INTRODUCTION — When ultrasound examination suggests fetal growth restriction (FGR), prenatal care involves accurately determining gestational age, confirming the suspected diagnosis, determining the cause and severity of FGR, counseling the parents, closely monitoring fetal growth and well-being, and determining the optimal time for and route of delivery. FGR resulting from intrinsic fetal factors, such as aneuploidy, congenital malformations, or infection, carries a guarded prognosis that often cannot be improved by any intervention. FGR related to uteroplacental insufficiency has a better prognosis, but the risk for adverse outcome remains increased.

This topic will discuss the evaluation and management of singleton pregnancies complicated by FGR. The diagnosis of FGR and outcome of affected infants are reviewed separately. (See "Fetal growth restriction: Screening and diagnosis" and "Infants with fetal (intrauterine) growth restriction".)

FGR in twin pregnancies is also reviewed separately. (See "Selective fetal growth restriction in monochorionic twin pregnancies" and "Twin pregnancy: Routine prenatal care", section on 'Screening for fetal growth restriction and discordance'.)

INITIAL APPROACH

Confirm the diagnosis — The diagnosis of FGR is based on discrepancies between actual and expected sonographic biometric measurements for a given gestational age. Traditionally, it has been defined as <10^th percentile weight for gestational age on a singleton growth curve, as this establishes the diagnosis as being small for gestational age (SGA). An abdominal circumference below the 10^th percentile for gestational age can also be used to define FGR [1].

In our practice, when a fetus <10^th percentile weight for gestational age is identified, we monitor fetal growth and fetal physiology over time. Normal anatomy, normal growth trajectory, normal Doppler velocimetry of the umbilical artery and/or middle cerebral artery, and normal amniotic fluid volume suggest a constitutionally small fetus or a fetus that is minimally impacted from uteroplacental insufficiency or other pathologic factors that impair fetal growth. The importance of normal growth trajectory in predicting a favorable outcome was supported by a study of 216 fetuses with suspected FGR (defined as estimated fetal weight [EFW] <10^th percentile) that showed those with growth curves parallel to the growth curve of fetuses with EFW >10^th percentile had normal outcomes [2].

A weight <10^th percentile definition is clinically practical, but it alone does not distinguish the constitutionally small fetus that achieves its normal growth potential and is not at increased risk of adverse outcome from the similarly small fetus whose growth potential is restricted and is at increased risk of perinatal morbidity and mortality. This definition also does not account for the fetus who is not SGA but is not achieving its growth potential because of intrinsic or environmental restrictions to normal growth. Distinguishing the constitutionally small fetus from the fetus with pathologic growth restriction is the first challenge for the clinician. This knowledge can help to avoid unnecessary interventions for pregnancies with a constitutionally small fetus and lead to appropriate monitoring and delivery timing of the growth-restricted fetus.

Characteristics that support a diagnosis of a constitutionally small fetus include:

●Modest smallness (ie, estimated weight between the 5^th and 10^th percentiles)

●Normal growth velocity across gestation

●Normal physiology (ie, normal amniotic fluid volume and umbilical artery Doppler)

●Abdominal circumference growth velocity above the 10^th percentile

●Appropriate size in relation to maternal characteristics (height, weight, race/ethnicity)

Maternal characteristics have a major influence on fetal growth potential. For example, when race/ethnicity was taken into account, the 5^th percentiles for White, Hispanic, Black, and Asian populations at 39 weeks of gestation were 2790, 2633, 2622, and 2621 grams, respectively, in a prospective study of over 2300 healthy women with low-risk, singleton pregnancies from 12 medical centers in the United States [3]. Using biometric standards derived solely from the group of White fetuses, as many as 15 percent of the fetuses from other groups would have been classified as growth restricted (<5^th percentile).

Using a lower threshold to define FGR may help distinguish the small fetus at increased risk of adverse outcome from the small fetus at low risk. As discussed above, FGR has been defined as <10^th percentile weight for gestational age, even though most fetuses with weights between the 5^th and 10^th percentiles are constitutionally small and have normal neonatal outcomes [4,5]. Use of a lower threshold for defining pathologic FGR is supported by several studies.

●In the Prospective Observational Trial to Optimize Pediatric Health (PORTO), which included over 1100 pregnancies between 24+0 and 36+6 weeks of gestation with nonanomalous fetuses and EFW <10^th percentile on ultrasound examination, only 2 percent of fetuses at the 3^rd to 10^th percentile (5 of 254) experienced adverse perinatal outcome, while 6.2 percent of those <3^rd percentile (51 of 826) had an adverse outcome and all eight mortalities were in this group [4]. The combination of EFW <3^rd percentile and abnormal umbilical artery Doppler was a strong and consistent predictor of adverse outcome: 16.7 percent of these fetuses developed intraventricular hemorrhage, periventricular leukomalacia, hypoxic ischemic encephalopathy, necrotizing enterocolitis, bronchopulmonary dysplasia, sepsis, or death. Abnormal Doppler in this study included both pulsatility index >95^th percentile and absent or reversed end-diastolic flow. An abnormal growth trajectory over time was another factor that predicted perinatal complications (eg, preterm birth, preeclampsia, neonatal morbidity) [6].

●In a retrospective study, composite fetal morbidity at <5^th percentile versus 5^th to 10^th percentile was 39 and 13 percent, respectively (odds ratio 2.41, 95% CI 1.5-3.8) [5].

Although using <5^th percentile as a threshold for FGR captures the group of fetuses at highest risk for adverse outcome, those at the 5^th to 10^th percentile still need to be monitored closely since not all are constitutionally small.

Determine the cause — The genetically predetermined growth potential of the fetus can be impaired as a result of maternal, placental, or fetal processes (table 1). Despite the following history/physical examination, pregnancy imaging, and laboratory evaluations, the reason(s) for growth impairment cannot always be determined antenatally.

Maternal examination — To determine the cause of FGR, a complete history and physical examination is performed to assess for maternal disorders that have been associated with restricted fetal growth (table 1). Although preeclampsia and chronic hypertension are commonly considered causal factors, a study of hypertensive pregnant women by the National Institute of Child Health and Human Development (NICHD) found that severe preeclampsia, particularly with onset <34 weeks of gestation, was associated with growth restriction as early as 22 and 23 weeks, but mild preeclampsia and chronic hypertension were not [7].

Fetal survey — A detailed fetal anatomic survey should be performed in all cases since approximately 10 percent of FGR is accompanied by congenital anomalies [8] and 20 to 60 percent of malformed infants are SGA [9]. Anomalies associated with FGR include omphalocele, gastroschisis, diaphragmatic hernia, skeletal dysplasia, and some congenital heart defects. Although omphalocele and gastroschisis can be associated with FGR, the abdominal circumference is distorted by the anomaly in these cases, limiting the utility of biometric measurements in diagnosis of FGR. Congenital portosystemic shunts are rare, but have been associated with FGR [10].

Triploidy of maternal origin is characterized early, severe asymmetrical FGR and triploidy of paternal origin is characterized by mild FGR with either proportionate head size or slight microcephaly [11]. Most fetuses have structural anomalies, most commonly involving the brain, heart, and limbs.

A fetal echocardiogram is indicated if results of an expert (level II) ultrasound examination suggest any uncertainty that the heart is normal.

Fetal genetic studies — Fetal genetic studies are indicated in any of the following settings because of the increased risk of an abnormality:

●FGR that is all of the following: Early (<24 weeks), severe (<5^th percentile), and symmetrical.

Although the finding of symmetrical FGR prior to 24 weeks of gestation is associated with a high risk of aneuploidy, this is no longer the case later in gestation [12]. After 24 weeks, we do not screen for fetal genetic abnormalities if anatomy is normal and FGR is asymmetric since the yield would be low, the etiology is most likely a maternal or placental disorder, and pregnancy termination is generally not an option. In a systematic review of 14 observational cohort studies including 874 apparently isolated FGR cases, the mean rate of chromosome anomalies was 6.4 percent (range 0 to 26 percent), and no abnormal karyotypes were found in the two studies of apparently isolated FGR diagnosed in the third trimester (32 pregnancies) [13].

●FGR with major fetal structural abnormalities.

