Fertility and Sterility On Air - TOC: May 2025

Take a sneak peak at this month's Fertility & Sterility! Articles discussed this month are:  

4:08 Classification system of human ovarian follicle morphology: recommendations of the National Institute of Child Health and Human Development - sponsored ovarian nomenclature workshop 

12:32 Impact of Prednisone on Vasectomy Reversal Outcomes (iPRED Study): Results from a Randomized, Controlled Clinical Trial

21:38 Triggering oocyte maturation in IVF treatment in normal responders: a systematic review and network
meta-analysis

33:57 Parental Balanced Translocation Carriers do not have Decreased Usable Blastulation Rates or Live Birth Rates Compared to Infertile Controls 

45:28 A re-look at the relevance of TSH and thyroid autoimmunity for pregnancy outcomes: Analyses of RCT data from PPCOS II and AMIGOS

View Fertility and Sterility May 2025, Volume 123, Issue 5: 
https://www.fertstert.org/issue/S0015-0282(25)X0004-2

View Fertility and Sterility at https://www.fertstert.org/

Welcome to Fertility and Sterility On Air, the podcast where you can stay current on the latest global research in the field of reproductive medicine. This podcast brings you an overview of this month's journal, in-depth discussion with authors and other special features. F&S On Air is brought to you by Fertility and Sterility family of journals in conjunction with the American Society for Reproductive Medicine, and is hosted by Dr. Kurt Barnhart, Editor-in-Chief, Dr. Eve Feinberg, Editorial Editor, Dr. Micah Hill, Media Editor, and Dr. Pietro Bortoletto, Interactive Associate-in-Chief.

Back to another episode of Fertility and Sterility On Air. We are in May 2025, volume 123, number five. I'm Micah Hill, the Media Editor, and I'm joined today by Kate Devine, one of our Associate Editors and Podcast Host.

Good morning, Kate. Good morning, Micah. How are you doing this morning? I am doing wonderful.

I am excited for my favorite day of the week, month, when we get to talk about the science and F&S. And we are joined by Kurt Barnhart, our Fearless Editor-in-Chief. Good morning, Kurt.

Good morning. Fearless, that's an adjective for an Editor-in-Chief I didn't know about. But anyway, I'm enjoying the tulips and everything blooming, so spring is here, and so is our podcast.

Outstanding. We're going to dive right in because we've got some great content to cover today. The views and reviews this month is from Editorial Editor Laura Rienzi on the goal of rescuing immature oocytes using IVM.

This article has some of the leading researchers in doing IVM currently, and they're looking at can we apply this technology as it emerges to rescuing eggs that are immature during conventional IVF. It's a great article. It dives into the science.

I really encourage you to read that. Kate, you actually have the inkling this month on progesterone and FETs. Can you give us a high-level view of why you wrote this inkling with Allison Eubanks, our Fellowship Editor this year? Yeah.

So, you know, this is obviously a topic that's near and dear to me, and much of the research that I'm interested in conducting is looking at how to optimize frozen embryo transfer outcomes. And gosh, it's so long ago now, but people are still talking about and trying to work around, I would say, the intramuscular progesterone versus vaginal progesterone randomized controlled trial, which is what we call the SUSTAIN study. And while I very much wish that we had found otherwise, the study does pretty clearly suggest, in my opinion, patients will have fewer live births and more miscarriages, both early and clinical losses, in the absence of intramuscular progesterone.

So there have been a number of other opinion pieces and reanalyses that have come up in various publications, and so we just wanted to say that, look, the data are the data. And that was kind of the thought behind publishing that. And in general, there are so many questions in our field where we're lacking level one evidence.

We're really lacking a clear answer to a question, and those are ones that we should really focus on. I would love to see more trials that can result in more patient-friendly options in terms of the mode of delivery of progesterone in program frozen embryo transfer cycles. And in fact, many of us are committed to looking more rigorously at subcutaneous progesterone as an option, perhaps in addition to other forms.

That said, we want to give our patients our best outcomes, and to me, we have to keep using progesterone in oil for now. Thank you for writing that. It's an interesting debate that's unfolded in FNS and FNS Reports, so I encourage you to read this article from Eubanks and Divine.

Kate, we're going to jump right into the seminal contribution, which you have. This is an NIH working group classifying the nomenclature of ovarian follicle morphology. Tell us about the seminal contribution in May.

Thanks, Micah. Yeah, so this is a classification system of human ovarian follicle morphology, and so this is a seminal contribution. I'm very pleased that our fearless editor-in-chief and others, as the case may be, nominated it as such because I really think in order to be helpful to our field, it's going to need all the attention it can get for wide adoption.

So essentially, the paper details recommendations from an international NICHD, so NIH-sponsored nomenclature workshop. It started in March of 2021. Authors were from the NIH, Oregon, Yale, Northwestern, Hopkins, Cornell, and even Yonsei University of Seoul, South Korea.

They met with the goal of developing a consensus on histologic human ovarian follicle staging. They wanted to also give guidance on how to calculate follicle density and assess for changes in follicle morphology and follicle density that may be owing to tissue fixation. So it's a pretty technical paper that's well worth the read.

