ABSTRACT
Objectives
Monochorionic diamniotic (MCDA) twins with selective fetal growth restriction (sFGR) and either continuous (cAREDF) or intermittent (iAREDF) absent or reversed end‐diastolic flow in the umbilical artery face significant fetal and neonatal risks. This study evaluated fetal and neonatal outcomes in these cases and compared outcomes for the larger twin following selective reduction (SR) vs expectant management (EM).
Methods
This was an international retrospective cohort study of MCDA twin pregnancies with sFGR and cAREDF or iAREDF from five fetal medicine centers over 7 years (2016–2022). Patients were included based on an estimated fetal weight (EFW) discordance of ≥ 20% combined with cAREDF or iAREDF. We collected demographic and antenatal characteristics, longitudinal ultrasound and management data, and key perinatal outcomes, including gestational age (GA) at birth, survival rate and severe neonatal morbidity. Outcomes for the larger twin following SR vs EM were compared using logistic regression with standardization, inverse probability weighting and augmented inverse probability weighting to adjust for confounders. Average treatment effects (risk differences) were calculated for three composite outcomes (live birth at ≥ 32 weeks; live birth at ≥ 32 weeks and absence of severe neonatal morbidity; and live birth without a GA limit and absence of severe neonatal morbidity) and overall live birth.
Results
Data were analyzed from 363 MCDA twin pregnancies (726 fetuses) with sFGR, which were diagnosed initially as having either cAREDF (n = 124) or iAREDF (n = 239). The umbilical artery flow pattern changed in 59% of pregnancies during gestation. Smaller twins with cAREDF at the final ultrasound scan before demise, delivery or intervention had a 70% survival rate, with 29% of survivors experiencing severe neonatal morbidity. In contrast, larger twins in this group had an 87% survival rate and a 26% risk of severe neonatal morbidity among survivors. For smaller twins with iAREDF at the final ultrasound scan, the survival rate was 83% and 22% of survivors were affected by severe neonatal morbidity, whereas larger twins had an 87% survival rate and a 13% risk of severe neonatal morbidity among survivors. The combined risk of adverse outcomes (fetal or neonatal demise or severe neonatal morbidity) was 52% for smaller twins with cAREDF and 37% for those with iAREDF. Among 37 cases of spontaneous fetal demise, 24 (65%) were double demise. Severe cerebral injury following single fetal demise occurred in approximately 30% of survivors. SR was associated with a 32–34% higher probability of the larger twin being liveborn ≥ 32 weeks compared with EM. This benefit seemed to align with later GA at birth and reduced rate of severe neonatal morbidity, despite similar rates of live birth.
Conclusions
MCDA twins with sFGR and cAREDF or iAREDF are at high risk for demise and severe morbidity, particularly the smaller twin with cAREDF. Compared with EM, SR significantly improves the chance of the larger twin being born at ≥ 32 weeks and surviving without severe morbidity, which may be influenced by increased GA at birth. These data should inform patient counseling and management decisions. © 2025 The Author(s). Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.
Keywords: AREDF, cAREDF, Gratacós classification, iAREDF, monochorionic twin, selective fetal growth restriction, sFGR
INTRODUCTION
Managing selective fetal growth restriction (sFGR) with continuous absent or reversed end‐diastolic flow (cAREDF) (Gratacós Type 2) or intermittent absent or reversed end‐diastolic flow (iAREDF) (Gratacós Type 3) in the umbilical artery (UA) 1 presents a complex decision‐making process, akin to navigating between Scylla and Charybdis, a metaphor for the challenge of choosing the lesser of two evils 2 . With expectant management (EM), the risk of unanticipated fetal demise must be balanced against the potential for severe neonatal morbidity due to very preterm birth. These challenges are intensified by the substantially higher risk of fetal and neonatal complications in cases with cAREDF and iAREDF compared to that in cases with positive end‐diastolic flow (pEDF) (Gratacós Type 1) 1 , 3 , 4 .
To address these risks, alternative management strategies have been explored. Laser coagulation of placental anastomoses aims to protect the larger twin while preserving both fetuses, although it may increase the risk of demise in the smaller twin compared to EM, without improving outcome for the larger twin 5 , 6 , 7 , 8 , 9 , 10 , 11 . Selective reduction (SR) of the smaller twin also aims to protect the larger twin from exsanguination if the smaller twin dies. However, there is a lack of direct comparison between SR and EM in terms of outcomes for the larger twin.
The rarity of cAREDF and iAREDF cases, relative to pEDF cases, has led to a shortage of large, observational cohort studies assessing their management and outcome. The variability in UA flow further complicates the development of consistent management protocols and accurate patient counseling 8 , 12 . This has resulted in significant differences in clinical practice, including variation in the timing of delivery and overall management strategy 3 , 12 , 13 , 14 , 15 , 16 , 17 .
The primary aim of this study was to describe the fetal and neonatal outcomes of monochorionic diamniotic (MCDA) twin pregnancies complicated by sFGR with cAREDF or iAREDF. Secondly, this study assessed and compared outcomes for the larger twin following SR vs EM. By providing robust data derived from an international multicenter collaboration, this study aimed to improve patient counseling and support clinical practice for sFGR cases with cAREDF or iAREDF.
METHODS
Study population
This retrospective, multicenter cohort study (the ALIGN Study), covering the period between January 2016 and December 2022, was conducted internationally across five referral centers (Leiden University Medical Center, Leiden, The Netherlands; University Hospitals Leuven, Leuven, Belgium; Karolinska University Hospital, Stockholm, Sweden; BCNatal, Barcelona, Spain and Mount Sinai Hospital, Toronto, Canada), all of which manage complex MCDA twin pregnancies. Included were consecutive MCDA twin pregnancies complicated by sFGR with cAREDF or iAREDF. We categorized UA end‐diastolic flow based on its observed pattern, corresponding to Gratacós Type 2 or Type 3 1 . sFGR was defined as an estimated fetal weight (EFW) discordance of ≥ 20% 18 , calculated as (EFW of larger fetus – EFW of smaller fetus) / EFW of larger fetus × 100. Pregnancies with a lethal fetal anomaly, higher‐order multiple pregnancies and pregnancies with concomitant twin–twin transfusion syndrome (TTTS) or twin anemia–polycythemia sequence (TAPS) at the time of sFGR diagnosis were excluded. Part of this cohort has been included in previous publications 4 , 15 , 19 , 20 . The study protocol was approved by research ethics boards in all participating countries.
EM generally involved weekly ultrasound surveillance up to 28 weeks' gestation, with some centers increasing to twice‐weekly monitoring afterward. Two centers routinely admitted patients for inpatient surveillance after 28 weeks, but three did not. Planned delivery was scheduled at 32 weeks for cases with cAREDF and at 34 weeks for those with iAREDF, although fetal deterioration could necessitate an earlier delivery. Fetal indications for earlier delivery were similar across centers and included non‐reassuring cardiotocography, concerning non‐stress test or new signs of fetal deterioration (Table S1). All centers opted for Cesarean section, although three centers allowed for a trial of labor under specific conditions. SR was discussed as an option when the smaller twin exhibited imminent signs of demise or a predicted poor outcome. These criteria were generally consistent across participating centers and are detailed in Table S1.
