Piraye Beim, PhD
Inflammation, Biomarkers, and the Origins of Endometriosis
Endometriosis Foundation of America
Medical Conference 2019
Targeting Inflammation:
From Biomarkers to Precision Surgery
March 8-9, 2019 - Lenox Hill Hospital, NYC
https://www.endofound.org/medicalconference/2019
So good morning, my name is Piraye Beim and in 2009 I founded Celmatix, a precision medicine company focused on women's health. In 2012, I was diagnosed with Endometriosis, and in 2018 I had the privilege of joining Dr. Seckin on the board of the Endometriosis Foundation of America, so like many of you in the room, I'm wearing multiple hats.
So I wanted to start by first welcoming all of the scientists and physicians who joined us today, with my EFA hat on. And then with my scientist hat, I wanted to thank you to the EFA for hosting this important event and letting me share a little bit about our work today. And then with my patient hat on, I want to say thank you to everyone in the room for helping improve the lives of hundreds of millions of women around the world with this condition. And then finally, I want to acknowledge and thank the senior scientist from Celmatix, whose work I will share with you today, and who helped prepare this presentation. So my colleagues Dr. Emlin Parfait, Dr. Katarina Clementi, and Dr. Genevieve Galarneau contributed to this work, thank you, guys.
Okay. So as we mark the tenth year anniversary of the EFA this year, one of the things we've been reflecting on is, what has changed for the Endometriosis field over the last 10 years? What still needs to change, and then what changes are we likely to see over the next 10 years?
When the EFA was founded a decade ago, there were already passionate investigators around the world like Grant Montgomery, Joe Lisamson, Carina Zondervan, and others who were working to demonstrate that Endometriosis had a heritable genetic basis.
Building on this initial work GWA studies and targeted association studies that leveraged multiple different OMX methodologies further helped to elucidate the etiology of this complex disease. And next came leveraging these insights to develop diagnostic methods, that has subsequently evolved into work to identify biomarkers related to response to treatment. And then finally, now, we're exploring the use of these biomarkers as novel drug targets, and we're going to hear a lot about these exciting developments today.
But one central theme that's been reinforced through all of this work is the important role that inflammation plays in Endometriosis, and that's one of the reasons that that is the theme of our 10-year anniversary research conference. And you can't really talk about inflammation and Endometriosis without first taking a step back and talking a little bit about the immune system.
So we know that virtually all women experience retrograde menstruation, however, it's like that the clearance of this tissue and debris is usually mediated by cytotoxic cells. However, if these cells become tolerant to the debris, or are disrupted molecularly in some way, this may open a window for Endometriosis lesions to develop. Endometriosis is also characterized by chronic inflammation suggesting that the immune system is constantly responding to stimuli, and by doing so, it further creates a favorable environment for the growth of the lesions.
Normally natural killer cells, Anti-lymphocytes, are able to recognize atopic cells in response by inducing cytotoxicity. When these cytotoxic cells are defective, or they cannot be properly recruited to the site of the lesion, or they can be recruited, but they're not able to induce apoptosis, the atopic tissue is not removed, and the lesion can develop. Cytotoxic cells can also be inhibited by T Regulatory Cells, also known as T Regs, that are immune cells to induce tolerance to specific stimuli, and here I'm showing you different genes that have been identified that help regulate these processes.
So one of the hypothesis then is that, if we have variants in these genes, such as PTPN22, FOXP3, FCRL3, and IL10, which control T Reg development and function, these variants could cause a number ... the number, or the activity of the T Regulatory cells to increase, and this could lead to a reduction in cytotoxic activity, and therefore create a more permissive environment for Endometriotic lesions to develop.
On the other hand, a lack of immune response atopic cells may also cause the development of lesions. And on the other hand, it's been observed that women with Endometriosis show a continuous activation of other parts of the immune system that generates chronic inflammation. So for example, pro-inflammatory cytokines like IL1B, TNF, and IL6, show higher expression in Endometriosis patients, and they've also been shown to affect the ectopic endometrial cells by promoting the expression of factors like integrins and metalloproteases that can facilitate cell adhesion.