They rightly state that the rationale for the construction of these recommendations is that we need consensus to facilitate communication among clinicians and reproductive medicine researchers of ovarian pathophysiology, and this is a common theme we see in our field that lots of people are doing work about the same thing, but they're all talking about it in different ways, which makes it very difficult to assess the literature en masse and to really kind of advance the field collectively and in collaboration. And so the authors wanted to streamline the terminology, and they really wanted to further our understanding of the process by which human females progress from somewhere around 500,000 to a million immature eggs encased in ovarian follicles at birth down to about 1,000 at the time of menopause, really a fascinating process that we don't understand well at all yet, and we know that we ovulate only around 500 over the course of the lifespan. So how that process of loss of ovarian reserve happens with aging is not well understood or not well enough understood in my opinion, and the hope is that this terminology, a common terminology will help facilitate that.

I have to take a slight diversion and give a shout out to Dr. Jim Segars, who sat with me at the microscope when I was a fellow because my thesis involved copious H&E stained ovarian sections and counting follicles from mice in this case, and this is on ovarian follicles, but I have a combination of deep appreciation for these authors and also maybe a slight small amount of PTSD. So at any rate, what the authors were able to do is they took Gougeon's previously published classification system of eight classes of follicles up to 22 total follicle classifications, and they include descriptions of both normal and abnormal morphology. They described luteinized unruptured follicles as well as various types of corporal lutea.

So the 22 includes all of those various stages of all the way from what we would think of as an immature germinal vesicle all the way up to normal and abnormal corporal lutea. They also provide an atlas for reference as well as reviewing the expression of various receptors and cellular and molecular markers that have been described at the various stages. So definitely provides a valuable resource for ovarian biologists, this paper, and I would strongly recommend it to REI fellows studying for their boards as well because these are things that you need to know.

So the group went on to provide guidance on how to estimate follicle count and density from 2D cross sections of 3D ovaries, and they established a new follicle designation in addition to the other 22 that they called tangential, which is just a practical designation and again thinking back to the days sitting back at the microscope, this is when you just see kind of it's out of plane, the follicle that you see under the microscope, and so you're unable to assess whether and to which type the follicle belongs in the classification otherwise, and they also give recommendations as to how to deal with these. They include how ovarian samples intended for evaluation of follicular structures and density should be handled in order to minimize artifact, and they also give some guidance as to how to identify follicle changes that may be attributable to sample handling and or fixation. So as to not erroneously designate a follicle due to something that's just because of the way it was fixed.

Honestly, this paper got me thinking that I really hope in the future imaging modalities assisted by AI can recreate 3D images with sufficient resolution and precision to obviate the need for this method of estimation, which is always going to be an estimation, and also to get around what's the elephant in the room, and the authors do acknowledge it, that most of the time, thankfully, we don't have whole ovaries from young girls and women that have been donated to science, that most of the time, the ovaries are remaining in situ, and so we don't have what they are recommending, which is to count every fifth section at five micrometer thickness in order to achieve a global picture of ovarian reserve while minimizing the risk of double counting. So again, I'm happy that this was assigned a seminal article because the recommendations are only going to be as helpful as the degree to which adoption is achieved, and I do think that standardization, as always, will help present and future scientists conducting related research to do better work and make more progress. I have some concerns, I have to say, that their zeal towards completeness and precision may have created a system whose applicability is somewhat challenged by its complexity, 22 follicle designations is a lot, but time will tell, and overall, I think it's a great contribution.

They address the importance of disseminating the system and lay out their current and future plans to do so. They're collaborating with the Human Biomolecular Atlas Program, which I like the acronym, it's called HUBMAP, and that will also incorporate omics data, including single-cell RNA-seq. So I hope that this will help us learn a lot more about human ovarian biology.

Thanks, Kate. I like maybe the fearless adjective does meet here because my main reason for picking this as a seminal contribution is FNS is more than just a clinical journal, it really is a scientific journal and although this might not affect our daily practice today as a clinician, knowing the precise molecular and biological way follicles grow is really, really important to all of us and it's going to advance the science. And I do think it's a good read.

I remember reading as a fellow that this was the most logical and intuitive story and once you get that story of folliculogenesis and things, it really makes a lot of sense. So I'm pleased to publish it and hopefully it will advance the science. The HUBMAP is a great opportunity.

It's been going on for a couple of years. We have a site at Penn with Kate O'Neill who's just fantastic. The idea that we finally convinced the NIH years ago, not today, it would be harder today, we finally convinced the NIH that we need to map human female anatomy rather than just livers and lungs was a huge bonus.

So I wanted to make a shout out for that as well. Thank you, Kate. Fantastic.

And I echo what Kate said. If you're a fellow, definitely read this. I wouldn't be surprised if you see some things out of this popping up on your board examinations.

It's an important article and very well summarized. Thank you, Kate. Kurt, we're jumping to you.

We have an RCT in the andrology section on men. Tell us about this one. Yeah, I was real pleased to see this one as well.

This one, for you keeping track at home, the impact of prednisone on vasectomy reversal outcomes, the IPRED study. Every study has to have an acronym and I'm not sure I get this one, but it's called the IPRED study. And it's the result of a randomized control trial by Landon Trost and Joshua Savage.

It's out of the male fertility clinics in Utah and in Brigham Young and in Rochester, the Mayo Clinic in Rochester. As far as I can tell, and this is important, this is an unfunded study. And I'm not sure this study is going to affect our practice on a daily basis, but it's a really good story.

So basically, the story here is that they wanted to evaluate the safety and efficacy of prednisone on pregnancy rates and sperm concentrations up to a year after vasectomy reversal. Yeah, a little bit of a specific niche there, but the question was based on some relatively old data. And it really was a, you know, a unique, pragmatic way to not unique, a very pragmatic way of finding out what was standard of care was actually working.

