Prevention of Preterm Birth

Abstract

Despite attempts to reduce the rates and risks of prematurity, effective primary interventions are still lacking. Progesterone supplementation has been recommended for women with prior preterm birth (PTB). However, the ideal formulation, optimal route of administration, effective dose, and long-term safety of the drug are still uncertain. This prospective, randomized, nonblinded, and non–placebo-controlled trial was undertaken to assess the efficacy and tolerability of vaginal progesterone gel compared with intramuscularly administered (IM) progesterone in reducing the rate of recurrent PTB before 34 weeks’ gestation. Eligible women were of any age and parity, had a singleton pregnancy, were at 14 to 18 weeks’ gestation, and had a history of 1 or more midtrimester previous PTBs or a cerclage placed in a previous pregnancy but not in the current pregnancy. For the vaginal gel, each applicator delivered 1.125 g Crinone containing 90 mg (8% gel) progesterone in a base, administered by the patient each day. For IM progesterone, each 1-mL ampoule contained 250 mg hydroxyprogesterone caproate. Patients in the IM group received a 2-week supply; 1 injection was given at the time of the visit, and she was then advised to visit the injection room or a clinic the following week. Routine antenatal care was provided, and compliance with the regimen was determined. Progesterone treatment began at 14 to 18 weeks’ gestation following randomization and continued until 36 completed weeks of gestation, the occurrence of premature rupture of membranes, or PTB. If a patient had PTB during the study, she was treated and remained in the study until delivery if her membranes remained intact. The study drug was continued during treatment for PTB. The primary outcome was delivery before 34 weeks’ gestation. The secondary outcomes were PTB at 34 to 37 weeks’ gestation and neonatal birth weight, death, and need for neonatal intensive care unit admission. A χ2 test or Fisher exact test was used to test for differences in qualitative data; the Student t test was used for quantitative data. P < 0.05 was considered significant. A Kaplan-Meier estimate was used to detect the probabilities of patients not delivering preterm with the use of the 2 different forms of progesterone. The 2 survival curves were compared using a log-rank test. P ≤ 0.05 indicated a significant difference between the curves. Of 547 women eligible for the study, 518 consented, with 262 randomized to vaginal progesterone and 256 to IM progesterone. The groups were similar in age, parity, body mass index, gestational age at randomization, mode of conception, cervical length, compliance, and history of prior PTB. Vaginal progesterone was associated with a lower percentage of deliveries before 34 weeks’ gestation than the IM preparation (P = 0.02; odds ratio [OR], 0.58; 95% CI, 0.37–0.89). This association was also found at 28 to 32 weeks (P = 0.04; OR, 0.46; 95% CI, 0.23–0.93). No statistically significant difference was observed between the groups regarding deliveries at other gestational weeks. A significant difference was observed in the probability of delivery among patients who used vaginal progesterone compared with those who used IM progesterone (P of log-rank test = 0.0023). Adverse effects were reported in 14.1% and 7.5% of patients in the IM and vaginal groups, respectively (P = 0.017; OR, 2.01; 95% CI, 1.12–3.63). The groups did not differ significantly in the other parameters. The infants of mothers given IM progesterone had a significantly higher rate of neonatal intensive care unit admission than those in the vaginal progesterone group (P = 0.006; OR, 0.53; 95% CI, 0.34–0.82). Other outcomes did not differ significantly between the groups. Vaginal progesterone was more effective than IM progesterone in reducing the number of deliveries before 34 weeks and at 28 to 32 weeks’ gestation and had fewer adverse effects. Because a prophylactic regimen to prevent PTB requires a long period of progesterone use, less invasive forms of administration are preferred. Data on the optimal route, dose, and duration of progesterone are needed; whether a dose-response relationship is in play between progesterone and its action to reduce PTB is uncertain. Data from other populations are needed to confirm these results.