of Cardiology in January 2002. He has lectured widely, both in the USA and abroad. He and his lovely wife Lori are the proud parents of 3 offspring, Christine and Anthony who are now married, and Jonathan. Both Tony and Lori are wonderful people and fun to be around. William C. Roberts, MD†(hereafter WCR): Dr. DeMaria, I appreciate your willingness to talk to me and therefore to the readers of The American Journal of Cardiology. We’re onboard the “Silver Whisper” ship traveling around the Italian peninsula. Could we start by my asking you to discuss your parents and siblings, early upbringing, and some early memories? *This series of interviews is underwritten by an unrestricted grant from Bristol-Myers Squibb. †Executive Director, Baylor Cardiovascular Institute, Baylor University Medical Center, Dallas, Texas, 75246. 204 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 FIGURE 1. Anthony Nicholas DeMaria, MD, during the interveiw (photo by WCR). Anthony Nicholas DeMaria, MD‡ (hereafter AND): I was born in Elizabeth, New Jersey, but I actually grew up in a town not far away, called Bayonne, New Jersey, which like Jersey City and Hoboken, is in the New York metropolitan area. My dad owned a fruit and vegetable market. He went to the farmer’s market very early 5 days a week for produce. He worked extraordinarily long hours, opening the store at 6 A.M. after coming back from the farmer’s market, and he closed it at 10 P.M. 7 days a week. He made his own sausage. During Christmas time, we sold Christmas trees. It was in the selling of Christmas trees that I learned that the price for anything was variable and subject to negotiation. We sold flowers at Easter and for Mother’s Day. I worked in the store with my dad. My mom was a homemaker (Figure 2). She took care of my 3 younger sisters and me. Bayonne was, and is, a largely an ethnic community. If I knew somebody’s address, I could pretty much tell what their nationality was and what church they went to. I went through the local schools and then went off to college in Worcester, Massachusetts. ‡ Director, Cardiovascular Center, Judith and Jack White Chair in Cardiology Professor of Medicine, University of California, San Diego, California, 92103-8411. 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.10.004 FIGURE 3. AND’s father’s grocery store. FIGURE 2. AND and his mother and sister. WCR: How big was Bayonne? AND: Bayonne had a population of 60,000 packed into a peninsula that was 3 miles long and at its very widest, about 1 mile wide. On one end it was connected to Staten Island, New York, via the Bayonne Bridge, and on the other end, it continued to Jersey City. We could almost see the Statue of Liberty from my house. WCR: Did you get to Manhattan or Staten Island very much? AND: Not very much. The Verrazano Narrows Bridge, which connects Staten Island to Brooklyn, hadn’t been built yet. It was alleged (and I’m sure it’s true) that there were people who had been born on Staten Island that never in their lives had left the island, even for 1 day. Going to New York City was an adventure. The traffic was worse, of course, than in Bayonne. The people were a little different. We mostly stayed on the Jersey side. WCR: You are the oldest of 4 siblings. You were born in 1943. What’s the difference in your siblings’ ages? AND: Each of us was born roughly 2 years apart. We spanned 8 to 10 years between the 4 of us. WCR: What was your daddy’s store like? AND: It was very small and fascinating. I have a picture of it in my office (Figure 3). When my father died and we went through his things, we found a crumpled up photograph of the store. I had to smooth it out to have it made into a nice big print. The whole store couldn’t have been 1,000 square feet. It was mostly fruits and vegetables with canned goods on 1 side and a big freezer in the corner at the back. It contained a little delicatessen, where we sliced lunch meats. His major business was selling to people who wanted very fresh fruits and vegetables or alternatively, to people who wanted delivery. The latter is what I did most of the time. Customers would call in orders (a pound of this and a sack of sugar) and we’d put together the order and then I’d deliver it to their houses. It was pretty convenient. WCR: On a bicycle? AND: Remember those bicycles with the small wheels in front? The rear wheel was the usual size and the front wheel was very small with a huge basket above it. I could usually fit 3 or 4 orders in the basket and deliver them. It was a marvelous education in life. My dad died prematurely at 59 years of age from an aortic dissection. My mom is still alive and my sisters Barbara, Carol, and Adele are doing well. WCR: When was your father born? AND: He was born in 1916 and died in 1976. WCR: He worked up to the day he died? AND: My father had 2 jobs at the time he died. The store had burned down. The building that the store was in was very old. We were home on a Sunday and somebody came by and told us the store was on fire. The building had kerosene heaters. This was firstgeneration America. It was not the suburbs. Because my father had records in the store of who owed him money, he ran into the building. I ran to the building, a judgment that compounded the stupidity of the moment, but I convinced him to get out; shortly thereafter the building went up in a blaze. WCR: Neither of you got burned? AND: Neither one of us got hurt. We got away in time. WCR: You were how old then? AND: I was 23. My dad then went to work for the city of Bayonne. He was in charge of all their parking meters and helped with zoning issues, etc. He’d do that all day, and then in the evening he went to some truck terminals in Jersey City and loaded trucks until about 2, and then come home and go to bed.(Jersey City was a port.) He died of aortic dissection while loading a truck. WCR: Did your father smoke? INTERVIEW/ANTHONY NICHOLAS DEMARIA 205 AND: Yes, cigars, continuously. I think he was born with a cigar in his mouth. He’d send me down to the store to get his cigars. I’d get Dutch Master’s panatelas. That’s all he smoked. WCR: How far was the store from your home? AND: A block or maybe a block and a half. WCR: What was your home like? AND: My home was pretty good. We were, by the standards of our neighborhood, well off. We had our own house. It was a small split-level house. My father had it built. There was a lot in the area that was empty. He and my uncle purchased the lot and then contracted out the building of 2 houses next to one another. WCR: Did you have a room of your own? AND: I did. It was in the attic. The house had 2 bedrooms and an attic. My parents slept in one and my 3 sisters had the master bedroom. That was the only one all 3 could fit in. My parents slept in a smaller bedroom. My dad, who was very handy, put a floor in the attic and plastered the ceiling. You could only stand up in the middle of the room because the roof slanted. But it was mine! I was the only one living in it! It was a different time, Bill. My folks thought it was important for a son to go to college. That was preached to me from day 1. They repeated over and over again that I needed to go to college so that I didn’t wind up running a fruit and vegetable market. But my sisters didn’t seem to matter so much. If they went to college that was nice; it was icing on the cake. For the first born son, it was absolutely necessary. WCR: Were your parents first-generation Americans? AND: Yes. Both my parents were born in the USA. Their parents had just come over. I never got to know my grandparents because I couldn’t speak their language. WCR: Where were they from? AND: On my father’s side, from Italy (the Naples area), and on my mother’s side, Poland. They couldn’t speak to one another. WCR: How did they meet? AND: They met at one of the beaches in Jersey. My father was black-haired with a darker complexion. My mother was blonde with very fair skin. I guess opposites attract. My mother was from Elizabeth, New Jersey. That is why I was born there. My dad was from Bayonne. I was only born in Elizabeth. We always lived in Bayonne. WCR: How far is Bayonne from Elizabeth? AND: Two stops on the Jersey Turnpike, maybe 10 miles. WCR: Where does your mother live now? AND: She lives in the same house, in Bayonne, New Jersey. My uncle still lives next door. Two other uncles, my father’s brothers, lived across the street until they died. The whole family on my dad’s side lived in that area. WCR: You had a lot of cousins and family? AND: Tons of cousins. There were lots and lots of DeMaria kids in that 3- to 4-block area. WCR: It sounds like your father wasn’t home very much because he was always at the store or working 206 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 a second job. But you could always go down to the store and see him. Did you work every day after school after you reached a certain age? How did that work out? AND: No. When I got into high school my parents made it clear to me that my job in life was to get to college. They didn’t want me to do anything that would interfere with that. I only worked in the store summers and on weekends. During the week, my job was to study and get good grades. My mother did not graduate from high school. My dad did graduate from high school and actually could have gone to college. He had that opportunity, but he had to go into the family business. At that time my paternal grandfather was a peddler. He had 4 sons. He had horses and wagons and as a kid, I often rode the horse and wagon. They’d go up and down the streets of Bayonne with fruits and vegetables. By those standards, my father was a big success. He got his own store! He didn’t have to peddle anymore. Frankly, if I didn’t want to work weekends, I didn’t have to. If I ever went to him and told him that I had to study, he’d say, “That’s all right. You do whatever you want. Go study. You don’t have to do work here.” In contrast, my kids grew up competitive swimmers and I was the starter of the swim team. Every weekend Lori and I packed up the car and went to a swim meet. I tried to see every little league and soccer game my kids were in. I would make the time. My dad was just too busy. He had to be in the store. WCR: What was your father like? Were you and your father close? AND: My dad was very smart. He was kind of an overwhelming presence. In our neighborhood, he managed to get his own house. Very few had their own house. Families lived in rooms. He was very facile politically. A lot of Italians in our neighborhood came to the store to seek his advice on various matters. If they had a problem with the city, he knew who to call to get a pothole or whatever fixed. He was active politically. If neighbors needed things done, he knew how to get them done. He spoke Italian fluently, as well as English. He was sought after as a political ally for the elections because a lot of people voted the way he recommended that they should. In our neighborhood, he was a big “chimichanga.” I was always quite awed by the respect he was afforded, and by the fact that we lived in our own house and I could go to a private Catholic school. WCR: You had to pay tuition? AND: Yes. As a kid I thought my dad was about the biggest success in the world. I wasn’t sure that I could ever quite achieve what he had achieved. If he had had the opportunities that I had, only God knows what he would have accomplished. He was the real McCoy! WCR: It sounds like he was very resourceful and that he was a magnet that attracted people to him. AND: Yes. He was the eldest brother and that’s why he couldn’t go to college. He had to go into the business because the other brothers were a bit younger. He was the glue that held that family together. He was very resourceful and was very handy. JANUARY 15, 2005 He managed by hook or by crook. I’ll never know in a million years where he learned it. I couldn’t contract out a house, but my father knew how to hire plumbers and stuff like that. What he didn’t know, he knew friends who would provide him with the information he needed. He knew everybody. He was very active in the community. WCR: Did you take family vacations? AND: Not really. When I was nearly out of high school, my parents rented a bungalow in Seaside for a couple of weeks during the summer, but I worked in the store. WCR: What did you do in the summertime other than work at your father’s store? AND: Nothing. I played the summer sports—Little League, Babe Ruth league. WCR: What was dinner like at home? It sounds like your father rarely made it. AND: He didn’t make it very often. During the summers, he would close the store at noon on Sundays and we’d pack up the car with a picnic lunch and go out to one of the lakes. In those days people swam off Staten Island at a place called South Beach where it was still swimmable. I remember well those excursions to South Beach or to Lake Hopatcong. WCR: It sounds like your father was the dominant figure around the house as well as at work with his male colleagues. AND: Yes. My father ran the house and everything and my mother ran my father! WCR: What was your mother like? AND: She was and is a very supportive person. She was a homemaker. She had great aspirations for all her kids. She demanded that we achieve at whatever we did. My sisters took ballet lessons and she’d be with them, making sure that they could dance on their toes. She was a meticulous housekeeper. She had her own circle of friends. Behind the scenes she was the person who took care of the family because my father was always at the store or somewhere. My mother got us ready for school and carted us off to wherever we had to get carted off to. WCR: You had a car? AND: Yes, our family had a 1948 Mercury, James Dean’s car. WCR: What about the safety in the neighborhood? Were you safe in the neighborhood where you lived? AND: Yes. All of Bayonne was ethnic. It was all shuffled together. There was no real slum. The rich area, so to speak, wasn’t all that rich and it wasn’t all that big. There wasn’t that much that separated the “haves” from the “have-nots.” It was all mixed up. That’s true of where I lived. Our house was on a nice street. I could go 2 blocks away where my father’s store was and there were the low-income “housing projects,” consisting of very big box apartment buildings. Bayonne had 2 sets of projects that went on for several city blocks. One of them started on 49th and Broadway and went up to 50th. My dad’s store was between 48th and 49th on Broadway. We lived on 48th off Avenue C, which was 1 block up. The projects were the toughest part of town. I can’t say that I ever worried very much about safety. Our house was never burglarized. WCR: Your father’s store was never robbed? AND: Yes, it was a couple times. There wasn’t much to rob, however. The robbers could steal salami or something. He didn’t have a large amount of money there. Everybody knew everybody. There was no way to commit a crime and not have everybody in the whole damn town knowing it. WCR: You spoke English in your home? AND: My parents were very adamant about that. My name is mispronounced. The correct pronunciation is not “de Marry A;” it’s “day Maria”. When my grandfather came over, the Americans called him “DeMaria” and the concept was that you had to assimilate into English as quickly as possible. My mother and father had to speak English to each other because, although my mother was fluent in Polish and my father was fluent in Italian, neither knew the other’s language, so they had to communicate in English. That’s what we did. WCR: What was your mother’s name before marriage? AND: It was Tess Kupis. That’s not what the actual name was. It was shortened. It was a long Polish name. When her parents came through Ellis Island, the name had 17 or so vowels and it was shortened. Everyone called her Tess. WCR: Do you remember dinner at night in your home when you were growing up? Your daddy was at the store. What was it like? Did you, your mother, and 3 sisters talk about what happened in school or what? AND: It wasn’t like my father was never there. He occasionally came home for supper. We didn’t have a family dinner exactly. It just depended on who was doing what. If one of the kids had a recital or had to go off to Cub Scouts or Girl Scouts or got home late from track practice, then we ate when we got there. My mother, who I hope will never read this, was not a great cook. She has many virtues in this world, but being a great cook was not one of them. WCR: But you could always go to the store and get something to eat if you wanted to. AND: Yes. We were all skinny as can be. When I was in high school, I became aware of women and things like that. I realized then that I was the “90pound weakling.” I went off to college at 6-feet tall and 122 pounds. I was just skin and bones. WCR: What do you weigh now? AND: More than that. WCR: Are you still 6-feet tall? AND: No, I shrunk a little bit. I’m 5 feet 11 inches. My intervertebral disks are desiccating. So are yours. WCR: I’m sure they are. Was your family strongly religious? Did you go to mass every Sunday? AND: My mother always did. There was a children’s mass and we had to go to that one. Those who went to Catholic school sat in one section at mass and those who went to public school in another section. It was fairly regimented. We always went. My dad always ascribed to religion, but he often missed mass until after the store burned down and then he attended INTERVIEW/ANTHONY NICHOLAS DEMARIA 207 mass regularly and indeed became an usher, helping attendees find seats and later passing the collection baskets. He was very active. They were both pretty religious by today’s standard. WCR: Did you have prayers before you ate meals at home? AND: No. WCR: Were there books around the house? Did your mother read much? AND: She read some but not a lot. It was not a very literary environment. WCR: Were there teachers in grammar school, junior high, or high school who had an impact on you? AND: Yes. I attended public grammar school. (My father felt very strongly about that.) Bayonne, however, was about 75% Catholic. Every parish had its own school. The concept was that your best chance of getting to college was to go to one of the Catholic schools. My dad, however, insisted that I go to public grammar school because he thought it was very important to learn how to live with people with different cultural backgrounds than our own. After grammar school, my parents switched the children to Catholic private schools, believing they provided a better chance of getting into college than the public schools. It wasn’t that kids who went to Bayonne High School didn’t get into college, but fewer of them went on to college, whereas most kids who went to the Catholic schools were targeting college. Then the question was, “How do you get into a Catholic high school?” We lived in a town of 75% Catholics and there was only 1 men’s and 1 women’s Catholic high school, and they couldn’t take all the Catholic high schoolers. There were other Catholic high schools in Jersey City, but most were even more competitive. You had to take an entrance exam. I don’t know what the acceptance rates of those Catholic high schools were in those days, but it was very competitive to get in. In the eighth grade I was assigned an elderly teacher named Mrs. Murphy. Early on she asked, “How many of you children in this class are going to try to go to Catholic high school and take the entrance exams?” There were about 8 of us in my class of 25 students. Our parents had decided that we were going to try to go to a Catholic high school. She said, “I’m going to come in 1 hour early 3 days a week and we are going to start preparing for the exams.” She did. The tests to see if you could qualify were given in the spring. Until April or May she came early 3 days each week to go through vocabulary with us. I don’t want to sound like I was in any way disadvantaged, because I wasn’t. My parents had a standard vocabulary. They didn’t have advanced language skills. Mrs. Murphy went over these things. We practiced taking the test. It was multiple choice, and we hadn’t taken one of those to that point. I took the exam. I was not the smartest of those 8, but I got into the Catholic High School in Bayonne and none of the others did. I was very fortunate. I’m convinced that I never would have gotten in if Mrs. Murphy hadn’t worked with us. She was at the stage of her life where she had a mission. Her own children were grown. Her life was teaching 208 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 us kids. She was a spectacular schoolteacher. In seventh grade, sometimes our teacher would come in and say, “Open your book and read this chapter.” There was a lot of free study time. Mrs. Murphy was very structured. I never did get a chance to thank her. She was delighted that somebody made it. Only 1 person also made it the year before. WCR: You started Catholic high school in the ninth grade. Was there a big change? Was it much harder than the public grammar school and Catholic junior high that you’d gone to? AND: Yes. Where I lived it was heavily Catholic. In Bayonne alone there were 7 parish schools. In the Catholic high school, my classmates were on average smarter and often came from families where the parents were schoolteachers, doctors, lawyers, and businessmen. That made it tougher for me. They had a bit of an edge coming in. WCR: They also had all come from parish schools? AND: Pretty much. WCR: Were you active athletically? AND: I was a good long-distance runner. In my freshman year in high school I finished second in the conference. I won a lot of races. As I got older, I didn’t improve as much as the others. There were a lot I could beat as a freshman, but by the time we were seniors, they had just gotten better than me. I achieved some success at track and I played basketball for a couple of years. It was clear that I was not as good at basketball as I was at track. I never made the starting team. WCR: What races did you run? AND: Cross-country in the fall and I was a miler or half-miler in the spring. WCR: Cross-country was usually how many miles? AND: It was 2.5 miles. We’d run it in a park. They would set up a course so that you had to go up and down some hills and through some fields. It wasn’t run around a track. In the winter and spring, we ran on the track. WCR: You were second in the long-distance run as a freshman. AND: That was my pinnacle. That was in our region of the state. In Bayonne, I won the mile and half-mile a couple of times. Once a year there would be a track meet for Bayonne High School, Marist (the Catholic High School I went to), and for other people who went to other schools who lived in Bayonne. Being the champion of Bayonne wasn’t that great. WCR: How many were in your high school per class? AND: About 90. WCR: How did you finish in the standings at the end of your senior year in high school? AND: I did okay. WCR: You were first in your class? AND: No. They never gave out a ranking exactly. At my graduation they gave awards to 7 people. To get the top award you had to have gotten an A in every single course for 4 years. To get the second award, you had to get at least a BП© in every course for 4 JANUARY 15, 2005 years. I got the second award. I missed getting the top award by 2. WCR: Did you have to study hard for those grades or did they come easy for you? AND: I had to study hard. WCR: Were there teachers in high school that had a major impact on you? AND: Yes, there were a number of them. I had religious brothers. They are like priests, but they don’t say mass. One of the brothers was my track coach and he helped me a lot. He was a good friend, but got transferred after a year. In my senior year, I had a brother, Hugh Arthur, a French/Canadian guy who scared everybody to death. He ran a tight ship. His field was mathematics. He also taught religion and something else. I was assigned to his homeroom class in my senior year. He was hard as nails, but he was actually a very warm person underneath it all. He gave me a lot of good advice. As seniors in high school, we would get distracted and get into a bit of mischief. I would never claim that I was the best student in the world. WCR: Was it a co-ed high school? AND: No. WCR: All boys. Where were the girls? AND: They were down the street at Holy Family. WCR: You dated them in high school? AND: Yes. I went to mixers and dances. WCR: It sounds like high school was a good experience. AND: Yes. I enjoyed it. WCR: When it came time to go to college, you really couldn’t get much advice from your parents or from your father’s brothers because they’d never been themselves. Were there many members of your mother’s family in the immediate neighborhood? Did you know your mother’s family? AND: They lived in Elizabeth, New Jersey. We saw them from time to time. She had a brother and a sister. Each of them had 1 child, a boy. We didn’t see much of them. We weren’t close. My grandmother on my mother’s side died relatively young from a subarachnoid hemorrhage. My maternal grandfather lived about 90 years. He couldn’t speak English very well, although he worked for decades for Esso (Standard Oil) in Elizabeth. I couldn’t communicate with him much, because I didn’t speak Polish. WCR: You didn’t know either of your grandparents on your father’s side? AND: I did. They lived in Bayonne. WCR: You had 4 grandparents. AND: Yes. But it was hard to communicate. WCR: On your mother’s side? AND: On my mother’s side and my father’s side. My paternal grandmother spoke halting English. My paternal grandfather could speak a fair bit, and I could communicate with him. WCR: When it came time to pick a college, how did that work out? AND: The brothers pushed us toward Catholic colleges. Many of my classmates went to St. Peter’s College in Jersey City. A batch went to Seton Hall. Some went to Rutgers. I can’t remember anybody going to an Ivy League school. I wanted to get away from home, but not too far. I told the brothers that maybe I’d go to New England or down south. They came back with Holy Cross or Georgetown. I applied to St. Peters and Seton Hall; everybody did. While I was sifting through that, I learned about Dartmouth. That seemed to me like it might fit the bill. I wrote off for their catalog, but I was roundly discouraged from pursuing that by everybody. For whatever reason, they said it wouldn’t be a good alternative. I wound up going to Holy Cross. WCR: You went to Holy Cross in Worcester, Massachusetts. Your parents must have been extremely proud of you for having gotten into such a good college? AND: Yes. It was a good college to get into. Most of my classmates stayed locally. I had a lot of really smart classmates. There was a school called the Stevens College of Engineering in Hoboken. It’s still there, but now it’s called something else. A number of my classmates went there. It too was considered quite a good school. WCR: Hoboken? That’s where Frank Sinatra was from. AND: That’s right. My dad knew Frank Sinatra. They dated cousins before my father married. WCR: What did your father say about Frankie? AND: He said he was a nice guy. My father was dating a woman and she had a cousin. Her cousin was dating Frank Sinatra. They double-dated a number of times. They got to be friends. He wasn’t anything then, just a skinny Italian kid from Hoboken. My dad lived in Hoboken before his family moved to Bayonne. Frank Sinatra married Nancy, but my father didn’t go with the other cousin anymore. They drifted apart. By the time my father was dating someone else, Frank Sinatra had started to make it big. He was playing a nightclub in New York City. My father wanted to do something special for New Year’s Eve. On a whim he called the nightclub and asked to speak to Frank Sinatra and told them it was Tony DeMaria calling. According to my father, Frank got on the telephone and asked how he was doing? My father asked Frank, ”Do you think you can get me a ticket to your show?“ And Frank did it; he got him a table for his show on New Year’s Eve in a nightclub in New York City. In Bayonne, it just doesn’t get any better than that. WCR: You are a junior. AND: The tradition was that you name the first son after the grandfather. My grandfather’s name was Nicolas. My mother just wouldn’t abide by naming me Nicholas. They compromised on Anthony, which was my father’s name, but Nicolas is my second name, after my grandfather. Of course, I went back to the tradition so my firstborn was named Anthony after my father. The fact that it happened to be my name was a side bonus. WCR: Did you go to Holy Cross to interview before you enrolled? AND: They didn’t have interviews. Maybe they INTERVIEW/ANTHONY NICHOLAS DEMARIA 209 did, but they didn’t call me for an interview. I just applied. I remember that the brothers, when I told them which college I was interested in, advised me to apply to Fairfield in Fairfield, Connecticut, Providence College in Providence, Rhode Island, and at Holy Cross. One weekend my father took off from work. (I’ll carry this to my grave because my father never took off from work.) He got one of his brothers to mind the store. We left on a Friday evening and we drove to visit Fairfield, Providence, and Holy Cross in Massachusetts. That drive was the most time I’d ever spent with my father alone. Shortly thereafter, I got a letter of acceptance from all 3 of them. We decided I’d go to Holy Cross, because it had the best reputation at the time. WCR: Holy Cross is how far from Bayonne? AND: Two hundred miles. It was a 5-hour drive. WCR: How many students were at Holy Cross when you were there? AND: Sixteen hundred. WCR: About 400 in each class. It’s a beautiful college. AND: It was a nice school. It was on a hill overlooking the city. Surprisingly, for a small steel belt town in the middle of Massachusetts, Worcester had quite a number of colleges: Worcester Poly Tech, Clark University, Anna Maria (the Catholic women’s college), Holy Cross, and Assumption College. It was a nice setting, but it was very strict. I had to go to mass every morning for the first 2 years I was there. They took attendance. WCR: You started college what year? AND: In 1960. WCR: You were 17. How did college strike you? That was the first time you’d ever been away from home, right? AND: Yes. WCR: How did it work out? AND: I was nervous. It was like a big filtering process. Going from Washington Public School to Marrist High School, they only took so many into my high school. Going from my high school to college, there were guys from all over the country. The major Catholic colleges in the USA then were Notre Dame, Georgetown, and Holy Cross. There were a lot of valedictorians there. Holy Cross was known for premed education. Even in high school I understood that getting into medical school was tough. One reason I went to Holy Cross was because they had a reputation for getting people into medical schools. When I got there, there were 150 of us who entered as premed (150 of 400 were designated as premed). They called us together at the beginning of school and gave premed orientation in a large auditorium. They told us that of the 150 students who were there, there were lots of valedictorians of their respective high schools. As I got further along, there were fewer guys like me from Bayonne and more who had come from professional families. It was very strict. They told us that, they “never had a person that they gave an unqualified recommendation for medical school that wasn’t accepted to an American medical school.” They fol210 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 lowed that up by saying, they “had never recommended more than 55 or 60.” You didn’t need to be a rocket scientist to realize that there were 150 of us, and the most they ever recommended was 60, so there were a lot of us who would not be recommended. Once that was clear, everybody looked at each other and said, “Who is going to be left standing at the end of 4 years?” That was the atmosphere that permeated the place. Folklore had it that at one time, they had put all the premeds in 1 dormitory on campus that the rest of the student body named “The House of Death.” They didn’t do that when I was there. It was very competitive as a premed. But it was a good education. It prepared me for medical school. Medical school was quite a bit easier than Holy Cross. WCR: It sounds like you had already decided that you wanted to be a physician before you entered college. How did that come about? You had no physicians in your family. AND: I’ll tell you a true story. My mother and father had been targeting me to go to college from the time I could remember and telling everybody I was going to go to college. They weren’t shy about it; they told everybody. It got to be senior year and the whole neighborhood (half of which were my family) was wondering what I would do. Everybody knew each other. Sure enough, I got accepted. My grandmother and my aunts and uncles were asking, “What are you going to take up in college?” I didn’t know. How do you know? I liked history. It was easy and I did well in it in school. The first thing I said was, “Maybe I’m going to do history.” That drew a blank expression. They didn’t know any historians in Bayonne. They were saying, “History. History. What is that?” That bombed. I rummaged around a bit and decided biology was pretty interesting and I had always enjoyed that. So I told them “Biology.” That didn’t ring any bells either. Honest Bill. This is the truth. I felt the pressure here. One day my grandmother came to me. “Anthony, Anthony. What are you going to take up in college?” And I said, “Maybe I’ll be a doctor.” “Aahhhh,” was her response. All my worries were gone. I just said “doctor” and they were happy. I didn’t know if I wanted to be a doctor, but I figured I could change along the way. It got everybody off my back telling them what I was going to do. Nobody would believe for a minute that you would go to college to get an education. What good was that? My dad said to me one time when I was flailing around trying to think what I wanted to do, “Think of college like a train. You are going to get on the train. When you get off you want to be somewhere. That train is going to take you. You have to figure out where it is you want to be when the train stops.” WCR: That’s beautiful. AND: Once I got to Holy Cross, the question was no longer what I wanted to do. The question was could I do it? It became a challenge. Was I going to be standing at the end of 4 years to get their recommendation? During orientation week they said, “Everybody needs to go to confession to get in the proper spirit, to take communion, etc.” You went to mass JANUARY 15, 2005 every morning so you might as well go to communion, too. I went to confession with one of the Jesuits from New England. I say my thing, “Bless me Father, etc.” He said, “What did you say? What did you say?” I said it again. This went on for 3 or 4 minutes. He told me, “Son, I want to tell you something. I can’t understand a word you’re saying. You’re going to waste your money on an education. Nobody’s going to understand you. You’ve got to get rid of this accent.” I had a heavy New Jersey accent. There weren’t a lot of us there at the time. WCR: Did you go on scholarship? AND: Partial. WCR: Partial scholarship from the very beginning? AND: Yes. They didn’t offer me a scholarship right off when I was accepted, but my father wrote them a letter. He asked them if there was any chance that I could get some financial help. They sent back an application form where he had to put down his income and expenses. Sure enough, they came back and they gave me a 50% scholarship. WCR: That was tuition? AND: Maybe room and board and tuition. They also got me a government loan. WCR: How did Holy Cross work out? You said it was very rigid. You were scared because the competition was vigorous. You hadn’t met all these students who had come from professional households, many of whom, I presume, were fairly wealthy also. AND: I can’t say it was a wealthy place, but many of the others had more varied experiences than I had. They had often been exposed to more things. However, there were a number of guys like me. Holy Cross wasn’t a ritzy school compared to Georgetown. College is a filtering process. To get into the place, nearly everybody was in the top 20% of their high-school class. If you’re in the top 20% of your class, you are better than 80% of the other guys. When you get into college, everybody is in the top 20% and things are tougher. WCR: You enjoyed the challenge though? You really poured it on from a studying standpoint. AND: It was my first time away from home. For the first time I was footloose and fancy free. I goofed off a bit and was slow academically. I caught up. It was a good college life and I’ve no complaints. I had to work very hard. When I got into medical school, compared with Holy Cross, it was easier. WCR: What was your major in college? AND: Liberal arts. If you were premed, you either took a bachelor of science in biology or a bachelor of arts. I was a bachelor of arts. I didn’t have a major. I took the science courses that I needed for medical school. It was pretty prescribed and I took all liberal arts courses—languages, philosophy, history, etc. WCR: Were there any teachers in college that had a major impact on you? AND: In my freshman year I had a faculty advisor, Father Leo McCarthy. I talked to him. He provided some encouragement. That was really important for me early on to get me through what was a fairly challenging transition. Many of the other students came from elite Jesuit high schools. They had come from an environment that was similar to Holy Cross. Maybe they had a bit of a running start. It took me a little longer to get going. Father McCarthy transitioned me through that pretty well. I did pretty well. I obviously was in the top 50 premeds because I was recommended and got accepted to medical school. WCR: Did you enjoy Worcester? Did you go into Boston much? AND: You could only go off campus until 11 P.M. If you were leaving the campus, you had to sign out, you had to sign back in, and they did bed check. Somebody came around to make sure you were in your room. (This was a long time ago.) During the week, the lights had to be out at a certain time. That was it. If you had to cram for a test or something, you had to go to the bathroom and get in one of the stalls and sit in there. On the weekends, you had to sign out on Friday and Saturday. I don’t think you could go off campus on Sunday. Most of the time students stayed on campus. It was an all-male school; there were no women. We were not allowed to have cars until we were seniors. The first year, all premeds just studied. Taking Saturday night off meant you went to the dining area and watched a movie on campus. If you wanted to go to Boston, you had to hitchhike. You could sign out for the weekend, but you had to give them an address. I didn’t have any addresses. It wasn’t a big social school. WCR: Who did you date up there? AND: There was a women’s college. There were a lot of students there in Worcester. When you got to Boston, there were even more students. A lot of us had girlfriends back home and we corresponded with our girlfriends. When you got a chance, you’d go back home. That was that. WCR: Did you go back home much during the school year? Christmas or Easter? AND: Always Christmas and Easter and usually Thanksgiving. That was a long weekend. There was a guy from my hometown of Bayonne who was also at Holy Cross. He was a senior in my first year. He was the son of one of the orthopedic surgeons in Bayonne. Whenever he was going to Bayonne, he’d call me and if I could do it, then I’d ride in his car to Bayonne and give him some money for the cost of the gas. WCR: Were you sick much as a child? Did you go to a doctor very often? Do you remember any doctor visits before you went away to school? AND: Yes. I used to have a lot of bronchitis. In high school, my mother took me to a chest doctor in Jersey City. He gave me what was the equivalent of puffers in those days and some nose drops. In retrospect, it may have been allergies. In college I had pneumothoraxes. That put me in the infirmary or the hospital. WCR: These were spontaneous? AND: Yes. I was a skinny kid. It scared the hell out of me the first time I had it. WCR: How did you end up in college? Do you have any idea? You mentioned that of the 150 premeds, you finished in the top 50. Did they give you any idea where you stood in your class at Holy Cross? INTERVIEW/ANTHONY NICHOLAS DEMARIA 211 AND: I made the Dean’s list a number of times. I don’t know where I finished, but I would say of the 50 guys who went to medical school, I was probably in the middle of that group. WCR: It’s also fair to say that those 150 premeds probably finished up higher than the 250 who started freshman year and weren’t premed. AND: Oh yes. Of course, there were examples of very bright guys in other areas, but, as a group, the premeds were generally at the top of the list. WCR: What did you do during the summers when you were in college? AND: I worked in my dad’s store for a year, I collected tolls on the New Jersey Turnpike for 2 years, and I drove a truck for a dairy distributor. WCR: That paid pretty well? AND: It paid more than I’d ever made working in the store. If you got a summer job collecting tolls on the New Jersey Turnpike, you got the worst shifts in the worst places. I’d wind up working the Lincoln Tunnel booth on Sunday when everybody was coming back from the Jersey shore. I worked a lot of nights. Those shifts screwed up my summers. My father called me and said that the guy who distributed dairy products to him was looking for a truck driver, somebody to distribute to different stores. I did that for a couple of summers. I worked Monday through Friday, 8 to 5 P.M. WCR: When it came time to apply to medical school, did you apply to several medical schools? What happened? AND: Everybody was really worried, even the top guys. It has always been competitive to get into medical school and we all had that anxiety. I applied to Seton Hall University, the local medical school in New Jersey. A bunch of my colleagues had applied as well. The Dean of Admissions at Seton Hall actually made a trip to Holy Cross. I’m sure he went to other schools in New England. He was there for 1 day and interviewed 6 of us. This happened in the first couple of weeks we were back in school as seniors. About a week after that, I got a letter in the mail saying that I’d been accepted to medical school. In the first month of my senior year I knew I had a place. My focus thereafter drifted a bit. You asked me about Boston before. I spent a whole lot more time in Boston. I applied to a couple of other places, too, but frankly I couldn’t have afforded to go elsewhere. Seton Hall was perfect for me because I lived with my folks again. It was the only sensible thing to do. They didn’t give me a scholarship, but they gave me a big loan for school. Between living at home and having a loan, it was financially feasible. WCR: How far from home was Seton Hall? AND: I didn’t have to stop at a red light from my house to there. Seton Hall was at the Jersey City Medical Center. I was in one of the first classes. The medical school didn’t last all that long. Ultimately, the state took it over and it became the New Jersey College of Medicine. WCR: It was private when you went there? AND: Yes. It was run by Seton Hall University, 212 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 which is a diocesan school. I went there for 2 years. It was too expensive for the diocese to run Seton Hall. The state of New Jersey took it over as a state medical school. I graduated from the New Jersey College of Medicine. WCR: You were back living at home after 4 years at very sophisticated (in a way) Holy Cross College. How did you like being back at home even though you were going to medical school at the time? AND: It worked out well for me because I was back in my old attic room. I was able to focus. When I was away at college, some of the distractions of college life got to me. The reason I bring up Ned Weyman is that we were very similar. He was 2 years ahead of me. I got to know Ned because his roommate and I were hospitalized together once. I had a pneumothorax and Ned’s roommate had a thoracic injury playing lacrosse. Here we were, these 2 college students in this hospital with a bunch of student nurses. That provided some distraction for a good part of the year after that. It seemed like in medical school I’d better not be distracted. Living at home wasn’t like being at college, where someone was playing cards all the time or somebody was going down to get a beer or any of those things. It worked pretty well. My mother made all the meals and did all the laundry for me. I must admit, in retrospect, that it was a good thing. I suspect that I had stayed at home for college, I might have wound up at the very top of my class. I didn’t mind living at home at all. The first couple of years of medical school were pretty rigorous. I had a girlfriend at the time. It was convenient. I’d call her and we’d go to movies. Living at home worked out well. WCR: How many students were in your medical school class? AND: One hundred. WCR: How many medical schools are in New Jersey? AND: There was only 1 then. I think there are 3 now. Rutgers has 1. They are all part of the same system. There’s 1 in New Brunswick called Rutgers. There’s 1 in Jersey City and there may be 1 in Camden. WCR: What is your medical school called now? AND: The College of Medicine and Dentistry of New Jersey. It was Seton Hall when I started and the New Jersey College of Medicine when I finished. WCR: It moved from Jersey City to Newark. Were there any surprises in medical school for you? When you first entered medical school did anything really strike you that struck you as very unusual? AND: Candidly, when I finished grammar school, I was one of the smartest kids. When I finished high school I was one of the smartest, right near the top. I went to college and I didn’t finish with the top grades. I dropped down a bit. That shook me up a little. For the first time coming out of college, I wasn’t coming out in the top 5% of the premed class. My batteries were a little charged for medical school. I was rip roaring ready to go. I spent the first 2 years of medical school making sure I was in the top 10 or so students in medical school. However, I did go through a period, which started at the end of my first year medical school and hit home in the second year of medical JANUARY 15, 2005 school, when I really wasn’t sure that I wanted to study medicine. There were not any counselors at that stage of the game, but I was sent to talk to the Dean by a professor. In medical school, I had been pulled out of the standard classes to be in a special group that was doing special studies on the side. I got to know Professor Bernard Briody, another hard-ass kind of guy. Everyone was a little timid around him. He had come up with this idea to pick some medical students and not bother them with the usual day-to-day drudgery of memorizing stuff and put them in some special classes. I was put in that class and got to know him. We had a special relation. I told him that I wasn’t sure I wanted to do medicine. He sent me to the Dean. The Dean was a little bit surprised because most of the people he was seeing that weren’t sure about going further didn’t have good grades. That was not a problem for me. To make a long story short, he said, “Don’t do anything dumb. The first 2 years of medical school aren’t exactly representative of what medicine is all about. You finish the second year and then take the summer off. Do something totally different and come back in the fall and start again in your clinical training and see what it’s like. Take another look at it.” That seemed like a reasonable idea. The question was what was I going to do in the summer. One classmate was a lifeguard at the Jersey shore at one of the state beaches. He said, “There’s a lifeguard who has helped with the first aid station who has had the job for a couple of years and he’s not coming back this year. He’s going to do something else. His name is Carl Pepine. Why don’t you take his job?” I did. I took Carl Pepine’s job on the Jersey Shore. I went down there to work to get away from medicine and to let things sift out. While working there, I met a woman who worked in one of the hamburger stands at the shore. WCR: Lori? AND: Yes, she’s now my wife. We started dating. I returned for my third year of medical school and it was quite a bit different. WCR: What were the characteristics of Lori that attracted you to her when you were a lifeguard? AND: Here I was on the lifeguard squad in good physical shape. I worked out every day. I was tanned. Most people knew that I was a medical student. Even if you have a bad case of the uglies, you are going to do okay under those circumstances. How many bronzed lifeguard medical students are having trouble finding dates? Not many. WCR: Were you living at home? AND: No. I was living at the shore with another medical student. When I first got there, I worked on the beach all day and went home at night. I was kind of bored. He was volunteering at some community hospital. I wasn’t going to do that. There was a big singles’ bar called “The Surf Club,” with 14 finger bars. It was enormous. I was having a beer there and asked them if they needed any help. I was anxious to work at night. The owner said, “As a matter of fact, we need a bouncer.” By this time I weighed about 165 pounds, but I was still not Gunga Din. I asked the guy to look at me. He said, “Number one, everyone wears casual clothes except the bartenders and bouncers who wear white shirt and tie. So, people are going to know you are a bouncer, and just that fact alone means no one is going to mess with you even though you don’t look like anything. They are going to think you must know karate or something to be a bouncer. Virtually everybody that you are going to need to take care of is going to be drunk.” I became a bouncer for about 1 month and then a bartender position opened up. If there’s one thing better than being a bronzed lifeguard medical student, it’s being a bronzed lifeguard medical student tending bar at the Surf Club with all these singles around. I had all of that going on and here’s this woman I’m taking a shine to. I had noticed Lori on the beach and she seemed like a pretty classy lady and a nice person. I tried to chat her up a little bit and she didn’t give me the time of day. I wanted to tell her, “Lady, don’t you realize? There are 500 women who would drop dead to get a date with me. You don’t seem to give a damn.” Finally, by hook or by crook (I don’t know how I worked it), I got a date with her and we hit it off. WCR: How long after that did you get married? AND: The next year. WCR: You got married after your junior year? AND: We got married in the summer between junior and senior year. We got engaged as my junior year in medical school was coming to an end. We thought we would get married when I graduated. WCR: Where was Lori going to school? AND: She was going to Rider College in the middle of New Jersey. That was the grand scheme. As the summer wore on, we just didn’t want to wait for another year. Neither of us wanted a big wedding anyway, so we decided that we would get married and live midway between Newark (where my medical school was) and the Princeton area (where Rider was). The idea was that we were going to live in between the 2, so I’d commute about 45 minutes and she’d commute about 45 minutes. I’d finish medical school; she’d finish college. We went to our families and announced that we were going to get married. We pulled out all the stops and got married in August. The problem was that being good Catholics, Lori got pregnant instantaneously. That screwed everything up. She took a job substitute teaching instead of going back to college during my senior year at medical school. We had Christine. (Lori went back and finished college, when I was on the faculty at UC Davis. Her degrees are from UC Davis.) When I started my internship we had a baby already. WCR: When you went back to medical school to begin your junior year after the summer of relaxing, meeting Lori, etc., medical school became very attractive to you again. How did that come about? AND: The first 2 years of medical school were really like graduate school in biology. It was heavy basic science, the only clinical thing we did, of any kind, was Introduction to Physical Diagnosis, where we examined patients. The rest was biochemistry, physiology, and immunology, etc. After getting to the INTERVIEW/ANTHONY NICHOLAS DEMARIA 213 clinical rotations in the third year, I became very happy with medicine. It came pretty easily to me. I didn’t have to work so hard at it. I enjoyed interacting with people. I liked the medicine part of medical school. WCR: When going through your rotations, did you have any idea what kind of doctor you wanted to be? As you rotated through surgery, pediatrics, medicine, and all the rest, was your decision to go into medicine easy for you? AND: I knew I didn’t want to be a surgeon and there was nothing else that appealed to me. The need for dexterity and being in the operating room to fix things up just didn’t appeal to me. I flirted a little with orthopedics. I thought orthopedics might be interesting. I got a telephone call from one of the people in medical school asking me what I was interested in. I told them orthopedics. They arranged for me to do orthopedics for one of my very first rotations. That dispelled my interest in orthopedics. I knew I wanted to do medicine, and it was just a question of what kind of medicine to do. WCR: Did any clinicians in your last 2 years in medical school have a particular impact on you? AND: Carol Leevy, Chairman of Medicine, a hepatologist, par excellence, was a world-class investigator. He’d lived in the Hudson County area and was very attached to it. When the medical school made the transition from Jersey City to Newark, there was a lot of disorder as you might imagine. Although a very productive researcher, he was called upon to grab the reins of the Department of Medicine through the transition period and he did. WCR: Were you first in your graduating class in medical school? AND: No. I’m sure I wasn’t first. I wasn’t the valedictorian speaker. For the life of me, I can’t remember who was. I wasn’t too far from first. WCR: When it came to picking an internship, how did you reason that out? AND: I was married and had a baby. My becoming a physician exceeded everyone’s wildest expectations. I not only went to college, I went to medical school. I became a doctor. The entire neighborhood was abuzz. When it came to internship, there was a guy who really was much revered by all the medical students. His name was Gustave Laurenzi. He was a pulmonologist. He really made an impression. He was flamboyant, had a big mustache, had a lot of dramatic qualities, and, of course, was very smart. Harold Jeghers, Chairman of Medicine at the New Jersey College of Medicine, had been recruited to go to Worcester, Massachusetts, to a hospital called St. Vincent’s to participate in the planning of the University of Massachusetts Medical School. He recruited Gustave Laurenzi to be the chairman there. Laurenzi went around to the students that he knew, and made a pitch to consider going to that hospital for internship. It was, to a large degree, the power of his personality coupled with my goal of going into practice. I figured I’d do an internship and residency and open an office and be a practicing doctor. Having a wife and child also was a 214 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 significant factor. The idea of going to a hospital where they had a house-staff apartment building next door was fairly attractive. Five of my classmates and I took an internship at that hospital. They filled their internship program. It was the biggest bolus they ever had gotten from 1 medical school. For me, there was a little extra in it because I already knew Worcester. I’d gone to college there. It was like going back to my college town. It was very convenient. I figured for being in private practice, it might even have an advantage because that’s what private practice was all about. I made the decision and went with Laurenzi. From the standpoint of clinical medicine, it was a great internship. We were deluged with patients and worked every other night and every other weekend. The hours were long beyond belief. The “clinical material” was everywhere. They had a solid group of very busy practitioners and they had a full-time faculty of about 12, a full-time staff in every specialty. It was a place where sick people were taken care of. Laurenzi and all the full-time people did some research. The Chief of Cardiology was Richard Myler, who subsequently went to San Francisco and coronary angioplasty fame. At the end of that year, I was drafted. I applied to go into the United States Public Health Service. I initially contacted them because I was going to volunteer for an Indian reservation. I figured that would be an idyllic place. By this time my wife was pregnant again with our second child. I thought that might be an interesting way to spend 2 years because Vietnam was red hot. I had to discharge the service obligation. They came back with an opening for a medical resident in the Staten Island Public Service Hospital. I told them, “You got it.” I was a resident at the Public Health Service Hospital. That turned out to be phenomenal. The hospital was busy as could be. There were many junior faculty from medical schools who had joined the faculty before the Vietnam war, but were still eligible for the draft. We were eligible until age 35. They were pulling guys out of junior faculty positions to go back in the service. There were a lot of guys who signed up to do 2 years in the Public Health Service. We had a ton of patients and we had an interesting accumulation of young academic people. We didn’t have a lot of grand professors, but there were some lifetime guys who were pretty good. It was a really great experience. Of course, in Staten Island, Tony Damato had established a cardiac unit, where they did the first His bundle electrocardiogram. They gave birth to the field of clinical electrophysiology. There were a ton of guys going through there: Ken Rosen, Mike Rosen, Bruce Goldreyer, John Gallagher (a classmate of mine at Holy Cross), Masoud Aktar, and Mark Josephson. WCR: You finished your medical residency as a Public Health Service officer, which fulfilled your draft and also paid you reasonably well. When did you decide you wanted to become a cardiologist? AND: The pay wasn’t great, but that’s basically true. There wasn’t any way Staten Island was a bonanza! It was good medical training, I was exposed to JANUARY 15, 2005 a lot of new developments going on in medicine, and I met many of the up-and-comers in cardiology. Cardiology was a huge thing at Staten Island because of all the His bundle work and the electrophysiology, etc. That tweaked my interest in cardiology. There was also some thinking that the Public Health Service might need subspecialists, particularly cardiologists. It sounded good to me and I thought maybe that was the way I ought to go. WCR: How did it come about that you went to Davis, California, to do your cardiology fellowship? AND: Another senior resident and I both decided on cardiology. As fate would have it, we were kind of desirable properties; because of our being at Staten Island, we had a fair bit of exposure to intracardiac electrophysiology when no one else had. My fellow resident interviewed at Indiana with Charles Fisch and was accepted into Fisch’s program. Later, Fisch mentioned to Dean Mason, who had just gone to UC Davis to be the Division Director of Cardiology, that he had hired a senior resident from the Public Health Service and heard that there was another guy looking for a cardiology fellowship. I got a telephone call from Dean Mason who said, “I’m starting a cardiology program out here and I think you might like it. Why don’t you come out and take a look?” He even offered to pay my way out. My wife got a baby sitter for the kids and we both flew out to Davis on a cold, rainy New York day in March. When we got to Davis it was sunny and warm. Dean Mason was an impressive person and we said, “Fine.” We were there for 36 hours and even rented a house for our coming in July. Because of being with Damato and all those electrophysiology guys early on, we were pretty desirable. I actually had the opportunity to go to Duke, to Yale, and to Emory. I went to Davis. That was one of the best decisions I ever made. But it was truly uninformed. A knowledgeable person probably would not have turned down the others. Willis Hurst also called me and offered me a job on the telephone. He’d written a big cardiology textbook by that time. I was on rounds at the Public Health Service Hospital and a nurse came up to me and said, “Dr. DeMaria, there’s a Dr. Hurst from Atlanta on the phone.” I thought, Holy Mackerel! But, I said “no” to all of that and went to this fledgling embryonic program in Davis. I was in the first fellowship class. Some of it was the charisma of Dean Mason. A lot of it was the attraction of being in California for the first time. It was just a gut feeling. I sat down with Dean Mason and he said to me, “What do you want to do in cardiology?” I said, “I’m going into practice. I’m going to open an office and practice on the Jersey shore where my wife lives because we like it there. We thought it might be fun to come to California for a couple of years and be trained, before we settled back East and I go into practice.” That’s exactly what I told him. WCR: This was during the interview? AND: Yes. It shows you what bad shape he was in; he took me anyway. WCR: How did he respond to that? AND: He didn’t say much of anything. As I say, I had a zero degree of sophistication in terms of what I was doing. I turned down all the prestigious places to go to Davis. No one even knew where Davis was. I had no idea you were supposed to say that you wanted to do research. Nobody told me that. I just told him what I thought. I figured honesty was the best policy. WCR: Had you written any papers by that time? AND: No. WCR: Research hadn’t even occurred to you yet? AND: Correct, not in the least. I got the intern’s award during my internship. There was a competitive test in the Public Health Service Hospital and I finished on top of that also. WCR: Who had Dean Mason hired by the time you arrived at Davis? AND: He’d hired Ezra Amsterdam from Boston, Rashid Massumi from George Washington, and Jim Spann and Bob Zelis, both former colleagues of Dean’s at the NIH. WCR: Were you the only fellow? AND: No. There were 3 of us. WCR: Who were the other 2? AND: Richard Miller, who went on to be chief of cardiology at Baylor College of Medicine in Houston and Louis Vismara. WCR: How did it work out for you at Davis? AND: I went there with the intention of learning how to be a cardiologist for 2 years while enjoying northern California and exploring the West Coast. Of course, the environment I was exposed to was totally research oriented. With Dean setting the agenda, I’d no sooner arrived there than I realized that the most important thing was that in September (a couple of months later) there was a deadline for research submissions to the Annual Scientific Sessions of the American College of Cardiology and what was I going to do to get some research submissions. The next most important thing was a grant deadline. I was entrained by the whole momentum of the place. Zelis and Amsterdam were junior guys who were very ambitious. Massumi, the electrophysiologist, was a very harddriving individual. I just got swept up in it. I couldn’t help but being swept up in it. I enjoyed it. I got derailed a little bit early on. I had a basic research project where we tried to study the effect of asynchrony of contraction on intrinsic myocardial performance. I’d go to the dog lab every day. It was a brutally difficult protocol to implement. We did cardiopulmonary bypass and put a button in the mitral valve position in the dog and put a balloon in the left ventricle. I didn’t like this very much. Dean Mason called me in one day and said, “I’ve just gotten this machine to work with that I think is going to be an important tool in cardiology. We are one of the few places that have it. I want you to take it on as your project. It’s called the radarkymogram. ” The radarkymogram! Do you remember that? It was a device that could track an epicardial border on fluoroscopy. I had to do a fluoroscopy and get the epicardial border and put a marker on it and it would trace it. From that you were supposed to derive information about cardiac function and everything under the sun. He told me, INTERVIEW/ANTHONY NICHOLAS DEMARIA 215 “That’s your project.” I said, “All right, Chief.” I worked like crazy for about 3 months on the project. One of my first publications was a review publication on radarkymography, a little known field. I told him that the radarkymography was just a bunch of bologna. It was the epicardial area and I couldn’t make much out of it, despite what the manuscript said. While I was trying to make it work, I read a paper written by Harvey Feigenbaum in Indianapolis about trying to do the same thing with ultrasound. I went to Dean Mason, which is a hutzpah kind of thing, and told him that I didn’t think the radarkymogram was ever going to make it, but there was a technique with ultrasound that might make it, so that we could see the inside, not the outside of the heart. I wanted to learn that. I wanted to go to Indianapolis where my buddy from the Public Health Service was. I could stay at his house and find out how to do this from Feigenbaum. Dean Mason said, “If I were you, I wouldn’t waste my time with this ultrasound thing. I can tell how bad mitral stenosis is and that’s all it is used for.” I prevailed upon him, telling him I really wanted to do this. I felt it could amount to something. He was going to humor me and said, “I’ll make a deal with you. I have a drug study that I need to get done. If you agree to do the drug study for me and get it done, then you can go to Indianapolis and I’ll take some of the money from the drug study and I’ll buy the echo machine.” An M-mode echo only cost $10,000 in those days. I told him we had a deal. The drug study turned out to be perhexiline, which came and died because nobody could figure out how it worked. Subsequently, somebody figured out that it was a calcium channel blocker. The drug was probably a decent drug all the while. In those days, Harvey Feigenbaum was “the king.” He had people come to his lab for 1 week. He claims that I gave the title “1-week wonder” to his training program. Walter Henry from the NIH went there. Rich Popp had been a fellow under Feigenbaum. Almost anybody who was anybody in echocardiography had made a pilgrimage to Indianapolis. I went for 1 week to learn echocardiography. During the breaks, when we weren’t examining patients, I read papers. Harvey had assembled every manuscript that had ever been written on cardiac ultrasound in 1 loose-leaf notebook. There were only about 10 of them plus a big symposium in Acta Medica Scandinavia. That was the world’s literature on echocardiography! I read the world’s literature in 1 week. In addition to that, Harvey had managed to get the first connection of an echo machine with a strip chart recorder. For the first time, instead of taking 1 Polaroid picture and getting a couple of beats on the picture, now he could get a strip chart that he could scan around. Holy mackerel, this was really going to amount to something! I went back to Davis on fire with all this stuff and talked to Dean Mason. He still thought I was a bit nuts, but he was a man of his word. He got the Smith-Kline echo machine for me. The only place we could find to do echocardiography in the hospital was in physical medicine and rehabilitation. Dean managed to get room in the basement, way down in God’s 216 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 never-never land, to put the machine in to do echocardiography. I felt like the lone ranger. I was trying to find patients to perform echo on to get some examples and to get some practice. Everybody summarily ignored me. One day during professor’s rounds, Dean saw a woman in the coronary care unit. This elderly woman had been admitted to the hospital 3 times over the course of about 8 weeks with severe unremitting chest pain, profound ST T-wave changes, and mitral regurgitation. She was old and didn’t want catheterization. They were reluctant to catheterize her, because to do a coronary angiogram in 1971 was a big deal. After rounds, I got my echo machine and took it up to the coronary care unit to get an example of what an echo looks like in a patient with mitral regurgitation. Of course she had the most incredible florid flagrant hypertrophic cardiomyopathy. The next day on rounds I showed the echo and told them what she had. Dean Mason became a believer! He said, “My God. This is unbelievable.” The echo showed clearly what she had and everybody else had missed. There was still a lot of skepticism. I was struggling to convince people, and 1 day Zelis called me up and said, “I know you are looking for patients. I’ve got a good mitral stenosis over here. It’s not quite classic, but it would be a nice one for you to echo. I’ve scheduled her for cath.” I told him to send her over. She had an atrial septal defect. From that point on, Zelis never catheterized another patient who hadn’t had an echo first. Then we applied for a big grant on disease of the muscle. I was going to do echoes in patients with muscular dystrophy. We got the NIH grant. Now I had enough money to hire a technician. Up to this point I had been everything—the sonographer, the machine washer, everything. Alex Newman, a PhD student, was washing glassware in the cath lab. Everybody said he was really good but was just trying to figure out what he wanted to do with his life. I didn’t know him very well, but I talked to him saying, “I’ve got a technician position in ultrasound. It pays more than washing glassware in the cath lab.” He immediately became my echo tech. The 2 of us started doing research. We wrote early papers on what nitroglycerin does to the heart, what mitral prolapse does, and Wolff-Parkinson-White, and we performed the first myocardial contrast echocardiography. We were spinning and spinning. Those were magic years. With that, Dean Mason called me and said, “You need to be on the faculty.” He got me a full-time faculty position. I went home and talked to Lori, “This is kind of fun. I like this stuff. I think we ought to keep doing it.” She said, “I like living here in California.” I told her, “Let’s keep doing it and if worse comes to worse, I can always go into practice. I’ll do it for a couple of years.” It was a magic time in Davis. It was a really productive place. Rich Miller was publishing like crazy. Dean was working 24 hours a day. There was a ton of stuff going on. I kept thinking that maybe I ought to go into practice. One day a guy called from Palos Verdes, California, looking for somebody to run cardiology at their hospital and asked if I would be JANUARY 15, 2005 FIGURE 4. AND receiving a plaque from Dr. Dean T. Mason, at that time President of the American College of Cardiology, for organizing the program of the Annual Scientific Sessions the year that Dr. Mason was the College’s president. interested. I went down and interviewed for the job and they offered it to me. I was very ambivalent. I decided that I would ask for the sky—a high salary, help with my mortgage, all kinds of stuff. I would ask for the moon and if they gave it to me, then of course I couldn’t turn it down. If they said they wouldn’t do it, I’d stay in academics. That was the way I could resolve my ambivalence. I asked for the moon. After a while they called me to apologize for the time that had gone by. They were pleased to tell me that they would grant me everything I requested, 100% of the things I requested. At that point I realized that practice was just not what I wanted to do. I never looked back after that. I was very lucky, because I got to know Harvey Feigenbaum by being a 1-week wonder. As he taught courses and presided over meetings, I was one of the people he would include. He helped me a lot. Of course, I owe everything to Dean (Figure 4). He was mentor par excellence. Dean was just spectacular. Harvey helped a lot. Lo and behold, after a time Dean was president of the American College of Cardiology and somebody had to run the Annual Scientific Session. He was going to have Richard Miller run it, but Rich left for Baylor to be Chief of Cardiology. Here he was, a couple of months into his presidency and didn’t have anybody to run the meeting. He grabbed me and said, “You’ve got to be Chairman of the Annual Scientific Session for the American College of Cardiology.” What the hell did I know? I was still wet behind the ears. We did it together. That year the whole program was laid out on the floor of my house. I had stacks of papers in different categories. I met a lot of people and got to know them very well. That helped. WCR: How did the move to Lexington come about? AND: That was kind of a “2 for 1.” I was a pretty young guy. I was 37. They were talking to me about being Chief of Cardiology. I liked the man who was recruiting me, John Thompson, Chairman of Medicine. I was fascinated by the idea, so I visited there. Then Davis had a problem with the cardiac surgical program. There was a fair amount of tension in the air at Davis. At some point in time, Dean was no longer Chief of Cardiology and they asked me to take the job. I couldn’t do it. I left. Lexington was a good place to go. The Chairman of Medicine was young and energetic. He was rebuilding a department of medicine. They were enormously generous in providing a recruitment package. It was a time for me to see if I could do things my way. WCR: That was 1981? AND: Yes. Lexington is a nice town, a great place to live. By this time, my wife and I had 3 kids and we liked living in small towns. I had good reason to go there. I was flattered. I was arrogant enough to think at 37 I could run my division. I was too stupid to realize that I was probably too young to do that. WCR: You recruited some good people there. AND: Yes. I took Steve Nissen with me from Davis. He was a resident at Davis and had already signed up for the cardiology fellowship program there. When I left, he came with me. John Waters came also. After I got to Lexington, I recruited some really good people—Tom Wisenbaugh, a really good hemodynamist Cindy Grines and Jon Elion who did really well. A guy called me during Christmas vacation. My secretary said, “There’s some fellow candidate who wants to talk to you. He happens to be visiting his in-laws in Louisville. He might want to do a fellowship at Kentucky. He wants to know if he can come over.” I said, “What the hell. Things are slow. It’s Christmas.” Paul Grayburn came over to visit. I was blown away. I thought he was terrific. I quickly sent him to talk to a couple of other guys who were around. I offered him a job on the spot. WCR: He told me you called his home and his wife answered the telephone. She, being from Louisville, accepted for him. When he got home that night she told him where he was going for his cardiology fellowship. AND: That’s actually the way it went down. He knew darn well when he left that he was getting the fellowship. Usually, after the interview you wait and see all the candidates and make a decision. When he was leaving, I told him not to take any other job unless he talked to me first. He must have known that he was accepted. The way, we selected fellows was that we’d get all of our applications and the faculty would sit around to discuss them at length and offer jobs. In this particular instance, I made it fairly clear to Paul that I was going to have the faculty sit around and deliberate this, but all the votes weren’t going to be equal on that. WCR: Paul did a little bit like you did. He interned at a private hospital in Dallas, thinking he was going to go into practice. AND: He reminded me of myself. The thing about Paul that really gets me is that when he applied to Kentucky, he was exactly the way I was when I applied to Davis. He hadn’t done a lick of research. I just knew that there was no way in God’s earth he could ever get an interview at University of Texas Southwestern. Jim Willerson was running the program INTERVIEW/ANTHONY NICHOLAS DEMARIA 217 FIGURE 5. AND as a Master in Cardiology in 1988. there and would have looked at his resume just the way he would have looked at mine. This guy comes to Lexington and I say, “You’ve got a job here.” He comes and he trains and we work with him for 3 years, teaching him everything we knew. I said this guy is going to stay here and he’s going to be a star and really build this program. Willerson, with 1 telephone call, recruited him back to Dallas. I thought that Willerson wouldn’t have given him the time of day 3 or 4 years earlier. Now all of a sudden Willerson says, “You need to come back to Dallas.” Grayburn went back. We lost him. WCR: Your experience in Lexington turned out to be a very good one? AND: It was a terrific experience (Figure 5). We made a name for ourselves in digital angiography. Steve Nissen started his intravascular ultrasonic imaging career there. He and I worked together very closely for 11 years. In the best tradition of mentorstudent relations, the student has now surpassed the mentor. Cindy Grines did a couple of really important studies and then moved on to work with Bill O’Neill again. We had a pretty good clinical volume. We trained a lot of good doctors. Many went into academics for only a few years and then drifted off regrettably. That was one of my regrets. It was very hard to leave Lexington. 