These pro-inflammatory markers have also been shown to promote proliferation of ectopic tissue, hence the growth of the lesion. Sorry, I'm messing up my ... I'm trying to do two things at once. Hence, the growth of the lesion and they promote the expression of angiogenic factors that in turn can lead to the revascularization of lesions essential for lesion survival.
So the evidence shows that Endometriosis is characterized by inflammation and that inflammation can also help the survival of lesions. But how does inflammation also impact reproductive function potentially? It's been shown that the levels of pro-inflammatory cytokines such as IL1B and IL6 are increased in the follicular fluid of women with Endometriosis. And because Endometriosis patients compared to healthy controls have been shown to have a lower rate of successful IVF outcomes, oocytes of lower quality and defective steroidogeneses, it has been suggested that inflammation of follicles may also be affecting ovarian function in oocyte development.
So we know that surgical visualization and pathological evaluation of suspected lesions are still the gold standards for diagnosis, but this high bar is in part why so many women go undiagnosed for so long. So in this next section, I'll talk a little bit about work from several groups to help the biological insights that we just discussed in the previous section translate eventually into a broader toolkit of diagnostic tools.
Here, for example, is a study lead by Linda Griffith, Dough Loughenberger, and Eric Kvikstad, I hope I pronounced his name correctly. They identified a combination of cytokines that were discriminating between the presence and absence of Endometriosis. And we'll also hear today work done by Hugh Taylor to use micro RNA's in a similar way.
The evidence base has also been growing on the genetic factors related to Endometriosis risk. So one of the initiatives that my colleagues and I at Celmatix have been focused on, is mapping the strength of these associations, and the reproducibility of these observations to a framework for gaining confidence in the clinical validity, and the utility of revealing these genetic risk factors for managing Endometriosis in the clinic.
Here I show you one framework we borrowed, which was developed by the Clingen Consurtium for ranking the strength of genetic associations with Mendelian Disease. The genetic underliers of Endometriosis are more complex than Mendelian Disorders, but there's still some solid principles from Clingen and organizations like ACMG that we can repurpose.
So one of the initiatives that we did, was at first we leveraged natural language processing AI methods to informatically refine the 24 million publications in pubment down to the 4,500 published studies relating a genetic factor to Endometriosis. That pretty much represents the body of work in our field on genetic associations. We then bio-informatically map these to over 600 unique genetic loci that have to date been linked to this condition. And then from there, we rank them using these frameworks that I discussed, in order to measure the potential effect of a given gene in our confidence in that association.
Then we looked at the variants within those genes that might be impacting their function, and using a statistical model and another set of guidelines, we ended up at the end of this process identifying 12 variants in nine genes that are likely to increase a woman's risk of being diagnosed with Endometriosis. Interestingly, genes containing the higher risk variants had functional roles in both DNA replication and repair, like XRCC1, and immune system function. In fact, four of the five genes, FCRL3, IL1A, PTPN2=2, and ICAM1 play roles in inflammation and immune system regulation, which makes sense.
We've been working with Dr. Goldstein and Dr. Seckin at the Seckin Clinic to start to now understand the utility of revealing the existence of these markers in women who are coming into their clinic. This is really preliminary work, we're not ready to publish this, but for this audience, I wanted to show just how it's trending. And what we were very excited to see is that a majority of the patients who came in with suspected Endometriosis tested ... so about 65% of them tested positive for at least one of these high-risk variants. And then when we looked at the patients who then had positive surgical confirmation, we saw that number actually go up to about 75% of the patients who had positive confirmation on pathology, had at least one of these markers.
So in addition to evaluating the existing body of work in our field, scientists at Celmatix have also been working to contribute through partnerships, for example, with the International Endometriosis Genomics Consortium and other groups to expanding our understanding of this disease.
So one of the initiatives that we're really proud of was partnering with the Consumer Genetics company 23 And Me, to leverage their vast data set to be able to do GWAS's on an unprecedented scale. So the first GWAS that we did on Endometriosis, we explored over nearly 300,000 women who were of European ancestry. We performed a case-control study where we looked at women who had reported having been diagnosed with or treated for Endometriosis compared to women who weren't.
So there were 37,183 cases, and then there were 251,258 controls. The samples were genome typed on a genome typing array targeting between 500,000 and 955,000 snips, and up to 13 million snips were also imputed using phase one of the 1000 Genomes Project as a reference population. We adjusted our statistical model for age, the first five principal components, and for the genome type, array version used.
