Clinical Opinion: The diagnosis and management of suspected fetal growth restriction: an evidence-based approach

Abstract

This study reviewed the literature about the diagnosis, antepartum surveillance, and time of delivery of fetuses suspected to be small for gestational age or growth restricted. Several guidelines have been issued by major professional organizations, including the International Society of Ultrasound in Obstetrics and Gynecology and the Society for Maternal-Fetal Medicine. The differences in recommendations, in particular about Doppler velocimetry of the ductus venosus and middle cerebral artery, have created confusion among clinicians, and this review has intended to clarify and highlight the available evidence that is pertinent to clinical management. A fetus who is small for gestational age is frequently defined as one with an estimated fetal weight of <10th percentile. This condition has been considered syndromic and has been frequently attributed to fetal growth restriction, a constitutionally small fetus, congenital infections, chromosomal abnormalities, or genetic conditions. Small for gestational age is not synonymous with fetal growth restriction, which is defined by deceleration of fetal growth determined by a change in fetal growth velocity. An abnormal umbilical artery Doppler pulsatility index reflects an increased impedance to flow in the umbilical circulation and is considered to be an indicator of placental disease. The combined finding of an estimated fetal weight of <10th percentile and abnormal umbilical artery Doppler velocimetry has been widely accepted as indicative of fetal growth restriction. Clinical studies have shown that the gestational age at diagnosis can be used to subclassify suspected fetal growth restriction into early and late, depending on whether the condition is diagnosed before or after 32 weeks of gestation. The early type is associated with umbilical artery Doppler abnormalities, whereas the late type is often associated with a low pulsatility index in the middle cerebral artery. A large randomized clinical trial indicated that in the context of early suspected fetal growth restriction, the combination of computerized cardiotocography and fetal ductus venosus Doppler improves outcomes, such that 95% of surviving infants have a normal neurodevelopmental outcome at 2 years of age. A low middle cerebral artery pulsatility index is associated with an adverse perinatal outcome in late fetal growth restriction; however, there is no evidence supporting its use to determine the time of delivery. Nonetheless, an abnormality in middle cerebral artery Doppler could be valuable to increase the surveillance of the fetus at risk. We propose that fetal size, growth rate, uteroplacental Doppler indices, cardiotocography, and maternal conditions (ie, hypertension) according to gestational age are important factors in optimizing the outcome of suspected fetal growth restriction.

Introduction

The diagnosis, surveillance, and time of delivery of fetuses with suspected fetal growth restriction (FGR) are major issues in obstetrical practice.^1,2 The Society for Maternal-Fetal Medicine (SMFM)³ and the International Society for Ultrasound in Obstetrics and Gynecology (ISUOG)⁴ have recently issued clinical guidelines concerning FGR management. Furthermore, several professional organizations have made recommendations, and the differences and similarities of the recommendations have been previously reviewed in the American Journal of Obstetrics & Gynecology.¹ This clinical opinion reviewed the key concepts on the diagnosis, surveillance, and time of delivery of patients with suspected FGR.

Diagnosis of Suspected Fetal Growth Restriction

Fetal growth restriction is frequently associated with a fetus failing to reach its genetic and biologic growth potential and is a consequence of several causes, with placental dysfunction being one of the most common causes. Before the publication of the Delphi consensus criteria in 2016 (Table 1),⁵ the most common definition of FGR was based on the evaluation of fetal size (estimated fetal weight [EFW] or abdominal circumference [AC]) and its statistical deviation from specific reference ranges or standards. Fetal size <10th percentile has most frequently been used to define a small for gestational age (SGA) fetus¹ and is most commonly defined using the classic Hadlock formula.⁶ This definition derives from that proposed in 1967 by Battaglia and Lubchenco,⁷ which classified newborns as SGA, appropriate for gestational age (AGA), and large for gestational age (LGA) based on the deviation of birthweight from the mean and falling <10th or >90th percentiles, SGA and LGA, respectively. Their definition identified a newborn below the absolute birthweight cutoff of 10th percentile as SGA but did not distinguish between the pathologically small and normally (or constitution-ally) small newborn. A birthweight that identifies SGA is undoubtedly associated with an increased risk of adverse outcome.^8,9 However, basing the definition of abnormal growth on a weight threshold alone has limitations.

TABLE 1.

