 So we'll be getting started for our working lunch. And again, the purpose of this next session is to have a little bit of a report back and synthesis of the discussions that happened when the working groups had breakouts yesterday. So for each of our working groups, there's two co-chairs and one of them will be coming up and giving a presentation. We will cut them off after 15 minutes but they don't need to feel that they use the full time. And so first, we're gonna have Lauren Trepanier from the University of Wisconsin in Madison where she's a professor in the Department of Medical Sciences in School of Veterinary Medicine, presenting on behalf of both her and Dr. Wen-Hung Chung. Good afternoon. I guess I would probably get parole if I, it was in that position right now, right, except for lunch. So I just wanna summarize, hopefully, accurately, what the first working group, the smoke mic is supposed to be working but it's not. It's coming in and out, so I'm taking it out. Okay. So I wanna summarize what the, now you can't see me behind the computer. I wanna summarize what the first working group talked about yesterday, focusing on basic research gaps. And I think that our charge was similar to some of the other working groups, looking at key gaps in research knowledge, barriers to execution, maybe some out-of-the-box or missing research perspectives, what additional genetic studies are needed, what's most ready for translation, resources, and then promising areas over the next five years. So as far as key gaps, one of the first we came up with was defining the cellular processes that lead to the development of drug neo-antigens prior to MHC presentations. So trying to understand what drives drug to the MHC molecule that then it gets presented to the immune system. And we kept that purposefully broad, but in the meeting we talked about specific things that you would do to address that. The second would be characterizing how specific culprit drugs activate immune responses outside of MHC restriction. Because certainly MHC is in everything. Characterization of cofactors that drive immunogenicity, whether that's a viral co-infection or organ dysfunction that leads to high plasma concentrations or metabolic polymorphisms that are maybe less of a driver, but contribute in the absence of MHC or contribute to the presence of antigen presentation with the appropriate MHC allele. And then we talked a lot about validation of early diagnostic and prognostic markers of the first onset of clinical signs and that because these are rare infections, rare diseases, you can't really get baseline data on all patients that are treated, but the goal would be to have a strong surveillance system that you could have collection of appropriate samples at very early in clinical onset before the skin is completely soft and things you can actually get maybe due to immunohistochemistry, you can get appropriate PBMCs, perhaps urine and plasma samples, things like that, as well as DNA to try to find diagnostic and prognostic markers. And then because it's such a catastrophic disease, drug re-challenge, even though it's a gold standard for diagnosis for adverse drug reactions, it's not possible. So some reliable, confirmatory and safe in vitro challenge tests. And I think we all know what the barriers to execution are for rare diseases, especially ones that maybe have a complex phenotype is first of a critical mass of patients that are well phenotyped. We've talked about that a lot already. And then in order to get those patients and get appropriate phenotyping, you need to see them early. So you need a sentinel surveillance system that can catch patients early. And then appropriate biobank samples from early onset of disease, so not just DNA, but also potentially PBMCs and RNA later, urine samples, plasma samples, marginal biopsy samples that are gonna give you good tissue for histopathologic assessment. We talked a lot about animal models. Certainly they're lacking. There wasn't a lot of enthusiasm for jumping right in with them because we know that many of these reactions are highly HLA related and animals obviously lack HLA restriction. Yes, you can humanize animals, but since there's other things going on besides HLA, those models may not be successful and initial attempts have not been successful. And so we sort of came to the consensus that animal models could be used to test specific pathways but only after you understand the human phenotype better. And so they should be used in a more narrow range rather than just saying, we're gonna try to reproduce this whole syndrome in animals. As far as missing research perspectives, one question was, are drugs being recognized in similar ways to viral antigens? Does viral co-infection trigger these reactions? Do these reactions trigger viral recredescence? And what is it about the epitopes that are being seen? Do they mimic or cross-react with viral antigens? And then the need to look just beyond T-cells. So we know that these reactions are probably mediated by CD8 T-cells, but what are the other players? NK cells, T-regulatory cells, dendritic cells, and then looking at checkpoint blockade molecules to see if they are down-regulated or up-regulated in these patients. We also talked about chemoinformatics. So looking at drugs that have been associated with TEN and looking at their structural motifs, and certainly this has been done