INTRODUCTION —
Selective fetal growth restriction (sFGR) refers to growth restriction of one fetus of a monochorionic twin pair, primarily caused by discordant placental sharing. By comparison, placenta-mediated growth restriction in dichorionic twins and singletons is due to uteroplacental insufficiency [1]. sFGR is a common complication of monochorionic twins and associated with significant risks for perinatal morbidity and mortality [2-4].
This topic will discuss the pathophysiology, diagnosis, and management of sFGR. The diagnosis and management of growth restriction in dichorionic twins and singletons are reviewed separately.
●(See "Fetal growth restriction: Screening and diagnosis" and "Fetal growth restriction: Evaluation".)
PATHOPHYSIOLOGY —
The placental "vascular equator" is an imaginary line that can be drawn along the vascular anastomoses between twins sharing a monochorionic placenta. This line can be used to estimate the percentage of placental territory belonging to each fetus. In sFGR, the placental territory is distributed unequally between twins: one twin benefits from having the majority share of the placenta while the growth-restricted twin is supported by a smaller portion of the placenta (figure 1 and picture 1) [5-8].
Other anatomic factors also contribute growth restriction. For example, a velamentous placental cord insertion can impair the twin's ability to access its placental territory. Velamentous placental cord insertion is found in up to approximately 30 percent of monochorionic twins and is strongly associated with both sFGR (odds ratio [OR] 9.24, 95% CI 2.05-58.84) and growth discordance ≥25 percent (OR 6.81, 95% CI 1.67-34.12) [9-12].
The number and size of the vascular anastomoses between the twins also play a role as they affect volume exchange and transfer of nutrients and oxygen between twins [13-16]. Increasing discordance in placental share increases net volume flow between the twins, leading to more interdependent circulations [10,14,17-19]. Large bidirectional anastomoses, particularly arterioarterial anastomoses, can compensate for unequal placental sharing and mitigate the adverse effect of a small territory on growth [20,21]. As a result, clinical deterioration is often less predictable and slower than would be anticipated in FGR, complicating singleton or dichorionic twin pregnancies [22,23].
INCIDENCE —
sFGR affects 10 to 15 percent of monochorionic twin pregnancies [24,25].
CLINICAL PRESENTATION —
Crown-rump length (CRL) discordancy of a monochorionic pregnancy on a first-trimester ultrasound can be the first sign of developing growth restriction or discordancy [26]. However, most cases are initially detected in the early second trimester as a result of routine sonographic monitoring of monochorionic pregnancies (see 'Diagnostic criteria' below). Initial identification in the third trimester is less common and associated with more favorable outcomes [27,28].
A common protocol for monitoring monochorionic twins is described in the table (table 1) and will identify sFGR, twin-twin transfusion syndrome (TTTS), and twin anemia-polycythemia sequence (TAPS).
Accurate determination of chorionicity and gestational age are crucial to appropriately diagnose, monitor, and manage monochorionic pregnancies. If CRLs are discordant, the CRL of the larger twin is used for assignment of gestational age in naturally conceived pregnancies. (See "Diagnosis and outcome of first-trimester growth delay", section on 'Twin pregnancies' and "Twin pregnancy: Overview", section on 'Assessment of chorionicity and amnionicity' and "Twin pregnancy: Overview", section on 'Determination of gestational age'.)
DIAGNOSIS
Diagnostic criteria — The diagnosis of sFGR is based on fetal biometric measurements, growth discordance, and umbilical artery (UA) Doppler parameters. Definitions of sFGR vary among guidelines but, at a minimum, include an estimated fetal weight (EFW) <10th percentile for one twin plus EFW discordance ≥25 percent [29]. These International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) criteria are used by many clinicians and identify pregnancies at higher risk for adverse outcome.
