Use of Oral Progestogen in Women With Threatened Miscarriage in the First Trimester

A Randomized Double-blind Controlled Trial

Diana Man Ka Chan; Ka Wang Cheung; Jennifer Ka Yee Ko; Sofie Shuk Fei Yung; Shui Fan Lai; Mei Ting Lam; Dorothy Yuet Tao Ng; Vivian Chi Yan Lee; Raymond Hang Wun Li; Ernest Hung Yu Ng


Hum Reprod. 2021;36(3):587-595. 

In This Article


Our study showed that use of oral progestogen in women with threatened miscarriage in the first trimester did not reduce the miscarriage rate or improve the live birth rate. This was in contrast to the subgroup analysis of the Cochrane meta-analysis (Wahabi et al., 2018), which found that treatment of miscarriage with oral progestogens compared to placebo (Turgal et al., 2017) or no treatment (El-Zibdeh and Yousef, 2009; Pandian, 2009) reduced the risk of miscarriage. The latest meta-analysis (Li et al., 2020), which included the recent large randomized trial (Coomarasamy et al., 2019), also showed the use of oral progestogen reduced risk of miscarriage and increased live birth rate.

In the Cochrane meta-analysis (Wahabi et al., 2018), three studies (El-Zibdeh and Yousef, 2009; Pandian, 2009; Turgal et al., 2017) out of the seven included trials using oral progestogen in threatened miscarriage. However, high risk of bias was noted with a lack of blinding in studies. Small sample sizes [n = 146 (El-Zibdeh and Yousef, 2009) and n = 191 (Pandian, 2009)] and relatively higher miscarriage rates in the control group [25.0% (El-Zibdeh and Yousef, 2009) and 28.4% (Pandian, 2009)] were noted in some of these included trials. The study by Alimohamadi et al. (2013) was a randomized double-blind controlled trial of 160 women but there were no clinically significant differences in the miscarriage rate between the oral progestogen and placebo groups. Other two studies were also small in size [n = 83 (Turgal et al., 2017) and n = 60 (Yassaee et al., 2014)] and not double-blinded. There was again no significant difference in the rate of miscarriage between the two groups. Similarly, the latest meta-analysis (Li et al., 2020), including the PRISM trial (Coomarasamy et al., 2019), showed the use of oral progestogen reduced risk of miscarriage (RR 0.58, 95% CI 0.42–0.80; P = 0.001) and increased live birth rate (RR 1.17, 95% CI 1.04–1.31; P = 0.008), but not with vaginal progesterone: the conclusion was in contrast to our results. However, the result of the oral progestogen group in the Li et al. (2020) meta-analysis was based on three small randomized trials with poor study methodology.

In the PROMISE trial (Coomarasamy et al., 2015), vaginal progesterone in the first trimester of pregnancy did not result in a significantly higher rate of live births among women with a history of unexplained recurrent miscarriages. In the PRISM trial (Coomarasamy et al., 2019), among women with bleeding in early pregnancy, vaginal progesterone administered during the first trimester also did not result in a significantly higher rate of live births than placebo. These results echo our study findings after oral hormone administration. However, the PROMISE and PRISM trials studied the effect of vaginal micronized progesterone, which has an identical molecular structure to natural progesterone. We differed by investigating the effect of oral synthetic progestogen on women presenting with the first-trimester miscarriage.

We are aware that for women with three or more previous miscarriages, there was a 15% increase in live birth rate (72% vs 57%; RR 1.28; 95% CI 1.08–1.51; P = 0.004) with use of vaginal progesterone in the PRISM trial (Coomarasamy et al., 2019). However, this was a secondary analysis of a small subgroup of 183 women and its recommendation on this specific group of women was still uncertain.

One of the strengths of our study was that it was a randomized double-blind controlled trial. Four subgroup analyses were performed and all revealed no significant differences in the miscarriage rate between treatment and placebo groups. Moreover, we included women with early pregnancy of uncertain viability and this enhances the generalizability of the results.

Our study has limitations. The miscarriage rate instead of the live birth rate was chosen as the primary outcome, although we trace the live birth rate and obstetric outcomes. Our sample size was larger than published trials using oral progestogens but not adequate to demonstrate a smaller difference in the miscarriage rate between the progestogen and placebo groups. The primary outcome was not available in nine and eight women in the progestogen group and the placebo group, respectively. We assumed that all these women had a miscarriage. However, a significant difference in the primary outcome between the two groups in favour of the progestogen group was found only when all nine women in the progestogen group did not have miscarriage and all eight women in the placebo group had miscarriage but this is very unlikely. We were unable to save all tissue masses for chromosomal studies after miscarriage. Women were recruited from the Early Pregnancy Assessment clinics which ran in the morning during weekdays and those with heavy bleeding would be admitted into wards through the Department of Accident and Emergency. We did not exclude women with multiple pregnancy, which increased the risk of miscarriage although there was only one set of twin pregnancy in the placebo group. Women subjectively graded the severity of vaginal bleeding as mild, moderate and severe, rather than using an objective measure e.g. pictorial chart.

The issue of compliance was addressed, as women often miss drugs on some occasions in reality. Nevertheless, 70% of women in the progestogen group had a drug compliance of >80% in our study. The most common reported side effect was nausea and vomiting, occurring in up to one-third of women in both groups with no significant difference between the two groups. This could be due to pregnancy itself rather than side effect of the intervention. There were also no significant differences in the secondary outcomes including obstetric complications. Thus, the use of oral progestogen in the first trimester overall appeared to be safe. Regarding its safety in pregnancy, despite some early suggestions that progestogens may increase the risk of congenital developmental disorders (Goujard and Rumeau-Rouquette, 1977; Nora et al., 1978), evidence from subsequent large prospective studies and meta-analyses indicates that any such teratogenic effects are unlikely (Katz et al., 1985; Resseguie et al., 1985; Raman-Wilms et al., 1995). A recent review of maternal use of dydrogesterone during pregnancy also found no evidence for an increased risk of congenital malformations (Queisser-Luft, 2009) and was not able to detect any long-term complications of dydrogesterone use in pregnancy. Miscarriage has multiple causes. Therefore, giving progesterone or progestogen blindly will not be beneficial and other diagnostic tools are necessary to guide treatment of this common problem.