Fertility and Sterility On Air - TOC: June 2026

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

(03:58) Views and Reviews: Data driven analysis of fertility optimization

(06;29) Fertile Battle: The dilemma: Is it reasonable to freeze immature oocytes and, if so, when?

(08:16) ASRM Pages: Recurrent pregnancy loss: a Committee Opinion & Recurrent Implantation Failure (RIF): A Committee Opinion

09:26) Birth outcomes after in vitro fertilization among cancer survivors 

(17:14) Intra-Patient Variability in the Number of Retrieved Oocytes and Ovarian Response Categories Between Consecutive IVF Cycles

(31:16) Racial/Ethnic Variation in Ovarian Responsiveness and Live Birth After In Vitro Fertilization 

(38:16) Spindle dynamics and chromosome segregation in human preimplantation embryos 

(44:47) One-Quarter of Men with AZFc Deletions Have Sperm within the Ejaculate at the Time of Planned Micro-TESE

(53:22) Association between hereditary breast cancer gene variants and aneuploidy rates in IVF cycles using PGT-M

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 discussions with authors and other special features. FNS On Air is brought to you by the 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, Dr. Pietro Bortoletto, Interactive Associate-in-Chief, and Associate Editor, Dr. Kate Devine.

Welcome to another episode of Fertility and Sterility On Air. I'm your host, Micah Hill, the Media Editor of FNS. I'm joined by Pietro Bortoletto.

Good morning, Pietro. How are you? Hey, Micah. It's a fun group today.

I know our listeners will be surprised to hear that we have a couple of new faces on the podcast today. I don't want to spoil it, but it's me, Micah, and your very own inaugural FNS Editorial Fellow, Alison Eubanks. Hi, Alison.

Hi. Thank you for having me. We're so excited.

You're usually someone who's been living behind the scenes in doing stuff to further the mission of the journal, helping with the editorial process. It is really cool to have you on the mic today. Thank you.

It's exciting to be here. Pietro said just before we started recording, you're only filling in the shoes for Editor-in-Chief Kurt Barnhart, so no pressure. Easy to do.

This is going to be a lot of fun to hear from you. Even Kate, don't bring much to the podcast. It'll be fine.

Was I supposed to prepare? We'll see. We'll find out. Only one way to know.

Actually, we have the call for the next Editorial Fellow. We're changing the name to Editorial Scholar because you don't have to be a fellow. You don't even have to be a physician.

You could be an embryologist. You could be a urologist. You could be post-fellowship.

I've heard from about a dozen people that have seen this online and are applying, but, Alison, give us a 30-second plug. We actually extended you to a second year because you were doing so well. What is the Editorial Fellowship or Editorial Scholar, and what did you like about it? Why should people apply? It's been an ability for me to really see behind the scenes of how the journal functions.

It's been interesting to see how articles are chosen or moved from one sister journal to another. I think in terms of learning, understanding what makes a good article, what makes an article that we are unable to publish is really important. Obviously, learning just from the scholars that are on the editorial board is incredible, but I think it just gives you a glimpse into what tough decisions have to be made behind the scenes in order to publish a journal as strong as fertility and sterility for those that would like to work more in academic medicine or someday take over on the editorial board.

Great opportunity. It's a great innovation from FNS, and thank you, Pietra, actually, who spearheaded and brought this up within the journals. Thank you, sir.

Well, imitation is the highest form of flattery, and I saw our peer journals coming up with these positions to bring in new talent, and I feel like we're the lucky ones, Micah. We managed to find Dr. Eubanks, and like you said, we extended her for a whole other year because she was so darn good at her job. But it's time to get her on the actual editorial board and bring on the next editorial scholar to join us and continue to further the mission of the journal.

Absolutely. All right, let's dive in. We've got some great stuff to talk about for this month.

We are in June 2026, volume 125, issue number 6. The views and reviews this month is from our very own editorial editor, Eve Feinberg, Data-Driven Analysis of Fertility Optimization. It's what ASRM causes optimizing natural fertility. It's what other folks call restorative reproductive medicine, or RRM.

There's a series of four articles. The first one looks at diet, lifestyle, and human fertility. The second, optimization of conception with intercourse for women with infertility.

The third, surgical interventions to optimize fertility, and my favorite, menstrual tracking technologies and fertility evaluating accuracy, utility, and time to pregnancy. This was done by a great group of people with very good epidemiologic experience, and they go through all of the articles. Just summarizing it, at a very high level, most of these studies look at natural fertility in an 18 to 24-month time span.

They don't have control groups, so that's probably the main downside of this. Then they compare that 18 to 24 months of what is a highly curated patient population. They have good sperm.

They're young. Their tubes are open. Essentially, their definition of infertility is unexplained at that time point, and compare it to national data of one IVF cycle to claim that restorative reproductive medicine, or RM, is equivalent or even better than IVF.

Obviously, from a methodologic standpoint, none of those comparators are correct. This breaks that down. There's actually an impressed meta-analysis from a group of other epidemiologists, including Jack Wilkinson.

If any of you know Jack, he's not at all—he's very critical of IVF in many ways and very rigorous in his methodologic approach. In that meta-analysis, they actually, I think, don't include any studies because none of them meet the criteria of what would be needed for a high-quality meta-analysis of randomized clinical trials. I definitely recommend you look at that.

The reason this is important, and ASRM is paying attention to it, is that there are efforts at the state level in various states to use these types of study to actually try to reduce access to IVF. My main point would be that in the United States, we actually utilize IVF at the lowest rate of any of the modernized countries in the world. If anything, we need more access to IVF.

I'm a big fan of anything we can do to help people get pregnant without IVF, but we should not use data like this to restrict access to IVF. I encourage you to read that. The fertile battle is actually really interesting as well from Catherine Rakowski.

Should we freeze all oocytes when we're doing fertility preservation, not just the M2s? I think my head embryologist's head would explode if we had this debate because it'd be more work for the lab, but as technologies mature, such as IVM, are we discarding eggs that might be of value in the future? I think that's a really good topic to debate. Petra, I know you have some interest in this area. Just high-level, quick thoughts on that.

