Ep 60: Guest Mike Atkins

[Music] Malano-Mama says it's an educational resource. The concept represents our opinions and does not constitute medical advice. If you're a patient with question specific to your situation, please consult your medical practitioner. Hi everybody, welcome to Malano-Mama Matters. I'm Sutton Epithel. And I'm James Locke and Hi. Today we have another special guest, Mike Atkins, joining from Georgetown University. Thank you Mike so much for joining us today. It's my pleasure to be here. I'd love to talk to you guys about Malano-Mama and the immunotherapy. And maybe a little kidney, you might dip into a little kidney today. You know what I'm saying? That's fine with me as well. We think it's always fun for our guests and for our audience to hear from our guests. Maybe just a little bit about their background, where they did training. Are you an East Coaster, your whole life? I'm a little bit like Jeff. When Jeff was on, he told us about his sort of ping-ponging back and forth around the country. Can you tell us about your background? Sure. So I grew up in the Midwest in a town called Milwaukee, which is north of Chicago in Wisconsin for those of you who aren't in the United States. And I think from a family of committed cheeseheads, everybody else in my family went to the University of Wisconsin. And I went out east to Boston for college at Tufts University. And I stayed in Boston for 40 years. I went to Tufts University, Tufts Medical School, did my internship in residency and chief residency or Tufts New England Medical Center. I did my fellowship in Hemant at Tufts and was on the faculty there for 10 years without moving, without ever having to interview for any job or anything like that. But then I moved across town to Bethesville Deaconis in the Dana Farber, Harvard Cancer Center because of the opportunity presented there to work with really talented scientists across the Harvard system. And I had a lot of Bethesville Deaconis. I led the kidney cancer program. I founded the kidney cancer program and the kidney cancer score and co led the melanoma program and was deputy director of the melanoma sport of the first melanoma sport ever, ever funded. I also led the biologic therapy program where we were heavy into high dose, into looping to treatment for patients with melanoma and kidney cancer, probably the largest volume of any place in the world. And I was probably treating more patients than the NCI surgery branch at our, hey, day. And, and so we were very committed immunotherapists, but we knew that there was couldn't get above the 15% ceiling with high dose, I helped us. And I spent a lot of time trying to figure out how to make a new therapy work better for patients and we thought about combining it with chemo therapies, combining it with targeted therapies, combining with anti-angogenic therapies. Just, we could get more responses, but we couldn't get more cures. And so we delighted when the checkpoint inhibitors came along and we saw a ability to treat not just more patients with melanoma and kidney cancer, but also potentially working in patients with other types of cancer. And this was different. And I said, if I'm going to actually do this, I need to break away from the current mold and go to another institution to try to lead this effort and introduce this effort in other institutions. So I moved to Georgetown 12 years ago, where I'm deputy director of the Cancer Center and a way that our immunotherapy initiatives and able to stay involved with the immunotherapy space, particularly my colleagues in Boston. And within the melanoma community, which is a very tight knit community from decades of banging our head against the wall to actually be able to participate in this revolution in cancer therapy. So maybe if I could ask a first question and I suppose it's about cytokines because you know, back in the day, that was really the only game in town I suppose for immunotherapy. So my question would be now, 2024 going into 2025. Do you see any role currently for cytokine treatment that sort of part what and then part two would be what do you make of some of these sort of reformulated dial two that we said a bit of in the last few years. Do you think they may find a place or not? Or does there great questions things I spent a lot of time thinking about not use how those aisle two since the last Boston. So now that close to that date and before, but I do think that aisle two still has a role, particularly in cellular therapy, it's part of the child program. I don't think it's something that we would give commonly to patients with metastatic melanoma or the third or fifth is in there. But I think it was very instructive because it told us that we could cure some patients with solid tumors that the immune system actually works and attempt to feel the live while we're trying to figure out how to break through the 15% response barrier. I think that we knew that from the work of Steve Rosenberg and till that there were T cells in the tumor that be nice to tumor and that we could get them expanded with enough with aisle two we might see a response. But we didn't know what was keeping them from being more active in the tumor micro and if you took them home, they were able to recognize and kill the cancer, but living in the tumor. I think those didn't just work the checkpoints discovery opened up that box. So getting back to our experience with high osi2 and how that informed these new engineered aisle two's we saw early on and I was just looking at this data for because I was talking to some people about the old and days. And we saw that when we gave high osi2 not only did we activate. Non specific T cells throughout the body, which was responsible for the toxicity, but we also activated T regulatory cells and we saw that not only CDAT cells went up in the blood, but T regulatory cells went up in the blood. And we looked like the patients who responded were those that had