Endofound Medical Conference 2017
"Breast, Ovary and Endometriosis"
October 28, 2017 - Lotte New York Palace Hotel
“The Abnormal Embryos That Aren’t”
Norbert Gleicher, MD , FACOG, FACS
Center for Human Reproduction
Thank you, Dr. Seckin, for inviting me. Thank you to the foundation. You heard a pretty remarkable story here. I'm going to give you the scientific side, especially if I can show a few slides
It's just this green here over here ...
I have conflicts but none in regards to today's topic. What you just heard is a hypothesis. What you heard is a fantastic hypothesis. But what you just heard also re-emphasizes that the hypothesis does not yet make facts. In many ways, that is probably one of the most important lessons to learn from what I will be telling you, because the history of PGS, preimplantation genetic screening, now renamed PGTA, preimplantation genetic testing for aneuploidy. Has been around really clinically for over 15 years without ever proving the hypothesis.
We are now on what I call a version three. Version one, we used to biopsy embryos at the cleavage stage six to eight cells. We did that for a few years. The whole industry grew around this procedure. Then, the American Society for Reproductive Medicine and SRE finally, after many years of use, came to the conclusion that what PGS really had promised, namely that eliminating chromosomally abnormal embryos prior to transfer would improve IVF outcomes for the remaining embryos. That hypothesis was not proven by that first form of PGS, which I call PGS 1.0.
Once the ASRM came up with the opinion that PGS was not effective in improving IVF outcomes and reducing miscarriages, the newly grown PGS industry very quickly found PGS 2, because in the meanwhile, some of the diagnostic platforms had improved. The idea was taking now in PGS 2 the embryo biopsy from cleavage stage at day six to eight to day five, six blastocyst stage. You could get more genetic materials. Therefore, with better platforms, the hypothesis was we would get more accurate results.
But what in all of this was forgotten was the basic biology of the embryo. Even though when I first heard the PGS hypothesis from Yury Verlinsky who really developed it, I was fascinated because it seemed to make such logical sense. I hired a PhD and we created in those days, I was in Chicago, probably one of the early PGS laboratories in the country.
The longer we were working with the concept, the more obvious it became to us that the biological background, in other words what an early preimplantation stage embryo really is, cannot ever fulfill this hypothesis. The reason being that we can simply not determine by a single embryo biopsy whether on day three cleavage stage or day five, six blastocyst stage, whether that embryo is really chromosomally normal or not. I will quickly explain to you why that is.
I also want to make clear to you that if that one biopsy cannot tell us reliably whether the embryo is normal or not, we also cannot reliably make the decision whether to transfer that embryo or to discard it. If we cannot make reliably that decision, then there are only two consequential possibilities. Either we have a lot of false positives, and that really means that we are throwing out a lot of healthy embryos, or we are putting a lot of bad embryos in.
As I will show you, the reason why this has become so damaging, in my opinion and in now the opinion of increasingly many other colleagues, is that we have been, as you just heard for almost two decades, discarding tens, if not hundreds, of thousands of healthy embryos. That is not only bad medicine because, obviously, we all practice under the first rule of clinical practice, which is do no harm. But this is particularly tragical because the biggest harm came to the patients who needed our help most, in other words, to older women, to women with low ovarian reserve who have few embryos.
If we throw out a few normal embryos in a young woman who has 15 beautiful embryos and she still has 10 beautiful embryos left, she will get pregnant. But if we throw out embryos in a case like this, where it took many, many cycles to collect 15 embryos and all of them turned out supposedly to be abnormal, that patient's chance to get pregnant with her own eggs is gone. That's what has been going on unfortunately for a long time and unfortunately still going on because the PGS industry is still trying to push the procedure.
Now, quickly, the biological explanation. I told you already the history. We have gone through various phases. The person who really raised the concept first was Yury Verlinsky in Chicago, was a close friend of mine. He unfortunately passed. Then came PGS 1. You see the years, how long this is already being used clinically. Then came PGS 2.0 where we moved to blastocyst stage biopsy and where, in the US, the procedure really exploded in utilization. Some data suggests that in 2016, over 20% of all IVF cycles in the US were accompanied by a PGS procedure.
Now, we have PGS 3.0 not because our genetics colleagues who are pushing this procedure changed their minds, but because their position was no longer maintainable once we and colleagues started reporting healthy babies born after transfer of those so-called abnormal embryos. They suddenly, in July of last year, changed their reporting policy. Until July of last year, every PGS report was either-or. It was either chromosomally normal or it was chromosomally abnormal.