Delphi consensus criteria for the definition of early and late fetal growth restrictions5

Early FGR: GA<32 wk, in the absence of congenital anomalies	Late FGR: GA>32 wk, in the absence of congenital anomalies
AC or EFW of <third percentile or UA-AEDF Or  1. AC or EFW of <10th percentile combined with  2. Uta-PI of >95th percentile and/or  3. UA-PI of >95th percentile	AC or EFW of <third percentile Or at least 2 of 3 of the following:  1. AC or EFW of <10th percentile  2. AC or EFW crossing percentiles of >2 quartiles on growth percentiles  3. CPR of <5th percentile or UA-PI of >95th percentile

AC, abdominal circumference; AEDF, absent end-diastolic flow; CPR, cerebroplacental ratio; EFW, estimated fetal weight; FGR, fetal growth restriction; GA, gestational age; PI, pulsatility index; UA, umbilical artery; UtA, uterine artery.

Fetal Growth Restriction: Which Thresholds are Important?

The definition of SGA makes no distinction between a constitutionally small but healthy fetus and a fetus with growth restriction. In a given population of fetuses, 10% will have a birthweight <10th percentile, and this group is considered to be SGA. From a biologic point of view, it is implausible to assume that all fetuses whose size is <10th percentile suffer from growth restriction and that none with a weight >10th percentile do. In a population of just under 1.2 million pregnancies in the Netherlands from 2002 to 2008 with >5000 perinatal deaths, perinatal mortality progressively increased as birthweight reduced from the 80th birthweight percentile.¹⁰ Although mortality was the highest below the 2.3rd birthweight percentile (Figure 1), the inflection point of the curve increased progressively starting from the 20th to 50th percentile, suggesting increased perinatal risk exists even above the 10th percentile of fetal size. This finding was consistent with an earlier study of 82,361 infants born at term in 1 US institution during an 8-year period where infants with a birthweight of less than the third percentile were at the highest risk of morbidity and mortality.¹¹ In the multicenter Irish Prospective Observational Trial to Optimize Pediatric Health (PORTO) study of fetal SGA, an adverse outcome was most closely related to an EFW of <3rd percentile; however, this was not true for other biometric cutoffs (EFW of <5th percentile and ACs of <3rd, 5th, and 10th percentile).¹² Similarly, the risk of fetal death for each week of gestation is the highest at birthweights of <3rd percentile, which are 3 times greater than birthweights between the 3rd the 5th percentiles and 4 to 7 times greater than birthweights between the 5th and 10^th percentiles.^10,13 In the same study by Pilliod et al,¹³ the “number needed to deliver” to prevent 1 fetal death at 34 weeks of gestation was 3.7 and 6.1 times higher at the 3rd to 5th percentile and 5th to 10th percentile, respectively, than the 3rd percentile; by 39 weeks of gestation, the difference was 6.0 and 13.7 times higher. The risk of fetal death in SGA fetuses is incrementally cumulative for each week beyond 37 weeks: 21 per 10,000 ongoing pregnancies with SGA fetuses (at 37 weeks), 26 per 10,000 ongoing pregnancies with SGA fetuses (at 39 weeks), and 60 per 10,000 ongoing pregnancies with SGA fetuses (>40 weeks).¹⁴

FIGURE 1. Perinatal mortality according to birthweight percentile.

Adapted from Vasak et al.¹⁰

The Dutch Multicenter Disproportionate Intrauterine Growth Intervention Trial at Term (DIGITAT) study was a randomized controlled trial (RCT) of induced labor vs expectant management of fetuses with an EFW or AC of <10th percentile or a “flattening of the third-trimester growth curve” as judged by the clinician. A birthweight of <2.3^rd percentile was the strongest predictor of an adverse 2-year neurodevelopmental outcome.^15,16 Finally, in one of the largest prospective longitudinal studies on late suspected FGR, infants with a birthweight Z-score of <−2 (in other words, the 2.3rd percentile) had a significantly higher risk of composite adverse outcome (relative risk [RR], 2.9; 95% confidence interval [CI], 1.8–4.0).¹⁷ Taken together, these data showed that the 10th percentile, although a convenient cutoff from a population statistical standpoint, is not Adapted from Vasak et al.¹⁰ suited as a “stand-alone” to define suspected FGR as it potentially exposes a significant proportion of healthy fetuses to unnecessary intervention in late preterm period. If one were to use a single biometric “cutoff ” to define suspected FGR, the 3rd percentile would be more appropriate.¹⁸