●FGR with soft ultrasound markers associated with an increased risk of aneuploidy, such as thickened nuchal fold/choroid plexus cyst and abnormal hand positioning. (See "Prenatal genetic evaluation of the fetus with anomalies or soft markers", section on 'Approach to the evaluation of the fetus with "soft markers" and no structural anomalies'.)

●Placental findings suggestive of a partial molar pregnancy. Triploidy of paternal origin is characterized by an enlarged placenta with cystic changes; the enlarged placenta is a partial mole in 60 to 80 percent of cases [11]. Triploidy of maternal origin results in earlier and more severe FGR, and the placenta is thin.

The American College of Obstetricians and Gynecologists suggests genetic counseling and offering diagnostic testing for patients with diagnosis of:

●FGR before 32 weeks or

●FGR in combination with polyhydramnios or fetal malformation [14].

FGR with a normal microarray could be due to a single gene disorder, particularly if FGR is severe (<1^st percentile) and/or associated additional ultrasound findings (eg, short long bones; microcephaly; cardiac defects; relative macrocephaly; abnormalities of the face, hands, and/or genitalia). Gene panels and whole exome sequencing for evaluation of FGR have not been validated, have low diagnostic yield, and are not recommended [12]. However, testing for a single gene disorder is reasonable with appropriate pre- and post-test counseling in cases with high suspicion (eg, positive family history for a single gene disorder, skeletal dysplasia). Consultation with a genetics professional is advised.

Most cases of confined placental mosaicism (CPM) do not result in FGR, but CPM is detected after delivery in approximately 10 percent of placentas associated with otherwise idiopathic FGR [15-17]. CPM is associated with an increased frequency of placental infarcts and decidual vasculopathy, presumably because some abnormal karyotypes can adversely affect placental function. In pregnancies complicated by FGR, approximately one-third of placentas with an infarct have underlying CPM [18]. We do not perform chorionic villus sampling in the second or third trimester to identify this abnormality antepartum because antenatal diagnosis would not change pregnancy management and the procedure is associated with a small risk of pregnancy complications. (See "Chorionic villus sampling", section on 'Confined placental mosaicism'.)

Trisomy 18 — Ultrasound examination has high sensitivity for identifying trisomy 18, as high as 100 percent, when performed by an experienced ultrasonographer at 19 to 20 weeks of gestation in a fetus with multiple structural anomalies characteristic of the syndrome. (See "Sonographic findings associated with fetal aneuploidy", section on 'Trisomy 18 (Edward syndrome)'.)

If ultrasound examination strongly suggests trisomy 18 (positive predictive value depends on the number and types of ultrasound findings), we use a cell-free DNA test to screen for trisomy 18. If cell-free DNA testing is positive for trisomy 18 in this specific setting, we do not believe amniocentesis is essential to confirm the diagnosis. We counsel the patient regarding false positives and negatives and will perform amniocentesis for confirmation if the patient chooses this approach. In a 2016 meta-analysis, the positive predictive value of cell-free DNA for trisomy 18 in a general obstetric population and a high-risk population was 37 and 84 percent, respectively [19], and would be higher when associated with ultrasound findings characteristic of the syndrome. However, if a karyotype has not been obtained, it should be obtained for confirmation after delivery (or termination) to determine whether the trisomy was the result of a parental balanced translocation, as this will impact the recurrence risk in future pregnancies.

If cell-free DNA is negative for trisomy 18, we perform genetic amniocentesis to obtain amniocytes for microarray analysis; microarray has a significantly higher diagnostic yield than conventional karyotype [20,21]. The most common pathogenic copy number variants are 22q11.1 duplication, Xp22.3 deletion, and 7q11.23 deletion [12]. We use the same approach if ultrasound examination reveals abnormalities for which detailed genetic analysis is indicated.

In most clinical settings, the combination of positive cell-free DNA results and ultrasound findings do not provide sufficient diagnostic certainty to allow omission of fetal karyotype/microarray by genetic amniocentesis if pregnancy termination is planned because of suspected aneuploidy alone. In addition, a negative cell-free DNA test does not exclude the possibility of a pathogenic chromosome abnormality not targeted by the test but associated with FGR. (See "Sonographic findings associated with fetal aneuploidy", section on 'Trisomy 13 (Patau syndrome)' and "Prenatal diagnosis of chromosomal imbalance: Chromosomal microarray", section on 'Benefits compared with conventional karyotype'.)

Work-up for infection — Infections such as cytomegalovirus (CMV), toxoplasmosis, rubella, varicella, herpes simplex, syphilis, and zika virus may be associated with FGR. Sonographic markers for fetal infection are often nonspecific and include echogenicity and calcification of the brain and/or liver, and hydrops. (Refer to individual topic reviews on specific fetal infections during pregnancy for a description of the ultrasound findings)

We do not perform routine TORCH serology in the evaluation of isolated FGR. We obtain serology studies when ultrasound findings are suggestive of an intrauterine infection (ie, sonographic markers of fetal infection are present in addition to FGR) or a careful maternal history and physical examination suggest the possibility of maternal infection and vertical transmission. If the patient chooses to have diagnostic fetal testing for isolated FGR, we perform amniocentesis for polymerase chain reaction (PCR) of amniotic fluid for CMV.

Use of maternal TORCH serology should be limited because serology is imprecise for determining the time of occurrence of the primary infection and can potentially lead to unnecessary invasive procedures with associated parental anxiety and cost. Due to its low yield, the Society for Maternal-Fetal Medicine recommends against TORCH serology for isolated FGR and recommends PCR of amniotic fluid for CMV in women with unexplained FGR who elect to have diagnostic testing [1]. In a systematic review including 2538 pregnancies in eight studies, 496 TORCH serologies were performed in patients with FGR (EFW <10^th percentile) [22]. Of this group, 12 were positive (2.4 percent, 95% CI 1.4-4.2%) and 10 congenital infections were detected (eight CMV, two parvovirus B19), but only 2 of the 10 fetuses had isolated FGR (CMV infection was found in both cases). The studies included in this review did not differentiate between early and late FGR and there were no cases of rubella or herpes simplex virus.

Malaria in pregnancy can also cause FGR and should be considered in endemic areas. The fetal effects of malaria in pregnancy are reviewed separately. (See "Malaria in pregnancy: Epidemiology, clinical manifestations, diagnosis, and outcome".)

Maternal COVID-19 does not appear to be associated with an increased prevalence of FGR [23,24]. However, data on perinatal outcomes when the infection is acquired in early pregnancy are limited, and any condition that results in prolonged maternal hypoxia or placental dysfunction places the fetus at risk for growth restriction. (See "The placental pathology report", section on 'Chronic histiocytic intervillositis'.)

Work-up for antiphospholipid syndrome — Antiphospholipid antibody syndrome (APS) is an acquired thrombophilia. Although early-onset placental insufficiency is one of the clinical criteria for the diagnosis of APS by expert consensus (table 2), a link between antiphospholipid antibodies alone and FGR has not been established, and there is insufficient evidence to support screening all women with FGR for these antibodies [25]. Indications for screening include a past history of fetal loss and prior unexplained arterial or venous thromboembolism. (See "Antiphospholipid syndrome: Pregnancy implications and management in pregnant women", section on 'Adverse pregnancy outcomes defining APS' and "Diagnosis of antiphospholipid syndrome", section on 'When to suspect the diagnosis' and "Diagnosis of antiphospholipid syndrome", section on 'Diagnostic evaluation'.)

Assessment for inherited thrombophilic disorders is not recommended as evidence for an association between the inherited thrombophilias and FGR is weak. (See "Inherited thrombophilias in pregnancy", section on 'Adverse pregnancy outcome risk'.)

PREGNANCY MANAGEMENT

Overview — The following discussion applies to pregnancy management of FGR in structurally and chromosomally normal fetuses. Most of these cases are caused by uteroplacental insufficiency. Management of FGR associated with congenital or chromosomal anomalies depends on the specific abnormality and is beyond the scope of this review.

The optimal management of the pregnancy with suspected growth restriction related to uteroplacental insufficiency consists of serial ultrasound evaluation of:

●Fetal growth velocity

●Fetal behavior (biophysical profile [BPP])

●Impedance to blood flow in fetal arterial and venous vessels (Doppler velocimetry)

Specific tests, timing, and frequency are discussed in the sections below. (See 'Fetal weight and amniotic fluid assessment' below and 'Frequency of nonstress tests and biophysical profiles' below and 'Our approach' below.)

