ASSESSING AND QUANTIFYING PLACENTAL DYSFUNCTION IN RELATION TO PREGNANCY OUTCOME IN PREGNANCIES COMPLICATED BY REDUCED FETAL MOVEMENTS A thesis submitted to the University of Manchester for the degree of PhD in the Faculty of Medical and Human Sciences 2015 Dr Lucy Elizabeth Higgins MBChB (Hons) School of Medicine TABLE OF CONTENTS LIST OF TABLES............................................................................................................. 6 LIST OF FIGURES........................................................................................................... 8 LIST OF ABBREVIATIONS........................................................................................... 10 SCIENTIFIC ABSTRACT.......................................................................... ...................... 14 LAY ABSTRACT.............................................................................................................. 15 DECLARATION............................................................................................................... 16 COPYRIGHT STATEMENT........................................................................................... 17 DEDICATION................................................................................................................... 18 ACKNOWLEDGEMENTS............................................................................................... 19 PREFACE.......................................................................................................................... 20 CONTRIBUTIONS FROM COLLABORATORS......................................................... 21 PUBLICATIONS ARISING FROM THIS WORK....................................................... 22 A NOTE ON ALTERNATIVE FORMAT THESES...................................................... 23 CHAPTER 1: INTRODUCTION.................................................................................... 24 1.1 Stillbirth............................................................................................................................... 24 The global burden of stillbirth.................................................................................. 24 Classifying the causes of stillbirth........................................................................... 24 Identification of the at-risk fetus............................................................................. 25 1.2. Reduced fetal movements........................................................................................... 29 Defining reduced fetal movement........................................................................... 30 Causes and associations of reduced fetal movement..................................... 32 Current management following reported reduced fetal movements........................................................................................................................ 34 1.3 Assessment of placental structure and function in utero............................... 37 Placental size and structure...................................................................................... 37 Placental function.......................................................................................................... 46 In utero placental assessment in reduced fetal movement pregnancies....................................................................................................................... 58 1.4 Defining and addressing the research question................................................. 59 Page 2 of 253 HYPOTHESIS................................................................................................................... 62 Aims.............................................................................................................................................. 62 Objectives................................................................................................................................... 62 CHAPTER 2: PLACENTAL FEATURES OF LATE-ONSET ADVERSE PREGNANCY OUTCOME.............................................................................................. 63 2.1 Abstract................................................................................................................................ 64 2.2 Introduction....................................................................................................................... 65 2.3 Materials and methods.................................................................................................. 67 Placental structure........................................................................................................ 67 Placental function.......................................................................................................... 67 Statistical analysis.......................................................................................................... 71 2.4 Results.................................................................................................................................. 72 2.5 Discussion........................................................................................................................... 79 2.6 Acknowledgements........................................................................................................ 84 2.7 Statement of author contributions.......................................................................... 84 2.8 Supplementary data....................................................................................................... 85 CHAPTER 3: PLACENTAL VOLUME AND BIOMETRY MEASUREMENT IN THE THIRD TRIMESTER: A VALIDATION STUDY............................................... 86 3.1 Abstract................................................................................................................................ 87 3.2 Introduction....................................................................................................................... 88 3.3 Materials and methods.................................................................................................. 90 Sonographic assessment of placental size and shape.................................... 90 Modelling of placental shape and tissue density.............................................. 91 Validation of sonographic placental biometry, volume, weight and fetoplacental ratio.......................................................................................................... 91 Assessment of sonographic reliability.................................................................. 95 Statistical analysis.......................................................................................................... 95 3.4 Results.................................................................................................................................. 97 3.5 Discussion........................................................................................................................... 104 3.6 Acknowledgements........................................................................................................ 107 3.7 Statement of author contributions.......................................................................... 107 Page 3 of 253 CHAPTER 4: FETOPLACENTAL ARTERIAL DOPPLER VALIDATION IN HUMAN PREGNANCY..................................................................... .............................. 108 4.1 Abstract................................................................................................................................ 109 4.2 Introduction....................................................................................................................... 110 4.3 Materials and methods.................................................................................................. 112 Sonographic assessment of the fetoplacental arterial circulation............ 112 Ex vivo examination of placental vascularity and arterial function............................................................................................................................... 114 Assessment of fetoplacental arterial Doppler reliability.............................. 116 Statistical analysis.......................................................................................................... 116 4.4 Results.................................................................................................................................. 118 4.5 Discussion........................................................................................................................... 127 4.6 Acknowledgements........................................................................................................ 132 4.7 Statement of author contributions.......................................................................... 132 4.8 Supplementary data....................................................................................................... 133 CHAPTER 5: PLACENTAL HORMONE CONCENTRATIONS: WHAT ASPECT(S) OF PLACENTAL FUNCTION DO THEY REFLECT?........................... 134 5.1 Abstract................................................................................................................................ 135 5.2 Introduction....................................................................................................................... 137 5.3 Materials and methods.................................................................................................. 139 Hormone selection............................................................................................................ 139 Fresh tissue processing.................................................................................................. 139 Assessment of placental microstructure................................................................. 140 Assessment of transcription......................................................................................... 141 Assessment of hormone concentration in biofluids........................................... 142 Statistical analysis............................................................................................................. 142 5.4 Results.................................................................................................................................. 144 5.5 Discussion........................................................................................................................... 151 5.6 Acknowledgements........................................................................................................ 155 5.7 Statement of author contributions.......................................................................... 155 CHAPTER 6: PLACENTAL ASSESSMENT PREDICTS ADVERSE PREGNANCY OUTCOME AFTER REDUCED FETAL MOVEMENT: A Page 4 of 253 COHORT STUDY................................................................................ ............................. 156 6.1 Abstract................................................................................................................................ 157 6.2 Introduction....................................................................................................................... 159 6.3 Materials and methods.................................................................................................. 160 Participant recruitment.................................................................................................. 160 Placental assessment in utero...................................................................................... 162 Outcome definition........................................................................................................... 163 Statistical analysis............................................................................................................. 163 6.4 Results.................................................................................................................................. 165 6.5 Discussion........................................................................................................................... 172 6.6 Acknowledgements........................................................................................................ 175 6.7 Statement of author contributions.......................................................................... 175 6.8 Supplementary data....................................................................................................... 175 CHAPTER 7: DISCUSSION AND FUTURE WORK.................................................. 187 7.1 Introduction....................................................................................................................... 187 7.2 Placental features of late-onset adverse pregnancy outcome..................... 188 7.3 Validation of size, vascular and endocrine placental biomarkers.............. 190 7.4 Placental assessment in the prediction of adverse pregnancy outcome. 195 7.5 Future work....................................................................................................................... 200 7.6 Conclusion.......................................................................................................................... 202 REFERENCES................................................................................................................... 203 WORD COUNT: 64,068 Page 5 of 253 LIST OF TABLES TABLE 1: Placental features associated with stillbirth, fetal growth restriction and reduced fetal movements............................................. TABLE 2: 28 Baseline and delivery characteristics of study participants and their offspring.................................................................................................... 73 TABLE 3: Ex vivo placental endocrine function....................................................... 78 TABLE 4: The placental ultrasound validation study cohort............................ 98 TABLE 5: Comparison of sonographically estimated placental biometry to ex vivo directly measured placental size........................................... TABLE 6: 99 Comparison of sonographically estimated placental volume and weight to ex vivo directly measured placental volume and weight.................................................................................................................... 100 TABLE 7: Reliability of third trimester placental size assessment................. 101 TABLE 8: The fetoplacental arterial Doppler study cohort................................ 119 TABLE 9: Relationships between placental arterial Doppler waveforms and villous vascularity indices................................................................... 121 TABLE 10: Relationships between placental arterial Doppler waveforms and vascular functional assessment........................................................ 123 TABLE 11: Intra- and inter-observer reliability of placental arterial Doppler impedance indices......................................................................... 124 TABLE 12: The placental hormone concentration study cohort........................ 145 TABLE 13: Relationship of circulating maternal serum placental hormone concentrations and placental structure................................................. 148 TABLE 14: Relationship of circulating maternal serum placental hormone concentrations placental endocrine function...................................... 149 TABLE 15: Comparison of women participating and declining to participate in the FEMINA2 trial and their pregnancy 161 outcomes.............................................................................................................. TABLE 16: Odds ratios for adverse pregnancy outcome by individual differentially distributed variables.......................................................... TABLE 17: Components and comparison of proposed predictive models.................................................................................................................. TABLE 18: Test performance characteristics of proposed predictive Page 6 of 253 166 170 models................................................................................................................... 171 TABLE 19: Comparison of placental features between placentas of stillborn, liveborn growth restricted and adverse outcome reduced fetal movements pregnancies................................................... 189 TABLE 20: Placental biomarkers validated within the studies of this thesis...................................................................................................................... 191 SUPPLEMENTARY TABLES SUPPLEMENTARY TABLE 1: Primers used to study placental transcription of key placental hormones.............................................. 85 SUPPLEMENTARY TABLE 2: Enzyme-Linked Immunosorbant Assay kits used to quantify hormone content of tissue lysate and explantconditioned media........................................................................................... SUPPLEMENTARY TABLE 3: Relationship of Doppler 85 site- specific vascular impedance to subsequent birth weight centile…............ 133 SUPPLEMENTARY TABLE 4: Variable reduction......................................................... 176 SUPPLEMENTARY TABLE 5: STARD guideline checklist......................................... 184 Page 7 of 253 LIST OF FIGURES FIGURE 1: Fetoplacental response to in utero stress........................................ FIGURE 2: Assessment of umbilical cord insertion position and coiling index................................................................................................................. FIGURE 3: 33 42 Quantification of flow and resistance in arteries by Doppler waveform analysis..................................................................................... 45 FIGURE 4: Ex vivo placental macrostructure........................................................ 74 FIGURE 5: Ex vivo placental microstructure......................................................... 75 FIGURE 6: Ex vivo placental arterial function....................................................... 76 FIGURE 7: Measurement of placental biometry using two-dimensional ultrasound..................................................................................................... FIGURE 8: Measurement of placental volume using two- and threedimensional ultrasound.......................................................................... FIGURE 9: 93 94 Visual representation of the accuracy of placental volume estimation methods................................................................................... 99 FIGURE 10: Intra-observer reliability of placental size estimates................. 102 FIGURE 11: Inter-observer reliability of placental size estimates................. 103 FIGURE 12: Doppler waveforms acquired from different sites in the 113 fetoplacental arterial circulation......................................................... FIGURE 13: Relationship of arterial flow impedance to proximity to the 120 placental microvasculature.................................................................... FIGURE 14: Relationship of intraplacental artery Doppler impedance to 122 villous vascularity...................................................................................... FIGURE 15: Intra-observer reliability of placental arterial Doppler 125 impedance estimates................................................................................ FIGURE 16: Inter-observer reliability of placental arterial Doppler 126 impedance estimates................................................................................ FIGURE 17: Relationship between maternal circulating hormone 146 concentration and gestational age...................................................... FIGURE 18: Relationship between maternal circulating hormone 147 concentration and individualised birth weight centile.............. FIGURE 19: Hormone release time course from villous explants.................. Page 8 of 253 150 FIGURE 20: Flow of participants through the FEMINA2 160 study................................................................................................................ FIGURE 21 Receiver operator characteristic curve comparison.................. 169 SUPPLEMENTARY FIGURES SUPPLEMENTARY FIGURE 1: Predictive model generation................................ Page 9 of 253 186 LIST OF ABBREVIATIONS AC Abdominal circumference AFI Amniotic Fluid Index AGA Appropriate for gestational age aOR Adjusted odds ratio APO Adverse pregnancy outcome AUC Area under the curve BE Base excess BMI Body mass index CD31 Cluster of differentiation-31 CI Confidence interval CK7 Cytokeratin-7 CM Explant conditioned medium CoV Coefficient of variance CPA Chorionic plate artery CPAD Chorionic plate artery Doppler CTG Cardiotocograph CU Cerebroumbilical D Depth DiamPAT Diameter at peak active tension DNA Deoxyribonucleic acid DV Ductus venosus EC50 Effective concentration for 50% response ECL Echogenic cystic lesions EDF End diastolic flow EFW Estimated fetal weight EFWc Estimated fetal weight centile ELISA Enzyme linked immunosorbant assay Est Estimated FEMINA2 Second fetal movement intervention and assessment trial FGR Fetal growth restriction Flt-1 Fms-like tyrosine kinase-1 F-PlGF Free placental growth factor Page 10 of 253 FPR fetoplacental ratio FPRV Fetoplacental volume ratio FPRW Fetoplacental weight ratio Gest Gestation GROW Gestation-related optimal weight hCG human chorionic gonadotrophin hPL human placental lactogen IBC Individualised birth weight centile ICC Intraclass correlation coefficient ID Identity IPA Intraplacental artery IPAD Intraplacental artery Doppler IQR Interquartile range L Length LR+ Positive likelihood ratio LR- Negative likelihood ratio MCA Middle cerebral artery MoM Multiples of the median MP Multiplanar MP5 Multiplanar 5mm slicing interval MP10 Multiplanar 10mm slicing interval MRI Magnetic resonance imaging mRNA Messenger ribonucleic acid N Number NICE National Institute of Clinical and Health Excellence NICU Neonatal intensive care unit NPO Normal pregnancy outcome NPV Negative predictive value OR Odds ratio PAPP-A Pregnancy associated plasma protein-A PAT Peak active tension PCR Polymerase chain reaction PI Pulsatility index Page 11 of 253 PlGF Placental growth factor PPV Positive predictive value PSV Peak systolic velocity PV Placental volume PW Placental weight qPCR Quantitative polymerase chain reaction RCOG Royal College of Obstetricians and Gynaecologists RFM Reduced fetal movements RI Resistance index RNA Ribonucleic acid ROC Receiver operating curve characteristics RR Relative risk Rs Spearman rank correlation coefficient RT Reverse transcription SFH Symphysiofundal height SGA Small for gestational age SNP Sodium nitroprusside TBP TATA-box protein TR Trophoblast ratio U46619 Thromboxane A2 mimetic UAD Umbilical artery Doppler UAD-A Umbilical artery abdominal insertion Doppler UAD-F Umbilical artery free loop Doppler UAD-P Umbilical artery placental insertion Doppler UK United Kingdom USA United States of America UtAD Uterine artery Doppler Vmax Maximal response to vasoactive agent VOCAL Virtual organ computer-aided analysis W Width 2-ΔCt Relative transcription rate 2D Two-dimensional 3D Three-dimensional Page 12 of 253 3DPD Three-dimensional power Doppler Page 13 of 253 SCIENTIFIC ABSTRACT of thesis entitled “Assessing and quantifying placental dysfunction in relation to pregnancy outcome in pregnancies complicated by reduced fetal movements” submitted by Lucy Elizabeth Higgins for the Degree of PhD from the University of Manchester, February 2015 Currently there is no test to accurately predict stillbirth. It is proposed that better identification of placental disease in utero may aid stillbirth prediction and prevention. Pregnancies complicated by reduced fetal movement (RFM) have increased risk of stillbirth. We hypothesised that RFM is a symptom of placental dysfunction associated with adverse pregnancy outcome (APO) and that this placental abnormality can be detected antenatally and used to identify fetuses at highestrisk of APO. We tested this hypothesis by: 1) comparison of ex vivo placental structure and function between APO RFM pregnancies and their normal outcome RFM counterparts, 2) comparison of in utero estimates of placental size, vascularity, vascular and endocrine functions obtained from placental ultrasound, Doppler waveform analysis and maternal circulating placentallyderived hormone concentrations, to their ex vivo correlates and 3) examination of the predictive potential of placental biomarkers at the time of RFM. Ex vivo placentas from APO RFM pregnancies, compared to normal outcome RFM counterparts, were smaller (diameter, area, weight and volume, p<0.0001), less vascular (vessel number and density, p≤0.002), with arteries that were less responsive to sodium nitroprusside (p<0.05), and with aberrant endocrine function (reduced tissue content and/or release of human chorionic gonadotrophin (hCG), human placental lactogen (hPL) and soluble fms-like Tyrosine Kinase-1 (sFlt-1), p<0.03). Placental volume (PV) ex vivo correlated with sonographic estimated PV (p<0.004), hPL, hCG and placental growth factor (PlGF) concentrations in the maternal circulation (p<0.03). Ex vivo villous vessel number and density correlated with Doppler impedance at the umbilical artery free-loop (UAD-F, p=0.02) and intraplacental arteries (p<0.0001) respectively, whilst UAD -F impedance correlated with arterial thromboxane sensitivity (p<0.04). Examination of placental structure and function at the time of presentation with RFM identified 15 independently-predictive biomarkers. Three potential predictive models, incorporating measures of placental size (PlGF), endocrine function (sFlt-1), arterial thromboxane sensitivity and villous vascularity (UADF), were proposed. Using these models, sensitivity for APO was improved from 8.9% with baseline care (assessment of fetal size and gestation) to up to 37.5% at a fixed specificity of 99% (p<0.05). This series of studies shows that antenatal placental examination is possible and improves identification of pregnancies at highest risk of stillbirth in a high-risk population by up to 29%. Therefore such tests merit further development to prospectively assess their ability to predict and prevent stillbirth itself. Page 14 of 253 LAY ABSTRACT of thesis entitled “Assessing and quantifying placental dysfunction in relation to pregnancy outcome in pregnancies complicated by reduced fetal movements” submitted by Lucy Elizabeth Higgins for the Degree of PhD from the University of Manchester, February 2015 Stillbirth (the death of an infant before birth) occurs in one in every 250 -300 pregnancies after 28 weeks of pregnancy in high-income countries. There is currently no good test to predict and prevent this. Stillbirth is associated with disease of the placenta (afterbirth) in six in every ten cases. It is proposed that better identification of placental disease before birth can improve the prediction, and prevention, of stillbirth. Pregnancies where there is a reduction in the baby’s movements (RFM) are two to three times more likely to experience stillbirth. We asked whether examination of the placenta before birth is possible, and whether it can help identify pregnancies at highest risk of adverse outcome (APO) after RFM. We answered this question by comparing the placentas of APO RFM pregnancies to those of their normal outcome counterparts after delivery to identify differences that may form the basis of useful tests. Ultrasound measurement of the placenta and its blood flow, and placental hormone concentrations in the mother’s blood were compared to direct measurements obtained from the same placentas after birth. Finally, we added antenatal placental examination to the assessment of pregnancies at the time of RFM and examined their ability to predict subsequent APO. Placentas from APO RFM pregnancies were smaller (length, width, area, volume and weight), with poorer blood supply (reduced number and density of blood vessels), and abnormal function (reduced relaxation of placental arteries, and altered production and release of placental hormones). A number of these placental features were reflected in examination of the placenta before birth. Placental size was related to ultrasound estimation of its volume, length and width, but also to the level of some hormones arising from the placenta in the maternal blood. Placental blood supply was related to assessment of blood flow resistance between the baby and placenta and within the placenta itself, whilst resistance to blood flow from baby to placenta reflected how sensitive the placental arteries were to chemicals that cause arteries to tighten. Using these placental tests, in a group of 300 pregnancies affected by RFM, three more pregnancies out of every 10 that resulted in APO could be predicted by the additional measurement of two placental hormones (placental growth factor and soluble fms-like tyrosine Kinase-1) and three measures of placental blood flow resistance between baby and the placenta, in addition to normal care. This series of studies shows that examination of the placenta before birth is possible and that it improves the ability to identify pregnancies that are at the highest risk of stillbirth following RFM. These tests should be further developed to assess their ability to predict and prevent stillbirth itself. Page 15 of 253 DECLARATION No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning. Page 16 of 253 COPYRIGHT STATEMENT i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “Copyright”) and she has given The University of Manchester certain rights to use such Copyright, including for administrative purposes. ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. iii. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the “Intellectual Property”) and any reproduction of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions. iv. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://www.campus.manchester.ac.uk/medialibrary/policies/intellec tual-property.pdf), in any relevant Thesis restrictions deposited in the University Library, The University Library’s regulations (see http://www.manchester.ac.uk/library/aboutus/regulations) and in The University’s policy on presentation of Theses. Page 17 of 253 DEDICATION To Matt, who is undoubtedly my most significant finding. Page 18 of 253 ACKNOWLEDGEMENTS First and foremost, I express my sincere gratitude to my supervisory team, Alex Heazell, Ed Johnstone and Colin Sibley. Your input and insight throughout this project have been invaluable. You have each gone above and beyond the “call of duty” in many ways in the course of my mentorship, undoubtedly shaping my future for the better. Alex, your dedication to stillbirth prevention has been a constant source of inspiration. Ed, your sonographic expertise, enthusiasm and straight talking have helped me to explore further than seemed possible or sensible! And to Colin, thank you for taking the gamble of appointing a very naïve Foundation doctor to the obstetric clinical academic training programme; you opened the doors of academic opportunity to me and have guided me ever since. Thanks are of course given to the midwives of St. Mary’s Hospital, Manchester who assisted by participant referral and placental collection and to the members of the Maternal and Fetal Health Research Centre (MFHRC), for tolerating a “Medic” in their laboratories, patiently and expertly training me in various laboratory techniques. Particular thanks are extended to Bernadette Baker, Sylvia Lui and Rebecca Jones (PCR training), Susan Greenwood (villous explant culture direction), and Tracey Mills and Mark Wareing (wire myography guidance). I am also grateful to the technical staff of MFHRC, including Alan Redfern, who variously provided support during project. Jenny Myers is acknowledged for her statistical expertise and mentorship. Statistical advice from Dr. Stephen Roberts, Centre for Biostatistics, is also recognised. This work was financially supported by a Manchester Biomedical Research Centre Research Fellowship and an Action Medical Research Training Fellowship. Finally, without the unfailing support (and proof-reading skills) of my family and friends, particularly Matthew Brice, the work presented in this thesis would not have been possible. Thank you for keeping me motivated, helping to find solutions to the many challenges overcome during the project and for accepting, even encouraging, my workaholic tendencies. Page 19 of 253 PREFACE LH is a Clinical Trainee in Obstetrics and Gynaecology who is currently Out of Programme for Research at the beginning of ST4. She has followed a Clinical Academic Training Programme (North West Postgraduate Medical Deanery) since Foundation Training (University of Birmingham) and completed an NIHR Clinical Research Fellowship in the Maternal and Fetal Health Research Centre, University of Manchester in August 2011. Page 20 of 253 CONTRIBUTIONS FROM COLLABORATORS All participant recruitment, sonographic examinations, laboratory experiments and statistical analyses were performed by LH with the exception of: Chapters 2 and 4 where Naa Addo calculated luminal areas from images produced by LH. Chapters 2 and 5 where LH and Nicolas Rey de Castro performed immunostaining of villous tissue for CK7 and quantified villous and trophoblast areas. Chapter 6 where Alan Redfern performed some maternal serum enzyme linked immunosorbant assays. Page 21 of 253 PUBLICATIONS ARISING FROM THIS WORK Review articles: 1. Higgins LE, Johnstone ED, Heazell AEP. Management of Reduced Fetal Movements. Fetal and Maternal Medicine Reviews 2013;24(4):201-231 Published abstracts: 1. Higgins L, Sibley CP, Heazell AEP, Johnstone ED. Placental volume can be accurately measured using two- and three-dimensional ultrasound near term. Arch Dis Child Fetal Neonatal Ed 2013;98:Suppl 1 A88 2. Higgins LE, Sibley CP, Wareing M, Johnstone ED, Heazell AEP. Umbilical artery PI/RI fail to detect aberrant placental arterial function in reduced fetal movement pregnancies. Reprod Sci 2014;21(3)suppl:T157 Page 22 of 253 A NOTE ON ALTERNATIVE FORMAT THESES The “Alternative Format” thesis allows a postgraduate student to present their doctoral studies in a format that is already published, or suitable for subsequent publication, in peer-reviewed journals. The work so presented must be the independent or collaborative work of the student and represent an original contribution to the field of research. This format has been selected for this thesis as the work contained within it falls naturally into five sections suitable for publication. Thus, the flow of the thesis is similar to that of a Traditional Format thesis. The process of preparing this data for presentation in manuscript form has focussed its emphasis on the key points of clinical interest, provided invaluable training in the processes of scientific writing and peer review and will enable more rapid dissemination of the findings of the thesis studies. The contributions of the student (LH) and her collaborators are specifically identified in the section “Contributions from collaborators”. Page 23 of 253 CHAPTER 1: INTRODUCTION 1. Stillbirth The global burden of stillbirth Stillbirth describes the birth of an infant showing no sign of life, being differentiated from late miscarriage by arbitrary gestational age criteria dependent on country-specific or international definitions (Lawn et al., 2011). Each year over 3,000 infants are stillborn in the United Kingdom (Office for National Statistics, 2014) contributing to an estimated 2.6 million stillbirths globally, although the true burden of stillbirth in low- and middle-income countries is likely underestimated (Cousens et al., 2011). As one facet of perinatal mortality, the rapid decline in international stillbirth rates in the mid to late 20th century has slowed in recent decades (Flenady et al., 2011a). This contrasts with significant improvements in neonatal survival (UNICEF, 2014) afforded by modern neonatal care during the same time period (Gregory, 2014). Classifying the causes of stillbirth The drive to reduce stillbirth rates is hindered by the lack of a single unified, clinically useful classification system to assign cause of death in cases of stillbirth. As many as 47% of stillbirths are labelled as “unexplained” dependent on the classification system used (Vergani et al., 2008); one problem with labelling stillbirths as “unexplained” is that this leads to the belief that they are “unpreventable”. This is despite reviews of perinatal care in cases of stillbirth citing suboptimal management as a contributory factor in the death of the infant in up to 37% of cases (Saastad et al., 2007). Other classification systems reduce the proportion of “unexplained” stillbirths within the same population to as few as 16%, with 40% of stillbirths being caused by placentally-related conditions including fetal growth restriction (FGR) and placental abruption when assessed by the TULIP classification system (Vergani et al., 2008). Critically, whilst no in utero treatment is currently available to improve fetal health, one quarter of all stillborn infants in high-income countries die at 37 or more weeks gestation (Zeitlin, 2013) when delivery would have been an option to prevent their death Page 24 of 253 without causing significant fetal or maternal morbidity (Stock et al., 2012). However, no good test to identify fetuses at highest risk of stillbirth currently exists (Haws et al., 2009). Identification of the at-risk fetus The 2011 Lancet Stillbirth series, produced by the International Stillbirth Alliance, evoked a renaissance in the drive to reduce international stillbirth rates. In low- and middle-income countries, efforts have focused on provision of basic antenatal and intrapartum care, in high-income countries such as the UK the focus has been turned on to the identification and management of the “at risk” fetus, particularly in relation to babies at risk of FGR (Flenady et al., 2011b). Yet, according to research by the Stillbirth Collaborative Research Network Writing Group (2011), less than one fifth of the population attributable risk of stillbirth can be explained by risk factors that can be ascertained at the beginning of pregnancy. Alternative strategies to identify the at-risk fetus include the identification of small for gestational age (SGA) fetuses and identification of placental disease. These will now be considered in further detail. Identification of the small for gestational age fetus Around 40% of all stillborn infants have a birth weight <10th centile (Gardosi et al., 2005). Therefore the ability to detect the SGA fetus antenatally is desirable in order to target efforts to prevent stillbirth. While not all SGA fetuses are FGR, there is evidence that even in the absence of other in utero markers purported to reflect placental insufficiency (e.g. oligohydramnios, abno rmal umbilical artery Doppler (UAD)) SGA infants are at three-fold increased risk of perinatal death (Moraitis et al., 2014). It is also known that antenatal SGA detection results in improved perinatal outcomes including reduced perinatal mortality, primar ily accounted for by increased stillbirth rate amongst undiagnosed SGA fetuses (20/573 (3.5%) vs. 6/681 (0.9%); undiagnosed SGA odds ratio (OR) 4.2) (Lindqvist and Molin, 2005). However, 46-90% of all SGA infants are not correctly identified whilst in utero (Jahn et al., 1998; Lindqvist and Molin, 2005; Page 25 of 253 Mattioli et al., 2010; Fratelli et al., 2013; Chauhan et al., 2014; Monier et al., 2014), preventing intervention to avert stillbirth related to SGA size. In the United Kingdom, the National Institute for Health and Care Excellence (NICE) recommends use of symphysiofundal height (SFH) measurement to detect the SGA fetus during routine care of the low-risk pregnant woman (2008). This estimates uterine size by measurement from a fixed point (the symphysis pubis) to the uterine fundus. Widely variable sensitivities (from 27% to 81 %) but high specificities (85% to 94%) of SFH for SGA detection are reported (National Institute for Health and Care Excellence, 2008), likely due to significant intra- and interobserver variability, even when measured using standardised technique (Morse et al., 2009). Alternatively, fetal size can be assessed by two-dimensional (2D) ultrasound based upon measurement of the fetal abdominal circumference or estimation of fetal weight (EFW) from measurement of combinations of the following biometric parameters: abdominal circumference, femur length, biparietal diameter and demonstrates head circumference. Sonographic fetal size superior performance compared to SFH assessment measurement (Hargreaves et al., 2011), but significant variability persists (Dudley, 2005). After 35 weeks gestation as many as one in three estimations differed from true birth weight by >10% (Kurmanavicius et al., 2004; Dudley, 2005; Burd et al., 2009; Melamed et al., 2009; Hargreaves et al., 2011). Interest has recently turned to three-dimensional (3D) ultrasound to improve EFW accuracy (Schild et al., 2008). However, 3D ultrasound is not currently used in routine practice. Whilst sonographic assessment of fetal size is generally considered superior to non-sonographic assessment, the practicalities (in terms of cost, infrastructure and uncertain management of pregnancies following diagnosis of SGA) of implementing whole population sonographic EFW screening are considerable (National Institute for Health and Care Excellence, 2008). Critically, a Cochrane systematic review and meta-analysis of ultrasound in the low-risk pregnancy concluded that there was insufficient evidence of reduction in perinatal Page 26 of 253 mortality to support its use (Bricker et al., 2008), although the study was underpowered to have detected significant reduction in stillbirth (N=2,834), particularly as no intervention protocol after detection of sonographic concerns was pre-specified. Identification of placental disease Ultimately, a single assessment of fetal size does not assess the rate of increase (growth) and will not identify fetuses that are currently appropriate for gestational age (AGA) size but whose growth trajectory is declining. As fetal growth is related to the ability of the placenta to provide sufficient nutrients and oxygen to the fetus, placental examination might identify non-SGA FGR fetuses or may overcome inaccuracies in sonographic fetal size assessment. Placental examination following stillbirth provides relevant information regarding cause of death and management of future pregnancies in 88% of cases (Kidron et al., 2009) and halves the number of “unexplained” stillbirths (Heazell and Martindale, 2009). A systematic review of studies examining the placenta following stillbirth reports causal placental abnormalities in 11–65% of all stillbirths, predominantly lesions of maternal- and fetal-placental underperfusion (Ptacek et al., 2014). Thus, the placenta is a potential target for antenatal assessment of stillbirth risk. A range of abnormalities have been described in the placentas of stillborn infants, including many features common to the placentas of liveborn FGR infants (summarised in Table 1). In particular, placentas from stillborn infants are smaller in terms of weight (Worton et al., 2014), and less vascular with increased thrombosis, infarction and fibrin deposition (Pinar et al., 2014). Although it is not possible to directly assess placental function after stillbirth, evidence of altered placental function is provided from increased fetoplacental weight ratios (FPR = birth weight / placental weight) (Worton et al., 2014) and retrospectively inferred from first and second trimester maternal circulating placentally-derived hormone concentrations (Gagnon et al., 2008). Further insight can be gained from examination of ex vivo functional abnormalities in Page 27 of 253 placentas of liveborn FGR infants, including potential effect on nutrient transport and placental vascular function (Table 1) which demonstrates altered amino acid, glucose, ion and oxygen transport (summarised by Sibley et al. (2005)) and deranged placental arterial function (Mills et al., 2005; Mills, 2011). TABLE 1: Placental features associated with stillbirth, fetal growth restriction and reduced fetal movements. Stillbirth FGR RFM Placental Structure Trimmed placental weight ↓ Mean placental diameter Unreported ↓ ↓ Placental “roundness” Unreported Unreported Central cord insertion ↓ ↓ ↑ ↑ ↓ ↑ ↑ ↓ ↓ ↑ ↔/↑ ↔/↓ ↑ ↑ ↓ ↓ ↔ ↓ ↓ ↑ Endocrine profile* ↑ ↓/↑* Amino acid transfer Not applicable Lipid transfer Not applicable Glucose transport Not applicable Ion transport Metabolism Placental vascular function Inflammation Not applicable Unreported Not applicable ↔/↑ ↓/↑* ↓ ↓ ↔ ↓/↑ Villous vascularity Infarction / thrombosis Fibrin deposition Trophoblast area Exchange barrier thickness Syncytial nuclear aggregates Placental Function Fetoplacental weight ratio Proliferation Apoptosis ↑ ↑ Unreported Unreported ↓ Unreported ↑ ↑ ↓* ↓ Unreported Unreported Altered* Altered Unreported Altered* Unreported ↑ ↔/↓ ↑ ↑ ↑ ↔ Collated from studies by Gagnon et al. (2008), Out et al. (1991), Pinar et al. (2014), Tantbirojn et al. (2009) and Worton et al. (2014) in pregnancies complicated by stillbirth, studies by Biswas and Ghosh (2008) and (2008), Chen et al. (2002), Daayana et al. (2004), Junaid et al. (2014), Mills et al. (2005) and (2011), Roje et al. (2011), Sibley et al. (2005), Smith et al. (1997), Spinillo et al. (2012) and Vermedovska et al. (2011) in pregnancies complicated by fetal growth restriction (FGR), and studies by Dutton et al. (2012), Girard et al. (2014), Heazell et al. (2012), Warrander et al. (2012) and Winje et al. (2012) in reduced fetal movements (RFM). Key: ↑ = increased, ↓ = decreased, ↔ = unaltered, * largely inferred from studies of maternal or cord blood or explant conditioned medium. Page 28 of 253 Antenatal placental examination is not currently part of routine antenatal care in the UK and will be considered in more detail in section 1.3. 1.2. Reduced fetal movements The International Stillbirth Alliance has identified the “timely evaluation of women reporting decreased fetal movements” as a priority for reducing stillbirth rates in high-income countries (Flenady et al., 2011b). Yet there is a lack of good evidence to guide this evaluation and subsequent management (Whitworth et al., 2011). Consequently, much variation exists between professionals’ views of appropriate investigation and management (Heazell et al., 2008; Flenady et al., 2009) and suboptimal management of reduced fetal movement (RFM) contributes to the occurrence of stillbirth (Evers et al., 2011). Following presentation with RFM and a live baby, women experience a two - to three-fold increased rate of subsequent stillbirth and SGA birth weight (O'Sullivan et al., 2009; Pagani et al., 2014a; Pagani et al., 2014b). Retrospective analysis of stillbirth cases describes that 54.7% of cases are preceded by maternal perception of RFM for a period of hours to days (Efkarpidis et al., 2004). Evidence of increased failure rate of neonatal hypothermia for hypoxic ischaemic encephalopathy (Bonifacio et al., 2011) and neurodevelopmental delay in infants whose mothers reported RFM (James et al., 2000) further suggests that RFM is associated with a prolonged adverse intrauterine environment. Studies of formal fetal movement counting frequency report that, of stillbirths occuring in fetal movement counting populations, up to 100% of subsequently bereaved mothers first presented to maternity services with RFM whilst the fetus was alive and were “falsely reassured” by subsequent monitoring (Pearson and Weaver, 1976; Leader et al., 1981; Sadovsky et al., 1983; Valentin et al., 1986; Moore and Piacquadio, 1989). This indicates that a window of opportunity for intervention may exist between maternal perception of RFM and fetal death in utero. However, RFM itself demonstrates low specificity for stillbirth; one or more episode of RFM will be reported in 4.0-18.1% of all pregnancies (Froen, 2004), Page 29 of 253 with approximately 1 in 5 of these women reporting multiple episodes (Dutton et al., 2012). Yet the rate of adverse pregnancy outcome (APO) amongst these pregnancies is <30% (O'Sullivan et al., 2009; Dutton et al., 2012). Thus additional tests of fetoplacental wellbeing are required to improve the evaluation of these pregnancies. In light of this, the Royal College of Obstetricians and Gynaecologists (RCOG) issued guidance on the management of RFM in 2011 (Whitworth et al., 2011). Defining reduced fetal movement One of the challenges in preventing stillbirth associated with RFM is differentiating “normal” from “abnormal” fetal activity; a reduction in movement frequency may be defined numerically or subjectively. A number of studies have attempted to apply thresholds for the “minimum” number of fetal movements perceived in a given time period; these limits range from as many as 10 movements in two hours (Moore and Piacquadio, 1989) to as few as 12 movements in 24 hours (Pearson and Weaver, 1976; Sadovsky et al., 1983) or even absence of movements for 24 hours (Leader et al., 1981). This no doubt contributes to confusion regarding appropriate advice to give women regarding fetal movements (Heazell et al., 2008; Flenady et al., 2009), and affects the timing within any potential window of opportunity for intervention at which pregnant women would present to maternity services. Systematic review of studies of fetal movement counting concluded that there was insufficient evidence to support a reduction in stillbirths, hospital admissions, operative delivery or early neonatal compromise (low five-minute Apgar score) with formal fetal movement counting (Mangesi and Hofmeyr, 2007). The meta-analysis was heavily influenced by the inclusion of a large cluster-randomised controlled study with a conservative definition of RFM as absent movements for at least 24 hours or less than 10 movements in 12 hours for two consecutive days, including 67,879 participants (Grant et al., 1989). The methodologies employed in this study by Grant et al. resulted in the final intention-to-treat analysis being underpowered; individuals randomised to both Page 30 of 253 arms of the study mixed within the same hospital resulting in roughly 50% noncompliance with counting in the active arm, and active counting by one in three women in the control arm. The potential impact of formal fetal movement counting was further reduced by failure to pre-specify the management of women presenting with RFM. Nevertheless, the rate of stillbirth in the entire study population decreased from a (conservatively estimated pre-study) 4 per 1000 pregnancies to just 2.9 per 1000 pregnancies, the lowest rate ever reported in the United Kingdom. Although no single formal movement counting technique has been found to be superior to other proposed counting techniques (Heazell and Froen, 2008), a consistent observation in studies of fetal movement counting is a reduction in the stillbirth rate in the study population as a whole (Neldam, 1983; Westgate and Jamieson, 1986; Grant et al., 1989; Moore and Piacquadio, 1989). Whilst critics may claim that this is simply a “Hawthorne effect” (Braunholtz et al., 2001), it appears that maternal and professional awareness of the importance of RFM may be beneficial in averting stillbirth. Indeed, in this case the “Hawthorne effect” (awareness) could be considered the active intervention. Low mater nal awareness of fetal activity has itself been linked to an increased risk of adverse pregnancy outcome particularly SGA birth weight (Saastad et al., 2008). Therefore, national guidance promotes maternal awareness of the typical pattern of an individual baby’s movements, and rapid reporting of subjective abnormalities of fetal activity rather than formal fetal movement counting (National Institute for Health and Care Excellence, 2008; Whitworth et al., 2011). A Norwegian study (Tveit et al., 2009; Saastad et al., 2010) demonstrated that delays in reporting RFM were reduced (without increase in the number of presentations with RFM) along with a reduction in the overall stillbirth rate when consistent information regarding RFM is provided to prospective mo thers. This, combined with a pre-specified package of care following presentation with RFM, is the primary basis of the Promoting Awareness of Fetal Movements to Reduce Fetal Mortality Stillbirth, a Stepped Wedge Cluster Randomised Trial Page 31 of 253 (AFFIRM) study (NCT01777022), currently recruiting in the United Kingdom and Eire (National Library of Medicine, 2014). Causes and associations of reduced fetal movement The frequency of fetal movement, or maternal perception thereof, is reportedly reduced in the context of maternal obesity or an anterior placenta (Tuffnell et al., 1991), fetal lie with the spine anterior (Fisher, 1999), maternal drug consumption including alcohol, nicotine, opiates, barbiturates, benzodiazepines and other hypnotic/anxiolytics (Hijazi and East, 2009), and following corticosteroid administration (Mulder et al., 1997; Jackson et al., 2003). Rarely an episode of RFM can be caused by fetomaternal haemorrhage (Wilcock and Kadir, 2004; Mahendru et al., 2007). However, when women present with RFM of a viable fetus, the most commonly encountered pathological condition is FGR (summarised by Froen (2004)). A recent randomised controlled study of fetal movement awareness demonstrated an increased detection rate of FGR (Saastad et al., 2011). It is hypothesised that in circumstances of limited oxygen or nutrient supply to the fetus, energy is conserved through the limitation of voluntary movement (Figure 1) (Maulik, 1997). In this hypothesis RFM reduces the metabolic demands of skeletal muscle and allows diversion of oxygen and nutrients to the essential organs. As such, caution should be exercised in the attribution of maternal report of RFM to factors such as placental position without exclusion of potentially pathologic causes. The placenta in reduced fetal movements As might be expected from the associations of RFM with FGR and stillbirth, recent studies have demonstrated that placentas from RFM pregnancies display a range of structural and functional abnormalities similar to those observed in the same conditions (summarised in Table 1) (Dutton et al., 2012; Warrander et al., 2012; Winje et al., 2012; Girard et al., 2014). Compared with placentas from control pregnancies, those from liveborn RFM pregnancies weigh less, are smaller in terms of disc area and diameter, have Page 32 of 253 more frequent non-central cord insertion sites and higher proportions of abnormal appearance placental tissue (Warrander et al., 2012). In terms of microstructure, RFM placentas display increased area of infarction, number of FIGURE 1: Fetoplacental response to in utero stress. Adapted from “Doppler Velocimetry for Fetal Surveillance: Adverse Perinatal Outcome and Fetal Hypoxia.” IN: Zalud, I. and Maulik, D. (Eds.) Doppler ultrasound in obstetrics and gynaecology. New York: Springer-Verlag, 1997. syncytial nuclear aggregates and increased villous nuclear proliferation, an d a reduction in the number of vessels per villus and trophoblast area (Warrander et al., 2012). There is also preliminary evidence of altered placental function in RFM pregnancies, with an increased fetoplacental ratio (FPR), reduced system A (placental neutral amino acid transporter) activity (Warrander et al., 2012) and evidence of imbalance between pro- and anti-inflammatory factors in maternal serum and villous tissue (Girard et al., 2014). It can also be inferred that the RFM placental phenotype, particularly functional phenotype, is more pronounced in association with adverse pregnancy outcome (APO) with reduced system A activity (Warrander et al., 2012), lower maternal serum concentrations of human chorionic gonadotrophin (hCG), human placental lactogen (hPL) and progesterone (Dutton et al., 2012), and alterations of the progesterone metabolic pathway (Heazell et al., 2012). Page 33 of 253 However, such studies may have been biased by limiting placental analysis to those delivered within seven days of RFM; unrestricted placental analysis failed to corroborate the majority of Warrander et al.’s findings, but did report an increased incidence of composite “abrupt” maternoplacental insufficiency lesions (including infarction, abruption and haemorrhage) in RFM pregnancy placentas, particularly in those delivered within one week (OR 2.4 and OR 3.0 respectively) (Winje et al., 2012). Furthermore pregnancies with first and second trimester evidence of placental dysfunction (as defined by low first trimester maternal circulating pregnancy associated plasma protein A (PAPP-A) concentrations or high first or second trimester uterine artery Doppler (UtAD) studies) are more likely to present with RFM in the third trimester (OR of RFM respectively ≥1.6 and ≥1.5 respectively) (Pagani et al., 2014a; Pagani et al., 2014b). These data suggest that RFM is linked to stillbirth via common placental pathology. Further work is required to confirm whether the placental phenotypes of RFM pregnancies with normal outcome (NPO) and APO are truly different, and could therefore be a potential target for antenatal assessment of relevant placental structure and function in RFM pregnancies. Current management following reported reduced fetal movements Investigation following maternal report of RFM aims to confirm fetal viability and to identify the potentially compromised fetus before irreversible harm occurs, whilst avoiding unnecessary intervention in the healthy fetus. Confirming fetal viability and excluding acute fetal compromise Fetal viability is traditionally confirmed by the presence of audible fetal heart pulsations but has low predictive value for ongoing fetal viability (National Institute for Health and Care Excellence, 2008). Next, the pattern of fetal heart activity is examined over a period of time by a cardiotocograph (CTG); ultimately a “normal” CTG trace is similarly unable to discriminate between fetuses at risk of future deterioration with meta-analysis failing to demonstrate improvement in perinatal mortality with use of antenatal CTG (Grivell et al., 2010). However, as this review included mortality data for only 1,627 pregnancies it was Page 34 of 253 underpowered to detect small, but significant differences in mortality rate. The incidence of abnormal CTG tracings in the RFM population is low (2-8% in two studies of 305 and 524 pregnancies respectively with RFM at ≥28 weeks gestation) (Daly et al., 2011; Dutton et al., 2012) with low sensitivity (4%) but reasonable negative predictive value (73.3%) for the detection of placental insufficiency (oligohydramnios or asymmetric growth restriction) (Heazell et al., 2005). Strategies to improve the sensitivity of antenatal CTG for acute fetal compromise include computerised interpretation, (static and ambulatory) fetal electrocardiogram and fetal stimulation tests (Tan and Sabapathy, 2001; Tan and Smyth, 2001; Neilson, 2006; Grivell et al., 2012a). Whilst computerised CTG interpretation was reported to reduce perinatal mortality compared to traditional CTG in a recent meta-analysis (relative risk, RR 0.2) the confidence intervals of this estimate were wide (0.4 to 0.88), largely reflecting the small number of pregnancies studied (N=469) (Grivell et al., 2012a). Computerised CTG interpretation has not been applied in the context of RFM and is not recommended in current national guidelines (Whitworth et al., 2011). Assessment of fetal wellbeing Following assessment of fetal size (as previously discussed) by SFH or ultrasound fetal biometry, biophysical assessment may be performed in RFM pregnancies undergoing sonographic assessment. Biophysical assessment considers the functional status of the fetus and placenta, aiming to identify fetuses that are compromised in utero. By doing so, obstetric intervention can be rationalised to minimise the risks of both perinatal mortality and inappropriate iatrogenic prematurity for each pregnancy. The three most common biophysical tests applied in the context of RFM are assessment of amniotic fluid volume, UAD impedance assessment and biophysical profile, of which only amniotic fluid volume assessment is recommended by the RCOG (Whitworth et al., 2011). As the UAD is commonly considered to be a test of placental vascular function, it will be considered separately later in this chapter, in section 1.3. Page 35 of 253 Amniotic fluid volume may be quantified sonographically by maximum pool depth or four-quadrant amniotic fluid index (AFI). A recent systematic review demonstrated that use of AFI resulted in a higher level of obstetric intervention without corresponding improvement in fetal outcomes, although the study was underpowered to detect improvement in perinatal mortality (N=1,689), and AFI more accurately predicted caesarean section for fetal distress (RR 1.45; 95% confidence intervals 1.07 – 1.97) (Nabhan and Abdelmoula, 2008). In the context of RFM, liquor volume is not independently predictive of APO (Tveit et al., 2009; Dutton et al., 2012). Less commonly, a biophysical profile (BPP) is performed (Heazell et al., 2008). This is a scoring system based upon CTG and sonographic findings (AFI and presence/type of fetal movements). Low BPP scores are associated with fetal acidaemia on cordocentesis (Vintzileos et al., 1987) but the BPP score is rarely abnormal before other fetal monitoring modalities such as the UAD (Baschat et al., 2001), thus BPP monitoring is unlikely to identify additional “at risk” fetuses. BPP assessment has not been specifically applied in the context of RFM and its use is not recommended in its management (Whitworth et al., 2011). Identification of an SGA fetus, or one with evidence of biophysical compromise, at presentation with RFM should prompt detailed obstetric assessment and creation of an individualised management plan (Whitworth et al., 2011). Currently, the only available intervention to prevent stillbirth is delivery. Such decisions should be guided by gestational age, to inform assessment of the riskbenefit ratio of delivery. However, as discussed, the currently-available monitoring modalities lack sufficient sensitivity and specificity for future fetal compromise, thus delivery (or lack thereof) has the potential to result in avoidable harm in many cases (MacKay et al., 2010; Evers et al., 2011). Better tests to rationalise delivery decisions are therefore required. Given the increased frequency of placental lesions in APO RFM pregnancies it is logical to consider tests of placental structure and function in this circumstance (Heazell et al., 2015). These might identify fetoplacental compromise at an earlier stage, Page 36 of 253 allowing intensification of fetal monitoring, administration of corticosteroids as required, and planning of delivery to reduce perinatal mortality. 