Fetal Size VS Growth Velocity Rate

The second limitation of any definition based purely on a point estimate of fetal size is that it does not take into account that fetal growth velocity may slow while the absolute fetal size is still >10th percentile (Figure 2), particularly in late suspected FGR. Fetal growth and fetal size are not the same: as opposed to absolute fetal size, fetal growth occurs over time, is equivalent to velocity, and therefore requires longitudinal evaluation.²¹ In postnatal life, a flattening of the growth velocity curve might alert the physician to possible pathologic growth failure of the neonate or infant.^22,23 It is entirely logical that this concept would apply during fetal life when the growth rate is particularly rapid.²³ In a prospective single-center cohort study on universal third-trimester sonography, an EFW of <10th percentile was significantly associated with any neonatal morbidity, which included poor immediate birth outcome, severe morbidity, and death (RR, 1.6; 95% CI, 1.2–2.1).¹⁹ However, when AC growth velocity was considered, the risk of adverse outcomes only increased for fetuses that in addition to an EFW of <10^th percentile were at the lowest decile of the AC growth velocity (RR, 2.5; 95% CI, 1.7–3.6); (Figure 2). Together with abnormally raised umbilical artery (UA) Doppler pulsatility index (PI), AC growth velocity had the strongest relationship in a stratified risk analysis of neonatal composite adverse outcome in SGA.

FIGURE 2. Patterns of fetal growth in SGA and FGR.

The figure illustrates 3 growth velocity pattern scenarios: (1) small for gestational age (blue circles), (2) suspected FGR with an EFW of <3rd percentile (red triangles), and (3) deceleration of growth velocity (orange diamonds). If clinical judgment is based only on time point evaluation of the fetal size, then the 2 last examinations (arrows: the blue circles and red triangles) would be both classified as FGR and the orange diamonds would be classified as appropriate for gestational age. However, the blue circles indicate a fetus with a maintained growth velocity, which indicates a constitutionally small fetus, whereas the orange diamonds indicate a fetus with a deceleration of growth velocity, although above the 10th percentile. The study by Sovio et al¹⁹ showed that in fetuses with a birthweight of <10th percentile (blue circles and red triangles), only fetuses with the AC growth velocity in lowest deciles (red triangles) have a significantly higher risk of adverse outcome than the fetuses in the control group; moreover, there was no difference between fetuses with a birthweight of <10th percentile and maintained growth velocity (blue circles) and fetuses in the control group in terms of adverse outcomes. To identify suspected FGR, Doppler evaluation of uteroplacental-fetal circulation is important (Figure 3).

AC, abdominal circumference; EFW, estimated fetal weight; FGR, fetal growth restriction; GA, gestational age.

Considering suspected FGR only in terms of fetal AC or EFW size of <10th percentile assumes that fetal size predicts fetal well-being in terms of oxygenation and acid-base status.²⁴ Approximately 60% to 70% of fetal deaths, especially toward term, occur in AGA fetuses,^10,25,26 and a recent case-control study showed that 88% of stillbirths had histologic features of hypoxia on postmortem or placental examination.²⁷ Individualized growth assessment is a method for the evaluation of fetal growth and neonatal growth outcome in which each fetus is its own control, based on estimates of individual growth potentials.²¹ This type of longitudinal evaluation addresses the concept of possible fetal growth failure even if the EFW is >10th percentile²⁸ (Figure 2), hence identifying potential “at-risk” fetuses that might otherwise be missed. Even within a population of fetuses whose EFW was <10th percentile recruited within the PORTO study, a recent reanalysis has shown that 38% showed a normal fetal growth pattern.²⁹ Using individualized growth assessment, different growth trajectories were associated with fetal Doppler and cardiac function and delivery outcome.²⁹

Although a reduction in fetal growth velocity may not be sufficient as a stand-alone criterion for suspected FGR,³⁰ it should alert the clinician to consider additional monitoring parameters, such as Doppler velocimetry as an important adjunct to identify fetal compromise or uteroplacental impairment (Figure 3).

FIGURE 3. Uteroplacental-fetal vascular components evaluated with Doppler velocimetry.