Data collection
Demographic information and gestational age (GA) at first presentation with sFGR were collected. We also recorded the following longitudinal ultrasound parameters at each visit: EFW; UA Doppler flow pattern, measured at the placental cord insertion of the smaller fetus, with recurrent long recordings at slow sweep speed to assess properly for the presence of intermittent flow patterns; ductus venosus (DV) Doppler flow pattern; presence of oligohydramnios in the smaller twin (defined as deepest vertical amniotic fluid pocket < 2 cm) or polyhydramnios in the larger twin (defined as deepest vertical amniotic fluid pocket > 8 cm at or before 20 weeks' gestation and > 10 cm after 20 weeks' gestation 21 ); progression to TTTS or TAPS 22 ; and diagnosis of single (sIUD) or double (dIUD) intrauterine demise. Management data included: details of any fetal intervention; complications during pregnancy; and delivery characteristics, including GA at birth, onset of labor, mode of delivery and survival.
For liveborn neonates, collected data included: sex; birth weight; birth‐weight discordance, calculated as (birth weight of larger fetus – birth weight of smaller fetus) / birth weight of larger fetus × 100; and occurrence of severe neonatal morbidity. Severe neonatal morbidity was defined as any of the following: bronchopulmonary dysplasia, defined according to Jobe and Bancalari 23 ; patent ductus arteriosus needing medical therapy or surgical closure; necrotizing enterocolitis ≥ Grade 2; retinopathy of prematurity ≥ Stage 3; ischemic limb injury; amniotic band syndrome; severe cerebral injury, defined as intraventricular hemorrhage of Grade 3 or greater, cystic periventricular leukomalacia of Grade 2 or greater, porencephalic or parenchymal cysts, or other severe cerebral lesions associated with adverse neurological outcome; and culture‐proven early‐onset sepsis. Neonatal death was assessed until 28 days after birth. Cerebral injury following sIUD was categorized as either focal or diffuse according to Spruijt et al. 24 , with an additional subdivision for severe cerebral injury 25 .
Statistical analysis
For the primary aim of describing fetal and neonatal outcomes, descriptive statistics are presented as mean ± SD for symmetrically distributed continuous variables and median (interquartile range) for non‐symmetrically distributed continuous variables. Categorical variables are expressed as n (%). We stratified data according to UA flow type at the time of diagnosis (baseline characteristics) and at the final ultrasound scan before delivery, demise or fetal intervention (all other outcomes). This approach accounts for the dynamic nature of UA flow type, which can influence survival outcome. A switch in UA flow type was defined as either a deterioration (from pEDF to cAREDF or iAREDF, or from iAREDF to cAREDF) or multiple changes in flow type during pregnancy. UA flow type was determined from the ultrasound report for each patient. We also assessed changes in UA flow type from diagnosis (cAREDF or iAREDF) to the scan, including normalization to pEDF or progression to TTTS or TAPS. To calculate the overall risk of adverse outcome, we assessed a composite variable encompassing fetal or neonatal demise and severe neonatal morbidity. If data on severe neonatal morbidity were missing, the composite outcome was coded as missing for GA at birth ≤ 34 weeks, and computed as negative for GA at birth > 34 weeks (given the low risk of severe morbidity with increasing GA at birth). Comparisons between the cAREDF and iAREDF groups were made using Pearson's chi‐square test for categorical variables and the Mann–Whitney U‐test or unpaired t‐test for continuous variables, as appropriate. For comparisons across multiple groups, the Kruskal–Wallis test for continuous variables, one‐way ANOVA or chi‐square test was used. Statistical significance was set at a two‐sided P‐value < 0.05.
For the secondary aim of comparing outcomes for the larger twin after SR vs EM, a subset of the cohort was selected, consisting of all cases with either cAREDF or iAREDF at ≤ 24 weeks' gestation, with known outcomes and ultrasound parameters. Cases with cAREDF and those with iAREDF were included in both management groups (SR and EM). The GA limit of ≤ 24 weeks was chosen because of legal restrictions on performing SR in some countries. Cases managed with laser coagulation for subsequent TTTS/TAPS and cases of termination of pregnancy (TOP), with the exception of those performed following sIUD, were excluded from the subset. To address causal inference in this subset, we employed logistic regression with standardization, inverse probability weighting and augmented inverse probability weighting. These exposure modeling methods estimated the average treatment effect (ATE) in terms of risk difference for three clinically relevant composite outcomes for the larger twin: live birth ≥ 32 weeks' gestation (composite outcome 1), live birth ≥ 32 weeks and absence of severe neonatal morbidity (composite outcome 2), and live birth without a GA limit and absence of severe neonatal morbidity (composite outcome 3). The analyses accounted for four confounders reflecting the severity of sFGR: UA Doppler flow type, EFW discordance, absent or reversed DV a‐wave and oligohydramnios in the smaller twin. Interaction terms between these confounders and management type (SR vs EM), along with flexible forms of continuous covariates, were explored using model fit measures for composite outcome 1. Three methods of SR were used; however, given the small number of cases for each, we chose to analyze SR collectively rather than according to individual SR method. The selected ultrasound parameters correspond to the hypothetical decision point for assessing management options, specifically the second ultrasound scan after confirming abnormal UA flow or the last ultrasound scan before demise if only one was available. The outcome models included a natural spline with two degrees of freedom for EFW discordance and an interaction term between management type (SR vs EM) and EFW discordance. Bootstrapping (1000 samples) was used to provide standard errors and 95% CI for the ATE estimates.
Statistical analysis was performed using IBM SPSS Statistics (version 25.0; IBM Corp, Armonk, NY, USA) and R (R Foundation for Statistical Computing, Vienna, Austria) 26 .
RESULTS
A total of 363 MCDA twin pregnancies (726 fetuses) qualified for inclusion, of which 124 were diagnosed initially with cAREDF and 239 with iAREDF (Table 1). Throughout pregnancy, UA Doppler flow pattern switched in 213 (59%) pregnancies, with most (105/213; 49%) switches occurring only after 22 weeks' gestation. In 79 (37%) cases, switches occurred only at or before 22 weeks, whereas in 29 (14%) cases, changes were observed both before and after 22 weeks' gestation. Switches in UA Doppler flow type were more common in pregnancies diagnosed initially with iAREDF (150/239; 63%) compared to those with cAREDF (63/124; 51%) (P < 0.001). Based on the comparison of initial vs final UA flow type, iAREDF cases were more likely to improve to pEDF (53/239; 22%) than were pregnancies with cAREDF at inclusion (6/124; 5%) (P < 0.001). Deterioration from iAREDF to cAREDF occurred in 28 (12%) of the initial iAREDF cases, and 22 (18%) of the initial cAREDF cases improved to iAREDF.
Table 1.