1.3. Assessment of placental structure and function in utero Two potential approaches to identifying placental disease in utero can be applied: “predictive” or “diagnostic”. In this context the term “predictive” applies to the attempted identification of placental disease in the first and second trimesters of pregnancy where the placental pathology associated with stillbirth may have begun but has not yet significantly diverged from “normal” and has not yet affected fetal wellbeing. The term “diagnostic” refers to the identification of established placental disease, which may already be exerting subclinical impairment of fetal wellbeing. Where no intervention exists to ameliorate the pathology causing deviation of the test result from “normal”, the benefit of predictive testing is limited. In the third trimester, when the risks of delivery may be outweighed by the potential risks of the fetus remaining in an adverse intrauterine environment (Rosenstein et al., 2012), diagnostic testing may be more beneficial. Placental size and structure In utero placental size and structure assessment Placental growth restriction precedes FGR by a period of up to three weeks (Wolf et al., 1989). Thus, examination of placental size (alone or in comparison to fetal size) may enable improved prediction of APO compared with EFW evaluation alone. Placental size can be quantified biometrically by placental diameter or depth, and by volume or weight. Placental weight and volume A number of sonographic methodologies for estimating placental volume (PV) in utero have been described including 2D ultrasound techniques based upon formulae for an ellipse (Suri et al., 2013) or elliptical shell (Azpurua et al., 2010) and 3D ultrasound techniques of rotational tracing (Virtual organ computer - Page 37 of 253 aided analysis (VOCAL); tracing the placental outline within a captured 3D volume at pre-specified rotation angles) (de Paula et al., 2008) and slicing (Hafner et al., 2001). No studies report the in utero sonographic determination of FPR. Only one study has attempted to validate sonographic PV estimates (using the 2D shell technique) with placental size (weight, PW) ex vivo (Azpurua et al., 2010), reporting significant correlation (Rs=0.80, N=38) in the third trimester, primarily in preterm deliveries. No studies have correlated estimated and true PV. Furthermore, the reliability of PV estimates is not reported for the 2D methodologies and for 3D methodologies first and second trimester intraclass correlation coefficients (ICC) vary between 0.59 (Jones et al., 2011) and ≥0.88 (Deurloo et al., 2007a; Cheong et al., 2010). Despite these reliability concerns, there is evidence of reduced PV (primarily estimated by VOCAL) in pregnancie s resulting in SGA birth weight from as early as the first (Law et al., 2009; Collins et al., 2013b; Suri et al., 2013) and second (Arleo et al., 2014) trimesters of pregnancy. Application of diagnostic PV estimation in the third trimester of pregnancy has been attempted (de Paula et al., 2008), reporting a significant positive correlation between VOCAL estimated PV and birth weight. No studies have examined the value of third trimester sonographic PV estimates (alone or in combination with EFW as a sonographic FPR) in the prediction of stillbirth or associated APO. The theoretical limitations of third trimester sonographic PV estimation are difficulties in capturing the entire placenta in a single image or volume (leading to variable underestimation of placental size or potentially unreliable extrapolation), and lack of clarity at the maternoplacental interface resulting in inclusion of a variable amount of tissue at this boundary. The placenta (and its maternal boundary) can be easily identified on Magnetic Resonance Imaging (MRI) (Gowland, 2005). Use of MRI is purported to overcome many of the potential limitations of ultrasound in assessment of placental structure and size, but there are no published reports correlating MRI-estimated and true ex vivo PV or PW. It is reported that the intra- and inter-observer reproducibility of MRI-estimated placental volume is higher than for Page 38 of 253 sonographic estimation methods (ICC≥0.98) (Damodaram et al., 2010). No studies of MRI-estimated FPR have been reported. MRI monitoring of placental wellbeing is unlikely to be of immediate routine clinical application. While safe in pregnancy (the only persisting safety concern being that of fetal acoustic damage, for which there is no strong evidence) (Denison et al., 2012), it is expensive with limited availability and lower acceptability compared to ultrasound. Placental diameter, depth and derived factors Quantification of placental biometry (length, width, depth and area) in vivo has been attempted for over three decades (Hoogland et al., 1980; Jauniaux et al., 1994; Toal et al., 2007; Toal et al., 2008a; Toal et al., 2008b; Proctor et al., 2009; McGinty et al., 2012; Schwartz et al., 2012; Suri et al., 2013; Milligan et al., 2014). The longest placental diameter is commonly referred to as placental length, whilst placental width is defined as the longest placental diameter perpendicular to the length (Yampolsky et al., 2013). No consensus has been reached concerning the plane of measurement (mid-placental, maternoplacental or fetoplacental border) or nature (linear or curvilinear) of the measurement (Toal et al., 2007; McGinty et al., 2012; Schwartz et al., 2012; Suri et al., 2013; Milligan et al., 2014). High inter-observer ICCs (>0.92) are reported for placental biometry measurements at midgestation (McGinty et al., 2012; Milligan et al., 2014). However, no studies report validation of these estimates against their ex vivo correlates. First and second trimester placental length positively correlates with gestational age (McGinty et al., 2012; Schwartz et al., 2012; Milligan et al., 2014) and birth weight centile (Suri et al., 2013), being smaller in pregnancies resulting in SGA birth weight (McGinty et al., 2012; Schwartz et al., 2012). Although increased technical difficulty may be anticipated in obtaining these measurements in the third trimester of pregnancy (Deurloo et al., 2007a; Schwartz et al., 2012) potentially resulting in lower reliability of such placental biometric estimates, Page 39 of 253 this might be offset by larger differences in placental size in the third trimester, given that placental biometry is divergent from early gestation. Studies of the relationship of estimated placental depth (by ultrasound or MRI) to APO are conflicting. Depth increases with gestational age (Jauniaux et al., 1994; Karthikeyan et al., 2012; Mathai et al., 2013) and EFW (Karthikeyan et al., 2012; McGinty et al., 2012; Suri et al., 2013), but thickening of the placenta is also associated with stillbirth and FGR (Elchalal et al., 2000; Fisteag-Kiprono et al., 2006; Damodaram et al., 2010; Porat et al., 2013). This may reflect differences between physiological (proportionate to other measures of placental size) and pathological (disproportionate) increase in placental depth and may confound detection of an association between placental depth and SGA birth (McGinty et al., 2012). Single point estimates of placental depth at midgestation demonstrate high inter-observer agreement (correlation coefficient 0.98, N=21) (McGinty et al., 2012) although this may ignore useful information that can be inferred from depth variability, which ex vivo has been correlated with FPR (Yampolsky et al., 2011). Thus, consideration of the proportionality of placental depth (i.e. its relationship to placental diameter) and its variability is likely to improve discrimination between normally and abnormally-thickened placentas. Directly-estimated midtrimester 2D sonographic placental area (Hoogland et al., 1980; Suri et al., 2013), and circumference (Jauniaux et al., 1994) are positively correlated with EFW. Being independent of placental depth (which can falsely inflate placental volume), they may more accurately reflect maternofetal transfer area and possibly identify asymmetric chorion regression (Yampolsky et al., 2008; Alwasel et al., 2010; Alwasel et al., 2013; Cotter et al., 2014). Direct estimation of such indices is precluded later in pregnancy by the frequent inability to obtain a single image of the entire placental disc (Schwartz et al., 2012). Thus, in the third trimester the area and shape of the placental disc can only be quantified from measurements of the placental length and width (“ellipticity” = width / length, estimated area = π*(0.5*length)*(0.5*width)) with potential loss of useful information about the intrauterine environment. Using this technique (ellipticity) Suri et al. (2013) were unable to demonstrate Page 40 of 253 significant correlation of placental shape in the first trimester with eventual birth weight centile. While it is possible that continued divergence of placental diameters throughout gestation might improve predictive value of third trimester placental ellipticity, recent data by Haeussner et al. (2013) failed to demonstrate relationships between birth weight and ex vivo placental shape variability. Umbilical cord structure and insertion eccentricity Whilst the frequency of ex vivo velamentous or marginal umbilical cord insertion is increased in placentas from liveborn FGR infants and stillborn infants, les s evidence exists to support the association of less extreme non-central umbilical cord insertions (which are more frequently seen in RFM (Warrander et al., 2012)) with APO (Biswas and Ghosh, 2008; Pinar and Carpenter, 2010). However, cord eccentricity is associated with lower FPR, and therefore may reflect “inefficiency” of the placenta to support fetal growth (Yampolsky et al., 2008; Yampolsky et al., 2009), perhaps by common aetiology. The best method of quantifying cord eccentricity (ex vivo) has not been established (Figure 2A); there are no reports directly comparing these quantifying techniques. McGinty et al. (2012) sonographically measured the cord insertion to placental margin distance at 18 – 24 weeks gestation but failed to detect a significant relationship with SGA birth weight despite reporting reasonable reproducibility (inter observer correlation coefficient of 0.80, N=21). Although there has not been an assessment of the relative reliabilities of the differing cord eccentricity quantification measures, it may be hypothesised that methods relating insertion site to the nearest placental edge are likely to be more easily applicable and reproducible in vivo than the other techniques as this is a more easily identifiable comparative point in 2D sonography. Other structural considerations of the umbilical cord include cord thickness and coiling index. The umbilical cord diameter and cross-sectional area increase into the third trimester in AGA fetuses, but in FGR the cord is often thinner, principally with a reduction in the contribution of Wharton’s Jelly (Proctor et al., Page 41 of 253 FIGURE 2: Assessment of umbilical cord insertion position and coiling index. Umbilical cord insertion eccentricity can be assessed (A) relative to the placental margin (i) (McGinty et al., 2012), the intersection of the placental length (L) and width (W) (ii) (Pathak et al., 2010) or to the geometric centre of the placenta (iii) (Yampolsky et al., 2009). Umbilical cord coiling index (B) can be assessed as 1/inter-coil distance (Degani et al., 1995) or the coil number / cord length (Strong et al., 1994). 2013). A reduction in Wharton’s Jelly leaves the cord vessels more vulnerable to compression and occlusion. However, as thin umbilical cords are associated with smaller placental size and eccentric cord insertion ex vivo (Proctor et al., 2013), cord diameter or area measurement is unlikely to offer significant additional predictive value over and above that provided by placental size and cord eccentricity assessment. Similarly, the umbilical coiling pattern is associated with histopathological placental disease, chronic fetal vascular occlusion and stillbirth (Ernst et al., 2013). The coiling index has been variously defined (Figure 2B) such that methodologies must be carefully reviewed prior to interpretation of reported relationships between low or high “coiling index” and Page 42 of 253 outcomes of interest. There is controversy regarding the associations of cord coiling with stillbirth, emergency delivery for fetal distress and FGR (de Laat et al., 2005), therefore the discriminative value of the umbilical coiling index in predicting composite adverse outcomes associated with both hypo- and hypercoiling is likely to be clinically limited (Jessop et al., 2014) and will not be considered further. Placental texture The first placental texture examinations (thought to reflect placental calcification) were described by Grannum et al. (1979) but do not correlate with subsequent stereological assessments of the ex vivo placenta (Yin et al., 2009). Despite this, a modest increase in the incidence of low or SGA birth weight (OR 3) and neonatal death (OR 4.5) is reported in pregnancies with Grannum grade III placentas at 36 weeks (McKenna et al., 2005; Chen et al., 2012) and reporting of the placental grading at 34-36 weeks resulted in significant reductions antepartum stillbirth in a single randomised controlled trial (N=2,000) (Proud and Grant, 1987). However, the technique displays low and variable interobserver reliability (Kappa 0.24 – 0.69, N=55 (Sau et al., 2004) and 0.34, N=90 (Moran et al., 2011)) and is therefore unsuitable for general use. Focal “lesions” of the placenta known as echogenic cystic lesions (ECLs) have been described, representing areas of villous thrombosis (Proctor et al., 2010). They are associated with poor prognosis in fetuses with severe, early-onset FGR (Viero et al., 2004). In “high risk” pregnancies the absence of ECLs as part of a multi-modal placental assessment protocol at midtrimester was associated with reduced odds of both SGA birth weight (OR 0.20) and stillbirth (OR 0.05) (Toal et al., 2007). However, the individual predictive value of ECLs is not discussed (Toal et al., 2007; Costa et al., 2008; Toal et al., 2008a; Toal et al., 2008b; Proctor et al., 2009). As ECLs are reported in only 8-27% of “high risk pregnancies”, the specificity and sensitivity will not be clinically useful in the general population. Page 43 of 253 MRI signal intensity of placental tissue reduces with gestation, presumed due to loss of placental tissue density in the process of normal placental maturation (Blaicher et al., 2006; Wright et al., 2011), but not in FGR pregnancies (Damodaram et al., 2010). T2 (spin-spin) relaxation times correlate to the volume and ratio of fibrin deposition within villous tissue (Wright et al., 2011), which is known to be increased in FGR and stillbirth (Pinar et al., 2014). An increased (and highly reproducible, ICC≥0.93 (Damodaram et al., 2010)) percentage of abnormal signal intensity placental tissue is also seen in earlyonset pathologies that lead to stillbirth (Messerschmidt et al., 2011). This is purported to represent maternal and fetal vascular histological lesio ns (Linduska et al., 2009). However, this texture measure also correlated with the degree of fetoplacental Doppler abnormality, and is therefore unlikely to provide significant predictive benefit independent of placental vascular assessment. Placental vascularity Doppler ultrasound allows assessment of the rate and direction of blood flow in vessels. By the ratios of flow velocities at different points in the cardiac cycle impedance to blood flow can be inferred. 2D arterial Doppler flow velocity waveforms are commonly described in terms of the peak systolic flow velocity (PSV), ratio of PSV to minimum end diastolic flow velocity (EDF) (SD ratio), ratio of PSV-EDF difference to the PSV (resistance index, RI) or ratio of PSV -EDF difference to the average flow velocity across the cardiac cycle (Pulsatility index, PI) (Figure 3). In an electrical circuit model of impedance to flow within the fetoplacental circulation, it is shown that the number of downstream branches (vessels) within the circuit is the primary determinant of impedance, rather than vessel calibre, fetal heart rate or blood pressure (Thompson and Trudinger, 1990). Thus impedance quantified from 2D Doppler waveforms of the umbilical artery are likely to indirectly assess villous vascularity, with the gestation dependent reduction in UAD impedance (Acharya et al., 2005b) reflecting the expansion of the placental vascular tree during maturation (Jackson et al., 1992). If this Page 44 of 253 FIGURE 3: Quantification of flow and resistance in arteries by Doppler waveform analysis. Impedance to flow can be quantified as the Pulsatility index (PI), resistance index (RI) or systolic-diastolic ratio (SD), calculated according to ratios of the peak systolic velocity (PSV), end diastolic flow velocity (EDF) and time average maximum velocity (TA). expansion does not occur, as in early-onset FGR (Kingdom et al., 1997), basal impedance is higher and will rise sharply with superadded luminal occlusion. Conversely, in late-onset FGR, the placental vascular tree is more adequately expanded, and large degrees of luminal occlusion are required before impedance rises significantly. In animal models, approximately one third of the placental vasculature must be embolised before frank abnormality of the UAD waveform is seen (Morrow et al., 1989). Therefore, the sensitivity of 2D fetoplacental Doppler for the detection of decreased villous vascularity might be improved by Doppler sampling at sites closer to the placental microvasculature, where the downstream placental tree is proportionately smaller. This will be considered in more detail in the following section. The technique of 3D power Doppler (3DPD) is proposed to assess tissue vascularity by calculation of the Vascular Index (VI; the proportion of voxels in an area of interest demonstrating movement), Flow Index (FI; the mean velocity of that movement) and the Vascular Flow Index (VFI; the combination of both VI and FI) (Pairleitner et al., 1999). A single study has examined the correlation between 3DPD indices and the number of fetal capillaries per villus on chorionic villous samples (N=23, all chromosomally normal) reporting higher correlation Page 45 of 253 coefficients (0.48–0.56) than might be expected given the high proportion of placental blood flow accounted for by the maternal circulation (Rizzo et al., 2011). Restriction of the area of interest to “sonobiopsies” sited within the placental body, away from the basal plate, has been attempted to remove the contribution of maternoplacental blood flow from 3DPD indices (Jones et al., 2011) but is associated with significant variability and has not been validated. There are early indications that placental 3DPD indices are lower both at the time of (Pomorski et al., 2012), and before (Bozkurt et al., 2010; Hafner et al., 2010), the clinical diagnosis of FGR. However, the method is frequently criticised for its poor reliability (Martins et al., 2012) induced by the wide range of patientand machine-related factors that affect 3DPD indices (Raine-Fenning et al., 2008a; Raine-Fenning et al., 2008b). This variability hampers comparison between readings in different individuals or in the same individual across time. Attempts have been made to standardise the technique such as use of sub -noise gain (Collins et al., 2012c) or normalisation to the 3DPD values obtained from a near-by loop of umbilical cord (Welsh et al., 2012a), but consensus has not been reached. Similarly placental MRI examination is not yet able to offer reliable assessment of placental vascularity; T1 (spin-lattice) and T2 relaxation times of placental tissue bear no significant relationship to maternal- or fetal-component villous vascular volume (Wright et al., 2011). Thus, neither technique is yet suitable for widespread implementation in assessment of placental vascularity. Placental function Placental vascular function Fetoplacental artery Doppler Despite evidence of ex vivo placental arterial dysfunction in FGR (Mills et al., 2005), controversy exists over whether UAD impedance indices directly relate to placental arterial function (Mills et al., 2005; Wareing et al., 2005). This may be because the functional differences are relatively subtle, less important determinants of impedance than branching of the arterial tree, because the UAD Page 46 of 253 measures aspects of structure and function that are unaltered in the particular pathology of late-onset FGR, or because the variability of these measures exceeds the magnitude of difference in impedance between healthy and unhealthy placentas. To aid interpretation (and optimisation) of UAD impedance quantification it is important to understand what the UAD is reflecting in order to determine the appropriate clinical response to its result The traditional UAD sampling site is a free floating loop in the middle third of the umbilical cord length (Bhide et al., 2013), although researchers have obtained and quantified Doppler waveforms from alternative sites including the abdominal and placental insertion points of the umbilical artery (Sonesson et al., 1993; Acharya et al., 2005a; Figueras et al., 2006; Khare et al., 2006; Cohen et al., 2014). It has been proposed that use of fixed-point sampling sites might reduce the variability of impedance estimates from the UAD with initial promising results (Figueras et al., 2006). Alternatively, impedance to flow can be quantified in arteries closer to the purported site of increased resistance (the placental microvasculature) including chorionic plate arteries (CPAs) and intraplacental arteries (IPAs) (Hsieh et al., 1991; Kirkinen et al., 1994; Rotmensch et al., 1994; Jaffe, 1996; Jaffe and Woods, 1996; Lacin et al., 1996; Yagel et al., 1999a; Yagel et al., 1999b; Mu et al., 2002; Haberman et al., 2004). Whilst the raw impedance indices at these sites failed to differ between pregnancies with normal and adverse outcome (Hsieh et al., 1991; Rotmensch et al., 1994), the relative impedence at these sites (compared with other vessels) was predictive of outcome (Jaffe and Woods, 1996; Haberman et al., 2004). Despite this, Doppler surveillance from these vessels has not been adopted in clinical practice. As a simple technique, without intensive training and resource requirements, adaptations of this technique deserve further consideration, alongside validation studies to establish their impedance indices truly reflect in terms of placental structure and function. Spatio-temporal imaging correlation allows the user to compare relative changes in the previously discussed 3DPD indices across the cardiac cycle (so-called fourdimensional power Doppler (4DPD)) (Kudla and Alcazar, 2012; Welsh et al., Page 47 of 253 2012b). As these indices are being compared to each other (acquired with constant patient and machine variables) they act as an internal comparator and this may offer some means of reducing patient-factor related variability (Kudla and Alcazar, 2012; Miyague et al., 2013), although reliability remains borderline at ICCs of 0.70–0.80 (Kudla and Alcazar, 2012; Welsh et al., 2012b). The technique has not been used in published studies to predict APO. As for 3DPD, 4DPD requires significant work to standardise the technique and delineate the biological correlations of 4DPD indices before it could be employed clinically. Examination of umbilical venous flow has been proposed to improve detection of increased subclinical intraplacental vascular impedance (Boito et al., 2002; Tchirikov et al., 2002; Tchirikov et al., 2009). Umbilical venous flow rate <20 th centile is associated with increased risk of intrapartum fetal distress in AGA “low risk” pregnancies (OR 1.7), but the assessment of umbilical arterial flow impedance, particularly in relation to cerebral blood flow (further discussed later in this chapter), was more predictive of such APO (Prior et al., 2014b). Therefore umbilical venous flow will not be considered further. Maternoplacental arterial Doppler Given the increased incidence of maternoplacental underperfusion lesions in placentas of liveborn FGR (Parra-Saavedra et al., 2014c) and stillborn infants (Kidron et al., 2009), researchers have begun to investigate the potential of the UtAD waveform as a screening tool for APO. The UtAD is proposed to reflect the efficiency of extravillous trophoblast migration and remodelling of the maternal spiral arteries (Lin et al., 1995; Prefumo et al., 2004). Pagani et al. (2014a) describe 1.5-fold increased odds of stillbirth per UtAD PI multiple of the median (MoM) in a general population of N=17,649 pregnancies with a 0.3% stillbirth rate. Furthermore, step-wise reduction in the risk of FGR has been described in pregnancies where high resistance UtAD impedance resolves by the early second trimester (Gomez et al., 2008), late second trimester (Campbell et al., 2000; Kurdi et al., 2004), early third trimester (Ghi et al., 2010) or not at all (Maroni et al., 2011). Together, this data suggest that third trimester UtAD impedance Page 48 of 253 assessment may assist identification of fetuses at risk of stillbirth (particularly in terms of placentally-mediated stillbirth). Strategies have been proposed to increase the sensitivity of maternoplacental vascular assessment. 2D Doppler of the spiral arteries at their entry to the intervillous space (so-called “arterial jets”) (Jurkovic et al., 1991; Matijevic et al., 1995; Murakoshi et al., 1996; Kurjak et al., 1997; Kurjak and Kupesic, 1997; Deurloo et al., 2007b; Ozkaya et al., 2007; Collins et al., 2012a; Collins et al., 2012b) may reflect subtle increases in downstream impedance more sensitively than the UtAD. However studies do not consistently report significant differences in 2D spiral artery impedance between pregnancies destined for NPO or APO (Deurloo et al., 2007b; Ozkaya et al., 2007). Furthermore, gestation-related loss of pulsatility from these spiral artery jets (Collins et al., 2012a) is likely to result in a high failure rate of spiral artery impedance estimation in the third trimester, limiting its usefulness as a clinical test. 3DPD of the sub-placental myometrial layer is a further technique proposed to reflect spiral artery conversion (Dar et al., 2010; Hafner et al., 2010). A subtly lower “myometrial vascular index” is reported in pregnancies resulting in SGA birth weight (AGA 32 vs. SGA 30 arbitrary units) but has not been statistically tested (Hafner et al., 2010). The same concerns regarding the reproducibility of these 3DPD indices persist with this technique. Finally, reduced uterine artery flow rate per kilogram of birth weight is reported in SGA infants (Rigano et al., 2010; Ferrazzi et al., 2011). Attempts to quantify uterine arterial flow using ultrasound or MRI are thwarted by high variability in mean flow rate estimates (312cm/s (Palmer et al., 1992) vs. 970cm/s (Konje et al., 2001) at 36-38 weeks gestation), with additional complication for MRI assessment introduced by variable vascular anatomy (Pates et al., 2010). As uterine flow volume is correlated with UtAD PI (Rigano et al., 2010), its estimation is unlikely to offer additional predictive benefit over the quantification of mean UtAD PI alone. Page 49 of 253 Fetal blood redistribution The phenomenon of preferential blood diversion to the essential organs is thought to prioritise delivery of oxygenated and nutrient rich blood to the fetal brain and is termed brachiocephalic blood diversion. In severe early-onset FGR it is described that in 75% of infants predictable fetoplacental arterial Doppler deterioration (including the umbilical and middle cerebral arteries) preceded deterioration in the biophysical profile (Baschat et al., 2001) and was associated with fetal hypoxia (Baschat et al., 2000), thus assisting decision-making regarding delivery. Even in the absence of frankly abnormal MCA or umbilical artery Doppler impedance, brachiocephalic blood diversion can be detected in the divergence of impedance in the brain (MCA) and placental (umbilical) circulations, often referred to as the cerebroumbilical (CU) ratio (Bahado-Singh et al., 1999; Ozeren et al., 1999; Baschat and Gembruch, 2003; Bano et al., 2010). Recently the CU ratio has been demonstrated to be clinically useful in detecting AGA fetuses with subclinical acute and chronic placental compromise (Prior et al., 2013; Prior et al., 2014a); pregnancies with CU ratios <10 th centile at the beginning of labour are six times more likely to require emergency caesarean delivery for non-reassuring fetal status and five times more likely to experience fetal distress during labour compared with those with CU ratio >10th centile. However, its benefit in predicting umbilical artery acidaemia and delivery for suspected fetal compromise has been questioned in post-mature pregnancies (D'Antonio et al., 2013). There remain technical considerations in applying the CU ratio in the third trimester of pregnancy. Firstly, significant differences exist in published reference ranges (particularly for the RI) of the MCA Doppler (Kurmanavicius et al., 1997; Bahlmann et al., 2002), suggesting that technical considerations in its performance (such as degree of pressure on the fetal head, presence or absence of fetal activity, ultrasound machine settings) can materially influence recorded impedance to flow in the MCA. Secondly, such reference ranges for MCA Doppler parameters are frequently based on small, cross-sectional studies (Ebbing et al., 2007) and rarely including fetuses of gestations above 40 weeks (Palacio et al., 2004). Finally, fetal head engagement in the late third trimester of pregnancy Page 50 of 253 may impair acquisition of MCA Doppler waveforms, although in a series of N=400 early-labour term pregnancies no mention is made of MCA acquisition failures (Prior et al., 2013). Comparison of cerebral impedance in the MCA or alternative cerebral arteries to flow against impedance at alternative sites in the placental arterial circulation may improve the detection of brachiocephalic blood diversion. It has been described that vasodilation can be detected in the anterior cerebral artery (ACA) (Dubiel et al., 2002; Figueroa-Diesel et al., 2007; Cruz-Martinez et al., 2010) or posterior cerebral arteries (PCA) (Figueroa-Diesel et al., 2007; HernandezAndrade et al., 2008) before the MCA, although the clinical predictive value of these cerebral arterial Doppler impedances (alone or in relation to placental arterial resistance) have not yet been compared to the MCA in terms of clinical predictive value. Acquisition of the ACA and PCA Doppler waveforms is more challenging than that of the MCA (Benavides-Serralde et al., 2010) and is therefore likely to be less readily applicable. 3DPD assessment of fetal (regional) cerebral perfusion has been attempted (Nardozza et al., 2009; Rossi et al., 2011; Milani et al., 2012), with early indications that this technique may detect cerebral vasodilation earlier than the CU ratio (Cruz-Martinez et al., 2010), but clinical utility remains limited by the general concerns of 3DPD reproducibility. Brachiocephalic blood diversion is achieved through a series of “shunts”: the ductus venosus (DV; diverts blood away from the liver), foramen ovale (directs blood from the DV to the fetal left ventricle), ductus arteriosus (directs blood from the left ventricle to the cerebral and coronary circulations) and the umbilical arteries (diverts blood to the placenta for re-oxygenation and nutrient/waste exchange). Each shunt is accompanied by one or more “watershed” zones in which fetal Doppler waveform analysis may identify fetal blood redistribution (Baschat, 2006). For example, the directionality of blood flow in the aortic isthmus may reverse in situations of significant brachiocephalic blood diversion and predict adverse perinatal outcome in severe early-onset FGR (Del Rio et al., 2008). However, researchers have failed to demonstrate significant differences in the aortic isthmus Doppler between SGA and AGA Page 51 of 253 fetuses in the absence of concurrent UAD abnormality (Kennelly et al., 2012). In the coronary arteries pulsatile flow becomes detectable earlier in early-onset FGR (28 weeks gestation) than it does in control pregnancies (34 weeks gestation), with slightly higher diastolic blood flow velocities (Baschat et al., 2003) but abnormal coronary artery blood flow represents late-stage fetal decompensation and is therefore unlikely to significantly improve detection of fetal compromise before traditional markers (Baschat et al., 1997). Other fetal arteries may show evidence of increased resistance, designed to divert blood away from non-essential organs such as the gut (coeliac axis and mesenteric arteries), kidney (renal arteries) and the lower body (femoral and iliac arteries) as summarised by Baschat (2004). Doppler assessment of these arteries and of the coronary and aortic circulations requires specialist sonographic training. Thus, it is unlikely that these techniques would be easily and reliably applied in the third trimester of pregnancy at present. Assessment of the DV Doppler waveform deserves further consideration. Severe early-onset FGR fetuses display greater diversion of placental venous return along the DV (Bellotti et al., 2004). In a similar manner to the CU resistance ratio, this venous flow ratio may identify subclinical blood diversion before FGR is otherwise clinically detectable, but the normal range and predictive value of this DV / umbilical vein flow ratio has not been examined in AGA fetuses. This technique requires further optimisation before implementation in clinical care, particularly in relation to flow rate estimation in a prism-shaped vessel. Furthermore DV Doppler waveform acquisition is unreliable in the presence of fetal breathing due to negative intrathoracic pressure (which occurs in 30% of the time in term fetuses (Patrick et al., 1980)) and fetal movement impairs waveform acquisition. Thus, it is not ready for widespread clinical implementation. Placental endocrine function The placenta is an active secretory organ, producing a range of hormones and other substances essential for the regulation of fetal growth and the function and Page 52 of 253 maintenance of the placenta. In the absence of aneuploidy or structural defects, pregnancies with abnormally high or low maternal circulation concentrations of fetoplacental hormones are at increased risk of APO including fetal loss (late miscarriage and stillbirth), FGR and other conditions of placental origin (including placental abruption and hypertensive disorders of pregnancy) (Dugoff et al., 2005; Gagnon et al., 2008; Poon et al., 2008; Lao et al., 2009). Much of this data has been derived from retrospective analysis of maternal serum collected in the first and second trimesters of pregnancy for the purposes of Down’s syndrome screening, although some third trimester studies have been reported (Chappell et al., 2002; Levine et al., 2004; Romero et al., 2008; Romero et al., 2010; Benton et al., 2012; Chappell et al., 2013). Such “biochemical tests of placental function” have not been systematically applied to the prediction of APO in the general population as their test performance characteristics are largely too poor to support their use as a standalone test. Indeed the Cochrane review of the same, failed to find evidence to support their implementation in pregnancy screening to prevent stillbirth , and remarks that consideration of such testing is only of “historical interest” (Neilson, 2012). This conclusion was based upon the findings of a single study of third trimester oestriol measurement (N=622), and was underpowered to detect small, but significant differences in stillbirth rates. The study (and subsequent “meta-analysis”) power was further reduced by the arbitrary definition of abnormal oestriol levels and failure to specify post-test management protocols for study participants (Duenhoelter et al., 1976). Despite this, interest in placental endocrine testing has begun to enjoy a resurgence of interest with the measurement of third trimester placental growth factor (PlGF) and human placental lactogen (hPL) in research contexts (Dutton et al., 2012; Chappell et al., 2013; Heazell et al., 2013) and first trimester PAPP-A in the clinical prediction of SGA birth weight risk (Robson et al., 2014). Yet what these circulating hormone concentrations reflect in terms of placental structure and function is largely unknown, impairing any potential benefit that might be gained from their measurement. Page 53 of 253 However, this aspect of placental monitoring deserves further interrogation, particularly as an element of multi-modal placental screening, as it is relatively non-invasive and requires no additional training, relying upon facilities already present in modern healthcare systems. Hormone selection is important to optimise the predictive value of the test. For example placental-dependency of the production of a given hormone is essential for it to function as a potential placental structure or function marker, thus oestriol concentr ations (being dependent on fetal adrenal function (Ishimoto and Jaffe, 2011)) are more likely to reflect fetal, rather than placental, wellbeing. Other prospective markers (such as Placental Protein-13 and Placental Growth Hormone) appear to primarily predict development of preeclampsia rather than SGA or stillbirth (Papadopoulou et al., 2006; Chafetz et al., 2007; Mittal et al., 2007; Cowans et al., 2008; Huppertz et al., 2008; Schneuer et al., 2012; Sifakis et al., 2012). Recent studies have demonstrated reduced predictive value of potentially useful markers (in early pregnancy) when measured in the third trimester (e.g. PAPPA) (Dutton et al., 2012) or are difficult to study due to instability during storage (e.g. disintegrin and metalloproteinase-12) (Cowans et al., 2010). However, a number of potential placentally-derived hormones deserve further consideration in the late pregnancy. Human placental lactogen hPL promotes fetal growth by increasing fetal nutrient availability (Newbern and Freemark, 2011) and release of insulin like growth factor-1 from the fetal liver. Maternal circulating hPL concentrations increase with fetal weight (Sorensen et al., 2000), placental weight (Gordon et al., 1977; Furuhashi et al., 1984a; Dutton et al., 2012) and liver volume (Murao, 1991), and thus may reflect fetoplacental growth. Whilst hPL is not currently routinely measured in modern obstetric practice, prior to the introduction of CTG and ultrasound monitoring of pregnancies it was a leading candidate serum marker for prediction of APO (Letchworth et al., 1978; Isouard, 1979; Morrison et al., 1980), predicting stillbirth and acute fetal compromise with 87% sensitivity and 95% specificity at levels <10th population centile (Morrison et al., 1980). Further development of Page 54 of 253 this test was neglected following the introduction of newer fetal monitoring techniques. Progesterone Whilst progesterone induces myometrial quiescence (Anderson et al., 2009), its function in human pregnancy and labour is unknown (Tulchinsky et al., 1972b). A significant positive relationship between maternal circulating progesterone concentration and placental, but not fetal, weight has been described (Spellacy et al., 1975; Dutton et al., 2012) and alteration of progesterone metabolism is seen in metabolomic studies of maternal serum in pregnancies resulting in APO, particularly SGA birth, (Dutton et al., 2012; Heazell et al., 2012). As placental growth restriction is described to occur prior to FGR, analysis of progesterone as a marker of placental size, particularly in comparison to fetal size, may offer potential to predict FGR early in its pathway. Human chorionic gonadotrophin hCG is a glycoprotein hormone secreted by the syncytiotrophoblast (Midgley and Pierce, 1962). It promotes the differentiation of cytotrophoblasts into syncytiotrophoblast (Shi et al., 1993) and thus influences the maintenance and function of the maternal-fetal barrier. It is one of the most studied hormones in terms of predicting APO: both low first trimester and high second trimester maternal serum hCG concentrations are associated with APO, particularly FGR and stillbirth (Benn et al., 1996; Yaron et al., 1999; Duric et al., 2003; Spencer et al., 2006; Tavor et al., 2007; Gagnon et al., 2008). Placental villous tissue hCG release is oxygen sensitive (Crocker et al., 2004), suggesting that maternal serum hCG concentrations might be altered by changes in the intrauterine environment. Yet its predictive value in later pregnancy has not been extensively investigated (Dutton et al., 2012), potentially due to fetal gender-specific hCG concentration profiles in late pregnancy (Boroditsky et al., 1975; Chatterjee et al., 1976; Furuhashi et al., 1984a; Chen et al., 1993; Steier, 1999; Steier and Freedman, 2004; Edelstam et al., 2007) and the absence of reported relationship between Page 55 of 253 serum hCG concentration and either placental or birth weight (Spellacy et al., 1975). Placental growth factor and soluble fms-like tyrosine kinase-1 PlGF is a member of the vascular endothelial growth factor (VEGF) family with pro-angiogenic effects, predominantly promotion of non-branching angiogenesis in the second half of pregnancy (Kaufmann et al., 2004). Third trimester maternal serum PlGF concentrations are lower in pregnancies resulting in SGA birth weight (Romero et al., 2008), with even lower concentrations associated with extreme “smallness” which is more likely to be true FGR (Benton et al., 2012). As reduced villous tissue PlGF content is reported in placentas from SGA births (Shibata et al., 2005), the relationship between circulating PlGF concentration and placentally-related pregnancy outcome is not simply accounted for by placental size. Circulating PlGF exerts its biological effect via a membrane-bound receptor known as fms-like tyrosine kinase-1 (Flt-1). Cells are protected from excess Flt-1 activation by the binding of PlGF to a soluble form of this receptor (sFlt-1), which lacks the transmembrane and tyrosine kinase domains and therefore renders bound PlGF functionally inactive (Kendall et al., 1996). Indeed, it is proposed that the true benefit of PlGF measurement in the third trimester results from the indirect measurement of sFlt-1 “effect”, particularly in the context of preeclampsia (Sela et al., 2008). Thus, the balance between PlGF and sFlt-1 is potentially more important than the concentration of either factor alone (Chaiworapongsa et al., 2013). However, in the context of prediction of SGA birth weight, third trimester maternal circulating sFlt-1 concentrations do not differ from those of AGA pregnancies (Romero et al., 2008), suggesting a primary pathogenic role for low PlGF concentrations. Page 56 of 253 Placental nutrient transport, metabolism and oxygenation Nutrient transport and metabolism Whilst ex vivo placental transport has been studied in some detail (Jansson et al., 1993; Glazier et al., 1997; Jansson et al., 1998; Norberg et al., 1998; Paolini et al., 2001; Cetin et al., 2002; Jansson et al., 2002; Johansson et al., 2002; Johansson et al., 2003; Strid et al., 2003; Tagliabue and Del Sole, 2014), in vivo human placental nutrient transport and metabolism remains poorly studied (Ceitin, 2003). Recently, the technique of positron emission tomography scanning has enabled monitoring of the distribution of an isotope around the body and is currently employed in the staging and monitoring of cancer (Tagliabue and Del Sole, 2014). Using this technique (in conjunction with spatiotemporal imaging), placental transfer of molecules such as methyl- α-aminoisobutyric acid (a substrate of system A) could be examined (Abramowicz and Sheiner, 2007). If an isotope such as fluorine-18 labelled fluorodeoxyglucose is used, it can also provide information on the metabolic state of tissues, in this instance the placenta. However, the technique has not been applied in pregnancy and it is unlikely that such radiolabelled isotope-based studies would be incorporated into routine obstetric care due to obvious concerns regarding toxicity. It is also possible to obtain information regarding the metabolic environment of the placenta by using proton resonance spectroscopy, an MRI technique that detects differences in the MRI signal caused by the environment of the proton in vivo. Differences have been observed in placentas of FGR fetuses in preliminary studies using this technique (Denison et al., 2012). A related technique, phosphorous-31 nuclear MRI, examines the in vivo environment of phosphorous, and thus may provide a non-invasive method for examining tissue content of high-energy bonds, for example adenosine triphosphate (Alves et al., 2012). This technique requires further optimisation before it can be applied to the examination of human pregnancy. The clinical utility of both techniques remains limited by the previously discussed barriers to use of MRI as a primary assessment tool in late pregnancy. Page 57 of 253 Placental oxygenation and gaseous transfer A number of adaptations of the maternal and fetal placental circulations optimise oxygen transfer across the placenta, including the development of vasculosyncytial membranes to reduce transport distance and the higher concentration, and oxyygen affinity, of fetal haemoglobin compared with circulating “adult” haemoglobin molecules (Carter, 2009). Yet the FGR fetus is frequently described as relatively hypoxic, as evidenced by lower umbilical arterial pH (Tagliabue and Del Sole, 2014). A number of techniques have been applied in preliminary studies of in vivo placental oxygenation assessment: transabdominal placental absorbance of infrared light (Kakogawa et al., 2005), and oxygen enhanced or blood oxygen level dependent contrast MRI (Huen et al., 2013). Use of hypoxia sensitive probes in conjunction with positron emission tomography (Krohn et al., 2008) has not been applied in pregnancy and is limited by the previously discussed safety concerns of positron emission tomography in pregnancy, compounded by exacerbation of cellular hypoxia by the persistence of hypoxia sensitive probes. However, as the oxygenation status of the placenta (as opposed to fetus) in FGR pregnancies has been proposed to be either hypoxic or relatively hyperoxic according to the underlying pathology (Ahmed et al., 2000), it is unlikely that placental oxygenation assessment would identify additional at-risk fetuses, although it may provide information about the specific phenotype of the disease. In utero placental assessment in reduced fetal movement pregnancies To date, in vivo placental structure, nutrient transport and placental oxygenation remain unstudied in RFM pregnancies, although there is preliminary evidence of altered lipid (including progesterone) metabolism in these pregnancies (Heazell et al., 2012). Slightly more is known about in vivo vascular and endocrine assessment of the RFM placenta. Four studies report results of UAD waveform analysis in RFM pregnancies (Dubiel et al., 2002; Korszun et al., 2002; Froen et al., 2008b; Dutton et al., 2012) with rates of frankly abnormal UAD waveforms (impedance >95 th centile or Page 58 of 253 abnormalities of EDF) varying between 0.3-6.7% (Dubiel et al., 2002; Dutton et al., 2012). This low incidence is in keeping with the relationship of UAD abnormalities to early-, but not late-onset placental pathologies. Furthermore, an abnormal first or second trimester UtAD waveform is associated with an increased incidence of RFM, and within the RFM cohort is an independent predictor of SGA birth weight (OR 2.6 for first trimester UtAD impedance >99 th centile and OR 4.1 per second trimester UtAD impedance MoM) (Pagani et al., 2014a; Pagani et al., 2014b). Yet in the only published study of contemporaneous UtAD assessment in RFM pregnancies an abnormal waveform (notching and/or impedance >95th centile) was observed in only 1.4% of all RFM pregnancies (Korszun et al., 2002). No other in utero placental vascular interrogation techniques have been used in the assessment of pregnancies following RFM. There is limited evidence that third trimester maternal circulating placentallyderived hormone concentrations (including hPL, hCG and progesterone) are lower in RFM pregnancies resulting in APO (stillbirth, SGA birth weight, preterm delivery or neonatal intensive care admission (NICU)) (Dutton et al., 2012). After adjustment for confounding associations, only hPL concentrations remained predictive of APO, supporting the findings of Leader et al. (1980) who reported hPL assessment was complementary to fetal movement assessment in prediction of adverse pregnancy outcome (stillbirth, NICU admission, umbilical artery pH ≤7.1, IBC <10). A subsequent pilot randomised controlled trial of hPL -triage (<0.8 MoM) in the management of RFM pregnancies ≥36 weeks gestation demonstrated reduction in the rate of APO (stillbirth, SGA birth weight, umbilical artery acidaemia or NICU admission) from 29% to 12% (Heazell et al., 2013). Maternal circulating PlGF and sFlt-1 concentrations have not yet been assessed in the context of RFM. 1.4. Defining and addressing the research question Further research, both basic science and clinical, is required to understand the causes of stillbirth and RFM and their relation to placental dysfunction, and the biological basis of placental biomarkers, in order to guide appropriate Page 59 of 253 management of pregnancies following presentation with RFM. These were identified as key priorities for reduction of stillbirth rates in high-income countries by the International Stillbirth Alliance in 2011 (Flenady et al. 2011b). The ability to accurately identify placental dysfunction prior to the onset of clinically recognisable FGR or fetal compromise would allow optimisation of pregnancy outcome by determining the appropriate time and place for delivery. Given the diversity of causes of stillbirth and of ex vivo observed placental pathology in late pregnancy, it is unlikely that any single placental marker would display sufficient test characteristics to be used alone in the prediction of APO following RFM. In light of the data regarding ex vivo placental examination and in vivo placental interrogation in RFM pregnancies, incorporation of structural, vascular and endocrine placental markers into a multimodal placental screening process may improve prediction of APO in the RFM population. If antenatal placental structure and function assessment is to prove clinically useful, it is important to ensure that: i) The tests are targeted to aspects of placental anatomy and physiology that are aberrant in pregnancies at highest risk of APO including stillbirth. ii) It is understood what aspect of placental structure and function the results reflect and why they relate to APO. iii) The tests are able to identify the at-risk fetus prior to the onset of frank fetal compromise detected by current tests of fetal wellbeing. Based upon the evidence presented here, sonographic size of the placenta, Doppler measures of feto- and maternoplacental blood flow impedance and brachiocephalic blood diversion, and concentrations of placentally-derived substances in the maternal serum are proposed as potentially useful biomarkers for the assessment of placental structure and function in utero that could be readily applied in current high-income countries’ healthcare systems. Biomarkers requiring the use of 3DPD, MRI or administration of (radiolabelled) isotopes are considered unsuitable for immediate application to clinical care in Page 60 of 253 the near future either due to need for further technique development or safety concerns. Page 61 of 253 Hypothesis RFM is a symptom of placental dysfunction which may progress to cause APO including stillbirth; this placental abnormality can be detected antenatally and can be used to improve identification of fetuses at highest-risk of APO. Aims 1. To compare placental structure and function between pregnancies ending in NPO and APO after presentation with RFM. 2. To develop antenatal placental assessment techniques that measure relevant aspects of placental structure and function. 