When the Math Ain’t Mathing: How Too Many Mosaics Can Hurt Pregnancy Rates

In this episode of Taco Bout Fertility Tuesday, Dr. Mark Amols breaks down one of the most overlooked problems in modern fertility care: the hidden harm of overcalling mosaic embryos.

We begin with a surprising math puzzle from a 2023 study by Dr. Jonah Bardos in Fertility & Sterility, where the lab reporting the highest euploid rate also had the highest live birth rate—a result that shouldn’t happen if “more abnormal embryos were slipping through.” Instead, the math points to something deeper.

Dr. Amols explains what mosaicism really is, why true biological mosaicism is rare, and how PGT-A testing can create artificial mosaics from noise, amplification variability, and overly sensitive algorithms. When labs generate inflated mosaic rates—sometimes 10–20%—they push potentially strong embryos out of the euploid category and unintentionally create selection bias.

Through clear examples and memorable analogies, you’ll learn how calling too many mosaics doesn’t protect patients—it can actually lower pregnancy rates by changing which embryos get transferred first. You’ll also hear why high-level mosaics can be clinically useful, while overcalling low-level mosaics can distort decision-making and delay success.

If you’ve ever felt overwhelmed or confused by mosaic results on a PGT report, this episode will help you understand what’s signal, what’s noise, and why more information isn’t always better in fertility care.

Stay tuned at the end for ways to support the podcast and for a preview of an upcoming episode on high-level mosaics.

Thanks for tuning in to another episode of 'Taco Bout Fertility Tuesday' with Dr. Mark Amols. If you found this episode insightful, please share it with friends and family who might benefit from our discussion. Remember, your feedback is invaluable to us – leave us a review on Apple Podcasts, Spotify, or your preferred listening platform.

Stay connected with us for updates and fertility tips – follow us on Facebook. For more resources and information, visit our website at www.NewDirectionFertility.com.

Have a question or a topic you'd like us to cover? We'd love to hear from you! Reach out to us at TBFT@NewDirectionFertility.com.

Join us next Tuesday for more discussions on fertility, where we blend medical expertise with a touch of humor to make complex topics accessible and engaging. Until then, keep the conversation going and remember: understanding your fertility is a journey we're on together.