These represent the key elements of fetal assessment and guide pregnancy management decisions. The purpose is to identify those fetuses who are at highest risk of perinatal demise and who may benefit from delivery. Observational data suggest that adoption of a protocol to detect and manage FGR can reduce stillbirth [26].

The temporal sequence of biophysical and Doppler changes in the peripheral and central circulatory systems of the growth-restricted fetus are summarized below and in the figure (figure 1) [27-31]. Progression through the entire sequence does not always occur before delivery; Doppler abnormalities in some growth-restricted fetuses progress slowly or not at all or along a different pathway [32]. The sequence is most likely to progress when FGR and Doppler abnormalities are identified in the second trimester and the Doppler indices worsen within the first two weeks of Doppler monitoring [33]. Progressive Doppler changes do not always occur because FGR is not a homogeneous entity and different phenotypes of FGR behave differently. For example, FGR due to preeclampsia will have a distinct clinical course that is different from idiopathic FGR [34].

The general sequence of Doppler and biophysical changes in FGR is:

●A reduction in umbilical venous flow is the initial hemodynamic change. Venous flow is redistributed away from the fetal liver and towards the heart. Liver size decreases, causing a lag in fetal abdominal circumference, which is the first biometric sign of FGR.

●Umbilical artery Doppler index increases (diminished end-diastolic flow) due to increased resistance in the placental vasculature.

●Middle cerebral artery (MCA) Doppler index (eg, pulsatility index) decreases (increased end-diastolic flow), resulting in preferential perfusion of the brain (brain-sparing effect).

●Increasing placental vascular resistance results in absent and then reversed end-diastolic flow in the umbilical artery.

●MCA peak systolic velocity increases secondary to an increase in the PCO₂ and decrease in the PO₂ in blood delivered to the fetal brain [35].

●MCA pulsatility index (MCAPI) normalizes or abnormally increases as diastolic flow falls due to loss of brain-sparing hemodynamic changes.

●As cardiac performance deteriorates due to chronic hypoxia and nutritional deprivation, absent or reversed end-diastolic flow in the ductus venosus and pulsatile umbilical venous flow may develop.

●Lastly, tricuspid regurgitation and reversed flow at the aortic arch develop, which can be preterminal events.

Near the end of this sequence, biophysical changes usually become apparent: The nonstress test (NST) becomes nonreactive, the BPP score falls, and late decelerations accompany contractions. However, the cardiovascular (Doppler) and behavioral (BPP) manifestations of fetal deterioration in FGR fetuses can occur largely independent of each other, resulting in discordant Doppler and BPP findings [36].

The figures (figure 2 and figure 3) show the Doppler changes observed from the time the diagnosis is made until delivery or fetal demise in pregnancies with idiopathic FGR (ie, placental insufficiency with no other maternal or fetal pathology) and those with FGR with preeclampsia diagnosed at <32 weeks of gestation.

Ambulatory monitoring — Women with pregnancies complicated by FGR may maintain normal activities and are usually monitored as outpatients. There are no data on which to base indications for hospitalization. We and other experts [37] consider hospitalization for selected women who need daily or more frequent maternal or fetal assessment (eg, daily BPP score because of reversed diastolic flow). Hospitalization provides convenient access for daily fetal testing and allows prompt evaluation and intervention in the event of decreased fetal activity or other complications, but there is no evidence that hospitalization or bed rest improves fetal growth or outcome [38].

Although decisions about selecting women for ambulatory versus in-hospital care should be made on a case-by-case basis, we generally admit women with absent/reversed flow of the umbilical artery and estimated fetal weight (EFW) >350 grams. This cutoff is based on our experience of good short- and long-term outcomes in such cases. However, the cutoff weight for monitoring varies among institutions.

Fetal weight and amniotic fluid assessment — Fetal weight estimates are calculated using various published equations and formulae. The computed weight is then plotted on a population-based or customized growth curve, which allows the clinician to determine when the EFW is below the 10^th percentile (table 3) and to monitor growth velocity [39]. Persistent growth deficiency in multiple examinations over many weeks strengthens the likelihood of FGR. Conversely, normal growth velocity in a small fetus suggests a constitutionally small but normal fetus.

Frequency — Serial sonograms are generally obtained at three- to four-week intervals to ascertain the growth velocity when the fetus has mild FGR (eg, EFW near the 10^th percentile, normal amniotic fluid volume, normal Doppler findings); a three-week interval is appropriate for the fetus with features of moderate or severe disease (eg, EFW ≤5^th percentile, oligohydramnios, abnormal Doppler findings). (See "Fetal growth restriction: Screening and diagnosis", section on 'Customized growth curve' and "Prenatal assessment of gestational age, date of delivery, and fetal weight", section on 'Sonographic assessment of gestational age'.)

Doppler velocimetry — Doppler velocimetry of the umbilical artery is an excellent tool for fetal assessment in FGR when the etiology is placental dysfunction related to progressive obliteration of the villus vasculature. As described above, placental vascular changes lead to fetal hemodynamic changes that can be evaluated by umbilical artery Doppler. The most common Doppler indices used in clinical practice for the umbilical artery are:

●Pulsatility index (PI = peak systolic velocity - end-diastolic velocity/time-averaged maximum velocity) [40]

●Resistance index (RI = peak systolic velocity – end-diastolic velocity/peak systolic velocity) [41]

●S/D ratio: Peak systolic velocity/end-diastolic velocity [42]

The pulsatility index is preferred because it gives a better estimate of the characteristics of the waveform than the RI or S/D ratio [43].

Reference ranges for the three umbilical artery indices were established by data obtained in a longitudinal study [44]. (See "Doppler ultrasound of the umbilical artery for fetal surveillance".) Doppler of the ductus venosus, middle cerebral artery (MCA), and other fetal vessels also may provide information about fetal hemodynamic status.

The changes in the umbilical artery waveform with advancing gestation for appropriate for gestational age fetuses are shown in the following waveforms (waveform 1). By comparison, in FGR, the umbilical artery diastolic flow decreases, may become absent, and in the most severe cases there is reversed diastolic flow, as shown in the following waveforms (waveform 2).

Our approach — In fetuses with estimated weight <10^th percentile and/or an abdominal circumference <10^th percentile, we obtain the pulsatility index of both the umbilical artery and middle cerebral artery (MCA) weekly. If one of the two is abnormal (>95^th percentile for the umbilical artery; <10^th percentile for the MCA), we diagnose FGR with placental insufficiency. We have observed that the MCA Doppler, in some FGR, can be the first parameter to become abnormal, followed by the umbilical artery. The technique that we use to sample the umbilical artery is that reported in the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) guideline on use of Doppler velocimetry in obstetrics [43].

If consecutive Doppler results are normal, we decrease the frequency of Doppler examination to two-week intervals. The two-week interval is reasonable for the fetus with EFW ≥5^th percentile, normal growth velocity, normal amniotic fluid volume, and no maternal risk factors for placental dysfunction. The Society for Maternal-Fetal Medicine suggests umbilical artery Doppler studies every one to two weeks initially, and, if normal, the interval between examinations can be lengthened to two- to four-week intervals [1].

If umbilical artery diastolic flow is present but decreased (abnormal umbilical artery pulsatility index [UAPI] >95^th percentile), we perform weekly Doppler evaluation to look for progression to absent or reversed flow.

Absent or reversed end-diastolic flow in the umbilical artery can be a sign of impending fetal cardiovascular and metabolic deterioration. If either of these abnormal patterns is identified, delivery should be considered (see 'Timing of delivery' below). The decision to deliver in this setting is based on gestational age as long as daily NST or BPP testing is normal. The absence of abnormal flow patterns in the ductus venosus has been used to support the decision to extend such a pregnancy and may enable the pregnancy to be prolonged for as long as two weeks. The Society for Maternal-Fetal Medicine has advised against routine use of this test for clinical management [1]

Although the use of venous Doppler interrogation remains largely investigational [45,46], an increasing number of maternal-fetal medicine specialists, including the authors, are using this tool to avoid very preterm delivery in fetuses with absent or reversed end-diastolic arterial flow in the umbilical artery and reassuring antepartum fetal testing (NST, BPP). In these pregnancies, the absence of abnormal flow patterns in the ductus venosus is used to support the decision to extend the pregnancy to 30 to 33 weeks, if the NST and BPP remain reassuring.