Maternal and baseline characteristics of 363 monochorionic diamniotic twin pregnancies, according to umbilical artery (UA) flow type at diagnosis of selective fetal growth restriction (sFGR)
| Characteristic | cAREDF (n = 124) | iAREDF (n = 239) | P |
|---|---|---|---|
| Maternal age (years)* | 32.3 ± 5.0 | 31.4 ± 4.8 | 0.045 |
| Maternal BMI (kg/m2)† | 25.0 ± 5.6 | 24.2 ± 4.1 | 0.29 |
| Nulliparous | 73 (58.9) | 139 (58.2) | 0.90 |
| Spontaneous conception | 60/77 (77.9) | 168/210 (80.0) | 0.70 |
| Maternal comorbidity | 37/122 (30.3) | 47/235 (20.0) | 0.03 |
| Non‐lethal fetal anomaly | |||
| At least one twin | 26/122 (21.3) | 40/238 (16.8) | 0.30 |
| Smaller twin only | 22/26 (84.6) | 26/40 (65.0) | 0.11 |
| Larger twin only | 3/26 (11.5) | 10/40 (25.0) | 0.28 |
| Both twins | 1/26 (3.8) | 4/40 (10.0) | 0.51 |
| GA at diagnosis (weeks) | 17 + 6 (16 + 2 to 20 + 4) | 19 + 5 (17 + 1 to 23 + 2) | < 0.001 |
| EFW discordance (%) | 35.4 (30.6–42.5) | 31.1 (26.2–37.2) | < 0.001 |
| TTTS | 13 (10.5) | 16 (6.7) | 0.89 |
| TAPS | 1 (0.8) | 1 (0.4) | 0.55 |
| Management‡ | |||
| Expectant management | 57 (46.0) | 193 (80.8) | < 0.001 |
| Ending in TOP in both twins | 2/57 (3.5) | 5/193 (2.6) | 0.72 |
| Selective reduction | 48 (38.7) | 28 (11.7) | 0.02 |
| Ending in TOP in larger twin | 1/48 (2.1) | 1/28 (3.6) | 0.70 |
| Other§ | 1 (0.8) | 5 (2.1) | 0.22 |
| Ending in TOP in both twins | 1/1 (100) | 2/5 (40.0) | 0.27 |
Data are given as mean ± SD, n (%), n/N (%) or median (interquartile range).
Data missing for two cases in group with continuous absent or reversed end‐diastolic flow (cAREDF) in the UA and 17 cases in the group with intermittent absent or reversed end‐diastolic flow (iAREDF) in the UA.
Data missing for 52 cases in cAREDF group and 87 cases in iAREDF group.
Excluding 31 cases that underwent laser coagulation (19 for subsequent twin–twin transfusion syndrome (TTTS)/twin anemia–polycythemia sequence (TAPS) and 12 solely for sFGR).
Four intrauterine transfusions, two amnioreductions. BMI, body mass index; EFW, estimated fetal weight; GA, gestational age; TOP, termination of pregnancy.
Baseline characteristics
Maternal and baseline characteristics are detailed in Table 1. Patients with cAREDF twins were slightly older and more likely to have comorbidity compared with those with iAREDF twins. As expected, cAREDF twins were diagnosed earlier and exhibited greater EFW discordance compared with iAREDF twins.
Most patients were managed expectantly, with 31% (n = 113) undergoing a fetal intervention (SR, n = 76; laser coagulation of placental anastomoses, n = 31; other, n = 6). Twelve pregnancies ended in TOP (Table S2). Among the SR cases, bipolar cord occlusion (BCO) was the most commonly used method (n = 46), followed by radiofrequency ablation (RFA) (n = 26) and then intrafetal laser ablation (n = 4). RFA was more frequently employed up to 18 weeks' gestation (50% of all RFA cases) compared with BCO (20% of all BCO cases). The median GA at which SR was performed was 19 + 1 (interquartile range, 17 + 6 to 21 + 0) weeks. Loss of the larger twin after SR occurred in only four (5%) cases, one following RFA and three after BCO.
Of the 31 laser coagulations performed, 19 were for subsequent TTTS/TAPS and 12 were solely for sFGR. Following laser coagulation for sFGR, all larger twins survived, but five (42%) of the smaller twins underwent sIUD (Table S3). Other interventions included four intrauterine transfusions (all for suspected fetal anemia of the larger twin after sIUD of the smaller twin) and two amnioreductions. The complete surveillance guidelines for each center are detailed in Table S1.
Outcome according to final umbilical artery flow type in expectantly managed cohort
Survival outcomes varied significantly according to UA flow type. Smaller twins with cAREDF had the lowest survival rate (70%) and those that survived were born at the earliest GA, corresponding to the highest risk of severe neonatal morbidity (29%) (Table 2). Overall, the smaller twin with cAREDF had a 52% combined risk of adverse outcome (including fetal or neonatal demise or severe neonatal morbidity). Larger twins in this group had an 87% survival rate and a 26% risk of severe neonatal morbidity among survivors. Smaller twins with iAREDF had a survival rate of nearly 83%, a 22% risk of severe neonatal morbidity and a 37% combined risk of adverse outcome, whereas larger twins had an 87% survival rate and a 13% risk of severe neonatal morbidity. Accordingly, the proportion of smaller twins surviving to 28 days without severe neonatal morbidity was 48% for cAREDF and 63% for iAREDF cases. When categorizing UA flow type into normalized (pEDF) or persistently abnormal (cAREDF or iAREDF), the survival rate for smaller twins with persistently abnormal flow was 79%, compared with 98% in those whose UA flow normalized (P < 0.001) (Table S4).
Table 2.