Given inconsistencies for the definition of sFGR in the literature, an expert consensus using the Delphi methodology developed the following criteria that also may be used for diagnosis [29,30]:
●EFW <3rd percentile of one twin
or
●At least two of the following:
•EFW <10th percentile for one twin
•Abdominal circumference <10th percentile for one twin
•UA pulsatility index (PI) >95th percentile for the smaller twin
•Weight discordance ≥25 percent
Percent weight discordance is calculated using the following equation:
(EFW larger twin weight – EFW smaller twin weight) / (EFW larger twin weight × 100)
The UA Doppler waveform reflects the relative size differences in placental share and the magnitude of hemodynamically significant vascular communications between the twins [31]. In monochorionic twins with sFGR, it may show diminished diastolic flow, persistently absent/reversed diastolic flow, or intermittently absent/reversed diastolic flow due to transmitted waveforms from the larger into the smaller twin's cord via one or more large arterioarterial anastomoses [31]. These Doppler patterns can be observed from very early in pregnancy, may differ between the two UAs and can vary over time [22].
Doppler technique — Proper acquisition of the UA waveform is required for accurate assessment and interpretation.
●Reproducibility of measurements is greater if the measurement is obtained in a free loop of the umbilical cord with an insonation angle close to 0 degrees during fetal quiescence.
●Focal zone, gain, and pulse repetition frequency should be adjusted to maximize the frame rate.
●The sweep speed should initially be set to visualize four to six waveforms and should take up approximately 75 percent of the Doppler screen [32]. To evaluate variation or oscillation of end-diastolic velocity in the UA, the sweep speed needs to be slowed to demonstrate the classic pattern in >6 waveforms (waveform 1).
●Maternal breath holding may be required to exclude interference during measurement of the waveform [31,33].
●Sampling both UAs can be helpful for complete assessment as the flow pattern may differ [22]. However, interarterial anastomoses, if present, can equalize blood flow between arteries even though the placental territories supplied by the UAs are different [34].
●UA waveforms are best identified close to the placental cord insertion where intertwin anastomoses on the placenta have the most impact on the UA Doppler flows, particularly in the setting of close placental cord insertions. Evaluation for proximal placental cord insertion (distance between cord insertions below the fifth percentile, or 3.3 to 4.0 cm across gestation) is prudent as intermittent absent or reversed end-diastolic velocity (as in type 3 sFGR, discussed below) is more common in this setting [12,35].
CLASSIFICATION —
The pattern of the umbilical artery (UA) waveform and end-diastolic velocity of the smaller fetus are used to classify sFGR into three types (waveform 2) [36-38]. This is an important designation to make since it is used to guide counseling and management decisions. The classification was based on early data of the clinical course and included pregnancies managed by selective feticide. Subsequent data suggest that type II sFGR has the worst prognosis (highest mortality) [39].
Type 1 sFGR — Type 1 sFGR is characterized by persistently forward UA end-diastolic velocity without variation in the waveform and with normal or elevated resistance. It has a more stable course than types 2 and 3 sFGR and typically has a favorable outcome; the mean gestational age at birth was 35.4 weeks in one large series [36]. In a meta-analysis of observational studies including 786 monochorionic pregnancies complicated by all classes of sFGR, expectantly managed type 1 sFGR was associated with a lower risk for an unanticipated fetal demise and a higher rate of intact survival than type 2 and type 3 sFGR (table 2) [39].
Late-onset sFGR (ie, identified after 26 weeks) is typically type 1. Although these fetuses usually have a benign course, they are at increased risk for hemoglobin differences at birth (ie, twin anemia-polycythemia sequence [TAPS]), which occurs in up to 38 percent of cases [6].
Type 2 sFGR — Type 2 sFGR is characterized by fixed absent or fixed reversed UA end-diastolic velocity without variation of the waveform in the smaller twin. It is associated with a high risk for midtrimester deterioration of the growth-restricted fetus; the mean gestational age at birth was 30.7 weeks in the same large series described above [36]. Further subclassification of type 2 sFGR into 2a (normal middle cerebral artery [MCA] and ductus venosus [DV] Doppler) and 2b (middle cerebral artery-peak systolic velocity [MCA-PSV] >1.5 multiples of median [MoM] or DV absent/reversed a-wave) has been proposed to stratify the prognosis for cases with coexisting twin-twin transfusion syndrome (TTTS) that undergo laser surgery [40].
