Ep 67: SITC white paper on ICI resistance

[Music] Melanoma Matsas is an educational resource and the opinions we express are our own, do not reflect medical advice. If you have questions specific to your own situation, please contact your own clinician. This podcast may be available in both audio and visual, so please check your platform. Hi everybody, welcome to Melanoma Matsas, I'm James Larkin. Senator Fidel. Today we're going to be thinking a little bit about the SITC Society for Immunotherapy of Cancer, consensus definitions for resistance to immune checkpoints inhibitors, and actually we're probably mainly going to focus on the paper from Journal of Immunotherapy of Cancer from 2023 where parent Cleger is the lead author, because I guess it's come up hasn't it? A lot, what we've been talking about, interpreting original trials, this whole business about resistance and sensitivity, so we thought it was worthwhile spending a bit time on that. So before we do that, I spray could to you is have you ever walked out of the cinema or theatre, I suppose, and why and can you remember what you were watching and if you've done it a lot, maybe one of the more memorable ones. I ever walked out, nothing comes to mind actually being seated and then leaving. I'm pretty at least back in the day during my med school days and residency, I was pretty famous for being able to fall asleep in the theatre, so not walking out but sort of like being, I mean I could just fall asleep and not tell you what happened. It happened in the last year or two, I was with my trainer and his wife, also a trainer at our gym in Houston at their gym. We went and saw the Superman movie and we were, it was one of the theaters we can drink alcohol, so we were eating popcorn, drinking alcohol, and I probably lasted an hour. The seats were fully like lay flat seats and I always bring a blanket from home because you know let's just call it what it is a nap. So I but I haven't left anything out of being turned off by it or uninterested, I'll just at that point take a nap. It did make me think though years ago with my friend Shavon in Houston, this would have been before 2013 because I was in my old bungalow house. We got tickets to a radio head concert and so she had, it's the opposite of your question, have I ever walked out? It's being so late to something that it basically ended. She had just gotten a dog or was really heavily contemplating, so she wanted to come over, play with my Choey, we were chatting about the dog, maybe about how does it work, how do you travel when going home? Well, so when you say my Choey, that's not a euphemism, that's your dog right just so we can. That's not a euphemism. Yes, we were flying with my Choey, my wolf wolf Choey. Anyway, so we got to talking, we finally, we were like okay concert starts at 830, so they'll be an opening act, who's the opening act, okay they'll probably do four songs, and then set change, I'm gonna go, we probably have 45 minutes to an hour, so let's get down there by 939, 939, whatever. And so we go, we take a Uber or whatever, we get downtown, we go to our seats and there's some people in our seats, and so we said oh excuse me, I think you're in our seats, you're in our seats, and they literally said something like you've missed the entire, what do you mean? Like you can't possibly be showing up now for this, and we were kind of sort of like mind your business, what's the big deal, we didn't want to see opening act, whatever, and then we are there, we probably caught two songs, and the concert was over, radio head. Do you know which songs they were? Well, Creep was the last one they did, but no, I don't remember the next to last one, there's another one that I like, maybe Carmel Police was the last one, anyway, but it was pretty like the concert end, and we thought like oh is this just a set change, and then everybody started clearing out, and we thought oh we really did miss the whole, call we really timed it very badly, so we missed everything, that's different from your question, which is. Yeah, so I don't think I told this poor related radio head story, so they're from Oxford, and when I was in Oxford doing clinical, 93 to 96, and then afterwards, I was there, I did one house job in Oxford, my girlfriend at the time is a Rhodes scholar actually from South Africa, used to have a job. I remember that story. That's right, she used to have a job working in a Thai restaurant, which was radio head's favorite Thai restaurant, which is called Chang Mai Kitchen in Oxford, and it was great, and she used to get like, you know, take out some stuff like that from there as part of the button, so quite often you would go in there to like me to her after she was finishing work or something like that, and then radio head, this is before they were that famous as well, they're just be sitting there, all dressed in black, having their, you know, their meal at Chang Mai Kitchen, so. They're dropping me to that Thai. Yeah, I don't know if they've ever commented on that, I mean, on obsessive-fan, so they must have done that so I went into the, and I hope I'm not breaking confidence as they're. Right, okay, so I don't know if you want that. But have you watched me hear your walk out stories? Yeah, yeah, well the only one I can clearly remember is 12 years of slave, I don't know when that came out maybe 10 years ago, something like that, 15 even, and Emma, my wife and I were watching, I think it would have been Liverpool, visiting family, Emma's family, with I think my brother and sister and I, and so the four of us there evening showing, I guess we were tired and you know, kids were young probably, and so we'd probably wanted a relaxing evening in our defence, but we started