Outcomes of 243 monochorionic diamniotic twin pregnancies that were managed expectantly, according to final umbilical artery (UA) flow type* or development of twin–twin transfusion syndrome (TTTS)/twin anemia–polycythemia sequence (TAPS)
| Outcome | pEDF (n = 57) | cAREDF (n = 57) | iAREDF (n = 124) | Subsequent TTTS/TAPS (n = 5) | P |
|---|---|---|---|---|---|
| GA at diagnosis of sFGR (weeks)† | 19 + 1 (16 + 4 to 24 + 0) | 20 + 3 (18 + 0 to 24 + 3) | 22 + 0 (18 + 3 to 25 + 3) | 21 + 3 (19 + 0 to 28 + 6) | 0.03 |
| GA at birth (weeks)† | 33 + 6 (32 + 1 to 34 + 5) | 30 + 5 (29 + 2 to 32 + 0) | 31 + 6 (30 + 1 to 33 + 0) | 30 + 5 (29 + 4 to 33 + 1) | < 0.001 |
| EFW discordance (%)†, ‡ | 29.3 (25.0–36.0) | 37.8 (29.8–42.7) | 31.1 (24.8–35.6) | 36.8 (25.4–48.8) | < 0.001 |
| BW discordance (%)† | 29.7 ± 9.2 | 39.8 ± 12.6 | 30.8 ± 9.6 | 32.7 ± 19.7 | < 0.001 |
| Live birth | |||||
| Smaller twin | 55/56 (98.2) | 37/53 (69.8) | 99/120 (82.5) | 5 (100) | < 0.001‡‡ |
| Larger twin | 55/56 (98.2) | 46/53 (86.8) | 104/120 (86.7) | 5 (100) | 0.08‡‡ |
| Severe neonatal morbidity§ | |||||
| Smaller twin | 7/53 (13.2) | 10/34 (29.4) | 19/88 (21.6) | 1 (20.0) | 0.33 |
| Larger twin | 1/53 (1.9) | 11/43 (25.6)** | 12/96 (12.5)†† | 0 (0) | 0.004 |
| Neonatal death | |||||
| Smaller twin | 1/55 (1.8) | 2/37 (5.4) | 6/95 (6.3) | 0 (0) | 0.60 |
| Larger twin | 0/55 (0) | 0/46 (0) | 1/100 (1.0) | 0 (0) | 0.79 |
| Combined adverse outcome¶ | |||||
| Smaller twin | 8/55 (14.5) | 26/50 (52.0) | 41/111 (36.9) | 1 (20.0) | < 0.001 |
| Larger twin | 2/55 (3.6) | 18/50 (36.0) | 28/113 (24.8) | 0 (0) | < 0.001 |
| Mode of delivery† | |||||
| Vaginal birth | 13/55 (23.6) | 1/37 (2.7) | 4/99 (4.0) | 0 (0) | < 0.001 |
| Cesarean section | 42/55 (76.4) | 36/37 (97.3) | 95/99 (96.0) | 5 (100) | < 0.001 |
Data are given as median (interquartile range), mean ± SD, n/N (%) or n (%). Expectant management of subsequent TTTS/TAPS was rare and led to induction of labor in all cases due to gestational age (GA) at diagnosis; given the small cohort, outcomes in this group should be interpreted with caution. Terminations of pregnancy were not included in this table.
At last ultrasound scan before delivery, demise or intervention.
Pregnancies delivering two liveborn twins.
Based on average estimated fetal weight (EFW) discordance across all individual ultrasound reports.
Within each UA flow type, P < 0.001 for smaller vs larger twin, except for TTTS/TAPS group, for which P = 0.18.
Fetal or neonatal demise or severe neonatal morbidity.
Risk of severe neonatal morbidity for larger twin among pregnancies with continuous absent or reversed end‐diastolic flow (cAREDF) in the UA and two liveborn twins was 7/35 (20.0%).
Risk of severe neonatal morbidity for larger twin among pregnancies with intermittent absent or reversed end‐diastolic flow (iAREDF) in the UA and two liveborn twins was 11/92 (12.0%).
Comparisons between four flow types and between small and large twins were statistically significant, but not between cAREDF and iAREDF alone. BW, birth weight; pEDF, positive end‐diastolic flow in the umbilical artery; sFGR, selective fetal growth restriction.
Characteristics and consequences for surviving twin after spontaneous demise
Among the cases of spontaneous fetal demise (n = 37), 65% were dIUDs, most of which (67%) presented with iAREDF before demise (Tables 3 and 4). EFW discordance was significantly greater in pregnancies that underwent sIUD or dIUD compared with those that delivered two liveborn twins. Frequently observed ultrasound parameters among the cases of spontaneous demise included absent or reversed DV a‐wave flow and oligohydramnios in the smaller twin, evident in both cAREDF and iAREDF pregnancies.
Table 3.
Ultrasound characteristics and outcomes of monochorionic diamniotic twin pregnancies with single or double spontaneous intrauterine demise or two live births, according to final umbilical artery (UA) flow type* or development of twin–twin transfusion syndrome (TTTS)/twin anemia–polycythemia sequence (TAPS)
| Spontaneous demise | Two liveborn twins | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Characteristic | All (n = 37) | pEDF (n = 1) | cAREDF (n = 15) | iAREDF (n = 21) | All (n = 214) | pEDF (n = 56) | cAREDF (n = 42) | iAREDF (n = 102) | TTTS/TAPS (n = 14) | P † | P ‡ | P § |
| GA at diagnosis of sFGR (weeks) | 16 + 6 (15 + 5 to 18 + 5) | 18 + 6 (NA) | 16 + 6 (15 + 0 to 17 + 6) | 16 + 5 (15 + 5 to 18 + 5) | 20 + 3 (17 + 5 to 24 + 4) | 19 + 2 (16 + 4 to 24 + 2) | 20 + 1 (17 + 5 to 24 + 2) | 22 + 0 (18 + 3 to 25 + 2) | 17 + 3 (16 + 3 to 20 + 5) | < 0.001 | < 0.001 | < 0.001 |
| EFW discordance (%) | 33.3 (29.0–44.0) | 35.9 (NA) | 32.8 (29.3–51.2) | 33.3 (28.9–42.4) | 31.4 (25.7–37.2) | 27.2 (24.3–33.3) | 35.6 (29.7–43.2) | 31.0 (25.3–35.9) | 46.0 (29.8–55.2) | 0.02 | 0.86 | 0.04 |
| Absent or reversed DV flow | 15/36 (41.7) | 0 (0) | 8 (53.3) | 7/20 (35.0) | 21/206 (10.2) | 3/55 (5.5) | 8/40 (20.0) | 8/98 (8.2) | 2/13 (15.4) | < 0.001 | 0.03 | 0.001 |
| Oligohydramnios in smaller twin | 14 (37.8) | 0 (0) | 9 (60.0) | 5 (23.8) | 22/200 (11.0) | 3 (5.4) | 9 (21.4) | 10 (9.8) | NA¶ | < 0.001 | 0.006 | 0.07 |
| Polyhydramnios in larger twin | 2 (5.4) | 0 (0) | 2 (13.3) | 0 (0) | 21/200 (10.5) | 9 (16.1) | 3 (7.1) | 9 (8.8) | NA¶ | 0.23 | 0.47 | — |
Data are given as median (interquartile range), n/N (%) or n (%). Nine terminations of pregnancy following single intrauterine demise were not included in this table.
At last ultrasound scan before delivery, demise or intervention.
Spontaneous demise (all pregnancies) vs two liveborn twins (all pregnancies).
Spontaneous demise (continuous absent or reversed end‐diastolic flow (cAREDF) in the UA) vs two liveborn twins (cAREDF).
Spontaneous demise (intermittent absent or reversed end‐diastolic flow (iAREDF) in the UA) vs two liveborn twins (iAREDF).
Not applicable (NA) as oligo‐ and polyhydramnios are inherent to TTTS/TAPS. DV, ductus venosus; EFW, estimated fetal weight; GA, gestational age; pEDF, positive end‐diastolic flow in the UA; sFGR, selective fetal growth restriction.
Table 4.
Characteristics and outcomes of 37 monochorionic diamniotic twin pregnancies with single (sIUD) or double (dIUD) spontaneous intrauterine demise
| Characteristic | sIUD (n = 13) | dIUD (n = 24) |
|---|---|---|
| UA flow type* | ||
| pEDF | 0 (0) | 1 (4.2) |
| cAREDF | 8 (61.5) | 7 (29.2) |
| iAREDF | 5 (38.5) | 16 (66.7) |
| Severe neonatal morbidity (larger twin) | 3/11 (27.3) | NA |
| Neonatal death (larger twin) | 0 (0) | NA |
Data are given as n (%) or n/N (%). Nine terminations of pregnancy following sIUD were not included in this table.