What do you think about that dichotomy? Yeah, that's a great question, Mike. I'll tell you what we do for our oncofertility patients and for our non-oncofertility fertility preservation patients. We are routinely freezing M1s and M2s.

I think all of us believe that M1s in the near or very imminent future, we will have better tools to be able to use them. Knowing that the time where people come back to actually use these eggs is sometimes a three, four, five-year time horizon for the 40 to 50 percent that do come back to use them. Four to five years from now is a totally different world in reproductive medicine.

We're going to have lots of new tools at our disposal. We're not routinely freezing GVs. I think that time horizon for making them useful is still so far away that it probably doesn't make sense to hold on to them, but I would love to be proven wrong.

I know that there are companies that are actively working on tools to be able to do the IVM. Full disclosure, I am a medical advisor for one of them, but I think that this is something that the field is really, really excited about. I think patients would stand to benefit if we could help GVs become useful eggs for patients down the road.

Great. Timely fertile battle from Catherine Rakowski. Definitely recommend you read that.

Then finally, ASRM pages, two practice committee documents. I think we've all been anticipating. The first is recurrent pregnancy loss.

Ruth Lottie was a special good shepherd and expert for the practice committee on helping to put this together. This was last done by ASRM in 2012, so very timely. There is a little bit of a shift in the paradigm of how we're recommending to do things.

A couple of new tests that weren't in the old one, such as CD138 staining with an endometrial biopsy, such as sperm DNA fragmentation on the male, although what do you do on the back end if they're not technically infertile and don't want to do IVF, I think is an interesting question. Then proceeding immediately if you have tissue to testing that tissue, because if the karyotype's abnormal, you got your answer for that pregnancy loss. Some pretty big shifts with the ASRM recommendations.

I recommend you read that. Second one, a brand new practice committee document on recurrent implantation failure led by Bruce Peer on the practice committee. I think this is something we've talked about a lot in the journal over the last five years.

Two really, really important new practice documents from ASRM. Definitely must read. All right, Pietro, we got to jump into the actual articles here, and you have the seminal contribution, birth outcomes after IVF among cancer survivors.

What did we learn from the seminal contribution this month? Thanks, Micah. I'm really excited. I got two killer articles this month to talk about, and I don't know, it's maybe because I assigned them to myself selfishly, or because there's just a lot of good science in this month's journal.

Well played. We'll never know, but I want to talk to you a little bit about a really important counseling paper, which is why I think Kurt highlighted it as a seminal contribution. We counsel cancer patients about fertility preservation all the time, but when they actually come back years later and use IVF to build a family, what are their actual pregnancy outcomes? Until now, we've been working with essentially little to no US data and only a small handful of international studies.

This is actually the first US analysis of IVF birth outcomes in cancer survivors. It's the largest study globally, and it includes 648 live births from cancer survivors out of 135,000 total IVF births. I'll lead with the punchline.

The bottom line is actually quite reassuring, but let's talk a little bit about how we got there. About 1.4 million reproductive age women in the US are living with a cancer diagnosis. IVF in the general population is already associated with higher rates of preterm births, somewhat lower birth weights than spontaneous or unassisted conceptions, which makes it hard to know whether cancer adds further risk on top of IVF, or is this all just coming from IVF? Like I told you, the prior evidence base was pretty thin.

There was one great Danish study that suggested higher risk, but it compared IVF cancer patients to non-IVF, non-cancer patients. If your statistical brain is going off, that's a nice co-mingling of some variables in populations that may not necessarily be totally apples to oranges. Two small Chinese studies show no difference, but none had really a meaningful number of patients to be definitive.

What did this group do? This group linked three big data sets across nine states. They linked SART cores, they linked state cancer registries, and state birth certificate data. This was data collected from 2004 up until 2018, so I think we could reasonably say that this straddles modern IVF practice in the middle of there obviously being the switch from slow-freeze to vitrification.

They compared two groups of cancer survivors to IVF patients without cancer. One was the fertility preservation group, so egg retrieval before cancer treatment, which resulted in 112 live births, and IVF after cancer treatment, which meant egg retrieval after treatment was completed, and this resulted in 536 live births. So this is important.

There's a teaching pearl here. The smartest design choice in this paper is that they used three different comparison groups instead of one. They used IVF for male factor only.

This is our proxy for known female infertility problems. IVF for prior tubal ligation, again an established proxy for established female fertility, and then IVF for all other indications combined. So if you can only use one comparison group and it happens to have unusual characteristics, your results could be misleading.

By showing the same answer across three different comparators, you've actually stress tested this finding a little bit. Think of it almost like a how do I do a sensitivity analysis. So let me give it to you in a bunch of different groups and really show that this finding is robust.

I thought that was a very cool methodologic decision on their part, and one that's like, oh okay, you knew exactly what I was going to poke holes in here on the podcast and you've already gotten ahead of it. So what's their big takeaway? The headline finding is that with one exception, birth outcomes in cancer survivors using IVF look the same as IVF outcomes for any other indication. When you adjust for maternal age, race, education, and other factors, women who used IVF after cancer treatment did not have significantly different rates of preterm birth, low birth weight, SGA, gestational hypertension, or cesarean delivery.

There was one consistent difference though, about a 30% lower risk of LGA babies in cancer survivors. It's actually a reassuring finding, not a concerning one. LGA is associated with birth injuries and soldier dystocia, so lower risk is actually better here.

And one final clinical pearl for fellows, this is now the largest dataset in the world on this question, and the clinical takeaway is straightforward. When you're counseling cancer survivors about returning for IVF and embryo transfer, you can use the same general counseling framework that you do for any other patient. You don't need a separate, more pessimistic version of this conversation.

