Video - Medical Conference - Hugh S Taylor

I think there are a large number of surgeons out there, who in practice treat endometriosis and who really are not as familiar as they should be with the varied presentations of endometriosis, and who do not do the aggressive resections that are really necessary to completely remove the disease and prevent recurrence. We know that it appears in many different colors, and sometimes these different presentations are associated with various degrees of pain. It is not always these classic blue lesions that we see that are the painful lesions. These vesicular lesions are often overlooked, but can often be a source of pain and fertility issues. Red lesions are often overlooked. Again, no surprise to any of the experienced surgeons in this room that we can have these varied presentations. These white areas are most commonly overlooked and I often see them when I am referred a patient who has undergone surgery for endometriosis. These types of lesions have not been resected.

Our staging system is not really very accurate. The staging we have for endometriosis does not correlate with pain or fertility. I think we do not fully understand the underlying biology of the disease and that accounts for this to some degree, that the lesions that we score are quite high in the staging system that we have, are not necessarily the ones that are causing most of the problems. When it comes to infertility, there really is not a role for medical suppression. We know that medical therapy does not restore fertility, with the exception of one report that suggests that perhaps long-term treatment with Lupron prior to IVF may help improve IVF outcome. There are no other studies that suggest, other than in that setting, that medical treatment will be helpful in restoring fertility. Not only that, they are not only ineffective, but because they suppress ovulation they are just delaying fertility, and someone ages. The longer we delay fertility perhaps the more detrimental it can be in the long run. Clearly there is no role for medical therapy in infertility treatment endometriosis.

How about surgical therapy? Well, there have been two large prospective randomized trials that have been conducted, one in Italy, and this is the Canadian Study that I have here, the Endocan study. They looked at women with mild-to-moderate endometriosis and either did a diagnostic laparoscopy, or removed the endometriosis at the time of surgery and looked at subsequent pregnancy rates. In this study, you can see this looks at pregnancy rate on the Y axis and the X axis looks at weeks from randomization up to 36 weeks.

You can see that compared to a diagnostic laparoscopy that actually treating the endometriosis resulted in a pregnancy rate that went from just under 18 percent to almost 31 percent. Clearly a significant improvement, at least in which in this study.

The smaller Gruppo Italiano study did not show the same effect, but again it was a much smaller study. But when we think about that and put this in perspective, if we look at the pregnancy rate per month in these women, the control group had about 2.5 percent getting pregnant per month, fairly low, and after intervention, although we doubled that rate, it is still quite low, 5 percent per month is a very low pregnancy rate. Why is that? Even if very good surgeons remove, the disease – which may not be the right word! – why are they less successful than we would like them to be? And how does endometriosis affect fertility?

Well, clearly there are many ways that endometriosis affects fertility and I will not go into them all today. I want to focus today on some of the research we have done as to how endometriosis affects the uterus in the endometrium and endometrial receptivity, that I think is illustrative of why these surgical therapies perhaps do not work as well as we would like them to.

When we talked to you about a gene, there are many genes that are affected, but one gene that I have studied for the last 20 years, that is near and dear to my heart, is the HOXA10 gene. This is the gene that is absolutely required for an embryo to attach in the uterus. If this gene is not expressed, in a mouse if you knock out the gene, the uterus will be not be receptive to embryos.

Even normal embryos placed in that uterus will never implant, yet this mouse will continue to make very good healthy embryos that can be transferred into another mouse and implant just fine. It is clearly an endometrial receptivity defect. Just like in women with endometriosis, or many other types of infertility, if you look at the endometrium in these mice, it looks completely normal. If you look at it under a microscope the histology is completely normal. It is a molecular defect that we are beginning to understand, that I have been doing work on for many years.

But suffice it to say now that this gene is absolutely essential for an embryo to implant. In humans, we looked and it varies with the menstrual cycle and it goes up right at the time of implantation in the epithelial cells, and it is regulated by estrogen and progesterone. It goes up there with estrogen and then further up with progesterone after ovulation. When we look at it again across the menstrual cycle here, the X axis is the menstrual cycle, P1 and P2 are early and late proliferative phase, S1, S2, S3 are early, mid, and late secretory phase, and the time of implantation is right here, mid-secretory phase.

We see this increase in the expression of this gene in women. It looks like it is playing an important role in that implantation process in women. We looked at its expression when we finally got a good antibody that would work.

What most of this increase is is that it turns on in the glands. The brown staining of the nuclei is the expression of this gene, it is in the stroma through most of the menstrual cycle but the glands here, you can see right in that window of implantation, this gene has turned on. So that increase you see is really mostly a big jump in expression, complete off to on in the endometrial epithelium that makes it receptive to the embryo.

Well, what happens in women with endometriosis? Here is the study we published many years ago where we looked again at the expression of this gene through the menstrual cycle. The X axis again is going across the menstrual cycle. The blue bars are normal fertile controls, and as I showed you earlier, right in the mid secretory phase, this expression goes up. But in women with endometriosis, shown here in red, that increase in expression fails to occur, at least not to the some degree in all women when they have endometriosis, suggesting that there may be some sort of implantation defect in these women with endometriosis, as has been shown in some clinical studies. There may be an endometrial receptivity defect.

We turned back some of our animal models to really get at the cause and effect, what is causing this defect that we see in these women. We looked in a mouse model, and in a nonhuman primate model, to explore how this works. First we made endometriosis in mice in much the same way that I described to you earlier when we were looking for stem cells migrating to endometriosis. We just take endometrium and we sutured into the peritoneal cavity of these mice, and make experimental endometriosis. These are mice that did not have pre-existing endometriosis; mice do not spontaneously generate endometriosis. This was normal endometrium we put in their peritoneal cavity. When we went to look at gene expression in their uterus, we saw that it was dramatically affected. Just normal endometrium placed in the peritoneal cavity has an effect on the uterus. You do not have to be somehow predisposed by having an abnormal endometrium to begin with.

