The literature certainly has favored the view that vaginal and IM progesterone is equivalent in their results for luteal phase support. Yet some of the most successful American programs still use IM progesterone, and in fact, most American women prefer the injections, despite the pain, to vaginal creams or pills or gels that have to be inserted two to three times per day. So what do the studies tell us? First i will go over the papers favoring equivalence of IM and vaginal progesterone. Then I will go over the studies that show the opposite. Finally I will try to interpret the reasons for these conflicting results, and why some of the most prestigious high quality clinics in America, quite to the surprise of the rest of the world, still prefer IM progesterone.
Reviewing the studies on vaginal vs. IM progesterone, Berger and Phillips wrote recently in the journal of assisted reproduction and genetics ((2012) 29:237-242, the results with pregnancy outcomes in oocyte donation recipients using vaginal gel versus intramuscular progesterone replacement, and found no statistically significant difference in live pregnancy outcomes. Silberberg, similarly in 2012 in fertility and sterility (2012) 97:344-348, compared the results with vaginal Crinone gel vs. intramuscular progesterone in oil in a large prospective trial, and actually reported a higher pregnancy rate with vaginal progesterone (70 per cent compared to 64 per cent) and the same with live birth ( 52 per cent vs. 45 per cent). Such high pregnancy rates however cause some concern about selection criteria for that study. Nonetheless, both these recent papers support the original paper of Gibbons in fertility and sterility (1998) 69:96-101, which showed that vaginal progesterone gel was as effective as IM progesterone in producing clinical and ongoing pregnancies in their donor egg program.
Doody et al in 2009 compared endometrin to progesterone gel and found no difference, but this was in a fresh transfer, non-donor group, and it was just a comparison of one vaginal progesterone brand to another (Fertil Steril 2009, 91: 1012-1017). The meta-analysis of Zarutskie and Phillips in the same year concluded in general that "vaginal P is comparable to administration of IM P for luteal phase support" (Fertil Steril 2009, 92:163-169). Teva has just come out with another vaginal progesterone product, a progesterone vaginal ring that can be placed around the cervix. Perloe et al argued that this new ring is as reliable as Crinone, but Teva has never done a study comparing their new ring to IM progesterone. This weekly ring is supposed to be more convenient than using vaginal gels 2 to 3 times per day.
Yet several good quality studies show IM progesterone to still be the superior "gold standard" for luteal support. As early as 1999, Chantilis et al found the pregnancy rate with Crinone to be similar to IM progesterone, but the pregnancy loss was much higher with Crinone than with IM progesterone (Fertil Steril 1999; 72: 823-829). Just this year, Feinberg et al reported that in frozen cycles, adding IM progesterone to endometriun significantly improved results (Fertil Steril 2013; 99:174-178). Similarly at the end of 2012, Kaser et al also in frozen cycles, observed significantly higher pregnancy rates with IM progesterone over Crinone, 49 per cent to 44 per cent (Fertil Steril 2012;98: 1464-1469). Furthermore the largest study to date, presented by McKeeby et al from the Shady Grove Clinic in Maryland, to the New England Fertility Society in April 2013, reviewing 4,536 frozen embryo cycles, using sophisticated multiple logistic regression analysis for all factors, showed much lower pregnancy rates and live birth rates with vaginal compared to IM progesterone. Shady Grove actually apologized, saying this was not the outcome they had hoped for, but nonetheless, it was the fact. The live birth rate for slow freeze was 25 per cent compared to 17 per cent, and for vitrified embryos was 45 per cent compared to 36 per cent. So they concluded that they simply have to use IM progesterone over vaginal progesterone even though that was not their wish. This latest huge Shady Grove study was very damning to the cause of vaginal progesterone.
So how do we explain these conflicting studies with those who favor vaginal compared to those who favor IM progesterone for luteal support? My best guess is that if you get enough progesterone absorbed vaginally, it can be comparable to IM progesterone. But just like androgen skin gel replacement in men, you can get highly variable absorption rates and highly variable quality of patient application of these gels. So in some studies vaginal gels will not be as good as IM progesterone, and in other studies, vagina gels will be equivalent to IM progesterone. For certain, any clinics which report incidents of the patient having a period or some spotting before their negative pregnancy test, would have to admit their luteal support is not adequate, and that never happens with IM progesterone, but often happens with vaginal progesterone. So many of the largest and most respected IVF centers in the United States prefer the security of IM progesterone over the variability with vaginal progesterone. That is my best guess as to why some studies show equivalence and many clinics prefer vaginal progesterone, while other very good clinics in the U.S. prefer IM progesterone.
Dr Sherman SilberInfertility Center of St Louis
St Louis, Missouri 63017
This email address is being protected from spambots. You need JavaScript enabled to view it.