Although pregnancies with type 2 sFGR are anticipated to have a predictable pattern of deterioration and a longer latency period between diagnosis and deterioration than type 3 sFGR, they have the worst prognosis when expectantly managed due to the significant risk of single fetal demise and preterm birth, as shown in the table (table 2) [39,41].
Type 3 sFGR — Type 3 sFGR is characterized by a pathognomonic UA waveform that has a variable flow pattern that cycles between forward, absent, and reversed flow over a short interval (waveform 1), which is termed intermittent absent/reversed end-diastolic flow (iAREDF). This results from a large artery-to-artery (AA) anastomosis on the placental surface and represents the bidirectional volume flow across these vessels. It is more commonly observed in the UA of the smaller fetus since the interface of the two waveforms is shifted toward the smaller twin and the AA anastomosis has a larger proportionate impact on the fetus with smaller placental share [16]. An AA anastomosis allows perfusion of oxygen and nutrients from the larger fetus to a portion of the smaller twin's placenta; consequently, type 3 sFGR is associated with the largest degree of placental territory discordance [13,19,36,42,43].
These cases have the most unpredictable clinical course, and unanticipated fetal demise can occur in a short interval, even after a reassuring ultrasound assessment. Comparative data are shown in the table (table 2) [39]. In a large multicenter study of 328 pregnancies with type 3 sFGR, single fetal demise occurred in 5.8 percent and double demise occurred in 4.9 percent of expectantly managed cases [44]. Normalization of UA Doppler flow occurred in 13.7 percent of cases and was associated with an improvement in growth of the smaller twin [23].
Deterioration of UA Doppler to persistent absent or reversed flow is a poor prognostic sign, especially in cases diagnosed at an early gestational age. In a study of 245 cases of type 3 sFGR, fetal death occurred in 7 of 12 fetuses (58 percent) diagnosed before 16+5 weeks with Doppler deterioration and 7 of 43 fetuses (16 percent) diagnosed after 16+5 weeks with Doppler deterioration, but only 13 of 190 fetuses (7 percent) without Doppler deterioration [45]. These features may help individualize counseling, surveillance, and management.
These pregnancies are also at the highest risk for neurologic morbidity, particularly of the larger twin, most likely attributable to the more unstable hemodynamic environment [33]. Only 61.9 percent (95% CI 38.4-81.9) had intact survival in a meta-analysis [39].
APPROACH TO DIFFERENTIAL DIAGNOSIS OF sFGR, TTTS, AND TAPS —
A systematic approach is required to arrive at the correct diagnosis and initiate a management plan given the substantial overlap between sFGR, twin-twin transfusion syndrome (TTTS), and twin anemia-polycythemia sequence (TAPS) (table 3) and the possibility that FGR is unrelated to monochorionicity [46]. All of the characteristics that define these conditions need to be evaluated and the final diagnosis should include a description of the entire spectrum of findings. Even when sFGR diagnosed, when TTTS and TAPS were not present at diagnosis of sFGR, TTTS subsequently developed in 10 percent of cases and TAPS subsequently developed in 3 percent of cases in a retrospective series of 177 sFGR cases [47].
●Rule out other causes of sFGR (anatomic and/or chromosomal abnormalities, infection) – Once sFGR or discordant fetal growth is suspected, a detailed anatomic survey should be performed to assess for structural fetal anomalies that may complicate up to 7 percent of monochorionic twin pairs and contribute to abnormal growth [48-51]. Among sFGR cases, a major fetal anomaly in at least one twin (usually the smaller twin) has been noted in 16 percent of cases [47]. (See "Fetal growth restriction: Evaluation", section on 'Perform a detailed fetal anatomic survey'.)