some defect in their T regulatory cells where they didn't expand and that sort of linked the response to aisle two to the aisle immunity that we also saw associated with response my first one of my first papers I wrote talked about thyroiditis. And with a high dose aisle to therapy and it's association with response and so I find it amusing that that child has been rediscovered. But the way we thought to improve the therapeutic index for high osi2 after we couldn't do it by combining it with chemotherapy and combining it with inner fear ons and combining it with other targeted therapies was to see if we could. Selectively activate CDAT cells and so. And naively we thought that maybe having non alpha aisle twos would do that because the high affinity receptor tended to be on T regs and if we didn't bind to the high affinity receptor maybe that would actually. And then we were all for CDATs to be activated without T regs the trouble with that assumption and it was naive when we think about it is there was no selectivity for doing that within the tumor micro environments and so you're going to activate CDAT cells in the periphery and in the tumor. And then we had to do so in a way that was not toxic it was going to be low dose aisle to which was not effective. And so after a lot of failed attempts that I think were in retrospect predictable I think the field is moved towards trying to selectively activate tumor specific T cells within the tumor micro environment was cytokines. We're looking at masks, cytokines, we're looking at IL-2 that's targeted through cells within the tumor microenvironment, either through targeting PD-1 or targeting hopefully CD-8s, which may be more concentrated in the tumor microenvironment, or there was a nice presentation on James Gully at SIT-C, looking at a bi-specific that brought IL-2 into the tumor microenvironment by targeting a specific V-Beta cane on TCRs in that was enriched in the tumor microenvironment. And I do think that those types of approaches and a number of companies are now working on it, with others much support, is I think they should be receiving will ultimately work, and will be supportive of not only increasing the activity of checkpoint inhibitors, but also may help with T-cell engages, anything where we want to get more T-cells working in the tumor microenvironment. So I have not given up on those. I wonder if I can make the insurance relational? Oh sorry, are we having that? My internet is going in and out a little bit. Sorry if I'm jumping on new James. No, it's not. I'm going to do it. Thanks, thanks. The interrelational IL-2 that we saw at ASCA, the Darimune, you know, it's interestingly called a pivotal trial, but that was just the name of it. It wasn't necessarily it's stature. I wonder what you think about that interrelational working on T-regs and CD-8s in the tumor, or is that too sort of pedantic approach to just shove the IL-2 inside a tumor? Well, we used to do that. We did IL-2 injections into tumors like 30 years ago and it did work on that tumor, but I don't think you get there. He sells to be active and go to distant sites with that approach. And so, you know, I think that maybe finds or some things where it's really not there. I think the best of ultimately, I think you want to get to all tumor. Yeah. Great. Go on, James. No, we're having a few technical glitches, but we'll keep going anyway. So, this is a bit of an unscripted question, but it's something Sapner and I talk about a little, at least a little. It's about complete responses to therapy. So, if I'm honest, much of the side-to-kind stuff was a little bit before my time, but I was sort of talk growing up by Martin Paul that if you're treating someone with side-to-kinds, then you really need to have a complete response to therapy to have, you know, properly durable benefit, shall we say. So, do you think that's right? It's the first part of the question, and the second bit is, do you think we should be thinking exactly the same thing with other immunotherapies, checkpoints, inhibitors being the obvious example? Do we need to really be chasing CRs? That's what Sapner and I call it, this idea of chasing a CR up front to really get the durable benefit. Well, yes and no. First of all, I don't think that a CR on a CT scan is, or a good PR on a CT scan might be a CR. And so, I don't think the CT scan should be the final arbiter of whether you've eliminated the last tumor cell. We know that from the neoregiment treatments, but that is true when you give things systemically as well. And back in the day when we were calling King CRs, we weren't even using CT scans. We were using like test X-rays or Steve Rosenberg used to use tomograms or things like that. So, as you get better and better at imaging, the ability to actually eliminate everything on a scan becomes less and less. So, I think a CR is necessary. You want to get rid of the last tumor cells, but not a resist CR or radiographic CR. We just need better ways of proving that, which is why we've tended to, when we wanted to stop therapy and there was still something there and still something there on the PET scan, pretended to biopsy or insect delusion in nature. It was no tumor there than we assume that all the disease was probably gone. And I think, you know, in the future, we'll have alternatives to biopsy. The tumor to show that the tumor is gone hopefully with sensitive CT DNA or other approaches. But that is the goal to get rid of the tumor, put the patient in remission, stop the treatment, and get them back to their pre-cancer life. All viet informed by their experience that they went through in facing their cancer. And we call those patients who have stopped therapy. Their disease is still under control, not just survivors, but drivers because they're back contributing to society in a robust way. I mean, I wonder if that just sort of begs the question, Mike, we all know your IL-2 paper, obviously melanoma and kidney