for other drug hypersensitivities in general, but specifically for TEN, and see if there's motifs that are in common or motifs that track particularly with certain HLA associations that might help with drug development as you go forward. And then we talked about additional genetic studies. We didn't spend a lot of time on this because there was a consensus that, yes, we need to look for more HLA associations because they're drug-specific and they're ethnicity-specific. And so get a wider toolbox for advising different populations. OXCAM type NSAIDs were high on everyone's list, Lomotrogene, of course, and then ingredients in cough medicines which have been associated with TEN, particular ocular complications, but it's not known which ingredients are actually important. And as far as most ready for translation, this is maybe a little disappointing, but really, as far as translation in the next year or so, we really felt that predictive genetic testing is really the only thing ready for translation is sort of already in translation. And that some of the other things that we talked about as far as targeting certain molecules like granulicin and things is just not ready for primetime yet. I think we all agree that we need an expanded consortium with international leadership. The United States is, sounds like we're behind the APOL partially because we don't have single-payer healthcare, that may be part of the problem, and our healthcare is very fractionated. We talked about a public-private funding model with shared responsibility from pharma, talking about not just the deaths of patients, but the death of drugs, so that there's buy-in from the drug development side and the drug safety side as well as from the physician-patient side and from the FDA side. And this infrastructure needs to provide the ability to screen and biobank patients in their early onset phase and then an international registry of patients that has standardized phenotyping, ideally based on some of these early phase biomarkers that could be developed and then molecular signatures within the biopsy itself. And these would provide the basis for adequately-powered clinical trials. So what's promising over the next five years? Investigation of differences between macular-popular rash and STS-TEN as therapeutic darkens and Lars French is already working on this, sort of, what are the triggers that take a benign course into a catastrophic course? And then biomarkers in the acute phase to facilitate diagnosis, so what are specific phenotypes? Obviously, they need histopathology, but there may be subtypes or there may be markers, I mean, histochemistry or QPCR or other things within the biopsy that would allow you to clearly define different subtypes. Using those biomarkers to then go back and look at prognosis, so who will progress? Which patients need more intensive therapy? What are negative prognostic indicators? And then, hopefully, treatment targets. So if you know what's changing in real time in that acute setting, you may know what you need to target to check that fulminant reaction that's occurring. And I think in the larger group, was talked about massive parallel sequencing of HLAs linked to medical records outcomes, so you've got this background of genotyping that's already in place as you go forward rather than trying to work backwards after the reaction occurs. And the buy-in for that is that, you know, as someone pointed out yesterday, these HLAs have relevance for many, many diseases, not just rare adverse drug outcomes, and so there should be a drive for that to happen. And then we also talked about predictive tests in addition to HLAs. So pathway analysis of GWAS data, which we're near talked about. So not just looking at the high-hit HLAs, but what's the collective impact of polymorphisms and say, you know, P450s, phase two enzymes, transporters, things that are gonna affect the amount of drug that's actually available to adduct and or associate non-covalently with an HLA. And then there's clear evidence that plasma drug concentrations actually probably are important, you know, so we're sort of going back to metabolism again. And so again, in that cute phase when you're collecting samples, looking at urine, looking at plasma, when you still may have washout from the drug, may help you understand whether there are metabolic drivers as well as HLA drivers. So I'll sit down and then open the floor for discussion. Open for questions. So there's a plethora of ideas and things to be done, but a challenge, and we've heard that several times today, is funding this. Several of us are doing drug research on drug eruptions, but have to run something else in the lab because it's just not sustainable, working on rare patient data and getting publications out within a year. And also, the Biobanking Initiative going, is not just from DNA, but you said RNA, fresh tissue, PBMCs, is a lot in it. And one question, I'm talking to Japanese, I don't know if there's any Japanese here, but yeah. They have a system from what I heard for TN and I noted that down where the pharma companies contribute from the forefront out, a part of each medication for the sickness leave or long-term sickness leave and employment of the patients and also for research center in Switzerland, I don't know how it's here, but when you buy a MacBook Air, you pay when you buy already for the destruction and recuperation of