So very briefly, the story is that you can go back to the 1980s when there's a case, not a case series, maybe a surgeon's report that basically said when you can give prednisone 40 milligrams for six weeks, it's a lot of prednisone, to his success rates, and he felt that the pregnancy rates were higher, the return to sperm was better. Although this was a relatively unsophisticated trial without really true statistics and everything else, it basically was adopted and became the standard of care. Later studies began looking at the prevention in other studies, and many people started adopting it, especially when I'm going to go back in time when we were worried about anti-sperm antibodies.

Do you remember that? I remember as a fellow, we were very concerned about tests and treatments for anti-sperm antibodies. And then I can't tell you the date, but just gone. We just don't even measure it.

We don't even think about it anymore. Anyway, that's a sidetrack. And then more recently, there was other studies, a lot of retrospective series, a second study that showed that there was perhaps a greater than 50 percent decrease in sperm concentrations if you didn't get prednisone and it became the standard.

But this group, we just basically said we sought to perform and enroll a trial to evaluate this prednisone postoperatively. And they wanted pregnancy as the primary outcome, not sperm counts, which had been the case in the most of the other ones. So they hypothesized a trial.

Now, the trial is relatively small. It only randomized 100 people into four groups. The four groups were high dose prednisone for six weeks.

High dose prednisone PRN, if they saw any reason that the sperm counts were not coming back. Low dose prednisone, which they put in and admit was they kind of made up just so they have a low dose prednisone. It wasn't really a standard of care and no prednisone.

And off they went at a single center enrolling men in a pseudo scientific way. They had built the tables preoperatively and where the people fit in the tables based on the characteristic would give them which arm to put them in. Interestingly, the criteria were age 18 to 65.

I didn't look closely to see how many people were at each end of that extreme, but that struck me as kind of an odd age inclusion for those getting vasectomy reversals. And they had to have prior paternity. The results were halfway through the study or at least interim in the study, they noticed that two prednisone arms were having lower pregnancy rates and they stopped enrolling to them after enrolling around 15 or 17 people.

And they just continued enrolling the placebo and the low dose prednisone. So they didn't actually meet their sample size requirements, but they they stopped in a sense, stopped the intervention arm because of lower pregnancy rates. And lo and behold, at the end of the day, they found out pretty definitively and statistically significantly that both arms of the high dose prednisone with vasectomy reversal were reducing pregnancy rates by a significant amount, despite the secondary outcomes being the same.

The sperm concentration was the same. The number of people that had sperm, meaning had a successful operation, was the same. And it's a pretty clear idea that the clinical endpoint was worse, whereas the surrogate endpoint was the same.

So if you just looked at sperm concentrations, the surrogate outcome, like any of us do for clinical outcome, they would have been wrong and actually they were potentially hurting people. So I'd like that's the study for those very, very reasons. It was perhaps not the most sophisticated RCT we've ever seen before, but it really had some novelty behind it and some really interesting thought processes behind it.

Unfortunately, it goes into what my fellow just tell me, another RCT of telling us what we shouldn't do rather than an RCT telling us what we should do. But it still gets rid of, you know, 20, 30 year old biased data by expert opinion that people have adopted that we find out is wrong. It's wrong probably for two reasons.

One, they were looking at the wrong outcome, the sperm concentration rather than the clinical pregnancy rates. And secondly, it does show you that we can adopt things based on plausibility or someone's expert rather than really quality data. Again, advocating the need.

That's why we do randomized trials. That's why Kate did the trial for IM versus subacute progesterone or vaginal progesterone, because we want it to be one thing, but it might not be that thing. The truth might not be so easy to come up with.

So that's why I was glad to see it in FNS. It taught me a lesson. I'm not going to do a vasectomy reversal tomorrow, so it's not going to change my clinical care.

But a pregnancy rate between, you know, 24% versus 30% is significant, both clinically and statistically. So I was I was glad that these authors came up with this study. And again, I think it was just their idea.

I can't tell from the paper. There's there's no conflict of interest. There's no there's no funding.

There's no one pushing them. They just really went after what they hoped would be the right answer. Such a neat story that you tell, Kurt, as soon as I saw this article in press, I took it to our urologist at Shady Grove because I was curious and they, you know, they explained sort of why this idea was out there.

But if I just Google it and ask or ask the AI on ChatGPT, it says that high dose steroids inhibit wound healing. And I think as surgeons, we all kind of get that. We're worried when we have our post-surgical patients on high dose steroids.

So it's such a good story about how observational data sometimes is completely different than what we find when we put it to a clinical trial. Yeah, that was a theory, believe it or not. ChatGPT is not wrong, although I'd like to say it was wrong.

The idea was that they didn't want their anastomosis to to to heal. In other words, that was the whole idea someone came up with. If we give steroids, we can keep it open and it stops this, the, you know, the whatever, the scarring down of the anastomosis.

I also want to make another plug for future podcasts here. This was done with male in semen analysis. All of their post-op was male in semen analysis.

Now, admittedly, all they can get is concentration, which is why they didn't report on motility and other things like that. But there are a lot of papers I'm seeing now that we're going to be talking about in the podcast in the near future about do men need to really come into the office for semen analysis? So we're going to have another change in care, which we're going to be discussing a lot about, I think. I love this paper and, you know, really rigorously done.

Yes, it's a small study. They followed these guys for a long time to to to really assess their semen concentration, as you say, and also their pregnancy rates. And I got thinking this the same thing as as you heard about the anti-sperm antibodies.

