Diagnosis and management of fetal growth restriction: the <scp>ISUOG</scp> guideline and comparison with the <scp>SMFM</scp> guideline

Abstract

Both published in 2020, the guidelines of the Society for Maternal–Fetal Medicine (SMFM)1 and those of the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG)2 on fetal growth restriction (FGR) make recommendations in respect of the diagnosis, monitoring and delivery timing of FGR pregnancies. The SMFM guidance1, being a document aimed primarily at USA practitioners, is authored by three USA-based authors. As ISUOG is an international society, the authorship of the ISUOG FGR guideline is drawn globally (Europe, the Americas and Asia-Pacific)2. There are important differences between the two guidelines, which are in many places contradictory. Thus, instead of achieving a highly desirable unanimity in respect of the three main aspects of FGR management, namely diagnosis, monitoring and delivery timing, it seems that an international consensus still remains elusive. In this Opinion, we debate the relative merits of the conflicting recommendations provided in the ISUOG and SMFM guidelines on FGR, given the major impact these differences might have on the management of FGR pregnancies. The SMFM guideline defines FGR as estimated fetal weight (EFW) or abdominal circumference (AC) < 10th percentile1. Though simple to apply, this definition classifies small-for-gestational-age (SGA) fetuses as FGR and would potentially label 10% of all babies as being growth restricted. While there is no doubt that EFW or AC < 10th percentile is associated with increased risk of adverse outcomes3, 4, this definition takes perinatology back to the 1990s, when there was intractable controversy about whether SGA represents a surrogate for FGR and, if not, why. The SMFM definition of FGR as fetal size below the 10th percentile does not take into account the fetal growth trajectory or functional indices, such as Doppler of the uteroplacental and fetoplacental circulations5. By contrast, the ISUOG guidelines recommend the adoption of the Delphi consensus criteria for the diagnosis of FGR (Table 1)6, which seek to differentiate FGR from SGA and include both fetal growth trajectory and Doppler findings as additional considerations. The Delphi consensus was based on the opinions of leading world experts in FGR management, some of whom were from the Americas. The risk of a ‘catch all’ definition based on the 10th percentile is that a proportion of babies whose size is below this cut-off are labeled as growth restricted and undergo enhanced monitoring and potentially intervention when it is not necessary7. In this context, if one were to pick an appropriate percentile at which the risk of adverse outcome is significant then this would most likely be the 3rd percentile rather than the 10th percentile7-10. Importantly, any size threshold that does not take into account the fetal growth pattern carries the risk of missing fetuses whose growth trajectory slows and who are therefore at risk of adverse outcome, even if their absolute size is greater than the 10th percentile5, 10-16. Furthermore, using the SMFM definition, only one-third of babies that are stillborn at or near term would be considered to be growth restricted17. Indeed, unrecognized FGR represents the single most common risk factor for stillbirth18. Finally, early-onset (before 32 weeks) and late-onset (after 32 weeks) FGR have different clinical manifestations and characteristics19-22. The diagnostic criteria used in term FGR do not apply well at, for example, 26 weeks, as the Doppler profile is quite different6, 19, 20, 23. Though it is possible to argue (as the SMFM does) that AC or EFW < 10th percentile de facto defines FGR and that Doppler assessment is irrelevant because only the fetal size is important, there is little doubt that Doppler criteria better identify those fetuses that show signs of placental impairment or dysfunction8, 12, 19, 24, 25 and are at higher risk of adverse outcome26. A fetal size cut-off defines a different population in early gestation from that in term or late preterm pregnancy. A study comparing the SMFM and ISUOG definitions of FGR in a high-risk cohort, published in this issue of the Journal27, showed that, although the SMFM definition had greater sensitivity for predicting birth weight < 10th percentile, the ISUOG definition had a higher specificity. It should be borne in mind, however, that the finding of higher sensitivity using the SMFM guideline was inevitable, given the outcome was prediction of birth weight < 10th percentile based on prenatal ultrasound showing AC or EFW < 10th percentile. Both the ISUOG and the SMFM guidelines make recommendations in respect of monitoring FGR pregnancies and timing delivery using umbilical artery (UA) Doppler. With respect to monitoring frequency, the ISUOG guideline suggests that, in the presence of absent (AEDF) or reversed (REDF) end-diastolic flow in the UA, monitoring before 34 weeks should be performed every 2–3 days2. The SMFM recommends hospitalization and surveillance with cardiotocography (CTG) 1–2 times per day when UA-REDF is detected1. Given that there is no good evidence underlying this guidance, both SMFM and ISUOG provide appropriately cautious advice. With respect to delivery timing, both guidelines graduate their recommendations for delivery based on gestational age and the degree of abnormality of the UA Doppler waveform (REDF, AEDF and raised impedance) in broadly similar ways after 30 weeks, with neither suggesting that FGR should be left undelivered beyond 39 weeks. The difference is that the ISUOG guideline recognizes the absence of evidence for basing delivery timing on UA Doppler whereas the SMFM guideline classifies as Grade 1B the recommendation for delivery prior to 37 weeks in the presence of UA Doppler abnormalities (strong recommendation, moderate-quality evidence). This is important as there is no evidence from randomized controlled trials (RCTs) that informs what are usually very carefully balanced decisions and where clinical experience and expertise may be in the