cancer with the plateaus of CRs and PRs. And really, you know, if you get that at three years, that's what God is thinking into, you know, response at three years can be quite durable, regardless of its CRPR. But I guess it begs the question, what were you using back then to call a CR? We were using CT scans that weren't as good as the scans we had. And we didn't have things like central review either. It was just us going to the radiology suite, pulling out the scans which were on films and measuring them. Technical difficulties. People are frozen and hoping for me. No, we're, this is like a real one. When I was, when I, yeah, going to go into. No, don't worry about the screen. Yeah, we don't need to worry about the screen freezing. It's an art in fact. We just, we just as long as we can keep talking we should. Yeah, now when I was training, we still very much one of my jobs as a junior doctor was, you know, at the back of the ward ran carrying the envelope with the old fashioned scans on. And then when I asked to bring the scans up to the front and put them up, yeah, fun memories of that. So, listen to maybe come a little bit more up to date. I'm sure people want to hear about DreamSeek, which has been an incredibly important trial, I think, over the last several years in terms of how we treat that's a static mother man. So, be great if you could tell, say a few words about that, maybe the background and the results. And some of the subgroups, I think, it's all about this at six. You just go on very interesting. Yeah, so, clearly huge advances in the treatment of metastatic melanoma, particularly B-ref, utin melanoma with combination B-ref mechanivators, which ended to play the progression free and overall survival curves and the combination checkpoint inhibitors, which had less than the effect on the early part of the curve, but tended to raise the tail. So, in 2015, just as both of these combinations were approaching approval, we launched the DreamSeek trial, which took patients with metastatic B-ref mutant melanoma, randomized them to either getting combination Nebripe upfront or combination B-ref Mac in this case, Tab Tram. And if they progressed and met the eligibility criteria, they crossed over to the alternative therapy. And one of the key things about DreamSeek is because we thought those curves might cross, we didn't think a hazard ratio was the appropriate endpoint. We thought we should look at a landmark overall survival. So, we chose two-year overall survival as the goal as the primary endpoint. And we estimated, and this is one of the things I'm proudest of, we based our statistics on the fact that we thought the Nivo Ipe would have 70% two-year overall survival and that the Tab Tram would have 50% overall survival. So, we predicted a 20% difference in two-year overall survival. So, the study was there to do. We were on like amendment 23. We were accruing about one patient a month. We would probably still be doing this study except for the Data Safety Monitor Committee was looking at the data and they saw this statistically meaningful clinically significant 20% difference in overall survival at two years, 72% versus 52% favoring the immunotherapy first time. So that stopped the study. About 90% to crude only 59% of the patients had passed the two-year time point. And then we struggled to get that published because we hadn't completed the trial. So one of the things that we thought about was why was immunotherapy better? And there are four major reasons in my view, why immunotherapy first was better. First, it was better in all the subgroups. Even the subgroups of patients where we thought targeted therapy did best those with low LDH, good performance status, less than three menestay, excites of disease immunotherapy was better. The duration of response was much longer with immunotherapy, 80% of 88% of the responders to immunotherapy were still responding. And only half the patients who responded to targeted therapy were responding. Third, there were the relapses that we were seeing were more frequent in the CNS with targeted therapy than with immune therapy. So immune therapy was working on that subclinical brain metastasis preventing those relapses while targeted therapy didn't do as well as that something that we verified subsequently in the Checkmate 204 study. And then finally, targeted therapy worked as well after immune therapy as it did in the front line while immunotherapy didn't appear to work as well in patients who progressed on targeted therapy. And because of all that, it was stacked so that the immunotherapy first would be the better arm. So we wanted to see if we could identify because the curves crossed and there were some people who died on the immunotherapy arm that there were more deaths on the immunotherapy arm in the first 10 months than on the targeted therapy arm. We wanted to identify who those were and clinically those were patients with aggressive disease who never got targeted therapy at least on protocol. But we wanted to look a little more closely at that. We did a modified forest plot where we looked at PFS for the initial treatments to your PFS and looked at the difference into your PFS. And immunotherapy was better two year PFS than targeted therapy in every clinical category which was surprising to me. This is in the cooperative group. This is not just a selected institution. And for overall survival, although the difference was not quite as robust for many groups, there was a significant difference in overall survival for the immunotherapy first. And no group was the dot for the difference favoring even non significantly targeted therapy first. I want to pause for a second because another thing that surprised me is that the response rates were the same to immunotherapy and targeted therapy. And that's really an artifact of how we when we chose to measure the responses. In the targeted therapy studies, responses were measured at eight weeks and 