Their cancer history, once they're cleared to pursue pregnancy, does not appear to add a meaningful risk to IVF-related birth outcomes. And I think that's generally a useful piece of evidence for the oncofertility conversation, which is already so emotionally charged and loaded. You don't want to then say, hey, we can use this, but it's going to potentially result in an even riskier pregnancy or neonatal outcome for you.

So I love this paper. As someone who used to lead the oncofertility program for a large academic network, I can say that this was kind of data that I so desperately wanted. I'm a big fan of this paper.

I think this is essential reading for fellows, essential reading for people who lead oncofertility programs. And if you're an oncofertility nurse listening to this podcast, which I know we have several, this is a paper that you should print out, have by your desk, and make sure the docs you work with are also reading. Micah, Allison, were you as jazzed up about this paper as I was? I love null studies.

There's this myth, I think, that people have to try to find something that's positive in the study and spin that as their primary outcome. And actually, what you said is an interesting thing, your little stat pearl. I agree.

It was beautiful in this study. The mistake some people make is dividing it up into three different comparisons, tripling the number of comparisons they have, finding one sub-analysis that's statistically significant because they did so many comparisons, and saying, well, that's the thing we were looking for when that wasn't a priority what they were. That's not what these people were doing.

They were beating up the null hypothesis. They were stress testing it in as many ways as they could, and it was still no difference. It was such reassuring data to see.

So this is not a study I would have thought to do. You with your MFM wife, this is probably something you've thought a lot about with your past role, but this is incredibly reassuring data for our patients. I agree.

Very good. Completely. And big shout out to the epidemiologists.

If you look at the author list, this is coming from the Department of Epidemiology at UNC. Leave it to the epidemiologist to come up with a really clever design, but also leave it to the epidemiologist to involve some really killer clinicians, Jenny Mercerow, Katie Cameron, Steve Spandorth, or Val Baker to help inform how this paper makes sense for the clinicians that are using it on the receiving end or the counseling end. Yeah, I think the diagnosis of cancer and the reproductive age is such a taxing emotionally process that it's so nice to have a rigorous evidence-based study that is overall reassuring.

I think one of the interesting outcomes of this study for me was that the C-section rate remained high in cancer survivors despite the other outcomes being overall reassuring, which we know is not devastating by any means. We all know C-sections are so notoriously difficult as an outcome to interpret because it's practice patterns, culture, patient preference, prior surgical history, and so I found that to be an interesting outcome. We love a good counseling paper here on the podcast, and this is one that you should all add to your back pocket.

Micah, I'm really excited about your next article. This is one that I would have assigned to myself had I not already had two killer articles and I wanted to share the love. Intrapatient variability in the number of retrieved eggs and ovarian response between consecutive IVF cycles.

I feel like there's so much counseling related to this exact topic that I do. What does this new paper tell us? I'm so excited that you gave me this paper. I think we're going to do this as the Journal Club at MRSI.

I'm confirming that today with the leaders there, so shout out Chicago in July. Come hear about us, talk about this. Also, we actually did this as a Journal Club just two days ago.

It was a joint Journal Club with Johns Hopkins, the NIH, Jim Segars, the program director at Johns Hopkins, and he was my program director at the NIH when I was a fellow, and Jared Robbins from ASRM flew in, taught us the embryo simulator course, and then we did this. Allison actually presented this article, so I'm not going to steal any of her thunder because it was one of the best fellow Journal Club presentations I've ever heard. This might be one of my favorite papers of the entire year.

I was really excited, and it's a little bit geeky to dive into this. The lead authors are from Israel, Hochberg and Horvitz. It's got some friends of the show, Nikos Palizos from Europe, Eduardo Harriton from the West Coast of the US, some really smart people that have done a lot of stuff with predictive modeling.

Their objective was to examine intrapatient variability in the number of oocytes retrieved across consecutive IVF cycles, so that means within the same patient, and we'll talk about how they whittled that down to make it a really cool analysis. This was a retrospective cohort study, and patients had to be doing the cycle within six months, and it had to be the same exact IVF cycle, and we'll talk about that here. Their primary outcome was the average percentage change in oocyte yield from one cycle to the next, as well as the R-squared, or the coefficient of determination.

In other words, how much did one cycle correlate with that second cycle? So how did they whittle down these patients to get them into pairs, so you can do this with inpatient analysis or intrapatient analysis? Their IVF cycle had to be within six months. It had to be the identical protocol, so GnRH antagonist versus microdose flare versus long gluteal, whatever your protocol du jour is for the patient. The gonadotropin type had to be the exact same, so mixed protocol with the two same drugs, just FSH only with the same drug, that had to be the same, and the exact same starting dosage.

They had to be freeze-only cycles, or freeze-all cycles, as they call it. At first, I put question marks by that when I read it, but I think that's just because of any possibility that a pregnancy that is then lost might affect the subsequent stimulation. I think that was probably why they did that.

Had to have at least one egg in each cycle, and if available, they had to have the same physician to manage both cycles when that data was available. So this means they went from a data set of 50,000 egg retrievals all the way down to 800, but in this case, I actually think that's better. Yeah, you're losing power from numbers, but you're making those numbers more meaningful, and you're getting a really good potential for a paired analysis.

They did something else called the Ideal Foresight Model, and Allison may have some comments on this as we get into it, but essentially, this evaluates the theoretical maximal ceiling at which any model could be predictive. And if you think about it, if you're having the same patient where everything is exactly the same, any difference between the two shouldn't be due to the model itself, because you haven't changed any inputs or any variables into that model. So if anything at all is different between the two cycles, that represents within the same person variability from one cycle to the next.

Sure, we'd all like to see an R-squared of one, which is a perfect line where the two cycles are always exactly the same in the patients. That's not biologic reality for anything in medicine, and so there are several ways to do this, but the Ideal Foresight Model is one way that essentially estimates that maximal ceiling at which any model could predict things. I think Allison has some thoughts on that for this.

There was no difference between the two groups, except for age. I guess you would expect a few things to be different just because the patients are older, but that just sort of shows us that they chose the right patients because two cycles were the same from their baseline characteristics. Table 2 is really where the meat of these data are.