bits of it. And then I was wondering when you were discussing from a FDA perspective, for you, it's also an issue having good quality data to be able to make the label good and that's gonna come with financing. If there's no way to try and get the companies to put a little center to off every package of pills going into that, it's in their interest, it's in the interest of everyone from the front. Yeah, we had talked about that a fair amount about public-private cooperation and so the various regulatory agency, the various research funding agencies of the various countries and their regulatory agencies but also pharma, it seems like an obvious source of funds and I'm not sure the logistics of making that happen in our political climate, but it seems like that would have to be part of the puzzle. Because as pity they come running when they end phase one or phase two have a severe adverse eruption and then they wanna invest lots and really have results in a month. That's not the way it's gonna work. So Cynthia, were you? Yeah, I was just gonna ask if Taiwan could, because they have the Taiwan Relief, Drug Relief Foundation, can you tell us a little bit more how do the drug companies contribute into that? Yeah, but virtually most of the funding are for a patient, not for research. So far, only not so many fundings that they can do for research purpose. So, but we did raise some funding for the drug company over every year, so we get more than three million or four million US dollar per year. But I think in the future we can push the government to do some research from the foundation. Linear. So we discussed this and funding is crucial here for not only for initiation, but also for maintaining any resources that you have as well as undertaking studies on the samples that you have with the new technologies which are coming through. And as you know, all those new technologies increase the cost of each experiment that you do. But the majority of drugs that actually cause problems are generic, off-patent. And so therefore this is not a problem just for farmer, it's actually a problem for healthcare, it's a problem for regulators. It's probably unlikely that one funding agency can fund all of this. And so therefore it is probably important to have a global consortium to be able to do this. So NIH working for example in the UK with Wellcome Trust, MRC with the Japanese funding agencies, et cetera. And that way that may be the only way to be able to lead some kind of sustainable funding in an area like this. Obviously there may be some small areas which may be focused and so on which could be funded by NIH, MRC, Wellcome Trust, et cetera. But here in terms of a global perspective that one really needs to think about how one can develop, get the funding in, but also maintain that resource that is developed from that. Okay. So maybe, come as no surprise that people recognize the need for funding and that you're going to tell us that we should fund it. I think it would be very helpful to spend our 15 minutes really kind of focusing on what is it. And what are the high priority studies? What are the key needs? What are the things that are obstacles as Lauren kind of laid out for us, I think very nicely, but in terms of a consensus of the group, where do you see the real pressure points where we could make some real progress? Neal. Okay. Well, since I was recognized by the chair before you said that, I just want to make a comment. I think puts it in some perspective and the disclaimer is this will never happen here. In Sweden, there's no fault insurance that the manufacturers pay into and in Japan too sort of that sort of description. In other words, if a patient has suffered harm from a drug or a vaccine, there's a central fund of money to help recognize that. And what it does is it recognizes the physician who prescribed the drug can still talk to the patient. They don't have to run home and hide under the bed. They can keep their relationship with the patient going. They can initiate the patient's participation in that fund to be recipient and if adjudicated, it's usually settled in about six months guaranteed. And it's not years and years and years of litigation hoping that you're going to win some big lottery. So what it does is it keeps the patient physician relationship there, collects wonderful data, funded by the industry, no fault anywhere in the world, you can't go sue somewhere else. You can see the obstacles here. It's actually, it's intensely sensible. It's a good use of money and extra money could be used for research. But if there's any lawyers watching, you're probably already calling your congressman saying don't ever let this happen. And because that's who would end up losing in the end is the lawyers. So thank you, Neil, but I think I'm gonna come back to Terry's point and bring the question and maybe I'll start with Wenhong by just by asking you to chime in. As you think about the items that came up, if you wanted to pick your top one, I know that's always a pressure thing to do from the discussion yesterday. What would you lead with? I think most important resource issue we should do is to get a good sample by banking over the sample, especially from the very early stage. So it's very difficult because most of the sample we, now we call it is expected to call the patient back to get the problem. But if you were trying to find out the biomarker, the