1905
Determinants of 2-Year Outcome After
Coronary Angioplasty in Patients With
Multivessel Disease on the Basis of
Comprehensive Preprocedural Evaluation
Implications for Patient Selection
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
Stephen G. Ellis, MD; Michael J. Cowley, MD; Germano DiSciascio, MD;
Ubeydullah Deligonul, MD; Eric J. Topol, MD; Thomas M. Bulle, MD;
Michel G. Vandormael, MD; and the Multivessel Angioplasty Prognosis Study Group*
Background. To assess the likelihood of intermediate-term event-free survival (freedom from
death, coronary artery bypass surgery, and myocardial infarction) in patients with multivessel
coronary disease undergoing coronary angioplasty, 350 consecutive patients from four clinical
sites were carefully evaluated and followed for 22±10 months.
Methods and Results. Eight clinical variables were evaluated at the clinical sites, and 23
angiographic variables describing the number, morphology, and topography of coronary
stenoses were evaluated at core angiographic laboratory. Most patients had Canadian
Cardiovascular Society class III or IV angina (72%), two-vessel coronary disease (68%), and
well-preserved left ventricular function (mean ejection fraction, 58+12%; range, 18-85%).
Follow-up was complete in 99%o of patients. At 2 years, event-free survival was 72%, overall
survival was 96%, freedom from bypass surgery was 82%, and freedom from nonfatal
myocardial infarction without surgery was 96%o. Sequential Cox proportional hazards regression analyses allowing stepwise entry of variables prospectively coded as simple, as of
intermediate complexity, or as complex found event-free survival to be independently predicted
by low Canadian Cardiovascular Society angina class, no diabetes, no proximal left anterior
descending stenoses, and the sum of stenosis simplified risk-territory scores of 15 or less. In the
absence of class IV angina and these risk factors, 2-year event-free survival was 87% and overall
survival was 100%. In the presence of two or more of these risk factors, event-free survival was
less than 50%.
Conclusions. Recognition of risk factors for poor long-term outcome in this setting may
improve clinical decision making and provide a framework on which to base meaningful
subgroup analyses in randomized trials assessing the efficacy of coronary angioplasty.
(Circulation 1991;83:1905-1914)
O-
the basis of multiple studies reporting that
percutaneous transluminal coronary angioplasty (PTCA) is both feasible and relatively
safe in patients with multivessel coronary disease,'-5
this form of therapy has now been advocated as an
alternate form of revascularization to coronary artery
bypass surgery, which has been the standard treatment
for many such patients.6 Initial reports of follow-up
after PTCA have shown primary procedural success in
83-93%, relief from angina pectoris in 81-92%, need
for late bypass surgery in 3-12%, and late mortality in
0-5% of patients after 7-37 months.1-5,7-10 However,
From the Division of Cardiology, Department of Internal Medicine, University of Michigan Medical Center (S.G.E., E.J.T.),
Ann Arbor, Mich.; St. Louis University Medical Center (U.D.,
M.G.V.), St. Louis, Mo.; Medical College of Virginia (M.J.C.,
G.DiS.), Richmond, Va.; and University of Alabama (T.M.B.),
Birmingham, Ala.
Supported by National Institutes of Health grant HL-38529-03
and a grant from Medtronic Inc., Minneapolis, Minn.
Address for reprints: Stephen G. Ellis, MD, Department of
Cardiology, 1500 East Medical Center Drive, UH B1F245, Ann
Arbor, MI 48109-0022.
*A listing of the Principal Investigators and Coinvestigators is
presented in the "Appendix."
Received April 5, 1990; revision accepted January 22, 1991.
n
1906
Circulation Vol 83, No 6 June 1991
careful subgroup analysis to establish criteria by which
to estimate the likelihood of long-term success in an
individual patient, given the diverse clinical syndromes,
natural history, coronary anatomy, and left ventricular
function of patients with multivessel disease, is not
available.
Therefore, to aid in choosing appropriate therapy,
and provide a rational basis for subgroup analysis in
ongoing or anticipated randomized trials, we prospectively evaluated 31 clinical and angiographic variables
in 350 consecutive patients with multivessel disease
undergoing coronary angioplasty to determine how
intermediate-term outcome might best be predicted.
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
Methods
Patient Population
Cineangiograms from 100 consecutive patients
with stable or unstable angina and multivessel coronary disease undergoing attempted PTCA on or after
January 1, 1986, were requested from each of the
four participating institutions. Patients with prior
coronary bypass surgery were excluded; only patients
likely to have 1-year or longer follow-up were included. Such patients reflected those accepted for
coronary angioplasty at each of the clinical sites in
1986 and 1987. Because of a change of study personnel at one clinical center, only 50 qualifying patients
were enrolled from that site; thus, results from 350
patients were analyzed.
Angioplasty Procedure
The techniques of coronary angioplasty used and
initial in-hospital follow-up have been described.'
Angiographic Analysis
One of two experienced angiographers at the
angiographic core laboratory without knowledge of
clinical outcome reviewed the diagnostic and procedural angiograms to determine suitability for study
entry and code for 18 stenosis and 13 patient variables. All quantitative measurements were made using
hand-held calipers, and percent diameter stenosis
was obtained from the mean of measurements in
orthogonal projections at end diastole, when possible. Thirty patients were excluded from the study
because of nonvisualization of a contralateral coronary artery (n=17), presence of one-vessel disease
(n=7), inadequate stenosis visualization (n=4), and
an angioplasty not performed during the study period
(n =2). Additional patients were then requested until
each institution's quota was met.