And Mike spoke to our same actual urologist that Micah did about them. And, you know, I think it was it was basically just a practice guideline that that said to drop it out of the workup. But I I did a literature search, too, and I didn't find anything that has ever sort of disproven it as a potential biologic plausibility as to our explanation as to why men who even when they have sperm return to the ejaculate post vasectomy do not have the same pregnancy rates as as men do pre ejaculate.

So pre vasectomy rather. So, you know, it's a really interesting concept and and one that I sometimes talk about with my patients, too. But I would love to see that see more in the literature looking at whether there is a correlation between post vasectomy reversal, you know, agglutination or, you know, anti-sperm antibodies and lack of pregnancy post reversal, even in the setting of a of a good semen sample.

So I digress. But, you know, I find it a very interesting possibility. It didn't amaze you that there are actually couples in this world that are getting pregnant without IVF after a vasectomy? I'm always other options.

Well, Kate, some fellow out there just got their thesis when they listen to this, so hopefully we'll see that published in FNS in a few years. We're going to move on to assisted reproduction, and the next article is one that I have. And so if you remember, back in January, we learned about a relatively new type of study called the emulated clinical trial.

This month, we're going to talk, I think, for the first time that I remember, about a new type of meta-analysis called the network meta-analysis, and we'll explain what that means in a minute. So this study is titled Triggering Oocyte Maturation and IVF Treatment of Healthy Responders, a Systematic Review and Network Meta-Analysis. It's from a group of authors from the UK, the US and Australia.

First author, Beebeejaun and senior author, Sunkara. So this objective was to compare four different types of trigger in healthy responders, meaning people where you're expecting a good number of eggs. So they're looking at HCG versus agonist trigger versus double triggers where you give the same medicine twice versus dual triggers where you use both HCG and an agonist.

They looked only at randomized controlled trials. They primarily looked at high integrity trials based upon the criteria that have been set out by Ben Moll and some of his colleagues in assessing data integrity. So what is a network meta-analysis? We're all used to the regular types of meta-analysis where we have, let's say, six RCTs.

They're comparing A versus B. We combine those six RCTs statistically to synthesize a combination result from all of them to give us what we hope is the best estimate of the actual effect of that intervention or comparison. In a network meta-analysis, we can compare things that don't have direct comparisons. So let's say we have six studies comparing A to B, we have five studies comparing B to C, but nothing comparing A to C. A network meta-analysis attempts to synthesize that data and allow us to compare using clinical trial data A to C, even though no clinical trials actually exist to do that.

A couple of points of this. Some people believe that because we don't have these direct observations, even though we're taking these from RCTs, we should consider those results observational. We should consider the results of a network meta-analysis to be observational, whereas if you have a direct meta-analysis, you can consider those to be randomized trial data when you synthesize these.

The important concept, because I spent an entire weekend going down the rabbit hole of learning what a network meta-analysis is, there's this concept called transitivity. And what does that mean? It means that all the indirect comparisons where those trials don't actually exist, but we're making those inferences, we have to assume that everything is randomized equally between those indirect studies except the actual intervention. And why is this important? Well, let's say the studies that compare HCG to agonist are all in high responders at risk of OHSS, and that's why those studies were done.

But studies comparing HCG to dual trigger are in patients that have a history of low maturity. And so that's why those studies were done. It may not be that in that case we can compare A to C because we don't have transitivity.

In other words, the patient groups aren't the same, the interventions aren't quite the same. And so we can't assume that those two comparisons are truly randomized. And so the other term that's been coined for that when we feel like we don't have that is incoherence or inconsistency.

In other words, the effect modifiers are not balanced between the two arms that we're indirectly trying to compare. So that's the basis of what they did. They used random effects models, which I think is appropriate for meta-analysis where you're using indirect comparisons because we don't truly have randomized data between the two.

So that's appropriate. I'm curious to what Kurt and Kate think of this statement. They said in studies where non-blinded outcomes were assessed, they scored the risk of detection bias at low when the outcome measure was objective.

In other words, if we're saying clinical pregnancy is the outcome, that's objective. If we're not blinded, that can't bias us in any way. So we're going to say it's low.

I happen to disagree with that. I think anyone who's done a clinical trial knows there's a lot of ways you could allow yourself to be biased if you're not blinded to the data, even if the data is a hard objective outcome. But I have read that statement a lot in papers over the last couple years.

So I just want to pause here and get Kurt and Kate's feeling on that statement. I'm still thinking about you and Logan at your fire pit going down the rabbit hole on the network meta-analysis. Micah and his dog like to prepare for the podcast together.

I agree with you 100 percent. I don't think that they're just because it happens to be something that is concrete and can't be and is immutable, I guess, that it can't be biased. I think that that could take us in a lot of really wrong directions.

I'm interested in what Kurt has to say about that, too. I'll be Switzerland. I think it's clearly more objective than other outcomes we're talking about.

So I don't disagree with the concept. But I also admit that there's lots of ways that we can turn a blind eye to one case or look more closely at another case if we know what's going on. But for the purpose of the analysis, I think I would make it lower.

But again, that's why you do a well done meta-analysis with sensitivity analysis and look at things in different ways. So, yeah, I'll play both sides. Well, I think that's fair, Kurt.

I disagree with it conceptually, but I don't think it affected the outcomes here or what they ultimately observed. So I think that's well said. What they did do really well that I like, which is something that Kurt has published on and instituted in F&S, is that we should look when we do our meta-analysis at high integrity trials first.

That should be the primary outcome. If you want to add in trials that have concerns over the methodology or the data integrity, you can do that. But that should be the sensitivity analysis.

