Endometriosis Foundation of America 2014
Endometriosis and Cancer: Is There a Connection?
- Robert Kurman, MD
Well thank you very much Harry for the nice introduction. I would like to thank Dr. Seckin for inviting me to come here. It is a real pleasure. I find myself in a bit of an unusual situation in that I cannot remember the last time I have given a talk where most of the people, at least many of the people in the audience are not physicians. In putting the talk together I tried to structure it so that it would be aimed at predominantly them recognizing that there were going to be some physicians in the audience. I hope it is not too complex for people that are not physicians and too simplistic for the people that are physicians. That may be a bit tricky.
It reminds me of, in terms of maintaining your interest, it reminded me of a story as I was putting this together of a scientist who is being shepherded through the afterlife by an unworldly figure. The scientist is completely disoriented; he does not know if he is in heaven or if he is in hell and despite multiple inquiries this guide remains silent. He is finally led into a room, it looks like a library. He sees a beautiful young woman seated there reading his scientific papers. He smiles, he turns to his ghostly guide and says, “Hah! Now I know where I am, this must be heaven. What a reward to spend eternity in the company of this beautiful woman surrounded by my life’s work”. The ghostly guide turns to him, the figure turns to him, and he says, “No, this isn’t heaven. This is hell. And it’s not your reward it’s her punishment”. So I hope you do not find at the end of this talk that you are in the company of, the fate of that poor woman.
Endometriosis and ovarian cancer: You have heard a little bit about it already, is there a connection? The way that I have done this is I talk a little bit about endometriosis, which I know many of you are already familiar with anyway. Then I will talk a little bit about cancer, ovarian cancer and then try to put the two together.
In terms of endometriosis, and I will be presenting it to you really from the standpoint of my background as a pathologist. It affects about 10 percent of the female population. It is defined as the presence of endometrial tissue outside of the endometrium and myometrium. The endometrium is the lining of the uterus and the myometrium is the muscular wall of the uterus. It is somewhat controversial but I think, I know Harry did not seem to agree with it, most people I believe do think that endometriosis develops from retrograde menstruation. Having said that there are ways that it must develop with other mechanisms as you will see in a few minutes, I do not think there is one explanation that answers it all.
This is some interesting work done by a molecular biologist that took endometrial tissue, the lining of the uterus, from women with endometriosis and compared it to looking at molecular analysis of women’s endometrium that did not have endometriosis. It was found that there were certain molecular alterations in the endometrium of the women with endometriosis that were not there in the women without endometriosis. This suggests that there is something abnormal about that endometrium of the women who have endometriosis that allows that endometrium to develop into endometriosis by retrograde menstruation. It explains why most women obviously do not have endometriosis despite having retrograde menstruation. I think it is a very fascinating area that really needs to be investigated further.
To just again show you a little bit of the anatomy, the lining of the endometrium, the myometrium, the fallopian tube with the finger like extensions at the end called the fimbria, which envelope the ovary at the time of ovulation. You see the ovary bisected to show the developing follicle where the egg develops and gets extruded into the fallopian tube. You have seen much more beautiful videos of retrograde menstruation by Dr. Seckin than I am showing you. Suffice to say that the retrograde, the blood, comes out of the peritoneum and into the peritoneal cavity. From there it can go on to implant on the ovary, the cul-de-sac and many other sites within the abdominal cavity, the pelvic and abdominal cavity. I am just listing some of the site, rectovaginal septum also refers to the cul-de-sac, the uterine serosa and it can be on the bowel serosa; as you know a variety of locations.
This slide shows you where the relative – I did not give you the frequencies – but basically most endometriosis, as you know, involves the ovary and the cul-de-sac or rectovaginal septum and that can easily be explained by the mechanism of retrograde menstruation. I think it probably does explain those cases. But as you just go down the list you can see for example lungs, nervous systems, soft tissue and bone, there is no way that retrograde menstruation can account for endometriosis in those locations. There has to be another mechanism by which endometriosis develops but I would venture to say that the vast majority does develop by retrograde menstruation. Just a couple of pictures to show you, you have seen nicer ones before, of an endometriosis in this case involving the ovary. Here it is involving the bowel.
So now let us shift gears and talk about ovarian cancer in very general terms, after that we will link it with endometriosis. It is without a doubt the deadliest gynecologic malignancy as you see here the fifth leading cause of cancer death among women in the United States. There are roughly 22,000 new cases every year in the US and approximately 14,000 deaths in the US. I had a lot of…to show you how to contrast ovarian cancer to these other major forms of cancer where the mortality of these cancers – it says 15 straight years since 2005 – the mortality has been decreasing and for the most part still continues to be that way. But you will notice absent from this list is ovarian cancer. This shows you again it can be extended out further to 2014 even though that slide ends in 2000. The overall mortality rate of ovarian cancer despite the addition of radical surgery, and the other things up there listed are various types of cytotoxic chemotherapy, has really not appreciably changed over all this time.
