Résumé
Background: The number of twin pregnancies continues to increase worldwide as both the number of pregnancies obtained by medically assisted reproduction and age at first pregnancy keep rising. Preterm delivery is the major complication associated with twin pregnancies. The effectiveness of preventive treatments such as progesterone or cervical cerclage for women with a short cervix is doubtful in twin pregnancies. The effectivity of cervical pessaries in preventing preterm birth and its associated morbidity and mortality is also controversial.Objective: We sought to investigate if the Arabin pessary reduces adverse neonatal outcomes in twin pregnancies with a short cervix.Study design: This open-label, multicenter, randomized controlled trial on twin pregnancies with a cervical length of <35 mm compared pessary placement at 16+0 to 24+0 weeks' gestation with standard care alone. The primary endpoint was a composite of adverse neonatal outcomes, namely peripartum or neonatal death or significant neonatal morbidity before hospital discharge, defined as at least 1 of the following complications: bronchopulmonary dysplasia, intraventricular hemorrhage grade III to IV, periventricular leukomalacia, necrotizing enterocolitis grade II or higher, culture-proven sepsis, and retinopathy requiring treatment. A sample size of 308 pregnancies was planned to ensure 80% power to compare the proportions of women with at least 1 infant with an adverse neonatal outcome. The intention-to-treat analysis after multiple imputation of missing data, was supplemented with a secondary analysis that controlled for gestational age and cervical length, both at inclusion. The primary endpoint was also compared between randomization groups in the per-protocol population, which excluded patients with prespecified major protocol violations (mostly cervical cerclage and/or progesterone after inclusion). Secondary endpoints included preterm birth, spontaneous preterm birth, and pessary side effects.Results: In total, 315 women were randomized to either receive a pessary (n=157) or standard management (n=158). Overall, 10.8% (34 women) of participants had a missing value for the primary endpoint, mostly (79%) because of the lack of paternal consent for neonatal data collection. In the intention-to-treat analysis, the adverse neonatal outcome occurred in 16.8% of the pessary group vs in 22.5% of the control group (risk ratio, 0.69; 95% confidence interval, 0.39-1.23; P=.210). The per-protocol analysis did not show any significant difference between groups (risk ratio, 0.78; 95% confidence interval, 0.47-1.28; P=.320). The occurrence of preterm birth or spontaneous preterm birth did not differ significantly between groups. No serious side effects were associated with pessary use.Conclusion: Pessary use in our study did not significantly reduce adverse neonatal outcomes in twin pregnancies with a short cervix.