Traditionally, it's been done in the opposite way. The problem with that is that's what ends up in the abstract and the conclusion. And that's what most people read.

And they don't realize that the high quality trial data that we have trust in actually shows a different finding. So I'm glad to see, Kurt, that what you have envisioned as our fearless leader has actually come into play in this study. So congratulations to the authors on that.

I really think that's the right way to do it. If we can digress for a second, I worry about this. And the only thing that really can call into question is who gets to call what's high quality and what's not.

But that's science. A meta-analysis should not just be a mathematic algorithm you put data in and out. This is where you think about your science.

And this is where you should have healthy scientific debate. And someone can come back to you and say, you should have included this study or excluded this study. It's almost like designing your study.

So I think it's much better, much more rigorous and will lead to better quality. Yeah, I wholeheartedly agree with you. And I think this is one place where, as you're implying, you could introduce bias on the authors because they could exclude studies that don't show maybe what they think they want to find.

And so pre-registering your meta-analysis like this one did and saying what your plan is to assess those things and making it clear helps reduce that bias on the back end. But I definitely think this is the way to go. And I appreciate that these authors did that as well.

The other thing I learned in addition to what we talked about with network meta-analysis and transitivity and incoherence of data is something called the SUCRA, a surface area, a surface under the cumulative ranking area. So a SUCRA basically says we have four different types of treatments here, the four different types of triggers. How likely is it for each one that they're the best, that they're ranked number one as we do this, that they're ranked number two, number three and number four? So just think of a graph.

Let's say that trigger A is 60 percent likely when you compare all these studies to be the best study. So it's going to be at point six on its plot. Number one.

And then it's 20 percent likely to be the second best study. Well, point six plus point two is 80 percent. So now it's at point eight on the graph.

And then it's 20 percent to be third. It's at 100 percent. And then fourth is zero percent.

It's still at 100 percent. So every study is going to be 100 percent as you go along. But the study I just described is going to have this big area under the curve.

If triggered C or D is zero percent likely to be one, zero percent to 30 percent, three and 70 percent four, it's going to have a much different curve. That's going to have a much lower area under this curve. So this is a new concept to me that makes sense.

There are some real limitations to this, though. This is just ranking their probability that it's number one, two, three and four. It doesn't actually tell you whether there's any difference between those or not.

And in this case, we actually see that, yes, something is ranked number one. In the case of this study, we're going to find out it's dual trigger, but it's not statistically better than the other treatments. So the SUCRA doesn't take that into account.

It also doesn't consider data quality and it doesn't consider risks and benefits and costs. So, for example, something might be most likely to be number one in a trigger for live birth, but also most likely to be number one for OHSS, which they looked at. And so the SUCRA can't really tease that out for us.

So those are the things I learned about in this study that are new to me. And I really appreciate those concepts. But what did they find? That's what we all want to know.

Bottom line, they have 24 different comparisons for these four different treatment groups, and they're looking at clinical pregnancy, live birth, OHSS, miscarriage and oocyte yield. All the triggers pretty much do the same overall between these 24 comparisons. I think clinically we probably all feel that way for at least the majority of our patients and that this is looking at good responders.

There might be a benefit to dual trigger, but the big caveat is there's one RCT that data has very large confidence intervals. So we really need more studies to know if the dual trigger and maybe that FSH and endogenous LH bolus that we get is truly helpful. So I think it confirms probably what we already know clinically.

But my interest in this study was all the really cool new methods that I got to learn about that I didn't know before. That was an amazing summary, Micah. I learned a lot from that and really well done analysis here and I think clinically useful, too.

I mean, of course, the devil's in the details. As you mentioned, with dual trigger, really what we're thinking about is, is it going to help those patients with a history of poor maturity? And this is, you know, we're unable to kind of look at it with that precision in a very in that specific population from this meta analysis. But it's just another example of how we do things to treat ourselves.

And it's good to have, you know, good science like this to bring us back down to earth as to as to reason. That was terrific, Micah. I want to delve back a little bit into the integrity thing because it's just a window of opportunity here.

I mean, the real problem, not just in our field, but in a lot of the medical subsexualities is the trials that are put out there are what we call me to trials for personal fame, meaning that a lot of people jump on a bandwagon and say, in my population, this worked better or growth hormone worked better or dual trigger worked better. And it's just not credible data. It's and a lot of these, as you've heard me say, are actually made up there that the data never happened.

It's just somebody trying to get fame for their practice or fame for themselves. And it's really hard to sift through that. But at the end of the day, when you do a big meta analysis like this, especially network one, if if that's true, if much of this data is really not credible and is just trying to contribute your name to literature without really making a true scientific breakthrough, it ends up to be no, it ends up to be coming back to the mean, except for the first two or three trials on a subject.

That's why I'm worried about this dual trigger. No one would ever publish from their clinic in wherever I use a dual trigger and it lowered my pregnancy rates. So it's, you know, for all the reasons I said, the reason to get your name in literature is positive.

No one's going to get fame and fortune from saying, I did a really good study and what you're proposing doesn't work. So it's still drifts. Everything drifts towards positivity in our field.

And we just have to just be careful of that is what I'm saying. That's well said. What I liked about it, it was just thinking through the emulated trial and trying to balance the effect modifiers or confounders and how they're trying to do the same thing here, but at a meta analysis level.

So similar concepts, a little bit of a different methodologic approach to the same problems that studies always try to control for. Kurt, we're going to jump into genetics now, and you have a study on balanced translocation patients and looking at their PGTA results and live birth results. Tell us about this study.