So the average or the mean change from one cycle to the next was 62 percent. That's huge. Jaw dropping.

Yeah. My jaw is dropping. That is way more than I would have expected.

Now, is mean the right one to use? The median is 40 percent. Still a bigger number than I would have expected. That's still a big number, but it's not as much.

Now think about it. When is egg yield ever parametric? When is it ever normally distributed? It can't be, right? Because you average, let's say, 10 eggs on patients. The most eggs I've ever had personally is 88.

I know a doctor that had 108. If you're a listener and you've had more than that, please hit us up on social media and say, this is the most I've ever heard of. You can't have less than zero eggs, and so you have to have this massive right tail skew.

I actually think the median is a much better way to do it, and that lowers that number from 63 percent to 40 percent. But that's still a lot of variability from one cycle to the next. The interesting thing about this data set is as patients got older, that variability actually spread.

In my mind, that was counterintuitive, although to Allison it was intuitive, so good on her for that. The older you get, the more likely you are, and as I thought about it, to have this variability spread, because you're going to have a lot of patients with DOR over 40, but you're going to have a lot that aren't in DOR and still respond very well. So your spread gets bigger.

Your variability gets bigger. The other thing that's interesting is over half of the patients were over 40, so that's giving a lot of data to this. Now the one thing I will take from Allison that I'm going to steal from you from two days ago, because of how they did it by percent change, the more a low responder you are, the bigger the percent change you're going to give.

Think about it. If you had one egg in the first cycle and two eggs in the second cycle, that's 100 percent change. If you had 10 and then 9, that's a 10 percent change.

If you had 20 and then 19, that's a 5 percent change. So they're all getting one egg difference, but by analyzing it by percent change, if we have a patient population that's older and maybe has more DOOR, we're overemphasizing that spread. So that's just a caveat to take from this.

Their R-squared from the ideal foresight modeling was 0.83 percent, so not one, but actually, you know, that's the theoretic ceiling, the best that a model could do. So Allison, I'm actually going to pause here because I think you had some comments on this at our journal club. I think you felt like that maybe is a little bit of an overestimation of what the potential theoretic ceiling would be from one cycle to the next within a given patient.

Am I stating that accurately? Yeah, I think so. What's important about their ideal foresight model to remember is that the model regresses each cycle oocyte count against the mean of the pair. So if you have a patient that has two cycles, one gets eight eggs and the other one gets 12, you have to find the mean of the pair for the model, which is 10.

And the remaining deviation between the 10 and the actual outcome of 8 and 12 is that variability that can't be explained by the measured parameters. But by including the cycles in the model, it creates a part-whole correlation from that problem. So you are essentially inflating that R-squared a little bit.

I felt like this would be said the other day too, but I would interpret this loss as like a true predictive model and more of a conceptual kind of upper bound on the reproducibility. Table 3 is also a really good table. So here they're switching from looking at the mean and median percent change to grouping the patients into four groups.

What we would actually care about, how many eggs did they get clinically? Is this a meaningful shift or not? Poor was 1 to 3 eggs, suboptimal was 4 to 9, normal was 10 to 14, and high was 15 or more. And this is from published parameters that define cycle outcomes from the number of eggs by that. So you basically make a 4 by 4 table and put what they got in the first cycle on one axis, what they got in the second cycle on the other axis, and see how often are patients clinically switching groups.

And even though we had this pretty big 40% median, 63% mean change in yield, from an actual changing groups, 53% of the patients stayed in the same group. But that's still almost half that switched groups. Now the good thing is, those patients almost always only switched one group, only 29 out of 800 patients switched two groups, and no one went to the extremes.

No single patient went from poor to high or high to poor. So it was only 3% of the patients that changed more than two groups from a clinical standpoint. Those other half that shifted, shifted by just one clinically relevant group.

So what does all of this mean? I mean, this was from a statistical standpoint. It really challenges my thought that patients do very similar from cycle to cycle. But we've all seen it where patients have variability.

You do everything the exact same, and they get fewer eggs. So the variability was more than I expected. But on the good side, 53% of patients don't change.

And of those that do, it's generally just changing from a clinically meaning group by one, only 3% shift in a big direction. The change was bidirectional, and I think that's important. If they had seen everybody getting better, that would mean that even though we use the same protocol, we learned something from the first cycle.

Maybe we triggered them a day earlier because we had a bunch of degenerative eggs, or we triggered them a day later because we had too many M1s in the first cycle. So maybe we learned something. This is showing the change was bidirectional, meaning this was not the management.

And they actually did some things to show that it wasn't that. You know, does this have meaning for live birth? Maybe. I don't know.

I mean, we see this. I have a patient who's 42 in the last month whose first cycle was amazing. We got four embryos.

They were all aneuploid. Second cycle was a bad stem. It was the exact same protocol.

Max dose, micro dose, flare, exact same protocol. She only got one embryo. That's euploid.

So does this actually predict live birth? Maybe not. But I think this just shows us that there's a lot more variability within a patient from month to month than what I have been taught to expect and what I teach the fellows. So I'll open it up here to questions to you guys, because this is a fascinating paper.

I loved it. I think there was a couple of things I wish they had done, and maybe they still had the opportunity to do in subsequent papers. But one of the gut feelings I've had about why second cycles are often better than first cycles, yes, there's like a, I've tweaked this and I've made these adjustments, but I think sometimes it's just proximity to that first stimulation.

We all know that when we expose the ovaries to exogenous gonadotropins, the first response, you get what you get and you don't get upset. But in the second one, it feels like we're often getting a little bit more response in that second cycle, even when we do nothing at all. I think this paper adds some data to that, which is really nice.

I would have loved to have seen time from first stimulation to second stimulation and looking that at one month apart, two months apart, three months apart, four months, five months, six months, to see if there was any degradation or any pattern that happened in that subsequent response. And it might be like a priming mechanism that could be at work potentially. Yeah, correct.

