The late vanishing of a co-twin contributes to adverse perinatal outcomes in the surviving singleton

STUDY QUESTION: Is the vanishing of a co-twin after or before the ultrasonic registration of cardiac activity at approximately 6-8 weeks of gestation associated with adverse perinatal outcomes? SUMMARY ANSWER: The timing of the demise of a co-twin after the registration of cardiac activity is an independent risk factor for adverse perinatal outcomes. WHAT IS KNOWN ALREADY: A significant body of evidence has confirmed that vanishing twin (VT) pregnancies are associated with higher levels of risk for preterm birth (PTB), low birthweight (LBW), small-for-gestational age (SGA) and perinatal mortality, compared with singleton pregnancy. However, the impact of co-twin vanishing, before or after the presence, of cardiac activity, on perinatal outcomes has yet to be extensively investigated. STUDY DESIGN, SIZE, DURATION: We retrospectively reviewed the medical records of 38 876 singletons delivered from ART cycles between 2006 and 2018, at the Peking University Third Hospital. PARTICIPANTS/MATERIALS, SETTING, METHODS: In total, 35 188 singletons were delivered from the singleton pregnancy group, 2256 singletons from the VT pregnancy group after that cardiac activity was noted, and 1432 singletons were delivered from the VT pregnancy group before cardiac activity could be registered. Using the Poisson model, the adjusted risk ratio (aRR) was used to estimate the incidence of PTB, LBW, SGA and perinatal mortality, in the pregnancies of two types of VT compared with singleton pregnancies after correction for potential confounding factors. MAIN RESULTS AND THE ROLE OF CHANCE: The vanishing of a co-twin after the registration of cardiac activity was associated with an increased risk of perinatal mortality when compared with the group of singleton pregnancies (0.5% vs 0.2%; P = 0.006); this association still existed after adjustment for potential confounders (aRR 2.19, 95% CI 1.12-4.30; P = 0.023). Furthermore, it was significantly associated with a higher risk of PTB (all cycles aRR 2.00, 95% CI 1.77-2.24; P<0.001; fresh transfer aRR 2.06, 95% CI 1.78-2.38; P<0.001; frozen transfer aRR 1.87, 95% CI 1.52-2.28; P<0.001), LBW (all cycles aRR 2.47, 95% CI 2.12-2.88; P<0.001; fresh transfer aRR 2.50, 95% CI 2.07-3.02; P<0.001; frozen transfer aRR 2.39; 95% CI 1.83-3.12; P<0.001) and SGA (all cycles aRR 1.56, 95% CI 1.35-1.80; P<0.001; fresh transfer aRR 1.53, 95% CI 1.29-1.81; P<0.001; frozen transfer aRR 1.62, 95% CI 1.24-2.11; P<0.001). However, prior to the presence of cardiac activity, the vanishing of a co-twin was not associated with a higher risk of perinatal mortality (all cycles aRR 0.71, 95% CI 0.17-2.92; P = 0.636; fresh cycles aRR 0.51, 95% CI 0.07-3.70; P = 0.502; frozen cycles aRR 1.29, 95% CI 0.17-9.66; P = 0.803), PTB (all cycles aRR 1.11, 95% CI 0.91-1.34; P = 0.301; fresh cycles aRR 1.10, 95% CI 0.87-1.39; P = 0.447; frozen cycles aRR 1.13, 95% CI 0.81-1.58; P = 0.467), LBW (all cycles aRR 1.19, 95% CI 0.91-1.55; P = 0.207; fresh cycles aRR 1.08, 95% CI 0.77-1.51; P = 0.668; frozen cycles aRR 1.45, 95% CI 0.93-2.25; P = 0.100) and SGA (all cycles aRR 1.09, 95% CI 0.89-1.35; P = 0.405; fresh cycles aRR 0.97, 95% CI 0.75-1.26; P = 0.839). Pregnancies involving the two types of VT were significantly different in terms of PTB (all cycles aRR 1.80, 95% CI 1.45-2.24; P<0.001; fresh cycles aRR 1.88, 95% CI 1.44-2.45; P<0.001; frozen cycles aRR 1.65, 95% CI 1.13-2.40; P = 0.009), LBW (all cycles aRR 2.08, 95% CI 1.55-2.79; P<0.001; fresh cycles aRR 2.32, 95% CI 1.61-3.36; P<0.001; frozen cycles aRR 1.65, 95% CI 1.01-2.70; P = 0.046) and SGA (all cycles aRR 1.70, 95% CI 1.36-2.11; P<0.001; fresh cycles aRR 1.87, 95% CI 1.42-2.45; P<0.001). LIMITATIONS, REASONS FOR CAUTION: The present data are not able to differentiate between co-twin demise occurring in the first or second trimester. Because the second trimester ultrasound scan is not an integral aspect of IVF assessment, this information was not available in the database. WIDER IMPLICATIONS OF THE FINDINGS: Adverse perinatal outcomes in ART babies can be avoided by replacing one embryo at a time. It is possible to apply selective single embryo transfer strategy for all while maintaining acceptable success rates.