watching 12 years of slave and it just got to that horrible scene where, well I don't know, probably the film's full of horrible scenes I don't know because we got to the first horrible scene and we just looked at each other and you know, is this so we always spend that evening we walked out. All four of you guys are just you and Emma. Just so much, and then we went and got you know, some pastor in a bottle of red wine and we had a nice evening. As in David. I'm sure it's an amazing film, but it was just so traumatic and it just didn't really feel like it was going to be the kind of evening we wanted to have and we haven't tried to, I don't think we'd even try and watch it again actually, because I'm not even sure if you get through that scene really easily and I'm not criticizing people that do and have and maybe it's my duty to watch that as well, but it just, it was horrible. It would be one where after, would potentially be hard to have words. Yeah. After watching something like that you just have to see this. Yeah, yeah, yeah, I'm not I'm not dismissing film or anything it's just that at least then we just couldn't face it. Side note Michael Passbender, one of my favourites, one of my whole past favourites. Well, maybe we could have him as a guest if we can think of something that you know, it's a wider interest. We did a full community film one time. Sorry. Can we make that a medical somehow make that medical? Yeah, whatever let's put that on the sort of listed potential episodes. Anyway, so maybe just to shout about this paper, I suppose the background is important, because I mean, it might be I've said this a million times before it's probably the biggest challenge we now having clinic in Malanama, which is resistance to checkpoints inhibitors. I guess I maybe think it back utaneous Malanama obviously. So we see it all the time. You have the various different scenarios, I suppose one is post adjuvant and then another one is metastatic primary resistance, never only benefit from treatment, secondary resistance, you get some benefit from treatment and then the benefit from treatment wears off. So we all know what it is, everyone is treating patients. And I think the problem was that it wasn't really codified in a way that was useful particularly for doing clinical trials and having definitions. I mean, the biology is heterogeneous in my view, we know that already. That's not really the issue here, I think the issue is about defining populations so that for clinical trials in terms of trial entry, we have a sense of what those populations are and also when defining output of trials, we can gauge what we think that means depending on the characteristics the patients going in. And I haven't said this to you before, but I immediately thought of platinum, resistance and sensitivity in a varying cancer, I guess days of platinum and my varying cancer, a lot of time and effort. Actually Martin did a lot of this Martin goal was defining all of this and I think for a varying cancer I remember when I was training. This obviously was the main issue you were thinking about with patients. What's their platinum sensitivity? Yeah and then how do you think about subsequent lines of treatment and so on and so forth. So those are my background reflections I guess. So as you said there's primary resistance. They're secondary resistance but there's a key part here which is drug exposure required. So in order to be called secondary resistance you need a few things you need initial benefit to therapy. So that's a CRPR I don't need or stable disease. Last thing six months right the stable disease has to last more than six months not necessarily the CR or PR but the drug exposure has to be six months. So then what do we call somebody with drug exposure of four months who had a partial response but lost it? So they wouldn't fit this definition very easily with that. So then I mean you were just talking about table two which summarises it nicely. This is the Harret Cleareder lead author paper and just a sort of I suppose book and that I'll comment on it for primary resistance. It says a minimum of two cycles of treatment and they're talking about combinations here just to be clear at both drugs. Brackets six to twelve weeks or actually parentheses six to twelve weeks have been really picky. Do you know about that distinction? Say that one more time. Brackets versus parentheses. Please. The limiting. Brackets are square. So you have to comment and mistake that. Look it up afterwards honestly. I'm not making that up. Yeah because I would have said the opposite in the US parentheses are rounded. Brackets are curly. Wow. Curly is an amazing discovery. But yeah so in British English that's the distinction I believe I think I'm right about anyway. So primary resistance you need to have had a minimum two cycles of treatment which generally will end up being six to twelve weeks and then the best response to that is progressive disease for stable disease lasting less than six months with the requirement for doing confirmed risk and at least four weeks after PD if the patient is stable. So I suppose you're making the point that there's something in the middle. Yeah I mean as early as this what happens if you have more than two cycles exposure or two doses exposure to ICI but not six months worth of exposure. What are we calling that? You know we all have had this scenario where somebody gets two doses of ICI gets toxicity of its combination therapy in the advanced setting stops. How are we characterizing their benefit if there was one or if there wasn't one. And then same with adjuvant therapy they might have taken four months of adjuvant therapy stopped for whatever reason are they primary resistant or they