The angiographic definitions used by the Multivessel Angioplasty Prognosis Study (MAPS) Group have
been described previously.'1 The following definitions are, however, particularly germane to the results and interpretation of this analysis.
Stenosis-Related Variables
Primary target stenosis risk score. A single primary
target stenosis was identified on the basis of severity
TABLE 1. Lesion-Specific Characteristics of Type A, B, and C
Lesions: American College of Cardiology/American Heart
Association Task Force Stenosis Classification System
Type A lesions (high success, 85%: low risk)
Discrete (less than 10-mm length)
Concentric
Readily accessible
Nonangulated segment of less than 45°
Smooth contour
Little or no calcification
Less than totally occlusive
Not ostial in location
No major branch involvement
Absence of thrombus
Type B lesions (moderate success, 60-85%; moderate risk)
Tubular (10-20-mm length)
Eccentric
Moderate tortuosity of proximal segment
Moderate angulated segment of 46-89°
Irregular contour
Moderate-to-heavy calcification
Total occlusions of less than 3 months old
Ostial in location
Bifurcation lesions requiring double guide wires
Some thrombus present
Type C lesions (low success, 60%; high risk)
Diffuse (more than 2 cm length)
Excessive tortuosity of proximal segment
Extremely angulated segments of more than 90°
Total occlusion of more than 3 months old
Inability to protect major side branches
Degenerated vein grafts with friable lesions
Class B stenoses were divided into Bi (one adverse characteristic) and B2 (two or more adverse characteristics).
Reproduced with permission (American College of Cardiology:
JAm Coll Cardiol 1988;12:538).
and morphological characteristics of the stenoses,1"
their jeopardized territories, and the presumed viability of the myocardium subserved. The American
College of Cardiology/American Heart Association
(ACC/AHA) stenosis classification (see Table 1) was
prospectively modified to subdivide B stenoses into
Bi (one adverse characteristic) and B2 (two or more
adverse characteristics) stenoses on the basis of prior
work, suggesting a cumulative importance of multiple
adverse lesion characteristics.12 The risk scores for A,
Bl, B2, and C stenoses were assigned 1, 2, 3, and 4
points, respectively.
Jeopardized teritory. The size of all terminal (e.g.,
distal left anterior descending, diagonal branch)
branches of the coronary tree were scored as small,
moderate, or large depending on vascular distribution.
Branches scored as small were assigned one point,
branches scored as moderate were assigned two
points, and branches scored as large were assigned
three points. Branches supplied by collateral vessels
were classified as jeopardized if the vessel supplying
Ellis et al Long-term Outcome After Multivessel PTCA
1907
Left Main Coronary Artery
Left Circumflex
Right C
Left Anterior
Descending
50%, short,
bifurcation
\
90%, irregular,
tubular
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
FIGURE 1. Stylized coronary artery diagram with illustrative coronary stenosis
scores.
Left ventricular function: mild anterior hypokinesis
Sum of coronary territories = 15
Individual Stenosis Scores
Left anterior descending
Right
RISk
Risk
3
2
Simplified
Territory
2
1
Territory
Contractility
2
3
5/15x 100 =33
3/15 x 100 = 20
Overall Scores
Primary Target Stenosis Risk Score:
Sum of Stenoses Risk Score:
Primary Target Stenosis Risk-Territory Score:
Sum of Stenoses Simplified Risk-Territory Score:
Sum of Stenoses Risk-Territory Score:
Primary Target Stenosis Risk-Territory-Contractility Score:
Sum of Stenoses Risk-Territory-Contractility Score:
3
3+2 =5
3 x 33 = 99
(3 x 2) + (2 x 1) = 8
(3 x 33) * (2 x 20) = 139
3 x 33 x 2 = 198
(3 x 33 x 2) + (2 x 20 x 3) = 318
collaterals had a stenosis that was dilated. The jeopardized territory equaled the sum of branches distal to
the stenosis dilated divided by the sum of all branches
of the coronary tree (see Figure 1).
Multivessel disease. Multivessel coronary disease
was defined as a 50% or greater diameter stenosis in
two or three of the three epicardial vessels (or their
surgically bypassable branches). When the right coronary artery was nondominant and failed to supply
any left ventricular myocardium, the first two moderate- or large-sized obtuse marginal branches of the
circumflex were thought to supply one vascular territory, and the distal obtuse marginal branches and
the posterior descending coronary artery were
thought to supply a different vascular territory.
Segmental contractility. Segmental left ventricular
myocardial contractility was scored normal (three
points), mildly hypokinetic (two points), severely hypokinetic (one point), and akinetic or dyskinetic (no points).
Patient-Related Variables
Primary target stenosis risk-teritory score. The primary target stenosis risk-territory score was defined
the product of the modified ACC/AHA score for
the primary target stenosis multiplied by the jeopardized territory.
Sum of all stenoses simplified risk-teritory scores.
The simplified risk-territory score was defined as the
modified ACC/AHA score multiplied by 2 if the
stenosis was located in the proximal or mid left
anterior descending, proximal circumflex, or proximal or mid right coronary artery and multiplied by 1
for all other vessel sites. The sum of all stenoses
simplified risk-territory scores was the sum of the
simplified score for all stenoses of 50% or greater.