218 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 WCR: How did the San Diego opportunity come about? AND: John Ross called me. The institution had recognized that they needed to strengthen the clinical program in cardiology. They were looking for someone to be Chief of Clinical Cardiology. John asked me about it. We liked Lexington, but San Diego was a bigger and more visible program and I admired John very much. The university had a lot of attraction about it, and being back in California, where we had spent 10 years earlier, helped. I went and looked at the position. I was leaning toward not going. I was happy in Lexington and was going to stay put. While I was looking, John made a decision to step down as Chief of Cardiology. That left the division head open and they made me a good offer. It was time. I stayed at Davis for 10 years; I stayed in Lexington for 11 years. It was time to move. So I recruited my assistant Kate Greathouse, and my lab manager, Oi Ling Kwan, both of whom I have worked with for Пѕ20 years and depend on greatly, to go west (Figure 6). Some people move, some people don’t. I moved. WCR: You went to San Diego in 1992. How many faculty do you have there now in cardiology? AND: It’s a relatively small program. We have 28 people on full-time faculty. WCR: How many fellows do you have? JANUARY 15, 2005 FIGURE 6. AND with his wife Lori and devoted assistants. AND: It varies from year to year. We have 3 regular training positions. WCR: Per year? AND: Per year. That’s 9. Then we have 2 interventional and 2 electrophysiology fellows. We have a number of people who are there just on research fellowships. WCR: If you take all the faculty positions and all the fellows, the secretaries, and the technicians, etc., you are talking about 150 people? AND: I’d think closer to 125. As of July 1, I’ve transitioned. WCR: You’ve stepped down as head the division of cardiovascular medicine. You are now director of the Cardiovascular Center. What is that? AND: Shortly after I got there, about 7 years ago, Stuart Jamieson, Ken Chien, and I wanted to have a cardiovascular center that would bring together all the people at our institution who were working in cardiovascular disease. We were spread out all over the place. We are not only in different areas—La Jolla versus the Hillcrest area of San Diego— but even within La Jolla at the Basic Science building and VA Hospital. We had approached the dean about a Cardiovascular Center to make it a comprehensive entity. WCR: That’s across departments? AND: Yes. Along the way, Ken actually became head of the Institute for Molecular Medicine. There’s an institute at UCSD dedicated to molecular medicine—all aspects—not only cardiology (Ken is mostly cardiology), but cancer, immunology, the whole ball of wax. He’s still associated with the cardiovascular center, but not in a primary position. We formed the group. The triumvirate of Stuart Jamison, Chief of Cardiac Surgery and me. As a research director, we have Kirk Knowlton. He bridged the gap between Ken Chien and me very nicely. Kirk is an excellent research guy, but he has clinical activities as well. We were working on the center raising money. About 8 to 10 months ago, Ed Holmes, the Dean, called me and said, “Listen, if you can raise $30 million, I can put in $70 million. The hospital will assume debt. Let’s build a cardiovascular center and make it inpatient and outpatient. We’ll focus on clin- ical and translational research. We’ll house cardiac surgery, adult cardiology, the pulmonary vascular program, pulmonary hypertension of all types a stroke program (Pat Lyden), all cardiovascular radiology, preventive cardiology (Michael Criqui from the department of preventive medicine), peripheral vascular surgery, and peripheral vascular medicine. The idea was that we would construct a building that would have inpatient and outpatient facilities and clinical research facilities (but not basic science laboratories, not hoods, not even small animals). The catheterization and electrophysiology labs would have adequate backup space so investigators could tie them up doing a research study. It’s not just a hospital. It will have all the clinical facilities that you would want to have available to do research projects—appropriate space for noninvasive imaging, cardiovascular, radiology, electrophysiology cath, or whatever you want for clinical research. Ed Holmes thought it was a full-time job. It’s probably close to that. More importantly, this position will oversee adjudication resources— how is space allotted, how are offices allocated, how is priority for hospital beds decided. Those issues on space and money required, in Ed Holmes’ opinion, and which I couldn’t disagree with, that someone had to be free of a vested interest in those decisions. The question was what did I want to do? I thought it was a neat challenge to build a cardiovascular center, and one that would fit in very well with my lifestyle at this point in time. Being Editor of the Journal of the American College of Cardiology obviously occupies a lot of my effort. The Dean and I negotiated a bit and came to an amicable agreement. As of July, I have accepted that responsibility. Kirk Knowlton will assume the duties of Chief of Cardiology. The goal is to have the building up and running in 4 years. It should take us 1 year to plan it out and another 2 to 3 to get it built. WCR: How are you enjoying the editorship of the Journal of the American College of Cardiology? AND: I love it. It’s been a great privilege to have that opportunity. It is a unique opportunity to influence what people read, and therefore, perhaps what they think. We have a sacred responsibility to the authors to treat their work with the respect that they put into it. I feel like I have my finger on the pulse of cardiology. Anything that is going on is being submitted, and usually substantially before it surfaces as something important. It’s a learning experience. It’s an opportunity to influence my profession. It’s gives me a certain creative opportunity in writing editor’s pages and making decisions on how to put manuscripts together and organize them in certain ways. It’s been a challenge in handling this transition from a purely paper journal to one that’s now paper and on-line, and will be progressively on-line. WCR: What were some surprises you as you took over the editorship? AND: I really hadn’t anticipated the degree of serendipity that’s involved in what gets accepted for publication and what doesn’t. We always joked from my earliest days with Dean Mason that the biggest INTERVIEW/ANTHONY NICHOLAS DEMARIA 219 variable was the reviewer. I fully anticipated that we were going to have to allot for variability on the part of a reviewer, but even with that I have been amazed at the frequency with which critiques are opposite. In addition to reviewer variation are other things that come into play. We went back for a period of 1 year and found that in just ПЅ40% of our manuscripts, 1 reviewer ranked it in the top 2 (of 5) categories and the other rejected it. That was in almost 40% of the manuscripts. I was totally unprepared for that much divergence in the way 2 learned clinicians would assess a given manuscript. The reality of the volume and timelines of work have also been striking. Every day, more manuscripts turn up. It’s got to be handled in a timely fashion. It’s not something you can put on the side. The camaraderie with the associate editors has been an unexpected side benefit. Working with Ori Ben-Yehuda, Wilbur Lew, Sam Tsimikas, Greg Feld, Kirk Knowlton, and Barry Greenberg has bonded us closely. WCR: I guess you get the question a lot and I’ve gotten it a lot through the years: How much time do you spend a week on the Journal of the American College of Cardiology? AND: About 25 hours a week. It varies up and down. This week I won’t spend any. In an average week I usually dedicate 11/2 entire weekend days at least, and usually a bit more to just doing journal things. I dedicate about 1 working day during the week to preparing and handling our weekly meeting, and then there’s a couple of hours each day depending on different days. Nobody works a 40-hour week in our business. We’re in a competitive business being editors in this day and age. If you and I are competing for the best papers, then one of the things we sell is rapid response. One thing we have not talked about is probably worth talking about. Part of my years in Lexington was being President of the American College of Cardiology. I had come to that position by Dean Mason making me the Scientific Sessions program chairman and then I got involved a lot in the Government Relations Committee and private sector relations. At a relatively young age (I think I was the youngest president ever), I got the opportunity to be president. When they called me and asked me if I was willing to be nominated, I said, “I’m too young for this, but, on the other hand, you may never ask me again.” I went and did it. WCR: You were how old? AND: I was 45 when I was president (1988 to 1989), but I was nominated at age 42. That was a marvelous experience. The year I was president we laid the foundation in a major way for some of the advocacy programs, which I think have been very meaningful to clinical cardiologists, while preserving the important educational mission and the support for scholarship. I had the opportunity to travel for the college. It’s a bit of a bully pulpit to tell people what you think cardiology should be like in those President’s Pages. I found that thoroughly enjoyable. If you had told me when I sat in front of Dean Mason, or even when I graduated medical school, “You know 220 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 what? You’re graduating from New Jersey College of Medicine. Someday you’re going to be president of the echo society, president of the American College of Cardiology, and the Editor of the Journal of the American College of Cardiology.” I would have told you, “You’ve got to be out of your mind. That could never happen.” WCR: Of all the things you’ve done professionally, what are you most proud of? AND: The people I’ve trained. Of the people I’ve had a chance to influence, some are famous—Steve Nissen and Paul Grayburn. Many are important people in medical schools—Bill Bommer at Davis, Michael Smith at Kentucky, Ari Ben Yehuda, Ajit Raisinghani, Shami Mahmud, and Sam Tsimikas here at UCSD. I’ve trained many foreign physicians. I’ve had the opportunity to have colleagues from countries around the world who’ve worked in the lab for 1 or 2 years and gone on to be major investigators. Far and away, that’s been the thing I’ve enjoyed the most. I most enjoy when fellows write their first papers and hold them in their hands. It doesn’t get much better than that. WCR: What happened to your 3 sisters? AND: Two are single and still live in New Jersey. Adele is in business and Carol is a schoolteacher. My oldest sister Barbara Hupe lives in Houston, Texas now. She’s also a schoolteacher. Her husband died an untimely death of a brain tumor. WCR: So 2 of the 3 are teachers? AND: Right. WCR: Did all 3 go to college? AND: Yes. WCR: That’s quite remarkable. AND: Especially for my sisters, because the philosophy at the time was men needed to go to college, but for women, it was a luxury. They went. WCR: Tell me about your immediate family. AND: My wife is the best thing about me. Everybody likes Lori. She’s marvelous. We have been partners in everything I’ve done, including the new UCSD Cardiovascular Center. She’s wise, insightful, and a tower of strength. WCR: What was her premarital name? AND: Horn. We’ve been married 37 years now. I’ve been very fortunate. That’s the best decision I ever made in my life. Going to Davis was the second best. Doing echocardiography was probably my third best decision. Our oldest daughter Christine describes herself as a recovering lawyer, who is now getting a PhD in sociology. She and her husband Matt have a son, Jackson, and are having a second child. Christine went to the University of Kentucky Law School. Our middle child Anthony is a very successful lawyer in Fresno, California. He and his wife Kim have 3 children, Alec, Blake, and Cameron. He’s phenomenally successful. He wins almost all his cases. He did very well in law school at USD Law School. Our son Jonathan is involved in a number of different activities. He’s very bright but hasn’t quite decided where he wants his career path to go yet (Figures 7 and 8). JANUARY 15, 2005 FIGURE 7. AND and his wife and three offspring. Anaheim. They flew my mother to Los Angeles and put her in a limo, and brought her to the hotel in Anaheim. I got a call that my mother was at the hotel. I went to the lobby to meet her. I said, “Mom, welcome. How’re you doing?” She said, “I’m fine. I’m fine.” Very soon she asked, “Where’s the telephone? I’ve got to make a telephone call.” I was taken aback. She called her best friend, Rose Yazwinski, and said, “Rose, you’re not going to believe this. They had a car just for me, a very big one! Rose, you could put 4 or 5 people in the car. I was the only one in it.” I honestly believe that that limo made my mother think that maybe I’d done all right. And all those years they suspected that I just couldn’t quite make it out of medical school! WCR: Tony, many thanks for pouring your soul out. It was great. MOST IMPORTANT PUBLICATIONS SELECTED BY AND FIGURE 8. AND and his wife with their grandchildren. WCR: Is there anything you’d like to talk about that we have neglected to talk about? AND: Two of my favorite stories. The first story happened when I was at Davis, my father called me and said, “Anthony, I’ve got to talk to you man to man. The people around the neighborhood are beginning to ask questions. How come Anthony is hanging around the medical school so long? It’s getting embarrassing. They wonder if you haven’t finished?” I told him, “Dad, I’m not going to do practice. I’m going to be an academic.” He said, “Anthony, you’ve got a wife and 2 kids. You’ve got to stop fooling around here. I’m calling you to tell you this very day that Dr. Fabrielle, who used to be on 46th and Avenue C, died, and his wife came in to tell me that you can have the practice if you want it.” I told him, “Dad, I’m going to stay at the medical school.” The second story. When as President of the American College of Cardiology, I invited my mother to come to the annual meeting. (The College lets the president bring their family to see their child conduct the meeting at the Convocation.) The meeting was in 1. Massumi RA, Mason DT, Amsterdam EA, DeMaria AN, Miller RR, Scheinman MM, Zelis R. Ventricular fibrillation and tachycardia after intravenous atropine for treatmentof bradycardias. N Engl J Med 1972;287:336 –338. 14. DeMaria AN, King JF, Bogren HG, Lies JE, Mason DT. The variable spectrum of echocardiographic manifestations of the mitral valve syndrome. Circulation 1974;50:33– 41. 15. DeMaria AN, Vera Z, Amsterdam EA, Mason DT, Massumi RA. Disturbances of cardiac rhythm and conduction inducted by exercise: diagnostic, prognostic and therapeutic implications. Am J Cardiol 1974;33:732–736. 16. DeMaria AN, Vismara LA, Auditore K, Amsterdam EA, Zelis R, Mason DT. Effects of nitroglycerin on left ventricular cavitary size and cardiac performance determined by ultrasound in man. Am J Med 1974;57:754 –760. 18. DeMaria AN, King JF, Salel AF, Caudill CC, Miller RR, Mason DT. Echography and phonography of acute aortic regurgitation in bacterial endocarditis. Ann Intern Med 1975;82:329 –335. 20. DeMaria AN, Lies JE, King JF, Miller RR, Amsterdam EA, Mason DT. Echocardiographic assessment of atrial transport, mitral movement, and ventricular performance following electroversion of supraventricular arrhythmias. Circulation 1975;51:273–282. 21. DeMaria AN, Vismara LV, Miller RR, Neumann A, Mason DT. Unusual echographic manifestations of right and left heart myxomas. Am J Med 1975;59: 713–720. 31. DeMaria AN. The syndrome of mitral valve prolapse: problems and perspectives. Ann Intern Med 1976;85(4):525–526. 32. DeMaria AN, Amsterdam EA, Vismara LA, Neumann A, Mason DT. Arrhythmias in the mitral valve prolapse syndrome: prevalence, nature and frequency. Ann Interm Med 1976;84:656 – 660. 33. DeMaria AN, Miller RR, Amsterdam EA, Markson W, Mason DT. Mitral valve early diastolic closing velocity in the echocardiogram: relation to sequential diastolic flow and ventricular compliance. Am J Cardiol 1976;37:693–700. 36. DeMaria AN, Vera Z, Neumann A, Mason DT. Alterations in ventricular contraction pattern in the Wolff-Parkinson-White syndrome: detection by echocardiography. Circulation 1976;53:249 –257. 59. DeMaria AN, Neumann A, Lee G, Mason DT. Echocardiographic identification of the mitral valve prolapse syndrome. Am J Med 1977;62:819 –29. 97. DeMaria AN, Neumann A, Lee G, Fowler W, Mason DT. Alterations in ventricular mass and performance induced by exercise training in man evaluated by echocardiography. Circulation 1978;57:237–244. 101. DeMaria AN, Oliver LE, Bogren HG, George L, Mason DT. Apparent reduction of aortic and left heart chamber size in atrial septal defect. Am J Cardiol 1978;42:545–550. 119. Sahn DJ, DeMaria AN, Kisslo J, Weyman A. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978;58(6):1072–1083. 125. DeMaria AN, Bommer W, Neumann A, Weinert L, Bogren H, Mason DT. Identification and localization of aneurysms of the ascending aorta by crosssectional echocardiography. Circulation 1979;59:755–761. 129. DeMaria AN, Neumann A, Bommer W, Weinert L, Grehl T, Amsterdam EA, Mason DT. Left ventricular thrombi by cross-sectional echocardiography. Ann Inten Med 1979;90:14 –18. 130. DeMaria AN, Neumann A, Schubart PJ, Lee G, Mason DT. Systemic correlation of cardiac chamber size and ventricular performance determined with echocardiography and alterations in heart rate in normal persons. Am J Cardiol 1979;43:1–9. INTERVIEW/ANTHONY NICHOLAS DEMARIA 221 148. DeMaria AN, Bommer W, Joye JA, Lee G, Bouteller J, Mason DT. Value and limitations of cross-sectional echocardiography of the aortic valve in the diagnosis and quantification of valvular aortic stenosis. Circulation 1980;62:304 –312. 150. DeMaria AN, Bommer W, Lee G, Mason DT. Value and limitations of two-dimensional echocardiography in assessment of cardiomyopathy. Am J Cardiol 1980;46:1224 –1231. 185. Low RI, Takeda P, Lee G, Mason DT, Awan NA, DeMaria AN. Effects of diltiazem-induced calcium blockade upon exercise capacity in effort angina due to chronic coronary artery disease. Am Heart J 1981;101:713–718. 195. Woythaler JN, Singer SL, Kwan OL, Meltzer RS, Reubner B, Bommer W, DeMaria AN. Accuracy of echocardiography versus electrocardiography in detecting left ventricular hypertrophy: comparison with post-mortem mass measurements. J Am Coll Cardiol 1982;2:305–311. 200. Nissen SE, Booth D, Waters J, Fassas T, DeMaria AN. Evaluation of left ventricular contractile pattern by intravenous digital subtraction ventriculography: comparison with cineangiography and assessment of interobserver variability. Am J Cardiol 1983;52(10):1293–1298. 204. DeMaria AN, Bommer W, Kwan OL, Riggs K, Smith MD, Waters J. In vivo correlation of thermodilution cardiac output and videodensitometric indicatordilution curves obtained from contrast two-dimensional echocardiograms. J Am Coll Cardiol 1984;3:999 –1004. 207. Johnson GL, Kwan OL, Cottrill CM, DeMaria AN. Detection and quantitation of right ventricular outlet obstruction secondary to aneurysm of the membranous ventricular septum by combined two-dimensional echocardiography: continuous-wave Doppler ultrasound. Am J Cardiol 1984;53:1476 –1477. 208. Johnson GL, Kwan OL, Handshoe S , Noonan J, DeMaria AN. Accuracy of combined two-dimensional echocardiography and continuous wave Doppler recording in the estimation of pressure gradient in right ventricular outlet obstruction. J Am Coll Cardiol 1984;3:1013–1018. 213. Smith MD, Kwan OL, Reiser HJ, DeMaria AN. Superior intensity and reproducibility of SHU-454, a new right heart contrast agent. J Am Coll Cardiol 1984;3:992–998. 219. Friedman BJ, Waters JS, Kwan OL, DeMaria AN. Comparison of magnetic resonance imaging and echocardiography in determination of cardiac dimensions in normals. J Am Coll Cardiol 1985;5:1369 –1376. 221. Smith MD, Dawson PL, Elion JL, Booth DC, Handshoe R, Kwan OL, Earle GF, DeMaria AN. Correlation of continuous wave Doppler velocities with cardiac catheterization gradients: an experimental model of aortic stenosis. J Am Coll Cardiol 1985;6:1306 –1314. 229. Friedman BJ, Drinkovic N, Miles H, Shih WJ, Mazzoleni A, DeMaria AN. Assessment of left ventricular diastolic function: comparison of Doppler echocardiography and gated blood pool scintigraphy. J Am Coll Cardiol 1986;8(6): 1348 –1354. 230. Grayburn PA, Smith MD, Handshoe R, Friedman BJ, Handshoe S, DeMaria AN. Comparative accuracy of pulsed Doppler, echo and auscultation in the detection of aortic insufficiency. Ann Int Med 1986;l04:599 – 605. 235. Nissen SE, Elion JL, Booth DC, Evans J, DeMaria AN. Value and limitations of computer analysis of digital subtraction angiography in the assessment of coronary flow reserve. Circulation 1986;73:562–571. 240. Smith MD, Handshoe R, Handshoe S, Kwan OL, DeMaria AN. Comparative accuracy of 2-D echocardiography and Doppler pressure half-time methods in assessing severity of mitral stenosis in patients with and without prior commissurotomy. Circulation 1986;73:100 –107. 241. Smith MD, Kwan OL, DeMaria AN. Value and limitations of continuouswave Doppler echocardiography in estimating severity of valvular stenosis. JAMA 1986;255:3145–3151. 244. Wisenbaugh T, Nissen S, DeMaria AN. Mechanics of postextrasystolic potentiation in normal subjects and patients with valvular heart disease. Circulation 1986;74:10 –20. 248. DeMaria AN, Wisenbaugh TW. Identification and treatment of diastolic dysfunction: role of transmitral Doppler recordings. J Am Coll Cardiol 1987;9: 1106 –1107. 251. Grayburn PA, Handshoe R, Smith MD, Harrison MR, DeMaria AN. Quantitative assessment of the hemodynamic consequences of aortic regurgitation by means of continuous wave Doppler recordings. J Am Coll Cardiol 1987;10:135–141. 252. Grayburn PA, Nissen SE, Elion JL, Evans J, DeMaria AN. Quantitation of aortic regurgitation by computer analysis of digital subtraction angiography. J Am Coll Cardiol 1987;10:1122–1127. 253. Grayburn PA, Smith MD, Gurley JC, Booth DC, DeMaria AN. Effect of aortic regurgitation on the assessment of mitral valve orifice area by Doppler pressure half-time. Am J Cardiol 1987;60:322–326. 260. Nissen SE, Elion JL, Grayburn P, Booth DC, Wisenbaugh TW, DeMaria AN. Determination of left ventricular ejection fraction by computer densitometric analysis of digital subtraction angiography: experimental validation and correlation with area-length methods. Am J Cardiol 1987;59:675– 680. 263. Rovai D, Nissen SE, Elion JL, Smith MD, L’Abbate A, Kwan OL, DeMaria AN. Contrast echo washout curves application of basic principles of indicator dilution theory and calculation of ejection fraction. J Am Coll Cardiol 1987;10: 125–134. 273. Gillespie MN, Booth DC, Friedman BJ, Cunningham MR, Jay M, DeMaria AN. fMLP provokes coronary vasoconstriction and myocardial ischemia in rabbits. Am J Physiol 1988;254:H481–H486. 274. Grayburn PA, Smith MD, Harrison MR, Gurley JC, DeMaria AN. Pivotal role of aortic valve area calculation by the continuity equation for Doppler 222 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 assessment of aortic stenosis in patients with combined aortic stenosis and regurgitation. Am J Cardiol 1988;61:376 –381. 275. Gurley JC, Nissen SE, Booth DC, Harrison M, Grayburn P, Elion JL, DeMaria AN. Comparison of simultaneously performed digital and film-based angiography in the assessment of coronary artery disease. Circulation 1988; 78(6):1411–1420. 279. Harrison MR, Smith MD, Nissen SE, Grayburn PA, DeMaria AN. Use of exercise Doppler to evaluate cardiac drugs: effects of propranolol and verapamil on aortic blood flow velocity and acceleration. J Am Coll Cardiol 1988;1(5): 1002–1009. 289. DeMaria AN, Smith MD, Harrison MR. Clinical significance of in-vitro and in-vivo experimental findings regarding Doppler flow velocity recordings. J Am Coll Cardiol 1989;13:1682–1686. 291. Grines CL, Nissen SE, Booth DC, Branco MC, Gurley JC, Bennett KA, DeMaria AN, KAMIT Study Group. A new thrombolytic regimen for acute myocardial infarction using half dose tissue plasminogen activator with full dose streptokinase. J Am Coll Cardiol 1989;14(3):573–580. 293. Harrison MR, Clifton GD, Sublett KL, DeMaria AN. Effect of heart rate on Doppler indexes of systolic function in humans. J Am Coll Cardiol 1989;14(4): 929 –935. 296. Rovai D, L’Abbate A, Lombardi M, Nissen SE, Marzilli M, Distante A, Ferdeghini EM, DeMaria AN. Nonuniformity of the transmural distribution of coronary blood flow during the cardiac cycle. In-vivo documentation by contrast echocardiography. Circulation 1989;79(1):179 –187. 297. Smith MD, Kwan OL, Elion JL, McClure R, DeMaria AN. Left heart opacification with peripheral venous injection of a new saccharide echo contrast agent in dogs. J Am Coll Cardiol 1989;13:1622–1628. 299. Spain MG, Smith MD, Graybum PA, Harlamert EA, O’Brien M, DeMaria AN. Quantitative assessment of mitral regurgitation by Doppler color flow imaging: angiographic and hemodynamic correlations. J Am Coll Cardiol 1989; 13:585–590. 300. Wisenbaugh TW, Harlamert EA, DeMaria AN. Relation of left ventricular filling dynamics to alterations in load and compliance in patients with and without pressure-overload hypertrophy. Circulation 1989;1(1):101–106. 303Berk MR, Xie G, Kwan OL, Knapp C, Evans J, Kotchen T, Kotchen JM, DeMaria AN. Reduction of left ventricular preload by lower body negative pressure alters Doppler transmitral filling patterns. J Am Coll Cardiol 1990;16(6): 1387–1392. 308. Garrahy PG, Kwan OL, Booth DC, DeMaria AN. Assessment of abnormal systolic intraventricular flow patterns by Doppler imaging in patients with left ventricular dyssynergy. Circulation 1990;82:95–104. 310. Grines CL, Booth DC, Nissen SE, Gurley JC, Bennett KA, O’Connor WN, DeMaria AN. Mechanism of acute myocardial infarction in patients with prior coronary artery bypass grafting and therapeutic implications. Am J Cardiol 1990;65:1292–1296. 312. Gurley JC, Nissen SE, Elion JL, Booth DC, DeMaria AN. Determination of coronary flow reserve by digital angiography: validation of a practical method not requiring power injection or electrocardiographic gating. J Am Coll Cardiol 1990;16:190 –197. 313. Nissen SE, Grines CL, Gurley JC, Sublett KL, Haynie DP, Diaz C, Booth DC, DeMaria AN. Application of a new phased-array ultrasound imaging catheter in the assessment of vascular dimensions: in-vivo comparison to cineangiography. Circulation 1990;81:660 – 666. 320. Grines CL, Booth DC, Nissen SE, Gurley JC, Bennett KA, DeMaria AN. Acute effects of parenteral beta-blockade on regional ventricular function of infarct and noninfarct zones after reperfusion therapy in humans. J Am Coll Cardiol 1991;17:1382–1387. 323. Grines CL, Nissen ST, Booth DC, Gurley JC, Chelliah N, Wolf R, Blankenship J, Branco MC, Bennett K, DeMaria AN, Kentucky Acute Myocardial Infarction Trial (KAMIT) Group. A prospective, randomized trial comparing combination half-dose tissue-type plasminogen activator and streptokinase with full-dose tissue-type plasminogen activator. Circulation 1991;84:540 –549. 325. Harrison MR, Clifton GD, Pennell AT, DeMaria AN. Effect of heart rate on left ventricular diastolic transmitral flow velocity patterns assessed by Doppler echocardiography in normal subjects. Am J Cardiol 1991;67:622– 627. 329. Nissen SE, Gurley JC, Grines CL, Booth DC, McClure R, Berk M, Fischer C, DeMaria AN. Intravascular ultrasound assessment of lumen size and wall morphology in normal subjects and patients with coronary artery disease. Circulation 1991;84:1087–1099. 330. Smith MD, Harrison MR, Pinton R, Kandil H, Kwan OL, DeMaria AN. Regurgitant jet size by transesophageal compared with transthoracic Doppler flow imaging. Circulation 1991;83:79 – 86. 338. King DL, Harrison MR, King D Jr, Gopal AS, Kwan OL, DeMaria AN. Ultrasound beam orientation during standard two-dimensional imaging: assessment by three-dimensional echocardiography. J Am Soc Echocardiogr 1992;5(6): 569 –576. 339. King DL, Harrison MR, King D Jr, Gopal AS, Martin R, DeMaria AN. Improved reproducibility of left atrial and left ventricular measurements by guided three-dimensional echocardiography. J Am Coll Cardiol 1992;20(5): 1238 –1245. 342. Rovai D, Ghelardini G, Lombardi M, Trivella MG, Nevola E, Taddei L, Michelassi C, Distante A, DeMaria AN. Myocardial washout of sonicated iopamidol reflects coronary blood flow in the absence of autoregulation. J Am Coll Cardiol 1992;20(6):1417–1424. JANUARY 15, 2005 355. Brotheron T, Pollard T, Simpson P, DeMaria AN. Classifying tissue and structure in echocardiograms: hierarchy of fuzzy-logic based neural networks automate analysis. Engineering Med Bio 1994;13(5):754 –759. 359. Xie GY, Berk MR, Smith MD, DeMaria AN. A simplified method for determining regurgitant fraction by Doppler echocardiography in patients with aortic regurgitation. J Am Coll Cardiol 1994;24(4):l041–1045. 360. Xie GY, Berk MR, Smith MD, Gurley JC, DeMaria AN. Prognostic value of Doppler transmitral flow patterns in patients with congestive heart failure. J Am Coll Cardiol 1994;24(I):132–139. 368. Nozaki S, DeMaria AN, Helmer GA, Hammond HK. Detection of regional left ventricular dysfunction in early pacing-induced heart failure using ultrasonic integrated backscatter. Circulation 1995;92:2676 –2682. 378. DeMaria AN, Lee TH, Leon DF, Ullyot DJ, Wolk MJ, Mills PS, Fay SC, Brown JH, Flatau CN, Bodycombe DP. Effect of managed care on cardiovascular specialists: involvement, attitudes and practice adaptations. J Am Coll Cardiol 1996;28(7):1884 –1895. 379. Giordano F, Ping P, McKirnan MD, Nozaki S, DeMaria AN, Dillman W, Mathieu-Costello O, Hammond HK. Intracoronary gene transfer of fibroblast growth factor-5 increases blood flow and contractile function in an ischemic region of the heart. Nature Med 1996;2(5):534 –539. 382. Kimura BJ, Russo RJ, Bhargava V, McDaniel M, Peterson KL, DeMaria AN. Atheroma morphology and distribution in proximal left anterior descending coronary artery: in vivo observations. J Am Coll Cardiol 1996;27(4):825– 831. 386. Yamagishi M, Nissen SE, Booth DC, Gurley JC, Koyama J, Kawano S, DeMaria AN. Coronary reactivity to nitroglyerin: intravascular ultrasound evidence for the importance of plaque distribution. J Am Coll Cardiol 1996;25:224 – 230. 398. Ohmori K, Cotter B, Kwan OL, Mizushige K, DeMaria AN. Relation of contrast echo intensity and flow velocity to the amplification of contrast opacification produced by intermittent ultrasound transmission. Am Heart J 1997; 134(6):1066 –1074. 406. Galiuto L, Ignone G, DeMaria AN. Heterogeneity of contraction and relaxation velocities of normal human myocardium. A quantitative study using pulsed wave tissue Doppler echocardiography. Am J Cardiol 1998;81(5):609 – 614. 415. Yasu T, Schmid-Schonbein GW, Cotter B, DeMaria AN. Flow dynamics of QW7437, a new dodecafluoropentane ultrasound contrast agent, in the microcirculation: microvascular mechanisms for persistent tissue echo enhancement. J Am Coll Cardiol 1999;34(2):578 –586. 423. Kimura BJ, Scott R, Willis CL, DeMaria AN. Accuracy and cost-effectiveness of single-view echocardiographic screening for suspected mitral valve prolapse. Am J Med 2000;108:331–333. 424. Masugata H, Cotter B, Peters B, Ohmori K, Mizushige K, DeMaria AN. Assessment of coronary stenosis severity and transmural perfusion gradient by myocardial contrast echocardiography: comparison of gray scale B-mode with power Doppler imaging. Circulation 2000;102:1427–1433. 432. Kimura BJ, Bocchicchio M, Willis CL, DeMaria AN. Screening cardiac ultrasonographic examination in patients with suspected cardiac disease in the emergency department. Am Heart J 2001;142(2):324 –330. 434. Lafitte S, Masugata H, Peters B, Togni M, Strachan M, Kwan OL, DeMaria AN. Comparative value of dobutamine and adenosine stress in the detection of coronary stenosis with myocardial contrast echocardiography. Circulation 2001; 103:2724 –2730. 435. Laffite S, Masugata H, Peters B, Togni M, Strachan M, Yao B, Kwan OL, DeMaria AN. Accuracy and reproducibility of coronary flow rate assessment by real-time contrast echocardiography: in vitro and in vivo studies. J Am Soc Echocardiogr 2001;14:1010 –1019. 436. Leistad E, Ohmori K, Peterson TA, Christensen G, DeMaria AN. Quantitative assessment of myocardial perfusion during graded coronary artery stenoses by intravenous myocardial contrast echocardiography. J Am Coll Cardiol 2001; 37:624 – 631. 437. Mahmud E, Raisinghani A, Keramati S, Auger W, Blanchard DG, DeMaria AN. Dilation of the coronary sinus on echocardiogram: prevalence and signifi- cance in patients with chronic pulmonary hypertension. J Am Soc Echocardiogr 2001;14:44 – 49. 438. Maisel AS, Koon J, Krishnaswamy P, Kasenegra R, Clopton P, Gardetto N, Morrissey R, DeMaria A. Utility of B-natriuretic peptide (BNP) as a rapid, point of care test for screening patients undergoing echocardiography for left ventricular dysfunction. Am Heart J 2001;141:367–374. 440. Masugata H, Lafitte S, Peters B, Strachan GM, DeMaria AN. Comparison of real-time and intermittent triggered myocardial contrast echocardiography for quantification of coronary stenosis severity and transmural perfusion gradient. Circulation 2001;104:1550 –1556. 442. Masugata H, Peters B, Lafitte S, Strachan GM, Ohmori K, DeMaria AN. Quantitative assessment of myocardial perfusion during graded coronary stenosis by rReal-time myocardial cContrast echo refilling curves. J Am Coll Cardiol 2001;37: 262–269. 443. Ohmori K, Cotter B, Leistad E, Bhargava V, Wolf PL, Mizushige K, DeMaria AN. Assessment of myocardial reperfusion by intravenous myocardial contrast echocardiography: analysis of the intensity and texture of opacification. Circulation 2001;103:2021–2027. 458. Kimura BJ, Amundson SA, Willis CL, Gilpin EA, DeMaria AN. Usefulness of a hand-held ultrasound device for bedside examination of left ventricular function. Am J Cardiol 2002;90:1038 –1039. 460. Lafitte S, Higashiyama A, Masugata H, Peters B, Strachan M, Kwan OL, DeMaria AN. Contrast echocardiography can assess risk area and infarct size during coronary occlusion and reperfusion: experimental validation. J Am Coll Cardiol 2002;39(9):1546 –1554. 461. Lubien E, DeMaria AN, Krishnaswamy P, Clopton P, Koon J, Kazanegra R, Gardetto N, Wanner E, Maisel AS. Utility of B-natriuretic peptide in detecting diastolic dysfunction: comparison with Doppler velocity recordings. Circulation 2002;105:595– 601. 464. Raisinghani A, Mahmud E, Sadeghi M, Peters B, Strachan GM, Huynh T, Blanchard D, DeMaria AN. Paradoxical inferior-posterior wall systolic expansion in patients with end-stage liver disease. Am J Cardiol 2002;89:626 – 629. 477. Kimura BJ, DeMaria AN. Time requirements of the standard echocardiogram: implications regarding limited studies. J Am Soc Echocardiogr 2003; 16(10):1015–1018. 478. Kimura BJ, Fowler SJ, Nguyen DT, Amundson SA, DeMaria AN. Detection of early carotid arterial atherosclerosis by briefly trained physicians using a hand-held ultrasound device. Am J Cardiol 2003;92(2):239 –240. 480. Kunichika H, Peters B, Cotter B, Masugata H, Kunichika N, Wolf PL, DeMaria AN. Visualization of risk-area myocardium as a high-intensity, hyperenhanced “hot spot” by myocardial contrast echocardiography following coronary reperfusion. J Am Coll Cardiol 2003;42(3):552–557. 482. Masugata H, Fujita N, Kondo I, Peters B, Ohmori K, Mizushige K, Kohno M, DeMaria AN. Assessment of right ventricular perfusion after right coronary artery occlusion. J Am Coll Cardiol 2003;41(10):1823–1830. 484. Raisinghani A, Wei K, Crouse LJ, Villanueva F, Feigenbaum H, Schiller NB, Weiss J, Zaqvi TZ, Siegel R, Monaghan M, Goldman JH, DeMaria AN. Myocardial contrast echocardiography with triggered ultrasound does not cause ventricular premature beats: evidence from PB127 studies. J Am Soc Echocardiogr 2003;10:1037–1042. 485. Wei K, Crouse L, Weiss J, Villanueva F, Schiller NB, Naqvi TZ, Siegel R, Monaghan M, Goldman J, Aggarwal P, Feigenbaum H, DeMaria A. Comparison of usefulness of dipyridamole stress myocardial contrast echocardiography to technetium-99m sestamibi single-photon emission computed tomography for detection of coronary artery disease (PB127 multicenter Phase 2 trial Results). Am J Cardiol 2003;91:1293–1298. 488. Nissen SE, Tuzeu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA, Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN. Effect of intensive compared with moderate lipid-lowering therapy on progression of coronary atherosclerosis—a randomized controlled trial. JAMA 2004;291(9): 1071–1080. INTERVIEW/ANTHONY NICHOLAS DEMARIA 223 Relation of Impaired Thrombolysis In Myocardial Infarction Myocardial Perfusion Grades to Residual Thrombus Following the Restoration of Epicardial Patency in ST-Elevation Myocardial Infarction Ajay J. Kirtane, MD, Aaron Weisbord, MD, Dimitrios Karmpaliotis, MD, Sabina A. Murphy, MPH, Robert P. Giugliano, MD, SM, Christopher P. Cannon, MD, Elliott M. Antman, MD, E. Magnus Ohman, MD, Matthew T. Roe, MD, MHS, Eugene Braunwald, MD, and C. Michael Gibson, MS, MD, for the TIMI Study Group Clinical and angiographic data were analyzed from 929 patients who had ST-elevation myocardial infarction and open epicardial arteries after fibrinolytic therapy. Residual angiographically evident thrombus was associated with more frequent Thrombolysis In Myocardial Infarction (TIMI) grade 2 flow (33.6% vs 26.8%, p ‫ ؍‬0.03), higher corrected TIMI frame counts (34 vs 31 frames, p ‫ ؍‬0.0003), and lower TIMI myocardial perfusion grades (43.0% vs 32.0% TIMI myocardial perfusion grades 0/1, p ‫ ؍‬0.001) among all patients and among patients who had TIMI grade 3 flow (33.5% vs 26.0% TIMI myocardial perfusion grades 0/1, p ‫ ؍‬0.043). In multivariate analyses, angiographically evident thrombus was associated with higher corrected TIMI frame counts and worsened myocardial perfusion independent of clinical and angiographic covariates, including TIMI grade 3 flow. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:224 –227) atients who have acute coronary syndromes may have nonocclusive and angiographically evident P thrombus (AET) in the culprit coronary artery despite reperfusion therapy. In patients who have non–STelevation acute coronary syndromes, visualization of thrombus by angioscopy and by angiography have been associated with high levels of troponin, and AET has been associated with impaired myocardial perfusion.1,2 Residual thrombus after coronary stenting has been associated with recurrent ischemic events.3 One potential pathophysiologic link between AET and adFrom the Department of Medicine, Beth Israel Deaconess Medical Center and the TIMI Study Group, Brigham & Women’s Hospital, Harvard Medical School, Cambridge, Massachusetts; the University of North Carolina, Chapel Hill, North Carolina; and the Duke Clinical Research Institute, Durham, North Carolina. This study was supported in part by grants from Millennium Pharmaceuticals, Cambridge, Massachusetts, and Schering-Plough Research Institute, Kennilworth, New Jersey; (Integrilin and Tenecteplase in Acute Myocardial Infarction); Merck and Co., Blue Bell, Pennsylvania (FASTER); Aventis Pharma, Antony, France (ENTIRE Thrombolysis In Myocardial Infarction 23); Centocor and Eli Lilly Inc., Malvern, Pennsylvania; and Indianapolis, Indiana (Thrombolysis In Myocardial Infarction 14). Dr. Gibson address is: 350 Longwood Avenue, First Floor, Boston, Massachusetts 02115. E-mail: [email protected]. Manuscript received June 8, 2004; revised manuscript received and accepted September 7, 2004. 224 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 verse outcomes is microvascular dysfunction secondary to distal embolization and the release of vasoactive substances that may arise in patients who have a greater thrombus burden.4 We hypothesized that the presence of AET in patent arteries after fibrinolytic therapy in patients who have ST-elevation myocardial infarction would be associated with impaired myocardial perfusion, independent of epicardial flow. ••• Clinical and angiographic data were drawn from patients with ST-elevation myocardial infarction who had been enrolled in the Thrombolysis In Myocardial Infarction (TIMI) 14,5 Integrilin and Tenecteplase in Acute Myocardial Infarction,6 Enoxaparin and Thrombolysis Reperfusion for Acute Myocardial Infarction Treatment [ENTIRE] TIMI 23,7 and Fibrinolytics and Aggrastat with ST segment resolution [FASTER]8 trials of fibrinolytic therapy for ST-elevation myocardial infarction. Angiography was performed at 60 minutes after fibrinolytic administration. Angiograms were analyzed off-line in a core laboratory (PERFUSE Core Laboratory). Analysis was restricted to patients who had been enrolled in these trials in whom the TIMI myocardial perfusion grade (TMPG) could be ascertained and in whom the culprit artery was angiographically patent as defined by TIMI flow grade 2 or 3 at 60 minutes after administration of fibrinolytic therapy (929 patients of 1,268 patients who had open arteries). Quantitative coronary angiography and assessments of corrected TIMI frame count9 and TMPG10 were performed as previously described. All analyses were performed with STATA 7.0.11 All continuous variables are reported as medians with 25% and 75% interquartile values unless otherwise specified. Nonparametric Wilcoxon’s rank sum test and Student’s t test were used as appropriate. Chisquare test was used for analysis of categorical variables. Multivariate linear and logistic regression models were constructed using impaired epicardial flow and myocardial perfusion by corrected TIMI frame count and TMPG, respectively, as outcome variables, and incorporating AET, potential confounders (p ПЅ0.10 in univariate analyses), other known covariates of corrected TIMI frame count, and TMPG in the dataset, and treatment with a glycoprotein IIb/IIIa inhibitor. 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.007 lesion (Table 1). The minimal luminal diameter of the culprit lesion did Thrombus not differ significantly between paNo. of tients who had AET and those who Characteristic Patients No Yes p Value did not. Among patients who had Left anterior descending AET, there was a larger percentage culprit vessel of TIMI grade 2 flow (33.6% vs П© 325 39.3% 27.9% ПЅ0.001 26.8%, p П 0.03; Table 2) and corПЄ 604 60.7% 72.1% No. of narrowed rected TIMI frame counts were coronary arteries higher (slower). When analysis was 1 486 51.2% 56.2% 0.34 restricted to the 656 patients who had 2 288 32.5% 29.8% TIMI grade 3 flow, corrected TIMI 3 141 16.3% 14.0% Lesion length (mm) 891 9 (7–13) 11 (7–15) 0.004 frame counts remained higher among Reference segment 894 2.7 (2.3–3.2) 2.9 (2.5–3.4) 0.0002 patients who had AET (median (mm) [25%, 75%] of 29 frames [22, 34] vs Minimum luminal 894 1.0 (0.8–1.3) 1.0 (0.7–1.3) 0.67 26 frames, p П 0.02). diameter (mm) The presence of AET in the inStenosis (%) 894 62.0 (52.5–71.6) 64.8 (54.2–73.1) 0.009 farct-related artery after fibrinolytic Values are percentages or medians (25% to 75% interquartile ranges). therapy was associated with impaired myocardial perfusion as assessed by TMPG (43.0% vs 32.0% TMPG 0/1, p П 0.001; Figure 1). TABLE 2 Angiographic Characteristics of Flow and Perfusion Similarly, any abnormality in TMPG Thrombus (grades 0/1/2) was observed more frequently among patients who had No. of No Yes AET (46.7% vs 34.6%, p ПЅ0.001). Patients (n П 566) (n П 341) Chi-square p Value When analysis was further restricted TIMI frame count at 31 (22–41) 34 (25–50) 0.0003 to patients who had TIMI grade 3 60 min* flow, the association between AET TIMI flow grade at 60 min and impaired myocardial perfusion 2 273 26.8% 33.6% 4.9 0.03 remained, with TMPG 0/1 observed 3 656 73.1% 63.4% in 33.5% of patients who had AET TMPG at 60 min versus 26.0% of patients who did not 0 208 21.5% 23.9% 18.3 ПЅ0.001 1 128 10.5% 19.1% have AET (p П 0.043; Figure 2). 2 28 2.6% 3.7% AET remained independently as3 565 65.4% 53.3% sociated with a higher corrected 0/1 336 32.0% 43.0% 11.5 0.001 TIMI frame count (p П 0.01) after 2/3 593 68.0% 57.0% adjustment for TIMI risk index, time 0/1/2 364 34.6% 46.7% 13.5 ПЅ0.001 3 565 65.4% 53.3% to treatment, gender, history of hypercholesterolemia, culprit location *Values are medians (25% to 75% interquartile ranges). in the left anterior descending coronary artery, reference segment diameter, percent stenosis of the culprit AET 60 minutes after fibrinolytic administration lesion, lesion length, treatment with a glycoprotein was detected in 37.8% of patients (351 of 929). A IIb/IIIa inhibitor, and TIMI grade 3 flow. In a similar history of hypercholesterolemia was more frequent model restricted to patients who had TIMI grade 3 among patients who had AET (27.6% vs 21.5%, p П flow, AET was independently associated with a higher 0.03). There was a slightly larger percentage of men corrected TIMI frame count (p П 0.03). AET was also independently associated with imwho had AET among the patients (82.9% vs 77.8%, p П 0.06). There were no significant associations be- paired TMPG after adjustment for clinical and angiotween AET and other baseline characteristics, includ- graphic covariates, including TIMI risk index, treatment ing age, active smoking, previous aspirin use, history with a glycoprotein IIb/IIIa inhibitor, and TIMI grade 3 of hypertension, history of diabetes, family history of flow (odds ratio 1.5 for TMPG 0/1, 95% confidence coronary artery disease, previous coronary artery dis- interval 1.1 to 2.0, p П 0.004; odds ratio 1.6 for TMPG ease (angina, revascularization, or myocardial infarc- 0/1/2, 95% confidence interval 1.2 to 2.1, p П 0.002). tion), time to treatment, white blood cell counts (n П When reference segment diameter, percent stenosis of 584), C-reactive protein levels (n П 415), or TIMI risk the culprit lesion, lesion length, and corrected TIMI index [heart rate П« (age/10)2/systolic blood pressure]. frame count were added to the model, AET remained Angiographic characteristics associated with AET independently associated with impaired TMPG (odds were noninvolvement of the culprit left anterior de- ratio 1.4 for TMPG 0/1, 95% confidence interval 1.0 to scending artery, longer lesions, reference segment di- 1.9, p П 0.026; odds ratio 1.5 for TMPG 0/1/2, 95% ameters, and percent diameter stenosis of the culprit confidence interval 1.1 to 2.0, p П 0.010). TABLE 1 Angiographic Lesion Characteristics BRIEF REPORTS 225 In patients who have ST-elevation myocardial infarction treated with fibrinolytic therapy or percutaneous coronary intervention, thrombus formation can result in distal embolization of microthrombi and platelet/fibrin aggregates into the microcirculation.13 The role of thrombus in distal embolization during percutaneous coronary intervention has been demonstrated by Fukuda et al,14 who described an association between the presence of thrombus as assessed by intravascular ultrasound and angiographically observed distal embolization during percutaneous coronary intervention. However, even before percutaneous coronary inFIGURE 1. Association of AET with TMPG among all patent arteries. Percentages of patervention or fibrinolytic therapy, mitients who had (A) TMPG 0/1 or (B) TMPG 0/1/2. croembolization may occur among patients who have acute coronary syndromes.15 Thus, the finding of an association between residual AET and impaired TMPG suggests that residual AET after fibrinolytic therapy may be a surrogate of a larger preexisting thrombotic burden or a more organized, thromboresistant clot that may result in worsening microvascular dysfunction. Although distal embolization may be 1 mechanistic explanation for our observations, residual intracoronary thrombus after fibrinolytic therapy may be a marker of a more intense prothrombotic and inflammatory process with resultant microcirculatory dysfunction. Fibrinolytic therapy results in thrombin generation and platelet activation in patients FIGURE 2. Association of AET with TMPG among patients who had TIMI flow grade 3. who have ST-elevation myocardial Percentages of patients who had (A) TMPG 0/1 or (B) TMPG 0/1/2. infarction.16,17 These 2 processes can contribute to thrombus propagation ••• and to the release of procoagulant and vasoconstrictor AET after fibrinolytic therapy for ST-elevation substances that may further perpetuate endothelial myocardial infarction has been associated with lower dysfunction and microvascular dysfunction.4 In anirates of epicardial patency and slower epicardial mal models, arterial thrombus has been associated flow.12 This study extends previous observations and with impaired release of endothelium-dependent redemonstrates that, even among patients who have had laxing factor and with higher indexes of arterial resissuccessful restoration of epicardial patency after fi- tance.18 Although we did not observe any associations brinolytic therapy, residual AET is associated with between AET and white blood cell counts or C-reacimpaired myocardial perfusion in addition to slower tive protein levels, this analysis was restricted to paepicardial flow. These associations were independent tients who had patent epicardial arteries Х…60 minutes of severity of residual stenosis and other characteris- of fibrinolytic administration, a subgroup previously tics of the culprit lesion. Moreover, after adjustment demonstrated to have a less intense systemic inflamfor TIMI grade 3 flow and corrected TIMI frame count matory response than patients who had occluded arin addition to baseline and lesion characteristics, AET teries.19 remained independently associated with impaired The combined effects of microembolization, inmyocardial perfusion. These findings support the hy- flammation, and vasoconstriction may contribute to pothesis that AET is associated with increased micro- the observed associations across thrombus, impaired vascular obstruction/vasoconstriction, perhaps related corrected TIMI frame count, and TMPG. An increase to greater distal embolization, even among patients in “downstream” microvascular resistance may also who have achieved successful epicardial reperfusion. explain the independent association between AET and 226 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 higher (slower) corrected TIMI frame count, even when adjusted for lesion characteristics that affect epicardial flow. Insofar as AET is associated with impaired myocardial perfusion, strategies aimed at decreasing the risks of distal embolization, vasoconstriction, and impaired myocardial perfusion warrant further evaluation. Although the negative preliminary results of the Enhanced Myocardial Efficacy and Recovery by Aspiration of Liberalized Debris [EMERALD] trial of a distal protection device in ST-elevation myocardial infarction may temper enthusiasm,20 distal protection devices do not offer protection against embolization into proximal arterial branches and do not decrease vasoconstrictor release. Further, embolization can occur with deployment of these devices, particularly if passing the device beyond the culprit lesion is difficult. 1. Okamatsu K, Takano M, Sakai S, Ishibashi F, Uemura R, Takano T, Mizuno K. Elevated troponin T levels and lesion characteristics in non-ST-elevation acute coronary syndromes. Circulation 2004;109:465– 470. 2. Wong GC, Morrow DA, Murphy S, Kraimer N, Pai R, James D, Robertson DH, Demopoulos LA, DiBattiste P, Cannon CP, Gibson CM. Elevations in troponin T and I are associated with abnormal tissue level perfusion: a TACTICSTIMI 18 substudy. Treat Angina with Aggrastat and Determine Cost of Therapy with an Invasive or Conservative Strategy-Thrombolysis in Myocardial Infarction. Circulation 2002;106:202–207. 3. Ross MJ, Herrmann HC, Moliterno DJ, Blankenship JC, Demopoulos L, DiBattiste PM, Ellis SG, Ghazzal Z, Martin JL, White J, Topol EJ. Angiographic variables predict increased risk for adverse ischemic events after coronary stenting with glycoprotein IIb/IIIa inhibition: results from the TARGET trial. J Am Coll Cardiol 2003;42:981–988. 4. Willerson JT, Golino P, Eidt J, Campbell WB, Buja LM. Specific platelet mediators and unstable coronary artery lesions. Experimental evidence and potential clinical implications. Circulation 1989;80:198 –205. 5. Antman EM, Giugliano RP, Gibson CM, McCabe CH, Coussement P, Kleiman NS, Vahanian A, Adgey AA, Menown I, Rupprecht HJ, et al. Abciximab facilitates the rate and extent of thrombolysis: results of the Thrombolysis In Myocardial Infarction (TIMI) 14 trial. The TIMI 14 Investigators. Circulation 1999;99:2720 –2732. 6. Giugliano RP, Roe MT, Harrington RA, Gibson CM, Zeymer U, Van de Werf F, Baran KW, Hobbach HP, Woodlief LH, Hannan KL, et al. Combination reperfusion therapy with eptifibatide and reduced-dose tenecteplase for STelevation myocardial infarction: results of the Integrilin and Tenecteplase in Acute Myocardial Infarction (INTEGRITI) Phase II Angiographic Trial. J Am Coll Cardiol 2003;41:1251–1260. 7. Antman EM, Louwerenburg HW, Baars HF, Wesdorp JC, Hamer B, Bassand JP, Bigonzi F, Pisapia G, Gibson CM, Heidbuchel H, et al. Enoxaparin as adjunctive antithrombin therapy for ST-elevation myocardial infarction: results of the ENTIRE-Thrombolysis in Myocardial Infarction (TIMI) 23 Trial. Circulation 2002;105:1642–1649. 8. Ohman EM, Van de Werf F, Antman EM, Califf RM, Lemos JAD, Gibson CM, Oliverio RL, Harrelson L, McCabe C, DiBattiste P, et al. Tenecteplase and tirofiban in ST-segment elevation acute myocardial infarction: results of a randomized trial. Am Heart J 2005; in press. 9. Gibson CM, Cannon CP, Daley WL, Dodge JT Jr, Alexander B Jr, Marble SJ, McCabe CH, Raymond L, Fortin T, Poole WK, Braunwald E. TIMI frame count: a quantitative method of assessing coronary artery flow. Circulation 1996;93: 879 – 888. 10. Gibson CM, Cannon CP, Murphy SA, Ryan KA, Mesley R, Marble SJ, McCabe CH, Van De Werf F, Braunwald E. Relationship of TIMI myocardial perfusion grade to mortality after administration of thrombolytic drugs. Circulation 2000;101:125–130. 11. STATA Statistical Software. College Station, TX: STATA Corp., 2000. 12. Gibson CM, Murphy S, Menown IB, Sequeira RF, Greene R, Van de Werf F, Schweiger MJ, Ghali M, Frey MJ, Ryan KA, et al. Determinants of coronary blood flow after thrombolytic administration. TIMI Study Group. Thrombolysis in Myocardial Infarction. J Am Coll Cardiol 1999;34:1403–1412. 13. Saber RS, Edwards WD, Bailey KR, McGovern TW, Schwartz RS, Holmes DR Jr. Coronary embolization after balloon angioplasty or thrombolytic therapy: an autopsy study of 32 cases. J Am Coll Cardiol 1993;22:1283–1288. 14. Fukuda D, Tanaka A, Shimada K, Nishida Y, Kawarabayashi T, Yoshikawa J. Predicting angiographic distal embolization following percutaneous coronary intervention in patients with acute myocardial infarction. Am J Cardiol 2003;91: 403– 407. 15. Heusch G, Schulz R, Baumgart D, Haude M, Erbel R. Coronary microembolization. Prog Cardiovasc Dis 2001;44:217–230. 16. Merlini PA, Bauer KA, Oltrona L, Ardissino D, Spinola A, Cattaneo M, Broccolino M, Mannucci PM, Rosenberg RD. Thrombin generation and activity during thrombolysis and concomitant heparin therapy in patients with acute myocardial infarction. J Am Coll Cardiol 1995;25:203–209. 17. Gurbel PA, Serebruany VL, Shustov AR, Bahr RD, Carpo C, Ohman EM, Topol EJ. Effects of reteplase and alteplase on platelet aggregation and major receptor expression during the first 24 hours of acute myocardial infarction treatment. GUSTO-III Investigators. Global Use of Strategies to Open Occluded Coronary Arteries. J Am Coll Cardiol 1998;31:1466 –1473. 18. Reil TD, Moore WS, Kashyap VS, Nene SS, Gelabert HA, QuinonesBaldrich WJ. The effects of thrombus, thrombectomy and thrombolysis on endothelial function. Eur J Vasc Endovasc Surg 2000;19:162–168. 19. Barron HV, Cannon CP, Murphy SA, Braunwald E, Gibson CM. Association between white blood cell count, epicardial blood flow, myocardial perfusion, and clinical outcomes in the setting of acute myocardial infarction: a Thrombolysis In Myocardial Infarction 10 substudy. Circulation 2000;102:2329 –2334. 20. Stone GW, Webb J, Cox DA, Brodie BR, Quershi M, Dulas D, Kalynch A, Turco M, Schultheiss HP, Rutherford B, et al. Primary angioplasty in acute myocardial infarction with distal protection of the microcirculation: principal results from the prospective, randomized EMERALD trial. J Am Coll Cardiol 2004;43(5 suppl A):285A. BRIEF REPORTS 227 Angiography and Revascularization in Patients With Heart Failure Following Fibrinolytic Therapy for ST-Elevation Acute Myocardial Infarction Amir Kashani, MS, MD, C. Michael Gibson, MS, MD, Sabina A. Murphy, MPH, Marc S. Sabatine, MD, MPH, David A. Morrow, MD, MPH, Elliott M. Antman, MD, and Robert P. Giugliano, MD, SM We evaluated the use of coronary angiography and clinical outcomes among patients who had heart failure and were enrolled in the Intravenous Novel Plasminogen Activator (NPA) for the Treatment of Infarcting Myocardium Early study, a large international trial of fibrinolytic therapy in ST-elevation myocardial infarction. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:228 –233) he Second Intravenous Novel Plasminogen Activator (NPA) for the Treatment of Infarcting MyoT cardium Early trial included patients Х†18 years of 1 age who presented within 6 hours of symptom onset with ST elevation in 2 contiguous leads or new left bundle branch block. Patients in cardiac shock (Killip’s class IV) were excluded. Other major exclusions were increased bleeding risk, previous stroke, systolic blood pressure Х†180 mm Hg, or diastolic blood pressure Х†110 mm Hg. A total of 15,078 patients was enrolled from 1997 to 1998 in 855 hospitals in 35 countries. ••• Patients from the trial were categorized into 1 of 4 mutually exclusive groups based on the development of heart failure (HF) and use of angiography during the index hospitalization: group 1, no HF and angiogram; group 2, no HF and no angiogram; group 3, HF and angiogram; and group 4, HF and no angiogram. Twelve patients did not have angiographic data and were excluded from analyses in these groups. A secondary analysis explored the severity of HF categorized in 4 groups2: (1) shock, defined as Killip’s class IV, systolic blood pressure ПЅ90 mm Hg despite adequate ventricular filling or systolic blood pressure Х†90 mm Hg that required inotropic agents or intra-aortic balloon pump after volume expansion; (2) severe HF, defined as HF (not satisfying shock criteria) that required new use of, or increase in the dose of, a diuretic and Х†1 objective signs of HF (i.e., radiographic evidence of pulmonary edema, rales on examination that did not From the Rochester General Hospital, Rochester, New York; and the TIMI Study Group, Brigham & Women’s Hospital, Boston, Massachusetts. Dr. Giugliano’s address is: TIMI Study Group, 350 Longwood Avenue, 1st Floor Offices, Boston, Massachusetts 02115. E-mail: [email protected]. Manuscript received August 5, 2004; revised manuscript received and accepted August 26, 2004. 