Umbilical artery — Doppler velocimetry of the umbilical artery is the primary surveillance tool for monitoring pregnancies in which FGR is suspected [47]. It has been well established by numerous randomized trials that monitoring umbilical artery Doppler can significantly reduce perinatal death, as well as unnecessary delivery of the preterm growth-restricted fetus. A systematic review of 18 trials comparing the use of Doppler with no Doppler in high-risk pregnancies showed a 29 percent reduction in perinatal deaths (odds ratio 0.71, 95% CI 0.52-0.98; 1.2 versus 1.7 percent; number needed to treat 203), and significantly fewer labor inductions and cesarean deliveries [48].

Normal diastolic flow is infrequently associated with significant perinatal morbidity or mortality and is strong evidence of fetal well-being, thus this finding provides support for delaying delivery when it is important to achieve further fetal maturity. In the PORTO study, growth-restricted fetuses with normal umbilical artery Doppler had lower perinatal mortality than those with abnormal Doppler (2 of 698 [0.3 percent] versus 6 of 418 [1.4 percent]) and a lower rate of overall adverse outcome (9 of 698 [1.3 percent] versus 48 of 418 [11.5 percent]) [49]. Abnormal Doppler was defined as a pulsatility index >95^th percentile or absent/reversed end-diastolic flow. However, normal umbilical artery pulsatility index may be less reassuring when FGR is <3^rd percentile as these cases were at higher risk for cesarean delivery for nonreassuring fetal status than cases of FGR ≥3^rd with normal Doppler in one study [50].

When 30 percent of the villous vasculature ceases to function, an increase in umbilical artery resistance leading to reduced end-diastolic flow is consistently seen [51] and is a weak predictor of adverse outcome [52]. When 60 to 70 percent of the villous vasculature is obliterated, umbilical artery diastolic flow is absent or reversed, and fetal prognosis is poor [51]. Reversed diastolic flow is associated with poorer neonatal outcomes than absent diastolic flow. As an example, in a study of 143 FGR pregnancies with either reversed or absent umbilical artery flow, mortality was over fivefold higher with reversed flow [53]. (See "Doppler ultrasound of the umbilical artery for fetal surveillance".)

Ductus venosus — Doppler interrogation of the ductus venosus provides information about the hemodynamic status of the fetus. Flow in the venous circulation is forward and uniform in normal fetuses. Changes in the venous circulation in the growth-restricted fetus, including absent or reversed flow in the ductus venosus (absent or reversed a-wave) or pulsatile umbilical venous flow, are late findings, generally occurring approximately two weeks after changes are observed in the arterial circulation.

With progressively increasing umbilical arterial resistance, fetal cardiac performance can become impaired and central venous pressure increases, resulting in reduced diastolic flow in the ductus venosus and other large veins. Vasodilatation of the ductus venosus further diverts nutrient- and oxygen-rich blood to the heart but enhances retrograde transmission of atrial pressure. The ductus venosus resistance index increases, ultimately with loss of the a-wave. An absent or reversed ductus venous a-wave indicates cardiovascular instability and can be a sign of impending acidemia and death [54,55]. Because the time from identification of ductus venosus reversed flow to fetal demise or nonreassuring fetal testing is variable, a transitional phase likely exists between the presence of end-diastolic forward flow and absent or reversed end-diastolic flow [56].

Although overall sensitivity and specificity for fetal pH <7.20 are only 65 and 95 percent, respectively [54], the duration of the absent or reversed ductus venous a-wave needs to be taken into account and appears to impact outcome independently of gestational age. Each day of this Doppler abnormality doubles the odds of stillbirth, and fetal survival for more than one week is unlikely [55]. In early FGR (<30 weeks of gestation), the sensitivity of an absent or reversed ductus venous a-wave for fetal demise, neonatal death, and neonatal survival has been reported to be 88, 78, and 32 percent, respectively [57]. Of 32 neonatal survivors in this study, 11 (34 percent) had absent or reversed ductus venous a-wave for a median of 11 days before delivery. The authors proposed a new index (s-wave/isovolumetric a-wave [SIA] index), hypothesizing that a quantitative assessment of the ductus venosus may be better than the qualitative assessment.

The Trial of Randomized Umbilical and Fetal Flow in Europe (TRUFFLE) [58] (see 'Timing of delivery' below) demonstrated no immediate neonatal benefit from delaying delivery until ductus venosus monitoring showed significant abnormalities (absent or reversed flow) and only a possible marginal benefit in neurodevelopment at two years of age [58]. In pregnancies in which delayed delivery is planned, a cohort analysis of data from the Growth Restriction Intervention Trial (GRIT) and TRUFFLE found that the rate of survival without impairment at age 2 years was higher in pregnancies monitored using ductus venosus Doppler plus computerized cardiotocography versus computerized cardiotocography alone (84 versus 77 percent) [59]. It is important to note that computerized cardiotocography is not used in most centers in the United States and that the BPP was not used in GRIT and TRUFFLE [58].

Middle cerebral artery — Doppler interrogation of the MCA also provides information about the hemodynamic status of the fetus. The fetal brain in uncomplicated pregnancies has a high resistance circulation. With progressive hypoxia, blood flow increases to compensate for the decrease in available oxygen (brain-sparing effect). This results in a reduction in the Doppler parameters used to assess blood flow through the MCA: the peak systolic to end-diastolic blood flow velocity ratio (S/D), resistance index, and pulsatility index [27-29,60-62]. Subsequent normalization of the indices may occur when the autoregulatory response becomes dysfunctional [63].

There is no convincing evidence that interrogation of the MCA Doppler alone is useful in guiding clinical decisions about timing of delivery, although MCA Doppler alterations may be useful as an adjunct to umbilical artery Doppler interrogation for assessing the severity of hypoxia and predicting neonatal outcome.

Cerebroplacental ratio — The cerebroplacental Doppler ratio (CPR) is the MCAPI (or resistance index) divided by the UAPI (or resistance index); a low CPR indicates fetal blood flow redistribution (brain sparing). CPR was initially described for detecting FGR fetuses [64]. Following a few initial studies, it was abandoned because it did not appear to provide more information than the umbilical artery alone. In the last 10 years, however, many additional studies measured this ratio to predict the perinatal outcome in FGR pregnancies and reported widely variable estimates of its accuracy [65-67]. Several threshold CPR values have been proposed for predicting adverse outcome (<1, <1.05, ≤1.08, <5^th percentile) [60,64-66].

We suggest not routinely performing CPR in pregnancies with FGR (or appropriate for gestational age fetuses), but believe more studies are needed to define whether CPR has a role in monitoring FGR and which pregnancies might benefit from its use. The most recent meta-analysis of this issue, which included individual participant data from both high- and low-risk pregnancies (10 centers provided 17 datasets for over 21,000 participants), concluded CPR at any threshold added no predictive value for adverse perinatal outcome beyond UAPI, irrespective of gestational age or fetal size [68]. These findings are similar to those from a large prospective observational study (n >47,000) not included in the analysis [67] and a smaller study [60].

Some studies not included in the meta-analysis have suggested CPR is predictive of outcome. As an example, in the PORTO study [49], which included >1100 ultrasound-dated singleton FGR pregnancies, the rate of serious adverse neonatal outcome with low CPR (<1) was 18 percent (27 of 146) versus 2 percent (14 of 735) when CPR was higher [69]. In a follow-up study of children in this trial comparing neurodevelopmental outcomes at three years of age in 136 children with FGR plus abnormal umbilical artery Doppler and normal CPR >1 with 41 children with FGR plus both abnormal umbilical artery Doppler and CPR <1, the latter group had significantly poorer neurodevelopmental outcomes [70].

The opposite of the CPR ratio (ie, the UAPI divided by the MCAPI) has also been suggested as an indicator of perinatal outcome; a high value is pathologic. There is no evidence, however, that this ratio provides more information than the CPR or the umbilical artery alone.