mother and fetus' best interest. A meta-analysis of RCTs on the application of UA Doppler vs no UA Doppler in high-risk pregnancies showed reduced perinatal mortality in women in whom UA Doppler was performed28; however, to our knowledge, no RCT on delivery timing based on UA Doppler is available to date. The SMFM guideline recommends not to use middle cerebral artery (MCA) Doppler in the routine management of early- or late-onset FGR (Grade 2B recommendation). It is not clear to what particular aspect of clinical management this recommendation refers. A low MCA impedance indicates redistribution of fetal cardiac output preferentially to crucial fetal organs, such as the brain, heart and adrenal glands, in response to fetal hypoxemia and/or hypercapnia29-32. There is abundant evidence showing that cerebral blood-flow redistribution is associated with a spectrum of adverse short- and long-term outcomes33-44. Moreover, in late-onset FGR, cerebral blood-flow redistribution may be the only Doppler sign to suggest placental dysfunction19, 36, 45. In view of these data, the ISUOG guideline recommends that signs of cerebral blood-flow redistribution should be considered as a criterion to identify late FGR, in line with the Delphi consensus criteria, and that twice-weekly surveillance might be carried out in its presence based on available data21. The association between two phenomena does not imply a causative link, and there is still no strong evidence for the use of MCA in the context of triggering delivery46. However, the issues of identifying an ‘at-risk’ baby and of enhanced monitoring are distinct from determining timing of delivery and, in contemporary practice, the evaluation of cerebral blood-flow redistribution should not in our view be neglected as it is justified in having a role in diagnosis and monitoring, especially in late FGR. Apart from the dissimilarities in respect of FGR diagnosis, the greatest and arguably most clinically important difference between the SMFM and ISUOG guidelines is that the SMFM guideline does not differentiate between management of early-onset and late-onset FGR. In early FGR, the risks of prematurity must be balanced against prolonged intrauterine fetal exposure to hypoxemia and acidemia, both of which are associated with perinatal morbidity and mortality47. Therefore, gestational age requires consideration when interpreting test results with a view to decide on elective delivery. Extremely low values of short-term variation (STV) in computerized analysis of CTG (cCTG) is a reliable predictor of metabolic acidemia at delivery or intrauterine death48-55. Increased pulsatility in the ductus venosus (DV) reflects myocardial decompensation resulting in increased end-diastolic ventricular pressure and decreased forward flow velocities during atrial contraction (a-wave) as early changes and absent or reversed flow as late changes of its waveform56. These changes might also reflect a progressive dilatation of the DV as an adaptive response to worsening hypoxemia56. DV abnormalities represent a relatively late biophysical sign in early FGR and are associated with significantly increased risk of stillbirth23, 48, 57, 58. The TRUFFLE (Trial of Randomized Umbilical and Fetal Flow in Europe) multicenter study assessed whether changes in the fetal DV could be used as indication for delivery instead of CTG STV alone in pregnancies with early-onset FGR59. Pregnancies were allocated randomly to one of three arms in which the timing of delivery was determined based on reduced STV on cCTG, early DV Doppler changes or late DV Doppler changes. The primary outcome of survival without neurodevelopmental impairment at 2 years of age was significantly more common in infants of women who were assigned to delivery according to late DV changes (95%; 95% CI, 90–98%) compared with those assigned to delivery according to cCTG (85%; 95% CI, 78–90%; P = 0.005). Importantly, safety-net criteria for delivery applied to all patients irrespective of the randomization group. These comprised cut-off ‘rescue’ values of STV in the two DV arms and, additionally, spontaneous repeated unprovoked heart rate decelerations on CTG and UA Doppler showing REDF from 30–32 weeks and AEDF from 32–34 weeks onwards in all three arms. The safety-net criteria indicated delivery in 33%, 23% and 15% of cases in the late-DV, early-DV and cCTG groups, respectively59. Thus, a key message from the TRUFFLE study is that the best outcome in early FGR is obtained when fetuses are monitored and delivery timing is decided based on both DV and cCTG60, where only cCTG can allow an objective and reproducible measurement of fetal heart rate STV. As the cCTG safety-net criteria were an integral part of the TRUFFLE study protocol, they are also central to the recommendations of the ISUOG guidelines for monitoring and timing of delivery of pregnancies with early-onset FGR. Clinical guidelines should justify their recommendations based on data acquired and conclusions drawn from studies providing the highest quality of evidence, i.e. well-performed randomized trials (Grade A evidence). It is therefore surprising that the SMFM guideline on FGR recommends against both Doppler assessment of the DV and use of cCTG STV in the routine clinical management of early-onset FGR, and classifies the recommendation to not use Doppler and cCTG as Grade 2B. In this context, it is questionable whether it is appropriate to refer to ‘routine’ clinical management of early-onset FGR, which is a rare disease requiring specialist input61. Early FGR is not a ‘routine’ condition and should not be managed in a clinical setting lacking expertise in maternal and fetal Doppler investigations. We appreciate that the SMFM guideline was developed based on many considerations, not least in relation to taking into account training, ultrasound and cCTG resources. However, guidelines should primarily aim to improve clinical care based on the best available evidence and technology rather than by maintaining the ‘status quo’ or recommending management that is available in a routine maternity setting.

Topics & Concepts