I call it greasing the wheels to patients. I tell them like we've lubed up the gears and now the next time we get back on the bike, we often see things just be a little bit more efficient in the ovarian response. I would have loved to have seen that from this data.

Maybe they just didn't have enough stratification across those six months to actually to do it that way. We talked about in a perfect world, how would you get a model that really better fit, that didn't take into account smaller changes in variability or jumps and things. And I think you'd have to do more and more and more cycles for the same patient without changing a single thing to really get that to be more accurate.

But I think one of the things that was brought up by Jared Robins was that while it's our instinct to change plans for a patient, it actually is probably more unsettling for the patient. And it implies that we could have done something better the first time, which I think is a really interesting perspective when you're talking about counseling patients that while it feels right, oh, it didn't go well, we'll just change this thing. Sometimes we know it's going to make a difference or it's not going to make a difference, but it feels like the right thing to do to show the patient that we're going to do something different.

But the way he was framing it, I think was really important in that it almost also says that we had room to play with in the beginning and we didn't do the best we could do the first time, which I think is a really interesting perspective on counseling. And I think this study really drives that home of stay the course, understand that there's going to be some differences for things we don't understand at this point, but staying calm and keeping on is really the answer here. Yeah, I actually emailed Eduardo and Nico to ask them, so what's actually more predictive in that second cycle? Is it their first cycle or is it their AFC AMH? Because I've always taught my fellows, because it was what I was taught, that the biggest predictor is your prior response.

And this maybe challenges that. They haven't actually done that in analysis, but you can do models of models and you can run models against each other and see which is more predictive. So they told me they would get back with that data.

I would love to see that because it would challenge my paradigm and I love things that challenge my paradigm. That's how we all grow. Fantastic article.

Thank you, Pietro, for assigning it to me. Allison, we're staying in the world of ART and we're moving on to race, ethnicity, and ART outcomes. So my first paper, Racial and Ethnic Variation in Ovarian Responsiveness and Cumulative Life Birth After IVF, it was led by groups out of Cedars-Sinai and they collaborated with several other institutions, Vial Cornell, UCSF, USC, and they used SART and national registry data.

Then they chose to investigate kind of this broader clinical problem that we've known about for years and really see if they could identify where the concerning areas are. So they were looking at Black patients undergoing IVF consistently have lower live birth rates compared with White patients. But considering that IVF is a multi-step process, their question, you know, where in that pathway does that disparity emerge? The study specifically asked whether differences in ovarian responsiveness could explain this disparity.

They used SART coerced data from 2017 through 2019 and analyzed over 246,000 autologous retrieval cycles with available data. They excluded donor cycles, gestational carrier cycles, without enough stimulation or outcome data. And they used something called an Ovarian Sensitivity Index or OSI.

This was the primary biologic measure and that's calculated as the number of OSITs retrieved divided by the total gonadotropin dose times 1,000. So conceptually, OSI measures ovarian efficiency, how much ovarian output you get per unit of stimulation medication. And I think it's important because raw OSI count alone can sometimes be misleading.

A patient getting 12 OSITs after 1,500 units of gonadotropins is biologically different from somebody getting 12 OSITs after 5,000 units. And so that OSI attempts to correlate those figures. So they found that black patients had slightly lower fertilization rates, but actually higher usable blastocyst rates.

And even in PGTA cycles, the live birth disparity persisted. So stimulation response, embryo development, and aneuploidy alone are not fully explaining the differences here. Clinically, I think the paper redirects attention downstream towards implantation biology, uterine factors, environmental exposures, potentially chronic stress.

I also appreciated that the authors were careful not to biologize race itself. Race is, you know, as we know, a social construct. And these findings really reflect structural environmental inequities rather than intrinsic biological differences.

The strengths, I think, obviously, massive national data set, cumulative live birth linkage. You know, that's the outcome that we all want to talk about. And they really had a thoughtful framework to try and localize where that disparity emerges within a very complex, multi-step IVF care.

I think, you know, the biggest limitation was that about a third of the cycles were missing race or ethnicity data. The authors use multiple imputation, which is statistically appropriate, but imputation can't fully substitute for complete demographic data collection. I think overall, the paper was really valuable, not because it fully explains what's going on in the IVF disparities, but because it convincingly suggests that ovarian responsiveness alone is probably not the primary driver.

I think those are all good points. Pietro? This is, I think, speaks to Rick Paulsen's constant feedback that we hear whenever we're talking about these papers is that, show me the data, prove it to me that this is actually real and not something that we feel. And Rick, I think this one's for you.

This is some nice data to help kind of add to this growing understanding that we can't compare races and ethnicities across the board and that there probably are some inherent differences with how we do through ART. But the novel aspect of this paper, because there's been dozens that show similar disparities, and so that's not new in and of itself, but it's the OSI, that Ovarian Stimulation Index, that they used. At the end of the day, when they looked at it, the AUC was 95% confidence interval was 0.51 to 0.61. So almost just better than flipping a coin, barely.

So that as a measure matters. But I have to imagine there's a ton of noise in that. Pietro, what's your maximum dose for a low responder that you'll use? I put my Cornell hat on and I really believe in the low dose approach to poor responding patients.

I do not flood them with medication. I will pretty much never go above a cumulative 450 IU dose a day for a patient unless their BMI is above, I don't know, 45. Then I will kind of go above and beyond that.

But for poor responding patients, I'm routinely doing 150, 150 and some oral Clomid or some oral electrozole. I give them what their ovary can handle, not any more, not any less. Yeah, I saw some patient, I was just running results for the book recently, and somebody else's patient, and they were on, I think, 450 of FSH and 300 of Menopur.

And, you know, they're getting like three follicles growing. At that point, the receptors are saturated. So I actually like your answer, but that changes the OSI.

You're going to get the same egg yield, but you're going to add a tropon dose in one patient's three times higher. So the OSI is going to vary greatly based upon what a provider's maximum dose is. And so I think that just adds noise to that OSI, which is probably why that area under the curve is basically a coin flip or just barely better than a coin flip.

