From SREI Fellows Retreat - Fellows vs Faculty Debate: Luteal Support in Programmed FET Cycles

Article: Intramuscular progesterone optimizes live birth from programmed frozen embryo transfer: a randomized clinical trial

Questions to be addressed:

1. How do routes of administration differ in progesterone levels in the serum and endometrium?
2. Does measuring luteal progesterone and adjusting progesterone administration impact outcomes?
3. Should all FET cycles utilize IM progesterone?

All right, welcome to Fertility and Sterility Journal Club Global, coming to you live from the 2024 National SREI Fellows Symposium here in Park City, Utah. I'm Rick Paulson, I'm Editor-in-Chief of Fertility and Sterility Reports, the open access journal of the fertility and sterility family of journals. And today's topic is going to be the administration of progesterone, specifically the question of whether intramuscular progesterone is needed in the context of medicated cycles of frozen embryo transfer.

And the springboard for today's discussion is a paper from 2021 entitled Intramuscular Progesterone Optimizes Live Birth from Programmed Frozen Embryo Transfer, a Randomized Clinical Trial by Drs. Devine, who's here, Richter, Jehandideh, Widra, and McKeeby. It was published in Fertility and Sterility in 2021, and it's a topic that doesn't seem to get old, which is why we're going to be debating it here.

We have a panel that is going to be on the pro side and a panel that is gonna be on the con side. And yes, these are three faculty members and these are three fellows, so we'll have a faculty versus fellows. So without further ado, I would like for each of the panelists to please introduce themselves, tell us who you are and where you are from.

Good morning, everyone. My name is Irene Su and I am at the University of California, San Diego. And I'm gonna be arguing on the affirmative that I am all the way, and we have drinks on the line for this.

Kenan Omurtag, WashU, our default is IM Prog, and yeah, I'm on the pro side. Good morning, I'm Tracy Harrison. I'm a physician at Kaiser Permanente, San Diego and an adjunct faculty at UCSD as well.

Also on the pro side, arguing for intramuscular progesterone. And on the side that is going to protect women's bottoms. Hi, I'm Hanna Kim.

I'm at Cleveland Clinic and we are going to be arguing that the route of progesterone does not have to be IM. And I'm Jacob Christ, I'm at UCSF and I'm team PV all the way. Obviously a great deal of enthusiasm.

All right, so here's how it's gonna go. We're going to alternate members from the two panels. Each panelist will have three to five minutes to state their case.

We will alternate pro and con going back and forth. Dr. Su will lead us off. Okay, so I'm gonna give a brief background to the study and start with whenever you think about a piece of evidence you get to think about, well, how good is it? And there are three ways to think about this.

How valid is the way it's designed? What were the results? And does it really apply to our patients? And because this study, I think, has a lot of validity in the way that it was designed. And because clearly IM in any version is better, then obviously the right answer is that you need some IM progesterone. So for background, there's about 400,000 ART cycles in the U.S. per year.

And of those cycles, if you look at the 2022 start data, more than 80% of these cycles will have something to freeze. That is why how we transfer in frozen embryo transfers is something that is of clinical importance to both us in practice and also to our patients. This is an RCT that's three-armed and designed in a non-inferiority fashion.

And so the interventions here were either IM progesterone daily at 50 milligrams per day, IM progesterone every three days with 200 milligrams of endometrin twice daily vaginally, or PV endometrin twice daily vaginally alone. The population are generally a pretty big population of IVF patients undergoing FET with BB embryos, 18 to 49. And there's some exclusion criteria that Dr. Harrison will talk about.

In the program FET cycles, all patients started with IM estradiol valerate every three days. And then once the lining was at seven millimeters or more, then they underwent whichever assigned group that they were in. The primary outcome is live birth.

So when you think about the validity of a study, you start with asking, are the patients kind of the same in all three groups? And you achieve this by randomization. They did this really well. They did this permuted, meaning you couldn't really tell who was gonna be in what group next.

And it was stratified by one of the three centers or one of the three laboratories that was a part of the shady group network. And so that allowed us to generate a similarly prognosed patient group in each of the three arms. So that was great.

You wanna know whether patients were blinded to the, and providers were blinded to the assigned group. And this happened because it was a study coordinator who knew which group next. And it wasn't the patient or the provider who thought, oh, IM is better, so I'm gonna put my better prognosis patient there.

They had very few loss to follow-up. Follow-up was complete. It was more than 91 to 95% in each group.

So you didn't really have differential dropout by group. So that was excellent. And they were able to account for this both in their intention to treat analysis and in their protocol analysis.

It was a nicely powered study. So they knew that at the start, they would expect standard of care to give you about 45% chance of live birth. In a non-inferiority study, when they say the lower margin is 10%, that means that in any of the other groups, if your live birth rate is less than 35%, you're starting at 45%, it's less than 35%, this is gonna be considered inferior.

That's a little bit wide. But you will see that the magnitude of difference in PV progesterone, which is not so good, was way beyond that. So even if they took a lower non-inferiority margin, made it narrower, we would have seen the same result.

And then finally, they did an interim analysis at 50% accrual. And that's really important because if your intervention is so crappy that at 50% in, you had to stop the trial for that arm, that means that there was significant harm to the patient and this is what happened. So briefly, in this study, there were 390, the goal was 390 per group.

And at baseline, on average, patients are about 33. On average, their BMI is about 26. And on average, 1.2 embryos were transferred.

And the take home point is the live birth rate in the IM progesterone group was 44%. In the IM progesterone plus PV group, 46%. And in the vaginal progesterone group, 27%.

And that did not matter, it was per protocol. Two thumbs down. This is not what you wanna do, is to do messy vaginal progesterone and then lose 15% in your live birth rate.