Thank you, Micah. This is yet another complex genetic study that really takes a good read to understand exactly what was proposed and exactly what was found. And then when you do that, you find out, wow, that was really an elegant paper.

And it then actually did teach me something. So this paper is called Parental Balanced Translocation Carriers Do Not Have a Decreased Usable Blastulation Rate or Live Birth Rates Compared with Infertile Couples. First authored by Kyle Nguyen Le and senior author Emily Osman, done by Cooper University and RMA New Jersey.

And it's actually a well-known study proposing to do exactly what it says. So the theory here is that carriers of translocations, and we all remember there's two types of translocations, there's Robertsonian and balanced translocations, are going to have trouble potentially making gametes and therefore potentially making embryos. And it's theorized that if you have this difficulty, even though you might make gametes, eggs and sperm, you might have a lower number of blastocysts and therefore a lower number of usable blastocysts.

By the way, usable blastocysts in this paper is defined as able to perform PGT testing. I can go to the specifics, but a certain grade and a certain advancement of the blastocysts. And they wanted to basically say, is this a true statement, whether they actually do have lower pregnancy rates, lower blastulation rates, lower usable blastulation rates? I like that their primary outcome was pregnancy rates, just like the paper I said before.

We really care about pregnancy, not just how many blastocysts somebody makes. And they went through a very busy practice with a very large number and made a couple of comparisons. They basically said, let's look at all the people that went through PGT-SR compared to those that had another reason for preimplantation genetic testing.

And that would be just your, quote unquote, standard infertility patient that went through PGT-A. And then therefore they tried to balance those groups. They say matched.

It's truly not a match. What they did was they just picked somebody that had roughly the same age and by roughly the same age, they meant fit into the same SART characteristic. Mild criticism.

I wish they controlled for age within the SART characteristic categories, but that doesn't seem to affect the study here. Then they were able to look at all those outcomes. So let's go back to the beginning.

Those are translocations, either balanced or Robinsonians. Clinically, we all know, present with sometimes recurrent pregnancy loss, sometimes allopaspermia, depending if the carrier is a male or a female, and perhaps something called implantation failure, which we don't know how to define. And I had never heard this, but one of the clinical rules of thumb that was coming out of this was that you don't always look for this balanced translocation.

So if somebody goes through your clinic and tries to make a blastocyst and doesn't, has blastulation failure, so to speak, therefore the recommendation, although I hadn't seen an ASRM, was to go do a carrier type on them to see if they actually were a translocation carrier. And that was one of the very specific ideas behind this paper was, is that recommendation worth it or not? So the answer was, let's look to see if that's the fundamental premise of this, that they are making lower blastulation, lower blasts or usable blasts or lower pregnancy rates. Is that actually even true or not? With me on the hypothesis? I told you I'd get confused on these papers.

We're tracking. I like it. All right.

So data from 2012 to 2022, a good 10 year data. They did do an appropriate two to one frequency match where they took all of their cases that had PGTSR, which was around 255. And they matched it to around, not around, 738 patients that had PGTA of similar ages.

It was about, I guess, very similar to most clinical practices. It's 11.5 percent Robinsonians and the rest were reciprocal. And the carriers were a little bit more predominantly in women, about 142 in women as opposed to 115 in men.

And then they compared all of those outcomes. And I wasn't very specific in most of those outcomes because the bottom line for the paper is they pretty much were similar. So what happened was the number of oocytes was about the same.

The number of 2PNs also about the same. The fertilization rate about the same. The mean blastulation rate was a little bit lower in the translocation group.

It was it was just under 60 percent as opposed to 62 percent. But that was statistically significant. But the mean number of usual blastocysts, again defined as growing far enough along with a grade and stage that's capable of doing a biopsy, was the same.

And then despite having a slightly lower number of euploid embryos, 1.7 opposed to 2.5, they had relatively similar pregnancy rates. So the bottom line of the paper is that, well, twofold. One, they didn't see big differences between all of these outcomes, between those that were carriers of translocations.

And then they went farther to say we also didn't see differences if the carrier was a man or if the carrier was a woman. And they also didn't see big differences whether the carrier was a Robinsonian or reciprocal. So big study to show that there aren't a lot of differences, but it took away a lot of the premises that we had built upon ourselves to say that these patients should be treated differently or we should be looking for this disorder with a carrier type in the patients themselves, which can be costly and time consuming.

And they ended up saying, really, you just need to go ahead and treat as you normally would, which is also not a bad outcome of a study. Well, let's talk about that first. There's a couple of things that I want to bring to your attention that are little Barnhart rabbit hole isms.

But what do you guys think of this study overall? I was thrilled to see it and shout out to Kyle, who's our current research fellow and soon to be first year fellow at EVMS. So congratulations to him on what I think is a great paper and to the whole that whole team. This is something that I see people do in practice all the time.

So when the rate of flash relation is lower than expected or when you put embryos don't implant or even sometimes when patients have a quote unquote pattern, not a repetitive deletion or duplication, but just the same chromosome represented in multiple abnormal results on their PGTA reports. And so obviously, karyotyping takes a lot of time. And if not covered by insurance, and this isn't for infertility anyway, it's not a standard indication.

It's quite expensive for patients as well. And it's a conversation that we have pretty, pretty frequently. So I liked that the data were reassuring and that this team is kind of answering right away your call for publishing negative studies as well.

And so the one quibble I had, and it's really not a quibble, this is still very meaningful data, is that there's obviously going to be differences in these populations that we can't account for. So the indications that they use to karyotype are the same indications that are standard. So basically recurrent pregnancy loss or basically they're looking for azoospermia or severe oligospermia.