Nonstress test and biophysical profile — Either the NST with amniotic fluid volume determination or the BPP or a combination of both tests is reasonable for monitoring fetal well-being. The value of these tests is based primarily on two lines of evidence: (1) observational studies that reported lower rates of fetal death in pregnancies that underwent fetal testing than among historic controls with the same indication for testing but no fetal testing and (2) the same or lower rates of fetal death in tested pregnancies (primarily high risk) than in a contemporary untested general obstetric population (primarily low risk). Meta-analyses of randomized trials of these interventions in high-risk pregnancies did not demonstrate a reduction in perinatal death, although most of the trials were of low quality; a benefit cannot be excluded given the limitations of the available data [71,72]. (See "Nonstress test and contraction stress test" and "Biophysical profile test for antepartum fetal assessment".)

We use a combination of the NST and BPP in addition to Doppler velocimetry to monitor FGR as these tests evaluate both acute and chronic fetal physiologic parameters. The tests are relatively easy to perform, and fetal death within one week of a normal test score is rare [73]. If the NST is performed without a BPP, amniotic fluid volume assessment should also be performed weekly. Chronic placental insufficiency results in both FGR and oligohydramnios, and observational studies have reported that pregnancies complicated by FGR and oligohydramnios have a modestly increased risk of perinatal mortality [74,75]. Conversely, normal amniotic fluid volume is infrequently associated with either FGR or fetal demise, unless the cause is a congenital malformation or aneuploidy. (See "Oligohydramnios: Etiology, diagnosis, and management".)

Frequency of nonstress tests and biophysical profiles

●For pregnancies with FGR based on estimated weight and/or abdominal circumference between 3^rd and 10^th percentile, normal growth velocity, and normal Doppler indices, we perform NSTs or BPPs on a weekly basis starting at 32 weeks gestation. However, more data are needed to determine whether antenatal testing in this group results in better outcomes.

●For pregnancies with FGR that is severe based on estimated weight and/or abdominal circumference <3rd percentile or with oligohydramnios, preeclampsia, decelerating growth velocity, increasing umbilical artery Doppler index, or other concerning findings, we perform testing twice per week (usually one NST and one BPP).

●For pregnancies with FGR based on estimated weight and/or abdominal circumference <10^th percentile and absent or reversed diastolic flow, we perform daily testing (NST or BPP) because these fetuses can deteriorate rapidly.

In a retrospective study of 27 FGR fetuses <32 weeks of gestation with BPP 8/8, 3 died within 24 hours of the BPP, 12 delivered within 24 hours of the BPP and were acidemic at birth, and the remaining 12 had normal umbilical artery blood gases at delivery [76]. Given that over 50 percent of the fetuses in this study with normal BPPs died or were acidemic within 24 hours, we monitor FGR pregnancies <32 weeks that have either absent or reversed flow of the umbilical artery with a combination of an NST every 12 hours and a daily BPP until delivery.

Antenatal corticosteroids — Ideally, a course of antenatal betamethasone is given to pregnancies <34+0 weeks of gestation in the week before preterm delivery is anticipated. Administration at 34+0 to 36+6 weeks does not appear to decrease the need for respiratory support and increases the rate of neonatal hypoglycemia [77] but is recommended by some guidelines [14]. Timing is estimated based on multiple factors, including the severity of FGR, Doppler findings, comorbid conditions, and rate of deterioration in fetal status. Administration of antenatal betamethasone is reviewed in detail separately. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery".)

The efficacy of antenatal steroids for reducing neonatal morbidity and mortality in the preterm growth-restricted neonate remains controversial, with two large studies showing conflicting results [78,79]. Until definitive information is available, a course of betamethasone should be administered as multiple series have found that both spontaneous and indicated preterm deliveries are more common in growth-restricted fetuses [80-82].

Three studies observed that growth-restricted fetuses with absent end-diastolic flow often show transient improvement in blood flow after betamethasone administration [83-85]. Fetuses that did not show increased end-diastolic flow appeared to have poorer neonatal outcomes. The reason sicker fetuses are unable to mount a vascular response to betamethasone administration is unclear. One action of glucocorticoids is to enhance the tropic effect of catecholamines on heart muscle. It is hypothesized that inotropy does not improve in sicker fetuses because they have impaired cardiac wall compliance.

Maternal interventions — There is no convincing evidence that any intervention in healthy women improves the growth of growth-restricted fetuses. Numerous approaches have been tried in small randomized trials, including maternal nutritional supplementation, oxygen therapy, and interventions to improve blood flow to the placenta, such as plasma volume expansion, low-dose aspirin, bed rest, and anticoagulation [38,86-90].

Use of a phosphodiesterase-5 enzyme inhibitor (eg, tadalafil, sildenafil) [91] or a statin appeared promising and was under investigation [92,93]. However, a multicenter Dutch trial of sildenafil for treatment of poor prognosis early-onset growth restriction was halted early because of a higher than expected rate of pulmonary hypertension in the intervention group with no benefit in the primary outcome (perinatal mortality or major neonatal morbidity) at the time the trial was stopped [94]. A concurrent trial in Australia and New Zealand reported sildenafil had no effect on fetal growth velocity after diagnosis of growth restriction before 30 weeks but no adverse effects on newborns [95]. In a meta-analysis of phosphodiesterase-5 inhibitors in FGR (six trials using sildenafil, one using tadalafil), the intervention improved uteroplacental, but not fetal cerebral, blood perfusion [96].

In hypertensive pregnant women, antihypertensive therapy does not improve fetal growth [97].

In smokers, an intensive smoking cessation program may be of value and has other pregnancy and health benefits [98,99]. (See "Cigarette and tobacco products in pregnancy: Impact on pregnancy and the neonate".)

Timing of delivery — There is little consensus about the optimum time to deliver the growth-restricted fetus. The following key randomized trials attempted to answer the question of when to intervene in these pregnancies, without a clear conclusion:

●The Growth Restriction Intervention Trial (GRIT) randomly assigned pregnant women between 24 and 36 weeks with FGR to immediate (n = 273) or delayed (n = 274) delivery if their obstetrician was uncertain about when to intervene [100]. Forty percent of these pregnancies had absent or reversed end-diastolic umbilical artery flow. In the delayed delivery group, delivery occurred when the obstetrician was no longer uncertain about intervening, which took a median 4.9 days.

The immediate delivery group had fewer stillbirths (2 versus 9 with delayed delivery) but more neonatal and infant deaths (27 versus 18), especially when randomization occurred before 31 weeks. Follow-up data up to age 13 years showed no differences between groups in cognition, language, motor, or parent-assessed behavior scores on standardized tests; follow-up was achieved in approximately 70 percent of survivors [101,102]. Cognition scores were close to the standardized normal range.

These data suggest that delaying delivery of the very preterm growth-restricted fetus in the setting of uncertainty results in some stillbirths, but immediate delivery produces an almost equal number of neonatal deaths, and neither approach results in better long-term neurodevelopmental outcome. Although widely cited, these studies are difficult to evaluate due to the lack of standard criteria for intervention.

●The Disproportionate Intrauterine Growth Intervention Trial At Term trial (DIGITAT) randomly assigned 650 pregnant women over 36.0 weeks of gestation with suspected FGR to induction of labor or expectant monitoring [103-105]. The primary outcome was a composite measure of adverse neonatal outcome (death before hospital discharge, five-minute Apgar score <7, umbilical artery pH <7.05, or admission to the intensive care unit) [103]. Neonatal morbidity was analyzed separately using Morbidity Assessment Index for Newborns (MAIN) score [104].

The induction group delivered 10 days earlier and weighed 130 grams less (mean difference -130 grams, 95% CI -188 to -71) than the expectantly managed group, but had statistically similar composite adverse outcome (6.1 versus 5.3 percent with expectant management) and cesarean delivery rates (approximately 14 percent) [103]. Developmental and behavioral outcomes at two years of age were also similar for both groups [105]. The authors concluded that both approaches were reasonable, and the choice should depend on patient preference. However, in a subanalysis of the data, they also reported that neonatal admissions were lower when growth restricted fetuses were delivered after 38 weeks of gestation [104], which suggests a benefit of deferring delivery until 38 weeks of gestation, as long as the fetus is closely monitored and in the absence of other indications for an early delivery.