We found in the study that 14 loci reached genome-wide significance in our study. KDR, the VEGF receptor, GRAB1, and ID4 will be among those very familiar to this audience. Seven of our loci, however, are novel and are labeled here in red. At least three of these loci have functional links to inflammatory mechanisms, which strengthens the idea that these processes are perhaps not just a consequence of Endometriosis, but potentially part of the mechanisms driving the disease. So I'll tell you a little bit about those loci now.
So one of the novel loci, NGF, has previously been associated at genome-wide significance with severity of Dysmenorrhea in another study that was performed on 23 female participants of European ancestry, and as we know, painful menstruation is one of the symptoms of Endometriosis. The expression levels of nerve growth factor, NGF, are higher in the peritoneal fluid of women with Endometriosis. Blocking NGF significantly decreases outgrowth in Endometriotic lesions in certain models, suggesting that this locus may be involved in inflammation and self-proliferation in those aspects of the disease ... Sorry, it's progressing itself by self here.
Okay. The suggestion is supported by nerve growth factors ability to influence both neuronal cell function and immune cell activity. The up-regulation of NGF described in inflamed tissues of many diseases can regulate innervation and neuronal activity of peripheral neurons inducing the release of immune active peptides and neurotransmitters, but can also directly influence innate and adaptive immune responses. NGF has a variety of effects that can either pro-inflammatory or anti-inflammatory. This apparent contradiction can be explained by considering NGF as part of an endogenous mechanism that while activating immune responses, also activates pathways necessary to dampen the immune response and limit tissue damage. This may be disruptive ... This feedback may be disrupted in patients with Endometriosis carrying variants in the NGF locus.
The second locus is near the gene HAY2 and also the gene NCOA7. This locus has been associated at genome-wide significance with Endometrial and Breast Cancer, which is a theme that we've hit upon before within the foundation and is an EQTL for both HAY2 and NCOA7. HAY2 is required for vascular genesis, and we've seen that double knockouts of HAY1 and HAY2 result in lethal vascular defects that affect the placenta, the yolk sac, and the embryo itself, likely due to impaired arterial fate determination and maturation. Single knockout mines of HAY2 have cardiac defects, a piece of info that can potentially be interesting when we talk about how Endometriosis may be a canary in the coal mine, working off of Dr. Stacey Missmer's findings that Endometriosis is a risk factor for cardiovascular disease.
Interestingly HAY2 is a transducer of notch signaling, and you'll be hearing later from Dr. Faz La Bas, I tried, about how notch one mediates inflammation in ectopic Endometriotic lesions and how we might take advantage of this mechanism therapeutically.
Sorry, it's still progressing itself. Okay.
The next locus is near the gene WT1, which encodes a transcription factor that is highly expressive in the Endometrium, the Testes, and the Ovary, and that a central role in the normal development of the urogenital system. WT1 is selectively expressed in neurons of deep Endometriosis and has also been shown to be downregulated in Endometriotic stromal cells compared to Endometrial stromal cells. Beyond this though, WT1 also has wide-ranging functions in development, homeostasis in disease, and perhaps most relevant to Endometriosis is its role in the epithelial to mesenchymal transition, which is thought to be a prerequisite in the establishment of Endometriotic lesions.
Through its role in regulating EMT, WT1 is also important for several aspects of cardiac development, specifically required for epithelial to mesenchymal transformation that forms epicardium-derived cells that will contribute to the formation of coronary vessels.
I'm sorry that my ... I'm not pushing the thing, but it's progressing on its own, so what can we do. Alright, next slide. Okay. Alrighty. Now it won't progress. Okay.
So as I mentioned, one of the next steps that we've seen that's very exciting for the field is that, many investigators are now leveraging these biological insights to start to think about how we can move our field into more targeted interventions from a therapeutic standpoint, and that's a direction that Celmatix has been headed in as well. Last year at patient day, somebody raised their hand and said, "You accumulated all of this data, are you thinking about leveraging it to help expand the toolbox of therapeutic interventions in the field?" And at the time it was a little premature, but we said yes we have some announcements to make soon. And in December, we announced Celmatix biosciences, which is dedicated to exactly this, which is helping to fertilize the field and through the identification and validation of novel targets and mechanisms of action.