If it was chromosomally normal, you transferred. If it was chromosomally abnormal, you discarded embryos. That was it. There was no no other solution. We transferred our first so-called chromosomally abnormal embryos in 2012. Our first healthy baby was delivered in 2013, quietly. We, in 2014, went public after we had two healthy babies. It's when Dr. Braverman and Dr. Vidali in New York joined us and started doing the same thing.
We have today over 10 healthy pregnancies from roughly 16 or 17 transfers. But worldwide, we have already close to 100 healthy pregnancies from transfers of so-called chromosomally abnormal or now sometimes also called mosaic embryos, because with PGS 3 as of July last year, suddenly the PGS industry is reporting three possibilities. Embryo can be either normal, can be what is called mosaic which means that that one biopsy that is obtained, the five to six cells, contains more than one cell line, one of which is abnormal, or can be aneuploid.
That's already a progress because the PGS community at least allowed for the possibility that in those mosaic biopsies, you could transfer those embryos, or at least some of these embryos. The only problem with PGS 3 is that like the two prior incarnation of PGS, the selection, the criteria for the diagnosis are completely pulled out of hot air. There is no validation for how the industry defines a chromosomally normal embryo and there is no validation for how the industry defines a chromosomally abnormal embryo.
Paradoxically, the new classification defines a normal embryo as having less than 20% DNA that is aneuploid. Now, superficially, you might think that makes sense. Very low aneuploidy in your single biopsy? Okay, that's normal. But the reason why it is 20% is not because nature tells us something, tells us that if it's under 20%, we are having healthy babies. If it's over 20%, it's less likely to have a healthy baby.
The reason why it's 20% is because all the diagnostic platforms but one until now cannot even detect mosaicism in a single biopsy. There's only one platform, the most recently used plan which is called NGS, next-generation sequencing, which can detect a second cell line in your biopsy. But it can detect it only above 20% DNA load. There's no biological logic for saying what is normal and what is mosaic.
The only reason why there is a cutoff between normal and mosaic is because 19% is not detectable, but 21% is detectable. To follow those new guidelines, a biopsy sample with 19% of DNA load that is abnormal would be considered "normal" and can be without restriction transferred. But if the biopsy specimen has 21% DNA load, you should be probably discarding it. But maybe under certain circumstances, you can transfer it.
It's even more bizarre on the other extreme. I told you there are now three classifications, normal, mosaic, and then there's again aneuploid. Aneuploid is defined by over 80% of DNA aneuploidy in a single biopsy. Where is that 80% coming from? The 80% is coming from the following logic as put in print by one of the most prominent PGS experts in the country, because until he explained it, nobody knew where the 80% comes from.
Here was his explanation. He said the average biopsy contains five cells, the average embryo biopsy, the average trophectoderm biopsy. Therefore, 80% represents for cells. Makes sense, right? It doesn't, because first of all, the medical literature suggests that the average trophectoderm biopsy really has six to seven cells. If you calculate 80% there, you find yourself already far below the 80% cutoff in the mosaic range. Okay?
But every embryologist who has ever done an embryo biopsy at blastocyst stage, a trophectoderm biopsy, will tell you that we never know how many cells we have biopsied. Moreover, from the first day on that we have been doing embryo biopsies for this purpose, we know that the biopsy itself breaks up cells and cells lose the DNA content. That's the reason why not infrequently, a cell or an egg is reported as no DNA. We cannot measure DNA volume from a biopsy.
To put a diagnostic cutoff point at 80%, simply paste on the theory that that represents one other five cells is ludicrous. It's simply unbelievable that the commercial diagnostic laboratory would set its criteria based on absolutely nothing. But that is exactly what has been happening and what is still going on, unfortunately.
Richard Paulson, who may be familiar to most of you is the current president of ASRM, was also extensively interviewed in this article in New York Magazine. He recently, infertility, sterility, the references here, presented the mathematical model in which he took existing data and made the conservative estimate of what the number of probable false positive results in our current PGS practices.
As you can see here, and I'm quoting him verbatim, he came to the conclusion that we may have been and may be throwing out up to 40% of completely normal embryos. Now, our opinion at our center, which has been following and working in this arena now for 15 years, is that this is a conservative estimate. We think that the number is probably closer to 50%. The reason ... Oops, so sorry for that.
A call from home.
This must have been one of the PGS proponents. The reason why this whole PGS concept really cannot work is best shown in this little drawing. You see on the left the scheme of a blastocyst. This is the outer layer which is kind of like a football. This is called the trophectoderm. The trophectoderm, however, has nothing to do with the baby. The trophectoderm becomes the placenta.