The figure represents the main uteroplacental-fetal vascular components that can be evaluated with Doppler velocimetry and that play a role in the diagnosis, surveillance, and/or time of delivery of fetuses with suspected FGR. A, The umbilical arteries reflect the impedance to blood flow on the fetal side of the placenta. With increasing GA, umbilical impedance becomes progressively reduced. With progressive damage of the chorionic villi, the umbilical artery impedance will increase until absent end-diastolic flow and reversed end-diastolic flow supervene. The umbilical artery PI does not correlate with fetal hypoxemia. B, The uterine artery blood flow reflects the resistance to the blood flow at the maternal side of the placenta. Similarly, to the umbilical artery uterine impedance decreases with GA in physiological states, whereas the resistance to the blood flow remains elevated in abnormal placental invasion and is predictive of FGR especially when associated with preeclampsia. C, The ductus venosus represents a fetal shunt that carries oxygenated blood at high velocity from the umbilical vein, through the foramen ovale into the left atrium and then left ventricle of the heart; the crista dividens separates this oxygenated blood from the deoxygenated blood of the inferior vena cava that passes through the right ventricle and systemic circulation. Physiologically, it appears tri-phasic with low impedance during the a-wave. An absent or reversed ductus venosus a-wave is an expression of an increased end-diastolic intracardiac pressure because of increased resistance in afterload and/or an expression of progressive dilatation of the ductus venosus to increase the delivery of the oxygenated blood directly to the myocardium and fetal brain. Changes in the ductus venosus waveform are associated with increased risk of perinatal mortality and morbidity. D, The middle cerebral artery normally shows high impedance to blood flow. Cerebral blood flow redistribution is a fetal adaptive response to hypoxemia and/or hypercapnia and can be identified as low impedance in the middle cerebral artery. For further discussion on Doppler velocimetry and indices, please refer to the recently published International Society of Ultrasound in Obstetrics and Gynecology guideline by Bhide et al.²⁰

FGR, fetal growth restriction; GA, gestational age; PI, pulsatility index.

Early and Late Fetal Growth Restriction

There are 2 widely described phenotypes of suspected FGR, early and late, commonly distinguished by time of diagnosis before and after 32 weeks of gestation, respectively.³¹ The main differences in etiology and Doppler findings that can assist in the identification of early and late suspected FGRs are shown schematically in Figure 4. Increased Doppler PI in the umbilical and uterine arteries, a sign of placental insufficiency, is much more frequent in early suspected FGR and comparatively rare in late suspected FGR.^32,33 Doppler signs of cerebral blood flow redistribution reflect fetal hypoxemia and aid in the identification of late suspected FGR.^31,34,35 Doppler velocimetry abnormalities of these vessels in suspected fetuses with FGR are associated with an increased risk of adverse perinatal outcomes.^{17,19,32,35–40} Therefore, where fetal size alone is not sufficient to identify suspected FGR, Doppler velocimetry of the uterine, umbilical, and middle cerebral arteries evaluates the full spectrum of fetal vascular compensatory abnormalities attributable to uteroplacental impairment across gestation and is crucial for a more accurate diagnosis of both early and late FGRs. The Trial of Umbilical and Fetal Flow in Europe (TRUFFLE) group expert consensus defined early FGR as fetal AC of <10th percentile and UA-PI of >95th percentile.⁴¹ A similar definition was described by the Irish consortium’s PORTO study based on the results of their study of suspected SGA and FGR fetuses.¹² Moreover, it is important to recognize that a lack of evidence for middle cerebral artery Doppler in delivery timing for suspected FGR does not disqualify this modality as important for its identification. Furthermore, the Delphi consensus criteria (on which definitions the ISUOG guideline is based) were more predictive of adverse neonatal outcomes than an EFW of <10th percentile, though both performed poorly.⁴²

FIGURE 4. Different clinical and biophysical characteristics of early and late suspected FGR.

The figure illustrates the different clinical and biophysical characteristics of early and late suspected FGR. Early suspected FGR (<32 weeks of gestation) is a rare condition and is characterized by placental insufficiency and reduced placental vascular perfusion on maternal (uterine arteries) and fetal interface (umbilical arteries). The fetus is usually very small, and profound Doppler changes might be present, with absent or reverse ductus venosus a-wave, which are associated with an increased risk of perinatal mortality and morbidity. The main challenge is the management because of prematurity and/or maternal hypertension (70% of cases). Late suspected FGR (>32 weeks of gestation) is a frequent condition caused by placental dysfunction. Vascular perfusion at the maternal and fetal interface is usually normal. The fetus is not necessarily very small, and signs of cerebral blood flow redistribution (low impedance in the middle cerebral artery and/or altered ratio to the umbilical artery) might be the only Doppler sign. The main challenge is the diagnosis.