circular target, we need some very early samples. It's very, very difficult for most of the researcher to do that. So I think we need to have a core whole and the perspective and the kind of the international consulting to collect the more sample size. I think it's more, it sure has many, many other people in risk of doing that. I think there's some infrastructure here that you could just go forward and say, all right, this is, these are the types of biomarkers that have promised for development. Plasma proteomics, just doing IHC, QPCR on biopsy samples, looking at PBMC, cytotoxicity, things like that. These are the samples that we would need and focus funding on study those kind of studies that are gonna develop a bank of biomarkers. The goal would be to have therapeutic targets because you're never gonna eliminate the disease completely, but if you have a good therapeutic target, then you could at least decrease morbidity and mortality. Sir, I think it would be easier if we do it in a clinical trial of different interventions, like because right now we do not have exactly the efficacy data of various treatment options, IVHG, cyclosporine, corticosterone, and if you run the trial in Taiwan and then in that trial, you can get all the samples because when we enroll the patient into the clinical trial interventions, this is, we need all the samples from them, right? And we monitor, we can monitor them through until the final sequelae after the Steven Tens was treatment, either patient die or they have complications or survive. And in that case, we will be easier to get, I mean, the biomarkers if there is one. So I think a large clinical trial to test the efficacy of various options may be good. Thank you, Mark. Yeah, I just carry on with that. I think a clinical trial is good, but one of the things that struck me about the presentations yesterday is that the identification of the disease is probably way too late in terms of really looking at effective interventions. And this is not specifically related to the basic science group, but I think in terms of takeaways for today, this would relate to some of the things that we talked about in group three about surveillance and identifying early markers of individuals that are in an early stage of Steven's Johnson Tens where, and if we could in fact accomplish that, then that would I think support a more effective enrollment into a clinical, into an intervention trial. Yeah, I was actually just thinking the same thing, and I don't know if some of the people around the room who are really involved in this can talk about what you see as challenges of getting earlier onset identification of cases. Mark. I wanted to follow up on actually what Mark had talked about yesterday, that the numbers are too small to actually get the cases of interest from a clinical trial, and also you get them too late. So by the time you get them, the thunderstorm has already happened. So you're just getting the footprint of what already has happened, which is a lot of different injury markers. They're not necessarily the harbingers, but what you can have, what was mentioned yesterday is the positive, the patients who have the positive marker, for example, carbamazepine in that marker, 1502, the patients who don't get the reaction. And so you can look for markers of people who don't get a reaction and ask what in that mix of, you know, from a biological perspective, might explain the protection effect. You know, if you had an animal model, you could study, for SJS, you could study it serially on a timeline from initiation of exposure and then every step of the way. But since you don't have that, which would then be ideal, then you could kind of find the early marker, the harbinger. But what you could do is look at individuals at a later point in the clinical trial who don't get a reaction or will get a minor skin reaction who have some adaptation. Lars? So if I had to choose from all, I think it's pretty clear that it's immunologically driven. It's pretty clear also that some of the patients who have a defined HLA, which is a risk factor, some of them, despite that, won't develop the disease. So a co-factor is needed. And I think research for that co-factor would be really important. If we could target that, we would have a lower incidence and we'd probably maybe be able to turn it off. HIV is a co-factor, and ampicillin EBV is a co-factor. So there's some evidence. It may be a pathogen. It may be signaling through innate pathways. And I think that's really something that's been under-researched and has to be looked at. And do you think the appropriate time to look at that is as the early surveillance and when the reaction's ramping up or before the patient is treated? Yeah, I think we have to really get as much samples as we can. Early samples, PBMC and skin samples, and then run them with the techniques, beautiful techniques we have to try and get leads. But there's quite some in vitro stuff you can do too, with hypothesis-driven and not non-biased, really, hypothesis-driven on PBMCs or on mixtures of PBMCs and keratinocytes. And some of this you may be able to do by having the patients, knowing the patients and going back to them six months afterwards and getting the PBMCs and getting keratinocytes from hair follicles or from biopsy, putting them in culture and working with this. We know that we can do the lymphocyte transformation