228 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 resolve with coughing, or mean pulmonary capillary wedge pressure Х†18 mm Hg and cardiac index Х…2.2 L/min/m2); (3) mild HF, defined as not meeting HF criteria 1 or 2 and objective signs of HF at presentation that resolved with therapy before discharge, required diuretics after randomization, or had objective findings of HF (but not the 2 together); and (4) no HF, defined as no evidence of HF on presentation or during index hospitalization. Data were obtained from trial case record forms. The log-rank test was used to compare cumulative mortality rates through 1 year. Multivariate logistic models were constructed using backward selection, with statistical significance set at a p value ПЅ0.05. The z score was used to identify variables with the greatest predictive power. A multivariate model to evaluate the association between use of angiography and 1-year mortality rate was developed. The Thrombolysis In Myocardial Infarction risk score for ST-elevation myocardial infarction (STEMI), a simple bedside tool that produces an integer score that provides risk stratification with respect to mortality rate in patients who have STEMI,3 was used to adjust for differences in baseline risk. A propensity score to predict the use of angiography was developed using baseline patient and hospital characteristics associated with performance of an angiogram before discharge. Thrombolysis In Myocardial Infarction risk score, angiographic propensity score, and baseline hospital characteristics were used in a stepwise fashion to calculate the adjusted hazard decrease in 1-year mortality rate associated with the use of angiography in the population stratified by the presence/absence of HF during the index admission. Of 15,078 patients enrolled, 3,420 (23%) developed HF during the index admission after fibrinolysis, and 5,667 (38%) underwent angiography before discharge. Among patients who had HF during the index admission, 75% developed HF on hospital days 0 to 1, and the remaining 25% developed HF between day 2 and discharge (median 9 days). Baseline characteristics for the 4 groups of patients who were stratified by the presence of HF and use of angiography during the index admission are presented in Tables 1 (patient characteristics) and 2 (hospital characteristics). Patients who underwent angiography before discharge were more often younger, were men, had previous percutaneous cor0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.092 TABLE 1 Baseline Characteristics HF During Index Admission (n П 3,420, 23%) No HF During Index Admission (n П 11,658, 77%) Angiogram Variable Age Х†65 yrs Women Weight ПЅ67 kg Systemic hypertension Diabetes mellitus Current smoker Previous PCI Previous angina pectoris Previous myocardial infarction Previous heart failure Previous renal insufficiency Previous medications вђ¤ Blockers Diuretics Angiotensin-converting enzyme inhibitors/ angiotensin receptor blockers Aspirin Findings at presentation Creatinine Пѕ1.5 mg/dl Atrial fibrillation Anterior myocardial infarction Peak creatine kinase П® SD Heart rate Пѕ100 beats/min Systolic blood pressure ПЅ100 mm Hg Angiogram Present (n П 1,010) Absent (n П 2,410) All (n П 3,420) Present (n П 4,657) Absent (n П 6,989) All* (n П 11,646) 50% 27% 19% 36% 20% 43% 8% 24% 21% 5% 2% 64% 33% 25% 37% 17% 37% 3% 32% 24% 11% 2% 60% 31% 23% 37% 18% 38% 4% 30% 23% 9% 2% 33% 20% 17% 30% 13% 46% 7% 17% 15% 1% 0.9% 39% 25% 21% 28% 13% 48% 3% 20% 14% 2% 0.8% 37% 23% 19% 29% 13% 47% 5% 19% 14% 1% 0.8% ПЅ0.001 ПЅ0.001 ПЅ0.001 ПЅ0.001 ПЅ0.001 ПЅ0.001 0.67 ПЅ0.001 ПЅ0.001 ПЅ0.001 ПЅ0.001 19% 14% 16% 19% 20% 18% 19% 18% 17% 17% 8% 13% 13% 9% 12% 15% 8% 12% ПЅ0.001 ПЅ0.001 ПЅ0.001 24% 26% 25% 21% 17% 18% ПЅ0.001 8% 6% 53% 2,247 П® 2,152 13% 4% 13% 8% 53% 2,383 П® 2,131 14% 4% 11% 7% 53% 2,343 П® 2,138 13% 4% 5% 2% 39% 1,485 П® 1,563 6% 2% 6% 3% 39% 1,755 П® 1,651 6% 2% 5% 3% 39% 1,646 П® 1,621 6% 2% BRIEF REPORTS Data are percentage of patients, unless otherwise indicated. *Twelve patients did not have data regarding the use or nonuse of angiography and were excluded from analyses. †HF present versus absent. All 4-way (yes HF/yes angiogram, yes HF/no angiogram, no HF/yes angiogram, no HF/no angiogram) p values are ПЅ0.001 except peak creatine kinase П® SD (p П 0.001). PCI П percutaneous coronary intervention. p Value†Ͻ0.0001 ПЅ0.001 ПЅ0.001 ПЅ0.0001 ПЅ0.001 ПЅ0.001 229 TABLE 2 Hospital Characteristics Variable Geographic region North America Latin America Eastern Europe Western Europe Hospital size (no. of beds) ПЅ300 beds 300–700 beds Пѕ700 beds City (Пѕ100,000 population) hospital On-site angiography available 24-h PCI available On-site coronary bypass available Teaching hospital Lytic by cardiologist‡ HF During Index Admission (n П 3,420, 23%) No HF During Index Admission (n П 11,658, 77%) Angiogram Angiogram Present (n П 1,010) Absent (n П 2,410) All HF (n П 3,420) Present (n П 4,657) Absent (n П 6,989) All No HF* (n П 11,646) 31% 2% 8% 59% 12% 2% 27% 60% 18% 2% 21% 59% 27% 3% 7% 62% 14% 3% 26% 60% 20% 3% 19% 59% 25% 47% 29% 72% 20% 47% 33% 84% 22% 47% 31% 81% 25% 47% 28% 71% 23% 49% 29% 79% 24% 48% 28% 76% ПЅ0.001 65% 46% 52% 69% 47% 56% ПЅ0.0001 31% 32% 15% 21% 20% 24% 37% 34% 16% 23% 24% 27% ПЅ0.001 0.0001 68% 60% 65% 56% 66% 57% 68% 65% 64% 60% 66% 62% 0.84 ПЅ0.001 p Value†Ͻ0.001 0.001 *Twelve patients did not have data regarding the use or nonuse of angiography and were excluded from analyses. †HF present versus absent. All 4-way (yes HF/yes angiogram, yes HF/no angiogram, no HF/yes angiogram, no HF/no angiogram) p values ПЅ0.001 except teaching hospital (p П 0.002). ‡ Hospitals in which the cardiologist is the physician who most commonly administers fibrinolytic therapy. Abbreviations as in Table 1. TABLE 3 Multivariate Analyses Predicting Patients Who Underwent Angiography Variable On-site PCI availability 24 h (vs none) Day only (vs none) Geographic region Eastern Europe (vs Western Europe) North America (vs Western Europe) Latin America (vs Western Europe) Rural/local hospital (ПЅ100,000 population) Age ПЅ65 yrs Hospitals with lytic administered by cardiologist* Previous PCI No previous HF Hospital size Medium 300–700 beds (vs large 700 beds) Small 300 beds (vs large 700 beds) Previous hypertension Men Creatinine ПЅ1.5 mg/dl OR 95% CI z Score 5.0 3.2 4.5–5.6 2.8–3.5 28.6 19.3 0.2 0.18–0.23 2.2 1.9–2.4 0.6 0.5–0.8 2.1 1.9–2.4 23.7 13.9 4.0 14.4 1.6 1.6 1.4–1.7 1.4–1.7 10.2 9.9 2.1 2.4 1.7–2.5 1.8–3.2 7.2 5.9 1.3 1.1–1.4 4.5 1.2 1.0–1.4 2.4 1.2 1.2 1.3 1.1–1.3 1.1–1.3 1.1–1.6 4.0 3.8 3.3 Only variables with p values Х…0.001 are shown. The model was also adjusted for the following variables: teaching hospital (p П 0.002), previous aspirin use (p П 0.007), anterior myocardial infarction (p П 0.02), current smoking (p П 0.02), atrial fibrillation (p П 0.03), previous angina (p П 0.045), weight Пѕ67 kg (p П 0.08), and previous myocardial infarction (p П 0.08). *Hospitals in which the cardiologist is the physician who most commonly administers fibrinolytic therapy. CI П confidence interval; OR П odds ratio. Other abbreviation as in Table 1. onary intervention, did not have HF, and presented at North American hospitals with on-site interventional facilities (Table 3). 230 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 FIGURE 1. Rates of angiography and revascularization before discharge in patients who had HF and those who did not. The third pair of bars represents revascularization rates among patients who had an angiogram. A total of 3,538 patients (23%) was revascularized; among these patients, 2,969 (20%) previously had percutaneous coronary intervention only, 610 (4%) previously had coronary artery bypass grafting only, and 53 (0.4%) previously had both procedures. Patients who had HF were less likely to have undergone angiography (30% vs 40%) and revascularization (19% vs 25%) than patients who did not have HF (p ПЅ0.001 for the 2 groups). However, once an angiogram had been performed, rates of revascularization were similar (63% vs 62%, p П 0.93; Figure 1). Rates JANUARY 15, 2005 FIGURE 2. Geographic variation in rates of angiography and revascularization. of invasive procedures varied widely and depended on hospital characteristics (Table 2) and geographic location (Figure 2). Procedures were performed fourfold to 10-fold more frequently in the United States (angiography 79%, revascularization 59%) than in Canada and Europe. There was a greater difference in rates of angiography and revascularization among patients who had HF than among those who did not who had been enrolled in Europe compared with patients in North America. Mortality was fourfold to sevenfold higher at 1 year in patients who had HF than in those who did not (Figure 3). Across the HF spectrum, mortality rates at 1 year were lower among patients who underwent angiography before discharge: shock (41% vs 75%), severe HF (11% vs 21%), mild HF (6% vs 16%), and no HF (2% vs 7%). Use of angiography before discharge was associated with a lower mortality at 1 year in an unadjusted analysis among patients who did not have HF (hazard ratio 0.33, 95% confidence interval 0.27 to 0.40) and among those who had HF during the index admission (hazard ratio 0.50, 95% confidence interval 0.42 to 0.60). After adjustment for baseline risk with the Thrombolysis In Myocardial Infarction risk score for STEMI, a propensity score for undergoing angiography before discharge, and dif- ferences in hospital characteristics, use of angiography before discharge continued to demonstrate a strong and independent association with lower 1-year mortality, regardless of the presence of HF during the index admission (Table 4). Causes of death at 1 year in patients who had been stratified by use of angiography are listed in Table 5. Among patients who died, recurrent myocardial infarction and cardiac rupture were more frequent causes of death in patients who did not undergo angiography during the index admission. Twenty-three percent of patients developed HF during the index hospitalization in this recent, large, international trial of fibrinolytic therapy for STEMI. Despite a fourfold higher mortality rate, patients who had HF were less likely to undergo angiography and revascularization than were patients who did not have HF. This observation may indicate physicians’ reluctance to use early and aggressive management secondary to the presence of greater co-morbidity in this patient population. However, our data demonstrate that patients who had HF and those who did not develop HF had higher survival rates if angiography had been performed before discharge, even after adjustment for baseline patient risk, a propensity score adjusting for the likelihood of angiography, and differences in hospital characBRIEF REPORTS 231 had been admitted to North American hospitals with on-site interventional facilities were most likely to undergo angiography before discharge. Although patients who have HF tend to have more co-morbidities, they also appear to benefit from an early invasive approach after fibrinolysis. We previously published a detailed comparison of patients who underwent angiography with patients who did not.4 ••• Our findings support recommendations based on previous studies5–7 and the 2004 STEMI guidelines8 that advocate an early invasive approach in HF after STEMI. We also associated an early invasive approach with longer survival rates across the HF spectrum. This observation is consistent with a similar analysis of primary percutaneous intervention for different degrees of HF.9 A German study associated routine early stenting in all patients after fibrinolysis (regardless of HF) with a 50% decreased rate of ischemic complications (51% vs 26%, p П 0.001).10 Our analysis of geographic practice variation has confirmed those of previous studies11,12 that demonstrated an apparent underuse of invasive procedures after fibrinolysis, particularly outside the United States. Even in American hospitals with on-site facilities, there appears to be a paradoxic tendency for higher rates of angiography in patients FIGURE 3. Cumulative mortality rates over 1 year in patients who have lower risk.4,13–15 We observed a similar who had HF versus those who did not (A) and among those who practice pattern among patients who had been strathad different degrees of HF during the index admission (B). Anified by the presence of HF in our analyses, espegio ‫ ؍‬angiography. cially in Europe, where patients who had HF were 25% to 40% less likely to undergo an invasive procedure than were patients who did not have HF. TABLE 4 Use of Inpatient Angiography to Predict 30-day Mortality Rate Our data suggest that such a conHazard Ratio servative approach is associated with for In-patient an increased incidence of deaths due Models Predicting Death Over 30 Days Angiography 95% CI p Value to reinfarction and cardiac rupture. Unadjusted (univariate) model Percutaneous coronary intervention HF absent 0.33 0.27–0.40 ПЅ0.0001 after fibrinolysis has been associated HF present 0.50 0.42–0.60 ПЅ0.0001 with a decrease in recurrent myocarAdjusted for TIMI risk score dial infarction,16 possibly due to the HF absent 0.36 0.29–0.44 ПЅ0.0001 HF present 0.57 0.48–0.69 ПЅ0.0001 additional mechanical stability that Adjusted for TIMI risk score and propensity the intervention provides to the culscore for angiography before discharge prit lesion. The observed decrease in HF absent 0.36 0.28–0.46 ПЅ0.0001 fatal cardiac rupture in patients who HF present 0.59 0.48–0.73 ПЅ0.0001 underwent revascularization is conAdjusted for TIMI risk score, propensity score, and hospital characteristics sistent with the notion that percutaHF absent 0.36 0.28–0.46 ПЅ0.0001 neous coronary intervention restores HF present 0.57 0.46–0.70 ПЅ0.0001 myocardial perfusion and improves Unadjusted and adjusted mortality at 1 year among patients who underwent angiography before tissue turgor, thus increasing the tendischarge. Results are shown stratified by the presence/absence of HF during the index admission. sile strength of the myocardium. This Patients who did not undergo angiography before discharge represent the referent group (OR 1.0). See mechanism has been postulated as an text for details. explanation for the very low rate of TIMI П Thrombolysis In Myocardial Infarction; other abbreviations as in Table 3. rupture observed with primary percutaneous coronary intervention.17–19 Improved myocardial perfusion into the healing infarct zone may also deteristics. In this worldwide clinical trial, young men crease the adverse effects of remodeling and has who previously had successful percutaneous coro- been postulated as a potential late benefit of an open nary intervention, had fewer co-morbidities, and artery.20 232 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 executive summary: a report of the ACC/AHA Task Force on Practice Guidelines (Committee to Revise the 1999 Guidelines on the Management of Patients With Acute Myocardial Infarction). Circulation 2004;110; Angiogram No Angiogram Excess Deaths/ e82– e292. 9. Zahn R, Schiele R, Schneider S, Gitt AK, WienberCause of Death (282 deaths) (1,238 deaths) p Value 1,000 Patients gen H, Seidl K, Voigtlander T, Gottwik M, Berg G, HF (n П 385) 24.8% 25.4% 0.88 П©21 Altmann E, et al. Primary angioplasty versus intravenous thrombolysis in acute myocardial infarction: can Reinfarction (n П 305) 15.6% 21.1% 0.04 П©20 we define subgroups of patients benefiting most from Rupture (n П 173) 6.4% 12.5% 0.002 П©13 primary angioplasty? Results from the pooled data of Intracranial hemorrhage 2.8% 7.0% 0.006 П©8 the Maximal Individual Therapy in Acute Myocardial (n П 95) Infarction Registry and the Myocardial Infarction RegOther cardiovascular (n П 133) 9.2% 8.6% 0.73 П©7 istry. J Am Coll Cardiol 2001;37:1827–1835. Dysrhythmia (n П 105) 8.2% 6.6% 0.36 П©5 10. Scheller B, Hennen B, Hammer B, Walle J, Hofer Other noncardiovascular 17.0% 7.9% ПЅ0.0001 П©2 C, Hilpert V, Winter H, Nickenig G, Bohm M. Bene(n П 146) ficial effects of immediate stenting after thrombolysis in acute myocardial infarction. J Am Coll Cardiol 2003; Unobserved (n П 75) 6.4% 4.6% 0.22 П©3 42:634 – 641. Unknown (n П 103) 9.6% 6.1% 0.048 П©3 11. Gupta M, Chang WC, Van de Werf F, Granger CB, Midodzi W, Barbash G, Pehrson K, Oto A, Toutouzas Numbers in the second and third columns represent the percentage of deaths due to the indicated P, Jansky P, Armstrong PW. International differences in causes. The last column represents the excess number of deaths among 1,000 patients who did not in-hospital revascularization and outcomes following undergo angiography versus those who did undergo angiography before discharge for each cause of acute myocardial infarction: a multilevel analysis of death. patients in ASSENT-2. Eur Heart J 2003;24:1640 – 1650. 12. Fox KA, Goodman SG, Anderson FA Jr, Granger CB, Moscucci M, Flather MD, Spencer F, Budaj A, Dabbous OH, Gore JM. From guidelines to clinical 1. The In TIME-II Investigators. Intravenous NPA for the Treatment of Infarcting practice: the impact of hospital and geographical characteristics on temporal Myocardium Early; InTIME-II, a double-blind comparison of single-bolus trends in the management of acute coronary syndromes. The Global Registry of lanoteplase vs accelerated alteplase for the treatment of patients with acute Acute Coronary Events (GRACE). Eur Heart J 2003;24:1414 –1424. 13. Stenestrand U, Wallentin L. Early revascularisation and 1-year survival in myocardial infarction. Eur Heart J 2000;21:2005–2013. 2. Kashani A, Giugliano RP, Altman E, Morrow DA, Gibson CM, Murphy S, 14-day survivors of acute myocardial infarction: a prospective cohort study. Braunwald E. Severity of heart failure, treatments, and outcomes after fibrinoLancet 2002;359:1805–1811. 14. Magid DJ, Calonge BN, Rumsfeld JS, Canto JG, Frederick PD, Every NR, lysis. Eur Heart J 2004;25:1702–1710. 3. Morrow DA, Antman EM, Charlesworth A, Cairns R, Murphy SA, de Lemos Barron HV. Relation between hospital primary angioplasty volume and mortality JA, Giugliano RP, McCabe CH, Braunwald E. TIMI risk score for ST-elevation for patients with acute MI treated with primary angioplasty vs thrombolytic myocardial infarction: a convenient, bedside, clinical score for risk assessment at therapy. JAMA 2000;284:3131–3138. 15. Pilote L, Miller DP, Califf RM, Rao JS, Weaver WD, Topol EJ. Determinants presentation: an intravenous nPA for treatment of infarcting myocardium early II of the use of coronary angiography and revascularization after thrombolysis for trial substudy. Circulation 2000;102:2031–2037. 4. Llevadot J, Giugliano RP, Antman EM, Wilcox RG, Gurfinkel EP, Henry T, acute myocardial infarction. N Engl J Med 1996;335:1198 –1205. 16. Gibson CM, Karha J, Murphy SA, James D, Morrow DA, Cannon CP, McCabe CH, Charlesworth A, Thompson S, Nicolau JC, et al. Availability of Giugliano RP, Antman EM, Braunwald E. Early and long-term clinical outcomes on-site catheterization and clinical outcomes in patients receiving fibrinolysis for associated with reinfarction following fibrinolytic administration in the ThromST-elevation myocardial infarction. Eur Heart J 2001;22:2104 –2115. 5. Hochman JS, Sleeper LA, Webb JG, Sanborn TA, White HD, Talley JD, Buller bolysis in Myocardial Infarction trials. J Am Coll Cardiol 2003;42:7–16. 17. Moreno R, Lopez-Sendon J, Garcia E, Perez de Isla L, Lopez de Sa E, Ortega CE, Jacobs AK, Slater JN, Col J, et al. Early revascularization in acute myocarA, Moreno M, Rubio R, Soriano J, Abeytua M, Garcia-Fernandez MA. Primary dial infarction complicated by cardiogenic shock. SHOCK Investigators. Should angioplasty reduces the risk of left ventricular free wall rupture compared with We Emergently Revascularize Occluded Coronaries for Cardiogenic Shock. thrombolysis in patients with acute myocardial infarction. J Am Coll Cardiol N Engl J Med 1999;341:625– 634. 6. Dauerman HL, Goldberg RJ, White K, Gore JM, Sadiq I, Gurfinkel E, Budaj 2002;39:598 – 603. 18. Sugiura T, Nagahama Y, Nakamura S, Kudo Y, Yamasaki F, Iwasaka T. Left A, Lopez de Sa E, Lopez-Sendon J. Revascularization, stenting, and outcomes of ventricular free wall rupture after reperfusion therapy for acute myocardial patients with acute myocardial infarction complicated by cardiogenic shock. Am J infarction. Am J Cardiol 2003;92:282–284. Cardiol 2002;90:838 – 842. 7. Hochman JS, Sleeper LA, White HD, Dzavik V, Wong SC, Menon V, Webb 19. Ohishi F, Hayasaki K, Honda T. Effect of thrombolysis on rupture of the left JG, Steingart R, Picard MH, Menegus MA, et al. One-year survival following ventricular free wall following acute myocardial infarction. J Cardiol 1996;28: early revascularization for cardiogenic shock. JAMA 2001;285:190 –192. 27–32. 8. Antman EM, Armstrong PW, Bates ER, Green LA, Hand M, Hochman JS, 20. Sadanandan S, Buller C, Menon V, Dzavik V, Terrin M, Thompson B, Lamas Krumholz HM, Kushner FG, Lamas GA, Mullany CJ, et al. ACC/AHA guideG, Hochman JS. The late open artery hypothesis—a decade later. Am Heart J lines for the management of patients with ST-elevation myocardial infarction: 2001;142:411– 421. TABLE 5 One-year Causes of Death in Patients Who Had Angiography Before Discharge and Those Who Did Not BRIEF REPORTS 233 Impact of ST-Segment Depression Resolution on Mortality After Successful Mechanical Reperfusion in Patients With ST-Segment Elevation Acute Myocardial Infarction Giuseppe De Luca, MD, PhD, Arthur C. Maas, MD, Arnoud W.J. van’t Hof, Jan Paul Ottervanger, MD, PhD, Jan C.A. Hoorntje, MD, PhD, A.T. Marcel Gosselink, MD, PhD, Jan-Henk E. Dambrink, MD, PhD, Menko-Jan de Boer, MD, PhD, and Harry Suryapranata, MD, PhD The aim of the present study was to evaluate the additional prognostic effect of ST-depression resolution in 610 patients who had ST-elevation myocardial infarction and underwent successful primary angioplasty (postprocedural Thrombolysis In Myocardial Infarction 3 flow and complete resolution of ST-segment elevation). Incomplete resolution of ST-segment depression (<70%) was observed in 50 patients (8.2%). These patients were older, had a higher Killip’s class at presentation, had larger infarcts, and had an increased 1-year mortality (10% vs 2%, p ‫؍‬ 0.0004). At multivariate analysis, incomplete resolution of ST-segment depression was an independent predictor of 1-year mortality (p ‫ ؍‬0.028). б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:234 –236) he main goal in the control of ST-segment elevation acute myocardial infarction is an early and T sustained patency of the infarct-related artery. 1–3 However, several studies have shown that patency of the epicardial vessel does not guarantee optimal myocardial perfusion.4 – 6 Electrocardiography represents a inexpensive and easy method to evaluate myocardial reperfusion.5–7 We5 previously showed that, among patients who had postprocedural Thrombolysis In Myocardial Infarction (TIMI) grade 3 flow, resolution of ST-segment elevation is an independent prognostic determinant after primary angioplasty. Despite the negative prognostic implication of additional ST-segment depression in patients who have ST-segment elevation acute myocardial infarction,8 –10 the prognostic role of resolved ST-segment depression has not been investigated. This study investigated the prognostic role of resolved ST-segment depression in patients whose ST-segment elevation acute myocardial infarction was successfully managed with primary angioplasty. ••• From the Department of Cardiology, ISALA Klinieken De Weezenlanden Hospital, Zwolle, The Netherlands. Dr. Suryapranata’s address is: ISALA Klinieken, De Weezenlanden Hospital, Department of Cardiology, Groot Wezeland 20, 8011 JW Zwolle, The Netherlands. Email: [email protected]. Manuscript received June 14, 2004; revised manuscript received and accepted September 1, 2004. 234 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, PhD, A total of 1,548 patients who had ST-segment elevation acute myocardial infarction underwent primary angioplasty at our institution between April 1997 and October 2001. Of 1,072 patients who had complete data on resolution of ST-segment elevation myocardial infarction, 900 showed baseline ST-segment depression. Our final patient population consisted of 610 patients who had baseline ST-segment depression and successful reperfusion, defined as postprocedural TIMI grade 3 flow and complete resolution of ST-segment elevation (ST-segment resolution Х†70%).7 Informed consent was obtained from each patient before angiography. All subjects presented Х…6 hours from symptom onset or between 6 and 24 hours if they had continuous symptoms and signs of ischemia (persistent or recurrent chest pain and/or persistent elevation or reelevation of ST segment). All patients received aspirin (500 mg intravenously) and heparin (10,000 IU intravenously) before the procedure. Because the benefits of glycoprotein IIb/IIIa inhibitors have been proved only recently,18 ПЅ5% of subjects received this additional therapy. Coronary angiography and ST-segment resolution were analyzed by an independent core laboratory (Diagram, Zwolle, The Netherlands). TIMI flow grade and myocardial blush grade were assessed after primary angioplasty, as previously described.6 Enzymatic infarct size was calculated by cumulative release of lactate dehydrogenase from serial measurements Х…48 hours after symptom onset. Analysis of ST-segment resolution was performed by comparison between baseline and 3-hour 12-lead electrocardiograms, as previously described.7 Patients were classified as having complete (Х†70%) and incomplete (ПЅ70%) resolution of ST-segment depression. Records of patients who visited our outpatient clinic were reviewed. For all other patients, information was obtained from the patient’s general physician or by direct telephone interview with the patient. For patients who died during followup, hospital records and necropsy data were reviewed. One-year follow-up data were available in all patients. Statistical analysis was performed with SPSS 10.0 (SPSS, Inc., Chicago, Illinois). Analysis of variance and chi-square test were used appropriately for continuous and categorical variables. Differences in event rates between groups during the follow-up period were assessed by the Kaplan-Meier method using the 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.008 TABLE 1 Demographic, Clinical, and Angiographic Characteristics According to Resolution of ST-segment Depression ST-depression Resolution Age (yrs) Men Diabetes mellitus Hypertension Previous myocardial infarction Killip’s class Пѕ1 Anterior wall acute infarction Ischemia time Пѕ4 h Multivessel coronary disease TIMI 0/1 flow before Myocardial blush grade 2/3 Stent Maximal baseline ST-depression (mm) Complete (n П 560) Incomplete (n П 50) p Value 59 П® 11 433 (77.3%) 40 (7.1%) 125 (22.9%) 49 (8.8%) 68 П® 9 33 (66.0%) 5 (10%) 15 (30%) 8 (16.0%) ПЅ0.0001 0.071 Пѕ0.1 Пѕ0.1 0.091 31 (5.5%) 215 (38.4%) 48 (16.0%) 16 (32.0%) 0.004 ПЅ0.1 177 (31.6%) 300 (53.6%) 17 (34.0%) 33 (66.0%) Пѕ0.1 0.091 369 (65.9%) 505 (90.2%) 37 (74%) 46 (92.0%) Пѕ0.1 Пѕ0.1 317 (56.6%) 2.22 П® 1.56 29 (58.0%) 2.37 П® 1.31 Пѕ0.1 Пѕ0.1 FIGURE 1. Kaplan-Meier survival curves according to resolution of ST-segment depression. TABLE 2 Predictors of One-year Mortality at Multivariate Analysis Relative risk (95% CI) log-rank test. Multivariate analysis was performed with Cox’s proportional hazard method to identify independent predictors of 1-year mortality rate. Stepwise selection of variables and estimation of significant probabilities were computed by a maximal likelihood ratio test. The chi-square value was calculated from the log of the ratio of maximal partial likelihood functions. The additional value of each category of variables added sequentially was evaluated on the basis of the increases in the overall likelihood statistical ratio. Incomplete resolution of ST-segment depression was observed in 50 patients (8.2%). These patients were older, had a higher Killip’s class at presentation, showed a higher prevalence of women, and had more previous infarctions (Table 1). Incomplete resolution of ST-segment depression was associated with larger infarcts (2,058 П® 1,658 vs 1,417 П® 1,253 U/L, p ПЅ0.0001) and higher 1-year mortality (10% vs 2%, relative risk 5.41, 95% confidence interval 1.88 to 15.58, p П 0.0004). Kaplan-Meier survival curves according to resolution of ST-segment depression are shown in Figure 1. At multivariate analysis, incomplete resolution of ST-segment depression was an independent predictor of 1-year mortality (p П 0.028; Table 2). ••• The main finding of the present study is that, among patients whose ST-segment elevation acute myocardial infarction was successfully managed with primary angioplasty, resolved ST-segment depression does provide additional prognostic information to resolved ST-segment elevation. Despite optimal restoration of anterograde flow in the infarct-related artery (TIMI grade 3 flow), suboptimal myocardial reperfusion may occur in a large percentage of patients.5,6 Evaluation of ST-segment resolution has been shown to be a reliable, cheap, and user-friendly method to Killip’s class Пѕ1 Anterior infarction Incomplete ST-depression resolution Age Пѕ65 yrs Myocardial blush grade 0/1 p Value 11.01 (4.92–30.2) ПЅ0.0001 3.52 (1.21–10.2) 0.021 3.58 (1.15–11.15) 0.028 3.69 (1.15–11.8) 1.83 (0.55–6.12) 0.028 Пѕ0.1 CI П confidence interval. analyze myocardial perfusion in patients who had acute myocardial infarction that was controlled by pharmacologic or mechanical reperfusion.5–7 Our previous report5 showed that, in patients who had postprocedural TIMI 3 flow, ST-segment resolution significantly added prognostic information in terms of long-term mortality rate. The mechanism of concomitant ST-depression in patients who have acute myocardial infarction remains a matter of debate. It may represent a “reciprocal” image of ST-segment elevation in the infarct zone,11,12 more extensive infarction,13,14 or additional ischemia beyond the infarct zone.15,16 Previous reports have shown that ST-segment depression has significant prognostic implications in patients who have ST-segment elevation acute myocardial infarction.8 –10 Thus, it is conceivable that resolution of ST-segment depression would provide additional prognostic information about resolution of ST-segment elevation. This is the first study to explore this issue in patients who had ST-segment elevation acute myocardial infarction and underwent primary angioplasty. To address this issue, we performed our analysis in patients who had postprocedural TIMI grade 3 flow and complete ST-segment resolution. We found that postprocedural incomplete resolution of ST-segment depression was associated with larger infarcts and subsequently with a higher mortality rate. This finding was confirmed at multivariate analysis. Analysis of resolved ST-segment elevation at a fixed period after reperfusion therapy (3 hours in our BRIEF REPORTS 235 study) may have limitations. Myocardial reperfusion is a dynamic process during which alternating episodes of ST-segment resolution may occur.17 Fluctuation of the ST segment has been described during or shortly after thrombolysis and primary angioplasty, and continuous ST-segment monitoring has been shown to be a valid alternative.17,18 1. The GUSTO Angiographic Investigators. The effects of tissue plasminogen activator, streptokinase, or both on coronary-artery patency, ventricular function, and survival after acute myocardial infarction. N Engl J Med 1993;329: 1615–1622. 2. Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Indications for fibrinolytic therapy in suspected acute myocardial infarction: collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients. Lancet 1994;343:311–322. 3. Rentrop KP. Restoration of anterograde flow in acute myocardial infarction: the first 15 years. J Am Coll Cardiol 1995;25:1S–2S. 4. van ’t Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction Study Group. Lancet 1997;350:615– 619. 5. van ’t Hof A, Liem A, Suryapranata H, Hoorntje JCA, de Boer MJ, Zijlstra F, on the behalf of the Zwolle Myocardial Infarction Study Group. Angiographic assessment of myocardial reperfusion in patients treated with primary angioplasty for acute myocardial infarction. Myocardial Blush Grade. Circulation 1998;97: 2302–2306. 6. Ito H, Tomooka T, Sakai N, Yu H, Higashino Y, Fujii K, Masuyama T, Kitabatake A, Minamino T. Lack of myocardial perfusion immediately after successful thrombolysis. A predictor of poor recovery of left ventricular function in anterior myocardial infarction. Circulation 1992;85:1699 –1705. 7. Schroder R, Dissmann R, Bruggemann T, Wegscheider K, Linderer T, Tebbe U, Neuhaus KL. Extent of early ST segment elevation resolution: a simple but strong predictor of outcome in patients with acute myocardial infarction. J Am Coll Cardiol 1994;24:384 –391. 8. Croft CH, Woodward W, Nicod P, Corbett JR, Lewis SE, Willerson JT, Rude RE. Clinical implications of anterior S-T segment depression in patients with acute inferior myocardial infarction. Am J Cardiol 1982;50:428 – 436. 9. Shah A, Wagner GS, Califf RM, Boineau RE, Green CL, Wildermann NM, Trollinger KM, Pope JE, Krucoff MW. Comparative prognostic significance of simultaneous versus independent resolution of ST segment depression relative to ST segment elevation during acute myocardial infarction. J Am Coll Cardiol 1997;30:1478 –1483. 10. Lembo NJ, Starling MR, Dell’Italia LJ, Crawford MH, Chaudhuri TK, O’Rourke RA. Clinical and prognostic importance of persistent precordial (V1V4) electrocardiographic ST segment depression in patients with inferior transmural myocardial infarction. Circulation 1986;74:56 – 63. 11. Camara EJ, Chandra N, Ouyang P, Gottlieb SH, Shapiro EP. Reciprocal ST change in acute myocardial infarction: assessment by electrocardiography and echocardiography. J Am Coll Cardiol 1983;2:251–257. 12. Mukharji J, Murray S, Lewis SE, Croft CH, Corbett JR, Willerson JT, Rude RE. Is anterior ST depression with acute transmural inferior infarction due to posterior infarction? A vectorcardiographic and scintigraphic study. J Am Coll Cardiol 1984;4:28 –34. 13. Wasserman AG, Ross AM, Bogaty D, Richardson DW, Hutchinson RG, Rios JC. Anterior ST segment depression during acute inferior myocardial infarction: evidence for the reciprocal change theory. Am Heart J 1983;106:516 –520. 14. Haraphongse M, Tanomsup S, Jugdutt BI. Inferior ST segment depression during acute anterior myocardial infarction: clinical and angiographic correlations. J Am Coll Cardiol 1984;4:467– 476. 15. Pichler M, Shah PK, Peter T, Singh B, Berman D, Shellock F, Swan HJ. Wall motion abnormalities and electrocardiographic changes in acute transmural myocardial infarction: implications of reciprocal ST segment depression. Am Heart J 1983;106(5 pt 1):1003–1009. 16. Tzivoni D, Chenzbraun A, Keren A, Benhorin J, Gottlieb S, Lonn E, Stern S. Reciprocal electrocardiographic changes in acute myocardial infarction. Am J Cardiol 1985;56:23–26. 17. Shah PK, Cercek B, Lew A, Ganz W. Angiographic validation of bedside markers of reperfusion. J Am Coll Cardiol 1993;21:55– 61. 18. Krucoff MW, Croll MA, Pope JE, Granger CB, O’Connor CM, Sigmon KN, Wagner BL, Ryan JA, Lee KL, Kereiakes DJ. Continuous 12-lead ST-segment recovery analysis in the TAMI-7 study. Performance of a noninvasive method for real-time detection of failed myocardial reperfusion. Circulation 1993;88:437– 446. Effects of Tirofiban and Statins on High-Sensitivity C-Reactive Protein, Interleukin-6, and Soluble CD40 Ligand Following Percutaneous Coronary Interventions in Patients With Stable Coronary Artery Disease Rabih R. Azar, MD, MSc, Georges Badaoui, MD, Antoine Sarkis, MD, Roland Kassab, MD, Elie SalamГ©, MD, Samira KlaymГ©, PhD, Roger Naman, MD, and Mirna Germanos, PharmD This study assessed the effects of tirofiban and statins on high-sensitivity C-reactive protein, interleukin-6, and soluble CD40 ligand after percutaneous coronary intervention in patients who had stable coronary artery disease. Tirofiban insignificantly limited the increase of soluble CD40 ligand after revascularization, especially in patients who had high levels of this marker at baseline (p ‫ ؍‬0.06), whereas statins significantly inhibited increases in interleukin-6 and, to a From the Division of Cardiology, the Laboratory of Histocompatibility, and the Division of Immunology, Hotel Dieu de France Hospital and the St-Joseph University School of Medicine, Beirut, Lebanon. This study was supported by research grant FM 65 from the Council of Research of St-Joseph University, Beirut, Lebanon. Dr. Azar’s address is: Division of Cardiology, Hotel Dieu de France Hospital, Achrafieh, Beirut, Lebanon. E-mail: [email protected]. Manuscript received July 9, 2004; revised manuscript received and accepted August 31, 2004. 236 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 lesser extent, high-sensitivity C-reactive protein without affecting the soluble CD40 ligand. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:236 –240) ercutaneous coronary intervention leads to activation of platelets and white cells and triggers P an acute inflammatory response that plays a major role in the pathogenesis of complications after this procedure.1,2 Antiplatelet therapy with tirofiban decreases ischemic complications of percutaneous intervention, especially in the setting of acute coronary syndromes, when inflammation is at its peak.3,4 However, effects of tirofiban on inflammation and platelet activation and their release of the proinflammatory mediator, soluble CD40 ligand (sCD40l), remain unknown. In addition, statins improve the outcome of patients who undergo percutaneous revascularization. The 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.093 TABLE 1 Baseline Clinical and Procedural Characteristics of Tirofiban and Saline Groups Variable Age (yrs) Men Diabetes mellitus Hyperlipidemia Current smoker Hypertension Family history of coronary artery disease Healed myocardial infarction Previous coronary bypass Current statin therapy 3-Vessel coronary artery disease Percent baseline stenosis Percent residual stenosis Average length of stent Maximum inflation pressure Tirofiban (n П 35) Saline (n П 35) 59 П® 9 30 (86%) 13 (37%) 29 (83%) 10 (29%) 19 (54%) 16 (46%) 63 П® 10 29 (83%) 8 (23%) 30 (86%) 15 (43%) 24 (69%) 10 (29%) 9 (26%) 5 (14%) 22 (63%) 4 (11%) 85 П® 12 1П®5 18 П® 7 13 П® 3 5 (14%) 5 (14%) 22 (63%) 4 (11%) 88 П® 10 2П®5 18 П® 4 14 П® 2 Data are numbers of patients (percentages) or mean П® SD. TABLE 2 Baseline Clinical and Procedural Characteristics in Statin and Control Groups Variable Age (yrs) Men Diabetes mellitus Hyperlipidemia Current smoker Hypertension Family history of coronary artery disease Healed myocardial infarction Previous coronary bypass 3-Vessel coronary artery disease Tirofiban during PCI Percent baseline stenosis Percent residual stenosis Average length of stent Maximum inflation pressure Statin (n П 44) Control (n П 26) 59 П® 10 35 (79%) 14 (32%) 41 (93%) 17 (39%) 24 (54%) 18 (41%) 63 П® 8 24 (92%) 7 (27%) 8 (69%)* 8 (31%) 19 (73%) 8 (31%) 10 (23%) 6 (14%) 5 (11%) 22 (50%) 86 П® 10 1П®5 18 П® 6 14 П® 3 4 (15%) 4 (15%) 3 (11%) 13 (50%) 86 П® 12 1П®5 19 П® 5 14 П® 2 *p П 0.008. Data are numbers of patients (percentages) or mean П® SD. PCI П percutaneous coronary intervention. mechanism of this benefit was hypothesized to be an anti-inflammatory activity,5–7 but this has not been directly demonstrated. This study investigated the effects of tirofiban and statins on serum levels of high-sensitivity C-reactive protein (hs-CRP), interleukin-6 (IL-6), and sCD40l after percutaneous coronary intervention in patients who have stable coronary artery disease. ••• Patients who had stable coronary artery disease and who were referred for elective percutaneous coronary intervention for a de novo lesion in a native coronary artery were screened to participate in this study. Patients who had proinflammatory conditions or were taking anti-inflammatory drugs were excluded. Informed consent was obtained from all patients, and the study protocol was approved by the institutional review board of the Hotel Dieu de France Hospital (Beirut, Lebanon). FIGURE 1. Mean changes ШЋ SD in levels of sCD40l according to sCD40l quartiles at baseline (p <0.001 by 1-way analysis of variance). TABLE 3 Pre- and Postprocedural Levels of All Markers in the Tirofiban and Saline Groups Tirofiban (n П 35) hs-CRP (mg/L) Baseline 48 h Mean change IL-6 (pg/ml) Baseline 48 h Mean change sCD40l (ng/ml) Baseline 48 h Mean change 7 П® 10 17.3 П® 16 10.3 П® 10 Saline (n П 35) 7.7 П® 6.9 19.6 П® 17.2 11.9 П® 14.7 3.3 П® 2.5 6 П® 4.5 2.7 П® 4 4.5 П® 4 7.2 П® 6.4 2.7 П® 4.4 4 П® 2.9 6.3 П® 3.1* 2.2 П® 3.6 4.5 П® 2.9 8П®4 3.5 П® 4.2 Data are mean П® SD. *p П 0.05 versus saline; p ПЅ0.01 for all values at 48 hours versus baseline. The study was randomized, open label, and prospective. At the time of percutaneous coronary intervention and before crossing the lesion with a guidewire, patients were randomized to receive an intravenous bolus and a 24-hour infusion of tirofiban or normal saline. Percutaneous coronary intervention was performed in the standard fashion and consisted of balloon angioplasty and stenting. Tirofiban was given at a dose of 10 вђ®g/kg as an intravenous bolus and as a 0.15-вђ®g/kg/min continuous infusion for 24 hours. Patients who were receiving statins at the time of the procedure were defined as statin users. Levels of hs-CRP, IL-6, and sCD40l were measured immediately before and 48 hours after the procedure, and cardiac troponin T was measured 24 hours after percutaneous coronary intervention. Study end points were changes in marker levels and in levels 48 hours after the procedure. BRIEF REPORTS 237 (Table 1) and between the statin and control groups (Table 2). Pretreatment with clopidogrel at a loading dose of 300 mg 18 to 24 hours before the procedure Statin Control was given to 94% of patients (equally between groups: (n П 44) (n П 26) tirofiban vs saline and statin vs control). There were hs-CRP (mg/l) no complications, and all patients had undetectable Baseline 7.3 П® 7.1 7.4 П® 10.8 troponin levels 24 hours after revascularization. 48 h 16.8 П® 13.3 21.3 П® 21 Angioplasty resulted in significant increases in hsMean change 9.4 П® 10.6 13.9 П® 15 CRP, IL-6, and sCD40l (150%, 69%, and 65%, reIL-6 (pg/ml) Baseline 3.9 П® 2.8 3.9 П® 4.2 spectively, vs baseline, p ПЅ0.01 for all). Changes in 48 h 5.6 П® 3.9* 8.3 П® 7.3 hs-CRP and IL-6 correlated closely together (r П 0.5, †Mean change 1.6 П® 3.9 4.3 П® 4.1 p ПЅ0.001) but did not correlate with changes in sCD40l (ng/ml) sCD40l (r П 0.2 for hs-CRP vs sCD40l and r П 0.05 Baseline 4.2 П® 3 4.5 П® 2.8 for IL-6 vs sCD40l, p П NS for the 2 comparisons). 48 h 7П®4 7.4 П® 3.4 Mean change 2.8 П® 3.8 2.9 П® 4.1 However, the increase in sCD40l correlated strongly †with the baseline level of this marker. Patients in the *p Х…0.05 and p П 0.008 versus controls; p ПЅ0.01 for all values at 48 upper quartile of sCD40l did not show a statistically hours versus baseline. Data are mean П® SD. significant change in this marker after the procedure, whereas those in the lower 3 quartiles showed a significant increase of this marker (p ПЅ0.001 for intergroup comparison; Figure 1). Effects of tirofiban are listed in Table 3. Levels of hs-CRP and IL-6 increased equally in the tirofiban and saline groups, but sCD40l increased less (p П NS) in the tirofiban group, resulting in a statistically lower level of sCD40l at 48 hours compared with the control group (p П 0.05). Because sCD40l did not significantly increase after angioplasty in patients who had high levels of this marker at baseline (highest quartile), the effect of tirofiban was tested in the 3 lower quartiles of sCD40l in which percutaneous coronary interFIGURE 2. Changes in hs-CRP according to baseline hs-CRP levels and statin use (p ‫؍‬ vention significantly increased this 0.005 for intergroup comparison in controls and p ‫ ؍‬NS in the statin group by marker. In these patients, the inhibi1-way analysis of variance). tory effect of tirofiban was more pronounced (mean changes 2.7 П® 3.7 vs 4.6 П® 3.5 ng/ml for tirofiban vs saBlood samples were immediately taken to the labo- line, p П 0.06), resulting in much lower postproceratory, where serum was frozen and stored at ПЄ80В°C dural levels of sCD40l in the tirofiban group (5.5 П® until batch analysis. hs-CRP was measured by nephe- 3.1 vs 7.7 П® 3.5 ng/ml, p П 0.02). lometry (N Latex CRP Mono, Dade-Behring, Marburg, Effects of statins are presented in Table 4. Statin Germany). IL-6 and sCD40l were measured with high- therapy did not affect the increase in sCD40l but did sensitivity quantitative immunoassays (R&D Systems, attenuate the increase of hs-CRP by 32% (p П NS) Abingdon, United Kingdom). Assay sensitivities were and of IL-6 by 63% (p П 0.008). However, when the 0.7 pg/ml for IL-6 and 4.2 pg/ml for sCD40l. Intra- and change in hs-CRP was measured according to baseline interassay variabilities were Х…6% for all tests. quartile levels of this marker, statin therapy signifiData were summarized as mean П® SD or as per- cantly attenuated the increase of hs-CRP by 71% in centages and compared with Student’s t test or chi- the subgroup of patients in the upper quartile (mean square test. One-way analysis of variance was used to change 9 П® 10 vs 31 П® 22 mg/L for statin vs control, compare changes in markers and effects of therapies p П 0.01). As shown in Figure 2, there was a statisin multiple subgroups. Pearson’s test was used to tically significant interaction between baseline levels study the correlation across changes in hs-CRP, IL-6, and sCD40l. All tests were 2-tailed, and a p value of hs-CRP, statin therapy, and increased hs-CRP after percutaneous revascularization. ПЅ0.05 was considered statistically significant. ••• Baseline clinical and procedural characteristics of This study demonstrates that, in patients who have the 70 patients who were enrolled in the study were similar between the tirofiban and saline groups stable coronary artery disease, percutaneous coronary TABLE 4 Pre- and Postprocedural Levels of All Markers in Statin and Control Groups 238 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 interventions induce a systemic inflammatory response similar to that seen in acute coronary syndromes. It also demonstrates platelet activation, as demonstrated by increased levels of sCD40l. This molecule is a potent proinflammatory and prothrombotic agent8 –11 and is a major link in interactions between platelets and white blood cells. However, there was no correlation between changes in hs-CRP and sCD40l or between IL-6 and sCD40l, suggesting that inflammation after percutaneous coronary intervention is independent of platelets and is primarily driven by activation of white blood cells. The effects of tirofiban and statins on these inflammatory markers provide additional support for this statement. This study suggests that tirofiban inhibits the release of sCD40l by platelets but does not attenuate the acute inflammatory response triggered by percutaneous coronary intervention. Several mechanisms may explain these findings. Although tirofiban blocks in vitro the release of sCD40l from platelets in a dose-dependent manner,12 the decrease in sCD40l observed in our trial was of borderline significance and may have been insufficient to suppress inflammation. In addition, tirofiban can theoretically attenuate the formation of platelet/leukocyte complexes by blocking the binding of the platelets’ glycoprotein IIb/IIIa receptor to the leukocyte’s MAC-1 receptor.13 However, these complexes can always form by binding the platelet’s P-selectin to the leukocyte’s P-selectin glycoprotein ligand-1.14 Leukocytes may also be directly activated by rupture of the culprit plaque (independent of their interaction with platelets). Blockage of the leukocyte MAC-1 receptor seems essential for suppression of inflammation. This receptor is not affected by tirofiban but is effectively blocked by abciximab,15,16 which has been shown to inhibit the increases in hs-CRP, IL-6, and tumor necrosis factor-вђЈ by 32% to 100% after percutaneous coronary intervention.17 Statins attenuated the acute inflammatory response triggered by percutaneous intervention. The greater effect of statins on IL-6 compared with hs-CRP most likely reflects the greater stability of CRP in the circulation (half-life of 19 vs 4 hours for IL-6).18 In addition, IL-6 is secreted by activated white blood cells and stimulates CRP synthesis by the liver.18 Thus, IL-6 is the first to decrease after inhibition of leukocytes, whereas a decrease in hs-CRP is usually delayed by an additional 24 to 48 hours. The mechanism of statins’ anti-inflammatory effect is independent of platelets, because these drugs did not attenuate the release of sCD40l and is most likely secondary to direct inhibition of leukocyte activation and to plaque stabilization.19 The implications of the present findings should be considered within the context of recently published trials. Chan et al6 demonstrated a survival benefit associated with statin therapy at the time of percutaneous coronary intervention. In their cohort of 5,052 patients, statin therapy before revascularization resulted in 53% and 67% decreases in mortality at 30 days and 6 months, respectively. In a subsequent report,7 they found that this benefit was mainly ob- served among patients within the highest hs-CRP quartile before intervention and hypothesized that it was secondary to the anti-inflammatory properties of statins. Our results support their conclusion. Our study has several limitations. First, the suppressive effect of tirofiban on the increase of sCD40l was of borderline significance (p П 0.06 after exclusion of the upper quartile group). However, levels of sCD40l after the procedure were significantly lower in patients who received tirofiban, whether or not patients in the upper sCD40l quartile were excluded. Second, we measured sCD40l levels 48 hours after angioplasty, whereas tirofiban infusion was given for only 24 hours. The magnitude of the decrease in sCD40l might have been greater if measurements had been done at 24 hours. Third, we limited our study to patients who had stable coronary artery disease. Fourth, statin use was not randomized. Although statins users and nonusers shared similar characteristics, there may have been unmeasurable potential confounders that affected the results. Fifth, the duration and doses of statins before revascularization were not collected. Statins might have been started shortly before percutaneous coronary intervention and/or the dosing may have been inadequate. However, in that case, the true anti-inflammatory effect of statins may be even more important than indicated by our findings. 1. Azar RR, McKay RG, Kiernan FJ, Seecharran B, Feng YJ, Fram DB, Wu AHB, Waters DD. Coronary angioplasty induces a systemic inflammatory response. Am J Cardiol 1997;80:1476 –1478. 2. Gaspardone A, Crea F, Versaci F, Tomai F, Pellegrino A, Chiariello L, Gioffre PA. Predictive value of C-reactive protein after successful coronary-artery stenting in patients with stable angina. Am J Cardiol 1998;82:515–518. 3. Platelet Receptor Inhibition in Ischemic Syndrome Management in Patients Limited by Unstable Signs and Symptoms (RRISM-PLUS) Study Investigators. Inhibition of the platelet glycoprotein IIb/IIIa receptor with tirofiban in unstable angina and non–Q-wave myocardial infarction. N Engl J Med 1998; 338:1488 –1497. 4. The RESTORE Investigators. Effects of platelet glycoprotein IIb/IIIa blockade with tirofiban on adverse cardiac events in patients with unstable angina or acute myocardial infarction undergoing coronary angioplasty. Randomized Efficacy Study of Tirofiban for Outcomes and Restenosis. Circulation 1997;96:1445–1453. 5. Albert MA, Danielson E, Rifai N, Ridker PM. Effect of statin therapy on C-reactive protein levels: the pravastatin inflammation/CRP evaluation (PRINCE): a randomized trial and cohort study. JAMA 2001;286:64 –70. 6. Chan AW, Bhatt DL, Chew DP, Quinn MJ, Moliterno DJ, Topol EJ, Ellis SG. Early and sustained survival benefit associated with statin therapy at the time of percutaneous coronary intervention. Circulation 2002;105:691– 696. 7. Chan AW, Bhatt D, Chew DP, Reginelli J, Schneider JP, Topol EJ, Ellis SG. Relation of inflammation and benefit of statins after percutaneous coronary interventions. Circulation 2003;107:1750 –1756. 8. AndrГ© P, Nannizzi-Alaimo L, Prasad SK, Phillips DR. Platelet-derived CD40L. The switch-hitting player of cardiovascular disease. Circulation 2002;106:896 – 899. 9. Schonbeck U, Libby P. The CD40/CD154 receptor/ligand dyad. Cell Mol Life Sci 2001;58:4 – 43. 10. Henn V, Slupsky J, Grafe M, Anagnostopoulos I, Forster R, Muller-Berghaus G, Kroczek RA. CD40 ligand on activated platelets triggers an inflammatory reaction of endothelial cells. Nature 1998;391:591–594. 11. AndrГ© P, Prasad SK, Denis CV, He M, Papalia JM, Hynes RO, Phillips DR, Wagner DD. CD40L stabilizes arterial thrombi by a beta-3 integrin-dependent mechanism. Nat Med 2002;8:247–252. 12. Nannizzi-Alaimo N, Alves VL, Phillips DR. Inhibitory effects of glycoprotein IIb/IIIa antagonists and aspirin on the release of sCD40l during platelet stimulation. Circulation 2003;107:1123–1128. 13. Weber C, Springer TA. Neutrophil accumulation on activated, surfaceadherent platelets in flow is mediated by interaction of Mac-1 with fibrinogen bound to alpha-IIb beta-3 and stimulated by platelet activating factor. J Clin Invest 1997;100:2085–2093. BRIEF REPORTS 239 14. Sarma J, Laan CA, Alam S, Jha A, Fox KAA, Dransfield I. Increased platelet 17. Lincoff AM, Kereiakes DJ, Mascelli MA, Deckelbaum LI, Barnathan ES, binding to circulating monocytes in acute coronary syndromes. Circulation 2002;105:2166 –2171. 15. Schwarz M, Nordt T, Bode C, Peter K. The GP IIb/IIIa inhibitor abciximab (c7E3) inhibits the binding of various ligands to the leukocyte integrin Mac-1 (CD11b/CD18, alphaMbeta2). Thromb Res 2002;107:121–128. 16. Mickelson JK, Ali MN, Kleiman NS, Lakkis NM, Chow TW, Hughes BJ, Smith CW. Chimeric 7E3 Fab (ReoPro) decreases detectable CD11b on neutrophils from patients undergoing coronary angioplasty. J Am Coll Cardiol 1999; 33:97–106. Patel KK, Frederick B, Nakada MT, Topol EJ. Abciximab suppresses the rise in levels of circulating inflammatory markers after percutaneous coronary revascularization. Circulation 2001;104:163–167. 18. Liuzzo G, Buffon A, Biasucci LM, Gallimore JR, Caligiuri G, Vitelli A, Altamura S, Ciliberto G, Rebuzzi AG, Crea F, et al. Enhanced inflammatory response to coronary angioplasty in patients with severe unstable angina. Circulation 1998;98:2370 –2376. 19. Bonetti PO, Lerman LO, Napoli C, Lerman A. Statin effects beyond lipid lowering—are they clinically relevant? Eur Heart J 2003;24:225–248. Value of Preprocedure Multislice Computed Tomographic Coronary Angiography to Predict the Outcome of Percutaneous Recanalization of Chronic Total Occlusions Nico R. Mollet, MD, Angela Hoye, MB, ChB, MRCP, Pedro A. Lemos, MD, Filippo Cademartiri, MD, Georgios Sianos, MD, Eugene P. McFadden, MB, Gabriel P. Krestin, MD, Patrick W. Serruys, MD, and Pim J. de Feyter, MD, PhD We performed multislice computed tomographic coronary angiography in 45 patients who had chronic total occlusions and were scheduled for percutaneous recanalization. Multivariate analysis identified a blunt stump (by conventional angiography), occlusion length >15 mm, and severe calcification (by multislice computed tomographic coronary angiography) as independent predictors of procedural failure. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:240 –243) ixteen-row multislice spiral computed tomographic (MSCT) coronary angiography has reS cently been shown to allow reliable noninvasive evaluation of coronary morphology.1–3 In the present study, we analyzed the potential of preprocedural MSCT coronary angiography to provide additional information and thus predict the procedural outcome in patients who had chronic total occlusion (CTO) and were referred for percutaneous coronary recanalization. ••• Forty-five patients referred for percutaneous recanalization of Х†1 CTO lesion underwent MSCT coronary angiography before the coronary procedure (median interval 29 days, interquartile range 9 to 53). The diagnosis of CTO was made on diagnostic angiograms that demonstrated complete occlusion of a major epicardial coronary artery, which was deemed to be of Х†3 months’ duration from the date from the previous angiogram, a clinical history of myocardial infarction, or onset of or a severe episode of prolonged anginal From the Department of Cardiology, Thoraxcenter and Department of Radiology, Erasmus Medical Center, Rotterdam, The Netherlands. Dr. de Feyter’s address is: Departments of Cardiology and Radiology, Thoraxcenter, Erasmus Medical Center, Bd 410, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands. E-mail: p.j.defeyter@ erasmusmc.nl. Manuscript received June 28, 2004; revised manuscript received and accepted September 2, 2004. 240 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 chest pain. In addition, inclusion into the study required a serum creatinine level ПЅ120 mmol/L, presence of sinus rhythm, and the ability to hold a breath for 20 seconds. The protocol was approved by the institutional review board, and all patients gave written informed consent. Conventional angiographic assessment was performed by observers who were unaware of the results of MSCT scans. Parameters previously reported to have prognostic importance for procedural failure were assessed: absence of anterograde flow through bridging collaterals, absence of a tapered stump, presence of severe calcification at the occluded segment, side branch at the occlusion site, and tortuosity of the vessel proximal to the occlusion (defined as an angle Пѕ45В° in any projection). Where possible, occlusion length was measured from the view with the longest lesion on quantitative coronary angiography as the distance between a stump and a distal vessel as visualized by anterograde filling through bridging collaterals. In addition, in some other patients, length was determined from the baseline angioplastic procedure film using a bilateral coronary injection. Twenty-two patients who had a heart rate Пѕ65 beats/ min before multislice spiral computed tomography received an oral dose of 100 mg of metoprolol 1 hour before scanning. All examinations were performed with a 16-row MSCT scanner (Sensation 16, Siemens, Forchheim, Germany; collimation 16 П« 0.75 mm, rotation time 420 ms, table feed 3.0 mm/rotation, tube voltage 120 kV, tube current 400 to 450 mA). After intravenous administration of 120 ml of nonionic contrast material (Visipaque 320, Amersham Health, Little Chalfont, United Kingdom), an automatic bolus-tracking technique triggered the start of MSCT scanning. Images were reconstructed with retrospective electrocardiographic gating during the mid- to end-diastolic phase to provide nearly motion-free image quality; additional reconstruc0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.009 Patients’ mean age П® SD was 57.0 П® 10.1 years, 40 (89%) were men, 10 (22%) were diabetic, and 14 (31%) had multivessel disease. ForSuccess Failure ty-seven CTO lesions were treated. (n П 26 lesions) (n П 21 lesions) p Value Angiographic measurement of CTO Mean total procedural time (min) 148 П® 61 148 П® 44 1.0 length was possible in 39 lesions Mean volume of contrast used (ml) 451 П® 258 453 П® 265 1.0 (83%), 31 (66%) from the diagnostic Mean no. of wires 2.0 П® 1.2 2.1 П® 0.8 0.9 film and an additional 8 (17%) from Wire type* Graphix Intermediate 20 (76.9%) 12 (57.1%) 0.2 bilateral injection and assessment of Choice PT Plus 5 (19.2%) 3 (14.3%) 0.8 retrograde collateral filling. The Crosswire NT 8 (30.8%) 10 (47.6%) 0.3 mean length of occlusion was longer Miracle 5 (19.2%) 12 (57.1%) ПЅ0.01 by MSCT coronary angiography Intraluminal 3 (11.5%) 4 (19.0%) 0.5 than by angiography (21.8 П® 18.6 vs Over-the-wire balloon support 15 (57.7%) 17 (81.0%) 0.1 14.6 П® 10.9 mm, respectively). Pro*Wire type not mutually exclusive. cedural data, including type and number of guidewires used, are presented in Table 1. The only differtion windows (e.g., early diastolic phase) were explored ence between success and failure was in the increased use of the Miracle wire in the failure group. Overall when necessary. All MSCT scans were analyzed off-line by op- mean procedural time was 148 П® 53 minutes, with a erators who were blinded to angiographic and pro- mean fluoroscopic time of 47 П® 24 minutes. Overall, cedural data. Parameters similar to those of conven- 45% of interventional procedures failed (Table 2). tional angiography were evaluated: a blunt rather Success versus failure was not dependent on the opthan tapered stump, severe calcification, side branch erator or choice of interventional strategy. At univarat the occlusion site, proximal tortuosity, and oc- iate analysis, the following were associated with proclusion length. Severe calcification was defined as cedural failure: clinical assessment (occlusion the presence of high-density plaques (Х†130 HU) duration Х†9 months), angiographic assessment (lack involving Пѕ50% of the coronary wall on a cross- of anterograde collateral filling, a blunt rather than a sectional image and localized within the occlusion tapered stump, and side branch at the occlusion site), and MSCT coronary angiographic assessment (a blunt stump or occluded segment. All procedures were performed by operators who rather than a tapered stump, severe calcification, and were highly experienced in the treatment of CTOs, occlusion length Пѕ15 mm; Table 2). When analyzed separately, the following “tradiwith the interventional strategy left to the discretion of the operator. Wires were used in a stepwise progres- tional” clinical and angiographic characteristics were sion, starting with a wire that had a relatively less identified as multivariate predictors of procedural failtraumatic tip (Graphix Intermediate, Boston Scientific ure: occlusion duration Пѕ9 months and stump morCorporation, Miami, Florida) or a hydrophilic wire phology (Table 3). Separate multivariate analysis that (Choice PT Plus, Boston Scientific Corporation, or assessed only MSCT coronary angiographic parameCrosswire NT, Terumo Corporation, Tokyo, Japan) ters identified the following predictors: occlusion and progressing to stiffer wires (Miracle, Asahi Intec, length Пѕ15 mm, severe calcification, and stump morNagoya, Japan) and specialized technologies (Safe- phology. Final best model testing for pooled clinical, Cross, Intraluminal Therapeutics, Carlsbad, New angiographic, and MSCT parameters identified a blunt Mexico).4,5 Procedural failure was defined as an in- rather than a tapered stump (by angiography), occlusion length Пѕ15 mm (by MSCT coronary angiograability to cross the occlusion with a guidewire. Multivariate logistic regression analyses were per- phy), and severe calcification (by MSCT coronary formed to identify angiographic and MSCT parame- angiography) as multivariate independent predictors ters associated with procedural failure (all univariate of procedural failure (Figure 1). predictors with a p value Х…0.1 were tested for their ••• multivariate predictive value, and final models were The current selection process of technically appropribuilt by backward stepwise selection). Angiographic ate candidates for percutaneous recanalization of CTO is parameters assessed were those identified in previous based on the evaluation of a relatively restricted number studies6: the occluded artery, duration of occlusion, of clinical and angiographic characteristics. In this study, multivessel disease, anterograde and retrograde collat- we show that noninvasive evaluation of patients who eral filling, type of stump, side branch at the site of have CTO by preprocedural MSCT coronary angiograocclusion, calcific deposits, vessel tortuosity, and oc- phy improves the ability to predict the outcome of a clusion length Пѕ15 mm. The predictive strengths of percutaneous recanalization attempt. Our findings indithe models were evaluated by means of the ПЄ2 log- cate that MSCT coronary angiography may aid in the likelihood statistic, and models’ lack of fit with the therapeutic decision making for patients who have CTO. Hosmer-Lemeshow test, and their global predictive In addition, accurate preprocedural characterization of accuracy were assessed by the C index (area under the CTO features may assist in outlining the therapeutic interventional strategy. receiver-operating characteristic curve). TABLE 1 Procedural Duration and Use of Contrast Material and Guidewires for Percutaneous Intervention of a CTO With Respect to Successful Versus Unsuccessful Recanalization BRIEF REPORTS 241 TABLE 2 Clinical, Angiographic, and MSCT Coronary Angiography Lesion Characteristics (n П 47) Angiography Frequency (%) Right coronary artery Left anterior descending artery Left circumflex artery Occlusion duration Х†9 mo Multivessel disease Anterograde filling Retrograde collateral filling Bridging collaterals Tapered stump Stump morphology not determinable Side branch at occlusion site Severe calcification Vessel tortuosity Occlusion length Пѕ15 mm Occlusion length not determinable Overall Failure Rate (%) MSCT Coronary Angiography OR (95% CI) p Value Frequency (%) Failure Rate (%) OR (95% CI) p Value 0.5 0.6 — — — — — — — — 43 43 50 40 1.45 (0.45–4.66) 0.72 (0.22–2.32) 15 51 34 60 72 38 60 0 43 63 38 32 44 39 25 — 0.92 4.72 0.64 0.27 0.92 0.68 0.12 (0.18–4.64) (1.36–16.39) (0.19–2.20) (0.08–0.94) (0.26–3.32) (0.21–2.25) (0.03–0.45) — 0.9 0.02 0.5 0.04 0.9 0.5 ПЅ0.01 — — — — — — — 60 15 — — — — — — 32 43 — — — — — — 0.16 (0.03–0.73) 0.25 (0.34–1.82) — — — — — — 0.02 0.2 57 34 23 26 17 59 50 64 67 50 4.37 1.39 2.75 3.39 1.29 (1.23–15.54) (0.41–4.65) (0.68–11.14) (0.85–13.48) (0.28–5.94) 0.02 0.2 0.08 0.7 45 38 21 51 0 52 67 50 63 — 1.76 4.44 1.31 4.72 0.3 0.02 0.7 0.02 — 100 45 — — — — (0.55–5.64) (1.27–15.61) (0.32–5.32) (1.36–16.39) — — — CI П confidence interval; MSCT П multislice computed tomographic; OR П odds ratio. TABLE 3 Angiographic and MSCT Coronary Angiographic Multivariate Predictors of Procedural Failure for Chronic Total Occlusion Variable Clinical/angiographic predictors Occlusion duration Пѕ9 mo Tapered stump Constant MSCT coronary angiography predictors Occlusion length Пѕ15 mm Severe calcification Stump morphology Blunt Tapered Not determinable Constant Clinical/angiographic П© MSCT coronary angiographic predictors Tapered stump* Occlusion length Пѕ15 mm Severe calcification Constant Coefficient 1.27 ПЄ1.93 0.24 1.86 2.49 — 1 (reference) ПЄ2.19 ПЄ2.65 ПЄ0.45 ПЄ2.43 2.17 2.03 ПЄ0.67 Wald’s Chi-square 3.30 7.46 0.13 5.21 6.51 5.63 — 5.23 3.46 0.26 7.98 6.16 5.18 0.74 DF p Value 1 1 1 1 1 2 — 1 1 1 1 1 1 1 0.07 ПЅ0.01 0.7 0.02 0.01 0.06 — 0.02 0.06 0.6 ПЅ0.01 0.01 0.02 0.4 OR (95% CI) HosmerLemeshow Test ПЄ2 Log Likelihood DF p Value C Index 50.0 2 0.66 0.80 44.2 6 0.99 0.84 41.0 5 0.60 0.85 3.56 (0.90–14.02) 0.15 (0.04–0.58) — 6.39 (1.30–31.41) 12.01 (1.78–81.1) — — 0.11 (0.02–0.73) 0.07 (0.00–1.15) — 0.09 (0.02–0.48) 8.77 (1.58–48.76) 7.62 (1.33–43.74) — *A ПЄ2 log-likelihood change in the global model if 1 variable is removed: tapered stump ПЄ10.7 (p ПЅ0.01 for change), occlusion length ПЄ8.2 (p ПЅ0.01), and calcification ПЄ6.6 (p П 0.01 for change). DF П degrees of freedom; other abbreviations as in Table 2. MSCT coronary angiography of CTOs adds important information compared with “conventional” coronary angiography. The length of the occluded segment has long been identified as an important predictor of failed recanalization. However, accurate measurement of lesion length using conventional angiography may be difficult, mainly due to foreshortening, calibration limita242 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 tions, and lack of visualization of the distal vessel in the absence of collateral filling. In the present study, lesion length could be measured in only 66% of diagnostic films. Conversely, MSCT coronary angiography allowed reliable 3-dimensional length measurement of coronary segments.7 In the present series, when angiographic and MSCT coronary angiographic occlusion lengths were JANUARY 15, 2005 phy for interventions scheduled to be performed shortly after scanning. However, the elective nature of CTO recanalization angioplasty allows a safe time lag between these procedures. In our study, multislice computed tomography was performed Пі1 month before coronary intervention. The relatively high radiation exposure during MSCT coronary angiography, reportedly between 6.7 and 13.0 mSv,8 –10 remains a matter of concern. However, MSCT coronary angiography may optimize therapeutic strategy (e.g., calcifications may require intraluminal techniques), resulting in shorter procedures. MSCT coronary angiography is currently feasible for selected patients, and further studies are needed to evaluate its value in a more general patient population. 1. Nieman K, Cademartiri F, Lemos PA, Raaijmakers R, Pattynama PM, de Feyter PJ. Reliable noninvasive coronary angiography with fast submillimeter multislice spiral computed tomography. Circulation 2002;106:2051–2054. 2. Ropers D, Baum U, Pohle K, Anders K, Ulzheimer S, Ohnesorge B, Schlundt C, Bautz W, Daniel WG, Achenbach S. Detection of coronary artery stenoses with thin-slice multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation 2003;107:664 – 666. FIGURE 1. (A, inset) CTO of the left anterior descending coronary artery with favor3. Mollet NR, Cademartiri F, Nieman K, Saia F, Lemos able invasive angiographic CTO characteristics, tapered stump, absence of calcificaPA, McFadden EP, Pattynama PMT, Serruys PW, Krestin GP, De Feyter PJ. Multislice spiral CT coronary angiogtions, and occlusion length <15 mm (arrowheads) in a patient who had been reraphy in patients with stable angina pectoris. J Am Coll ferred for percutaneous recanalization. Volume-rendered MSCT image that provides Cardiol 2004;43:2265–2270. a 3-dimensional overview of the coronary arteries, and (B) maximum intensity pro4. Ng W, Chen WH, Lee PY, Lau CP. Initial experience jection of the same MSCT image show a severely calcified occlusion stump (arrows). and safety in the treatment of chronic total coronary (C) Curved multiplanar reconstructed MSCT image of the left anterior descending occlusions with a new optical coherent reflectometryguided radiofrequency ablation guidewire. Am J Cardiol coronary artery shows a severely calcified stump (arrow) and an occluded segment 2003;92:732–734. (arrowheads). Collateral filling is clearly visible distal to the occlusion. This percuta5. Hoye A, Onderwater E, Cummins P, Sianos G, Serruys neous attempt at recanalization of the CTO was unsuccessful. P. Improved recanalization of chronic total coronary occlusions using an optical coherence reflectometry-guided guidewire. Catheter Cardiovasc Interv 2004;63:158 –163. 6. Puma JA, Sketch MH Jr, Tcheng JE, Harrington RA, measured, results of the MSCT coronary angiography HR, Stack RS, Califf RM. Percutaneous revascularization of chronic were “longer,” perhaps reflecting the inaccuracy of quan- Phillips coronary occlusions: an overview. J Am Coll Cardiol 1995;26:1–11. titative coronary angiography as previously described. 7. Ferencik M, Moselewski F, Ropers D, Hoffmann U, Baum U, Anders K, Moreover, MSCT coronary angiography allows evalua- Pomerantsev EV, Abbara S, Brady TJ, Achenbach S. Quantitative parameters of quality in multidetector spiral computed tomographic coronary imaging tion of the morphology of the occlusion trajectory, in- image with submillimeter collimation. Am J Cardiol 2003;92:1257–1262. cluding detailed delineation of coronary calcification. 8. Morin RL, Gerber TC, McCollough CH. Radiation dose in computed tomogLong occlusions and severe calcifications on MSCT cor- raphy of the heart. Circulation 2003;107:917–922. 9. Trabold T, Buchgeister M, Kuttner A, Heuschmid M, Kopp AF, Schroder S, onary angiograms were found to be important predictors Claussen CD. Estimation of radiation exposure in 16-detector row computed tomogof procedural failure, whereas neither feature was iden- raphy of the heart with retrospective ECG-gating. Rofo Fortschr Geb Rontgenstr tified as an independent predictor on conventional angio- Neuen Bildgeb Verfahr 2003;175:1051–1055. 10. Hunold P, Vogt FM, Schmermund A, Debatin JF, Kerkhoff G, Budde T, grams. Erbel R, Ewen K, Barkhausen J. Radiation exposure during cardiac CT: The need to use contrast material may pose a effective doses at multi-detector row CT and electron-beam CT. Radiology limitation to preprocedural MSCT coronary angiogra- 2003;226:145–152. BRIEF REPORTS 243 Angiographic and Clinical Outcomes of Polytetrafluoroethylene-Covered Stent Use in Significant Coronary Perforations Hung Ly, MD, Jean-Pierre S. Awaida, MSc, MD, Jacques LespГ©rance, Luc Bilodeau, MD Coronary perforations remain a rare but life-threatening complication of percutaneous coronary intervention. In the setting of complex coronary lesions, 14 of 25 consecutive perforations related to percutaneous coronary intervention were managed with stents covered with polytetrafluoroethylene (PTFE), and 11 were managed conventionally with reversal of anticoagulation and prolonged balloon inflation. Procedural success was achieved in 71.4% (10 of 14 perforations) of the PTFE arm compared with 27.3% (3 of 11 perforations) in the standard management arm (p ‫ ؍‬0.047). Smaller final percent diameter stenosis and postprocedural cessation of dye extravasation were achieved with PTFE-covered stents. No delayed cardiac tamponade or in-hospital mortality was reported among patients in the PTFE group. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:244 –246) ignificant coronary perforations remain a rare but life-threatening complication during percutaneous S coronary intervention (PCI) associated with urgent bypass surgery in Пѕ50% of cases and high mortality rates.1– 4 Stents covered with polytetrafluoroethylene (PTFE) have emerged as a new strategy to seal such perforations.5 This study reports angiographic and clinical outcomes of PCI-induced coronary perforations after using PTFE-covered stents. ••• The JOSTENT GraftMaster coronary stent (Abbott Vascular Devices, Redwood City, California) is a 6Fr compatible, balloon-expandable, slotted-tube stent with a layer of porous PTFE sandwiched between 2 coaxial 316-L stainless steel stents. Overall wall thickness reaches 0.30 mm, and the device crossing profile is 1.6 mm. Diameters from 2.5 to 5.0 mm may be achieved. Data related to all consecutive coronary perforations from databases of the catheterization laboratory of the Montreal Heart Institute (Montreal, Quebec, Canada) were included in this study. All coronary procedures were performed according to current standards of practice. All patients were pretreated with From the Department of Medicine, Division of Interventional Cardiology and the Department of Radiology, Montreal Heart Institute, and the University of Montreal, Montreal, Quebec, Canada. Dr. Bilodeau’s address is: Cardiac Catheterization Laboratory, Montreal Heart Institute, 5000 Belanger Street, Montreal, Quebec H1T 1C8, Canada. E-mail: [email protected]. Manuscript received April 18, 2004; revised manuscript received and accepted September 8, 2004. 244 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, and oral aspirin and intravenous unfractionated heparin. An activated clotting time of 300 seconds was chosen for standard PCI, whereas 250 seconds was preferred if an intravenous glycoprotein IIb/IIIa inhibitor was used as adjunctive therapy. In the event of a coronary perforation, immediate reversal of anticoagulants was performed. Choice of alternatives to seal perforations was left to the operator’s discretion and included deployment of a PTFE-covered stent as first-line strategy or prolonged balloon inflation with or without bare metal stenting. If a PTFE-covered stent was deployed, aspirin and clopidogrel were given for Х†3 months. Six-month clinical follow-up was scheduled. Procedural success was defined as successful delivery and deployment of the PTFE-covered or bare metal stent to the perforation site, complete cessation of dye extravasation at the site of rupture at the end of the procedure (after stenting or after prolonged balloon inflation), and postprocedural Thrombolysis In Myocardial Infarction grade 3 flow. Angiographic analysis was performed with a validated automated edge-detection system (CMS 5.2, MEDIS, Leiden, The Netherlands).6 Lesion morphology was classified according to the modified classification of the American College of Cardiology/American Heart Association.7 The Thrombolysis In Myocardial Infarction flow grade was evaluated as previously described.8 The balloon-to-artery ratio was defined as the ratio of maximal balloon diameter to the interpolated mean reference vessel diameter of the target vessel. Coronary perforations were defined as extravasation of dye (blood) from the coronary artery, and extent of perforation was reported according to the classification of Ellis et al2: type 1, defined as an extraluminal crater without extravasation (fully contained perforation); type 2, defined as a pericardial or myocardial blush without contrast jet extravasation (limited extravasation); and type 3, defined as brisk extravasation through a frank perforation with a Пѕ1-mm exit hole (Figure 1). Coronary perforations were considered significant if they were type 2 or 3. Statistical analysis was performed with STATA 8.0 (STATA Corp., College Station, Texas). Continuous variables are expressed as mean П® SD and were analyzed with Wilcoxon’s (Mann-Whitney) rank sum test. Dichotomous variables are reported as frequencies (percentages) and were analyzed with Fisher’s exact test for proportions. Between April 1999 and June 2003, 9,382 patients underwent PCI at the cardiac catheterization laboratory of the Montreal Heart Institute. During that period, 25 consecutive operator-diagnosed coronary per0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.010 FIGURE 1. Mid right coronary artery Ellis 3 type perforation. (A) Before and (B) after PTFE covered stent. (age, gender, co-morbidities, smoking status, and reasons for PCI) was noted across groups. Most perforaPTFE Stent Group tions resulted from balloon or stent Procedural Procedural Bare Metal oversizing during deployment (11 of Successes Failures Stent Group 14 patients in the PTFE group and 8 (n П 10) (n П 4) (n П 11) of 11 patients in the bare metal Age (yrs) 70.3 П® 10.5 72.0 П® 16.4 65.7 П® 10.1 group). The mean balloon-to-artery Men 2 (50) 6 (60) 7 (64) ratio was 1.37 П® 0.09. Other causes Diagnosis Stable angina 2 (20) 0 (0) 4 (36) of perforations were guidewire damUnstable angina 4 (40) 2 (50) 4 (36) age to the vessel (2 of 14 in the PTFE Myocardial infarction 4 (40) 2 (50) 2 (18) group and 3 of 11 in the bare metal GP IIa/IIIb inhibitors 5 (50) 1 (25) 4 (36) group) and rotational atherectomy (1 AHA lesion complexity A1/B1 1 (10) 2 (50) 0 (0) of 14 in the PTFE group and 0 of 11 B2 3 (30) 0 (0) 6 (55) in the bare metal group). C 6 (60) 2 (50) 5 (45) Baseline angiographic characterisRVD (mm) 3.13 П® 0.71 2.79 П® 0.38 2.51 П® 0.81 tics were similar across groups (target MLD (mm) 0.93 П® 0.18 0.88 П® 0.30 0.66 П® 0.17 vessel, bifurcation lesion, eccentricity, BAR Пѕ1.1 7 (70) 1 (25) 6 (55) Perforation type* tortuosity, presence of thrombus, le1 1 (10) 0 (0) 0 (0) sion ulceration, or calcification), and 2 2 (20) 2 (50) 5 (45) chronic total occlusions characterized 3 7 (70) 2 (50) 6 (55) a minority of treated lesions (1 in the Postprocedural TIMI grade 3 flow 10 (100) 3 (75) 8 (73) Cessation of dye extravasation 0 (0) 1 (25) 6 (55) PTFE group vs 3 in the bare metal Postprocedural increase in cardiac 0 (0) 1 (25) 5 (45) group). Smaller postprocedural mean enzymes percent diameter stenoses were Values are mean П® SD or numbers of patients or lesions (percentages). achieved in the PTFE group than in the *Classification according to Ellis et al.2 standard management group (69.1% to AHA П American Heart Association; BAR П balloon-to-artery ratio; GP П glycoprotein; MLD П 12.7% vs 74.8% to 35.4%, respecminimal luminal diameter; RVD П reference vessel diameter; TIMI П Thrombolysis In Myocardial tively, p П 0.002). Infarction. Procedural success was achieved in more patients who had been treated with a PTFE-covered stent than in the forations were observed (0.3%); all were included in standard treatment group (10 of 14 patients, 71.4%, vs 3 this analysis. Decision to use PTFE-covered stents to of 11 patients, 27.3%, respectively, p П 0.047; Table 1). seal ruptures was left to the operator’s discretion. Use Reasons for failure of the PTFE-covered stent included of the PTFE-covered stent was attempted in a sub- persistence of dye extravasation despite successful stent group of 14 patients. Selected clinical and angio- deployment in 1 patient and lack of stent delivery to the perforation site in 3 patients, despite aggressive guiding graphic data are listed in Table 1. No difference in baseline clinical characteristics support and use of a buddy-wire technique. TABLE 1 Clinical and Angiographic Data BRIEF REPORTS 245 Although comparable rates of final Thrombolysis In Myocardial Infarction grade 3 flow were achieved, most notable was the more common persistence of postprocedural dye extravasation in the standard management arm (Table 1). Cardiac tamponade (immediate or delayed) occurred in 1 patient in the PTFE arm compared with 4 patients in the standard management group (p П 0.133). Emergency referral for cardiac surgery was required in 1 patient in the PTFE arm and 2 patients in the standard management arm. After excluding patients who had been referred for cardiac surgery, postprocedural increases in cardiac enzyme were documented more frequently in the standard management arm (5 patients in the bare metal group vs 1 patient in the PTFE group, p П 0.056). The only in-hospital death from perforation (due to hemodynamic collapse) was in the standard management arm. At 6-month follow-up, 13 of 14 patients in the PTFE group remained asymptomatic or had mild angina on exertion (Canadian Cardiovascular Society class 1); 1 patient underwent clinically driven target site revascularization for restenosis. In comparison, 3 of 10 survivors in the standard management group were readmitted for acute coronary syndromes, whereas the other patients remained asymptomatic or had mild to moderate angina on exertion (Canadian Cardiovascular Society class 1 or 2). Of these patients, 2 received medical therapy and 1 patient was referred for surgical revascularization for persistent anginal symptoms. ••• In this historical case-control study, we found that PTFE-covered stents were significantly better than bare metal stents at sealing coronary perforations. They were associated with greater procedural success, more frequent cessation of dye extravasation after PCI, larger postprocedural minimal luminal diameter, and favorable in-hospital and 6-month clinical outcomes. Previous management algorithms for coronary perforations have restricted the PTFE-covered stent to a second-line strategy if conventional options were unsuccessful.9,10 Use of PTFE-covered stents for procedure-related coronary perforations has been shown to be feasible. Briguori et al5 reported a 100% success rate in their series of 11 patients who had been treated with these stents compared with a 40% success rate in 246 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 17 patients who had been treated with bare metal stents to seal coronary ruptures refractory to prolonged balloon inflation and anticoagulation reversal. Clinical outcomes showed a significant benefit of covered stents over bare metal stents for all major adverse cardiac events, except for mortality rate (18% vs 88%, p ПЅ0.001). In-stent restenosis was documented in 29% of patients whose treatment was successful during the 14-month follow-up period. The present study reports a procedural success rate of 71% with 1 patient who developed in-stent restenosis. Although limited by a small sample, the results from this study show that PTFE-covered stents are a safe and feasible first-line strategy to seal significant coronary perforations in the setting of complex coronary lesions PCI. 1. Stankovic G, Orlic D, Corvaja N, Airoldi F, Chieffo A, Spanos V, Montorfano M, Carlino M, Finci L, Sangiorgi G, Colombo A. Incidence, predictors, inhospital, and late outcomes of coronary artery perforations. Am J Cardiol 2004; 93:213–216. 2. Ellis SG, Ajluni S, Arnold AZ, Popma JJ, Bittl JA, Eigler NL, Cowley MJ, Raymond RE, Safian RD, Whitlow PL. Increased coronary perforation in the new device era. Incidence, classification, management, and outcome. Circulation 1994;90:2725–3270. 3. Gunning MG, Williams IL, Jewitt DE, Shah AM, Wainwright RJ, Thomas MR. Coronary artery perforation during percutaneous intervention: incidence and outcome. Heart 2002;88:495– 498. 4. Fasseas P, Orford JL, Panetta CJ, Bell MR, Denktas AE, Lennon RJ, Holmes DR, Berger PB. Incidence, correlates, management, and clinical outcome of coronary perforation: analysis of 16,298 procedures. Am Heart J 2004;147:140 – 145. 5. Briguori C, Nishida T, Anzuini A, Di Mario C, Grube E, Colombo A. Emergency polytetrafluoroethylene-covered stent implantation to treat coronary ruptures. Circulation 2000;102:3028 –3031. 6. Reiber H, Van de Zwet PMJ, Von Land CD, Koning G, Meurs BV, Buis B, Van Voorthuisen AE. Quantitative coronary arteriography: equipment and technical requirements. In: Reiber JHC, Serruys P, eds. Advances in Quantitative Coronary Arteriography. Norwell, MA: Kluwer Academic Publishers, 1993:75– 111. 7. Ellis SG, Vandormael MG, Cowley MJ, DiSciascio G, Deligonul U, Topol EJ, Bulle TM. Coronary morphologic and clinical determinants of procedural outcome with angioplasty for multivessel coronary disease. Implications for patient selection. Multivessel Angioplasty Prognosis Study Group. Circulation 1990;82: 1193–1202. 8. TIMI Study Group. The Thrombolysis in Myocardial Infarction (TIMI) trial. Phase I findings. N Engl J Med 1985;312:932–936. 9. Dippel EJ, Kereiakes DJ, Tramuta DA, Broderick TM, Shimshak TM, Roth EM, Hattemer CR, Runyon JP, Whang DD, Schneider JF, Abbottsmith CW. Coronary perforation during percutaneous coronary intervention in the era of abciximab platelet glycoprotein IIb/IIIa blockade: an algorithm for percutaneous management. Catheter Cardiovasc Interv 2001;52:279 –286. 10. Mulvihill NT, Boccalatte M, Sousa P, Farah B, Laborde JC, Fajadet J, Marco J. Rapid sealing of coronary perforations using polytetrafluoroethylene-covered stents. Am J Cardiol 2003;91:343–346. JANUARY 15, 2005 Role of Postoperative Vasopressor Use in Occurrence of Atrial Fibrillation After Coronary Artery Bypass Grafting Vikrant Salaria, MD, Nirav J. Mehta, MD, Syed Abdul-Aziz, Syed M. Mohiuddin, MD, DSc, and Ijaz A. Khan, MD Vasopressor use is common after coronary artery bypass grafting surgery. This study evaluated the role of postoperative vasopressor use as a predictor of occurrence of atrial fibrillation after coronary artery bypass grafting and demonstrates that vasopressor use is an independent predictor of such an occurrence. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:247–249) he most effective way of preventing complications of postoperative atrial fibrillation (AF) is to preT vent their occurrence. Age, atrial ischemia, intraoperative elements, postoperative fluctuations in autonomic tone, and postoperative sympathetic activation have been suggested as causative factors of AF after coronary artery bypass grafting (CABG) surgery.1–12 Sympathomimetic (adrenergic) agents are commonly used as vasopressors after CABG. Use of these agents sets a higher catecholamine stage, which, theoretically, may trigger postoperative AF. Practically, it is not known whether use of vasopressor agents after CABG is a predictor of an occurrence of postoperative AF. This study evaluated the role of postoperative vasopressor use as a predictor of AF after CABG. ••• The study protocol was approved by the institutional review board of Creighton University (Omaha, Nebraska). All patients who were Пѕ18 years of age and underwent CABG at Creighton University Medical Center between June 1999 and December 2000 were evaluated for inclusion into the study. Patients who had previous AF and those who were using antiarrhythmic agents (except вђ¤ blockers) before or after CABG were excluded. One hundred thirty-one patients comprised the study population. Data collection was retrospective. AF was defined as an irregular heart rate without P waves on a 12-lead electrocardiogram or continuous rhythm monitoring. Postoperative AF was defined as development of AF any time after surgery during the index hospitalization. The occurrence of postoperative AF was determined. Data were collected and tabulated for the following variables: postoperative vasopressor use, age, gender, hypertension, preoperative echocardioFrom the Division of Cardiology, Creighton University School of Medicine, Omaha, Nebraska; and the Division of Cardiology, University of Maryland School of Medicine, Baltimore, Maryland. Dr. Khan’s address is: Division of Cardiology, University of Maryland School of Medicine, 22 South Greene Street, S3B06, Baltimore, Maryland 21201. E-mail: [email protected]. Manuscript received May 11, 2004; revised manuscript received and accepted September 7, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, graphic parameters (left atrial size, left ventricular ejection fraction, and left ventricular hypertrophy) that were defined according to standard criteria, cardioplegia use, aorta cross-clamp time, bypass time, perioperative вђ¤-blocker use, postoperative cardiac pacing (atrial), postoperative serum levels of magnesium, and length of stay in an intensive care unit. These variables were examined for the prediction of postoperative AF. Continuous variables were expressed as means П® SD and analyzed with Student’s unpaired t test. Categorical variables were expressed as numbers and percentages and analyzed with chi-square statistics or Fisher’s exact test, as appropriate. Stepwise logistic regression analysis was used for multivariate analysis to determine independent predictors for occurrence of postoperative AF. A p value Х…0.05 was considered statistically significant. SPSS 7.0 (SPSS Inc., Chicago, Illinois) was used for statistical analyses. Among 131 patients, postoperative AF occurred in 30, with an incidence of 23%. Vasopressor agents were used in 74 patients (56%). The vasopressors used were dopamine, dobutamine, and phenylephrine. Dopamine was used in 21 patients (16%), dobutamine in 40 (31%), and phenylephrine in 13 (10%). Postoperative AF occurred in 22 of 74 patients (30%) who used vasopressors and 8 of 57 patients (14%) who did not use vasopressors (p П 0.03). Baseline and clinical characteristics with univariate analyses are presented in Table 1. By univariate analyses, postoperative vasopressor use (p П 0.03) and age (p ПЅ0.001) were predictors of developing AF after CABG (Table 1). In contrast, lack of vasopressor use (p П 0.02) and cardiac (atrial) pacing (p ПЅ0.001) were predictors of not developing AF after CABG. Among 131 patients, CABG was performed off-pump in 43 patients. There was no significant difference in the occurrence of AF with off-pump versus on-pump CABG (10 of 43 patients, 23% vs 20 of 88 patients, 23%, p П NS). According to stepwise logistic regression analysis (Table 2), vasopressor use was an independent predictor of developing postoperative AF (odds ratio 2.783, 95% confidence interval 1.073 to 7.220, p П 0.03). Age was the only other independent predictor (odds ratio 1.085, 95% confidence interval 1.038 to 1.135, p ПЅ0.001). Among various vasopressor drugs, use of those with predominantly вђ¤-1–adrenergic affinity was associated with a higher incidence of postoperative AF (dopamine 43% and dobutamine 30% vs phenylephrine 8%, p П 0.05; Table 3). ••• Our study demonstrates that use of vasopressors is an independent predictor of development of AF after 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.011 247 TABLE 1 Atrial Fibrillation After Coronary Artery Bypass Grafting Variables Postoperative vasopressor use Age (yrs) Men/women Hypertension Left ventricular hypertrophy Left ventricular ejection fraction (%) Left atrial diameter (mm) Off-pump surgery On-pump surgery Cardioplegia use Aorta cross-clamp time (min) Bypass time (min) Perioperative вђ¤-blocker use Postoperative cardiac (atrial) pacing Postoperative cardiac index (L/m2) Postoperative serum magnesium (mEq/L) Hours in intensive care unit AF Present (n П 30) AF Absent (n П 101) p Value 22 (73%) 74.6 П® 9.3 19 (63%)/11 (37%) 26 (87%) 15 (50%) 47.0 П® 16.7 42.7 П® 5.8 10 (33%) 20 (67%) 29 (97%) 74.9 П® 65.0 133.3 П® 92.4 27 (90%) 21 (70%) 2.67 П® 0.66 1.95 П® 0.27 130.5 П® 107.4 52 (52%) 64.8 П® 12.2 74 (73%)/27 (27%) 87 (86%) 37 (37%) 46.3 П® 13.5 41.8 П® 5.4 33 (33%) 68 (67%) 97 (96%) 66.7 П® 26.4 110.4 П® 44.8 96 (95%) 100 (100%) 2.60 П® 0.62 1.89 П® 0.28 97.3 П® 80.4 0.02 ПЅ0.001 0.29 0.94 0.19 0.82 0.55 0.88 0.88 0.87 0.59 0.29 0.31 ПЅ0.001 0.59 0.31 0.07 Values are numbers of patients (percentages) or mean П® SD. operative trauma and from being in an intensive care unit, which are potent stimulators of the sympathetic Variables B SE of B OR (95% CI) p Value* nervous system. That the postoperative state is hyperadrenergic has been Vasopressor use 1.024 0.486 2.783 (1.073–7.220) 0.03 proved by marked postoperative inAge (yrs) 0.082 0.023 1.085 (1.038–1.135) ПЅ0.001 Constant ПЄ7.578 1.721 ПЅ0.001 creases in norepinephrine levels in mixed venous samples after sur*Overall model significance, p ПЅ0.001. gery,11,14 because the predominant CI П confidence interval; OR П odds ratio. source of circulating norepinephrine in humans is from sympathetic nerve endings.15 A significant association has also been shown between postoperative mixed TABLE 3 Incidence of Atrial Fibrillation by Vasopressor Agent venous norepinephrine levels after cardiac surgery and Incidence of Agents Used development of postoperative AF, suggesting that Vasopressor AF by Agent Within AF Group sympathetic activation may be an important factor in most such patients in whom postoperative AF develNone 8/57 (14%) 8/30 (27%) Dopamine 9/21 (43%)* 9/30 (30%) ops.11 In another study that analyzed heart rate variDobutamine 12/40 (30%)* 12/30 (40%) ability, a shift in autonomic balance that resulted in Phenylepherine 1/13 (8%)* 1/30 (3%) enhanced sympathetic tone and a loss of vagal tone *p П 0.05 for dopamine or dobutamine versus phenylepherine. was demonstrated before the onset of postoperative AF.12 In brief, the postoperative hyperadrenergic state is a strong factor responsible for the precipitation of CABG. Moreover, the incidence of postoperative AF AF after surgery because AF usually resolves spontawas higher with the use of dopamine and dobutamine, neously once the provocative hyperadrenergic state is which are predominantly вђ¤-1–adrenergic receptor no longer present. Use of adrenergic vasopressors agonists, than with the use of phenylephrine, which in would enhance the hyperadrenergic postoperative therapeutic doses is predominantly an agonist of вђЈ-adrenergic receptors and lacks direct stimulatory effects state after CABG and result in a greater chance of on atrial myocardium. This is the first study to eval- developing AF. Age has been consistently shown as an indepenuate the role of vasopressor use in the development of dent predictor of postoperative AF,1–5 and the present AF after CABG. These findings indicate that vasostudy substantiates this finding. Age-related structural pressor use after CABG should be limited to a real and hemodynamic changes in atrial dilation and atrial necessity. substrate and higher atrial pressures due to age-related Excessive use of catecholamines enhances trigslow ventricular relaxation might be plausible mechgered activity and automaticity, the triggers for initiation of AF.13 Further, catecholamines shorten atrial anisms of higher incidences of postoperative AF in refractoriness, but in a nonhomogenous pattern, thus older patients.16,17 Further, advancing age has been favoring continuation of AF.13 A postoperative state is associated with increased sympathetic outflow and clearly a hyperadrenergic condition that results from increased levels of circulating norepinephrine.18 TABLE 2 Independent Predictors of Atrial Fibrillation After Coronary Artery Bypass Grafting 248 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 1. Mathew JP, Parks R, Savino JS, Friedman AS, Koch C, Mangano DT, Browner 10. Abreu JE, Reilly J, Salzano RP, Khachane VB, Jekel JF, Clyne CA. Comparison WS. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. JAMA 1996;276:300 –306. 2. Funk M, Richards SB, Desjardins J, Bebon C, Wilcox H. Incidence, timing, symptoms, and risk factors for atrial fibrillation after cardiac surgery. Am J Crit Care 2003;12:424 – 433. 3. Aranki SF, Shaw DP, Adams DH, Rizzo RJ, Couper GS, VanderVliet M, Collins JJ, Cohn LH, Burstin HR. Predictors of atrial fibrillation after coronary artery surgery: current trends and impact on hospital resources. Circulation 1996;94:390 –397. 4. Mathew JP, Fontes ML, Tudor IC, Ramsay J, Duke P, Mazer CD, Barash PG, Hsu PH, Mangano DT. A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA 2004;291:1720 –1729. 5. Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993;56:539 –549. 6. Taylor GJ, Malik SA, Colliver JA, Dove JT, Moses HW, Mikell FL, Batchelder JE, Schneider JA, Wellons HA. Usefulness of atrial fibrillation as a predictor of stroke after isolated coronary bypass grafting. Am J Cardiol 1987; 60:905–907. 7. Villareal RP, Hariharan R, Liu BC, Kar B, Lee VV, Elayda M, Lopez JA, Rasekh A, Wilson JM, Massumi A. Postoperative atrial fibrillation and mortality after coronary artery bypass surgery. J Am Coll Cardiol 2004;43:742–748. 8. Crosby LH, Pifalo WB, Woll KR, Burkholder JA. Risk factors for atrial fibrillation after coronary artery bypass grafting. Am J Cardiol 1990;66:1520 –1522. 9. Kolvekar S, D’Souza A, Akhatar P, Reek C, Garratt C, Spyt T. Role of atrial ischemia in development of atrial fibrillation following coronary artery bypass surgery. Eur J Cardiothorac Surg 1997;11:70 –75. of frequencies of atrial fibrillation after coronary artery bypass grafting with and without use of cardiopulmonary bypass. Am J Cardiol 1999;83:775–776. 11. Kalman JM, Munawar M, Howes LG, Louis WJ, Buxton BF, Gutteridge G, Tonkin AM. Atrial fibrillation after coronary artery bypass grafting is associated with sympathetic activation. Ann Thorac Surg 1995;60:1709 –1715. 12. Dimmer C, Tavernier R, Gjorgov N, Nooten GV, Clement DL, Jordaens L. Variation of autonomic tone preceeding onset of atrial fibrillation after coronary artery bypass grafting. Am J Cardiol 1998;82:22–25. 13. Waldo AL. Mechanisms of atrial fibrillation, atrial flutter, and ectopic atrial tachycardia. Circulation 1987;75(suppl 3):37– 40. 14. de Leeuw PW, van der Starre PJ, Harinck-de Weerdt JE, de Bos R, Tchang PT, Birkenhager WH. Humoral changes during and following coronary bypass surgery: relationship to postoperative blood pressure. J Hypertens 1983; 1:52–54. 15. Goldstein DS, McCarty R, Polinsky RJ, Kopin IJ. Relationship between plasma norepinephrine and sympathetic neural activity. Hypertension 1983;5: 552–559. 16. Manyari DE, Patterson C, Johnson D, Melendez L, Kostuk WJ, Cape RD. Atrial and ventricular arrhythmias in asymptomatic active elderly subjects: correlation with left atrial size and left ventricular mass. Am Heart J 1990;119: 1069 –1076. 17. Aronow WS. The older man’s heart and heart disease. Med Clin North Am 1999;83:1291–1303. 18. Hoeldtke RD, Cilmi KM. Effects of aging on catecholamine metabolism. J Clin Endocrinol Metab 1985;60:479 – 484. Reaching Recommended Lipid and Blood Pressure Targets With Amlodipine/Atorvastatin Combination in Patients With Coronary Heart Disease Jean-Francois Dorval, MD, Todd Anderson, MD, Jean Buithieu, MD, Sammy Chan, MD, Stuart Hutchison, MD, Thao Huynh, MD, Jean Jobin, PhD, Eva Lonn, MD, Paul Poirier, MD, Lawrence Title, MD, Ann Walling, MD, Thang Tran, MSc, Ghyslain Boudreau, PhD, Francois Charbonneau, MD, and Jacques Genest, MD The effects of combined atorvastatin and amlodipine on blood pressure (BP) and low-density lipoprotein (LDL) cholesterol levels were investigated in 134 patients with documented coronary heart disease treated for 1 year. BP at baseline was 128 ШЋ 15/79 ШЋ 9 mm Hg and was controlled by the treating physician; no calcium channel blockers were allowed. Baseline means for plasma cholesterol were 6.4 ШЋ 1.1 mmol/L (147 ШЋ 39 mg/dl), triglycerides 2.0 ШЋ 0.9 mmol/L (177 ШЋ 88 mg/dl), LDL cholesterol 4.4 ШЋ 1.0 mmol/L (170 ШЋ 39 mg/dl), and high-density lipoprotein cholesterol 1.2 ШЋ 0.3 mmol/L (46 ШЋ 12 From McGill University Health Center, Montreal, QuГ©bec; Department of Cardiovascular Sciences and the Libin Cardiovascular Institute of Alberta, Calgary, Alberta; Vancouver General Hospital, Vancouver, British Columbia; Toronto Hospital Network, Toronto, Ontario; UniversitГ© Laval, Institut de Cardiologie de QuГ©bec, QuГ©bec; Department of Medicine, Division of Cardiology, McMaster University, Hamilton, Ontario; Queen Elizabeth II Medical Center, Halifax, Nova Scotia; Jewish General Hospital, Montreal, Quebec; Pfizer Canada, Montreal, Quebec; and Foothills Medical Centre, Calgary, Alberta, Canada. This study was supported by an unrestricted grant from Pfizer. Dr. Genest’s address is: Division of Cardiology, McGill University Health Center/Royal Victoria Hospital, 687 Pine Avenue West, MontrГ©al, Quebec, H3A 1A1, Canada. E-mail: [email protected]. Manuscript received July 7, 2004; revised manuscript received and accepted August 31, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 mg/dl). Patients were all given atorvastatin 10 mg, then increased to 80 mg if the LDL cholesterol was <2.5 mmol/L (100 mg/dl). At 3 months, patients were randomized to amlodipine 10 mg or placebo. Plasma LDL cholesterol was decreased by 50%, and the LDL cholesterol target of <2.5 mmol/L was achieved in 81% of the patients. BP targets were achieved in 69% of the atorvastatin Ш‰ placebo group, versus 96% in the atorvastatin Ш‰ amlodipine group (p ‫ ؍‬0.0002). With use of combination atorvastatin Ш‰ amlodipine at doses ranging from 10 to 80 mg and 5 to 10 mg, respectively, recommended therapeutic goals were reached in most select subjects with coronary artery disease who were concomitantly receiving aspirin and antihypertensive therapy. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:249 –253) n the present study, we investigated the effects of atorvastatin and amlodipine in subjects with estabIlished and stable coronary heart disease (CHD) requiring a statin to decrease low-density lipoprotein (LDL) cholesterol and who had adequate blood pressure control at baseline, as judged by the treating physician. We used a conditional titration with atorvastatin 10 or 80 mg to reach a LDL target value of Х…2.5 mmol/L, 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.012 249 atorvastatin was increased to 80 mg/ day. At 3 months, patients were ranAtorvastatin П© Atorvastatin П© domized to amlodipine 10 mg/day or Amlodipine Placebo Overall Group placebo, and atorvastatin and amloVariable (n П 64) (n П 70) (n П 134) dipine/placebo therapy was continAge (yrs) 58 П® 8 56 П® 9 57 П® 9 ued for 9 months. The dose of amloMen/women 55/9 60/10 115/19 dipine could be downtitrated at any 28.1 П® 3.1 28.9 П® 3.8 28.5 П® 3.5 Body mass index time to 5 mg/day if patients had 2 (kg/m ) symptoms related to hypotension. Diabetes mellitus 2 (3.1%) 10 (14.3%) 12 (9.0%) Smoking (current 81% 84% 83% After 12 months, all study-driven П© ex) treatment was discontinued and paBP (mm Hg) 127 П® 15/79 П® 9 128 П® 14/79 П® 8 128 П® 15/79 П® 9 tients were followed for an addiData are expressed as mean П® SD or number (%). tional 2 months. Plasma lipid and lipoprotein lipid measurements were determined at each institution using standardized in keeping with current practice guidelines.1– 4 Amlo- techniques. LDL cholesterol was calculated according dipine 10 mg was used for blood pressure (BP) control to the Friedewald formula. Apolipoprotein B was deaccording to current recommendations5–7 and was termined by nephelometry. BP was taken in triplicate, downtitrated to 5 mg if symptoms of hypotension with the subject sitting for Х†5 minutes, in accordance developed. Previous reports have indicated a low with the Canadian Hypertension Education Program prevalence of reaching recommended targets for CHD guidelines.6 prevention in high-risk subjects8 –12; recent trials point All continuous variables were analyzed at basetoward more aggressive LDL cholesterol targets in line and at months 3 and 12 for the intent-to-treat very high-risk subjects.13 and per-protocol populations. Analysis of covariance methods with effects for treatment, adjusted ••• Study subjects were selected in 11 centers across for center and baseline values, were performed. To Canada. The study was designed to assess the effects evaluate the effect before and after the treatment, a of atorvastatin and the combination of atorvastatin and paired Student’s t test was used within treatment amlodipine on vascular endothelial function measured comparisons. Continuous data are reported as mean by flow-mediated vasodilation. Inclusion criteria were П® SEM when using analysis of covariance and men and women aged Пѕ30 years with documented mean П® SD for raw data. For categorical variables, CHD (determined by coronary angiography Х†70% which included the target rate for BP and LDL, the cross-sectional stenosis), including patients after cor- Cochran-Mantel-Haenzsel chi-square test, stratified onary revascularization, by nuclear or stress echocar- according to center, was performed. All statistical diography or by a history of documented myocardial tests were 2-sided, and a p value ПЅ0.05 was coninfarction. For inclusion into the study, study subjects sidered statistically significant. No adjustments for had to be in stable condition for Х†3 months, without multiple comparisons were made. One hundred thirty-four subjects were randomized changes in current therapy, except those allowed within the protocol. The lipid entry criteria were a into the study (115 men and 19 women; mean age LDL cholesterol Пѕ3.2 mmol/L (123 mg/dl), triglyc- 57 П® 9 years). Baseline characteristics of all study erides ПЅ4.5 mmol/L (400 mg/dl), and no lipid-lower- subjects are listed in Table 1. As previously stated, at ing medications for 2 months. No patient was allowed study entry (baseline) all patients received atorvastatin to take calcium blocking drugs, hormone replacement 10 mg, which was uptitrated to 80 mg/day if the LDL therapy, or statins at entry. Antihypertensive medica- cholesterol was Пѕ2.5 mmol/L (100 mg/dl). Randomtion, antiplatelet agents, and antianginal agents were ization to amlodipine 10 mg or placebo was done at required to be stable for Х†3 months before entry into month 3 and BP was taken shortly thereafter. No the study. Choice of BP therapy was left to the dis- significant differences in baseline characteristics were cretion of the treating physician as long as protocol- observed in subjects subsequently randomized to amdriven restrictions were respected. Patients were ex- lodipine or placebo. The total duration of the study cluded if they had a coronary event, revascularization was 12 months; 56 of 64 subjects in the atorvastatin П© procedure, or a cerebrovascular accident in the pre- amlodipine group and 65 of 70 in the atorvastatin П© ceding 3 months. Diabetic subjects had to be well placebo group completed the study. At the baseline evaluation, 83% of patients reported a controlled for Х†3 months, as reflected by a hemoglobin A1c level Х…10% of total hemoglobin. No patient history of cigarette smoking, although only 21% were current smokers; the rate of smokers at 12 months was had renal or hepatic disease. All patients were asked to discontinue lipid-modi- 22%. Sixty-two percent of subjects were ex-smokers and fying agents and calcium blockers for 2 months, after only 16% never smoked. Mean body mass index (weight which time baseline lipid and lipoprotein cholesterol [kilograms]/height[square meters]) was 28.5 П® 3.5 at levels were obtained and all patients were started on baseline and 28.7 П® 3.6 at 12 months. The number of atorvastatin 10 mg/day. If the LDL cholesterol was subjects with a body mass index ПЅ25 kg/m2 was 14% Пѕ2.5 mmol/L (100 mg/dl) after 6 weeks, the dose of and at 12 months 14%, with no differences between TABLE 1 Baseline Characteristics of Study Population 250 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 TABLE 2 Lipid, Lipoprotein Lipids, and Apolipoprotein B Levels at Baseline, at Three and 12 Months on Study Medication, and After a Two-month Washout Lipids Cholesterol mmol/l mg/dl Triglycerides mmol/l mg/dl LDL cholesterol mmol/l mg/dl HDL cholesterol mmol/l mg/dl Cholesterol HDL cholesterol mmol/l mg/dl Non-HDL cholesterol mmol/l mg/dl Apolipoprotein B mg/dl Baseline 3 Mo 12 Mo 6.4 П® 1.1 147 П® 49 4.0 П® 0.6* 155 П® 23 3.9 П® 0.6* 150 П® 23 2.0 П® 0.9 177 П® 80 1.4 П® 0.3* 124 П® 27 1.4 П® 0.7* 124 П® 62 4.4 П® 1.0 170 П® 36 2.2 П® 0.5* 85 П® 19 2.2 П® 0.5* 85 П® 19 1.2 П® 0.3 46 П® 11 1.2 П® 0.3 46 П® 11 1.2 П® 0.3 46 П® 11 5.7 П® 1.5 220 П® 58 3.6 П® 0.8 139 П® 31 3.6 П® 0.9* 139 П® 35 5.2 П® 1.0 202 П® 39 1.4 П® 0.3 2.9 П® 0.6* 112 П® 23 0.80 П® 0.1* 2.8 П® 0.5* 108 П® 19 0.8 П® 0.1* *p ПЅ0.0001 versus baseline. Overall, there were no significant differences between amlodipine and placebo groups. FIGURE 1. Effect of atorvastatin on plasma LDL cholesterol (C) levels. groups. Body mass index remained stable throughout the study. At 12 months, in the atorvastatin П© placebo group, 26% of patients (17 of 65) were taking atorvastatin 10 mg and 74% (48 of 65) were taking 80 mg. In the atorvastatin П© amlodipine group, 21% (12 of 56) were taking atorvastatin 10 mg and the remaining subjects (79% [44 of 56]) were taking 80 mg. In the atorvastatin П© amlodipine group, 91% of patients continued to take 10 mg of amlodipine throughout the study, whereas only 9% (5 of 56) required downtitration to 5 mg. Baseline total cholesterol was 6.4 П® 1.0 mmol/L (147 П® 39 mg/dl) in the atorvastatin П© amlodipine and 6.4 П® 1.2 mmol/L (148 П® 45 mg/dl) in the FIGURE 2. Effects of amlodipine (Amlo) and placebo (Plac) on BP. Ator ‫ ؍‬atrovastatin. atorvastatin П© placebo groups. An overall decrease in total cholesterol by 2.4 П® 0.1 mmol/L was observed in the combined groups (p ПЅ0.0001). A similar overall decrease in LDL cholesterol was seen, from 4.4 П® 1.0 mmol/L (170 П® 39 mg/dl) to 2.2 П® 0.5 mmol/L (85 П® 19 mg/dl) (Table 2 and Figure 1). There was a trend toward a slightly lower LDL cholesterol in the atorvastatin П© placebo compared with the atorvastatin П© amlodipine groups (2.0 П® 0.1 vs 2.2 П® 0.1 mmol/L, p П 0.005) at 12 months; this represented an absolute decrease in LDL cholesterol of 2.2 П® 0.1 and 2.0 П® 0.1 mmol/L in each group, respectively, representing a 50% decrease in LDL cholesterol. Plasma triglycerides decreased from 2.0 П® 1.0 mmol/L (177 П® 88 mg/dl) in both groups to 1.4 П® 0.7 mmol/L (117 П® 56 mg/dl), with no significant differences between both groups. High-density lipoprotein (HDL) cholesterol levels did not change with atorvastatin treatment (1.2 П® 0.3 at baseline vs 1.2 П® 0.3 mmol/L at 12 months). Data at 3 months was virtually identical to 12-month values for cholesterol, triglycerides, and LDL/HDL cholesterol. Overall apolipoprotein B levels decreased from a baseline level of 1.35 П® 0.25 to 0.76 П® 0.14 mg/dl, a 43% decrease (p ПЅ0.0001), with no significant difference between treatment groups. The overall cholesterol/HDL cholesterol ratio decreased from 5.6 П® 1.4 to 3.6 П® 1.0 mg/dl (p ПЅ0.0001; Table 2)., with no significant difference between treatment groups. The major LDL cholesterol target of ПЅ2.5 mmol/L (100 mg/dl) was achieved overall in 81% of study subjects (97 of 120). Interestingly, there were slightly fewer patients reaching an LDL cholesterol of ПЅ2.5 mmol/L in the atorvastatin П© amlodipine group (39 of 56 [71%], 95% confidence intervals 57% to 82%) than in the atorvastatin П© placebo group (58 of 65 [89%], 95% confidence intervals 79% to 96%) (p П 0.01). This may be related to the slight increase in patients taking the 80-mg dose of atorvastatin in the atorvastatin П© placebo group (78% vs 74%) and to the absolute 0.3 mmol (11 mg/dl) difference among LDL cholesterol levels between the atorvastatin П© placebo group and the atorvastatin П© amlodipine group (2.3 П® 0.5 vs BRIEF REPORTS 251 TABLE 3 Blood Pressure at Baseline, Three Months, 12 Months, and After a Two-month Washout Time Baseline (mm Hg) 3 mo (mm Hg) 12 mo (mm Hg) Atorvastatin П© Amlodipine Atorvastatin П© Placebo 127 П® 15/79 П® 9 117 П® 11/74 П® 9*†118 П® 10/76 П® 7*†128 П® 14/79 П® 8 127 П® 16/79 П® 7 130 П® 17/80 П® 9 *p ПЅ0.0001 versus baseline; †p ПЅ0.0001 versus atorvastatin П© placebo. FIGURE 3. Percentage of patients reaching LDL cholesterol (C) targets (<2.5 mmol/L) taking atorvastatin 10 or 80 mg, and reaching BP targets taking amlodipine 5 or 10 mg or placebo. Abbreviations as in Figure 2. 2.0 П® 0.4 mmol/L, p П 0.005). The percentage of subjects reaching a cholesterol/HDL cholesterol ratio ПЅ4.0 was 11.3% in the atorvastatin П© amlodipine group versus 2.9% in the atorvastatin П© placebo group at baseline (p П 0.066). At month 12, these rates were 70.9% and 73.8%, respectively. Overall, 72.5% of all study participants reached a HDL cholesterol target of ПЅ4.0 compared with 6.8% at baseline. At baseline, non-HDL cholesterol was 5.17 П® 0.84 mmol/L (200 П® 32 mg/dl) in the atorvastatin П© amlodipine group and 5.30 П® 1.13 mmol/L (205 П® 44 mg/dl) in the atorvastatin П© placebo subjects (p П NS). At 12 months, non-HDL cholesterol was 2.9 П® 0.1 mmol/L (109 П® 27 mg/dl) and 2.7 П® 0.1 mmol/L (102 П® 27 mg/dl), respectively (p П 0.048). Mean BP was 128 П® 15/79 П® 9 mm Hg at baseline and was similar in both groups: 127 П® 15/79 П® 9 mm Hg in the atorvastatin П© amlodipine group and 128 П® 14/79 П® 8 mm Hg in the atorvastatin П© placebo group. Diuretics were administered to 12% of patients, вђ¤-adrenergic blockers were given to 75%, angiotensin-converting enzyme inhibitors to 54%, and angiotensin receptors blockers to 5%. Other antihypertensives (вђЈ-adrenoreceptor blockers) were administered to 1 patient and accounted for ПЅ1% of prescriptions (calcium blockers were not allowed at the time of randomization). At the end of the study, BP was 118 П® 10/76 П® 7 mm Hg in the atorvastatin П© amlodipine group (an absolute decrease of 8 П® 2/3 П® 1 mm Hg compared with baseline values, p ПЅ0.0001), and was 252 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 130 П® 17/80 П® 9 mm Hg in the atorvastatin П© placebo group. The differences in systolic and diastolic pressures at 12 months between the atorvastatin П© amlodipine and the atorvastatin П© placebo groups was significant at the p ПЅ0.0001 and 0.0006 levels, respectively (Figure 2). We used the treatment goals of the Canadian Hypertension Education Program in the analysis: at baseline, 75% of the patients were at BP target in the atorvastatin П© amlodipine group compared with 70% in the atorvastatin П© placebo group (p П 0.503). At 12 months, 96% (54 of 56) of subjects in the atorvastatin П© amlodipine group were at or below Canadian Hypertension Education Program specified BP targets, whereas 69% (45 of 65) were at or below target in the atorvastatin П© placebo group (p П 0.0002) (Table 3 and Figure 3). The medications were well tolerated. Myalgias were reported in 6 of 70 patients (8.6%) in the atorvastatin П© amlodipine group and in 1 of 74 patients (1.4%) in the atorvastatin П© placebo group. Elevations in creatine kinase (Пѕ3 times the upper limit of normal) occurred in 3 of 70 (4.3%) and in 2 of 74 (2.7%) subjects. There were 4 of 70 cases (4.3%) of symptomatic hypotension in the atorvastatin П© amlodipine group, resulting in a decrease in the amlodipine dosage and in 1 subject (1.4%) in the atorvastatin П© placebo group. Angina occurred in 5 of 70 subjects (7.1%) in the atorvastatin П© amlodipine group and in 9 of 74 (12.3%) in the atorvastatin П© placebo group. One patient in each group developed acute myocardial infarction. ••• Despite a relatively normal BP at entry (127 П® 15/79 П® 9 mm Hg), 25% of patients in the atorvastatin П© amlodipine group and 30% of patients in the atorvastatin П© placebo groups were not at target at baseline (p П 0.503 between-group assignment). With the addition of amlodipine, 96% of patients were at target at 3 and 12 months compared with 69% in the placebo group (p ПЅ0.0001). Target LDL cholesterol levels have remained a source of controversy; an arbitrary level of ПЅ2.5 mmol/L (100 mg/dl),1,3 a single dose of statin,14 or a 45% to 45% reduction in LDL cholesterol15 are all evidenced based. The use of other cardioprotective medications are also part of evidencebased medicine in high-risk prevention strategies.16 –19 The mean LDL cholesterol achieved at 1 year is less than the recommended level, yet is in keeping with novel studies13,20 that address the issue of aiming for a LDL cholesterol value lower than recommended. In our study, 81% of all patients reached a LDL cholesterol of ПЅ2.5 mmol/L. The rates of LDL cholesterol target at 3 and 12 months were 79% and 71%, respectively, in the atorvastatin П© amlodipine group and 87%, and 89% in the atorvastatin П© placebo group. The data presented here suggest that intensive patient follow-up can lead to very high rates of reaching multiple recommended targets for the prevention of cardiovascular disease in high-risk patients (Figure 3). We used a combination of atorvastatin 10 or 80 mg and amlodipine 5 or 10 mg. Therapy with atorvastatin 10 mg П© amlodipine 10 JANUARY 15, 2005 mg was administered to 26% of patients, atorvastatin 80 mg П© amlodipine 5 mg to 9%, and atorvastatin 80 mg П© amlodipine 10 mg to 70%. The combination of the 2 medications was well tolerated. There were no unexpected adverse effects reported in either treatment group. A hard cardiovascular end point (acute myocardial infarction) occurred in 1 subject in each treatment arm (2 of 134 [1.5%]) during the 12-month treatment period. Although not designed to assess cardiovascular end points, the low event rate is a reflection of baseline treatment according to current recommendations for secondary prevention.4 1. Expert Panel on Detection, Evaluation, and Treatment of High Blood Choles- terol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285:2486 –2497. 2. Genest J, Frohlich J, Fodor G, McPherson R, and the Working Group on Hypercholesterolemia and Other Dyslipidemias. Recommendations for the management of dyslipidemia and the prevention of cardiovascular disease: summary of the 2003 update. Can Med Assoc J 2003;169:921–924. 3. De Backer G, Ambrosioni E, Borch-Johnsen K, Brotons C, Cifkova R, Dallongeville J, Ebrahim S, Faergeman O, Graham I, Mancia G, et al, and the European Society of Cardiology Committee for Practice Guidelines. European guidelines on cardiovascular disease prevention in clinical practice: third joint task force of European and other societies on cardiovascular disease prevention in clinical practice (constituted by representatives of eight societies and by invited experts). Eur J Cardiovasc Prev Rehabil 2003;10(suppl):S1–S10. 4. Genest J, Pedersen TR. Prevention of cardiovascular ischemic events: high-risk and secondary prevention. Circulation 2003;107:2059 –2065. 5. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ, and the National Heart, Lung, and Blood Institute Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure, and the National High Blood Pressure Education Program Coordinating Committee. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003;289:2560 –2572. 6. Hemmelgarn BR, Zarnke KB, Campbell NR, Feldman RD, McKay DW, McAlister FA, Khan N, Schiffrin EL, Myers MG, Bolli P, et al, and the Canadian Hypertension Education Program, Evidence-Based Recommendations Task Force. The 2004 Canadian Hypertension Education Program recommendations for the management of hypertension: part I– blood pressure measurement, diagnosis and assessment of risk. Can J Cardiol 2004;20:31– 40. 7. European Society of Hypertension-European Society of Cardiology Guideline Committee 2003. European Society of Hypertension—European Society of Car- diology guidelines for the management of arterial hypertension. J Hypertens 2003;21:1011–1053. 8. Wolf-Maier K, Cooper RS, Kramer H, Banegas JR, Giampaoli S, Joffres MR, Poulter N, Primatests P, Stegmayr B, Thamm M. Hypertension treatment and control in five European Countries, Canada and the Unites States. Hypertension 2004;43:10 –17. 9. Touyz RM, Campbell N, Logan A, Gledhill N, Petrella R, Padwal R. Canadian Hypertension Education Program. The 2004 Canadian recommendations for the management of hypertension: part III—lifestyle modifications to prevent and control hypertension. Can J Cardiol 2004;20:55–59. 10. EUROASPIRE I and II Group. European Action on Secondary Prevention by Intervention to Reduce Events. Clinical reality of coronary prevention guidelines: a comparison of EUROASPIRE I and II in nine countries. EUROASPIRE I and II Group. European Action on Secondary Prevention by Intervention to Reduce Events. Lancet 2001;357:972–973. 11. Vanuzzo D, Pilotto L, Ambrosio GB, Pyorala K, Lehto S, De Bacquer D, De Backer G, Wood D, and the EUROASPIRE Study Group. Potential for cholesterol lowering in secondary prevention of coronary heart disease in Europe: findings from EUROASPIRE study. European Action on Secondary Prevention through Intervention to Reduce Events. Atherosclerosis 2000;153: 505–517. 12. Pearson TA, Laurora I, Chu H, Kafonek S. The lipid treatment assessment project (L-TAP): a multicenter survey to evaluate the percentages of dyslipidemic patients receiving lipid-lowering therapy and achieving low-density lipoprotein cholesterol goals. Arch Intern Med 2000;160:459 – 467. 13. Grundy SM, Cleeman JL, Merz CN, Brewer HB Jr., Clark T, Hunninghake DB, Pastanak RC, Smith SC Jr., Stone NJ, for the Coordinating Committee of the National Cholesterol Education Program. Implication of recent clinical trials for the National Cholesterol Education Program adult Treatment Panel III Guidelines. Circulation 2004;110:227–239. 14. Genest J, Libby P, Gotto AM. Lipoprotein disorders and cardiovascular risk. In: Braunwald E, Libby P, Zipes D, eds. Braunwald’s Heart Disease. Philadelphia: WB Saunders, 2004;1013–1033. 15. Bucher HC, Griffith LE, Guyatt GH. Systematic review on the risk and benefit of different cholesterol-lowering interventions. Arterioscler Thromb Vasc Biol 1999;19:187–195. 16. Freemantle N, Cleland J, Young P, Mason J, Harrison J. Beta blockade after myocardial infarction: systematic review and meta regression analysis. BMJ 1999;26318:1730 –1737. 17. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 2000;342:145–153. 18. Fox KM, and the EURopean trial On reduction of cardiac events with Perindopril in stable coronary Artery disease Investigators. Efficacy of perindopril in reduction of cardiovascular events among patients with stable coronary artery disease: randomised, double-blind, placebo-controlled, multicentre trial (the EUROPA study). Lancet 2003;362:782–788. 19. Pfeffer MA, McMurray JJ, Velazquez EJ, Rouleau JL, Kober L, Maggioni AP, Solomon SD, Swedberg K, Van de Werf F, White H, et al, and the Valsartan in Acute Myocardial Infarction Trial Investigators. Valsartan, captopril, or both in myocardial infarction complicated by heart failure, left ventricular dysfunction, or both. N Engl J Med 2003;349:1893–1906. 20. Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA, Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN, and the REVERSAL Investigators. Effect of intensive compared with moderate lipidlowering therapy on progression of coronary atherosclerosis: a randomized controlled trial. JAMA 2004;291:1071–1080. BRIEF REPORTS 253 Benefits of Niacin by Glycemic Status in Patients With Healed Myocardial Infarction (from the Coronary Drug Project) Paul L. Canner, PhD, Curt D. Furberg, MD, PhD, Michael L. Terrin, Mark E. McGovern, MD The Coronary Drug Project, conducted during 1966 to 1974, was a randomized, double-blind, placebo-controlled trial of 5 lipid-modifying agents in 8,341 men with previous myocardial infarction.1 Among the 5 drug treatment regimens, only niacin significantly reduced the risk of (1) cardiovascular events during a mean follow-up of 6.2 years and (2) total mortality during 6.2 years with study treatment plus an additional 9 years of post-trial follow-up (Figure 1).2,3 Cardiovascular and total mortality outcomes in the niacin and placebo groups are presented by baseline glycemic status and by change in glycemic status from baseline to year 1. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:254 –257) iacin increases plasma glucose levels in some patients through mechanisms that are not comN pletely understood. As a result, caution has been 4 urged in the use of niacin among patients with elevated fasting glucose levels or overt diabetes.5,6 Whether the effect on glucose control results in any disadvantage with respect to cardiovascular events or mortality risk has not been studied in detail. A previously reported finding may cast some light on the matter. In the Coronary Drug Project (CDP), the niacin–placebo difference in 15-year total mortality was similar between patients with baseline fasting plasma glucose (FPG) levels ПЅ100 mg/dl (48.6% vs 53.5%) and in patients with FPG levels Х†100 mg/dl (56.1% vs 63.9%).3 Patients treated with insulin were not eligible for enrollment in the CDP. The present investigation examines in detail the effects of niacin on cardiovascular events and total mortality in patient subgroups defined by glycemic status. ••• Event rates were obtained for recurrent nonfatal definite myocardial infarction (MI) and for coronary heart disease death or MI at 6 years, and for total mortality at 15 years by baseline fasting and 1-hour plasma glucose levels and by changes in fasting and 1-hour plasma glucose levels from baseline to year 1. Treatment group comparisons were made using the From the Maryland Medical Research Institute, Baltimore, Maryland; Wake Forest University, Winston-Salem, North Carolina; and Kos Division of Medical Affairs, Weston, Florida. This report was supported by an unrestricted research grant from Kos Pharmaceuticals, Inc. Dr. Canner’s address is: Maryland Medical Research Institute Inc., 600 Wyndhurst Avenue, Baltimore, Maryland 21210. E-mail: [email protected]. Manuscript received June 3, 2004; revised manuscript received and accepted August 31, 2004. 254 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, and FIGURE 1. CDP: results for cardiovascular end points at 6 years and mortality at 15 years. *p <0.005; †p <0.05. CHD ‫ ؍‬coronary heart disease; MI ‫ ؍‬nonfatal definite MI. Cox proportional-hazards model (SAS software, version 8.2, SAS Inc., Cary, North Carolina),7 which yielded a hazard ratio for the niacin–placebo comparison of event rates in each glucose subgroup. A hazard ratio of ПЅ1.0 indicates a lesser event rate with niacin than with placebo. The hazard ratios for the glucose subgroups were evaluated for homogeneity using a Cox regression model with 6-year MI, 6-year coronary heart disease death or MI, or 15-year total mortality as the dependent variable and treatment group (niacin coded 1 or placebo coded 2), glucose category (e.g., ПЅ95, 95 to 104, 105 to 125, Х† 126 mg/dl coded 1, 2, 3, or 4, respectively), and the product of treatment group and glucose category codes as the independent variables. If the z value for the estimated regression coefficient for this product (or interaction) term is reasonably close to zero and not close to П®1.96, it can be inferred that the subgroup hazard ratios are homogeneous or do not differ significantly from one another. A negative interaction z value indicates that the hazard ratios are getting smaller (i.e., favoring niacin) as the baseline glucose level or the change in glucose levels at year 1 increases. A positive interaction z value indicates that the niacin effect is becoming less advantageous with increasing glucose levels. In analyses based on change in fasting or 1-hour plasma glucose levels from baseline to year 1, nonfatal events occurring before the year 1 visit are excluded. This analysis included all but 2 patients from the original niacin (n П 1,119) and placebo (n П 2,789) cohorts. When transferring the CDP data from 3 magnetic tapes to more modern electronic storage devices, 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.013 TABLE 1 Coronary Drug Project: Mean Glycemia in the Niacin and Placebo Groups for the Cohort of Patients Completing Closeout Visit (CV) Two Fasting Plasma Glucose (mg/dl) Time Baseline Yr 1 Yr 2 Yr 4 CV1* CV2†1-h Plasma Glucose (mg/dl) Niacin (n П 567) Placebo (n П 1,447) p Value Niacin (n П 542) Placebo (n П 1,382) p Value 101 107 107 108 108 105 100 101 102 104 106 107 NS ПЅ0.001 ПЅ0.001 ПЅ0.01 NS NS 168 179 179 183 188 171 169 164 165 170 177 177 NS ПЅ0.001 ПЅ0.001 ПЅ0.001 ПЅ0.001 ПЅ0.05 *CV1 П first closeout visit (i.e., the visit at which study treatment was ended). †CV2 П Пі4 months after CV1. TABLE 2 Percentages of Events by Baseline Glucose Levels in the Niacin and Placebo Groups in the Coronary Drug Project Fasting Plasma Glucose (mg/dl) Event* MI at 6 yrs CHD death or MI at 6 yrs Total mortality at 15 yrs ПЅ95 (n П 424, 1,026)‡ 95–104 (n П 409, 1,028) 105–125 (n П 216, 552) Х†126 (n П 70, 181) z Value for Interaction†11.1 vs 15.4 (0.70) 24.8 vs 32.4 (0.74) 50.5 vs 53.9 (0.91) 9.8 vs 12.9 (0.76) 27.9 vs 29.6 (0.94) 47.7 vs 55.5 (0.82) 10.6 vs 14.3 (0.75) 26.4 vs 31.7 (0.84) 56.5 vs 63.9 (0.86) 7.1 vs 15.5 (0.43) 37.1 vs 44.8 (0.76) 71.4 vs 77.9 (0.83) ПЄ0.44 0.47 ПЄ0.63 1-h Plasma Glucose (mg/dl) MI at 6 yrs CHD death or MI at 6 yrs Total mortality at 15 yrs ПЅ140 (n П 299, 765)‡ 140–179 (n П 386, 919) 180–219 (n П 246, 654) Х†220 (n П 182, 443) 9.7 vs 13.1 (0.73) 25.4 vs 27.8 (0.89) 46.5 vs 49.8 (0.91) 11.1 vs 14.6 (0.73) 25.1 vs 32.3 (0.75) 50.8 vs 56.4 (0.86) 9.8 vs 13.6 (0.74) 28.0 vs 31.5 (0.92) 49.2 vs 61.2 (0.76) 10.4 vs 16.7 (0.61) 31.9 vs 39.5 (0.79) 65.9 vs 70.9 (0.86) ПЄ0.47 ПЄ0.23 ПЄ0.66 *Event rates given as percent in niacin group versus percent in placebo group (hazard ratio). †z value between П® 1.96 indicates homogeneity or no statistical difference (p Х†0.05) in hazard ratios among the various glucose levels. ‡ Sample sizes for the niacin and placebo groups, respectively. CHD П coronary heart disease; MI П nonfatal definite MI. data for the last patient on each of the first 2 tapes were lost. Therefore, the actual placebo cohort for the present analysis numbers 2,787. One-hour plasma glucose determinations were not obtained at baseline for 6 patients taking niacin and 6 taking placebo due to problems with the glucose challenge. Analyses of glucose change from baseline to year 1 excluded patients who either died, withdrew during the first year, or missed the first annual follow-up visit. Thus, 965 and 2,471 patients in the niacin and placebo groups, respectively, contributed to the analyses of change in FPG levels, and 959 and 2,460 correspondingly contributed to the analyses of change in 1-hour plasma glucose levels. At baseline, about 42% of patients in the CDP had elevated FPG levels (Х†100 mg/dl), and approximately 39% had impaired glucose tolerance as indicated by elevated 1-hour plasma glucose levels (post-glucose load) Х†180 mg/dl.2 Approximately 6% used oral hypoglycemic agents or had urine positive for glucose. During the trial, niacin increased fasting and 1-hour plasma glucose levels (Table 1); however, there was little difference at study end between the niacin and placebo groups in the percentage of patients who were newly prescribed insulin (1.3% vs 0.9% placebo; p П NS) or oral hypoglycemic agents (4.3% vs 4.6%; p П NS).2 Compared with placebo, niacin reduced the risk of 6-year recurrent MI and coronary heart disease death or MI and 15-year total mortality similarly in patients at all levels of baseline FPG, including those with FPG levels Х†126 mg/dl (the current definition of diabetes) or 1-hour plasma glucose levels of Х†180 or Х†220 mg/dl (Table 2 and Figures 2 and 3). The beneficial effect of niacin on cardiovascular events and total mortality was not diminished, even among those with the highest baseline fasting or 1-hour plasma glucose levels. Compared with placebo, niacin reduced the risk of 6-year recurrent MI and coronary heart disease death or MI and 15-year total mortality similarly in patients at all levels of change in fasting or 1-hour plasma BRIEF REPORTS 255 FIGURE 2. CDP at 6 years: MI by baseline FPG. z Value for interaction ‫ ؍‬؊0.44 (p ‫ ؍‬0.66), indicating homogeneity of treatment effects across all levels of FPG. Abbreviation as in Figure 1. FIGURE 3. CDP at 15 years: total mortality by baseline FPG. z Value for interaction ‫ ؍‬؊0.63 (p ‫ ؍‬0.53), indicating homogeneity. TABLE 3 Percentages of Events by Changes in Baseline Glucose Levels at Year One in the Niacin and Placebo Groups in the Coronary Drug Project Change in Fasting Plasma Glucose (mg/dl) Event* MI at 6 yrs CHD death or MI at 6 yrs Total mortality at 15 yrs ПЅ0 (n П 318, 1,165)‡ 0 –9 (n П 346, 941) Х†10 (n П 301, 365) z Value for Interaction†7.9 vs 12.7 (0.62) 23.0 vs 26.2 (0.87) 49.4 vs 55.0 (0.88) 9.0 vs 12.9 (0.68) 20.2 vs 26.2 (0.75) 45.4 vs 51.9 (0.82) 9.0 vs 15.1 (0.55) 20.6 vs 34.5 (0.55) 49.8 vs 64.1 (0.69) ПЄ0.21 ПЄ1.93 ПЄ1.14 Change in 1-h Plasma Glucose (mg/dl) MI at 6 yrs CHD death or MI at 6 yrs Total mortality at 15 yrs ПЅ0 (n П 395, 1,357)‡ 0–29 (n П 270, 670) Х†30 (n П 294, 433) 7.1 vs 13.5 (0.51) 20.8 vs 27.3 (0.74) 48.4 vs 53.5 (0.87) 8.9 vs 12.4 (0.71) 21.5 vs 26.9 (0.78) 47.4 vs 53.7 (0.86) 10.5 vs 13.2 (0.78) 21.4 vs 29.6 (0.70) 47.3 vs 62.6 (0.66) 1.49 ПЄ0.18 ПЄ1.80 *Event rates given as percentage in niacin group versus percentage in placebo group (hazard ratio). †z value between П® 1.96 indicates homogeneity or no statistical difference (p Х†0.05) in hazard ratios among the various glucose levels. ‡ Sample sizes for the niacin and placebo groups, respectively. Abbreviations as in Table 2. glucose levels from baseline to year 1 (Table 3 and Figures 4 and 5). As with baseline fasting and 1-hour plasma glucose levels, the effect of niacin on cardiovascular events and mortality was not diminished, even among patients with the largest increases in fasting or 1-hour plasma glucose levels from baseline to year 1. Recurrent MI by change in 1-hour plasma glucose levels was an exception to this trend (Table 3). Some caution must be exercised in interpreting the results based on glucose change, because these are variables defined at follow-up rather than at baseline.8 Glucose change in the CDP is positively correlated with level of adherence to study protocol prescription in patients taking niacin, but not in those taking placebos. Also, as reported elsewhere,8 the level of adherence was inversely correlated with total mortality in the CDP. Nevertheless, analyses of events by glucose change for patients at full adherence for the first 256 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 year of follow-up gave results very similar to those in Table 3 without regard to adherence. ••• The results of this study confirm and extend the evidence from the CDP relating to the effects of niacin in reducing the risk of cardiovascular events and mortality.2,3 Among patients receiving placebo, elevated baseline fasting and 1-hour plasma glucose levels were both positively associated with total mortality.9 This can also be seen in the present article by comparing the placebo group data across each line of Table 2 for 15-year mortality. At the same time, it has been reported from both the CDP and other studies that niacin treatment, especially at high doses, may elevate plasma glucose levels in some patients.2,5,10 The present results show that the increase in glucose levels with niacin did not translate into any disadvantage with respect to cardiovascular events or mortality risk. JANUARY 15, 2005 FIGURE 4. CDP at 6 years: MI by change in year 1 FPG. z Value for interaction ‫ ؍‬؊0.21 (p ‫ ؍‬0.83), indicating homogeneity. Abbreviation as in Figure 1. in this subgroup had a baseline FPG Х†126 mg/dl; thus, the numbers are too small to perform any meaningful analyses. In the aggregate, the findings detailed in this report demonstrate that niacin has favorable effects on clinical outcome in patients with evidence of abnormal glucose metabolism or overt diabetes, although patients treated with insulin were excluded from enrollment in the CDP. Recently, there has been a resurgence of interest in using niacin for the treatment of dyslipidemia, especially for raising low levels of high-density lipoprotein cholesterol.11 Although niacin is the most effective agent for raising low levels of high-density lipoprotein cholesterol, it can cause hyperglycemia in some patients, and its use in patients with abnormal glucose metabolism or overt diabetes mellitus has previously been cautioned.6 Such caution is not supported by these analyses of the CDP. 1. The Coronary Drug Project Research Group. The Coronary Drug Project. Design, methods, and baseline results. Circulation 1973;47(suppl I):I-1–I-50. 2. The Coronary Drug Project Research Group. Clofibrate and niacin in coronary heart disease. JAMA 1975;231:360 –381. 3. Canner PL, Berge KG, Wenger NK, Stamler J, Friedman L, Prineas RJ, FIGURE 5. CDP at 15 years: total mortality by change in year 1 FPG. z Value for interaction ‫ ؍‬؊1.14 (p ‫ ؍‬0.25), indicating homogeneity. In the CDP, niacin produced average reductions of 10% in serum cholesterol and 26% in serum triglycerides over the first 5 years of follow-up despite modest compliance with the study medication.2 Niacin is known to raise low levels of high-density lipoprotein cholesterol.6 In the CDP, only 492 patients (138 taking niacin and 354 taking placebos) at baseline and none at follow-up had high-density lipoprotein cholesterol determinations. Only 6 patients taking niacin Friedewald W, for the Coronary Drug Project Research Group. Fifteen year mortality in Coronary Drug Project patients: long-term benefit with niacin. J Am Coll Cardiol 1986;8:1245–1255. 4. McGovern ME. Use of nicotinic acid in patients with elevated fasting glucose, diabetes, or metabolic syndrome. Br J Diabetes Vasc Dis 2004;4:78 – 85. 5. Garg A, Grundy SM. Nicotinic acid as therapy for dyslipidemia in non–insulindependent diabetes mellitus. JAMA 1990;264:723–726. 6. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation 2002;106:3143–3421. 7. Cox DR. Regression models and life-tables. J R Stat Soc B 1972;34:187–202. 8. The Coronary Drug Project Research Group. Influence of adherence to treatment and response of cholesterol on mortality in the Coronary Drug Project. N Engl J Med 1980;303:1038 –1041. 9. The Coronary Drug Project Research Group. The prognostic importance of plasma glucose levels and of the use of oral hypoglycemic drugs after myocardial infarction in men. Diabetes 1977;26:453– 465. 10. Berge KG, Molnar GD. Effects of nicotinic acid on clinical aspects of carbohydrate metabolism. In: Altschul R, ed. Niacin in Vascular Disorders and Hyperlipidemia. Springfield, IL: Charles C Thomas, 1964:136 –155. 11. Miller M. Niacin as a component of combination therapy for dyslipidemia. Mayo Clin Proc 2003;78:735–742. BRIEF REPORTS 257 Occurrence of Acute Myocardial Infarction in Worcester, Massachusetts, Before, During, and After the Terrorists Attacks in New York City and Washington, DC, on 11 September 2001 Robert J. Goldberg, PhD, Frederick Spencer, MD, Darleen Lessard, MS, Jorge Yarzebski, MD, MPH, Craig Lareau, BS, and Joel M. Gore, MD The Worcester Heart Attack Study is an ongoing population-based investigation examining changes over time in the incidence and case-fatality rates of acute myocardial infarction (AMI) in residents of metropolitan Worcester, Massachusetts.1–3 Using this large population-based database of patients with independently validated AMI, we examined the impact of the September 11th tragedy on the occurrence rates of AMI in residents of greater Worcester presenting to all area hospitals during 2001 compared with rates in the previous 10 years. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:258 –260) esidents of the Worcester metropolitan area hospitalized with a primary or secondary discharge R diagnosis of acute myocardial infarction (AMI) (International Classification of Disease-9th Revision [ICD-9] code 410) from all Worcester standard metropolitan statistical area hospitals during 1991, 1993, 1995, 1997, 1999, and 2001 formed the study population of this report. The present study is part of an ongoing, population-based investigation examining changes over time in the incidence, hospital, and longterm case-fatality rates of residents of the Worcester metropolitan area hospitalized with confirmed AMI.1–3 ••• Although additional earlier study years were included in this community-wide investigation (1975 to 1990), we restricted the study sample to patients hospitalized with validated AMI during the decade before and during 2001. This time period was chosen to examine relatively contemporary and decade-long trends in the occurrence of AMI. This period was also chosen to provide a more systematic and extensive contrast of the usual, or endemic, magnitude of AMI between 1991 and 1999 and evaluation of the possible “epidemic” effects of the World Trade Center attacks on September 11, 2001. These 6 annual periods (1991, 1993, 1995, 1997, 1999, and 2001) were originally From the Department of Medicine, Division of Cardiovascular Medicine, University of Massachusetts Medical School, Worcester, Massachusetts. This report was supported by Grant RO1 HL35434 from the National Heart, Lung, and Blood Institute, Bethesda, Maryland. Dr. Goldberg’s address is: Division of Cardiovascular Medicine, Department of Medicine, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Massachusetts 01655. E-mail: [email protected]. Manuscript received May 11, 2004; revised manuscript received and accepted August 31, 2004. 