International Medical Conference Endometriosis 2025:
Endometriosis 2025: Your Mother Should Know, Your Doctor Should Know Better!
Endometriosis Should Be Updated for the Limitations of Trial Evidence - Philippe Koninckx, MD, PhD
Good evening. Sorry for not being in New York. Despite all intelligent pathophysiology of end meteorologists remains announced, I have no disclosures. Who I am is cleared. We learn from each other. This is the Merc of many. I do thank my quarter, Frank Dan Martin for the many discussions with many people all over the world and you learn when you learn languages how subtle difference becomes important. My message is be a clinician, be a clinician because the endo pathophysiology of endometriosis we do know but we do not accept. This is what I think is important. Reside to something different. Artificial intelligence limitation and bin of the human brain. Be weak Clinicians we observe, compare and interpret. Learn Fromm experience, predict we are aware of biases and it is completely different from all the, let's say, ways of doing statistics about the data. What we have to add in clinical medicine is experience heuristics and intuition.
It's something that's very clear that if you see for instance this kind of dark alley, you be prudent. Maybe it's called intuition to return a ball at 200 kilometers an hour that you cannot think and to know how you do a gun fight, you have to have experience. What is important there you have to train. It becomes less and less conscious and this is why good surgeons can be poor teachers. I'm not going to talk about the different types of intelligence, but we ought to have experience istic based in learning by experience. It's very clear that we update the past by new experience. You see a girl and you think that she like me and then she smiles and then she update your opinion. Well, what is important to this kind of update is that it's about past experience. It's uncertain, it's a probability and it's multivariate and if you look clearly at it, this is the same as used in Google search, self-driving, car hedge funds, weather prediction, artificial intelligence, but the clinician thinks multivariate and sigmoidal the indication for surgery of cystic co variant and the material.
If you ask to a group of clinicians, when do you do surgery for pain? You see that below three in 10 nobody does surgery and above seven older surgery. And the same thing for this and age of women. The clinician thinks multivariate and M, when you ask two clinicians with experience when to do surgery for deep endometriosis, this below pain three and 10, nobody does surgery and above seven almost everybody is going to do surgery. This is a typical sigmoidal curve which can be expressed in a sigmoidal family. We come back to this in a second, but that's what a clinician does daily in his mind. We combine pain symptoms, size, age, imaging aspect, antis, patient wishes. But you have to realize that all the decisions we take are binary for diagnosed, for surgery, for treatment. And binary means that we do or we do not.
And for all it's like when you go to the ware and you buy something or you buy or you don't buy, which means that for this you cannot use linear regression, logistic regression, which is almost sigmoid. What is important with the SMO decision is that there is an area of uncertainty where you beat exams, do additional exams come back in a month. But the most important is that the yes no of all the decisions we have in evidence-based medicine should be changed in a no rather, no, I don't know rather, yes, yes. It's a different way of looking at things. And when doing this, I was very surprised to find out that artificial intelligence neural networks are exactly doing the same. Initially they were doing linear regression and then they came to seek modal artificial intelligence, it compliment or brain power perfect to correct codes for programming.
It's perfect for language and memory it out, speech when it's complex, but the limitations they analyze it permits the analysis and neuronal networks for imaging for buying behavior and prediction is always based in driving edge. For example. This is you see the complexity to predict when a custom is going to buy how much is spent, how much is long time is there, how much let's say is information before next. This is for those interested in mathematics. For those interested in mathematics realize that a SMO curve and a log function is exactly the same but in the other direction. And then we come to evidence-based medicine and would say how it went wrong. Why most published research finding are fault 90 to 95% is the estimation. The clinical medicine was reduced to evidence. Such a randomized controlled trial, the limitation of the randomized control trial, real life where ignored the limitation of traditions, statistics were ignored and you see in evidence-based medicine there are no based in statistic and then because it didn't work, there was an interpretation by experts in evidence-based medicine statistic and those are judging surgery.
And then we should not be surprised that the guidelines for endometriosis are different When formulated by AOC X-Ray A of the Canada, they haven't tried today to look at the pathophysiology of endometriosis. I would say it is known, it's very clear and had tried to explain mensuration these assumption. This is still the most believed hypothesis and normal cells in an abnormal environment. I think you have to postulate some genetic epigenetic changes and then it become an abnormal cell still in an abnormal environment. It's heredi and the me is very clear. This is the predisposition. We know this. You can read faster than I can read it. Hereditary means DNA or epigenetic. An epigenetic is something that cannot not be explained by changing in DNA and is a stable change and you have to realize that all our cells have the same DNA and then we come to epigenetics is the cell differentiation.