Primary target stenosis rsk-territory-contractility score.
The primary target stenosis risk-territory-contractility
score was the product of the modified ACC/AHA score
of the primary target stenosis multiplied by the jeopardized territory multiplied by segnental contractility.
Sum of all stenosis risk-territory-contractility scores:
This score was defined as the sum of all individual
stenoses' risk-territory-contractile scores.
Clinical variables. The clinical variables assessed
were age, gender, diabetes, hypertension, hyperchoas
1908
Circulation Vol 83, No 6 June 1991
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
lesterolemia, current smoking status, Canadian Cardiovascular Society angina class, and prior myocardial infarction.
Follow-up. Follow-up data were provided by the
individual clinical sites after patient and/or physician
contact by telephone or mail and supplemented by
review of hospital records when necessary. Follow-up
at each site was performed by a cardiovascular research nurse or physician.
TABLE 2. Patient Characteristics
Clinical characteristics (n=350)
Age (mean and range, yr)
Male gender (%)
Prior myocardial infarction (%)
Canadian Cardiovascular
Society angina class
Current smoking (%)
Diabetes mellitus (%)
Definitions of Outcomes
Procedural success. Procedural success was defined as
a reduction in diameter stenosis in one or more
stenoses to less than 50% stenosis associated with
clinical improvement and no major ischemic complications (death, bypass surgery, or myocardial infarction).
Intermediate-term event-free survival. Intermediateterm event-free survival was defined as freedom from
death, bypass surgery, and myocardial infarction at
the time of latest follow-up.
Statistical Analysis
All data were entered into the MAPS data bank at
the data coordinating center. Data are expressed as
mean+ 1 SD unless otherwise indicated. Cox proportional hazards regression analyses were used to determine the clinical and angiographic correlates of
event-free survival, freedom from death, freedom
from bypass surgery, and freedom from myocardial
infarction. For each of these analyses, patients were
censored at the time of their death from obvious
noncardiac causes. If the cause of death could not be
determined, it was considered to be cardiac in origin.
For each of these analyses, clinical and angiographic
variables were prospectively divided into those that
were simple, intermediate, or complex to determine.
Preliminary Cox analyses were performed within
each of these three groups of variables. A final sequential Cox analysis was then performed, entering
first the significant variables from the simple group
and then the variables from the intermediate and
complex variable groups, but only if they provided
significant additional information. This method of
analysis was chosen to evaluate the usefulness of the
variables in the clinical setting. The variables defined
as simple were age, Canadian Cardiovascular Society
angina class, diabetes, gender, hypercholesterolemia,
hypertension, number of diseased vessels, number of
stenoses of 50% or greater, prior myocardial infarction, proximal left anterior descending coronary artery
stenosis of 50% or greater, and smoking. The variables
defined as intermediate in complexity were the primary target stenosis risk score, the sum of all stenosis
risk scores, and the sum of all stenosis simplified riskterritory scores. The variables defined as complex
were the primary target stenosis risk-territory score,
the sum of all stenosis risk-territory scores, the primary target stenosis risk-territory-contractility score,
and the sum of all stenosis risk-territory-contractility
scores. Analyses were performed using SAS software
from SAS Institute Inc., Cary, N.C.
Systemic hypertension (%)
Hypercholesterolemia (%)
Angiographic characteristics
Diseased vessels (n)
Stenoses .50% (n)
Class A stenoses (n)
Class Bi stenoses (n)
Class B2 stenoses (n)
Class C stenoses (n)
Chronic total occlusions (n)
Left ventricular ejection fraction
(mean and range, %)
58.4+11.1 (30-87)
71.4
48.7
31-+-1.1
50.3
19.1
51.5
34.0
2.3+0.5
3.1± 1.1
0.8±0.7
1.1+0.8
0.8±0.6
0.5 ±0.6
0.2±0.4
57.6±11.6 (18-85)
Results
Patient Characteristics
Cineangiograms from 350 patients (1,100 stenoses)
were analyzed. Characteristics of the study patients
are enumerated in Table 2. Of importance, 72% of
the patients had severe angina (Canadian Cardiovascular Society class III or IV), the majority of patients
(68%) had two-vessel coronary disease, left ventricular function was usually well preserved (ejection
fraction, 58+12%), and 1.9+1.0 stenoses per patient
were dilated. The stenoses dilated were of mixed
complexity, with 29% having characteristics suggestive of expected high success and low complication
rates (ACC/AHA type A), 61% having characteristics of expected moderate success and complication
rates (ACC/AHA type B), and 10% having characteristics suggestive of expected low success and high
complication rates (ACC/AHA type C).
Overall Clinical Outcome
Two hundred ninety-nine of 350 patients (85.4%)
had at least one stenosis successfully dilated, no
complications, and an improvement in functional
status. Major procedural complications occurred in
30 of 350 patients (8.6%) (death, four of 350, 1.1%;
nonfatal emergency bypass surgery, 20 of 350, 5.7%;
nonfatal myocardial infarction without bypass surgery, six of 350, 1.7%).