Who wants that? That's the end of my starting arguments. All right, thank you, Dr. Sue. I'm gonna use a timer so that I am ready for each of you.

Okay, all right. So I wanna give a little more background about progesterone in pregnancy and the different methods for replacement. So little is known about normal variation or ideal progesterone levels in the luteal phase of the menstrual cycle or in unassisted pregnancies given the pulsatile nature of progesterone release.

So a progesterone level greater than three is expected after ovulation, but this is considered more of a qualitative marker rather than quantitative. And some studies have shown progesterone levels maintain above 10 during unassisted pregnancies was associated with optimal outcomes. However, viable pregnancies have been seen with progesterone levels less than five.

So this variability makes it difficult to define cutoff levels during IVF. Progesterone replacement in FET cycles can be provided through a few different routes. So medications can either be IM, vaginal, oral, or subcutaneous.

Intramuscular progesterone has been used since the 1980s and is known to achieve high serum concentrations. However, absorption can be affected by BMI, and this route is notoriously cumbersome and painful for patients. Vaginal progesterone comes as vaginal inserts, tablets, gels, or capsules.

The vaginal route has been shown to provide higher bioavailability at the endometrium. And oral progesterone has low bioavailability, but there are some forms that have been shown to be equivalent to vaginal that aren't available in the U.S. Subcutaneous progesterone also exists, and some studies have shown it equally effective, but it's also not available in the U.S. at this time. So IM and PV have been used most commonly in IVF with adequate support of pregnancy, but there's mixed data on the optimal regimen or equivalence of the methods.

So while this was a well-done study, a major limitation we identified is the regimen of PV progesterone used. So like Dr. Su mentioned, they used 200 milligrams PVBID, and that choice seemed to be made from some data comparing 50 to 200 milligrams PV, both QD and BID. One of these studies was by Dr. Paulson looking at serum and endometrial concentrations at different doses, and Jacob will talk more about it.

But another of these studies was performed as a trial from Faring, who's the manufacturer, and it did show that lower doses did not result in secretory changes to the endometrium on endometrial biopsy, whereas at 200 milligrams BID, it did show that secretory changes occurred, but when you look closer at the data in their trial, it showed that secretory changes on biopsy were not achieved in 40% of patients on BID dosing. So this leaves patients at a significant risk of inadequate endometrial transformation. Use of 200 milligrams BID PV is generally lower or different than the dose used in other studies, or in places where it's considered the standard of care.

So Hannah will talk more about some of those studies. That's our major concern about the study, but there were a couple of other things that we wanted to mention. One is that patients had different exposure times from initiation of progesterone to the time of the transfer, and there are reasons behind that, which maybe we'll talk a little about, but about the effect on the endometrium being sooner in PV, but patients who used PV only had their transfer on the afternoon of day five, whereas patients who had IM only had it on the afternoon of day six, so it ended up being about 12 hours difference.

Additionally, patients were able to be included if they had three prior failed FET cycles or up to three prior clinical pregnancy losses following FETs, so this included patients with recurrent pregnancy loss and recurrent implantation failure, but we didn't see data on how that distribution looked between groups, and I think that's particularly important because pregnancy loss was the primary driver of the poor live birth rate in the PV group. It also seemed that patients in the PV-only arm had twice the dropout rate for administration error before the per-protocol analysis, so it did make us curious about how that was assessed, and Jacob again will talk about the issues with serum progesterone levels, but patients in the PV group were three times more likely to have progesterone levels less than three and had a lower mean level than described in studies using VIB dosing, so only 17% of them had a progesterone level above 10, so it made us wonder if there was possibly more administration errors or issues that weren't identified, and maybe there's less room for error using PV at the lower doses than with IM. And then lastly, patients were able to be included more than once, and so even if they had a prior cycle that was in the study, they were able to be included again but assigned to a different group, so I don't think we saw data on how many crossover events happened, and it didn't seem like the repeated measures were accounted for in the statistics, but they did mention a sub-analysis on unique patients was performed.

So overall, this is a great addition to trying to understand both the optimal and individual regimens of progesterone supplementation for programmed cycles, however the findings should not be overstated to insinuate there's no PV-only regimen without IM that can provide equivalent results. All right. I see we should have voted before we started doing the debate.

All right, on the pro side. Okay, so I'm going to talk about the patient survey data, and I think it's important. There's a lot of conversation about IM progesterone versus any vaginal progesterone because of the, you know, idea that the IM is more cumbersome, notoriously, quote-unquote, as was stated earlier, and what do the patients actually say, and the study actually included survey data about that, and the conclusions are what you might expect, but I think there's some practical things that we need to think about.

So just some things to share. They surveyed 742 post-transfer patients. The response rate was 75%.

Vaginal progesterone was considered more convenient to administer. Most patients using injectable progesterone reported that they needed or preferred assistance, and then few patients administering only vaginal suppositories wanted help. So in addition, patients reported that administering IM progesterone was significantly more stressful than vaginal administration, but in the end, interestingly, less than 8% of respondents in any treatment group noted any dissatisfaction with their medications.

However, the vaginal group was twice as more satisfied. Okay? So let me ask you a simple question. If I told you I could give you something that would be a little less, that will give you the best chance of success versus something that will have a lower chance of success, but will be more satisfying to administer, which one would you choose? Especially when you're investing so much mentally, physically, and financially in this journey, what has been your observation of what your patients have said? Have they said, yeah, I'll take the more satisfactory, messier thing that causes more discharge, and maybe more phone calls, although that was not described, I'll be fair.