It's almost like origami when you have made something in an origami and you have to redo it later, it's much more easy. And this I think for me is the transmission of the epigenetics. Very unclear whether it is reversible but it's not everything because fish land from other fish. If your mother is a bist you have more probability of being a bist and nobody understand how the transmission is made. Endometriosis and endometriosis is endometrial like tissue is the definition of 100 years ago when pathology was the only technique we had and is a problem I have with all microscopy and histology, the accuracy is not known since the gold standards. The microscopy cannot detect all the genetic and epigenetic changes and on top of this it is an artificial intelligence like impression with descriptive words with nobody understands what it really means. Same thing for cancer and has no time to discuss the visual illusion and bias.
Look at these images. Who has the longest legs? They have exactly the same length of flex. This is visual illusion and problem of histology it looks like and the I would say it's a fast tissue then I would not be surprised that today a lot of mutations cancer drive a mutation. There is something which already their in pregnancy. So we are not surprised to find that if you have a predisposition for endometriosis or endometriosis you have higher risk of not having physiological changes and have hypertension in pregnancy. And these are one of the reasons why the junctional zone and the basal layer should be taken together in order to understand endometriosis including the mobility of stromal cells. And if you have endometriosis you will find stromal cells of the endometriosis in the endometrial. The upright position changed a lot. This is all animals have the yellow part, the orange part, this is RA and the rest is there.
The rest of the omet is because we are standing upright. This explains that this RA is important for contractions, cellular and so on and is important in order to have for fertility and it's also changed in any materials. That's why Samsung implantation theory, it's not enough. It's clearly incomplete and it's in incompatible with a series of aspects which you all are fully aware of, does not explain effect of dioxin and it occurs a man and a woman without a uterus. It was already there before puberty and before pregnancy. Pregnancy outcome is different equipment and the OSIS was more for days, more prematurely, more hypertension and after doing surgery, removing it, it does not change. Same thing for bowel complications on, and this is similar for a look of aspects before and after surgery. C one to five was high before and after surgery becomes slow.
But natural kill cell they are low before surgery decreased and after surgery they remain low, which means that it suggests that a lot of the defects were there before surgery. It's something is there before puberty. It explains easily the dioxin which has an effect directing the DNA, the classic data from rear about exposure to the is monkey to dioxin. What for me is important is that endometriosis is clonal one cell developing to a clone of cells. This photic endometriosis deep and typical lesions also most important that if you find 10 difficult typical lesions, 10 difficult lesions are 10 different and then it works like this. Retrograde menstruation, it's normal to have implantation and you get subtle lesions and microscopic lesions. But these are mostly normal cells and look, pathology and personally are more and more and this intended to believe that some of abnormal cells were already abnormal in the end and then developed endometriosis.
This is predisposition. Infertility change in the endometrial immunology is all there before pubic and then something happens in the uterus or in the peritoneal cavity because of oxidative stress rate you get mensuration bleeding in the lesion infection, whatever. But something happens. The cell change and this explains that. You get clonal before different pathology and the lesion heterogene understands that like a break of a car you have a dual break. If one goes down you still have a break, you can have a lot of lesions before it becomes visible depression. So this is the end result, inherited defects and then because of radiation pollution, oxidative stress, we have more lesions and of sudden it's too much. And you start with endometriosis, which becomes different lesions. Well it's very important I think is infection. Infection are a potential co-factor in order to change these cells. Very clear you have more infections if done with you have an alter microbiome and in the end, end materials lesions you have H more HP, V and molecules.
Much more difficult is the transmural migration of the microbiome from the bowel to the peritoneal cavity. But after all, microbiome is something what I think is going to be important. So this becomes the model is added theary, the predisposition and then infection ative stress. You get genetic epigenetic incidents and you get the tumor which has to grow two subtle typical CY deep lesions, four different lesions and the growth and the natural history impaired cavity androgenic factors are not going to review this. These are is all data which we know very well the growth and natural history of those abnormal cells, which were already there. But growth is self-limiting. This is the data from lun and you see that when you women with of age 20 and 40, there is not a big difference. Even if they had pregnancies, one or two children doesn't make a difference, which means it has to start to grow and then stop to grow.