3. To examine the predictive value of placental assessment in RFM. Objectives 1. To examine placental macrostructure, microstructure, placental vascularity and arterial function and endocrine function in ex vivo placentas following maternal report of RFM (Chapter 2). 2. To develop and validate sonographic methods to assess placental volume, length, width, depth and fetoplacental ratio in utero (Chapter 3). 3. To develop and validate sonographic methods to assess placental vascularity and arterial function (Chapter 4). 4. To investigate the relationship between maternal serum concentrations of placentally-derived hormones and placental structure and endocrine function (Chapter 5). 5. To perform fetoplacental assessment in a cohort of RFM pregnancies and identify combinations of measurements that may improve identification of the at-risk fetus (Chapter 6). Page 62 of 253 CHAPTER 2: PLACENTAL FEATURES OF LATE-ONSET ADVERSE PREGNANCY OUTCOME Higgins LE, Addo N, Rey de Castro N, Wareing M, Greenwood SL, Jones RL, Sibley CP, Johnstone ED, Heazell AEP. A manuscript prepared for submission to PLOS ONE Page 63 of 253 2.1 Abstract Objective: Currently, no investigations reliably identify placental dysfunction in late pregnancy. To facilitate the development of such investigations we aimed to identify placental features that differ between normal and adverse outcome in late pregnancy in a group of pregnancies with reduced fetal movement. Methods: Following third trimester presentation with reduced fetal movement (N=100), placental structure ex vivo was measured. Placental function was then assessed in terms of (i) chorionic plate artery agonist responses and lengthtension characteristics using wire myography, and (ii) production and release of placentally-derived hormones (by quantitative polymerase chain reaction and enzyme linked immunosorbant assay of villous tissue and explant conditioned culture medium). Results: Placentas from pregnancies ending in adverse outcome (N=23) were ~25% smaller in weight, volume, length, width and disc area (all p<0.0001) compared with those from normal outcome pregnancies. Villous and trophoblast areas were unchanged, but villous vascular density was reduced (median (interquartile range): normal outcome 13 (12–15) vs. adverse outcome 10 (10– 12) vessels/mm2, p=0.002). Adverse outcome pregnancy placental arteries were relatively insensitive to nitric oxide donated by sodium nitroprusside compared to normal outcome pregnancy placental arteries (50% Effective Concentration 12 (6-24) vs. 30 (19-50) nM, p=0.02). Adverse outcome pregnancy placental tissue contained less human chorionic gonadotrophin (55 (24–102) vs. 20 (11– 50) mIU/mg, p=0.007) and human placental lactogen (27 (9–50) vs. 11 (6–14) mg/mg, p=0.006) and released more soluble fms-like tyrosine kinase-1 (5 (2–15) vs. 21 (13–29) ng/mg, p=0.01) compared with normal outcome pregnancy placental tissue. Conclusion: These data provide a description of the placental phenotype of adverse outcome in late pregnancy. Antenatal tests that accurately reflect elements of this phenotype may improve its prediction. Keywords: Reduced fetal movement, Fetal growth restriction, Stillbirth, Placenta, Phenotype, Biomarker, Adverse pregnancy outcome Page 64 of 253 2.2 Introduction According to the 2010 Europeristat review of perinatal mortality in 31 countries approximately 26% of all stillbirths occurred at or beyond 37 weeks gestation and a further 22% occurred between 32- 36+6 weeks gestation (Zeitlin, 2013). Similar patterns are seen in the United States (MacDorman, 2012) and New Zealand (Ministry of Health, 2012). Critically, stillbirth rates in high-income countries have shown little improvement in recent decades (Flenady et al., 2011b). This may reflect that current antenatal screening strategies (including sonographic fetal size assessment and umbilical artery Doppler) do not reduce perinatal mortality in the general obstetric population (Bricker et al., 2008). A simple and widely employed method to monitor fetal wellbeing is maternal perception of fetal activity (Heazell and Froen, 2008; Flenady et al., 2009). Pregnancies with reduced fetal movement (RFM) have an increased risk of adverse pregnancy outcome (APO), including two- to three-fold increased risk of fetal growth restriction (FGR) and stillbirth (O'Sullivan et al., 2009; Pagani et al., 2014a), with increased incidence of neurodevelopmental problems (Bonifacio et al., 2011; James et al. 2000). RFM is thought to occur as a compensatory response to limited delivery of oxygen and nutrients to the fetus by a dysfunctional placenta (Maulik, 1997; Warrander and Heazell, 2011). In a systematic review and related Delphi exercise, the International Stillbirth Alliance highlighted RFM as a key priority area for preventing stillbirth in high income countries (Flenady et al., 2011b). Meta-analysis suggests that although using specific “alarm limits” and formal counting of fetal movement does not decrease perinatal mortality (Hofmeyr and Novikova, 2012), maternal and professional awareness of the importance of fetal activity (and subsequent investigation to identify fetoplacental compromise) is associated with reductions in overall perinatal mortality and other APOs in study populations (Froen et al., 2008a). In support of the hypothesis that RFM represents fetal compensation to placental dysfunction, placentas from pregnancies complicated by RFM (hereafter referred to as RFM placentas) display a range of pathologies similar to those observed Page 65 of 253 following stillbirth (Warrander et al., 2012; Winje et al., 2012; Girard et al., 2014) (Table 1) in terms of reduced gross placental size, reduced villous vascularity, increased villous thrombosis and infarction and altered maternal circulating placentally-derived hormone concentrations. However, these studies largely focused on structural, rather than functional, changes and have not distinguished between RFM pregnancies with normal pregnancy outcome (NPO) or APO. Warrander et al. (2012) demonstrated evidence of reduced amino acid transporter activity in placentas from APO pregnancies, but to date there has been no assessment of placental vascular or endocrine function according to outcome in RFM. This is important as established clinical investigations to detect placental dysfunction (in terms of Doppler assessments of materno - and fetoplacental blood flow and maternal serum hormone measurements) focus on these facets of placental physiology (Pasupathy et al., 2008). This study tested the hypothesis that the placental phenotype associated with APO differs from women who have an NPO after presentation with RFM. We compared the ex vivo placental biometry, vascularity, and vascular and endocrine function of RFM pregnancies with and without APO. Page 66 of 253 2.3 Materials and methods Ethical approval was received from Greater Manchester North West Research Ethics Committee (11/NW/0650) and all work was conducted in accordance with the Declaration of Helsinki 1975 (revised 2013). Women presenting to St. Mary’s Hospital, Manchester, UK with perceived reduction in fetal activity after 28 weeks gestation in singleton pregnancies were asked to participate in the study. Pregnancies complicated by pre-pregnancy hypertension or diabetes, or known fetal abnormalities were excluded. Written informed consent to examination of placental tissue after delivery was obtained. APO was defined as previously described (Heazell et al., 2013) if any of following criteria were met: stillbirth or neonatal death, small for gestational age (SGA) birth weight (individualised birth weight centile (IBC) <10 derived by Bulk Centile Calculator (UK) 6.7 software, Gestation Network, Birmingham, UK), five minute Apgar score <7, umbilical arterial pH <7.1 or base excess <-10.0 or admission to neonatal intensive care unit within 24 hours of birth. Placentas were collected within one hour of delivery. Unless otherwise stated, all chemical reagents used were supplied by SigmaAldrich (Poole, UK). Placental structure Placentas were trimmed of their extraplacental membranes and umbilical cord, weighed and photographed, chorionic plate facing upward, alongside a scale bar. Their volume was calculated by fluid displacement. Depth was manually measured at the apparent deepest point of the placental body. Using Image ProPlus 6.0 imaging software (Media Cybernetics, Marlow, UK) placental photographs were analysed to establish placental disc area, length (longest diameter), width (longest diameter perpendicular to length) and minimum distance from cord insertion to placental edge (cord distance). Cord distance was divided by placental length to generate the “cord ratio” in order to take into account overall placental size. Page 67 of 253 Villous biopsies (1cm3) from the edge and middle of the placental body and beside the umbilical cord insertion were fixed in 10% neutral buffered formalin (18 hours at 4°C) before paraffin-embedding and sectioning. Tissue was immunoperoxidase stained using mouse monoclonal antibodies against the trophoblast marker cytokeratin 7 (CK7; 0.9μg/ml; Dako, Ely, UK), and the endothelial cell marker cluster of differentiation-31 (CD31) (0.16μg/ml; Dako) with non-immune mouse immunoglobulin G at corresponding concentration as a negative control and imaged as previously described (Warrander et al., 2012). Total villous and trophoblast areas, the number of vessels (CD31 positive structures) and the combined vascular luminal area per field of view were quantified as previously described (Hayward et al., 2011). The mean values for each index across 10 images per tissue section were taken to represent each tissue biopsy, and the median “biopsy” value was taken to represent each placenta. Placental function Vascular function Small (200-500μm) chorionic plate arteries (CPAs) were dissected from the placenta. Arterial segments (N=8 per placenta) were loaded onto wire myographs (610M Danish Myo Technology A/S, Aarhus, Denmark). Tissue baths contained physiological saline solution [119mmol/L NaCl, 25mmol/L NaHCO 3, 4.69mmol/L KCl, 2.4mmol/L MgSO4, 1.6mmol/L CaCl2, 1.18mmol/L KH2PO4, 6.05mmol/L D-glucose, 0.034mmol/L Ethylenediaminetetraacetic acid; pH7.4] bubbled with 5% O2 at 37°C. Resting vessel diameters were measured at 0.9 of L5.1kPa according to the method previously described by Mills et al. (2005), adapted from Wareing et al. (2002). Vessel responses to vasoactive agonists and length-tension characteristics were analysed using Myodata 2.01 software (Myonic Software, Aarhus, Denmark). Vessel responses to vasoactive agonists were assessed according to the protocol described by Mills et al. (2005), to construct concentration-response curves to Page 68 of 253 the thromboxane A2 mimetic U46619 (10-10M – 10-5.7M; Calbiochem, EMD Millipore, Billerica, USA) and the nitric oxide donor sodium nitroprusside (SNP: 10-10M – 10-4M). Constriction responses were expressed as the maximal pressure generated (V max; kPa), the concentration of U46619 at which vessel segments achieved 50% of their maximal response (effective concentration, EC 50; nM) and the area under the concentration-response curve (AUC; arbitrary units). Relaxation responses were similarly expressed as V max (residual constriction percentage, normalised to time control), EC50 and AUC. Arterial length-tension characteristics were determined according to the protocol described by Wareing et al. (2002). Passive tension accumulation was quantified by Tau (the time to doubling of tension, expressed as 1/k, where k is the rate constant of the exponential curve) whilst active tension generation was quantified by AUC. Individual vessels’ peak active tension (PAT, mN/mm 2) values were recorded, along with the diameter at which this was achieved (DiamPAT, % of normalised diameter). Endocrine function To assess placental endocrine function, the production and release of human chorionic gonadotrophin (hCG), human placental lactogen (hPL), progesterone, placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1) were assessed in placental villous tissue. These hormones were selected due to previous work demonstrating altered concentrations in antenatal maternal serum in pregnancies complicated by APO following RFM (Dutton et al., 2012; Heazell et al., 2012), FGR (Benton et al., 2012) or prior to stillbirth (Gagnon et al., 2008; Romero et al., 2010). Fresh villous tissue was washed in phosphate buffered saline and divided for RNA extraction (treated with RNAsave (Geneflow, Lichfield, UK) at 4°C for 18 hours), tissue hormone content analysis (lysed in 1.5ml distilled water at room temperature for 18 hours) or assessment of hormone release into culture media (CM) in a villous explant culture model as described previously; explants were maintained on Netwell permeable supports (Corning, via Sigma-Aldrich) at the liquid-gas interface in culture media [in 1L: 100ml CMRL-1066 (Gibco, Life Technologies, Paisley, UK), 10% heat-inactivated Page 69 of 253 fetal bovine serum, 100mg streptomycin sulphate, L-glutamine, 50IU/ml penicillin, 50μg/ml 50μg/ml gentamicin, 100μg/ml hydrocortisone, 0.1μg/ml retinoic acid, 1μg/ml insulin] under conditions of 6% O 2/5% CO2 and 37°C for seven days (Siman et al., 2001). RNA was extracted from fresh villous tissue using a mirVana miRNA isolation kit followed by removal of genomic DNA using a TURBO DNA-freeTM kit (both Ambion, Austin, Texas, USA). Nucleic acid concentration and contamination were assessed by spectroscopy using the NanoDrop 2000C (Thermo Fisher Scientific, Wilmington, USA). Reverse transcription (RT) was performed in triplicate using an AffinityScript Multi-temperature RT kit (Agilent, Santa Clara, USA) in an MX3000 thermocycler (Stratagene, La Jolla, USA). Real time quantitative polymerase chain reaction (qPCR) was then performed on RT triplicates using primers specific to TATA-box binding protein (Supplementary table 1) and Brilliant III Ultra-fast SYBR® QPCR mastermix (Agilent) with annealing at 60°C, followed by dissociation curve analysis. All qPCRs had efficiencies of 85 -105%. The cycle threshold (Ct) of each sample was calculated and gene expression calculated as 2-ΔCt. Triplicates with a coefficient of variance (CoV) <25% were pooled for further qPCR analysis. Subsequently qPCR was performed according to the same protocol, on pooled triplicates using primers for genes encoding TATA-box binding protein, human chorionic gonadotrophin (hCG), human placental lactogen (hPL), placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1) along with CYP11A1 (the gene encoding CYP450scc, the rate-limiting enzyme of progesterone synthesis (Tuckey and Sadleir, 1999)) (Supplementary table 1). TATA-box binding protein expression between NPO and APO placentas was compared to ensure it was an appropriate housekeeping gene (p>0.05). Relative mRNA expression levels were calculated relative to expression of TATA-box binding protein. Hormone concentrations in villous tissue lysate and explant conditioned media (CM) (discarded after the first 24 hours of culture then collected daily) were quantified in duplicate using commercially available colorimetric enzyme-linked immunosorbant assay (ELISA) kits (Supplementary Table 2). The optical density Page 70 of 253 (OD) of the well contents was read by a multiplate reader (BMG labtech, Aylesbury, UK) at a pre-specified wavelength (See Supplementary table 2). Sample content was quantified by comparison to a standard curve. The average concentration and CoV of sample duplicates were calculated; those with CoV above 10% were repeated. The protein content of tissue pellets and villous explants were calculated by dissolution in 4ml 0.3M sodium hydroxide and quantification with Bio-Rad Protein Assay (Bio-Rad Laboratories, Hempstead, UK). Lysate and CM hormone concentrations were then normalised to the concentration of a quality control run in every assay, and adjusted to tissue protein content. Statistical analysis Based on previous observations of placental arterial function, trophoblast area and villous vascularity in FGR (Daayana et al., 2004; Mills et al., 2005; Junaid et al., 2014) we calculated that a minimum of seven placentas per group would be required to observe similar magnitude differences in placental structure and function between NPO and APO pregnancies (power 80%, significance p<0.05). As both functional assessment techniques require fresh placental tissue, simultaneous use of placentas for both was not possible. Therefore, based on a conservative expected APO rate of at least 15% (Dutton et al., 2012) we calculated that 100 participants would be required to provide adequate power to the study; all placentas were examined macroscopically, with subgroup analysis based on the above power calculation for microscopic and functional analyses. Data were presented as median (interquartile range, IQR) and differences between groups were assessed by Mann-Whitney U Test using Prism 6 for Mac OS X (Graphpad software Inc., San Diego, USA). Page 71 of 253 2.4 Results Participants (N=100) were recruited between January 2012 and May 2014. Twenty-three pregnancies were classified as APO with the most frequent APO category being SGA (21, 91.3%), consequently the median IBC is lower in APO versus NPO pregnancies. No stillbirths or neonatal deaths occurred amongst the study population. Demographic and pregnancy outcome data are shown in Table 2. There were no statistically significant differences in maternal demographics, features of the RFM episode(s) or mode of delivery between pregnancies ending in APO compared to those with NPO. The median time from RFM to delivery was 5 (2-12) days. Placental structure Placentas from RFM APO pregnancies (N=23) were significantly smaller in terms of length, width, volume, weight and area compared with their NPO counterparts’ placentas (N=77) (Figure 4). There was no difference in placental depth (Figure 4E), the degree of cord eccentricity as measured by the cord ratio (Figure 4G) or the fetoplacental weight or volume ratios (Figure 4J-K). When compared to placental centile curves (Thompson et al., 2007) 21/23 (91.3%) and 15/23 (65.2%) of APO placentas weighed below the 10 th and 3rd centiles respectively, compared with 36/77 (46.8%) and 17/77 (22.1%) of NPO placentas (p≤0.0002). Villous area and trophoblast areas were not significantly different between NPO (N=23) and APO (N=11) placentas (Figure 5A-F). A reduced number and density of fetal vessels was observed in villous tissue of APO (N=9) vs. NPO (N=20) placentas (p≤0.002); there was a further trend towards reduced luminal area and ratio in APO placentas (both p=0.06) (Figure 5G-L). Placental vascular function In the subset of placentas used to evaluate vascular function, the only statis tically significant difference in pregnancy characteristics was that a higher proportion of APO placentas came from obese mothers (NPO 1/20 (5.0%) vs. APO 5/9 (55.6%), p=0.002). Eight (88.9%) APO pregnancies examined for ex vivo Page 72 of 253 TABLE 2: Baseline and delivery characteristics of study participants and their offspring. Maternal Characteristics Age (years) Ethnicity Caucasian Asian Black Other BMI (kg/m2) Parity Smoker RFM Characteristics Episode number Episode duration (hours) Gestation at presentation (weeks +days ) Delivery Characteristics Gestation (weeks +days ) Induction of labour Laboured Caesarean Male infant Outcome Characteristics Live birth* IBC IBC<10* Apgar score <7 at 5 minutes* Umbilical Arterial pH<7.10* Umbilical Base Excess<-10* NICU Admission* NPO N=77 APO N=23 p 29.4 (25.1 – 32.7) 28.0 (23.0 – 33.3) 0.50 54 (70.1) 11 (14.3) 7 (9.1) 5 (6.5) 26.3 (23.2 – 30.3) 0 (0 – 1) 9 (11.6) 15 (65.2) 4 (17.4) 0 (0) 4 (17.4) 25.2 (23.3 – 32.1) 0 (0 – 1) 5 (21.7) 1 (1 – 2) 36 (12 – 72) 38+6 (36+1 – 40+4) 1 (1 – 2) 60 (27 – 108) 38+4 (37+0 – 39+1) 0.93 40+1 (38+4 – 41+1) 48 (62.3) 60 (77.9) 15 (19.5) 40 (51.9) 39+1 (38+0 – 40+4) 14 (60.9) 15 (65.2) 7 (30.4) 12 (52.2) 0.076 77 (100) 41.7 (23.1 – 67.2) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 23 (100) 5.8 (3.0 – 9.0) 21 (91.3) 0 (0) 1 (4.3) 1 (4.3) 1 (4.3) 1.00 <0.0001 0.74 0.97 0.38 0.22 0.13 0.27 0.90 0.35 0.27 1.00 <0.0001 1.00 0.066 0.066 0.066 Normal outcome (NPO) and adverse outcome (APO) pregnancies differ by outcome characteristics alone. Key: BMI = Body Mass Index, IBC = Individualised Birth weight Centile, NICU = Neonatal intensive care unit. Data are presented as median (interquartile range) or number (%) and are compared by Mann Whitney U Test or Chi squared test (with Yates’ Correction as required). * denotes a variable included in the definition of APO. Page 73 of 253 C **** 40 30 20 10 0 NPO 6 10 0 NPO NPO Placental Weight (grams) Cord Ratio 400 200 0.3 0.2 0.1 0.0 APO APO Weight 0.5 0.4 Area (cm2) 0 Cord Ratio **** NPO 2 H Area 0 4 APO G 600 Depth **** 20 APO F E Width 30 Placental Width (cm) 50 Placental Length (cm) D Length Placental Depth (cm) A NPO **** 1200 1000 800 600 400 200 0 APO NPO APO B I J K Volume 1000 800 600 400 200 0 NPO APO Fetoplacental Volume Ratio 15 15 Fetoplacental Volume Ratio (g/cm3) **** Fetoplacental Weight Ratio 5cm$ Placental Volume (cm3) 1200 Fetoplacental Weight Ratio 10 5 0 NPO . Page 74 of 253 APO 10 5 0 NPO APO FIGURE 4: Ex vivo placental macrostructure. Ex vivo placentas from reduced fetal movement pregnancies with adverse (APO; N=23) and normal (NPO; N=77) outcome are biometrically different. Example placental photographs from NPO (A) and APO (B) pregnancies are shown at equal magnification (scale bar = 5cm); box and whisker plots of placental length (C), width (D), depth (E), area (F), cord ratio (G), weight (H), volume (I), fetoplacental weight ratio (J) and fetoplacental volume ratio (K). Data are presented as median (horizontal line), interquartile range (box limits) and range (vertical line) and compared by Mann Whitney U Test. **** denotes NPO vs. APO p<0.0001. FIGURE 5: Ex vivo placental microstructure. Example images of anti-cytokeratin 7 immunostained villous tissue from normal pregnancy outcome (NPO; N=20) and adverse pregnancy outcome (APO; N=9) placentas are shown in (A) and (B) respectively; box and whisker plots of villous area (C), trophoblast area (D) or their relative proportions (trophoblast ratio) (E). Example images of antCD31 immunostained villous tissue from NPO (N=23) and APO (N=11) placentas are shown in (F) and (G) with vessels identified by arrows; box and whisker plots of vessel number (H), density (I), luminal area (J) and luminal ratio (K). Images are captured at 40x original magnification with scale bar representing 200µm. The corresponding negative control image is inset in the top right corner. Data are presented as median (horizontal line), interquartile range (box limits) and range (vertical line) and compared by Mann Whitney U Test. *** denotes NPO v. APO p<0.001. placental arterial function were SGA, of whom four had an IBC<5. CPA resting diameters were well matched between APO and NPO placentas (NPO: 288 (190-411) vs. APO: 365 (240-436) μm). Constriction to U46619 by CPAs from APO placentas was not significantly different from those from NPO placentas (Figure 6A): AUC (NPO 17.3 (12.0–22.6) vs. APO 10.3 (7.2– Page 75 of 253 U46619 Concentration Response Curve 20 15 10 5 0 -5 B Constriction Remaining (%) Active Effective Pressure (kPa) A -11.0 -10.0 -9.0 -8.0 -7.0 -6.0 SNP Concentration Response Curve 120 100 80 * 60 40 20 0 -5.7 -10.0 -9.0 Log Concentration of U46619 (M) C Passive Tension Accumulation D 500 -4.0 3000 2000 1000 APO F Diameter (% of L0.9 5.1kPa) 5 Tension (mN/mm2) -5.0 4000 0 NPO Peak Active Tension 4 3 2 1 0 -6.0 Active Tension Generation Area under Curve (Arbitrary Units) Tau (1/k) 1000 E -7.0 5000 1500 0 -8.0 Log Concentration of SNP (M) NPO APO NPO APO Peak Active Tension Diameter 400 300 200 100 0 NPO APO FIGURE 6: Ex vivo placental arterial function. Chorionic plate arteries from adverse pregnancy outcome (APO; N=8 in grey) placentas show altered responses to vasoactive agonists but length tension characteristics are unchanged compared to normal pregnancy outcome (NPO; N=15 in black) counterparts; thromboxane mimetic, U46619 (A) and sodium nitroprusside (SNP) (B) concentration response curves, passive tension accumulation (C), active tension generation (D), peak active tension (E), and peak active tension diameter (F). Data are presented as median and IQR (A&B) or median (horizontal line), interquartile range (box limits) and range (vertical line) (C -F) and all data are compared by Mann Whitney U Test. * denotes NPO vs. APO p<0.05. 15.9)), Vmax (10 (7–13) vs. 7 (5–9), p=0.063), and EC50 (46 (32–66) vs. 49 (33–90). However, CPAs from APO placentas demonstrated impaired relaxation to nitric oxide donation by SNP in terms of AUC (NPO 441 (310– 530) vs. APO 543 (457–595), p=0.024), V max (54 (37–76) vs. 87 (57–98) %, p=0.047) and EC50 (12 (6–24), vs. 30 (19–50) nM, p=0.02) compared with those from NPO placentas (Figure 6B). Length-tension characteristics of the CPAs were similar in NPO and APO placentas (Figures 6C-F). Page 76 of 253 Placental endocrine function No differences were observed in the relative rates of mRNA transcription for the five genes of interest between NPO (N=21) and APO (N=14) placentas. However, fresh placental tissue from APO pregnancies contained less hCG and hPL protein, and in culture, released less hCG and more s-Flt-1 into CM compared with placental tissue from NPO pregnancies (Table 3). There were no differences between APO and NPO in the other hormones assessed. Page 77 of 253 TABLE 3: Ex vivo placental endocrine function. A Gene hCG NPO N=21 44.3 (5.6 – 74.2) hPL CYP11A1 PlGF sFlt-1 mRNA APO N=14 14.4 1931 1487 (467.3 – 2074) 23.9 29.2 (8.3 – 82.8) (12.0 – 76.5) 124.7 104.3 (52.1 – 167.2) (35.2 – 134.6) 21.3 19.1 (12.7 – 39.5) Hormone p 0.37 (1.7 – 47.8) (606.8 – 3556) (16.2 – 30.3) B 0.12 0.82 0.45 0.84 hCG (mIU/mg) hPL (mg/mg) Progesterone (ng/mg) PlGF (pg/mg) sFlt-1 (pg/mg) NPO N=28 55.1 (23.8 – 102.0) Lysate APO N=14 19.9 P 0.0067 (11.0 – 50.2) 27.0 10.7 (9.1 – 49.8) (5.9 – 14.1) 426.9 429.4 (323.8 – 605.2) (365.2 – 620.6) 277.3 133.1 (89.3 – 501.8) (87.7 – 308.2) 673.4 812.6 (348.6 – 1230) (496.4 – 2016) Explant-Conditioned Medium NPO APO p N=15 N=6 160.7 84.6 0.028 (119.6 – 295.2) 0.0062 0.88 0.22 0.36 (69.4 – 179.1) 89.5 76.6 (67.5 – 103.6) (54.0 – 95.9) 425.1 309.9 (314.5 – 555.3) (179.1 – 523.3) 342.6 329.1 (287.4 – 480.6) (205.3 – 727.1) 5473 20498 (1650 – 14814) (12978 – 28914) 0.40 0.17 0.59 0.013 Relative gene transcription (compared to expression of TATA-box Binding Protein) (A) and tissue content and release (B) of key placental hormones in placentas of normal outcome (NPO) and adverse outcome (APO) reduced fetal movement pregnancies. Key: hCG = human chorionic gonadotrophin, hPL = human placental lactogen, CYP11A1 = gene encoding CYP450scc (key synthetic hormone of progesterone), PlGF = placental growth factor, sFlt-1 = soluble fms-like tyrosine kinase. Values are presented as median (IQR) and compared by Mann Whitney U Test. Page 78 of 253 2.5 Discussion The data reported here support our hypothesis that there is an altered placental phenotype in pregnancies reporting RFM when there is an APO versus those with RFM and a NPO. In pregnancies complicated by RFM after 28 weeks gestation placentas from pregnancies ending in APO are lighter, smaller and less vascularised and also display aberrant vascular and endocrine function compared to NPO pregnancies. These structural and functional changes might reduce transplacental transfer of nutrients and oxygen to the fetus, impairing the ability of the placenta to support the optimal growth and development of the fetus, thereby putting the infant at increased risk of perinatal mortality and morbidity. Alternatively, they may confound the relationship between an alternative causal pathology and APO. In any case, such changes merit exploration as they could form the basis of future tests of placental health to assist the identification of the “at-risk” fetus. The principal reason for pregnancies in this study to be classified as APO was being SGA at birth (21% of the total study population); this is in excess of the rate in unselected populations (~11% (McCowan et al., 2013)) and in keeping with a previously described increased risk of SGA in pregnancies complicated by RFM (O'Sullivan et al., 2009). This supports the choice of RFM pregnancies as a population at increased risk of APO in late pregnancy. The absence of stillbirth in a pregnancy cohort of this size is unsurprising, given the background rate of 5.2 per 1,000 UK births during the time of this study (Cox et al., 2009). Given the preponderance of SGA babies amongst the APO cohort, the observed macro- and micro-structural phenotype in APO pregnancies is not unexpected (see Table 1), with the exception of the lack of reduction in villous area, which contrasts with that reported by Daayana et al. (2004) and lack of significant increase in fetoplacental weight ratios (Worton et al., 2014). Nevertheless, our observations provide further evidence that the pathophysiological processes underlying APO in RFM are similar to those observed in FGR and stillbirth. Our findings suggest that the Page 79 of 253 differences in placental structure seen between RFM and healthy controls (Warrander et al., 2012; Winje et al., 2012) are most pronounced in pregnancies associated with APO. Conversely, the vascular function profile observed in APO placentas contrasts with that described in studies of FGR (commonly defined for research purposes as IBC <5) which demonstrated an increased sensitivity to U46619 (Mills et al., 2005) and alteration of length-tension characteristics (Mills et al., 2011). Possible factors contributing to this difference may be the more physiological oxygen concentrations used in the current study (Wareing et al., 2002), as hyperoxia increases placental arterial constriction (Wareing et al., 2006), and differences in maternal BMI and differences in gestational age of samples. The majority of mothers of APO pregnancies in this study had a BMI>30 kg/m2; these women are a subpopulation with an increased risk of stillbirth (Yao et al., 2014). Interestingly, obesity is associated with the same insensitivity to nitric oxide donation observed in the CPAs of APO placentas examined in this study (Hayward et al., 2013). With regard to the effects of gestation, Mills et al. largely used early-onset FGR pregnancies requiring preterm delivery which contrasts with our population of predominantly late-onset FGR/placental dysfunction in pregnancies delivering at or around term (Mills et al., 2005 and 2011). This is consistent with observations of altered pathophysiology between earlyand late-onset FGR (Mifsud and Sebire, 2014). Critically, late-onset FGR/placental dysfunction is more reflective of the clinical picture that results in stillbirth in late pregnancy when abnormal umbilical artery Doppler waveforms are uncommon (Oros et al., 2011). Thus, these pathophysiological changes to placental vasculature in late pregnancy may not be reflected in the umbilical artery Doppler waveform. Further research is necessary to determine which clinical investigations best relate to these abnormalities of placental vessels. Disruption of placental endocrine function was evident by altered production and release of several placental hormones in APO placentas. Page 80 of 253 Reduced hPL protein content per unit of placental tissue described in this study mirrors the reduction in maternal serum hPL concentrations in APO (Spellacy et al., 1977; Letchworth et al., 1978; Isouard, 1979; Leader and Baillie, 1980; Morrison et al., 1980; Dutton et al., 2012) and the down regulation of hPL transcription in placental tissue from SGA infants (Mannik et al., 2010), although we were not able to confirm the latter finding in our study. Similarly, the reduced placental tissue hCG content and release mirrors the association of increased APO risk with low first trimester maternal serum hCG concentration (Gagnon et al., 2008), but contrasts with previous reports of enhanced hCG release from placental explants and primary cytotrophoblasts of FGR infants (Crocker et al., 2004; Newhouse et al., 2007). These reductions in hPL and hCG lysate content are not related to altered trophoblast area or transcription rate. Thus, they may reflect reduced protein synthesis, consistent with post-transcriptional regulation for example by miRNA (Krol et al., 2010; Fu et al., 2013) or posttranslational regulation such as the unfolded-protein response secondary to placental stress in FGR (Yung et al., 2008). Lack of corresponding reduction in lysate PlGF, sFlt-1 or progesterone content suggests that there may be differential effects on net production of individual hormones rather than a global reduction in hormone production or release in APO. Increased release of sFlt-1 during villous explant culture of APO placentas is in keeping with the findings of Nagamatsu et al. (2004), Nevo et al. (2006) and Gu et al. (2008) who observed that release of sFlt-1 is enhanced under hypoxic culture conditions. Collectively, these data (normalised to the amount of placental tissue) do suggest the presence of placental dysfunction rather than simply being an effect of reduced placental size. The major strengths of our study are the detailed, matched structural and functional characterisation of placental samples with short delivery to collection interval to reduce storage artefact (Garrod et al., 2013). Additionally, given the predominant growth and vascular development of the placenta in the first half of pregnancy (Kingdom et al., 2000; Orzechowski et al., 2014), the short interval between presentation with RFM Page 81 of 253 and delivery suggests that this placental phenotype was present at the time of presentation with RFM. Finally, the analyses of placental structure and function were restricted to those with the availability of potential noninvasive methods of in utero assessment (for example placental ultrasound, placental arterial circulation Doppler and maternal blood tests). Technologies under development, such as magnetic resonance imaging may in the future allow assessment of placental oxygen and nutrient transfer or metabolism (Wright et al., 2011; Denison et al., 2012; Huen et al., 2013) which could exploit findings of reduced amino acid transport in APO placentas (Warrander et al., 2012), but these were not considered in this study. Thus, the placental phenotype of APO described here has potential to be translated into clinical practice with minimal safety and acceptability concerns. The current study has a number of limitations. Outcomes were defined on the basis of a composite primarily influenced by birth weight (the end result of intrauterine growth) and not growth itself, although each component of the APO definition is related to increased perinatal mortality (Doctor et al., 2001; Malin et al., 2010; Iliodromiti et al., 2014) and is in keeping with that used by other obstetric studies (Hannah et al., 1996; Koopmans et al., 2009; Boers et al., 2010; Dutton et al., 2012; Barrett et al., 2013; Heazell et al., 2013). The limitation of such an outcome definition is reflected in the high proportion of placental weights <10th centile even amongst “normal” outcome RFM pregnancies in the cohort (46.8%). Secondly, PlGF and sFlt-1 were measured separately using commercially available ELISA kits, and not by bedside measurement techniques (Triage ® by Alere Inc., Waltham, USA and Elecsys ® by Roche Diagnostics Ltd, Burgess, UK) that might more readily be employed in clinical practice. Although there is a good correlatio n between PlGF measurements by these techniques (Nice et al., 2014), it cannot be assumed that the findings of this study are translatable to other measurement methods. Regarding vascular functional assessment, the technique of wire myography isolates the vessels from endocrine and paracrine factors from blood and from the support of surrounding tissue Page 82 of 253 that may influence the in vivo vascular phenotype; whole placental cotyledon perfusion studies may assist this in the future. Finally, the external validity of these placental findings, outside the context of RFM, may be questioned. However, when compared to Table 1, the pattern of placental findings described in RFM APO placentas in this study is similar to that of stillbirth and FGR in general, with or without preceding RFM. Thus, while the findings would need to be corroborated in other populations and at-risk groups, we believe them to be generally applicable. A previous Cochrane review found insufficient evidence to recommend or refute routine measurement of either biochemical (Neilson, 2012) or vascular placental markers (Bricker et al., 2008). For such tests to demonstrate significant benefit in preventing stillbirth, a standardised combined “test and treatment” intervention is required. The Reduced Movements Intervention Trial (ReMIT) pilot study shows that studies of this nature are feasible (Heazell et al., 2013). In late pregnancy a treatment (delivery) to prevent stillbirth is available; what is required now to make progress is an accurate test. This study highlights areas of placental health that could form the basis for the development of such tests and makes significant progress towards their development. However it also raises two important questions. Firstly, whether these aspects of placental health can be reliably assessed in utero during late pregnancy, and secondly whether multi-modal antenatal placental health assessment can identify fetuses at greatest risk of stillbirth. Only then can we hope to significantly redu ce placentally-related stillbirth rates. Page 83 of 253 2.6 Acknowledgements The authors would like to thank the women who participated in this study and the midwives at St Mary’s Hospital, Manchester for their assistance in participant recruitment and placental collection after birth. Dr L Higgins is supported by an Action Medical Research Training Fellowship and a Manchester NIHR Biomedical Research Centre fellowship. The study was also supported by Tommy’s - the Baby Charity. 2.7 Statement of author contributions The project was conceived by EDJ, AEPH, CPS and LH, and methodologies planned by LH with expert supervision from MW (myography), SLG (explant culture and ELISA) and RLJ (PCR). NRdC performed quantification of trophoblast area, NA performed quantification of luminal area; LH performed all other laboratory analyses. Statistical analysis and manuscript preparation was performed by LH. All authors were involved in writing the paper and had final approval of the submitted and published versions. Page 84 of 253 2.8 Supplementary data SUPPLEMENTARY TABLE 1: Primers used to study placental transcription of key placental hormones. Gene TBP i hCGi hPLe CYP11A1e PlGFi sFlt-1i Primer Sequence (5’ – 3’) F: CACGAACCACGGCACTGATT R: TGCAGCACGCGGGTCATGGT F: TCACTTCACCGTGGTCTCCG R: TGCAGCACGCGGGTCATGGT F: TCCTCAGGAGTATGT R: CACAGCTACCCTCTA F: TCCAGAAGTATGGCCCGATT R: CATCTTCAGGGTCGATGACATAAA F: GAACGGCTCGTCAGAGGTG R: ACAGTGCAGATTCTCATCGCC F: GGGAAGAAATCCTCCAGAAGAAGA R: GAGATCCGAGAGAAAACAGCCTTT Accession Number NM_001172085 NM_000737 NM_020991 NM_000781 NM_001207012 NM_001159920 Key: e = Eurofins Genomics, Ebersberg, Germany. I = Invitrogen, Paisley, UK. TBP = TATA-box binding protein (a placental housekeeping gene), hCG = human chorionic Gonadotrophin, hPL = human placental lactogen, CYP11A1 = gene encoding CYP450scc (key synthetic enzyme of progesterone), PlGF = placental growth factor, sFlt-1 = soluble fms-like tyrosine kinase-1. SUPPLEMENTARY TABLE 2: Enzyme-Linked Immunosorbant Assay kits used to quantify hormone content of tissue lysate and explant-conditioned media. Hormone hCG hPL Progesterone PlGF sFlt-1 Company DRG DRG DRG R&D R&D Product ID Optical Density (nM) EIA 1469 450 EIA 1283 450 EIA 1561 450 DPG00 540 DVR100B 540 Range of Detection 5-1000mIU/ml 0-20mg/L 0-40ng/ml 0 – 1000pg/ml 0-2000pg/ml Intraassay CoV 5.5% 5.2% 2.6% 5.0% 4.8% Key: hCG = human chorionic gonadotrophin, hPL = human placental lactogen, PlGF = placental growth factor, sFlt-1 = soluble fms-like tyrosine kinase-1, DRG = DRG International, Springfield, USA, R&D = R&D Systems, Abingdon, UK, CoV = coefficient of variance. Page 85 of 253 CHAPTER 3: PLACENTAL VOLUME AND BIOMETRY MEASUREMENT IN THE THIRD TRIMESTER: A VALIDATION STUDY Higgins LE, Simcox L, Sibley CP, Heazell AEP, Johnstone ED. A manuscript prepared for submission to the Journal of Ultrasound in Obstetrics and Gynecology Page 86 of 253 3.1 Abstract Objectives: To test the hypothesis that third trimester placental biometry and volume can be measured by two- and three-dimensional ultrasound in utero and to determine which method of measurement has the best biological correlation. Methods: Singleton pregnancies underwent placental ultrasound within seven days of delivery (n=87, 29 +3 – 41+5 weeks). Length and width (linear and curvilinear) and maximal depth were estimated using two-dimensional (2D) ultrasound. Placental volume (PV) was estimated using 2D (ellipse and shell) and three-dimensional (3D) (rotational and multiplanar) techniques. Measurements were compared to their true correlates following delivery. Intra- and inter-observer reliabilities of candidate placental size estimates were assessed by intraclass correlation coefficient (ICC). Results: Curvilinear placental length (Rs=0.24, p=0.031), width (Rs=0.27, p=0.013) and depth (Rs=0.31, p=0.0056) correlated well with ex vivo measurements. All methods of PV estimation were related to ex vivo volume (Rs≥0.32, p<0.01) but not placental weight (p>0.05); rotational estimation at 30° intervals demonstrated the strongest biological correlation (Rs=0.40, p=0.0004). ICCs for intra- and inter-observer reliability of placental size measurements were suboptimal (0.59-0.70 and 0.096–0.58 respectively). Conclusions: 2D and 3D ultrasound can be used to obtain biologically correlated measures of third trimester placental size. However, the reliability (particularly interobserver reliability) of these estimates needs to be improved prior to prospective studies to determine their predictive value. Key words: Placental volume, Placental weight, Tissue density, Twodimensional ultrasound, Three-dimensional ultrasound, VOCAL, Ellipse Page 87 of 253 3.2 Introduction Ex vivo examination of the placenta reveals relative reductions in size and increases in fetoplacental ratio (FPR) in pregnancies ending in stillbirth (Heazell and Martindale, 2009; Worton et al., 2014), fetal growth restriction (FGR) (Biswas and Ghosh, 2008; Balihallimath et al., 2013) and RFM pregnancies with adverse outcome [Unpublished work, Chapter 2] as compared to pregnancies with a normal outcome. Wolf et al. (1989) demonstrated that sonographically detectable placental growth restriction precedes the onset of FGR by three weeks or more. Two-dimensional (2D) ultrasound measures of placental diameter and thickness have been used as indicators of pregnancies at highest risk of adverse pregnancy outcomes of placental origin including FGR and stillbirth (Viero et al., 2004; Toal et al., 2007; Toal et al., 2008a; Toal et al., 2008b; Proctor et al., 2009). No studies have examined the biological correlation of the 2D placental biometry estimation techniques used in these studies i.e. whether in utero sonographic measurements of placental biometry have any relationship to direct measurements of ex vivo placental size after delivery. Sonographic placental volume (PV) using the three-dimensional (3D) ultrasound measurement technique Virtual Organ Computer Aided anaLysis (VOCAL) has previously been performed, demonstrating smaller first trimester PV in pregnancies ending in delivery of small for gestational age (SGA) infants (Law et al., 2009; Collins et al., 2013b) and early-onset FGR and hypertensive disorders (Rizzo et al., 2008; Bozkurt et al., 2010; Arakaki et al., 2014). In contrast, pregnancies complicated by late-onset FGR and preeclampsia failed to show an appreciable significant difference in first trimester PV (Rizzo et al., 2009; Odeh et al., 2011; Arakaki et al., 2014). Due to their nature (first trimester studies), no studies have determined the biological correlation of this VOCAL PV estimation technique. In contrast, a 2D PV estimation technique (calculated using a convex shell model) was validated in the third trimester by correlation against placental weight (Azpurua et al., 2010). In the second trimester this technique demonstrates Page 88 of 253 reasonable specificity (91%) but low sensitivity (19%) for SGA births (Staff et al., 2011), with pregnancies with PV<25 th centile being twice as likely to deliver an SGA infant or experience adverse outcome (Laine et al., 2012). Yet it is not known whether placental weight and PV are consistently related and therefore this validation may be inappropriate. This study tested the hypothesis that placental biometry and volume can be measured in utero using ultrasound in third trimester pregnancies with high ex vivo correlation and reliability. Page 89 of 253 3.3 Materials and methods Ethical approval was received from Greater Manchester North West Research Ethics Committee (11/NW/0650). Women with singleton pregnancies (in the absence of known fetal abnormality) of at least 28 weeks gestation undergoing third trimester ultrasound examination (for assessment of reduced fetal movements, suspected or confirmed fetal growth restriction or confirmation of fetal presentation) were asked to participate in the study by undergoing additional ultrasound measurements during their scan and donating their placenta to the study after delivery. Written informed consent was obtained. Sonographic assessment of placental size and shape All ultrasound examinations were conducted by a single RCOG-accredited sonographer (LH) using a Voluson E6 with a three-dimensional (3D) enabled RA4B 4-8Hz curvilinear probe (GE Healthcare) unless otherwise stated. Fetal weight was estimated by two-dimensional (2D) fetal biometry according to the Hadlock C formula (1985). The placental body was identified by 2D ultrasound, its predominant location defined as anterior, posterior, lateral or fundal and its longest plane identified. The ultrasound probe was angulated to include as much of the placental body as possible and minimise acoustic shadowing from the fetus and a 2D image and 3D volume (85° sweep) of the placenta were captured in this plane, The probe was rotated 90° to identify the longest perpendicular plane and a further 2D image was captured. This procedure was repeated three times. Unless otherwise stated, 2D images were analysed in real time whilst 3D images were analysed offline using 4Dview v.5 (GE Healthcare) ultrasound image analysis software. Throughout the manuscript, the prefix “ est” refers to an in utero estimate of a particular aspect of placental size. Biological correlation refers to the existence of a significant statistical correlation between estimated and true measurements (see “statistical analyses”). Biological relevance (relationship Page 90 of 253 of the measure to an outcome of interest such that a change in management may be effected by the result of the test) is not tested in this study. Modelling of placental shape and tissue density All placentas (regardless of scan to delivery interval) were trimmed of their extra-placental membranes and umbilical cord. Their weight (PW) and volume (PV; measured by volume displacement (Scherle, 1970)) were recorded and the placentas were photographed, chorionic plate facing upward alongside a scale bar. Placental depth (D) was measured directly at the apparent deepest point of the placental body. Using Image ProPlus v.6.0 (Media Cybernetics UK, Marlow, UK) placental photographs were analysed to quantify placental length (L; longest diameter of the placenta), width (W; longest diameter perpendicular to the placental length) and average diameter (A). These measurements were incorporated into formulae for the volume of an ellipse (PV = 4/3π x 0.5L x 0.5W x 0.5D), elliptical cylinder (PV = π x 0.5L x 0.5W x D) and circular cylinder (PV = π x (0.5 x A)2 x D). The modelled ex vivo PV determined by each formula was then correlated to true ex vivo PV. The model that best approximated true PV was carried forward into subsequent analyses. Tissue density was expressed as the ratio of placental weight to volume. Validation of sonographic placental biometry, volume, weight and fetoplacental ratio Subgroup analysis of those placentas delivered within seven days of ultrasound examination (the “Validation cohort”) was performed to valida te sonographic placental measurements. Systematic and random errors were calculated for the most accurate method of estimating each placental size measure. Page 91 of 253 Placental biometry estL, estW and estD were defined as for the ex vivo measurements described above. estL (from images of the longest plane of the placenta) and estW (from images of the longest perpendicular plane of the placenta) were estimated in three ways; (i) a straight line (or two straight lines meeting at the angle o f the placenta if the placental shadow was particularly curved) through the placenta from tip to tip (Figure 7A) (McGinty et al., 2012), (ii) a curvilinear line along the maternoplacental interface (Figure 7B) (Suri et al., 2013), and (iii) a curvilinear line through the middle of the placental body (Figure 7C). If the placental shadow extended beyond the edge of the image, real-time extrapolation to incorporate “missed tissue” was permitted for 2D measurements. estD was estimated at the visibly deepest point of the placental body, perpendicular to its plane (Figure 7D). The most accurate method of estimating estL, and estW, were carried forward into further analyses. Placental volume and weight estPV was then measured in four ways. Firstly it was calculated according to the most appropriate geometric formula (as assessed above) using 2D estW estL, and estD (Figure 8A). Next the placental arc was measured as described by Azpurua et al. (2010) (Figure 8B) and the measurements used to calculate estPV from the concave convex shell formula: 1/4π(T/6) x [4H(B-T) + B(B-4T) + 4T2] where B refers to the tip-to-tip distance across the base of the placental arc, H to the maximal height of the arc and T to the thickness of placental tissue at the maximal height of the arc. Finally, 3D volumes were analysed, starting with the placental body positioned in the centre of image. The placental outline was traced at both 30° (VOCAL 30⁰) and 15° (VOCAL 15⁰) rotation angles (Figure 8C) as previously described (de Paula et al., 2008; Artunc Ulkumen et al., 2014b; Artunc Ulkumen et al., 2014a), and at “slicing” intervals of 10mm (MP10) and 5mm (MP5) from the first to last “slice” that the placenta appeared until no placental tissue was visible (Figure 8D) in a modification of the multiplanar technique described by Page 92 of 253 A B C D D FIGURE 7: Measurement of placental biometry using two-dimensional ultrasound. Length and width were measured by three methods; straight line (A), curvilinear line at maternoplacental interface (B) and curvilinear