The baby itself comes from those cells, which is called the inner cell mass. Now, we are biopsying the trophectoderm and we are taking out of this trophectoderm, as I already explained, on the average six to seven cells, out of a trophectoderm that may probably contain somewhere between 120 to 150 cells at blastocyst stage, depending what we are biopsying the embryo.
First of all, we are basically biopsying the early placenta to find out something about the early baby. Now, we have known for 30 to 40 years the term pregnancies, completely normal term pregnancies. If you cut the placenta and put it under the microscope, has little islands of chromosomally abnormal cells. This is common knowledge. What varies in the literature is not the fact, just the percentage. In general, you find it in roughly 25% to 30% of placenta which, in my opinion, is probably an underestimate.
We are biopsying the early placenta to find out information on the early baby, even though we know that even the mature placenta still can be aneuploid. Maybe more importantly, there is now, I'm so sorry, there is now convincing data to suggest that this early trophectoderm in a very high percentage, and maybe universally maybe as a universal physiological function to help implantation, is frequently characterized by aneuploid islands of cells. Mosaicism, in other words. Mosaicism is extremely frequent in preimplantation stage embryos.
Think about it. We are taking five, six, seven cells at random from one point and making, based on that, one decision, one little biopsy of a few cells a decision about the whole globe. I can tell you and we published that, we had mathematicians at the Rockefeller Institute build models for us, who told us definitely that you can never, even on the best of circumstances, from a five, six, seven cell biopsy make conclusion on the presence of aneuploidy, of aneuploid clones anywhere else on a 130, 150 cell globe. Impossible mathematically, yet we are doing that.
Thirdly, there are studies both in animals and humans that show while there is in general reasonable similarity between biopsies in trophectoderm and in a cell mass, there's about the 25%, 30% discrepancy. Once again, we are biopsying the placenta to get the result here, and there is a significant discrepancy, but probably the most important point but all of this doesn't work. Probably also the most controversial point is that animal models have beautifully shown, and those experiments are obviously difficult to do in humans, the downstream from this blastocyst stage a few days down the pregnancy.
Particularly the inner cell mass has an amazing ability to self-correct. In other words, if there are indeed aneuploid cells in the inner cell mass, they commit suicide. It's called apoptosis. You heard about it by the prior speaker. The healthy cells continue growing. I think this is one of my next slides.
Here, you see the result. What you see here obviously in the mouse experiment is that if the inner cell mass is 100% aneuploid, chromosomally abnormal, there's no survival. If the inner cell mass is only one-third, meaning one-third of cells are normal, two-thirds are abnormal, you already get one-thirds of healthy pulps. The moment the inner cell mass at blastocyst stage is 50 or more percent normal, you have 100% healthy pulps. The same investigators also looked at the mechanisms. As I explained before, they found that there are pathways active in the inner cell mass that, as I said, further apoptosis and suicide of abnormal cells.
Now, how come then that some people are still claiming that PGS works? Even though most of the major proponents of PGS no longer claim it improves clinical pregnancy and life birth rates, there are still some who do. They constantly refer, and I'm sure you will see that in the marketing materials from the various genetic laboratories, they refer particularly to four studies, in the literature that were supposedly done as form of clinical trials and showed that PGS works. Those studies don't exist. The studies exist.
Those studies, with one exception which had other significant shortcomings, they used as a reference point embryo transfer, meaning they didn't tell what the clinical pregnancy or life birth rates are based on cycle start, but based on a woman reaching embryo transfer. They reported wonderful results. Of course, they reported wonderful results, because look what is happening when you do that. You start, let's say, with a general population of 100. Then, you're selecting your study because you don't want to take patients with too high FSH and too low AMH, so you select out against those.
Then, you culture your patients to blastocyst stage, extended culture. A lot of patients, particularly the poor prognosis patients, their embryos don't make it to blastocyst, so they don't show up in the final analysis. Then, on top of it, you do this PGS procedure and sometimes there won't be any normal embryos. Let's assume that you start with 100, and out of those 100, only 10 make it to embryo transfer. Eight of those implant and have a life birth, 8% clinical pregnancy rate.
But look at what it really is. 100 started, eight life [inaudible 00:27:55]. That's an 8% clinical pregnancy rate. This is how the PGS community has been propagating false outcomes now for almost 20 years. I encourage you to go back to the individual papers when you hear somebody making the point that indeed that were prospective randomized studies that proved that PGS improves clinical outcomes in IVF. This is the article and that appeared in New York Magazine. I thank you very much.