Infection can affect one or both fetuses and lead to growth restriction, so maternal history of signs/symptoms or ultrasound markers of infection other than isolated FGR require additional investigation, such as maternal serology for cytomegalovirus (CMV) and toxoplasmosis [52]. (See "Fetal growth restriction: Evaluation", section on 'Work-up for infection'.)
Rarely, monochorionic twins have discordant karyotypes. This is most often observed in the context of discordant fetal anomalies [53,54]. Screening (eg, cell-free DNA) or a procedure for diagnostic genetic testing for aneuploidy is recommended in anomalous fetuses when the findings will impact clinical decision-making [55,56]. (See "Prenatal genetic evaluation of the fetus with anomalies or soft markers".)
●Assess AFV to evaluate for TTTS – Amniotic fluid volume (AFV) is assessed to look for oligohydramnios/polyhydramnios sequence to diagnose or exclude coexistent TTTS [57]. Distinguishing TTTS complicated by growth restriction of the donor twin from sFGR can be difficult since in both situations the growth-restricted fetus may have oligohydramnios (table 3). However, once fluid criteria for TTTS are met (defined as maximum vertical pocket <2 cm for the donor and >8 cm for the recipient), appropriate stage-based management of TTTS is required, regardless of coexisting sFGR or twin weight discordance [58]. (See "Twin-twin transfusion syndrome: Screening, prevalence, pathophysiology, and diagnosis" and "Twin-twin transfusion syndrome: Management and outcome".)
●Determine MCA-PSV to evaluate for TAPS – The middle cerebral artery-peak systolic velocity (MCA-PSV) is determined in both twins to diagnose or exclude coexistent TAPS. Diagnostic criteria for TAPS has shifted from absolute cutoffs for the MCA-PSV >1.5 multiples of median (MoM) in one twin (suggestive of anemia) and <0.8 MoM in the other twin (suggestive of polycythemia) to intertwin discordance (delta MCA-PSV) of >0.5 MoM since this has an improved correlation with postnatal TAPS [59,60]. (See "Twin anemia-polycythemia sequence (TAPS)", section on 'Diagnostic criteria'.)
Fetal anemia in monochorionic twins can also be caused by parvovirus infection, alloimmunization, hemoglobinopathy (eg, alpha-thalassemia major), or massive fetomaternal transfusion. In contrast to TAPS, these other diagnoses have maternal laboratory findings suggestive of the diagnosis (eg, red blood cell antibodies, abnormal hemoglobin, Kleihauer-Betke showing a large fetomaternal bleed, parvovirus serology consistent with recent infection). TAPS and these other disorders can be associated with hydrops. (See "Parvovirus B19 infection during pregnancy" and "RhD alloimmunization in pregnancy: Overview" and "Hemoglobinopathy: Screening and counseling in the reproductive setting and fetal diagnosis" and "Spontaneous massive fetomaternal hemorrhage".)
PREGNANCY MANAGEMENT
Goal — The initial goal when managing pregnancies with sFGR is to identify those that can be safely managed expectantly versus those that might benefit from fetal intervention. The strategy for risk assessment and pregnancy management that has evolved considers sFGR classification, gestational age, prognosis, technical considerations, and patient values and preferences [46]. It is based on personal experience, expert opinion, and data from observational studies; randomized trials have not been performed.
Type 1 sFGR
●Monitoring – Expectant management with weekly ultrasound surveillance (umbilical artery [UA], middle cerebral artery [MCA]) beginning at diagnosis is the preferred approach for these mild cases identified in the second trimester, as described in the algorithm (algorithm 1). Weekly biophysical profile scoring (BPP) is added at 28 to 32 weeks.
If the UA pulsatility index (PI) increases to >95th percentile or the MCA PI falls below the 5th percentile, we would increase surveillance to twice weekly and also monitor for abnormalities in the ductus venosus (DV) waveform.
Worsening of the UA Doppler pattern in the growth-restricted fetus is observed in up to 26 percent of these cases [22]; however, this typically takes months to develop, and the likelihood for serious fetal deterioration (venous Doppler abnormalities, a low BPP score, or oligohydramnios [36,61,62]) or demise is low when the UA flow is consistently forward.