line through the placental body (C). Placental depth was measured perpendicular to the placental body at the maximal point (D). Broken line represents line of measurement. Cheong et al. (2010) in the first trimester, and Hafner et al. (2001) in the second trimester. The average of three estPVs for each methodology was taken per placenta to minimise random error. Placental weight was estimated (estPW) by multiplication of assessed above). The most accurate estPV estPV by placental tissue density (as and estPW techniques were taken forward into further analyses. Fetoplacental Ratio FPR was assessed to examine placental efficiency, dividing the EFW by estPV to generate a fetoplacental volume ratio ( estFPRv) and by the generate a fetoplacental weight ratio ( estFPRw). Page 93 of 253 estPW to A B C D FIGURE 8: Measurement of placental volume using two- and three-dimensional ultrasound. Schematic representations of placental volume estimation by two-dimensional (A&B) and threedimensional (C&D) ultrasound with accompanying axial and sagittal orthographic planes. Ellipse (A) and shell (B) volumes were calculated according to formulae for the volume of an ellipse and elliptical shell respectively. Rotational (C) and multiplanar (D) volumes were calculated using in -built formulae within 4D view v.5 (GE Healthcare) ultrasound image analysis software after tracing the placental outline at pre-specified rotation or slicing intervals. Broken line represents line of measurement. Page 94 of 253 Assessment of sonographic reliability Reliability of placental sonographic measures was assessed in a series of scans conducted by two RCOG-accredited sonographers (LH and LS) following the same methodologies described above. A series of three measurement sets were obtained (LH 1, LS, LH2) with each sonographer blinded to the values obtained in each previous assessment. Images were analysed offline in random order, blinded to participant identity and image pairing. Intra-observer reliability was assessed by comparison of LH 1 and LH2 values, whilst inter-observer reliability was assessed by comparison of LH1 and LS values (Bland and Altman, 2003). Statistical analysis Pregnancies included in the validation cohort were compared to the remainder of the total study cohort to assess for bias; data were compared by Mann-Whitney U-Test (for continuous data) and Chi squared test with Yates’ correction as required (for categorical data). Sonographic accuracy was assessed by the relationship between EFW and true birth weight in the validation cohort (Spearman Rank Correlation). Relationships between estimated and true placental size were tested by Spearman Rank Correlation. Statistical analysis was carried out using Prism 6 for Mac OS X (Graphpad Software Inc., San Diego, USA) and statistical significance of the existence of a relationship between estimated and true values was determined by a p value <0.05. The strength of that association was determined by the highest R s value of each set of measurements; where equal, the most accurate method of estimation was determined to be the o ne with smallest systematic error (Systematic error: (predicted – true)/true. Random error: standard deviation of systematic errors). The most accurate method of estimating each measurement was taken forward into subsequent analyses. Reliability (intra- and inter-observer) was assessed as the coefficient of variance (CoV; standard deviation of measures / average of measures), intra-class correlation coefficient (ICC), bias and 95% limits of agreement. Page 95 of 253 Based on an expected correlation of 0.68 between estPV and PV in term pregnancies and a 37.5% failure to obtain measurements in advanced gestation as published by Azpurua et al. (2010) a minimum of 22 placentas would be required to detect a significant correlation of similar magnitude with power of 80% at the level of p<0.05. Given the local proportion of pregnancies delivering within seven days of assessment for reduced fetal movements (32.6%, unpublished data) and a 10% loss to follow up rate, we estimated that at least 120 participants would need to be recruited to obtain sufficient matched in vivo and ex vivo datasets. Page 96 of 253 3.4 Results Recruitment and assessment of biological correlation Placentas from 129 participants were received; these pregnancies formed the whole study cohort. Table 4 summarises their maternal and pregnancy characteristics. Within the study cohort 87 (67.4%) pregnancies delivered within seven days of ultrasound examination and formed the sonographic validation cohort; one or more 2D placental measurements were not obtained in two cases (2.3%) and 3D volumes acquired were deemed inadequate in four cases (4.6%). EFW was highly correlated with true birth weight (Rs=0.77, p<0.0001) indicating that the sonographic accuracy in general was good. Ex vivo placental volume modelling and tissue density Each shape approximation generated a modelled PV with highly significant relationship to the true PV (Rs=0.74, p<0.0001 for each). Head to head comparison of the three modelling methods revealed significant degrees of systematic bias in modelled volume by elliptical cylinder (median 42.2% overestimation) and circular cylinder methods (182.8% overestimation), and that the ellipse formula was the most representative PV model (6.4% underestimation) (Figure 9A). PW and PV were highly significantly correlated (Rs=0.96, p<0.0001) with a median placental tissue density of 1.05g/cm3 (interquartile range 1.03–1.09g/cm3). Validation of sonographic estimated placental biometry, volume, weight and fetoplacental ratio Placental biometry Table 5 summarises the performance of each described method of measuring 2D placental biometry. Statistically significant relationships were observed between estL and L using only method 3, whilst for estW both methods 1 and 3 each correlated with W with statistical significance, however method 3 once more performed best. Page 97 of 253 estD was significantly related to the D. Systematic and random error for each measurement were significant, particularly for 13.35%, estD=97.83%. estD (Systematic error: Random error: estL=-14.76%, estW=- estL=17.14%, estW=18.32%, estD=49.66% respectively). TABLE 4: The placental ultrasound study cohort. Whole Cohort Subgroup-Analysis Scan to delivery interval (days) Any ≤7 >7 N 129 87 42 p Maternal Characteristics Age (years) 29.2 (25.0 – 33.1) 30.8 (26.0 – 33.5) 29.0 (23.6 – 33.2) 93 (72.1%) 17 (13.2%) 9 (7.0%) 10 (7.8%) 26.0 (23.1 –30.2) Parity (number) 0 (0 – 1) Pregnancy Characteristics 63 (72.4%) 11 (12.6%) 8 (9.2%) 5 (5.7%) 25.5 (23.1 – 29.5) 1 (0 – 1) 30 (71.4%) 6 (14.2%) 1 (2.4%) 5 (11.9%) 28.1 (22.9 – 31.6) 0 (0 – 1) 38+6 (37+2 – 40+1) 39+0 (38+2 – 40+3) 36+0 (32+3 – 38+2) 39 (30.2%) 45 (34.9%) 15 (11.6%) 12 (9.3%) 18 (14.0%) 28 (32.2%) 40 (46.0%) 9 (10.3%) 7 (8.0%) 3 (3.4%) 11 (26.1%) 5 (11.9%) 6 (14.3%) 5 (11.9%) 15 (35.7%) 3 (1 – 7) 2 (0 – 4) 20 (12 – 54) <0.0001 36 (14.5 – 61.2) 36 (14.3 – 63.2) 48 (15.5 – 60) 0.56 Ethnicity: Caucasian Asian Black Other BMI (kg/m2) Gestation at Scan (weeks +days ) Placental site: Anterior Lateral Posterior Fundal Not defined Scan to delivery interval (days) Individualised birth weight centile 0.31 0.34 0.30 0.12 <0.0001 <0.0001 Data are presented as median (interquartile range) or number (percentage). Subgroup analysis (by Mann-Whitney U Test and Chi Squared test) was performed between those delivered within seven days of ultrasound assessment (“Validation cohort”) and those who delivered after seven days. Individualised birth weight centile was calculated using Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK). Page 98 of 253 A Ex vivo volume estimation Modelled Volume (cm3) 2500 2000 1500 1000 Ellipse Elliptical Cylinder Circular Cylinder X=Y 500 0 0 250 500 750 True volume 1000 (cm3) In Vivo Volume Estimation B Estimated Placental Volume (cm3) 1000 800 Ellipse Shell VOCAL 30° VOCAL 15° MP 10mm MP 5mm X=Y 600 400 200 0 0 200 400 600 True Placental Volume 800 1000 (cm3) FIGURE 9: Visual representation of the accuracy of placental volume estimation methods. Ex vivo volume estimation (A); whilst demonstrating equal correlation coefficients (not shown), the line of best fit for the ex vivo ellipse model (solid black line) is closer to the true ex vivo placental volume (x=y) than that of the elliptical cylinder (solid grey line) or circular cylinder (broken grey line) model (N=129). In vivo volume estimation (B); compared with the two-dimensional ultrasound estimation methods (ellipse and Shell, shown in black), the three-dimensional ultrasound estimation methods (rotational (VOCAL) and multiplanar (MP)), shown overlapping in grey) are more closely related to the true ex vivo placental volume (x=y) (N=87). TABLE 5: Comparison of sonographically estimated placental biometry to ex vivo directly measured placental size. Measurement Length Width Depth Method 1 2 3 1 2 3 Maximum Rs 0.16 0.088 0.24 0.26 0.17 0.27 0.31 p 0.13 0.43 0.031 0.17 0.13 0.013 0.0056 Placental length and width were defined respectively as the longest diameter of the placenta and the longest diameter perpendicular to the length. Length and width were sonographically estimated by three methods 1) straight line(s) through the placental body from tip to tip, 2) curvilinear line along the maternoplacental interface from tip to tip and 3) curvilinear line through the centre of the Page 99 of 253 placental body from tip to tip. Placental depth was determined as the thickest vertical measurement perpendicular to the plane of the placental length. N=87, Spearman Rank Correlation. Placental volume and weight Table 6 summarises the performance of each method of estimating PV. estPV by the 2D elliptical model (using estL and estW by method 3 above) was more accurate than the previously published shell method. All 3D estPV methods demonstrated higher biological correlation than either 2D estPV method, with VOCAL 30⁰ performing with the highest statistical significance (Rs=0.40, p=0.0004) and demonstrating greater accuracy (particularly at smaller placental volumes) than the best 2D estPV technique (Figure 9B). Despite the observed correlation between PW and PV, neither estPW estPV nor correlated (by any method) significantly correlated with PW (p>0.05). In comparison to the 2D biometric measures, 3D systematic error was improved (estPV=0.60%, estPW=0.91%) but random error was greater (estPV=35.61%, estPW=34.38%). TABLE 6: Comparison of sonographically estimated placental volume and weight to ex vivo directly measured placental volume and weight. Measurement Volume Weight Method Ellipse Shell VOCAL 30° VOCAL 15° MP 10mm MP 5mm Ellipse Shell VOCAL 30° VOCAL 15° MP 10mm MP 5mm 2D or 3D ultrasound 2D 2D 3D 3D 3D 3D 2D 2D 3D 3D 3D 3D Rs p 0.35 0.32 0.40 0.37 0.36 0.38 -0.17 -0.19 -0.20 -0.20 -0.20 -0.16 0.0017 0.0039 0.0004 0.0009 0.0012 0.0007 0.13 0.087 0.081 0.073 0.082 0.15 Elliptical volume was estimated from two-dimensional (2D) sonographic estimates of length (L), width (W) and depth (D) (PV = 4/3π x 0.5L x 0.5W x 0.5D). Shell volume was estimated from 2D measurements of the placental arc (height (H), base (B) and thickness (T)) (PV = 1/4π(T/6) x [4H(BT) + B(B-4T) + 4T2). Three-dimensional (3D) rotational (VOCAL, rotational angle) and multiplanar (MP, slicing interval) measurement techniques were used to estimate PV using inbuilt formulae within 4Dview v.5 (GE Healthcare) ultrasound image analysis software after tracing the placental outline in multiple planes at pre-specified rotation or slicing intervals. N=87, Spearman Rank Correlation. Page 100 of 253 Fetoplacental ratio As estPW failed to correlate with PW, the biological correlation of was not assessed. estFPRv estFPRw by the VOCAL 30⁰ technique correlated with true FPRv (Rs=0.30, p=0.0063) with systematic error of 16.88% and random error of 48.46%. Assessment of sonographic reliability Reliability was assessed in 46 scans. Table 7 summarises the reliability indices of each placental measurement. For all measures the variability in measurements (intra- and inter-) was suboptimal with no ICC >0.75 (Khan and Chien, 2001). Intra-observer variability (Figure 10) was smaller than inter-observer variability (Figure 11) although the Bland Altman plots demonstrate wide limits of agreement for all measures both between and within observers by the wide scatter of points, while the degree of bias was unaffected by placental size (p>0.05). TABLE 7: Reliability of third trimester placental size assessment. Measure Length Width Depth Volume Intra Inter Intra Inter Intra Inter Intra Inter CoV (%) 8.3 14.0 9.5 10.8 12.4 13.0 14.4 26.2 ICC r 0.68 0.096 0.70 0.37 0.59 0.58 0.66 0.54 p <0.0001 0.39 <0.0001 0.098 0.0030 0.0090 0.0010 0.011 Intra- and inter-observer reliability was assessed in a series scans (N=46) by two observers. Variability between observations is displayed as the coefficient of variance (CoV) and intraclass correlation coefficient (ICC). Page 101 of 253 50 0 10 20 30 Average (cm) -50 -100 100 50 -50 -100 2 4 Average (cm) 100 50 0 5 6 8 10 15 20 25 800 1000 Average (cm) -50 -100 D Depth 0 Width Intra-observer difference (%) 100 C Intra-observer difference (%) B Length Intra-observer difference (%) Intra-observer difference (%) A Volume 100 50 0 -50 200 400 Average 600 (cm3) -100 FIGURE 10: Intra-observer reliability of placental size estimates. Bland-Altman plots demonstrate no systematic bias (p>0.05), but suboptimal within observer reliability of placental size estimate replicates as demonstrated by wide scatter in a series (N=46) of repeated measures; length (A), width (B), depth (C) , volume (D). The average of both readings from one observer is shown on the x-axis and the between reading difference (expressed as a percentage of the average of both readings) on the y -axis. The bias between observations is depicted by a solid grey line (where not visible, overlying y=0); the 95% limits of agreement are depicted by a broken line. Page 102 of 253 B 100 50 0 15 20 25 30 Average (cm) -50 -100 Inter-observer difference (%) C 100 50 -50 -100 2 4 Average (cm) 100 50 0 15 6 8 20 25 Average (cm) -50 -100 D Depth 0 Width Inter-observer difference (%) Inter-observer difference (%) Length Inter-observer difference (%) A Volume 100 50 0 -50 200 400 600 800 Average (cm3) -100 FIGURE 11: Inter-observer reliability of placental size estimates. Bland-Altman plots demonstrate no systematic bias (p>0.05), but suboptimal between observer reliability of placental size estimate replicates as demonstrated by wide scatter in a series (N=46) of scans by two observers; length (A), width (B), dept h (C), volume (D). The average of both observer’s readings is shown on the x-axis and the between observer difference (expressed as a percentage of the average of both readings) on the y -axis. The bias between observations is depicted by a solid grey line (where not visible, overlying y=0); the 95% limits of agreement are depicted by a broken line. Page 103 of 253 3.5 Discussion This study has established that placental biometry and volume can be measured with biological correlation using 2D and 3D ultrasound in third trimester pregnancies. Such measurements are clinically desirable as reduced placental size is related to FGR and stillbirth (Biswas and Ghosh, 2008; Worton et al., 2014). This study suggests that small placental size can be identified by placental ultrasound in the third trimester of pregnancy. The ellipse was identified by ex vivo modelling as the most representative geometric volume for PV estimation, which explains why the 2D ellipse has a higher biological correlation than the previously described shell (elliptical cylinder) method. An elliptical estPV trimester of pregnancy but utilised has previously been applied in the first estL and estW derived by method 2, shown here to be less strongly correlated to true L and W than method 3 in third trimester (Suri et al., 2013). Conversely, the shell estPV method was previously correlated to PW in a smaller cohort of 38 participants (with 23.7% overall failure rate) (Azpurua et al., 2010) compared to our validation cohort of 87 pregnancies (with 2.3-4.6% failure rate) in which estPV was validated by correlation to PV; we were unable to replicate the previously reported correlation between shell estPV and PW, despite establishing a constant relationship between PW and PV, with a tissue density close to 1.0 (as in liver (Overmoyer et al., 1987)). Indeed, no method of estPW or estPV demonstrated biological correlation with PW, although several methods approached statistical significance. While this may be a result of the relatively small sample size studied and suboptimal reproducibility (see below), we believe that this currently prohibits use of PW to validate estPV, and prevents generation of a dimensionless third trimester placental quotient (FPRw) as promoted by Collins et al. (2013a). While estimated and true placental size values are correlated, often with highly statistically significant values, they are not equal with 3D estPV being smaller and estD and 2D estPV Page 104 of 253 estL, estW, and larger than their ex vivo correlates. estD may potentially be inflated by placental blood (Porat et al., 2013) whilst the other measures may be artificially reduced by “missed” placental tissue. Furthermore the strength of the relationship is lower here than predicted from previous studies (Azpurua et al., 2010). The estPV range by all techniques in this study is similar to that published by de Paula et al. (2008). This implies that sonographic placental measurements should be compared against in vivo, rather than ex vivo reference curves. Our study further suggests that development of clinically useful in vivo reference curves may be impaired by the intra- and inter-observer variability in these measures. 2D ultrasound reliability data are limited with a singular study of second trimester placental biometry reporting very high reliability (ICCs ≥0.92) (Milligan et al., 2014). Similarly by the VOCAL technique is reported to demonstrate intra- and inter-observer ICCs≥0.88 in the first and second trimesters (Deurloo et al., 2007a; Cheong et al., 2010), and between 12 and 40 weeks gestation (heavily influenced by measurements obtained <28 weeks gestation) (de Paula et al., 2008). However, Jones et al. (2011) reported significantly lower reliability of estPV with the same technique at 12 weeks gestation (ICC 0.59) with wide limits of agreement, more in keeping with the results presented in this study. We believe that this reflects the stringent assessment methods used by ourselves and by Jones et al., (2011) combined with increasing difficulty in identification and imaging of the placental borders in a single volume at advanced gestation. Whether the degree of biological correlation and reliability demonstrated in our study is sufficient to detect relatively subtle differences (e.g. a less than three centimetre difference in placental length and width [Unpublished work, Chapter 2] in placental size remains to be seen, although both Pomorski et al. (2012) and Artunc Ulkumen et al. (2014a) were each able to detect a 92cm3 reduction in VOCAL estPV between third trimester FGR and control pregnancies. Formalised training in placental measurement techniques is likely to be required to improve consistency of size estimation if implemented in a clinical study (Fries et al., 2007). Page 105 of 253 The major strength of this study is the like-for-like validation of placental size estimation techniques with a short scan-to-delivery interval. This has resulted in the development of a collection of biologically correlated placental measurements that can be tested prospectively in future studies. Immediate validation of estimated mid-trimester placental biometry in ongoing pregnancies is not possible, but there is little reason to suspect that these methodologies would not remain valid at earlier gestations. The study design was kept as similar to the real-life clinical situation as possible, assessing placentas in advanced gestation, irrespective of placental site and maternal body mass index. The study population also incorporated the full range of fetal sizes, in an ethnically diverse cohort, making the study findings generally applicable in a wide range of health care settings. The primary limitations of the study are that each method of placental size estimation demonstrated reliability, particularly in relatively poor intra- and estL inter-observer measurement, with consequent implications for implementation in clinical practice, and the cross-sectional nature, sample size and high-risk population of the study, which prevent generation of centile charts. Conclusions With increasing interest in antenatal placental assessment to identify the potentially compromised pregnancy, this study provides evidence that placental size can be estimated in the third trimester with biological correlation. The biological relevance of these results remains untested, but may be limited by the suboptimal reliability of these measurements. Thus, reliability of in utero placental size estimates needs to be improved before implementation in prospective studies to determine if placental size assessment is useful in the prediction of pregnancy outcome. Page 106 of 253 3.6 Acknowledgements The authors would like to thank the women who participated in this study and the midwives at St Mary’s Hospital, Manchester for their assistance in participant recruitment and placental collection after birth. Dr L Higgins is supported by an Action Medical Research Training Fellowship and a Manchester NIHR Biomedical Research Centre fellowship. The study was also supported by Tommy’s - the Baby Charity. 3.7 Statement of author contributions The project was conceived by EDJ, AEPH, CPS and LH, and methodologies planned by LH with expert supervision from EDJ (sonography). LS performed ultrasound scans for analysis of interobserver reliability. LH performed all other ultrasound scans and all other analyses. Statistical analysis and manuscript preparation was performed by LH. All authors were involved in writing the paper and had final approval of the submitted and published versions. Page 107 of 253 CHAPTER 4: FETOPLACENTAL ARTERIAL DOPPLER VALIDATION IN HUMAN PREGNANCY Higgins LE, Heazell AEP, Simcox L, Wareing M, Naa A, Sibley CP, Johnstone ED. A manuscript prepared for submission to the journal Clinical Science Page 108 of 253 4.1 Abstract Background: The clinical utility of umbilical artery Doppler velocimetry in late gestation is limited by lack of understanding of what aspect(s) of placental vascular structure and function it reflects. We hypothesised that placental arterial Doppler waveforms reflect placental vascularity and arterial function. Methods: Impedance was quantified in 300 third trimester pregnancies by Doppler waveform analysis (pulsatility index and resistance index) at three points along the umbilical artery, in chorionic plate and intraplacental arteries. Site-specific impedance was correlated to villous vascularity (CD31 immunostaining) and placental arterial function (wire myography). Results: Impedance decreased with proximity to the placental microvasculature (p<0.0001), gestational age (p<0.0001) and birth weight centile (p<0.05). Intraplacental artery impedance correlated significantly with villous vascularity including: vessel number (R s=-0.6, p<0.0001), luminal area (Rs≤-0.3, p<0.02), luminal ratio (R s=-0.3, p=0.03) and mean vessel area (Rs=-0.3, p<0.05). Umbilical artery free-loop impedance correlated with sensitivity to the thromboxane mimetic U46619 (50% Effective Concentration: R s≤-0.4, p<0.04) and villous vessel density (R s=-0.4, p=0.20). Reliability (Intraclass correlation coefficient) was highest at the umbilical artery abdominal insertion (0.65-0.89) and lowest at the intraplacental arteries (0.4-0.57) Discussion: The stronger correlation between intraplacental artery impedance and villous vascularity compared with that of the traditional umbilical artery free-loop Doppler suggests that measurement of the intraplacental artery Doppler may improve clinical detection of late-onset placental disease. However, the reliability of these measures needs to be improved prior to prospective studies to determine their predictive value. Key words: Placenta, Doppler, Vascularity, Vascular Function, Resistance, Fetal Growth Restriction Page 109 of 253 4.2 Introduction The severe adverse consequences of early-onset (<32 weeks gestation (Savchev et al., 2014)) fetal growth restriction (FGR) in terms of perinatal morbidity and mortality (Lees et al., 2013) are well described. In early-onset FGR an abnormal umbilical artery Doppler (UAD) waveform is often seen (Oros et al., 2011) and the risk of perinatal mortality can be stratified according to the degree of UAD waveform abnormality (Soregaroli et al., 2002; Byun et al., 2009). In later pregnancy the significance of being small for gestational age (SGA), particularly in the absence of UAD abnormalities, is much debated; some groups argue this this clinical scenario is synonymous with being “constitutionally small” i.e. appropriately small for their growth potential and that FGR is being over diagnosed amongst these infants (O'Dwyer et al., 2014). This is despite an increasing body of evidence of an increased rate of stillbirth (De Jong et al., 2000), emergency delivery for fetal distress (Parra-Saavedra et al., 2014c), neonatal metabolic acidosis (Parra-Saavedra et al., 2014c), neonatal morbidity (Parra-Saavedra et al., 2014c) and even of longer-term adverse outcome such as delayed neurodevelopment (Savchev et al., 2013; Parra-Saavedra et al., 2014b) in these pregnancies. Placentas from late-onset SGA pregnancies demonstrate increased incidence of placental histopathological abnormalities, principally related to materno- and fetoplacental underperfusion (Parra-Saavedra et al., 2014c)[Unpublished work, Chapter 2]. Thus, many authors argue that SGA birth weight in late pregnancy should be considered synonymous with lateonset FGR (McCowan et al., 2005; Gardosi et al., 2009; Gardosi and Francis, 2009). Accordingly, assessment of placental blood flow should, in theory, identify abnormalities regardless of the gestation at onset of placental disease. There are a number of reasons why assessment of placental blood flow impedance may not be considered abnormal in late-onset FGR/SGA. Firstly, the choice of the 95th centile for vascular impedance as measured by the UAD, as the arbiter of “abnormality” is not evidence-based; indeed many placentally-derived stillbirth cases demonstrate slightly higher vascular Page 110 of 253 impedance than live born infants (although still <95 th centile) (Babovic et al., 2011). This supports the work of Joern et al. (1997) who determined the 60th centile as the optimal predictor of chronic placental insufficiency in the third trimester UAD. Secondly, the vessel (or site) being assessed may not be ideal; for example the maternal element of placental vascular disease may be more sensitively detected by examination of the maternal uterine artery Doppler waveform (Madazli et al., 2003; Parra-Saavedra et al., 2014a). Furthermore the fetal element of placental vascular disease may be masked by structural or functional adaptation in the umbilical or chorionic arterial tree between the site of increased resistance and the site of Doppler waveform acquisition. Critically, there is a lack of understanding of what the Doppler waveform reflects in terms of placental vascular structure and function (Mitra et al., 1997; Mitra et al., 2000; Sagol et al., 2002; Mills et al., 2005; Wareing et al., 2005). Thus, interpreting what changes in impedance mean in relation to disease pathophysiology is not currently possible. We hypothesised that Doppler waveforms of the placental arterial circulation reflect placental vascular structure and arterial function in utero in third trimester pregnancies, and that the strength of this association is proportional to the proximity of the measurement site to the villous vascular tree. We addressed this hypothesis through examination of the biological correlation and reliability of vascular impedance by Doppler sampling site, in relation to ex vivo measures of villous vascularity and placental arterial function. Page 111 of 253 4.3 Materials and methods Ethical approval was received from Greater Manchester North West Research Ethics Committee (11/NW/0650) and the study was conducted in accordance with the Declaration of Helsinki 1975 (revised 2013). Women with non-anomalous singleton pregnancies of ≥28 weeks gestation who were undergoing third trimester ultrasound examination for assessment of reduced fetal movements, suspected or confirmed FGR or confirmation of fetal presentation were approached to participate in the study. Written informed consent was obtained to undergo additional ultrasound measurements during their scan, and to donate their placenta to the study after delivery. Sonographic assessment of the fetoplacental arterial circulation All ultrasound examinations were conducted by a single RCOG-accredited sonographer (LH) using a Voluson E6 with RA4B 4-8Hz curvilinear probe (GE Healthcare). The abdominal insertion of the umbilical cord was identified and colour Doppler traces obtained from both umbilical arteries. The process was repeated at a free-loop of the middle third of the umbilical cord and again at the placental insertion. Next four Doppler traces were obtained from arteries running along the surface of the placenta (chorionic plate arteries, CPAs). Finally, in a similar manner four traces from arteries running from the chorionic plate into the body of the placenta (intraplacental arteries, IPAs) were recorded. Sampling sites and example Doppler traces are depicted in Figure 12. Doppler traces of at least three consecutive waveforms o f consistent morphology were quantified by Pulsatility Index (PI = Peak systolic velocity (PSV) – End diastolic velocity (EDF) / time averaged maximal velocity (TA)) and Resistance Index (RI = PSV – EDF / EDF) (Figure 3). For umbilical artery sites the lowest resistance traces of each pair were taken forward for analysis. For CPA and IPA Dopplers the mean values of PI and RI obtained Page 112 of 253 from all readings were calculated to reflect total placental resistance at that level of the placental arterial circulation. A B D C E FIGURE 12: Doppler waveforms acquired from different sites in the fetoplacental arterial circulation. Showing Doppler waveforms acquired from the abdominal insertion (A), free loop (B) and placental insertion (C) of the umbilical artery, from the chorionic plate arteries (D) and from the intraplacental arteries (E) at 36 weeks gestation. Throughout the manuscript, the umbilical artery Doppler will be referred to as UAD with the suffix “-F” referring to its middle-third free-loop (equivalent to the standard clinical UAD measurement), “-A” to its abdominal insertion, and “-P” to the placental insertion. CPAD and IPAD refer to the mean Page 113 of 253 waveform at the levels of the CPAs and IPAs respectively. Biological correlation refers to the existence of a significant correlation between Doppler estimated vascular impedance (PI or RI) and villous vascularity or CPA function indices. Ex vivo examination of placental vascularity and arterial function The subcohort of placentas that were delivered within seven days of sonographic examination (“Validation cohort”) was used to validate relationships between in vivo and ex vivo measurements of placental vascular health. Placental vascular structure validation Five 1cm3 villous biopsies were taken along a randomly selected plane bisecting the placenta via the umbilical cord insertion, fixed in 10% neutral buffered formalin (Sigma-Aldrich, Poole, UK) for 18 hours at 4°C and paraffin-embedded. Endothelial cells were identified by immunoperoxidase staining using mouse monoclonal anti-cluster of differentiation 31 (CD31) antibodies (0.16μg/ml; Dako) with equal concentration non-immune mouse immunoglobulin G (Sigma-Aldrich) as a negative control, as previously described (Warrander et al., 2012). Ten images per section were obtained using a Leitz Dialus 22 Microscope (Ernst Leitz GMBH, Wetziar, Germany) at 40x original magnification with a Qicam Fast 1394 camera (Qimaging, Surrey, Canada). Total villous area was calculated for each image using the histogram function of Image ProPlus 6.0 (Media Cybernetics UK, Marlow, UK); CD31 positive structures (vessels) per field of view were manually counted vascular luminal area quantified using the area of interest tool as previously described (Hayward et al., 2011). Villous vascularity was expressed as a) number of vessels per field of view, b) villous vessel density (number of vessels / villous tissue), c) total luminal area per field of view, d) luminal ratio (total luminal area / total villous area) and e) mean vessel luminal area. The mean value per tissue section was taken to represent villous vascularity in that tissue biopsy, and the median of all five “biopsy” values taken to represent villous vascularity in the placenta as a whole. Page 114 of 253 Placental arterial function validation To assess placental arterial function small CPAs (150-500μm diameter) were dissected from the placenta. Arterial segments per placenta were loaded onto Multi Myograph System 610M wire myographs (Danish Myo Technology A/S, Aarhus, Denmark), in vessel baths containing physiological saline solution [119mmol/L NaCl, 25mmol/L NaHCO3, 4.69mmol/L KCl, 2.4mmol/L MgSO4, 1.6mmol/L CaCl2, 1.18mmol/L KH2PO4, 6.05mmol/L Dglucose, 0.034mmol/L ethylenediaminetetraacetic acid; pH7.4] (SigmaAldrich) bubbled with 5% O2 at 37°C. Resting vessel diameters were measured at L 0.9 5.1kPa according to a normalisation method described by Mills et al. (2005) adapted from Wareing et al. (2002). Up to four arterial segments per placenta were used to assess each of vessel responses to vasoactive agonists and length-tension characteristics, analysed using Myodata 2.01 software (Myonic Software, Aarhus, Denmark). Vessel responses to the thromboxane A 2 mimetic U46619 (Calbiochem, EMD Millipore, Billerica, USA) and nitric oxide donor sodium nitroprusside (SNP, Sigma-Aldrich) were assessed according to the protocol described by Mills et al. (2005). Vessel agonist sensitivity was quantified by the concentration of drug at which vessel segments achieved 50% of their maximal response to either agonist (Effective concentration, EC 50, nM). Arterial length-tension characteristics were measured according to the protocol described by Wareing et al. (2002). Vessel compliance was quantified by the rate of passive tension accumulation, expressed as Tau (the time to doubling of tension, calculated as 1/k, where k is the rate constant of the exponential curve). Maximal vessel constrictive ability was expressed as the peak active tension (PAT, mN/mm2), and vascular physiological reserve by the diameter at which PAT was achieved (DiamPAT, % of normalised diameter). For each measure of vascular function the mean value of all four arterial segments per placenta was taken to represent placental arterial function. Page 115 of 253 Assessment of fetoplacental arterial Doppler reliability Reliability of placental arterial Doppler assessment was quantified in a series of scans conducted by two RCOG-accredited sonographers (LH and LS). Three sets of Doppler waveforms were obtained (LH 1, LS, LH2), with each sonographer blinded to the waveforms obtained in each previous assessment. Waveforms were quantified offline in random order, blinded to participant identity and waveform pairing. Intra-observer reliability was assessed by comparison of LH 1 and LH2 values, whilst inter-observer reliability was assessed by comparison of LH 1 and LS values (Bland and Altman, 2003). Statistical analysis Data analysis was performed using Prism 6 for Mac OS X (Graphpad software Inc., San Diego, USA). The validation cohort was compared to the remainder of the total study cohort by Mann-Whitney U Test (for continuous data) and Chi squared test with Yates’ correctio n as required (for categorical data). Sonographic accuracy was assessed by the relationship between estimated fetal weight (EFW) according to the Hadlock C formula (1985) and true birth weight in the validation cohort (Spearman Rank Correlation). Between-site differences in PI and RI were compared by Kruskal-Wallis test with Dunn’s post hoc test. Biological correlation between PI and RI and continuous variables was tested by Spearman Rank Correlation. Statistical significance was determined by p<0.05. The strength of that biological correlation was determined by the highest correlation coefficient (Rs) of each set of measurements. Reliability (intra- and interobserver) of PI and RI estimation at each Doppler sampling site was assessed as the coefficient of variance (CoV = standard deviation of measures / mean of measures), intra-class correlation coefficient (ICC), bias and 95% limits of agreement as recommended by Bland and Altman (2003). Based on studies correlating colour Doppler “vascularity index” and vascularity in solid tumours (Chen et al., 2002; Yang et al., 2002), it was Page 116 of 253 calculated that 40 placentas would need to be examined to validate relationships of a similar magnitude (R s=0.30–0.50) with power of 80%. Assuming a similar strength relationship between placental arterial Doppler and vascular function by wire myography as that described by Wareing et al. (2005) and Mills et al. (2005), a minimum of 25 placentas would need to be functionally examined to validate placental arterial Doppler against CPA function with 80% power. Page 117 of 253 4.4 Results Participants (N=300) were enrolled in the study between August 2012 and May 2014; Table 8 summarises their maternal and pregnancy characteristics. The validation cohort (N=40 delivered within seven days) differed from the remaining study participants, in terms of potential relevance to placental structure and function, only by gestational age (being on average 12 days more advanced in pregnancy at the time of examination). Failure to obtain Doppler waveforms of sufficient quality occurred with the following frequencies; UAD-A 32%, UAD-F 0%, UAD-P 10%, CPAD 2% and IPAD 4%. EFW was highly correlated to birth weight (Rs=0.79, p<0.0001) indicating that the sonographic accuracy of the study in general was good. PI and RI decreased from UAD-A to IPAD (p<0.0001)(Figure 13) with the IPAD PI and RI respectively being 25% (13–37%) and 19% (11–28%) lower compared to their matched UAD-A PI and RI. Gestational age was inversely related to both PI and RI at each Doppler site (Rs=-0.31 to -0.50, p<0.0001). EFW centile at the time of Doppler examination did not correlate with PI or RI at any site (p>0.05), although an inverse relationship was demonstrated between PI and RI and subsequent individualised birth weight centile (R s=0.13 to -0.21, p<0.048; Supplementary Table 3). Placental vascular structure validation Where present, all relationships between Doppler PI and RI and measures of villous vascularity were inverse in nature, with less vascular placentas associated with higher PI and RI in the placental arterial circulation (Table 9). All measures of villous vascularity except vessel density correlated significantly with IPAD PI and RI; the strongest relationship was seen between IPAD RI and vessel number per field of view (Figure 14). Other Doppler sampling sites demonstrated a less consistent relationship with villous vascularity. UAD-F PI and RI correlated with vessel density (PI: R s=0.36, p=0.022. RI: Rs=-0.38, p=0.015) whilst their relationships to vessel Page 118 of 253 TABLE 8: The fetoplacental arterial Doppler study cohort. Subgroup N Maternal Characteristics Age (years) Ethnicity: Caucasian Asian Black Other Not stated BMI (kg/m2) Parity (number) Whole cohort 300 Validation p 40 Nonvalidation 260 29 (25 – 33) 30 (25 – 34) 29 (25 – 33) 0.77 0.14 192 (64%) 50 (17%) 34 (11%) 17 (6%) 7 (2%) 26 (23 – 30) 0 (0 – 1) 22 (55%) 7 (18%) 5 (13%) 1 (3%) 3 (13%) 26 (23 – 29) 0 (0 – 1) Pregnancy Characteristics Gestation at Scan 37+4 39+0 (weeks +days ) (34+3 – 39+3) (38+0 – 40+3) Placental site: Anterior 101 (34%) 8 (20%) Lateral 85 (28%) 22 (55%) Posterior 56 (19%) 2 (5%) Fundal 44 (15%) 8 (20%) Not defined 14 (5%) 0 (0%) Scan to delivery interval 11 3 (days) (3 – 33) (1-4) Individualised birth 40 40 weight centile (18 – 68) (13 – 61) 170 (65%) 43 (17%) 29 (11%) 16 (6%) 4 (0%) 26 (23 – 30) 0 (0 – 1) 37+2 (34+3 – 39+1) 0.71 0.63 <0.0001 0.00022 93 (36%) 63 (24%) 54 (21%) 36 (14%) 14 (5%) 14 (5 – 37) 38 (18 – 69) <0.0001 0.37 Data are presented as median (interquartile range) or number (percentage). Subgroup analysis (by Mann-Whitney U Test and Chi Squared test) was performed between those included and not included in the validation cohort. Individualised birth weight centile calculated using Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK). number demonstrated a non-significant inverse trend (PI: R s=-0.29, p=0.079. RI: Rs=-0.28, p=0.084). CPAD PI correlated with number of vessels per field of view, but not significantly with vessel density (Vessel number: Rs=-0.32, p=0.048. Vessel density: R s=-0.30, p=0.064) and CPAD RI was unrelated to villous vascularity, showing only a trend to correlation with vessel number that did not reach statistical significance (R s=-0.29, p=0.06). Placental arterial function validation Of all Doppler sampling sites, Only UAD-F PI and RI consistently and Page 119 of 253 A **** **** 0.8 1.5 1.0 0.5 0.0 Resistance Index 1.0 Impedance Impedance B Pulsatility Index 2.0 0.6 0.4 0.2 UAD-A UAD-F UAD-P CPAD IPAD 0.0 UAD-A UAD-F UAD-P CPAD IPAD FIGURE 13: Relationship of arterial flow impedance to proximity to the placental microvasculature. Pulsatility index (A) and resistance index (B) are shown as median and interquartile range and compared by Kruskal-Wallis test. Both impedance measures are significantly different between all sampling sites (p<0.0001). Key: UAD = umbilical artery Doppler; -A = abdominal insertion, -F = free loop, -P = placental insertion. CPAD = chorionic plate artery Doppler. IPAD = intraplacental artery Doppler. **** indicates p<0.0001. significantly related to any marker of placental arterial function (Table 10), being inversely correlated to EC 50 U46619 (PI: Rs=-0.42, p=0.039. RI: Rs=0.46, p=0.022) such that placentas with increased sensitivity to thromboxanes (lower EC50 concentrations) had higher UAD-F PI and RI values. The IPAD PI, but not RI, inverse relationship with Tau reached statistical significance (PI: R s=-0.44, p=0.044. RI: Rs=-0.35, p=0.11) whilst several other correlations approached, but did not reach, statistical significance (Table 10). Placental arterial Doppler reliability assessment Intra- and inter-observer reliability of UAD-A, UAD-P, CPAD and IPAD PI and RI were assessed in 46 scans. Table 11 summarises the reliability indices of PI and RI at each sampling site. For all measures the variability in measurements (intra- and inter-) was significant, with only UAD-A PI and RI demonstrating an inter-observer ICC>0.75. Intra-observer variability (Figure 15) was less than inter-observer variability (Figure 16) although the limits of agreement were wide for all measures. Page 120 of 253 TABLE 9: Relationships between placental arterial Doppler waveforms and villous vascularity indices. UAD-A UAD-F UAD-P CPAD IPAD Vessel Number Vessel Density Total Luminal Area Luminal Ratio Mean Vessel Area (Number) (Vessels/mm2) (μm2) (% Villous area) (μm2) R2 p R2 p R2 p R2 p R2 p PI -0.061 0.75 -0.18 0.36 -0.054 0.78 -0.070 0.72 -0.046 0.81 RI -0.12 0.53 -0.23 0.22 -0.10 0.60 -0.11 0.56 -0.093 0.63 PI -0.29 0.073 -0.36 0.022 -0.068 0.68 -0.039 0.81 0.10 0.53 RI -0.28 0.084 -0.38 0.015 -0.025 0.88 0.0042 0.98 0.16 0.33 PI -0.22 0.19 -0.13 0.44 -0.15 0.38 -0.15 0.36 -0.096 0.57 RI -0.19 0.27 -0.14 0.42 -0.14 0.40 -0.16 0.35 -0.081 0.63 PI -0.32 0.048 -0.14 0.39 -0.27 0.10 -0.24 0.14 -0.22 0.18 RI -0.29 0.069 -0.14 0.39 -0.22 0.18 -0.19 0.24 -0.16 0.34 PI -0.60 <0.0001 -0.23 0.16 -0.38 0.017 -0.33 0.040 -0.32 0.049 RI -0.62 <0.0001 -0.26 0.11 -0.39 0.013 -0.35 0.030 -0.32 0.048 Pulsatility index (PI) and resistance index (RI) obtained from five placental arterial Doppler sampling sites within seven da ys of delivery were compared to villous vascularity in N=40 pregnancies by Spearman Rank Correlation. Bold text indicates the presence of statistically significant associations. Key: UAD = umbilical artery Doppler. –A = abdominal insertion. –F = free loop . –P = placental insertion. CPAD = chorionic plate artery Doppler. IPAD = intraplacental artery Doppler. Page 121 of 253 FIGURE 14: Relationship of intraplacental artery Doppler impedance to villous vascularity. N=40, Spearman Rank Correlation between intraplacental artery Doppler pulsatility index (closed circles, solid line) or resistance index (open circles, broken line) and vessel number per field of view (A), vascular density (B), total luminal area (C), luminal ratio (lumin al area / villous area) (D) and mean vessel area (E) from anti-CD31 immunostained villous tissue images captured at x40 original magnification. Page 122 of 253 TABLE 10: Relationships between placental arterial Doppler waveforms and vascular functional assessment. UAD-A UAD-F UAD-P CPAD IPAD EC50 U46619 EC50 SNP Tau PAT DiamPAT (nM) (nM) (1/k) (mN/mm2) (%) Rs p Rs p Rs p Rs p Rs p PI 0.045 0.98 -0.20 0.48 0.18 0.54 -0.18 0.53 0.47 0.094 RI 0.054 0.85 0.00 1.00 0.033 0.91 -0.15 0.61 0.50 0.072 PI -0.42 0.039 0.38 0.073 -0.011 0.96 -0.029 0.89 -0.040 0.85 RI -0.46 0.022 0.35 0.10 -0.084 0.69 -0.15 0.49 -0.15 0.46 PI 0.086 0.74 0.11 0.66 -0.10 0.69 -0.18 0.47 0.27 0.27 RI 0.018 0.94 0.11 0.67 -0.082 0.75 -0.24 0.34 0.22 0.37 PI 0.039 0.86 0.11 0.61 -0.29 0.16 -0.045 0.83 -0.14 0.51 RI -0.0061 0.98 0.074 0.73 -0.30 0.15 -0.074 0.72 -0.12 0.55 PI -0.057 0.81 -0.058 0.80 -0.44 0.044 -0.24 0.29 -0.18 0.43 RI -0.053 0.82 -0.047 0.84 -0.35 0.11 -0.012 0.31 -0.23 0.31 Pulsatility index (PI) and resistance index (RI) obtained from five placental arterial Doppler sampling sites within seven da ys of delivery were compared to chorionic plate artery functional indices in N=25 pregnancies by Spearman Rank Correlation. Bold text indicates the presence of statistically significant associations. Key: UAD = umbilical artery Doppler. –A = abdominal insertion. –F = free loop. –P = placental insertion. CPAD = chorionic plate artery Doppler. IPAD = intraplacental artery Doppler. EC 50 = concentration of drug required to achieve 50% of a vessel’s maximal response to the same drug. U46619 = thromboxane mimetic. SNP = sodium nitroprusside. Tau = rate of passive tension accumulation. PAT = peak active tension. DiamPAT = diameter at which PAT is achieved, as a percentage of the vessel’s resting diameter. Page 123 of 253 TABLE 11: Intra- and inter-observer reliability of placental arterial Doppler impedance indices. Measure Pulsatility Index UAD-A UAD-F UAD-P CPAD IPAD Resistance Index UAD-A UAD-F UAD-P CPAD IPAD CoV (%) ICC r p Intra Inter Intra Inter Intra Inter Intra Inter Intra Inter 9.62 9.61 10.44 10.29 10.77 9.44 8.49 9.79 9.58 12.66 0.77 0.76 0.75 0.50 0.71 0.62 0.67 0.48 0.55 0.41 <0.0001 <0.0001 <0.0001 0.019 0.0010 0.0020 <0.0001 0.021 0.0060 0.053 Intra Inter Intra Inter Intra Inter Intra Inter Intra Inter 6.23 6.91 5.99 6.02 6.33 5.40 5.07 5.90 6.99 8.78 0.89 0.65 0.66 0.55 0.73 0.61 0.68 0.60 0.57 0.40 <0.0001 0.0010 0.0010 0.0080 <0.0001 0.0020 <0.0001 0.0020 0.0040 0.058 Intra- and inter-observer reliability was assessed in a series (N=46) of scans by two observers. Variability between observations is displayed as the coefficient of variance (CoV) and intraclass correlation coefficient (ICC). Key: UAD = umbilical artery Doppler. –A = abdominal insertion. –F = free loop. –P = placental insertion. CPAD = chorionic plate artery Doppler. IPAD = intraplacental artery Doppler. Page 124 of 253 B 100 50 0 -50 0.5 1.0 1.5 Average UAD-F PI -100 C Intra-observer difference (%) 100 50 -50 -100 0.5 100 50 0 -50 0.5 1.0 Average IPAD PI 1.5 0.6 0.7 0.8 Average UAD-F RI -100 D IPAD PI 0 UAD-F RI Intra-observer difference (%) Intra-observer difference (%) UAD-F PI IPAD RI Intra-observer difference (%) A 100 50 0 -50 0.4 0.5 0.6 0.7 Average IPAD RI -100 FIGURE 15: Intra-observer reliability of placental arterial Doppler impedance estimates. Bland-Altman plots demonstrate no systematic bias (p>0.05), but suboptimal within-observer reliability of placental arterial resistance replicates as demonstrated by wide scatter in a series (N=46) of repeated measur es by one observer; umbilical artery free loop Doppler (UAD-F) pulsatility index (PI) (A), UAD-F resistance index (RI) (B), intraplacental artery Doppler (IPAD) PI (C) and IPAD RI (D). The mean of both readings from one observer is shown on the x-axis, and the between reading difference (expressed as a percentage of the mean of both readings) on the y-axis. The bias between observations is depicted by a solid line (where not visible, overlying y=0); the 95% limits of agreement are depicted by a broken line. Page 125 of 253 50 0 -50 0.8 1.0 1.2 1.4 Average UAD-F PI -100 D 100 50 0 -50 -100 0.5 Inter-observer difference (%) Inter-observer difference (%) 100 IPAD PI Inter-observer difference (%) C B UAD-F PI 1.0 Average IPAD PI 1.5 Inter-observer difference (%) A UAD-F RI 100 50 0 -50 0.5 0.6 0.7 0.8 0.9 1.0 Average UAD-F RI -100 IPAD RI 100 50 0 -50 0.4 0.5 0.6 0.7 Average IPAD RI -100 FIGURE 16: Inter-observer reliability of placental arterial Doppler resistance estimates. Bland-Altman plots demonstrate no systematic bias (p>0.05), but suboptimal between-observer reliability of placental arterial resistance replicates as demonstrated by wide scatter in a series (N=46) of scans performed by two observers; umbilical artery free loop Doppler (UAD-F) pulsatility index (PI) (A), UAD-F resistance index (RI) (B), intraplacental artery Doppler (IPAD) PI (C) and IPAD RI (D). The mean of both readings from both observers is shown on the x-axis, and the between reading difference (expressed as a percentage of the mean of both readings) on the y-axis. The bias between observers is depicted by a solid line (where not visible, overlying y=0); the 95% limits of agreement are depicted by a broken line. Page 126 of 253 4.5 Discussion In this study we show that vascular impedance can be quantified by Doppler waveform analysis at multiple sites in the placental arterial circulation, and that this impedance varies dependent on proximity to the placental microcirculation, gestation and subsequent birth weight centile. Furthermore, impedance indices from the IPAD were most closely related to villous vascularity and those from the UAD-F to thromboxane sensitivity of CPAs. The finding of an impedance gradient along the umbilical cord and into the placental arteries is in keeping with work by Sonesson et al. (1993), Acharya et al. (2005a), Khare et al. (2006) and Cohen et al. (2014) in the cord itself and may reflect the relative vascular pathway length from placental microvasculature to the Doppler sampling point (Kingdom et al., 2000). Far from being a passive conduit, the umbilical artery has been shown to contribute up to 30% of total vascular resistance in animal models; in humans its contribution is potentially higher due to the relatively longer cord length (Kingdom et al., 2000). Umbilical artery vascular responses to a range of stimuli including prostaglandins and thromboxanes have additionally been demonstrated (Tuvemo and Strandberg, 1975; Crichton et al., 1993). For accurate comparison of placental vascular impedance within and between pregnancies, and thus for prognostication, use of a fixed-point umbilical artery sampling site has been suggested (Figueras et al., 2006; Khare et al., 2006). Our data suggest that intra- and inter-observer reliability may be improved using UAD-A, although there is an increased failure rate of UAD-A waveform acquisition compared with the UAD-F. We have confirmed that measures of placental arterial impedance show an inverse relationship with gestational age, in keeping with the expansion of the peripheral placental vascular tree in the third trimester of pregnancy (Kingdom et al., 2000) and mirroring the findings of Thompson and Trudinger (1990) using mathematical modelling of resistance within Page 127 of 253 circuits. This may also explain why Martinez et al. (1995) failed to demonstrate a significant UAD-A to UAD-P resistance gradient in the first trimester (when the placental vascular tree is small and pathway length is short). Given that FGR is known to be associated with reduced peripheral placental vascularity (Junaid et al., 2014) it is somewhat surprising that no relationship at all was detected between PI or RI at any Doppler sampling site and EFW centile at the time of examination; a weak relationship with eventual birth weight centile was shown. This may mean that UAD measurements are predictive of final growth though this hypothesis requires prospective testing. We have demonstrated that IPAD vascular impedance reflects villous vascularity more closely and consistently than at the other Doppler sampling sites studied. Thus IPAD may offer improved detection of placental microvascular disease associated with late-onset FGR than the other Doppler sampling sites. In addition to reduced capillary volume and density (Mayhew et al., 2003; Junaid et al., 2014) [Unpublished work, Chapter 2], an increased incidence of thrombo-occlusive lesions is reported amongst FGR placentas (Sugimura et al., 2001), which would theoretically increase resistance in the microvasculature. However, animal studies demonstrate that a significant proportion of the placental vasculature must be occluded before impedance significantly rises (Morrow et al., 1989). Our study did not examine vascular patency; biopsies were not perfusion-fixed or stereologically processed and this may have reduced the power of the study to detect differences in these aspects of villous vascularity. Like Sagol et al. (2002), we observed a much less marked association between villous vascularity and impedance at the level of UAD-F, being only correlated with vascular density and thus the UAD-F waveform may relate more to total placental size (thus total villous vascularity) and the adaptation (e.g. by vessel dilatation) of the placental arterial tree proximal to the site of increased resistance, rather than its “unit vascularity”. Page 128 of 253 In contrast to the findings of Mills et al. (2005), we have demonstrated a significant association between UAD-F impedance and CPA thromboxane sensitivity; the placentas with CPAs most sensitive to thromboxanes are those with the highest UAD-F PI and RI values. The relative importance of CPA thromboxane and nitric oxide sensitivities in determining placental vascular resistance (in physiological and pathophysiological states) may explain why UAD-F monitoring performs better for early- rather than lateonset placental conditions (Oros et al., 2011). CPAs from pregnancies complicated by (predominantly early-onset) FGR demonstrate increased sensitivity and maximal response to thromboxane stimulation (Mills et al., 2005). However, studies of late-onset pregnancy complications have failed to replicate the same vascular phenotype [Unpublished work, Chapter 2]. Possible explanations for the lack of significant association between EC 50 U46619 and UAD-F vascular impedance in the study by Mills et al. (2005) may be the high proportion of severe early-onset FGR placentas (in which chronic exposure to increased levels of thromboxane(s) may have resulted in relative insensitivity (Read et al., 1999)), and the relative hyperoxic conditions (20% O2) under which CPA function was assessed (Wareing et al., 2006). Being distal to the CPAs, it is perhaps unsurprising that the IPAD does not reflect CPA agonist sensitivity, however impedance at the UAD -A, UAD-P and CPAD might reasonably have been expected to demonstrate a similar relationship to vascular function. The higher failure rate of Doppler waveform acquisition at the UAD-A and UAD-P and resulting lower sample number may be responsible for this, whilst differences in the PI and RI at the level of individual CPAs may have been too subtle to detect in the four vessel CPAD, becoming apparent only in UAD lying upstream of the CPAs. The IPAD (and to a lesser extent the CPAD) also showed a trend to significant inverse correlation with vessel distensibility (Tau). This is of further clinical interest and in keeping with the work of Mills et al. (2011) that reports increased passive tension accumulation (and therefore decreased vessel distensibility) in placentas of a mixed early- and late-onset FGR population. This may be Page 129 of 253 related to the thickness of vessel walls (Mitra et al., 1997; Mitra et al., 2000), and consequent luminal area. It is possible that a larger validation cohort might detect significance in these areas. The predominant strengths of this study are the well-described cohort with a high frequency of late-onset placental disease and combined structural and functional validation (with short scan to delivery interval). Risk prediction is notoriously difficult in this group, but may potentially have th e greatest impact on perinatal mortality. The improved understanding of what Doppler indices tell us about placental vascular structure and function could assist this risk prediction, although the study’s findings could be further strengthened by perfusion studies. Additionally, the sonographic methods employed in the study use standard and widely available two -dimensional ultrasound and are devoid of the limitations of three-dimensional power Doppler that has been used by other groups to examine in utero placental blood flow (Raine-Fenning et al., 2008b). The primary limitation of the study is the relatively high rate of failure to obtain UAD-A and UAD-P Doppler waveforms and the suboptimal intra- and inter-observer reliability at most Doppler sampling sites which reduced the power of this study to detect, and assess the relative strength of, relationships between Doppler waveforms at each site and placental vascular health and may also limit the clinical utility of these tests. Furthermore, centile charts for vascular impedance cannot be created from this dataset as the cohort was “skewed” towards placental disease (e.g. FGR) and cross sectional in nature. Finally, despite the association of late -onset FGR with maternal-side placental underperfusion (Parra-Saavedra et al., 2014a), validation of the uterine artery Doppler was not attempted. Examination of myometrial biopsies at the time of caesarean delivery was not feasible within this study design. Thus, this study focused primarily on assessment and validation of the fetal-side placental arterial circulation. It has been demonstrated that maternal underperfusion of the placenta leads to abnormalities of placental maturation, particularly affecting peripheral Page 130 of 253 villous vascularity (Stallmach et al., 2001) and that functional maternal and fetal placental perfusion matching occurs (Stock et al., 1980; Lang et al., 2002). Clinical Perspectives In contrast to the situation of early-onset FGR, the UAD-F is rarely abnormal in late-onset placental dysfunction related to perinatal morbidity and mortality. Critically, the vascular phenotypes of early- and late-onset FGR differ. We have demonstrated that the UAD-F primarily reflects placental arterial thromboxane sensitivity, which is not significantly altered in late onset FGR, whereas the IPAD more closely reflects villous vascularity, a predominant feature of this condition. The intra- and inter-observer reliability of IPAD is only moderate. Therefore, whilst measurement of IPAD in late pregnancy has potential to enhance detection of the at-risk pregnancy, its reliability needs to be improved prior to prospective studies to determine its predictive value before introduction to routine clinical care. Summary Statement In vivo fetoplacental arterial Doppler resistance measures are dependent on villous vascularity and chorionic plate arterial thromboxane sensitivity. Understanding these associations may improve detection of placental disease and guide future personalised treatment of placental-origin pregnancy complications, but require further development and prospective testing. Page 131 of 253 4.6 Acknowledgements The authors would like to thank the women who participated in this study and the midwives at St Mary’s Hospital, Manchester for their assistance in participant recruitment and placental collection after birth. Dr L Higgins is supported by an Action Medical Research Training Fellowship and a Manchester NIHR Biomedical Research Centre fellowship. The study was also supported by Tommy’s - the Baby Charity. 