Modelling birthweight as a function of gestational age at delivery in hypertensive disorder of pregnancy

Introduction Pre-eclampsia (PE) is a multifactorial syndrome with different clinical phenotypes. To address this heterogeneity, the syndrome has been subdivided, according to gestational age (GA), into early and late PE, but this classification is not completely satisfying [1]. The most-studied clinical phenotype is that caused by shallow trophoblastic invasion of the spiral arteries, with oxidative stress and release of placental factors, which lead to endothelial dysfunction and organ damage [2]. Doppler velocimetry of the uterine arteries (UtA-PI) and measurement of vascular growth factors can be used to screen for early vascular damage, associated with reduced placental growth and intrauterine growth restriction (IUGR) later in pregnancy [3]. This clinical phenotype is predominant in cases prior to 34 wks of GA, thus early PE can be regarded as a proxy for cases affected by this sequence of events. Anyhow, the disease does not end at 34 wks of GA, and less severe cases represent a large proportion of syndromic manifestations of PE in late gestation [1]. Cases affected by hypertensive disorders of pregnancy (HDP) including PE are frequently found to have normal neonates and placentas [4]. In these cases, endothelial dysfunction can be ascribed to the placental pro-inflammatory response, dyslipidemia or other pro-inflammatory conditions resulting in low-grade inflammation due to pre-existing maternal conditions, such as maternal metabolic syndrome and cardiovascular risk factors. These risk factors are not suppposed to operate only from 34 wks of GA onwards. The aim of this presentation is to verify whether a classification based only on fetal abdomen circumference (AC) and UtA-PI is suitable for defining two internally homogeneous subgroups of women affected by HDP, characterized by different effects on fetal growth, as can be inferred from the outcome, in terms of birthweight conditional on GA, and whether GA at the onset of HDP has a role in determining fetal growth. Data and methods This multicentre study includes all eligible cases of HDP consecutively admitted over a 12-month period (from December 2013) to seven Italian tertiary referral centres for maternal-fetal medicine. To be eligible for inclusion, the clinical records of each patient had to report two “key items”: fetal assessment of abdomen circumference (AC) and mean UtA-PI. Women with multiple pregnancy and with chromosomal or structural fetal abnormalities were excluded. A total of 902 patients met these inclusion/exclusion criteria. Data concerning maternal characteristics, pregnancy complications and outcome were retrieved from the archived clinical records [5]. Pregnant women were grouped into two phenotypes, independently of GA at the onset of HDP. The first phenotype, which was named HDP-IUGR (n=124) [5], is characterized by maternal HDP associated with fetal AC at admission below the 5th centile of Todros et al reference [6], and an abnormal UtA-PI, defined as mean of right and left UtA-PI>95th centile of Gomez et al reference [7]. The second phenotype, which was named HDP-AGA, i.e. fetuses whose AC is appropriate-for-gestational age (n=205), is characterized by maternal HDP associated with fetal AC≥5th centile and UtA-PI≤95th centile. The group was then subdivided in early and late-onset HDP, depending whether GA at the onset of disease was 95th centile, or AC at admission. Results Figure 1 shows birthweight as a function of GA at delivery in neonates classified by sex, maternal phenotype and GA at the onset of HDP. When the model includes only the term accounting for the effect of sex, birthweight of males is estimated to be higher than that of females by 4% (95%CL: 1%, 7%; p=0.0158). The introduction of the term expressing the effect of phenotype into the model consistently improves the goodness of fit: the coefficient of determination increased from 0.86 to 0.91 and the residual standard deviation decreased from 366 to 294 g. The children of women with phenotype HDP-IUGR show a birthweight lower by 23% (95%CL: 20%, 26%; p. Conclusions The use of the parametric nonlinear model presented in this study yielded a synthetic and faithful description of the relationship between birthweight and GA in over 3 hundred babies born to women suffering with hypertensive disorders of pregnancy, and delivered between the 25th and the 42nd week of GA. Furthermore the model provided estimates of the separate effects that sex, HDP phenotype, and GA at the onset of HDP exert on birthweight. Among babies born to HDP mothers, birthweight of males was found to be higher than that of by 4%: an identical difference was observed in the reference set used to trace INeS charts [10], this suggest that HDP impairs fetal growth independently of sex. The babies born to mothers with phenotype HDP-IUGR were found to have birthweight lower by about 25% than those born to mothers with phenotype HDP-AGA: this is not unexpected, fetal weight being partly determined by abdomen volume. Nonetheless, since delivery occurred, on the average, 3 weeks after the assessment of abdomen circumference, such a difference suggests the persistence of conditions unfavourable to fetal growth up to end of pregnancy. The third hypothesis that the model could test was whether GA at the onset of HDP affects the severity of fetal growth impairment. In this case we would expect that birthweight of babies born to mothers with early onset of HDP is considerably lower. Results showed that fetal growth follows the same pattern, independently of GA at the onset of HDP. References 1. Verlohren S, Melchiorre K, Khalil A, et al. Uterine artery Doppler, birth weight and timing of onset of pre-eclampsia: providing insights into the dual etiology of late-onset pre-eclampsia. Ultrasound Obstet Gynecol 2014; 44:293-298. 2. Cindrova-Davies T. Gabor Than Award Lecture 2008: pre-eclampsia – from placental oxidative stress to maternal endothelial dysfunction. Placenta 2009; 30:S55–S65. 3. Levine RJ, Maynard SE, Qian C, et al. Circulating angiogenic factors and the risk of pre-eclampsia. N Engl J Med 2004; 350:672-683. 4. Conde-Agudelo A, Belizan JM. Risk factors for pre-eclampsia in a large cohort of Latin American and Caribbean women. BJOG 2000; 107:75-83. 5. Ferrazzi E, Zullino S, Stampalija T, et al. Bedside diagnosis of two major clinical phenotypes of hypertensive disorders of pregnancy. Ultrasound Obstet Gynecol 2016; 48:224-231 6. Todros T, Ferrazzi E, Nicolini U, et al. Fitting growth curves to head and abdomen measurements of the fetus: A multicentric study. J Clin Ultrasound 1987; 15:95-105. 7. Gomez O, Figueras F, Fernandez S, et al. Reference ranges for uterine artery mean pulsatility index at 11–41 weeks of gestation. Ultrasound Obstet Gynecol 2008; 32:128-132. 8. von Bertalanffy L. Quantitative laws in metabolism and growth. Q Rev Biol 1957; 32:217-231 9. Marubini E. Mathematical handling of long-term longitudinal data. In Human Growth, vol.1: Principles and Prenatal Growth; Falkner F, Tanner JM(eds). Plenum Press: New York 1978; 209-225 10. Bertino E, Spada E, Occhi L, et al. Neonatal anthropometric charts: the Italian neonatal study compared with other European studies. J Pediatr Gastroenterol Nutr 2010; 51:353-361.