Or do you pick the thing that you have the most experience with in your clinic, and that the data demonstrates is better than vaginal alone? Than vaginal alone? I think the patients have said clearly what they prefer. Now, again, they're more satisfied when you look empirically and ask them specific questions. It's, you know, when you look at it through that vacuum, that's what happens.

And to the chagrin of some, I'm up here saying, you know, what the data shows is, and kind of parlaying the experience of the provider. But I think the facts are, progesterone, IM alone is better than vaginal alone. And that is irrefutable.

The satisfaction that comes with vaginal administration is not enough to justify an inferior result. So what this study has done is said, well, what if you combine them to alleviate the burden? We acknowledge that. And there is a role for that.

But how complicated is that? How effective is that? Does everyone get, can everyone, does everyone have to pay out of pocket for endometrin or another vaginal suppository? What has been your experience with patients who have called with vaginal suppositories? They're great, but do they all have access to them too? That's another dimension of this that's important to understand. So again, there's a lot of evidence to support the satisfaction that comes with vaginal progesterone. But again, what does, what is the, how do you maximize the outcome? Thank you.

Okay, so first of all, to answer Dr. Omurtag's question, what happens if I could tell you that perhaps this dose and this regimen of vaginal progesterone did not work, but there are other ways to go about this without having to do the cumbersome IM shots, okay? So from Dr. Devine's study, we saw that there was a statistical significant difference in the birth rates of women and men between 200 milligrams twice daily of the endometrin vaginally and 50 milligrams daily of IM progesterone. But this contrasts with many of the previous studies in the literature. And there has been more research available into fresh embryo transfer cycles, which have shown no difference in outcomes based on the route of progesterone used for luteal support.

Currently, however, there is a paucity of data in terms of frozen embryo transfers, especially with blastocyst transfers. In the literature that is available, there is mixed data, but that is difficult to compare because there's not a uniform regimen of vaginal progesterone formulations that is currently being used. So in a retrospective study that was looking at 920 patients that was published by Dr. Shapiro in 2014, 90 milligrams BID of crinogel was compared with 50 milligrams of IM progesterone with transfer of blastocyst stage embryos.

In this study, there was no difference that was seen in the outcome. So implantation rates were 46.4% versus like 45.6%. Clinical pregnancy rates were 61.7% versus 60.5%. And birth rates were 91, sorry, 49.1% versus 48.9%. And all of these p-values showed non-significance. Interestingly, in this study, 13.3% of pregnancies ended in miscarriage in the IM progesterone group compared to 11.8% of the vaginal progesterone group.

So that's kind of in direct contrast to what we have seen in Dr. Devine's study. So in another retrospective study that was done looking at 1,710 cycles by Dr. Backenson in 2020, crino 90 milligrams BID where it's compared to 50 milligrams of IM progesterone. In this study, there were similar findings as was seen in Dr. Shapiro's study.

So they found no statistical significance in outcomes for frozen blastocyst transfers. So live birth rate, biochemical pregnancies, spontaneous abortion, clinical pregnancies were all noted to be similar. So additionally, other studies from China and Europe have also shown no difference in pregnancy rates for patients using vaginal progesterone gel compared to IM progesterone.

As postulated in the study by Dr. Backenson, some of the discrepancies from Dr. Devine's study might be attributed to the different pharmacokinetics of varying progesterone preparations. So endometrin has been shown to produce higher tissue concentrations, reach steady state more rapidly, but has more rapid clearance while vaginal progesterone gel seems to have longer duration of effect. So we postulate that this might be a reason that we see such mixed data in the literature.

So currently in Europe, vaginal preparations of progesterone are accepted as the standard of care. And as of now, there is no good consensus on the optimal regimen to maximize success rates for frozen embryo transfers. Ultimately, from this study, from Dr. Devine's study, we can conclude that 200 milligrams BID of endometrin may not be that optimal regimen.

However, I think we really should be careful to not generalize these results to all different regimens of progesterone supplementations, as there have been many studies that show non-inferiority. All right. Pro.

Okay, so I'd like to start, and it's important that you're hearing my... Oh, sorry. So some of these appropriately selected exclusion criteria include use of a gestational carrier, trophectoderm, biopsied embryos, vitrified oocytes, and cleavage-stage embryos. These exclusion criteria contributed to uniform study population with similar prognosis for statistical analysis, and did not compromise the generalizability, because there's low biologic plausibility that these criteria impact progesterone levels, absorption, endometrial receptivity, implantation, or early pregnancy development.

Because of those factors, we have improved validity of the statistical analysis, and still retain generalizability to cycles that include those exclusion criteria. And limitations to the generalizability do include the exclusion of obese patients with BMI greater than 38 in this study. We know that obese patients are at risk of having lower circulating levels of progesterone, both due to mechanical difficulties with administering IM injections due to high adiposity, and also due to higher volume of distribution for the medications.

However, it has not been definitively shown that the vaginal administration of progesterone compensates for this issue in obese patients. We do acknowledge the other very important limitation to generalizability with regard to the dose of progesterone that was selected. It is true that there are smaller randomized studies and cohort studies that do not have the same level of validity in terms of study design when compared to this study.

Therefore, you cannot generalize the superiority, sorry, you can generalize the superiority of IM progesterone to the majority of our patients that are going to be undergoing these frozen embryo transfer cycles. Thank you. All right, and now the final argument on the con side, and then we're gonna open it up to the audience for a general discussion.

So on a fundamental biologic and pharmacologic level, I would argue that vaginal progesterone should be adequate to support a pregnancy, and that serum progesterone levels may not appropriately reflect endometrial progesterone content. Going back to work from Dr. Paulson and colleagues in the 1990s, we know that while vaginal progesterone does have lower serum levels than IM progesterone, vaginal progesterone reaches endometrial progesterone concentrations that are 10-fold higher than IM progesterone. The phenomenon is known as the uterine first pass effect whereby the endometrium preferentially uptakes the vaginal progesterone, and you have this direct uptake from the vagina into the uterus.