P = 0.41 for umbilical artery (UA) flow type in sIUD group and P < 0.001 for UA flow type in dIUD group. cAREDF, continuous absent or reversed end‐diastolic flow; iAREDF, intermittent absent or reversed end‐diastolic flow; NA, not applicable; pEDF, positive end‐diastolic flow.
Assessment of cerebral injury in the surviving twin after sIUD (including the cases that ended in TOP of the surviving twin following sIUD) revealed that 6/22 (27%) surviving larger twins developed severe cerebral injury, all of which resulted in TOP (Table S5). Of these cases, three had iAREDF as the final UA flow type, two had cAREDF and one progressed to TTTS. Focal brain injury occurred in three additional cases, leading to a 41% overall risk of cerebral injury following sIUD. Of the three cases with focal brain injury, no neurodevelopmental impairment was observed at follow‐up at 36 months after birth.
Outcome for larger twin after selective reduction vs expectant management
The subcohort selected for the comparison between SR and EM comprised 246 cases (Figure S1). Baseline characteristics were similar regarding maternal factors, with the exception of mode of conception (Table 5). There was a higher rate of non‐lethal fetal anomalies in at least one twin in the SR group, and, as anticipated, ultrasound parameters at baseline were more frequently abnormal in these pregnancies. For the larger twin, composite outcome 1 (live birth at ≥ 32 weeks' gestation) occurred in 49/65 (75%) of the SR group compared with 84/181 (46%) in the EM group (Table 6). The ATE, after adjusting for confounders with the prespecified methods, estimated a 32–34% higher probability of the larger twin being liveborn ≥ 32 weeks' gestation following SR compared with EM.
Table 5.
Baseline and ultrasound characteristics in subcohort of 246 monochorionic diamniotic twin pregnancies, according to whether they underwent expectant management or selective reduction
| Characteristic | Expectant management (n = 181) | Selective reduction (n = 65) | P |
|---|---|---|---|
| Baseline characteristics | |||
| Maternal age (years) | 31.2 ± 4.6 | 32.1 ± 5.1 | 0.20 |
| Maternal comorbidity | 36/178 (20.2) | 14/54 (25.9) | 0.82 |
| Nulliparous | 97 (53.6) | 39 (60.0) | 0.37 |
| Spontaneous conception | 131/153 (85.6) | 29/46 (63.0) | < 0.001 |
| Prenatal genetic testing | |||
| Normal NIPT* | 59/60 (98.3) | 18/20 (90.0) | 0.67 |
| Normal PND† | 5/5 (100) | 23/23 (100) | NA |
| Non‐lethal fetal anomaly | |||
| At least one twin | 27/179 (15.1) | 18/64 (28.1) | 0.02 |
| Smaller twin only | 18/27 (66.7) | 13/18 (72.2) | 0.35 |
| Larger twin only | 8/27 (29.6) | 3/18 (16.7) | 0.94 |
| Both twins | 1/27 (3.7) | 2/18 (11.1) | 0.11 |
| GA at diagnosis of sFGR (weeks) | 18 + 5 (16 + 4 to 21 + 1) | 18 + 0 (16 + 1 to 19 + 5) | 0.09 |
| Ultrasound parameters ‡ | |||
| GA at ultrasound (weeks) | 20 + 4 (18 + 2 to 22 + 3) | 18 + 2 (16 + 6 to 19 + 6) | < 0.001 |
| EFW discordance (%) | 31.0 (25.0–36.1) | 37.1 (31.5–42.4) | < 0.001 |
| UA flow type | |||
| cAREDF | 39 (21.5) | 37 (56.9) | 0.05 |
| iAREDF | 142 (78.5) | 28 (43.1) | < 0.001 |
| Oligohydramnios in smaller twin | 23 (12.7) | 24 (36.9) | < 0.001 |
| Absent or reversed DV flow | 24 (13.3) | 16 (24.6) | 0.06 |
Data are given as mean ± SD, n/N (%), n (%) or median (interquartile range). Subcohort includes all cases with continuous absent or reversed end‐diastolic flow (cAREDF) or intermittent absent or reversed end‐diastolic flow (iAREDF) diagnosed at ≤ 24 weeks' gestation, excluding those managed with laser coagulation for subsequent twin–twin transfusion syndrome or twin anemia–polycythemia sequence, or termination of pregnancy, except when performed following single intrauterine demise.
One result in each group was inconclusive, and one result in the selective reduction group was not specified in the patient chart.
Quantitative fluorescence polymerase chain reaction, array or whole‐exome sequencing.
Parameters were taken from ultrasound scan at hypothetical decision point, at which management options may have been evaluated. DV, ductus venosus; EFW, estimated fetal weight; GA, gestational age; NA, not applicable; NIPT, non‐invasive prenatal testing (based on cell‐free DNA analysis in maternal blood); PND, prenatal diagnosis; sFGR, selective fetal growth restriction; UA, umbilical artery.
Table 6.
Comparison of outcomes for larger twin and average treatment effects (ATE) after selective reduction vs expectant management in subcohort of 246 monochorionic diamniotic twin pregnancies
| Expectant management | Selective reduction | ATE (95% CI) (%) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Outcome | All (n = 181) | cAREDF (n = 39) | iAREDF (n = 142) | All (n = 65) | cAREDF (n = 37) | iAREDF (n = 28) | P * | P † | P ‡ | LR + standardization | IPW | Augmented IPW |
| Severe neonatal morbidity | 24/151 (15.9) | 10/32 (31.3) | 14/119 (11.8) | 3/59 (5.1) | 3/35 (8.6) | 0/28 (0) | 0.04 | 0.02 | 0.13 | — | — | — |
| GA at birth (weeks) | 31 + 2 (28 + 6 to 33 + 3) | 29 + 5 (28 + 1 to 32 + 2) | 32 + 0 (29 + 2 to 33 + 4) | 36 + 5 (31 + 5 to 38 + 3) | 36 + 4 (30 + 4 to 38 + 0) | 36 + 5 (33 + 1 to 38 + 3) | < 0.001 | < 0.001 | < 0.001 | — | — | — |
| Composite outcome 1§ | 84/181 (46.4) | 10/39 (25.6) | 74/142 (52.1) | 49/65 (75.4) | 27/37 (73.0) | 22/28 (78.6) | < 0.001 | < 0.001 | 0.01 | –32 (–47 to –16) | –34 (–52 to –16) | –34 (–49 to –16) |
| Composite outcome 2¶ | 75/175 (42.9) | 8/38 (21.1) | 67/137 (48.9) | 48/64 (75.0) | 26/36 (72.2) | 22/28 (78.6) | < 0.001 | < 0.001 | 0.004 | –36 (–51 to –19) | –37 (–53 to –21) | –38 (–53 to –21) |
| Composite outcome 3** | 123/175 (70.3) | 21/38 (55.3) | 102/137 (74.5) | 55/64 (85.9) | 31/36 (86.1) | 24/28 (85.7) | 0.01 | 0.004 | 0.2 | –17 (–30 to –0.2) | –17 (–30 to –0.4) | –16 (–30 to –0.2) |
| Live birth | 153/181 (84.5) | 32/39 (82.1) | 121/142 (85.2) | 59/65 (90.8) | 35/37 (94.6) | 24/28 (85.7) | 0.21 | 0.09 | 0.9 | –3 (–15 to 13) | –1 (–13 to 11) | 1 (–13 to 18) |
Data are given as n/N (%) or median (interquartile range).