●Timing of delivery – If fetal status remains reassuring, as it usually does, we suggest delivery at 34+0 to 35+6 weeks as pregnancies with sFGR are not "uncomplicated" twins and have a higher rate for unanticipated demise than uncomplicated monochorionic twins in which delivery may be delayed until 37+6 weeks [63]. Earlier delivery is indicated if standard maternal or fetal indications for delivery develop. (See "Twin pregnancy: Labor and delivery", section on 'Monochorionic/diamniotic twins'.)
The International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) suggests timing delivery in sFGR based on assessment of fetal well-being, interval growth, BPP, DV waveform, and/or nonstress testing and points out that the risk of fetal demise in these pregnancies is increased, so delivery might be indicated even before fetal deterioration becomes evident [29].
Type 2 and 3 sFGR
●Monitoring – The approach to moderate and severe sFGR is more complicated due to the higher rates of adverse outcomes, particularly fetal demise [2,48] (see 'Prognosis by severity of sFGR' below). Ultrasound surveillance (UA, MCA, and DV) is performed at least weekly beginning at diagnosis, with a weekly NST added at 28 weeks, as described in the algorithm (algorithm 2).
●Approach to fetal deterioration – Intervention for fetal deterioration is preferable to expectant management because demise of one twin can result in acute fetal transfusion and volume shifts, which lead to double fetal demise or neurologic damage in the surviving co-twin in up to 30 percent of cases [17,61,64-67]. Depending on the gestational age, the intervention may be an ultrasound-guided fetoscopic procedure or delivery. Although demise of the sFGR twin is certain after cord occlusion and likely after laser ablation, both procedures reduce the risk of neurologic damage in the surviving twin.
•Selective cord occlusion or laser ablation of anastomoses – Prior to the lower limit of neonatal viability, intervention with either selective fetal reduction using a cord occlusion technique [68-70] or fetoscopic laser ablation of intertwin placental vascular anastomoses can be considered in cases with fetal deterioration (deterioration of type 2 or type 3 sFGR, venous Doppler abnormalities [36,61,62], or oligohydramnios associated with the growth-restricted fetus [64,71]), as described in the algorithm (algorithm 2). Coexisting oligohydramnios has been associated with impending fetal death.
-Cord occlusion – If cord occlusion is performed, we perform ultrasound surveillance of the remaining twin weekly for the first two weeks after the procedure and then serial ultrasounds every four weeks to monitor fetal growth. Delivery is at term or earlier if standard obstetric indications arise. Selective cord occlusion is described separately. (See "Multifetal pregnancy reduction and selective termination", section on 'Monochorionic fetuses'.)
-Laser ablation – Fetoscopic laser ablation of anastomoses is associated with high mortality for the sFGR twin and does not guarantee survival for the normally grown (AGA) twin, but may protect this twin from the neurological consequences of co-twin demise. In a meta-analysis of mostly retrospective studies, postablation fetal death rates for the sFGR and AGA twins were 58 and 12 percent, respectively, compared with fetal death rates of 22 and 9 percent, respectively, with expectant management [72]. After laser ablation, 1 percent of AGA twins had abnormal neonatal neuroimaging compared with 9 percent in expectantly managed pregnancies. Although more normal neuroimaging in the laser group was not associated with a reduction in long-term neurodevelopmental impairment compared with expectantly managed group, the small number of cases and events and the observational design of the studies precludes a clear conclusion about the neurodevelopmental benefits of laser therapy for the AGA twin.
After the ablation, elevated MCA-peak systolic velocity (MCA-PSV) and abnormal DV appear to be strong predictors of demise of the sFGR twin [73]. Ablation may be more technically challenging compared with that for treatment of twin-twin transfusion syndrome (TTTS) and may not always be possible in cases of pure sFGR due to lack of polyhydramnios in the sac of the normally grown twin, the vascular equator may be obscured by the overlying donor requiring consideration for septostomy, and presence of large artery to artery anastomoses [64,74,75]. If performed and the procedure is considered complete, then these pregnancies are managed similarly to dichorionic twins. (See "Multifetal pregnancy reduction and selective termination", section on 'Monochorionic fetuses' and "Twin-twin transfusion syndrome: Management and outcome", section on 'Fetoscopic laser ablation of anastomotic vessels'.)
