Vitrified blastocyst transfer cycles with the use of only vaginal progesterone replacement with Endometrin have inferior ongoing pregnancy rates: results from the planned interim analysis of a three-arm randomized controlled noninferiority trial

Vaginal progesterone replacement with Endometrin only for vitrified blastocyst transfer resulted in a one- third decrease in ongoing pregnancy compared with patients randomized to intramuscular progesterone alone or in combination with Endometrin.

Volume 109, Issue 2, Pages 266–275


Kate Devine, M.D., Kevin S. Richter, Ph.D., Eric A. Widra, M.D., Jeffrey L. McKeeby, M.D.



To assess the noninferiority of vaginal P (Endometrin) compared with daily intramuscular P for replacement in programmed vitrified-warmed blastocyst transfer cycles and to assess the noninferiority of vaginal P in combination with intramuscular progesterone every third day compared with daily intramuscular P.


Three-arm randomized controlled noninferiority study. To enable early recognition of inferiority if present, an a priori interim analysis was planned and completed once ongoing pregnancy data were available for 50% of the total enrollment goal. The results of this interim analysis are presented here.


Assisted reproduction technology practice.


Women undergoing transfer of nonbiopsied high quality vitrified-warmed blastocyst(s) in a programmed cycle.


Vitrified-warmed blastocyst transfer with mode of P replacement determined by randomization to either: (1) 50 mg daily intramuscular P only; (2) 200 mg twice daily vaginal Endometrin; or (3) 200 mg twice daily Endometrin plus 50 mg intramuscular P every 3rd day.

Main Outcome Measure(s)

Live birth. The primary outcome of this interim analysis was ongoing pregnancy.


A total of 645 cycles were randomly assigned to one of the three treatment arms, received at least one dose of P replacement therapy according to this assignment and underwent vitrified-warmed blastocyst transfer. These cycles were included in the intention-to-treat analysis. The study team, including the statistician, were blinded to the identity of the treatment arms, which were randomly labeled “A,” “B,” and “C” in the dataset. Ongoing pregnancy occurred in 50%, 47%, and 31% of cycles in arms A, B, and C respectively. Although arm C had an rate of positive hCG equivalent to the other two arms, the rate of pregnancy loss for arm C was significantly higher than for either of the two arms, resulting in a more than one-third lower rate of ongoing pregnancy. There were no statistically significant differences for any outcome tested between arms A and B. Results of a per-protocol analysis were nearly identical to those of the intention-to-treat analysis. On completion of these analyses, arm C was revealed to be the vaginal P only arm.


Relative to regimens inclusive of intramuscular P, vaginal-only P replacement for vitrified-warmed blastocyst transfer results in decreased ongoing pregnancy, due to increased miscarriage, and should be avoided. Randomization to the vaginal-only arm was terminated with these findings. This trial is ongoing to assess the noninferiority of the vaginal plus every 3rd day intramuscular P arm compared with daily intramuscular P in terms of live birth.

Clinical Trial Registration Number

NLM identifier NCT02254577.

Read the full text here.


Go to the profile of Micah J Hill
over 4 years ago

Thank you Dr. Devine and colleagues for a very interesting study. These data are very helpful and I'm looking forward to a full discussion at the next F&S journal club.!  When do you expect the final two arms to be unblinded with live birth outcomes available?  

Go to the profile of Kate Devine
over 4 years ago

Thank you Dr. Hill!  We too are looking forward to discussing our findings at the live F&S journal club on March 15.  The last delivery outcome for the study should be known at the beginning of April 2018, and we plan to analyze these data straight away once the outcome data are complete.  Hopefully we will be able to report our findings at the national ASRM meeting this year.

Go to the profile of Sherif Awadalla
over 4 years ago

Interesting article.  I think the issue with the vaginal progesterone group is that it was not given long enough and at enough dosage.  We transfer of the 6th day of progesterone and use endometrin tid.

Our live birth rates for 2014 and 2015 without the use of PGS are greater than the clinical pregnancy rates you report.

Age       LBR               N

<35         42.6%         357

35-37       37.4%        115

38-40       39.1%         46

I realize this is an interim report and maybe I did not read closely enough but I did not see data on mean age of patients or implantation.

Sherif G Awadalla, M. D.

Institute for Reproductive Health

Cincinnati, OH

Go to the profile of Kate Devine
over 4 years ago

Thank you very much for your comments, Dr. Awadalla.   

PGS was an exclusion for the current study, which likely accounts for some of the difference you describe between your population and ours.  

We fully acknowledge that the optimal duration of progesterone exposure (whether IM or vaginal) has not been established.  That said, in the current study, the group that received both vaginal and IM progesterone had the same duration of progesterone exposure as the group that received vaginal progesterone only. However, the group receiving vaginal and IM progesterone had a significantly higher ongoing pregnancy rate.  The only difference between the two protocols was the addition of IM progesterone once every third day.  

As for the ages of the patients they are found in Table 2 and were a mean age of 33.2-33.5y at time of vitrification.   

Please logon to discuss for the live Journal Club on Thursday evening March 15, 2018 at 7p.  


Kate Devine, MD


Washington DC

Go to the profile of Benjamin D Webman
about 4 years ago

Dr. Devine:  This is a very interesting study!  Have you seen any research about combining IM progesterone with estrace ?   

Also  newly available S.C. progesterone seems like an opportunity for a follow-up.  Only one company with FDA approval, so your phone should be ringing with a Pharma companies who wish to test gel vs IM noninferiority.