AEDF, absent end-diastolic flow; BPP, biophysical profile; cCTG, computerized cardiotocography; FGR, fetal growth restriction; REDF, reversed end-diastolic flow; STV, short-term variation; UA, umbilical artery.

Monitoring of Fetal Growth Restriction

In early-onset suspected FGR, fetal deterioration is heralded by progressively increasing UA impedance, most commonly expressed as PI, resistance index or S-to-D (systolic-to-diastolic) ratio. This change may occur during many weeks before severe fetal cardiovascular and metabolic deterioration^43,44 and may be less predictable if pre-eclampsia supervenes.⁴⁵ The absence of umbilical end-diastolic velocities as a reflection of worsening uteroplacental insufficiency precede reversed end-diastolic velocity in the UA and eventually ductus venosus abnormalities^46–48 (Figure 3).

Strong evidence supports the use of ductus venosus Doppler in early suspected FGR for both prognostic and monitoring purposes.^49–54 The ductus venosus deviates oxygenated blood from the umbilical vein toward the heart at the confluence of the inferior vena cava with the right atrium (Figure 3), and the crista dividens preferentially streams this oxygenated blood toward the left atrium and cerebral vessels. The ductus venosus waveform commonly appears triphasic with the a-wave, a sensitive reflection of the atrial contraction. An augmented a-wave is provoked by increased preload, inotropic drive because of increased afterload, hypoxia, or a combination of these factors and/or as an expression of progressive dilatation of the ductus venosus to increase the delivery of the oxygenated blood directly to the systemic circulation.^55–57 In extreme cases, the a-wave reaches the baseline or is reversed; these findings presage an increased risk of fetal death (40% –70%), neonatal mortality^50,58,59; they represent late-stage acidemia independent of gestational age (GA) at delivery and frequently appear with the onset of spontaneous decelerations on cardiotocography.^46,49,60 When these advanced cardiovascular changes are observed, spontaneous heart rate decelerations, an abnormal biophysical profile followed by fetal death may be anticipated within 1 week in 40% to 70% of fetuses.^32,35,46,60

In late-onset suspected FGR, deterioration is far more subtle and is characterized by either a mild elevation of the UA Doppler PI, a decline in cerebral PI, or sometimes both.^24,36,61 This is the reason why a ratio between the middle cerebral artery and UA-PI, the so-called cerebroplacental ratio, or its inverse, the so-called umbilical-cerebral ratio, has been proposed. Abnormal Doppler findings in the middle cerebral artery, cerebroplacental or umbilical-cerebral ratio, reflect the hemodynamic phenomenon of blood flow redistribution to the fetal brain. Moreover, this phenomenon has been called “brain sparing”⁶² and is triggered by decreased fetal oxygenation as shown in both animals^63,64 and humans.^62,64–66 Cerebral vasodilatation can easily be evaluated by measurement of middle cerebral artery PI (Figure 3), which becomes progressively lower with worsening hypoxemia.

It is incontrovertible that there is a link between low middle cerebral artery Doppler impedance and adverse perinatal outcome. Cerebral blood flow redistribution is widely reported to be associated with poorer perinatal outcomes,¹⁷ including fetal death,³⁵ higher risk of cesarean delivery,^67–69 increased risk of abnormal neurodevelopment at birth⁷⁰ and at 2 or 3 years of age.^37,71,72 Recent systematic review and meta-analysis evidence support these findings.^39,40 The assessment of cerebral blood flow redistribution is particularly useful in identifying late suspected FGR, (Figure 4) and may improve the prediction of adverse outcomes.⁷³ Indeed, the alteration of the UA Doppler waveform is uncommon in late preterm and term suspected FGR^32,35 and does not predict adverse pregnancy outcomes accurately.^61,74 Hence, assessment of cerebral blood flow redistribution plays an important role to identify and monitor fetuses exposed to placental dysfunction and hypoxemia. Although it remains to be elucidated whether cerebral redistribution is causative for adverse outcomes (as opposed to being a marker for it) assessment of cerebral Doppler velocimetry remains important for diagnosis and surveillance purposes.