•Expectant management – For patients who decline these interventions, we continue to perform twice weekly Doppler surveillance of UA, MCA, and DV and begin BPPs with each ultrasound assessment performed at ≥28 weeks. If Doppler findings remain stable and BPPs are reassuring, outpatient monitoring until delivery is reasonable.
After the gestational age of neonatal viability, if Doppler findings worsen (eg, deterioration of the UA Doppler pattern, such as progressive reversed end-diastolic velocity or DV pulsatility index [DV PI] >95th percentile) or if oligohydramnios develops, we would increase surveillance with Dopplers/BPP to two to three times weekly and consider hospital admission for daily fetal monitoring with nonstress tests if delivery for fetal indications would be considered [76].
●Timing of delivery – We deliver type 2 and 3 sFGR pregnancies with stable Dopplers and reassuring fetal surveillance results at 30+0 to 34+0 weeks. We deliver those with UA reversed end-diastolic flow at 30+0 to 32+0 weeks and those with UA absent end-diastolic flow at 32+0 to 34+0 weeks, with earlier delivery for standard obstetric indications.
For type 2 and 3 sFGR pregnancies that are worsening after neonatal viability but have reassuring fetal surveillance results, we deliver at 32+0 weeks in the absence of complications necessitating earlier delivery [31,44,77]. Signs of worsening disease include progression from type 2 to type 3 sFGR, DV PI >95th percentile, development of absent or reversed a-wave in the DV waveform, and oligohydramnios in the sac of the smaller twin.
Antenatal corticosteroids — Betamethasone or dexamethasone is administered if fetal status deteriorates or prior to planned preterm delivery (see "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery"). UA Doppler flow may improve transiently after administration of betamethasone [78].
PROGNOSIS
Prognosis by severity of sFGR — Type 1 sFGR twin pregnancies have the most favorable prognosis. They have lower overall rates of fetal demise, neonatal mortality, and cerebral injury than type 2 and type 3 sFGR twin pregnancies and generally can be safely delivered at a more advanced gestational age than the other types [79]. Nearly all studies report that the smaller twin is at higher risk for adverse perinatal outcome.
The risks of fetal demise of one or both fetuses, progression, preterm birth, and survival derived from a systematic review are shown in the table (table 4) [2]. Heterogeneity among retrospective studies for the definition of sFGR, antenatal management protocols, and lack of standardized perinatal and long-term outcomes limit direct comparisons of outcomes by sFGR type [79].
Another review in which the majority of cases were type 1 sFGR (110 in 177) compared survival for isolated sFGR, sFGR associated with a major anomaly, and sFGR with subsequent development of twin-twin transfusion syndrome (TTTS) and reported survival rates of 91 percent for isolated, 70 percent with an anomaly, and 65 percent with TTTS [47]. Overall survival for type 1, 2, and 3 sFGR were 96, 55, and 83 percent, respectively.