And Dr. Paulson's work has shown that vaginal progesterone reaches steady state within 24 hours and has similar biologic effects on the secretory endometrium as IM progesterone. So from a biologic standpoint, vaginal progesterone should be adequate to support a pregnancy. The sustained trial does confirm earlier work showing that serum progesterone levels when comparing vaginal versus IM are different with vaginal having lower serum levels.

And while lower pregnancy rates and higher losses were seen with lower serum progesterone levels when checked on the day of HCG, this data conflicts with prior studies that have looked at the same outcome. Going back to 1999, Lightman and colleagues did a similar RCT where they compared vaginal progesterone to IM progesterone and that they found that while yes, vaginal progesterone did have lower serum concentrations than IM progesterone, there is no differences in clinical pregnancy rates or miscarriage rates. So that data suggests that yes, vaginal progesterone has lower serum concentrations, but that lower serum concentration doesn't necessarily translate to worse clinical outcomes.

Of note in that study, they used TID dosing of 200 milligrams of permetrium and they achieved serum progesterone levels of 15 nanograms per deciliter. In Dr. Paulson's work using 200 milligrams of vaginal progesterone BID, they had maximum serum concentrations of 13 nanograms per deciliter, whereas in Dr. Devine's study, the average serum progesterone was only 7.4 nanograms per deciliter in the vaginal progesterone arm. So even if we can't say that there's a direct association between the serum and the endometrial content, it does suggest that there's some difference going on between these studies.

Perhaps one, the BID dosing of vaginal progesterone was insufficient, or two, as Kerry mentioned, there may have been further issues with compliance in the vaginal progesterone arm that weren't fully accounted for. Also, it's worth noting that while 7% of the vaginal progesterone cycles had serum progesterone levels that were less than three, two to 3% of the IM arms also had serum progesterone levels that were less than three. So it's not like giving IM progesterone guarantees that you're gonna have higher serum levels.

Outside of the discrepancies between serum and endometrial progesterone concentrations, the clinical utility of serum progesterone to predict pregnancy outcomes is further limited by variations in assays. Most studies use a radioimmunoassay to assess progesterone levels, and those radioimmunoassays are not as accurate at lower serum concentrations of progesterone. Progesterone levels at varying times in relation to when progesterone is given.

So because of these differences between studies, there's no universally agreed upon threshold for serum progesterone to predict pregnancy outcomes. So overall, I think we can say that yes, vaginal progesterone in general is gonna have lower serum progesterone concentrations than IM progesterone, but I don't think we can say that a specific serum threshold adequately reflects endometrial content or universally predicts pregnancy outcomes, and that from just a fundamental pharmacologic and biologic basis, vaginal progesterone, when given at adequate doses, should be enough to support a pregnancy. All right.

Thank you. I was gonna give you extra time. If you quote one of my papers, you get extra time.

Are there questions from the audience for our panel? So I may have some questions prepared. Does anybody know how much progesterone is produced, come on, up to one of the microphones. Does anybody know what the production of progesterone is by the corpus luteum in the luteal phase? So we're reproductive endocrinologists.

It would be good, and we know we're giving people so much progesterone. What does the body make? What does a corpus luteum make? 25 milligrams a day. 25 milligrams a day.

I'll answer that question. I've prepared myself because I Googled it before I came up with this. So 25 milligrams is what the body makes, and we're giving him, let's see, 50 milligrams IM in one arm, and then we're giving him 200 milligrams vaginally BID.

That's 400 milligrams. So what we're saying is out of the 400 milligrams, if, what percentage would 25 be out of that one? 16. So if 8.25% of that absorbs into the circulation, that'll imitate what's happening by the corpus luteum.

I would also point out that the world's first IVF pregnancy, first egg donation pregnancy, was born in a cycle where the recipient received 50 milligrams BID vaginally. That was all the progesterone that was given at that time. Question.

I was gonna just start by saying this was a great discussion on both sides. Not to be a pain in the butt, I had a couple comments that I just wanted to share with the group. So Hannah had mentioned that there was a study that had shown there was no difference between live birth rates with people that had had vaginal progesterone versus IM progesterone in fresh transfer cycles.

Kind of to Dr. Paulson's point, you have a corpus luteum in those cycles. So I think to, while that data is interesting, I think to extrapolate it to the patients that are undergoing a programmed FET cycle, I don't think we can quite make that leap of faith. I do think that the pharmacokinetics of vaginal progesterone with looking at, it'd be interesting to see if the study had been repeated using TID dosing if we would have seen different results.

And the other thing that I'm just gonna bring up to the group, because I think it's interesting, we looked at the data at the difference in cost in IM progesterone versus endometrin in our clinic. We typically will do 12 weeks of progesterone. We're super mean for our programmed FET cycles at Penn, but if you look at what that costs for the 70 injections, it ends up being about $50 for IM progesterone versus about $3,500 for endometrin.

So it ends up being over 70 times more expensive for patients. They're obviously already undergoing something that's very expensive, but I think that's at least aside to the pro-IM progesterone group that you guys could have drilled down on, so you're welcome. That might take the sting out of the injection.

Any comments from the panel? I don't think the data Hannah shared was about fresh cycles. I think it was just a comment towards that there's more data on fresh cycles, but the studies that you shared were about the frozen cycles, I guess. Before Dr. Devine comments on her own paper, we have a question over here.

