Concurrent progestogen and cerclage to reduce preterm birth: a multicenter international retrospective cohort

BACKGROUND: Both progestogens and cerclage are individually effective in preterm birth prevention in high risk pregnancies. However, national and international guidelines cite a lack of data available to comment on the potential bene fit of concurrent progestogen therapy after cerclage has been placed. Studies to date have been small with mixed results regarding bene fit of concurrent progestogen with cerclage leaving uncertainty regarding best clinical practice. OBJECTIVE: This study aimed to evaluate whether cerclage with progestogen therapy was superior to cerclage alone in the prevention of spontaneous preterm birth in singleton pregnancies. METHODS: This is an international retrospective cohort study of singleton pregnancies, without major anomaly or aneuploidy, and with cerclage placed at 10 different institutions in the United States and Colombia from June 2016 to June 2020. Exclusion criteria were lack of documentation regarding whether progestogen was prescribed, unavailable delivery outcome, and pregnancy termination (spontaneous or induced) before 16 weeks ' gestation. The exposure of interest was progestogen use with cerclage placement, which included those who continued to use progestogen or who started progestogen after cerclage. The comparison group consisted of those without progestogen use after cerclage placement, which included those who had no progestogen use during the entire pregnancy or who initiated progestogen and then stopped it after cerclage placement. Progestogen type, cerclage indication, maternal baseline characteristics, and maternal/neonatal outcomes were collected. The primary outcome was spontaneous preterm birth at <37 weeks. The secondary outcomes were spontaneous preterm birth at <34 weeks, gestational age at delivery, and a composite neonatal outcome including >= 1 of the following: perinatal mortality, con firmed sepsis, grade III or IV intraventricular hemorrhage, retinopathy of prematurity, respiratory distress syndrome, and bronchopulmonary dysplasia. There were planned subgroup analyses by cerclage indication, progestogen type (vaginal progesterone vs 17-hydroxyprogesterone caproate), preterm birth history, and site. Continuous variables were compared in adjusted analyses with analysis of covariance, and categorical variables were compared with multivariable logistic regression, adjusting for potential confounders with adjusted odds ratio. A Cox regression survival curve was generated to compare latency to spontaneous delivery, censored after 37 weeks. RESULTS: During the study period, a total of 699 singletons met the inclusion criteria: 561 in the progestogen with cerclage group and 138 with cerclage alone. Baseline characteristics were similar, except the higher likelihood of previous spontaneous preterm birth in the progestogen group (61% vs 41%; P <.001). Within the progestogen group, 52% were on 17-hydroxyprogesterone caproate weekly, 44% on vaginal progesterone daily, and 3% on oral progesterone daily. Progestogen with cerclage was associated with a signi ficantly lower frequency of spontaneous preterm birth <37 weeks (31% vs 39%; adjusted odds ratio, 0.59 [0.39 -0.89]; P=.01) and <34 weeks (19% vs 27%; adjusted odds ratio, 0.55 [0.35 -0.87]; P=.01), increased latency to spontaneous delivery (hazard ratio for spontaneous preterm birth <37 weeks, 0.66 [0.49 -0.90]; P=.009), and lower frequency of perinatal death (7% vs 16%; adjusted odds ratio, 0.37 [0.20 -0.67]; P=.001). In planned subgroup analyses, association with reduced odds of preterm birth <37 weeks persisted in those on vaginal progesterone, those without a previous preterm birth, those with ultrasound- or examination-indicated cerclage, those who started progestogen therapy before cerclage, and in sites restricted to the United States. CONCLUSION: Use of progestogen with cerclage was associated with reduced rates of spontaneous preterm birth and early spontaneous preterm birth compared with cerclage alone. Although this study was not sufficiently powered for subgroup analysis, the strength of evidence for bene fit appeared greatest for those with ultrasound- or examination-indicated cerclage, and with vaginal progesterone. El resumen esta disponible en Espanol al final del art & iacute;culo.