Those guys were excluded from the study. So basically you're comparing an RPL population to an infertile population. And so that's not something that even with matching we can really kind of account for.

And I just wonder whether there's differences already in terms of the rate of blastulation among patients that have RPL versus those that basically are infertile. And again, we have to assume that the infertile population has normal karyotypes. Most of them were not karyotyped.

And that said, great paper. Somebody out there who has a huge, huge, huge patient population, please compare RPL patients undergoing IVF with or without a valence translocation. I'd love to see that, but I really will use this paper in my counseling.

Full of great research ideas today, Kate. I hope fellows around the country and the world are coming to me. At the risk of adding to our podcast, I do want to comment on two quick things about this paper and no fault of the authors.

But one is they use as pretty high. I don't remember if they called it primary or secondary was sustained implantation rate. So I just want to talk about this as an outcome real quickly.

These authors use it correctly. Sustained implantation rate was defined as the number of embryos with cardiac activity divided by the number of embryos transferred. Now, I just want to bring it up.

Most people or many people, I don't want to project too much, are using this incorrectly. Sustained implantation rate is not just the pregnancy made it to heart rate. It's actually got to do with the numbers of embryos transferred because sometimes, not much anymore, we used to transfer two embryos.

And how do you account for the idea that one of the two embryos would implant and the other didn't? So it's drifting is what I'm saying. Sustained implantation has become a surrogate marker for live birth and is the primary outcome for many trials for in our field. And I think that that's the wrong outcome.

I just wanted to say that explicitly. If you really are trying to account for multiple embryo transfers, OK, that's a good it's a good outcome. But it's not the primary outcome of a trial of infertility is your sustained implantation rate, meaning you made it to 12 weeks or 30 weeks or something like that.

So live birth is the right outcome. That's that was the first comment. The second comment is nomenclature.

They use the word equivalent about six or seven times in this paper. In fact, in the abstract, the usable blastulation rate, 42.7 versus 50 were equivalent. 47.2 and 50 are not equivalent.

So we need to get rid of the assumption that if you don't have statistical significance, therefore, the two arms are the same. If you had a big enough study, you might find a three percent difference in blastulation rate is statistically significant, but it's not equivalent. There's a method of my madness.

There's one paper we quote all the time, which is the paper that says that when you biopsy a blastocyst, the implantation rate with a biopsy blastocyst is equivalent to a non biopsy blastocyst. It's not equivalent. It's that the paper says it in its in its title and its title and its figure.

But it actually is lower. And you can do statistics around it. You can show that there is a lower.

So we have to stop using the word equivalent because it implies outcomes better than they might be. But it also is just wrong. It's just the wrong use of the word.

So I just wanted to go down that rabbit hole. Fireside stats with Kurt Barnhart. It's my favorite thing that we do on this show.

There are such things as equivalency trials, right? They're like non-inferiority trials, but even have a second margin. They have both the superiority and inferiority margin. It has to be within both of those.

And when you say the word equivalence, you're sort of implying that that's the study type you did. And that is not the study type that was done. Right.

These were superiority studies that didn't find a difference. Right. But but again, not finding something doesn't mean that that something is not there.

What what's the evidence of absence is not the absence of evidence. So I think I said that wrong. Absence of evidence is anyway.

You get the point. You can't just do a study and say, look, not statistically. Therefore, they're the same.

We never have to do the study again. That's just not correct. Fantastic.

All right. So as we're finding out, FNS is much more than just a IVF journal. We've talked about some basic science.

We've talked about some genetics. Now we're diving into reproductive endocrinology. The paper I have is called A Relook at the Relevance of TSH and Thyroid Autoimmunity for Pregnancy Analysis of RCTs from PPCOS-2 and the AMIGOS trial.

So this is from first author Kuokkanen and Pal, with a bunch of other wonderful authors from the Reproductive Medicine Network who are still doing secondary analyses and working with the CREST scholars to further our understanding of reproductive medicine. So they wanted to look specifically at sort of the interaction between thyroid autoimmunity and TSH levels. So looking at the interplay between both of those.

They call this a retrospective study of two RCTs. And I think that is the correct terminology. Even though some of the data was prospectively collected, the hypothesis was done well after.

And there's some missingness of data because not all the patients had all this blood work done. And some of it was done after the fact from frozen samples. So, again, remember that PPCOS was looking at patients with PCOS and AMIGOS was looking at patients with unexplained infertility.

And this is not an IVF trial. This is ovulation induction and intrauterine insemination. So they look specifically at a TSH over two and positive TPO antibodies.

And they recognize that it's a limitation potentially of using two as your TSH threshold. They did that based upon scatter plots of the interaction or the association between TSH and TPO. And they also did a secondary analysis looking at TSH of 2.5. Now, this was my favorite study that I read this month, just because of how well written it was and how critical the authors were and how they described the setup of their study and what they did.

And as the good shepherd myself of the ASRM practice committee guidance on TSH or subclinical hypothyroidism that came out, I was very interested in what they said. And in their introduction, they said additional they quote the practice committee saying that additional clinical trials evaluating the screening and treatment of SCH in the infertile populations could result in a change in our recommendation. And that is what we said in the practice committee.

And then they comment on that. This reflects not just the limitations of the existing data, but they convey the sense that the topic remains susceptible to being reshaped by future studies. And that's exactly correct.

The practice guidance was not saying that we have evidence that screening and treating subclinical hypothyroidism should not be done. We just don't have evidence that it should be done. And more studies are important.