Expectant management (all pregnancies) vs selective reduction (all pregnancies), unadjusted.
Expectant management (continuous absent or reversed end‐diastolic flow in the umbilical artery (cAREDF)) vs selective reduction (cAREDF), unadjusted.
Expectant management (intermittent absent or reversed end‐diastolic flow in the umbilical artery (iAREDF)) vs selective reduction (iAREDF), unadjusted.
Live birth + born at ≥ 32 weeks' gestation.
Live birth + born at ≥ 32 weeks' gestation + absence of severe neonatal morbidity.
Live birth + absence of severe neonatal morbidity. GA, gestational age; IPW, inverse probability weighting; LR, logistic regression.
The GA at birth for the surviving twin was higher after SR, with only 18/65 (28%) born before 32 weeks in the SR cohort compared with 102/181 (56%) in the EM cohort. The risk of severe neonatal morbidity was strongly related to GA at birth: 27% for liveborn twins delivered ≤ 32 weeks' gestation, and just 4% for those born > 32 weeks' gestation. Comparing the groups on composite outcome 2, which included both live birth at ≥ 32 weeks' gestation and the absence of severe neonatal morbidity, confirmed the advantage of the SR group (Table 6). When GA was excluded from the composite while retaining the absence of severe neonatal morbidity (composite outcome 3), SR continued to show a benefit, albeit with a lower ATE. The ATE for live birth alone showed no significant difference between the groups, suggesting that the main advantage of SR lies in prolonging the pregnancy and reducing severe neonatal morbidity.
DISCUSSION
Our international multicenter study of 363 MCDA twin pregnancies with sFGR and cAREDF or iAREDF found that over half (52%) of smaller twins with cAREDF experienced an adverse outcome and only 70% were liveborn, whereas smaller twins with iAREDF had a combined risk of adverse outcome of 37% and a live‐birth rate of 83%. This highlights the high risk of adverse outcome in sFGR pregnancies with abnormal UA flow, especially for smaller twins with cAREDF.
Most cases of spontaneous fetal demise presented as dIUD. Compared with pregnancies delivering two liveborn neonates, those that underwent spontaneous demise exhibited greater EFW discordance, more frequent abnormal DV a‐wave flow and a higher rate of oligohydramnios in the smaller twin, with 27% of surviving larger twins after sIUD developing severe cerebral damage.
Evaluation of the impact of SR on the outcome of the larger twin demonstrated that SR improved the chance of the larger twin being liveborn ≥ 32 weeks by 32–34% compared with EM. Although SR may not increase the survival rate of the larger twin significantly, it may improve long‐term outcome, as increased GA at birth reduces the associated risk of severe morbidity. However, SR is a sensitive and ethically complex decision that may not be acceptable to some families or in certain regions.
Survival rates in our cohort align with previous findings, which also report the lowest survival for sFGR twins with cAREDF. However, reported survival rates vary widely, from 45–100% 4 , 16 . The latest systematic review 3 reported a lower rate of spontaneous sIUD among cAREDF twins (8%) compared with our study (15%), potentially due to selection biases, variation in sFGR definitions and differences in the timing of UA flow classification. The 17.5% risk of combined sIUD and dIUD among iAREDF twins in our study is higher than the 11.3% reported by Shinar et al. 15 , who classified iAREDF at presentation, possibly including milder cases that later normalized to pEDF. Another possible explanation is the higher rate of dIUD in our cohort, which may reflect selection differences between the studies 15 .
Our findings for dIUD risk are consistent with the most comprehensive review of sIUD prognosis, which estimates a cotwin IUD risk of 41%, increasing to 61% if sIUD occurs before 28 weeks' gestation 27 . Although the review suggests that this risk may be overstated, we believe it might be understated, as using sIUD as an inclusion criterion may overlook cases that originally presented as dIUD. This does not imply that the dIUD risk always exceeds 60%; rather, it suggests that when the risk of imminent demise is considered to be high, the dIUD risk may indeed surpass 60%. Conversely, if sIUD has occurred without subsequent dIUD, the remaining risk of dIUD is likely to be lower. Counseling should include the risk of severe cerebral injury for the surviving twin after sIUD. Our 27% risk of severe cerebral injury for the surviving twin aligns with literature estimates (around 30%), although injury definitions vary 25 , 27 , 28 , 29 , 30 and long‐term outcome remains unknown. The potential benefit of intrauterine transfusion in mitigating this risk remains controversial, with limited evidence to suggest no major improvement in survival or reduction in risk of brain injury for the anemic twin after sIUD 31 , 32 .
Few studies have compared directly management strategies for sFGR with cAREDF or iAREDF. The largest recent study of Colmant et al. on cAREDF twins reported similar survival rates for the larger twin after EM (n = 45) and BCO (n = 50) (80% and 82%, respectively), with comparable GA at birth (32 and 35 weeks' gestation, respectively) and no differences in neurodevelopmental impairment, special care needs or sensory deficits between the two strategies 8 . Their surveillance frequency and management protocols generally aligned with ours. However, our study encompasses both cAREDF and iAREDF cases and adjusts for severity‐related confounders, offering a broader perspective.
Strengths and limitations
Our study provides the largest dataset to date for MCDA twins with sFGR and cAREDF or iAREDF, offering comprehensive data for counseling, and presents the first comparison of outcomes for the larger twin between SR and EM. Extensive ultrasound data were analyzed, focusing on UA flow type at initial and final diagnosis and its association with outcomes. The assessment of UA flow switch frequency concentrated solely on clinically relevant changes. Additionally, the use of various statistical methods for exposure modeling (comparing SR with EM) enhanced the robustness of our findings.
Nevertheless, our study is limited by its retrospective design. The number of available ultrasound reports varied between patients, which may have affected our ability to assess the timing of deteriorating parameters and could have led to under‐reporting. Furthermore, we lacked sufficiently detailed data to further subclassify cAREDF (continuously absent or continuously reversed) and iAREDF (mostly positive, absent or reversed), or to account for the role of middle cerebral artery Doppler measurements. We reported GA at diagnosis as an indicator of sFGR severity, but did not assess it as a predictor for survival. The onset of sFGR influences severity, but its precise impact, alongside variations in other ultrasound parameters, should be examined prospectively. This necessity is underscored by the high percentage of changing UA flows, which led us to report outcomes based on final UA flow type. Additionally, we did not evaluate the prospective risk of fetal death, which decreases as pregnancy progresses as a result of early demise, fetal intervention or iatrogenic delivery. When interpreting rates of survival without severe neonatal morbidity, it is important to note that non‐severe neonatal morbidity cannot be excluded. However, the clinical relevance of non‐severe morbidity is often limited. It was beyond the scope of this study to account for differences in protocols between neonatal intensive care units and their potential influence on neonatal outcome across centers.