And I won't get into how we do that, but the important thing is that, if you can plug a large data set like the one that we've accumulated through our reproductive atlas platform into a drug discovery pipeline, you can move away from the paradigm that if a particular program fails, that was a waste of time, so if a program fails, you can go back recursively into your AI engine, and you can learn. So that means that every dollar that we're spending in progressing these drug programs, can actually help us learn and improve the programs that we ultimately take into the clinic, and we're going to be making some announcements later this year about some work we're doing on that front. But I know there a lot of people in the audience as well who are going to tell us about work to hallucinate novel targets, so I think this is a very exciting next direction.
And this is really the vision, is that right now while we celebrate that there are a number of interventions and options for women, and we want to increase access to those options for women, we also want to be priming the pump and pushing the pipeline forward for moving into targeted interventions, which is really where fields like oncology and others have gone.
So the other thing that I wanted to reflect on before I conclude, is in the arc of where we're going and where do we still have a lot of things to overcome and what hasn't changed. So I just want to talk through some of these numbers, because we share them sometimes, and I think I'm preaching to the choir in this audience, but it's just breathtaking to think how unmet some of these needs are, and that's one of the things we think a lot of at the foundation is, how do we bridge this gap. So I'm going to throw some numbers out here.
So Endometriosis is a relatively common condition. We estimate now that it could be as high as 20% problems in women of this generation, which makes it basically as common as diabetes, and yet it gets very little funding. So according to the American Diabetes Association, the cost of Diabetes to the US economy was $327 billion including $90 billion on lost productivity and $237 billion in healthcare expenditure.
So according to the National Centers for Health Statistics, in 2019, NAH funding on Endometriosis is estimated to only be $6 million. So that's a 14% decrease from 2018 when it was seven million. And just to compare and contrast, Diabetes, got nearly $1 billion of NAH funding this year. So when you compare the cost of Diabetes to the US economy versus the cost of Endometriosis to the US economy, it's actually really staggering that the spending for Diabetes is 331 to one as a ratio, and the spending on Endometriosis is 13,000 to one. So these are some of the numbers that I think that we all need to start to have in our back pocket as we go out into the world, and make sure that we start to continue to really shout from the rooftops that this is a missed opportunity.
And I'll end on this note, which is one of the reasons that I think it's a huge missed opportunity, and as we reflect on who's going to be in this audience 10 years from now when we have our 20th anniversary for EFA, I think the audience is going to be much bigger and full of many different kinds of physicians and scientists. Because, what we're realizing as we're realizing that Endometriosis is fundamentally rooted in immune function and in inflammation, is that it really could be the canary coal mine for many different conditions down the road including Cancer, autoimmune disorders, allergies, cardiovascular disease, et cetera, and this is really important.
I've heard some people talk about how the NAH has a reverse pyramid, where we put all of the funding into the latest manifestation of these conditions, and a very, very low funding into the earliest manifestation, and/or the events that set these later problems into motion with maternal-fetal health, pediatric health, reproductive health, things like Endometriosis.
So I just also want to stress that, I think the more we can think outside of our field a little bit and find these linkages to things that do traditionally have more support, that will hopefully be catalytic to raise awareness on the importance of this but also to bring funding into the space.
And then finally, the last note that I wanted to end on is that many different disease areas are coming together to create a learning health system. So this is a buzzword that maybe some of you have heard, but I think this is an important role that EFA can help play in the field, which is that academia cannot solve this alone, pharma cannot solve this alone, a foundation cannot solve this alone, patient advocacy cannot solve this alone. A learning health system is really about bringing all of the different stakeholders to the table and making sure that we learn from each other. So as scientists have a hypothesis, we need to make sure that there's a clear way to translate that into the private sector and to translate that into novel diagnostics and novel treatments.