The goals of fetal monitoring are the prevention of fetal death and delivery of the fetus in the best possible condition. These goals can only be successfully achieved when the disease-specific features of the fetal condition are taken into consideration.⁷⁵ In early-onset suspected FGR, this requires Doppler evaluation of the UA and ductus venosus Doppler and to increase surveillance frequency when there is evidence of progressing cardiovascular dysfunction in informing decisions regarding in-patient vs outpatient monitoring.^49–52 In late-onset suspected FGR, Doppler examination of the middle cerebral artery and consideration of its ratio to UA is required to make similar decisions.^76–80 Accordingly, not evaluating these vessels in the specific clinical context of early- and late-onset suspected FGR is likely to result in suboptimal surveillance intervals and possibly increased risk of fetal death because features that indicate impending deterioration are not captured.

Delivery Timing in Fetal Growth Restriction

The timing of delivery in any condition that places the fetus at risk is indicated when the risk of remaining in utero exceeds those of delivery. In suspected FGR, this balances changes with advancing GA requiring adjustment of the delivery thresholds with advancing GA. The TRUFFLE study of early-onset suspected FGR was conducted in 20 European centers between 2005 and 2010,^41,81 and all women received surveillance using UA Doppler and computerized cardiotocography (cCTG). Participants were randomized to delivery according to 3 management arms: “early” changes (ductus venosus PI of >95th percentile), “late” changes (ductus venosus with absent or reversed “a”-wave), and cCTG short-term variation (STV) thresholds (Figure 5).^41,81 An absolute indication for delivery irrespective of allocated randomization arm, the so-called safety net, was represented by spontaneous repeated persistent unprovoked decelerations in all 3 arms and short-term variation <2.6 milliseconds (ms) at 26 0/7 to 28 6/7 weeks of gestation and <3 ms at 29 0/7 to 31 6/7 weeks of gestation in the ductus venosus arms, respectively. The protocol recommended delivery no later than 32 weeks of gestation with reversed umbilical end-diastolic flow and no later than 34 weeks of gestation with absent umbilical end-diastolic flow. The TRUFFLE study concluded that timing delivery based on ductus venosus Doppler measurement in conjunction with STV obtained from cCTG led to the best long-term (2-year neurodevelopmental) outcome in survivors.^41,81 This constitutes the strongest evidence in favor of the use of ductus venosus Doppler combined with cCTG for monitoring and triggering delivery in early suspected FGR.

FIGURE 5. Schematic representation of the TRUFFLE randomization and the “safety net”.

The inclusion criteria were singleton fetus between 26 0/7 and 31 6/7 weeks of gestation with an estimated fetal weight >500 g, an abdominal circumference <10th percentile, and an umbilical artery >95th percentile, with a normal ductus venosus pulsatility index and short-term variation. Chromosomal and congenital anomalies constituted an exclusion criteria.

TRUFFLE, Trial of Umbilical and Fetal Flow in Europe.

The importance of cCTG concerning the TRUFFLE study findings is that it cannot be generalized to a simple visual interpretation of the fetal heart rate trace. cCTG has been used successfully in identifying acidemia in fetuses with growth restriction not only in Europe^82–84 but also in the United States with prospective validation of its accuracy (Figure 6).^85,86 Although it is true that visual interpretation of cardiotocography based on late decelerations is moderately powerful in detecting or ruling out fetal hypoxia and antepartum acidemia in the setting of suspected FGR,^87–94 this results in large inter- and intraobserver variations. This prompted Dawes and Redman to develop a numeric analysis of antenatal fetal heart rate changes in the 1970s.⁹⁵ With this algorithm, the STV is defined as the mean difference in fetal pulse at intervals between successive epochs of 3.75 sec, after exclusion of decelerations.⁹⁶ Obvious advantages are greater reproducibility and the conversion of fetal heart rate variation into objective numeric data. STV increases with GA and so does the lower limit of normality.⁹⁷ In suspected FGR, STV (and the similar long-term variation) is correlated to fetal oxygen partial pressure obtained from umbilical cord blood raphy.^84,98 Very low STV is associated with fetal acidemia^{84,85,99–101} with a correlation coefficient of 0.72 between STV and UA pH in recordings made <4 hours before cesarean delivery.⁸⁵ With an STV ≤3 ms, 54% of fetuses with early suspected FGR had metabolic acidemia, compared with 10% of fetuses with an STV >3 ms.¹⁰⁰ In 1 large study, where STV was >3 ms, there was no case of fetal death.⁹⁹ When assessed longitudinally in suspected FGR, late decelerations and reduced fetal heart rate variation appeared at more or less the same time, most likely because of the onset of fetal hypoxemia (and academia).^82,94