Thank you for a very interesting talk. I had a quick question about generalizability. If both groups might talk about the estrogen regimen that was chosen and how that might perhaps affect the applicability to other regimens that might be used, thank you.

Anyone wanna pick that up? Yep. So as you all know, there are various ways, yeah, I think we're on, of replacing estradiol, whether it's orally, transdermally, or IM. But the nice thing about an RCT, as opposed to lots of retrospective cohort studies that in which you can't adjust for potential prognostic factors that you have not accounted for, is that it's actually the same across all three, right? And so then, therefore, the difference that you're seeing in the live birth rate can be attributed to the progesterone, and it doesn't matter whether it was IM, or vaginal, or transdermal, because it was the same across all three groups.

Comments over here? Estrogen, because you were happy with the estradiol regimen. And I think you might have mentioned already that patients that didn't achieve adequate endometrial thickness or required additional vaginal, or other routes of estrogen, or required greater than 20 days for endometrial preparation were not included in the study. So that did account for that.

I would say, because we were talking about generalizability, I think it does reduce the generalizability because we can't say that this regimen is gonna apply necessarily to other routes of estradiol administration, which are more commonly used in most IVF practices, but it does allow for uniformity in the study, which I agree is useful. Dr. DeVine. So just picking up from that last point on why did we use intramuscular estradiol, because I really didn't want to, is that it was historical, that essentially at our practice, the founding partners felt strongly about using intramuscular estradiol, and so we didn't wanna switch mid-study when I'd finally convinced them that it wasn't necessary, but we don't do that anymore.

And our findings in terms of pregnancy rates are fairly robust. I wanted to congratulate the fellows especially for a really excellent presentation, but both sides, I would say that you guys had the harder side of that debate, although you do have a, you do have a very sympathetic moderator. So what I would say is that everybody is right here.

So I think that what this paper showed is that 200 milligrams endometrin BID was inferior to both of the intramuscular containing regimens. Can the paper say that there is no vaginal regimen that could provide adequate success from frozen embryo transfer? Absolutely not. Can you do a 16-arm RCT where you compare every single protocol? Unfortunately not.

So there's a PLOS ONE paper that showed dihydrogestone orally plus PV was equivalent. Of course, it used cleavage stage embryos and women in their late 20s with a BMI of 21, so it's hard to extrapolate to a US population, but I would love to know whether that regimen would work equivalently well in our system. We know from all of Labarta's work that while vaginal progesterone can sometimes be inferior, especially in the case of lower serum progesterone concentrations, those cycles seem to be able to be rescued with sub-Q.

And we're working on, Dr. Paulson doesn't like that. So, but that said, we're working on getting a trial to get sub-Q approved in the US. I will say it delayed our IRB approval by about five months just to get the IRB to approve using 200BID, because that was above any FDA-approved dose in the US.

So we had to call, they called in two different experts, even though this was a commercial IRB, to comment specifically on the safety of that high of a dose, which I, at the time, still felt was absurd since they use up to 800 in Europe. But I don't think we would have been able to convince them to go too much higher than that. And again, we chose the 200BID based on steady state and AUC data and wanting to optimize patient compliance.

That said, I think you guys did an amazing job, and I hope that you all do RCTs to test some of these other regimens. I designed and started this study as a fellow, so you can too. All right, well, you brought up the topic of measuring serum progesterone levels in the luteal phase of these artificial cycles and whether one should act on that or not.

And so I'm gonna ask that question of the audience. But first, let me ask the question, does anyone's program use vaginal progesterone only in a medicated FET or egg donation cycle? So everyone uses some degree, you use vaginal alone? I'm putting one finger up, it's one faculty member. Okay, so that means that, so, I mean, Kate, this is the, yeah, so this is your accomplishment.

I mean, everybody is using IM progesterone. Everybody has bought into that. So the Labarta studies look at serum progesterone levels at the time of embryo transfer.

Does anybody's program measure serum progesterone levels at the time of embryo transfer and then adjust the dosing? They do not, you do, okay. Not routinely, well, let's pick it up from here. Does that, is that a thing? Should we be doing that? We measure them the day before the embryo transfer and then again, two days after the embryo transfer.

And we do give additional supplementation if it falls below our threshold. What's the threshold? It's 18. 18 nanograms per milliliter.

It's true, folks. So the patients are bathed in progesterone. You know, you'd be surprised, I feel like, or, I mean, even, so we, interestingly also, we don't do PV only, but we, if patients can't tolerate the IM, we do PVBID with prometrium POTID, although we know that has low bioavailability and I was trying to find where that data came from.

I talked to Dr. Devine about it and I wonder if it's from the study, they're looking at the other formula of progesterone that I'm going to mispronounce, didrogesterone, because they looked at that oral dosing with endometrin and found equivalents in one study and so sometimes we'll do the POPV regimen and patients do tend to have levels over 18 with both, but we'll adjust it if not. So 18, 18, how about, Kenan, you said you sometimes measure in? 20, 20 nanograms, question? No? I'm sorry, you are? 15, you cut off as 15 nanograms per mil. Okay, so these are pretty high, pretty high levels.

Yes, please, Paula. Why are you laughing? Okay, question. I was surprised, I was surprised.

You appeared seemingly out of nowhere, but there you go. I'm curious, for the programs that use IAM progesterone, which we use as well, for 12 weeks, do you ever transition in program cycles to vaginal and if so, when? Anybody on the panel? At our program at UCSD and affiliated private practices, we also use IAM progesterone as the primary mode and transition to vaginal progesterone after nine weeks. Vaginal only at nine weeks.

And between transfer, is it daily IAM progesterone or do you do the every three days? Daily IAM progesterone, we use 100 milligrams. 100 milligrams daily and no vaginal? Not initially. Not initially, and then you transition from one to the other, okay.