But live birth is such a hard thing to predict. It's so complicated. I don't think we're ever going to get any single measurement that has to do with the IVF stimulation or something we do with IVF that gets us over like 0.65, because it's just too complex.

And no one thing is ever going to predict that. Completely. I think that stats 101, right? Like the outcome and the exposure are too distant from each other.

This is where you draw out a DAG and you notice that there are a bunch of things between that outcome and that variable. And I often just think it's silly to look at live birth outcomes when we're talking about stimulation changes or stimulation decision points. They're just too far removed for me to make me feel like they're going to make a difference.

I actually think the fact that the AUC is modest is part of the strength of the paper. If OSI was this extremely powerful predictor of live birth, then you would expect there to be more power behind ovarian responsiveness, which I think we conceptually know is not true. And I think the modest AUCs kind of support that ovarian responsiveness alone is not going to explain your observed disparity that we all know to be true.

I think that's exactly what you should expect from a single ovarian metric trying to predict this multi-step reproductive outcome. So I felt that I was glad that to see the AUCs as kind of that modest coin flip. That makes sense.

Well, congrats to the authors from Penn for another rigorous SART study. Pietro, we're moving on to reproductive science. We don't often get reproductive science unless it's of really high quality within FNS.

What was this paper? Why did it make it in? What did we learn? Shout out to our friends from up north, from Montreal and Canadian colleagues publishing infertility and sterility. This is a paper looking at how human embers divide and why they sometimes make mistakes. So we know that early human embers make a lot of chromosomal mistakes.

That's why mosaic and anti-polar embers are so common. We know that through IVF. But until now, we've had almost no direct look at how those mistakes actually happen because human embers for research are incredibly hard to come by.

They're not an abundant resource. This group did something pretty unusual. They took 91 donated human embryos and filmed them dividing under a time-lapse microscope.

And the result, I think, is one of the clearest pictures we have of where aneuploidy comes from. And they point the finger at a specific culprit, the micronucleus, which is something that many of our listeners may not be familiar with. So as a bit of background, we know that when a cell divides, it has to copy its chromosomes and pull one complete set into each daughter cell.

And the machine that does that work is called the spindle. And think of it like a microscopic tug-of-war. The protein cables grab chromosomes, pull them apart.

And when the spindle does its job correctly, both daughter cells get the correct 46 chromosomes. But when it misfires, you get aneuploidy. Too many or too few chromosomes.

Most of what we know about this process and embers came from mouse studies. But mouse and human embryos handle cell division differently. So direct human data here actually really matters and adds a little bit more information.

So these 91 donated frozen embryos from 26 patients were a mixture of day 2, day 3, and day 5 embryos. These authors used two fluorescent dyes, one that lights up DNA and one that lights up the spindle cables and filmed embryos dividing in real time. And here's a fun teaching pearl that I didn't know.

The dye that everyone uses for this kind of procedure in mouse embryos actually didn't work in human embryos. The team had to test alternatives and found a different one that did. And the lesson here for the fellows and the scientists are that what works in one model system doesn't automatically translate into another.

And you always have to validate tools in your actual system before you commit to a project. So this is something that they found out the hard way. They're like, oh, if it works in mice, we can just get it to... Nope, didn't work.

But what did they find? So their key findings. Human embryos divide differently than mouse embryos. So in humans, the spindles built from two organizing centers that come from the sperm and it assembles it in a recognizable orderly way, similar to most adult cells in our body.

Whereas in mice, early embryos build their spindles a completely different way. So a lot of what we thought we knew from mouse models may not directly apply to human embryos. Another cool thing that I really dug was that sister cells actually stay physically connected for hours after dividing.

So after cell divides, you'd expect to see two daughter cells to fully separate and separate quickly. In human embryos, they actually don't. They remain tethered by a thin bridge of spindle material for almost three hours after division, which is new, novel, interesting information.

And I think the clinical implication for this is if sister cells are still physically connected for hours after dividing, what might that mean for trophectoderm biopsy? Are we sampling cells that are still in communication with their neighbors? This, of course, is speculative, but it's, I think, one of the kinds of questions that this paper opens up for people who are reading it critically. And then finally, chromosomal mistakes are also very common, and they happen at every single stage. About one in four dividing cells at day two, day three had some kind of error.

About one in three dividing cells at day five had some kind of error. Errors happen in fast divisions and slow divisions alike. This concept of, oh, because it's rushing, it's more error prone.

That's not the explanation here. This actually shows us that expected along the whole journey of this embryonic culture. Finally, this micronucleus, the micronucleus they find is a major driver for aneuploidy.

Micronucleus is what forms when a single chromosome gets left behind during division. So instead of joining the rest of the chromosomes in the new nucleus, it forms its own, its own little mini nucleus floating in the cytoplasm. In human embryos, that micronucleus doesn't rejoin the main nucleus in later divisions.

It gets passed intact into just one of the two daughter cells over and over and over and over again. This means that one daughter cell actually ends up with extra chromosome material and the other ends up missing it, which is exactly how a mosaic embryo is born. Cool.

The micronucleus, major driver of aneuploidy. So I think if you're a trainee listening to this and you're like, well, what does this mean for me clinically? I think when you're explaining a mosaic PGT result to a patient, mosaicism isn't random noise. It has a mechanism.

And the most important one this paper identifies is that chromosomes get left behind during cell division. They form its own micronucleus. That micronucleus gets passed unilaterally into one daughter line and that embryo now has cells with different chromosomal counts.

That's a much more concrete story than telling patients that errors happen. And I think it's the foundation for understanding why mosaicism behaves the way that it does. So this is a really cool paper.

And Micah, I think said it, whenever you see a science paper in FNS, you know that there's something that we're all learning about it that's really practice informing. And I think opens up a couple of really great research questions for people who are in this space. Micah, Allison, what did you think? Did you also have this like, oh my gosh, what a cool study, feeling that I had about it? Yeah, the whole micronucleus thing that you walked us through was just, you know, opening my mind to new mechanisms of how things are happening.