16 weeks because we wanted to have the CT scans for targeted therapy done for immunotherapy done at 12 weeks and 24 weeks. That's when we measured the responses for targeted therapy. And so there were plenty of patients who were responding at 12 weeks. But by the time they got to 24 weeks, they had new disease or the disease was growing or they had brain mitts who never counted as a response, who would have counted as a response if we had measured their scans at eight weeks and 16 weeks. So that's why you might see that being different than what was predicted. So we wanted to see whether we could identify because we had tumor tissue on patients, we could identify looking at RNA seek on the tumor tissue, a subset of patients who might do better with targeted therapy. And we looked at a lot of different things and immunotherapy did better in most of them. But one of the things that was curious is that if you had an angiogenesis high signature, you did poorly with both targeted therapy and immunotherapy with 25% two-year progression free survival. But when we looked at the patients with angiogenesis high at their overall survival at two years, it looked like the targeted therapy first was above the immunotherapy first. And we tried to figure out why. And so we looked at what happened with second-line targeted therapy versus immunotherapy. And the patients who got immunotherapy in the second line after progressing on targeted therapy tended to have, and these are the patients who actually got to the immunotherapy. Most of those patients, I suspect, didn't get to immunotherapy because they had aggressive disease. But the ones who got to immunotherapy had a response rate of 70%, compared to the 25% response rate that you have in the frontline with immunotherapy and targeted therapy did not work well after immunotherapy and those patients who still had aggressive disease. And so that suggested that in that group of patients with really aggressive disease, that a short course of targeted therapy, similar to what was done in succumbent with the sandwich arm might be the best approach to allow the tumor to shrink to allow the hypoxia to reduce and to allow the immunotherapy to fully be able to work. And so we're now going to validate that with the second half of the tumor specimens, but I think it does create an approach to patients who are presenting. Keep in mind that the patients with really aggressive disease probably didn't go on this trial because they didn't meet the eligibility criteria and those are patients we were probably giving targeted therapy to and we weren't sure when to switch to immune therapy or how we could do it, but now succumbent helps us know how to do that and dream seek helps us understand why that's a good approach. Yeah, I mean Mike, that was an absolute master class in dream seek for people who meet to know the ends and outs of it. I will say I think you were so careful when you designed the study to a lot of people it may have just seemed like well what do you give first type of study, but just the fact that you knew you had a really strong hunch that these curves would cross statistically you knew what to do there. And it's incredible, James and I have talked about it on here before, but there's other studies that have come out. If they had just understood rapid responses versus you know, takes a while to get a response and then durable, if they would have understood the crossing they would have known not to power for a median overall survival and a hazard ratio, it's a landmark. So we saw you know this asko with one of the targeted and you know is that you know for some reason they had an OS as their primary in quite which really didn't make sense particularly after dream seek. I think it also tells us you've taught us so much about dream seek over the years in terms of the patient population that went on. We know e-cog 2 you know patients if we're putting on an e-cog 2, we know they're really a three, we know they're you know slipping and we're getting them on study, but to your point the sickest of the sick are probably getting the target at therapy first. And as James was pointed out when we talked about dream seek, the dream seek probably actually under estimates the benefit just because of some artifacts that were built into the study, the wash out period, the you know the fact that you have to reach your immunotherapy in order to see what the benefit is etc. But I think all this angiogenesis stuff is really relevant for kidney cancer as well, right? James you seem to smoothed out with the connection so maybe I'll give it back to you if you want to ask a kidney cancer question or two. Yeah I mean for those who didn't see a talk at Sitzy it was obviously a title after my own heart as someone who also treats kidney cancer and malonymer, there's a few of us I guess around the world. Hopefully more actually I would say in the future because I think that's quite striking similarities actually between the diseases and the way we treat them and I was there at Asco last year actually in the audience when there was a discussion which you started which really is about in kidney cancer what do we do with first line metastatic treatment if we put to one side the issue of adjuvant therapy and kidney cancer for a moment. Is it ipi-nevo or should it be the Jeff targeted therapy plus a check point inhibitor and I have to say and I think you know this already I'm 100% with you on this particular question but it would be great if you could just say a bit about that for the people that listening because I think it's very instructive. Well so this is one of my great frustrations and I hope I'm making progress in this area. So first of all as everybody who treats melanoma knows the because the immune system can target multiple different antigens simultaneously and broaden and deepen over time it can eliminate the last tumor cell and produce cures and therefore the hallmark of an effective