258 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 selected because of the availability of grant funding and to examine changes over time in our principal study end points over an alternating yearly basis. All 11 hospitals in the Worcester standard metropolitan statistical area (2000 census population, 478,000) participated in this study. The medical records of greater Worcester residents with a discharge diagnosis of AMI from these hospitals were individually reviewed and validated according to pre-established criteria that have been previously described.1–3 The present report is based on the 6,072 residents of the Worcester metropolitan area who satisfied our predefined diagnostic criteria for AMI. Of these, 4,834 patients were hospitalized at all greater Worcester medical centers during the study years between 1991 and 1999, whereas 1,238 patients were hospitalized in 2001. The hospital records of patients with validated AMI were abstracted for information about patient’s age, gender, medical history, occurrence of several clinical complications during hospitalization, and AMI order (initial vs prior) and type (Q-wave or non–Q wave). Detailed descriptions of these factors have been provided in previous publications.1–3 Information on possible precipitating factors associated with onset of AMI was not collected. We examined the average daily occurrence of all AMIs to determine whether the frequency of AMI events increased in relation to the terrorist attacks on September 11 (Figure 1). In 2001, the average daily number of cases of AMI was 3.4 in the 9-month period ending on Sepember 10. On the day of September 11, there were 7 cases of AMI, 9 cases of AMI on September 12, and 13 additional cases during the remainder of the week. In the month after September 11 there were 122 confirmed cases of AMI (daily mean 4.1), and 322 (mean 3.5) in the 3-month period thereafter. In contrast, among patients hospitalized during the “unexposed” endemic or control period between 1991 and 1999, the average number of greater Worcester residents hospitalized with confirmed AMI on a daily basis was 2.7 over the period January 1 to September 10. On September 11, there were an average of 1.8 cases of AMI and an average of 2.6 cases on September 12 during this period. The average daily number of cases of AMI was 2.7 in the 1-week, 1-month, and 3-month period after September 11 in the control or comparison study years. Essentially similar trends were observed in patients who 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.094 FIGURE 1. Occurrence of acute myocardial infarction during selected time periods (Worcester Heart Attack Study). TABLE 1 Descriptive Characteristics of Patients With Acute Myocardial Infarction (AMI) Pre 9/11 (n П 3,436) Characteristic Age (median yrs) Age (yrs) ПЅ55 55–64 65–74 Х†75 Men Angina pectoris Diabetes mellitus Heart failure Hypertension Stroke AMI characteristics Initial Q wave Duration of prehospital delay (median h) 9/11 and 9/12 (n П 22) Month after 9/11 (9/12– 10/11) (n П 406) 3 Months After 9/11 (9/12– 12/11) (n П 1,208) According to Time Period of Hospitalization: 1991 to 1999 71 69 72 72 16.4% 17.2% 26.6% 39.8% 57.4% 26.4% 29.7% 20.3% 58.9% 10.4% 18.2% 18.2% 31.8% 31.8% 72.7% 27.3% 40.9% 27.3% 45.5% 0% 18.5% 16.8% 23.7% 41.1% 57.9% 21.4% 24.4% 19.2% 58.6% 9.9% 16.8% 17.0% 24.8% 41.5% 56.4% 22.4% 28.5% 18.6% 59.7% 10.4% 64.1% 37.5% 2.2 59.1% 27.3% 3.1 65.8% 38.2% 2.2 65.0% 35.9% 2.1 According to Time Period of Hospitalization: 2001 (n П 856) Age (median yrs) Age (yrs) ПЅ55 55–64 65–74 Х†75 Men Angina pectoris Diabetes mellitus Heart failure Hypertension Stroke AMI characteristics Initial Q wave Duration of prehospital delay (median h) (n П 16) (n П 122) (n П 322) 74 71 73 73 14.1% 13.6% 23.0% 49.3% 57.7% 23.8% 33.2% 24.0% 69.4% 12.1% 25.0% 18.8% 12.5% 43.8% 62.5% 18.8% 25.0% 12.5% 50.0% 0% 12.3% 18.9% 23.8% 45.1% 54.6% 24.6% 34.4% 24.6% 70.5% 11.5% 14.3% 18.9% 22.7% 44.1% 57.6% 20.5% 30.8% 24.2% 66.8% 11.5% 64.8% 22.4% 1.8 75.0% 18.8% 2.3 64.8% 18.0% 2.0 65.8% 19.3% 2.0 presented to all greater Worcester hospitals with an initial AMI. The risk of AMI was significantly increased on the day of September 11, 2001 (odds ratio [OR] 3.04, 95% confidence interval [CI] 1.13 to 8.17), and on September 11 and 12, 2001 (OR 2.85, 95% CI 1.49 to 5.45) compared with similar periods during the control or unexposed period. There was no significant increase in the risk of AMI during the week after September 11, 2001 (OR 1.23, 95% CI 0.81 to 1.87), the month after September 11 (OR 1.20, 95% CI 0.97 to 1.48), and the 3-month period thereafter (OR 1.05, 95% CI 0.91 to 1.22) compared with similar periods during the unexposed or control years. To examine the potential role of chance in the increased occurrence of AMIs on the day of September 11, 2001, and the day thereafter, we performed a simple count of the number of days that there were Х†7 confirmed cases of AMI on consecutive days in the unexposed cohorts. There were no times during the control years between 1991 and 1999 when the number of patients hospitalized with AMI on consecutive days was Х†16 as was seen on September 11 and 12, 2001. We examined the demographic and clinical characteristics of patients with AMI during different study periods to determine whether there were any differences in the distribution of these characteristics according to the time period of hospitalization (Table 1). Patients experiencing AMI on September 11 and 12, 2001, included a greater proportion of younger patients, women, and those with an initial AMI compared with patients hospitalized on the same days during the control years. Appropriate caution needs to be exercised in the interpretation of these data, however, given the small number of cases of AMI occurring on September 11 and 12. ••• The results of this multihospital, population-based study suggest that the impact of the tragic events in New York City and the Pentagon on September 11, 2001 may have resulted in a transient increase in the occurrence of AMIs in the greater Worcester population on September 11 and on the day thereafter. The Worcester metropolitan area is ПЅ200 BRIEF REPORTS 259 miles from New York City and both of the planes that crashed into the twin towers on September 11 originated from Boston, Massachusetts. The frequency of AMIs in the weeks to months after September 11, 2001, was similar to earlier periods of this year and similar to the frequency of AMIs observed during the control years of 1991 to 1999. Because of the relatively small number of hospitalized cases of AMI occurring in the metropolitan Worcester population on a daily basis, it is difficult and tenuous to extrapolate this pattern of events to a cause– effect association. There are limited published reports on the effects of natural disasters, such as earthquakes, on the occurrence or death rates associated with atherosclerotic cardiovascular disease.4 – 8 There is an even sparser literature describing the effects of war or battlefield-type situations on the risk of acute cardiac disease in the civilian or military population.9,10 We were able to identify a small number of descriptive studies that examined the impact of the events of September 11 on hospital admissions for cardiac disease, or on cardiovascular-associated mortality, in New York City.11,12 A terrorist attack may be viewed in a relatively similar context to that of an earthquake or natural disaster, in that there is generally little warning that the event will occur. Our results suggest that the terrorist attacks on September 11 may have resulted in a transitory increase in the number of heart attacks occurring in greater Worcester residents. Although unknown, and if real, this may have resulted in the occurrence of a considerable number of additional heart attacks nationally in the few days after the events of September 11. Conversely, the deleterious health effects associated with the World Trade Center and Pentagon attacks may have been limited to persons residing on the East Coast of the United States. Although the play of chance cannot be completely ruled out given the small number of AMIs occurring in the greater Worcester population on a daily basis, these attacks may have placed undue psychological and emotional strain and stress in at-risk patients, resulting in an increased triggering of acute coronary events. High-risk cardiac patients may be at particularly increased risk for adverse coronary events after an acute 260 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 life stressor.13 Although suggestive evidence for the role of different triggering agents in the precipitation of AMI has been obtained from previous studies,14 it remains unclear what may have caused this increase in the number of AMIs in the greater Worcester population on September 11 and the day thereafter. The shock of the initial event, uncertainty with regard to one’s future, and a number of additional factors may have contributed to the observed increase in acute coronary events. Our findings reinforce the need for the exploration of different mechanisms through which acute stress, or other potential triggering agents, may result in the onset of AMI. 1. Goldberg RJ, Gore JM, Alpert JS, Dalen JE. Recent changes in the attack rates and survival rates of acute myocardial infarction (1975–1981): The Worcester Heart Attack Study. JAMA 1986;255:2774 –2779. 2. Goldberg RJ, Gore JM, Alpert JS, Dalen JE. Incidence and case fatality rates of acute myocardial infarction (1975–1984): The Worcester Heart Attack Study. Am Heart J 1988;115:761–767. 3. Goldberg RJ, Yarzebski J, Lessard D, Gore JM. A two-decades (1975–1995) long experience in the incidence, in-hospital and long-term case-fatality rates of acute myocardial infarction: a community-wide perspective. J Am Coll Cardiol 1999;33:1533–1539. 4. Kloner RA, Leor J, Poole WK, Perritt R. Population-based analysis of the effect of the Northridge earthquake on cardiac death in Los Angeles County, California. J Am Coll Cardiol 1997;30:1174 –1180. 5. Suzuki S, Sakamoto S, Miki T, Matsuo T. Hanshin-Awaji earthquake and acute myocardial infarction. Lancet 1995;345:981–982. 6. Leor J, Kloner RA. The Northridge earthquake as a trigger for acute myocardial infarction. Am J Cardiol 1996;77:1230 –1232. 7. Trichopoulos D, Katsouyanni K, Zavitsanos X, Tzonou A, Dalla-Vorgia P. Psychological stress and fatal heart attack: the Athens (1981) earthquake natural experiment. Lancet 1983;1:441– 444. 8. Dobson AJ, Alexander HM, Malcolm JA, Steele PL, Miles TA. Heart attacks and the Newcastle earthquake. Med J Aust 1991;155:757–761. 9. Meisel SR, Kutz I, Dayan KI, Pauzner H, Chetboun I, Arbel Y, David D. Effect of Iraqi missile war on incidence of acute myocardial infarction and sudden death in Israeli civilians. Lancet 1991;338:660 – 661. 10. Rumboldt Z, Guinio L, Miric D, Polic S, Bozic I, Tonkic A. War-stress induced medical emergencies in South Croatia. Lancet 1993;341:965–966. 11. Chi JS, Speakman MT, Poole WK, Kandefer SC, Kloner RA. Hospital admissions for cardiac events in New York City after September 11, 2001. Am J Cardiol 2003;92:61– 63. 12. Chi JS, Poole K, Kandefer SC, Kloner RA. Cardiovascular mortality in New York City after September 11, 2001. Am J Cardiol 2003;92:857– 861. 13. Qureshi EA, Merla V, Steinberg J, Rozanski A. Terrorism and the heart: implications for arrhythmogenesis and coronary artery disease. Cardiac Electrophys Rev 2003;7:80 – 84. 14. Muller JE, Abela GS, Nesto RW, Tofler GH. Triggers, acute risk factors, and vulnerable plaques: the lexicon of a new frontier. J Am Coll Cardiol 1994;23: 809 – 813. JANUARY 15, 2005 Characteristics and Outcomes of Patients With Acute Myocardial Infarction and Angiographically Normal Coronary Arteries Alf Inge Larsen, MD, PhD, P. Diane Galbraith, BN, William A. Ghali, MD, MPH, Colleen M. Norris, PhD, Michelle M. Graham, MD, and Merril L. Knudtson, MD, for the APPROACH Investigators* This study on patients undergoing coronary angiography for acute myocardial infarction demonstrated that 2.8% of patients had angiographically normal coronary arteries and that these patients have a better prognosis than patients with angiographically verified coronary artery disease. The trend toward a higher prevalence of malignancy in this unique patient group raises the possibility of malignancy-induced hypercoagulability or inflammation as an underlying etiologic factor. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:261–263) he aims of the present study were (1) to evaluate a cohort of patients with acute myocardial infarction T (AMI) who underwent subsequent coronary angiography to determine the proportion of patients with normal coronary arteries and (2) to compare the clinical characteristics and outcomes of AMI patients with normal coronary arteries with those of patients with varying degrees of coronary artery disease (CAD). ••• The Alberta Provincial Project for Outcome Assessment in Coronary Heart disease (APPROACH) is a From the Cardiology Division, University of Bergen, Central Hospital, Stavanger, Norway; the Department of Community Health Sciences and Centre for Health and Policy Studies, Calgary, Alberta; University of Alberta, Edmonton, Alberta; Division of Cardiology, University of Alberta Hospital, Edmonton, Alberta; and Foothills Medical Centre, Calgary, Alberta, Canada. APPROACH was initially funded with a grant from the W. Garfield Weston Foundation. The ongoing operation has been made possible by contributions from Merck Frosst Canada Inc., Pointe-Claire, Dorval, Quebec; Monsanto Canada Inc.– Searle, Mississauga, Ontario; Eli Lilly Canada Inc., Toronto, Ontario; Guidant Corporation, Santa Clara, California; Boston Scientific Ltd., Mississauga, Ontario; Hoffmann-La Roche Ltd., Mississauga, Ontario; and Johnson & Johnson Inc.–Cordis, Markham, Ontario, Canada. Dr. Knudtson received partial support from the Libin Cardiovascular Trust Fund, University of Calgary, Calgary, Alberta, Canada. Dr. Ghali was supported by a Government of Canada Research Chair in Health Services Research and by a Health Scholar Award from the Alberta Heritage Foundation for Medical Research, Edmonton, Alberta. Miss Galbraith and Dr. Norris were supported by Canadian Cardiovascular Outcomes Research Team (www.torchalberta.ca) student fellowship, funded by an operating grant from the Canadian Institutes of Health Research and the Heart and Stroke Foundation, and Tomorrows Research Cardiovascular Health Professionals (www.torchalberta.ca) student fellowship. Dr. Knudtson’s address is: University of Calgary/Foothills Hospital, Division of Cardiology, 1403-29 Street NW, Calgary, Alberta, Canada, T2N 2T9. E-mail: [email protected]. Manuscript received April 4, 2004; revised manuscript received and accepted August 31, 2004. *See Appendix for members of the APPROACH Clinical Steering Committee. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 clinical data collection initiative in which data are prospectively compiled for all adult patients undergoing cardiac catheterization in the province of Alberta, Canada.1 The database and its contents are described in detail elsewhere.2– 4 The database also tracks subsequent revascularization procedures. The extent of CAD is documented in a computerized data template (HeartView, Heartware, Durham, North Carolina). For this analysis, patients were classified into 5 subgroups by severity of CAD: (1) normal coronary arteries (no coronary artery stenoses), (2) minor disease (ПЅ50% diameter stenoses), (3) low-risk disease (1- or 2-vessel disease with Пѕ50% diameter stenoses), (4) high-risk disease (2-vessel disease with proximal left anterior descending artery or 3-vessel disease), and (5) left main disease. We focused specifically on the subset of patients who underwent cardiac catheterization with an indication of AMI. This indication was defined in the registry by the clinician who performed the procedure, and then data were further enhanced by a post hoc data-merging procedure in which registry data were merged to International Classification of Disease-9th Revision (ICD-9) administrative data.5 This process provides a 2-level verification of the indication for catheterization. Mortality was ascertained through semi-annual merging with data from the Alberta Bureau of Vital Statistics Records. The emphasis of our analysis was a description of clinical characteristics of patients with no angiographically obvious coronary artery lesions, referred to herein as the “normal” group, but we also documented the characteristics of patients with increasing degrees of CAD. We used chi-square tests, Fisher’s exact tests, or t tests for statistical comparisons of clinical characteristics for the normal coronary arteries group versus all other groups combined. Survival analysis after AMI was estimated using the KaplanMeier method with log-rank tests to compare survival for the normal coronary arteries group versus all other groups combined. All statistical analyses were performed using SPSS software, version 11.0 (SPSS Inc., Chicago Illinois). From January 1995 to December 2000, 9,796 patients underwent coronary angiography with AMI listed as the indication for the procedure. Of these, 273 (2.8%, 95% confidence interval 2.5 to 3.1) had angiographically normal coronary arteries. Baseline characteristics of the patient groups according to increasing degrees of CAD are listed in Table 1. Compared with patients with identifiable 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.014 261 TABLE 1 Demographics and Risk Factors for Patients With Acute Myocardial Infarction (AMI) and Normal Coronary Arteries Compared With Patients With AMI and Identifiable Coronary Artery Disease Variable Mean age (yrs) Patients Пѕ75 yrs Patients Х…45 yrs Men Hypertension Hyperlipidemia Diabetes mellitus Smokers Current smoker Previous thrombolytic Previous PCI Cerebrovascular disease Congestive heart failure Peripheral vascular disease Creatine Пѕ200 mmol/L Dialysis Chronic lung disease Liver/gastroIntestinal Malignancy Ejection fraction ПЅ30 30–50 Пѕ50 Not done (unstable) Missing Normal (n П 273) ПЅ50 (n П 452) Low Risk (n П 4,942) High Risk (n П 3,430) Left Main (n П 699) p Value* 49.4 4.4% 39.6% 62.3% 34.1% 23.1% 3.7% 38.5% 31.1% 10.6% 0.0% 4.8% 11.0% 1.8% 2.2% 0.4% 16.5% 4.4% 5.1% 58.1 12.2% 17.5% 60.2% 45.6% 40.9% 13.1% 44.7% 38.3% 20.6% 2.0% 5.1% 15.5% 6.0% 1.5% 1.1% 15.7% 7.3% 3.3% 59.7 10.7% 11.7% 73.9% 43.7% 45.1% 14.9% 42.8% 42.8% 25.5% 5.7% 4.2% 14.3% 5.0% 2.0% 0.8% 9.3% 3.7% 3.0% 65.1 21.4% 4.5% 76.6% 55.9% 45.8% 26.4% 42.8% 31.7% 21.1% 7.8% 7.8% 25.5% 8.9% 4.7% 1.7% 11.8% 5.1% 3.3% 68.5 30.5% 2.0% 78.0% 56.4% 43.9% 25.8% 43.9% 26.6% 15.3% 5.6% 14.2% 36.1% 12.6% 7.0% 2.3% 16.7% 5.4% 5.3% 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.154 0.036 0.000 0.000 0.373 0.000 0.000 0.389 0.267 0.008 1.0 0.090 5.5% 10.9% 70.3% 4.0% 9.2% 2.4% 16.6% 67.9% 5.8% 7.3% 3.0% 24.7% 54.9% 8.5% 8.9% 7.7% 31.1% 40.4% 9.6% 11.2% 10.7% 34.9% 35.8% 11.0% 7.6% *The p values represent comparisons between normal versus all other groups combined. PCI П percutaneous coronary intervention. TABLE 2 Treatment Strategy and Mortality in Acute Myocardial Infarction (AMI) Patients With Normal Coronary Arteries Compared With AMI Patients With Identifiable Coronary Artery Disease Variable Treatment Medical Coronary bypass PCI Coronary bypass 30 d Coronary bypass 1 yr* PCI 30 d PCI 1 yr* Death 30 d Death 1 yr* Normal (n П 274) ПЅ50% (n П 452) Low Risk (n П 4,942) High Risk (n П 3,430) Left Main (n П 699) Total (n П 9,796) 100% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.8% 4.0% 90.7% 0.4% 8.9% 0.2% 0.4% 6.7% 8.2% 0.7% 2.9% 27.3% 4.2% 68.5% 3.2% 5.5% 64.3% 68.0% 1.7% 3.3% 25.5% 31.7% 42.9% 22.2% 34.2% 40.1% 42.9% 5.1% 9.4% 22.7% 67.4% 9.9% 59.4% 69.4% 9.7% 10.7% 11.6% 19.2% 31.3% 18.0% 50.7% 13.6% 19.8% 47.5% 50.5% 3.5% 6.6% *One-year data includes 30-day data. Abbreviation as in Table 1. CAD, the mean age was considerably lower in the normal coronary group (p ПЅ0.001). Compared with patients with identifiable CAD, prevalence of hypertension, hyperlipidemia, diabetes mellitus, previous thrombolytic therapy, and history of congestive heart failure and peripheral vascular disease occurred less often in the normal group. There was also an intriguing trend toward a difference in the prevalence of malignancy between normal patients and patients in all other groups. Patients in the latter groups were older and therefore may have been expected to have a higher prevalence of malignancy. The specific types of malignancy in the normal group were prostatic cancer (n П 2), skin cancer (n П 1), cervical cancer (n П 2), breast cancer (n П 3), malignant myeloma 262 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 (n П 1), Hodgkins’ disease (n П 1), and bladder cancer (n П 1). In 3 of the patients, the type of malignancy was not available. Table 2 demonstrates treatment strategy and outcomes at 30 days and 1 year for each of the patient groups. No patient with normal angiograms after AMI underwent a revascularization procedure within 1 year. Of the total group of patients with AMI who underwent angiography, 68.7% underwent revascularization. Mortality in normal patients was 1.8% at 30 days and 4.0% at 1 year. In all, 11 patients with normal coronary arteries died within the first year. These mortality rates were generally similar to those for patients with stenosis ПЅ50% and the low-risk anatomy groups, but signifiJANUARY 15, 2005 the probability that this is due to chance alone is relatively low. At a minimum, this finding warrants further exploration. Our other notable finding, diagnostic and etiologic considerations aside, is that patients presenting with the clinical syndrome of AMI with normal coranary arteries have a significantly better prognosis than do those with identified CAD. FIGURE 1. Kaplan-Meier survival curves for patients with AMI with normal coronary arteries compared with patients with AMI who have coronary disease. cantly less than the high-risk and left main groups (Table 2). Figure 1 shows survival curves for the group with normal coronary arteries versus all other groups combined extending to up to 7 years and confirms the general survival picture seen in the first year (log-rank test for difference in survival across groups, p П 0.017). The median time of follow-up to assess survival was 3.3 years (range 0 to 7.0). ••• These findings indicate that approximately 2.8% of patients undergoing coronary angiography for indication of an AMI have normal coronary arteries. Patients in this group are younger and less likely to have traditional risk factors for atherosclerosis, such as hypertension, hyperlipidemia, and diabetes mellitus, than are patients with documented coronary disease. Interestingly, there was a trend toward a higher prevalence of malignancy in the normal group than in other groups, although normal patients were younger and therefore would be expected to be at lower risk for malignancy. Other investigators6 –11 speculated on potential etiologies and/or mechanisms underlying the clinical syndrome of AMI with normal coronary arteries. Possible explanations include cocaine ingestion, coronary spasm, misdiagnosed myocarditis, and hypercoagulability states (both inherited and acquired). Our finding of a trend toward a higher prevalence of malignancy in the group with normal coronary arteries raises the possibility that malignancy, regardless of the type of cancer, with associated coagulation disorders and/or inflammation, may place these patients at higher risk of thromboembolic events. Another possibility is that aggressive cancer treatments, such as chemotherapy, radiation therapy, and major surgery, may predispose to physiologically stressed states and induce AMI.12–14 Our finding does not meet traditional statistical significance (p ПЅ0.05). However, at p П 0.09, Acknowledgment: We appreciate the assistance of the Calgary Health Region and the Capital Health Authority for supporting on-line data entry. We gratefully acknowledge personnel in the cardiac catheterization laboratories for their diligence in data collection and entry: Holy Cross Hospital, Foothills Medical Center, Royal Alexandra Hospital, and the University of Alberta Hospital. We thank Jason Badry, BSc, our systems analyst, who is responsible for software development and maintenance. APPENDIX Members of the APPROACH Clinical Steering Committee: William Hui, MD (chair); Stephen Archer, MD; Micheal Curtis, MD; William A. Ghali, MD; Michelle Graham, MD; Merril L. Knudtson, MD; Arvind Koshal, MD; Andrew Maitland, MD; L. Brent Mitchell, MD; and Ross Tsuyuki, MD. 1. Ghali WA, Knudtson ML, for the APPROACH Investigators. Overview of “APPROACH”—the Alberta Provincial Program for outcome assessment in coronary heart disease. Can J Cardiol 2000;16:1225–1230. 2. Ghali WA, Quan H, Brant R, van Melle G, Norris CM, Faris PD, Knudtson ML, for the APPROACH Investigators. A comparison of two methods for calculating adjusted survival curves from proportional hazards models. JAMA 2001;286:1494 –1497. 3. Ghali WA, Faris PD, Galbraith PD, Norris CM, Curtis MJ, Saunders LD, Dzavik V, Mitchell LB, Knudtson ML. Sex differences in access to coronary revascularization after cardiac catheterization: importance of detailed clinical data. Ann Intern Med 2002;136:723–732. 4. Graham MM, Ghali WA, Faris PD, Galbraith PD, Norris CM, Merril Knudtson, for the APPROACH Investigators. Survival after coronary revascularization in the elderly. Circulation 2002;105:2378 –2384. 5. Norris CM, Ghali WA, Knudtson ML, Saunders LD, Naylor CD. Dealing with missing data in observational health care outcome analysis. J Clin Epidemiol 2000;53:377–383. 6. Minor RL Jr, Scott BD, Brown DD, Winniford MD. Cocaine-induced myocardial infarction in patients with normal coronary arteries. Ann Intern Med 1991;115:797– 806. 7. Zurbano MJ, Heras M, Rigol M, Roig E, Epelde F, Miranda F, Sanz G, Escolar G, Ordinas A. Cocaine administration enhances platelet reactivity to subendothelial components: studies in a pig model. Eur J Clin Invest 1997;27:116 –120. 8. Pierron F, Panagides D, Bonnet JL, Yvorra S, Desfossez L, Bory M. Spasm of normal or irregular coronary arteries. Long-term outcome of 277 patients. Arch Mal Coeur Vaiss 1995;88:1819 –1825. 9. Da Costa A, Tardy-Poncet B, Isaaz K, Cerisier A, Mismetti P, Simitsidis S, Reynaud J, Tardy B, Piot M, Decousus H, Guyotat D. Prevalence of factor V leiden and other inherited thrombophilias in young patients with myocardial infarction and normal coronary arteries. Heart 1998;80:338 –340. 10. Mansourati J, Da Costa A, Munier S, Mercier B, Tardy B, Ferec C, Isaaz K, Blanc JJ. Prevalence of factor V Leiden in patients with myocardial infarction and normal coronary angiography. Thromb Haemost 2000;83:822– 825. 11. Caussin C, Ohanessian A, Lancelin B, Rahal S, Hennequin R, Dambrin G, Brenot P, Angel CY, Paul JF. Coronary plaque burden detected by multislice computed tomography after acute myocardial infarction with near-normal coronary arteries by angiography. Am J Cardiol 2003;92:849 – 852. 12. Zeymer U, Neuhaus KL. Acute coronary thrombosis and myocardial ischemia following chemotherapy of Hodgkin’s disease. Onkologie 1990;13:221–224. 13. Kruit WH, Punt KJ, Goey SH, de Mulder PH, van Hoogenhuyze DC, Henzen-Logmans SC, Stoter G. Cardiotoxicity as a dose-limiting factor in a schedule of high dose bolus therapy with interleukin-2 and alpha-interferon: an unexpectedly frequent complication. Cancer 1994;74:2850 –2856. 14. Scholz KH, Herrmann C, Tebbe U, Chemnitius JM, Helmchen U, Kreuzer H. Myocardial infarction in young patients with Hodgkin’s disease—potential pathogenic role of radiotherapy, chemotherapy, and splenectomy. Clin Invest 1993;71:57– 64. BRIEF REPORTS 263 Long-Term Safety and Efficacy of High-Dose Atorvastatin Treatment in Patients With Familial Hypercholesterolemia Sanne van Wissen, MD, PhD, Tineke J. Smilde, MD, PhD, Mieke D. Trip, MD, Anton F.H. Stalenhoef, MD, PhD, and John J.P. Kastelein, MD, PhD In the 2-year Atorvastatin versus Simvastatin on Atherosclerosis Progression extension study, patients with familial hypercholesterolemia who continued to take atorvastatin 80 mg for an additional 2 years had complete arrest of the progression of mean carotid intima-media thickness (0.89 mm at the start vs 0.90 mm at the end of the study, p ‫ ؍‬0.58). In contrast, patients previously taking simvastatin 40 mg had significant regression of intima-media thickness (0.95 mm at the start vs 0.92 mm at the end of the study, p ‫ ؍‬0.01). Therefore, both placebo- and statintreated patients with familial hypercholesterolemia are best treated with high-dose atorvastatin, a therapeutic regimen that induces atherosclerosis regression and is safe and well tolerated over a 4-year period. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:264 –266) reviously, we demonstrated that high-dose atorvastatin therapy was effective in atherosclerosis P regression in the 2-year Atorvastatin versus Simvastatin on Atherosclerosis Progression (ASAP) study.1 In this study, patients with familial hypercholesterolemia (FH) were treated with either atorvastatin 80 mg/day or simvastatin 40 mg/day. Patients with FH have sharply increased low-density lipoprotein (LDL) cholesterol levels, resulting from defects in the LDL receptor, which leads to an increased risk for cardiovascular disease. Patients with FH are therefore considered excellent models for atherogenesis. The effect of statin intervention on the atherosclerotic process was monitored in this trial by measuring intima-media thickness (IMT), a validated surrogate marker for future cardiovascular end points.2 After completing the first 2-year period of the ASAP trial, patients either continued with or were switched to atorvastatin 80 mg/day in a 2-year extension study. This provided us with the opportunity (1) to study change in IMT in patients previously randomized to simvastatin and (2) to obtain long-term follow-up IMT data for patients origiFrom the Department of Vascular Medicine, Academic Medical Center, University of Amsterdam, Amsterdam, and Department of Medicine, Division of General Internal Medicine, University Medical Center Nijmegen, Nijmegen, The Netherlands. This study was supported by Pfizer, Capelle aan de Yssel, The Netherlands. Dr. van Wissen’s address is: Academic Medical Center, Department of Vascular Medicine, Room F4 –159.2, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. E-mail: [email protected]. Manuscript received May 6, 2004; revised manuscript received and accepted September 7, 2004. 264 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 PhD, nally randomized to high-dose atorvastatin. Finally, this study provides the first report on 4-year safety data of the highest registered dose of atorvastatin. ••• Patients in the ASAP extension study were previously randomized to either atorvastatin 80 mg/day or simvastatin 40 mg/day as previously described.1 In short, participating patients (aged 30 to 70 years) were diagnosed with FH and had LDL cholesterol levels Пѕ4.5 mmol/L (174 mg/dl). After completion of the controlled and randomized period of the ASAP study, patients were invited to continue in a 2-year extension study with atorvastatin 80 mg/day. The ethics committees approved the protocol and written informed consent was obtained from every patient. The last visit of the ASAP study was the first of the extension study. Laboratory parameters were measured at 12 weeks after the start of the study and every 24 weeks thereafter. At each visit, a brief physical examination and a laboratory analysis was performed, compliance was checked, and standard dietary instructions were given. The ultrasound scanning protocol and its reproducibility have been described.1 Ultrasound examinations were performed every year with a Biosound Phase-2 real-time scanner (BiosoundEsaote, Indianapolis, Indiana) equipped with a 10-MHz transducer. Three segments of 10 mm of the intima-media were scanned bilaterally, ie., the distal portion of the common carotid artery, the carotid bifurcation, and the proximal portion of the internal carotid artery. Images were analyzed with a semiautomatic program (Eurequa, TSA Company, Meudon, France). The IMT readers were blinded to all clinical information, including the original treatment to which the patients were randomized. Differences between treatment groups in IMT change from baseline after 24 months and treatment differences in the percent changes from baseline of lipoprotein parameters were analyzed with analysis of covariance. Analyses were done on an intent-to-treat basis. The last observation carried forward was used for patients who did not complete the study or who had missing values. Statistical analyses were done with SAS software (version 6.12, SAS Inc., Cary, North Carolina). Of the 280 patients who completed the ASAP study, 255 entered the extension study. Reasons for not entering the extension study were desire to become pregnant (n П 1), travel distance too great (n П 2), physical or mental discomfort (n П 6), personal reasons (n П 5), alcohol abuse (n П 7), dislike of atorvastatin (n П 1), too little compliance (n П 2) and visual complaints (n П 1). Of the 255 patients in the extension study, 30 patients dis0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.015 plained of muscle ache. In 7 patients (3 in the atorvastatin group and 4 in the group previously treated with Atorvastatin Simvastatin simvastatin), aspertate aminotrans(n П 132) (n П 123) ferase and alanine transferase inStart End Start End creased to Пѕ3 times the upper limit of normal (Пѕ120 and 135 U/L, reTC (mg/dl) 215 П® 42 223 П® 46* 251 П® 46 220 П® 45* spectively). TG (mg/dl) 45 П® 59 53 П® 95* 54 П® 69 45 П® 56 HDL (mg/dl) 52 П® 16 50 П® 14* 51 П® 13 51 П® 12 Percent withdrawals due to adverse LDL (mg/dl) 143 П® 40 152 П® 42* 177 П® 45 150 П® 43* events was small for groups and not significant: 5 patients (3.8%) in the Values are expressed as mean П® SD. *p ПЅ0.05 for percent change within 1 treatment arm. atorvastatin group versus 3 patients HDL П high-density lipoprotein cholesterol; TC П total cholesterol; TG П triglycerides. (2.4%) in the simvastatin group. The primary results of ASAP have been previously published.1 In short, in the atorvastatin 80-mg group, mean carotid IMT decreased by 0.031 mm after 2 years (p П 0.002); in contrast, in the simvastatin 40 mg group, IMT increased by 0.036 mm (p ПЅ0.001).1 In the extension study, patients taking atorvastatin 80 mg for an additional 2 years (4 years total) exhibited a complete arrest of the progression of mean carotid IMT (0.89 mm at the start vs 0.90 mm at the end of the extension period; p П 0.58), whereas, in contrast, patients who were previously taking simvastatin 40 mg had significant regression of IMT after 2 years of treatment with atorvastatin 80 mg (0.95 mm at the start vs 0.92 mm at the end of the extension period; p П 0.01) (Figure 1). Mean carotid IMT between the 2 groups (0.90 vs 0.92 mm) no longer differed at 4 years (p П 0.06). In a stepwise regression FIGURE 1. Changes (mean ШЋ SE) in mean carotid IMT in the analysis, the change in IMT over 4 years was only ASAP study. Gray squares, atorvastatin; black triangles, simvarelated to baseline IMT, was negatively associated statin. with age (implying a greater increase in older patients) and gender (implying a greater increase in men). ••• continued (reasons were travel distance too great [n П We conclude that patients with FH are best treated 16] personal reasons [n П 8], increased transaminases [n П 4], muscle ache [n П 1], and skin rash [n П 1]). with high-dose atorvastatin therapy, a therapeutic regPatient characteristics between both study groups were imen that induces regression of atherosclerosis and is comparable with regard to demographic variables and safe and well tolerated over a period of Х†4 years. The results of the ASAP are supported by the recent vital signs. Changes in lipid and lipoprotein levels are listed in outcome of the Arterial Biology for the Investigation of Table 1. In patients previously treated with simvasta- the Treatment Effects of Reducing Cholesterol trial, tin 40 mg, total cholesterol and LDL cholesterol de- which assessed the effects of atorvastatin 80 mg versus creased significantly after the change to atorvastatin pravastatin 40 mg on IMT in a mixed hyperlipidemic 80 mg. In patients treated with atorvastatin 80 mg, population and also reported regression of carotid IMT in there was a small but significant increase in total the atorvastatin group.3 In another recent study, the Reversal of Atherosclerosis with Aggressive Lipid Lowercholesterol, LDL cholesterol, and triglycerides. The overall observed adverse event rates were ing trial, using intracoronary intravascular ultrasound in comparable between groups, although they showed a patients with cardiovascular disease, atorvastatin 80 mg trend toward being higher in the group previously also arrested progression of atherosclerosis in the corotreated with simvastatin. Muscle-related complaints nary arteries, whereas pravastatin 40 mg led to progreswere 9% versus 14% in the atorvastatin group and the sion of atherosclerotic vascular disease.4 Furthermore, group previously treated with simvastatin, respec- and more importantly, the use of high dosages of atortively. The percentages for abdominal complaints vastatin decreased the number of major cardiovascular were 18% and 29% for the atorvastatin and simvasta- events compared with moderate cholesterol reductions tin groups, respectively. All these differences were not with pravastatin 40 mg in patients with a recent acute coronary syndrome in the Pravastatin or Atorvastatin statistically significant. No significant differences with regard to laboratory Evaluation and Infection Therapy trial.5 However, folparameters were noticed. In 8 patients (4 in each low-up of these 3 last trials ran from 12 to approximately group), creatine kinase increased to Пѕ3 times the 24 months, whereas follow-up in the ASAP was 48 upper limit of normal (Пѕ540 U/L). No patient com- months.3–5 TABLE 1 Plasma Lipid Levels in Patients With Familial Hypercholesterolemia at Start and End of Extension Study BRIEF REPORTS 265 These results make it likely that the initial regression and subsequent arrest of progression due to 80 mg of atorvastatin will be maintained over protracted periods of time. This is compatible with current guidelines. We could not show further IMT regression in patients who were previously randomized to atorvastatin. It could be hypothesized that the maximum potential of statin treatment is reached in these FH heterozygotes. If this is the case, additional interventions may be required to optimize the treatment of dyslipidemia in this disorder. Such therapeutic modalities would include even more potent LDL lowering, increasing high-density lipoprotein cholesterol levels, or inhibition of very low-density lipoprotein production. Cholesterol absorption inhibitors (e.g., ezetimibe) may lower LDL cholesterol levels by an additional 20%,6 cholesteryl ester transfer protein inhibitors may increase high-density lipoprotein cholesterol by 30% to 40%,7 and peroxisome proliferator-activated receptor agonists decrease very low-density lipoprotein production and increase apolipoprotein A-I synthesis.8 1. Smilde TJ, van Wissen S, Wollersheim H, Trip MD, Kastelein JJP, Stalenhoef AFH. Effect of aggressive versus conventional lipid lowering on atherosclerosis progression in familial hypercholesterolaemia (ASAP): a prospective, randomised, double-blind trial. Lancet 2001;357:577–581. 2. Lekakis JP, Papamichael CM, Cimponeriu AT, Stamatelopoulos KS, Papaioannou TG, Kanakakis J, Alevizaki MK, Papapanagiotou A, Kalofoutis AT, Stamatelopoulos SF. Atherosclerotic changes of extracoronary arteries are associated with the extent of coronary atherosclerosis. Am J Cardiol 2000;85:949 –952. 3. Taylor AJ, Kent SM, Flaherty PJ, Coyle LC, Markwood TT, Vernalis MN. ARBITER: Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol: a randomized trial comparing the effects of atorvastatin and pravastatin on carotid intima medial thickness. Circulation 2002;106:2055–2060. 4. Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA, Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN. Effect of intensive compared with moderate lipid-lowering therapy on progression of coronary atherosclerosis: a randomized controlled trial. JAMA 2004;291:1071–1080. 5. Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, Belder R, Joyal SV, Hill KA, Pfeffer MA, Skene AM. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med 2004; 350:1495–1504. 6. Evans M, Roberts A, Rees A. The future direction of cholesterol-lowering therapy. Curr Opin Lipidol 2002;13:663– 669. 7. de Grooth GJ, Kuivenhoven JA, Stalenhoef AF, de Graaf J, Zwinderman AH, Posma JL, van Tol A, Kastelein JJ. Efficacy and safety of a novel cholesteryl ester transfer protein inhibitor, JTT-705, in humans: a randomized phase II doseresponse study. Circulation 2002;105:2159 –2165. 8. Fruchart JC, Duriez P, Staels B. Peroxisome proliferator-activated receptoralpha activators regulate genes governing lipoprotein metabolism, vascular inflammation and atherosclerosis. Curr Opin Lipidol 1999;10:245–257. Effect of Medical and Surgical Weight Loss on Endothelial Vasomotor Function in Obese Patients Noyan Gokce, MD, Joseph A. Vita, MD, Marie McDonnell, MD, Armour R. Forse, Nawfal Istfan, MD, PhD, Maria Stoeckl, BA, Izabella Lipinska, PhD, John F. Keaney, Jr., MD, and Caroline M. Apovian, MD We prospectively examined brachial artery endothelial function using vascular ultrasound in 41 obese subjects treated with medical or surgical (gastric bypass) weight loss interventions. Surgical intervention produced greater weight loss and more pronounced improvement in endothelial function than medical treatment alone. Improved endothelial function with weight loss correlated strongly with fasting glucose but not with alteration in blood pressure, lipids, degree of weight loss, or plasma resistin concentrations. These data demonstrate that weight loss in markedly obese patients improves endothelial function and glycemic control that may represent important mecha- From the Evans Department of Medicine, Cardiology and Endocrinology Sections, the Department of Surgery, Gastrointestinal Section, and the Whitaker Cardiovascular Institute, Boston University School of Medicine, Boston, Massachusetts. This work was supported by Grant HL 74097 to Drs. Gokce, Forse, and Apovian from the National Institutes of Health, Bethesda, Maryland. Dr. Gokce is the recipient of a mentored patient-oriented research career transition Award HL04425 from the National Institutes of Health, Bethesda, Maryland. Dr. Gokce’s address is: Section of Cardiology, Boston Medical Center, C-818, 88 East Newton Street, Boston, Massachusetts 02118. E-mail: [email protected]. Manuscript received June 2, 2004; revised manuscript received and accepted September 8, 2004. 266 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, nisms of the cardiovascular benefit associated with weight reduction. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:266 –268) everal recent studies have suggested that weight reduction improves endothelial function in obese S subjects, although the importance of mode and extent of weight loss remain largely unexplored.1–3 The purpose of this study was to determine the relative benefit of marked weight loss produced by bariatric surgery compared with more modest weight loss induced with medical therapy and to investigate the mechanisms of benefit. ••• We enrolled consecutive obese patients (body mass index [BMI] Х†30 kg/m2) receiving care at the Boston Medical Center Nutrition and Weight Management Center, which provides comprehensive dietary, medical, behavioral, and surgical treatments to promote weight loss. All subjects provided written informed consent and the study was approved by the Boston Medical Center Institutional Review Board. Each subject made a visit at baseline and Х†3 months after initiation of medical or surgical weight loss treatment. During each visit, endothelium-dependent flow-mediated dilation (FMD) and endothelium-independent, nitroglycerin (NTG)-mediated dilation of the brachial artery were examined in the upper arm using an es0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.016 loss were determined by the Boston Medical Center clinical laboratory. Medical Therapy Surgical Therapy Serum insulin levels were measured (n П 17) (n П 24) p Value using chemiluminometric methods (Quest Diagnostics, Cambridge, Age (yrs) 44 П® 9 44 П® 10 0.89 Body weight (kg) 112 П® 32 142 П® 26 0.002* Massachusetts), and insulin resis42 П® 10 50 П® 8 0.01* Body mass index (kg/m2) tance was estimated using the previHemoglobin A1C (%) 5.8 П® 0.8 6.1 П® 1.6 0.60 ously validated homeostasis model Women 82% 63% 0.30 assessment (HOMA).8 Plasma resisDiabetes mellitus* 24% 29% 0.74 Hypertension†41% 50% 0.81 tin concentrations were measured usHypercholesterolemia 35% 50% 0.54 ing a commercially available enTobacco use 24% 8% 0.21 zyme-linked immunosorbent assay Coronary artery disease 6% 4% 0.99 kit (ALPCO Diagnostics, Windham, Lipid lowering therapy 24% 21% 0.99 New Hampshire).9 ACE inhibitor or ARB therapy 18% 29% 0.48 All data are presented as mean П® *Diabetes mellitus was defined by history or fasting glucose Пѕ126 mg/dl. †SD, unless otherwise indicated. BaseHypertension was defined by history or blood pressure Пѕ140/90 mm Hg. line clinical characteristics were comACE П angiotensin-converting enzyme; ARB П angiotensin-receptor blocker. pared using unpaired t tests, and the chi-square or Fisher’s exact test as appropriate. The effect of weight loss on vascular, hemodynamic, and metabolic parameters within each group was examined using paired t tests. Treatment effects between different groups were examined using 2-way repeated measures analysis of variance (ANOVA) with post hoc Tukey’s comparison. Correlations were examined using linear regression analysis. We prospectively followed a total of 41 subjects who achieved Х†5% weight reduction with dietary, medical, and behavioral intervention (medical therapy group, n П 17) or gastric bypass surgery (surgical therapy group, n П 24) during the investigation period. Baseline clinical characteristics are listed in Table 1. As expected, the surgical treatment group had higher weight and BMI; otherwise, there were no significant differences between the 2 groups. Baseline brachial artery FMD was similar in the medical (7.3 П® 5%) and surgical (6.8 П® 4.0%) FIGURE 1. Effect of medical and surgical (gastric bypass) weight loss groups (p П 0.75) (Figure 1). Patients in the weight loss on brachial artery FMD in obese patients (n ‫؍‬ medical group sustained an average 10% (range 5% to 41). Examinations were performed at baseline (black bars) 18%) decrease in mean body weight and BMI, from 112 and after medical and surgical weight loss interventions П® 32 to 101 П® 29 kg (p ПЅ0.001) and from 42 П® 10 to (white bars). Surgical gastric bypass was more effective in improving brachial dilation than medical therapy (p ‫ ؍‬0.004). 38 П® 10 kg/m2 (p ПЅ0.001), respectively. This effect was Data are presented as mean ШЋ SEM. associated with an increase in FMD to 9.6 П® 4.3% (p П 0.08). As expected, surgery was more effective in protablished noninvasive, standardized, blinded method moting weight loss than medical or dietary interventions of vascular ultrasound as previously described.4 After alone. Bariatric surgery produced an average 25% (range the baseline visit, all subjects initiated a clinically 9% to 49%) decrease in mean body weight (142 П® 26 to 23 kg, p ПЅ0.001) and BMI (50 П® 8 to 37 П® 7 prescribed weight loss program that included thera- 107 П® 2 kg/m , p ПЅ0.001). Gastric bypass surgery was associated peutic medical and/or surgical intervention. Medical with a significant improvement in brachial artery FMD, therapy consisted of a comprehensive dietary, behavioral, and educational treatment regimen based on to 10.2% П® 4.0% (p П 0.006 compared with baseline), National Heart, Lung, and Blood Institute guide- and was more effective than medical therapy in improvlines.5,6 Increasing physical activity was encouraged, ing endothelial function (p П 0.004 by ANOVA) (Figure but specifically tailored exercise regimens were not 1). In a subset of patients who underwent examination of prescribed. We also followed patients with clinically non– endothelium-dependent dilator function, weight severe obesity defined as BMI Х†40 or Х†35 kg/m2 modification produced no significant effect on NTGwith co-morbid conditions who underwent gastric by- mediated dilation (12.6 П® 5.6% vs 11 П® 6.8%, p П pass (bariatric) surgery as weight reduction treat- 0.62). In 22 separate subjects with similar clinical demoment.7 The standard operation involved Roux-en-Y graphics (data not shown) who failed to lose weight, gastric bypass surgery with an isolated gastric pouch endothelial function remained unchanged (FMD 7.5 П® 4.3% at baseline vs 7.3 П® 4.9% at follow-up, p П 0.87). of ПЅ2 ounces and a 100-cm distal bowel limb. Biochemical analyses at baseline and after weight Correlation between percent weight reduction and FMD TABLE 1 Baseline Clinical Characteristics BRIEF REPORTS 267 was ПЄ0.17 (p П 0.30); percent change in BMI with FMD was ПЄ0.15 (p П 0.35). To investigate the potential mechanisms that could account for the beneficial effects of weight loss on endothelial function, we examined blood pressure, lipid profiles, fasting glucose and insulin, and HOMA before and after weight loss. Blood pressure did not change significantly with weight loss in this group of patients. Gastric bypass surgery resulted in significant decreases in total, low-density cholesterol, and highdensity lipoprotein cholesterol levels that were not observed with more modest medical weight loss intervention. In contrast, both weight reduction strategies significantly improved glycemic indexes. In the medical treatment group, fasting glucose and insulin decreased from 104 П® 24 mg/dl and 16 П® 10 вђ®IU/ml at baseline, to 93 П® 16 mg/dl and 11 П® 5 вђ®IU/ml after weight loss, respectively (p ПЅ0.05). Similarly, HOMA decreased from 4.5 П® 3.7 to 2.6 П® 1.5 (p ПЅ0.05) with medical weight loss. Surgical treatment was more effective in lowering glucose (119 П® 39 to 90 П® 18 mg/dl), insulin (22 П® 14 to 7 П® 5 вђ®IU/ml) and HOMA (6.5 П® 4.7 to 1.6 П® 1.5) than medical or dietary interventions alone (p ПЅ0.001 for all variables by ANOVA). The only significant correlation for improved endothelial function was fasting glucose (r П ПЄ0.46, p П 0.004). We observed no significant differences in resistin levels at baseline or after weight loss (4.3 П® 3.6 vs 4.8 П® 4.1 ng/ml, p П 0.29, n П 18). ••• This prospective study demonstrated that endothelium-dependent flow-mediated vasodilation is impaired in obese patients and improves significantly with weight loss intervention. Gastric bypass surgery was associated with more dramatic weight loss, more pronounced improvement in endothelial function, and greater metabolic changes than medical therapy alone. Improved vascular function correlated significantly with reduced glucose concentration but was not linked to magnitude of weight reduction per se. Thus, the study suggests that altered glycemic status associated with therapeutic weight loss may be an important determinant of vascular homeostasis in obese patients. Several previous studies have examined endothelial function in the setting of dietary and behavioral weight reduction strategies, but to our knowledge, the present investigation is the first to incorporate surgically treated patients. In 2 previous studies, the combination of medical weight loss and regular exercise significantly improved vascular reactivity1,2; however, the benefit may have been partly due to a therapeutic vascular effect of standardized exercise regimens.10 The mechanisms for the effect of weight reduction on endothelial function are incompletely understood and likely multifactorial, relating to changes in glucose tolerance, lipid profiles, and hypertension.5,11 In agreement with 2 previous studies,1,3 our study suggests that an alteration in glycemic status after weight loss may primarily account for the benefits on endothelial function and extends these findings to gastric bypass patients who sustained dramatic reductions in body weight. Our present investigation failed to provide evidence for a role of resistin, an adipocyte268 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 derived peptide hormone implicated in mechanisms of obesity-related insulin resistance12 as a key modulator of endothelial function. Our finding that surgical treatment produced greater weight loss and improved metabolic parameters and endothelial function to a greater extent than dietary interventions alone may be clinically relevant. It remains unknown whether improved endothelial function is sustained in the long term in these patients who are otherwise still considered obese and whether further weight reduction will translate into vascular benefit. Our present study has several limitations. The study was relatively small and not randomized. However, it would not be ethical to randomize patients to medical therapy when surgical treatment is clinically indicated. Only a subset of patients received NTG; thus, we cannot definitively exclude a change in endothelium-independent dilation, although no such effect was seen in previous similar studies.2,3 Last, our data suggest that surgical therapy was predominantly responsible for the improvement in endothelial function. We stress that dietary and medical interventions are critical components of effective weight loss strategies and that the strong trend for improved vascular function with medical therapy may have been significant in a larger study population. In summary, this prospective study demonstrated that weight loss significantly improves endothelial function in obese patients. Reversal of endothelial dysfunction may reduce the association between obesity and cardiovascular risk and serve as an important therapeutic target for weight loss strategies. 1. Sciacqua A, Candigliota M, Ceravolo R, Scozzafava A, Sinopoli F, Corsonello A, Sesti G, Perticone F. Weight loss in combination with physical activity improves endothelial dysfunction in human obesity. Diabetes Care 2003;26: 1673–1678. 2. Hamdy O, Ledbury S, Mullooly C, Jarema C, Porter S, Ovalle K, Moussa A, Caselli A, Caballero AE, Economides PA, Veves A, Horton ES. Lifestyle modification improves endothelial function in obese subjects with the insulin resistance syndrome. Diabetes Care 2003;26:2119 –2125. 3. Raitakari M, Ilvonen T, Ahotupa M, Lehtimaki T, Harmoinen A, Suominen P, Elo J, Hartiala J, Raitakari OT. Weight reduction with very-low-caloric diet and endothelial function in overweight adults: role of plasma glucose. Arterioscler Thromb Vasc Biol 2004;24:124 –128. 4. Vita JA. Nitric oxide-dependent vasodilation in human subjects. Methods Enzymol 2002;359:186 –200. 5. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults, NHLBI, 1998. Available at: http://www.nhlbi.nih. gov/guidelines/obesity/ob_gdlns/pdf. Accessed May 5, 2004. 6. Kiernan M, Winkleby MA. Identifying patients for weight-loss treatment: an empirical evaluation of the NHLBI obesity education initiative expert panel treatment recommendations. Arch Intern Med 2000;160:2169 –2176. 7. Brolin RE. Bariatric surgery and long-term control of morbid obesity. JAMA 2002;288:2793–2796. 8. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412– 419. 9. Lee JH, Chan JL, Yiannakouris N, Kontogianni M, Estrada E, Seip R, Orlova C, Mantzoros CS. Circulating resistin levels are not associated with obesity or insulin resistance in humans and are not regulated by fasting or leptin administration: cross-sectional and interventional studies in normal, insulin-resistant, and diabetic subjects. J Clin Endocrinol Metab 2003;88:4848 – 4856. 10. Vita JA, Keaney JF Jr. Exercise—toning up the endothelium? N Engl J Med 2000;342:503–505. 11. Schauer PR, Burguera B, Ikramuddin S, Cottam D, Gourash W, Hamad G, Eid GM, Mattar S, Ramanathan R, Barinas-Mitchel E, et al. Effect of laparoscopic Roux-en Y gastric bypass on type 2 diabetes mellitus. Ann Surg 2003; 238:467– 484. 12. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, Patel HR, Ahima RS, Lazar MA. The hormone resistin links obesity to diabetes. Nature 2001;409:307–312. JANUARY 15, 2005 Prevalence of Adequate Control of Increased Serum LowDensity Lipoprotein Cholesterol in Self-Pay or Medicare Patients Versus Medicaid or Private Insurance Patients Followed in a University General Medicine Clinic Raja Varma, MD, Wilbert S. Aronow, MD, Glenn Gandelman, Christopher Zammit, BS In a study of 514 patients with increased serum lowdensity lipoprotein (LDL) cholesterol followed in a general medicine clinic at a university hospital, the serum LDL cholesterol in patients with coronary heart disease (CHD), other atherosclerotic vascular disease, or diabetes mellitus was <100 mg/dl in 219 of 276 patients (79%) with Medicaid or private insurance and in 28 of 67 self-pay or Medicare patients (42%) without pharmaceutical coverage (p <0.001). The serum LDL cholesterol was <130 mg/dl in patients with 2Ш‰ risk factor and a 10-year risk for CHD of <20% or <160 mg/dl in patients with a 0 to 1 risk factor and a 10-year risk for CHD of <10% in 54 of 141 patients (38%) with Medicaid or private insurance and in 5 of 30 self-pay or Medicare patients (17%) (p <0.025). б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:269 –270) umerous studies have demonstrated that treatment of hypercholesterolemia in high-risk persons with N statins reduces cardiovascular morbidity and mortality. 1 The National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (NCEP III) guidelines recommend that the serum low-density lipoprotein (LDL) cholesterol be reduced to ПЅ100 mg/dl in persons with coronary heart disease (CHD), other clinical forms of atherosclerotic vascular disease, diabetes mellitus, and with 2П© risk factors that confer a 10-year risk for CHD of Пѕ20%.2 These guidelines also recommend that the serum LDL cholesterol be reduced to ПЅ130 mg/dl in persons with 2П© risk factors and a 10-year risk for CHD of Х…20%, and to ПЅ160 mg/dl in persons with a 0 to 1 risk factor and a 10-year risk for CHD of ПЅ10%.2 This study reports on the prevalence of adequate serum LDL cholesterol control in 514 unselected patients with an increased serum LDL cholesterol followed in the general medicine clinic at a university hospital between July 2003 and June 2004. ••• Charts of 514 unselected patients with increased serum LDL cholesterol followed in the general medicine clinic at Westchester Medical Center/New York From the Department of Medicine, Cardiology Division, Westchester Medical Center/New York Medical College, Valhalla, New York. Dr. Aronow’s address is: Cardiology Division, New York Medical College, Macy Pavilion, Room 138, Valhalla, New York 10595. E-mail: [email protected]. Manuscript received July 27, 2004; revised manuscript received and accepted September 7, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, MPH, and Medical College between July 2003 and June 2004 were reviewed according to a protocol designed by WSA. The 514 patients included 281 women and 233 men (mean age 57 П® 13 years, range 18 to 90). Of the 514 patients, 172 (34%) were white, 152 (30%) were Hispanic, 133 (26%) were African-American, and 57 (11%) were of other races. Of the 514 patients, 343 (67%) had CHD, other clinical forms of atherosclerotic vascular disease, or diabetes mellitus (target LDL cholesterol goal of ПЅ100 mg/dl), 87 (17%) had 2П© coronary risk factors with a 10-year risk for CHD of Х…20% (target LDL cholesterol goal of ПЅ130 mg/ dl), and 84 (16%) had a 0 to 1 coronary risk factor with a 10-year risk for CHD of ПЅ10% (target LDL cholesterol goal of ПЅ160 mg/dl). NCEP III guidelines were used to assess the adequacy of reaching target serum LDL cholesterol treatment goals using the last serum LDL cholesterol level recorded in the chart. Student’s t tests were used to analyze continuous variables and chi-square tests to analyze dichotomous variables. Table 1 shows the prevalence of adequate serum LDL cholesterol control in patients who had Medicaid or private insurance versus patients who did not have pharmaceutical coverage because they were self-pay or Medicare patients. Table 1 also lists the levels of statistical significance. Table 2 shows the prevalence of adequate control of serum LDL cholesterol in patients aged Х†65 versus ПЅ65 years, in men versus women, and in whites, Hispanics, African-Americans, and patients of other races. No significant differences were observed between these groups. Of the 417 patients who had Medicaid or private insurance, 335 (80%) were receiving statins versus 46 of 97 patients (47%) who were self-pay or Medicare patients (p ПЅ0.001). Table 3 shows the prevalence of adequate control of serum LDL cholesterol in patients treated with different statins. No significant differences were observed. Two hundred ninety-five of 361 patients (82%) treated with atorvastatin, simvastatin, or rosuvastatin and 11 of 20 patients (55%) treated with pravastatin, lovastatin, or fluvastatin achieved adequate control of serum LDL cholesterol (p ПЅ0.005). One of 82 Medicaid or private insurance patients (1%) not treated with a statin was treated with ezetimide and none of 51 self-pay or Medicare patients (0%) not treated with a statin was treated with other lipid-lowering medication. The 1 patient treated with ezetimide did not have adequate serum LDL cholesterol control. 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.095 269 population involving 591 centers, 76 of 2,856 patients (3%) with proved or probable CHD had a serum LDL choSelf-pay or lesterol ПЅ100 mg/dl.7 In the Lipid Medicare Treatment Assessment Project, 37% of 28/67 (42%)* 2,285 high-risk patients and 18% of 1,460 patients with CHD achieved adequate serum LDL cholesterol con†trol.8 In the Heart and Estrogen/Pro5/30 (17%) gestin Replacement Study, 262 of 2,752 postmenopausal women (10%) with CHD had a serum LDL cholesterol of ПЅ100 mg/dl.9 In the present study, the prevalence of a serum LDL cholesterol ПЅ100 mg/dl in patients with CHD, other clinical forms of atherosclerotic vascular disease, or diabetes mellitus was 79% in patients with Medicaid or private insurance versus 42% in self-pay or Medicare patients (p ПЅ0.001). In patients without CHD, other clinical forms of atherosclerotic vascular disease or diabetes mellitus, adequate serum LDL cholesterol control according to NCEP III guidelines was achieved in 38% of Medicaid or private insurance patients versus 17% of self-pay or Medicare patients (p ПЅ0.025). Adequate serum LDL cholesterol control was not significantly different in elderly versus younger patients, in men versus women, and in whites, Hispanics, African-Americans, and patients of other races. Data from the present study show a significantly lower prevalence of adequate serum LDL cholesterol control in patients who have no pharmaceutical coverage for medications prescribed by their physician. This problem needs to be addressed if we are to decrease the great amount of cardiovascular morbidity and mortality due to inadequate control of hypercholesterolemia. TABLE 1 Prevalence of Adequate Serum Low-density Lipoprotein (LDL) Cholesterol Control in 514 Patients With Hypercholesterolemia Medicaid or Private Insurance Variable LDL cholesterol ПЅ100 mg/dl in patients with CHD, other atherosclerotic vascular disease, or diabetes mellitus LDL cholesterol ПЅ130 mg/dl in patients with 2П© risk factors and 10-year risk for CHD Х…20% or ПЅ160 mg/dl in patients with 0–1 risk factor and 10-year risk for CHD ПЅ10% 219/276 (79%) 54/141 (38%) *p ПЅ0.001;†p ПЅ0.025. TABLE 2 Prevalence of Adequate Serum Low-density Lipoprotein (LDL) Cholesterol Control in Persons Х†65 Versus ПЅ65 Years, in Men Versus Women, and in Whites, Hispanics, African-Americans, and Persons of Other Races Variable No. Age Х†65 yrs Age ПЅ65 yrs Women Men Whites Hispanics African-Americans Other races 190 324 281 233 172 152 133 57 Adequacy of Serum LDL Cholesterol Control 115 191 170 136 105 88 80 33 (61%) (59%) (60%) (58%) (61%) (58%) (60%) (58%) TABLE 3 Prevalence of Adequate Serum Low-density Lipoprotein (LDL) Cholesterol Control in Persons Taking Different Statins Statin Atorvastatin Simvastatin Pravastatin Lovastatin Fluvastatin Rosuvastatin No. of Patients Adequacy of Serum LDL Cholesterol Control 201 159 11 7 2 1 171 123 6 4 1 1 (85%) (77%) (55%) (57%) (50%) (100%) ••• Harnick et al3 reported data from 270 patients with CHD seen at the Mount Sinai Medical Center (New York, New York), either in the cardiology clinic (n П 131) or in the cardiology private practice suite (n П 139). Of the 270 patients, 26 (22%) had a serum LDL cholesterol Х…100 mg/dl.3 In a primary care outpatient clinic in Southern Florida, 16 of 118 patients (14%) needing lipid-lowering therapy ever had a cholesterol-lowering drug prescribed during a 5-year study period.4 In 27 managed care plans, 1,072 of 3,018 patients (36%) with CHD had a serum LDL cholesterol of Х…100 mg/dl.5 In the Duke University Medical Center Outpatient Clinic, 348 of 518 patients with CHD (68%) had measurements of their serum LDL cholesterol.6 Of the 268 patients found to be at risk, 71 (26%) had a serum LDL cholesterol of Х…100 mg/dl.6 In a German outpatient 270 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 1. Aronow WS. Should the National Cholesterol Education Program guidelines be changed for persons at high risk for cardiovascular events? Preventive Cardiol 2004;7:71–72. 2. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285:2486 –2497. 3. Harnick DJ, Cohen JL, Schechter CB, Fuster V, Smith DA. Effects of practice setting on quality of lipid-lowering management in patients with coronary artery disease. Am J Cardiol 1998;81:1416 –1420. 4. Lai LL, Poblet M, Bello C. Are patients with hyperlipidemia being treated? Investigation of cholesterol treatment practices in an HMO primate care setting. Southern Med J 2000;93:283–286. 5. Latts LM. Assessing the results: phase 1 hyperlipidemia outcomes in 27 health plans. Am J Med 2001;110(suppl 6A):17S–23S. 6. Higgins PDR, Russo C, Scheurer M, Duvall WL. How well do we treat elevated LDL cholesterol? Results from a university residents’ clinic. NC Med J 2002;63:247–252. 7. Ruof J, Klein G, Marz W, Wollschlager H, Neiss A, Wehling M. Lipidlowering medication for secondary prevention of coronary heart disease in a German outpatient population: the gap between treatment guidelines and real life treatment patterns. Preventive Med 2002;35:48 –53. 8. Pearson TA, Laurora I, Chu H, Kafonek S. The Lipid Treatment Assessment Project (L-TAP). A multicenter survey to evaluate the percentages of dyslipidemic patients receiving lipid-lowering therapy and achieving low-density lipoprotein cholesterol goals. Arch Intern Med 2000;160:459 – 467. 9. Schrott HG, Bittner V, Vittinghoff E, Herrington DM, Hulley S, for the HERS Research Group. Adherence to National Cholesterol Education Program Treatment goals in postmenopausal women with heart disease. The Heart and Estrogen/Progestin Replacement Study. JAMA 1997;277:1281–1286. JANUARY 15, 2005 Clinical Variables Predicting Inappropriate Use of Implantable Cardioverter-Defibrillator in Patients With Coronary Heart Disease or Nonischemic Dilated Cardiomyopathy Dominic A.M.J. Theuns, MSc, A. Peter J. Klootwijk, MD, PhD, Maarten L. Simoons, MD, PhD, and Luc J. Jordaens, MD, PhD Inappropriate therapy is a common clinical problem in recipients of implantable cardioverter-defibrillators (ICDs). The present study evaluated whether clinical characteristics could predict inappropriate ICD therapy due to atrial tachyarrhythmias in a series of 260 patients. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:271–274) he implantable cardioverter-defibrillator (ICD) has become the standard therapy for life-threatening T ventricular tachyarrhythmias. Despite the accuracy 1–3 and effectiveness in the diagnosis and treatment of ventricular tachyarrhythmias, a substantial proportion of patients with ICDs experience inappropriate interventions. The reported incidence ranges from 8% to 40%.4 –7 Several studies have investigated clinical risk predictors for ventricular arrhythmia recurrence.8 –10 In contrast, clinical risk predictors for inappropriate ICD use have not been investigated. This study examined the variables that may predict which patients are more likely to receive inappropriate therapy. ••• The study population consisted of 326 consecutive patients who underwent first transvenous implantation at the Erasmus Medical Center (Rotterdam, The Netherlands). Of these, 57 patients were excluded because of participation in a prospective, randomized study of single- and dual-chamber detection algorithms. Another 9 were excluded because of the presence of hypertrophic cardiomyopathy. Thus, 260 patients were eligible for analysis. The patients were assigned to ICD therapy because of a history of cardiac arrest, spontaneous sustained ventricular tachycardia (VT), or nonsustained VT with subsequent inducible sustained VT. Data in the ICD registry are updated prospectively after each clinic visit. The prospectively collected clinical and functional variables for each patient include age, gender, the presence of coronary artery disease (including myocardial infarction and cardiomyopathy), the left ventricular ejection fraction (LVEF) as determined by nuclear isotopes, index arrhythmia, history of atrial From the Department of Cardiology, Thoraxcenter, Erasmus Medical Center, Rotterdam, The Netherlands. Dr. Theuns’s address is: Erasmus Medical Center, Department of Electrophysiology, Bd402, Dr. Molewaterplein 40, 3015 GD Rotterdam, The Netherlands. E-mail: [email protected]. Manuscript received June 15, 2004; revised manuscript received and accepted September 20, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 tachyarrhythmias documented in the clinical file, and pharmacologic treatment. The implanted devices were manufactured by Biotronik (Phylax AV, Tachos DR, and Belos VR-T; Biotronik GmbH & Company, Berlin, Germany), ELA Medical (Defender IV and Alto DR; ELA Medical, Paris, France), Guidant (Mini IV, Contak CD, Renewal I, and Renewal II; Guidant Corporation, St. Paul, Minnesota), and Medtronic (7227, 7250, 7271, and 7272; Medtronic, Inc., Minneapolis, Minnesota). The tachycardia detection rate was programmed according to the clinical presentation of each patient. For all patients, the detection enhancements were activated immediately after ICD implantation. In singleand dual-chamber devices, the stability criterion was programmed at 40 to 50 ms, and the onset criterion was programmed at 15% to 20%. In all dual-chamber devices, the respective dual-chamber detection algorithms were activated. Follow-up began at ICD implantation. At every follow-up visit (at 3-month intervals) or every visit prompted by ICD therapy, all stored data of tachyarrhythmia episodes were collected. Two independent researchers reviewed the stored electrocardiograms. In case of disagreement between the 2 reviewers about the stored electrocardiograms, a third reviewer was consulted and made a decision. For each episode, the date, type, morphology (monomorphic or polymorphic), and mean cycle length (CL) of the tachyarrhythmia and the type and outcome of delivered ICD therapy were recorded. A ventricular tachyarrhythmia was defined as an event with a sudden increase in rate combined with a change in electrocardiographic morphology from the baseline rhythm. If an atrial electrocardiogram was present, the presence of atrioventricular dissociation was used to classify a ventricular tachyarrhythmia. Therapy delivered for atrial arrhythmias (including atrial fibrillation, atrial flutter, atrial tachycardia, and sinus tachycardia) was defined as inappropriate. Continuous variables were evaluated using Student’s t test or analysis of variance. The chi-square test was used for the analysis of categorical variables. The actuarial event-free rates from ventricular and atrial tachyarrhythmias triggering ICD therapy were calculated according to the Kaplan-Meier method. Differences between pairs of actuarial curves were tested by the log-rank test. Relative risks expressed as hazard ratios with 95% confidence intervals (CIs) were based on a Cox proportional-hazards model. 