For our secondary objective (comparing outcomes of the larger twin after SR vs EM), limiting the GA at diagnosis to ≤ 24 weeks may have excluded milder cases, potentially lowering the ATE, although SR may be restricted in these cases. Additionally, ATE analysis for the composite outcome without a GA limit and for live birth alone should be interpreted cautiously due to lower event rates. We assume conditional exchangeability, with confounder selection based on the literature, further supported by observed differences between our EM and SR cohorts (Table 5) 10 , 12 , 19 , 33 , 34 .
To address these limitations, future research should prioritize longitudinal prospective studies assessing the predictive value of various ultrasound parameters and include long‐term follow‐up of twins. The prospective CONTRAST study (ClinicalTrials.gov identifier: NCT05952583), which is currently recruiting, exemplifies this approach. Such studies are essential for refining newly proposed subgroup classifications 35 and improving outcome prediction. This should enable clinicians to tailor management strategies, optimizing the timing and criteria for fetal intervention and delivery.
Conclusions
Our findings underscore the substantial risk of adverse outcomes of sFGR twins with cAREDF or iAREDF, including increased risks of dIUD and severe cerebral injury for the surviving twin following sIUD. SR of the smaller twin improves the prognosis for the larger twin by prolonging gestation and reducing severe neonatal morbidity, although the live‐birth rate of the larger twin remains comparable to that following EM. These results could guide clinicians in navigating between Scylla and Charybdis, by supporting patient counseling and informing clinical decision‐making for severe sFGR.
Supporting information
Appendix S1 ALIGN Study Group collaborators
Table S1 Management guidelines by center for selective fetal growth restriction with continuous or intermittent absent or reversed end‐diastolic flow in umbilical artery
Table S2 Demographic characteristics of cases that underwent termination of pregnancy
Table S3 Outcome after laser coagulation performed solely for selective fetal growth restriction
Table S4 Comparison of outcomes between cases with normalized vs persistently abnormal umbilical artery flow
Table S5 Cerebral impact for surviving twin after single fetal demise
Figure S1 Flowchart summarizing selection of cohort for secondary aim (comparison of outcomes for larger twin after selective reduction vs expectant management).
ACKNOWLEDGMENTS
We thank Claudine Stark and Britt Min for their statistical insights, which enhanced our understanding of the data. We are grateful to Mari Flodström for her exceptional speed and accuracy in data collection for the Swedish patients. Additionally, we thank all the specialists involved in scanning the twins, acknowledging the patience and skill required to obtain quality images, especially in complex twin pregnancies.
L.L. received a grant from the FWO Fonds Wetenschappelijk Onderzoek (1804723N), L.H. received funding from Region Stockholm, L.H. and A.N. were supported by the Center for Innovative Medicine (CIMED), and A.N. also received funding from the Swedish Society of Medicine (SLS‐984967), the Stichting Jo Kolk Studiefonds and LUMC Global.
Contributor Information
the ALIGN Study Group:
Femke Slaghekke, Monique C Haak, Enrico Lopriore, Lotte E van der Meeren, Johannes van der Merwe, and Roland Devlieger
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- 1. Gratacós E, Lewi L, Muñoz B, et al. A classification system for selective intrauterine growth restriction in monochorionic pregnancies according to umbilical artery Doppler flow in the smaller twin. Ultrasound Obstet Gynecol. 2007;30(1):28‐34. [DOI] [PubMed] [Google Scholar]
- 2. Homer . The Odyssey Murray AT. The Odyssey (English Translation). Harvard University Press; 2008. [Google Scholar]
- 3. Townsend R, D'Antonio F, Sileo FG, Kumbay H, Thilaganathan B, Khalil A. Perinatal outcome of monochorionic twin pregnancy complicated by selective fetal growth restriction according to management: systematic review and meta‐analysis. Ultrasound Obstet Gynecol. 2019;53(1):36‐46. [DOI] [PubMed] [Google Scholar]
- 4. Couck I, Ponnet S, Deprest J, Devlieger R, De Catte L, Lewi L. Outcome of monochorionic twin pregnancy with selective fetal growth restriction at 16, 20 or 30 weeks according to new Delphi consensus definition. Ultrasound Obstet Gynecol. 2020;56(6):821‐830. [DOI] [PubMed] [Google Scholar]
- 5. Gratacós E, Antolin E, Lewi L, et al. Monochorionic twins with selective intrauterine growth restriction and intermittent absent or reversed end‐diastolic flow (type III): feasibility and perinatal outcome of fetoscopic placental laser coagulation. Ultrasound Obstet Gynecol. 2008;31(6):669‐675. [DOI] [PubMed] [Google Scholar]
- 6. Koch A, Favre R, Viville B, et al. Expectant management and laser photocoagulation in isolated selective intra‐uterine growth restriction: a single‐center series. J Gynecol Obstet Hum Reprod. 2017;46(10):731‐736. [DOI] [PubMed] [Google Scholar]
- 7. Miyadahira MY, Brizot ML, Carvalho MHB, et al. Type II and III selective fetal growth restriction: perinatal outcomes of expectant management and laser ablation of placental vessels. Clinics (Sao Paulo). 2018;73:e210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Colmant C, Lapillonne A, Stirnemann J, et al. Impact of different prenatal management strategies in short‐ and long‐term outcomes in monochorionic twin pregnancies with selective intrauterine growth restriction and abnormal flow velocity waveforms in the umbilical artery Doppler: a retrospective observational study of 108 cases. BJOG. 2021;128(2):401‐409. [DOI] [PubMed] [Google Scholar]
- 9. Quintero RA, Bornick PW, Morales WJ, Allen MH. Selective photocoagulation of communicating vessels in the treatment of monochorionic twins with selective growth retardation. Am J Obstet Gynecol. 2001;185(3):689‐696. [DOI] [PubMed] [Google Scholar]
- 10. Peeva G, Bower S, Orosz L, Chaveeva P, Akolekar R, Nicolaides KH. Endoscopic placental laser coagulation in monochorionic diamniotic twins with type II selective fetal growth restriction. Fetal Diagn Ther. 2015;38(2):86‐93. [DOI] [PubMed] [Google Scholar]
- 11. Buskmiller C, Munoz JL, Cortes MS, Donepudi RV, Belfort MA, Nassr AA. Laser therapy versus expectant management for selective fetal growth restriction in monochorionic twins: a systematic review. Prenat Diagn. 2023;43(5):687‐698. [DOI] [PubMed] [Google Scholar]
