Progesterone flops for miscarriage, preterm birth prevention in PREEMPT

In pregnant women who present with first-trimester vaginal bleeding, treatment with progesterone throughout pregnancy does not appear to be beneficial in terms of lowering the risk of miscarriage or premature delivery, according to PREEMPT trial presented at this year’s American College of Obstetricians and Gynecologists (ACOG 2022) meeting.

Compared with placebo, progesterone yielded similar rates of spontaneous miscarriage <20 weeks (14.4 percent vs 16.0 percent; p=0.94), preterm birth (10.2 percent vs 12.3 percent; p=0.46), and stillbirth (0.9 percent vs 1.3 percent; p=0.65). [ACOG 2022, abstract A260]

There were likewise no differences observed in adverse maternal or newborn outcomes, or in time-to-event for miscarriage or birth between the groups.

“First-trimester bleeding is associated with miscarriage and preterm birth, for which progesterone has been suggested to improve outcomes,” said lead trial investigator Dr Haim Abenhaim of the Jewish General Hospital at McGill University, Montreal, US, who spoke to an audience at the meeting.

The findings suggest that progesterone is not the best drug to prescribe for the prevention of miscarriage or premature delivery in otherwise low-risk pregnancies with first-trimester bleeding, Abenhaim added.

Overall, the present data are in line with the report from the PRISM RCT trial, which examined the effect of progesterone use on the incidence of miscarriage in women with early pregnancy bleeding. Specifically, progesterone was not beneficial, falling short of preventing miscarriage when compared with placebo. However, treatment appeared to have a favourable effect in the subgroup of women with early pregnancy bleeding and a previous history of miscarriages. [Health Technol Assess 2020;24:1-70]

PREEMPT included 549 women with vaginal bleeding and live intrauterine pregnancy <14 weeks of gestation. They were randomized to receive 200-mg micronized progesterone or an identically appearing placebo. Treatment was vaginally administered every night, from presentation to 34 weeks of gestation.

Of the women, 16 withdrew or were lost to follow-up, leaving 264 patients in the progesterone group and 269 patients in the placebo group. Patient characteristics at baseline were comparable in the two treatment groups. None of the patients had multifetal gestations, cervical insufficiency, recurrent pregnancy loss, or bleeding unrelated to placentation.

There were 197 women in the progesterone groups and 190 in the placebo group who delivered live neonates at term (74.6 percent vs 70.6 percent; p=0.30).

Abenhaim acknowledged a couple of issues in PREEMPT, such as underreporting of subchorionic haemorrhage with radiologic confirmation. This, in turn, yielded a smaller population and a change in protocol to include patients with no identifiable secondary source of bleeding.

In addition, the COVID-19 pandemic caused the enrolment to be stopped. Nevertheless, Abenhaim and colleagues halted PREEMPT altogether, as interim analysis showed no likely benefit.