4.7 Statement of author contributions The project was conceived by EDJ, AEPH, CPS and LH, and methodologies planned by LH. LS performed the second observer scans for reliability assessment; LH performed all other sonographic assessments with EDJ providing expert sonographic supervision. NA performed quantification of luminal area; LH performed all other laboratory analyses with MW, AEPH and CPS providing expert laboratory supervision. Statistical analysis was performed by LH. All authors were involved in the preparation of the manuscript. Page 132 of 253 4.8 Supplementary data SUPPLEMENTARY TABLE 3: Relationship of Doppler site-specific vascular impedance to subsequent birth weight centile. IBC UAD-A UAD-F UAD-P CPAD IPAD R2 p PI -0.15 0.048 RI -0.19 0.011 PI -0.21 0.0002 RI -0.19 0.0009 PI -0.14 0.035 RI -0.14 0.032 PI -0.13 0.023 RI -0.15 0.012 PI -0.18 0.0023 RI -0.18 0.0035 Pulsatility index (PI) and resistance index (RI) obtained from five placental arterial Doppler sampling sites within seven days of delivery were compared to Individualised birth weight centiles (IBC) calculated using Gestation Related Optimal Weight (GROW) software version 6.7 (UK) (Gestation Network, Birmingham, UK) in N=300 pregnancies by Spearman Rank Correlation. Key: UAD = umbilical artery Doppler. –A = abdominal insertion. –F = free loop. –P = placental insertion. CPAD = chorionic plate artery Doppler. IPAD = intraplacental artery Doppler. Page 133 of 253 CHAPTER 5: PLACENTAL HORMONE CONCENTRATIONS: WHAT ASPECT(S) OF PLACENTAL FUNCTION DO THEY REFLECT? Higgins LE, Rey de Castro N, Greenwood SL, Jones RL, Sibley CP, Johnstone ED, Heazell AEP. A manuscript prepared for submission to the Journal of Clinical Endocrinology and Metabolism Page 134 of 253 5.1 Abstract Purpose: Measurement of placentally-derived hormones in the maternal circulation is attracting renewed interest in the prediction of adverse pregnancy outcome. However, its clinical utility is limited by lack of understanding of the hormones’ relationships to placental structure and function. We aimed to describe the relationship between placentally-derived circulating hormone concentrations and placental structure and function in the third trimester of pregnancy. Methods: Concentrations of human chorionic gonadotrophin (hCG), human placental lactogen (hPL), progesterone, placental growth factor (PlGF) and soluble fms-like tyrosine kinase (sFlt-1) measured by maternal serum ELISA were correlated to gestational age and birth weight centile (IBC) (N=268). The relationships between placental weight, volume, trophoblast area, gene transcription rate (assessed by qPCR), villous tissue lysate and explantconditioned medium hormone contents and serum hormone concentrations (each measured by ELISA) were examined in women who delivered ≤7 days from venepuncture (n=50). Results: I) hPL, progesterone, PlGF and sFlt-1 concentrations were related to gestational age (p≤0.0007). II) hCG, hPL, progesterone and PlGF concentrations were related to IBC (p≤0.01). III) hCG, hPL and PlGF concentrations also correlated with placental weight and volume (p≤0.03); only PlGF correlated with trophoblast area (p=0.03). IV) Weaker functional associations were detected between hCG and sFlt-1 concentrations and relative transcription rate (p=0.03) and tissue lysate content (p=0.007) respectively. Conclusions: Placental endocrine biomarkers provide biologically relevant information relating to placental macro- and microstructure and function. These tests merit prospective evaluation to determine whether they can improve detection of fetal compromise secondary to placental insufficiency. Key Words: human chorionic gonadotrophin; human placental lactogen; Progesterone; Placental growth factor; Soluble fms-like tyrosine kinase-1; Placental structure; Placental weight; Page 135 of 253 Trophoblast area; Transcription; Translation; Hormone secretion Page 136 of 253 5.2 Introduction It is estimated that 2.6 million babies are stillborn each year worldwide (Lawn et al., 2011). The bulk of these deaths occur in low- and middleincome countries, however stillbirth rates in high-income countries remain significant (1:250-300 pregnancies of at least 28 weeks gestation) and have not declined in recent decades (Flenady et al., 2011b). One quarter of stillbirths in high-income countries occur at 37 weeks or more gestation (with an additional 20% between 32-37 weeks gestation) (Zeitlin, 2013); in these pregnancies intervention (delivery) would have been an option to avoid stillbirth with good intact survival rates (Draper et al., 1999). Yet there is currently no good test to predict which babies are at highest risk of stillbirth and therefore who would benefit from early delivery (Smith, 2010). Placental disease is present in up to 65% of all stillbirths (Ptacek et al., 2014) and it has been proposed that better identification of placental dysfunction could decrease perinatal mortality (Heazell et al., 2015). Indeed retrospective analysis of cohorts undergoing early-pregnancy screening for Trisomy 21 has described associations between stillbirth and early pregnancy maternal circulating concentrations of placentally-derived hormones (including human chorionic gonadotrophin (hCG), PAPP-A, alfafetoprotein, unconjugated oestriol and inhibin) (Gagnon et al., 2008). Fewer studies have examined the “diagnostic” potential of placental biomarkers (including human placental lactogen (hPL), hCG, progesterone and placental growth factor (PlGF)) in third trimester prediction of adverse pregnancy outcome including fetal growth restriction (FGR) and stillbirth, at a time when delivery may be possible to prevent in utero fetal demise and any causative placental pathology is likely to be already present (Letchworth et al., 1978; Isouard, 1979; Morrison et al., 1980; Romero et al., 2008; Benton et al., 2012; Dutton et al., 2012; Chaiworapongsa et al., 2013; Chappell et al., 2013). Commonly these later gestation studies demonstrate lower maternal serum placentally-derived hormone concentrations in pregnancies ending in adverse outcome. Page 137 of 253 Whilst the Cochrane systematic review and meta-analysis (Neilson, 2012) of such tests failed to demonstrate significant benefit from the biochemical screening for stillbirth prevention (being underpowered to detect such benefit as it included one study evaluating a single hormone, oestradiol, in just 622 pregnancies), placental hormone tests are increasingly being incorporated into clinical care (Dutton et al., 2012; Chappell et al., 2013; Heazell et al., 2013). Without adequate understanding of what these hormone concentrations actually reflect in terms of placental health the clinical utility of measurement of these placental biomarkers (and any subsequent intervention) is limited, being at best not evidence-based, and at worst potentially leading to harm (Hibbard et al., 2010; MacKay et al., 2010). Rigorous systematic evaluation of the relationships between maternal circulating placental hormone concentrations and corresponding placental structure and function is required to further develop these assays as clinically useful tests. We hypothesised that maternal serum placental hormone profiles in the third trimester of pregnancy reflect both structure and endocrine function of the placenta. We aimed to test this hypothesis by systematic evaluation of the relationship between maternal circulating placentally-derived hormone concentrations and macroscopic and microscopic placental structure and the transcription, translation and hormone release function of placental tissue. Page 138 of 253 5.3 Materials and methods Ethical approval for the study was obtained from Greater Manchester North Research Ethics Committee (11/NW/0650) and all work was carried out in accordance with the Declaration of Helsinki 1975 (revised 2013). Pregnant women of ≥28 weeks gestation in singleton pregnancies without known fetal abnormality, hypertension (Chappell et al., 2013) or diabetes (Stewart et al., 1989) were approached from the antenatal assessment unit, induction of labour bay, caesarean section pre-operative assessment clinic or at presentation with maternal perception of reduced fetal movements. All women gave written consent prior to donating maternal blood and (if delivery occurred within seven days of venepuncture) placental tissue to research. Hormone selection The hormones examined in this study (hCG, hPL, progesterone, PlGF and sFlt-1) were selected due to their syncytiotrophoblast-dependent production (Midgley and Pierce, 1962; Kaplan and Grumbach, 1965; Shore et al., 1997; Clark et al., 1998; Tuckey, 2005), and their previously demonstrated predictive value (for placental-origin disease) in the third trimester of pregnancy (Benton et al., 2012; Dutton et al., 2012; Chappell et al., 2013). Alfafetoprotein and pregnancy-associated plasma protein-A (PAPP-A) were not examined, as their relationship to pregnancy outcome is largely confined to early pregnancy (Dutton et al., 2012). Fresh tissue processing Pre-delivery maternal blood was collected into serum separator tubes (BD Vacutainer, Franklin Lakes, US) and processed per manufacturer recommendations to obtain serum. Serum was stored at -80°C. Placentas were trimmed of their extraplacental membranes and umbilical cord, weighed and their volume was calculated by fluid displacement in isotonic saline solution. Placental weight and volume were taken to represent placental macrostructure. Two sets of villous biopsies (1cm 3) from the edge Page 139 of 253 and middle of the placental body and beside the umbilical cord insertion were obtained. One set was fixed in 10% neutral buffered formalin (18 hours, 4°C) before paraffin-embedding. The other set was washed in phosphate buffered saline and divided for RNA extraction (treated with RNAsave (Geneflow, Lichfield, UK) for 18 hours at 4°C before storage at -80°C), tissue hormone content analysis (lysed in 1.5ml distilled water at room temperature for 18 hours before lysate storage at -80°C) or assessment of hormone release in a villous explant culture model for seven days as described previously (Siman et al., 2001); explants were maintained on Netwell permeable supports (Corning, via Sigma-Aldrich) at the liquidgas interface in culture media [(in 1L) 100ml CMRL-1066 (Gibco, Life Technologies, Paisley, UK), 10% heat-inactivated fetal bovine serum, 100mg L-glutamine, 50IU/ml penicillin, 50μg/ml streptomycin sulphate, 50μg/ml gentamicin, 100μg/ml hydrocortisone, 0.1μg/ml retinoic acid, 1μg/ml insulin] under conditions of 6% O2/5% CO2 (normoxic conditions for the placenta in late pregnancy) and 37°C for seven days. Daily explant conditioned media (CM) was stored at -80°C from 24 to 166 hours. The protein contents of lysed and explanted villous tissues were calculated by dissolution in 4ml 0.3M sodium hydroxide and quantification with Bio -Rad Protein Assay (Bio-Rad Laboratories, Hempstead, UK). Assessment of placental microstructure Paraffin-embedded tissue blocks (N=3 per placenta) were immunostained using mouse monoclonal antibodies against the trophoblast marker cytokeratin 7 (CK7) (Daya and Sabet, 1991) (0.9μg/ml; Dako, Ely, UK), with mouse non-immune immunoglobulin G (Sigma-Aldrich) at the corresponding equal concentration as a negative control and imaged as previously described (Warrander et al., 2012). Sections were photographed at 40x original magnification using a Leitz Dialus 22 Microscope (Ernst Leitz GMBH, Wetziar, Germany) and Qicam Fast 1394 camera (Qimaging, Surrey, Canada). Ten randomly selected images were taken per tissue section. Using the histogram function of the Image ProPlus 6.0 imaging software (Media Page 140 of 253 Cybernetics UK, Marlow, UK) total villous and trophoblast areas were quantified as previously described (Hayward et al., 2011). The trophoblast ratio (TR) was calculated (TR = total trophoblast area / total villous area). Mean values for each index across 10 images per tissue section were taken to represent each villous biopsy, and the median “biopsy” value was taken to represent each placenta. Assessment of transcription RNA was extracted from fresh villous tissue using a mirVana miRNA isolation kit followed by removal of genomic DNA using a TURBO DNAfreeTM kit (both Ambion, Austin, Texas, USA). Nucleic acid concentration and contamination were assessed by spectroscopy using the NanoDrop 2000C (Thermo Fisher Scientific, Wilmington, USA). Reverse transcription (RT) was performed in triplicate using an AffinityScript Multi-temperature RT kit (Agilent, Santa Clara, USA) in an MX3000 Thermocycler (Stratagene, La Jolla, USA). Real time quantitative polymerase chain reaction (PCR) was then performed on RT triplicates using primers specific to TATA-box binding protein and Brilliant III Ultra-fast SYBR® QPCR mastermix (Agilent) with annealing at 60°C, followed by dissociation curve analysis. All PCRs had efficiencies of 85-105%. The cycle threshold (Ct) of each sample was calculated and gene expression calculated as 2-ΔCt. Triplicates with a coefficient of variance (CoV) ≤25% were pooled for further qPCR analysis. Subsequently, qPCR was performed according to the same protocol, on pooled triplicates using primers for genes encoding TATA-box binding protein (placental house-keeping gene), hCG, hPL, PlGF and sFlt-1 along with CYP11A1, the gene encoding CYP450scc, the rate-limiting enzyme of progesterone synthesis (Tuckey, 2005). Relative mRNA expression levels were calculated relative to expression of TATA-box binding protein. All primer sequences and accession numbers are provided in Supplementary Table 1. All RT and qPCR reactions were performed simultaneously for all samples. Page 141 of 253 Assessment of hormone concentration in biofluids Hormone concentrations in matched aliquots of maternal serum, tissue lysate and daily (2 – 7) CM were quantified in duplicate, using colorimetric enzyme-linked immunosorbant assay (ELISA) kits for hCG, hPL, progesterone, PlGF and sFlt-1 (Supplementary Table 2). The optical density (OD) of the well contents was read by a multiplate reader (BMG Labtech, Aylesbury, UK) at a pre-specified wavelength (see Supplementary Table 2). Sample concentrations were quantified by comparison to an OD standard curve. The average concentration and CoV of sample duplicates were calculated; those with CoV >10% were repeated. Lysate and daily CM hormone concentrations were then normalised to the concentration of a quality control sample, and to tissue protein content, with CM concentration further adjusted to CM volume and culture time period to calculate total release per 24 hours. Daily CM hormone content was plotted against time to assess the release time course and further combined to calculate total hormone release (24 to 166 hours) from villous explants. Statistical analysis Statistical analysis was performed using GraphPad Prism 6 for Mac OS X (GraphPad Software, San Diego, USA). Differences between groups were examined using Mann-Whitney U Test for continuous data and Chi squared or Yates’ correction as appropriate for categorical data. Maternal circulating concentrations of each hormone were correlated sequentially to gestation at venepuncture, subsequent individualised birth weight centile (IBC) (Bulk centile calculator v6.7 (UK), Gestation Network, Birmingham, UK), placental weight, placental volume, villous and trophoblast areas, trophoblast ratio, relative transcription level (2 -ΔCt), villous hormone content and total hormone release by Spearman Rank Correlation. Biological correlation was defined as the presence of a significant (p<0.05) correlation with serum concentration of the hormone. The strength of that relationship was determined by the highest R s value. Page 142 of 253 A power calculation was performed to establish the required size of validation cohort. We were unable to identify previous studies that had correlated tissue endocrine function to serum hormone levels, however based on studies of placental weight and hormone concentrations we anticipated that to detect a significant correlation between indices of placental structure or function and maternal serum hormone concentrations of Rs≥0.5, with power of 80% at the level of p<0.05 would require at least 23 participants. To account for sample attrition due to i) delay from delivery to placental collection >1 hour, ii) confirmed or suspected maternal or fetal infection and iii) explant culture infection necessitating non-analysis of these tissue samples we aimed to recruit until fifty matched samples were received. Mann-Whitney U-Test was used to assess differences between subcohorts. To enable all mRNA quantification to occur simultaneously (and reduce inter-run variability), analysis was restricted to the N=40 samples with highest purity on spectroscopy. Similarly analysis of villous microstructure was restricted to the first N=40 samples received. Page 143 of 253 5.4 Results Participants (N=268) were recruited between January 2012 and May 2014, of whom matched samples were received from 50 participants delivering within seven days of venepuncture to form the validation cohort; th eir demographic and pregnancy outcome details are presented in Table 12. Pregnancies included in the validation cohort differed from the remaining study participants by gestational age (being on average 17 days more advanced in pregnancy at the time of examination), and by median IBC, which was smaller (25.3 vs. 42.2) due to a preponderance of SGA babies (p=0.01). Within the whole study cohort, gestational age at venepuncture did not correlate with maternal serum concentrations of hCG, whilst hPL, progesterone and sFlt-1 serum concentrations correlated positively and PlGF correlated inversely (Figure 17). Likewise, IBC did not correlate with maternal serum concentrations of sFlt-1, whilst hPL, progesterone and PlGF showed a positive correlation and hCG correlated inversely (Figure 18). The participant details of those validation cohort members whose samples were excluded from the mRNA quantification (N=10), villous microstructure analysis (N=10) or hormone release assessment (N=20) (per specification in the statistical analysis section) were not significantly different except in terms of fetal gender, being more likely to be male (mRNA analysis: 9/10 vs. 15/40, p=0.0088. CM: 14/20 vs. 10/30, p=0.024) and to have a significantly higher mean IBC in those excluded from the mRNA analysis (37.6 (28.4– 68.6) vs. 18.3 (7.4–40.6), p=0.044). Table 13 summarises the relationships between maternal serum placental hormone concentrations and placental structure. In terms of placental macrostructure hPL and PlGF concentrations each positively correlated with placental volume and weight, whilst hCG concentration inversely correlated with placental volume, but was not significantly correlated with placental weight. Circulating progesterone concentrations show a trend to positive correlation with placental volume and weight, but fail to reach statistical significance. sFlt-1 concentration showed no relationship with placental Page 144 of 253 TABLE 12: The placental hormone concentration study cohort. Whole cohort N 268 Maternal Characteristics Age (years) 29 (26 – 33) Ethnicity: Caucasian 178 (66%) Asian 45 (17%) Black 29 (11%) Other 16 (6%) 2 BMI (kg/m ) 25.6 (22.8 – 29.5) Parity (number) 0 (0 – 1) Smoking: Yes 38 (14%) No 229 (85%) Not stated 1 (0.4%) Pregnancy Characteristics Gestation at 37+3 venepuncture (33+6 – 39+3) +days (weeks ) Delivery interval 12 (days) (4 – 37) Laboured: Yes 231 (86%) No 37 (14%) IBC 38.4 (17.7 – 66.4) IBC <10 39 (15%) Infant gender: Male 134 (50%) Female 134 (50%) Subgroup analysis Non-validation Validation cohort cohort 218 50 29 (26 – 33) 29 (26 – 33) p 0.72 0.17 142 (65%) 35 (16%) 28 (13%) 13 (6%) 25.5 (22.7 – 29.0) 0 (0 – 1) 36 (72%) 10 (20%) 1 (2%) 3 (6%) 26.3 (23.4 – 31.6) 0 (0 – 1) 27 (12%) 191 (88%) 0 (0%) 11 (22%) 38 (76%) 1 (2%) 36+3 (33+0 - 39+1) 38+6 (38+0 – 40+3) <0.0001 20 (7 – 46) 2 (1 – 4) <0.0001 0.088 0.40 0.20 0.34 190 (87%) 28 (13%) 42.2 (19.3 – 68.5) 26 (12%) 41 (82%) 9 (18%) 25.3 (8.9 – 48.3) 13 (26%) 111 (51%) 107 (49%) 23 (46%) 27 (54%) 0.0005 0.011 0.53 Data are presented as median (interquartile range) or number (percentage). Delivery interval calculated as number of days from venepuncture to delivery. Individualised birth weight centile (IBC) calculated using Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK) . Subgroup analysis (by Mann-Whitney U Test and Chi Squared test) was performed between those included and not included in the validation cohort. Page 145 of 253 B hCG 1500 Serum hPL concentration (mg/L) Serum hCG concentration (mIU/mL) A 1000 500 0 200 220 240 260 280 hPL *** 2000 1500 1000 500 0 300 200 Gestation (days) D Progesterone **** 6000 2000 200 220 240 260 260 280 300 280 300 PlGF **** 4000 4000 0 240 Gestation (days) Serum PlGF (pg/ml) Serum progesterone (ng/ml) C 220 280 3000 2000 1000 300 Gestation (days) 0 200 220 240 260 Gestation (days) E sFlt-1 **** Serum sFlt-1 (pg/ml) 25000 20000 15000 10000 5000 0 200 220 240 260 280 300 Gestation (days) FIGURE 17: Relationship between maternal circulating hormone concentration and gestational age. N=268, Spearman Rank Correlation between gestational age and maternal serum concentration of human chorionic gonadotrophin (hCG) (A), human placental lactogen (hPL) (B), pr ogesterone (C), placental growth factor (PlGF) (D) and soluble fms-like tyrosine kinase-1 (sFlt-1) (E). *** indicates p<0.001, **** indicates p<0.0001. structure. In terms of placental microstructure only maternal serum PlGF correlated to any marker of microstructure, being positively associated with trophoblast area, and showing a trend to positive correlation to the trophoblast ratio. Compared with the observed relationship between maternal serum placental hormone concentrations and placental macrostructure, the Page 146 of 253 B hCG ** 1500 Serum hPL concentration (mg/L) Serum hCG concentration (mIU/mL) A 1000 500 0 0 20 40 60 80 hPL **** 2000 1500 1000 500 0 100 0 20 40 IBC D Progesterone **** 6000 4000 2000 0 0 20 40 100 60 80 80 100 3000 2000 1000 100 IBC E 80 PlGF * 4000 Serum PlGF (pg/ml) Serum progesterone (ng/ml) C 60 IBC 0 0 20 40 60 IBC sFlt-1 Serum sFlt-1 (pg/ml) 25000 20000 15000 10000 5000 0 0 20 40 60 80 100 IBC FIGURE 18: Relationship between maternal circulating hormone concentration and individualised birth weight centile. N=268, Spearman Rank Correlation between individualised birth weight centile (IBC) calculated using Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK) and maternal serum concentration of human chorionic gonadotrophin (hCG) (A), human placental lactogen (hPL) (B), progesterone (C), placental growth factor (PlGF) (D) and soluble fms-like tyrosine kinase-1 (sFlt-1) (E). * indicates p<0.05, ** indicates p<0.01, **** indicates p<0.0001. relationship with placental endocrine function was less consistent (Table 14). Relative transcription rates of hCG correlated with circulating serum hCG concentration, whilst in tissue lysate, only sFlt-1 correlated with its corresponding circulating hormone concentration, with hCG (p=0.11) and PlGF (p=0.16) approaching statistical significance. The temporal release patterns of each hormone from villous explants are displayed in figure 19. Release of sFlt-1 was variable with tissue from some placentas releasing 10 fold more sFlt-1 into the media than others. In particular this appeared to be Page 147 of 253 TABLE 13: Relationship of circulating maternal serum placental hormone concentrations and placental structure Volume Weight Villous Area Trophoblast Area Trophoblast Ratio N=50 N=50 N=40 N=40 N=40 Rs p Rs p Rs p Rs p Rs p hCG -0.38 0.013 -0.23 0.10 -0.0011 0.99 0.083 0.61 0.084 0.61 hPL 0.42 0.0058 0.46 0.0008 0.18 0.26 0.061 0.71 -0.078 0.63 Progesterone 0.26 0.093 0.20 0.17 0.092 0.57 -0.074 0.65 -0.15 0.36 PlGF 0.35 0.025 0.31 0.031 0.012 0.94 0.34 0.031 0.23 0.16 sFlt-1 0.081 0.61 -0.033 0.82 -0.043 0.90 -0.055 0.87 -0.18 0.60 N=50, Spearman Rank Correlation between maternal serum concentration of human chorionic gonadotrophin (hCG), human placental lactogen (hPL), progesterone, placental growth factor (PlGF), and soluble fms-like tyrosine kinase-1 (sFlt-1) and measures of placental structure. Bold text indicates the presence of statistically significant associations (p<0.05). Page 148 of 253 TABLE 14: Relationship of circulating maternal serum placental hormone concentrations placental endocrine function mRNA Lysate Media N=40 N=50 N=30 Rs p Rs p Rs p hCG -0.34 0.027 -0.23 0.11 -0.12 0.56 hPL 0.15 0.36 0.017 0.91 0.21 0.29 Progesterone 0.14 0.39 0.16 0.26 0.29 0.12 PlGF -0.025 0.88 0.20 0.16 0.13 0.52 sFlt-1 -0.065 0.69 0.38 0.0070 0.086 0.66 Spearman Rank Correlation between maternal serum concentration of human chorionic gonadotrophin (hCG), human placental lactogen (hPL), progesterone, placental growth factor (PlGF), soluble fms-like tyrosine kinase-1 (sFlt-1) and measures of placental endocrine function. Bold text indicates the presence of statistically significant associations (p<0.05). associated with birth of an SGA baby (area under the hormone release curve: SGA 430,993 (272,263–592,910) vs. non-SGA 180,647 (40,153– 346,560), p=0.036). Hormone content in CM did not correlate with circulating hormone concentrations, with only progesterone demonstrating a trend to correlation (p=0.12). Page 149 of 253 A B hCG 40 hPL Release (mg) 100 hCG Release (mIU) hPL 80 60 40 20 30 20 10 0 0 0 24 48 72 96 0 12 4 14 8 16 0 24 Hours of Culture 72 96 0 12 4 14 8 16 4 14 8 16 Hours of Culture D Progesterone PlGF 100 200 PlGF Release (pg) Progesterone Release (ng) C 48 150 100 50 80 60 40 20 0 0 0 24 48 72 96 0 12 4 14 8 16 Hours of Culture E 0 24 48 72 96 0 12 Hours of Culture sFlt-1 sFlt-1 Release (pg) 5000 4000 3000 2000 1000 0 0 24 48 72 96 0 12 4 14 8 16 Hours of Culture FIGURE 19: Hormone release time course from villous explants. Data expressed as release per milligram of protein per 24 hour period (N=30; median and interquartile range) for human chorionic gonadotrophin (hCG) (A), human placental lactogen (hPL) (B), progesterone (C), placental growth factor (PlGF) (D) and soluble fms-like tyrosine kinase-1 (sFlt-1) (E). Page 150 of 253 5.5 Discussion This study has demonstrated that third trimester maternal circulating concentrations of placentally-derived hormones, particularly hPL and PlGF, relate to placental structure. There was also a weaker association of concentrations of hCG and sFlt-1 with per-unit placental function. As these hormones have been shown to predict adverse pregnancy outcome in other cohorts, they merit further investigation in the triage of third trimester pregnancies. We show that third trimester maternal serum concentrations of these hormones continue to rise throughout the third trimester, with the exception of PlGF (which significantly declines) and hCG which demonstrates no relationship with third trimester gestational age, likely due to the equal fetal gender distribution within the study cohort (Steier et al., 2004). With the exception of hPL (concentrations of which are reported to decline from approximately 30 weeks gestation (Furuhashi et al., 1984a)) these patterns are in keeping with the published literature (Tulchinsky et al., 1972a; Romero et al., 2008). Conversely, whilst we report correlations between IBC and hCG, hPL, progesterone and PlGF concentrations (but not sFlt-1) other studies have failed to demonstrate significant relationships between these hormone concentrations and birth weight itself, probably due to the confounding effect of gestational age (Spellacy et al., 1975; Furuhashi et al., 1984a). Circulating maternal hPL concentrations should be interpreted as reflective of placental macrostructure, not function as we confirmed that maternal serum hPL concentrations are positively correlated with placental weight (Furuhashi et al., 1984a; Dutton et al., 2012) and volume, but were unable to demonstrate any relationship with placental microstructure or function. As hPL promotes nutrient availability and transfer across the placenta and fetal insulin like growth factor-1 production (Handwerger and Freemark, 2000), lower circulating hPL concentrations would compound the inherent limited Page 151 of 253 nutrient transfer capacity of a small placenta, but might limit fetal nutrient demand (by limiting fetal growth) in compensation. Circulating PlGF concentrations can be interpreted to reflect, at least in part, placental volume and weight (in keeping with the findings of Shibata et al. (2005) in SGA and preeclamptic pregnancies) and aspects of placental microstructure (reduced trophoblast volume) that are altered in FGR and SGA pregnancies (reviewed by Mayhew et al. (2009)). A lower placental lysate PlGF concentration in SGA pregnancy placentas has also been reported (Shibata et al., 2005), in keeping with the trend to positive association between circulating PlGF concentration and tissue lysate PlGF content in this study. Lower circulating PlGF concentrations could reduce non-branching angiogenesis resulting in impaired maternofetal gas exchange (Kingdom and Kaufmann, 1997). The lack of association between any examined measure of placental structure and circulating sFlt-1 suggests that the excessive circulating sFlt-1 concentration observed in preeclampsia reflects either altered placental function (rather than structure) or a non-placental origin of circulating sFlt1. As we report a positive correlation between maternal circulating sFlt-1 concentration and placental tissue lysate content and circulating sFlt-1 concentrations are reported to decline rapidly following placental delivery (Reddy et al., 2009) the former is considered more likely. As a similar association with CM sFlt-1 content and serum sFlt-1 concentration is not seen, we hypothesise that release of sFlt-1 from placental tissue is normally controlled in order to limit the adverse anti-angiogenic effects of excess sFlt1, and that this becomes dysregulated in pathological pregnancy as suggested by the wide variation in per-unit sFlt-1 release between SGA and AGA pregnancies. Given the divergent relationships between PlGF and sFlt-1 with placental size and placental function respectively, measurement of their relative concentrations in the maternal circulation (through the PlGF/sFlt-1 ratio or free circulating PlGF concentration) may offer superior Page 152 of 253 prediction of adverse pregnancy outcome over measurement of either hormone alone. We were unable to corroborate the previously reported significant correlation between placental weight and circulating progesterone (Spellacy et al., 1975; Dutton et al., 2012) and hCG (Spellacy et al., 1975; Furuhashi et al., 1984a) concentrations, likely due to our smaller sample size (N=50 vs. N=268) and enrichment of our validation population with placental disease (as reflected in the relatively low median IBC). A significant inverse correlation was found between circulating hCG concentration and placental volume (Rs=-0.38); this discrepancy in relationship between placental weight and volume may reflect differing tissue density between functioning and non-functioning placental tissue, and requires further evaluation. Placental hCG production is regulated at the level of gene transcription (Hussa, 1980). It is therefore unsurprising that a significant relationship between circulating hCG concentration and relative placental hCG mRNA expression is observed (Rs=-0.34). The effects of hCG are diverse (Cole, 2010) but promote a favourable intrauterine environment, therefore “oversecretion” of hCG may be designed to improve placental efficiency. Given that fetal and placental sizes are intrinsically linked (Roland et al., 2012) and that the observed relationship of circulating fetoplacental hormone concentrations in maternal serum to per-unit placental function in this study is much weaker than that with placental macrostructure, it is reasonable to question whether the link between placentally-derived hormone concentrations and IBC and gestational age is simply due to confounding by placental size. However, as placental growth occurs predominantly in the first and second trimesters of pregnancy (Orzechowski et al., 2014), the continued correlation of circulating placentally-derived hormone concentrations, and indeed the divergent relationship of PlGF concentration with gestation (negative) and IBC (positive), with these variables suggests at least some independent contribution from placental function. Page 153 of 253 This study is underpowered to determine the relative contributions of gestation, placental structure and function by logistic regression. Furthermore, its cross-sectional nature and the relatively high rate of SGA birth prevent normograms. the generation Pre-parturition of third-trimester changes (in which maternal serum maternal serum placentally-derived hormone concentrations have been reported to increase or decrease in the hours and days leading up to the onset of labour (Furuhashi et al., 1984b)) may also have weakened observed associations between placental structure and function and circulating hormone concentrations as a large proportion of the study population laboured shortly after venepuncture. Finally, it is not known whether the antibodies in these ELISA kits recognise free or total PlGF (Maynard et al., 2003), or indeed whether they recognise all variants of sFlt-1 (Rajakumar et al., 2009). The major strength of this study is the concurrent validation of maternal serum concentrations of placentally-derived hormones against both placental structure and function. In light of the advanced knowledge of what biologically relevant information placental endocrine biomarkers are capable of providing (particularly that relating to placental macrostructure reflected by hPL and PlGF) improvements in antenatal care may be possible. Page 154 of 253 5.6 Acknowledgements The authors would like to thank the women who participated in this study and the midwives at St Mary’s Hospital, Manchester for their assistance in participant recruitment and placental collection after birth. Dr L Higgins is supported by an Action Medical Research Training Fellowship and a Manchester NIHR Biomedical Research Centre fellowship. The study was also supported by Tommy’s - the Baby Charity. 5.7 Statement of author contributions The project was conceived by EDJ, AEPH, CPS and LH, and methodologies planned by LH. NRdC performed quantification of trophoblast area. LH performed all other laboratory analyses with expert supervision from RLJ (PCR) and SLG (Explant culture and ELISA). Statistical analysis was performed by LH. All authors were involved in the preparation of the manuscript. Page 155 of 253 CHAPTER 6: PLACENTAL ASSESSMENT PREDICTS ADVERSE PREGNANCY OUTCOME AFTER REDUCED FETAL MOVEMENT: A COHORT STUDY Higgins LE, Myers JE, Sibley CP, Johnstone ED , Heazell AEP. A manuscript prepared for submission to the journal Obstetrics and Gynecology Page 156 of 253 6.1 Abstract Objective: To assess the value of placental assessment in the prediction of adverse pregnancy outcome after presentation with reduced fetal movements. Methods: Three hundred singleton pregnancies of 28–42 weeks gestation presenting with subjective reduction in fetal activity were assessed by ultrasound and endocrine measurements. Results: 56 (18.8%) of studied pregnancies ended in adverse outcome. 187 recorded variables were reduced based on univariate association with adverse pregnancy outcome (p<0.10). The remaining 49 variables were further reduced using logistic regression (after adjustment for gestational age and estimated fetal weight centile) to 15 variables with potential predictive value (p<0.05). Multiple logistic regression with backwards elimination generated three proposed predictive models, sequentially adding a) umbilical artery (UAD) free loop RI, UAD placental insertion PI >90th centile and free PlGF concentration (Model 1), b) plus UAD abdominal insertion PI >75 th centile (Model 2), and c) plus UAD free loop PI >95 th centile (Model 3) to the baseline model of care (gestational age and estimated fetal weight centile). ROC analysis of each proposed model demonstrated significantly improved performance (AUC≥0.86, p<0.05) compared with baseline care (AUC=0.75). At a fixed specificity of 99%, Model 3 demonstrated superior performance in terms of sensitivity (37.5%), positive and negative predictive values (90.0% and 87.5% respectively) and positive and negative likelihood ratios (39.9 and 0.63 respectively). Conclusion: This study supports the hypothesis that antenatal placental assessment might improve identification of pregnancies with reduced fetal movements at highest risk of adverse pregnancy outcome. Keywords: Reduced fetal movements, Adverse pregnancy outcome, Stillbirth, Fetal growth restriction, Placental volume, Placental length, Placental width, Placental depth, Placental area, Fetoplacental ratio, Umbilical cord eccentricity, Umbilical Page 157 of 253 artery Doppler, Uterine artery Doppler, Middle cerebral artery Doppler, Chorionic plate artery Doppler, intraplacental artery Doppler, human chorionic gonadotrophin, human placental lactogen, Progesterone, Placental growth factor, Soluble fmslike tyrosine kinase, Prediction, Sensitivity, Specificity. Précis Multi-modality placental assessment in utero improves the identification of pregnancies at highest risk of stillbirth by up to 28.6% after presentation with reduced fetal movements. Page 158 of 253 6.2 Introduction Up to one in 250 pregnancies in high-income countries ends in stillbirth (Flenady et al., 2011b), a quarter of which occur after 37 weeks gestation (MacDorman, 2012; New Zealand Ministry of Health, 2012; Zeitlin, 2013) and are potentially preventable by delivery. Currently there is no accurate clinical test to predict which pregnancies are at highest risk of fetal death (Smith, 2010). Placental assessment has been proposed as a means to improve the prediction of adverse pregnancy outcome (APO) (Heazell et al., 2015). Women who present with reduced fetal movements (RFM) represent a population with an increased risk of stillbirth, and the related condition of fetal growth restriction (FGR) (O'Sullivan et al., 2009; Pagani et al., 2014a; Pagani et al., 2014b). Placentas from RFM pregnancies with APO show a similar range of structural and functional abnormalities to those described in stillbirth and FGR [Unpublished work, Chapter 2]. Many aspects of placental structure and function that are altered in APO associated with RFM can be measured by ultrasound (such as placental diameter, volume and tissue vascularity) or by the concentration of placentally-derived hormones (such as human placental lactogen (hPL), placental growth factor (PlGF), soluble fms-like tyrosine kinase-1 (sFlt-1) and human chorionic gonadotrophin (hCG)) in the maternal circulation [Unpublished work, Chapters 3-5]. We aimed to test the hypothesis that antenatal assessment of placental structure and function would improve the prediction of pregnancies at highest risk of APO in a cohort of women presenting with RFM. Page 159 of 253 6.3 Materials and methods Participant recruitment The study received ethical approval from Greater Manchester North West Research Ethics Committee (11/NW/0650) and was conducted in accordance with the Declaration of Helsinki 1975 (revised 2013). Three hundred and forty-seven women with singleton pregnancies of ≥28 weeks gestation presenting with any subjective reduction in perceived fetal activity (Whitworth et al., 2011) between January 2012 and May 2014 were prospectively asked to participate in the study. Written informed consent was obtained from 300 women (86%). Exclusion criteria were: suspected immediate fetal compromise on cardiotocograph (CTG), fetal abnormality or pre-existing hypertension or diabetes. Figure 20 demonstrates the flow of patients through the study; their characteristics are summarised in Table 15. FIGURE 20: Flow of participants through the FEMINA2 study. NPO = normal pregnancy outcome. APO = adverse pregnancy outcome, diagnosed on the basis of one or more classifier of APO (shaded boxes). IBC = individualised birth weight centile. UA = umbilical artery. BE = base excess. NICU = neonatal intensive care unit. Page 160 of 253 TABLE 15: Comparison of women participating and declining to participate in the FEMINA2 trial and their pregnancy outcomes. Participants N=300 Maternal Characteristics Age (yea rs ) Ethnicity: Caucasian Asian Black Other BMI (kg/m 2) Parity (Number) RFM Characteristics Episode duration Non-Participants N=47 p 0.44 29 30 (25 – 32) (26 – 33) 204/300 (68.0%) 47/300 (15.7%) 33/300 (11.0%) 16/300 (5.3%) 25.3 0 17/47 (36.2%) 11/47 (23.4%) 12/47 (25.5%) 7/47 (14.9%) 25.9 (21.8 – 28.4) 0 (0 – 1) (0 – 2) 0.0002 (22.7 – 29.6) 0.94 0.039 48 38 (24 – 96) (21 – 84) 72/265 (27.1%) 37+0 6/39 (15.4%) 38+0 (33+3 – 39+2) (34+1 – 39+6) 117/272 (43.0%) 17/39 (43.6%) 0.94 40+1 40+3 0.17 (38+6 – 41+1) (39+5 – 41+2) Induction of labour Laboured 132/296 (44.6%) 258/296 (87.2%) 17/46 (37.0%) 42/46 (91.3%) 0.33 0.43 Caesarean Male infant Outcome Characteristics* Live birth IBC 65/296 (22.0%) 158/296 (53.3%) 9/46 (19.6%) 22/46 (47.8%) 0.71 0.48 295/296 (99.7%) 40 (18 – 68) 43/296 (14.5%) 4/296 (1.4%) 3/162 (1.9%) 3/156 (1.9%) 46/46 (100%) 37 (18 – 63) 5/45 (11.1%) 1/46 (2.2%) 1/27 (3.7%) 2/25 (8.0%) 0.28 0.83 8/296 (2.7%) 1/46 (2.2%) 0.77 (hours) Absent movements Gestation at presentation (weeks +days) Resolved Delivery Characteristics* Gestation (weeks +days) IBC < 10 5 min APGAR score < 7 Umbilical Arterial pH <7.10 Umbilical Artery Base Excess <-10 NICU Admission 0.32 0.12 0.20 0.70 0.82 0.92 0.29 Continuous data are expressed as median (interquartile range) and compared by Mann -Whitney U Test. Categorical data are expressed as number (%) and compared by Chi Squared test (with Yates’ correction as required). Statistical significance was set at the level of p<0.05. IBC = individualised birth weight centile (Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham , UK). NICU = neonatal intensive care unit. * two participants and one non-participant were lost to follow up, a further two participants were excluded from analysis of pregnancy outcome following postnatal diagnosis of fetal abnormality. Where clinical data is missing the denominator is accordingly reduced Page 161 of 253 Placental assessment in utero Fetal wellbeing was assessed in accordance with guidelines from the Royal College of Obstetricians and Gynaecologists (Whitworth et al., 2011) by a single sonographer of four years’ experience (LH): estimated fetal weight centile (EFWc) (Bulk centile calculator v6.7 (UK), Gestation Network, Birmingham, UK), four-quadrant Amniotic Fluid Index (AFI) (Magann et al., 2000) and quantification of vascular impedance at the middle third of the umbilical artery (UAD-F) by Pulsatility Index (PI) and Resistance Index (RI) (Acharya et al., 2005b) were recorded. These results were revealed to the clinical team responsible for the care of the study participant. Further ultrasound measurements were made to assess placental structure and function. Curvilinear placental length, width, maximal depth, VOCAL 30° volume (PV; 4Dview software v.5 (GE Healthcare)) and relative fetal and placental size in the fetoplacental ratio (FPR = EFW / PV) were calculated [Unpublished work, Chapter 3] along with vascular impedance at the umbilical artery abdominal (UAD-A) and placental (UAD-P) insertion points (Sonesson et al., 1993; Acharya et al., 2005a), chorionic plate arteries (CPAD) and intraplacental arteries (IPAD) [Unpublished work, Chapter 4]. Mean uterine artery PI, RI and notch status were ascertained transabdominally from left and right uterine artery Doppler (UtAD) waveforms (Gomez et al., 2008). Middle cerebral artery (MCA) vascular impedance and flow velocity were quantified by PI, RI and peak systolic velocity (PSV) at the proximal third with angle correction (Kurmanavicius et al., 1997; Bahlmann et al., 2002; Palacio et al., 2004; Ebbing et al., 2007; Morales-Rosello et al., 2014). Pre-delivery maternal blood was collected into serum separator tubes (BD Vacutainer, Franklin Lakes, US) and processed per manufacturer recommendations to obtain serum. Serum was stored at 80°C. Concentrations of human chorionic gonadotrophin (hCG), human placental lactogen (hPL), progesterone, placental growth factor (PlGF) and soluble fms-Like tyrosine kinase-1 (sFlt-1) were measured in maternal serum using ELISA kits in accordance with the manufacturer’s instructions Page 162 of 253 (Supplementary Table 2) as previously described [Unpublished work Chapter 5]. Outcome definition Pregnancy outcome was defined as adverse (APO) if any of the following events occurred: stillbirth or neonatal death, individualised birth weight centile (IBC) <10 (Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK), five minute Apgar score <7, umbilical artery pH <7.1 or base excess <-10 or admission to neonatal intensive care (NICU) within 24 hours of birth in accordance with other obstetric studies (Dutton et al., 2012; Heazell et al., 2013). Where delivery did not occur at the study site, outcome data was collected from the participant, their General Practitioner or delivering hospital. Statistical analysis Statistical analysis was performed using Stata 13 (StataCorp, College Station, USA). Characteristics and outcome data of participants and non-participants were compared by univariate analysis (Student’s t test for parametric data, Mann-Whitney U test for non-parametric data and Chi squared with Yates’ correction as required for categorical data). Where data were missing the denominator was reduced accordingly. Sonographic accuracy of fetal weight estimates was assessed by Spearman Rank correlation of EFWc and IBC. Scan to delivery centile difference was calculated (IBC – EFWc) and was compared between pregnancies with NPO and APO by Mann-Whitney U test. Participant demographics, past medical and obstetric histories, features of the RFM episode, sonographic and endocrine variables were analysed singularly or in combination (N=187); variable reduction was performed on the basis of univariate analysis with an a priori threshold of p<0.10 (Supplementary Table 5). Unadjusted odds ratios (OR) for APO for the remaining variables (log transformed where non-parametric) were calculated and adjusted (aOR) for “baseline” predictors; EFWc (as IBC is Page 163 of 253 included within the classification of APO and EFWc is a function of IBC) and gestation at recruitment (as many placental biomarkers are correlated to gestational age [Unpublished work, Chapter 3-5]). Those variables with statistically significant adjusted odds ratio (aOR) for APO (p<0.05) were taken forward into multiple logistic regression with backwards elimination to identify combinations of variables (predictive models) demonstrating superior receiver operating characteristic (ROC) curve AUC than “baseline” (EFWc and gestation at recruitment). Three models containing the fewest covariates (Supplemental Figure 1) were compared against each other and “baseline” by test characteristics (sensitivity, positive predictive value (PPV), negative predictive value (NPV), positive and negative likelihood ratios (LR+/-) and post test probabilities) at a fixed specificity of 99% aiming to achieve a positive likelihood ratio (LR+) >10 and negative likelihood ratio (LR-) <0.2. The study is reported in accordance with Standards for Reporting of Diagnostic accuracy studies (STARD) guidelines (Supplementary Table 6). Based on an expected APO rate of 20%, with 5% loss to follow up rate (Dutton et al., 2012), a cohort of 300 participants was anticipated to provide sufficient power to adjust individual APO risk by up to six potentially predictive risk factors/measurements. Page 164 of 253 6.4 Results Two (0.67%) participants were lost to follow up; a further two participants were excluded following postnatal diagnosis of significant fetal abnormality. This left 296 pregnancies for analysis, in whom 56 (18.8%) cases of APO occurred. Figure 20 presents the frequencies of each category of APO in the study cohort; categories were not mutually exclusive. Compared with nonparticipants, participants were more likely to be Caucasian (p=0.0002) and of lower parity (p=0.039) (Table 15). Despite EFWc and IBC being highly correlated (Rs=0.59, p<0.0001), across the entire study cohort the median centile loss from scan to delivery was 13.1 with median rate of decline being 1 centile per day; although APO pregnancies demonstrated greater overall centile decline between scan and birth (NPO 11.9 (-1.4–28.3) vs. APO 20.3 (5.5-40.6), p=0.008) the decline rate was unchanged (p=0.12). From a total of 187 variables, 49 variables were identified with univariate association with APO of p<0.10 (Supplementary table 2); contrary to that predicted by published literature and guidelines (Tuffnell et al., 1991; O'Sullivan et al., 2009; Whitworth, 2011; Dutton et al., 2012) the variables “previous RFM episodes” (p = 0.85), “RFM resolved” (p=0.81), anterior placental site (p=0.16), “clinical suspicion of SGA” (p=0.13), diastolic blood pressure (p=0.13), all features relating to the CTG (p≥0.13) and amniotic fluid volume (p≥0.12) all failed to reach the pre-specified cut-off value. ORs and aORs for APO are presented for each variable in Table 16; 15 were considered strongly independently associated with APO (aOR p<0.05). In contrast to the associations described in the previously literature (Dutton et al., 2012; Heazell et al., 2013; Pagani et al., 2014a; Pagani et al., 2014b) “significant past medical history” (p=0.07), “previous SGA birth” (p=0.34), all UtAD indices (p>0.07) and maternal serum hPL concentration (p=0.90) failed to demonstrate predictive value after adjustment for EFWc and gestational age. After multiple logistic regression with backwards elimination (Supplementary Figure 1) the three predictive models containing the fewest Page 165 of 253 TABLE 16: Odds ratios for adverse pregnancy outcome by individual differentially distributed variables. Variable Maternal Characteristics Obese Significant Past Medical History Previous SGA birth RFM Characteristics Post-mature presentation Duration (hours) Duration >24 hours Absent movements >12 hours Fetal Wellbeing