FIGURE 6. Main characteristics of computerized cardiotocography and STV.

The figure summarizes the main characteristics of computerized cardiotocography and STV. An example in which STV is useful is represented.

fHRV, fetal heart rate variability; STV, short-term variation.

This relationship was further analyzed in the TRUFFLE study. Of 217 fetuses delivered before 32 weeks of gestation because of abnormal cardiotocography or ductus venosus, abnormal fetal heart rate was the indication for delivery in 165 fetuses. Of those fetuses, STV was reduced in 41%, late fetal heart rate decelerations were present in 48%, and both alterations were present in 11%,¹⁰² emphasizing the value of both parameters in giving a high degree of confidence in decision-making. Hence, given the favorable outcome data of the TRUFFLE trial, both STV and occurrence of decelerations should be taken into account and delivery undertaken if 1 of the 2 alterations becomes abnormal. Although thresholds denoting neonatal acidemia are described, the exact values of STV that mandate delivery in the context of clinical management of suspected FGR are still unknown.¹⁰⁰

As far as UA Doppler is concerned, a Cochrane systematic review and meta-analysis explored Doppler ultrasound vs no Doppler or Doppler not revealed in high-risk pregnancy.¹⁰³ The pooled data of the use of UA Doppler velocimetry in high-risk pregnancies showed fewer perinatal deaths (RR, 0.71 [95% CI, 0.52–0.98]; 1.2% vs 1.7%; number needed to treat, 203 [95% CI, 103–4352]; evidence graded moderate).¹⁰³ Similarly, there were fewer fetal and neonatal deaths, although this did not reach statistical significance: fetal death (RR, 0.65; 95% CI, 0.41–1.04) and neonatal deaths (RR, 0.81; 95% CI, 0.53–1.24; evidence graded low).¹⁰³ These findings underline the importance of UA Doppler in monitoring high-risk pregnancies, including suspected FGR, but importantly, no evidence exists from an RCT to inform the timing of delivery with different degrees of abnormality of the UA Doppler waveform.

Hence, the relationship between the findings from fetal Doppler monitoring and findings from cardiotocography monitoring are not directly consistent with one another,¹⁰⁴ underlying the usefulness of utilizing both modalities to assess fetal condition.

Comparison Between Society for Maternal-Fetal Medicine and International Society of Ultrasound in Obstetrics and Gynecology Guidelines and Considerations on Suspected Fetal Growth Restriction

The key differences between the SMFM and ISUOG recommendations concerning the diagnosis, surveillance, and time of delivery of fetuses with suspected FGR are summarized in Table 2. ISUOG guidelines adopt the Delphi consensus diagnostic criteria for FGR, thus introducing the concept of growth velocity and Doppler evaluation to define and differentiate between early and late suspected FGR as different entities. The SMFM guideline adopts a definition of FGR as fetal size (EFW or AC) of <10th percentile.

TABLE 2.

Differences between Society for Maternal-Fetal Medicine and International Society of Ultrasound in Obstetrics and Gynecology recommendations in the diagnosis, surveillance, and time of delivery decision of fetuses with suspected fetal growth restriction

Variable Diagnosis of suspected FGR	SMFM recommendations Estimated fetal weight or abdominal circumference <10th percentile	ISUOG recommendations Delphi consensus criteria
	Surveillance
UA	Yes	Yes
Ductus venosus	No	Yes
Middle cerebral artery	No	Yes
Cardiotocography	Yes	Yes
Short-term variation	No	Yes
	Delivery timing
Ductus venosus	No	≥26 0/7 to 31 6/7 wk: ductus venosus a-wave absent or reverse
UA reverse end-diastolic flow	30–32 wk	>30 0/7 to 32 0/7 wk
UA absent end-diastolic flow	33–34 wk	>32 0/7 to 34 0/7 wk
UA pulsatility index >95th percentile	37 wk	≥36 0/7 to 37 6/7 wk
Middle cerebral artery	No	38 0/7 to 39 0/7 wk
Short-term variation	No	≥26 0/7 to 28 6/7 wk: <2.6 ms ≥29 0/7 to 31 6/7 wk: <3.0 ms ≥32 0/7 wk: <3.5 ms >34 0/7 wk: <4.5 ms