tests a few months after and that works and those patients will survive. So it's not only the acute phase, it's getting material also later on in vitro. I think regardless of fundamental infrastructure is needed and the capacity building is going to be key. And I think what maybe differentiates this from other initiatives is the need for broad international collaboration to get the cases. And I would be interested from Terry and others in terms of within the NIH framework and FDA framework and maybe broad collaboration across not just the NIH, but other FDA, other regulatory agencies internationally given the international presence here and the need for international collaborations are there ways to actually facilitate that or make that happen? Well, sure, I mean NIH does, we probably fund more research internationally than any other funding body funds outside of their country. So there are many good examples of collaborative projects that we've done as Munir mentioned with the Welcome Trust and the MRC and a whole variety of other agencies. I think that the bigger question may be how would one sustain such an effort because one of the challenges in the area of drug safety is that it never goes away. And once you get started down that road which is an important road to start down, it's not clear that the research agencies can sustain that effort. So are there ways then to work with other agencies? And we've talked about a few of them here today in terms of the US, the CDC, our Centers for Disease Control, our Agency for Research, Health and Quality. And those may be agencies that could pick up some of these efforts down the road. But I think again, first rather than perhaps sort of figuring out who's gonna fund this and how we make it funded, we need to define what the what is. So if an international collaboration is something that sounds like everybody is enthusiastic about, what are the key things this international collaboration would do and how is it different from those that already exist? Biobanking is sort of the fundamental sort of first step is getting samples that aren't currently being collected which includes PBMCs which is not and tissue and so there's been a lot of work on getting DNA and DNA banks but I think we need to broaden our repertoire of samples that are being collected. So DNA, RNA, serum, plasma, cells. Okay, so Moonier and then Lars. So we've heard that so many people have biobanks and collections already. I guess the question that you asked Terry is what is it, what's different? I think scale and upscale what we do and tremendously increase the scale of what we do by this global collaboration. That's number one in terms of difference. I think in terms of key factors that we need to look at, we need to look at the predictive biomarkers, the diagnostic and prognostic biomarkers and key for that particularly for the latter two will be identifying patients early. Now that is a challenge and you'll have lots of false positives but getting those patients identified earlier as was stated earlier today and we discussed that extensively yesterday will allow you to be able to identify those kind of diagnostic biomarkers which allows them to be able to undertake stratified trials in smaller numbers of patients which at the moment you can't do because it's such a rare disease so undertaking trial in this area is going to be immensely difficult. So having those kind of early markers which allow you to be able to stratify undertake kind of Bayesian type trials will be the only way to be able to then really test some of the therapeutic interventions. Lars. I just wanted to add on to these two comments that biobanking can't go in my opinion without a registry and that's as we discussed before with well phenotyped cases so that you can have the clinical data to it and then one big element if we had to have an alliance together is talk the same language. So harmonization for the clinical characteristics we're bringing in our registry so that we can really exchange well. Yeah and I think part of that is also thinking about expanding to areas with high numbers of cases that don't have the infrastructure in place yet and so sort of both retrospectively putting together groups that are working and then thinking about how to best move forward into new areas as well. Steve and then we're gonna move on to the next group. I think together with the early identification with patients in collection of biomedical samples whether it's plasma or cellular materials there needs to be centers that can receive them and work on different parts or touch different parts of the elephant and so Lauren and her summary of what their group did identified the various steps between drug administration, drug bioactivation, processing of the neoantigens all the way through to the immune response probably you're looking at four or five specialized centers that can work on each of the components that might be involved but there also needs to be a way of sending peripheral blood mononuclear cells from Kansas City, Missouri to Switzerland or wherever that aspect of the big picture is going to be worked upon. And so I think there's a lot of elements toward making this happen that would have to be worked out as well. Very good point, thank you. So we're gonna turn over now to the report back from working group two. I don't know if Howard or Srikabeth is going to the podium.