Assessment EFWc EFWc <10 EFWc <25 Placental Size Assessment Length Placental Efficiency Quotient Placental Vascular Assessment UAD-A PI UAD-A PI>75th centile (Acharya et al., 2005a) UAD-A PI>95th centile (Acharya et al., 2005a) UAD-A RI UAD-A RI>75th centile (Sonesson et al., 1993) UAD-A RI>95th centile (Sonesson et al., 1993) UAD-F PI Log? ✓ ✓ ✓ ✓ OR 95% CI p aOR 95% CI p 1.8 2.7 1.2 (1.0 -3.4) (1.1 – 6.4) (1.1 – 6.4) 0.07 0.03 0.03 1.5 2.4 1.6 (0.8 – 3.0) (0.9 – 6.0) (0.6 – 3.9) 0.23 0.07 0.34 0.4 1.9 2.0 4.1 (0.2 – 1.1) (1.0 – 3.6) (1.0 – 3.8) (1.4 – 11.9) 0.09 0.06 0.04 0.009 0.6 1.4 1.5 4.5 (0.2 – 1.8) (0.7 – 2.9) (0.8 – 3.1( (1.4 – 14.0) 0.35 0.35 0.22 0.01 0.1 7.0 6.6 (0.1 – 0.3) (2.8 – 17.6) (3.2 – 13.0) <0.0001 <0.0001 <0.0001 N/A 0.6 1.8 N/A (0.1 – 2.7) (0.5 – 6.9) N/A 0.50 0.40 0.9 0.3 (0.8 – 1.0) (0.0 – 1.6) 0.09 0.15 1.0 0.6 (0.9 – 1.1) (0.1 – 4.3) 0.43 0.65 8.5 3.6 (1.8 – 40.3) (1.6 – 8.0) 0.007 0.002 5.1 3.1 (0.9 – 28.5) (1.3 – 7.5) 0.06 0.01 (1.1 – 7.8) (32.4 – 46x104) (1.3 – 8.5) (0.7 – 95.0) (2.3 – 40.6) 0.03 0.001 0.01 0.09 0.002 2.0 934.5 2.1 10.5 9.8 (0.7 – 6.2) (4.4 – 20x104) (0.8 – 5.9) (0.8 – 139.0) (2.3 – 40.6) 0.21 0.01 0.15 0.07 0.002 2.9 3883.5 3.4 8.3 9.8 Page 166 of 253 UAD-F PI>95th centile (Acharya et 2005b) UAD-F RI UAD-F RI>95th centile (Acharya et 2005b) Abnormal UAD-F UAD-P obtained UAD-P PI>90th centile (Acharya et 2005a) UAD-P PI>95th centile (Acharya et 2005a) CPAD RI CPAD:UAD-A PI CPAD:UAD-F PI CPAD:UAD-F RI UAD-P:UAD-F PI UAD-P:UAD-F RI UtAD PI UtAD RI UtAD RI>95th centile (Gomez et al., 2008) UtAD score Bilateral high resistance UtAD Placental Endocrine Assessment Normalised hCG hPL PlGF Free PlGF al., 3.6 (1.6 – 8.0) 0.002 2.8 al., 622.9 14.3 (13.8 – 28x103) (2.8 – 72.8) 0.001 0.001 237.2 5.8 al., 3.4 0.4 2.8 (1.5 – 7.5) (0.1 – 1.1) (1.1 – 7.2) 0.002 0.06 0.03 al., 5.1 (1.4 – 18.5) 63.9 25.2 69.0 9.1 2.9 6.1 4.4 16.4 1.8 1.2 2.4 4.3 0.6 0.4 0.7 ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ (1.1 – 6.8) 0.02 (1.8 – 31x103 ) (1.0 – 33.9) 0.03 0.049 0.2 0.5 3.1 (1.0 – 5.9) (0.1 – 1.7) (1.1 – 8.9) 0.046 0.25 0.04 0.01 5.7 (1.3 – 24.7) 0.02 (0.5 – 7677.9) (0.6 – 1117.2) (1.9 – 2467.6) (1.1 – 74.2) (0.8 – 10.1) (0.8 – 48.5) (0.7 – 26.6) (0.5 – 535.0) (1.0 – 3.3) (1.0 – 1.5) (1.1 – 5.3) 0.09 0.10 0.02 0.04 0.10 0.09 0.10 0.11 0.07 0.04 0.03 1.9 30.3 36.8 6.6 5.3 5.33 2.1 3.9 1.5 1.2 2.2 (0.2 – 4045.7) (0.4 – 2163.2) (0.8 – 1747.4) (0.7 – 65.2) (0.5 – 51.9) (0.5 – 51.9) (0.3 – 16.5) (0.1 – 181.5) (0.7 – 2.9) (0.9 – 1.5) (0.9 – 6.0) 0.46 0.12 0.07 0.11 0.15 0.15 0.47 0.49 0.28 0.16 0.07 (1.3 – 14.1) (0.1 – 2.1) (0.2 – 1.0) (0.5 – 1.0) 0.02 0.40 0.06 0.08 2.3 0.9 0.2 0.5 (0.6 – 8.2) (0.2 – 5.2) (0.1 – 0.8) (0.3 – 0.8) 0.21 0.90 0.01 0.006 Page 167 of 253 Brachiocephalic Blood Diversion MCA:UAD-A PI MCA:UAD-A RI MCA:UAD-F PI MCA:UAD-F PI MoM MCA:UAD-F PI MoM <5th centile (MoralesRosello et al., 2014) MCA:UAD-F RI MCA:UAD-F RI <5th centile (Kurmanavicius et al., 1997) MCA:IPAD RI 0.4 0.0 0.5 0.2 3.7 (0.1 0 1.0) (0.0 – 0.4) (0.2 – 1.0) (0.0 – 0.8) (1.6 – 8.4) 0.05 0.007 0.05 0.03 0.002 0.4 0.1 0.6 0.4 2.6 (0.1 – 1.2) (0.0 – 0.8) (0.3 – 1.2) (0.1 – 1.6) (1.0 – 6.6) 0.09 0.03 0.16 0.17 0.04 3.0 3.0 (1.1 – 8.4) (1.1 – 8.4) 0.04 0.04 0.1 2.0 (0.0 – 0.7) (0.6 – 6.6) 0.03 0.24 0.2 (0.0 – 1.2) 0.07 0.3 (0.0 – 2.7) 0.30 Unadjusted (OR) and adjusted (aOR) odds ratios are presented with 95% confidence intervals (95% CI) for each variable (✓ indicates that the variable was log transformed). Adjustment was performed for gestation at recruitment and estimated fetal weight centile (EFWc). Variables in bold text indicate independently predictive variables (aOR p<0.05). For derivation of variables please see Supplementary Table 4. SGA = Small for gestational age (individualised birth weight ce ntile <10 (Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK)). UAD = Umbilical artery Doppler (-A = abdominal insertion, -F = free loop, -P = placental insertion). CPAD = chorionic plate artery Doppler, IPAD = intraplacental artery Doppler. UtAD = Mean uterine artery Doppler. MCA = Middle cerebral artery Doppler. MoM = mult iples of the median. Page 168 of 253 covariates were identified (Table 17); removal of further variables from Model 1 (F-PlGF, UAD-F RI and UAD-P PI>90 plus baseline care) resulted in loss of significance compared with the “baseline” model (Supplementary Figure 2). Participant non-consent to donation of blood (N=36) and failure to obtain UAD-P (N=73) or UAD-A (N=155) waveforms reduced the number of participants studied in models 1, 2 and 3 to 194, 130 and 130 respectively. ROC curve AUC’s of Models 1 – 3 were not significantly different from each other (p>0.05) but were higher than that of the “baseline” model (p<0.05) (Figure 21). It is notable that UAD-F RI remained in all the predictive models whilst UAD-F PI was excluded from Models 1 and 2 suggesting both independent predictive value of UAD-F in general, and superiority of RI over PI in contrast to published guidance (Whitworth et al., 2011). Test characteristics of each predictive model are presented at a fixed specificity of 99% (Table 18); Model 3 demonstrated highest sensitivity, PPV, NPV, LR+ and lowest LR-. Whilst the LR+ for each proposed model was >10, no model demonstrated an LR-<0.2, resulting in a significant (>10%), 0.50 0.25 0.00 Sensitivity 0.75 1.00 residual post-negative test probability of APO. 0.00 0.25 0.50 1-Specificity 0.75 1.00 Baseline: AUC 0.75 (0.64 - 0.86) Model 1: AUC 0.86 (0.79 - 0.94) Model 2: AUC 0.88 (0.81 - 0.95) Model 3: AUC 0.88 (0.80 - 0.95) Reference FIGURE 21: Receiver operator characteristic curve comparison. Demonstrating model performance for the baseline model and proposed models 1 – 3 (see text for model components). Each proposed model was superior to the baseline model (p<0.05) but equal to each other (p>0.05). AUC = area under receiver operating characteristic curve with 95% confidence intervals. Page 169 of 253 TABLE 17: Components and comparison of proposed predictive models. Model Logit(pAPO) = AUC p Baseline 14.3 – 5.1*Log(Gest) – 2.2*Log(EFWc) 0.75 - 1 19.7 – 9.4*Log(Gest) – 2.4*Log(EFWc) - 0.9*Log(F-PlGF) + 10.3*(UAD-F RI) + 1.1*(UAD-P PI>90) 0.86 0.036 2 26.0 – 14.6*Log(Gest) – 2.1*Log(EFWc) – 1.2*Log(F-PlGF) + 19.5*(UAD-F RI) + 2.2*(UAD-P PI>90) + 0.2*(UAD-A PI>75) 0.88 0.0038 3 37.6 – 18.3*Log(Gest) – 2.2*Log(EFWc) – 1.1*Log(F-PlGF) + 14.9*(UAD-F RI) + 2.2*(UAD-P PI>90) + 0.3*(UAD-A PI>75) + 1.3*(UAD-F PI>95) 0.88 0.0061 Compared with the baseline model, each proposed model displays significantly higher ROC AUC (p<0.05). Key: pAPO = probability of adverse pregnancy outcome. Logit (pAPO) = ln(pAPO/1-pAPO). Gest = gestation (days) at recruitment. EFW c = estimated fetal weight centile. F-PlGF = PlGF*(PlGF/sFlt-1). UAD = umbilical artery Doppler (-A = abdominal insertion, -F = free loop, -P = placental insertion). PI = pulsatility index. RI = resistance index. UAD-P PI>90 = >90 th centile (Acharya et al., 2005a). UAD-A PI>75 = >75 th centile (Acharya et al., 2005a). UAD-F PI>95 = >95th centile (Acharya et al., 2005b). Page 170 of 253 TABLE 18: Test performance characteristics of proposed predictive models. Model N Threshold Sensitivity PPV NPV p(APO) (%) (%) (%) LR+ LR- Negative post-test probability (%) Baseline 1 2 3 296 194 130 130 0.66 0.65 0.68 0.62 8.9 71.4 82.3 10.7 0.91 (3.9 – 19.3) (35.9 – 91.8) (77.5 – 86.3) (2.1 – 53.8) (0.85 – 1.00) 12.5 80.0 85.2 20.3 0.87 (5.0 – 28.1) (37.6 – 96.4) (79.4 – 89.6) (2.3 – 175.3) (0.77 – 1.00) 25 85.7 85.4 26.5 0.75 (12.5 – 44.9) (48.7 – 97.4) (78.1 – 90.5) (3.3 – 210.0) (0.60 – 0.96) 37.5 90.0 87.5 39.8 0.62 (21.2 – 57.3) (59.6 – 98.2) (80.4 – 92.3) (5.3 – 299.1) (0.46 – 0.86) 17.2 16.5 14.1 11.8 Test characteristics are presented at a fixed specificity of 99% and are displayed with 95% confidence intervals. p(APO) = probability of adverse pregnancy outcome. PPV/NPV = positive and negative predictive values. LR+/LR- = positive and negative likelihood ratios. Negative post-test probability = probability of APO following a negative test by each predictive model Page 171 of 253 6.5 Discussion This study’s findings support the hypothesis that antenatal assessment of placental size (PlGF concentration incorporated within F-PlGF [Unpublished work, Chapter 5]) and function (sFlt-1 concentration within F-PlGF and the UAD-F RI [Unpublished work, Chapter 5]) improves prediction of RFM pregnancies at highest risk of APO. We previously demonstrated that maternal circulating total PlGF concentration reflects macro- and microscopic placental structure and that circulating sFlt-1 concentration reflects placental endocrine function [Unpublished work, Chapter 5], thus their combination in the F -PlGF variable (F-PlGF = PlGF x (PlGF/sFlt-1)) assesses both placental size and function (particularly the hypoxia-responsive balance of angiogenic and anti-angiogenic factors). Here we conclude that the balance of PlGF and sFlt1 is potentially predictive of APO following RFM, supporting the findings of Griffin et al. [In press, (2015)] who demonstrated a small increase in prediction of SGA births when PlGF (measured by the bedside Triage® test (Alere Inc., Waltham, USA)) was combined with EFWc. We have previously shown that the UAD-F RI correlates with CPA thromboxane sensitivity [Unpublished work, Chapter 4], in keeping with the potential predictive value of the UAD-F for imminent delivery in RFM pregnancies reported by Korszun et al. (2002). Previous studies have reported enhanced prediction of APO amongst highrisk pregnancies using limited structural, vascular and endocrine placental assessment in the first (Pagani et al., 2014b) and second trimesters (Toal et al., 2008a; Pagani et al., 2014a). We demonstrate that such assessment later in pregnancy (when intervention, e.g. delivery, is possible) is also potentially useful. We were unable to replicate the potential predictive value of gestational age, hPL, CTG findings, diastolic blood pressure or maternal serum hPL concentration suggested in a previous study of pregnancy outcome following RFM by Dutton et al. (2012). This may be due to the Page 172 of 253 exclusion of prematurity from the APO definition in the current study. Furthermore we failed to corroborate, in the third trimester after adjustment for gestational age and EFWc, the predictive value of UtAD in RFM pregnancies reported by Pagani et al. (2014a; 2014b) in the first and second trimesters. Furthermore, we believe that the results of this study support more liberal use of ultrasound in the RFM population in general (as first and “resolved” episodes of RFM were not associated with improved outcome) and delayed re-assessment of fetal growth (after a period of two to three weeks, the recommended inter-scan interval for assessment of fetal growth (Robson et al., 2014)) in ongoing RFM pregnancies in view of significant residual APO probability and centile decline between scan and delivery, suggestive of subclinical placental insufficiency even in pregnancies with apparently “normal” outcome. A number of study limitations are recognised. Firstly the necessity of a composite definition of APO due to the number of participants required to directly examine stillbirth as a primary outcome (Smith, 2012). Secondly, models with more than five predictors (Models 2 and 3) exceed the event per variable rate of 10 and are likely to be over fitted. However, there was no significant difference in ROC AUC between the three models, which all need prospective validation. We have previously shown suboptimal interobserver reliability (intraclass correlation coefficient <0.75 (Khan and Chien, 2001)) of many of the placental variables measured in this study [Unpublished work, Chapters 3&4], which may have affected their predictive performance, resulting in exclusion of potentially useful factors from the predictive models. These techniques require refinement before being implemented on a wider scale. Finally, the potential impact of bias (e.g. referral/recruitment bias in the discrepancy between expected rate of RFM presentation and actual recruitment rate, and investigator bias introduced by investigator knowledge of EFWc or other features of the Page 173 of 253 presentation) is unknown due to lack of data regarding all RFM episodes at the study hospital in the same time period. Notwithstanding these potential limitations, the principal strength of this study is the multi-domain prospective assessment of in utero placental structure and function in the context of a common complaint of pregnancy, within a diverse population. We therefore believe these findings are widely applicable in the context of RFM, and potentially in other third trimester placental-origin pregnancy complications, although this requires confirmatory studies. Furthermore, the facility and expertise to measure these aspects of placental structure and function exist in high income countries’ health services worldwide making the model widely implementable. Prospective studies of these proposed models (with prescribed management protocols for those with “high” and “not high” risk scores) are now required to validate these findings and to assess the health economic implications of placental assessment for RFM. Page 174 of 253 6.6 Acknowledgements The authors would like to thank the women who participated in this study and the midwives at St Mary’s Hospital, Manchester for their assistance in participant recruitment. Dr L Higgins is supported by an Action Medical Research Training Fellowship and a Manchester NIHR Biomedical Research Centre fellowship. The study was also supported by Tommy’s - the Baby Charity. 6.7 Statement of author contributions The project was conceived by AEPH, EDJ, CPS and LH, and methodologies planned by LH. LH performed all sonographic assessments, laboratory and statistical analyses with EDJ providing expert sonographic supervision, and JEM providing statistical advice. All authors were involved in the preparation of the manuscript. Page 175 of 253 6.8 Supplementary data SUPPLEMENTARY TABLE 4: Variable reduction. Variable Maternal characteristics Age (years) t Height (cm) t Weight (kg) m BMI (kg/m2 ) m Obese c Caucasian c Black c Asian c Other ethnicity c Parity (number) m Nulliparity c Miscarriages (number) m Recurrent Miscarriage c Previous Caesarean Section c Number of previous Caesarean Sections Derivation Weight/ (Height)2 BMI ≥30 Parity = 0 ≥3 miscarriages - NPO APO p 29 ± 6 164 ± 7 69 (60 – 80) 25 (23 – 29) 53/240 (22%) 155/240 (65%) 33/240 (14%) 39/240 (16%) 13/240 (5%) 0 (0 – 1) 144/240 (60%) 0 (0 – 1) 7/240 (3%) 17/240 (7%) 0 (0 – 0) 30 ± 6 163 ± 6 74 (58-84) 27 (23 – 31) 19/56 (34%) 38/56 (68%) 4/56 (7%) 11/56 (20%) 3/56 (5%) 0 (0 – 1) 35/56 (63%) 0 (0 – 1) 0/56 (0%) 1/56 (2%) 0 (0 – 0) 0.52 0.27 0.33 0.20 0.063 0.48 0.18 0.68 0.86 0.69 0.73 0.35 0.20 0.32 0.17 35/240 (15%) 19/240 (8%) 2/240 (1%) 73 (64 – 84) 26/240 (11%) 16/240 (7%) 11/56 (20%) 10/56 (18%) 1/56 (2%) 74 (64 – 84) 4/56 (7%) 9/56 (16%) 0.35 0.024 0.52 0.87 0.41 0.023 15/240 (6%) 1/56 (2%) 0.18 m Significant obstetric history c Previous SGA c Previous stillbirth c Booking Gestation (days) m Late booker c Significant Past Medical History c Mental Health Disorder c Previous child IBC <10 Booking gestation >14 weeks Any medical condition (or treatment thereof) associated with APO Any mental health condition Page 176 of 253 Current alcohol intake c Current smoker c Current illegal drug use c Features of RFM episode Gestation (days) m Term gestation c Post-mature gestation c Episode number m Previous RFM episode(s)? c Duration (hours) m Duration >6 hours c Duration > 12 hours c Duration >24 hours c Duration >48 hours c Absent movements? c Absent (hours) m Absent >6 hours c Absent >12 hours c Absent >24 hours c Systolic blood pressure (mm/Hg) m Diastolic blood pressure (mm/Hg) m Urinalysis abnormal c Proteinuria c CTG normal c Baseline (bpm) t Variability <5bpm c Acceleration c Decelerations c Atypical/late decelerations c Gestation ≥37 weeks Gestation ≥40 weeks (National Institute for Health and Care Excellence, 2007) Page 177 of 253 3/240 (1%) 30/240 (13%) 0/240 (0%) 1/56 (2%) 8/56 (14%) 0/56 (0%) 0.72 0.41 1.00 261 (236 – 275) 126/240 (53%) 45/240 (19%) 1 (1 – 1) 51/237 (22%) 36 (18 – 72) 150/229 (66%) 138/229 (60%) 95/229 (41%) 50/229 (22%) 59/214 (28%) 0 (0 – 3) 34/214 (16%) 8/214 (4%) 5/214 (2%) 111 (103 – 120) 69 (60 -75) 40/152 (26%) 20/152 (13%) 186/208 (89%) 137 ± 8 0/208 (0%) 196/208 (94%) 13/208 (6%) 5/208 (7%) 257 (221 – 272) 27/56 (48%) 5/56 (9%) 1 (1-1) 11/54 (20%) 48 (24 – 72) 42/55 (76%) 39/55 (71%) 32/55 (58%) 14/55 (25%) 13/51 (25%) 0 (0 – 1) 11/51 (22%) 7/51 (14%) 2/51 (4%) 110 (102 – 122) 70 (62 – 80) 9/35 (26%) 6/35 (17%) 40/49 (82%) 136 ± 8 0/49 (0%) 44/49 (90%) 6/49 (12%) 1/49 (2%) 0.20 0.56 0.077 0.77 0.85 0.072 0.12 0.14 0.025 0.56 0.76 0.90 0.33 0.006 0.88 0.93 0.13 0.94 0.73 0.13 0.64 1.00 0.26 0.15 0.19 Movement frequency (number/min) m RFM resolved c Fetal wellbeing assessment Suspected SGA c Estimated fetal weight centile (EFWc) m EFWc <10 c EFWc <25 c Amniotic Fluid Index (AFI) cm m AFI <5 c AFI >95 c Anterior placenta c Lateral placenta c Posterior placenta c Fundal placenta c Placental structure Length (cm) t Width (cm) t Width:Length m Area (cm2) t Depth (cm) m Depth CoV (%) m Length:Depth m Width:Depth m Placental efficiency quotient (cm) m Volume (cm3) m FPR (g/cm3) m Maternally recorded movements during CTG recording / duration of CTG recording - 0.29 (0.12 – 0.69) 0.30 (0.11 – 0.57) 0.40 95/221 (43%) 21/51 (41%) 0.81 SFH or previous ultrasound AC / EFW <10th customised centile Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK) EFWc <10th centile EFWc <25th centile (Magann et al., 2000) AFI <5th centile (Magann et al., 2000) AFI >95th centile (Magann et al., 2000) Predominant placental location at time of ultrasound 16/240 (7%) 66 (44 – 86) 7/55 (13%) 40 (15 – 62) 0.13 <0.0001 9/240 (4%) 20/240 (8%) 11 (9 – 15) 1/240 (0%) 18/240 (8%) 83/240 (35%) 71/240 (30%) 48/240 (20%) 38/240 (16%) 12/56 (21%) 21/56 (38%) 11 (8 – 13) 1/56 (2%) 1/56 (2%) 25/56 (45%) 11/56 (20%) 11/56 (20%) 9/56 (16%) <0.0001 <0.0001 0.17 0.26 0.12 0.16 0.13 0.95 0.97 18.7 ± 2.7 15.8 ± 2.7 18.0 ± 3.0 15.1 ± 3.0 0.084 0.13 0.86 (0.77 – 0.93) 242 ± 68 4.3 (3.7 – 4.9) 0.88 (0.80 – 0.94) 227 ± 80 4.2 (3.6 – 5.1) 0.34 0.15 0.97 17.9 (12.3 – 23.2) 4.4 (3.6 – 5.1) 3.7 (3.0 – 4.4) 69 (53 – 87) 478 (364 – 602) 6.0 (4.4 – 8.7) 16.8 (13.3 – 23.3) 4.2 (3.4 – 5.1) 3.3 (3.0 – 4.3) 62 (51 – 76) 441 (321 – 581) 6.0 (4.7 – 7.1) 0.99 0.30 0.18 0.092 0.13 0.53 Longest placental diameter [Unpublished work, Chapter 3] Longest placental diameter perpendicular to length [Unpublished work, Chapter 3] Placental width / length π x ½Length x ½Width Maximal placental thickness perpendicular to length [Unpublished work, Chapter 3] Standard deviation / average of 5 depth measurements x 100 Placental length / depth Placental width / depth Placental length x width / depth VOCAL 30° [Unpublished work, Chapter 3] EFW / VOCAL 30° Page 178 of 253 Qualitative cord eccentricity c Cord ratio t Quantitative cord eccentricity c Placental vascular assessment UAD-A obtained c UAD-A PI t UAD-A PI>75 c UAD-A PI>90 c UAD-A PI>95 c UAD-A RI t UAD-A RI>75 c UAD-A RI>95 c UAD-F obtained c UAD-F PI t UAD-F PI>95 c UAD-F RI t UAD-F RI>95 c Abnormal UAD-F c UAD-P obtained c UAD-P PI t UAD-P PI>75 c UAD-P PI>90 c UAD-P PI>95 c UAD-P RI t Cord insertion point qualitatively closer to placental edge than centre of placental body Shortest distance from cord insertion point to placental edge / placental length Cord ratio<0.25 52/197 (26%) 10/44 (23%) 0.62 0.33 ± 0.10) 0.31 ± 0.09 0.49 40/170 (24%) 9/40 (23%) 0.89 Doppler waveform obtained from abdominal insertion of umbilical artery [Unpublished work, Chapter 4] UAD-A PI > 75th centile (Acharya et al., 2005a) UAD-A PI >90th centile (Acharya et al., 2005a) UAD-A PI >95th centile (Acharya et al., 2005a) UAD-A RI > 75th centile (Sonesson et al., 1993) UAD-A RI >95th centile (Sonesson et al., 1993) Doppler waveform obtained from middle third of umbilical artery [Unpublished work, Chapter 4] UAD-F PI >95th centile (Acharya et al., 2005b) UAD-F RI >95th centile (Acharya et al., 2005b) UAD-F PI or RI >95th centile (Acharya et al., 2005b) or absent/reversed end diastolic flow Doppler waveform obtained from placental insertion of umbilical artery [Unpublished work, Chapter 4] UAD-P PI > 75th centile (Acharya et al., 2005a) UAD-P PI >90th centile (Acharya et al., 2005a) UAD-P PI >95th centile (Acharya et al., 2005a) - 126/189 (67%) 32/45 (71%) 0.57 0.98 ± 0.24 44/126 (35%) 20/126 (16%) 13/126 (10%) 0.62 ± 0.09 15/126 (12%) 1/126 (1%) 240/240 (100%) 1.12 ± 0.25 20/32 (63%) 9/32 (28%) 7/32 (22%) 0.68 ± 0.10) 10/32 (31%) 2/32 (6%) 56/56 (100%) 0.005 0.005 0.11 0.079 0.0002 0.007 0.043 1.00 0.90 ± 0.20 17/240 (7%) 0.59 ± 0.08 2/240 (1%) 18/240 (8%) 1.00 ± 0.22 12/56 (21%) 0.63 ± 0.09 6/56 (11%) 12/56 (21%) 0.001 0.001 0.0007 <0.0001 0.002 183/193 (95%) 40/46 (87%) 0.055 0.82 ± 0.22 34/183 (19%) 15/183 (8%) 5/183 (3%) 0.55 ± 0.08 0.86 ± 0.22 10/40 (25%) 8/40 (20%) 5/40 (13%) 0.56 ± 0.09 0.34 0.36 0.026 0.007 0.40 Page 179 of 253 UAD-P RI>75 c UAD-P RI>95 c CPAD obtained c CPAD PI m CPAD PI CoV (%) m CPAD RI t CPAD RI CoV (%) m IPAD obtained c IPAD PI t IPAD PI CoV (%) m IPAD RI t IPAD RI CoV (%) m UAD-F:UAD-A PI m UAD-F:UAD-A RI m UAD-P:UAD-A PI m UAD-P:UAD-A RI t CPAD:UAD-A PI m CPAD:UAD-A RI t IPAD:UAD-A PI m IPAD:UAD-A RI t UAD-P:UAD-F PI t UAD-P:UAD-F RI t CPAD:UAD-F PI m CPAD:UAD-F RI t IPAD:UAD-F PI m IPAD:UAD-F RI m CPAD:UAD-P PI t CPAD:UAD-P RI m UAD-P RI > 75th centile (Sonesson et al., 1993) UAD-P RI >95th centile (Sonesson et al., 1993) Doppler waveform obtained from placental surface arteries [Unpublished work, Chapter 4] Standard deviation / average of 4 CPAD PI measurements x 100 Standard deviation / average of 4 CPAD RI measurements x 100 Doppler waveform obtained from intraplacental arteries [Unpublished work, Chapter 4] Standard deviation / average of 4 IPAD PI measurements x 100 Standard deviation / average of 4 IPAD RI measurements x 100 UAD-F PI / UAD-A PI UAD-F RI / UAD-A RI UAD-P PI / UAD-A PI UAD-P RI / UAD-A RI CPAD PI / UAD-A PI CPAD RI / UAD-A RI IPAD PI / UAD-A PI IPAD RI / UAD-A RI UAD-P PI / UAD-F PI UAD-P RI / UAD-F RI CPAD PI / UAD-F PI CPAD RI / UAD-F RI IPAD PI / UAD-F PI IPAD RI / UAD-F RI CPAD PI / UAD-P PI CPAD RI / UAD RI Page 180 of 253 18/183 (10%) 1/183 (1%) 215/220 (98%) 6/40 (15%) 1/40 (3%) 51/52 (98%) 0.34 0.24 0.71 0.78 (0.69 – 0.89) 14.6 (10.1 – 21.0) 0.53 ± 0.064 9.69 (6.51 – 14.2) 201/209 (96%) 0.80 (0.71 – 0.92) 15.6 (11.4 – 20.7) 0.55 ± 0.064 10.8 (7.36 – 13.6) 45/48 (94%) 0.26 0.41 0.088 0.44 0.46 0.72 ± 0.13 15.1 (9.5 – 21.9) 0.50 ± 0.06 11.1 (7.4 – 16.6) 1.10 (0.94 – 1.20) 1.05 (0.97 – 1.12) 1.17 (0.99 – 1.33) 1.13 ± 0.18 1.16 (1.02 – 1.35) 1.15 ± 0.17 1.31 (1.12 – 1.55) 1.23 ± 0.19 1.13 ± 0.26 1.09 ± 0.16 1.11 (0.98 – 1.25) 1.11 ± 0.14 1.22 (1.07 – 1.40) 1.16 (1.06 – 1.28) 1.04 ± 0.23 1.02 (0.94 – 1.10) 0.76 ± 0.11 13.5 (10.8 – 21.1) 0.52 ± 0.056 10.6 (6.8 – 15.6) 1.05 (0.89 – 1.21) 1.03 (0.97 – 1.14) 1.31 (1.06 -1.43) 1.17 ± 0.17 1.32 (1.08 – 1.40) 1.21 ± 0.23 1.39 (1.23 – 1.73) 1.29 ± 0.17 1.21 ± 0.23 1.14 ± 0.18 1.21 (1.07 – 1.32) 1.15 ± 0.16 1.30 (1.11 – 1.50) 1.19 (1.11 – 1.30) 1.05 ± 0.22 1.01 (0.95 – 1.12) 0.11 0.65 0.11 0.84 0.63 0.67 0.19 0.32 0.072 0.098 0.14 0.14 0.098 0.083 0.011 0.037 0.12 0.22 0.66 0.82 IPAD:UAD-P PI m IPAD:UAD-P RI m IPAD:CPAD PI m IPAD:CPAD RI m UtAD PI m UtAD PI>95 c UtAD PI CoV (%) m UtAD RI m UtAD RI>95 c UtAD RI CoV (%) m Any high resistance UtAD c Bilateral high resistance UtAD c Any notched UtAD c Bilateral notched UtAD c Notch score m UtAD score m UtAD score >3 c UAD-A:UtAD PI m UAD-A:UtAD RI m UAD-F:UtAD PI m UAD-F:UtAD RI m UAD-P:UtAD PI m UAD-P:UtAD RI m CPAD:UtAD PI m CPAD:UtAD RI m IPAD PI / UAD-P PI IPAD RI / UAD-P RI IPAD PI / CPAD PI IPAD RI / CPAD RI Average of left and right uterine artery PIs UtAD PI >95th centile (Gomez et al., 2008) Standard deviation / average of left and right uterine artery PIs x 100 Average of left and right uterine artery RIs UtAD RI >95th centile (Gomez et al., 2008) Standard deviation / average of left and right uterine artery RIs x 100 PI or RI >95th centile in either uterine artery (Gomez et al., 2008) PI or RI >95th centile in both uterine arteries (Gomez et al., 2008) Notched uterine artery Doppler waveform in either uterine artery Notched uterine artery Doppler waveform in both uterine arteries Score 1 for each notch Score 1 for each uterine artery PI or RI >95th centile (Gomez et al., 2008), score 2 for each notched uterine artery Doppler waveform UAD-A PI / UtAD PI UAD-A RI / UtAD RI UAD-F PI / UtAD PI UAD-F RI / UtAD RI UAD-P PI / UtAD PI UAD-P RI / UtAD RI CPAD PI / UtAD PI CPAD RI / UtAD RI Page 181 of 253 1.12 (0.99 – 1.29) 1.08 (0.99 – 1.19) 1.12 (0.99 – 1.24) 1.07 (0.98 – 1.14) 0.78 (0.63 – 1.02) 67/229 (29%) 22.3 (10.6 – 36.0) 1.08 (0.94 – 1.27) 1.07 (0.99 – 1.15) 1.07 (1.00 – 1.20) 1.05 (0.99 – 1.14) 0.87 (0.72 – 1.17) 22/55 (40%) 23.0 (12.6 – 42.7) 0.62 0.57 0.61 0.96 0.054 0.12 0.51 0.51 (0.45 – 0.58) 59/229 (26%) 14.3 (6.6 – 22.9) 0.54 (0.48 – 0.63) 21/55 (38%) 14.3 (8.2 – 26.0) 0.055 0.066 0.52 102/235 (43%) 22/229 (10%) 37/235 (16%) 3/229 (1%) 31/56 (55%) 11/55 (20%) 13/56 (23%) 2/55 (4%) 0.11 0.025 0.17 0.23 0 (0 – 0) 0 (0 – 1) 0 (0 – 0) 1 (0 – 2) 0.16 0.057 5/229 (2%) 0.77 (0.58 – 1.06) 0.81 (0.65 – 0.96) 0.84 (0.68 – 1.19) 0.84 (0.73 - 1.00) 0.95 (0.74 – 1.39) 0.91 (0.77 – 1.34) 1.01 (0.76 – 1.34) 0.96 (0.80 – 1.12) 3/55 (5%) 0.76 (0.62 – 1.27) 0.77 (0.71 – 0.95) 0.80 (0.70 – 1.29) 0.84 (0.74 – 1.03) 1.00 (0.77 – 1.49) 1.09 (0.83 – 1.41) 1.09 (0.83 – 1.41) 0.97 (0.85 – 1.15) 0.22 0.55 0.70 0.85 0.98 0.43 0.38 0.28 0.46 IPAD:UtAD PI m IPAD:UtAD RI m Placental endocrine assessment hCG (mIU/ml) m Normalised hCG (mIU/ml/cm3 ) m hPL (mg/ml) m Normalised hPL (mg/ml/cm3) m Progesterone (ng/ml) m Normalised Progesterone (ng/ml/cm3) m PlGF (pg/ml) m Normalised PlGF (pg/ml/cm3) m sFlt-1 (pg/ml) m Normalised sFlt-1 (pg/ml/cm3) m PlGF:sFlt-1 m Normalised PlGF:sFlt-1 (1/cm3) m F-PlGF (pg/ml) m Normalised Free PlGF (pg/ml/cm3) m Brachiocephalic blood diversion MCA PI m MCA PI <10 c MCA PI<5 c MCA RI t MCA RI<5 c MCA PSV (cm/s) t MCA PSV >90 c MCA PSV >95 c MCA:UAD-A PI t MCA:UAD-A RI t IPAD PI / UtAD PI IPAD RI / UtAD RI Maternal serum hCG concentration hCG concentration / placental volume Maternal serum hPL concentration hPL concentration / placental volume Maternal serum progesterone concentration Progesterone concentration / placental volume Maternal serum PlGF concentration PlGF concentration / placental volume Maternal serum sFlt-1 concentration sFlt-1 concentration / placental volume PlGF concentration / sFlt-1 concentration PlGF:sFlt-1 / placental volume x 1000 PlGF x (PlGF / sFlt-1) Free PlGF / placental volume MCA PI <10th centile (Palacio et al., 2004; Ebbing et al., 2007) MCA PI <5th centile (Palacio et al., 2004; Ebbing et al., 2007) MCA RI <5th centile (Kurmanavicius et al., 1997) MCA PSV >90th centile (Bahlmann et al., 2002) MCA PSV >95th centile (Bahlmann et al., 2002) MCA PI / UAD-A PI MCA RI/ UAD-A RI Page 182 of 253 1.10 (0.84 – 1.46) 1.01 (0.85 – 1.20) 1.14 (0.95 – 1.64) 1.04 (0.92 – 1.25) 0.23 0.34 378 (282 – 505) 0.79 (0.54 – 1.23) 728 (566 – 918) 1.54 (1.13 – 2.08) 814 (544 – 1128) 1.72 (1.12 – 2.52) 343 (212 – 570) 0.72 (0.46 – 1.29) 1800 (1039 – 3492) 3.76 (1.93 – 7.99) 0.20 (0.06 – 0.58) 0.41 (0.12 – 1.1) 71 (12 – 319) 0.12 (0.024 – 0.55) 425 (301 – 577) 1.19 (0.61 – 1.58) 665 (536 – 807) 1.60 (1.23 – 1.89) 721 (476 – 990) 1.61 (1.20 – 2.50) 262 (151 – 492) 0.53 (0.32 – 1.48) 2550 (892 – 4497) 5.22 (2.59 – 10.75) 0.10 (0.04 – 0.47) 0.21 (0.10 – 1.0) 19 (7 – 157) 0.048 (0.016 – 0.28) 0.11 0.035 0.098 0.59 0.30 0.61 0.041 0.30 0.43 0.12 0.13 0.24 0.087 0.15 1.62 (1.29 – 1.95) 41/173 (24%) 13/173 (8%) 0.77 ± 0.09 17/173 (10%) 52 ± 14 29/166 (17%) 13/166 (8%) 1.68 ± 0.49 1.25 ± 0.20 1.60 (1.37 – 1.82) 13/45 (29%) 5/45 (11%) 0.77 ± 0.08 7/45 (16%) 51 ± 13 6/44 (14%) 3/44 (7%) 1.47 ± 0.36 1.13 ± 0.16 0.90 0.47 0.44 0.80 0.27 0.81 0.54 0.82 0.049 0.0055 MCA:UAD-F PI t MCA:UAD-F PI MOM t MCA:UAD-F PI MOM <5 c MCA:UAD-F RI t MCA:UAD-F RI <5 c MCA:UAD-P PI t MCA:UAD-P RI t MCA:CPAD PI m MCA:CPAD RI t MCA:IPAD PI m MCA:IPAD RI t MCA PI / UAD-F PI MCA:UAD-F PI MoM (Morales-Rosello et al., 2014) MCA/UAD-F PI MOM <5 (Morales-Rosello et al., 2014) MCA RI /UAD-F RI MCA:UAD-F RI <5th centile (Kurmanavicius et al., 1997) MCA PI / UAD-P PI MCA RI / UAD-P RI MCA PI / CPAD PI MCA RI / CPAD RI MCA PI / IPAD PI MCA RI / IPAD RI 1.83 ± 0.53 0.98 ± 0.27 17/173 (10%) 1.31 ± 0.18 10/173 (6%) 2.01 ± 0.55 1.42 ± 0.21 1.97 (1.71 – 2.34) 1.45 ± 0.18 2.21 (1.82 – 2.57) 1.53 ± 0.20 1.66 ± 0.50 0.88 ± 0.27 13/45 (29%) 1.22 ± 0.18 7/45 (16%) 1.94 ± 0.58 1.39 ± 0.23 1.90 (1.59 – 2.49) 1.40 ± 0.18 1.90 (1.76 – 2.44) 1.46 ± 0.19 0.048 0.026 0.001 0.0016 0.029 0.50 0.50 0.56 0.16 0.18 0.071 The distributions of variables between groups were compared by univariate analysis. Non -parametric data ( m ) are presented as median (IQR) and compared by Mann-Whitney U Test, parametric data ( t) as mean ± standard deviation and compared by Students’ T Test and categorical data ( c ) as number (%) and compared by Chi squared test with Yates’ correction as required. Bold text denotes variables meeting a priori variable reduction criteria of p<0.10. Key: NPO = normal pregnancy outcome. APO = adverse pregnancy outcome. SGA = small for gestational age. IBC = individualised birth weight centile (Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK)). RFM = reduced fe tal movements. CTG = cardiotocograph. Bpm = beats per minute. EFWc = estimated fetal weight centile (Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham, UK)). CoV = coefficient of variance. VOCAL 30° = Virtual organ computer-aided analysis with 30° rotation angle. FPR = fetoplacental ratio. UAD = umbilical artery Doppler; –A = abdominal insertion. –F = free loop. –P = placental insertion. CPAD = chorionic plate artery Doppler. IPAD = intraplacental artery Doppler. MCA = middle cerebral artery Doppler. UtAD = uterine artery Doppler. PI = pulsatility index. RI = resistance index. PSV = peak systolic velocity. hCG = human chorionic gonadotrophin. hPL = human placental lactogen. PlGF = placental growth factor. sFlt-1 = soluble fms-like Tyrosine Kinase-1. Page 183 of 253 SUPPLEMENTARY TABLE 5: STARD guideline checklist. Section & Item Topic Title / Abstract / Keywords 1 Identify the article as a study of diagnostic accuracy (recommend MeSH heading 'sensitivity and specificity'). Introduction 2 State the research questions or study aims, such as estimating diagnostic accuracy or comparing accuracy between tests or across participant groups. Methods Participants 3 The study population: The inclusion and exclusion criteria, setting and locations where data were collected. 4 Participant recruitment: Was recruitment based on presenting symptoms, results from previous tests, or the fact that the participants had received the index tests or the reference standard? 5 Participant sampling: Was the study population a consecutive series of participants defined by the selection criteria in item 3 and 4? If not, specify how participants were further selected. 6 Data collection: Was data collection planned before the index test and reference standard were performed (prospective study) or after (retrospective study)? Test methods 7 The reference standard and its rationale. 8 Technical specifications of material and methods involved including how and when measurements were taken, and/or cite references for index tests and reference standard. 9 Definition of and rationale for the units, cut-offs and/or categories of the results of the index tests and the reference standard. 10 The number, training and expertise of the persons executing and reading the index tests and the reference standard. 11 Whether or not the readers of the index tests and reference standard were blind (masked) to the results of the other test and describe any other clinical information available to the readers. Statistical 12 Methods for calculating or comparing measures of diagnostic accuracy, and the statistical methods used to methods quantify uncertainty (e.g. 95% confidence intervals). 13 Methods for calculating test reproducibility, if done. Results Participants 14 When study was performed, including beginning and end dates of recruitment. Page 184 of 253 Page number Key words 4 6 6 Consecutive referrals (6). Potential bias discussed (14) 6 6&8 6&7 6&7 6 Unblinded: potential bias discussed (Page 14) 8 Not performed in this study 6 15 21 22 23 24 Clinical and demographic characteristics of the study population (at least information on age, gender, spectrum Table 1 & Suppl. Table 1 of presenting symptoms). The number of participants satisfying the criteria for inclusion who did or did not undergo the index tests 6 & Figure 1 and/or the reference standard; describe why participants failed to undergo either test (a flow diagram is strongly recommended). Time-interval between the index tests and the reference standard, and any treatment administered in 6 between. Distribution of severity of disease (define criteria) in those with the target condition; other diagnoses in Table 1 & Figure 1 participants without the target condition. A cross tabulation of the results of the index tests (including indeterminate and missing results) by the results Table 4 of the reference standard; for continuous results, the distribution of the test results by the results of the reference standard. Any adverse events from performing the index tests or the reference standard. No adverse events resulting from test performance Estimates of diagnostic accuracy and measures of statistical uncertainty (e.g. 95% confidence intervals). Table 4, Figure 2 & Suppl. Fig 1 How indeterminate results, missing data and outliers of the index tests were handled. Page 7 Estimates of variability of diagnostic accuracy between subgroups of participants, readers or centers, if done. Not performed in this study Estimates of test reproducibility, if done. Not performed in this study 25 Discuss the clinical applicability of the study findings. 16 Test results 17 18 19 20 Estimates Discussion Page 185 of 253 12 & 14 Combination 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ Ges t Recruit ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ EFWc ✓ ✓ ✓ ✓ ✓ ✓ ✓ UAD-F PI ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ F-Pl GF ✓ ✓ ✓ ✓ ✓ ✓ Abs ent >12 hours ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ UAD-A PI>75th centile (Acharya et al., 2005a ) ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ UAD-A RI ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ Pl GF ✓ ✓ ✓ ✓ ✓ UAD-P PI>95th centile (Acharya et al., 2005a ) ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ UAD-F PI>95th centile (Acharya et al., 2005b) ✓ ✓ ✓ MCA:UAD-F RI ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ UAD-F RI ✓ ✓ ✓ ✓ MCA:UAD-A RI ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ UAD-P PI>90th centile (Acharya et al., 2005a ) ✓ ✓ MCA:UAD-F PI MoM<5th centi le (Mora l es-Rosello et al., 2014) ✓ Abnormal UAD-F PI/RI (Acharya et al., 2005b) ✓ UAD-F RI>95th centile (Acharya et al., 2005b) AUC 0.75 0.93 0.94 0.94 0.94 0.91 0.90 0.90 0.90 0.89 0.82 0.88 0.82 0.88 0.86 0.77 0.76 0.80 p 0.0007 0.0005 0.0005 0.0006 0.005 0.006 0.002 0.002 0.002 0.06 0.006 0.07 0.04 0.04 0.10 0.07 0.20 Model ? 0 3 2 1 SUPPLEMENTARY FIGURE 1: Predictive model generation. Multiple logistic regression with backwards elimination was used to identify combinations of variables with superior predictive performance compared to the baseline (“0”) model (p<0.05). The variable(s) contributing least (greatest p value) to any given combination of variables was removed at each iteration; if this resulted in loss of significance compared to “baseline” the variable was replaced and the next least contributing variable removed. Th e final three significant iterations were identified as candidate models (1 – 3). Key: Gest Recruit = gestation at recruitment. EFWc = estimated fetal weight centile (Bulk centile calculator v6.7 (UK) (Gestation Network, Birmingham , UK). UAD = umbilical artery (-A = abdominal insertion, -F = free loop, -P = placental insertion). PI = Pulsatility index. RI = Resistance index. Abnormal UAD-F = UAD-F PI or RI >95th centile or absent/reversed end diastolic flow. PlGF = Placental growth hormone (F- = free; (PlGF / sFlt-1)xPlGF). Page 186 of 253 CHAPTER 7: DISCUSSION AND FUTURE WORK 7.1 Introduction Attempts to identify fetuses “at risk” of stillbirth have primarily focused on the detection of the SGA infant, but have largely failed to reduce stillbirth rates (Bricker et al., 2008; Grivell et al., 2012b). Such studies have been thwarted by inadequate study design and sample number to detect small but significant improvement in stillbirth rates (Smith, 2010; Smith, 2012). An alternative strategy is the identification of placental failure (Heazell et al., 2015) as up to 65% of stillbirths are associated (causally or otherwise) with placental disease (Ptacek et al., 2014). Research has demonstrated that similar placental features, particularly in terms of macrostructure, are seen ex vivo following stillbirth regardless of its cause (Out et al., 1991; Tantbirojn et al., 2009; Pinar et al., 2014; Worton et al., 2014). RFM is associated with a two- to three-fold increased risk of stillbirth (O'Sullivan et al., 2009; Pagani et al., 2014a) and RFM live births display similar placental pathology (Warrander et al., 2012; Winje et al., 2012) to those of stillborn infants. The International Stillbirth Alliance recommended the “timely evaluation of women reporting [reduced] fetal movements” to prevent stillbirth in high-income countries (Flenady et al., 2011b). The studies of this thesis directly address this research priority and support the hypotheses that RFM is a symptom of placental dysfunction, which may evolve to cause APO including stillbirth. In chapter two I demonstrated that the placentas of RFM pregnancies with APO differ from their NPO counterparts in a range of structural and functional features that are similar to those associated with stillbirth and FGR. In chapters three to five I demonstrated that aspects of placental macrostructure, vascularity and vascular function and endocrine function can be measured by a range of traditional and novel placental biomarkers. Finally, in chapter six I demonstrated that addition of placental biomarkers Page 187 of 253 to baseline assessment of pregnancies at presentation with RFM can improve prediction of APO. I will now discuss the findings, strengths and limitations of each study and make recommendations for future work to extend, confirm or refute the findings of this thesis. 7.2 Placental features of late-onset adverse pregnancy outcome Placentas from RFM pregnancies ending in APO are smaller, less vascular and demonstrate altered placental vascular and endocrine function compared to their NPO counterparts. Remarkable similarities are demonstrated between my findings and those of other researchers studying placentas from pregnancies complicated by stillbirth or by live-birth of an FGR infant (Table 19; RFM APO data derived solely from the studies of this thesis). This is particularly true in terms of placental size (Biswas and Ghosh, 2008; Worton et al., 2014), vascularity (Out et al., 1991) and key endocrine variables (hCG, hPL and sFlt-1) (Shibata et al., 2005; Gagnon et al., 2008; Romero et al., 2008; Chaiworapongsa et al., 2013), and builds on earlier work that showed placental transport was reduced in RFM pregnancies with APO (Warrander et al., 2012). Parameters measuring villous microstructure and vascular function were less consistent with the ex vivo placental phenotype of liveborn FGR infants (Table 19). RFM APO placentas did not demonstrate significant differences in villous area, trophoblast area and trophoblast ratio as would be expected from studies of placentas from FGR infants (Mayhew et al., 2003; Daayana et al., 2004). Similarly, instead of enhanced CPA constriction and relaxation in response to U46619 and SNP treatment and altered length-tension characteristics (as described by Mills et al. (2005 & 2011) in FGR placentas) we observed relative CPA SNP insensitivity and unaltered thromboxane response and length tension characteristics in RFM APO CPAs. Being of similar methodologies and sample sizes, it is likely that the observed differences relate to the placental pathology being studied; predominantly Page 188 of 253 early-onset FGR in the previously published studies and predominantly lateonset placental dysfunction in the current study. TABLE 19: Comparison of placental features between placentas of stillborn, liveborn growth restricted and adverse outcome reduced fetal movements pregnancies. Placental Structure Trimmed placental weight / volume Placental diameter Centrality of cord insertion Villous vascularity Trophoblast area Placental Function Fetoplacental weight ratio Third trimester endocrine profile - hCG - hPL - Progesterone - PlGF - sFlt-1 Placental vascular function - Thromboxane sensitivity - Nitric oxide sensitivity - Vessel compliance - Maximal constrictive ability - Physiological reserve Stillbirth FGR APO RFM ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↔/↓ ↓ ↓ ↔ ↓ ↔ ↑ ↑ ↔ Unreported Unreported ↓* ↓* Unreported Unreported ↓* ↓ ↑* ↔* N/A N/A N/A N/A N/A ↑ ↑ ↑ ↓ Unreported ↔*/↓ ↔/↓* ↔ ↓ ↔ ↔ ↓ ↔ ↔ ↔ The placental features of adverse pregnancy outcome (APO) following maternal report of reduced fetal movements (RFM) as described in Chapter two are similar to those of placentas of stillborn infants (Out et al., 1991; Gagnon et al., 2008; Tantbirojn et al., 2009; Chaiworapongsa et al., 2013; Pinar et al., 2014; Worton et al., 2014) and liveborn fetal growth restricted (FGR) infants (Smith et al., 1997; Chen et al., 2002; Daayana et al., 2004; Mills et al., 2005; Sibley et al., 2005; Biswas and Ghosh, 2008; Biswas et al., 2008; Gagnon et al., 2008; Mills et al., 2011; Roje et al., 2011; Vedmedovska et al., 2011; Spinillo et al., 2012; Junaid et al., 2014) particularly in relation to placental structure and inferred endocrine function. Key: * = inferred from maternal circulating hormone concentration. N/A = not applicable. ↑ = increased. ↓ = decreased. ↔ = unchanged. Bold indicates consistent placental findings between all three pregnancy groups. The predominant strengths of this study include; a short interval between RFM and delivery (suggesting that the ex vivo observed placental features were present in utero at the time of RFM), short interval between delivery and collection (minimising storage artefact (Garrod et al., 2013)) and structural and functional characterisation of placental samples which were restricted to aspects of placental structure and function with potential non invasive in utero assessment techniques. Its primary limitation is that the composite outcome, although adapted from that of previous obstetric Page 189 of 253 studies (Hannah et al., 2000; Boers et al., 2010; Dutton et al., 2012; Barrett et al., 2013; Heazell et al., 2013) to focus on placentally-derived APO e.g. excluding sepsis and (largely iatrogenic) prematurity, is principally influenced by fetal size (not growth). While the individual elements of this composite outcome are each linked to increased perinatal mortality (Malin et al., 2010; Iliodromiti et al., 2014; Moraitis et al., 2014) the definition is not specific for placental dysfunction. Furthermore, as all placentas studied came from liveborn fetuses it cannot be ruled out that differences may exist between the placentas of those with placental dysfunction that are born alive or die in utero. Overall, the findings of the detailed multi-domain assessment of placental structure and function described herein validate the choice of study population (RFM), as there is evidence of APO in combination with placental pathology. Furthermore, if these aspects of placental structure and function can be measured in utero, they might form the basis of clinically useful tests that could enable differentiation between pregnancies likely to end in successful or unsuccessful outcome after maternal perception of RFM or in other pregnancies at high-risk of placentally-mediated APO. 7.3 Validation of size, vascular and endocrine placental biomarkers Measurements reflective of placental size, vascularity and vascular and endocrine function (particularly the balance of pro- and anti-angiogenic factors in the placental environment in vivo) can be obtained in utero using 2D and 3D ultrasound, colour Doppler and maternal serum hormone concentration (the findings of chapters three to five are summarised in Table 20). Such validation of biomarkers by comparison with ex vivo measurements has not frequently been attempted (Spellacy et al., 1975; Furuhashi et al., 1984a; Mills et al., 2005; Yin et al., 2009; Azpurua et al., 2010; Wright et al., 2011). When attempts have been made associations have not always been found (Mills et al., 2005; Yin et al., 2009; Wright et al., Page 190 of 253 2011), for example sonographic placental Grannum grading which failed to relate to placental calcification or stereological assessments of the placenta (Yin et al., 2009). TABLE 20: Placental biomarkers validated within the studies of this thesis. Biomarker target Measurable in utero by Intraobserver ICC Interobserver ICC Placental Structure Ultrasound VOCAL 30° 0.66 0.54 Serum ELISA hCG hPL PlGF Not studied Not studied Placental diameter Length Width Curvilinear Villous vascularity Doppler IPAD PI & RI UAD-F PI & RI 0.68 0.70 0.55-0.57 0.66-0.75 0.096 0.37 0.40-0.41 0.50-0.55 Trimmed placental weight / volume Placental Function Third trimester maternal serum endocrine profile - hCG (lysate & CM content) - hPL (lysate content) - sFlt-1 (CM content) Placental vascular function - Nitric oxide sensitivity Unable to validate against examined in vivo measures Unable to validate against examined in vivo measures Showing placental features demonstrated to be altered in ex vivo placentas of APO RFM pregnancies, their correlation to examined in utero placental biomarkers and biomarker reliability. Optimal reliability is considered ICC≥0.75 (Khan and Chien, 2001). Key: ICC = intraclass correlation coefficient. VOCAL 30° = 3D volume analysis technique (Virtual Organ Computer Aided analysis) with 30° rotation angle. ELISA = enzyme linked Immunosorbant assay. hCG = human chorionic gonadotrophin. hPL = human placental lactogen. PlGF = placental growth factor. sFlt-1 = soluble fms-like tyrosine kinase-1. IPAD = intraplacental artery Doppler. UAD-F = free-loop (traditional) umbilical artery Doppler. PI = pulsatility index. RI = resistance index. CM = explant conditioned media. Where successful ex vivo validation has been previously reported, my findings are largely in keeping with, and extend, the overall trends, particularly regarding the relationship between placental macrostructure (volume) and sonographic estimated placental volume (Azpurua et al., 2010) and maternal serum placentally-derived hormone concentrations of hCG, hPL and PlGF, (Spellacy et al., 1975; Gordon et al., 1977; Furuhashi et al., 1984a; Shibata et al., 2005; Dutton et al., 2012) (Table 20). Failure to confirm the finding of Spellacy et al. (1975) relating progesterone concentration to placental weight is likely an artefact of small sample size in Page 191 of 253 my study (N=50 vs. N=268). It is notable that the correlation of serum hPL concentration to ex vivo placental volume is stronger than that of VOCAL 30° placental volume estimates (R s=0.46 vs. 0.40 respectively). Thus, hPL measurement may have the potential to be a more efficient placental biomarker than sonographic placental volume. This study was the first to attempt biological correlation of true and in utero estimated placental length, width, and FPR. Porat et al. (2013) examined the difference between in utero and ex vivo placental depth but correlation was not performed. Placental length, width, depth and FPR measurements demonstrate biological correlation but are not perfectly accurate (i.e. a 1:1 ratio between in vivo and ex vivo correlates) implying that sonographic placental measurements should be compared to in utero reference charts or normograms. With respect to vascular structure, the strongest biological correlation was demonstrated between IPAD impedance and villous vascularity. This is in keeping with the observations of Thompson and Trudinger (1990) where the number of branches of a “placental circuit” is the principle determinant of impedance; by examining impedance closer to the placental microvasculature, minor degrees of placental dysfunction have greater impact on the impedance of the smaller placental arterial tree downstream of the IPAD sampling site than they do on at sampling sites further upstream (UAD-F). Ex vivo placental examination was not able to validate all in utero estimates of placental structure and function (Table 20); this was particularly true of function biomarkers (endocrine and vascular). There was no relationship between the concentrations of hCG, hPL and sFlt-1 in maternal serum, placental lysate or CM. The correlation between UAD-F impedance and CPA thromboxane sensitivity demonstrated herein contrasts with previous studies that failed to demonstrate correlation between CPA thromboxane response and UAD-F PI and RI (Mills et al., 2005; Wareing et al., 2005), possibly due to the more physiological oxygen and arterial tension characteristics under which the current study was performed (Wareing et Page 192 of 253 al., 2002; Wareing et al., 2006). Correlations between UAD-F impedance and placental arterial nitric oxide sensitivity failed to reach significance (p=0.070.10), likely as a result of a relatively small sample size (N=25) and significant biological variability (SNP EC 50 IQR in CPAs from NPO placentas 6-24nM). In early-onset placental disease the combination of reduced branching angiogenesis (Jackson et al., 1992), increased circulating thromboxane concentrations (Read et al., 1999), and enhanced arterial constriction to thromboxane stimulation (Mills et al., 2005) is likely to result in increased baseline vascular impedance, and development of abnormalities of the UAD-F waveform with minor degrees of additional damage such as thrombosis (Sugimura et al., 2001). In contrast, in late-onset placental disease the well-expanded fetoplacental vascular tree (Burton and Jauniaux, 1995) is relatively immune to minor degrees of additional damage (Morrow et al., 1989), with lesions of maternal underperfusion (ParraSaavedra et al., 2013) predominating. Therefore, the UAD-F is less likely to demonstrate significant abnormality in “low risk” pregnancies with lateonset placental disease. Other in utero placental structure or function assessment techniques (maternal serum hCG, PlGF and sFlt-1 concentrations) were found to relate to aspects of placental function that were unchanged between the placentas of APO and NPO RFM pregnancies (respectively transcription rate and tissue content). These findings remain logical in light of other reports in the literature. The inverse correlation between maternal hCG serum concentration and relative transcription rate is in keeping with its production being controlled at the level of transcription (Hussa, 1980). A trend to positive association between PlGF serum concentration and tissue content is in agreement with the work of Shibata et al. (2005) who demonstrated correlation only in placental tissue from complicated pregnancies, not in placental tissue from normal pregnancies. Finally, the positive correlation between sFlt-1 serum concentration and tissue content supports the placenta being the primary source of circulating sFlt-1 in pregnancy (Reddy et al., 2009). Although these aspects of ex vivo placental Page 193 of 253 function examination were unaltered in APO RFM placentas, the findings confirm that examination of serum placental biomarkers provides information about in utero placental function as well as structure. For sonographic placental biomarkers (with the exception of UAD-A PI and resistance index (RI) and UAD-F PI) the intra- and inter-observer reliabilities are suboptimal for use as a clinical test (ICC <0.75 (Khan and Chien, 2001)). This will impede the development of clinically useful in utero reference charts and to reduce the discriminative power of these biomarkers. Reproducibility of sonographic placental size and vascular biomarkers in the studies of this thesis are considerably lower than those (where published) for similar techniques applied in the first and second trimesters of pregnancy (Deurloo et al., 2007a; de Paula et al., 2008; Cheong et al., 2010; Jones et al., 2011; Milligan et al., 2014). This is not unexpected, due to relatively increased technical difficulty in obtaining full placental visualisation after midtrimester (Deurloo et al., 2007a), and the considerable variability in ex vivo placental size (for example true placental volume ranged from 227 to 881cm3 in NPO RFM pregnancies). Where intraobserver reliability is significantly higher than the inter-observer reliability (e.g. length and width) operator training may be the key to improving test performance. When both reliabilities are suboptimal, the tests themselves require refinement. This will be discussed in more detail in the “future work” section. The predominant strengths of these studies are the validation of both structural and functional placental biomarkers (obtained using widely available technology) in late gestation using correlation rather than qualitative comparison, minimisation of type 2 error introduced by subsequent placental growth, maturation or damage between assessment and delivery by restriction of validation analyses to those pregnancies delivered within seven days, and rigorous assessment of intra- and interobserver reliability. The primary limitation is that most examined placental Page 194 of 253 biomarkers are insufficiently reliable in their current format for widespread clinical use. The ability to obtain reproducible antenatal measurements that reflect placental structure and function in the third trimester of pregnancy is key to reducing placentally-related stillbirths. The studies of this thesis confirm that abnormal ex vivo size, shape and function of placentas is associated with APO (Table 20); preliminary work from other populations indicate that multimodal in utero placental examination in the first and second trimesters can assist prediction of these adverse outcomes (Toal et al., 2007; Costa et al., 2008; Toal et al., 2008a; Toal et al., 2008b; Proctor et al., 2009; Porat et al., 2013). Here, in delineating the biological correlations and reproducibility of a range of potential placental structural and functional biomarkers in the third trimester, my studies represent an important step in understanding how best to perform and interpret antenatal placental assessments at a time when elective delivery is possible to prevent stillbirth. 