Fetal Growth Restriction Prediction: How to Move beyond

The actual burden and future burden of the small-for-gestational-age (SGA) babies turn their screening in pregnancy a question of major concern for clinicians and policymakers. Half of stillbirths are due to growth restriction in utero, and possibly, a quarter of livebirths of low- and middle-income countries are SGA. Growing body of evidence shows their higher risk of adverse outcomes at any period of life, including increased rates of neurologic delay, noncommunicable chronic diseases (central obesity and metabolic syndrome), and mortality. Although there is no consensus regarding its definition, birthweight centile threshold, or follow-up, we believe birthweight <10th centile is the most suitable cutoff for clinical and epidemiological purposes. Maternal clinical factors have modest predictive accuracy; being born SGA appears to be of transgenerational heredity. Addition of ultrasound parameters improves prediction models, especially using estimated fetal weight and abdominal circumference in the 3rd trimester of pregnancy. Placental growth factor levels are decreased in SGA pregnancies, and it is the most promising biomarker in differentiating angiogenesis-related SGA from other causes. Unfortunately, however, only few societies recommend universal screening. SGA evaluation is the first step of a multidimensional approach, which includes adequate management and long-term follow-up of these newborns. Apart from only meliorating perinatal outcomes, we hypothesize SGA screening is a key for socioeconomic progress.


Introduction
e intrauterine environment influence on fetus development is a well-known determinant of individual's long-term health and quality of life. From the initial description of 23 infants being born at term weighing less than 2000 g, Warkany et al. [1] introduced the idea of "intrauterine growth retardation" (IUGR). Soon, they were followed by others [2][3][4]. ey considered IUGR "all conditions leading to a marked reduction in size during intrauterine life" [1], mainly represented by reduced birthweight. Although all of them have described pregnancies and infants with a wide variation of phenotype, with and without hypertensive syndromes or morphologic anomalies, for instance, the turning points were to consider the environment in which the fetus was developing and the placenta role in this process.
In fact, human development goes far beyond genetic inheritance. Lessons learned from pregnancies subjected to smoking [5,6] or intermittent fasting [7], for instance, show how intrauterine growth is adjustable. Posttranslational changes in small-for-gestational-age (SGA) infants [8] reinforce the thrifty phenotype hypothesis [9]. According to Hales and Barker, the nutritional deficiency, especially regarding amino acids supply, would decrease the pancreatic beta-cells function and induce changes of the muscular, hepatic, and adipose tissue systems functioning, for example [9]. ese newborns are at higher risk of neonatal morbidity and mortality [6,[10][11][12][13][14][15][16][17][18]; in adolescence and adulthood, they present worse neurodevelopment [19,20] and metabolic [21,22] and cardiovascular [23] adverse outcomes. On the contrary, the placenta-a shared organ by both the mother and the fetus-is responsible for adjusting maternal supply to the fetus demands. Since it is difficult to realize which are the normal placental functioning patterns and the optimal fetal growth, it is reasonable to use the birthweight as a measure of the intrauterine environment [24] and SGA newborns as surrogates for fetal growth restriction (FGR) [10][11][12][13]. erefore, considering the long latency of some events, such as cognitive delays and cardiovascular diseases, SGA has impacts of public health magnitude, especially in lowand middle-income countries (LMICs) [13,25]. In this review, we will discuss the importance of SGA screening in pregnancy and which are the best approaches and moments to perform it.