Overall neurodevelopmental prognosis — A systematic review of the impact of sFGR and/or birth weight discordance (BWD) on long-term neurodevelopment in monochorionic twins found that sFGR/BWD twins appeared to be higher risk compared with uncomplicated monochorionic or dichorionic twins, with the smaller twin at highest risk [80]. For the smaller twin, long-term rates of moderate and severe neurologic morbidity were 3 and 6 percent, respectively; for the larger twin, these rates were 1 and 5 percent, respectively, but available data were limited. Another systematic review noted that severe cerebral injury (intraventricular hemorrhage ≥grade II, periventricular leukomalacia ≥grade II, porencephalic cysts and/or intraparenchymal bleeding) was present in 0 to 2 percent of type 1, 2 to 30 percent of type 2, and 0 to 33 percent of type 3 sFGR, and the smaller twin was at higher risk than the larger twin [79]. A subsequent cohort study of 47 monochorionic diamniotic twin pairs with sFGR born in the Netherlands evaluated neurodevelopment at a median age of 11 years (range 8 to 13) [81]. Smaller twins had a higher rate of mild neurodevelopmental impairment than larger twins (17 in 47 [36 percent] versus 5 in 47 [11 percent]; odds ratio [OR] 4.8, 95% CI 1.6-14.1). There was no difference in severe neurodevelopmental impairment between groups (2 in 47 [4 percent] in both groups)
Prognosis for sFGR with coexistent TTTS — In a meta-analysis of retrospective studies comparing pregnancies complicated by both sFGR and laser therapy for TTTS with those with TTTS alone [82]:
●The combination of sFGR and TTTS increased the overall risk of fetal loss (20.9 versus 14.4 percent with TTTS, OR 1.57, 95% CI 1.30-1.91; four studies, >3000 twins).
The donor twin in pregnancies with sFGR and TTTS was at the highest risk of loss (donor twin sFGR-TTTS: 25.7 percent, recipient twin sFGR-TTTS or donor twin TTTS alone: approximately 16 percent).
●The combination of sFGR and TTTS increased the risk for neurological morbidity (intraventricular hemorrhage grade III and IV or periventricular leukomalacia grade II: 12.4 versus 6.3 percent with TTTS alone; pooled OR 1.8, 95% CI 1.1-2.9; two studies, nearly 1000 twins).
The donor twin of pregnancies with sFGR and TTTS was at highest risk (12.2 versus 5.6 percent with TTTS alone; pooled OR 2.4, 95% CI, 1.1-5.2); the difference was not statistically significant for the recipient twin (11.9 versus 6.6 percent with TTTS alone; OR 1.56, 95% CI 0.60-4.05).
In pregnancies with sFGR and TTTS treated with laser therapy, survival without neurological impairment at two years was approximately 70 percent for both donor and recipient twins.
Prognosis of sFGR without coexistent TTTS — In a meta-analysis of retrospective and prospective studies of pregnancies with sFGR without TTTS, the incidence of stable, deteriorating, or improving umbilical artery (UA) Dopplers was [83]:
●Type 1 sFGR: 68 percent (95% CI 26-89), 23 percent (95% CI 7-40), and 9 percent (95% CI 0.0-100), respectively
●Type 2 sFGR: 40 percent (95% CI 18-81), 50 percent (95% CI 23-82), and 10 percent (95% CI 4-37), respectively
●Type 3 sFGR, 55 percent (95% CI 2-99), 23 percent (95% CI 9-43), and 22 percent (95% CI 6-54), respectively
Risk factors for demise of the smaller twin were earlier gestational age at diagnosis, larger intertwin weight discordance, deterioration of UA Dopplers for type 2 and 3 cases, and absent or reversed a-wave for type 2 and 3 cases. Development of TTTS was not significantly associated with smaller twin demise for type 2 and 3 cases.
FUTURE STUDIES —
A core outcome set is available for investigators studying the management of sFGR [84]. The Delphi survey resulted in the following 11 outcomes: live birth, gestational age at birth, birth weight, intertwin birth weight discordance (BWD), death of surviving twin after death of cotwin, loss during pregnancy or before final hospital discharge, parental stress, procedure-related adverse maternal outcome, length of neonatal stay in hospital, neurological abnormality on postnatal imaging, and childhood disability.
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" and "Society guideline links: Multiple gestation".)
SUMMARY AND RECOMMENDATIONS
●Incidence – Selective fetal growth restriction (sFGR) refers to growth restriction of one fetus of a monochorionic twin pair due to discordant placental sharing (figure 1 and picture 1). It affects 10 to 15 percent of these pregnancies. (See 'Introduction' above and 'Pathophysiology' above and 'Incidence' above.)