Now the good news is that women live longer as a result of these various surgical and medical techniques but eventually the overall mortality rate has not changed; about 30 percent survival with ovarian cancer. There is so much more attention focused on breast cancer largely because as you see here it is so much more common than ovarian cancer. But when you look at the number of deaths from these two types of cancers over the number of cases you will be struck by the enormously greater lethality of ovarian cancer compared to breast cancer. And that is another problem. Because there are so few survivors of ovarian cancer that you do not get the kind of advocacy that you get from breast cancer survivors. There are just not enough of them to go around and really encourage support for this very deadly disease.
What is the pathogenesis? By pathogenesis how this disease develops over the last 50 years. The thinking is that ovarian cancer is a single disease and that it begins in the ovary and spreads from there. Well there have really been some dramatic changes in that concept in the last, oh maybe, five to ten years and they are as follows: Ovarian cancer is now recognized to be a group of diseases a family of diseases. They are related but yet have very distinctive clinical, pathologic and molecular genetic characteristics. That has important implications in terms of detection, diagnosis and treatment. The newest findings strongly suggest that the tumor develops, it is not proven it is still somewhat controversial. But there is more and more compelling data to indicate that it begins in the fallopian tube or the endometrium and involves the ovary secondarily and this again has important clinical implications.
This is the family of diseases that constitute ovarian cancer: Serous carcinoma on the top. Dr. Seckin mentioned he had a patient with that recently. That accounts for certainly the majority of ovarian cancer and it accounts for the vast majority of deaths, maybe 90 percent of the deaths from this disease are from this particular cell type. After that you see clear cell carcinoma, endometrioid carcinoma, and mucinous carcinoma and then there are a whole lot of others that are much less common that I did not even list. You can also note that clear cell carcinoma and endometrial carcinoma are the ones that are really very closely linked to endometriosis. Certainly you might see a patient as Dr. Seckin mentioned who had serous carcinoma and endometriosis but that is likely a coincidental finding and I doubt that the two are related whereas the data to implicate a relationship between endometriosis is really very strong for endometrioid and clear cell carcinoma. Parenthetically, I mention endometrioid just to make sure you understand that this is a carcinoma that when you look at it under the microscope looks very much the same as cancer that arises in the uterus from the endometrium. But since it is in the ovary and not in the endometrium but it looks like a tumor in the endometrium it is called endometrioid.
Going on with this – so now we are going to look at the association of the ovarian cancer with endometriosis. Women with ovarian cancer and endometriosis compared to women with endometriosis but not carcinoma tend to be younger, often premenopausal, obese and have a history of unopposed estrogens. The endometriosis associated carcinoma is more often lower grade, lower stage – in other words it has not spread as far and therefore has a better prognosis than those that are not associated with endometriosis. Continuing – four percent may be a little bit high. I do not want to give you the impression that if a woman has endometriosis there is a real significant risk of developing cancer. There is but it is really pretty low. That includes both endometriosis involving the ovary or the pelvis. I think these figures are a little high. I got that from the chapter in the textbook that Harry mentioned, the Blasting textbook but I did not write that chapter, I think it is a little bit high.
The precursor lesion – where did the cancers actually develop? We talked about endometriosis that most of them do not develop into a cancer so what is going on here? We think that the immediate precursor is atypical endometriosis. By atypical, and I will show you some slides, the cellular changes look significantly worse than “normal endometriosis” if you will, normal endometriosis but not as bad as a cancer so it is somewhere in between. Atypical endometriosis is found in 60 percent of ovarian cancers associated with endometriosis but on the other hand it is only found in two percent of endometriosis without carcinoma. Again, providing some support to the concept that it is the atypical endometriosis that is really the precursor of these types of cancer.
And just to show you a gross example of an ovarian cancer. This was a clear cell carcinoma of the ovary. It has been opened and it is rising in a cysts, it is almost invariably developing endometrioid cysts, endometriotic cysts, endometriomas. Here you see a cartoon over the actual microscopic view. So that all that stuff on the bottom is the cancer and it is arising in the endometriotic cyst. And the yellow box shows you that the endometrium that separates the cancer on the right from the normal, if you will, endometriosis on the left is an area of atypicality. Again, it is a transition zone between the normal, between the endometriosis and the actual cancer itself.