Why We Should Screen for Fetal
Growth Restriction? e identification of FGR as a distinct pathophysiological entity is merged with preterm birth history. In the first half of the 20 th century, gestational age at birth and birthweight concepts overlapped; the World Health Organization recommended a birth weight of 2500 g or less to characterize prematurity [26]. However, several authors and clinicians were intrigued by "pseudopremature" newborns, who would be in chronic suffering due to placental insufficiency and would benefit from earlier delivery [2][3][4]. Only in 1961, the terminology IUGR was first cited [1]. Apart from only birthweight (<2000 g), Warkany et al. suggested that preterm infants whose birthweight were 40% below the expected for a given gestational age should be considered IUGR. Two years later, Battaglia and Lubchenco proposed to use the birthweight as a proxy for intrauterine development [27] and this is still a common practice in the 2000s [10][11][12][13], due to difficulties in defining and measuring fetal growth [28][29][30].
Currently, the birthweight <10 th centile, either by population-based or customized charts, is the most accepted definition for SGA infants [28]. is mathematical threshold was initially chosen due to (i) the increased neonatal mortality observed in this group when compared to those born between the 10 th and the 90 th centiles and (ii) the agreement on the 10 th centile among studies up to the 1960s [27].
ere are concerns that some of these infants are "constitutionally small," not at higher risk of (neonatal) adverse outcomes, and lower limits for SGA, such as ≤5 th [31], ≤3 rd or even ≤2, and 3 rd centile [32], are considered by some researchers. However, little is still known about the long-term health endpoints of the "constitutionally small" newborns. erefore, the 10 th centile seems the most suitable cutoff for epidemiological and clinical purposes, and it is the adopted threshold in this review. e SGA prevalence varies according to the reference standards applied; it tends to be higher with customized curves [11,12,14]. Using population-based charts, live births between 19.3% [13] and 27% [25] in LMIC could have been classified as SGA in 2000s. A majority of them were term-SGA (98% and 95.6%, respectively). is turns SGA the most important pregnancy-related syndrome since other pathological conditions, such as pregnancy hypertension and preterm birth, have markedly lower prevalence [11,12,14]. It is interesting to note, however, that these "great obstetrical syndromes" may share pathophysiological pathways [33], and it is possible that SGA may represent an underlying condition for the other ones. e "great obstetrical syndromes" are related to defective deep placentation [33], and studies on placental biomarkers point in this direction [34,35]. Not surprisingly, pathological placental findings have been related to SGA pregnancies, especially vascular malperfusions lesions, infarction, and chronic villitis of unknown etiology [36][37][38][39]. Vascular-mediated changes (e.g., decidual vasculopathy and single or multiple infarctions) usually coexist with Doppler (uterine (UtA), umbilical (UA), or middle cerebral (MCA) arteries) [37,39] or biochemical abnormalities (such as low levels of placental growth factor (PlGF) [35] or alphafetoprotein (AFP): pregnancy-associated plasma protein-A (PAPP-A) ratio >10 [34]). ese pathological and functional observations are similar to those found in pregnancies affected by hypertensive disorders, preterm deliveries, and stillbirth [40][41][42][43][44], then possibly reflecting an elementary chronic hypoxia mediating these outcomes.
Unfortunately, the higher risk of mortality goes beyond the neonatal period. Data from Sweden show a hazard ratio (HR) of 1.37 (95% CI 1.28-1.47) of death up to 18 years old, which increased to 2.61 (95% CI 2.19-3.10) for those neonates born <3 rd centile [59]. Additionally, growth restriction is 2 e Scientific World Journal associated with a lower Bailey score, especially in communication skills domain [19], sleep disorders [52], and hyperactivity [56]. If SGA fetuses experience any degree of brain-sparing effect, the delayed motor skills and cognitive development are even more pronounced [19,51]. Regarding metabolic repercussions, insulin and insulin resistance index (HOMA IR) are higher in SGA children at 6-8 years old and those born <3 rd centile also have higher levels of leptin [22]. Evidence from adults exposed to famine in utero shows increased odds for metabolic syndrome [21] and obesity [60] in SGA newborns, perhaps in a sex-specific manner, depending on childhood nutritional parameters (especially weight gain velocity). Proportionate biometric measurements at birth were the initial observations of Barker et al. who related the ponderal index, head circumference, and birthweight <2495 g to cardiovascular mortality [23]. Although maternal undernourishment is not synonymous of SGA infant and considering that the birthweight approach has changed over time, these findings mean that adequate fetal development is the standpoint for long-term health.
ere is greater visceral fat thickness (in women) [61], higher fat-free soft tissue mass [62], and increased trunk and abdominal fat mass proportion (of both sexes) [63] in adults born SGA.
ese epidemiological data ground current theories of epigenetic modifications in SGA infants, leading to enriched (i.e., with increased DNA methylation) pathways involved with fat, sugar, and protein metabolism [8].
erefore, timely recognition of SGA-still in pregnancy-is a real concern for obstetricians, perinatologists, health workers, and policymakers. Unfortunately, only a small proportion of SGA babies are suspected before birth [18,57], leading to a lack of appropriate short-and longterm follow-up of these newborns. SGA suspicion will provide adequate management of the mother and fetus/ newborn, including referencing to a specialized facility for antenatal care and delivery and individualized follow-up in childhood, adolescence, and adulthood.

When and How We Should Screen for Fetal
Growth Restriction?