Interesting, other? So Labarta studies found an association, as you're coming up, you can correct because I'm not sure that I remember them very carefully, but I flinched when you said that they were able to rescue these cycles. I take issue with the study design. It's not a great study design, for sure, because they used a historical control.

So they looked at, they did a series of studies, all either retrospective or prospective observational, and they looked at the serum progesterone concentrations following administration of vaginal progesterone in program cycles and correlating them with pregnancy outcomes and arrived upon a threshold of about nine nanograms per ml as associated with optimal outcomes. And then they went back and started adding sub-Q to those that fell below nine and compared them historically to those who had had above nine. So the control group is not optimal.

That said, there are a number of investigators that have found that low serum progesterone concentrations in vaginal cycles are associated with worse outcomes. We did look at serum progesterone concentrations. This was a pragmatic trial.

We did not time it precisely relative to when the last dose was, and it was at the time of the pregnancy, and so if these patients are not pregnant, that may also contribute to what their serum progesterone concentrations are. And so I do think that the measurement of serum progesterone concentrations in program cycles in the U.S. has gotten totally out of control and is probably completely unnecessary, especially in cycles where IM is administered. You know, there was really no association in our study between the serum concentration in IM-containing cycles and the lower serum progesterone concentrations in the PV arm was not enough to explain the difference.

And so we analyzed whether that effect compensated and it did not, and we presented those data as well. That said, I think it's a question that we should be answering because we shouldn't be bringing patients in to measure something that shouldn't be acted upon. And so if serum progesterone concentration in program cycles where patients are getting IM is not necessary, and then whatever else you do, if you double their dose and potentially increase their pain and discomfort, that's a bad thing.

So this is an easy question to answer and one that we definitely should be answering. One more thing that I just wanted to comment upon is it was said sort of as a blanket statement. There's higher endometrial concentrations in PV cycles.

It should be enough. But there's no data to say that that's true, right? So we don't know that only endometrial concentrations matter. We know that, yes, there's this first uterine pass effect, but there could be metabolites of serum progesterone that are important in early pregnancy, and there could be changes to the lower reproductive tract that change absorption of vaginal progesterone once someone becomes pregnant.

And there could be immunomodulatory effects of systemic progesterone that are important in early pregnancy. Kate, thanks for that. Dr. Bergen? Oh, I had a couple of sort of adjunct thoughts to that with respect to this paper and measuring serum progesterone.

So when you think about sort of measurements, what are you doing it for, right? One is to predict the outcome, and the second is potentially because you're gonna act on it. And exactly to Dr. Devine's point, there really has been no trial to say that if you acted on it by giving more oral, more vaginal, more IM, whether that's gonna affect your outcome. And then in terms of serum progesterone as a biomarker, in this study and others, kind of what you've heard a little bit already, what you want is some threshold that's gonna predict no baby or no outcome that you want.

And if you kind of look at this, every serum level, even when your serum level was less than one nanogram per milliliter of progesterone in this study, you still had babies. So what are you measuring it for if you're not going to change your intervention or if there's no demonstration that an additional intervention will help and if it doesn't predict the outcome that you want? So we agree that the Labarta study where they added additional progesterone after they measured a low serum level and then they had similar outcomes to good outcomes that we did not think that's a good study design or that that shows that the intervention made a difference. Yes, they could have given them a green M&M in place of the other progesterone.

If you don't control for that other group, yes. So that is why you saw me cringing previously. Sorry, Dr. Bergin.

Well, I don't think that the serum progesterone level that we don't have an optimal level or know how to adjust dosing necessarily, that the verdict is maybe still out on some of that. I do wonder if it can at least help us identify patients who are having administration issues or absorption issues. And so maybe there is a subset of patients that may benefit from having the levels checked sooner and being able to supplement or to change route or something.

And maybe you shouldn't do this for everyone just because there's a subset of people, but I do wonder if we should still find who those people are that would benefit from it, maybe. Yeah, well, so, but Dr. Christ, you said that you've mentioned all of those progesterone serum levels after vaginal administration. Seems like they were a lot lower than, what are we hearing? 12, 15, 18, 20.

Well, I think that those levels are very high that people are using as a guideline. There's a 2021 meta-analysis looking at vaginal progesterone, and those studies that use thresholds less than 10 did have significantly lower pregnancy rates. So I think using maybe 10 could be a marker, but I think certainly higher than that.

I don't think that there's good data to necessarily support that, especially if we're all using IM. The data for IM and serum levels is even worse than for vaginal, so. Christ.

Yes. I'd like to ask the audience a question. I want you to pretend you are the medical director of the IVF clinic that you work for, and you are posed with the question of what will your default programmed, no, your just default programmed FET progesterone regimen will be.

And you've done your due diligence. You've gotten feedback from the nurses. You've obviously consumed this conversation, the scholarship associated with it.

You've polled your colleagues who like to do whatever they want. You've polled your nurses who want everything but chaos. And you're trying to do what's best for the patients while balancing what's best for the division's interests at large and minimizing chaos.

So who would make the default, let me see a show of hands. We're gonna force you guys to make a commitment here. Who would pick IM only as the default regimen for programmed FETs in your practice? Show of hands.

This is IM only as opposed to IM plus vaginal, is that? Correct. Okay, IM only. Okay.

Who would pick IM plus PBS as the default? Okay, so that is an exercise that you will be confronted with in due time. So recognize that these decisions are not made in a vacuum, that you are not only balancing the mental, physical, and financial stress of your patients, but you're gonna have to be balancing the mental and physical and stress of your team and everything that goes with it. So that is what is ahead of you and that is very exciting because you guys are gonna have the tools or will be given the tools equipped to tackle that.

