protein levels in maternal serum have been associated with FGR caused by placental insufficiency, including lower levels of PlGF and increased levels of soluble fms‐like tyrosine kinase‐1 (sFLT‐1).48–50 Placentas associated with FGR tend to be small and are associated with a range of pathologies that may be a cause or consequence of impaired maternal or fetal vascular supply.51 These findings can include increased syncytial knots, intervillous fibrin deposition, villous infarcts, villous agglutination, distal villous hypoplasia, villous hypermaturation, marginal abruption, thrombosis, and chorangioma, among others. Compromised placental vascularization can also be seen as microvascular regression (particularly at the placenta periphery).52 However, such findings can occur in the absence of FGR, and further research is needed to clarify these relationships.
Genetic causes of fetal growth restriction
Many genetic conditions can be associated with FGR, but these are individually rare. The only relatively common known genetic cause of FGR is CPM, typically occurring as trisomy in some or most cells from the placenta, with a predominantly normal diploid fetus. CPM is present in approximately 10 percent of placentas associated with FGR pregnancies (after exclusion of constitutional chromosomal abnormalities).53–56 In the past, CPM was generally only diagnosed prenatally when chorionic villus sampling (CVS) was performed, with trisomy 16 being one of the most common trisomies involved when there is poor fetal growth. CPM can also be detected by noninvasive prenatal testing (NIPT), which is based on genetic analysis of DNA that is largely derived from the placental trophoblast and increasingly used routinely in prenatal assessment.57, 58
Although there is no identifiable placental pathology characteristic of placental (mosaic or nonmosaic) trisomy, certain features are more likely to be present. Placentas tend to be small, although fetal–placental weight ratio is often preserved.59 In early gestation, there may be trophoblastic irregularities reminiscent of hyperplasia, with a lacey appearance, and increased invaginations or inclusions of trophoblastic epithelium.60 In some cases of trisomy 16, ultrasound examination shows cystic changes, raising the clinical possibility of partial molar pregnancy but those changes are often not appreciated until after delivery of the placenta.61 Although not common, there are reports of other trisomies with a histological partial hydatidiform mole (PHM)‐like phenotype, including trisomies 7, 15, and 22.62
The frequency with which CPM is identified as the explanation for FGR may be dependent on the criteria used to diagnose FGR and the prevalence of environmental risk factors for FGR (smoking, insufficient maternal nutrition, etc.) in the population. In addition, CPM would be considerably more likely when placental insufficiency occurs in the presence of older maternal age, a strong risk factor for trisomy. However, CPM and placental pathology can occur in the absence of IUGR, and further research is needed to clarify these relationships.
Amniocentesis for chromosome analysis in the second trimester may be recommended if fetal growth is discrepant for gestational age. Similarly, FGR detected in the third trimester would lead to serious consideration of an amniocentesis, since discovery, for example, of a trisomy, may well influence the mode and management of delivery. Chromosome analysis of the placenta at term may be warranted if there are concerns for the baby's development, especially if cord blood is not obtained. Placental mosaicism may be associated with fetal uniparental disomy or the trisomy may not be entirely confined to the placenta (i.e. low level fetal mosaicism can be present). For example, trisomy 7 in the placenta can be associated with fetal maternal uniparental disomy (UPD) 7, which can cause Silver–Russell syndrome (SRS),63, 64 a condition characterized by severe FGR and dysmorphic facial features. If placental investigation is undertaken, it is important that specimens from multiple sites of the placenta are tested to diagnose the underlying mosaicism, as it is often present in only a subset of sampled sites.53, 65
Developmental considerations in confined placental mosaicism
Most of our knowledge of CPM stems from cases based on CVS. CPM is detected in 1–2 percent of pregnancies undergoing CVS, most commonly in the form of trisomy mosaicism.57, 66–68 Low levels of trisomy confined to the placenta typically do not have a significant effect on fetal growth and development. Although high levels of trisomy will generally affect placental function, chromosomal abnormalities that may be lethal to the fetus are often tolerated to some degree when confined to the placenta. For example, trisomy 16 can be present in very high levels in the placenta as long as the fetus is entirely (or mostly) diploid.69 Placentas associated with CPM16 are small and there is almost always FGR, as well as an increased risk for malformations, maternal PE, and PTB.69, 70 Importantly though, the babies born in conjunction with CPM16 typically do quite well after birth once separated from the abnormal placenta.71, 72
The level and distribution of the abnormal cells depends, in part, on how and when the mosaicism arose in development.73 Selection may also favor contribution of diploid cells to the embryo.74 Furthermore, some trisomies may be tolerated in the inner core of the villi, but not in the trophoblast component (e.g. trisomy 8). The fact that trisomy may be present at low levels or patchy in its distribution55, 65 (due to independent clonal development of each of the ∼50–70 villous trees) can lead to apparent “false‐positive” and “false‐negative” diagnostic results using CVS. CPM has been identified using NIPT testing. The level of trisomy detected by NIPT, compared with CVS, should represent more of an average across the placental trophoblast.75, 76 Surprisingly though, NIPT has also missed CPM even when high levels of placental trisomy were present.77, 78 This is possibly explained by trisomy‐specific differences in trophoblast shedding into maternal circulation76 or by absence of the trisomy from incorporation into the syncytiotrophoblast and EVTs, from which the “fetal” DNA likely originates.
Imprinting and fetal growth restriction
Disruption of imprinted genes, those differentially expressed depending on parental origin, has been investigated in FGR due to their important role in the placenta and in regulation of growth. Classic examples of the growth effects of imprinted genes in humans are the Beckwith–Wiedemann syndrome (BWS) and SRS that are associated with overgrowth and growth restriction, respectively.79 BWS is caused by alterations of the 11p15.5 region, which includes at least eight imprinted genes organized into two domains. Growth restriction in SRS has been associated with epimutations leading to reduced IGF2 expression, as well as uniparental disomy and chromosomal rearrangements involving chromosome 7 including the MEST/PEG1 and/or the GRB10 regions.80, 81
Reduced fetal growth has been associated with placental changes in DNA methylation at several imprinted differentially methylated regions (DMRs), including those associated with PLAGL1, PEG10, H19/IGF2, and ZNF331.82–90 However, the reported changes in DNA methylation at imprinted DMRs have typically been of small magnitude and have not always been reproduced in other studies. Many studies have used SGA (<10th percentile) as a surrogate for FGR, a potential source of confusion, and many affected pregnancies deliver preterm or in association with PE, which could confound results. Furthermore, it is not clear if altered expression of imprinted genes is a cause of FGR or a compensatory effect of other defects. Larger sample sizes with stricter phenotypic criteria may help clarify the relationship between altered placental imprinting and FGR.
Preeclampsia
Maternal PE frequently occurs with FGR and shares many similar pathologic features.91, 92 PE is part of the spectrum of hypertensive disorders