- 12. Mustafa HJ, Javinani A, Heydari MH, Saldaña AV, Rohita DK, Khalil A. Selective intrauterine growth restriction without concomitant twin‐to‐twin transfusion syndrome, natural history, and risk factors for fetal death: a systematic review and meta‐analysis. Am J Obstet Gynecol MFM. 2023;5(10):101105. [DOI] [PubMed] [Google Scholar]
- 13. Batsry L, Matatyahu N, Avnet H, et al. Perinatal outcome of monochorionic diamniotic twin pregnancy complicated by selective intrauterine growth restriction according to umbilical artery Doppler flow pattern: single‐center study using strict fetal surveillance protocol. Ultrasound Obstet Gynecol. 2021;57(5):748‐755. [DOI] [PubMed] [Google Scholar]
- 14. Valsky DV, Eixarch E, Martinez JM, Crispi F, Gratacós E. Selective intrauterine growth restriction in monochorionic twins: pathophysiology, diagnostic approach and management dilemmas. Semin Fetal Neonatal Med. 2010;15(6):342‐348. [DOI] [PubMed] [Google Scholar]
- 15. Shinar S, Xing W, Pruthi V, et al. Outcome of monochorionic twin pregnancy complicated by type‐III selective intrauterine growth restriction. Ultrasound Obstet Gynecol. 2021;57(1):126‐133. [DOI] [PubMed] [Google Scholar]
- 16. Emrani SE, Groene SG, Verweij EJ, et al. Gestational age at birth and outcome in monochorionic twins with different types of selective fetal growth restriction: a systematic literature review. Prenat Diagn. 2022;42(9):1094‐1110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Lewi L. Monochorionic diamniotic twins: what do I tell the prospective parents? Prenat Diagn. 2020;40(7):766‐775. [DOI] [PubMed] [Google Scholar]
- 18. Multifetal gestations: twin, triplet, and higher‐order multifetal pregnancies: ACOG practice bulletin, number 231. Obstet Gynecol. 2021;137(6):e145‐e162. [DOI] [PubMed] [Google Scholar]
- 19. Van Mieghem T, Lewi L, Slaghekke F, et al. Prediction of fetal death in monochorionic twin pregnancies complicated by type‐III selective fetal growth restriction. Ultrasound Obstet Gynecol. 2022;59(6):756‐762. [DOI] [PubMed] [Google Scholar]
- 20. Couck I, Van der Merwe J, Russo F, et al. Umbilical venous diameters and flow in monochorionic diamniotic twin pregnancies: association with placental sharing and fetal demise. Ultrasound Obstet Gynecol. 2022;60(4):514‐522. [DOI] [PubMed] [Google Scholar]
- 21. Senat MV, Deprest J, Boulvain M, Paupe A, Winer N, Ville Y. Endoscopic laser surgery versus serial amnioreduction for severe twin‐to‐twin transfusion syndrome. N Engl J Med. 2004;351(2):136‐144. [DOI] [PubMed] [Google Scholar]
- 22. Quintero RA, Morales WJ, Allen MH, Bornick PW, Johnson PK, Kruger M. Staging of twin‐twin transfusion syndrome. J Perinatol. 1999;19(8 Pt 1):550‐555. [DOI] [PubMed] [Google Scholar]
- 23. Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med. 2001;163(7):1723‐1729. [DOI] [PubMed] [Google Scholar]
- 24. Spruijt MS, van Klink JMM, de Vries LS, et al. Fetal and neonatal neuroimaging in twin‐twin transfusion syndrome. Ultrasound Obstet Gynecol. 2024;63(6):746‐757. [DOI] [PubMed] [Google Scholar]
- 25. Rondagh M, Lopriore E, de Vries LS, et al. The current knowledge of cerebral magnetic resonance imaging in monochorionic twins: a systematic review of the last 20 years. J Clin Med. 2023;12(23):7211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Team RC . R: A language and environment for statistical computing. 2021.
- 27. Mackie FL, Rigby A, Morris RK, Kilby MD. Prognosis of the co‐twin following spontaneous single intrauterine fetal death in twin pregnancies: a systematic review and meta‐analysis. BJOG. 2019;126(5):569‐578. [DOI] [PubMed] [Google Scholar]
- 28. Lanna MM, Consonni D, Faiola S, et al. Incidence of cerebral injury in monochorionic twin survivors after spontaneous single demise: long‐term outcome of a large cohort. Fetal Diagn Ther. 2020;47(1):66‐73. [DOI] [PubMed] [Google Scholar]
- 29. Shinar S, Harris K, Van Mieghem T, et al. Early imaging predictors of fetal cerebral ischemic injury in monochorionic twin pregnancy complicated by spontaneous single intrauterine death. Ultrasound Obstet Gynecol. 2022;59(4):497‐505. [DOI] [PubMed] [Google Scholar]
- 30. Moradi B, Badraqe N, Rahimi Sharbaf F, et al. Early detection of ischemic brain injuries by diffusion‐weighted imaging after radiofrequency ablation for fetal reduction in monochorionic pregnancies. J Clin Ultrasound. 2022;50(9):1288‐1296. [DOI] [PubMed] [Google Scholar]
- 31. Tedjawirja VN, van Klink JM, Haak MC, et al. Questionable benefit of intrauterine transfusion following single fetal death in monochorionic twin pregnancy. Ultrasound Obstet Gynecol. 2022;59(6):824‐825. [DOI] [PubMed] [Google Scholar]
- 32. Kontopoulos E, Quintero RA. Fetoscopic demonstration of feto‐fetal hemorrhage before actual demise of a monochorionic twin, successful antenatal treatment, and narrative review of the literature. J Matern Fetal Neonatal Med. 2023;36(1):2215898. [DOI] [PubMed] [Google Scholar]
- 33. Monaghan C, Kalafat E, Binder J, Thilaganathan B, Khalil A. Prediction of adverse pregnancy outcome in monochorionic diamniotic twin pregnancy complicated by selective fetal growth restriction. Ultrasound Obstet Gynecol. 2019;53(2):200‐207. [DOI] [PubMed] [Google Scholar]
- 34. Chmait SR, Monson MA, Korst LM, Llanes A, Chon AH. Selective fetal growth restriction type III: application of a recent expert consensus definition. J Ultrasound Med. 2022;41(7):1657‐1666. [DOI] [PubMed] [Google Scholar]
- 35. Chmait RH, Chon AH, Korst LM, Stephen Y, Llanes A, Ouzounian JG. Selective intrauterine growth restriction (SIUGR) type II: proposed subclassification to guide surgical management. J Matern Fetal Neonatal Med. 2022;35(6):1184‐1191. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1 ALIGN Study Group collaborators
Table S1 Management guidelines by center for selective fetal growth restriction with continuous or intermittent absent or reversed end‐diastolic flow in umbilical artery
Table S2 Demographic characteristics of cases that underwent termination of pregnancy
Table S3 Outcome after laser coagulation performed solely for selective fetal growth restriction
Table S4 Comparison of outcomes between cases with normalized vs persistently abnormal umbilical artery flow
Table S5 Cerebral impact for surviving twin after single fetal demise
Figure S1 Flowchart summarizing selection of cohort for secondary aim (comparison of outcomes for larger twin after selective reduction vs expectant management).
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.