●The Trial of Randomized Umbilical and Fetal Flow in Europe (TRUFFLE) assessed whether changes in the fetal ductus venosus Doppler waveform could be used to guide timing of delivery of growth-restricted fetuses with a high umbilical artery Doppler pulsatility index (>95^th percentile) instead of the conventional approach using cardiotocography short-term variation (STV) [58]. The primary outcome measure was survival without neurodevelopmental impairment at two years of age. Pregnancies were randomly assigned to one of three monitoring approaches: cardiotocography with delivery for reduced STV, ductus venosus monitoring with delivery for early ductus venosus changes (pulsatility index >95^th percentile), or ductus venosus monitoring with delivery for late ductus venosus changes (a-wave indicating absent or reversed flow). The proportion of infants surviving without neurodevelopmental impairment was 77 to 85 percent, with no significant differences among the three groups. Among survivors, delaying delivery until the development of late ductus venosus changes resulted in an improvement in survival without neurodevelopmental impairment (95 percent versus 91 percent for the early ductus venosus changes group and 85 percent in the reduced STV group); however, this came at the cost of a small increase in unexpected fetal demise (0 of 166 in the STV group versus 3 of 167 in the early ductal changes group versus 4 of 170 in the late ductal changes group). There were no differences in immediate neonatal composite morbidity or mortality.

These findings do not support a change in clinical practice, given that the improvement in neurodevelopment was offset, in part, by an increase in fetal demise. Moreover, the number of neurodevelopmentally impaired children in each group was small (7 in the late ductal changes group, 12 in the early changes group, and 20 in the STV group); thus a larger trial may have resulted in a different outcome. Lastly, it is not clear that the investigators adjusted neurodevelopmental outcome scores for the mother's educational level.

In a post hoc analysis of their data, the TRUFFLE group concluded both ductus venosus and cardiotocography evaluation are warranted since the majority of infants in the ductus venous groups were delivered for reduced STV or spontaneous decelerations in fetal heart rate rather than early or late pulsatility changes in the ductus venous [106]. These observations clearly demonstrate the need for additional study of the most appropriate methods to determine delivery timing in the very preterm (<32 weeks of gestation) FGR fetus.

In addition, a subsequent retrospective cohort study reported that infants with severe SGA (birth weight <3^rd percentile) who were iatrogenically delivered early for suspected FGR had poorer school outcomes in grades 3, 5, and 7 compared with infants with the same degree of growth restriction who were not suspected of having FGR and thus not delivered early for this indication (mean gestational age at birth: 37.9 versus 39.4 weeks) [107]. The authors suggested that clinicians delay delivery of pregnancies with suspected FGR when safe to do so, at least until 38 weeks of gestation, at which point previous studies have reported that the adverse effects of intervention diminish such that induction and expectant management with intensive maternal and fetal monitoring are associated with similar outcomes [103,105]. The findings of this retrospective study need to be interpreted with caution given that they may have been due to unmeasured confounding. Detailed information on fetal biometry, Doppler studies, or other markers of pathology were not recorded; the etiology for SGA was not known; stillborns and neonatal deaths were excluded; and there was no adjustment for many factors that contribute to academic success.

Our approach — We time the delivery of the growth-restricted fetus based on a combination of factors, including gestational age, Doppler ultrasound of the umbilical artery, BPP score, ductus venosus Doppler, and the presence or absence of risk factors for, or signs of, uteroplacental insufficiency. The goal is to maximize fetal maturity and growth while minimizing the risks of fetal or neonatal mortality and short-term and long-term morbidity. The greatest challenge related to timing of delivery is in the preterm fetus <32 weeks of gestation. Morbidity and mortality related to preterm delivery is relatively high before 32 weeks [108,109], and between 26 and 29 weeks of gestation each day in utero has been estimated to improve survival by 1 to 2 percent [110].

The following is a synopsis of our approach and is based on the evidence described above (see 'Nonstress test and biophysical profile' above and 'Doppler velocimetry' above and 'Timing of delivery' above) and similar to the approach of the American College of Obstetricians and Gynecologists [111]:

●Persistent reversed a-wave of the ductus venosus Doppler – We deliver these pregnancies immediately if ≥30+0 weeks of gestation. Before 30+0 weeks, we individualize the decision of delivery.

●Umbilical artery reversed diastolic flow – We deliver these pregnancies between 30+0 and 32+0 weeks of gestation or at diagnosis if diagnosed later (consult with maternal-fetal medicine).

●Umbilical artery absent diastolic flow – We deliver these pregnancies between 33+0 and 34+0 weeks of gestation or at diagnosis if diagnosed later (consult with maternal-fetal medicine).

If a course of antenatal betamethasone has not been administered yet, it is given upon diagnosis of reversed or absent diastolic flow.

●Umbilical artery abnormal Doppler (persistent pulsatility index >95^th percentile) – This is a weak predictor of fetal death. We perform a BPP two times per week and deliver these fetuses at 37+0 weeks (or at diagnosis if diagnosed later) or when the BPP becomes abnormal.

●Normal umbilical artery Doppler and:

•EFW <3^rd percentile – We deliver these pregnancies at 37+0 weeks of gestation or at diagnosis if diagnosed later, given severe FGR.

•EFW ≥3^rd and <10^th percentile – We deliver these pregnancies at 39+0 weeks of gestation or at diagnosis if diagnosed later if the pregnancy is uncomplicated and there are no concurrent findings. A normal Doppler provides strong evidence of fetal well-being, especially in the absence of severe FGR or concurrent findings.

Delivery should not be delayed beyond 39+0 weeks of gestation. The risk of fetal demise significantly increases at term, particularly as the severity of FGR increases. As an example, in a retrospective cohort study, the risk of fetal death at 39 weeks was estimated as 32 per 10,000 ongoing pregnancies for fetal weight <3^rd percentile, 23 per 10,000 ongoing pregnancies for fetal weight <5^th percentile, 13 per 10,000 ongoing pregnancies for fetal weight <10^th percentile, and 2 per 10,000 ongoing pregnancies for fetal weight ≥10^th percentile [112].

●FGR and concurrent conditions (oligohydramnios, maternal comorbidity [eg, preeclampsia, chronic hypertension]) – We deliver these pregnancies between 34+0 and 37+0 weeks of gestation.

Route of delivery — An unfavorable cervix is not a reason to avoid induction. We prefer mechanical ripening methods (insertion of a balloon catheter or laminaria), which may be safer than prostaglandins in this setting [113]. If the Bishop score is >6, we administer oxytocin without mechanical ripening.

In a secondary analysis of data from the DIGITAT and HYPITAT trials (pregnancies complicated by FGR and hypertension), induction of labor at term in women with median Bishop scores of 3 (range 1 to 6) was not associated with a higher rate of cesarean delivery than expectant management, and approximately 85 percent of women in both groups achieved a vaginal delivery [114]. Prostaglandins or a balloon catheter was used for cervical ripening. In a meta-analysis of observational studies of labor induction with misoprostol, dinoprostone, or mechanical methods in FGR, mechanical methods appeared to be associated with a lower occurrence of adverse intrapartum outcomes, but a direct comparison among methods could not be performed [115].

However, when the indication for delivery is persistent reversed flow of the umbilical artery, we give patients the option of a scheduled cesarean birth, especially when the cervix is unfavorable, because many of these fetuses will not tolerate labor and we wish to avoid adding an acute insult on a chronic hypoxic fetus.

INTRAPARTUM MANAGEMENT — Growth-restricted fetuses may exist in a state of mild-to-moderate chronic oxygen and substrate deprivation. Potential consequences include antepartum or intrapartum fetal heart rate abnormalities, passage of meconium with risk of aspiration, and neonatal polycythemia, impaired thermoregulation, hypoglycemia, and other metabolic abnormalities (see "Infants with fetal (intrauterine) growth restriction", section on 'Outcomes'). Consequently, it is important to optimize the timing of delivery (see 'Our approach' above), perform continuous intrapartum fetal monitoring to detect nonreassuring fetal heart rate patterns suggestive of progressive hypoxia during labor, and provide skilled neonatal care in the delivery room [116]. Umbilical cord blood analysis should be considered as a component of establishing baseline neonatal status. (See "Umbilical cord blood acid-base analysis at delivery", section on 'Indications for fetal acid-base analysis'.)