7.4 Placental assessment in the prediction of adverse pregnancy outcome When evaluated in a clinical setting, the sensitivity of predicting APO following RFM improved from as little as 8.9% with baseline care to as much as 37.5% with addition of placental biomarkers. This was achieved by measurement of impedance at the UAD-A, UAD-F, UAD-P and the balance of pro- and anti-angiogenic factors (PlGF and sFlt-1 respectively) in maternal serum, in addition to baseline care. However a significant post-negative test probability (≥11.8%) remained, reflected in the sub-optimal negative likelihood ratio for all predictive models (≥0.62). The true increase in sensitivity achieved by antenatal placental assessment in RFM may be as little as 1.9% due to the wide confidence intervals of the test performance parameters for each predictive model. This results from the relatively small study sample size (further reduced by failure to obtain Page 195 of 253 all results for every participant) with significant heterogeneity within NPO and APO groups, suboptimal placental biomarker reliability (ICC<0.75) and imprecision of the definition of “normal” or “adverse” pregnancy outcome (as discussed in earlier). Together these factors may also have resulted in type 1 error in the rejection of potentially useful placental biomarkers, e.g. IPAD impedance, which is reflective of villous vascularity but was excluded from analysis at the variable reduction stage (p=0.11) on the basis of an arbitrarily defined a priori threshold of univariate analysis p<0.10. Despite demonstrating smaller ex vivo placental size in stillbirth, FGR and APO RFM pregnancy placentas, no sonographic markers of placental size were found to be independently predictive of APO, although maternal serum total PlGF concentration was predictive and is shown to correlate with placental volume and weight. This contrasts with the findings of Costa et al. (2008) and Proctor et al. (2009) who describe that second trimester placental length of <10cm predicts delivery of an SGA infant. These studies were conducted on small numbers (N≤90) of midtrimester “high-risk” pregnancies, compared with the larger (N=300), late-gestation RFM cohort studied in this thesis which is more representative of the management dilemmas faced in the general obstetric population (Heazell et al., 2008; Flenady et al., 2009). In contrast to the negative findings of Froen et al. (2008b) and Dutton et al. (2012), significant predictive value of UAD-F monitoring (when considered as a continuous variable) in RFM pregnancies was demonstrated. Amongst late-pregnancy complications, this is likely to be primarily due to its correlation with villous vessel density (and a further trend to correlation with placental arterial nitric oxide sensitivity), which are altered in lateonset placental disease, rather than thromboxane sensitivity that is not altered. Conversely, I was unable to demonstrate independent predictive value of UtAD impedance variables despite reported associations between high UtAD impedance and APO following RFM (Pagani et al., 2014a; Pagani et al., 2014b), and between ex vivo placental lesions of maternal Page 196 of 253 underperfusion and late-onset FGR (Parra-Saavedra et al., 2014c). Suboptimal UtAD reproducibility (unstudied in this thesis) may be a contributing cause, with published first and second trimester intra- and inter-observer ICCs varying from 0.69-0.94 and 0.10 – 0.87 respectively (Hollis et al., 2001; Ferreira et al., 2014). Finally, inclusion of F-PlGF (as a continuous variable) in the proposed predictive models is interesting in light of the recent study by Griffin et al. [In press, (2015)]. Pregnancies clinically suspected to be SGA were triaged on the basis of maternal plasma free PlGF concentrations (Triage® test, Alere Inc., San Diego, US) above or below the 5 th centile. A small increase in sensitivity (57.7% vs. 69.2%) for prediction of SGA delivery was demonstrated when combined with ultrasound assessment (EFWc <10 or oligohydramnios). The authors concluded that this did not represent clinically significant improvement in risk prediction. However, this could be extrapolated to the prediction of one additional SGA birth for every nine cases screened. Due to the study design, ultrasound measurements were only available for 343/592 (57.8%) of studied pregnancies, of whom only 52 subsequently delivered an SGA infant (15.2%) and UAD-F measurements (which individually were relatively insensitive but highly specific for SGA birth) were excluded from this combined assessment analysis. The predominant strengths of this study are the trimodal, prospective assessment of in utero placental structure and function in an unselected population at a time when delivery could be considered to prevent stillbirth. There is unlikely to be any singular pathology underpinning all stillbirths, therefore it is equally unlikely that any single placental biomarker would accurately predict all stillbirths of placental origin. Trimodal placental assessment has not previously been attempted in the third trimester, and in earlier pregnancy has applied variously defined thresholds to mark individual pregnancies as “screen positive” on the basis of one or more aspects of placental structure or function instead of considering all variables continuously (Viero et al., 2004; Toal et al., 2007; Costa et al., 2008; Toal et Page 197 of 253 al., 2008a; Proctor et al., 2009). This does not allow for interplay between different aspects of placental structure and function. Such holistic assessment may be more sensitive than application of “rule of thumb”, for example in the national Down’s syndrome screening programme where combined screening has replaced single marker assessment (Canick, 2012). Another important strength of this study is that the facility to measure the placental biomarkers explored herein, with the potential exception of 3D placental volume, already exists in the NHS and other high-income countries’ healthcare systems. Therefore, these findings could be easily translated into clinical care within a short period of time. The primary limitation of the study is that it is underpowered to use stillbirth as a primary outcome, with a risk of over fitting in proposed models 2 and 3 as the event per variable rate is below 10. It is also likely that potentially useful placental biomarkers (such as IPAD impedance which correlated strongly with ex vivo villous vascularity) have been rejected due to type 2 error introduced by sub-optimal reliability and failure to obtain measurements in all participants. Finally the lack of available data regarding the total number and characteristics and outcomes of the wider population of women (and their babies) who presented with RFM during the study period prevents the assessment of potential selection bias, although the study population is similar to that of previous RFM cohorts (Dutton et al., 2012; Heazell et al., 2013; Pagani et al., 2014a; Pagani et al., 2014b). The implication of these findings is that placental dysfunction associated with increased risk of stillbirth can be prospectively identified in utero to improve prediction of APO. Therefore, care of the woman and her baby could be rationalised according to individualised risk. This may include increased frequency of fetal monitoring or earlier elective delivery for those with a “high-risk” placental profile. Equally, the significant post-negative test probability of APO supports more liberal use of serial ultrasound in RFM pregnancies to identify those with declining growth trajectory whom may also benefit from similar management. Full economic evaluations are Page 198 of 253 required to determine whether the investment in increased antenatal screening is economically justified. This, in turn, is reliant upon a realistic and comprehensive appreciation of the cost (individual and societal) of stillbirth, which is currently lacking (Mistry et al., 2013). Page 199 of 253 7.5 Future work In order to confirm, refute, optimise and further the findings of this thesis, the following areas for future work are prioritised: 1. Improving the definition of APO Adjudication by markers of chronic in utero hypoxia, for example umbilical cord blood erythropoietin concentration (Teramo et al., 2002), nucleated red blood cell count (Korst et al., 1996), or prelabour maternal blood concentration of hypoxia-induced fetal RNA (Whitehead et al., 2013). Adjudication by evidence of abnormal growth trajectory, for example (time-adjusted) scan to delivery fetal weight centile change, centile decline on serial third trimester 2D ultrasound EFW (data to support this may be obtained from the Pregnancy Outcome Prediction Study (Pasupathy et al., 2008)), or comparison to individual-growth potential, estimated from serial second trimester 3D EFW (Lee et al., 2009). Adjudication by evidence of suboptimal fetal adipose deposition, for example neonatal ponderal index (Landmann et al., 2006) or percentage body fat on air displacement plethysmography (Lee et al., 2012). 2. Improving the reliability of placental biomarkers Further investigation of anatomically defined Doppler sampling points such as UAD-A and UAD-P (Figueras et al., 2006; Khare et al., 2006). Evaluation of the potential benefit of wider ultrasound probe with extended angle range to minimise “missed” tissue volumes in assessment of in utero placental volume. Creation of a rigorous training programme with ongoing competence assessment for sonographers attempting in utero placental assessment, (similar to the programme developed for assessment of nuchal translucency (Fries et al., 2007)). Page 200 of 253 Creation of in utero reference charts for placental size measures, placental vascular biomarkers and placentally-derived hormone concentrations, constructed from measurements in adequately powered low-risk populations after optimisation of placental biomarker technique and appropriate operator training as above. 3. Assessment of the external validity of the study findings in other conditions associated with increased risk of stillbirth Small for gestational age fetuses at term Term pregnancies complicated by maternal diabetes with good glycaemic control Post-mature pregnancies 4. Assessment of the ability of antenatal placental assessment to predict and prevent stillbirth Assessment of the “number needed to assess” to identify and prevent (by delivery) one stillbirth Assessment of health economic implications of introducing placental assessment, including evaluation of the economic cost of stillbirth itself Page 201 of 253 7.6 Conclusion Ultimately, multimodal placental assessment at term gestation (as developed and feasibility-tested in the studies of this thesis) could be incorporated into the routine care of pregnant women in high-income countries, where (with improvements in test reliability, and standardised management protocol) it could identify as many as four in every 10 term pregnancies that are at increased risk of stillbirth. On a background of 26% of all stillbirths in high-income countries occurring at or beyond 37 weeks gestation (Zeitlin, 2013), combined with elective delivery, this screening strategy has the potential to prevent thousands (up to 40%) of term stillbirths in these countries. Page 202 of 253 REFERENCES Abramowicz, J. S. & Sheiner, E. (2007) In utero imaging of the placenta: importance for diseases of pregnancy. Placenta, 28 Suppl A, S14-22. Acharya, G., Wilsgaard, T., Berntsen, G. K., Maltau, J. M. & Kiserud, T. (2005a) Reference ranges for serial measurements of blood velocity and pulsatility index at the intra-abdominal portion, and fetal and placental ends of the umbilical artery. Ultrasound Obstet Gynecol, 26(2), 162-9. Acharya, G., Wilsgaard, T., Berntsen, G. K., Maltau, J. M. & Kiserud, T. (2005b) Reference ranges for serial measurements of umbilical artery Doppler indices in the second half of pregnancy. Am J Obstet Gynecol, 192(3), 937-44. Ahmed, A., Dunk, C., Ahmad, S. & Khaliq, A. (2000) Regulation of placental vascular endothelial growth factor (VEGF) and placenta growth factor (PIGF) and soluble Flt-1 by oxygen--a review. Placenta, 21 Suppl A, S16-24. Alves, T. C., Jarak, I. & Carvalho, R. A. (2012) NMR methodologies for studying mitochondrial bioenergetics. Methods Mol Biol, 810, 281309. Alwasel, S. H., Abotalib, Z., Aljarallah, J. S., Osmond, C., Alkharaz, S. M., Alhazza, I. M., Badr, G. & Barker, D. J. (2010) Changes in placental size during Ramadan. Placenta, 31(7), 607-10. Alwasel, S. H., Harrath, A. H., Aljarallah, J. S., Abotalib, Z. , Osmond, C., Al Omar, S. Y., Thornburg, K. & Barker, D. J. (2013) The velocity of fetal growth is associated with the breadth of the placental surface, but not with the length. Am J Hum Biol, 25(4), 534-7. Anderson, L., Martin, W., Higgins, C., Nelson, S. M. & Norman, J. E. (2009) The effect of progesterone on myometrial contractility, potassium channels, and tocolytic efficacy. Reprod Sci, 16(11), 1052-61. Arakaki, T., Hasegawa, J., Nakamura, M., Hamada, S., Mishina, M., Takita, H., Ichizuka, K. & Sekizawa, A. (2014) Prediction of early and late onset pregnancy induced hypertension using the three-dimensional Page 203 of 253 ultrasound placental volume and uterine artery Doppler findings. Ultrasound Obstet Gynecol. Arleo, E. K., Troiano, R. N., da Silva, R., Greenbaum, D. & Kliman, H. J. (2014) Utilizing two-dimensional ultrasound to develop normative curves for estimated placental volume. Am J Perinatol, 31(8), 683-8. Artunc Ulkumen, B., Pala, H. G., Uyar, Y., Koyuncu, F. M. & Bulbul Baytur, Y. (2014a) The alteration in placental volume and placental mean grey value in growth-restricted pregnancies assessed by 3D ultrasound (Growth Restriction & 3D Ultrasonography). J Obstet Gynaecol, 1-4. Artunc Ulkumen, B., Pala, H. G., Uyar, Y., Koyuncu, F. M. & Bulbul Baytur, Y. (2014b) The assessment of placental volume and mean gray value in preeclamptic placentas by using three-dimensional ultrasonography. J Matern Fetal Neonatal Med, 1-4. Azpurua, H., Funai, E. F., Coraluzzi, L. M., Doherty, L. F., Sasson, I. E., Kliman, M. & Kliman, H. J. (2010) Determination of placental weight using two-dimensional sonography and volumetric mathematic modeling. Am J Perinatol, 27(2), 151-5. Babovic, I., Tadic, J., Plesinac, S., Radojicic, Z. & Plecas, D. (2011) Doppler assessment between pathological examination of the placenta and late fetal intrauterine demise. Clin Exp Obstet Gynecol, 38(1), 43-5. Bahado-Singh, R. O., Kovanci, E., Jeffres, A., Oz, U., Deren, O., Copel, J. & Mari, G. (1999) The Doppler cerebroplacental ratio and perinatal outcome in intrauterine growth restriction. Am J Obstet Gynecol, 180(3 Pt 1), 750-6. Bahlmann, F., Reinhard, I., Krummenauer, F., Neubert, S., Macchiella, D. & Wellek, S. (2002) Blood flow velocity waveforms of the fetal middle cerebral artery in a normal population: reference values from 18 weeks to 42 weeks of gestation. J Perinat Med, 30(6), 490-501. Balihallimath, R. L., Shirol, V. S., Gan, A. M., Tyagi, N. K. & Bandankar, M. R. (2013) Placental morphometry determines the birth weight. J Clin Diagn Res, 7(11), 2428-31. Bano, S., Chaudhary, V., Pande, S., Mehta, V. & Sharma, A. (2010) Color doppler evaluation of cerebral-umbilical pulsatility ratio and its Page 204 of 253 usefulness in the diagnosis of intrauterine growth retardation and prediction of adverse perinatal outcome. Indian J Radiol Imaging, 20(1), 20-5. Barrett, J. F., Hannah, M. E., Hutton, E. K., Willan, A. R., Allen, A. C., Armson, B. A., Gafni, A., Joseph, K. S., Mason, D., Ohlsson, A., Ross, S., Sanchez, J. J. & Asztalos, E. V. (2013) A randomized trial of planned cesarean or vaginal delivery for twin pregnancy. N Engl J Med, 369(14), 1295305. Baschat, A. A. (2004) Fetal responses to placental insufficiency: an update. BJOG, 111(10), 1031-41. Baschat, A. A. (2006) The fetal circulation and essential organs-a new twist to an old tale. Ultrasound Obstet Gynecol, 27(4), 349-54. Baschat, A. A. & Gembruch, U. (2003) The cerebroplacental Doppler ratio revisited. Ultrasound Obstet Gynecol, 21(2), 124-7. Baschat, A. A., Gembruch, U. & Harman, C. R. (2001) The sequence of changes in Doppler and biophysical parameters as severe fetal growth restriction worsens. Ultrasound Obstet Gynecol, 18(6), 571-7. Baschat, A. A., Gembruch, U., Reiss, I., Gortner, L. & Diedrich, K. (1997) Demonstration of fetal coronary blood flow by Doppler ultrasound in relation to arterial and venous flow velocity waveforms and perinatal outcome--the 'heart-sparing effect'. Ultrasound Obstet Gynecol, 9(3), 162-72. Baschat, A. A., Gembruch, U., Reiss, I., Gortner, L., Weiner, C. P. & Harman, C. R. (2000) Relationship between arterial and venous Doppler and perinatal outcome in fetal growth restriction. Ultrasound Obstet Gynecol, 16(5), 407-13. Baschat, A. A., Muench, M. V. & Gembruch, U. (2003) Coronary artery blood flow velocities in various fetal conditions. Ultrasound Obstet Gynecol, 21(5), 426-9. Bellotti, M., Pennati, G., De Gasperi, C., Bozzo, M., Battaglia, F. C. & Ferrazzi, E. (2004) Simultaneous measurements of umbilical venous, fetal hepatic, and ductus venosus blood flow in growth-restricted human fetuses. Am J Obstet Gynecol, 190(5), 1347-58. Page 205 of 253 Benavides-Serralde, J. A., Hernandez-Andrade, E., Figueroa-Diesel, H., Oros, D., Feria, L. A., Scheier, M., Figueras, F. & Gratacos, E. (2010) Reference values for Doppler parameters of the fetal anterior cerebral artery throughout gestation. Gynecol Obstet Invest, 69(1), 33-9. Benn, P. A., Horne, D., Briganti, S., Rodis, J. F. & Clive, J. M. (1996) Elevated second-trimester maternal serum hCG alone or in combination with elevated alpha-fetoprotein. Obstet Gynecol, 87(2), 217-22. Benton, S. J., Hu, Y., Xie, F., Kupfer, K., Lee, S. W., Magee, L. A. & von Dadelszen, P. (2012) Can placental growth factor in maternal circulation identify fetuses with placental intrauterine growth restriction? Am J Obstet Gynecol, 206(2), 163 e1-7. Bhide, A., Acharya, G., Bilardo, C. M., Brezinka, C., Cafici, D., HernandezAndrade, E., Kalache, K., Kingdom, J., Kiserud, T., Lee, W., Lees, C., Leung, K. Y., Malinger, G., Mari, G., Prefumo, F., Sepulveda, W. & Trudinger, B. (2013) ISUOG practice guidelines: use of Doppler ultrasonography in obstetrics. Ultrasound Obstet Gynecol, 41(2), 23339. Biswas, S. & Ghosh, S. K. (2008) Gross morphological changes of placenta s associated with intrauterine growth restriction of fetuses: a case control study. Early Hum Dev, 84(6), 357-62. Biswas, S., Ghosh, S. K. & Chhabra, S. (2008) Surface area of chorionic villi of placentas: an index of intrauterine growth restriction of fetuses. J Obstet Gynaecol Res, 34(4), 487-93. Blaicher, W., Brugger, P. C., Mittermayer, C., Schwindt, J., Deutinger, J., Bernaschek, G. & Prayer, D. (2006) Magnetic resonance imaging of the normal placenta. Eur J Radiol, 57(2), 256-60. Bland, J. M. & Altman, D. G. (2003) Applying the right statistics: analyses of measurement studies. Ultrasound Obstet Gynecol, 22(1), 85-93. Boers, K. E., Vijgen, S. M., Bijlenga, D., van der Post, J. A., Bekedam, D. J., Kwee, A., van der Salm, P. C., van Pampus, M. G., Spaanderman, M. E., de Boer, K., Duvekot, J. J., Bremer, H. A., Hasaart, T. H., Delemarre, F. M., Bloemenkamp, K. W., van Meir, C. A., Willekes, C., Wijnen, E. J., Rijken, Page 206 of 253 M., le Cessie, S., Roumen, F. J., Thornton, J. G., van Lith, J. M., Mol, B. W. & Scherjon, S. A. (2010) Induction versus expectant monitoring for intrauterine growth restriction at term: randomised equivalence trial (DIGITAT). BMJ, 341, c7087. Boito, S., Struijk, P. C., Ursem, N. T., Stijnen, T. & Wladimiroff, J. W. (2002) Umbilical venous volume flow in the normally developing and growth-restricted human fetus. Ultrasound Obstet Gynecol, 19(4), 344-9. Bonifacio, S. L., Glass, H. C., Vanderpluym, J., Agrawal, A. T., Xu, D., Barkovich, A. J. & Ferriero, D. M. (2011) Perinatal events and early magnetic resonance imaging in therapeutic hypothermia. J Pediatr, 158(3), 360-5. Boroditsky, R. S., Reyes, F. I., Winter, J. S. & Faiman, C. (1975) Serum human chorionic gonadotropin and progesterone patterns in the last trimester of pregnancy: relationship to fetal sex. Am J Obstet Gynecol, 121(2), 238-41. Bozkurt, N., Basgul Yigiter, A., Gokaslan, H. & Kavak, Z. N. (2010) Correlations of fetal-maternal outcomes and first trimester 3-D placental volume/3-D power Doppler calculations. Clin Exp Obstet Gynecol, 37(1), 26-8. Braunholtz, D. A., Edwards, S. J. & Lilford, R. J. (2001) Are randomized clinical trials good for us (in the short term)? Evidence for a "trial effect". J Clin Epidemiol, 54(3), 217-24. Bricker, L., Neilson, J. P. & Dowswell, T. (2008) Routine ultrasound in late pregnancy (after 24 weeks gestation). Cochrane Database Syst Rev, (4), CD001451. Burd, I., Srinivas, S., Pare, E., Dharan, V. & Wang, E. (2009) Is sonographic assessment of fetal weight influenced by formula selection? J Ultrasound Med, 28(8), 1019-24. Burton, G. J. & Jauniaux, E. (1995) Sonographic, stereological and Doppler flow velocimetric assessments of placental maturity. Br J Obstet Gynaecol, 102(10), 818-25. Page 207 of 253 Byun, Y. J., Kim, H. S., Yang, J. I., Kim, J. H., Kim, H. Y. & Chang, S. J. (2009) Umbilical artery Doppler study as a predictive marker of perinatal outcome in preterm small for gestational age infants. Yonsei Med J, 50(1), 39-44. Campbell, S., Black, R. S., Lees, C. C., Armstrong, V. & Peacock, J. L. (2000) Doppler ultrasound of the maternal uterine arteries: disappearance of abnormal waveforms and relation to birthweight and pregnancy outcome. Acta Obstet Gynecol Scand, 79(8), 631-4. Canick, J. (2012) Prenatal screening for trisomy 21: recent advances and guidelines. Clin Chem Lab Med, 50(6), 1003-8. Carter, A. M. (2009) Evolution of factors affecting placental oxygen transfer. Placenta, 30 Suppl A, S19-25. Cetin, I. (2003) Placental transport of amino acids in normal and growthrestricted pregnancies. Eur J Obstet Gynecol Reprod Biol, 110 Suppl 1, S50-4. Cetin, I., Giovannini, N., Alvino, G., Agostoni, C., Riva, E., Giovannini, M. & Pardi, G. (2002) Intrauterine growth restriction is associated with changes in polyunsaturated fatty acid fetal-maternal relationships. Pediatr Res, 52(5), 750-5. Chafetz, I., Kuhnreich, I., Sammar, M., Tal, Y., Gibor, Y., Meiri, H., Cuckle, H. & Wolf, M. (2007) First-trimester placental protein 13 screening for preeclampsia and intrauterine growth restriction. Am J Obstet Gynecol, 197(1), 35 e1-7. Chaiworapongsa, T., Romero, R., Korzeniewski, S. J., Kusanovic, J. P., Soto, E., Lam, J., Dong, Z., Than, N. G., Yeo, L., Hernandez-Andrade, E., CondeAgudelo, A. & Hassan, S. S. (2013) Maternal plasma concentrations of angiogenic/antiangiogenic factors in the third trimester of pregnancy to identify the patient at risk for stillbirth at or near term and severe late preeclampsia. Am J Obstet Gynecol, 208(4), 287 e1-287 e15. Chappell, L. C., Duckworth, S., Seed, P. T., Griffin, M., Myers, J., Mackillop, L., Simpson, N., Waugh, J., Anumba, D., Kenny, L. C., Redman, C. W. & Shennan, A. H. (2013) Diagnostic accuracy of placental growth factor Page 208 of 253 in women with suspected preeclampsia: a prospective multicenter study. Circulation, 128(19), 2121-31. Chappell, L. C., Seed, P. T., Briley, A., Kelly, F. J., Hunt, B. J., Charnock-Jones, D. S., Mallet, A. I. & Poston, L. (2002) A longitudinal study of biochemical variables in women at risk of preeclampsia. Am J Obstet Gynecol, 187(1), 127-36. Chatterjee, M., Baliga, B. S. & Munro, H. N. (1976) Synthesis of human placental lactogen and human chorionic gonadotropin by polyribosomes and messenger RNA's from early and full term placentas. J Biol Chem, 251(10), 2945-51. Chauhan, S. P., Beydoun, H., Chang, E., Sandlin, A. T., Dahlke, J. D., Igwe, E., Magann, E. F., Anderson, K. R., Abuhamad, A. Z. & Ananth, C. V. (2014) Prenatal detection of fetal growth restriction in newborns classified as small for gestational age: correlates and risk of neonatal morbidity. Am J Perinatol, 31(3), 187-94. Chen, C. N., Cheng, Y. M., Lin, M. T., Hsieh, F. J., Lee, P. H. & Chang, K. J. (2002) Association of color Doppler vascularity index and microvessel density with survival in patients with gastric cancer. Ann Surg, 235(4), 512-8. Chen, K. H., Chen, L. R. & Lee, Y. H. (2012) The role of preterm placental calcification in high-risk pregnancy as a predictor of poor uteroplacental blood flow and adverse pregnancy outcome. Ultrasound Med Biol, 38(6), 1011-8. Chen, R. J., Huang, S. C., Chow, S. N. & Hsieh, C. Y. (1993) Human chorionic gonadotropin pattern in maternal circulation. Amniotic fluid and fetal circulation in late pregnancy. J Reprod Med, 38(2), 151-4. Cheong, K. B., Leung, K. Y., Li, T. K., Chan, H. Y., Lee, Y. P. & Tang, M. H. (2010) Comparison of inter- and intraobserver agreement and reliability between three different types of placental volume measurement technique (XI VOCAL, VOCAL and multiplanar) and validity in the invitro setting. Ultrasound Obstet Gynecol, 36(2), 210-7. Clark, D. E., Smith, S. K., He, Y., Day, K. A., Licence, D. R., Corps, A. N., Lammoglia, R. & Charnock-Jones, D. S. (1998) A vascular endothelial Page 209 of 253 growth factor antagonist is produced by the human placenta and released into the maternal circulation. Biol Reprod, 59(6), 1540-8. Cohen, J., Wauquiez, M. J. & Janse-Marec, J. (2014) Doppler analysis of the umbilical artery: why we must prefer the free loop of the cord. Minerva Ginecol, 66(5), 455-9. Cole, L. A. (2010) Biological functions of hCG and hCG-related molecules. Reprod Biol Endocrinol, 8, 102. Collins, S. L., Birks, J. S., Stevenson, G. N., Papageorghiou, A. T., Noble, J. A. & Impey, L. (2012a) Measurement of spiral artery jets: general principles and differences observed in small-for-gestational-age pregnancies. Ultrasound Obstet Gynecol, 40(2), 171-8. Collins, S. L., Stevenson, G. N., Noble, J. A. & Impey, L. (2012b) Developmental changes in spiral artery blood flow in the human placenta observed with colour Doppler ultrasonography. Placenta, 33(10), 782-7. Collins, S. L., Stevenson, G. N., Noble, J. A. & Impey, L. (2013a) Elsevier Trophoblast Research Award Lecture: Searching for an early pregnancy 3-D morphometric ultrasound marker to predict fetal growth restriction. Placenta, 34 Suppl, S85-9. Collins, S. L., Stevenson, G. N., Noble, J. A. & Impey, L. (2013b) Rapid calculation of standardized placental volume at 11 to 13 weeks and the prediction of small for gestational age babies. Ultrasound Med Biol, 39(2), 253-60. Collins, S. L., Stevenson, G. N., Noble, J. A., Impey, L. & Welsh, A. W. (2012c) Influence of power Doppler gain setting on Virtual Organ Computer aided AnaLysis indices in vivo: can use of the individual sub-noise gain level optimize information? Ultrasound Obstet Gynecol, 40(1), 75-80. Costa, S. L., Proctor, L., Dodd, J. M., Toal, M., Okun, N., Johnson, J. A., Windrim, R. & Kingdom, J. C. (2008) Screening for placental insufficiency in high-risk pregnancies: is earlier better? Placenta, 29(12), 1034-40. Cotter, S. L., Klika, V., Kimpton, L., Collins, S. & Heazell, A. E. (2014) A stochastic model for early placental development. J R Soc Interface, 11(97), 20140149. Page 210 of 253 Cousens, S., Blencowe, H., Stanton, C., Chou, D., Ahmed, S., Steinhardt, L., Creanga, A. A., Tuncalp, O., Balsara, Z. P., Gupta, S., Say, L. & Lawn, J. E. (2011) National, regional, and worldwide estimates of stillbirth rates in 2009 with trends since 1995: a systematic analysis. Lancet, 377(9774), 1319-30. Cowans, N. J., Spencer, K. & Meiri, H. (2008) First-trimester maternal placental protein 13 levels in pregnancies resulting in adverse outcomes. Prenat Diagn, 28(2), 121-5. Cowans, N. J., Stamatopoulou, A., Jaakohuhta, S. & Spencer, K. (2010) ADAM12 stability in first trimester maternal serum. Prenat Diagn, 30(6), 555-60. Cox, S., Golightly, S. & Sullivan, A. (2009) Confidential Enquiry into Maternal and Child Health (CEMACH) Perinatal Mortality 2007: United Kingdom. London: CEMACH. Crichton, C. A., Templeton, A. G., McGrath, J. C. & Smith, G. L. (1993) Thromboxane A2 analogue, U-46619, potentiates calcium-activated force in human umbilical artery. Am J Physiol, 264(6 Pt 2), H1878-83. Crocker, I. P., Tansinda, D. M., & Baker, P. N. (2004) Altered cell kinetics in cultured placental villous explants in pregnancies complicated by pre-eclampsia and intrauterine growth restriction. J Pathol 204(1):11-18. Cruz-Martinez, R., Figueras, F., Hernandez-Andrade, E., Puerto, B. & Gratacos, E. (2010) Longitudinal brain perfusion changes in near term small-for-gestational-age fetuses as measured by spectral Doppler indices or by fractional moving blood volume. Am J Obstet Gynecol, 203(1), 42 e1-6. D'Antonio, F., Patel, D., Chandrasekharan, N., Thilaganathan, B. & Bhide, A. (2013) Role of cerebroplacental ratio for fetal assessment in prolonged pregnancy. Ultrasound Obstet Gynecol, 42(2), 196-200. Daayana, S., Baker, P. & Crocker, I. (2004) An image analysis technique for the investigation of variations in placental morphology in pregnancies complicated by preeclampsia with and without intrauterine growth restriction. J Soc Gynecol Investig, 11(8), 545-52. Page 211 of 253 Daly, N., Brennan, D., Foley, M. & O'Herlihy, C. (2011) Cardiotocography as a predictor of fetal outcome in women presenting with reduced fetal movement. Eur J Obstet Gynecol Reprod Biol, 159(1), 57-61. Damodaram, M., Story, L., Eixarch, E., Patel, A., McGuinness, A., Allsop, J., Wyatt-Ashmead, J., Kumar, S. & Rutherford, M. (2010) Placental MRI in intrauterine fetal growth restriction. Placenta, 31(6), 491-8. Dar, P., Gebb, J., Reimers, L., Bernstein, P. S., Chazotte, C. & Merkatz, I. R. (2010) First-trimester 3-dimensional power Doppler of the uteroplacental circulation space: a potential screening method for preeclampsia. Am J Obstet Gynecol, 203(3), 238 e1-7. Daya, D. & Sabet, L. (1991) The use of cytokeratin as a sensitive and reliable marker for trophoblastic tissue. Am J Clin Pathol, 95(2), 137-41. De Jong, C. L., Francis, A., Van Geijn, H. P. & Gardosi, J. (2000) Customized fetal weight limits for antenatal detection of fetal growth restriction. Ultrasound Obstet Gynecol, 15(1), 36-40. de Laat, M. W., Franx, A., van Alderen, E. D., Nikkels, P. G. & Visser, G. H. (2005) The umbilical coiling index, a review of the literature. J Matern Fetal Neonatal Med, 17(2), 93-100. de Paula, C. F., Ruano, R., Campos, J. A. & Zugaib, M. (2008) Placental volumes measured by 3-dimensional ultrasonography in normal pregnancies from 12 to 40 weeks gestation. J Ultrasound Med, 27(11), 1583-90. Degani, S., Lewinsky, R. M., Berger, H. & Spiegel, D. (1995) Sonographic estimation of umbilical coiling index and correlation with Doppler flow characteristics. Obstet Gynecol, 86(6), 990-3. Del Rio, M., Martinez, J. M., Figueras, F., Bennasar, M., Olivella, A., Palacio, M., Coll, O., Puerto, B. & Gratacos, E. (2008) Doppler assessment of the aortic isthmus and perinatal outcome in preterm fetuses with severe intrauterine growth restriction. Ultrasound Obstet Gynecol, 31(1), 417. Denison, F. C., Semple, S. I., Stock, S. J., Walker, J., Marshall, I. & Norman, J. E. (2012) Novel use of proton magnetic resonance spectroscopy Page 212 of 253 (1HMRS) to non-invasively assess placental metabolism. PLoS One, 7(8), e42926. Deurloo, K., Spreeuwenberg, M., Rekoert-Hollander, M. & van Vugt, J. (2007a) Reproducibility of 3-dimensional sonographic measurements of fetal and placental volume at gestational ages of 11-18 weeks. J Clin Ultrasound, 35(3), 125-32. Deurloo, K. L., Spreeuwenberg, M. D., Bolte, A. C. & Van Vugt, J. M. (2007b) Color Doppler ultrasound of spiral artery blood flow for prediction of hypertensive disorders and intra uterine growth restriction: a longitudinal study. Prenat Diagn, 27(11), 1011-6. Doctor, B. A., O'Riordan, M. A., Kirchner, H. L., Shah, D. & Hack, M. (2001) Perinatal correlates and neonatal outcomes of small for gestational age infants born at term gestation. Am J Obstet Gynecol, 185(3), 6529. Draper, E. S., Manktelow, B., Field, D. J. & James, D. (1999) Prediction of survival for preterm births by weight and gestational age: retrospective population based study. BMJ, 319(7217), 1093-7. Dubiel, M., Gunnarsson, G. O. & Gudmundsson, S. (2002) Blood redistribution in the fetal brain during chronic hypoxia. Ultrasound Obstet Gynecol, 20(2), 117-21. Dudley, N. J. (2005) A systematic review of the ultrasound estimation of fetal weight. Ultrasound Obstet Gynecol, 25(1), 80-9. Duenhoelter, J. H., Whalley, P. J. & MacDonald, P. C. (1976) An analysis of the utility of plasma immunoreactive estrogen measurements in determining delivery time of gravidas with a fetus considered at high risk. Am J Obstet Gynecol, 125(7), 889-98. Dugoff, L., Hobbins, J. C., Malone, F. D., Vidaver, J., Sullivan, L., Canick, J. A., Lambert-Messerlian, G. M., Porter, T. F., Luthy, D. A., Comstock, C. H., Saade, G., Eddleman, K., Merkatz, I. R., Craigo, S. D., Timor-Tritsch, I. E., Carr, S. R., Wolfe, H. M. & D'Alton, M. E. (2005) Quad screen as a predictor of adverse pregnancy outcome. Obstet Gynecol, 106(2), 260-7. Page 213 of 253 Duric, K., Skrablin, S., Lesin, J., Kalafatic, D., Kuvacic, I. & Suchanek, E. (2003) Second trimester total human chorionic gonadotropin, alphafetoprotein and unconjugated estriol in predicting pregnancy complications other than fetal aneuploidy. Eur J Obstet Gynecol Reprod Biol, 110(1), 12-5. Dutton, P. J., Warrander, L. K., Roberts, S. A., Bernatavicius, G., Byrd, L. M., Gaze, D., Kroll, J., Jones, R. L., Sibley, C. P., Froen, J. F. & Heazell, A. E. (2012) Predictors of poor perinatal outcome following maternal perception of reduced fetal movements--a prospective cohort study. PLoS One, 7(7), e39784. Ebbing, C., Rasmussen, S. & Kiserud, T. (2007) Middle cerebral artery blood flow velocities and pulsatility index and the cerebroplacental pulsatility ratio: longitudinal reference ranges and terms for serial measurements. Ultrasound Obstet Gynecol, 30(3), 287-96. Edelstam, G., Karlsson, C., Westgren, M., Lowbeer, C. & Swahn, M. L. (2007) Human chorionic gonadatropin (hCG) during third trimester pregnancy. Scand J Clin Lab Invest, 67(5), 519-25. Efkarpidis, S., Alexopoulos, E., Kean, L., Liu, D. & Fay, T. (2004) Case-control study of factors associated with intrauterine fetal deaths. MedGenMed, 6(2), 53. Elchalal, U., Ezra, Y., Levi, Y., Bar-Oz, B., Yanai, N., Intrator, O. & Nadjari, M. (2000) Sonographically thick placenta: a marker for increased perinatal risk--a prospective cross-sectional study. Placenta, 21(2-3), 268-72. Ernst, L. M., Minturn, L., Huang, M. H., Curry, E. & Su, E. J. (2013) Gross patterns of umbilical cord coiling: correlations with placental histology and stillbirth. Placenta, 34(7), 583-8. Evers, A. C., Nikkels, P. G., Brouwers, H. A., Boon, J., van Egmond-Linden, A., Hart, C., Snuif, Y. S., Sterken-Hooisma, S., Bruinse, H. W. & Kwee, A. (2011) Substandard care in antepartum term stillbirths: prospective cohort study. Acta Obstet Gynecol Scand, 90(12), 1416-22. Ferrazzi, E., Rigano, S., Padoan, A., Boito, S., Pennati, G. & Galan, H. L. (2011) Uterine artery blood flow volume in pregnant women with an Page 214 of 253 abnormal pulsatility index of the uterine arteries delivering normal or intrauterine growth restricted newborns. Placenta, 32(7), 487-92. Ferreira, A. E., Mauad Filho, F., Abreu, P. S., Mauad, F. M. & Martins, W. P. (2014) The reproducibility of first and second trimester uterine artery pulsatility index by transvaginal and transabdominal ultrasound. Ultrasound Obstet Gynecol. Figueras, F., Fernandez, S., Eixarch, E., Gomez, O., Martinez, J. M., Puerto, B. & Gratacos, E. (2006) Umbilical artery pulsatility index: reliability at different sampling sites. J Perinat Med, 34(5), 409-13. Figueroa-Diesel, H., Hernandez-Andrade, E., Acosta-Rojas, R., Cabero, L. & Gratacos, E. (2007) Doppler changes in the main fetal brain arteries at different stages of hemodynamic adaptation in severe intrauterine growth restriction. Ultrasound Obstet Gynecol, 30(3), 297-302. Fisher, M. L. (1999) Reduced Fetal Movements: a research-based project British Jornal of Midwifery, 7(12), 733-737. Fisteag-Kiprono, L., Neiger, R., Sonek, J. D., Croom, C. S., McKenna, D. S. & Ventolini, G. (2006) Perinatal outcome associated with sonographically detected globular placenta. J Reprod Med, 51(7), 5636. Flenady, V., Koopmans, L., Middleton, P., Froen, J. F., Smith, G. C., Gibbons, K., Coory, M., Gordon, A., Ellwood, D., McIntyre, H. D., Fretts, R. & Ezzati, M. (2011a) Major risk factors for stillbirth in high-income countries: a systematic review and meta-analysis. Lancet, 377(9774), 1331-40. Flenady, V., MacPhail, J., Gardener, G., Chadha, Y., Mahomed, K., Heazell, A., Fretts, R. & Froen, F. (2009) Detection and management of decreased fetal movements in Australia and New Zealand: a survey of obstetric practice. Aust N Z J Obstet Gynaecol, 49(4), 358-63. Flenady, V., Middleton, P., Smith, G. C., Duke, W., Erwich, J. J., Khong, T. Y., Neilson, J., Ezzati, M., Koopmans, L., Ellwood, D., Fretts, R. & Froen, J. F. (2011b) Stillbirths: the way forward in high-income countries. Lancet, 377(9778), 1703-17. Fratelli, N., Valcamonico, A., Prefumo, F., Pagani, G., Guarneri, T. & Frusca, T. (2013) Effects of antenatal recognition and follow-up on perinatal Page 215 of 253 outcomes in small-for-gestational age infants delivered after 36 weeks. Acta Obstet Gynecol Scand, 92(2), 223-9. Fries, N., Althuser, M., Fontanges, M., Talmant, C., Jouk, P. S., Tindel, M. & Duyme, M. (2007) Quality control of an image-scoring method for nuchal translucency ultrasonography. Am J Obstet Gynecol, 196(3), 272 e1-5. Froen, J. F. (2004) A kick from within--fetal movement counting and the cancelled progress in antenatal care. J Perinat Med, 32(1), 13-24. Froen, J. F., Heazell, A. E., Tveit, J. V., Saastad, E., Fretts, R. C. & Flenady, V. (2008a) Fetal movement assessment. Semin Perinatol, 32(4), 243-6. Froen, J. F., Tveit, J. V., Saastad, E., Bordahl, P. E., Stray-Pedersen, B., Heazell, A. E., Flenady, V. & Fretts, R. C. (2008b) Management of decreased fetal movements. Semin Perinatol, 32(4), 307-11. Fu, G., Brkic, J., Hayder, H. & Peng, C. (2013) MicroRNAs in Human Placental Development and Pregnancy Complications. Int J Mol Sci, 14(3), 5519-44. Furuhashi, N., Tachibana, Y., Shinkawa, O., Hiruta, M., Takahashi, T. & Tanaka, M. (1984a) Simultaneous and serial measurement of serum levels of human placental lactogen, beta-human chorionic gonadotropin and unconjugated estriol levels in pregnant women. Tohoku J Exp Med, 144(2), 211-5. Furuhashi, N., Tachibana, Y., Shinkawa, O., Takahashi, T., Tanaka, M., Hiruta, M. & Suzuki, M. (1984b) Retrograde time-scale analysis of human placental lactogen, beta human chorionic gonadotropin, and unconjugated estriol levels in human maternal serum from the onset of spontaneous labor. Gynecol Obstet Invest, 18(5), 264-8. Gagnon, A., Wilson, R. D., Audibert, F., Allen, V. M., Blight, C., Brock, J. A., Desilets, V. A., Johnson, J. A., Langlois, S., Summers, A. & Wyatt, P. (2008) Obstetrical complications associated with abnormal maternal serum markers analytes. J Obstet Gynaecol Can, 30(10), 918-49. Gardosi, J., Chang, A., Kalyan, B., Sahota, D. & Symonds, E. M. (1992) Customised antenatal growth charts. Lancet, 339(8788), 283-7. Page 216 of 253 Gardosi, J., Clausson, B. & Francis, A. (2009) The value of customised centiles in assessing perinatal mortality risk associated with parity and maternal size. BJOG, 116(10), 1356-63. Gardosi, J. & Francis, A. (2009) Adverse pregnancy outcome and association with small for gestational age birthweight by customized and population-based percentiles. Am J Obstet Gynecol, 201(1), 28 e1-8. Gardosi, J., Kady, S. M., McGeown, P., Francis, A. & Tonks, A. (2005) Classification of stillbirth by relevant condition at death (ReCoDe): population based cohort study. BMJ, 331(7525), 1113-7. Garrod, A., Batra, G., Ptacek, I. & Heazell, A. E. (2013) Duration and method of tissue storage alters placental morphology - implications for clinical and research practice. Placenta, 34(11), 1116-9. Ghi, T., Contro, E., Youssef, A., Giorgetta, F., Farina, A., Pilu, G. & Pelusi, G. (2010) Persistence of increased uterine artery resistance in the third trimester and pregnancy outcome. Ultrasound Obstet Gynecol, 36(5), 577-81. Girard, S., Heazell, A. E., Derricott, H., Allan, S. M., Sibley, C. P., Abrahams, V. M. & Jones, R. L. (2014) Circulating cytokines and alarmins associated with placental inflammation in high-risk pregnancies. Am J Reprod Immunol, 72(4), 422-34. Glazier, J. D., Cetin, I., Perugino, G., Ronzoni, S., Grey, A. M., Mahendran, D., Marconi, A. M., Pardi, G. & Sibley, C. P. (1997) Association between the activity of the system A amino acid transporter in the microvillous plasma membrane of the human placenta and severity of fetal compromise in intrauterine growth restriction. Pediatr Res, 42(4), 514-9. Gomez, O., Figueras, F., Fernandez, S., Bennasar, M., Martinez, J. M., Puerto, B. & Gratacos, E. (2008) Reference ranges for uterine artery mean pulsatility index at 11-41 weeks of gestation. Ultrasound Obstet Gynecol, 32(2), 128-32. Gordon, Y. B., Grudzinskas, J. G., Lewis, J. D., Jeffrey, D. & Letchworth, A. T. (1977) Circulating levels of pregnancy-specific beta1-glycoprotein and human placental lactogen in the third trimester of pregnancy: Page 217 of 253 their relationship to parity, birth weight, and placental weight. Br J Obstet Gynaecol, 84(9), 642-7. Gowland, P. (2005) Placental MRI. Semin Fetal Neonatal Med, 10(5), 485-90. Grannum, P. A., Berkowitz, R. L. & Hobbins, J. C. (1979) The ultrasonic changes in the maturing placenta and their relation to fetal pulmonic maturity. Am J Obstet Gynecol, 133(8), 915-22. Grant, A., Elbourne, D., Valentin, L. & Alexander, S. (1989) Routine formal fetal movement counting and risk of antepartum late death in normally formed singletons. Lancet, 2(8659), 345-9. Gregory, E. C. W. M., M. F.; Martin, J. A. (2014) Trends in fetal and perinatal mortality in the United States, 2006 - 2012. In: Statistics, N. C. f. H. (ed.) NCHS data brief. Hyattsville, MD: National Centre for Health Statistics. Griffin, M. Seed, P. T., Webster, L., Myers, J., Mackillop, L., Simpson, N., Anumba, D., Khalil, A., Debrow, M., Sau, A., Hinshaw, K., von Dazelden, P., Benton, S., Girling, J., Redman, C. W. G., Chappell, L. C. & Shennan, A. H. (2015) Diagnostic accuracy of placental growth factor and ultrasound parameters to predict the small-for-gestational-age infant in women presenting with reduced symphysis-fundal height measurement. Ultrasound in Obstetrics and Gynaecology, In press. Grivell, R. M., Alfirevic, Z., Gyte, G. M. & Devane, D. (2010) Antenatal cardiotocography for fetal assessment. Cochrane Database Syst Rev, (1), CD007863. Grivell, R. M., Alfirevic, Z., Gyte, G. M. & Devane, D. (2012a) Antenatal cardiotocography for fetal assessment. Cochrane Database Syst Rev, 12, CD007863. Grivell, R. M., Wong, L. & Bhatia, V. (2012b) Regimens of fetal surveillance for impaired fetal growth. Cochrane Database Syst Rev, 6, CD007113. Gu, Y., Lewis, D. F. & Wang, Y. (2008) Placental productions and expressions of soluble endoglin, soluble fms-like tyrosine kinase receptor-1, and placental growth factor in normal and preeclamptic pregnancies. J Clin Endocrinol Metab, 93(1), 260-6. Page 218 of 253 Haberman, S., Bracero, L. A. & Byrne, D. W. (2004) Spectral Doppler index mapping of the umbilicoplacental circulation and pregnancy outcome. Gynecol Obstet Invest, 58(1), 1-7. Hadlock, F. P., Harrist, R. B., Sharman, R. S., Deter, R. L. & Park, S. K. (1985) Estimation of fetal weight with the use of head, body, and femur measurements--a prospective study. Am J Obstet Gynecol, 151(3), 333-7. Haeussner, E., Schmitz, C., von Koch, F. & Frank, H. G. (2013) Birth weight correlates with size but not shape of the normal human placenta. Placenta, 34(7), 574-82. Hafner, E., Metzenbauer, M., Stumpflen, I., Waldhor, T. & Philipp, K. (2010) First trimester placental and myometrial blood perfusion measured by 3D power Doppler in normal and unfavourable outcome pregnancies. Placenta, 31(9), 756-63. Hafner, E., Schuchter, K., van Leeuwen, M., Metzenbauer, M., Dillinger-Paller, B. & Philipp, K. (2001) Three-dimensional sonographic volumetry of the placenta and the fetus between weeks 15 and 17 of gestation. Ultrasound Obstet Gynecol, 18(2), 116-20. Handwerger, S. & Freemark, M. (2000) The roles of placental growth hormone and placental lactogen in the regulation of human fetal growth and development. J Pediatr Endocrinol Metab, 13(4), 343-56. Hannah, M. E., Hannah, W. J., Hewson, S. A., Hodnett, E. D., Saigal, S. & Willan, A. R. (2000) Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Term Breech Trial Collaborative Group. Lancet, 356(9239), 1375-83. Hannah, M. E., Ohlsson, A., Farine, D., Hewson, S. A., Hodnett, E. D., Myhr, T. L., Wang, E. E., Weston, J. A. & Willan, A. R. (1996) Induction of labor compared with expectant management for prelabor rupture of the membranes at term. TERMPROM Study Group. N Engl J Med, 334(16), 1005-10. Hargreaves, K., Cameron, M., Edwards, H., Gray, R. & Deane, K. (2011) Is the use of symphysis-fundal height measurement and ultrasound Page 219 of 253 examination effective in detecting small or large fetuses? J Obstet Gynaecol, 31(5), 380-3. Haws, R. A., Yakoob, M. Y., Soomro, T., Menezes, E. V., Darmstadt, G. L. & Bhutta, Z. A. (2009) Reducing stillbirths: screening and monitoring during pregnancy and labour. BMC Pregnancy Childbirth, 9 Suppl 1, S5. Hayward, C. E., Greenwood, S. L., Sibley, C. P., Baker, P. N. & Jones, R. L. (2011) Effect of young maternal age and skeletal growth on placental growth and development. Placenta, 32(12), 990-8. Hayward, C. E., Higgins, L., Cowley, E. J., Greenwood, S. L., Mills, T. A., Sibley, C. P. & Wareing, M. (2013) Chorionic plate arterial function is altered in maternal obesity. Placenta, 34(3), 281-7. Heazell, A. E., Bernatavicius, G., Roberts, S. A., Garrod, A., Whitworth, M. K., Johnstone, E. D., Gillham, J. C. & Lavender, T. (2013) A randomised controlled trial comparing standard or intensive management of reduced fetal movements after 36 weeks gestation--a feasibility study. BMC Pregnancy Childbirth, 13, 95. Heazell, A. E., Bernatavicius, G., Warrander, L., Brown, M. C. & Dunn, W. B. (2012) A metabolomic approach identifies differences in maternal serum in third trimester pregnancies that end in poor perinatal outcome. Reprod Sci, 19(8), 863-75. Heazell, A. E. & Froen, J. F. (2008) Methods of fetal movement counting and the detection of fetal compromise. J Obstet Gynaecol, 28(2), 147-54. Heazell, A. E., Green, M., Wright, C., Flenady, V. & Froen, J. F. (2008) Midwives' and obstetricians' knowledge and management of women presenting with decreased fetal movements. Acta Obstet Gynecol Scand, 87(3), 331-9. Heazell, A. E. & Martindale, E. A. (2009) Can post-mortem examination of the placenta help determine the cause of stillbirth? J Obstet Gynaecol, 29(3), 225-8. Heazell, A. E., Sumathi, G. M. & Bhatti, N. R. (2005) What investigation is appropriate following maternal perception movements? J Obstet Gynaecol, 25(7), 648-50. Page 220 of 253 of reduced fetal Heazell, A. E. P., Worton, S. A., Higgins, L. E., Ingram, E., Johnstone, E. D., Jones, R. L., Sibley, C. P. (2015) Recognition of Placental Failure is Key to Saving Babies' Lives. Placenta, In Press. Hernandez-Andrade, E., Figueroa-Diesel, H., Jansson, T., Rangel-Nava, H. & Gratacos, E. (2008) Changes in regional fetal cerebral blood flow perfusion in relation to hemodynamic deterioration in severely growth-restricted fetuses. Ultrasound Obstet Gynecol, 32(1), 71-6. Hibbard, J. U., Wilkins, I., Sun, L., Gregory, K., Haberman, S., Hoffman, M., Kominiarek, M. A., Reddy, U., Bailit, J., Branch, D. W., Burkman, R., Gonzalez Quintero, V. H., Hatjis, C. G., Landy, H., Ramirez, M., VanVeldhuisen, P., Troendle, J. & Zhang, J. (2010) Respiratory morbidity in late preterm births. JAMA, 304(4), 419-25. Hijazi, Z. R. & East, C. E. (2009) Factors affecting maternal perception of fetal movement. Obstet Gynecol Surv, 64(7), 489-97; quiz 499. Hofmeyr, G. J. & Novikova, N. (2012) Management of reported decreased fetal movements for improving pregnancy outcomes. Cochrane Database Syst Rev, 4, CD009148. Hollis, B., Mavrides, E., Campbell, S., Tekay, A. & Thilaganathan, B. (2001) Reproducibility and repeatability of transabdominal uterine artery Doppler velocimetry between 10 and 14 weeks of gestation. Ultrasound Obstet Gynecol, 18(6), 593-7. Hoogland, H. J., de Haan, J. & Martin, C. B., Jr. (1980) Placental size during early pregnancy and fetal outcome: a preliminary report of a sequential ultrasonographic study. Am J Obstet Gynecol, 138(4), 4413. Hsieh, F. J., Kuo, P. L., Ko, T. M., Chang, F. M. & Chen, H. Y. (1991) Doppler velocimetry of intraplacental fetal arteries. Obstet Gynecol, 77(3), 478-82. Huen, I., Morris, D. M., Wright, C., Parker, G. J., Sibley, C. P., Johnstone, E. D. & Naish, J. H. (2013) R1 and R2 * changes in the human placenta in response to maternal oxygen challenge. Magn Reson Med, 70(5), 1427-33. Page 221 of 253 Huppertz, B., Sammar, M., Chefetz, I., Neumaier-Wagner, P., Bartz, C. & Meiri, H. (2008) Longitudinal determination of serum placental protein 13 during development of preeclampsia. Fetal Diagn Ther, 24(3), 230-6. Hussa, R. O. (1980) Biosynthesis of human chorionic gonadotropin. Endocr Rev, 1(3), 268-94. Iliodromiti, S., Mackay, D. F., Smith, G. C., Pell, J. P. & Nelson, S. M. (2014) Apgar score and the risk of cause-specific infant mortality: a population-based cohort study. Lancet. Ishimoto, H. & Jaffe, R. B. (2011) Development and function of the human fetal adrenal cortex: a key component in the feto-placental unit. Endocr Rev, 32(3), 317-55. Isouard, G. (1979) Value of total serum oestriol and human placental lactogen in the assessment of fetal-placental function. Aust N Z J Obstet Gynaecol, 19(2), 69-76. Jackson, J. R., Kleeman, S., Doerzbacher, M. & Lambers, D. S. (2003) The effect of glucocorticosteroid administration on fetal movements and biophysical profile scores in normal pregnancies. J Matern Fetal Neonatal Med, 13(1), 50-3. Jackson, M. R., Mayhew, T. M. & Boyd, P. A. (1992) Quantitative description of the elaboration and maturation of villi from 10 weeks of gestation to term. Placenta, 13(4), 357-70. Jaffe, R. (1996) Multigate spectral Doppler velocimetry for fast assessment of intraplacental fetal circulation: a preliminary study. J Ultrasound Med, 15(4), 309-12. Jaffe, R. & Woods, J. R. (1996) Doppler velocimetry of intraplacental fetal vessels in the second trimester: improving the prediction of pregnancy complications in high-risk patients. Ultrasound Obstet Gynecol, 8(4), 262-6. Jahn, A., Razum, O. & Berle, P. (1998) Routine screening for intrauterine growth retardation in Germany: low sensitivity and questionable benefit for diagnosed cases. Acta Obstet Gynecol Scand, 77(6), 643-8. James, D. K., Telfer, F. M., Keating, N. A., Blair, M. E., Wilcox, M. A. & Chilvers, C. (2000) Reduced fetal movements and maternal medication - new Page 222 of 253 pregnancy risk factors for neurodevelopmental disability in childhood. J Obstet Gynaecol, 20(3), 226-34. Jansson, T., Scholtbach, V. & Powell, T. L. (1998) Placental transport of leucine and lysine is reduced in intrauterine growth restriction. Pediatr Res, 44(4), 532-7. Jansson, T., Wennergren, M. & Illsley, N. P. (1993) Glucose transporter protein expression in human placenta throughout gestation and in intrauterine growth retardation. J Clin Endocrinol Metab, 77(6), 1554-62. Jansson, T., Ylven, K., Wennergren, M. & Powell, T. L. (2002) Glucose transport and system A activity in syncytiotrophoblast microvillous and basal plasma membranes in intrauterine growth restriction. Placenta, 23(5), 392-9. Jauniaux, E., Ramsay, B. & Campbell, S. (1994) Ultrasonographic investigation of placental morphologic characteristics and size during the second trimester of pregnancy. Am J Obstet Gynecol, 170(1 Pt 1), 130-7. Jessop, F. A., Lees, C. C., Pathak, S., Hook, C. E. & Sebire, N. J. (2014) Umbilical cord coiling: clinical outcomes in an unselected population and systematic review. Virchows Arch, 464(1), 105-12. Joern, H., Klein, A., Kuehlwein, H. & Rath, W. (1997) Critical comparison of indices and threshold values for assessing placenta performance using Doppler ultrasound. Ultrasound Med Biol, 23(8), 1179-83. Johansson, M., Glazier, J. D., Sibley, C. P., Jansson, T. & Powell, T. L. (2002) Activity and protein expression of the Na+/H+ exchanger is reduced in syncytiotrophoblast microvillous plasma membranes isolated from preterm intrauterine growth restriction pregnancies. J Clin Endocrinol Metab, 87(12), 5686-94. Johansson, M., Karlsson, L., Wennergren, M., Jansson, T. & Powell, T. L. (2003) Activity and protein expression of Na+/K+ ATPase are reduced in microvillous syncytiotrophoblast plasma membranes isolated from pregnancies complicated by intrauterine growth restriction. J Clin Endocrinol Metab, 88(6), 2831-7. Page 223 of 253 Jones, N. W., Raine-Fenning, N. J., Mousa, H. A., Bradley, E. & Bugg, G. J. (2011) Evaluating the intra- and interobserver reliability of threedimensional ultrasound and power Doppler angiography (3D-PDA) for assessment of placental volume and vascularity in the second trimester of pregnancy. Ultrasound Med Biol, 37(3), 376-85. Junaid, T. O., Brownbill, P., Chalmers, N., Johnstone, E. D. & Aplin, J. D. (2014) Fetoplacental vascular alterations associated with fetal growth restriction. Placenta, 35(10), 808-15. Jurkovic, D., Jauniaux, E., Kurjak, A., Hustin, J., Campbell, S. & Nicolaides, K. H. (1991) Transvaginal color Doppler assessment of the uteroplacental circulation in early pregnancy. Obstet Gynecol, 77(3), 365-9. Kakogawa, J., Sumimoto, K., Ho, E. & Kanayama, N. (2005) Transabdominal measurement of oxygenation of the placenta by near-infrared spectroscopy. Semin Thromb Hemost, 31(3), 297-301. Kaplan, S. L. & Grumbach, M. M. (1965) Serum chorionic "growth hormoneprolactin" and serum pituitary growth hormone in mother and fetus at term. J Clin Endocrinol Metab, 25(10), 1370-4. Karthikeyan, T., Subramaniam, R. K., Johnson, W. & Prabhu, K. (2012) Placental thickness & its correlation to gestational age & foetal growth parameters- a cross sectional ultrasonographic study. J Clin Diagn Res, 6(10), 1732-5. Kaufmann, P., Mayhew, T. M. & Charnock-Jones, D. S. (2004) Aspects of human fetoplacental vasculogenesis and angiogenesis. II. Changes during normal pregnancy. Placenta, 25(2-3), 114-26. Kendall, R. L., Wang, G. & Thomas, K. A. (1996) Identification of a natural soluble form of the vascular endothelial growth factor receptor, FLT1, and its heterodimerization with KDR. Biochem Biophys Res Commun, 226(2), 324-8. Kennelly, M. M., Farah, N., Hogan, J., Reilly, A., Turner, M. J. & Stuart, B. (2012) Longitudinal study of aortic isthmus Doppler in appropriately grown and small-for-gestational-age fetuses with normal and abnormal umbilical artery Doppler. Ultrasound Obstet Gynecol, 39(4), 414-20. Page 224 of 253 Khan, K. S. & Chien, P. F. (2001) Evaluation of a clinical test. I: assessment of reliability. BJOG, 108(6), 562-7. Khare, M., Paul, S. & Konje, J. C. (2006) Variation in Doppler indices along the length of the cord from the intraabdominal to the placental insertion. Acta Obstet Gynecol Scand, 85(8), 922-8. Kidron, D., Bernheim, J. & Aviram, R. (2009) Placental findings contributing to fetal death, a study of 120 stillbirths between 23 and 40 weeks gestation. Placenta, 30(8), 700-4. Kingdom, J., Huppertz, B., Seaward, G. & Kaufmann, P. (2000) Development of the placental villous tree and its consequences for fetal growth. Eur J Obstet Gynecol Reprod Biol, 92(1), 35-43. Kingdom, J. C., Burrell, S. J. & Kaufmann, P. (1997) Pathology and clinical implications of abnormal umbilical artery Doppler waveforms. Ultrasound Obstet Gynecol, 9(4), 271-86. Kingdom, J. C. & Kaufmann, P. (1997) Oxygen and placental villous development: origins of fetal hypoxia. Placenta, 18(8), 613-21; discussion 623-6. Kirkinen, P., Kurmanavichius, J., Huch, A. & Huch, R. (1994) Blood flow velocities in human intraplacental arteries. Acta Obstet Gynecol Scand, 73(3), 220-4. Konje, J. C., Kaufmann, P., Bell, S. C. & Taylor, D. J. (2001) A longitudinal study of quantitative uterine blood flow with the use of color power angiography in appropriate for gestational age pregnancies. Am J Obstet Gynecol, 185(3), 608-13. Koopmans, C. M., Bijlenga, D., Groen, H., Vijgen, S. M., Aarnoudse, J. G., Bekedam, D. J., van den Berg, P. P., de Boer, K., Burggraaff, J. M., Bloemenkamp, K. W., Drogtrop, A. P., Franx, A., de Groot, C. J., Huisjes, A. J., Kwee, A., van Loon, A. J., Lub, A., Papatsonis, D. N., van der Post, J. A., Roumen, F. J., Scheepers, H. C., Willekes, C., Mol, B. W. & van Pampus, M. G. (2009) Induction of labour versus expectant monitoring for gestational hypertension or mild pre-eclampsia after 36 weeks gestation (HYPITAT): a multicentre, open-label randomised controlled trial. Lancet, 374(9694), 979-88. Page 225 of 253 Korst, L. M., Phelan, J. P., Ahn, M. O. & Martin, G. I. (1996) Nucleated red blood cells: an update on the marker for fetal asphyxia. Am J Obstet Gynecol, 175(4 Pt 1), 843-6. Korszun, P., Dubiel, M., Kudla, M. & Gudmundsson, S. (2002) Doppler velocimetry for predicting outcome of pregnancies with decreased fetal movements. Acta Obstet Gynecol Scand, 81(10), 926-30. Krohn, K. A., Link, J. M. & Mason, R. P. (2008) Molecular imaging of hypoxia. J Nucl Med, 49 Suppl 2, 129S-48S. Krol, J., Loedige, I. & Filipowicz, W. (2010) The widespread regulation of microRNA biogenesis, function and decay. Nat Rev Genet, 11(9), 597610. Kudla, M. J. & Alcazar, J. L. (2012) Spatiotemporal image correlation with spherical sampling and high-definition flow: new 4-dimensional method for assessment of tissue vascularization changes during the cardiac cycle: reproducibility analysis. J Ultrasound Med, 31(1), 7380. Kurdi, W., Fayyad, A., Thakur, V. & Harrington, K. (2004) Delayed normalization of uterine artery Doppler waveforms is not a benign phenomenon. Eur J Obstet Gynecol Reprod Biol, 117(1), 20-3. Kurjak, A., Dudenhausen, J. W., Hafner, T., Kupesic, S., Latin, V. & Kos, M. (1997) Intervillous circulation in all three trimesters of normal pregnancy assessed by color Doppler. J Perinat Med, 25(4), 373-80. Kurjak, A. & Kupesic, S. (1997) Doppler assessment of the intervillous blood flow in normal and abnormal early pregnancy. Obstet Gynecol, 89(2), 252-6. Kurmanavicius, J., Burkhardt, T., Wisser, J. & Huch, R. (2004) Ultrasonographic fetal weight estimation: accuracy of formulas and accuracy of examiners by birth weight from 500 to 5000 g. J Perinat Med, 32(2), 155-61. Kurmanavicius, J., Florio, I., Wisser, J., Hebisch, G., Zimmermann, R., Muller, R., Huch, R. & Huch, A. (1997) Reference resistance indices of the umbilical, fetal middle cerebral and uterine arteries at 24-42 weeks of gestation. Ultrasound Obstet Gynecol, 10(2), 112-20. Page 226 of 253 Lacin, S., Demir, N., Koyuncu, F. & Goktay, Y. (1996) Value of intraplacental villous artery Doppler measurements in severe preeclampsia. J Postgrad Med, 42(4), 101-4. Laine, K. S., A.; Murzakanova, G.; Gustafsson, J.; Barbero, M. L.; Ye, S.; Kliman, H. J. (2012) Second trimester estimated placental volume (EPV) screening is predictive of SGA and LGA newborns and adverse pregnancy outcomes. Ultrasound in Obstetrics & Gynecology, 40(Suppl. 