All right, well, you kind of answered the question as to whether all FET cycles should utilize IM progesterone. I think we had a very good show of hands as to what everyone would choose. So maybe this would be a good time for us.

First of all, are there other questions from the audience? Because then I'll have the panel wrap up. So now each panel member is going to have one, oh, did I see it? I missed it. Oh, sorry.

Can you come up to one of the microphones? Because we're recording this for posterity. Thank you all for a great discussion. I'm a fellow at a practice where the default protocol is IM injection only.

But my question to the panel is what about those patients who cannot tolerate IM injection? What should be my take home message for a protocol? And natural cycle is not an option for them. So what should be my regimen? And if I should measure serum progesterone, what is the cutoff and what adjustment do I do? Excellent question. Okay, so who wants to pick up the question? So there you go, they cannot tolerate.

Let's say an allergy and is amenorrheic, so it cannot have a natural cycle. Okay, so it's your patient, they're doing an FET for the first time, you've already done the consult, you've told them about the, you've set expectations of what the progesterone IM is gonna be, and then they're getting ready, they're getting a little bit cold feet, the nurse says, hey, they're not comfortable with this, they're scared, they want an alternative. So I think the data here presents an alternative that is IM plus PVS.

So the take-home message, I think, from this conversation is that's the alternative that you can comfortably present. She cannot tolerate IM. Oh, she cannot tolerate IM, okay.

She is deathly allergic to every form of oil that progesterone comes in. Missed that. I mean, practically speaking, what I do, and I think what most people will find themselves doing is you're gonna rely on the gel data that was talked about here as your default.

You're gonna say, look, I understand that you can't tolerate this. This is what the data demonstrates. There is a viable alternative, and we can offer that to you.

However, it's gonna come at probably some considerable out-of-pocket cost. Are you willing to confront that? And they're gonna say, you know what, I'll make it work with IM, or they're gonna say absolutely not. That is, I think if you poll anyone who's been in practice for five years or more, that would be the experience.

If any faculty wanna co-sign that comment, that would be helpful, please. But that's been my experience. Also the option of ovulation induction.

Right. So gonadotropins or letrozole. Well, if they're POF, then they're not gonna ovulate.

I would say, so unless they were, as Dr. Paulson just alluded, POF, and have a deathly allergy to every vehicle that intramuscular progesterone could possibly come in, then yeah, then you have to use PV only. And at that point, I would do 800 milligrams a day vaginally, and I would check the serum concentration, although the data are not perfect. I think in that situation, it might matter, especially if they only have one embryo to work with.

But it doesn't come up very often that there's someone that you can't stem to get a corpus luteum, and they can't tolerate even once every third day in some vehicle. So the topic of today was not the issue of using natural cycles for FET, but if you do natural cycles, do you supplement them with additional progesterone? You give them IM progesterone after that? No, you just let them be natural? You give them. You make them ovulate in a totally normal cycle, that they could have gotten pregnant, and you say, we're gonna give you a little extra.

Question. Thank you, everybody, for a wonderful discussion. I had a follow-up question from Dr. Devine's point of giving 800 a day.

Would we be able to use a less expensive alternative like prometrium instead of endometrium in such a case, and would we think that it would have similar clinical effects? I have one number for you. 25 milligrams. Corpus luteum makes 25 milligrams, okay? You're gonna give somebody, let me see if I understand this, 800 milligrams vaginally, and you're concerned about how much of it is gonna be absorbed.

You need to absorb 25 milligrams of it. I can't even do, what is the math? Like 5% or something like that? So my point is that we're, my own feeling is we're hugely, hugely, hugely overdosing our patients with progesterone, which is perfectly okay, because progesterone probably has saturation kinetics. Once you get above that saturation, who cares? So you can give them extra, extra, extra.

We would certainly absolutely use prometrium, micronized vaginal progesterone. Micronized progesterone is designed for oral use, also absorbs vaginally very nicely. The capsule melts, and then the powder absorbs that way.

You can use endometrion. The gel, you can't use that much. The gel, right, Shapiro, that was 90 BID, that's 180 milligrams that was given vaginally.

They found no difference between IM and that dose. So there's lots and lots of other data that vaginal is more than enough, but we're all stuck, I think, using IM. We do it, too.

I am not holier than thou. Thank you. All right, are we ready for a wrap-up? Okay, let's have a wrap-up.

All right, we're gonna alternate one minute each from the two sides. Maybe we're gonna let the con side. No, we'll let the pro side go first, because then the con side has the last word.

All right, one minute. A little bit of a repeat of yesterday. Choose the choice that gives you the most choices, and starting with IM gives you the choice of either doing IM alone or IM with PB supplementation.

And I think in this case, we can all ponder about biological plausibility, that permethrin at this dose should be the same. 800 milligrams at this dose should be better or the same. And the thing is, in short of high-quality data, you're really subjecting patients to a data-free space, and at least looking at the study, if you are 15% to 18% less live birth rate, I am not sure that it is conscionable to offer vaginal progesterone, except in very specific, rare circumstances.

Ouch, okay. Okay, overall, I think it's safe to say that as of now, what we know and what we're used to is IM progesterone, and we can always say, why change something when it's working, right? However, it's clear that other formulations could work, and the patients have clearly spoken in this study that they prefer vaginal formulations. So really, I guess this is just an urge for everybody to be careful to not generalize these results to all different regimens of progesterone, and I guess this is a call for future high-quality research to look more into this question.

Pro, so, Dr. Suh, you think that all these Europeans that are using only vaginal is unconscionable? Dr. Paul, the average live birth rate in Europe is significantly lower than that of the U.S. That it is, that is true. All right. And that is unconscionable.