If antenatal testing (nonstress test or biophysical profile) is normal, a trial of labor with continuous intrapartum monitoring is reasonable [117]. However, the frequency of cesarean delivery for nonreassuring fetal heart rate tracing is increased, given the increased prevalence of chronic hypoxia and oligohydramnios among these fetuses. In one study, variable decelerations lasting greater than 60 seconds, with depth greater than 60 beats per minute (bpm) or nadir less than 60 bpm, were more common in the growth-restricted group, while the rates of late decelerations, prolonged decelerations, or bradycardia were similar for both groups [118].

The risk of fetal heart rate abnormalities related to hypoxia is highest among fetuses with abnormal Doppler velocimetry. In one large series, no fetus with normal Doppler velocimetry was delivered with metabolic acidemia associated with chronic hypoxemia [119]. Compared with growth-restricted fetuses with normal umbilical artery Doppler ratios, growth-restricted fetuses with a systolic/diastolic ratio >90^th percentile for gestational age had significantly lower umbilical artery and vein pH values at birth (artery 7.23±0.08 versus 7.25±0.1; vein, 7.31±0.01 versus 7.34±0.09), an increased likelihood of nonreassuring fetal heart rate patterns (26 versus 9 percent), more admissions to the neonatal intensive care unit (41 versus 31 percent), and a higher incidence of respiratory distress (66 versus 27 percent).

For fetuses less than 32 weeks of gestation, magnesium sulfate is given before delivery for neuroprotection. When use of magnesium sulfate was studied specifically in pregnancies with growth-restricted fetuses, a decrease in significant neurodevelopmental impairment and death was observed [120]. (See "Neuroprotective effects of in utero exposure to magnesium sulfate".)

PROGNOSIS

Fetal, newborn, and childhood outcomes — Fetal demise, neonatal death, neonatal morbidity, and abnormal neurodevelopmental outcome are more common in growth-restricted fetuses than in those with normal growth [121]. The prognosis worsens with early-onset FGR, increasing severity of growth restriction (figure 4), and absent or reversed end-diastolic flow on umbilical artery Doppler [122].

In a systematic review of studies of FGR diagnosed before 32 weeks of gestation (n = 2895 pregnancies delivered after 2000), the frequencies of fetal and neonatal death were 12 and 8 percent, respectively [123]. The most common neonatal morbidities were respiratory distress syndrome (34 percent), retinopathy of prematurity (13 percent), and sepsis (30 percent). Among children who underwent neurodevelopmental assessment, 12 percent were diagnosed with cognitive impairment and/or cerebral palsy. The quality of evidence was generally rated as very low to moderate, except for three large, well-designed, randomized trials. When all pregnancies with FGR are considered, regardless of gestational age at diagnosis, the risk of fetal death is approximately 1.5 percent at <10^th percentile and 2.5 percent at <5^th percentile for gestational age [124,125]. (See "Infants with fetal (intrauterine) growth restriction".)

Longer term outcomes — An association has been observed between poor fetal growth, early accelerated postnatal growth, and later development of obesity, metabolic dysfunction, insulin sensitivity, type 2 diabetes, and cardiovascular and renal diseases (eg, coronary heart disease, hypertension, chronic kidney disease). This association has been attributed to partial resetting of fetal metabolic homeostasis and endocrine systems in response to in utero nutritional deprivation. The combination of prematurity and severe FGR increases the risk of long-term neurodevelopmental abnormalities and decreased cognitive performance. (See "Infants with fetal (intrauterine) growth restriction", section on 'Adult chronic disorders'.)

Numerous studies have also demonstrated fetal, neonatal, and long-term cardiac remodeling, which may be associated with cardiovascular morbidity and mortality later in life [126-129]. Concerns regarding an increased frequency of long-term metabolic and cardiovascular abnormalities raise the question whether preterm delivery may prevent these abnormalities. There are no data to support early delivery for this indication, but timing of delivery based on prevention of long-term morbidities related to impaired in utero growth needs careful study.

Maternal — The birth of a newborn with idiopathic growth restriction may be predictive of an increased long-term maternal risk for cardiovascular disease (coronary artery disease, myocardial infarction, coronary revascularization, peripheral arterial disease, transient ischemic attack, stroke). A systematic review of 10 cohort studies found a consistent trend of an increased risk of cardiovascular disease-related morbidity and mortality in patients with a history of birth of an SGA infant compared to those with no such history (range of odds ratios 1.09 to 3.50) [130]. Pooling was not performed because of variations in the exposure definition among studies.

RECURRENCE RISK — There is a tendency to repeat small for gestational age (SGA) deliveries in successive pregnancies [126,127,131,132]. As an example, a prospective national cohort study from the Netherlands reported that the risk of a nonanomalous SGA birth (<5^th percentile) in the second pregnancy of women whose first delivery was "SGA" versus "not SGA" was 23 and 3 percent, respectively [127]. The odds of recurrence increase markedly as the number of previous SGA infants increases [133].

Furthermore, uteroplacental insufficiency may manifest in different ways in different pregnancies. Growth restriction, preterm delivery, preeclampsia, abruption, and stillbirth can all be sequelae of impaired placental function. The association between the birth of an SGA infant in a first pregnancy and stillbirth in a subsequent pregnancy was illustrated by analysis of data from the Swedish Birth Register (table 4) [134]; subsequent studies from the United States and Australia reported similar findings [135,136]. The highest risk of stillbirth was in women who delivered a preterm SGA infant. Another series suggested a sibling delivered after the birth of an SGA infant (even if mildly SGA) was at increased risk of sudden infant death syndrome [137].

Prevention in subsequent pregnancies — In subsequent pregnancies, we address any potentially treatable causes of FGR (eg, cessation of smoking and alcohol intake, chemoprophylaxis and mosquito avoidance in areas where malaria is prevalent, balanced energy/protein supplementation in women with significant nutritional deficiencies; refer to individual topic reviews). Avoiding a short or long interpregnancy interval may also be beneficial. (See "Interpregnancy interval: Optimizing time between pregnancies".)

Low-dose aspirin may be effective when FGR is secondary to preeclampsia since aspirin appears to reduce the risk of developing preeclampsia in women at moderate to high risk of developing the disorder. In a meta-analysis of 45 randomized trials of low dose aspirin for prevention of preeclampsia and FGR in women at high risk, aspirin prophylaxis markedly reduced the incidence of FGR (relative risk [RR] 0.56, 95% CI 0.44-0.70) compared with placebo/no treatment [138]. Low-dose aspirin is not recommended in the absence of risk factors for preeclampsia [139]. These data are reviewed in more detail separately. (See "Preeclampsia: Prevention", section on 'Low-dose aspirin'.)

Anticoagulation with unfractionated heparin or low molecular weight heparin does not reduce the risk of recurrent placenta-mediated late pregnancy complications, such as growth restriction. In a 2016 meta-analysis using individual patient data from randomized trials of low molecular weight heparin (LMWH) therapy versus no LMWH for women with any prior placenta-mediated pregnancy complications, the intervention did not significantly reduce the incidence of the primary composite outcome (early-onset or severe preeclampsia, SGA <5^th percentile, abruption, pregnancy loss ≥20 weeks of gestation): 62 of 444 (14 percent) versus 95 of 443 (22 percent), RR 0.64, 95% CI 0.36-1.11 [140]. These data support avoidance of anticoagulation in women with previous placenta-mediated disease, given the lack of clear benefit and potential risks of anticoagulation, cost, and inconvenience. The combination of low dose aspirin and LMWH does not appear to be more effective than aspirin alone [141].

The body of evidence suggests that long-chain polyunsaturated fatty acid supplements, antihypertensive therapy of hypertensive women, beta-mimetics, and bedrest do not prevent FGR [38,97,142,143]. However, in a trial conducted at a single institution in Spain that randomly assigned 1221 pregnancies deemed at high risk of having an SGA fetus to either a structured Mediterranean diet, mindfulness-based stress reduction, or usual care from about 20 weeks of gestation to delivery, both interventions appeared to significantly reduce the percentage of newborns with birth weight <10^th percentile (14, 15.6, and 21.9 percent, respectively) [144]. This trial had many limitations. The most important was baseline imbalances among the three groups. The "usual care" group had more patients with autoimmune disease and chronic hypertension, a high number of cigarette smokers, and a higher rate of alcohol consumption, which are risk factors for preeclampsia and SGA. Therefore, the favorable findings in the intervention groups may reflect a relatively worse outcome in the higher risk control group rather than beneficial effects of the interventions. A second limitation is represented by a lack of control in the number of visits and follow-up in the usual care group. Finally, the population included few patients with obesity or metabolic conditions, patient groups in whom previous data have suggested that these interventions are not effective [145,146]. We agree with the authors that this study needs to be replicated to assess the effectiveness of these interventions before making a clinical recommendation. However, both interventions are considered healthy behavioral changes.

