Identifying an optimal inter-twin weight discordance cutoff for predicting donor survival after fetoscopic laser surgery for twin-twin transfusion syndrome.

BACKGROUND: Increased inter-twin growth discordance has been associated with adverse pregnancy outcomes. An inter-twin growth discordance of at least 25% is often considered as a clinical threshold for pathological discordance in unselected monochorionic-diamniotic (MCDA) twins, but an optimal cutoff for predicting survival outcomes following fetoscopic laser surgery (FLS) for twin-twin transfusion syndrome (TTTS) is unclear. OBJECTIVES: This study aimed to (1) compare outcomes of FLS for TTTS based upon on the presence or absence of discordance ≥25%, (2) evaluate the association of different inter-twin discordance thresholds on donor twin survival after FLS, and (3) identify an optimal discordance cutoff for the prediction of donor twin survival. STUDY DESIGN: Retrospective review of MCDA twins with TTTS that underwent FLS at a single center from 2009-22. The primary outcome was donor twin survival at first sonogram after FLS. Pregnancies with pre-operative inter-twin discordance ≥ 25% were compared to those < 25%. Differing thresholds of inter-twin discordance were compared to determine predictive accuracy for donor twin survival, and discordance was examined as a continuous variable to determine an optimal cutoff for donor twin survival prediction. RESULTS: 169 MCDA pregnancies underwent FLS and met inclusion criteria and 62 (37%) had inter-twin discordance ≥25%. Patients with inter-twin discordance ≥25% were younger (32 ± 5 years vs 30 ± 6 years, p=0.03). Gestational age at TTTS diagnosis and at FLS were similar between groups. Donor twin selective fetal growth restriction (sFGR) was more likely with discordance ≥25% (74.2% vs 23.4%, p<0.01). Pre-FLS Quintero stage differed between the two groups (p=0.014), with higher likelihood of stage III and IV disease in the ≥25% group. Discordance ≥25% was associated with lower donor twin survival at first post-FLS sonogram (75.8% vs 89.7%, p=0.02) and at newborn hospital discharge (52.8% vs 73.6%, p=0.04). After adjusting for age, Quintero stage, and donor FGR, donor twin survival at first sonogram remained lower in patients with discordance ≥ 25% (p=0.01). Discordance cutoffs of ≥25% and ≥30% were significantly associated with decreased donor twin survival at first sonogram post-FLS (p=0.03 and p<0.001, respectively) and this was true after adjusting for donor fetal growth restriction and pre-FLS Quintero stage. In fitting a logistic regression model, discordance as a continuous variable was significantly associated with predicting donor twin survival (p=0.02). The area under the curve (AUC) for predicting donor twin survival was 0.63 (95% CI: 0.49, 0.74, p=0.06). Based on the Youden index, the optimum cut point was a discordance of 29%, which had a sensitivity of 0.75, specificity of 0.58, positive predictive value of 0.91, and negative predictive value of 0.29 for predicting donor twin survival. CONCLUSION: Discordance ≥25% was associated with lower donor twin survival after FLS for TTTS; however, inter-twin discordance alone is a poor predictor of donor twin survival after FLS for TTTS. If it is to be considered for use in clinical practice, a cutoff value of 29% was most predictive of donor twin survival in this cohort and may be useful for preoperative counseling on expected outcomes.