Mean duration of follow-up for patients with eventfree survival was 22±10 months. Three patients (0.9%)
were lost to follow-up. The event-free survival by
life-table analysis was 72.3% at 24 months. The incidence rates of death, bypass surgery, and myocardial
infarction without bypass surgery were 4.5%, 17.8%,
and 3.8%, respectively (see Figure 2). An additional
Ellis et al Long-term Outcome After Multivessel PTCA
1909
20
1
r-
CUMUlatiVe 15Events (%)
-
-
.I..
.
---
10
~~~Bypass
~~Surgery
FIGURE 2. Plot of actuarial late out350 patients.
1-,~
~ ~ ~ ~ ~ ~~et
ycada
1~~~~~~~~~~~Ifrto
come in the
5.(1Av
i
0
2
6
4
8
10
12
16
14
18
20
22
24
Months
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
16.2% of patients had at least one additional PTCA
during follow-up at a mean 9.8±+8.0 months.
Correlates of Event-Free Survival
Graphic and tabular representations of the correlates of event-free survival are shown in Figure 3 and
Table 3, respectively. Of the simple variables, proximal left anterior descending coronary artery stenosis
(p=0.O06), diabetes (p=0.0O8), and Canadian Cardiovascular Society angina class (p =0.08, but multivariate p =0.04) were significantly correlated with
event-free survival. Of the intermediate variables, only
the sum of the simplified risk-territory scores was
significantly correlated with outcome (p =0.001). Of
the complex variables, only the sum of all stenosis risk-
territory
scores
provided independently prognostic
information (p =0.003). In the sequential Cox analysis, only diabetes, angina class, sum of simplified risk
scores, and proximal left anterior descending coronary
artery stenosis provided independent prognostic information. When all variables had equal access to the
model, the sum of all stenosis risk-territory scores was
the only variable to enter the regression equation. In
the absence of diabetes, proximal left anterior descending coronary artery stenosis, rest pain, and sum
of simplified risk territory score of 16 or more, patients
(n=70) had a 2 -year event-free survival rate of 86.9%.
Two-year event-free survival rates for patients with
two or more of these adverse characteristics ranged
from 39% to 58%.
1.0
1.0 -
Event-Free 0.9
Suirvival
1
0
No Proximal
LAD
Disease
Event-Free 0.8.
Survival
0.8
0.7-
....Proximal LAD
Disease
L
-
0.5J
~~~~~~~~~~~Univariate p 0.006
11
=
(i
4
6
8
12
10
14
16
18
22
20
Univariate p = 0.08
:1~
,
2
24
0
2
6,
4
8
10
Months
14
12
16
18
20
2,2
24
Months
1.01
1.0
Event-Free 0.9
Survival
0.9.
ibee
Event-Free
Survival 0.8,
Diabetes
0.7'
F1I -1o
0.81
0.7'
0.6
~~~~~~~~~~~1-1
L-1
~
~
l_------
0.6
0.5i1
~
~
~
~
~
--
0.5J
Univariate p
Univariate p - 0.001
0.008
n
n
0
2
4
6
10
8
12
14
16
18
20
22
0
24
2
4
6
8
10
Months
Plots
Proximal
risk-territory
i
0.6-
0.5-
panel:
~~~~~~~~~~ill
a -~~~~~~~~~~1zzz.
I---
0.7
0.6-
FIGURE 3.
0~~~~~
11
0.9
1
L
of correlates of event-fr-ee
left
scores.
anterior
survival,
descending coronary
Bottom left
panel:
12
14
16
18
20
22
Months
Diabetes.
lop
right panel:
artery (LAD)
Canadian Cardiovascular
stenoses.
Bottom
Society angina class. Top
right panel:
Sum
of
stenosis
left
simplified
Circulation Vol 83, No 6 June 1991
1910
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
TABLE 3. Correlates of Event-Free Survival (Sequential Analysis)
Partial
P3 SEM X2
Variable
P
0.559 0.258 4.68 0.031
Diabetes mellitus
0.240 0.119 4.04 0.044
Angina class
Sum of simplified risk
0.045 0.022 4.33 0.037
territory scores
Proximal left anterior descending
0.358 0.202 2.87 0.085
coronary artery stenosis
TABLE 4. Correlates of Survival
Correlates of Internediate-Term Survival
The correlates of intermediate-term survival are
shown graphically in Figure 4 and in tabular form in
Table 4. Of the simple variables, left ventricular
ejection fraction (p <0.0001), diabetes mellitus
(p=0.002), and proximal left anterior descending
coronary artery stenosis (p=0.01) were independently predictive. Of the intermediate variables, only
the sum of all stenosis simplified risk-territory scores
was correlated with survival (p=0.01). Of the complex variables, the sum of all stenosis risk-territory
scores (p=0.009) and the sum of all stenosis risk'
territory-contractility scores (p=0.001) were independently predictive of survival. In the sequential
Cox model, left ventricular ejection fraction, diabetes
mellitus, and presence of a proximal left anterior
descending coronary artery stenosis were the only
variables independently correlated with outcome.
1.0-
-I,
,
SEM
X2
p value
to add
0.055
-1.288
0.018
0.536
9.53
5.77
0.002
0.016
-1.183
0.552
4.59
0.032
Correlates of Myocardial Infarction
The correlates of myocardial infarction are shown
graphically in Figure 4 and in tabular form in Table 5.
Of the simple variables, only diabetes mellitus
(p=0.006) was correlated with the risk of infarction.