Management of subsequent pregnancies — Accurate dating by early ultrasonography is important to establish gestational age and intermittent ultrasound examinations are used to monitor fetal growth. It has been reported that FGR in a previous pregnancy is not an indication for antepartum fetal surveillance with nonstress tests, biophysical profiles, or umbilical artery Doppler velocimetry at the next pregnancy [147]. However, FGR in the first pregnancy is associated with increased risk of stillbirth in the subsequent pregnancy even if the fetus is appropriate size for gestational age [134]. Therefore, guidance from the American College of Obstetricians and Gynecologists suggests antenatal fetal surveillance in pregnancies after a previous stillbirth [148,149]. Timing and frequency is reviewed separately. (See "Stillbirth: Incidence, risk factors, etiology, and prevention", section on 'Principles of pregnancy management'.)

FETUSES WITH A REDUCTION IN GROWTH RATE BUT NOT FGR — Although a significant decrease in the fetal weight percentile between 30 and 38 weeks may be of concern, the presence of normal umbilical artery diastolic flow and amniotic fluid volume are reassuring in fetuses with estimated fetal weight that remains ≥10^th percentile for gestational age. Because of the reduction in fetal growth rate, we follow these fetuses with weekly biophysical profiles and consider delivery at 38 weeks of gestation reasonable, with preterm delivery for standard obstetric indications. An alternative is to increase fetal testing to twice a week (one nonstress test [NST] and one biophysical profile [BPP]) with delivery at 39 weeks.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Fetal growth restriction".)

SUMMARY AND RECOMMENDATIONS

●Overview — Evaluation and management of suspected fetal growth restriction (FGR) involves:

•Accurate determination of gestational age

•Confirming the diagnosis

•Distinguishing between the constitutionally small and the growth-restricted fetus

•Monitoring the fetal weight trajectories

•Managing maternal comorbidities

•Serial assessment of fetal well-being

•Preterm delivery when indicated

●Initial approach

•An estimated fetal weight <10^th percentile or abdominal circumference <10^th percentile signifies a small for gestational age (SGA) fetus. It is then the clinician's task to distinguish between the constitutionally small fetus that achieves its normal growth potential and is not at increased risk of adverse outcome from the similarly small fetus whose growth potential is restricted and is at increased risk of perinatal morbidity and mortality (ie, FGR). (See 'Confirm the diagnosis' above.)

•We perform a detailed fetal anatomic survey in all cases of FGR since major congenital anomalies are frequently associated with failure to maintain normal fetal growth. (See 'Fetal survey' above.)

•Evaluation of the fetal karyotype/microarray is indicated if FGR is associated with structural anomalies, ultrasound markers of aneuploidy, or early severe FGR (<5^th percentile before 24 weeks of gestation). (See 'Fetal genetic studies' above.)

•We do not perform routine TORCH serology in the evaluation of isolated FGR. We obtain serology when ultrasound findings are suggestive of an intrauterine infection (ie, sonographic markers of fetal infection are present in addition to FGR) or a careful maternal history and physical examination suggest the possibility of maternal infection and vertical transmission. If the patient chooses to have diagnostic fetal testing for isolated FGR, we perform amniocentesis for polymerase chain reaction (PCR) of amniotic fluid for CMV. (See 'Work-up for infection' above.)

•Antiphospholipid antibody screening is performed in patients with FGR and a past history of fetal loss or prior unexplained arterial or venous thromboembolism. Assessment for inherited thrombophilic disorders is not recommended as evidence for an association between the inherited thrombophilias and FGR is weak. (See 'Work-up for antiphospholipid syndrome' above.)

●Antepartum management

•Serial ultrasound evaluation of (1) fetal growth, (2) fetal behavior (biophysical profile [BPP] or nonstress test [NST] with assessment of amniotic fluid volume), and (3) impedance to blood flow in fetal vessels (Doppler velocimetry) represent the key elements of fetal assessment and guide pregnancy management decisions. The purpose is to identify those fetuses that are at highest risk of in utero demise and neonatal morbidity and thus may benefit from preterm delivery. The frequency is based upon the severity of findings and whether the examinations are being done to monitor fetal well-being (one to seven times per week) or fetal growth (every three to four weeks). (See 'Fetal weight and amniotic fluid assessment' above and 'Nonstress test and biophysical profile' above and 'Umbilical artery' above.)

•We recommend Doppler velocimetry of the umbilical artery for monitoring pregnancies with suspected growth restriction (Grade 1A). Delivery prompted by abnormal umbilical artery Doppler reduces the frequency of perinatal death. Normal umbilical artery Doppler findings are reassuring and thus potentially allow prolongation of pregnancy and reduction in the number of unnecessary preterm deliveries. (See 'Our approach' above.)

Doppler assessment of the fetal venous circulation can provide additional information for decision making in the very preterm fetus, but it remains investigational. (See 'Ductus venosus' above.)

•We recommend one course of antenatal betamethasone between 24 and 34 weeks of gestation in the week before delivery is expected (Grade 1A) Earlier administration is indicated if delivery and aggressive neonatal intervention are planned (see "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery"). Timing is estimated based on multiple factors, including the severity of FGR, Doppler findings, comorbid conditions, and rate of deterioration in fetal status. (See 'Antenatal corticosteroids' above.)

●Delivery

•We time the delivery of the growth-restricted fetus based on a combination of factors, including gestational age, umbilical artery Doppler, BPP score, ductus venosus Doppler, and the presence or absence of risk factors for, or signs of, uteroplacental insufficiency. The goal is to maximize fetal maturity and growth while minimizing the risks of fetal or neonatal mortality and short-term and long-term morbidity. For pregnancies with FGR and normal BPP scores or NSTs (see 'Our approach' above):

-Persistent reversed a-wave of the ductus venosus Doppler – Deliver immediately if ≥30+0 weeks of gestation. Before 30 weeks, individualize the decision of delivery. We give patients the option of a scheduled cesarean birth because many of these fetuses will not tolerate labor. (See 'Route of delivery' above.)

-Umbilical artery reversed diastolic flow – Deliver between 30+0 and 32+0 weeks of gestation or at diagnosis if diagnosed later (consult with maternal-fetal medicine).

-Umbilical artery absent diastolic flow – Deliver between 33+0 and 34+0 weeks of gestation or at diagnosis if diagnosed later (consult with maternal-fetal medicine). If a course of antenatal betamethasone has not been administered yet, it is given upon diagnosis of reversed or absent diastolic flow.

-Umbilical artery abnormal Doppler (persistent pulsatility index >95^th percentile) – We perform fetal testing two times per week (one NST and one BPP) and deliver at 37+0 weeks (or at diagnosis if diagnosed later) or when fetal testing becomes abnormal.

-Normal umbilical artery Doppler and EFW <3^rd percentile – Deliver at 37+0 weeks of gestation or at diagnosis if diagnosed later, given severe FGR.

-Normal umbilical artery Doppler and EFW ≥3^rd and <10^th percentile – Deliver at 39+0 weeks of gestation or at diagnosis if diagnosed later if the pregnancy is uncomplicated and there are no concurrent findings.

-FGR and concurrent conditions (oligohydramnios, maternal comorbidity [eg, preeclampsia, chronic hypertension]) – Deliver between 34+0 and 37+0 weeks of gestation.

●Recurrence risk – There is a tendency to repeat SGA or low birth weight deliveries in successive pregnancies. Growth restriction, preterm delivery, preeclampsia, abruption, and stillbirth can all be sequelae of impaired placental function that may manifest in different ways in different pregnancies. (See 'Recurrence risk' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Robert Resnik, MD, who contributed to an earlier version of this topic review.