FGR, fetal growth restriction; ISUOG, International Society of Ultrasound in Obstetrics and Gynecology; SMFM, Society for Maternal-Fetal Medicine; UA, umbilical artery.

In general, the SMFM guideline relies particularly on UA Doppler and conventional cardiotocography in monitoring and delivery planning of both early and late suspected FGRs, recommending against the use of middle cerebral artery and ductus venosus Doppler. Concerning surveillance of suspected FGR, the ISUOG guideline recommends different monitoring modalities (umbilical and middle cerebral artery, ductus venosus Doppler, and cCTG) depending on the context in early and late suspected FGR.

Finally, important differences exist regarding the timing of delivery decision in early suspected FGR. Ductus venosus Doppler alteration and low STV on cCTG are recommended in the ISUOG guideline, but explicitly not in SMFM guideline. These additional tools may finesse the delivery guidance based on UA Doppler after 30 weeks of gestation on which the guidelines otherwise largely agree.

Conclusion

A substantial body of evidence has been gathered in the past 4 decades that has helped to clarify the clinical features of suspected FGR across GA and the role of fetal surveillance parameters to predict the clinical course and fetal status at the time of testing. Nevertheless, important differences remained in the recommended management of suspected FGR. Management is a composite process consisting of diagnosis, surveillance, and time of delivery: Doppler evaluation of uteroplacental and fetal vessels all play a separate and defined role in each of these. The evidence underlying fetal placental Doppler as essential components in the management of FGR formed a cornerstone of the ISUOG guideline,⁴ but not in the SMFM guideline,³ and we consider this to be an important omission. The major areas of contention were the definition of suspected FGR; the use of Doppler parameters for surveillance, quite separately, for decision for delivery; and the value of cCTG vs conventional cardiotocography assessment. These differences might have a profound effect on clinical management, namely, frequency of follow-up, time of delivery, and consequent fetal death rate, and short-and long-term outcomes for these fetuses and infants.

Evaluation of a single biometric fetal measurement defines the size and not fetal growth. Fetoplacental Doppler and evaluation of fetal growth velocity are important tools to identify uteroplacental insufficiency associated with suspected FGR, allowing its differentiation from a constitutionally small fetus.

Level 1 evidence supports the application of UA and ductus venosus Doppler in the monitoring of early FGR. High-quality observational studies support the application of the middle cerebral artery and/or its ratios to UA in the surveillance of late suspected FGR. Collectively, these Doppler parameters can predict fetal deterioration and guide optimal surveillance intervals.

cCTG analysis of fetal heart rate variation and, in particular, the STV provided an objective basis on which to evaluate the fetal condition. Low STV was associated with fetal acidemia and fetal death, and there was strong evidence that monitoring early suspected FGR with ductus venosus Doppler and STV from cardiotocography leads to optimal perinatal outcome.

Glossary

AC

abdominal circumference

AGA

appropriate for gestational age

CI

confidence interval

cCTG

computerised Cardiotocography

CPR

cerebro-placental ratio

DIGITAT

Disproportionate Intrauterine Growth Intervention Trial at Term

(A)(R)EDF

(Absent)(Reversed) End-diastolic flow

EFW

estimated fetal weight

FGR

fetal growth restriction

GA

gestational age

ISUOG

International Society of Ultrasound in Obstetrics and Gynecology

LGA

large for gestational age

MCA

middle cerebral artery

msec

milliseconds

pH

per hydrion

PI

pulsatility index

PORTO

Prospective Observational Trial to Optimize Pediatric Health

RR

relative risk

SGA

small for gestational age

SMFM

Society of Maternal-Fetal Medicine

STV

short-term variation

TRUFFLE

Trial of Umbilical and Fetal Flow in Europe

UA

Umbilical artery

UCR

Umbilical-cerebral ratio