And so that's why they did this. And so I'm glad that that was conveyed. So they end up with four groups of patients.

But the main two that they compare are the abnormal abnormal. So you have a TSH over two and you have TPO positive antibodies versus the opposite. Both of those results are normal.

They do have the two other groups, which are normal, abnormal and abnormal normal. But those end up not being very much different. They do some very nice statistics, including GLM models, which account for all the confounders within the study.

And then as we sort of talked about in some other studies, they do some nice sensitivity analyses, which really helps you drill down on how confident you are that the findings are robust across various potential confounders or modifiers. And this included looking at TSH over 2.5. And they make a really nice comment that I like. Statistically, they say that we would expect in the magnitude of any associations to increase, but we might lose statistical power because we'll have fewer patients.

And that sort of makes sense as you raise that threshold of what you found. Well, the summary of the study is that the patients who had both the TSH over two and positive antibodies had a lower chance of clinical pregnancy, 19 percent versus 29. Again, remember, this is not an IVF study that they're looking at.

Pregnancy loss was higher at 32 percent versus 11. Live birth was lower at 14 percent versus 26. And preterm birth was higher at 18 percent versus 3 percent.

So all of those outcomes were worse in patients that had this combination. The one thing I will note is that we get down to some pretty small numbers here. So preterm birth is three cases versus seven cases.

The P value is very, very low. But if you had one more case or one fewer case in either arm, it would make that not statistically significant. So there is some potential for type one error here if one case was different in either direction.

But nonetheless, pretty consistent findings across all of their outcomes. And these held up in the GLM models adjusting for confounders. Now, one thing that's really interesting that I think is important to remember here, you would say, so why don't we go back to treating these patients with levothyroxine? Well, remember, when these two studies were done, PPCOS2 and AMIGOS, we were treating these patients with levothyroxine.

And in fact, half the patients in all of the study were on levothyroxine. That's right. Half the patients in this studies were on treatment.

And there were four times more likely to be on treatment in the abnormal arm. So they're seeing worse outcomes in these patients and they're almost all on treatment. And so they observe this and they say this clearly shows that treating these patients probably isn't associated with better outcomes.

And maybe if there is a mechanism that's causing these worse outcomes. And to me, this was the most interesting thing. It's a separate mechanism from the action of the hormone on the receptor itself, which is sort of what I've always argued.

It's never to me from a biologic plausibility standpoint. I know there's really smart people that see it the other way. I've never quite understood why giving levothyroxine in the setting of the antibody would change the outcome of that antibody.

But there's good arguments. But they at least suggest that this and they give some hypotheses of some other mechanisms by which this could happen. So I thought that was a really interesting finding.

And we can talk about that here. And the other thing I liked at the end, again, this is just exquisitely well-written and they really define what they found in a good way. They say we underscore that our findings should not be interpreted as justification for treating these women.

Rather, it's our belief that these findings should guide further studies looking at the positive or putative causative mechanisms that might explain this relationship that's been observed. They say maybe this is why it's observed in some studies and not others. And their findings should be hypothesis generating.

And for me, when I see authors who are really critical of their own data and don't try to overstate their findings, but frame it in an actual way that's healthy and helpful for us for future studies, I feel like that study probably was very rigorously done. And so for fellows out there, I would encourage you to read this study, even just to see how to do a really well-written paper from the experts at the RMN. This is sort of what I guess I would expect from an author group like this.

Kate, Kurt, thoughts on the many things that I just said about that study and those findings? I was just so happy that both of these two papers were in this issue, Micah. They're so well-suited for your expertise. So great summary.

And I agree. And I might even go, it's hard to think of a mechanism for this, too. So to your point about biologic plausibility, I recommend caution here.

But the other potential mechanism that came to mind to me is that the exposed group were four times more likely to be on levothyroxine. So it's also possible that it's actually harmful rather than helpful. But not an argument here.

And I think you said this well, and it's important to swing the pendulum back to testing everyone because we don't really have anything to offer them, even if these findings were to were to hold up. And again, it's meant to be hypothesis generating. Yeah, I would just add that I've been frustrated at this for a long time because I think we're just chasing lab values here.

And if there really is evidence that our chasing lab values is harmful, that's a really good thing to know. And not only can we save time, energy and a lot of angst by ourselves, the cost of testing and the patients don't like it either. They get obsessed about whether they're taking half doses or how many days and things.

So I think that maybe this is just the wrong rabbit hole to be chasing. So another fantastic study, as always, there are many other papers that are very educational and teach us a lot, including our video articles and our research letters. So I encourage you to look through the whole journal and read all the articles that are important to you.

A shout out to all of our listeners. Please like and subscribe to our podcast wherever you download your podcast from. And we look forward to talking to you again next month.

Enjoy a vigorous discussion. This is a lot of fun. I hope it's very valuable to you all.

I know we keep saying the fellows should read it, but everybody should be reading these articles. That's why we have this journal. Well said.

Absolutely. This concludes our episode of Fertility and Sterility on Air, brought to you by Fertility and Sterility in conjunction with the American Society for Reproductive Medicine. This podcast is produced by Dr. Molly Kornfield and Dr. Adriana Wong.

This podcast was developed by Fertility and Sterility and the American Society for Reproductive Medicine as an educational resource in service to its members and other practicing clinicians. While the podcast reflects the views of the authors and the hosts, it is not intended to be the only approved standard of practice or to direct an exclusive course of treatment. The opinions expressed are those of the discussants and do not reflect fertility and sterility or the American Society for Reproductive Medicine.