No Dabetes
EF 40 59%
-
Survival .
i~1
0.9-
/
1.0
EF.60%
11
Suvial
Partial
Variable
Left ventricular ejection
fraction
No diabetes mellitus
No proximal left anterior
descending coronary
artery stenosis
0.8
0.8
iL..
-1
EF 40%
0.7
0.7
11
Univariate p 0.0001
<
m.
u,
0
.
.
2
4
-
6
0
.
10
8
12
14
20
18
16
22
Univariate p = 0.002
0i
2
6
4
10
8
24
Months
Survival
12
16
14
18
20
22
24
Months
1.0
No Proximail
1 .C
aDLAD Disease
0.'
Proximal LAD
Disease
Freedom 0.9
from CABG
:
L~~~~~~~~~
|~~~~~~~~~~
'
*-----
"
0.81
Class
BA
L_______-------I
Class B2
------------
0.8
Class C
0.7
Univariate p
0.7
Univariate p = 0.01
.
0
2
0
4
6
8
0o 0
2
4
6
8
.1.
10
12
.
14
16
18
20
=
0.003
22
24
Months
*
10
12
14
16
18
20
22
24
Months
1.00
No Diabets
Freedom
from
0.95
Infarction
1
0.90
L
-
FIGURE 4. Plots of correlates of outcome. Top left panel: left
ventricular ejection fraction (EF) and overall survival. Top right
panel: Diabetes and overall survival. Middle left panel:proximal
left anterior descending coronary artery (LAD) stenoses and
survival. Middle right panel: primary target stenosis risk score
and bypass surgery (CABG). Bottom left panel: diabetes and
myocardial infarction.
0.85
Univariate p - 0.006
2
0
4
6
8
10
12 14
Months
16
18
20
22 24
Ellis et al Long-term Outcome After Multivessel PTCA
TABLE 5. Correlates of Infarct-Free Survival
(Sequential Analyses)
Variable
No diabetes mellitus
p3
-1.666
SEM
0.606
X2
7.56
p
0.006
No intermediate or complex variables were predictive of infarction. Therefore, in the sequential Cox
analysis, only diabetes mellitus was predictive of the
risk of myocardial infarction.
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
Correlates of Bypass Surgery
The correlates of bypass surgery are shown graphically in Figure 4 and in tabular form in Table 6. No
simple variables were predictive of the risk of bypass
surgery. Of the intermediate variables, the primary
target stenosis risk score (p=0.003) was correlated
with outcome. Of the complex variables, only the risk
territory score of the primary target stenosis was
correlated with outcome. In the sequential Cox analysis, only the primary target stenosis risk score was
independently predictive of outcome.
Correlates of Late PTCA
In multivariate testing, only patient age (coefficient=-0.005, p=0.01) and baseline left ventricular
ejection fraction (coefficient=0.004, p =0.05) were correlated with late PTCA. Neither the presence of disease in the left anterior descending coronary artery nor
the extent of the disease was predictive of late PTCA.
Late Functional Outcome
At the time of last follow-up, of 144 patients with
positive stress tests before PTCA, 52 (36.1%) had
positive testing during follow-up and had undergone
further revascularization, 17 (11.8%) had positive
tests but were treated medically, 60 (41.7%) had
negative tests, and 15 (10.4%) had nondiagnostic
tests or no stress test results available. Of 194 patients with nondiagnostic or no stress testing before
PTCA (175 patients had class IV angina before
PTCA), 65 (33.5%) had positive testing and had
undergone further revascularization, 14 (7.2%) had
positive tests but were treated medically, 86 (44.3%)
had negative tests, and 29 (14.9%) had nondiagnostic
tests or no results available. Of 12 patients with
negative stress tests within 30 days of PTCA, no
patient had late revascularization, three patients
(25%) had positive stress tests but were treated
medically, six patients (50%) had negative stress
tests, and three patients (25%) had nondiagnostic or
no stress test results available.
TABLE 6. Correlates of Bypass-Free Survival
(Sequential Analyses)
Variable
Primary target risk
score
0.440
SEM
X2
p
0.148
8.81
0.003
1911
Discussion
Despite increased use of coronary angioplasty to
treat patients with multivessel coronary disease and
several studies reporting a 64-91% event-free
(death, bypass surgery, or myocardial infarction)
7-37-month survival in carefully selected patients,457,1013 no study has addressed the issue of
defining criteria by which to assess the likelihood of
intermediate- or long-term event-free survival in a
given patient.
The present study, in which 350 consecutive and
carefully characterized patients with multivessel disease undergoing coronary angioplasty were followed
for nearly 2 years, was designed to identify the
predictors of and define the nature of late outcome
after PTCA in this patient population.
Four conclusions are noteworthy. First, of the
27.6% of patients experiencing a major adverse outcome within 2 years of the index angioplasty, approximately two thirds had bypass surgery and survived
without myocardial infarction, whereas the rates of
death and myocardial infarction within 2 years were
low- 4.5% and 3.8%, respectively.