But then the private sector has to stay closely aligned with foundations and academics, to make sure that the things that we're learning can feedback and generate a novel hypothesis, and that we can turn this into a self-kind of driving engine. And that's one of the things that I wanted to kind of throw out there, is what are we doing? We're all individually contributing, and we're in our silos, and we're working together, but how do we start to bring the pieces of the puzzle together. And one thing that I'm very proud of, is that we partnered ... the EFA partnered with the World Endometriosis Foundation at the ASRM conference in October to bring a number of pharmaceutical companies into a meeting. Five of the top pharmaceutical companies that care about women's health, and care about Endometriosis, came to the meeting, and we had a very open dialogue about how do we start to work on helping pharma where a lot of the dollars are, quite frankly, to understand the pain points of the patient, of the investigators, but then through doing that and through helping to seed and maybe pass some of those profits forward, to actually improve the quality of the products and who they're targeting to, and possibly grow their own markets.
So anyway, I'll end on that note, but thanks again for inviting me.
While we're pulling up questions, would you back up to the last economic slide?
Sure.
The one that had all four, and then we'll take questions. I just want to look at the slide again while she's answering questions. Yes, sir?
Thank you for your talk, this is not on. Can you hear me?
I can hear you.
Let's try this one. I've got a better one now, okay. So thank you for your talk, and I'm glad that targeting inflammation has been a center focus of this meeting. I have been working in this field for the last 10 years. Immune dysfunction and Endometriosis is the theme that I have been working on, we published a new article. One question I get from readers when we publish our work is, inflammation is important, but is it a cause or effect of the disease? And second question we get is, what is the source of inflammatory markers, are the lesions making it or are the perigonal microenvironment is responsible? And in my opinion, both are fair questions. We have proved that Endometriosis lesions actually make cytokines, you do a surgery, and the levels drop precipitously.
Then the second question comes, what is the clinical utility of those cytokine biomarkers? And I think one way of doing that is, it's about time that we treat each stage, and each form of the disease separately, and probably we'll be able to establish better clinical cutoffs. What you are doing I think is an important step towards that, but I don't know in terms of the clinical utility if you guys are thinking on some novel ways of addressing this question.
Yeah, so thank you. And I think that it's premature to say if it's chicken or the egg, but my guess is that's going to be chicken and the egg. That these things are, like we said, in these regulatory loops, so there might be an underlying condition, may be inflamed by an environmental factor. So maybe there's a genetic predisposition inflamed by an environmental factor, probably further inflamed by anatomical factors that maybe promote a higher degree of retrograde menstruation or greater persistence, or something like that, that sets it into motion and then it turns into this kind of self-fulfilling cycle that either you can't ... once the brakes are turned on, and you hyper respond, it's much harder to turn it off.
I think part of why we're taking a multi omegs and a big data approach to trying to create a model of Endometriosis, but also in thinking about how we intervene, is that it's such a complex set of conditions. And the algorithm for each woman is slightly different. In one of the slides I didn't show, I'll show it on patient day, is just how Endometriotic lesions, as you know, can implant in lots of different parts of the body. So there's so much stochasticity as well built into why a particular woman ends up with a particular phenotype. So I think it's going to, like all complex diseases, it's an exciting field to be in because there are so many things to work out, but it's also going to forever spark these controversies because people are going to fall into these camps of, "I think this is the important thing, or, I think this is the important thing."
One of the things at Celmatix, because we take a data lens on this is, we're agnostic. We think all of it is contributing in some way, and it's really about building these models that can add each of these little contributors, so that's one.
With respect to the utility, I don't think there's ... I don't think any biomarker, any test is going to be the golden bullet, because as you said, this is a multi-stage disease. Sometimes it's something that looks very different as a teenager than somebody who is post-menopausal. And I think that one of the challenges that we have is, we have such a late start in doing anything about it, that the hammers that we have we're trying to apply to everything, right? So that's always ... There's always going to be some shortcoming.
But I think as we start to expand the menu, then I think that we'll eventually, when we really are in a precision medicine world for Endometriosis, we'll have certain things that ... certain biomarkers that are helpful for identifying women at risk, and employing preventative measures. And then for women who are either missed out on those opportunities, because maybe they didn't have access or awareness, or did not respond to them, then we'll have the second line, et cetera. Surgery will always be part of the algorithm, I think drugs will be part of the algorithm, but anyway, it's a complex answer, but yeah.
Okay. Now we're done. Thank you.
Okay.
Appreciate that.
Thank you so much.
