Removal versus retention of cerclage in preterm premature rupture of membranes: a randomized controlled trial

The decision of whether to retain or remove a previously placed cervical cerclage in women who subsequently rupture fetal membranes in a premature gestation is controversial and all studies to date are retrospective. We performed a multicenter randomized controlled trial of removal vs retention of cerclage in these patients to determine whether leaving the cerclage in place prolonged gestation and/or increased the risk of maternal or fetal infection.

A prospective randomized multicenter trial of 27 hospitals was performed. Patients included were those with cerclage placement at ≤23 weeks 6 days in singleton or twin pregnancies, with subsequent spontaneous rupture of membranes between 22 weeks 0 days and 32 weeks 6 days. Patients were randomized to retention or removal of cerclage. Patients were then expectantly managed and delivered only for evidence of labor, chorioamnionitis, fetal distress, or other medical or obstetrical indications. Management after 34 weeks was at the clinician's discretion.

The initial sample size calculation determined that a total of 142 patients should be included but after a second interim analysis, futility calculations determined that the conditional power for showing statistical significance after randomizing 142 patients for the primary outcome of prolonging pregnancy was 22.8%. Thus the study was terminated after a total of 56 subjects were randomized with complete data available for analysis, 32 to removal and 24 to retention of cerclage. There was no statistical significance in primary outcome of prolonging pregnancy by 1 week comparing the 2 groups (removal 18/32, 56.3%; retention 11/24, 45.8%) P = .59; or chorioamnionitis (removal 8/32, 25.0%; retention 10/24, 41.7%) P = .25, respectively. There was no statistical difference in composite neonatal outcomes (removal 16/33, 50%; retention 17/30, 56%), fetal/neonatal death (removal 4/33, 12%; retention 5/30, 16%); or gestational age at delivery (removal mean 200 days; retention mean 198 days).

Statistically significant differences were not seen in prolongation of latency, infection, or composite neonatal outcomes. However, there was a numerical trend in the direction of less infectious morbidity, with immediate removal of cerclage. These findings may not have met statistical significance if the original sample size of 142 was obtained, however they provide valuable data suggesting that there may be no advantage to retaining a cerclage after preterm premature rupture of membranes and a possibility of increased infection with cerclage retention.