And so we come to this viewpoint that as lesions develop, you get immunologic reaction, fibrosis, more and more fibrosis and you end with some kind of war of trenches. The enemy cannot get out and you cannot get. In conclusion today it's more than one disease. It's heterogeneous, it's clonal, it's self limiting growth. After initiation, the impact is independent of the endometriosis lesion rather it's a predisposition iv, progesterone resistance, revers and irreversible. It's not very clear what it is. What is an inherited predisposition and the future will be why these changes and which causes the changes. What causes the changes. The future I think is artificial intelligence and under the sand why and which is going to ditch. Causes are going to changes. For this you need observational data and very large database important to grasp. The problem we are facing is you see that these are forced almost 5,000 women and endometriosis, but typical subst, deep and ovarian are always associated.
This complexity has to be taken into account. Adolescence, primary disciplinary error from the first cycle suggests it the disorder that was already there when when the lesion starts. We have a major increase in laparoscopies around age 25, 10 years before this was the puberty. And this is a model like many other. You start with developing end materials during puberty is highest risk and then progressively with H there is decrease of initiation. The nice thing is that with this model you can prevent and materials, you can reduce pollution, maybe more difficult, but you reduce oxidative stress and you exchange the peritoneal microbiome. And this could be with antioxidant change, the foot intake exercise, continuous oral contraception. I think for the first time we have the possibility to change the probability of starting endometriosis. If I would be young, I would have a close look at the vagina microbiome uterine cavity.
I think we should have much more attention to vaginal infections and understanding vaginosis, phages, vaginal immunology and form. That's where I think the future lies for understanding of the material delay in diagnosis. I don't think it's a problem. I don't know it's a problem because when it started it's already too late. I think we mainly should prevent end materials just to see how far this can lead us. The fibrosis around an endometriosis lesion belongs to the body, so no reason to take it out. But also like in oncology, the cells around the bad cells, they are give the impression of being antic but they are not the cell scores it, which means that also those extensions of other materials in the lesion are probably not necessary to be taken out. It's like a snake. If you cut off the hatch, the snake will die panel distance.
I think it's very important it is there but we don't know why. We know that for the clear lesions and the subtle and the typical lesions, more than half 60, 70% are absolutely painless. And I think these are the new things about endometriosis, although they are more than 20 years old. This explains also that Indian Chinese food supplements, diet eggs exercise is something which is going to change the endometriosis. And to finish I would like to say a few words which I think is very important. Today we do not have data about endometriosis. Very little, no animal mogul. We know very, very little solid data and then we do surgery. There are no data, there are no rules, there are no checks. This is the wild twist of surgery. The wild twist of surgery is written in one of our last papers. When you do life surgery, life surgery, you see all the different very good surgeons demonstrating, explaining their technique and their preferences.
But they are very different. I am not going to retweet you all the differences which are there and we just still a discussion. But wild twist is very different surgery and take for instance adhesions, ais prevention. We know we know since 40 years don't use g, have rinsing, have a clear discussion. And then you see a lot of surgeons are using gause and do not have wrenching in all the not to D disturb the planes. That's why evidence-based management of endometriosis should us experience. Because what is here in roles, this is research, this is the investigated group. When you, at the experience of all the clinicians, you extend to the population with basic learning from the past. You have heuristics trained to do surgery and have decision made. These are the skills you have in tuition. And I do not have time to discuss the black swans heuristics.
We have to understand you need a teacher to learn surgery because a teacher is going to avoid bad habits. You need skills then for instance, not dying. Have to know that you learn when you're sleeping. Sleep is needed to consolidate and the learning is sixmo. And when this is important, very well known in the far rest in the police academy and so on, high cognitive load and fast decision like in surgery, that's where you need theistic. And so we come to guidelines. When I look at the RA guidelines and the osis, really there are few heart practical data with recommendation. As LAPROSCOPY is no longer the gold standard. Adolescent are no different from adults. I haven't seen the evidence for that. So we have to update what we know with experience because we decide every day about all aspects and how we decide it's multivariate and each factor we use, it's not only multivariate but sig for each.
When we add experience, we also can have data about unnecessary or surgery, not well done surgery without a strategy. Poor preparation for accident at these formation. So this is, we can have data of what research, as we all know we are things of this should not do. And with your help, let's say all the clinicians, we can change documenting experience. We update evidence-based medicine guidelines with collective or collective experience. So we have data about the population and when we do this, we are going to understand better the diagnosis, understand better therapy. And most important for me, you can say bye-bye while west surgery there will be rules for surgery and it'll not be the personal decision how to do. Thank you for your attention.
