American Journal of Obstetrics & Gynecology, Volume 224, Issue 5, P552, May 01, 2021 – 2021-01-19
Claire J. Mcgnnon1, MBBS, Vinh Q. Dang, MD2, Ben W. Mol, professor1,3
Published: January 19, 2021
1Monash Medical Centre 10 246 Clayton Rd, Clayton Victoria 3168 Australia
2IVFMD, My Duc Hospital 4 Nui Thanh, Tan Binh District, Ho Chi Minh City, Vietnam
3Monash University 246 Clayton Rd, Clayton Victoria 3168 Australia
We read with interest the recent randomized clinical trial comparing vaginal progesterone
versus placebo in women with twin pregnancies for the prevention of spontaneous preterm birth (1). We want to congratulate the authors on this large study that answers the question as to whether high dose vaginal progesterone reduces preterm birth in unselected women with a twin pregnancy: it does not. However, we have concerns regarding the study design, which downplays the potential of progesterone in women with short cervix.
Our concern is the primary outcome: spontaneous preterm birth between 24+0 and 33+6 weeks gestation, reported as a RR of 1.35 with a 95% confidence interval of 0.88 to 2.05. The choice not to include women who deliver before 24 weeks makes the groups on the primary endpoint less comparable, specifically because 2.6% of the progesterone group delivered before 24 weeks as compared to 4.4% in the placebo group. In our opinion, the primary outcome should have been any delivery before 34 weeks, which occurred in 16.7% and 15.8% respectively in this study, with RR 1.10 (0.80 to 1.51). It is not surprising that delivery between 24-34 weeks occurs more in women with compliance above 80% (RR 1.73, 1.04 to 2.91), which is not a sign of harm, but probably due to the fact that with good compliance progesterone prevents preterm birth before 24 weeks (we cannot extract these data from the paper).
This becomes even more important when the authors report the treatment effect in women with a cervical length below 30mm. Out of the massive number of comparisons in the paper, the most relevant one is shown in supplementary table 7. While in women with a cervical length of 30 mm or more the endpoint of delivery < 34 weeks is not influenced by progesterone (RR 1.1 (0.82 – 1.5)), there is a clear benefit from progesterone (RR 0.44, 95% CI 0.17 to 0.91) in women with a cervix <30 mm.
Claire McGannon, MD 1
Vinh Q. Dang, MD 2
Ben W. Mol, professor 1 3
- Rehal A, Benkő Z, De Paco Matallana C, et al. Early vaginal progesterone versus placebo in 94 twin pregnancies for the prevention of spontaneous preterm birth: a randomized, double-blind 95 trial [published online ahead of print, 2020 Jun 26]. Am J Obstet Gynecol. 2020;S0002- 96 9378(20)30675-X. doi:10.1016/j.ajog.2020.06.050