Clinical Factors.
Clinical risk assessment is the first approach to antenatal care. A detailed maternal history at booking can identify several risk factors and guide referencing to tertiary care facilities. Single maternal clinical factors demonstrate poor prediction accuracy (Table 2), and, as a result, are generally considered in a multidimensional model. Smoking, although less prevalent in the early years of the 21 st century, still demonstrates effects on fetal growth [5,6,48] and is the most common maternal variable to compose a prediction model. Lower maternal stature and weight appear associated with SGA in some studies [11,14,48] but showed only 43% and 73% of sensitivity, respectively [64]. Body mass index (BMI) and maternal weight gain throughout pregnancy demonstrate an area under the (AUC) receiver operating characteristic (ROC) curve of 0.56 and 0.60, respectively [64]. e performance of symphysial-fundal height (SFH) measurement in predicting SGA newborns increases with gestational age [68], but it is not different to Leopold's maneuvers (RR1. 32, 95% CI 0.92-1.90) [69]. However, since it is inexpensive and already part of the routine obstetrical examination, Cochrane reviewers advise its use and health professionals should associate it with some other technique or evaluation of fetal growth.
Other maternal factors have been combined differently, evidencing how SGA syndrome can be heterogeneous in distinct settings. In a multicenter international nulliparous cohort, a family history of coronary heart disease, maternal birthweight <3000 g, infertility, college student, smoking at the 2 nd trimester, proteinuria, daily vigorous exercise, and diastolic blood pressure ≥80 mmHg, combined with the protective factors rising random glucose, recreational walking (≥4x/week), and Rhesus negative blood group, provided an AUC of 0.63 [6]. is same AUC (0.66, 95% CI 0.61-0.70) was achieved by combining maternal age and height, smoking, previous SGA infant, and chronic hypertension in Spain [70]. In the United Kingdom, a logistic regression model included maternal height, weight, parity, ethnic background, smoking, and previous history of preeclampsia or SGA [71]. In this model, maternal factors evaluation between 35 and 37 w have had similar AUC for delivery within two weeks (0.744; 95% CI 0.731-0.756) and term delivery (0.712; 95% CI 0.700-0.725) for SGA without preeclampsia.
PlGF has consistently lower levels in SGA pregnancies, in 2 nd and 3 rd trimesters [82][83][84], especially for BW < 5 th or <10 th centiles. For higher sFlt-1/PlGF ratios, there is better AUC for preeclampsia-associated SGA [40,41]. ese findings point in the direction of angiogenesis-mediated pathophysiology of SGA, especially when there are Doppler abnormal parameters [37]. Unfortunately, PlGF shows poor accuracy to be implemented in clinical practice: the combined AUC was 0, 66 (95% IC 0, 44-0, 87) for FGR prediction [65]. Perhaps, this finding is due to the diverse PlGF measurements and FGR definitions used by the studies included in the systematic review, which considered either After all, better accuracy was achieved by combining multiple maternal, ultrasonographic, and biochemical clinical factors. In an international cohort of nulliparous women [79], PlGF has had an AUC of 0.84 (95% CI 0.78-0.89) for hypertensive-SGA when combined with smoking, proteinuria, uterine artery Doppler, PAPP-A, and triglycerides. In the 2 nd trimester (19-24 w), PlGF and AFP, combined with maternal factors and fetal biometry, made up an AUC of 0.96 for birth below 32 weeks in SGA newborns [31].

Conclusions
Fetal growth restriction is related to adverse outcomes in the perinatal period, childhood, and adulthood; the estimated actual burden of SGA [13,25] might be even higher in the next few years. Starting antenatal care at early pregnancy leads to adequate risk management and additional evaluation assessment, with US or biomarkers. e "inverted pyramid" of prenatal care claims attention to the early pregnancy risk evaluation [85], and we strongly believe screening is the first step towards a better disease diagnosis and management. Screening for FGR is a major cornerstone for coordinating care from pregnancy to the postpartum period, which affects both maternal and fetal/neonatal outcomes [86]. e low velocity in which stillbirth and neonatal death rates have decreased in the past 30 years is an "unfinished agenda" [86].
Although the cost-effectiveness of short-term pregnancy-related adverse outcomes is still a matter of debate [87], little is known about the future consequences of a health policy devoted to primary prevention of pregnancyassociated illness in a long-term [49,59,88]. On the contrary, the lack of definition of a high-risk group of women that could benefit from a more directed approach delays scientific and clinical evaluation of SGA. As maternal factors have a different magnitude between settings and placental biomarkers are not a reality in most LMIC countries, currently, the 3 rd trimester US seems the best approach for SGA prediction [44]. In near future, we envision an integrated approach of pregnant women at booking [85], aiming a transgenerational [49] effect of long-term health, both at individual and populational levels.

AC:
Abdominal circumference AFP: Alpha-fetoprotein AUC: Area under the curve BW: Birth weight CPR: Cerebral-placental ratio DOR: Diagnostic odds ratio DR: Detection rate EFW: Estimated fetal weight FGR: Fetal growth restriction FL: Femur length FPR: False positive rate HC: Head circumference hPL: Human placental lactogen HR: Hazard ratio IUGR: Intrauterine growth restriction LMIC: Low-and middle-income countries MCA: Middle cerebral artery MRI: Magnetic resonance imaging NT: Nuchal translucency PAPP-A: Placental protein-A PlGF: Placental growth factor POP: Pregnancy outcome prediction RI: Resistance index ROC: Receiver operator characteristic SFH: Symphysial fundal height SGA: Small for gestational age UA: Umbilical artery US: Ultrasound UtA-PI: Uterine artery pulsatility index UVBF: Umbilical vein blood flow VEGF: Vascular growth factor.

Disclosure
is manuscript is part of the PhD thesis of Debora F Leite under the tutorial of Jose G. Cecatti, presented to the Postgraduate Program of Obstetrics and Gynecology from the University of Campinas, Brazil. e content is solely the responsibility of the authors and does not necessarily represent the official views of CAPES. It did not influence the content of the manuscript. CAPES had no role in the authors' decision of drafting or submitting this manuscript.

Conflicts of Interest
e authors declare that there are no conflicts of interest at all.

Authors' Contributions
DFBL has proposed the review and drafted the first manuscript. JGC has supervised and checked the drafting. Both authors have read and agreed with this submission.