●Identification of sFGR – Beginning in the early second trimester, monochorionic twin pregnancies should routinely undergo serial ultrasound examinations to monitor for development of sFGR as well as twin-twin transfusion syndrome (TTTS) and twin anemia-polycythemia sequence (TAPS). An example of a monitoring protocol is depicted in the table (table 1). (See 'Clinical presentation' above.)
●Diagnosis – The diagnosis of sFGR is based on any of the following (see 'Diagnosis' above):
•Estimated fetal weight (EFW) of one twin <10th percentile plus EFW discordance ≥25 percent, or
•EFW of one twin <3rd percentile, or
•At least two of the four following criteria:
-EFW <10th percentile for one twin
-Abdominal circumference <10th percentile for one twin
-Weight discordance ≥25 percent
-Umbilical artery (UA) pulsatility index (PI) >95th percentile for the smaller twin
●Classification – The pattern of the UA waveform and end-diastolic velocity of the smaller fetus are used to classify affected fetuses as type 1, type 2, or type 3 sFGR (waveform 2), which predicts the anticipated clinical course (table 4) and risk profile. Type 2 sFGR has the worst prognosis due to the high risk of single fetal demise and preterm birth. Proper Doppler technique is essential to accurately assess the UA Doppler pattern. (See 'Classification' above and 'Doppler technique' above.)
●Diagnostic evaluation – In a high proportion of cases, sFGR coexists with TTTS, TAPS, or discordant fetal anomalies. Due to substantial overlap between these disorders (table 3), a systematic approach to evaluation is required to arrive at the correct diagnosis and initiate management planning. (See 'Approach to differential diagnosis of sFGR, TTTS, and TAPS' above.)
●Antenatal corticosteroids – Antenatal corticosteroids are administered for standard indications (ie, if fetal status deteriorates preterm or prior to planned preterm delivery). The evidence for steroid administration prior to preterm birth is available separately. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery".).
●Monitoring and delivery – The goal in twin pregnancies complicated by sFGR is to identify those that can be safely managed expectantly versus those that might benefit from fetal intervention. Our approach to decision-making is based on sFGR type and described in the algorithms: type 1 sFGR (algorithm 1), type 2, and type 3 sFGR (algorithm 2). (See 'Pregnancy management' above.)
•Type 1 sFGR – For stable type 1 sFGR, we suggest expectant management rather intervention (Grade 2C). These pregnancies are monitored closely but are at low risk of progressing to type 2 or 3 sFGR. We suggest delivering these pregnancies at 34+0 to 35+6 weeks rather than later, in the absence of complications necessitating earlier delivery (Grade 2C). (See 'Type 1 sFGR' above.)
•Type 2 or 3 sFGR – For stable type 2 and 3 sFGR, we suggest expectant management rather intervention (Grade 2C). These pregnancies are monitored closely and are at increased risk for worsening disease. We suggest delivering type 2 and 3 sFGR pregnancies with stable Dopplers and reassuring fetal surveillance results at 30+0 to 34+0 weeks rather than later (Grade 2C). We deliver those with UA reversed end-diastolic flow at 30+0 to 32+0 weeks and those with UA absent end-diastolic flow at 32+0 to 34+0 weeks, with earlier delivery for standard obstetric indications. (See 'Type 2 and 3 sFGR' above.)
For type 2 and 3 sFGR with signs of worsening disease prior to neonatal viability, we suggest intervention rather than expectant management (Grade 2C). The choice of intervention (cord occlusion versus laser ablation of anastomoses) depends on technical feasibility and patient values and preferences. (See 'Type 2 and 3 sFGR' above.)
For type 2 and 3 sFGR pregnancies with signs of worsening disease after neonatal viability and reassuring fetal surveillance results, we suggest delivery at 32+0 weeks (Grade 2C). Signs of worsening disease include progression from type 2 to type 3 sFGR, ductus venosus pulsatility index (DV PI) >95th percentile, development of absent or reversed a-wave in the DV waveform, and oligohydramnios in the sac of the smaller twin. (See 'Type 2 and 3 sFGR' above.)