Second, there was a wide patient-to-patient variation in event-free survival, overall survival, and freedom from bypass surgery and myocardial infarction
that were highly correlated with simple and readily
obtainable clinical and angiographic characteristics
(see Figures 3 and 4). Event-free survival was independently predicted by three simple (Canadian Cardiovascular Society angina class, diabetes, and presence of a significant stenosis in a proximal left
anterior descending coronary artery) and one intermediately complex (sum of all stenosis simplified risk
territory scores) variables. Patients with silent or
class I angina had an 88% event-free 2-year survival
compared with 78% for patients with class II or III
angina and 67% for patients with angina at rest
(p=0.04). This difference first became apparent at
the time of the initial procedure and continued to
increase throughout follow-up. It is probably explained in part by the higher incidence of ischemic
procedural complications reported in patients with
unstable ischemic syndromes,14,15 by the higher frequency of restenosis in patients with unstable angina,16 and by a lesser sense of need to refer patients
with minimal symptoms for bypass surgery. Patients
with proximal left anterior descending coronary artery stenoses had a much lower event-free survival
rate (56% versus 81%,p=0.006) and a poorer overall
survival rate (91.5% versus 97.5%, p=0.01) than
patients without this finding. This can probably be
explained by the 1.5-2.5-fold increase in the likelihood of restenosis after PTCA at this site compared
with other native vessel sites,16'17 by the critical
importance of this site in terms of the impact on left
ventricular function in the event of its occlusion'18
and by the effect of stenoses at this site in determining survival in patients with multivessel disease.19
Patients with diabetes had considerably worse event-
1912
Circulation Vol 83, No 6 June 1991
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
free and overall survival rates as well as an increased
likelihood of myocardial infarction compared with
their nondiabetic counterparts. Diabetes appears to
predispose to a higher rate of angioplasty-induced
complications, particularly in patients with multivessel coronary artery disease'1; may predispose to
restenosis after angioplasty20; and is related to a
generally poor outcome in patients with coronary
artery disease.21 The adverse prognostic impact of
diabetes in patients in this study is striking. Finally, as
might be expected from the results of other studies
assessing prognostic factors in patients with coronary
artery disease,22 left ventricular ejection fraction was
an important correlate of overall survival.
Third, highly complex analyses evaluating stenosis
morphology, topography, and regional left ventricular contractility in the area subserved by the stenosis
were not required to assess outcome. Although the
sum of all stenosis risk-territory scores was the best
overall predictor of event-free survival, it added no
independent prognostic value once the impact of
simpler variables was considered. Two variables of
intermediate complexity added independent prognostic information and should be considered in clinical decision making. The sum of all stenosis simplified risk-territory scores-the additive effect of the
stenosis risk score (class A, one point; class Bi, two
points; class B2, three points; class C, four points)
multiplied by 2 for proximal and mid artery stenoses
and by 1 for distal or branch stenoses for all 50% or
greater stenoses (see Figure 2) -was a significant
correlate of event-free survival even after all simpler
variables were considered. Only 54% of patients with
scores of 16 or greater were free of death, bypass
surgery, or myocardial infarction at 2 years, whereas
event-free survival was 72% for patients with scores
of 11-15 and 85% for patients with scores of 10 or
less (p=0.003). This score incorporates an assessment of the cumulative risk of ischemic angioplastyrelated complications12 as well as a rough assessment
of the extent of myocardium potentially at risk for
such complications. Similar jeopardy scores have
been shown to have great prognostic value in assessing long-term outcome in patients with coronary
artery disease22 and in assessing a risk of cardiac
death after vessel closure with PTCA.23 By assessing
all significant stenoses, it also includes a rough estimate of the overall risk of restenosis for a given
patient.20 In addition, the primary target stenosis risk
(modified ACC/AHA) score was the only variable
correlated with long-term need for bypass surgery.
The major impact of this variable was seen within the
first month of the procedure and thus reflects the
need for emergency bypass surgery and for bypass
surgery performed because of failed but uncomplicated PTCA. The patients with primary target
stenoses classified as A or Bi had a 5% early and a
12% overall need for bypass surgery, compared with
9% and 20% respectively for class B2 stenoses, and
16% and 28%, respectively for class C stenoses
(p=O0.003).
Fourth, the often-used descriptor of coronary artery disease severity- "number of diseased vessels"-although a significant univariate predictor of
outcome, was not as important as angina class, diabetes, and proximal left anterior descending coronary
artery disease and did not enter as a significant
independent correlate of outcome in the Cox model.
Implications
Ideal multivessel disease patients for coronary angioplasty. Based on the findings of the present study, the
patient without diabetes, proximal left anterior descending coronary artery stenoses, or rest pain; with
a sum of stenosis simplified risk-territory score of 15
or less; and in need of revascularization would be
considered ideal for referral for PTCA. On the basis
of the findings of the present study, their expected
2-year event-free survival rate would be 87%, with an
overall survival at 2 years of 100% (n=70).
Poor candidates for coronary angioplasty. Patients
with two or more of these risk factors, particularly if
their left ventricular ejection fraction was less than
40%, should be referred for PTCA only under unusual circumstances because of an expected 2-year
event-free survival rate of less than 50% and a
mortality rate of more than 25%. Patients with only
one risk factor should be considered individually.
However, given the adverse impact on both eventfree and overall survival rates of proximal left anterior descending coronary artery stenoses and the
excellent expected outcome after internal mammary
artery grafting in that setting,24 the finding of important stenoses in that coronary segment in patients
with multivessel disease should be considered a relative contraindication to PTCA.
Limitations. In considering the results, certain limitations of the present study should be kept in mind.