My other thought was I would love to see this as a video article. I mean, this had to be a full text article because there's so much information, but I want to see these images. I want to see these, you know, time sped up.

So maybe they can present a sister study of this as a video article so we can see what this looks like. Yeah, we'll take it as a video abstract, guys. People like dyes and time lapse imaging.

Yeah. The point is video articles are not just for surgery. We've had some really cool video articles like one on IVM out of Belgium and how you do the retrievals and what the processes look like in the lab.

We definitely should have some basic science video articles. We definitely should have some embryology driven video articles. So if you're listening and you have something cool that you've seen and it's novel and it hasn't been published before as a video, submit that to us because we want to publish that.

Speaking of killer counseling papers, Allison, you have another one that if you are getting ready to study for your boards or to have a nice male factor population in front of you, the AZF deletion and the different subtypes of AZF deletions and how often are we to find sperm. Tell us a little bit more about this paper, Allison, specifically AZFC deletions. Yeah, so this was a research letter from several different institutions and it focuses on men with AZFC Y chromosome micro deletions.

So, as a reminder, AZFA and B deletions, which carry extremely poor sperm retrieval potential. AZFC deletions are unique because that residual spermatogenesis can persist. So clinically, these are the men where we sometimes see cryptozoospermia, intermittent zoospermia, fluctuating semen analysis over time and then found incidentally sometimes.

I know I've found a couple incidentally. So again, traditionally, many of these patients we think of as the ones that go on directly to microtest C once they have kind of a persistent AZ spermia in the setting of IVF. So, this is a retrospective single study 1996 through 2022.

They included 72 men with complete AZFC deletions who either had a zoospermia or cryptozoospermia and were planning on IVF with microtest C. The methodological point for this paper, really the premise of the paper, is that every patient provided a semen sample the morning of the surgery. So, if modal sperm suitable for ICSI were identified in the ejaculate, the microtest C was actually canceled. The primary outcome was simply how often the surgery could be avoided, which I think is a huge question.

They looked at predictors of ejaculated sperm presence and compared IVF outcomes between ejaculated sperm and testicular sperm. The headline finding was that about 28% of the patients had usable ejaculated sperm identified on the day of the planned surgery and 26% ultimately avoided their microtest C altogether. I think this is important to consider also that they didn't just limit it to cryptozoospermic men.

Even among the men thought to be persistently azoospermic, about 14% of them still had sperm identified in the ejaculate. Men with ejaculated sperm tended to have lower FSH levels and larger testicular volume, although those predictors became less informative when they were compared with men with any sperm retrieval, ejaculated or testicular, versus the retrieval failure. I think the fertilization data was also really interesting.

Ejaculated sperm had higher fertilization rates compared with testicular sperm, which I think to me biologically makes sense given greater sperm maturity, but pregnancy and live birth rates were ultimately similar. Clinically, I think the paper challenges the way we often think about azoospermia, especially in these AZFC patients. The authors are arguing that azoospermia may be more intermittent than we previously assumed.

Because of that, we should be repeating semen sample analysis potentially before surgery and ideally avoid a surgery we don't need to do. The paper also indirectly reinforces the importance of aggressive sperm cryopreservation strategies when sperm are identified in these patients. Availability is clearly fluctuating over time.

I think the strengths here are that they did have a relatively large cohort for a rare genetic subgroup. Management was fairly standardized and the intervention is obviously immediately clinically actionable. Limitations are certainly that it's a little retrospective, very specialized, so both a strength and a weakness.

This is a pretty elite center that's doing this, so how can you apply this to a smaller center that can make last-minute changes? I think clinically also relevant. Prior azoospermia does not always necessarily mean no ejaculated sperm will be present on the day of retrieval and that you have to have a microtest. After reading this, I'd actually be curious, as a fellow, I don't usually make these decisions myself at this point, but are there any centers that are doing this that are routinely obtaining a fresh sample immediately before the microtest in these patients or any azoospermic patients? Actually, we had someone who had set up their own new practice in the last year and was asking all of these questions about everything present.

Pietro, you trained under Dr. Schlegel. Have you implemented this? Or you may not have had a patient yet with this, I don't know. I lived through this as a fellow, and the way the Cornell tessie batches went is that patients were kind of started once a quarter as a large cohort of patients going through their IPF cycle, and there was basically 30 days in which the operating room was always up and open and ready to perform tessies as the female partners were ready to have their egg retrieval.

I think this speaks to the quality and the attention to detail that the folks got at the Cornell program. These are cash paid patients. These are patients who are paying out of pocket for procedures and are waiting a long time to get them done, and the center, I think, is willing and able to avoid procedures for patients.

For one out of four patients in this scenario who did not have to undergo anesthesia, did not expose themselves to the risk of hypogonadism post-procedure, this is really cool, and I think a testament to if you can do things in a patient-centered way and give them good outcomes. Logistically, it is incredibly challenging. Cornell was in a unique setup where the tessie operating room was across the hall from the retrieval operating room on the same floor sharing the same PACU.

The collection room was down the hall, and their ability to have someone collect, have the sperm analyzed, and then either walk back into the operating room or not is a kind of a unique setup. Most centers do not do all three of those things on site, and they fragment those bits and pieces. So we have not yet had an AZF mutation patient come through, but we work with a couple of really outstanding academic urologists in the area who are able to allow patients to have these more frequent checks before procedures to avoid surgeries like this.

So fascinating. I had never even thought about this. I think the other interesting thing is, so you mentioned that you can spare them a procedure 25% of the time.

If I was a man and you can spare me a testicular dissection, 25% chance, yeah, I'll take that. But the reverse was also true. Of those 25%, if they had to do a second cycle, 20% of them needed a tessie in that second cycle.

They did not have sperm. Even though they had sperm last time, they did not have sperm the next time, showing that it is a little bit intermittent or hit or miss. In other words, their semen analysis before or what happened last cycle doesn't always inform us what's going to happen the next time they come.