immunotherapy is this tail on the progression free survival or overall survival or duration of response or most importantly treatment free survival curves and that's second nature for people who treat melanoma to look at the right hand side of the curve the tail rather than focusing on the left hand side of the curve but G.U. oncologist who treat prostate cancer mostly don't understand that and they're stuck looking at the left hand side of the curve and so I also have the advantage of studying angiogenesis in addition translation in patients with kidney cancer and what we saw in our mouse models and reproduced in patients is looking at perfusion MRIs when you give a VEDGEF receptor TKI to a patient with kidney cancer you see in the tumor and in function within three days the blood vessels disappear within the tumor and the reason why that happens is because kidney cancer has a VHL mutation and therefore it is able to have VEDGEF around without causing hypoxia and therefore it never needs to develop a sophisticated angiogenic drive the only way you get that angiogenic drive is after you block VEDGEF make the tumor a futile hypoxic and then you cause up regulation of other factors which lead to the resistance to the VEDGEF receptor TKI is what we called angiogenic escape and to me that's the worst time to try to fully activate the immune system is in this hypoxic tumor micro environment and so we were involved in some of those early IOTKI trials I led the first phase one trial of PEMBO plus exitnib and we saw this beautiful response rate 70 plus percent we saw a median TFS in this really favorable risk population of close to two years and if that was great and I was following these patients and then I had a few patients where I had to stop the TKI for one reason or know that they were having a stroke or they needed surgery or something like that and their disease exploded and so I'm thinking where is the immune response in this response because in immune therapy that would not happen and so this looked like it was just accentuated VEDGEF targeted therapy and so we looked at those curves and those curves look like TKI curves they don't look like immunotherapy curves there's no tail on those curves as more time goes by they just get lower and lower and those patients are still on treatment they're on treatment for the rest of their lives putting them on a TKI which is a great treatment is a one way dead end street you can't cure them anymore and so we then tried to focus on just making the immune therapy better and so the first thing we did after the luvaripi trial and which I think vastly underestimated the efficacy of nivo-ripe and kidney cancer because people didn't know how to use it and they didn't know how to continue using it after they had a toxicity and they didn't know when to switch to a targeted therapy and patients who weren't responding which is what happened and why the favorable risk patients did so poor on that trial but we did the nivo only trial what we call the HCRN trial where we biopsy to metastatic lesion before we gave patients nivo-lamab and then we if they progressed we biopsy again to prove their risk cancer there and now it always was and then we switched them to nivo-ripe and we've learned a lot from this trial we've learned that first nivo as a decent response rate it works well in favorable risk patients that the nivo-ripe values on 214 were probably because those patients never got targeted therapy as we've now shown and that I think the whole IMDC categorization which was developed for VEGF receptor TKI's doesn't apply to immune therapy we knew from a rial 2 days that those favorable risk patients responded probably the best to IL2 and if this VEGF was actually synergizing with anti-PD1 you would have seen those patients do the best but their survival was no better than starting with synonym and so we also learned about all the things associated with resistance to immune therapy and things like PD1 on T-regs or exhausted CD8's that have lag 3 and Kim 3 on them or SLAMF7 expression on subset of CD8's or whether they're not there's for shared lymphoid structures in the tumor and some of those gave us ideas for how we could improve frontline immune therapy and we're now pursuing that in clinical trials through our clinical trials group and hopefully we can raise the bar above NIVO it be cure more patients and convince these physicians who think that they need to get the best initial response and prevent early progression and that's what their patients want instead of curing them we can convince them that they should give a pure immune therapy first and that's the best way to cure patients and that should be not just for ear me or poor or so comatally patients but for all patients and less they're in some sort of clinical crisis yeah well that was that was a tour of course a wonderful summary of I think where we are today and also the history of immune therapy to treat cancer over the last several decades so Mike I know your schedule is extremely busy so I'd like to thank you for joining us today can I close with one thing yes so we talked about melanoma and kidney cancer but I think this approach is relevant for a lot of cancers and I just wish the people who are treating those other cancers with immunotherapy would consider giving the best immunotherapy first if it's in the first line metastatic setting or as Sapna has shown in the neoadjuvant setting we should be going for cures not for improving median pfs and I wish the FDA would value that more because that's what patients want it may not be what pharma wants but that's what patients want and as advocates for our patients we should want that to yeah sorry I know what what a great way to finish so thank you so much it's been wonderful to have you with us and I hope the technical problems didn't cause too much trouble for you so thank you very much good to see you my pleasure take care of you guys thank you you