First, the majority of these patients had two-vessel
coronary disease and well-preserved left ventricular
function and do not reflect the distribution of coronary artery involvement and functional impairment
of all patients in multivessel coronary disease.25 The
importance of left ventricular function and the extent
of the disease in the determination of late outcome
defined herein suggest that long-term rates of adverse events would be higher than those reported if
the study population was more representative of all
patients with multivessel disease. Second, to a certain
extent, assessment of coronary artery morphology is
subjective,26 and the use of these data to assist in
clinical decision making requires strict attention to
the details of the definitions used. Third, this is not a
randomized comparison of treatments, and definitive
statements regarding the efficacy of coronary angioplasty versus bypass surgery or other forms of therapy
must await carefully randomized comparisons of prospectively defined subgroups of patients in this heterogeneous population. Fourth, improved results
with angioplasty may be seen with lower-profile balloons or the concomitant use of intracoronary
stents,27 laser balloon angioplasty,28 atherectomy,29
Ellis et al Long-term Outcome After Multivessel PTCA
and other evolving adjunctive technologies. Finally,
this assessment does not include the possible impact
of completeness of revascularization on outcome
because complete revascularization is dependent on
the procedural result and therefore cannot be determined with certainty before beginning the procedure.
Conclusions
The considerable patient-to-patient variation in late
outcome after coronary angioplasty for patients with
multivessel coronary disease appears to be highly correlated with readily identifiable clinical and anatomic
risk factors. Identification of such risk factors should
improve clinical decision making and provide a framework on which to base prospective subgroup evaluations in randomized trials comparing standard coronary
angioplasty with other treatment modalities in patients
with multivessel coronary disease.
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
Acknowledgments
The authors would like to express their gratitude
to Ms. Lin Nicholson and Mr. Kevin Handy for their
expert organizational assistance.
Appendix
Multivessel Angioplasty Prognosis Study Group
Medical College of Virginia, Richmond, Va. Michael J.
Cowley, MD (Principal Investigator); Germano DiSciascio, MD, George W. Vetrovec, MD, Kim M. Kelley,
RN (Coinvestigators).
St. Louis University, St. Louis, Mo. Michel G. Vandormael, MD (Principal Investigator); Ubeydullah
Deligonul, MD, Kathy Galan, RN, Sue Taussig, RN
(Coinvestigators).
University of Alabama, Birmingham. Thomas M.
Bulle, MD (Principal Investigator); Joan Anderson,
RN (Coinvestigator).
University of Michigan, Ann Arbor. Stephen G.
Ellis, MD (Principal Investigator); Eric J. Topol,
MD, Vicky Savas, MD (Coinvestigators).
Angiographic Core Laboratory. Stephen G. Ellis,
MD (Principal Investigator); Thomas M. Bulle, MD,
Darrell DeBowey, MS (Coinvestigators).
Data Coordinating Center. Stephen G. Ellis, MD
(Principal Investigator); M. Anthony Schork, PhD,
Robert Bagen, PhD (Coinvestigators).
References
1. Cowley MJ, Vetrovec GW, DiSciascio G, Lewis SA, Hirsh PD,
Wolfgang TC: Coronary angioplasty of multiple vessels: Shortterm outcome and long-term results. Circulation 1985;72:
1314-1320
2. Hartzler GO, Rutherford BD, McConahay DR, Johnson WL,
Giorgi LV: "High-risk" percutaneous transluminal coronary
angioplasty. Am J Cardiol 1988;61:33G-37G
3. Detre K, Holubkov R, Kelsey S, Cowley M, Kent K, Williams
D, Myler R, Faxon D, Holmes D, Bourassa M, Block P,
Gosselin A, Bentivoglio L, Leatherman L, Dorros G, King S,
Galichia J, Al-Bassam M, Leon M, Robertson T, Passamani E:
Percutaneous transluminal coronary angioplasty in 1985-1986
and 1977-1981. N Engl J Med 1988;318:265-270
4. Deligonul U, Vandormael MG, Kern MJ, Zelman R, Galan K,
Chaitman BR: Coronary angioplasty: A therapeutic option for
1913
symptomatic patients with two and three vessel coronary
disease. JAm Coll Cardiol 1988;11:1173-1179
5. Stertzer SH, Shaw RE, Myler RK, O'Donnel MJ: Coronary
angioplasty in a setting of multivessel disease: Current status
and future directions. Cardiol Clin 1989 (in press)
6. CASS Principal Investigators and Their Associates: Myocardial infarction and mortality in the Coronary Artery Surgery
Study (CASS) randomized trial. N Engl J Med 1984;310:
750-758
7. Dorros G, Stertzer SH, Cowley MJ, Myler RK: Complex
coronary angioplasty: Multiple coronary dilatations. Am J
Cardiol 1984;53:126C-130C
8. Hartzler GO: Percutaneous coronary angioplasty in patients
with multivessel disease, in Jang GD (ed): Angioplasty. New
York, McGraw-Hill, 1986, pp 221-336
9. Finci L, Meier B, De Bruyne B, Steffenino G, Divernois J,
Rutishauser W: Angiographic follow-up after multivessel percutaneous transluminal coronary angioplasty. Am J Cardiol
1987;60:467-470
10. Detre K, Holubkov R, Kelsey S, Bourassa M, Williams D,
Holmes D, Dorros G, Faxon D, Myler R, Kent K, Cowley M,
Cannon R, Robertson T, Coinvestigators of the National
Heart, Lung, and Blood Institute's Percutaneous Transluminal Coronary Angioplasty Registry: One-year follow-up results
of the 1985-1986 National Heart, Lung, and Blood Institute's
Percutaneous Transluminal Coronary Angioplasty Registry.