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.017 271 TABLE 1 Patients’ Clinical Characteristics (n П 260) Characteristic Men Age (yrs) LVEF (%) Underlying cardiac disease Coronary artery disease Dilated cardiomyopathy History of atrial tachyarrhythmias Index arrhythmia Ventricular fibrillation VT Nonsustained VT Pharmacologic treatment at discharge Amiodarone вђ¤ blockers Digoxin ACE inhibitor Diuretic Lipid-lowering drug Value 216 (83%) 60 П® 13 31 П® 14 184 (71%) 61 (24%) 79 (29%) 78 (30%) 125 (48%) 57 (22%) 97 120 57 189 158 129 (37%) (46%) (23%) (73%) (61%) (50%) ACE П angiotensin-converting enzyme. Covariates previously identified to be independently associated with the occurrence of inappropriate ICD therapy were used in the multivariate model. A 2-tailed p value ПЅ0.05 was considered significant. The clinical characteristics are listed in Table 1. Coronary artery disease was present in 70% of the patients, nonischemic cardiomyopathy (excluding hypertrophic cardiomyopathy) in 30%, with dilation in 21%. Twenty-four percent received single-chamber devices, 53% dual-chamber devices, and 23% dualchamber devices with cardiac resynchronization capability. The programmed mean detection CL of the VT zone was 377 П® 52 ms. During a mean follow-up of 22 П® 16 months (range 1 to 60), 107 patients (41%) experienced Х†1 episode of sustained ventricular tachyarrhythmia, triggering ICD therapy. The actuarial event-free rates for ventricular tachyarrhythmias were 66.0%, 55.8%, and 45.0% at 1, 2, and 4 years, respectively. The mean CL of monomorphic VT was 333 П® 62 ms; for polymorphic VT or ventricular fibrillation, it was 223 П® 26 ms. A total of 37 patients (14%) experienced inappropriate ICD therapy due to atrial tachyarrhythmias. The actuarial event-free rates for inappropriate therapy were 87.0%, 83.6, and 80.8%, at 1, 2, and 4 years, respectively (Figure 1). Nineteen patients experienced inappropriate therapy for atrial fibrillation at least once, and 18 patients received inappropriate therapy for sinus or atrial tachycardia. There was no significant difference between actuarial event-free rates for inappropriate therapy triggered by atrial fibrillation (94.4% and 88.8%) or atrial or sinus tachycardia (92.7% and 91.3%) at 1 and 4 years, respectively. The mean ventricular CL during atrial fibrillation was 319 П® 44 ms (range 260 to 400); for atrial or sinus tachycardia, the mean ventricular CL was 374 П® 48 ms (range 300 to 480). Clinical variables for patients with and without inappropriate device therapy are presented in Table 2. Age, gender, LVEF, underlying cardiac disease, and 272 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 FIGURE 1. Actuarial event rates for inappropriate device therapy. pharmacologic treatment did not differ between those with and without inappropriate device therapy. The incidence of inappropriate device therapy was greater after a history of atrial tachyarrhythmias (p П 0.003). Additionally, inappropriate ICD therapy was noted more frequently in patients who received appropriate device therapy (p П 0.001). To evaluate clinical predictors of inappropriate device therapy, variables were entered in a Cox proportional-hazards model (age, pharmacologic treatment, type of ICD, LVEF, coronary artery disease, cardiomyopathy, a history of atrial tachyarrhythmias, and recurrent VT). This analysis revealed a history of atrial tachyarrhythmias and recurrent VT with a CL Х†350 ms triggering device therapy as independent clinical predictors of inappropriate ICD therapy. The relative risk was 2.4 (95% CI 1.2 to 4.8, p П 0.01) for a history of atrial tachyarrhythmias. This was supported by lower actuarial event-free rates for inappropriate device therapy for patients with a history of atrial tachyarrhythmias compared with patients without such a history (62.5% vs 88.2% at 4 years, p П 0.002; Figure 2). The relative risk increased to 3.1 if patients had recurrent VT with a CL Х†350 ms that triggered device therapy (95% CI 1.5 to 6.3, p П 0.002). To address the question of whether long interval programming explained this greater risk, we analyzed a detection interval of Х†350 ms as a tachycardia criterion. In multivariate analysis, this detection interval was not identified as an independent predictor for inappropriate therapy, with a relative risk of 2.2 (95% CI 0.6 to 7.5, p П 0.21). Although not significant, the programmed detection CL of the VT zone tended to be shorter in patients with inappropriate therapy compared with those without inappropriate therapy (386 П® 46 vs 371 П® 44 ms, p П 0.06). Proportionally, patients with dual-chamber devices experienced more inappropriate therapy compared with those with single-chamber devices (6% vs 17%, p ПЅ0.05). ••• The present study evaluated whether clinical characteristics could predict inappropriate ICD therapy JANUARY 15, 2005 tion for atrial tachyarrhythmias. Given the epidemiology of atrial fibrillation, No Inappropriate ICD atrial tachyarrhythmias are common in Therapy (n П 223) p Value ICD recipients, of whom most have structural heart disease.13 More re186 (83%) NS 60 П® 13 NS cently, it has been reported that a his31 П® 14 NS tory of paroxysmal atrial fibrillation predicted a higher recurrence rate of 159 (71%) NS atrial tachyarrhythmias.14 52 (23%) NS 57 (26%) 0.003 More notable is the association of recurrent VT with a CL Х†350 ms with 66 (30%) NS an increased risk for inappropriate 106 (47%) NS therapy. The association between ven51 (23%) NS tricular tachyarrhythmias and paroxys85 (38%) NS mal atrial tachyarrhythmias in ICD re101 (45%) NS cipients has been established in 46 (21%) NS previous studies.15–17 Slow ventricular 159 (74%) NS tachyarrhythmias were associated with 135 (63%) NS 112 (52%) NS left ventricular dysfunction (LVEF ПЅ40%) and class III antiarrhythmic drug therapy.18 Another aspect to be addressed is the programmed detection interval. In case of slow ventricular tachyarrhythmias, an overlap with the ventricular rate of atrial tachyarrhythmias is present. The programmed detection interval was not identified in multivariate analysis as an independent predictor. Whether device selection should depend on the knowledge of a history of atrial tachyarrhythmias is an open question, because inappropriate therapy occurs equally in patients with single- and dual-chamber devices. The addition of an atrial lead might improve the specificity of arrhythmia discrimination but introduces potential surgical and technical problems. This should be balanced against the potential minor advantages, such as the presence of slow ventricular tachyarrhythmias. TABLE 2 Comparison of Patients With and Without Inappropriate Device Therapy Characteristic Men Age (yrs) LVEF (%) Underlying cardiac disease Coronary artery disease Dilated cardiomyopathy History of atrial tachyarrhythmias Index arrhythmia Ventricular fibrillation VT Nonsustained VT Pharmacologic treatment Amiodarone вђ¤ blockade Digoxin ACE inhibitor Diuretic Lipid-lowering drug Abbreviation as in Table 1. Inappropriate ICD Therapy (n П 37) 30 (81%) 61 П® 12 31 П® 14 25 (68%) 9 (24%) 19 (51%) 12 (32%) 19 (52%) 6 (16%) 12 (32%) 19 (51%) 11 (30%) 30 (82%) 23 (62%) 17 (46%) FIGURE 2. Actuarial event rates for inappropriate device therapy for patients with and without a history of atrial tachyarrhythmias. due to atrial tachyarrhythmias. The observations noted in this study are: (1) a history of atrial tachyarrhythmias is an independent predictor of inappropriate therapy and (2) recurrent VT with a CL Х†350 ms that triggered device therapy is associated with an increased risk for first inappropriate therapy. In this study, the incidence of inappropriate ICD therapy was 14%. Most patients experienced inappropriate ICD interventions in the first year after device implantation, regardless of primary and secondary prevention. These findings agree with studies reporting on inappropriate ICD therapy.11,12 In the Antiarrhythmics Versus Implantable Defibrillators Trial, atrial tachyarrhythmias were responsible for inappropriate therapy in 22% of patients and 16% of all treated episodes.12 It is no surprise that patients with a history of atrial tachyarrhythmias are at risk for inappropriate interven- 1. Moss AJ, Hall WJ, Cannom DS, Daubert JP, Higgins SL, Klein H, Levine JH, Saksena S, Waldo AL, Wilber D, et al. Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia. Multicenter Automatic Defibrillator Implantation Trial Investigators. N Engl J Med 1996;335:1933–1940. 2. The Antiarrhythmics Versus Implantable Defibrillators (AVID) Investigators. A comparison of antiarrhythmic-drug therapy with implantable defibrillators in patients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med 1997;337:1576 –1583. 3. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert JP, Higgins SL, Brown MW, Andrews ML. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002;346:877– 883. 4. Grimm W, Flores BF, Marchlinski FE. Electrocardiographically documented unnecessary, spontaneous shocks in 241 patients with implantable cardioverter defibrillators. Pacing Clin Electrophysiol 1992;15:1667–1673. 5. Nunain SO, Roelke M, Trouton T, Osswald S, Kim YH, Sosa-Suarez G, Brooks DR, McGovern B, Guy M, Torchiana DF, et al. Limitations and late complications of third-generation automatic cardioverter-defibrillators. Circulation 1995;91:2204 –2213. 6. Kuhlkamp V, Dornberger V, Mewis C, Suchalla R, Bosch RF, Seipel L. Clinical experience with the new detection algorithms for atrial fibrillation of a defibrillator with dual chamber sensing and pacing. J Cardiovasc Electrophysiol 1999;10:905–915. 7. Sticherling C, Schaumann A, Klingenheben T, Hohnloser SH. First worldwide clinical experience with a new dual chamber implantable cardioverter defibrillator: advantages and complications. Europace 1999;1:96 –102. 8. Costeas XF, Link MS, Foote CB, Homoud MK, Wang PJ, Estes NA. Predictors of ventricular tachycardia recurrence in 100 patients receiving tiered therapy defibrillators. Clin Cardiol 2000;23:852– 856. 9. Raitt MH, Klein RC, Wyse DG, Wilkoff BL, Beckman K, Epstein AE, Coromilas J, Friedman PL, Martins J, Ledingham RB, et al. Comparison of BRIEF REPORTS 273 arrhythmia recurrence in patients presenting with ventricular fibrillation versus ventricular tachycardia in the Antiarrhythmics Versus Implantable Defibrillators (AVID) trial. Am J Cardiol 2003;91:812– 816. 10. Raitt MH, Dolack GL, Kudenchuk PJ, Poole JE, Bardy GH. Ventricular arrhythmias detected after transvenous defibrillator implantation in patients with a clinical history of only ventricular fibrillation. Implications for use of implantable defibrillator. Circulation 1995;91:1996 –2001. 11. Gradaus R, Block M, Brachmann J, Breithardt G, Huber HG, Jung W, Kranig W, Mletzko RU, Schoels W, Seidl K, et al. Mortality, morbidity, and complications in 3344 patients with implantable cardioverter defibrillators: results from the German ICD Registry EURID. Pacing Clin Electrophysiol 2003;26:1511–1518. 12. Klein RC, Raitt MH, Wilkoff BL, Beckman KJ, Coromilas J, Wyse DG, Friedman PL, Martins JB, Epstein AE, Hallstrom AP, et al. Analysis of implantable cardioverter defibrillator therapy in the Antiarrhythmics Versus Implantable Defibrillators (AVID) Trial. J Cardiovasc Electrophysiol 2003;14:940 –948. 13. Schmitt C, Montero M, Melichercik J. Significance of supraventricular tachyarrhythmias in patients with implanted pacing cardioverter defibrillators. Pacing Clin Electrophysiol 1994;17:295–302. 14. Wolpert C, Jung W, Spehl S, Schimpf R, Omran H, Schumacher B, Esmailzadeh B, Tenzer D, Mehra R, Luderitz B. Incidence and rate characteristics of atrial tachyarrhythmias in patients with a dual chamber defibrillator. Pacing Clin Electrophysiol 2003;26:1691–1698. 15. Tavernier RH, Kimman GP, Theuns DAMJ, Klootwijk AP, Kazmierczak J, Jordaens LJLM. Comparison of events and survival in dual- versus single chamber ICDs. Neth Heart J 1999;6:604 – 608. 16. Gronefeld GC, Mauss O, Li YG, Klingenheben T, Hohnloser SH. Association between atrial fibrillation and appropriate implantable cardioverter defibrillator therapy: results from a prospective study. J Cardiovasc Electrophysiol 2000;11: 1208 –1214. 17. Stein KM, Euler DE, Mehra R, Seidl K, Slotwiner DJ, Mittal S, Markowitz SM, Lerman BB. Do atrial tachyarrhythmias beget ventricular tachyarrhythmias in defibrillator recipients? J Am Coll Cardiol 2002;40:335–340. 18. Bansch D, Castrucci M, Bocker D, Breithardt G, Block M. Ventricular tachycardias above the initially programmed tachycardia detection interval in patients with implantable cardioverter-defibrillators. Incidence, prediction and significance. J Am Coll Cardiol 2000;36:557–565. Design of the SHock Inhibition Evaluation with Azimilide (SHIELD) Study: A Novel Method to Assess Antiarrhythmic Drug Effect in Patients With an Implantable Cardioverter-Defibrillator Craig M. Pratt, MD, Paul Dorian, MD, Hussein R. Al-Khalidi, PhD, Jose M. Brum, MD, Martin Borggrefe, MD, Daljit S. Tatla, PhD, Johannes Brachmann, MD, Robert J. Myerburg, MD, David S. Cannom, MD, Michael J. Holroyde, PhD, Michael van der Laan, MD, and Stefan H. Hohnloser, MD, on behalf of the SHIELD Investigators This report presents the rationale and study design details of the SHock Inhibition Evaluation with Azimilide study, which is recruiting 624 patients with implantable cardioverter-defibrillators (ICDs) who are at risk for lifethreatening ventricular arrhythmia, randomized to azimilide 75 mg, azimilide 125 mg, or placebo and followed for 1 year. The objective of this study is to determine the effect of azimilide versus placebo on the symptomatic ventricular arrhythmia burden using a unique statistical analysis based on the unusual temporal distribution of symptomatic ICD therapies. The primary efficacy end points are time to all-cause shocks and time to all-cause shocks plus symptomatic ventricular arrhythmic events triggering antitachycardia pacFrom The Methodist DeBakey Heart Center and the Section of Cardiology, Department of Medicine, Baylor College of Medicine, Houston, Texas; the Division of Cardiology, St. Michael’s Hospital, Toronto, Ontario, Canada; Health Care Research Center, Procter & Gamble Pharmaceuticals, Cincinnati, Ohio; Klinikum Mannheim, Universtatsklinikum, Mannheim, Germany; Landkrankenhaus Coburg, II Medizinische Klinik, Coburg, Germany; University of Miami School of Medicine, Miami, Florida; Los Angeles Cardiology Associates, Los Angeles, California; Procter & Gamble Pharmaceuticals, Egham, United Kingdom; and J. W. Goethe University, Medizinische Klinik IV, Kardiologie, Frankfurt, Germany. This study was sponsored by Procter & Gamble Pharmaceuticals, Cincinnati, Ohio. Dr. Pratt’s address is: The Methodist DeBakey Heart Center, 6565 Fannin Street, F1001, Houston, Texas 77030. E-mail: [email protected]. Manuscript received July 14, 2004; revised manuscript received and accepted August 31, 2004. 274 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 ing measured from randomization. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:274 –276) he SHock Inhibition Evaluation with Azimilide (SHIELD) study is designed to expand our knowlT edge of the efficacy of azimilide in reducing symptomatic ventricular arrhythmias and implantable cardioverter-defibrillator (ICD) therapies, total ventricular arrhythmias, and “electrical storms” in patients with ICDs. ••• The SHIELD study is a randomized, double-blind, placebo-controlled, parallel-group study of 624 patients assigned to placebo or oral azimilide 75 or 125 mg and followed for 365 days. The scientific organization of the SHIELD study includes the Executive Steering Committee, the Event Committee, and the Data and Safety Monitoring Board. A list of investigators is presented in the Appendix. The major inclusion criterion is to include patients with ICDs capable of arrhythmia discrimination algorithms (e.g., high rate, sudden onset, or rate stability). Patients who have received their first ICDs will be randomized Х…72 days after documented episodes of sustained ventricular tachycardia (VT), cardiac arrest, or ventricular fibrillation that occurred within a 6-week time period (42 days) before the implantation of the ICDs. Patients with ICDs fitted for previous cardiac arrest or ventricular fibrillation are required to have left ventricular ejection fractions Х…40% to qual0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.096 ify for the study. Patients who have an ICD implanted for Пѕ90 days and had an ICD shock triggered by spontaneous VT or ventricular fibrillation must be randomized Х…180 days after this shock. Major cardiac exclusion criteria include New York Heart Association class IV congestive heart failure, heart transplantation, unstable angina pectoris, recent myocardial infarction (within 30 days), a history of Torsades de Pointes VT, QTc values Пѕ440 ms (with a QRS interval Х…120 ms), or JTc values Пѕ320 ms (if QRS Пѕ120 ms). Noncardiac exclusions are standard in nature. Patients with QTc values Пѕ525 ms (with QRS intervals Х…120 ms) or JTc values Пѕ400 ms (with QRS intervals Пѕ120 ms) will be withdrawn from the study. Patients with absolute neutrophil counts Х…1,000 cells/вђ®l at any time will be withdrawn. In an attempt to minimize heterogeneity regarding ICD programming and to minimize the effect of different devices, investigators are required to program each ICD in a standardized fashion: antitachycardia pacing (ATP) must be programmed “on” for all patients at the time of randomization and set for a minimum of 2 attempts of ATP in the lowest detection zone; the “floor” and “ceiling” for this zone are prescribed according to the slowest documented or induced VT. The first shock strength is to be set no less than the smallest successful defibrillation energy at implantation or the VT before randomization. At least 1 discriminator to reduce the likelihood of inappropriate therapy must be enabled. All ICD therapy events will be adjudicated by the independent Event Committee (Appendix) blinded to treatment assignment. ICD interrogation is mandatory at all scheduled visits and unscheduled visits (caused by symptomatic arrhythmias and/or ICD therapies). At each of these visits, patients are questioned regarding the occurrence of arrhythmia symptoms before the interrogation of the ICD. The date of occurrence and the nature of these symptoms are documented at each visit. The primary efficacy end points are (1) time to all-cause shocks and (2) time to all-cause shocks plus symptomatic ventricular arrhythmic events terminated by ATP measured from randomization. The secondary efficacy end point is time to all ventricular arrhythmias terminated by ICD shocks or ATP. An important tertiary end point is time to all-cause electrical storms, where an “electrical storm” is defined as Х†3 separate episodes of ventricular arrhythmia terminated by ICD shocks or ATP within a time period of 24 hours. Assuming a placebo incidence of 4 events per patientyear for all-cause shocks,1 a total of 624 patients followed for 365 days is required to detect a 30% reduction (hazard ratio 0.7) in the frequency of allcause shocks with 90% power at a significance level of 0.25%.2 The Andersen-Gill mean intensity model is the primary statistical technique to evaluate the observed differences in the primary end point between randomized treatment groups.3 All events for each patient are thus counted, not only the first event. This model compares the distribution of all the interevent intervals in the placebo and drug treatment groups and adjusts for the correlated events over time. Patients are FIGURE 1. The cumulative relative frequency of interevent intervals is plotted against the values of interevent intervals in days in a patient with 21 appropriate shocks during 380 days of follow-up. The abscissa represents the time interval between successive shocks, not a representation of follow-up time. For this patient, 70% (cumulative relative frequency 0.7) of all interevent intervals were <1 day (24 hours). The dashed (“fitted”) line represents the expected cumulative frequency if interevent intervals were randomly distributed, that is, followed an exponential distribution. The observed and fitted distributions are significantly different (p <0.01). stratified by ejection fraction (Х…40% vs Пѕ40%), вђ¤-blocker use, and ICD type (new vs existing). The unique features of the SHIELD study design include the exploration of a dose range of azimilide on symptomatic ventricular arrhythmias and ICDdelivered therapies, the total ventricular arrhythmia burden, and electrical storms in a patient population with uniform programming of ICD settings to reduce device variability. This latter feature is of crucial importance, because end points related to device therapy are determined by the device settings as well as the detected arrhythmia and can be confounded by variability in programming arrhythmia detection and therapies. An additional innovative aspect of the SHIELD study relates to the inclusion of all symptomatic ventricular arrhythmic events, not just the first event in the primary end point, to capture the total symptomatic arrhythmia burden (ICD shocks and symptomatic ventricular arrhythmias terminated by ATP) and, similarly, all ventricular arrhythmic events, as the secondary end point. In contrast, previous ICD trials assessing sotalol and metoprolol focused only on time to first ICD shock.4,5 This is not an optimal end point to assess the effect of an antiarrhythmic therapy, because it does not consider either all symptomatic events (shocks and symptomatic ventricular arrhythmias terminated by ATP) or all ventricular arrhythmias. In addition, the “time to first event” approach is a valid end point only when the events are distributed according to a Poisson process, that is, randomly, with the same distribution as the interevent intervals.6 The Poisson process provides a mathematic model for completely random events for which the cumulative BRIEF REPORTS 275 Andersen-Gill mean intensity model is a semiparametric regression model that applies to recurrent events and generates robust standard error estimators, using all accumulated interevent intervals. APPENDIX FIGURE 2. The cumulative relative frequency of interevent intervals is plotted against the values of the interevent intervals in days. There are 86 appropriate shocks (events) in 18 patients. For this population, half (cumulative relative frequency 0.5) of all interevent intervals were <1 day (24 hours). See also legend in Figure 1. interevent intervals (the time intervals between each consecutive pair of ICD therapies) follow an exponential distribution. As illustrated in Figure 1, the observed cumulative relative frequency of appropriate ICD shocks (data were adapted from the placebo patients in the azimilide pilot study1 to remove any treatment effect on the distribution of interevent intervals) are clustered and nonrandomly distributed and do not follow an exponential distribution. Under a Poisson process (assuming that shocks occur at random), this patient would be expected to have, on average, an appropriate shock every 20 days. However, Figure 1 shows that about 70% of this patient’s shocks occurred Х…24 hours after the initial shock. Similarly, in our pilot study, the cumulative placebo population interevent intervals (Figure 2) are also clustered, illustrating a greater risk for arrhythmia recurrence shortly after each event. For events with this type of distribution, the time to the first event is an inadequate depiction of the overall antiarrhythmic drug effect on ICD therapies. The total symptomatic arrhythmia burden is the best representation of the disease severity and the efficacy of therapy. However, the simple calculation of total number of events per patient or total events per patient per year, does not adjust for the correlation between events within patients, because most of the events occur within a short time after the previous event. The 276 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 Executive Steering Committee: Paul Dorian, Craig Pratt, Martin Borggrefe, Johannes Brachmann, Stefan Hohnloser, David Cannom, and Robert Myerburg. Data and Safety Monitoring Board: John Cairns, Leone Greene, William E. Wilkinson, and Carina BlomstrГ¶m Lundqvist. Event Committee: Igor Singer, Jesus Almendral, Tony Simmons, Patrick Tchou, Vilma Torres, Kenneth Stein, Imran Niazi, Francois Philippon, Walid Saliba, and David Martin. Investigators: Etienne Aliot, Barry Alpert, Dietrich Andresen, Fernando Arribas, John Bailey, Malcolm Bersohn, Ulrika Birgersdotter-Green, Tristram Bahnson, Madgy Basta, John T. Beard, David Benditt, Joseph Bissett, Louis Blier, Martin Borggrefe, Johannes Brachmann, Christopher Brown, Joseph Brugada, Jeffrey Buetikofer, David Cannom, Mark Carlson, Anthony Chang, Don Chilson, Ted Chow, Lai Chow-Kok, Jacques Clementy, James Coman, Stuart Connolly, Sean Connors, James Cook, Steven Kalbfleisch, Richard Corbelli, Pierre Cosnay, George Crossley, Ken Curry, Joel Cutler, James Daubert, Luc De Roy, Rainer Dietz, John DiMarco, Paul Dorian, Kenneth Ellenbogen, Martin Emert, Roger Freedman, Eli Gang, J. Anthony Gomes, Martin Gottwik, Ricardo Ruiz Granell, Martin Green, Arnold Greenspon, Thomas Guarnieri, Hartmut Gulker, Moshe Gunsburg, Charles Haffajee, Mark Hamer, David Henderson, Steven Higgins, Ellen Hoffmann, Stefan Hohnloser, Andrew Hordes, Alberto Interian, Mohammed Jazayeri, Jose Joglar, Roy John, Luc Jordaens, Salem Kacet, Steven Kalbfleisch, Mohammed Yousuf Kanjwal, Stephen Keim, Shane Kimber, Helmut Klein, Zdzislawa Kornacewicz-Jach, Volker KГјhlkamp, Herve Le Marec, Antoine Leenhardt, Robert Leman, Bruce Lerman, Samuel Levy, Albert Lin, Ignacio Lozano, Andrzej Lubinski, Richard Luceri, Berndt Luderitz, Jean-Claude Mabo, Paul Maccaro, Michael Markel, Frank McGrew, Craig McPherson, A. Meijer, Thomas Meinertz, William M. Miles, Ralph Mletzko, Foad Moazez, Michael Mollerus, George Monir, David Nabert, Imran Niazi, Brendan O’Cochlain, Gearaoid O’Neill, Eraldo Occhetta, Michael Oeff, Brian Olshansky, Sudha Pai, John Payne, Christian Perings, Heinz Pitschner, Edward Platia, Scott Pollak, AR Ramdat Misier, Eric Rashba, Pratap Reddy, Paolo Rizzon, Francisko Alzueta Rodriguez, Marc Roelke, Lawrence Rosenthal, Andrea Russo, Sanjeev Saksena, Robert Sangrigoli, Massimo Santini, Martin Jan Schalij, Alexander Schirdewan, Claus Schmitt, Wolfgang Schols, Peter Schwartz, Richard Sheahan, Hue-Teh Shih, Karlheinz Seidl, Tony Simmons, Igor Singer, Nicholas Stamato, Jonathan Steinberg, Laurence Sterns, Hanna Szwed, Mario Talajic, Patrick Tchou, Paul Touboul, Maria Trusz-Gluza, Isabelle Van Gelder, Yves Vandekerckhove, Ernst Vester, Juergen Vogt, Abdul Wase, Peter Wells, Eric Wever, Teresa Widomska-Czekajska, Arthur Wilde, Stephen Winters, Deborah Wolbrette, Kevin Wolfe, Seth Worley, and Manfred Zehender. 1. Singer I, Al-Khalidi H, Niazi I, Tchou P, Simmons T, Henthorn R, Holroyde M, Brum J. Azimilide decreases recurrent ventricular tachyarrhythmias in patients with implantable cardioverter defibrillators. J Am Coll Cardiol 2004; 43:39 – 43. 2. Signorini D. Sample size for Poisson regression. Biometrika 1991;78: 446 – 450. 3. Lin DY, Wei LJ, Yang I, Ying Z. Semiparametric regression for the mean and rate functions of recurrent events. R Stat Soc Ser B 2000;62:711–730. 4. Pacifico A, Hohnloser SH, Williams JH, Tao B, Saksena S, Henry PD, Prystowsky EN. Prevention of implantable-defibrillator shocks by treatment with sotalol. d,l-Sotalol Implantable Cardioverter-Defibrillator Study Group. N Engl J Med 1999;340:1855–1862. 5. Seidl K, Hauer B, Schwick NG, Zahn R, Senges J. Comparison of metoprolol and sotalol in preventing ventricular tachyarrhythmias after the implantation of a cardioverter/defibrillator. Am J Cardiol 1998;82:744 –748. 6. Rose MS, Gillis AM, Sheldon RS. Evaluation of the bias in using time to the first event when the inter-event intervals have a Weibull distribution. Stat Med 1999;18:139 –154. JANUARY 15, 2005 Refeeding Normalizes the QT Rate Dependence of Female Anorexic Patients FrГ©dГ©ric Roche, MD, PhD, Jean-Claude BarthГ©lГ©my, MD, PhD, Norbert Mayaud, MD, Vincent Pichot, PhD, David Duverney, PhD, Natacha Germain, MD, FranГ§ois Lang, MD, and Bruno Estour, MD We postulated that dynamic alterations in QT interval adaptation could characterize patients with anorexia nervosa (AN) and could be restored after weight gain. To assess ventricular repolarization features, we evaluated the QT dynamicity along RR intervals from 24-hour electrocardiographic data of patients with AN before and after refeeding. Ten young women with AN (19 ШЋ 3 years) were included in the study. The QT/RR slope was found significantly enhanced compared with normals (ШЉ1.82 ШЋ 0.62 vs ШЉ1.40 ШЋ 0.30; p <0.05). This slope returned to normal range values after refeeding. The QT/RR slope was significantly correlated with the body mass index (r ‫ ؍‬0.59; p <0.007) in the patient group. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:277–280) he clinical manifestations of anorexia nervosa (AN) have been well described for several deT cades. Its mortality rate is the highest of any major psychiatric disorder, with many deaths occurring suddenly.1 Although reports have suggested an association between sudden death in anorexic patients and QT prolongation on routine electrocardiography (ECG), the incidence and the clinical consequences of such abnormalities remain controversial.2– 4 That a primary cardiac abnormality may be present at least in some cases is suggested by findings of repolarization abnormalities and QT-interval prolongation under a variety of circumstances associated with severe weight loss.5 QT-rate dependence is 1 of the major properties of ventricular repolarization, with its circadian and autonomic modulations.6 Susceptibility to severe ventricular arrhythmias has been associated with the alteration of the rate-dependence adaptation of repolarization after myocardial infarction as well as in acquired or congenital long-QT syndrome. We thus confirmed that dynamic alterations in QT-interval adaptation7 and its link with abnormalities in autonomic control could characterize patients with AN. To assess ventricular repolarization features, we compared QT parameters and their dynamicity along RR intervals from the 24-hour electrocardiographic From Service de Physiologie Clinique et de l’Exercice – Groupe PPEH, CHU Nord, Service d’Endocrinologie, CHU Bellevue, and Service de Psychiatrie, CHU Bellevue, FacultГ© de MГ©decine Jacques Lisfranc, UniversitГ© Jean Monnet, Saint-Etienne, France. Dr. Roche’s address is: EFCR, CHU Nord, Niveau 6, F-42055 Saint-Etienne Cedex 2, France. E-mail: [email protected]. Manuscript received May 18, 2004; revised manuscript received and accepted September 7, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 data of patients with and without this serious medical eating disorder. Furthermore, the improvement of these parameters was tested after medically controlled refeeding. ••• Ten young women (mean age 19 П® 3 years) with diagnosed AN fulfilling the Diagnostic and Statistical Manual of Mental Disorders (fourth edition) criteria for AN were included in the study. All patients were hospitalized because of clinical signs of malnutrition, with a mean body mass index (BMI) of 14.7 П® 2.3 kg/m2. Their body weights averaged ПЄ34.2 П® 6.9% of their ideal body weights. The study procedure was fully explained to the subjects, who gave their informed consent to participate. All patients were reevaluated after 5 П® 2 months of medically controlled refeeding, which allowed a significant increase in BMI to 17.6 П® 4.2 kg/m2 (p ПЅ0.003 vs baseline) and a reduction of body weight deficit reaching ПЄ21.6 П® 9.2% (p ПЅ0.003 vs baseline). Reference values for 24-hour electrocardiographic recordings and QT analysis were obtained from a control group of 10 healthy female medical students aged 20 П® 3 years (mean BMI 23.1 П® 1.9 kg/m2). Standard 3-channel Holter tape recorders were used to acquire the data (Del Mar Reynolds Medical, Inc., Irvine, California). The Holter electrocardiographic system allowed the extraction of the list of RR intervals with a precision of 1/256 second. Heart rate variability (HRV) was first evaluated using time-domain analysis. To perform the analysis, only intervals between normal beats were considered. Several classic parameters were calculated8: the rootmean-square of successive differences (the square root of the mean of the sum of the square differences between adjacent normal RR intervals), the SD of all normal RR intervals in the entire recording, and the proportion of consecutive beats with RR intervals showing a Пѕ50-ms difference. Spectral analysis of HRV was thus performed by the fast-Fourier transform method using sliding 256-point Hanning windows.8 The interpolation mode was linear. Frequencies in the band from 0.04 to 0.14 Hz were considered low frequencies (LFs), and frequencies from 0.15 to 0.40 Hz were considered high frequencies (HFs). The LF/HF ratio was calculated. The LF and HF powers of HRV were further computed in normalized units. A RR interval was considered to be a pause when it reached or exceeded 2.5 seconds. Bradycardia was defined as a heart rate (HR) ПЅ40 beats/min. Isolated and grouped ventricular ectopic activities were analyzed. Nonsustained ventricular tachycardia was de0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.018 277 the time of hospital admission for the AN group are listed in Table 1. Periods of bradycardia were noted in 6 of 10 patients and pauses were observed in AN Before AN After Controls 10% of patients with AN. Frequent Parameter (n П 10) (n П 10) (n П 10) supraventricular premature beats Minimal HR (beats/min) 46 П® 9 61 П® 16†55 П® 4 (Х†10/hour) were observed in 10% and Maximal HR (beats/min) 130 П® 14 141 П® 10 145 П® 20 frequent ventricular premature beats Day, mean HR (beats/min) 73 П® 12* 92 П® 16 82 П® 10 were also noted in 1 patient. Neither Night, mean HR (beats/min) 56 П® 11* 73 П® 19 67 П® 11 VPCs Х†10/h (%) 1 (10) 0 1 (10) high-grade nor frequent (Х†10/hour) Nonsustained VT (%) 0 0 0 ventricular ectopy was evident in the Sinoatrial block (%) 1 (10) 0 0 patients, but 1 presented with idioven†Bradycardia (%) 6 (60)* 0 0 tricular accelerated rhythm (1 salve, 4 *p ПЅ0.05, AN before versus controls; †p ПЅ0.05, AN before versus AN after refeeding. beats). After refeeding, none of the paVPCs П ventricular premature complexes; VT П ventricular tachycardia. tients with AN presented with bradycardia, pauses, or significant ventricular arrhythmias. Patients with AN fined as Х†6 consecutive ventricular ectopic complexes presented with lower HRs than controls during the night at a rate Пѕ120 beats/min and lasting ПЅ30 seconds. and the day. The maximal and the minimal HRs were The severity of ventricular arrhythmias was evaluated affected in anorexic patients, but the differences did not by the mean number of ventricular premature com- reach statistical significance. All parameters returned plexes per hour (threshold values Х†10 per hour). close to the control group values after refeeding. The Supraventricular tachycardia was defined as Пѕ3 con- relative bradycardia observed in patients with AN before secutive supraventricular ectopic complexes at a rate refeeding was associated with an enhanced autonomic tone, as the SD of all normal RR intervals in the entire Пѕ120 beats/min. The Holter tapes were analyzed for QT intervals by recording was found to be significantly greater and the a computer-assisted method (QT Analysis System; HFs as well as the HFs in normalized units tended to be Del Mar Reynolds Medical, Inc.). Data for QT anal- greater but without significant difference in this small ysis was derived from a good-quality Holter recording population. The autonomic tone indexes tended to go with a 256-Hz sample rate and containing no more back to a normal range value after refeeding. Mean QT than 20 minutes of total artifact. T-wave amplitude duration and corrected QT values did not differ statistimust be Х†0.2 V to ensure consistent discrimination cally during the 24-hour period in subjects with AN between QT peak and QT end. The operator (FR) first (Table 2). No abnormal 24-hour mean QTc duration was performed a careful Holter analysis before attempting observed in the control group, and only 1 anorexic paQT analysis to be sure to detect and correctly label all tient showed a mean QTc reaching 440 ms after the cardiac beats. The Holter channel corresponding to refeeding procedure. The QT/RR slope was found to be lead V5 on the surface ECG was used. The operator the most significant discriminating parameter of the cardefined a template QT interval by the beginning of the diac repolarization analysis observed between patients QRS and the end of the T wave on 1 beat. The end of with AN at baseline and controls. QT length related to the T wave was defined as the point at which the HR was found to be longer for lower HRs in patients down-slope of the T wave crossed the isoelectric line. with AN compared with normal controls. This QT/RR The waveform analyzer was programmed to sample relation was significantly correlated in the AN group QRS complexes at a specific time, and the RR and QT (before and after refeeding) with the BMI (r П 0.59, intervals were measured automatically. Mean П® SD p ПЅ0.007), body weight deficit (r П 0.58, p ПЅ0.008), QT was calculated from all the beats measured. Cor- minimal HR (r П 0.75, p ПЅ0.001), mean HR (r П 0.76, rected QT values at 60 beats/min estimated from Ba- p ПЅ0.001), SD of all normal RR intervals in the entire zett’s quadratic correction formula were calculated recording (r П ПЄ0.60, p ПЅ0.03), LF/HF ratio (r П 0.38, from all QT values for each hour of analysis. QT-rate p ПЅ0.05), and HF in normalized units (r П ПЄ0.49, dependence graph plots for each HR range against QT p ПЅ0.03). were individually constructed and showed the least••• squares fit regression lines and formulas for these AN leads to abnormal beat-to-beat changes in venpoints. tricular repolarization, characterized by an enhanced reComparisons of clinical status and QT parameters lation between QT duration and RR intervals. This relabetween patients with AN and those without were tion is significantly correlated with the bradycardia based on unpaired Student’s t tests for continuous data associated with the eating disorder. Furthermore, these and on chi-square tests for categorical data. The paired abnormalities seem to be reversible after medically conStudent’s t test was used to compare parameters at trolled refeeding. The role of autonomic cardiac regulafollow-up. Statistical significance was defined at the p tion in ventricular repolarization adaptation seems to be ПЅ0.05 level. Pearson’s product moment procedure very important9: pharmacologic sympathetic activation was used to determine the correlation between vari- or blockade has been shown to greatly alter QT length ables and the QT/RR shape. and QT-rate adaptation. This suggests a direct influence Twenty-hour-hour electrocardiographic findings at of cholinergic activity on the repolarization of ventricular TABLE 1 Evolution of the 24-hour Electrocardiographic Holter Parameters of the Population With AN Before and After Refeeding and Comparison Controls 278 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 focused on QT dispersion in patients with AN.14 –16 This method, however, has several operator-dependent limitations, and its prognostic impliAN Before AN After Controls cations remain uncertain in high-risk Measurement (n П 10) (n П 10) (n П 10) populations, such as patients who SDNN (ms) 207 П® 64* 148 П® 76†173 П® 20‡ have had myocardial infarctions. UnpNN50 (%) 27 П® 15 16 П® 16 23 П® 4 usual autonomic imbalance characRMSSD (ms) 57 П® 16 41 П® 27 52 П® 8 terizes the severe starvation status. 2,982 П® 1,467 2,195 П® 1,776 3,868 П® 1,016 Total PSD (ms2/Hz) VLF PSD (ms2/Hz) 1,378 П® 579 877 П® 625 1,978 П® 684 Kollai et al,17 using baroreflex sensiLF PSD (ms2/Hz) 602 П® 253 487 П® 327 1,066 П® 419 tivity, and others,18 –20 using valiHF PSD (ms2/Hz) 873 П® 705 646 П® 839 576 П® 171 dated HRV analysis, have reported 2 LFnu PSD (ms /Hz) 45 П® 18 55 П® 18 55 П® 11 enhanced parasympathetic control of HFnu PSD (ms2/Hz) 49 П® 17 45 П® 18 31 П® 9 HR probably associated with sympaLF/HF ratio 1.4 П® 1.4 1.6 П® 1.1 2.0 П® 1.0 QT/RR slope ПЄ1.8 П® 0.7* ПЄ1.1 П® 0.8†Ϫ1.4 П® 0.3 thetic withdrawal in patients with QT mean (ms) 374 П® 29 345 П® 55 374 П® 28 AN. It should be noted that in our QT SD (ms) 39 П® 10* 31 П® 7†31 П® 5 population, HF values as well as norQTc Bazett (ms) 388 П® 20* 410 П® 28 412 П® 14 malized HF power spectral density *p ПЅ0.05, AN before versus controls; †p ПЅ0.05, AN before versus AN after refeeding; ‡p ПЅ0.05, tended to be greater in patients with AN after versus controls. AN before treatment but without staHFnu П HF computed in normalized units; LFnu П LF computed in normalized units; pNN50 П tistically significant differences beproportion of consecutive beats with RR intervals showing a Пѕ50-ms difference; PSD П power spectral density; RMSSD П root mean square of successive differences; SDNN П SD of all normal RR intervals cause of the small sample of patients. in the entire recording; VLF П very LF. None of the HRV parameters, except SDNN, seems to have significantly altered after refeeding. myocardium independently from rate adaptation of QT In conclusion, the QT/RR relation is found to be intervals after the elimination of cardiac parasympathetic enhanced in patients with AN, reflecting in part the activity. Because the prolongation of QT interval has specific autonomic imbalance encountered in this popubeen associated with sudden cardiac death, the relation lation, and this abnormal pattern seems to be reversible between semistarvation periods, the development of after refeeding. The clinical implications of such findings non–rate-related QT-interval prolongation, and subse- need to be discussed, because an equivalent enhancequent clinical outcomes have been examined by several ment of the QT/RR slope has been described in patients investigators first during therapeutic starvation for with life-threatening ventricular arrhythmias. morbid obesity and then in patients with AN.10 QT prolongation had been described by Rasmussen and Andersen11 in 12 of 22 women in association with Acknowledgment: We are indebted to Christian weight reduction after gastroplasty. Thurston and Marks2 Roche, MD, Maryse Victoire, BSc, and Delphine described minor degrees of QT-interval prolongation in 5 Maudoux, BSc, for their expert technical assistance. of 9 women with AN but did not correct for HR, and Isner et al10 retrospectively reported the prolongation of QTc intervals in 3 adult anorexic women who died 1. American Psychiatric Association. Practice guidelines for eating disorders. suddenly. The absence of histologic abnormalities and of Am J Psychiatry 1993;150:212–228. 2. Thurston J, Marks P. Electrocardiographic abnormalities in patients with electrolyte imbalance within the heart suggests that some anorexia nervosa. Br Heart J 1974;36:719 –723. extracardiac influence may explain the potential QT pro- 3. Isner JM, Roberts WC, Heymsfield SB, Yager J. Anorexia nervosa and sudden longation observed in such patients. It is not inconceiv- death. Ann Intern Med 1985;102:49 –52. 4. Cooke RA, Chambers JB, Singh R, Todd GJ, Smeeton NC, Treasure J, able that autonomic inflow to the heart might be de- Treasure T. QT interval in anorexia nervosa. Brit Heart J 1994;72:69 –73. ranged, particularly under conditions of stress, and this 5. Pringle TH, Scobie IN, Murray RG, Kesson CM, Maccuish AC. Prolongation should limit the rate adaptation of the QT interval in such of the QT interval during therapeutic starvation: a substrate for malignant arInt J Obesity 1983;7:253–261. conditions. In contrast, it should be noted that other rhythmias. 6. Viitasalo M, Karjalainen J. QT intervals at heart rates from 50 to 120 beats per 12,13 investigators have been unable to document QT- minute during 24-hour electrocardiographic recordings in 100 healthy men. interval prolongation in their patients with AN. In our Circulation 1992;86:1439 –1442. 7. Roche F, Estour B, Kadem M, Millot L, Pichot V, Duverney D, Gaspoz JM, population, we could not confirm a significantly large BarthГ©lГ©my JC. Alteration of the QT rate dependence in anorexia nervosa. PACE prevalence of abnormal QT prolongation. Although the 2004;27:1099 –1104. mean QTc interval, which should be considered as the 8. Task Force of the European Society of Cardiology and the North American of Pacing and Electrophysiology. Heart rate variability: standards of most discriminating parameter in the evaluation of de- Society measurements, physiological interpretation, and clinical use. Circulation 1996; polarization in several heart disorders, was not signifi- 93:1043–1065. cantly altered by AN in our population, only 1 patient 9. Ahnve S, Vallin H. Influence of heart rate and inhibition of autonomic tone on QT interval. Circulation 1982;65:435– 439. with AN had an abnormal QTc interval, which was the 10. Isner JM, Roberts WC, Heymsfield SB, Yager J. Anorexia nervosa and evident after refeeding only. sudden death. Ann Intern Med 1985;102:49 –52. This study is the first to report a 24-hour mean QT 11. Rasmussen LH, Andersen T. The relationship between QTc changes and nutrition during weight loss after gastroplasty. Acta Med Scand 1985;21:271–275. duration and variation in a population of patients with 12. Powers PS, Schocken DD, Feld J, Holloway JD, Boyd F. Cardiac function during AN before and after refeeding. Recent studies have weight restoration in anorexia nervosa Int J Eating Disord 1991;10:521–530. TABLE 2 QT and HRV Measurements Obtained Using 24-hour Electrocardiographic Holter Monitoring in Patients With AN (Before and After Refeeding) and Control Group BRIEF REPORTS 279 13. Gottdiener JS, Gross HA, Henry WL, Borer JS, Ebert MH. Effects of self-induced starvation on cardiac size and function in anorexia nervosa. Circulation 1978;58:421– 433. 14. Galetta F, Franzoni F, Cupisti A, Belliti D, Prattichizzo F, Rolla M. QT interval dispersion in young women with anorexia nervosa. J Pediatr 2002;140:456 – 460. 15. Swenne I, Larsson PT. Heart risk associated with weight loss in anorexia nervosa and eating disorders: risk factors for QTc interval prolongation and dispersion. Acta Paediatr 1999;88:304 –309. 16. Swenne I. Heart risk associated with weight loss in anorexia nervosa and eating disorders: risk factors for QTc interval prolongation and dispersion. Acta Paediatr 2000;89:447– 452. 17. Kollai M, Bonyhay I, Jokkel G, Szonyi L. Cardiac vagal hyperactivity in adolescent anorexia nervosa. Eur Heart J 1994;15:1113–1118. 18. Petretta M, Bonaduce D, Scalfi L, De Filippo E, Marciano F, Migaux ML, Themistoclakis S, Ianniciello A, Contaldo F. Heart rate variability as a measure of autonomic nervous system function in anorexia nervosa. Clin Cardiol 1997; 20:219 –224. 19. Rechlin T, Weis M, Ott C, Bleichner F, Joraschky P. Alterations of autonomic cardiac control in anorexia nervosa. Biol Psychiatry 1998;43:358 –363. 20. Kreipe RE, Goldstein B, Deking DE, Tipton R, Kempski MH. Heart rate power spectrum analysis of autonomic dysfunction in adolescents with anorexia nervosa. Int J Eating Disord 1994;16:159 –165. Left Ventricular Structure and Function in Sedentary and Physically Active Subjects With Left Ventricular Hypertrophy (the LIFE Study) Kurt Boman, MD, PhD, Mona Olofsson, BS, BjГ¶rn DahlГ¶f, MD, PhD, Eva Gerdts, Markku S. Nieminen, MD, PhD, Vasilios Papademetriou, MD, PhD, Kristian Wachtell, MD, PhD, and Richard B. Devereux, MD, PhD Patients with left ventricular hypertrophy had higher heart rates, left ventricular mass, and left atrial size independent of gender, blood pressure, or body mass index than physically active patients. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:280 –283) the echocardiographic substudy of the Losartan for End Point Reduction in HypertenIsionnIntervention (LIFE) study, we examined left ventricular (LV) structure and function in sedentary and physically active hypertensive patients with LV hypertrophy. Our hypothesis was that physically active patients would have more favorable LV structure and function than sedentary patients. ••• The present analysis was conducted in 958 patients aged 55 to 80 years with stage II or III hypertension enrolled in the LIFE echocardiographic substudy who underwent yearly echocardiograms during 3 years of antihypertensive treatment1– 4 and had data on exercise status at baseline. Mean seated systolic and diastolic blood pressures after a 1- to 2-week placebo run-in period were 160/95 to 200/115 mm Hg. Exercise status was obtained by questionnaire and was categorized into 3 groups: group 1 (sedentary): never exerFrom the Department of Medicine, SkellefteГҐ County Hospital and UmeГҐ University, SkellefteГҐ; the Department of Medicine, Sahlgrenska University Hospital/Г–stra, and University of Gothenburg, Gothenburg, Sweden; the Department of Cardiology, Haukeland University Hospital, Bergen, Norway; the Department of Cardiology, Helsinki University Central Hospital, Helsinki, Finland; the Department of Cardiology, Veterans Administration Hospital, Washington, DC; the Department of Medicine, Glostrup University Hospital, Glostrup, Denmark; and Weill Medical College of Cornell University, New York, New York. This study was supported in part by a grant from Merck & Co., Inc., West Point, Pennsylvania. Dr. Boman’s address is: SkellefteГҐ County Hospital and UmeГҐ University, 93186 SkellefteГҐ, Sweden. E-mail: [email protected]. Manuscript received March 30, 2004; revised manuscript received and accepted September 1, 2004. 280 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, PhD, cise; group 2: exercise Х…30 minutes twice per week; and group 3 (physically active): exercise Пѕ30 minutes twice per week. The LIFE echocardiographic substudy was carried out in echocardiographic centers in Denmark, Finland, Iceland, Norway, Sweden, the United Kingdom, and the United States. Geographic distribution, blood pressure, body mass index, and prevalences of diabetes and vascular disease resembled the entire LIFE population, with the exception of enrolling more men and African-Americans. Patients gave informed consent, and ethical committees in participating countries approved this study. Before enrollment, all patients had screening electrocardiograms showing LV hypertrophy by either genderadjusted Cornell voltage-duration product Х†2,440 mV П« ms or Sokolow-Lyon voltage Пѕ38 mV.5 Exclusion criteria included myocardial infarction or stroke Х…6 months before the study, current heart failure or known LV ejection fraction ПЅ40%, significant aortic stenosis, or overt renal insufficiency (serum creatinine Пѕ160 вђ®mol/L or 1.8 mg/dl). Echocardiographic procedures for this study have been previously described.1– 4 End-diastolic LV dimensions were used to calculate LV mass by an anatomically validated formula (r П 0.90 vs necropsy LV mass).6 LV hypertrophy was considered present when LV mass/ body surface area was Пѕ116 g/m2 for men and Пѕ104 g/m2 for women.7 Relative wall thickness, calculated as diastolic posterior wall thickness/LV internal radius,8 was increased when this ratio was Пѕ0.430.9 Normal geometry was present when LV mass index and relative wall thickness were normal, increased relative wall thickness and normal LV mass index identified concentric LV remodeling, increased LV mass index with normal relative wall thickness identified eccentric LV hypertrophy, and increases of the 2 variables identified concentric LV hypertrophy.10 Endocardial fractional shortening (percent) was calculated from LV internal dimensions in diastole and systole.11 To assess LV contractility, we used the 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.019 TABLE 1 Patient Characteristics According to Exercise Status Variable Age (yrs) Women Body mass index (kg/m2) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Pulse pressure (mm Hg) Heart rate (beats/min) Smoker Framingham risk score (%) Diabetes mellitus Prevalence of myocardial infarction and/or angina pectoris Cerebral vascular disease Peripheral vascular disease Group 1 (n П 212) Group 2 (n П 236) Group 3 (n П 510) ANOVA p Value 66.4 П® 7.5 42.5% 28.6 П® 5.2 174 П® 15 99 П® 10 75 П® 16 74 П® 11 26.4% 23.8 П® 9.6 17.9% 15.1% 66.2 П® 7.2 46.8% 27.6 П® 5 174 П® 14 98 П® 9 76 П® 16 74 П® 11 19.1% 22.9 П® 9.6 10.2% 12.8% 65.6 П® 6.6 38.4% 26.6 П® 3.6 173 П® 14 98 П® 9 75 П® 15 71 П® 11 18.0% 22.6 П® 9.2 9.0% 13.1% 0.285 0.087 ПЅ0.001 0.600 0.690 0.777 ПЅ0.001 0.035 0.257 0.002 0.727 12.7% 8.0% 10.6% 4.3% 5.5% 4.9% 0.002 0.158 Group 1 (sedentary), never exercise; group 2, exercise Х…30 minutes twice per week; group 3 (physically active), exercise Пѕ30 minutes twice per week. ANOVA П analysis of variance. relation between mid-wall shortening and mid-wall circumferential end-systolic stress measured at the level of the LV minor axis,12,13 using a previously reported method to identify the location in the LV wall at end-systole of the surface between the inner and outer myocardial shell volumes at end-diastole.14 Circumferential end-systolic stress, as the primary measure of myocardial afterload, was estimated at the mid-wall from LV linear dimensions using a cylindrical model.13,15 Previously published equations relating endocardial and mid-wall shortening to circumferential end-systolic stress in 140 normotensive adults were used to derive predicted endocardial fractional and mid-wall shortening.12 Stress-corrected endocardial and midwall shortening were then calculated as the ratios to the predicted value. SPSS version 11.0.1 (SPSS, Inc., Chicago, Illinois) was used for data management and analysis. Continuous variables are expressed as mean П® SD and categorical data as percentages. Differences in continuous and categorical variables among the 3 exercise groups were assessed by 1-way analysis of variance followed by the ScheffГ© post hoc test and chi-square analyses, respectively. Bivariate correlations between physical activity and LV geometric variables were assessed with Spearman’s correlation coefficients. A full factorial model was used for analysis of covariance; multiple simple contrasts were obtained after adjustment of the confidence intervals with Sidak’s method, whereby main effects were adjusted for potential confounders (body mass index, heart rate, smoking, and the effects of coexisting disease as diabetes or previous cerebral vascular disease). Twotailed p ПЅ0.05 was considered statistically significant. A total of 958 patients who had echocardiograms and provided information on exercise habits at enrollment in the LIFE Study constituted the present study population (Table 1). No significant difference (p П 0.081) in exercise categories was observed between the 397 women and 561 men. Group 1 included 90 women and 122 men, group 2 included 111 women and 125 men, and group 3 included 196 women and 314 men. Body mass index was significantly greater in groups 1 and 2 than in group 3 (28.6 and 27.6 vs 26.6 kg/m2, p ПЅ0.001 and p П 0.008, respectively), without a significant difference between groups 1 and 2 (p П 0.067). Body surface area was larger in group 1 than in group 3 (1.92 vs 1.89 m2, p ПЅ0.001) but did not differ between groups 1 and 2 (p П 0.067). Mean body weight was greater in group 1 than in groups 2 and 3 (82.0 vs 78.4 and 76.9 kg, p П 0.024 and p ПЅ0.001, respectively) but was statistically indistinguishable between groups 2 and 3 (p П 0.385). There were no significant differences among the groups in height or systolic and diastolic blood pressures. LV mass was greater in group 1 than in groups 2 and 3 (p П 0.015 and p П 0.040, respectively), without a significant difference between groups 2 and 3. LV mass/ height2.7 was greater in group 1 than in group 3 (p П 0.004), but LV mass/body surface area did not differ significantly among the groups (Table 2). Heart rate during the echocardiograms was lower in group 3 than in group 1 (p П 0.001). Interventricular septal thickness was lower in group 3 than in group 1 (p П 0.012), with an intermediate mean value in group 2. The LV posterior wall was thicker in group 1 than in groups 2 and 3 (p П 0.045 and p П 0.016, respectively). Left atrial diameter was larger in group 1 than in groups 2 and 3 (p П 0.006 and p П 0.005, respectively). No differences were found in LV internal dimension or relative wall thickness among the 3 groups, as a result of which there was also no significant intergroup difference in the patterns of LV geometry. There were no significant differences among the 3 groups in LV systolic function measured as endocardial fractional shortening, ejection fraction, circumferential systolic stress, or mid-wall shortening. The atrial filling fraction was greater in group 1 than in groups 2 and 3 (p П 0.003 and p П 0.018, respectively), but there were no significant differences in other measures of LV diastolic function among the BRIEF REPORTS 281 TABLE 2 LV Structure and Function and Left Atrial Size According to Exercise Status Group 1 (n П 212) Variable Interventricular septum thickness (cm) Posterior wall thickness (cm) LV internal dimension in diastole (cm) LV mass (g) LV mass/surface area (g/m2) LV mass/height2.7 (g/m2.7) Relative wall thickness Left atrial dimension (cm) Fractional shortening (%) Ejection fraction (%) Midwall shortening (%) Circumferential end-systolic stress (kdynes/cm2) Stress-corrected mid-wall shortening (%) Isovolumic relaxation time (ms) Mitral valve E/A ratio Mitral valve deceleration time (ms) Atrial filling fraction Stroke volume (ml) Cardiac output (ml/min) Cardiac index (ml/min/m2) Total peripheral resistance (dynes В· sec В· cmПЄ5) Total peripheral resistance index (dynes В· sec В· cmПЄ2) 1.18 1.09 5.3 244 126 58.8 0.42 4.1 33.1 60.8 15.2 182 95.7 116 0.86 211 0.44 76.3 5,313 2,782 1,948 3,696 П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® 0.2 0.2 0.6 68 30 15.7 0.08 0.6 5.9 8.7 2.3 53 13.7 24 0.41 63 0.12 17.6 1,378 701 609 1,120 Group 2 (n П 236) 1.16 1.06 5.2 229 122 56 0.41 3.9 33.4 61.2 15.3 182 96.1 113 0.84 210 0.43 76.2 5,254 2,815 1,962 3,659 П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® 0.1 0.1 0.6 51 25 12 0.07 0.6 6.1 8.9 2.2 48 13.1 23 0.41 64 0.11 16.3 1,274 664 552 1,052 Group 3 (n П 510) ANOVA p Value П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® 0.012 0.011 0.249 0.011 0.208 0.004 0.123 0.001 0.823 0.806 0.120 0.730 0.109 0.303 0.685 0.102 0.001 0.052 0.241 0.384 0.388 0.518 1.15 1.06 5.3 232 123 55.3 0.41 3.9 33.4 61.3 15.6 185 97.7 115 0.87 220 0.41 79.1 5,136 2,739 2,010 3,758 0.1 0.1 0.6 53 24 11.4 0.06 0.5 5.6 8.1 2 48 12.5 23 0.33 67 0.1 17.4 1,252 666 558 1,028 Explanation of groups in Table 1. Abbreviation as in Table 1. groups, including the isovolumic relaxation time, E- or A-wave velocities, E/A ratio, and deceleration time. No difference was found in stroke volume, cardiac output, and total peripheral resistance among the groups. Multivariate analysis in the general linear model showed that LV mass was greater in group 1 than in groups 2 and 3 (adjusted mean 241 vs 229 and 232 g, p П 0.032), independent of body mass index, male gender, and pulse pressure (all p ПЅ0.001). In similar models, ventricular septal thickness was significantly (p П 0.007) less in physically active than in sedentary patients, as was LV posterior wall thickness (p П 0.015), whereas the adjusted mean atrial filling fraction was smaller in physically active than in sedentary patients (p П 0.026). With addition of age to these covariates, mean adjusted LV mass was smaller in patients in group 2 and tended to be smaller in group 3 than in group 1 (229 and 232 vs 241 g, p ПЅ0.05 and p П 0.12, respectively). In similar models, interventricular septal thickness was smaller in physically active than sedentary patients (p П 0.015), as was LV posterior wall thickness (1.06 vs 1.09 cm, p П 0.024). The adjusted mean atrial filling fraction remained significantly smaller in physically active than in sedentary patients (0.41 vs 0.44, p П 0.04). ••• To our knowledge, this is the largest study showing that patients with hypertension and sedentary lifestyles have increased LV mass due to larger LV wall thicknesses. It extends the results of smaller studies that reported the regression of pathologic LV hypertrophy with exercise training in selected hypertensive patients.16 The finding that LV mass was greater in sedentary patients independent of the effect of body mass index is especially important in view of the strong association between relative body weight and arterial pressure and the known benefit of weight loss 282 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 for regression of LV hypertrophy. There was a significant association between body mass index and LV mass. Hinderliter and coworkers17 found that exercise and weight loss reduced blood pressure and induced favorable changes in LV structure in a cohort of overweight, sedentary men and women. Blood pressure decreased by 7/6 mm Hg in the weight management group and by 3/4 mm Hg in the aerobic exercise group. In association with these decreases in blood pressure, participants in the intervention groups had significant decreases in relative wall thickness (p П 0.003), posterior wall thickness (p П 0.05), and septal thickness (p П 0.004), with a trend toward a decrease in indexed LV mass (p П 0.08) relative to the control group.17 In our study, the difference between physically active and sedentary hypertensive patients in LV mass was due to greater ventricular septal and posterior wall thicknesses, without a significant betweengroup difference in LV chamber size. In the present study, the greater LV hypertrophy, due to increased LV wall thicknesses, in sedentary patients was not associated with abnormalities of LV systolic function. In contrast, the atrial filling fraction of diastolic transmitral blood flow was significantly greater in sedentary patients. The abnormalities of LV geometry and filling associated with a sedentary lifestyle among patients in the LIFE Study occurred despite the absence of significant differences among groups in arterial blood pressure. Although hypertension is the most important cause of LV hypertrophy, the relation between the degree of LV hypertrophy and blood pressure at rest is generally weak and less close than that with pressure measured by ambulatory monitor. Papademetriou et al18 found that systolic blood pressure at a low level of exercise (5 METs), corresponding to usual daily acJANUARY 15, 2005 tivities, was the closest correlate of LV hypertrophy in hypertensive patients. Endurance exercise training can reduce blood pressure in older adults with mild (grade I) hypertension. It also improves exercise capacity and the quality of life and can induce a modest but significant regression of LV hypertrophy and remodeling in older hypertensive adults. 1. Wachtell K, Rokkedal J, Bella JN, Aalto T, DahlГ¶f B, Smith G, Roman MJ, Ibsen H, Aurigemma GP, Devereux RB. Effect of electrocardiographic left ventricular hypertrophy on left ventricular systolic function in systemic hypertension (the LIFE Study). Am J Cardiol 2001;87:54 – 60. 2. Wachtell K, Bella JN, Liebson PR, Gerdts E, DahlГ¶f B, Aalto T, Roman MJ, Papademetriou V, Ibsen H, Rokkedal J, et al. Impact of different partition values on prevalences of left ventricular hypertrophy and concentric geometry in a large hypertensive population: the LIFE study. Hypertension 2000;35:6 –12. 3. Wachtell K, Smith G, Gerdts E, DahlГ¶f B, Nieminen MS, Papademetriou V, Bella JN, Ibsen H, Rokkedal J, Devereux RB. Left ventricular filling patterns in patients with systemic hypertension and left ventricular hypertrophy (the LIFE study). Am J Cardiol 2000;85:466 – 472. 4. Devereux RB, Bella JN, Boman K, Gerdts E, Nieminen MS, Rokkedal J, Papademetriou V, Wachtell K, Wright J, Paranicas M, et al. Echocardiographic left ventricular geometry in hypertensive patients with electrocardiographic left ventricular hypertrophy: the LIFE Study. Blood Press 2001;10:74 – 82. 5. DahlГ¶f B, Devereux RB, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Julius S, Kjeldsen S, Kristianson K, Lederballe-Pedersen O, et al. The Losartan Intervention For Endpoint Reduction (LIFE) in Hypertension Study. Am J Hypertens 1997;10:705–713. 6. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450 – 458. 7. Devereux RB, DahlГ¶f B, Levy D, Pfeffer MA. Comparison of enalapril versus nifedipine to decrease left ventricular hypertrophy in systemic hypertension (the PRESERVE trial). Am J Cardiol 1996;78:61– 65. 8. Reichek N, Devereux RB. Reliable estimation of peak left ventricular systolic pressure by M-mode echographic-determined end-diastolic relative wall thickness: identification of severe valvular aortic stenosis in adult patients. Am Heart J 1982;103:202–209. 9. Roman MJ, Pickering TG, Schwartz JE, Pini R, Devereux RB. Association of carotid atherosclerosis and left ventricular hypertrophy. J Am Coll Cardiol 1995;25:83–90. 10. Ganau A, Devereux RB, Roman MJ, de Simone G, Pickering TG, Saba PS, Vargiu P, Simongini I, Laragh JH. Patterns of left ventricular hypertrophy and geometric remodeling in essential hypertension. J Am Coll Cardiol 1992;19: 1550 –1558. 11. Gutgesell HP, Paquet M, Duff DF, McNamara DG. Evaluation of left ventricular size and function by echocardiography. Results in normal children. Circulation 1977;56:457– 462. 12. de Simone G, Devereux RB, Roman MJ, Ganau A, Saba PS, Alderman MH, Laragh JH. Assessment of left ventricular function by the midwall fractional shortening/end-systolic stress relation in human hypertension. J Am Coll Cardiol 1994;23:1444 –1451. 13. Gaasch WH, Zile MR, Hoshino PK, Apstein CS, Blaustein AS. Stressshortening relations and myocardial blood flow in compensated and failing canine hearts with pressure-overload hypertrophy. Circulation 1989;79:872– 883. 14. Shimizu G, Hirota Y, Kita Y, Kawamura K, Saito T, Gaasch WH. Left ventricular midwall mechanics in systemic arterial hypertension. Myocardial function is depressed in pressure-overload hypertrophy. Circulation 1991;83: 1676 –1684. 15. Gaasch WH, Battle WE, Oboler AA, Banas JS Jr, Levine HJ. Left ventricular stress and compliance in man. With special reference to normalized ventricular function curves. Circulation 1972;45:746 –762. 16. Hagberg JM, Park JJ, Brown MD. The role of exercise training in the treatment of hypertension: an update. Sports Med 2000;30:193–206. 17. Hinderliter A, Sherwood A, Gullette EC, Babyak M, Waugh R, Georgiades A, Blumenthal JA. Reduction of left ventricular hypertrophy after exercise and weight loss in overweight patients with mild hypertension. Arch Intern Med 2002;162:1333–1339. 18. Papademetriou V, Notargiacomo A, Sethi E, Costello R, Fletcher R, Freis ED. Exercise blood pressure response and left ventricular hypertrophy. Am J Hypertens 1989;2:114 –116. Comparison of Angiotensin-Converting Enzyme Inhibitors in the Treatment of Congestive Heart Failure Karen Tu, MD, MSc, Muhammad Mamdani, PharmD, Douglas Lee, MD The effectiveness of the different angiotensin-converting enzyme (ACE) inhibitors in the treatment of patients with congestive heart failure (CHF) was compared by performing a retrospective cohort study using linked administrative databases on elderly patients admitted to the hospital for the treatment of CHF. Relative to those initiated on enalapril, no significant differences in the combined end point of readmission to the hospital for CHF or mortality were observed among users of lisinopril, ramipril, or other From the Institute for Clinical Evaluative Sciences, Toronto, Ontario; University Health Network, Toronto Western Hospital Family Medicine Centre, Toronto, Ontario; University Health Network, Division of Cardiology, Toronto, Ontario; and the Departments of Family and Community Medicine, Family Healthcare Research Unit, Health Policy Management and Evaluation, Pharmacy, and Medicine, University of Toronto, Toronto, Ontario, Canada. Dr. Lee holds research fellowships from the Heart and Stroke Foundation of Canada and the Canadian Institutes of Health Research, Ottawa, Ontario, Canada. Dr. Tu’s address is: Institute for Clinical Evaluative Sciences, G106-2075 Bayview Avenue, Toronto, Ontario M4N 3M5, Canada. E-mail: [email protected]. Manuscript received July 8, 2004; revised manuscript received and accepted August 26, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MA, MPH, Alex Kopp, BA, and ACE inhibitors. In terms of effectiveness for the treatment of patients with CHF, the findings of this study suggest a class effect among ACE inhibitors. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:283–286) ngiotensin-converting enzyme (ACE) inhibitors have been demonstrated to reduce morbidity and A mortality in the treatment of patients with congestive heart failure (CHF).1 Although the classic and strongest evidence supporting the use of ACE inhibitors in CHF has involved 1 specific agent, enalapril,2,3 the findings are often extrapolated to other ACE inhibitors, and a class effect is often presumed despite differences in pharmacokinetic and pharmacodynamic properties among agents.4,5 The publication of the Heart Outcomes Prevention Evaluation Study,6 a study showing the cardiovascular protective effects of ramipril, resulted in a dramatic increase in the use of ramipril in patients with CHF, although such patients were excluded from the trial.7 Furthermore, there have been no published head-to-head trials comparing the effectiveness of the different ACE inhibitors in pa0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.097 283 TABLE 1 Hospitalizations, Procedures, and Drug Use History Variable Enalapril (n П 2,976) Lisinopril (n П 1,192) Ramipril (n П 1,138) Other ACE Inhibitors (n П 1,447) 79.6 54% 6% 78.4 54% 4% 78.7 51% 5% 79.3 53% 6% 7% 15% 3% 4% 8% 16% 3% 4% 9% 15% 4% 4% 7% 15% 3% 5% 1% ПЅ1% 20% 2% ПЅ1% 2% 1% 23% 3% ПЅ1% 3% 1% 25% 4% ПЅ1% 1% ПЅ1% 22% 2% ПЅ1% 3% 5% 11% 18% 2% 1% 23% 21% 31% 17% 16% ПЅ1% 21% 9% 29% 25% 18% 1% 2% 4% 6% 13% 19% 2% ПЅ1% 22% 23% 30% 18% 15% ПЅ1% 20% 12% 26% 26% 19% ПЅ1% 2% 4% 5% 12% 18% 2% 1% 23% 24% 32% 16% 14% ПЅ1% 18% 12% 27% 25% 15% ПЅ1% 1% 3% 5% 11% 19% 2% ПЅ1% 24% 19% 34% 17% 15% ПЅ1% 22% 9% 31% 25% 19% 1% 2% Age, mean (yrs) Women Long-term care Hospitalizations Acute myocardial infarction Coronary heart disease Renal failure Stroke Procedures Coronary angiography Coronary bypass Echocardiography Coronary angioplasty Cardiac valve surgery Previous drug use вђЈ blockers Antiarrhythmics Anticoagulants Antidiabetics Antiplatelets Antirheumatics Aspirin вђ¤ blockers Calcium channel blockers Digoxin Diuretics Hydralazine Inhalers Lipid-lowering drugs Loop diuretics Nitrates Nonsteroidal anti-inflammatory drugs Other antihypertensives Spironolactone TABLE 2 Readmissions or Mortality, Readmissions Alone, or Mortality Alone Variable Readmissions or mortality Follow-up time, mean (d) Death or readmission Death or readmission per 100 person-yrs Crude HR (95% CI) Adjusted HR (95% CI) Readmissions Follow-up time, mean (d) Readmission Readmission per 100 person-yrs Crude HR (95% CI) Adjusted HR (95% CI) Mortality Follow-up time, mean (d) Death Death per 100 person-yrs Crude HR (95% CI) Adjusted HR (95% CI) Enalapril 259 669 (22%) 31.8 1.00 1.00 259 390 (13%) 18.5 1.00 1.00 280 360 (12%) 15.8 1.00 1.00 Lisinopril Ramipril 286 290 (24%) 31.0 285 266 (23%) 30.0 1.00 (0.88–1.15) 1.08 (0.94–1.23) 286 175 (15%) 18.7 0.97 (0.84–1.12) 1.06 (0.92–1.24) 285 170 (15%) 19.1 Other ACE Inhibitors 247 324 (22%) 33.1 1.04 (0.91–1.19) 1.02 (0.90–1.17) 247 211 (15%) 21.6 1.06 (0.88–1.26) 1.11 (0.92–1.32) 1.08 (0.90–1.29) 1.19 (0.99–1.45) 1.15 (0.97–1.36) 1.13 (0.96–1.34) 313 158 (13%) 15.5 0.98 (0.81–1.18) 1.09 (0.90–1.31) 311 137 (12%) 14.1 0.90 (0.74–1.09) 0.97 (0.78–1.20) 268 158 (11%) 14.9 0.95 (0.78–1.14) 0.94 (0.78–1.13) CI П confidence interval; HR П hazard ratio. tients with CHF. Whether the effects are unique to specific ACE inhibitors or all ACE inhibitors show equivalent benefits has not been firmly established.8,9 284 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 The objective of this study was to examine the relative association between ACE inhibitors and readmission and mortality rates of patients with CHF. JANUARY 15, 2005 ••• We performed a retrospective cohort study using administrative databases, which were linked anonymously through encrypted unique patient identifiers, on the Пѕ1.4 million elderly residents in Ontario. We identified patients discharged alive from the hospital with CHF from April 1, 1997, to March 31, 2000, using the International Classification of Diseases, Ninth Revision, Clinical Modification code 428.x recorded in the Canadian Institute for Health Information hospital discharge database, which contains detailed information on all hospitalizations in Ontario. A recent chart abstraction study validating the accuracy of coding for CHF admission in our databases revealed positive predictive values of 90% to 96%, depending on the criteria used to define CHF.10 We excluded patients ПЅ66 years of age and those with admissions for CHF in the 3 years before the index date. To minimize the potential for survival treatment selection bias,11 we limited our analysis to patients surviving Х†30 days after discharge and those initiated on ACE inhibitors during this interval. To minimize the potential for channeling bias, we limited our analysis to ACE inhibitor users. Furthermore, we excluded patients who received ACE inhibitors or angiotensin II receptor blockers in the year before the hospital admission to limit our analysis to treatmentnaive patients. Enalapril users served as the reference group. The Ontario Drug Benefit database records prescriptions filled by elderly residents aged Х†65 years in Ontario and served as the source of drug use data. Patients were divided into those initiated on enalapril, lisinopril, ramipril, or other ACE inhibitors (benazepril, captopril, cilazapril, fosinopril, perindopril, quinapril, and trandolapril). For each of the study drug cohorts, we defined the duration of exposure as the period of continuous, exclusive enrollment in the study medication group after the index date. The “days supply” variable of the pharmacy claims database allowed us to estimate the intended duration of each prescription. If patients were dispensed drugs before the end of this period, the excess drug supplies were carried over to the next prescription’s day supply estimation. Patients were allowed a 20% grace period on the previous day supply to refill the next prescriptions. If they did not refill their prescriptions for the study drugs within these successive time windows, they were deemed to have discontinued the study drugs. The primary outcome of interest was the combined end point of readmission for CHF as a primary diagnosis or mortality. The secondary outcomes were CHF readmission alone and mortality alone. Patients were followed for Х…2 years and were censored if they were readmitted for CHF, died, switched medications, discontinued the initial ACE inhibitor, or reached the end of the observation period (March 31, 2002). Time-to-event analyses were conducted for CHF hospitalization using Cox proportional-hazards models with the enalapril group as the reference, controlling for all covariates outlined in Table 1, as well as the fiscal quarter of the index date to account for temporal effects. As an overall measure of co-morbidity, we controlled for the number of distinct drugs dispensed in the year before the index date,12 the Charlson co-morbidity index,13 and the presence of diabetes.14 The proportional-hazards assumption for each exposure variable was assessed in each analysis. All analyses were performed using SAS for Unix, version 8.2 (SAS Institute Inc., Cary, North Carolina). All statistical tests were performed at the 5% level of significance and were 2 sided. We identified 6,753 patients with newly diagnosed CHF admitted to the hospital who survived Х†30 days after discharge and were initiated on ACE inhibitors during this time. The average age of the patients receiving the different types of ACE inhibitors was 79.2 years, and in total, 53.3% of the cohort was women and 5.5% were long-term care patients (Table 1). There did not appear to be any consistently meaningful clinical differences among the study drug groups with respect to the covariates, as outlined in Table 1. Relative to enalapril users, no significant differences in the combined end point of readmission for CHF or mortality were observed in users of lisinopril (adjusted hazard ratio 1.08, 95% confidence interval 0.94 to 1.23), ramipril (adjusted hazard ratio 1.06, 95% confidence interval 0.92 to 1.24), or the other ACE inhibitors (adjusted hazard ratio 1.02, 95% confidence interval 0.90 to 1.17). No significant differences among groups were observed in the secondary end points of readmission for CHF or mortality independently (Table 2). ••• We observed no significant difference among the various ACE inhibitors for patients with newly diagnosed CHF initiated on ACE inhibitors in terms of CHF readmission or mortality, readmission alone, or mortality alone. Regardless of the varying pharmacologic properties of ACE inhibitors,5 including differences in chemical structure, lipophilicity, half-life, and tissue-binding capacity, our results suggest a class effect of ACE inhibitors in the treatment of patients with CHF. Several limitations must be acknowledged. Our analysis considered the duration of the prescriptions, but we were unable to determine actual patient compliance with their medications or dosages taken. Our databases did not contain important information on other variables that may have been associated with the outcome of interest, including body mass index, dietary salt intake, alcohol consumption, hypertension, and smoking. It is possible that patients varied in the proportion of left ventricular dysfunction or the cause of CHF (ischemic, nonischemic, diastolic, systolic), but it is unlikely that this measurement was substantially different among the different drug groups. Although bias is often problematic in observational studies, the distribution of the observed clinical covariates among the study groups was very similar, and the unadjusted and adjusted hazard ratios were virtually identical, implying that this was probably not a major concern for this study. In attempts to minimize bias in our study through strict exclusion criteria, the number of BRIEF REPORTS 285 patients in each study group may have been relatively low, which may have compromised statistical power in detecting a clinically meaningful difference. However, given the results of our study, we estimate Х†80% power to detect a clinically meaningful relative risk reduction of 30% in the primary outcome among groups. Our study is the first direct comparison of ACE inhibitors in terms of heart failure effectiveness. Our results suggest no significant differences among patient ACE inhibitors in clinically meaningful outcomes for treating patients with CHF. Thus, when prescribing ACE inhibitors, consideration should likely be given to dosing convenience and cost. 1. Garg R, Yusuf S. Overview of randomized trials of angiotensin-converting enzyme inhibitors on mortality and morbidity in patients with heart failure. JAMA 1995;273:1450 –1456. 2. The CONSENSUS Trial Study Group. Effects of enalapril on mortality in severe congestive heart failure. N Engl J Med 1987;316:1429 –1435. 3. The SOLVD Investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. N Engl J Med 1992;395:293–302. 4. Kostis JB. Pharmacological differentiation of angiotensin-converting enzyme inhibitors. J Human Hyperten 1989;3:119 –125. 5. Brown NJ, Vaughan DE. Angiotensin-converting enzyme inhibitors. Circula- tion 1998;97:1411–1420. 6. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 2000;342:145–153. 7. Tu K, Mamdani MM, Jacka RM, Forde NJ, Rothwell DM, Tu JV. The striking effect of the Heart Outcomes Prevention Evaluation (HOPE) on ramipril prescribing in Ontario. Can Med Assoc J 2003;168:553–557. 8. Furberg CD, Pitt B. Are all angiotensin-converting enzyme inhibitors interchangeable? J Am Coll Cardiol 2001;37:1456 –1460. 9. Sica DA. Class effects of angiotensin-converting enzyme inhibitors. Am J Manag Care 2000;6:S85–S108. 10. Jong P, Gong Y, Liu PP, Austin PC, Lee DS, Tu JV. Care and outcomes of patients newly hospitalized for heart failure in the community treated by cardiologists compared with other specialists. Circulation 2003;108:184 –191. 11. Glesby MJ, Hoover DR. Survivor treatment selection bias in observational studies: examples from the AIDS literature. Ann Intern Med 1996;124:999 –1005. 12. Schneeweiss S, Seeger JD, Maclure M, Wang PS, Avorn J, Glynn RJ. Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data. Am J Epidemiol 2001;154:854 – 864. 13. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40:373–383. 14. Hux JE, Ivis F, Flintoft V, Bica A. Diabetes in Ontario: determination of prevalence and incidence using a validated administrative data algorithm. Diabetes Care 2002;25:512–516. Simplified Peak Power Reserve in Patients With an Implantable Cardioverter-Defibrillator and Advanced Heart Failure William T. Katsiyiannis, MD, Alan D. Waggoner, MHS, Benico Barzilai, MD, Brian F. Gage, MD, MS, Jose M. Sanchez, MD, Joseph G. Rogers, MD, Bruce D. Lindsay, MD, and Marye J. Gleva, MD The prognostic ability of simplified peak power (SPP) reserve, a novel measure of left ventricular systolic performance, was prospectively studied in patients with advanced heart failure (HF) and implantable cardioverter-defibrillators. Reduced SPP reserve identified patients who are at high risk for experiencing progressive HF. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:286 –288) ome patients with left ventricular dysfunction and ventricular arrhythmias derive limited benefit S from the placement of implantable cardioverter-defibrillators (ICDs), because they die of progressive heart failure (HF). The accurate identification of this group could guide the selection of patients who may benefit from more directed HF therapy. Unfortunately, the tools that are currently available to stratify this population are imprecise. Patients with the most advanced HF symptoms are more likely to die from From the Cardiovascular Division, Department of Internal Medicine, Washington University School of Medicine, St. Louis, Missouri. Dr. Katsiyiannis was supported by the Michael Bilitch fellowship of the North American Society of Pacing and Electrophysiology, Natick, Massachusetts. Dr. Katsiyiannis’s address is: Minneapolis Heart Institute, 920 East Twenty-Eighth St., Suite 300, Minneapolis, Minnesota 55407. E-mail: [email protected]. Manuscript received June 2, 2004; revised manuscript received and accepted September 8, 2004. 286 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 progressive pump failure as opposed to arrhythmic death. Contractile reserve determined by invasively measured peak power has been used as a prognostic indicator in patients with HF.1 Ventricular reserve using noninvasive simplified peak power (SPP) is easily obtained, afterload independent, can be preload adjusted, and may add additional predictive power to traditional prognostic measures. We conducted a prospective study to determine whether noninvasive SPP reserve can identify ICD candidates with rapidly progressive HF. Our hypothesis was that in patients with ICDs in New York Heart Association functional class III HF, those with limited SPP reserve would have a greater number of adverse HF events. ••• We prospectively enrolled patients in New York Heart Association class III HF who underwent ICD implantation for an American College of Cardiology– American Heart Association class I indication at Barnes-Jewish Hospital–Washington University School of Medicine. Patients with either ischemic or nonischemic cardiomyopathy were included. Patients were excluded for the inability to complete a dobutamine stress echocardiogram (i.e., the development of chest pain, arrhythmia, or hypotension or hypertension necessitating the early discontinuation of the study). A cohort of agematched control patients without a history of HF or 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.020 SPP was calculated according to the methods described by Armstrong SPP Reserve and colleagues,2 where peak power is equal to the product of peak aortic ПЅ1.5 Пѕ1.5 flow and mean arterial pressure. Peak Characteristic (n П 8) (n П 10) p Value aortic flow was defined as the prodAge (yrs) 66 58 0.41 uct of peak aortic velocity and aortic African-American 0 (0%) 2 (20%) 0.48 annulus area. Aortic velocity was Men 6 (75%) 7 (70%) 0.81 Hypertension 3 (38%) 8 (80%) 0.145 measured by continuous-wave DoppDiabetes mellitus 4 (50%) 6 (60%) 0.67 ler in the apical 5-chamber view. The Coronary artery disease 7 (88%) 7 (70%) 0.59 aortic annulus diameter was meaLeft ventricular ejection fraction (%) 27 П® 13 37 П® 8 0.06 sured from the parasternal long-axis Mitral deceleration time (ms) 166 П® 42 221 П® 81 0.11 Sodium (mmol/L) 140 П® 3 139 П® 3 0.53 view at rest. Mean arterial pressure Creatinine (mg/dl) 2.0 П® 1.4 1.1 П® 0.2 0.08 was obtained by a sphygmomanomAngiotensin-converting enzyme inhibitor 8 (100%) 9 (90%) 1.0 eter at the brachial artery. SPP reor angiotensin receptor blocker serve, as described in the following вђ¤-adrenergic blocker 5 (63%) 5 (50%) 0.66 equations, equals the difference beAntiarrhythmic 3 (38%) 4 (40%) 1.0 Digoxin 4 (50%) 5 (50%) 1.0 tween SPP at maximal dobutamine Diuretic 8 (100%) 9 (90%) 1.0 stress and SPP at rest divided by the 2 0.7 П® 0.4 5.7 П® 2.7 0.0001 SPP reserve (W/ml ) square of end-diastolic volume, an estimate of preload: (1) SPP reserve П SPP maximum dobutamine ПЄ SPP baseline; (2) SPP П aortic flow П« mean arterial pressure; (3) aortic flow П aortic annulus area П« peak aortic velocity; and (4) mean arterial pressure П ([2 П« systolic pressure] П© diastolic pressure)/3. Patients were followed in the Washington University Medical Center Arrhythmia Clinic every 3 months. At each follow-up visit, clinical information was obtained, including a history of shocks, ICD interrogation, HF hospitalizations, or cardiac transplantation. Mortality data were collected from hospital records and family interviews. The composite end point of HF hospitalizations, cardiac transplantation, and all-cause mortality was analyzed by the Kaplan-Meier method. Analyses were performed with SPP reserve dichotomized at 1.5 W/ml2. This value was chosen a priori on the basis of a previous study.1 Continuous variables were compared using the unpaired Student’s t test, and categorFIGURE 1. Kaplan-Meier curves for the composite end point in ical variables were compared using Fisher’s exact test. patients with ICDs and New York Heart Association class III HF. Analyses were performed using SPSS version 10.0 for Windows (SPSS, Inc., Chicago, Illinois) statistical software. Twelve age-matched normal control patients unarrhythmia was also enrolled and underwent noninvasive SPP reserve measurements only. The Human Studies derwent dobutamine stress echocardiography and Committee of Washington University School of Medi- the determination of SPP reserve. Their mean age cine approved all aspects of the study, and written in- was 61 years. Their mean baseline ejection fraction formed consent was obtained from all participants before was 73 П® 11%, and their mean SPP reserve was 35.2 П® 19 W/ml2. Eighteen patients in New York enrollment. Dobutamine stress echocardiograms were performed Heart Association class III HF who had ICDs were beginning at an initial dose of 5 вђ®g В· kgПЄ1 В· minПЄ1 for 3 enrolled. Their mean age was 61 years. Their mean minutes, followed by 10 вђ®g В· kgПЄ1 В· minПЄ1 for 3 min- baseline ejection fraction was 32 П® 11%, and their utes, with subsequent dose increments of 10 вђ®g В· kgПЄ1 В· mean SPP reserve was 3.5 П® 3.2 W/ml2. minПЄ1 every 3 minutes, up to a maximum of 40 вђ®g В· SPP reserve discriminated HF patients from conkgПЄ1 В· minПЄ1. The infusions were discontinued when the trols without overlap (mean 3.5, range 0.05 to 10.34 vs patients achieved 85% of their target heart rate for their mean 35.2, range 21.8 to 51.3, respectively; p age group or if chest pain, ST-segment depression, or ПЅ0.0001). The study population had a mean follow-up new regional wall motion abnormalities developed. of 15.5 months. There were no differences in the Echocardiograms at rest and dobutamine stress echocar- baseline characteristics of the 2 groups of patients diograms were used to determine the left ventricular with HF and ICDs: those with adequate (Пѕ1.5 W/ml2) and poor (ПЅ1.5 W/ml2) SPP reserve (Table 1). There ejection fraction and SPP reserve. TABLE 1 Characteristics of Patients With HF and ICDs BRIEF REPORTS 287 was 1 death, 1 heart transplantation, and 3 HF hospitalizations in the group with poor SPP reserve and no end points in the group with adequate SPP reserve. There were 4 ICD shocks in the group with poor SPP reserve and 1 ICD shock in the group with adequate SPP reserve. Kaplan-Meier analysis of the composite end point of death, heart transplantation, or HF hospitalization showed a significantly (p П 0.02) greater event rate in the group with poor SPP reserve (Figure 1). SPP reserve discriminated the 2 groups of patients with HF and ICDs without overlap, whereas the ejection fraction showed considerable overlap. The mean SPP reserve of the group with death or transplantation was 0.63 П® 0.4 W/ml2, and the mean SPP reserve of the surviving patients was 4.9 П® 2.3 W/ml2 (p П 0.02). predict survival in patients with ICDs and is not easily measured in patients with more advanced HF and those unable to reach aerobic threshold. Therefore, an ideal prognostic tool would have the predictive power of VO2 max, be noninvasive, and be easy to use in patients with advanced HF. One such instrument that has emerged as a novel echocardiographic measure of ventricular function is SPP reserve. It has an advantage over other echocardiographic measures in that it is not significantly affected by afterload, can be adjusted for preload,8 and correlates well with VO2 max.2 The present study demonstrates that SPP reserve identifies patients with advanced HF and ICDs who are more likely to have progressive HF. ••• For ICDs to show a continued survival benefit in patients with advanced HF, the risk for sudden cardiac death from life-threatening ventricular arrhythmias must be sufficiently greater than the risk for dying from other causes.3 As ICD indications continue to expand, it is increasingly important to discriminate patients who are less likely to die of HF, thus identifying a subpopulation more likely to enjoy a survival benefit from ICD therapy. The prognostic values of the left ventricular systolic ejection fraction, mitral deceleration time, and serum sodium concentration for mortality have been well established.4 – 6 These measurements are insensitive and are relatively poor predictors of survival in patient with advanced HF. Peak oxygen consumption stress testing (VO2 max) has greater prognostic ability than the left ventricular ejection fraction and has been useful for the risk stratification of ambulatory patients who may benefit from cardiac transplantation.7 It has not been used to 288 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 1. Marmor A, Schneeweiss A. Prognostic value of noninvasively obtained left ventricular contractile reserve in patients with severe heart failure. J Am Coll Cardiol 1997;29:422– 428. 2. Armstrong GP, Carlier SG, Thomas JD. Estimation of cardiac reserve by peak power: validation and initial application of a simplified index. Heart 1999;82: 357–364. 3. Fogoros RN. Impact of the implantable defibrillator on mortality: the axiom of overall implantable cardioverter-defibrillator survival. Am J Cardiol 1996; 78(suppl):57A– 61A. 4. Rihal CS, Nishimura RA, Hatle LK. Systolic and diastolic function in patients with clinical diagnosis of dilated cardiomyopathy: relation to symptoms and prognosis. Circulation 1994;90:2772–2779. 5. Xie GY, Berk MR, Smith MD. Prognostic value of Doppler transmitral flow patterns in patients with congestive heart failure. J Am Coll Cardiol 1994;24: 132–139. 6. Lee WH, Packer M. Prognostic importance of serum sodium concentration and its modification by converting enzyme inhibition in patients with severe chronic heart failure. Circulation 1986;73:257–267. 7. Mancini DM, Eisen H, Kussmaul W. Value of peak exercise oxygen consumption for optimal timing of cardiac transplantation in ambulatory patients with heart failure. Circulation 1991;83:778 –786. 8. Sharir T, Feldman MD, Haber H. Ventricular systolic assessment in patients with dilated cardiomyopathy by preload-adjusted maximal power. Validation and noninvasive application. Circulation 1994;89:2045–2053. JANUARY 15, 2005 Effect of Hormone Therapy on Mortality Rates Among Women With Heart Failure and Coronary Artery Disease Kirsten Bibbins-Domingo, PhD, MD, Feng Lin, MS, Eric Vittinghoff, PhD, Elizabeth Barrett-Connor, MD, Stephen B. Hulley, MD, MPH, Deborah Grady, MD, and Michael G. Shlipak, MD, MPH Randomized, controlled trial data from the Heart and Estrogen-progestin Replacement Study were used to evaluate the effect of estrogen plus progestin use on all-cause mortality in women with heart failure and coronary disease. Over the 4.1-year follow-up, estrogen plus progestin use had no effect on all-cause mortality (hazard ratio 1.0, 95% confidence interval 0.7 to 1.4, p ‫ ؍‬0.8) in women with heart failure and coronary disease. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:289 –291) bservational studies suggest that hormone therapy may reduce mortality risk in women with O heart failure. We used data from the Heart and Estrogen-progestin Replacement Study (HERS), a randomized, double-blind, placebo-controlled trial of the effects of postmenopausal hormone therapy on cardiovascular outcomes in women with established coronary artery disease, to evaluate the effect of randomization to estrogen plus progesterone or placebo on all-cause mortality in women with heart failure and coronary artery disease. Of the 2,763 women with coronary artery disease enrolled in HERS, 644 women had heart failure. Over the 4.1 years of follow-up in this double-blinded trial, estrogen plus progestin use had no effect on all-cause mortality (hazard ratio [HR] 1.0, 95% confidence interval [CI] 0.7 to 1.4, p П 0.8) in women with heart failure and coronary artery disease. ••• HERS methods have been previously described.1,2 Briefly, participants were postmenopausal women with known coronary artery disease defined by previous myocardial infraction, coronary artery bypass surgery, percutaneous transluminal coronary angioplasty, or angiographic evidence of Х†50% narrowing of Х†1 major coronary artery. Women were excluded if they were Пѕ79 years of age or had had hysterectomies, had From the Division of General Internal Medicine, San Francisco General Hospital, University of California, San Francisco, San Francisco, California; the Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, California; the Department of Family and Preventive Medicine, University of California, San Diego, San Diego, California; and the Section of General Medicine, San Francisco VA Medical Center, University of California, San Francisco, San Francisco, California. Dr. Bibbins-Domingo’s address is: Box 1364, San Francisco General Hospital, University of California, San Francisco, San Francisco, California 94143-1364. E-mail: [email protected]. Manuscript received July 14, 2004; revised manuscript received and accepted September 3, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MPH, coronary events Х…6 months before randomization, had serum triglyceride levels Пѕ3.39 mmol/L (300 mg/dl), had used hormones Х…3 months before the study, or had a history of conditions that would contraindicate estrogen therapy.2 Women who were found to be in New York Heart Association class IV at baseline physical examination were also excluded from HERS. Participants in HERS were randomly assigned within the 20 HERS clinical centers to conjugated equine estrogens 0.625 mg plus medroxyprogesterone acetate 2.5 mg in 1 tablet daily (n П 1,380) or a placebo of identical appearance (n П 1,383). The HERS randomized, controlled trial was conducted over 4.1 years. After the trial, women were unblinded to treatment assignment and observed for an additional 2.7 years on average (HERS II).3 Women were defined as having heart failure on the basis of a self-reported history of heart failure at entry into HERS (n П 343) or diagnoses of heart failure during HERS follow-up on the basis of signs or symptoms at annual (n П 143) visits or hospital admissions for heart failure (n П 158). Participants were included in this analysis and contributed follow-up information only after they met 1 of these eligibility criteria. The left ventricular ejection fraction was determined by chart review and was recorded in 442 women (69%) in this analysis. We used an unadjusted, intention-to-treat Cox model to assess the effect of assignment to hormone therapy on mortality risk during the HERS main trial. Using the methods given in Therneau and Grambsch,4 the 114 deaths during the HERS trial provided 80% power to detect HRs of 1.7 in 2-sided tests with an вђЈ level of 5%. We performed an unadjusted intentionto-treat Cox model using data from the HERS main trial and the HERS II follow-up period. We also undertook an “as-treated” analysis in which the follow-up time for women receiving hormone therapy was ascribed to placebo when they stopped using hormone therapy, and the follow-up time for women receiving placebo was ascribed to hormone therapy if they used open-label hormone therapy. This “astreated” analysis was adjusted for other factors associated with mortality in women with heart failure and coronary artery disease (New York Heart Association classification; atrial fibrillation; diabetes; creatinine clearance ПЅ60 ml/min; Пѕ1 myocardial infarction, age Пѕ70 years, current tobacco use, limited exercise, left ventricular ejection fraction, race, left bundle branch block, and the use of aspirin, angiotensin-converting enzyme inhibitors, вђ¤ blockers, statins, calcium chan0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.021 289 TABLE 1 Baseline Characteristics of 644 Women With Heart Failure and Coronary Disease Characteristic Age (mean yrs) Race White Black Atrial fibrillation Diabetes Previous myocardial infarction 1 Пѕ1 Systolic blood pressure Пѕ140 mm Hg Creatinine clearance ПЅ60 ml/min Total cholesterol (mean mg/dl) Low-density lipoprotein cholesterol (mean mg/dl) Body mass index (mean kg/m2) Exercise less than peers Smoking Former smoker Current smoker Left ventricular ejection fraction Х…50% Пѕ50% New York Heart Association classification II III or IV Angiotensin-converting enzyme inhibitor use вђ¤-blocker use Aspirin use Digoxin use Estrogen Plus Progestin (n П 324) 67.7 П® 6.6 271 41 12 112 (84) (13) (4) (35) Placebo (n П 320) 67.9 П® 6.6 269 37 5 93 (84) (12) (2) (29) 159 (49) 20 (6) 136 (42) 177 (55) 229.5 П® 41.7 147.3 П® 37.6 160 (50) 28 (9) 129 (40) 194 (61) 227.1 П® 42.6 143.2 П® 39.9 29.7 П® 6.0 104 (32) 29.4 П® 5.7 142 (44) 155 (48) 39 (12) 166 (52) 42 (13) p Value 0.7 0.8 0.09 0.1 0.5 0.7 0.1 0.5 0.2 0.5 0.001 0.4 95% CI 0.7 to 1.2, p П 0.4 for HERS plus HERS II). Results were also unchanged when we restricted the analysis to the pure randomized design, including only the 343 women with heart failure at the time of randomization (HR 0.9, 95% CI 0.5 to 1.4, p П 0.6 for HERS; HR 0.9, 95% CI 0.6 to 1.3, p П 0.5 for HERS plus HERS II). Of the 442 women with recorded left ventricular ejection fractions, 282 (60%) had left ventricular ejection fractions Х†50%. We tested for interactions in each of the models described previously and found no evidence for interaction between hormone treatment and the left ventricular ejection fraction as determinants of survival. ••• Two recent observational studies that used data collected from 89 (28) 93 (29) 138 (43) 122 (38) clinical trials of interventions for 0.8 heart failure have found that the 42 (13) 41 (13) use of postmenopausal hormone 9 (3) 9 (3) therapy was associated with de99 (31) 100 (31) 0.8 creased mortality in women with 103 (32) 100 (31) 0.9 heart failure.5,6 Postmenopausal 238 (74) 231 (72) 0.7 hormones have been postulated to 76 (24) 71 (22) 0.7 be of particular benefit to women with heart failure through their effects on neurohormonal activanel blockers, digoxin, and diuretics). We tested for the tion,7–10 endothelial function,9,11 and cardiac repossibility of interactions between hormone therapy modeling.12 Animal studies have found that and the left ventricular ejection fraction in each of the estrogen is associated with reductions in pre- and models described previously using a significance level afterload.13–16 of p ПЅ0.2 to identify interactions. The investigators of previous observational studOf the 2,763 women in HERS, 644 had heart ies that have reported an association between postfailure. Of these, 324 had been randomly assigned at menopausal hormone therapy and reduced mortality the beginning of HERS to receive estrogen plus pro- in women with heart failure have called for a rangestin and 320 had been assigned to placebo. The domized, controlled trial to confirm these results. women assigned to hormone therapy did not differ Our study uses data from a randomized, controlled significantly in their demographic characteristics, co- trial that compared the cardiovascular effects of morbid conditions, tobacco use, and medication use estrogen plus progestin versus placebo, and we from the women assigned to placebo (Table 1). found no survival advantage associated with horWomen in the placebo group were significantly more mone use in the subset of trial participants with likely than those in the hormone group to report ex- heart failure. Our study has the major advantage of ercising less often than their peers. a randomized, double-blind, placebo-controlled Over the 4.1-year HERS randomized trial, 114 of trial design, allowing the evaluation of the hormone the 644 women with heart failure died (18%). Hor- effect independent of measured and unmeasured mone use did not affect mortality rates during the trial confounding factors. (Table 2). During the 2.7 years of the HERS II obserThe likely explanation for the difference between vation period after the conclusion of the trial, an the previous observational studies of hormone therapy additional 99 women with heart failure died (total in women with heart failure and the results of our mortality 25%). Hormone use was not associated with randomized trial is the inability of observational studmortality risk during the HERS II observation period ies to exclude or adjust for residual confounding facor during the combined 6.8 years of the HERS trial tors that put hormone users at less risk than nonusers. plus HERS II follow-up (Figure 1). Another possibility for the different results of previous The use of an as-treated analysis and adjustment studies is that our study population may have had less for other risk factors did not change the findings (HR severe heart failure than those in the heart failure 0.9, 95% CI 0.6 to 1.4, p П 0.7 for HERS; HR 0.9, intervention trials, and many in our study had heart 290 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ 0.5 VOL. 95 JANUARY 15, 2005 TABLE 2 HR Associated With Estrogen Use During the HERS and HERS II Follow-up Periods Period HR (95% CI) p Value HERS trial (4.1 yrs) HERS-II observation study (2.7 yrs) Combined HERS and HERS-II study (6.8 yrs) 1.0 (0.7–1.4) 0.9 (0.6–1.3) 0.8 0.6 0.9 (0.7–1.2) 0.6 power to detect the small benefit of hormone use on mortality in women with heart failure. Our study does not address the impact of hormone use in women with nonischemic heart failure. However, the absence of an effect of hormone treatment in HERS women with ischemic heart failure, coupled with the results of large randomized trials showing that hormone therapy increases the risk for coronary artery disease events, stroke, and pulmonary embolism1,18 argues for caution in the pursuit of another trial of hormone therapy for women with cardiovascular disease. 1. Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B, Vittinghoff E. FIGURE 1. Kaplan-Meier survival curve by the use of estrogen plus progestin over the HERS plus HERS II follow-up period. failure with preserved systolic function. We cannot directly address whether a potential mortality benefit of hormone therapy is limited to those with severe heart failure with depressed systolic function. Our study population also had coexisting coronary artery disease. In a trial of bucindolol for the treatment of patients with heart failure, the use of postmenopausal hormone therapy was associated with reduced mortality in women with nonischemic heart failure but not those with ischemic disease.5 However, this interaction has been questioned17 and has not been consistently observed.6 A limitation of our study is that the assessment of the left ventricular ejection fraction was based on chart review at baseline, the measures were not standardized, and a significant percentage was missing. In calling for a randomized, controlled trial of hormone therapy in women with heart failure, the investigators of these observational studies suggested that such trials offer clear advantages to observational data when exploring the association between therapies and outcomes. We believe that our finding of no mortality benefit of hormone therapy in women with heart failure using data from a randomized, controlled trial makes it less likely that such a trial would find a positive effect, thus making its design and conduct more difficult to justify. Although we cannot exclude the possibility of a small positive or negative effect of hormone use on mortality, the CIs that we observed suggest that only a small effect (ПЅ30% benefit or ПЅ40% harm) could have been missed by our study. Therefore, a very large trial would be required to have sufficient Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS) Research Group. JAMA 1998;280:605– 613. 2. Grady D, Applegate W, Bush T, Furberg C, Riggs B, Hulley SB. Heart and Estrogen/progestin Replacement Study (HERS): design, methods, and baseline characteristics. Control Clin Trials 1998;19:314 –335. 3. Grady D, Herrington D, Bittner V, Blumenthal R, Davidson M, Hlatky M, Hsia J, Hulley S, Herd A, Khan S, et al. Cardiovascular disease outcomes during 6.8 years of hormone therapy: Heart and Estrogen/progestin Replacement Study follow-up (HERS II). JAMA 2002;288:49 –57. 4. Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. Berlin, Germany: Springer-Verlag; 2000:62. 5. Lindenfeld J, Ghali JK, Krause-Steinrauf HJ, Khan S, Adams K, Goldman S, Peberdy MA, Yancy C, Thaneemit-Chen S, Larsen RL, et al. Hormone replacement therapy is associated with improved survival in women with advanced heart failure. J Am Coll Cardiol 2003;42:1238 –1245. 6. Reis SE, Holubkov R, Young JB, White BG, Cohn JN, Feldman AM. Estrogen is associated with improved survival in aging women with congestive heart failure: analysis of the vesnarinone studies. J Am Coll Cardiol 2000;36:529 –533. 7. Komesaroff PA, Sudhir K, Esler MD. Effects of estrogen on stress responses in women. J Clin Endocrinol Metab 1999;84:4292– 4293. 8. Weitz G, Elam M, Born J, Fehm HL, Dodt C. Postmenopausal estrogen administration suppresses muscle sympathetic nerve activity. J Clin Endocrinol Metab 2001;86:344 –348. 9. Lieberman EH, Gerhard MD, Uehata A, Walsh BW, Selwyn AP, Ganz P, Yeung AC, Creager MA. Estrogen improves endothelium-dependent, flowmediated vasodilation in postmenopausal women. Ann Intern Med 1994;121: 936 –941. 10. Nogawa N, Sumino H, Ichikawa S, Kumakura H, Takayama Y, Nakamura T, Kanda T, Mizunuma H, Kurabayashi M. Effect of long-term hormone replacement therapy on angiotensin-converting enzyme activity and bradykinin in postmenopausal women with essential hypertension and normotensive postmenopausal women. Menopause 2001;8:210 –215. 11. Saitta A, Altavilla D, Cucinotta D, Morabito N, Frisina N, Corrado F, D’Anna R, Lasco A, Squadrito G, Gaudio A, et al. Randomized, double-blind, placebocontrolled study on effects of raloxifene and hormone replacement therapy on plasma no concentrations, endothelin-1 levels, and endothelium-dependent vasodilation in postmenopausal women. Arterioscler Thromb Vasc Biol 2001; 21:1512–1519. 12. Lim WK, Wren B, Jepson N, Roy S, Caplan G. Effect of hormone replacement therapy on left ventricular hypertrophy. Am J Cardiol 1999;83: 1132–1134. 13. Sharkey LC, Holycross BJ, Park S, Shiry LJ, Hoepf TM, McCune SA, Radin MJ. Effect of ovariectomy and estrogen replacement on cardiovascular disease in heart failure-prone SHHF/Mcc-fa cp rats. J Mol Cell Cardiol 1999;31:1527– 1537. 14. Nekooeian AA, Pang CC. Estrogen restores role of basal nitric oxide in control of vascular tone in rats with chronic heart failure. Am J Physiol 1998; 274:H2094 –H2099. 15. Nekooeian AA, Pang CC. Effects of estrogen on venous function in rats with chronic heart failure. Am J Physiol Heart Circ Physiol 2000;278:H1941– H1947. 16. Sudhir K, Chou TM, Mullen WL, Hausmann D, Collins P, Yock PG, Chatterjee K. Mechanisms of estrogen-induced vasodilation: in vivo studies in canine coronary conductance and resistance arteries. J Am Coll Cardiol 1995; 26:807– 814. 17. Petitti DB. New hope for hormone replacement and the heart? J Am Coll Cardiol 2003;42:1246 –1248. 18. Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML, Jackson RD, Beresford SA, Howard BV, Johnson KC, et al. WGftWsHI. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 2002;288:321–333. BRIEF REPORTS 291 Aortic Dimensions in Patients With Bicuspid Aortic Valve Without Significant Valve Dysfunction Moreno Cecconi, MD, Marcello Manfrin, MD, Alessandra Moraca, MD, Raffaele Zanoli, PhD, Pier Luigi Colonna, MD, Maria Grazia Bettuzzi, MD, Stefano Moretti, MD, Domenico Gabrielli, MD, and Gian Piero Perna, MD The dimensions of the entire aorta at different anatomic levels were measured by transthoracic 2-dimensional echocardiography in 162 consecutive patients with isolated bicuspid aortic valves (BAVs) without significant aortic valve dysfunction. Aortic dilation involved the aortic root and the ascending aorta but was not present in the descending and abdominal aorta. A significant increase in the dimensions of the aortic arch was found in patients with BAVs aged >40 years. Ascending aortic diameter and the extension of aortic dilation were significantly correlated with age, but no correlation was found between aortic dimensions and aortic valve morphology. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:292–294) icuspid aortic valves (BAVs) are associated with dilation of the aortic root and ascending aorta. B However, limited data exist regarding the dimensions 1–12 of the aorta at other anatomic levels in patients with BAVs. In addition, the influence of age and BAV morphology on aortic dilation is not clearly determined.7,11 The purpose of this study was to analyze the size of the entire aorta at different anatomic levels in a large group of consecutive patients with isolated BAVs without significant aortic valve dysfunction to better characterize the associated aortic dilation and to determine whether it is related to age and aortic valvular morphology. ••• Three hundred six consecutive patients with unequivocal diagnosis of BAVs on 2-dimensional transthoracic echocardiography were retrospectively identified. To eliminate any factor influencing the aortic dimensions, patients were excluded if they had aortic stenosis, more than mild aortic regurgitation, or any other congenital anomalies of the heart or great vessels, including aortic coarctation, mitral and tricuspid valve disease, aortic dissection, Marfan’s syndrome, a family history of Marfan’s syndrome, a history of aortic surgery, systemic arterial hypertension, and a history of systemic arterial hypertension. Patients with left ventricular systolic dysfunction (left ventricular ejection fraction on 2-dimensional echocardiography Х…50%) were also excluded. Of the 306 patients initially screened, 144 were excluded. A total of 162 patients (113 males and 49 females) aged 1 to 68 years (mean age 24 П® 20) were thus included in the From Ospedale Cardiologico “G. M. Lancisi,” Ancona; and the Polytechnic University of Marche, Ancona, Italy. Dr. Cecconi’s address is: Via Guazzatore 66, 60027 Osimo (AN), Italy. E-mail: [email protected]. Manuscript received March 30, 2004; revised manuscript received and accepted August 31, 2004. 292 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 study. One hundred sixty-two subjects without known heart disease and with normal 2-dimensional echocardiographic and Doppler studies, matched for age, gender, and body surface area, were included as a control group. Comprehensive transthoracic M-mode, 2-dimensional, and Doppler echocardiographic studies were performed using commercially available equipment. Aortic valve morphology and, particularly, the number and the opening motion of the cusps during systole were carefully examined using multiple cardiac views. A BAV was identified according to previously described criteria.13,14 On the basis of commissural orientation and cusp position in the parasternal short-axis view, 2 morphologic types of BAVs were identified15,16: type 1, with the commissures oriented in a medial–lateral direction, resulting in anterior and posterior cusps, and type 2, with the commissures oriented in an anteroposterior direction, resulting in left and right cusps. The presence and position of raphe, when visualized, was reported. Cusps were examined with regard to relative size, and an eccentricity index was obtained according to previously published criteria.16 A BAV was considered eccentric if the eccentricity index was Х†1.2.16 The presence and degree of aortic valvular disease was assessed by Doppler echocardiography. Aortic stenosis was defined as present if the peak aortic flow velocity, assessed by continuouswave Doppler, was Х†2.5 m/s.17 Aortic regurgitation was graded according to a previously described algorithm.18 Patients with trivial aortic regurgitation were classified as having mild aortic regurgitation. A comprehensive evaluation of the entire aorta was performed for patients with BAVs and controls using multiple views, including left parasternal, left high parasternal, right parasternal, right high parasternal, apical, subcostal, suprasternal, and abdominal views. The aortic dimensions were assessed at end-diastole at the following levels: (1) the annulus, (2) the mid-point of the sinuses of Valsalva, (3) the sinotubular junction, (4) the ascending aorta at the level of its largest diameter, (5) the transverse arch, (6) the descending aorta posterior to the left atrium, and (7) the abdominal aorta just distal to the origin of the renal arteries. Measurements were made using 2-dimensional imaging perpendicular to the long axis of the aorta in views showing the largest aortic diameter. The maximal luminal width was measured at each aortic level. Three to 5 measurements were made and averaged at each aortic level, and results were indexed for body surface area. The presence and prevalence of aortic dilation was established in comparison with the aortic dimensions measured in the control group; dilation 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.098 0.7 1.1 1.3 1.2 0.8 0.9 0.9 П® П® П® П® П® П® П® 11.7 17.3 15.1 16.8 12.8 11.5 9.9 1.1†3‡ 3‡ 3.1‡ 2.1* 1.1 1.4 П® П® П® П® П® П® П® 12.8 20.7 18.2 22.1 13.8 11.7 9.6 0.7 0.9 1 1.1 0.8 1.2 1.1 П® П® П® П® П® П® П® 11.9 16.7 14.8 16 12.9 11.4 8.9 0.9* 2.0†2.4†3.4‡ 2.3 1.6 1.1 П® П® П® П® П® П® П® 12.8 18.2 16.4 21.2 13.2 10.8 8.7 1.8 2.4 2.1 2.5 2 1.5 1.5 П® П® П® П® П® П® П® 12.2 16 13.6 15.4 11.8 9.6 8.3 1.9 2.9* 3.1* 4.3* 1.4 1.3 1.5 П® П® П® П® П® П® П® 13.5 18.6 16.3 18.7 12 10.1 8.2 2.2 2.7 1.8 1.6 1.7 1.6 1.1 П® П® П® П® П® П® П® 15.3 19.5 16.4 18.1 13.3 11.1 9.6 2.8 3.8 3.5 5.1†3.2 1.3 1 Controls (n П 27) BAV (n П 27) 7–12 0–6 Values are expressed as mean П® SD. *p ПЅ0.05; †p ПЅ0.01; ‡p ПЅ0.001. Age Group (yrs) was defined as a diameter greater than the mean П© 2 SDs of the value found in controls for a given aortic level. The extension of dilation was defined as the number of dilated aortic segments. To analyze the relation between age and aortic dimensions, the patients were divided in 5 groups: group 1, aged 0 to 6 years; group 2, aged 7 to 12 years; group 3, aged 13 to 18 years; group 4, aged 19 to 40 years; and group 5, aged Пѕ40 years. We used this division because of the discontinuous distribution of ages in our study population. Data are presented as mean П® 1 SD. Differences in mean values between patients with BAVs and controls were assessed using the independent-samples Student’s t test; differences in variances were accounted for using Levene’s test. A 2-tailed p value of ПЅ0.05 was considered statistically significant. Because of the nonlinear relation between aortic size and body surface area, the relation between age and ascending aortic dimensions was assessed only in patients with BAVs aged Пѕ12 years. The relations among ascending aortic dimensions, age, systolic and diastolic blood pressures, and peak aortic flow velocity were assessed by multiple regression analysis. Baseline characteristics of patients with BAVs and controls are listed in Table 1. Of the entire BAV group, 58 patients (36%) had normally functioning valves, whereas 104 patients (64%) had mild aortic regurgitation. The morphologic analysis of BAVs showed that 89 patients (55%) had type 1 and 73 patients (45%) type 2 BAVs. Raphes were identified in 91 patients (56%); they TABLE 3 Aortic Dimensions at Different Levels in Patients With BAVs and Controls According to Age Values are expressed as mean П® SD. *p ПЅ0.01; †p ПЅ0.001. П® П® П® П® П® П® П® 4 5.1 4.1 4.7 3.8 2.9 2.5 П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® П® 14.4 19.5 16.7 18.6 14.4 12.3 10.5 Controls (n П 24) 4.7 6.5* 6.3†7.3†4.7 3.2 3 BAV (n П 24) П® П® П® П® П® П® П® Controls (n П 32) 15.4 22 19.5 23.7 15.1 12.4 10.2 BAV (n П 32) Annulus (mm/m2) Sinuses of Valsalva (mm/m2) Sinotubular junction (mm/m2) Ascending aorta (mm/m2) Aortic arch (mm/m2) Descending aorta (mm/m2) Abdominal aorta (mm/m2) Controls (n П 162) 13–18 Variable Patients With BAVs (n П 162) BAV (n П 40) TABLE 2 Aortic Dimensions at Different Levels in Patients With BAVs and Controls 15.9 20.8 17.9 22.1 14.8 11.8 9.6 19 – 40 Values are expressed as mean П® SD or number of patients. *p ПЅ0.001. 3.3 5.5 4.7 5.7 4.5 4 3.5 1.3 П® 0.2 20.9 27.6 23.2 26 20.4 16.4 14 1.7 П® 0.5* 5.3 8.2 8.3 9.4* 6.5 4.1 4 68 П® 13 22.1 30.1 27 31.9 20.5 16.4 13.8 65 П® 12 Annulus (mm/m2) Sinuses of Valsalva (mm/m2) Sinotubular junction (mm/m2) Ascending aorta (mm/m2) Aortic arch (mm/m2) Descending aorta (mm/m2) Abdominal aorta (mm/m2) 23.6 П® 19.8 113/49 1.44 П® 0.5 111 П® 20 BAV (n П 39) 23.6 П® 19.8 113/49 1.43 П® 0.5 108 П® 18 Controls (n П 40) Controls (n П 162) Variable Age (yrs) Male/female Body surface area (m2) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Peak aortic velocity (m/s) Patients With BAVs (n П 162) Пѕ40 Variable Controls (n П 39) TABLE 1 Baseline Characteristics of Patients With BAVs and Controls BRIEF REPORTS 293 abdominal aorta. We also found dilation of the aortic arch in patients aged Пѕ40 years. This is the first study to demonstrate the involvement of the aortic arch in patients with BAVs. Previous studies6,7,10 showed that the greater aortic dimension in patients with BAVs was at the level of the sinuses of Valsalva, with comparable size at the sinuses and the proximal ascending aorta. We found that the largest aortic segment was the ascending aorta. Discrepancies with previous data are probably due to the different site of sizing of the ascending aorta; we measured the ascending aortic size at the maximum diameter, which is usually at or near the mid-ascending aorta instead of the proximal ascending aorta, as done in previous studies. These results are in accordance with those of 2 recent studies.9,11 Consequently, the maximum aortic dimensions and the prevalence of aortic dilation found in this study are greater than those previously reported.2,6,7,10 Acknowledgment: We thank Francesco Antonini Canterin, MD, for his critical review and suggestions. FIGURE 1. Extension of aortic dilatation in patients with BAVs according to age. Bars, means; error bars, 95% confidence intervals of means. were located in the anterior cusp in type 1 BAVs and in the right cusp in type 2 BAVs. Eccentric BAVs were found in 86 patients (53%). The aortic dimensions at different anatomic levels are listed in Table 2. In the BAV group, no significant difference in ascending aortic diameter was found between male and female patients (23.3 П® 7.7 vs 24.7 П® 5.6 mm/m2) and between those with and without mild aortic regurgitation (24.1 П® 7.6 vs 23.5 П® 7.1 mm/m2). No correlation was found between aortic dimensions and systolic blood pressure, diastolic blood pressure, and peak aortic flow velocity. Aortic dimensions in the age groups are listed in Table 3. In patients with BAVs who were aged Пѕ12 years, age was significantly correlated with ascending aortic diameter (r П 0.73, p ПЅ0.0001). In the entire BAV group, age was significantly correlated with the extension of aortic dilation (r П 0.42, p ПЅ0.001). Figure 1 shows the extension of aortic dilation in patients with BAVs according to age. The prevalence of dilation of the ascending aorta, assessed in relation to normal values found in the control group, was 28% in group 1, 44% in group 2, 42% in group 3, 65% in group 4, and 77% in group 5. Only age showed a significant independent correlation with ascending aortic dimensions in multiple regression analysis (p ПЅ0.01). No relation was found between the morphologic characteristics of BAVs and aortic dimensions. In particular, the ascending aortic diameter was not statistically different in type 1 and type 2 valves (23.5 П® 6.1 vs 24.4 П® 7.6 mm/m2), in valves with and without raphes (23.9 П® 7.2 vs 23.5 П® 6.6 mm/m2), and in eccentric and symmetric valves (24.2 П® 7.9 vs 23.4 П® 5.9 mm/m2). ••• The results of this study confirm that aortic dilation in patients with BAVs involves the aortic root and the ascending aorta, but is not present in the descending and 294 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 1. Pachulski RT, Weinberg AL, Chan KL. Aortic aneurysm in patients with functionally normal or minimally stenotic bicuspid aortic valve. Am J Cardiol 1991;67:781–782. 2. Hahn RT, Roman MJ, Mogtader AH, Devereux RB. Association of aortic dilation with regurgitant, stenotic and functionally normal bicuspid aortic valves. J Am Coll Cardiol 1992;19:283–288. 3. Burks JM, Illes RW, Keating EC, Lubbe WJ. Ascending aortic aneurysm and dissection in young adults with bicuspid aortic valve: implications for echocardiographic surveillance. Clin Cardiol 1998;21:439 – 443. 4. Ando M, Okita Y, Morota T, Takamoto S. Thoracic aortic aneurysm associated with congenital bicuspid aortic valve. Cardiovasc Surg 1998;6:629 – 634. 5. Sabet HY, Edwards WD, Tazelaar HD, Daly RC. Congenitally bicuspid aortic valves: a surgical pathology study of 542 cases (1991 through 1996) and a literature review of 2,715 additional cases. Mayo Clin Proc 1999;74:14 –26. 6. Nistri S, Sorbo MD, Marin M, Palisi M, Scognamiglio R, Thiene G. Aortic root dilatation in young men with normally functioning bicuspid aortic valves. Heart 1999;82:19 –22. 7. Keane MG, Wiegers SE, Plappert T, Pochettino A, Bavaria JE, St. John Sutton MG. Bicuspid aortic valves are associated with aortic dilatation out of proportion to coexistent valvular lesions. Circulation 2000;102(suppl):III-35–III-39. 8. Fedak PWM, Verma S, David TE, Leask RL, Weisel RD, Butany J. Clinical and pathophysiological implications of a bicuspid aortic valve. Circulation 2002; 106:900 –904. 9. Alegret JM, Duran I, Palazon O, Vernis JM, Ameijide A, Rabassa A, Masana L. Prevalence of and predictors of bicuspid aortic valves in patients with dilated aortic roots. Am J Cardiol 2003;91:619 – 622. 10. Nkomo VT, Enriquez-Sarano M, Ammash NM, Melton LJ III, Bailey KR, Desjardins V, Horn RA, Tajik AJ. Bicuspid aortic valve associated with aortic dilatation. A community-based study. Arterioscler Thromb Vasc Biol 2003; 23:351–356. 11. Novaro GM, Tiong IY, Pearce GL, Grimm RA, Smedira N, Griffin BP. Features and predictors of ascending aortic dilatation in association with a congenital bicuspid aortic valve. Am J Cardiol 2003;92:99 –101. 12. Ferencik M, Pape LA. Changes in size of ascending aorta and aortic valve function with time in patients with congenitally bicuspid aortic valves. Am J Cardiol 2003;92:43– 46. 13. Brandenburg RO Jr, Tajik AJ, Edwards WD, Reeder GS, Shub C, Seward JB. Accuracy of 2-dimensional echocardiographic diagnosis of congenitally bicuspid aortic valve: echocardiographic-anatomic correlation in 115 patients. Am J Cardiol 1983;51:1469 –1473. 14. Chan KL, Stinson WA, Veinot JP. Reliability of transthoracic echocardiography in the assessment of aortic valve morphology: pathological correlation in 178 patients. Can J Cardiol 1999;15:48 –52. 15. Roberts WC. The congenitally bicuspid aortic valve: a study of 85 autopsy cases. Am J Cardiol 1970;26:72– 83. 16. Beppu S, Suzuki S, Matsuda H, Ohmori F, Nagata S, Miyatake K. Rapidity of progression of aortic stenosis in patients with congenital bicuspid aortic valves. Am J Cardiol 1993;71:322–327. 17. Otto CM, Lind BK, Kitzman DW, Gersh BJ, Siscovick DS. Association of aortic valve sclerosis with cardiovascular mortality and morbidity in the elderly. N Engl J Med 1999;341:142–147. 18. Padial LR, Oliver A, Vivaldi M, Sagie A, Freitas N, Weyman AE, Levine RA. Doppler echocardiographic assessment of progression of aortic regurgitation. Am J Cardiol 1997;80:306 –314. JANUARY 15, 2005 Is Transesophageal Echocardiography Overused in the Diagnosis of Infective Endocarditis? Molly Thangaroopan, MD, Because of its greater sensitivity, transesophageal echocardiography (TEE) is often misused as a screening tool for the exclusion of infective endocarditis (IE) in patients with small clinical probability of the disease. This study examined the role of using TEE exclusively at a Canadian tertiary care center for the diagnosis of IE and determined which clinical variables are most often associated with positive or negative echocardiographic results supporting or refuting the diagnosis. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:295–297) uke’s criteria, which form the basis of current guidelines, incorporate echocardiographic docuD mentation of vegetations in establishing a diagnosis of infective endocarditis (IE).1–3 Because of its greater sensitivity, transesophageal echocardiography (TEE) is often misused as a screening tool for the exclusion of IE in patients with small clinical probability of the disease.4 – 6 This study examined the role of using TEE exclusively at a Canadian tertiary care center for the diagnosis of IE and determined which clinical variables are most often associated with positive or negative echocardiographic results supporting or refuting the diagnosis. ••• From August 31, 2001, to August 30, 2002, 104 consecutive inpatients referred to the University of Alberta Hospital Adult Echocardiography Laboratory for the primary indication of “rule out IE” were reviewed. The study protocol was approved by the Ethics Committee at the University of Alberta and Stollery Children’s Center. Patients’ charts were reviewed to extract clinical findings associated with IE, including intravenous drug use, the presence of an indwelling catheter, high-risk invasive procedures, a history of IE, immunocompromised status, and highrisk predisposition to heart disease. Clinical findings such as fever (Пѕ38.0В°C), documentation of a new regurgitant murmur, and peripheral manifestations (Roth spots, Janeway lesions, Osler’s nodes, splinter hemorrhages) and splenomegaly were also included. Laboratory data were collected, including the number of blood cultures, the number of positive blood culture results, micro-organisms identified, and alternative potential microbiologic evidence of infection. A disFrom the Division of Cardiology, University Health Network, Toronto, Ontario; and the Division of Cardiology, University of Alberta Hospital, Edmonton, Alberta, Canada. Dr. Choy’s address is: Division of Cardiology, 2C2, WM Health Sciences Center, 8440-112 Street, Edmonton, AB T6G 2B7, Canada. E-mail: [email protected]. Manuscript received May 26, 2004; revised manuscript received and accepted August 31, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 and Jonathan B. Choy, MD charge diagnosis of endocarditis and alternative diagnoses were obtained from medical records. Each case was categorized by the von Reyn criteria into high, medium, or low probability of endocarditis using clinical criteria based on previously published definitions.7 TEE was performed using a Philips Sonos 5500 ultrasound machine (Philips Medical Systems, Andover, Massachusetts) with a 4.0- to 7.0MHz multiplane transducer. TEE was performed by American Society of Echocardiography level III– trained cardiologists. Statistical analysis was performed using SAS version 8.2 (SAS Institute Inc., Cary North Carolina). Fisher’s exact test was used to compare the frequencies of prespecified clinical criteria in the 2 groups, positive versus negative echocardiographic results, with a p value of 0.05 defining significance. Forward logistic regression was used to determine independent predictors of having positive echocardiographic results for vegetations. From August 31, 2001, to August 30, 2002, 114 inpatients underwent TEE in our institution for suspected IE. Ten patients already known to have vegetations were having repeat TEE for reassessment and were excluded. Table 1 lists the demographic characteristics of the patients. Seventy-seven patients (74%) had fevers and 70 (67%) had positive blood culture results, of whom only 20 (19%) had persistently positive blood culture results. According to the von Reyn criteria, 87 patients were low, 5 were intermediate, and 12 were high probability for IE. In the high-probability group, 9 of 12 echocardiographic results (75%) were positive for vegetations (Figure 1). Three of 5 patients (60%) in the intermediateprobability group had positive electrocardiographic results. In contrast, only 2 of 87 patients (2.2%) in the low-probability group had vegetations verified on TEE, both of whom had 3 characteristics in common: positive blood culture results for Staphylococcus aureus (S. aureus) infection, immunosuppression, and fever (Пѕ38.0В°C). The remainder of the patients in the lowprobability group had negative echocardiographic results and firm alternative diagnoses for the source of suspected infection. The most common sources of infections in the low-risk group were soft tissue or bone (46%), abscess (15%), pneumonia (14%), and line sepsis (10%). The remaining discharge diagnoses were connective tissue disorder (7%), malignancy (5%), and viral illness (2%). Of the 14 positive echocardiographic results, the causative micro-organism from blood cultures was S. aureus (n П 5), Staphylococcus epidermidis (n П 1), Enterococcus fecalis (n П 4), Streptococcus viridans (n П 2), Streptococcus bovis (n П 1), and Streptococcus mitis (n П 1; Table 2). Gram negative rods were not 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.022 295 TABLE 1 Baseline Demographic Characteristics of Patients Who Underwent TEE for Suspected IE Parameter TABLE 3 Clinical Criteria in Patients With and Without Evidence of Endocarditis on TEE Value Mean age (yrs) Men Fever Positive blood culture results Persistently positive blood culture results Immunocompromised Intravenous drug user Indwelling catheter Predisposing heart disease Previous endocarditis 54 70 77 70 20 54 8 46 24 7 (17–89) (67%) (74%) (67%) (19%) (52%) (8%) (44%) (23%) (7%) Clinical Criterion Previous IE Fever Intravenous drug use Indwelling catheter Rheumatic heart disease Immunosuppression New regurgitant murmur* Persistently positive blood culture results* Streptococcus aureus cultured Streptococcus viridans cultured Peripheral manifestations Positive blood culture results Positive (n П 14) 2 14 3 6 1 7 8 10 (14) (100) (21) (43) (7) (50) (57) (71) Negative (n П 90) 5 63 5 40 6 47 9 10 (6) (70) (6) (44) (7) (52) (10) (11) Univariate p Value NS 0.02 0.07 NS NS NS ПЅ0.001 ПЅ0.001 5 (35) 25 (28) NS 2 (14) 2 (2) 0.09 9 (64) 14 (100) 11 (12) 56 (62) ПЅ0.001 0.004 Values are numbers (percentages). *Independent predictors by multivariate analysis were new regurgitant murmur (p П 0.001) and persistently positive blood culture results (p ПЅ0.001). FIGURE 1. Comparison of positive and negative TEE results according to risk group. TABLE 2 Type of Organism Cultured With Positive Blood Culture Results Organism Cultured S. aureus Staphylococcus epidermis Enterococcus fecalis Streptococcus bovis Streptococcus viridans Gram negative bacilli Streptococcus pneumonia Fungal Other Positive TEE Results Negative TEE Results 5 1 4 0 2 0 0 0 2 24 6 6 1 2 10 2 2 5 found to be a causative agent for endocarditis. There were 2 cases of prosthetic valve endocarditis involving the aortic valve, 1 case of pacemaker lead tip infection, 6 cases of indwelling catheter–associated endocarditis, and 3 cases in patients with a history of intravenous drug use. Of the high-risk patients, 4 (33%) had peripheral embolization (i.e., septic pulmonary emboli, splenic abscess, and infarct); 3 (25%) had central nervous system complications, including stroke, brain abscess, and mycotic aneurysm with rupture; and 1 (8.3%) had prosthetic valve dehiscence and aortic ring abscess with subsequent right coronary artery embolism resulting in acute myocardial infarction and cardiogenic shock. The univariate predictors of positive TEE results were fever, new regurgitant murmur, peripheral man296 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 ifestations, positive blood culture results, persistently positive blood culture results, intravenous drug use, and Streptococcus viridans bacteremia (Table 3). In a forward stepwise logistic regression model, only persistently positive blood culture results (p ПЅ0.001) and new regurgitant murmur (p П 0.001) remained significant independent predictors of positive TEE results. The presence of an intravascular catheter, rheumatic heart disease, and S. aureus infection were not considered independent predictors of positive TEE results. ••• We found that the most TEE studies were performed in patients with small risk for IE, and the results were therefore negative. Many of these TEE studies could probably have been avoided with more judicious screening and the appropriate application of clinical criteria to select patients with a moderate to large risk for IE. Because of the large incidence of complications in our high-risk group (67%), our results support the use of TEE in high-risk patients to acquire prognostic information and for assessment of the risk for complications related to IE, such as heart failure, risk for embolization, valve dehiscence, and abscess formation. TEE in high-risk patients offers additional information, such as characteristics of the vegetation (e.g., size and mobility), the presence of a perivalvular abscess, prosthetic valve dehiscence, or multivalvular involvement. These have been shown in previous studies to help with the prediction of prognosis and tailor the course of management accordingly.8 TEE in the intermediate-risk group, in contrast, likely has the greatest diagnostic utility, confirming the results of previous studies.9,10 In the low-risk group, our study suggests that most TEE studies are unnecessary unless S. aureus bacteremia, immunosuppression, and fever are present. S. aureus bacteremia is a condition associated with a JANUARY 15, 2005 high mortality rate if inadequately treated.11 In the current era, the most common mode of acquisition is hospital-acquired S. aureus bacteremia, not community acquired, as specified in Duke’s criteria.12 With the increasing use of intravascular catheters and prosthetic materials and attendant increases in the risk for staphylococcal infections related to foreign material, the revision of Duke’s original criteria is required to better select patients for TEE with suspected IE who currently are being misclassified as having a small probability of having this syndrome.13 In conclusion, it is well known that diagnostic tests, such as myocardial perfusion imaging, performed in patients with very small or large clinical likelihood of coronary disease offer little additional information that alters patient management.14 Similarly, in patients with either large or small clinical likelihood of IE, the practical role of TEE for diagnostic purposes is also small. Although Duke’s criteria essentially suggest that echocardiography be used as a screening test for IE, our study does not support this strategy. Acknowledgment: We would like to thank EPICORE at the University of Alberta for aid in statistical analysis. 1. Durack DT, Lukes BS, Bright DK, for the Duke Endocarditis Service. New criteria for the diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994;96:200 –209. 2. Olaison L, Hogevik H. Comparison of the von Reyn and Duke criteria for the diagnosis of infective endocarditis: a critical analysis of 161 episodes. Scand J Infect Dis 1996;28:399 – 406. 3. Bayer AS, Bolger AF, Taubert KA, Wilson W, Steckelberg J, Karchmer AW, Levison M, Chambers HF, Dajani AS, Gewitz MH, et al. Diagnosis and management of infective endocarditis and its complications. Circulation 1998;98: 2936 –2948. 4. Erbel R, Rohmann S, Drexler M. Improved diagnostic value of echocardiography in patients with infective endocarditis by transesophageal approach: a prospective study. Eur Heart J 1988;9:43–53. 5. Birminghan GD, Rahko PS, Ballantyne F. Improved detection of infective endocarditis with transesophageal echocardiography. Am Heart J 1992;123: 774 –781. 6. Shapiro SM, Young E, De Guzman S. Transesophageal echocardiography in the diagnosis of infective endocarditis. Chest 1994;105:377–382. 7. von Reyn CF, Levy BS, Arbeit RD, Friedland G, Crumpacker CS. Infective endocarditis: an analysis based on strict case definitions. Ann Intern Med 1981; 94:505–518. 8. Sanfilippo AJ, Picard MH, Newell JB, Rosas E, Davidoff R, Thomas JD, Weyman AE. Echocardiographic assessment of patients with infectious endocarditis: prediction of risk of complications. J Am Coll Cardiol 1991;18:1191–1199. 