These images, and I know there is probably not another pathologist in this room other than me, and the surgeons may be a little bit more, I am sure, familiar with this than the lay audience but suffice it to say if you look at the upper left corner that is normal, if you will, endometriosis. And I draw your attention to that very superficial layer of epithelium on the top and as your eye goes to the field on the right and the lower fields you will notice something different, there is more of it, it looks a little different and that is atypical endometriosis. In fact, it can get so atypical, that is blood in all the red stuff, red blood, they kind of go together, that strip going diagonally across the field is the epithelium of either the most atypical endometriosis, or even for a pathologist who is skilled at looking at this it is hard to distinguish from maybe the earliest form of cancer. Clear cell cancer that has not yet invaded the underlying tissue, it is that close. So you really get the sense of a transition.
Again, just to go over our cartoon the retrograde menstruation leading to endometriosis, the endometriosis in turn leading to atypical endometriosis, the atypical endometriosis then going on to develop either a clear cell or endometrioid carcinoma. Undoubtedly hormones and inflammation will play a role with these cancers as it does with the other types of ovarian cancers as well.
There are some very interesting molecular genetic findings now that have been reported in the last few years. Ultimately, all cancers are genetic in origin. There are abnormalities that occur in our DNA that lead, as a result of mutations and other changes, that ultimately lead to the development of a cancer, This is where the whole field is going, trying to understand those molecular genetic alterations because once you understand those changes then you will know how to detect disease in a much more accurate way than we do right now and even more importantly, treated appropriately.
These are some of the molecular changes that have been identified. This mutation of ARID1A is a very important one that was recognized by workers in Vancouver and in our group at Hopkins. ARID1A is called a tumor suppressor gene, again I do not have time to go into all the details of it. Suffice it to say that about 40 to 50 percent of clear cell and endometrioma carcinomas contain that mutation. The really fascinating thing that we found is that you can see the identical mutation that you see in the cancer in that atypical endometriosis right next to the cancers, I showed you in that previous cartoon. Interestingly enough, not in the endometriosis that is more distant. In normal endometriosis you do not see the mutation and in atypical endometriosis you do. This goes along with the whole concept of carcinogenesis, which is an accumulation of a certain number of mutations that ultimately take you from the normal, through the atypical, the borderline to ultimately carcinoma. Again, it provides compelling evidence that atypical endometriosis is the immediate precursor based on mutations say of this particular gene. There is typically more than one gene, there are usually five, six, seven genes that need to be altered or mutated for a cancer to develop.
This schematically tries to illustrate what I just said. On the left you have typical endometriosis, normal endometriosis, if you will, then atypical endometriosis, then a very, very low grade kind of tumor. It is not really quite cancer yet but it is very close to it. And then finally on the extreme right a low grade, in this case, endometrioid carcinoma and these can very occasionally go on to high grade endometrioid carcinoma, which is really bad but most cases of carcinoma endometrioid type associated with endometriosis tend to be low grade.
What does all this mean in terms of our looking to the future and its clinical applications? Well, what we want to do of course is reduce the risk, the burden of ovarian cancer. As I mentioned at the start, radical surgery and chemotherapy have really not improved overall survival. They have lengthened the disease free interval but they have not achieved a real improvement in overall survival. Screening studies also have failed to provide a survival benefit and it is certainly not recommended to be used in a clinical situation or being looked at in research situations at this point.
We think, perhaps, a better approach is prevention. What do I mean by prevention? Well, if we can unequivocally show that ovarian carcinoma begins in the fallopian tube or the endometrium and then involves the ovary secondarily, well perhaps young women at high risk – and here I am specifically thinking of women who have what are called germline mutations that have inherited mutations in a particular gene called BRACA, which is the breast cancer gene. You look at their family history and they have sisters, or mothers, or aunts who also have either breast or ovarian cancer, they are considered at very high risk of developing ovarian cancer and breast cancer; 50 percent for ovary and even higher for breast cancer. They often undergo risk reducing salpingo oophorectomy, removal of the ovaries and tubes to prevent development of ovarian cancer.
Perhaps, on the other hand, if we prove that these cancers really develop first in the endometrium for endometrium clear cell and in the fallopian tube for serious carcinoma then if we remove the tubes but leave the ovary and the uterus behind you have preserved their fertility, you have preserved their hormone function and hopefully, better quality of life. Possibly oophorectomy when they get older.