1), 55-170. Landmann, E., Reiss, I., Misselwitz, B. & Gortner, L. (2006) Ponderal index for discrimination between symmetric and asymmetric growth restriction: percentiles for neonates from 30 weeks to 43 weeks of gestation. J Matern Fetal Neonatal Med, 19(3), 157-60. Lang, U., Baker, R. S., Khoury, J. & Clark, K. E. (2002) Fetal umbilical vascular response to chronic reductions in uteroplacental blood flow in late term sheep. Am J Obstet Gynecol, 187(1), 178-86. Lao, M. R., Calhoun, B. C., Bracero, L. A., Wang, Y., Seybold, D. J., Broce, M. & Hatjis, C. G. (2009) The ability of the quadruple test to predict adverse perinatal outcomes in a high-risk obstetric population. J Med Screen, 16(2), 55-9. Law, L. W., Leung, T. Y., Sahota, D. S., Chan, L. W., Fung, T. Y. & Lau, T. K. (2009) Which ultrasound or biochemical markers are independent predictors of small-for-gestational age? Ultrasound Obstet Gynecol, 34(3), 283-7. Lawn, J. E., Blencowe, H., Pattinson, R., Cousens, S., Kumar, R., Ibiebele, I., Gardosi, J., Day, L. T. & Stanton, C. (2011) Stillbirths: Where? When? Why? How to make the data count? Lancet, 377(9775), 1448-63. Leader, L. R. & Baillie, P. (1980) Comparison of fetal movements and human placental lactogen as predictors of fetal outcome. A prospective study. S Afr Med J, 58(15), 609-10. Leader, L. R., Baillie, P. & Van Schalkwyk, D. J. (1981) Fetal movements and fetal outcome: a prospective study. Obstet Gynecol, 57(4), 431-6. Lee, W., Balasubramaniam, M., Deter, R. L., Yeo, L., Hassan, S. S., Gotsch, F., Kusanovic, J. P., Goncalves, L. F. & Romero, R. (2009) New fetal weight Page 227 of 253 estimation models using fractional limb volume. Ultrasound Obstet Gynecol, 34(5), 556-65. Lee, W., Riggs, T., Koo, W., Deter, R. L., Yeo, L. & Romero, R. (2012) The relationship of newborn adiposity to fetal growth outcome based on birth weight or the modified neonatal growth assessment score. J Matern Fetal Neonatal Med, 25(10), 1933-40. Lees, C., Marlow, N., Arabin, B., Bilardo, C. M., Brezinka, C., Derks, J. B., Duvekot, J., Frusca, T., Diemert, A., Ferrazzi, E., Ganzevoort, W., Hecher, K., Martinelli, P., Ostermayer, E., Papageorghiou, A. T., Schlembach, D., Schneider, K. T., Thilaganathan, B., Todros, T., van Wassenaer-Leemhuis, A., Valcamonico, A., Visser, G. H. & Wolf, H. (2013) Perinatal morbidity and mortality in early-onset fetal growth restriction: cohort outcomes of the trial of randomized umbilical and fetal flow in Europe (TRUFFLE). Ultrasound Obstet Gynecol, 42(4), 400-8. Letchworth, A. T., Slattery, M. & Dennis, K. J. (1978) Clinical application of human-placental-lactogen values in late pregnancy. Lancet, 1(8071), 955-7. Levine, R. J., Maynard, S. E., Qian, C., Lim, K. H., England, L. J., Yu, K. F., Schisterman, E. F., Thadhani, R., Sachs, B. P., Epstein, F. H., Sibai, B. M., Sukhatme, V. P. & Karumanchi, S. A. (2004) Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med, 350(7), 672-83. Lin, S., Shimizu, I., Suehara, N., Nakayama, M. & Aono, T. (1995) Uterine artery Doppler velocimetry in relation to trophoblast migration into the myometrium of the placental bed. Obstet Gynecol, 85(5 Pt 1), 7605. Lindqvist, P. G. & Molin, J. (2005) Does antenatal identification of small-forgestational age fetuses significantly improve their outcome? Ultrasound Obstet Gynecol, 25(3), 258-64. Linduska, N., Dekan, S., Messerschmidt, A., Kasprian, G., Brugger, P. C., Chalubinski, K., Weber, M. & Prayer, D. (2009) Placental pathologies in fetal MRI with pathohistological correlation. Placenta, 30(6), 5559. Page 228 of 253 MacDorman, M. F. (2012) Fetal and Perinatal Mortality, United States, 2006 [Online]. Hyattsville, MD: National Center for Health Statistics. Available: http://www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_08.pdf [Accessed 23/10/2014]. MacKay, D. F., Smith, G. C., Dobbie, R. & Pell, J. P. (2010) Gestational age at delivery and special educational need: retrospective cohort study of 407,503 schoolchildren. PLoS Med, 7(6), e1000289. Madazli, R., Somunkiran, A., Calay, Z., Ilvan, S. & Aksu, M. F. (2003) Histomorphology of the placenta and the placental bed of growth restricted foetuses and correlation with the Doppler velocimetries of the uterine and umbilical arteries. Placenta, 24(5), 510-6. Magann, E. F., Sanderson, M., Martin, J. N. & Chauhan, S. (2000) The amniotic fluid index, single deepest pocket, and two-diameter pocket in normal human pregnancy. Am J Obstet Gynecol, 182(6), 1581-8. Mahendru, A., Gajjar, K. & Bashir, T. (2007) Idiopathic chronic fetomaternal haemorrhage presenting with reduced fetal movements. J Obstet Gynaecol, 27(8), 848-9. Malin, G. L., Morris, R. K. & Khan, K. S. (2010) Strength of association between umbilical cord pH and perinatal and long term outcomes: systematic review and meta-analysis. BMJ, 340, c1471. Mangesi, L. & Hofmeyr, G. J. (2007) Fetal movement counting for assessment of fetal wellbeing. Cochrane Database Syst Rev, (1), CD004909. Mannik, J., Vaas, P., Rull, K., Teesalu, P., Rebane, T. & Laan, M. (2010) Differential expression profile of growth hormone/chorionic somatomammotropin genes in placenta of small- and large-forgestational-age newborns. J Clin Endocrinol Metab, 95(5), 2433-42. Maroni, E., Youssef, A., Arcangeli, T., Nanni, M., De Musso, F., Contro, E., Kuleva, M., Bellussi, F., Pilu, G., Rizzo, N. & Ghi, T. (2011) Increased uterine artery pulsatility index at 34 weeks and outcome of pregnancy. Ultrasound Obstet Gynecol, 38(4), 395-9. Martinez, J. M., Comas, C., Ojuel, J., Borrell, A., Puerto, B., Sentis, J. & Fortuny, A. (1995) The influence of the site of Doppler recording on umbilical Page 229 of 253 artery pulsatility index during the first trimester. Ultrasound Obstet Gynecol, 5(5), 325-7. Martins, W. P., Lima, J. C., Welsh, A. W., Araujo Junior, E., Miyague, A. H., Filho, F. M. & Raine-Fenning, N. J. (2012) Three-dimensional Doppler evaluation of single spherical samples from the placenta: intra- and interobserver reliability. Ultrasound Obstet Gynecol, 40(2), 200-6. Mathai, B. M., Singla, S. C., Nittala, P. P., Chakravarti, R. J. & Toppo, J. N. (2013) Placental thickness: its correlation with ultrasonographic gestational age in normal and intrauterine growth-retarded pregnancies in the late second and third trimester. J Obstet Gynaecol India, 63(4), 230-3. Matijevic, R., Meekins, J. W., Walkinshaw, S. A., Neilson, J. P. & McFadyen, I. R. (1995) Spiral artery blood flow in the central and peripheral areas of the placental bed in the second trimester. Obstet Gynecol, 86(2), 28992. Mattioli, K. P., Sanderson, M. & Chauhan, S. P. (2010) Inadequate identification of small-for-gestational-age fetuses at an urban teaching hospital. Int J Gynaecol Obstet, 109(2), 140-3. Maulik, D. (1997) Doppler Velocimetry for Fetal Surveillance: Adverse Perinatal Outcome and Fetal Hypoxia. In: Zalud, I. M., D. (ed.) Doppler ultrasound in obstetrics and gynaecology. New York: Springer-Verlag. Mayhew, T. M. (2009) A stereological perspective on placental morphology in normal and complicated pregnancies. J Anat, 215(1), 77-90. Mayhew, T. M., Ohadike, C., Baker, P. N., Crocker, I. P., Mitchell, C. & Ong, S. S. (2003) Stereological investigation of placental morphology in pregnancies complicated by pre-eclampsia with and without intrauterine growth restriction. Placenta, 24(2-3), 219-26. Maynard, S. E., Min, J. Y., Merchan, J., Lim, K. H., Li, J., Mondal, S., Libermann, T. A., Morgan, J. P., Sellke, F. W., Stillman, I. E., Epstein, F. H., Sukhatme, V. P. & Karumanchi, S. A. (2003) Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest, 111(5), 649-58. Page 230 of 253 McCowan, L. M., Harding, J. E. & Stewart, A. W. (2005) Customized birthweight centiles predict SGA pregnancies with perinatal morbidity. BJOG, 112(8), 1026-33. McCowan, L. M., Thompson, J. M., Taylor, R. S., North, R. A., Poston, L., Baker, P. N., Myers, J., Roberts, C. T., Dekker, G. A., Simpson, N. A., Walker, J. J. & Kenny, L. C. (2013) Clinical prediction in early pregnancy of infants small for gestational age by customised birthweight centiles: findings from a healthy nulliparous cohort. PLoS One, 8(8), e70917. McGinty, P., Farah, N., Dwyer, V. O., Hogan, J., Reilly, A., Turner, M. J., Stuart, B. & Kennelly, M. M. (2012) Ultrasound assessment of placental function: the effectiveness of placental biometry in a low-risk population as a predictor of a small for gestational age neonate. Prenat Diagn, 32(7), 620-6. McKenna, D., Tharmaratnam, S., Mahsud, S. & Dornan, J. (2005) Ultrasonic evidence of placental calcification at 36 weeks gestation: maternal and fetal outcomes. Acta Obstet Gynecol Scand, 84(1), 7-10. Melamed, N., Yogev, Y., Meizner, I., Mashiach, R., Bardin, R. & Ben-Haroush, A. (2009) Sonographic fetal weight estimation: which model should be used? J Ultrasound Med, 28(5), 617-29. Messerschmidt, A., Baschat, A., Linduska, N., Kasprian, G., Brugger, P. C., Bauer, A., Weber, M. & Prayer, D. (2011) Magnetic resonance imaging of the placenta identifies placental vascular abnormalities independently of Doppler ultrasound. Ultrasound Obstet Gynecol, 37(6), 717-22. Midgley, A. R., Jr. & Pierce, G. B., Jr. (1962) Immunohistochemical localization of human chorionic gonadotropin. J Exp Med, 115, 289-94. Mifsud, W. & Sebire, N. J. (2014) Placental pathology in early-onset and lateonset fetal growth restriction. Fetal Diagn Ther, 36(2), 117-28. Milani, H. J., de Sa Barreto, E. Q., Araujo Junior, E., Haratz, K. K., Rolo, L. C., Nardozza, L. M. & Moron, A. F. (2012) Assessment of cerebral circulation in normal fetuses by three-dimensional power Doppler ultrasonography. Eur J Radiol, 81(3), 514-21. Page 231 of 253 Milligan, N., Rowden, M., Wright, E., Melamed, N., Lee, Y. M., Windrim, R. C. & Kingdom, J. C. (2014) Two-dimensional sonographic assessment of maximum placental length and thickness in the second trimester: a reproducibility study. J Matern Fetal Neonatal Med, 1-7. Mills, T. A., Greenwood, S. L., Johnstone, E. D. & Wareing, M. (2011) Mechanical and receptor mediated responses of placental chorionic plate arteries are altered in fetal growth restriction. Reproductive Sciences, 19(3 (Supplement)), F-291. Mills, T. A., Wareing, M., Bugg, G. J., Greenwood, S. L. & Baker, P. N. (2005) Chorionic plate artery function and Doppler indices in normal pregnancy and intrauterine growth restriction. Eur J Clin Invest, 35(12), 758-64. Mistry, H., Heazell, A. E., Vincent, O. & Roberts, T. (2013) A structured review and exploration of the healthcare costs associated with stillbirth and a subsequent pregnancy in England and Wales. BMC Pregnancy Childbirth, 13, 236. Mitra, S. C., Seshan, S. V. & Riachi, L. E. (2000) Placental vessel morphometry in growth retardation and increased resistance of the umbilical artery Doppler flow. J Matern Fetal Med, 9(5), 282-6. Mitra, S. C., Venkataseshan, V. S., von Hagen, S., Barton, P. T., Delshad, G. & Gil, J. (1997) Morphometric study of the placental vessels and its correlation with umbilical artery Doppler flow. Obstet Gynecol, 89(2), 238-41. Mittal, P., Espinoza, J., Hassan, S., Kusanovic, J. P., Edwin, S. S., Nien, J. K., Gotsch, F., Than, N. G., Erez, O., Mazaki-Tovi, S. & Romero, R. (2007) Placental growth hormone is increased in the maternal and fetal serum of patients with preeclampsia. J Matern Fetal Neonatal Med, 20(9), 651-9. Miyague, A. H., Raine-Fenning, N. J., Polanski, L., Martinez, L. H., Araujo Junior, E., Pavan, T. Z. & Martins, W. P. (2013) Assessing repeatability of 3D Doppler indices obtained by static 3D and STIC power Doppler: a combined in-vivo/in-vitro flow phantom study. Ultrasound Obstet Gynecol, 42(5), 571-6. Page 232 of 253 Monier, I., Blondel, B., Ego, A., Kaminiski, M., Goffinet, F. & Zeitlin, J. (2014) Poor effectiveness of antenatal detection of fetal growth restriction and consequences for obstetric management and neonatal outcomes: a French national study. BJOG. Moore, T. R. & Piacquadio, K. (1989) A prospective evaluation of fetal movement screening to reduce the incidence of antepartum fetal death. Am J Obstet Gynecol, 160(5 Pt 1), 1075-80. Moraitis, A. A., Wood, A. M., Fleming, M. & Smith, G. C. (2014) Birth weight percentile and the risk of term perinatal death. Obstet Gynecol, 124(2 Pt 1), 274-83. Morales-Rosello, J., Khalil, A., Morlando, M., Papageorghiou, A., Bhide, A. & Thilaganathan, B. (2014) Changes in fetal Doppler indices as a marker of failure to reach growth potential at term. Ultrasound Obstet Gynecol, 43(3), 303-10. Moran, M., Ryan, J., Higgins, M., Brennan, P. C. & McAuliffe, F. M. (2011) Poor agreement between operators on grading of the placenta. J Obstet Gynaecol, 31(1), 24-8. Morrison, I., Green, P. & Oomen, B. (1980) The role of human placental lactogen assays in antepartum fetal assessment. Am J Obstet Gynecol, 136(8), 1055-60. Morrow, R. J., Adamson, S. L., Bull, S. B. & Ritchie, J. W. (1989) Effect of placental embolization on the umbilical arterial velocity waveform in fetal sheep. Am J Obstet Gynecol, 161(4), 1055-60. Morse, K., Williams, A. & Gardosi, J. (2009) Fetal growth screening by fundal height measurement. Best Pract Res Clin Obstet Gynaecol, 23(6), 80918. Mu, J., Kanzaki, T., Tomimatsu, T., Fukuda, H., Fujii, E., Takeuchi, H. & Murata, Y. (2002) Investigation of intraplacental villous arteries by Doppler flow imaging in growth-restricted fetuses. Am J Obstet Gynecol, 186(2), 297-302. Mulder, E. J., Derks, J. B. & Visser, G. H. (1997) Antenatal corticosteroid therapy and fetal behaviour: a randomised study of the effects of Page 233 of 253 betamethasone and dexamethasone. Br J Obstet Gynaecol, 104(11), 1239-47. Murakoshi, T., Sekizuka, N., Takakuwa, K., Yoshizawa, H. & Tanaka, K. (1996) Uterine and spiral artery flow velocity waveforms in pregnancyinduced hypertension and/or intrauterine growth retardation. Ultrasound Obstet Gynecol, 7(2), 122-8. Murao, F. (1991) Measurements of the fetal liver size, hormonal level and pregnancy outcome. Gynecol Obstet Invest, 32(3), 153-6. Nabhan, A. F. & Abdelmoula, Y. A. (2008) Amniotic fluid index versus single deepest vertical pocket as a screening test for preventing adverse pregnancy outcome. Cochrane Database Syst Rev, (3), CD006593. Nagamatsu, T., Fujii, T., Kusumi, M., Zou, L., Yamashita, T., Osuga, Y., Momoeda, M., Kozuma, S. & Taketani, Y. (2004) Cytotrophoblasts upregulate soluble fms-like tyrosine kinase-1 expression under reduced oxygen: an implication for the placental vascular development and the pathophysiology of preeclampsia. Endocrinology, 145(11), 483845. Nardozza, L. M., Araujo Junior, E., Simioni, C., Torloni, M. R. & Moron, A. F. (2009) Evolution of 3-D power Doppler indices of fetal brain in normal pregnancy. Ultrasound Med Biol, 35(4), 545-9. National Institute for Health and Clinical Excellence. (2007) Intrapartum care: care of healthy women and their babies during childbirth. Clinical Guideline Number 65. London: National Institute for Health and Clinical Excellence. National Institute for Health and Care Excellence. (2008) Antenatal Care: Routine Care for the Healthy Pregnant Woman. Clinical Guideline Number 62. National Institute for Health and Care Excellence. National Library of Medicine. (2014) Promoting Awareness of Fetal Movements to Reduce Fetal Mortality Stillbirth, a Stepped Wedge Cluster Randomised Trial (AFFIRM) [Online]. Available: http://www.clinicaltrials.gov/ct2/show/NCT01777022 [Accessed 14/12/2014]. Page 234 of 253 Neilson, J. P. (2006) Fetal electrocardiogram (ECG) for fetal monitoring during labour. Cochrane Database Syst Rev, (3), CD000116. Neilson, J. P. (2012) Biochemical tests of placental function for assessment in pregnancy. Cochrane Database Syst Rev, 8, CD000108. Neldam, S. (1983) Fetal movements as an indicator of fetal well-being. Dan Med Bull, 30(4), 274-8. Nevo, O., Soleymanlou, N., Wu, Y., Xu, J., Kingdom, J., Many, A., Zamudio, S. & Caniggia, I. (2006) Increased expression of sFlt-1 in in vivo and in vitro models of human placental hypoxia is mediated by HIF-1. Am J Physiol Regul Integr Comp Physiol, 291(4), R1085-93. Newbern, D. & Freemark, M. (2011) Placental hormones and the co ntrol of maternal metabolism and fetal growth. Curr Opin Endocrinol Diabetes Obes, 18(6), 409-16. Newhouse, S. M., Davidge, S. T., Winkler-Lowen, B., Demianczuk, N. & Guilbert, L. J. (2007) In vitro differentiation of villous trophoblasts from pregnancies complicated by intrauterine growth restriction with and without pre-eclampsia. Placenta, 28(10), 999-1003. New Zealand Ministry of Health. (2012) Fetal and Infant Deaths 2008 and 2009 [Online]. Wellington: Ministry of Health. Available: http://www.health.govt.nz/system/files/documents/publications/fe tal-infant-deaths-2008-2009.pdf [Accessed 23/10/2014]. Nice, D. Hayden, K., Higgins, L. E., Johnstone, E. D. & Heazell, A. E. P. (2014) Human placental lactogen and placental growth factor can differentiate small for gestational age from appropriately grown infants. Placenta, 35(Supplement), P1.28-N. Norberg, S., Powell, T. L. & Jansson, T. (1998) Intrauterine growth restriction is associated with a reduced activity of placental taurine transporters. Pediatr Res, 44(2), 233-8. O'Dwyer, V., Burke, G., Unterscheider, J., Daly, S., Geary, M. P., Kennelly, M. M., McAuliffe, F. M., O'Donoghue, K., Hunter, A., Morrison, J. J., Dicker, P., Tully, E. C. & Malone, F. D. (2014) Defining the residual risk of adverse perinatal outcome in growth-restricted fetuses with normal umbilical artery blood flow. Am J Obstet Gynecol, 211(4), 420 e1-5. Page 235 of 253 O'Sullivan, O., Stephen, G., Martindale, E. & Heazell, A. E. (2009) Predicting poor perinatal outcome in women who present with decreased fetal movements. J Obstet Gynaecol, 29(8), 705-10. Odeh, M., Ophir, E., Maximovsky, O., Grinin, V. & Bornstein, J. (2011) Placental volume and three-dimensional power Doppler analysis in prediction of pre-eclampsia and small for gestational age between Week 11 and 13 weeks and 6 days of gestation. Prenat Diagn, 31(4), 367-71. Office for National Statistics. (2014) Births in England and Wales, 2013 [Online]. Available: http://www.ons.gov.uk/ons/publications/re- reference-tables.html?edition=tcm%3A77-317529 [Accessed 13/12/2014]. Oros, D., Figueras, F., Cruz-Martinez, R., Meler, E., Munmany, M. & Gratacos, E. (2011) Longitudinal changes in uterine, umbilical and fetal cerebral Doppler indices in late-onset small-for-gestational age fetuses. Ultrasound Obstet Gynecol, 37(2), 191-5. Orzechowski, K., Thomas, D., McNamara, C. J. & Miller, R. C. (2014) Volumetric assessment of longitudinal placental growth. Obstet Gynecol, 123 Suppl 1, 164S. Out, H. J., Kooijman, C. D., Bruinse, H. W. & Derksen, R. H. (1991) Histopathological findings in placentae from patients with intrauterine fetal death and anti-phospholipid antibodies. Eur J Obstet Gynecol Reprod Biol, 41(3), 179-86. Overmoyer, B. A., McLaren, C. E. & Brittenham, G. M. (1987) Uniformity of liver density and nonheme (storage) iron distribution. Arch Pathol Lab Med, 111(6), 549-54. Ozeren, M., Dinc, H., Ekmen, U., Senekayli, C. & Aydemir, V. (1999) Umbilical and middle cerebral artery Doppler indices in patients with preeclampsia. Eur J Obstet Gynecol Reprod Biol, 82(1), 11-6. Ozkaya, U., Ozkan, S., Ozeren, S. & Corakci, A. (2007) Doppler examination of uteroplacental circulation in early pregnancy: can it predict adverse outcome? J Clin Ultrasound, 35(7), 382-6. Page 236 of 253 Pagani, G., D'Antonio, F., Khalil, A., Akolekar, R., Papageorghiou, A., Bhide, A. & Thilaganathan, B. (2014a) Association between reduced fetal movements at term and abnormal uterine artery Doppler indices. Ultrasound Obstet Gynecol, 43(5), 548-52. Pagani, G., D'Antonio, F., Khalil, A., Papageorghiou, A., Bhide, A. & Thilaganathan, B. (2014b) Association between reduced fetal movements at term and first trimester markers of impaired placental development. Placenta, 35(8), 606-10. Pairleitner, H., Steiner, H., Hasenoehrl, G. & Staudach, A. (1999) Threedimensional power Doppler sonography: imaging and quantifying blood flow and vascularization. Ultrasound Obstet Gynecol, 14(2), 139-43. Palacio, M., Figueras, F., Zamora, L., Jimenez, J. M., Puerto, B., Coll, O., Cararach, V. & Vanrell, J. A. (2004) Reference ranges for umbilical and middle cerebral artery pulsatility index and cerebroplacental ratio in prolonged pregnancies. Ultrasound Obstet Gynecol, 24(6), 647-53. Palmer, S. K., Zamudio, S., Coffin, C., Parker, S., Stamm, E. & Moore, L. G. (1992) Quantitative estimation of human uterine artery blood flow and pelvic blood flow redistribution in pregnancy. Obstet Gynecol, 80(6), 1000-6. Paolini, C. L., Marconi, A. M., Ronzoni, S., Di Noio, M., Fennessey, P. V., Pardi, G. & Battaglia, F. C. (2001) Placental transport of leucine, phenylalanine, glycine, and proline in intrauterine growth-restricted pregnancies. J Clin Endocrinol Metab, 86(11), 5427-32. Papadopoulou, E., Sifakis, S., Giahnakis, E., Fragouli, Y., Karkavitsas, N., Koumantakis, E. & Kalmanti, M. (2006) Increased human placental growth hormone at midtrimester pregnancies may be an index of intrauterine growth retardation related to preeclampsia. Growth Horm IGF Res, 16(5-6), 290-6. Parra-Saavedra, M., Crovetto, F., Triunfo, S., Savchev, S., Peguero, A., Nadal, A., Gratacos, E. & Figueras, F. (2014a) Association of Doppler parameters with placental signs of underperfusion in late-onset Page 237 of 253 small-for-gestational-age pregnancies. Ultrasound Obstet Gynecol, 44(3), 330-7. Parra-Saavedra, M., Crovetto, F., Triunfo, S., Savchev, S., Peguero, A., Nadal, A., Parra, G., Gratacos, E. & Figueras, F. (2013) Placental findings in late-onset SGA births without Doppler signs of placental insufficiency. Placenta, 34(12), 1136-41. Parra-Saavedra, M., Crovetto, F., Triunfo, S., Savchev, S., Peguero, A., Nadal, A., Parra, G., Gratacos, E. & Figueras, F. (2014b) Neurodevelopmental outcomes of near-term small-for-gestational-age infants with and without signs of placental underperfusion. Placenta, 35(4), 269-74. Parra-Saavedra, M., Simeone, S., Triunfo, S., Crovetto, F., Botet, F., Nadal, A., Gratacos, E. & Figueras, F. (2014c) Correlation between placental underperfusion, histologic signs and perinatal morbidity in late-onset small for gestational age fetuses. Ultrasound Obstet Gynecol. Pasupathy, D., Dacey, A., Cook, E., Charnock-Jones, D. S., White, I. R. & Smith, G. C. (2008) Study protocol. A prospective cohort study of unselected primiparous women: the pregnancy outcome prediction study. BMC Pregnancy Childbirth, 8, 51. Pates, J. A., Hatab, M. R., McIntire, D. D., Cunningham, F. G. & Twickler, D. M. (2010) Determining uterine blood flow in pregnancy with magnetic resonance imaging. Magn Reson Imaging, 28(4), 507-10. Pathak, S., Hook, E., Hackett, G., Murdoch, E., Sebire, N. J., Jessop, F. & Lees, C. (2010) Cord coiling, umbilical cord insertion and placental shape in an unselected cohort delivering at term: relationship with common obstetric outcomes. Placenta, 31(11), 963-8. Patrick, J., Campbell, K., Carmichael, L., Natale, R. & Richardson, B. (1980) Patterns of human fetal breathing during the last 10 weeks of pregnancy. Obstet Gynecol, 56(1), 24-30. Pearson, J. F. & Weaver, J. B. (1976) Fetal activity and fetal wellbeing: an evaluation. Br Med J, 1(6021), 1305-7. Pinar, H. & Carpenter, M. (2010) Placenta and umbilical cord abnormalities seen with stillbirth. Clin Obstet Gynecol, 53(3), 656-72. Page 238 of 253 Pinar, H., Goldenberg, R. L., Koch, M. A., Heim-Hall, J., Hawkins, H. K., Shehata, B., Abramowsky, C., Parker, C. B., Dudley, D. J., Silver, R. M., Stoll, B., Carpenter, M., Saade, G., Moore, J., Conway, D., Varner, M. W., Hogue, C. J., Coustan, D. R., Sbrana, E., Thorsten, V., Willinger, M. & Reddy, U. M. (2014) Placental findings in singleton stillbirths. Obstet Gynecol, 123(2 Pt 1), 325-36. Pomorski, M., Zimmer, M., Florjanski, J., Michniewicz, J., Wiatrowski, A., Fuchs, T. & Milnerowicz-Nabzdyk, E. (2012) Comparative analysis of placental vasculature and placental volume in normal and IUGR pregnancies with the use of three-dimensional Power Doppler. Arch Gynecol Obstet, 285(2), 331-7. Poon, L. C., Zaragoza, E., Akolekar, R., Anagnostopoulos, E. & Nicolaides, K. H. (2008) Maternal serum placental growth factor (PlGF) in small for gestational age pregnancy at 11(+0) to 13(+6) weeks of gestation. Prenat Diagn, 28(12), 1110-5. Porat, S., Fitzgerald, B., Wright, E., Keating, S. & Kingdom, J. C. (2013) Placental hyperinflation and the risk of adverse perinatal outcome. Ultrasound Obstet Gynecol, 42(3), 315-21. Prefumo, F., Sebire, N. J. & Thilaganathan, B. (2004) Decreased endovascular trophoblast invasion in first trimester pregnancies with highresistance uterine artery Doppler indices. Hum Reprod, 19(1), 206-9. Prior, T., Mullins, E., Bennett, P. & Kumar, S. (2013) Prediction of intrapartum fetal compromise using the cerebroumbilical ratio: a prospective observational study. Am J Obstet Gynecol, 208(2), 124 e16. Prior, T., Mullins, E., Bennett, P. & Kumar, S. (2014a) Prediction of fetal compromise in labor. Obstet Gynecol, 123(6), 1263-71. Prior, T., Mullins, E., Bennett, P. & Kumar, S. (2014b) Umbilical venous flow rate in term fetuses: can variations in flow predict intrapartum compromise? Am J Obstet Gynecol, 210(1), 61 e1-8. Proctor, L. K., Fitzgerald, B., Whittle, W. L., Mokhtari, N., Lee, E., Machin, G., Kingdom, J. C. & Keating, S. J. (2013) Umbilical cord diameter Page 239 of 253 percentile curves and their correlation to birth weight and placental pathology. Placenta, 34(1), 62-6. Proctor, L. K., Toal, M., Keating, S., Chitayat, D., Okun, N., Windrim, R. C., Smith, G. C. & Kingdom, J. C. (2009) Placental size and the prediction of severe early-onset intrauterine growth restriction in women with low pregnancy-associated plasma protein-A. Ultrasound Obstet Gynecol, 34(3), 274-82. Proctor, L. K., Whittle, W. L., Keating, S., Viero, S. & Kingdom, J. C. (2010) Pathologic basis of echogenic cystic lesions in the human placenta: role of ultrasound-guided wire localization. Placenta, 31(12), 1111-5. Proud, J. & Grant, A. M. (1987) Third trimester placental grading by ultrasonography as a test of fetal wellbeing. Br Med J (Clin Res Ed), 294(6588), 1641-4. Ptacek, I., Sebire, N. J., Man, J. A., Brownbill, P. & Heazell, A. E. (2014) Systematic review of placental pathology reported in association with stillbirth. Placenta, 35(8), 552-62. Raine-Fenning, N. J., Nordin, N. M., Ramnarine, K. V., Campbell, B. K., Clewes, J. S., Perkins, A. & Johnson, I. R. (2008a) Determining the relationship between three-dimensional power Doppler data and true blood flow characteristics: an in-vitro flow phantom experiment. Ultrasound Obstet Gynecol, 32(4), 540-50. Raine-Fenning, N. J., Nordin, N. M., Ramnarine, K. V., Campbell, B. K., Clewes, J. S., Perkins, A. & Johnson, I. R. (2008b) Evaluation of the effect of machine settings on quantitative three-dimensional power Doppler angiography: an in-vitro flow phantom experiment. Ultrasound Obstet Gynecol, 32(4), 551-9. Rajakumar, A., Powers, R. W., Hubel, C. A., Shibata, E., von Versen-Hoynck, F., Plymire, D. & Jeyabalan, A. (2009) Novel soluble Flt-1 isoforms in plasma and cultured placental explants from normotensive pregnant and preeclamptic women. Placenta, 30(1), 25-34. Read, M. A., Leitch, I. M., Giles, W. B., Bisits, A. M., Boura, A. L. & Walters, W. A. (1999) U46619-mediated vasoconstriction of the fetal placental Page 240 of 253 vasculature in vitro in normal and hypertensive pregnancies. J Hypertens, 17(3), 389-96. Reddy, A., Suri, S., Sargent, I. L., Redman, C. W. & Muttukrishna, S. (2009). Maternal cirulating levels of activin A, inhibin A, sFlt-1 and endoglin at parturition in normal pregnancy and pre-eclampsia. PLoS ONE 4(2):e4453 Rigano, S., Ferrazzi, E., Boito, S., Pennati, G., Padoan, A. & Galan, H. (2010) Blood flow volume of uterine arteries in human pregnancies determined using 3D and bi-dimensional imaging, angio-Doppler, and fluid-dynamic modeling. Placenta, 31(1), 37-43. Rizzo, G., Capponi, A., Cavicchioni, O., Vendola, M. & Arduini, D. (2008) First trimester uterine Doppler and three-dimensional ultrasound placental volume calculation in predicting pre-eclampsia. Eur J Obstet Gynecol Reprod Biol, 138(2), 147-51. Rizzo, G., Capponi, A., Pietrolucci, M. E., Capece, A. & Arduini, D. (2009) Firsttrimester placental volume and vascularization measured by 3dimensional power Doppler sonography in pregnancies with low serum pregnancy-associated plasma protein a levels. J Ultrasound Med, 28(12), 1615-22. Rizzo, G., Silvestri, E., Capponi, A., Servadei, F., Pietrolucci, M. E., Capece, A., Pisa, R. & Arduini, D. (2011) Histomorphometric characteristics of first trimester chorionic villi in pregnancies with low serum pregnancy-associated plasma protein-A levels: relationship with placental three-dimensional power doppler ultrasonographic vascularization. J Matern Fetal Neonatal Med, 24(2), 253-7. Robson, S. C., Martin, W. L. & Morris, R. K. (2014) The Investigation and Management of the Small-for-Gestational-Age Fetus. Green-top Guideline. 2nd Edition ed. London: Royal College of Obstetricians and Gynaecologists. Roje, D., Tomas, S. Z., Prusac, I. K., Capkun, V. & Tadin, I. (2011) Trophoblast apoptosis in human term placentas from pregnancies complicated with idiopathic intrauterine growth retardation. J Matern Fetal Neonatal Med, 24(5), 745-51. Page 241 of 253 Roland, M. C., Friis, C. M., Voldner, N., Godang, K., Bollerslev, J., Haugen, G. & Henriksen, T. (2012) Fetal growth versus birthweight: the role of placenta versus other determinants. PLoS One, 7(6), e39324. Romero, R., Chaiworapongsa, T., Erez, O., Tarca, A. L., Gervasi, M. T., Kusanovic, J. P., Mittal, P., Ogge, G., Vaisbuch, E., Mazaki-Tovi, S., Dong, Z., Kim, S. K., Yeo, L. & Hassan, S. S. (2010) An imbalance between angiogenic and anti-angiogenic factors precedes fetal death in a subset of patients: results of a longitudinal study. J Matern Fetal Neonatal Med, 23(12), 1384-99. Romero, R., Nien, J. K., Espinoza, J., Todem, D., Fu, W., Chung, H., Kusanovic, J. P., Gotsch, F., Erez, O., Mazaki-Tovi, S., Gomez, R., Edwin, S., Chaiworapongsa, T., Levine, R. J. & Karumanchi, S. A. (2008) A longitudinal study of angiogenic (placental growth factor) and antiangiogenic (soluble endoglin and soluble vascular endothelial growth factor receptor-1) factors in normal pregnancy and patients destined to develop preeclampsia and deliver a small for gestational age neonate. J Matern Fetal Neonatal Med, 21(1), 9-23. Rosenstein, M. G., Cheng, Y. W., Snowden, J. M., Nicholson, J. M. & Caughey, A. B. (2012) Risk of stillbirth and infant death stratified by gestational age. Obstet Gynecol, 120(1), 76-82. Rossi, A., Romanello, I., Forzano, L., Fachechi, G. & Marchesoni, D. (2011) Evaluation of fetal cerebral blood flow perfusion using power Doppler ultrasound angiography (3D-PDA) in growth-restricted fetuses. Facts Views Vis Obgyn, 3(3), 175-80. Rotmensch, S., Liberati, M., Luo, J. S., Kliman, H. J., Gollin, Y., Bellati, U., Hobbins, J. C. & Copel, J. A. (1994) Color Doppler flow patterns and flow velocity waveforms of the intraplacental fetal circulation in growth-retarded fetuses. Am J Obstet Gynecol, 171(5), 1257-64. Saastad, E., Ahlborg, T. & Froen, J. F. (2008) Low maternal awareness of fetal movement is associated with small for gestational age infants. J Midwifery Womens Health, 53(4), 345-52. Saastad, E., Tveit, J. V., Flenady, V., Stray-Pedersen, B., Fretts, R. C., Bordahl, P. E. & Froen, J. F. (2010) Implementation of uniform information on Page 242 of 253 fetal movement in a Norwegian population reduced delayed reporting of decreased fetal movement and stillbirths in primiparous women - a clinical quality improvement. BMC Res Notes, 3(1), 2. Saastad, E., Vangen, S. & Froen, J. F. (2007) Suboptimal care in stillbirths - a retrospective audit study. Acta Obstet Gynecol Scand, 86(4), 444-50. Saastad, E., Winje, B. A., Stray Pedersen, B. & Froen, J. F. (2011) Fetal movement counting improved identification of fetal growth restriction and perinatal outcomes--a multi-centre, randomized, controlled trial. PLoS One, 6(12), e28482. Sadovsky, E., Ohel, G., Havazeleth, H., Steinwell, A. & Penchas, S. (1983) The definition and the significance of decreased fetal movements. Acta Obstet Gynecol Scand, 62(5), 409-13. Sagol, S., Sagol, O. & Ozdemir, N. (2002) Stereological quantification of placental villus vascularization and its relation to umbilical artery Doppler flow in intrauterine growth restriction. Prenat Diagn, 22(5), 398-403. Sau, A., Seed, P. & Langford, K. (2004) Intraobserver and interobserver variation in the sonographic grading of placental maturity. Ultrasound Obstet Gynecol, 23(4), 374-7. Savchev, S., Figueras, F., Sanz-Cortes, M., Cruz-Lemini, M., Triunfo, S., Botet, F. & Gratacos, E. (2014) Evaluation of an optimal gestational age cutoff for the definition of early- and late-onset fetal growth restriction. Fetal Diagn Ther, 36(2), 99-105. Savchev, S., Sanz-Cortes, M., Cruz-Martinez, R., Arranz, A., Botet, F., Gratacos, E. & Figueras, F. (2013) Neurodevelopmental outcome of full-term small-for-gestational-age infants with normal placental function. Ultrasound Obstet Gynecol, 42(2), 201-6. Scherle, W. (1970) A simple method for volumetry of organs in quantitative stereology. Mikroskopie, 26(1), 57-60. Schild, R. L., Maringa, M., Siemer, J., Meurer, B., Hart, N., Goecke, T. W., Schmid, M., Hothorn, T. & Hansmann, M. E. (2008) Weight estimation by three-dimensional ultrasound imaging in the small fetus. Ultrasound Obstet Gynecol, 32(2), 168-75. Page 243 of 253 Schneuer, F. J., Nassar, N., Khambalia, A. Z., Tasevski, V., Guilbert, C., Ashton, A. W., Morris, J. M. & Roberts, C. L. (2012) First trimester screening of maternal placental protein 13 for predicting preeclampsia and small for gestational age: in-house study and systematic review. Placenta, 33(9), 735-40. Schwartz, N., Wang, E. & Parry, S. (2012) Two-dimensional sonographic placental measurements in the prediction of small-for-gestationalage infants. Ultrasound Obstet Gynecol, 40(6), 674-9. Sela, S., Itin, A., Natanson-Yaron, S., Greenfield, C., Goldman-Wohl, D., Yagel, S. & Keshet, E. (2008) A novel human-specific soluble vascular endothelial growth factor receptor 1: cell-type-specific splicing and implications to vascular endothelial growth factor homeostasis and preeclampsia. Circ Res, 102(12), 1566-74. Shi, Q. J., Lei, Z. M., Rao, C. V. & Lin, J. (1993) Novel role of human chorionic gonadotropin in differentiation of human cytotrophoblasts. Endocrinology, 132(3), 1387-95. Shibata, E., Rajakumar, A., Powers, R. W., Larkin, R. W., Gilmour, C., Bodnar, L. M., Crombleholme, W. R., Ness, R. B., Roberts, J. M. & Hubel, C. A. (2005) Soluble fms-like tyrosine kinase 1 is increased in preeclampsia but not in normotensive pregnancies with small-forgestational-age neonates: relationship to circulating placental growth factor. J Clin Endocrinol Metab, 90(8), 4895-903. Shore, V. H., Wang, T. H., Wang, C. L., Torry, R. J., Caudle, M. R. & Torry, D. S. (1997) Vascular endothelial growth factor, placenta growth factor and their receptors in isolated human trophoblast. Placenta, 18(8), 657-65. Sibley, C. P., Turner, M. A., Cetin, I., Ayuk, P., Boyd, C. A., D'Souza, S. W., Glazier, J. D., Greenwood, S. L., Jansson, T. & Powell, T. (2005) Placental phenotypes of intrauterine growth. Pediatr Res, 58(5), 82732. Sifakis, S., Akolekar, R., Kappou, D., Mantas, N. & Nicolaides, K. H. (2012) Maternal serum placental growth hormone at 11-13 weeks gestation Page 244 of 253 in pregnancies delivering small for gestational age neonates. J Matern Fetal Neonatal Med, 25(9), 1796-9. Siman, C. M., Sibley, C. P., Jones, C. J., Turner, M. A. & Greenwood, S. L. (2001) The functional regeneration of syncytiotrophoblast in cultured explants of term placenta. Am J Physiol Regul Integr Comp Physiol, 280(4), R1116-22. Smith, G. C. (2010) Predicting antepartum stillbirth. Clin Obstet Gynecol, 53(3), 597-606. Smith, G. C. (2012) Researching new methods of screening for adverse pregnancy outcome: lessons from pre-eclampsia. PLoS Med, 9(7), e1001274. Smith, S. C., Baker, P. N. & Symonds, E. M. (1997) Increased placental apoptosis in intrauterine growth restriction. Am J Obstet Gynecol, 177(6), 1395-401. Sonesson, S. E., Fouron, J. C., Drblik, S. P., Tawile, C., Lessard, M., Skoll, A., Guertin, M. C. & Ducharme, G. R. (1993) Reference values for Doppler velocimetric indices from the fetal and placental ends of the umbilical artery during normal pregnancy. J Clin Ultrasound, 21(5), 317-24. Soregaroli, M., Bonera, R., Danti, L., Dinolfo, D., Taddei, F., Valcamonico, A. & Frusca, T. (2002) Prognostic role of umbilical artery Doppler velocimetry in growth-restricted fetuses. J Matern Fetal Neonatal Med, 11(3), 199-203. Sorensen, S., von Tabouillot, D., Schioler, V., Greisen, G., Petersen, S. & Larsen, T. (2000) Serial measurements of serum human placental lactogen (hPL) and serial ultrasound examinations in the evaluation of fetal growth. Early Hum Dev, 60(1), 25-34. Spellacy, W. N., Conly, P. W., Cleveland, W. W. & Buhi, W. S. (1975) Effects of fetal sex and weight and placental weight on maternal serum progesterone and chorionic gonadotropin concentrations. Am J Obstet Gynecol, 122(3), 278-82. Spellacy, W. N., Cruz, A. C., Gelman, S. R. & Buhi, W. C. (1977) Fetal movements and placental lactogen levels for fetal-placental evaluation. A preliminary report. Obstet Gynecol, 49(1), 113-5. Page 245 of 253 Spencer, K., Cowans, N. J., Avgidou, K. & Nicolaides, K. H. (2006) Firsttrimester ultrasound and biochemical markers of aneuploidy and the prediction of impending fetal death. Ultrasound Obstet Gynecol, 28(5), 637-43. Spinillo, A., Gardella, B., Bariselli, S., Alfei, A., Silini, E. & Dal Bello, B. (2012) Placental histopathological correlates of umbilical artery Doppler velocimetry in pregnancies complicated by fetal growth restriction. Prenat Diagn, 32(13), 1263-72. Staff, A. C., Laine, K., Lahmami, G., Gustafsson, J., Surbehan, H., Barbero, M. L. & Kliman, H. J. (2011) Second trimester estimated placental volume (EPV) screening is predictive of SGA and LGA at birth. Placenta, 32(9), A5:PL3.NI4. Stallmach, T., Hebisch, G., Meier, K., Dudenhausen, J. W. & Vogel, M. (2001) Rescue by birth: defective placental maturation and late fetal mortality. Obstet Gynecol, 97(4), 505-9. Steier, J. A., Bergsjo, P. B., Thorsen, T. & Myking, O. L. (2004) Human chorionic gonadotropin in maternal serum in relation to fetal gender and utero-placental blood flow. Acta Obstet Gynecol Scand, 83(2), 170-4. Steier, K. J. (1999) Take the time to teach to ensure the legacy and future of our profession. J Am Osteopath Assoc, 99(12), 605. Steier, L. & Freedman, G. (2004) Next adhesive endodontics: a single-step obturation and post technique. Compend Contin Educ Dent, 25(10A), 827-8, 830. Stewart, M. O., Whittaker, P. G., Persson, B., Hanson, U. & Lind, T. (1989) A longitudinal study of circulating progesterone, oestradiol, hCG and hPL during pregnancy in type 1 diabetic mothers. Br J Obstet Gynaecol, 96(4), 415-23. Stillbirth Collaborative Research Network Writing Group (2011). Association between stillbirth and risk factors known at pregnancy confirmation. JAMA, 306(22), 2469-79. Page 246 of 253 Stock, M. K., Anderson, D. F., Phernetton, T. M., McLaughlin, M. K. & Rankin, J. H. (1980) Vascular response of the fetal placenta to local occlusion of the maternal placental vasculature. J Dev Physiol, 2(5), 339-46. Stock, S. J., Ferguson, E., Duffy, A., Ford, I., Chalmers, J. & Norman, J. E. (2012) Outcomes of elective induction of labour compared with expectant management: population based study. BMJ, 344, e2838. Strid, H., Bucht, E., Jansson, T., Wennergren, M. & Powell, T. L. (2003) ATP dependent Ca2+ transport across basal membrane of human syncytiotrophoblast in pregnancies complicated by intrauterine growth restriction or diabetes. Placenta, 24(5), 445-52. Strong, T. H., Jr., Jarles, D. L., Vega, J. S. & Feldman, D. B. (1994) The umbilical coiling index. Am J Obstet Gynecol, 170(1 Pt 1), 29-32. Sugimura, M., Ohashi, R., Kobayashi, T. & Kanayama, N. (2001) Intraplacental coagulation in intrauterine growth restriction: cause or result? Semin Thromb Hemost, 27(2), 107-13. Suri, S., Muttukrishna, S. & Jauniaux, E. (2013) 2D-Ultrasound and endocrinologic evaluation of placentation in early pregnancy and its relationship to fetal birthweight in normal pregnancies and preeclampsia. Placenta, 34(9), 745-50. Tagliabue, L. & Del Sole, A. (2014) Appropriate use of positron emission tomography with [(18)F]fluorodeoxyglucose for staging of oncology patients. Eur J Intern Med, 25(1), 6-11. Tan, K. H. & Sabapathy, A. (2001) Fetal manipulation for facilitating tests of fetal wellbeing. Cochrane Database Syst Rev, (4), CD003396. Tan, K. H. & Smyth, R. (2001) Fetal vibroacoustic stimulation for facilitation of tests of fetal wellbeing. Cochrane Database Syst Rev, (1), CD002963. Tantbirojn, P., Saleemuddin, A., Sirois, K., Crum, C. P., Boyd, T. K., Tworoger, S. & Parast, M. M. (2009) Gross abnormalities of the umbilical cord: related placental histology and clinical significance. Placenta, 30(12), 1083-8. Tavor, O., Shohat, M. & Lipitz, S. (2007) The relationship between perinatal outcome of singleton pregnancies and isolated highly elevated levels Page 247 of 253 of maternal serum human chorionic gonadotropin at mid-gestation. Isr Med Assoc J, 9(7), 509-12. Tchirikov, M., Rybakowski, C., Huneke, B., Schoder, V. & Schroder, H. J. (2002) Umbilical vein blood volume flow rate and umbilical artery pulsatility as 'venous-arterial index' in the prediction of neonatal compromise. Ultrasound Obstet Gynecol, 20(6), 580-5. Tchirikov, M., Strohner, M., Forster, D. & Huneke, B. (2009) A combination of umbilical artery PI and normalized blood flow volume in the umbilical vein: venous-arterial index for the prediction of fetal outcome. Eur J Obstet Gynecol Reprod Biol, 142(2), 129-33. Teramo, K. A., Schwartz, R., Clemons, G. K. & Widness, J. A. (2002) Amniotic fluid erythropoietin concentrations differentiate between acute and chronic causes of fetal death. Acta Obstet Gynecol Scand, 81(3), 24551. Thompson, J. M., Irgens, L. M., Skjaerven, R. & Rasmussen, S. (2007) Placenta weight percentile curves for singleton deliveries. BJOG, 114(6), 71520. Thompson, R. S. & Trudinger, B. J. (1990) Doppler waveform pulsatility index and resistance, pressure and flow in the umbilical placental circulation: an investigation using a mathematical model. Ultrasound Med Biol, 16(5), 449-58. Toal, M., Chaddha, V., Windrim, R. & Kingdom, J. (2008a) Ultrasound detection of placental insufficiency in women with elevated second trimester serum alpha-fetoprotein or human chorionic gonadotropin. J Obstet Gynaecol Can, 30(3), 198-206. Toal, M., Chan, C., Fallah, S., Alkazaleh, F., Chaddha, V., Windrim, R. C. & Kingdom, J. C. (2007) Usefulness of a placental profile in high-risk pregnancies. Am J Obstet Gynecol, 196(4), 363 e1-7. Toal, M., Keating, S., Machin, G., Dodd, J., Adamson, S. L., Windrim, R. C. & Kingdom, J. C. (2008b) Determinants of adverse perinatal outcome in high-risk women with abnormal uterine artery Doppler images. Am J Obstet Gynecol, 198(3), 330 e1-7. Page 248 of 253 Tuckey, R. C. (2005) Progesterone synthesis by the human placenta. Placenta, 26(4), 273-81. Tuckey, R. C. & Sadleir, J. (1999) The concentration of adrenodoxin reductase limits cytochrome p450scc activity in the human placenta. Eur J Biochem, 263(2), 319-25. Tuffnell, D. J., Cartmill, R. S. & Lilford, R. J. (1991) Fetal movements; factors affecting their perception. Eur J Obstet Gynecol Reprod Biol, 39(3), 165-7. Tulchinsky, D., Hobel, C. J., Yeager, E. & Marshall, J. R. (1972a) Plama estradiol, estriol, and progesterone in human pregnancy. II. Clinical applications in Rh-isoimmunization disease. Am J Obstet Gynecol, 113(6), 766-70. Tulchinsky, D., Hobel, C. J., Yeager, E. & Marshall, J. R. (1972b) Plasma estrone, estradiol, estriol, progesterone, and 17- hydroxyprogesterone in human pregnancy. I. Normal pregnancy. Am J Obstet Gynecol, 112(8), 1095-100. Tuvemo, T. & Strandberg, K. (1975) Effects and interactions of oxygen and prostaglandins on the tone of the isolated human umbilical artery. Ups J Med Sci, 80(3), 131-4. Tveit, J. V., Saastad, E., Stray-Pedersen, B., Bordahl, P. E., Flenady, V., Fretts, R. & Froen, J. F. (2009) Reduction of late stillbirth with the introduction of fetal movement information and guidelines - a clinical quality improvement. BMC Pregnancy Childbirth, 9, 32. UNICEF. (2014) Neonatal Mortality [Online]. http://data.unicef.org/child-mortality/neonatal Available: [Accessed 13/12/2014]. Valentin, L., Marsal, K. & Wahlgren, L. (1986) Subjective recording of fetal movements. III. Screening of a pregnant population; the clinical significance of decreased fetal movement counts. Acta Obstet Gynecol Scand, 65(7), 753-8. Vedmedovska, N., Rezeberga, D., Teibe, U., Melderis, I. & Donders, G. G. (2011) Placental pathology in fetal growth restriction. Eur J Obstet Gynecol Reprod Biol, 155(1), 36-40. Page 249 of 253 Vergani, P., Cozzolino, S., Pozzi, E., Cuttin, M. S., Greco, M., Ornaghi, S. & Lucchini, V. (2008) Identifying the causes of stillbirth: a comparison of four classification systems. Am J Obstet Gynecol, 199(3), 319 e1-4. Viero, S., Chaddha, V., Alkazaleh, F., Simchen, M. J., Malik, A., Kelly, E., Windrim, R. & Kingdom, J. C. (2004) Prognostic value of placental ultrasound in pregnancies complicated by absent end-diastolic flow velocity in the umbilical arteries. Placenta, 25(8-9), 735-41. Vintzileos, A. M., Gaffney, S. E., Salinger, L. M., Campbell, W. A. & Nochimson, D. J. (1987) The relationship between fetal biophysical profile and cord pH in patients undergoing cesarean section before the onset of labor. Obstet Gynecol, 70(2), 196-201. Wareing, M., Akef, H., Greenwood, S. L., Bugg, G. & Baker, P. N. (2005) Umbilical artery Doppler waveform indices from normal pregnant women are related to vasodilatation of placental chorionic plate small arteries. J Obstet Gynaecol, 25(3), 248-52. Wareing, M., Crocker, I. P., Warren, A. Y., Taggart, M. J. & Baker, P. N. (2002) Characterization of small arteries isolated from the human placental chorionic plate. Placenta, 23(5), 400-9. Wareing, M., Greenwood, S. L. & Baker, P. N. (2006) Reactivity of human placental chorionic plate vessels is modified by level of oxygenation: differences between arteries and veins. Placenta, 27(1), 42-8. Warrander, L. K., Batra, G., Bernatavicius, G., Greenwood, S. L., Dutton, P., Jones, R. L., Sibley, C. P. & Heazell, A. E. (2012) Maternal perception of reduced fetal movements is associated with altered placental structure and function. PLoS One, 7(4), e34851. Warrander, L. K. & Heazell, A. E. (2011) Identifying placental dysfunction in women with reduced fetal movements can be used to predict patients at increased risk of pregnancy complications. Med Hypotheses, 76(1), 17-20. Welsh, A. W., Collins, S. L., Stevenson, G. N., Noble, J. A. & Impey, L. (2012a) Inapplicability of fractional moving blood volume technique to standardize Virtual Organ Computer-aided AnaLysis indices for Page 250 of 253 quantified three-dimensional power Doppler. Ultrasound Obstet Gynecol, 40(6), 688-92. Welsh, A. W., Hou, M., Meriki, N. & Martins, W. P. (2012b) Spatiotemporal image correlation-derived volumetric Doppler impedance indices from spherical samples of the placenta: intraobserver reliability and correlation with conventional umbilical artery Doppler indices. Ultrasound Obstet Gynecol, 40(4), 431-6. Westgate, J. & Jamieson, M. (1986) Stillbirths and fetal movements. N Z Med J, 99(796), 114-6. Whitehead, C. L., Teh, W. T., Walker, S. P., Leung, C., Larmour, L. & Tong, S. (2013) Circulating MicroRNAs in maternal blood as potential biomarkers for fetal hypoxia in-utero. PLoS One, 8(11), e78487. Whitworth, M. K., Fisher, M. & Heazell, A. E. P. (2011) Reduced Fetal Movements. Green-top Guideline Number 57. London: Royal College of Obstetricians and Gynaecologists. Wilcock, F. M. & Kadir, R. A. (2004) Fetomaternal haemorrhage--a cause for unexplained neonatal death, presenting with reduced fetal movements and non-reactive fetal heart trace. J Obstet Gynaecol, 24(4), 456-7. Winje, B. A., Roald, B., Kristensen, N. P. & Froen, J. F. (2012) Placental pathology in pregnancies with maternally perceived decreased fetal movement--a population-based nested case-cohort study. PLoS One, 7(6), e39259. Wolf, H., Oosting, H. & Treffers, P. E. (1989) A longitudinal study of the relationship between placental and fetal growth as measured by ultrasonography. Am J Obstet Gynecol, 161(5), 1140-5. Worton, S. & Heazell, A. E. P. (2014) Decreased placental weight centile and increased birthweight: placental weight ratios in stillbirth suggests placental insufficiency even in stillbirths of "unknown" cause. Placenta, 35(9), A15-a16. Wright, C., Morris, D. M., Baker, P. N., Crocker, I. P., Gowland, P. A., Parker, G. J. & Sibley, C. P. (2011) Magnetic resonance imaging relaxation time measurements of the placenta at 1.5 T. Placenta, 32(12), 1010-5. Page 251 of 253 Yagel, S., Anteby, E. Y., Shen, O., Cohen, S. M., Friedman, Z. & Achiron, R. (1999a) Placental blood flow measured by simultaneous multigate spectral Doppler imaging in pregnancies complicated by placental vascular abnormalities. Ultrasound Obstet Gynecol, 14(4), 262-6. Yagel, S., Anteby, E. Y., Shen, O., Cohen, S. M., Friedman, Z. & Achiron, R. (1999b) Simultaneous multigate spectral Doppler imaging of the umbilical artery and placental vessels: novel ultrasound technology. Ultrasound Obstet Gynecol, 14(4), 256-61. Yampolsky, M., Salafia, C. M., Misra, D. P., Shlakhter, O. & Gill, J. S. (2013) Is the placental disk really an ellipse? Placenta, 34(4), 391-3. Yampolsky, M., Salafia, C. M., Shlakhter, O., Haas, D., Eucker, B. & Thorp, J. (2008) Modeling the variability of shapes of a human placenta. Placenta, 29(9), 790-7. Yampolsky, M., Salafia, C. M., Shlakhter, O., Haas, D., Eucker, B. & Thorp, J. (2009) Centrality of the umbilical cord insertion in a human placenta influences the placental efficiency. Placenta, 30(12), 1058-64. Yampolsky, M., Salafia, C. M., Shlakhter, O., Misra, D. P., Haas, D., Eucker, B. & Thorp, J. (2011) Variable placental thickness affects placental functional efficiency independent of other placental shape abnormalities. J Dev Orig Health Dis, 2(4), 205-11. Yang, W. T., Tse, G. M., Lam, P. K., Metreweli, C. & Chang, J. (2002) Correlation between color power Doppler sonographic measurement of breast tumor vasculature and immunohistochemical analysis of microvessel density for the quantitation of angiogenesis. J Ultrasound Med, 21(11), 1227-35. Yao, R., Ananth, C. V., Park, B. Y., Pereira, L. & Plante, L. A. (2014) Obesity and the risk of stillbirth: a population-based cohort study. Am J Obstet Gynecol, 210(5), 457 e1-9. Yaron, Y., Cherry, M., Kramer, R. L., O'Brien, J. E., Hallak, M., Johnson, M. P. & Evans, M. I. (1999) Second-trimester maternal serum marker screening: maternal serum alpha-fetoprotein, beta-human chorionic gonadotropin, estriol, and their various combinations as predictors of pregnancy outcome. Am J Obstet Gynecol, 181(4), 968-74. Page 252 of 253 Yin, T. T., Loughna, P., Ong, S. S., Padfield, J. & Mayhew, T. M. (2009) No correlation between ultrasound placental grading at 31-34 weeks of gestation and a surrogate estimate of organ function at term obtained by stereological analysis. Placenta, 30(8), 726-30. Yung, H. W., Calabrese, S., Hynx, D., Hemmings, B. A., Cetin, I., CharnockJones, D. S. & Burton, G. J. (2008) Evidence of placental translation inhibition and endoplasmic reticulum stress in the etiology of human intrauterine growth restriction. Am J Pathol, 173(2), 451-62. Zeitlin, J. M., A.; Delnord, M. (2013) European Perinatal Health Report. Health and Care of Pregnant Women and Babies in Europe in 2010. Euro-Peristat project with SCPE and Eurocat. Zlatnik, F. J., Varner, M. W. & Hauser, K. S. (1979) Human placental lactogen: a predictor of perinatal outcome? Obstet Gynecol, 54(2), 205-10. Page 253 of 253
© Copyright 2026 Paperzz