And I would tell you that the practice pattern at Cornell was everyone who had a non-obstructive azoospermia diagnosis who was undergoing microtessie was offered the opportunity to produce a sample the morning of. Because lo and behold, you often were able to cancel procedures. And you know that feeling you get when it's like Friday afternoon and they cancel that last case? Oh, it's intoxicating.

It's good for nurses, good for doctors, good. And here it's actually great for patients. So really, really killer practice setup for men getting pretty high intensity, high complexity care.

Yeah. Well, I learned something from this. I had never even thought about this or heard about this.

We do have an OR setup where we have two or three ORs going. And if we're doing fresh microtessie combined to the egg retrieval, that's going on in the other room. So we could do this in our practice.

You're going to have to open up your andrology at 630. So men will start producing specimens about an hour before the procedure happens. And yeah, exactly.

It's cool. It works. But you need scale for it and you need a team that's willing to be able to offer it for patients.

Yeah. Maybe it's the old adage of, you know, you buy a red car and then all of a sudden you only notice other red cars. But it does feel like this persistent theme in a lot of these articles is this underlying variability that we know to be true, but kind of have now put into evidence-based medicine into numbers of this.

Just because you look bad on one outcome does not mean the next one will be bad. Or, you know, good on one outcome does not mean the next one will be bad. And I think that's kind of pervasive through a lot of these articles today.

Love it. I love it when there's a theme to the FNS on-air podcast. Speaking of killer counseling articles, Micah, take us home on yet another one.

That is a niche population, but patients who are coming through with a PGTM indication and they have a hereditary breast cancer gene variant. How does their IBF cycle outcome look? That's a great question, Petro. And we have this addressed in a research letter, which is becoming rapidly my favorite section of the journal because it's a very digestible, quick, single question answered very succinctly.

So we have these homologous recombination repair genes, such as the BRCA genes, which obviously can get mutations in patients. And that can cause cancers. And the mechanism of that cancer causing is genetic instability.

And genetic instability could also lead to errors in meiosis. And that could, in theory, lead to more aneuploid embryos. And so we have a good patient population to test this in.

They're undergoing IVF for PGTM for their hereditary cancers. And so the hypothesis is maybe they'll have higher aneuploidy as well. Obviously, they'll have fewer usable embryos because they're going to have some you'll rule out from the PGTM itself.

But you also have random aneuploidy. So essentially, are these patients at risk of higher aneuploidy than that that is random based upon age? So this was a retrospective cohort study. They use next generation sequencing.

This was done from a group out in California and through Cooper Genomics. Shout out to Jenna Miller, who teaches our fellow and was the Cooper Genomics rep on this. She does a great job of looking at these kind of questions.

They also looked at mosaicism as well as aneuploidy. So less than 20% was euploid. 20% to 40% low level mosaic.

41% to 80% high. Greater than 80% was aneuploid. They looked at BRCA1 and 2. But there are others of these homologous recombination repair genes, such as TP53, ATM, PolB2, and Check2.

So they looked at those. We obviously know that age is the biggest driver of aneuploidy. And so they divided their patients by SART age groups to control for age.

So overall, 68,000 PGT cycles reported here from a large PGT lab. 450 of those met criteria for the study population, those who were being tested for one of those six genes that I mentioned. As you look at the SART age groups, aneuploidy is similar between those who had these mutations versus the overall larger general patient population until you get all the way to age over 42.

It was actually lower. And you actually see this trend. As they get older, the patients with these mutations, actually the distance between them and the PGTA normal group started to spread.

It became statistically significant over 42. So 55%, if they had one of these breast cancer mutation genes, aneuploidy over the age of 42, versus 75% in the general population. And I like that number because it fits right in with Jason Frenasiek's famous graph of aneuploidy with age right at 75 as you hit age 42.

Now, does that mean these are actually protective of aneuploidy? No, I wouldn't go that far. How many patients were in that group in the four? So they probably just had four patients that randomly had a good response. If you think of type one error, we're always setting it at 0.5 generally.

So it's not that error gets more likely with small studies or small ends. But it just means if you have one random event and one patient that's really good or really bad, and you only have four patients, that can really change your outcome. So I think that's just random noise, type one error just based upon the small size.

They did a sub analysis of just BRCA genes, saw the same thing, no difference between the two groups until you got over 42. This time, three of those patients were represented and had statistically lower aneuploidy with these breast cancer genes. Again, I think that's just noise.

So this is a wonderful paper to be able to counsel our patients who are coming through with these breast cancer genes to do PGTM. Yes, they're going to have fewer embryos because the PGTM combined with the PGTA is going to exclude more embryos. But their risk of PGTA aneuploidy is not any higher than any other patient population.

They should be reassured by that data. It's just their normal age-related risk. You know, kind of connecting to the papers, too, that in this paper, the mosaicism rates are basically similar.

And that, I think, relates directly to the spindle paper earlier, too, and potentially, again, reinforcing that some of those mosaicism is probably coming from some of those ongoing mitotic errors and micronuclear formulation rather than meiotic problems up front, which, again, I think is an interesting connection between those two papers, too. Absolutely. Well, this was a wonderful time to get and learn from you guys.

I actually learned a lot this month, a lot of things that I didn't know or weren't even on my radar. And some of these are very clinically applicable. Some of these are very practical from a counseling standpoint.

Allison, thank you for filling in for us on late notice, as others were unavailable. We really appreciate your insight and your leadership with the Editorial Fellowship. Oh, it was my pleasure.

Thank you for having me. And if you want a little bit more about one of these articles, like Micah said, the Journal Club Global was recorded just this past week and should be available online right around the time that this episode is dropping. It's another opportunity to dig in to what was a really killer article published in Fertility and Sterility.

Thank you for listening. Hit us up on social media with your thoughts, and we look forward to talking with you again next month. 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, Dr. Adriana Wong, Dr. Elena HogenEsch, Dr. Selena Park, Dr. Carissa Pekny, and Dr. Nicholas Raja.