In terms of the origin of endometrium and clear cell carcinoma there have been several epidemiologic studies that have kind of confirmed this idea that has shown that tubal ligation indeed reduces the risk of ovarian cancer, and in a couple of studies very specifically for endometrioid and clear cell. If you imagine for a moment I am the uterus, fallopian tubes…if you do a tubal ligation just cut this tube here you obviously prevent the retrograde menstruation so you reduce the risk of endometrioid and clear cell carcinoma. But if you still have the remainder of the tube, which is where the serous cancers begin you have not really changed much of the risk there and that is why we think the whole tube should really be removed.
Going on with this idea of prevention, women who are at high risk perhaps we ought to be thinking about other ways of reducing their risk by – and there have been several small studies that really have not been proven yet but they are tantalizing, very interesting – aspirin, statins, metformin have been shown to reduce the risk of developing ovarian cancer and a very strong argument for oral contraceptives, which I am going to show you in this slide over here. This is a meta-analysis so they took multiple studies in the literature, put them altogether to get this huge number of patients that were analyzed, 100,000 women, and it was found that OCP or oral contraceptives pills reduced the risk, as you see, by a tremendous reduction in risk of ovarian cancer over women who had not use it and you see a very significant p<0.0001 value, that is highly significant. Interestingly, that reduction in risk persisted for over 30 years after OCP use ceased. By modeling it they extrapolated that based on the use of a number of women this is worldwide data now, perhaps since it was used over 50 years maybe 200,000 new cases of ovarian cancer have been prevented and maybe 100,000 deaths. And looking into the future that with current usage it could significantly reduce 30,000, and I am talking worldwide, ovarian cancers every year.
In terms of prevention there is another group of women who fit into this scenario. They are over 40, they have benign uterine disease, say fibroids or abnormal bleeding and the gynecologist recommends that you need to do a hysterectomy to be cured of them and of course that will. But since you are in your 40s, you really do not need your ovaries anymore, why do not we take out your ovaries and tubes that is BSO, bilateral salpingo oophorectomy, TAH is total abdominal hysterectomy, and by taking out the ovaries and tubes your ovaries specifically will reduce your risk of ovarian cancer. This is an interesting study recently published from the Nurses’ Health Study in Boston where they looked at women who had bilateral salpingo oophorectomy versus ovarian conservation in this group of women. Benign disease, they have a hysterectomy either to remove the ovaries and tubes or leave them in and see what happens. Indeed, when you take out the ovaries and tubes there is a marked reduction in ovarian cancer.
But are we paying a price for having performed that procedure and reduced the risk of ovarian cancer? Well, it is interesting. There is no survival benefit from removing ovaries and tubes at any age. That in fact you get increased mortality from coronary heart disease, from lung cancer, from colorectal cancer – all causes 13 percent. We really need to consider these other issues when we make these kinds of recommendations.
Perhaps women who are undergoing this type of therapy, perhaps all they need is a bilateral salpingo oophorectomy with preservation of the ovaries at least for some time and then possibly, when they are older remove them, in their 50s and really post-menopausal when these other causes of mortality do start diminishing. It is not a trivial issue because there are about 300,000 women in the US who undergo elective oophorectomy each year.
Well, I mentioned that early detection and early screening studies not really being of any benefit and I would like to just present very briefly a new detection test that was developed in conjunction with Bert Vogelstein at Hopkins. He is a preeminent molecular biologist and has made enormous contributions to the field of cancer and molecular genetics. He and his colleagues have developed a highly sensitive exome sequencing panel in which the panel consists of genes that have been commonly mutated in ovarian and endometrial cancer. What we did was we looked at women who were coming to hysterectomy either for ovarian or endometrial cancer, took that tumor tissue, analyzed it for what mutations were present and compared it to a ThinPrep, which is the newest approach I am sure you all know of doing the pap smear. We do not put it on the slide now. We take some scraping, we put it into a fixative so we have a liquid sample. That liquid sample can be analyzed for papillomavirus DNA and all sorts of other kinds of DNA and look for the mutations in that fluid and see if they correspond to the ones in the ovarian or endometrial cancer. And indeed they did. In 100 percent of the endometrial cancers we found the exact same mutation in that cytology fluid and not quite 50 percent of the women with ovarian cancer.
I think that is very exciting. It is obviously very, very, very preliminary data and we are in the process of trying to study much larger numbers of patients and do other kinds of studies to see if we can really detect those changes early enough to really make an impact. Identify the tumor early and therefore really make a significant role in reducing the mortality of this disease.
In closing, I am going to end with a quote not by a gynecologist or a pathologist but by a great 20th century statesman in terms of where we are now with ovarian cancer. He did not talk about ovarian cancer he talked about certain events that were occurring in WWII and this was Winston Churchill. He made the statement that I am going to show you in a moment shortly after the first allied victory at Al Alamein in North Africa and this is what he said, “Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning”.
Thank you.