These were normal mice to begin with. Just having endometrium in your peritoneal cavity causes changes in the gene expression in the uterus. Why that is, what the signal, is we are not sure, but that same gene that I told you about earlier, HOXA10, is decreased when we cause these mice to have endometriosis in the uterus. These mice then had a lower implantation rate, and lower pregnancy rate. We looked at several other genes that are involved in implantation as well, another HOX gene, IGFBP-1, all dramatically decreased when we caused endometriosis.

So, just causing endometriosis made the uterus less receptive to embryos, but why would that be? I told you this gene is regulated by estrogen, and progesterone, and women with endometriosis typically have fairly normal estrogen and progesterone levels. That is not the defect. If this gene is regulated by estrogen and progesterone, why would not it be expressed at normal levels in women with endometriosis?

So, we looked at epigenetic regulation of the gene. The normal regulators are there, why isn’t it responding? I will explain for those who are not familiar with epigenetics, a little background as to what that is. You know we used to think that everything about heredity had to do with just DNA sequence. The nuclear-type mutations really are what caused disease or genetic disease. Disease susceptibility was really only involving the DNA sequence. If you did not mutate it, you had the same DNA. We now know that, that is a vast oversimplification, that DNA can be modified, even in adults. DNA on CG base pairs can have methyl groups added, which can influence how the gene behaves, how it is expressed. The histones which package the DNA can also be modified. It will change the accessibility of that DNA, how readily it can be turned on and expressed, and this is illustrated here. We have again the histone and the DNA. If we add methyl groups or if we modify the histone tail, it causes these histones to compact the DNA more, make it less accessible, less likely to be turned on. It is not expressed as well. So, we wondered if these types of changes might be occurring in women with endometriosis in their endometrium.

Again, just to show you how important epigenetic changes are, I put in this slide.  Not only because it is beautiful, the butterfly, but it really illustrates an important point that that caterpillar that gave rise to that butterfly, they have the very same DNA, the very same sequences. It is this type of epigenetic difference, this is the very same organism with the same DNA, and it is these epigenetic changes that can have dramatic differences. We think again that, that is probably what is happening in the endometriosis patients.

Our HOX gene, again the one that I am using here is an example, but there are other genes involved as well, has lots of these CG base pairs that are clustered together in what we call CpG Islands. These are areas that have a lot of these base pairs and are very susceptible to these epigenetic modifications. We looked at our mouse model, looking here at methylated genes in the mice with endometriosis, this was mostly methylated.  The unmethylated was almost nonexistent. In the controls, there is a small number that are methylated, mostly unmethylated.

So, it shows that is regulating our mouse model. When we did this again a little more quantitatively, you can see it is a dramatic difference in the amount of methylation. In our baboon model, getting a little closer to humans, we created endometriosis again by taking endometrium or menstrual debris, and putting it in the peritoneal cavity. These animals develop lesions that look very much like what see in human’s endometriosis. Turn off the same gene the expression goes down over time.  It took months for that to happen. This happened, it is 12 months, it just really reached statistical significance by 12 months.

This something that progresses slowly, probably very similar, we do not diagnose endometriosis in women probably until it has existed for at least a year, probably a number of years. But this type of decrease in expression was occurring in this monkey model. We looked at the methylationist gene for epigenetic modifications. We looked at those three clusters with CpG Islands that I showed you. In one of them here that we call F1, here are the animals that had endometriosis, here are there are animals that were our controls that we did not induce endometriosis. The dark black shows an area that is fully methylated, the gray shows an area that is partially methylated, and essentially no methylation on the controls. There are a couple of these CpG Islands that were heavily methylated in the animals with endometriosis. So somehow creating that endometriosis in the peritoneal cavity causes these epigenetic changes in the uterus that inhibit implantation. Sun-Wei Guo’s group in Shanghai has looked at this same process in women with endometriosis, and showed indeed that women with endometriosis have increased methylation of this gene, which would lead to the lower expression that we had reported earlier.

So, why do our surgical therapies not work as well as we might like them to? Again, we can double the pregnancy rate per month, but it still does not restore them to normal. I think part of that reason is that there are these epigenetic changes, which are permanent, or relatively permanent.

Once that DNA is modified, it stays that way. We can be removing all of the endometriosis, and yet that change in the uterus will not revert simply by treating the endometriosis, really removing it all. We can still have a residual fertility defect. How should we optimize fertility treatment? Well, certainly again, based on the data from the Endocan study, without treatment the rates are low.

Surgical treatment, again we can double those pregnancy rates, but we have other fertility treatment that we often use that can considerably increase pregnancy rates. When somebody with significant endometriosis wants to conceive, I think we ought to talk about using all of the available things that we have in our armamentarium, including ART, and surgical treatment both, to help augment pregnancy rates. Surgical treatment may be not be enough to completely restore fertility.

I probably have run over time. ______is for the most part irreversible epigenetic changes in the endometrium. It is a chronic disease that the endometrial effects may not be curable. Again, we can remove the disease, but these may be long-lasting. I do not want to end on this pessimistic note, and will tie this back to my last talk. Even though we cannot reverse the epigenetic changes in the endometrium, we are looking now at ways to recruit new stem cells into the endometrium to replace that damaged endometrium. Perhaps we can get new cells in there that do not have this epigenetic damage and we can restore that fertility making endometrium more receptive again.

Again, I would like to thank the people in my laboratory who have done all this work and the NIH for support.

Thank you.