Cervical Pessary vs. Vaginal Progesterone for Preventing Premature Birth in IVF Twin Pregnancies

This will be a randomized controlled trial.

Women with twin pregnancies, at 16-22 weeks of gestation, will be invited to participate into the study.

Subjects meeting the study criteria will be randomized into two groups: (1) treated with cervical pessary (Arabin) or (2) treated with 400mg vaginal progesterone, once daily. Randomization will be done by third party via telephone, using a computer generated random list, with a variable block size of 2, 4 or 8. Apart from randomization, patients will be examined and treated according to local protocol.

Patients in pessary group will have an Arabin pessary placed within a week after randomization. A pessary certified by European Conformity (CE0482, MED/CERT ISO 9003/ EN 46003; Dr. Arabin, Witten, Germany) will be inserted through the vagina of the woman in the recumbent position and will be placed upward around the cervix. The research-team members who inserted the Arabin pessary have experience with Arabin pessary for singleton pregnancy before.

Patients in progesterone group will use vaginal progesterone (Cyclogest 400mg) once daily before bedtime, starting from the day of randomization onwards. They will be given a monitoring sheet and instructed to note everyday the date of using. If they forget one dose of any night, and remember it in the next morning or afternoon, they will use immediately the forgotten dose and continue with the dose of that day at night. If one dose is missed until the next evening, there will be no compensation use, they will only use the dose of the next day. Any change in using medication should be noted in the monitoring sheet.

In both groups, intervention will be stopped at 36 weeks of gestation or at delivery. All the participants will have follow-up visits every 4 weeks. If patients develop (threatened) preterm labor, they will receive treatment as routine practice.

Statistical analyses will be by intention to treat. For dichotomous endpoints, we will calculate rates. These will be compared by calculating a relative risk and a 95% confidence interval. Between-group differences in non-continuous variables will be assessed using the χ2-test. Results of continuous variables were given in mean ± SD or in percentage. Between-group differences of continuous variables were assessed with the Student's t-test. We will consider correlation between neonatal endpoints when we analyse at the level of the child. We assessed time to delivery by Cox proportional hazard analysis and Kaplan-Meier estimates, and compared results with a log-rank test. We plan an exploratory subgroup analysis in women with a cervical length of less than the 25th percentile (according to the distribution in all twins), as well as 25th - 50th percentile, 50th - 75th percentile and > 75th percentile. We also plan an exploratory subgroups analyses for chorionicity. A p-value < 0.05 will be considered to indicate a statistically significant difference. The analysis will be done with statistical Package for Social Sciences version 19 (SPSS, USA).

Sample size has been set at 290. This was incorporated in an amendment of the protocol, and was approved by the IRB on 22 Sept 2016.

Inclusion criteria

To be eligible for enrolment into this trial, each female subject must fulfil all of the following criteria at the start of enrolment, unless specified otherwise:

• Women with a twin pregnancy (mono- and di-chorionic)
• 16 0/7 to 22 0/7 weeks of gestation
• Maternal age ≥ 18 yrs
• Cervical length less than 38 mm
• Informed consent
• Not participating in another PTB study at the same time

Exclusion criteria

To be eligible for enrolment in this study each subject must not meet any of the following criteria:

• History of cervical surgery
• Cervical cerclage in place
• Twin-to-twin transfusion syndrome
• Stillbirth or major congenital abnormalities in any of the fetus
• Severe vaginal discharge, acute vaginitis
• Premature rupture of membranes
• Premature labor