Circulation 1989;80:421-428
11. 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. Circulation 1990;82:1193-1202
12. Ellis SG, Roubin GS, King SB, Douglas JS, Weintraub WS,
Thomas RG, Cox WR: Angiographic and clinical predictors of
acute closure after native vessel coronary angioplasty. Circulation 1988;77:372-379
13. DiSciascio G, Cowley MJ, Vetrovec GW, Kelly KM, Lewis SA:
Triple vessel coronary angioplasty: Acute outcome and longterm results. JAm Coll Cardiol 1988;12:42-48
14. de Feyter PJ, Serruys PW, Wijns W, van den Brand M:
Emergency PTCA in unstable angina pectoris refractory to
optimal medical treatment. N Engl J Med 1985;313:342- 346
15. de Feyter PJ, Serruys PW, Soward A, van den Brand M, Bos
E, Hugenholtz PG: Coronary angioplasty for early postinfarction unstable angina. Circulation 1986;74:1365-1370
16. Leimgruber PP, Roubin GS, Hollman J, Cotsonis GA, Meier
B, Douglas JS, King SB, Gruentzig AR: Restenosis after
successful coronary angioplasty in patients with single-vessel
disease. Circulation 1986;73:710-717
17. Whitworth HB, Pilcher GS, Roubin GS, Gruentzig AR: Do
proximal lesions involving the original of the left anterior
descending artery have a higher restenosis rate after coronary
angioplasty? Circulation 1985;72(suppl III):III-398
18. Kumpuris AG, Quinones MA, Kanon D, Miller RR: Isolated
stenosis of left anterior descending or right coronary artery:
Relation between site of stenosis and ventricular dysfunction
and therapeutic implications. Am J Cardiol 1980;46:15-20
19. European Coronary Surgery Study Group: Long-term results
of prospective randomized study of coronary bypass surgery in
stable angina pectoris. Lancet 1982;2:1173-1180
20. Myler RK, Shaw RE, Stertzer SH, Clark DA, Fishman J,
Murphy MC: Recurrence after coronary angioplasty. Cathet
Cardiovasc Diagn 1987;13:77-86
21. Russek HI: The "natural" history of severe angina pectoris
with intensive medical therapy alone: A five-year prospective
study of 133 patients. Chest 1974;65:46-51
22. Califf RM, Phillips HR, Hindman MC, Mark DB, Lee KL,
Behar VS, Johnson RA, Pryor DB, Rosati RA, Wagner GS,
Harrell FE: Prognostic value of a coronary artery jeopardy
score. JAm Coll Cardiol 1985;5:1055-1063
23. Ellis SG, Roubin GS, King SB, Douglas JS, Shaw RE, Stertzer
SH, Myler RK: In-hospital cardiac mortality after acute closure after coronary angioplasty: Analysis of risk factors from
8,207 procedures. JAm Coll Cardiol 1988;11:211-216
1914
Circulation Vol 83, No 6 June 1991
24. Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormastic
M, Williams GW, Golding LAR, Gill CC, Taylor PC, Sheldon
WC, Proudfit WL: Influence of the internal-mammary-artery
graft on 10-year survival and other cardiac events. N Engl J
Med 1986;314:1-6
25. Chaitman BR, Bourassa MG, Davis K, Rogers WJ, Tyras DH,
Berger R, Kennedy JW, Fisher L, Judkins MP, Mock MB,
Killip T: Angiographic prevalence of high-risk coronary artery
disease in patient subsets (CASS). Circulation 1981;64:
360-367
26. Ellis S, Alderman EL, Cain K, Wright A, Bourassa M, Fisher
L, Participants of the Coronary Artery Surgery Study (CASS):
Morphology of left anterior descending coronary territory
lesions as a predictor of anterior myocardial infarction: A
CASS Registry study. JAm Coll Cardiol 1989;13:1481-1491
27. Sigwart U, Urban P, Golf S, Kaufmann U, Imbert C, Fischer
A, Kappenberg L: Emergency stenting for acute occlusion
after coronary balloon angioplasty. Circulation 1988;78:
1121-1127
28. Sinclair NI, Dear WE, Safian RD, Plokker TM, Spears JR,
LBA Study Group: Acute closure post-PTCA successfully
treated with laser balloon angioplasty. Circulation 1989;
80(suppl II):II-476
29. Robertson GC, Hinohara T, Selmon MR, Johnson DE, Simpson JB: Directional coronary atherectomy, in Eric J. Topol
(ed): Textbook of Interventional Cardiology. Philadelphia, WB
Saunders Co, 1990, pp 563-579
KEY WORDS * percutaneous transluminal coronary angioplasty
* long-term outcome * multivessel disease
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
Determinants of 2-year outcome after coronary angioplasty in patients with multivessel
disease on the basis of comprehensive preprocedural evaluation. Implications for patient
selection. The Multivessel Angioplasty Prognosis Study Group.
S G Ellis, M J Cowley, G DiSciascio, U Deligonul, E J Topol, T M Bulle and M G
Vandormael
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
Circulation. 1991;83:1905-1914
doi: 10.1161/01.CIR.83.6.1905
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1991 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/83/6/1905
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/