I think we have to be pragmatic, practical, and then, again, setting appropriate expectations about what the experience is with the products that we give patients, with the spirit of what we think is in their best interest. There's a lot of variation. It's important to understand what's your default, and the default is clearly IM with options.

But it's appropriate to be cautious about vaginal supplementation and complementation with your IM product, and again, vaginal alone in the circumstances that are more niche, and I think there's more to be said about the role of vaginal progesterone as we move forward. I did wanna, no, that's all I'll say. Thank you.

Con. One closing statement. I may give another one on your behalf, but go ahead.

It's not my role. So I just wanted to summarize all of our pro comments so far that this is a very well-designed study with appropriately selected patient population that is generalizable to the vast majority of our patients, exclusive of those who have catastrophic allergies and no ovaries. And so in that circumstance, I think that we've demonstrated that for cost reasons, efficacy reasons, IM progesterone-containing regimens are superior to vaginal only.

Okay, we can't let you show us up, actually. We'll wing it. I think we have to really use our brains and think about the physiology here and think about what makes sense as well as look at the data.

And so I think we have gone over some of the ways that these medications are absorbed and the way that they affect the endometrium and whatnot and how they're processed. And so I think that we really need to be thoughtful about that when creating new regimens and thinking about what makes sense for our patients. And yeah, we really owe it to our patients to make sure that we're giving them IM injections because that is truly what is best for them.

And I'll add one more. I think I agree, this is a fantastic study. It's very generalizable, but I think we do need to also remember that these randomized trials, these large trials, they kind of regress everything to the mean, so there are gonna be subpopulations that potentially would benefit from a vaginal or do better with vaginal.

And so when considering how these apply to your patients, there may be times where vaginal could be better. And it is, given the biologic plausibility, warrants further research. All right, well, as much as I wanna wrap it up, I just have to put on my hat as the professor.

So generally, when you talk about randomized clinical trials, we all search for them, we try to do them, we admire people that do them, we really base a lot of our decisions on the basis of that kind of good data. But when you do a randomized clinical trial, the only thing that it does is it prevents you from being biased. It doesn't prevent nature from being biased or from random chance to present you with a biased sample that you could not control for, because you randomized it.

You did it, you did everything you possibly could. There's a number of examples in medical history where an RCT was performed where the control group had a suddenly inferior outcome to what you would expect. And the study group, which would have had the same kind of outcome like everyone else, suddenly seemed that much better.

And there's a great story out of the MFM literature. Some of you were reproductive endocrinologists. You guys have heard about 17-hydroxyprogesterone caproate.

250 milligrams weekly was given to patients to see if it prevented preterm labor. There was a huge RCT. The huge RCT showed an improvement in the group that took the 17-hydroxyprogesterone caproate because the control group had an unusually high rate of preterm labor.

And 250 milligrams weekly, okay? If they had Googled how much progesterone is produced by the placenta at term, they would have said, oh, it's 250 milligrams daily. We're adding 250 milligrams weekly, so I'm adding one-seventh of the additional progesterone. They might have questioned the physiological plausibility of that, but they went through with it, and because the difference was statistically different, 17-hydroxyprogesterone caproate was approved by the FDA, but the FDA said, but you have to do a follow-up study.

So they did a follow-up study, took another seven or eight years, showed no difference, and it was withdrawn. The problem here was this particular RCT was very well done. The miscarriage rate in the control group is out of sight.

It's 30% or 35%. We used vaginal progesterone for a very long time in our practice. We never saw miscarriage rates that high.

That is very, very unusual. The problem was that after the interim analysis, Kate, you guys decided not to enroll any more patients into the vaginal group anymore. And so that second trial that was done, that disproved 17-hydroxyprogesterone caproate would never have been done had they said, this group is so inferior, we have to throw that out.

So you can't, there's other examples where this occurs. When you see an RCT that shows statistical significance, look at the control group and ask yourself, is this my experience? Is this what I normally see in the control group? If it isn't, and something is unusual about it, just be wary, the data might be great. Maybe they need to be repeated, or maybe they need to be verified.

Why did you stop enrolling patients into that? I'm here to answer you. So, you know, first of all, it was part, and it should always be, part of the a priori statistical analysis plan to do this interim analysis. To look at the data is to change the data.

You cannot look at your data unless you have pre-specified it. And we wanted to, because actually, the reason that we did this study was twofold. One, we wanted to provide better patient experience, and we were hoping that vaginal progesterone would be sufficient.

Two, we had looked at thousands and thousands of cycles retrospectively, and so had many of our colleagues, and seen much higher biochemical pregnancy loss rates with vaginal only. And so that was something that was kind of buzzing around, is this observation that people are having gonna bear out if we look at it systematically? Lastly, when we looked and designed the interim analysis, in order to avoid spurious conclusions of inferiority, we set our p-value at .015. The magnitude of difference was also massive. So yes, as you just eloquently described, the magnitude of difference is not really as substantive as the probability that you have erroneously concluded a difference that's not there.

But the chances of that in this study was 1.5%. Okay, thank you very much. Thank you all for coming today. I'm glad to be here.

I just, I wonder if we should let the audience vote on who won. If you had to pick who won. We're gonna vote, who had won.

I don't think this is a slightly biased question. And I believe the deal was that all of the attendings would buy all of the fellows a round of drinks if we won. I don't know, that's what I heard.

All right, who votes for team pro? Oh. I don't know, I think it's gonna be close. Who votes for team con? Thank you for your enthusiasm, for your participation and fascination.

And thanks to all of our faculty and fellows and our panel members and everyone for participating. Thank you.