9. Kuruppu JC, Corretti M, Mackowiak P, Roghmann MC. Overuse of transthoracic echocardiography in the diagnosis of native valve endocarditis. Arch Int Med 2002 1992;162:1715–1720. 10. Lindner JR, Case RA, Dent JM, Abbott RD, Scheld WM, Kaul S. Diagnostic value of echocardiography in suspected endocarditis: an evaluation based on the pretest probability of disease. Circulation 1996;93:730 –736. 11. Rosen AB, Fowler VG, Corey GR, Downs SM, Biddle AK, Li J, Jollis JG. Cost effectiveness of transesophageal echocardiography to determine the duration of therapy for intravascular catheter associated staphylococcus aureus bacteremia. Ann Intern Med 1999;130:810 – 820. 12. Fowler VG, Li J, Corey R, Boley J, Marr KA, Gopal AK, Kong LK, Gottlieb G, Donovan CL, Sexton DJ, et al. Role of echocardiography in evaluation of patients with Staphylococcus aureus bacteremia: experience in 103 patients. J Am Coll Cardiol 1997;30:1072–1078. 13. Lamas CC, Eykyn SJ. Suggested modifications to the Duke criteria for the clinical diagnosis of native valve and prosthetic valve endocarditis: analysis of 118 pathologically proven cases. Clin Infect Dis 1997;25:713–719. 14. Diamond GA, Denton TA, Berman DS, Cohen A. Prior restraint: a Bayesian perspective on the optimization of technology utilization for diagnosis of coronary artery disease. Am J Cardiol 1995;76:82– 86. Symptomatic Patients Have Similar Outcomes Compared With Asymptomatic Patients After Carotid Artery Stenting With Emboli Protection Michael H. Yen, MD, David S. Lee, MD, Samir Kapadia, MD, Ravish Sachar, Deepak L. Bhatt, MD, Christopher T. Bajzer, MD, and Jay S. Yadav, MD In a single-center cohort of 174 consecutive patients, we sought to evaluate whether the use of emboli protection devices (EPDs) results in equivalent rates of adverse events in symptomatic and asymptomatic patients after carotid artery stenting (CAS) with EPDs. Death or stroke occurred in 3.3% in the symptomatic group and in 3.5% of the asymptomatic group at 30 days (p ‫ ؍‬NS). At 6 months, there was also no significant difference in the rate of stroke or death between the groups. Unlike surgical revascularization, symptomatic patients did not have a greater risk From the Cleveland Clinic Foundation, Cleveland, Ohio. Dr. Yadav’s address is: Department of Cardiovascular Medicine / F25, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195. E-mail: [email protected]. Manuscript received May 25, 2004; revised manuscript received and accepted August 31, 2004. В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, for stroke and death compared with asymptomatic patients after CAS with EPDs. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:297–300) everal randomized clinical trials have shown that surgical revascularization is superior to medical S management for symptomatic and asymptomatic patients with severe carotid stenoses.1,2 However, most patients in these trials had few medical co-morbidities or anatomic features that are known to increase the risk for major adverse clinical events after carotid endarterectomy (CEA).3 In the major surgical studies, symptomatic patients had substantially higher stroke and death rates than asymptomatic patients, which is reflected in the published CEA practice guidelines.4,5 The introduction of carotid artery stenting (CAS) with emboli protection devices (EPDs) offers an alternative, less invasive method of intervention for patients considered to be at high risk for CEA.6,7 We sought to 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.09.023 297 TABLE 1 Adverse Outcomes at 30 Days and 6 Months Outcomes 30-d Stroke Intracranial hemorrhage Death Stroke or death 6-mo Stroke Intracranial hemorrhage Death Stroke or death Symptomatic (n П 62) Asymptomatic (n П 112) 1 1 1 2 (1.6%) (1.6%) (1.6%) (3.2%) 1 0 3 4 (0.9%) (0%) (2.7%) (3.6%) 1 1 2 3 (1.6%) (1.6%) (3.2%) (4.8%) 1 0 5 6 (0.9%) (0%) (4.5%) (5.4%) FIGURE 1. Kaplan-Meier 30-day and 6-month estimates of survival free of stroke or death in symptomatic and asymptomatic patients who underwent CAS with EPDs. determine whether the hazard associated with symptomatic disease in patients who undergo CEA persists in those treated with CAS in the setting of EPDs. ••• We prospectively followed 174 consecutive patients who underwent CAS of 193 arteries with EPDs from January 2000 to September 2002. Symptomatic patients were defined as patients who had stenoses Х†60% as determined by the North American Symptomatic Endarterectomy Trial criteria1 and had either ipsilateral transient ischemic attacks or strokes Х…6 months before their carotid stenting procedures. A transient ischemic attack was characterized by distinct focal neurologic dysfunction or monocular blindness 298 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 that resolved Х…24 hours after onset. Strokes were defined as the persistence of neurologic symptoms or signs for Пѕ24 hours. All asymptomatic patients had carotid stenoses Х†80%. A neurologist evaluated all patients with complete neurologic examinations and a standardized stroke scale assessment before CAS. All patients who underwent CAS gave written informed consent and were enrolled in protocols for CAS approved by the Cleveland Clinic Foundation Institutional Review Board. Procedural details regarding the CAS procedure have been previously described.8 The AngioGuard emboli capture guidewire (Cordis Corporation, Miami Lakes, Florida) was used in 137 patients, GuardWire (Medtronic Vascular, Santa Rosa, California) in 22 patients, FilterWire EX (Boston Scientific Corporation, Natick, Massachusetts) in 11 patients, Accunet (Guidant Corporation, Indianapolis, Indiana) in 3 patients, and Emboshield (Abbott Vascular Devices, Redwood City, California) in 1 patient. Procedural success was defined as a ПЅ30% residual stenosis as determined by angiography immediately after stent placement. The primary end point was death or any stroke at 30-day and 6-month follow-up. Postoperative stroke events were defined as any neurologic deficit persisting for Пѕ24 hours and were adjudicated by a neurologist who did not participate in the CAS procedure. Mortality was determined by the Social Security Death Index. Event-free survival between patient groups was assessed using Kaplan-Meier analysis, and differences between the curves were assessed by the log-rank test. Categorical values are expressed as values and percentages, whereas continuous variables are presented as mean П® SD. Differences between categorical variables were analyzed with the chi-square test or Fisher’s exact test. For continuous variables, 2-tailed t tests were used to assess differences. A p value ПЅ0.05 was considered statistically significant. Statistical analyses were performed using SPSS statistical software version 9.0 (SPSS, Inc., Chicago, Illinois). In our cohort, 36% of the patients were symptomatic (n П 62) and 64% were asymptomatic (n П 112). Baseline characteristics were similar between the 2 study groups except for a higher incidence of previous transient ischemic attacks (81% vs 13%, p П 0.001) and a lower incidence of previous CEA (16% vs 34%, p П 0.03) in the symptomatic group. Procedural success was achieved in 100% of patients in the 2 groups. There were no significant differences between the groups in either the 30-day or 6-month outcomes (Table 1 and Figure 1). There was only 1 stroke in each group at 6-month follow-up, each of which was attributed to periprocedural embolic events (Table 2). There was 1 myocardial infarction in each group at 30-day follow-up and, at 6-month follow-up, there was 1 myocardial infarction in the symptomatic group and 4 in the asymptomatic group (p П NS). ••• Although symptomatic and asymptomatic patients benefit from surgical revascularization for significant JANUARY 15, 2005 results compare favorably with those of recent reports in a global registry7 Time After and other studies,18 in which the 30Carotid Index day stroke and death rates were relaAge (yrs) Artery Procedure Event tively low in symptomatic and asympSymptomatic tomatic patients who receive cerebral 58 Right internal Periprocedural Right hemispheric embolic stroke protection at time of CAS. 76 Right internal 3.5 d Intracranial hemorrhage/death There are certain limitations with 83 Right internal 38 d Death (cardiac) this single-center study. Patients who 86 Left internal 143 d Death (unknown cause) Asymptomatic underwent CAS with EPDs did not 78 Left internal 1h Left hemispheric embolic stroke have concomitant transcranial Dopp80 Left internal 2.6 d Death (presumed cardiac) ler to compare the degree and fre69 Left internal 25 d Death (cardiac) quency of microembolization be69 Right internal 22 d Death (after cardiac bypass) tween the 2 groups at the 3 phases of 77 Left internal 36 d Death (unknown, after cardiac bypass) 62 Left internal 45 d Death (acute renal failure) the procedure associated with increased embolic events: predilatation, stent deployment, and postdilacarotid stenosis, the risk for stroke or death from CEA tation.19 Because it was not required of patients to is increased in symptomatic patients. In a systemic have a more sensitive evaluation of embolic events overview of 25 trials of CEA in symptomatic and after the procedure, such as diffusion-weighted magasymptomatic patients, Rothwell et al9 reported that netic resonance imaging, potential silent neurologic the overall risk for stroke and/or death within 30 days embolic events were not detected.20 Although the Anin symptomatic patients was 5.18% (95% confidence gioGuard device was the most commonly used at our interval 4.30 to 6.06) compared with 3.35% (95% center, this study was not designed to evaluate differconfidence interval 2.38 to 4.31) in asymptomatic ences in efficacy among various cerebral EPDs. patients. This difference in short-term outcome may Whether similar results can be attained in a lower risk be possibly explained by the increased risk for embo- patient cohort with severe carotid stenoses remains to lization during CEA in symptomatic patients. be determined. Last, given the low incidence of major Atherosclerotic plaque in patients with significant adverse clinical events in our cohort, our study was carotid artery stenosis frequently has vulnerable char- underpowered to demonstrate statistical equivalence acteristics, such as ulceration, thrombus, platelet ag- between symptomatic and asymptomatic patients. gregates, or fibrin clots that may predispose the plaque However, the low event rate in either group provides to embolize during either surgical or percutaneous reassuring data regarding the safety of CAS with revascularization, resulting in periprocedural cerebral EPDs. ischemia.9,10 Previous studies have demonstrated that In summary, the use of emboli protection to premicroembolism as assessed by transcranial Doppler vent periprocedural and postprocedural neurologic occurs more frequently in symptomatic than in asympevents results in equivalent major adverse events in tomatic patients during CEA.11 The higher rate of symptomatic patients with significant carotid stenoses stroke after CEA in the surgical trials of symptomatic patients may be related to a greater burden of unstable who are considered to be at high risk for CEA comatherosclerotic plaques in symptomatic patients, lead- pared with asymptomatic patients with similar coing to increased risk for embolization during revascu- morbidities. By preventing embolization into the celarization.12,13 Before the development of EPDs for rebral circulation, the use of EPDs during CAS CAS, percutaneous revascularization was associated appears to neutralize the increased risk for stroke in with a greater risk for microembolization as detected symptomatic patients that was reported in the CEA by transcranial Doppler than CEA.14 Recent studies trials. Whether CAS with EPDs can be safely perthe completion of have suggested that this greater embolic risk during formed in lower risk patients awaits 15 large, randomized clinical trials. CAS can be attenuated with cerebral embolic protection,15 with a corresponding decrease in major clinical events.7,16 The decrease in cerebrovascular events with the use of EPDs in CAS has recently been con- 1. North American Symptomatic Carotid Endarterectomy Trialists’ Collaborative Group. Beneficial effect of carotid endarterectomy in symptomatic patients with firmed by the Endarterectomy Versus Angioplasty in high-grade carotid stenosis. N Engl J Med 1991;325:445– 453. Patients with Symptomatic Severe Carotid Stenosis 2. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study trial, in which the Safety Committee recommended Group. Endarterectomy for asymptomatic carotid artery stenosis. JAMA 1995; 273:1421–1428. the cessation of unprotected CAS because of the three- 3. Tu JV, Wang H, Bowyer B, Green L, Fang J, Kucey D. Risk factors for death fold greater 30-day rate of stroke compared with CAS or stroke after carotid endarterectomy: observations from the Ontario Carotid Endarterectomy Registry. Stroke 2003;34:2568 –2573. with EPDs.17 4. Moore WS, Mohr JP, Najafi H, Robertson JT, Stoney RJ, Toole JF. Carotid In the present study, we have demonstrated that CAS endarterectomy: practice guidelines. Report of the Ad Hoc Committee to the Joint with EPDs results in similar rates of stroke or death in Council of the Society for Vascular Surgery and the North American Chapter of the symptomatic compared with asymptomatic patients. International Society for Cardiovascular Surgery. J Vasc Surg 1992;15:469 – 479. 5. Biller J, Feinberg WM, Castaldo JE, Whittemore AD, Harbaugh RE, Only 1 periprocedural embolic stroke occurred in either Dempsey RJ, Caplan LR, Kresowik TF, Matchar DB, Toole J, et al. Guidegroup in the setting of cerebral distal protection. Our lines for carotid endarterectomy: a statement for healthcare professionals from TABLE 2 Adverse Clinical Events BRIEF REPORTS 299 a special writing group of the Stroke Council, American Heart Association. Stroke 1998;29:554 –562. 6. Roubin GS, New G, Iyer SS, Vitek JJ, Al-Mubarak N, Liu MW, Yadav J, Gomez C, Kuntz RE. Immediate and late clinical outcomes of carotid artery stenting in patients with symptomatic and asymptomatic carotid artery stenosis: a 5-year prospective analysis. Circulation 2001;103:532–537. 7. Wholey MH, Al-Mubarek N. Updated review of the global carotid artery stent registry. Catheter Cardiovasc Interv 2003;60:259 –266. 8. Yadav JS, Roubin GS, Iyer S, Vitek J, King P, Jordan WD, Fisher WS. Elective stenting of the extracranial carotid arteries. Circulation 1997;95:376 –381. 9. Rothwell PM, Slattery J, Warlow CP. A systematic comparison of the risks of stroke and death due to carotid endarterectomy for symptomatic and asymptomatic stenosis. Stroke 1996;27:266 –269. 10. Valton L, Larrue V, le Traon AP, Massabuau P, Geraud G. Microembolic signals and risk of early recurrence in patients with stroke or transient ischemic attack. Stroke 1998;29:2125–2128. 11. Stork JL, Kimura K, Levi CR, Chambers BR, Abbott AL, Donnan GA. Source of microembolic signals in patients with high-grade carotid stenosis. Stroke 2002;33:2014 –2018. 12. Golledge J, Greenhalgh RM, Davies AH. The symptomatic carotid plaque. Stroke 2000;31:774 –781. 13. Liapis CD, Kakisis JD, Kostakis AG. Carotid stenosis: factors affecting symptomatology. Stroke 2001;32:2782–2786. 14. Jordan WD Jr, Voellinger DC, Doblar DD, Plyushcheva NP, Fisher WS, McDowell HA. Microemboli detected by transcranial Doppler monitoring in patients during carotid angioplasty versus carotid endarterectomy. Cardiovasc Surg 1999;7:33–38. 15. Hobson RW II. CREST (Carotid Revascularization Endarterectomy Versus Stent Trial): background, design, and current status. Semin Vasc Surg 2000;13: 139 –143. 16. Kastrup A, Groschel K, Krapf H, Brehm BR, Dichgans J, Schulz JB. Early outcome of carotid angioplasty and stenting with and without cerebral protection devices: a systematic review of the literature. Stroke 2003;34:813– 819. 17. Mas JL, Chatellier G, Beyssen B. Carotid angioplasty and stenting with and without cerebral protection: clinical alert from the Endarterectomy Versus Angioplasty in Patients With Symptomatic Severe Carotid Stenosis (EVA-3S) trial. Stroke 2004;35:e18 – e20. 18. Cremonesi A, Manetti R, Setacci F, Setacci C, Castriota F. Protected carotid stenting: clinical advantages and complications of embolic protection devices in 442 consecutive patients. Stroke 2003;34:1936 –1941. 19. Al-Mubarak N, Roubin GS, Vitek JJ, Iyer SS, New G, Leon MB. Effect of the distal-balloon protection system on microembolization during carotid stenting. Circulation 2001;104:1999 –2002. 20. Schluter M, Tubler T, Steffens JC, Mathey DG, Schofer J. Focal ischemia of the brain after neuroprotected carotid artery stenting. J Am Coll Cardiol 2003;42:1007–1013. Prognostic Importance of Isolated T-Wave Abnormalities Takuya Yamazaki, MD, Jonathan Myers, Computerized electrocardiograms recorded on 46,950 male veterans were analyzed to demonstrate the prognostic value of T-wave amplitude in a general medical population. There were 3,926 cardiovascular deaths over 6 years. Multivariate survival analysis allowed the development of a T-wave amplitude graphic that provides clinicians with a simple method of estimating the relative risk for cardiovascular mortality from T-wave amplitude in limb lead I. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:300 –304) lthough population studies have highlighted the prognostic value of static T-wave abnormalities in A asymptomatic community cohorts, the clinical im1–3 pression is that such findings are labile and nonspecific. This is at least partially because most of the studies have included ST measurements and have been based on older electrocardiographic technology and scoring systems designed for epidemiologic research. This study was performed in a general medical population to evaluate simple T-wave amplitude measurements from computerized 12-lead electrocardiograms (ECGs) and to compare the prognostic power of isolated T-wave abnormalities to other established electrocardiographic measurements and interpretations. ••• All ECGs obtained at the Palo Alto Veterans Affairs Medical Center from March 1987 to January From the Division of Cardiovascular Medicine, Stanford University Medical Center and Veterans Affairs Health Care System, Palo Alto, California. Dr. Froelicher’s address is: Veterans Affairs Palo Alto Health Care System, 3801 Miranda Ave, 111C, Palo Alto, California 94304. E-mail: [email protected]. Manuscript received July 14, 2004; revised manuscript received and accepted August 31, 2004 300 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 PhD, and Victor F. Froelicher, MD 2000 were digitally recorded and stored in the General Electric MUSE ECG management system (GE Healthcare, Waukesha, Wisconsin). When a patient had Пѕ1 ECG in the database, only the earliest ECG was considered. Computerized measurements from the ECGs as well as several computerized electrocardiographic interpretations were extracted. ECGs obtained in an inpatient or outpatient setting or in the emergency room were coded because they could possibly be from acute clinical events. Women (n П 4,616) and patients with ECGs exhibiting electronic pacing (n П 309) or Wolff-Parkinson-White syndrome (n П 44) were excluded from all analyses. To make the target population specific for isolated repolarization abnormalities, a subset was created by excluding ECGs exhibiting atrial fibrillation, diagnostic Q waves, ST depression (Х…0.5 mm), left ventricular hypertrophy, QRS duration Пѕ120 ms, left bundle branch block, right bundle branch block, and intraventricular conduction delay. This primary analysis group consisted of 27,335 patients, and those remaining (n П 14,662) with abnormal ECGs were considered separately. Standardized computerized electrocardiographic criteria as described by the General Electric 12-lead electrocardiographic analysis program were used for diagnostic Q waves and bundle branch blocks (see MUSE/12-lead ECG Physician Program Manual at http://www.gemedicalsystems.com). Although these interpretations are usually read by a cardiologist before being released as “confirmed,” we used only the computer interpretation. T-wave amplitude in 12 leads was taken as the first deflection (either positive or negative) after the QRS complex and the maximum deviation from the PR isoelectric baseline. In addition to simple voltage measurements, the Minnesota Code 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.099 TABLE 1 Demographics and Electrocadiographic Findings of the Total Study Cohort Classified as to CV Death Status Variable Age (yrs) Body mass index Inpatient Heart rate (beats/min) QRS duration (ms) Minor T-wave change (MC 5-3, 5-4) Major T-wave change (MC 5-1, 5-2) Minor ST change (MC 4-3) Major ST change (MC 4-1, 4-2) QT interval (ms) QTc (ms) QT dispersion (ms) QTc dispersion (ms) Total (n П 41,997) CV Death (n П 3,926) Free of CV Death (n П 38,071) p Value* 56.9 П® 14.4 27.3 П® 5.4 12,246 (29.2%) 74.4 П® 16.7 96.1 П® 16.7 5,583 (13.3%) 3,378 (8.0%) 24,366 (58.0%) 5,723 (13.6%) 386.5 П® 41.5 422.9 П® 27.4 29.1 П® 21.2 30.7 П® 22.3 67.7 П® 11.2 26.8 П® 5.3 1,661 (42.3%) 77.9 П® 17.5 103.9 П® 22.7 824 (21.0%) 757 (19.3%) 2,336 (59.5%) 918 (23.4%) 391.2 П® 45.9 437.6 П® 32.3 28.2 П® 24.1 30.2 П® 25.9 55.7 П® 14.2 27.4 П® 5.5 10,585 (27.8%) 74.1 П® 16.6 95.3 П® 15.7 4,759 (12.5%) 2,621 (6.9%) 22,030 (57.9%) 4,805 (12.6%) 386.0 П® 41.0 421.4 П® 26.3 29.1 П® 20.9 30.7 П® 21.9 ПЅ0.0001 0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 0.05 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.01 0.22 *p values are for comparisons between the subjects who experienced CV death and those free from CV death. MC П Minnesota code. TABLE 2 Demographic and Clinical Characteristics of Primary Normal and Abnormal ECG Groups Variable CV death Annual mortality Age (yrs) Body mass index Inpatient Heart rate (beats/min) QRS duration (ms) QT interval (ms) QTc (ms) QT dispersion (ms) QTc dispersion (ms) Minor T-wave change (MC 5-3, 5-4) Major T-wave change (MC 5-1, 5-2) Minor ST change (MC 4-3) Major ST change (MC 4-1, 4-2) Atrial fibrillation Left ventricular hypertrophy Pathologic Q wave QRS duration Пѕ120 ms Left bundle branch block Right bundle branch block Intraventricular conduction delay Primary Normal Group (n П 27,335) Abnormal ECG Group* (n П 14,662) 1,431 (5.2%) 0.7 53.6 П® 14.0 27.2 П® 5.4 7,364 (26.9%) 73.0 П® 15.2 91.5 П® 9.4 384.3 П® 37.7 417.4 П® 22.7 29.6 П® 19.0 31.1 П® 19.8 2,817 (10.3%) 799 (2.9%) 16,286 (59.6%) — — — — 2,495 (17.0%) 2.9 63.0 П® 13.1 27.5 П® 5.6 4,882 (33.3%) 77.0 П® 19.0 104.7 П® 22.7 390.6 П® 47.6 433.2 П® 32.0 28.1 П® 25.0 29.9 П® 26.4 2,766 (18.9%) 2,579 (17.6%) 8,080 (55.1%) 5,723 (39.0%) 1,266 (8.6%) 2,004 (13.7%) 5,431 (37.0%) — — — 565 (3.9%) 1,615 (11.0%) 1,435 (9.8%) p Value ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0005 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 ПЅ0.0001 *Includes ECGs exhibiting atrial fibrillation, pathological Q waves, left ventricular hypertrophy, ST depression, left bundle branch block, right bundle branch block, intraventricular conduction delay, and QRS duration Пѕ120 ms. Abbreviation as in Table 1. was determined using computer algorithms (4-1 and 4-2 as major ST depression, 4-3 as minor ST depression, 5-1 and 5-2 as a major T-wave abnormalities, and 5-3 and 5-4 as minor T-wave abnormalities).4 Left ventricular hypertrophy was defined using the Romhilt-Estes point score as Пѕ3.5 The Social Security Death Index and the California Health Department Service were used to ascertain vital status as of December 30, 2000. The main outcome measure was cardiovascular (CV) mortality. Descriptive statistics were used to find mean values for continuous variables and to test for normality. Bivariate associations between those who experienced CV death and all others were tested using chi-square tests for categorical data and Student’s t test for con- tinuous variables. A p value ПЅ0.05 was considered significant. Cox proportional-hazards testing was performed to assess the significance and independence of predictors of CV mortality, and models were adjusted for age and heart rate. The T-wave amplitude of standard leads, as well as other established electrocardiographic indicators, was considered in the model. Limb lead aVR was excluded from all analysis because it is a reciprocal lead, and the remaining 11 leads were considered to determine the representative lead. Multivariate models included T-wave amplitude, QRS duration, QT interval, QT dispersion, and minor ST depression (Minnesota code 4-3) for the analyses in the primary normal electrocardiographic group. The BRIEF REPORTS 301 had the greatest predictability, limb lead I was the most powerful predictor in the 2 groups. If lead I was Hazard Ratio for removed, lead V5 took its place. For CV Death each 1-mm (0.1-mV) decrease in amVariable (95% CI) p Value plitude, the risk for CV death inPrimary normal ECG group creased 33%. The histogram of TT-wave ampitude in limb lead I†0.71 (0.68–0.74) ПЅ0.0001 wave amplitude in the primary group QT interval (per 5 ms) 1.03 (1.03–1.05) ПЅ0.0001 was distributed as a bell-shaped QRS duration (per 5 ms) 1.04 (1.01–1.06) ПЅ0.0001 Minor ST depression — NS curve with average amplitude of 1.8 QT dispersion — NS П® 1.1 mm. Abnormal ECG group Other electrocardiographic mea†T-wave ampitude in limb lead I 0.82 (0.80–0.84) ПЅ0.0001 surements, such as ST and T changes QRS duration Пѕ120 ms 1.52 (1.39–1.67) ПЅ0.0001 based on the Minnesota code, QRS Pathologic Q wave 1.48 (1.36–1.61) ПЅ0.0001 Left ventricular hypertrophy 1.36 (1.22–1.51) ПЅ0.0001 duration, QT interval, and QT disQT interval — NS persion, were also significantly assoQT dispersion — NS ciated with outcome when consid*Data are from the Cox proportional-hazards model with age and heart rate adjustment. The variables ered univariately adjusted for age are listed in the order of the model selection in both groups. and heart rate. Because limb lead I †T-wave amplitude was evaluated as a continuous variable in millimeters. demonstrated significant ability to predict CV death, it was entered into multiple models to examine the efentire analysis was repeated in the abnormal electro- fect and interaction with other electrocardiographic cardiographic group with the common electrocardio- characteristics (Table 3). It was found to be signifigraphic abnormalities and indicators in the model. The cantly and independently associated with CV mortalsubset analysis was also repeated only in outpatients ity, with a hazard ratio of 0.71 (95% confidence into remove all possible ECGs associated with acute terval [CI] 0.68 to 0.74) for the normal ECG group. clinical events. Surprisingly, in the abnormal ECG group, simple TRelative risks (RRs) were calculated for candidate wave amplitude outperformed commonly used elecamplitudes to determine criteria for practical clinical trocardiographic indicators, including QRS duration, utility. The RR of each threshold was calculated to diagnostic Q waves, left ventricular hypertrophy, and demonstrate incremental risk per each downward step ST depression (hazard ratio 0.82, 95% CI 0.80 to in amplitude. Kaplan-Meier survival curves were plot- 0.84). The predictive value of simple T-wave amplited accordingly. tude was demonstrated in the absence of other elecDemographics of the total study cohort (41,997 male trocardiographic abnormalities or when they were veterans, mean age 57 years) are listed in Table 1. Most present. In addition, even when T-wave amplitude were outpatients when their ECGs were obtained. relative to QRS voltage was taken into account, the There were 3,926 CV deaths during a mean follow-up results were similar. of 6 years. Patients who experienced CV death were RRs for lead I T-wave amplitude thresholds were older and had higher heart rates and longer QRS plotted using a T-wave amplitude of 1.0 to 2 mm as duration and QT intervals compared with patients free the reference group. Examination of this plot showed from CV death. ST- and T-wave abnormalities were that the RR did not decrease significantly for T waves more prevalent in those who experienced CV death. The characteristics of electrocardiographic indica- Пѕ1 mm, nor did the risk increase significantly for T tors and measurements in the subsets with otherwise waves ПЅ0.5 mm. These thresholds were then used for primary and abnormal ECGs are listed in Table 2. The Figure 1, which provides a simple illustration of our primary normal ECG group consisted of 27,335 male main findings for clinical application showing the RR veterans (mean age 54 years) who experienced 1,431 of each T-wave amplitude cut point with Пѕ1.0 mm as CV deaths during a mean follow-up of 6 years. Av- the reference group. There was a progressive increase erage annual CV mortality was 0.7% per year in the of risk as amplitude decreased, flattened, and inverted. primary group compared with 2.9% in the abnormal From Пѕ1.0 to 0.5 to 1.0 mm, the RR CI was 1.2 to 1.6; group. The abnormal ECG group had greater mortal- to 0.0 to 0.5 mm, the RR was CI 1.9 to 2.5; and to an ity, age, heart rate, QRS duration, and prevalence of inverted T wave, the RR CI was 2.9 to 3.8 (p ПЅ T-wave abnormalities. Most inpatient ECGs were for 0.0001). Survival plots are shown in Figure 2 using the routine presurgical evaluation, and hospital status did T-wave amplitude cut points of ПЅ0 mm П 1, 0 to 0.5 not affect the results of analyses. Adjusting for age and heart rate in the Cox model, mm П 2, 0.5 to 1.0 mm П 3, and Пѕ1.0 mm П 4. A the T-wave amplitude in each lead was considered significant and quantitative separation was demonunivariately. All leads but leads III and V2 were sig- strated among those groups with amplitude categories nificantly associated with CV mortality in the primary in the primary and abnormal groups. The annual CV and abnormal ECG groups. When the leads were mortality for inverted T waves was 3.3% in the normal considered in the same model to determine which lead ECG group and 5.4 in the abnormal ECG group, TABLE 3 Multivariate Analyses of Electrocardiographic Criteria Adjusted for Age and Heart Rate* 302 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 FIGURE 1. RR of T-wave amplitude, using 0.5 to 1.0 mm as the reference group. From >1.0 to 0.5 to 1.0 mm, RR CI was 1.2 to 1.6; to 0.0 to 0.5 mm, RR CI was 1.9 to 2.5; and to inverted T wave, RR CI was 2.9 to 3.8 (p <0.0001). FIGURE 2. Kaplan-Meier survival curves for cut points of T-wave amplitude in limb lead I in the primary (normal ECG) and abnormal ECG groups. In all analyses, the stratification according to T-wave amplitude discriminated among groups of subjects with significantly different mortality rates; that is, the survival rate was lower as T-wave amplitude decreased (p <0.001). whereas the annual CV mortality associated with amplitude Пѕ1.0 mm was 0.5% and 1.6%, respectively. ••• T-wave abnormalities are found in up to 15% of apparently healthy subjects and more frequently in general hospital and clinical settings. In our clinical study, 21% of patients were found to have T-wave abnormalities in the entire cohort, with 13% of those with otherwise normal ECGs and 40% of those with abnormal ECGs. From the Framingham study, isolated T-wave flattering or inversion carried a significant increased risk for morbidity and mortality.1 Dutch researchers used the Minnesota code to classify the ST-T changes of nearly 10,000 apparently healthy men and women.2 They found that ST changes were significantly associated with CV mortality, and isolated T-wave abnormalities had borderline significance. The prognostic impact of isolated minor ST and/or T-wave abnormalities has also been demonstrated in the Chicago Western Electric Company Study cohort, which included annual ECGs over a 5-year baseline period.6 It was found that the repeated occurrence of minor ST-T abnormalities was associated with greater risk than abnormalities that were only present transiently.7 Epidemiologists have used data from the Multiple Risk Factor Intervention Trial to demonstrate the prognostic significance of isolated minor T-wave abnormalities after excluding other coexisting abnormalities.8 Researchers in The Netherlands reported the predictive value of T-wave amplitude and ST-segment level using lead I in a community population.9 T-wave amplitude in lead I had a hazard ratio of 2 for ПЅ0.5 mm and 1/3 for 1.5 mm after adjustment for clinical risk factors. Our study confirms the findings of these community epidemiologic studies but extends the importance of T-wave amplitude to general medical populations. A recent review focused on the ventricular repolarization components on surface ECGs, including ST segments and T and U waves.10 The investigators concluded that the T wave represents the transmural dispersion of repolarization. Clinicians consider T-wave abnormalities as “nonspecific” because so many factors affect T-wave amplitude. Serum electrolytes, medications, hyperventilation, intracranial lesions, and recent food ingestion can all produce T-wave amplitude variation. Large T-wave amplitude can be seen in athletes,11 whereas the experimental blocking of parasympathetic activity by the administration of atropine results in a decrease of T-wave amplitude.12 A high prevalence of early repolarization (elevated ST and T wave) has been demonstrated in those with spinal cord injuries, most likely because of enhanced vagal tone or the loss of sympathetic tone.13 Our ECGs were obtained from inpatients and outpatients and represent findings from a very broad range of patients similar to the type seeking general medical evaluation. This contrasts with previous studies of patients, which focused on patients with specific conditions or considered community epidemiologic cohorts. Our sample is representative of everyday medical practice, in which electrocardiography is commonly used as the first assessment tool for possible cardiac disease. Although baseline clinical and laboratory data were not available, it has been well validated that electrocardiography14 and specifically the T wave9 add additional prognostic information to baseline clinical and laboratory data. Furthermore, requiring other tests, comparing T-wave amplitude with the results of other tests, or using other test results as surrogate end points would be very biased because these tests are performed for more specific clinical reasons other than electrocardiography. 1. Kannel WB, Anderson K, McGee DL, Degatano LS, Stampfer MJ. Nonspecific electrocardiographic abnormality as a predictor of coronary heart disease: the Framingham Study. Am Heart J 1987;113:370 –376. 2. De Bacquer D, De Backer G, Kornitzer M, Myny K, Doyen Z, Blackburn H. Prognostic value of ischemic electrocardiographic findings for cardiovascular mortality in men and women. J Am Coll Cardiol 1998;32:680 – 685. 3. Harlan WR, Cowie CC, Oberman A, Mitchell RE, MacIntyre NR. Prediction of subsequent ischemic heart disease using serial resting electrocardiograms. Am J Epidemiol 1984;119:208 –217. BRIEF REPORTS 303 4. Prineas RJ, Crow RS, Blackburn H. The Minnesota Code Manual of Electrocar- 9. Dekker JM, Schouten EG, Klootwijk P, Pool J, Kromhout D. ST segment and diographic Findings. Boston, Massachusetts: John Wright PSG; 1982:223–229. 5. Romhilt DW, Estes EH Jr. A point-score system for the ECG diagnosis of left ventricular hypertrophy. Am Heart J 1968;75:752–758. 6. Daviglus ML, Liao Y, Greenland P, Dyer AR, Liu K, Xie X, Huang CF, Prineas RJ, Stamler J. Association of nonspecific minor ST-T abnormalities with cardiovascular mortality: the Chicago Western Electric Study. JAMA 1999;10: 281:530 –536. 7. Greenland P, Xie X, Liu K, Colangelo L, Liao Y, Daviglus ML, Agulnek AN, Stamler J. Impact of minor electrocardiographic ST-segment and/or T-wave abnormalities on cardiovascular mortality during long-term follow-up. Am J Cardiol 2003;91:1068 –1074. 8. Prineas RJ, Grandits G, Rautaharju PM, Cohen JD, Zhang ZM, Crow RS, MRFIT Research Group. Long-term prognostic significance of isolated minor electrocardiographic T-wave abnormalities in middle-aged men free of clinical cardiovascular disease (the Multiple Risk Factor Intervention Trial [MRFIT]). Am J Cardiol 2002;90:1391–1395. T wave characteristics as indicators of coronary heart disease risk: the Zutphen Study. J Am Coll Cardiol 1995;25:1321–1326. 10. Yan GX, Lankipalli RS, Burke JF, Musco S, Kowey PR. Ventricular repolarization components on the electrocardiogram: cellular basis and clinical significance. J Am Coll Cardiol 2003;42:401– 409. 11. Bjornstad H, Storstein L, Meen HD, Hals O. Electrocardiographic findings in athletic students and sedentary controls. Cardiology 1991;79:290 –305. 12. Annila P, Yli-Hankala A, Lindgren L. Effect of atropine on the QT interval and T-wave amplitude in healthy volunteers. Br J Anaesth 1993;71:736 –737. 13. Marcus RR, Kalisetti D, Raxwal V, Kiratli BJ, Myers J, Perkash I, Froelicher VF. Early repolarization in patients with spinal cord injury: prevalence and clinical significance. J Spinal Cord Med 2002;25:33–38. 14. Ashley EA, Raxwal VK, Froelicher VF. The prevalence and prognostic significance of electrocardiographic abnormalities. Curr Probl Cardiol 2000;25:1–72. Relation Between Effects of Adenosine on Brachial Artery Reactivity and Perfusion Pattern in Patients With Known or Suspected Coronary Artery Disease Deval Mehta, MD, Gurpreet Baweja, MD, Rajesh Venkataraman, MD, Gilbert J. Zoghbi, MD, Thein Htay, MD, Jaekyeong Heo, MD, Navin C. Nanda, and Ami E. Iskandrian, MD This study examined the changes in brachial artery diameter and flow velocity in response to intravenous adenosine and compared the results to cuff occlusion and single-photon emission computed tomographic (SPECT) images. The change in diameter was less with adenosine than with cuff occlusion. There was no correlation between the presence of abnormal SPECT images and the responses to adenosine or cuff occlusion in either diameter or flow velocity. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:304 –307) he purpose of this study was to examine the changes in the brachial artery dimensions and flow T velocity during adenosine infusion and compare the results to those obtained after cuff occlusion of the brachial artery and to myocardial perfusion imaging. ••• The study group consisted of patients with known or suspected coronary artery disease (CAD) referred to our stress test laboratory for adenosine stress single-photon emission computed tomographic (SPECT) imaging for clinical indications. Pertinent demographic data and medical histories were documented at enrollment in the study. Our institutional review board approved the study protocol, and all patients signed informed consent forms From the Division of Cardiovascular Diseases, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama; and the Sarver Heart Institute, Tucson, Arizona. This study was supported by an unrestricted grant from Fujisawa Healthcare, Inc., Deerfield, Illinois. Dr. Iskandrian’s address is: University of Alabama at Birmingham, 318 LHRB, 1900 University Blvd., Birmingham, Alabama 35294-0006. E-mail: [email protected]. Manuscript received June 28, 2004; revised manuscript received and accepted August 26, 2004. 304 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MD, before participating in the study. There were no complications related to the study. Brachial artery reactivity was assessed at the time of adenosine SPECT imaging, using the technique described by Celermajer et al.1 Patients fasted for Х†6 hours, and all medications were withheld the morning of the study. Patients were placed in a supine position, and the right arm was used for the brachial artery measurements. The brachial artery was scanned longitudinally 2 to 5 cm above the elbow with a 10.0MHz linear array transducer using the Acuson-Aspen ultrasound system (Siemens, Palo Alto, California) to find the clearest images of the anterior and posterior walls of the artery. This area was marked on the skin, and all subsequent measurements were performed at the same position. The baseline brachial artery diameter and peak velocity (using color Doppler-guided pulsewave Doppler) were measured, along with blood pressure and heart rate. These measurements were repeated at 1-minute intervals for 5 minutes during intravenous adenosine infusion at a dose of 0.140 mg В· kgПЄ1 В· minПЄ1. Five minutes after the completion of adenosine infusion, a blood pressure cuff was placed around the arm proximal to the position of the ultrasound transducer and inflated to a pressure of 50 mm Hg greater than the measured systolic pressure. After 5 minutes of occlusion, the cuff was deflated, and diameter and velocity measurements were performed immediately and repeated every minute for a total of 5 measurements. All studies were recorded on S-VHS tapes. Off-line quantitative analysis was performed using calibrated electronic calipers to measure brachial artery diameter. The maximal end-diastolic diameter was measured at the same position for all the readings. Adenosine-mediated change in diameter was calcu0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.100 TABLE 1 Baseline Characteristics of the Patients Category Age (yrs) Men Body mass index (kg/m2) Diabetes mellitus Systemic hypertension Hyperlipidemia* Smoker Systolic blood pressure at rest (mm Hg) Systolic blood pressure peak (mm Hg) Heart rate at rest (beats/min) Heart rate peak (beats/min) Total (n П 93) Normal SPECT Images (n П 59) Abnormal SPECT Images (n П 34) p Value 60.4 П® 12.7 56 (60%) 27.1 П® 8.5 43 (46%) 73 (79%) 41 (44%) 27 (29%) 135 П® 32 140.0 П® 41.0 75 П® 13.0 88 П® 16 59.0 П® 12.8 32 (54%) 27.7 П® 7.5 28 (47%) 45 (76%) 21 (36%) 14 (24%) 138 П® 35 143 П® 47.3 76.0 П® 13 89 П® 16 62.9 П® 12.5 24 (71%) 26.1 П® 10.1 15 (44%) 28 (82%) 20 (59%) 13 (38%) 130 П® 25 135 П® 27 73 П® 13 86 П® 16 0.16 0.13 0.36 0.83 0.60 0.03 0.16 0.20 0.36 0.25 0.33 *Total cholesterol Пѕ200 mg/dl or taking lipid-lowering medicines. Results are mean П® SD or numbers and percentages (in parentheses). FIGURE 1. Correlation between adenosine- and cuff occlusion– induced changes in brachial artery diameter. lated as: (peak diameter during adenosine infusion ПЄ baseline diameter)/baseline diameter П« 100. Change in diameter after cuff occlusion was calculated as: (peak diameter after cuff occlusion ПЄ baseline diameter)/baseline diameter П« 100. The percentage changes in velocity with adenosine and after cuff occlusion were also measured in a similar fashion. Intraobserver variation assessments for the brachial artery measurements were performed in 29 patients using correlation coefficients, and the variation was found to be small (r П 0.98, p ПЅ0.001). The initial 40 studies were performed by 1 observer and the remaining by a second observer, both of whom were blinded to the SPECT results. Patients were injected with technetium-99m tetrofosmin (10 to 35 mCi) at the 3-minute point during the adenosine infusion. Gated SPECT images were obtained 60 minutes later using an ADAC dual-head gamma camera (ADAC Laboratories, Milipitas, California) according to methods previously described FIGURE 2. Correlation between adenosine- and cuff occlusion– induced changes in brachial artery velocity. FIGURE 3. Time distribution of peak dilation in brachial artery with adenosine and cuff occlusion. from our laboratory.2 Images at rest were obtained later during the day or on a separate day in patients with abnormal stress images. The images were interpreted without knowledge of the brachial reactivity results. The images were interpreted as normal or BRIEF REPORTS 305 Brachial artery images were of suboptimal quality in 5 of 98 patients, and these were excluded. The Normal SPECT Abnormal SPECT baseline characteristics of the reImages Images maining 93 patients are listed in TaCategory (n П 59) (n П 34) p Value ble 1. Thirty-four patients had known Baseline diameter (mm) 4.1 П® 0.1 4.0 П® 0.1 0.79 CAD. Twenty-four had a history of Percentage diameter change with 8.6 П® 6.4 8.9 П® 6.4 0.84 myocardial infarction, 18 had undercuff occlusion gone coronary bypass surgery, and Percentage diameter change with 4.6 П® 5.7 5.1 П® 6.9 0.70 adenosine 10 had undergone percutaneous corBaseline velocity (cm/s) 58.5 П® 32.0 67.2 П® 34.8 0.23 onary intervention. There was a high Percentage velocity change with 21.0 П® 18.0 26.0 П® 26.5 0.79 prevalence of hypertension, diabetes, cuff occlusion smoking, obesity, and family history Percentage velocity change with 21.0 П® 21.8 19.4 П® 19.6 0.79 of CAD in patients with normal or adenosine abnormal images. The baseline diameter was larger in men than women (4.2 П® 0.7 vs 3.8 П® 1.0 mm, p П 0.04), but the change in diameter was similar in the 2 genders with adenosine (5.9 П® 7.3% in men vs 4.0 П® 5.1% in women, p П 0.15) and after cuff occlusion (8.9 П® 7.0% in men vs 8.6 П® 5.9% in women, p П 0.85). There was no correlation between age and the changes in diameter with adenosine (r2 П 0.002, p П 0.66) or after cuff occlusion (r2 П 0.013, p П 0.28). The changes in diameter with adenosine were not different in patients with and without diabetes (4.7 П® 6.0% vs 4.8 П® 6.3% p П 0.95), hypertension (4.4 П® 5.7% vs 6.3 П® 7.4%, p П 0.23), or hyperlipidemia (3.8 П® 5% vs 5.5 П® FIGURE 4. Cuff occlusion- and adenosine-mediated brachial ar6.7%, p П 0.16). The changes in diameter after cuff tery vasodilatation in patients with (n ‫ ؍‬19) and without (n ‫؍‬ occlusion were also not different in patients with and 74) evidence of ischemia on their adenosine myocardial perfusion images. without diabetes (8.2 П® 7.1% vs 9.2 П® 5.7%, p П 0.45), hypertension (8.5 П® 5.8% vs 9.7 П® 8.2%, p П 0.43), or hyperlipidemia (8.6 П® 7.4% vs 8.6 П® 5.8%, p П 0.85). The percentage change in diameter was less with adenosine than with cuff occlusion (4.8 П® 6.1% vs 8.7 П® 6.4%, p ПЅ0.0001). There was a weak correlation between the change in diameter with adenosine and with cuff occlusion (r2 П 0.20, p ПЅ0.001; Figure 1). The changes in velocity were similar with adenosine and cuff occlusion (20.4 П® 21% vs 23 П® 21.5%, p П 0.32). There was also a weak correlation between adenosine- and cuff occlusion–induced changes in velocity (r2 П 0.17, p FIGURE 5. Prevalence of abnormal SPECT images in different ПЅ0.001; Figure 2). The maximum change in diameter quartiles of patients on the basis of adenosine- and cuff occluoccurred in 65 patients (70%) Х…2 minutes after adenosion–induced brachial artery dilatation. sine infusion and in 71 patients (76%) Х…1 minute after the release of cuff occlusion (Figure 3). Thirty-four patients had abnormal SPECT scan abnormal, showing either reversible defect (ischemia), fixed defect (scar), or both, as previously described. results and 19 patients had evidence of ischemia. The The extent of ischemia was assessed as the number of baseline measurements of diameter and velocity and their responses to adenosine and cuff occlusion were coronary artery territories with reversible defects. Continuous variables are reported as mean П® SD similar in patients with normal (n П 59) and abnormal and discrete data as percentages. Continuous variables (n П 34) SPECT images (Table 2). The changes in were analyzed by the unpaired Student’s t test and brachial artery diameter were also similar in patients discrete variables by the chi-square test. Bonferroni with and without ischemia (Figure 4). The patients were divided into quartiles on the correction was used to adjust for multiple comparisons. Pearson’s method was used to estimate the cor- basis of the changes in diameter with adenosine or relation coefficients between continuous variables. A cuff occlusion. There was no difference in the number p value ПЅ0.05 was considered statistically significant. of patients with abnormal SPECT images in the highAll data were entered into an Excel spreadsheet (Mi- est or lowest quartiles with either adenosine (p П 0.7) crosoft Corporation, Redmond, Washington), and the or cuff occlusion (p П 0.3; Figure 5). In the 19 patients with myocardial ischemia on the analysis was performed using the SPSS version 10.1 SPECT images, there was no correlation between the (SPSS, Inc., Chicago, Illinois) statistical package. TABLE 2 Comparison Between Patients With Normal and Abnormal SPECT Images 306 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 JANUARY 15, 2005 extent of ischemia and baseline diameter (r П 0.04, p П 0.69), change in diameter with adenosine (r П 0.16, p П 0.88), change in diameter with cuff occlusion (r П 0.14, p П 0.19), baseline velocity (r П 0.05, p П 0.67), and change of velocity with adenosine (r П 0.05, p П 0.62) or with cuff occlusion (r П 0.15, p П 0.16). ••• This is the first report on brachial artery response during intravenous adenosine infusion performed simultaneously at the time of adenosine gated SPECT imaging. Our data show that the response of diameter to adenosine is less than that with cuff occlusion, suggesting a smaller increase in flow with adenosine than with cuff occlusion. Moreover, the adenosineinduced changes in velocity were only a fraction of those observed in the coronary circulation.3–5 These differences between different vascular beds might be due to differences in receptor subtype (A-2a) density in the coronary versus systemic circulation or blood levels of adenosine in coronary versus brachial artery after intravenous infusion.6 In our study, we measured diameter and velocity. The percentage change in velocity was much greater with adenosine and cuff occlusion than the change in diameter. However, the SD for the velocity measurement was much wider than that for the change in diameter measurements. This may suggest that the change in diameter may be more robust and reproducible than the change in velocity measurements. The arm position remained stable during the adenosine study, perhaps making such measurements more reproducible. Previous studies have demonstrated endothelial dysfunction in patients with established CAD7 as well as those with preclinical atherosclerosis, such as those with coronary risk factors.8 –11 We did not find a significant difference in adenosine- or cuff occlusion– induced change in diameter or velocity in patients with and without diabetes mellitus, hypertension, or hyperlipidemia. Most previous studies compared pa- tients with 1 or 2 risk factors with healthy subjects, whereas most of our patients had multiple risk factors. Previous studies have suggested that endothelial dysfunction worsens with an increasing number of risk factors.11,12 Thus, the presence of multiple risk factors in our cohort may explain the lack of difference in the brachial artery responses in different subgroups. 1. Celermajer DS, Sorensen KE, Gooch VM, Spiegelhalter DJ, Miller OI, Sullivan ID, Lloyd JK, Deanfield JE. Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. Lancet 1992;340: 1111–1115. 2. Iskandrian AS, Heo J, Nguyen T, Beer S, Cave V, Cassel D, Iskandrian BB. Tomographic myocardial perfusion imaging with technetium-99m teboroxime during adenosine-induced coronary hyperemia: correlation with thallium-201 imaging. J Am Coll Cardiol 1992;19:307–312. 3. Wilson RF, Wyche K, Christensen BV, Zimmer S, Laxson DD. Effects of adenosine on human coronary arterial circulation. Circulation 1990;82:1595–1606. 4. Iskandrian AS, Verani MS, Heo J. Pharmacologic stress testing: mechanism of action, hemodynamic responses, and results in detection of coronary artery disease. J Nucl Cardiol 1994;1:94 –111. 5. Shryock JC, Belardinelli L. Adenosine and adenosine receptors in the cardiovascular system: biochemistry, physiology, and pharmacology. Am J Cardiol 1997;79:2–10. 6. Neunteufl T, Katzenschlager R, Hassan A, Klaar U, Schwarzacher S, Glogar D, Bauer P, Weidinger F. Systemic endothelial dysfunction is related to the extent and severity of coronary artery disease. Atherosclerosis 1997;129:111–118. 7. Balletshofer BM, Rittig K, Stock J, Lehn-Stefan A, Overkamp D, Dietz K, Haring HU. Insulin resistant young subjects at risk of accelerated atherosclerosis exhibit a marked reduction in peripheral endothelial function early in life but not differences in intima-media thickness. Atherosclerosis 2003;171:303–309. 8. Celermajer DS, Sorensen KE, Georgakopoulos D, Bull C, Thomas O, Robinson J, Deanfield JE. Cigarette smoking is associated with dose-related and potentially reversible impairment of endothelium-dependent dilation in healthy young adults. Circulation 1993;88:2149 –2155. 9. Clarkson P, Celermajer DS, Powe AJ, Donald AE, Henry RM, Deanfield JE. Endothelium-dependent dilatation is impaired in young healthy subjects with a family history of premature coronary disease. Circulation 1997;96:3378 –3383. 10. Jensen-Urstad K, Johansson J, Jensen-Urstad M. Vascular function correlates with risk factors for cardiovascular disease in a healthy population of 35-year-old subjects. J Intern Med 1997;241:507–513. 11. Celermajer DS, Sorensen KE, Bull C, Robinson J, Deanfield JE. Endotheliumdependent dilation in the systemic arteries of asymptomatic subjects relates to coronary risk factors and their interaction. J Am Coll Cardiol 1994;24:1468 –1474. 12. Hashimoto M, Kozaki K, Eto M, Akishita M, Ako J, Iijima K, Kim S, Toba K, Yoshizumi M, Ouchi Y. Association of coronary risk factors and endotheliumdependent flow-mediated dilatation of the brachial artery. Hypertens Res 2000; 23:233–238. BRIEF REPORTS 307 Usefulness of Real-Time Three-Dimensional Echocardiography for Reliable Measurement of Cardiac Output in Patients With Ischemic or Idiopathic Dilated Cardiomyopathy Sean M. Fleming, MB, MD, Barry Cumberledge, FSCST, Christoph Kiesewetter, Gareth Parry, MB, and Antoinette Kenny, MB, MD The determination of stroke volume (SV) is a potentially important application of real-time 3-dimensional echocardiography (RT3DE). SV measurements by thermodilution were compared with values obtained using transthoracic RT3DE in a sequential cohort of patients who underwent assessment for potential cardiac transplantation. There was a strong correlation between echocardiographically derived SV and catheterization data (r ‫ ؍‬0.95, n ‫ ؍‬14). On average, RT3DE appeared to underestimate SV by 7.5 ml (SD ‫ ؍‬5.8) or 17% (SD ‫ ؍‬12%). A role for RT3DE in the measurement of SV in severe heart failure is suggested. б®Љ2005 by Excerpta Medica Inc. (Am J Cardiol 2005;95:308 –310) hree-dimensional echocardiography provides accurate measurements of left ventricular (LV) volT ume and function and has been shown to have particular applicability in distorted and remodeled ventricles.1,2 The clinical application of 3-dimensional echocardiography, however, has been limited by prolonged acquisition, reconstruction, and data analysis times.3 Real-time 3-dimensional echocardiography (RT3DE) is now commercially available. By eliminating acquisition and reconstruction delays, RT3DE offers the potential to bring 3-dimensional echocardiography into the clinical mainstream. We examined the application of RT3DE in patients who underwent assessment for cardiac transplantation. We aimed to compare stroke volume (SV) quantification by RT3DE with our clinical standard method, bolus thermodilution. ••• Seventeen sequential patients who underwent right-sided cardiac catheterization at the Department of Cardiac Transplantation at Freeman Hospital (Newcastle upon Tyne, United Kingdom) were enrolled. Images suitable for 3-dimensional analysis were obtained from 14, and these patients formed the study cohort. In this clinically based study, SV as estimated using bolus thermodilution was taken as the clinical From the Departments of Echocardiography and Cardiac Transplantation, Freeman Hospital, High Heaton, Newcastle Upon Tyne, United Kingdom. Dr. Fleming’s address is: Department of Echocardiography, Freeman Hospital, High Heaton, NE7 7DN, United Kingdom. E-mail: [email protected]. Manuscript received April 7, 2004; revised manuscript received and accepted August 31, 2004. 308 В©2005 by Excerpta Medica Inc. All rights reserved. The American Journal of Cardiology Vol. 95 January 15, 2005 MB, MD, reference method. In each patient, repeated injections were performed until 3 measurements were obtained within a 15% range. The average value was taken as the cardiac output and divided by the heart rate at the time of measurement to give SV. Each patient underwent right-sided heart catheterization as part of a workup for potential cardiac transplantation. Threedimensional echocardiographic volumes were obtained using the Philips Sonos 7500 and x4 transducer (Philips Medical Systems, Andover, Massachusetts). This transducer has a fully sampled matrix array with 3,000 physical image-formation channels and a frequency range of 1.6 to 4 MHz. For “full-volume” acquisition of the left ventricle, a series of 4 electrocardiographic-triggered subvolumes are first acquired during suspended respiration. These are then synchronized with the R wave to render a time-aligned full volume data set with a sector size of 60В° by 60В°. This process is automated, giving a total acquisition time equivalent to 4 cardiac cycles. Subsequent rendering is almost instantaneous, allowing the immediate playback and inspection of the 3-dimensional volume. All 3-dimensional data sets were acquired in the cardiac catheterization laboratory, immediately after thermodilution measurements were taken and with the patient in the left lateral decubitus position. At the time of data acquisition, each data set was visually inspected to ensure that image quality was acceptable, that there were no obvious breathing or movement artifacts, and that as much as possible of the left ventricle was contained within the image pyramid. In total, a median of 5 data sets were recorded for each patient. Data sets that were clearly inadequate were not stored. Analysis was performed off-line using semiautomated border detection software (TomTec Imaging Systems, Munich, Germany). Three-dimensional volumes were analyzed using an 8-plane apical rotational method.4 End-diastolic volume (EDV) and end-systolic volume (ESV) were measured and SV estimated by subtraction. All measurements were performed blinded to the cardiac catheterization data. When dealing with a large and distorted ventricle, it may be difficult to fit the entire ventricle within the 3-dimensional data set. Using a 16-segment model, and in the case of each participant, the presence or absence of each wall segment within the full-volume pyramid at end-diastole was noted on a binary scale. The total number of imaged segments for the cohort as a whole and the median number for each participant are reported. 0002-9149/05/$–see front matter doi:10.1016/j.amjcard.2004.08.101 FIGURE 1. Absolute difference plot: SV estimated by thermodilution less SV by RT3DE is plotted against mean SV. The continuous line represents the average absolute difference between the methods (in milliliters), and the dashed lines represent 95% confidence intervals for this average. Statistical analysis was performed using Excel (Microsoft Corporation, Redmond, Washington). Descriptive data are reported as medians and interquartile ranges (IQRs). For comparison between methods, the correlation coefficient is reported. Bland-Altman plots were constructed, and absolute and percentage differences П®2 SDs are reported. All 3-dimensional measurements were performed twice by the first investigator (SMF). These results were averaged to provide the measurements of LV volumes used in the study. Intraobserver error is reported as the average and SE of the percentage difference between each pair of measurements. Analysis was repeated by another investigator (CK) for a random 10 studies, and interobserver error is reported in a similar fashion. In total, 14 patients (10 men) participated (median age 50 years, range 28 to 62). Seven patients had ischemic cardiomyopathy and 7 had idiopathic dilated cardiomyopathy. Seven patients were in New York Heart Association symptom class IV, and 7 were in class III. At standard transthoracic echocardiographic scanning, the median end-diastolic dimension was 66 mm (IQR 60 to 75) and the median ejection fraction was 12.5% (IQR 12.5% to 21%). Four patients had significant valvular regurgitation (1 moderate mitral regurgitation, 2 severe mitral regurgitation, 1 moderate to severe aortic regurgitation). No patient had significant pulmonary regurgitation. The cardiac rhythm was sinus for all but a single participant, who had atrial fibrillation. At right-sided cardiac catheterization, the median pulmonary artery wedge pressure was 30 mm Hg (IQR 24 to 32), the median mean pulmonary artery pressure was 39.5 mm Hg (IQR 28.5 to 46), the median cardiac output was 3.25 L/min (IQR 2.9 to 3.7), and the median SV was 41 ml (IQR 35 to 57). At RT3DE, the median EDV was 218 ml (IQR 160 to 281), the median ESV was 191 ml (IQR 119 to 220), the median SV was 36 ml (IQR 29 to 50), and the median LV ejection fraction was 16% (IQR 13% to 23%). Intraobserver average percentage differences were 2.1% (SE 0.9%) for EDV, 1.5% (SE 1.4%) for ESV, and 2.3% (SE 4.2%) for SV (n П 14). Interobserver average percentage differences were 2.9% (SE 2.9%) for EDV, 4.8% (SE 4.1%) for ESV, and 1.6% (SE 6.8%) for SV (n П 10). In total, using a 16segment model of the left ventricle, 90.2% of the segments (202 of 224) were contained within the data set at end-diastole. Of the missing segments, 15 of 22 were apical segments, with the most commonly missed segment being the apical inferior segment (inadequately imaged in 9 patients). The median number of segments contained within the pyramid was 14.5 per patient (IQR 13.25 to 15.75). There was a strong correlation between SV by RT3DE and SV by thermodilution (r П 0.95). On the Bland-Altman plot (Figure 1), SV by RT3DE underestimated SV by thermodilution on average by 7.5 ml (95% confidence interval 4 to 19). Expressed as a percentage of the mean value, the average difference was 17% (ПЄ5% to 39%). When patients with significant valvular regurgitation were excluded from the analysis, a marginally tighter correlation between SV by RT3DE and SV by thermodilution (r П 0.96, n П 10) emerged. However, the average underestimate on the Bland-Altman plot of 9 ml (95% confidence interval 2 to 20) was greater than that seen in the entire cohort. Expressed as a percentage of the mean value, the average difference in this subgroup was 21% (ПЄ1% to 43%). ••• To date, 3-dimensional echocardiography has largely implied the reconstruction of 3-dimensional data sets from multiple 2-dimensional planes. The acquisition of such data sets is cumbersome, time consuming, and prone to movement and breathing artifacts.3 Furthermore, such artifacts may not be apparent at the time of acquisition, because off-line reconstruction is required before a data set can be viewed. This impracticality, rather than concerns over accuracy, has limited the clinical application of 3-dimensional echocardiography. The assessment of LV function derived from such reconstructed data sets represents a clear advance on 2-dimensional imaging and has been extensively validated against different gold standards.5–10 In the development of 3-dimensional echocardiography, attention has more recently focused on developing transducer technology. This has culminated in the introduction of so-called matrix arrays, which feature a third dimension to the scanning array, allowing true real-time 3-dimensional scanning. Initial research work with a 256-element “sparse-array” matrix has established the clinical potential of this method.1,2,11–13 However, limitations, as set out by Sugeng et al4 in their review of real-time imaging, such as poor image quality, a relatively low frame rate, and the inability to produce 3-dimensional images on-line, have precluded its routine use. The recent release of a commercial system for RT3DE, however, offers the potential to bring 3-dimensional echocardiography into the clinical mainstream. Now, 3-dimensional images of the left ventricle can be acquired and rendered BRIEF REPORTS 309 on-line. In this study, we present what we believe to be the first study examining a commercially available RT3DE system in a clinically relevant patient group. We show that images suitable for analysis were produced in 14 of 17 unselected patients (83%) with severe heart failure. We demonstrate excellent reproducibility; in these 14 patients, inter- and intraobserver mean absolute errors were small (ПЅ5%) in a blinded analysis. In addition, we demonstrate a close correlation between RT3DE and our clinical reference method (r П 0.95). Excluding participants with significant left-sided regurgitation resulted in a marginal improvement in the observed correlation. However, overall, a difference between the methods emerged, with RT3DE appearing to underestimate SV as assessed by thermodilution. We chose thermodilution as a reference method, because at our institution it is the method of choice for assessing cardiac function in the setting of potential cardiac transplantation, and we wished to make a clinically based comparison. Our choice of thermodilution as a reference method, however, is a potential limitation to this study. There is evidence that thermodilution tends to overestimate SV by approximately 20%.14 This overestimation may be particularly evident in severe heart failure.15 Two previous studies reported SV estimation by 3-dimensional echocardiographic techniques in comparison with thermodilution in humans. The 2 studies used a transesophageal approach and off-line reconstruction and analysis. In the 2 studies, a similar difference between methods as found in our study was noted, with 3-dimensional echocardiography seeming to underestimate SV by 8 ml in patients with coronary artery disease (n П 12) and by 6.4 ml in an intensive care setting (n П 20).9,10 Taking all 3 studies together, a remarkably consistent difference is seen between 3-dimensional echocardiography and thermodilution, with the echocardiographic method appearing to underestimate the invasively derived result by approximately 20%. Although we do not make a comparison with a gold standard reference method, it is notable that our study is consistent with previous transesophageal studies and that the reported percentage difference in all three 3-dimen- 310 THE AMERICAN JOURNAL OF CARDIOLOGYа®ѕ VOL. 95 sional echocardiographic studies approximates the known overestimation of thermodilution. 1. Qin JX, Jones M, Shiota T, Greenberg NL, Tsujino H, Firstenberg MS, Gupta PC, Zetts AD, Xu Y, Ping Sun J, et al. Validation of real-time three-dimensional echocardiography for quantifying left ventricular volumes in the presence of a left ventricular aneurysm: in vitro and in vivo studies. J Am Coll Cardiol 2000;36: 900 –907. 2. Lee D, Fuisz AR, Fan PH, Hsu TL, Liu CP, Chiang HT. Real-time 3-dimensional echocardiographic evaluation of left ventricular volume: correlation with magnetic resonance imaging—a validation study. J Am Soc Echocardiogr 2001;14:1001–1009. 3. Siu SC, Rivera JM, Handschumacher MD, Weyman AE, Levine RA, Picard MH. Three-dimensional echocardiography: the influence of number of component images on accuracy of left ventricular quantification. J Am Soc Echocardiogr 1996;9:147–155. 4. Sugeng L, Weinert L, Lang RM. Left ventricular assessment using real time three dimensional echocardiography. Heart 2003;89(suppl):iii29 –iii36. 5. Siu SC, Rivera JM, Guerrero JL, Handschumacher MD, Lethor JP, Weyman AE, Levine RA, Picard MH. Three-dimensional echocardiography in vivo validation for left ventricular volume and function. Circulation 1993;88:1715–1723. 6. Siu SC, Levine RA, Rivera JM, Xie SW, Lethor JP, Handschumacher MD, Weyman AE, Picard MH. Three-dimensional echocardiography improves noninvasive assessment of left ventricular volume and performance. Am Heart J 1995;130:812– 822. 7. Jiang L, Vazquez de Prada JA, Handschumacher MD, Vuille C, Guererro JL, Picard MH, Joziatis JT, Fallon JT, Weyman AE, Levine RA. Quantitative three-dimensional reconstruction of aneurysmal left ventricles. In vitro and in vivo validation. Circulation 1995;91:222–230. 8. Buck T, Hunold P, Wentz KU, Tkalec W, Nesser HJ, Erbel R. Tomographic three-dimensional echocardiographic determination of chamber size and systolic function in patients with left ventricular aneurysm. Circulation 1997;96:4286 – 4297. 9. Kuhl HP, Franke A, Janssens U, Merx M, Graf J, Krebs W, Reul H, Rau G, Hoffmann R, Klues HG, et al. Three-dimensional echocardiographic determination of left ventricular volumes and function by multiplane transesophageal transducer: dynamic in vitro validation and in vivo comparison with angiography and thermodilution. J Am Soc Echocardiogr 1998;11:1113–1124. 10. Muller S, Bartel T, Baumann G, Erbel R. Preliminary report: evaluation of three-dimensional echocardiographic volumetry by simultaneous thermal dilution in coronary heart disease. Cardiology 1996;87:552–559. 11. Shiota T, McCarthy PM, White RD, Qin JX, Greenberg NL, Flamm SD, Wong J, Thomas JD. Initial clinical experience of real-time three-dimensional echocardiography in patients with ischaemic and idiopathic dilated cardiomyopathy. Am J Cardiol 1999;84:1068 –1073. 12. Schmidt MA, Freidlin RZ, Ohazama CJ, Jones M, Laurienzo JM, Brenneman CL, Norman JE, von Ramm OT, Panza JA. Anatomic validation of a novel method for left ventricular volume and mass measurements with use of real-time 3-dimensional echocardiography. J Am Soc Echocardiogr 2001;14:1–10. 13. Mondelli JA, Di Luzio S, Nagaraj A, Kane BJ, Smulevitz B, Nagaraj AV, Greene R, McPherson DD, Rigolin VH. The validation of volumetric real-time 3-dimensional echocardiography for the determination of left ventricular function. J Am Soc Echocardiogr 2001;14:994 –1000. 14. Russell AE, Smith SA, West MJ, Aylward PE, McRitchie RJ, Hassam RM, Minson RB, Wing LM, Chalmers JP. Automated non-invasive measurement of cardiac output by the carbon dioxide rebreathing method: comparison with dye dilution and thermodilution. Br Heart J 1990;63:195–199. 15. Espersen K, Jensen EW, Rosenborg D, Thomsen JK, Eliasen K, Olsen NV, Kanstrup IL. Comparison of cardiac output techniques: thermodilution, Doppler, CO2 rebreathing and the direct Fick method. Acta Anaesth Scand 1995;39:245–251. JANUARY 15, 2005
© Copyright 2024 Paperzz