Transcript

0:01

Today, we're talking about when the math ain't mathing, and why in fertility, more genetic information can actually lower pregnancy rates. I'm Dr. Mark Amols, and this is Taco about Fertility Tuesday. I want to start today with a math problem. Not a medical one, but a math one. And when you hear it, your brain should immediately say, wait, that doesn't make sense. See, there was a study published in Fertility and sterility in 2023 by Jonah Bardos and colleagues. They looked at embryos created from donor eggs. Young, healthy donors. So egg quality was basically controlled. Same eggs, same embryos. The only thing that changed was which PGT. A lab analyzed the biopsy. Here's the math problem. One lab reported a much higher euploid rate, about 74%. Other labs report euploid rates closer to 50 to 60%. So, okay, fine. Different labs, different pipelines. But here's where the math stops mathing. That same lab, the one reporting the highest Euplant rate, also had the highest live birth rate per Euploid transfer. Think about that. Hmm? If that lab were just being sloppy, if they were calling abnormal embryos normal, pregnancy rates should have gone down, not up. You'd expect more failed transfers, more miscarriages. But that didn't happen. The pregnancy rates were better. So now we have a problem. Either one lab is magically better at embryos, or something about how embryos are being labeled is changing, which embryos actually get transferred. And this is where mosaic embryos enter the story. When the patients hear the word mosaic, what they usually think, they think the embryo is broken. Half normal, half abnormal. But here's the uncomfortable truth. When a PGT report says mosaic, we often don't know what it actually represents. That embryo could either be a normal embryo with some testing noise making it look abnormal. It could be an abnormal embryo with fewer normal cells. Or it could be a true mosaic embryo, and PGT can't tell us which it is. And when it comes to true biological mosaicism, it's rare. Best estimates have it around 2.5percent of embryos, not 10 to 15%. So we know that mosaicism isn't always real, and it's created by other things. Because we take so few cells, noise is unavoidable. The data. So in the end, all PGTA can show us is that that signal from 5 to 10 cells landed somewhere in the middle. But here's where it gets crazy. Some labs report mosaic rates of 10, 12, even 15%. One study even showed as high as 20%. So did biology change no. Did math change? Nope. So then why do some labs have higher mosaic rates? Well, it comes down to different things. It's all the same technology, next gen sequencing. But there can be differences in, for example, the way they amplify the DNA, even in the algorithms they use, or even the cutoffs. Essentially. Some labs are very sensitive to noise, and they label the noise as mosaic. other labs can filter it out. Same embryo, different answers. So usually a lab that's going to over call mosaics is usually going to be a low level mosaic. These are the ones that usually just implant fine when it had some noise around them. Now, here's where things get really important. Let's say you have two embryos. One embryo looks beautiful, great morphology, but it's labeled low level mosaic. The other embryo looks okay, it's not amazing, but it's labeled, euploid. Which embryo do most clinics transfer first? Well, of course it's the euploid embryo. But if many low level mosaics are actually normal embryos mislabeled by noisy testing, we just pushed a better embryo to the back of the line. Think about that. This is not a biology problem. This is a selection bias problem. Ah. And suddenly that study makes sense. The lab with the highest euploid rate wasn't hurting patients. They were probably just not downgrading good embryos into mosaic categories, which means euploid embryos they transferred were, on average, better embryos. If you ever heard my podcast about different gradings, I talked about the Gartner system, which is ABC grading, and then the Marius system, which has an ABCDEF grading. Now, what I talked about in that is that if everything had to be labeled by three letters, it's very difficult to separate things. For example, the best looking guy in the world is Brad Pitt. Of course, number one. Then I can't even be number two. At best, I can be a three because that's the worst category. No way I can be a number two. But if you change that grading one through six, and Brad Pitt's still number one, of course, then I get to be like a four. I'm not even a six anymore at the very bottom, because there's more range here where we can start adjusting it. Now, again, I might have given myself a little higher number, but who wouldn't lie about their score? Now let's use that same analogy. Now for embryos, let's grade all the embryos one through six. The best embryos are one. The worst embryos are six. Now we test embryos and we Find out which ones are euploid, which ones are aneuploid, and which ones are going to be mosaic. And for our analogy, we're going to add great personality, meh and bad personality. Now we're going to wipe all the bad personalities. I don't even care if you're a one, you're not going to be transferred. At that point you got a bad personality. But if you have a really good personality and you're one fantastic. What if you're meh? Well, we'll call meh the mosaics. Now, purely for this experiment, we're going to say Brad Pitt's personality is meh. Now, we all know that's a lie. He's also amazing, but we're going to say it's meh. Let's say number two is Tom Cruise. Good looking, but his personality comes back great. Well, what's going to happen? We're going to go with Tom Cruise even though Brad Pitt has a better score because we're not even going to look at the Mets because they're mosaic. And that is where it lies. The problem there is a selection bias. Mike, not using the those other ones because we're picking some factor like mosaic, we could potentially be deselecting some of the best embryos. And just like mosaicism, which is not actually true, we know personalities can be mistaken. And so of course Brad Pitt has a great personality. We were wrong the whole time. We should have put him back first. Of course he's the best. So here's the part that's going to make people uncomfortable. More information is only helpful if it improves decision. High level mosaic calls can actually help. They can prevent us from throwing away embryos. That might work. However, interestingly, which a podcast I'll do in the future, the actual chances of discarding an embryo as aneuploid, that's really high. Mosaic is very low. That will lead to a live birth with over calling. Low level mosaics can do the opposite. It can distort embryo ranking, delay transfers and lower pregnancy rates, even when we think we're being cautious. So when the math's not mathing, it's usually because we're mistaking the noise for signal and, infertility. That mistake doesn't just confuse patients, it can cost them pregnancies. I can't tell you how many times I've had patients want to use a different PGT company because they say, well, they're going to catch more of the mosaics and they think that's a good thing. Again, there's a time and place. But as you can see, having something called mosaic when it's not mosaic can actually lead to problems such as selection bias. So here's the takeaway. Mosaic embryos aren't the enemy, and genetic testing isn't the enemy either. The real problem is when we confuse precision with accuracy. Just because a test can detect a small difference doesn't mean those differences matter or that labeling them improves outcome. In fertility, the goal isn't to identify every abnormal cell. The goal is to identify the embryo most likely to become a healthy baby. When the lab over calls mosaicism, especially low level mosaicism, we don't just create anxiety. We change which embryos get transferred. And when we change transfer order based on noisy labels, we create selection bias. That's how you end up with a situation where the math ain't mathed and where the lab that calls more embryos normal actually gets better pregnancy rates. Not because they're reckless, but because they're not letting noise override biology. Remember, natural to the 5% in PGT A, we're seeing up to 15, 20% mosaicism. That's, not natural. So if you were a patient, the question isn't is mosaicism real? It is. The real question is how much of what we're calling mosaic is actually signal and how much of it is just static. Because in fertility, more labels don't always mean better care. Sometimes the smartest medicine is knowing when not to over interpret the data. If, this episode helped you think differently about your embryos or your treatment, that's exactly why I did this. The easiest way to support this show is to share it, rate it, or leave a review. Especially if you know someone who's feeling overwhelmed by genetic testing results. The more people who hear this, the more we can replace fear with understanding. I greatly appreciate everyone listening to this podcast and I hope you enjoy these episodes. If you like this episode, look forward to the next episode I'm going to do in the future about high level mosaics. And most of all, don't forget to come back next week on Taco Belt Fertility Tuesday.