don't meet the drug exposure to be secondary. And so like there is still a gray area despite this really good effort and I guess my question really goes into how we're developing drugs in this PD1 experienced setting. So now we have lots of small biotexts. We've talked on the podcast about many interesting signal in phase one. Maybe you and I can banter back and forth about this. What do we think the bar is for primary resistance response rate or PFS? Yeah I'll go for response rate I think I've said this before I think PFS is by definition potentially more difficult to interpret certainly normal randomized trials because you've got the natural history of the disease to take into account which isn't controlled for a response is a perturbation to use that word maybe in the natural history of the disease by definition okay because you've got tumour that is shrinking hopefully because of your intervention but that's a slightly that's another issue just to think about. So for me primary resistance it's a checkpoint inhibition in melanoma I would be happy probably with a response rate of at least 15 or 20% I think with median-dure response situation I think of at least six months so that made my kind of bar I think for an interesting signal not saying it's right but that's what I'm thinking. So because we've done an episode on PD1 resistance I don't know if at this point we've published it but potentially by the point by the time we published this episode we will have published that where we talked about the Dan Olson the INS-PERS-DISILVA non-randomized data and then the S-16-16 randomized SWOT data at least for those it's pretty consistent and the S-16-16 was for primary resistant PD1 patients. So as per this CITC definition that would be at least two exposures to checkpoint two consecutive exposures and progressive disease or stable disease lasting six months. Okay so the single agent response rate there is about 10% and it's considered the null hypothesis for this so kind of uninteresting just to get 10%. So I think you're 15% bars right but we have seen the signal as high as 30% because it's something like 27% in S-16-16 and right about there in the Dan Olson, Lotus, Ippi, Pembro, etc. So I think 30% is where the bar is for primary resistance if you're developing a drug you know till life you lose out if you will was fourth line therapy median fourth line and was still getting 3035% response rate. So I think that's where we might need to be there. Now more complicated as secondary resistance because you you say something that is incredibly fair I don't know how many drug companies are listening that in the secondary resistance these are patients who've had a benefit. It's the duration of drug exposure was at least six months so they've had stability of their tumor burden or a true perturbation as you described it and now you're going to do something more and a lot of times it's with a PD1 backbone continuing and you add something to it and then what is it that you say about that response rate that we need to be careful of when we see it for that population? The fact of residual sensitivity to anti-PD1, anti-PDL1 for the sake of argument would have thought I think we will know in those kinds of patients that if you'd given the patient anti-PD1 or anti-PDL1 again let's just say one of therapy for the sake of discussion doesn't need to be there. There will be some activity that's in there and so then how much is that activity? Well difficult to know and it will depend a lot on the patient can at least 16 or how long have they been off treatment so in other words if you stop the treatment after a period of time some of those people are still going to have drug on board certainly three months and maybe to some extent at six months so that's important. If you've got if you've got people who let's say definitely over a year since their last treatment I think you can be confident there's probably not much drug on board but then that doesn't mean there isn't some biological effect of the drug on board so all I'm saying is that let's call these partially sensitive by my definition if you want to call it that I think a lot of those patients you're going to be seeing a 10% or greater level of benefit based on response rate if you would just give them monotherapy again with anti-PD1 or anti-PDL1. So then if for that population going on to their next line they are potentially PD1 still sensitive I guess I have two things to say we should really be careful about the gap from prior PD1 so is that gap less than three months where they're still potentially drug on board or is it greater than do we want these patients coming directly off of that therapy onto this whatever next line therapy is or do we want a gap? And the second thing I would say is do you then think no right answer here no wrong answer do you then surmise that the the ORR for a secondary resistant PD1 patient should be higher lower or the same? Higher definitely higher I think for the reasons that I've said already beginning by some of my comment about primary resistance as well and I'm saying like you know 15-20% something like that I think maybe I'm deliberately a bit conservative about that because I'm sort of thinking let's say you're doing a trial with a new mobility of treatment so it's not a checkpoints inhibitor back bone for me that level of signal is interesting and what I'm saying in that primary resistance setting is that if you've got a signal that's looking like that that is actually something I want to know more about I would not throw that away with a new therapy in that situation because we're talking about looking for signal set up. Agree agree that that's a signal something in the 15% response rate primary resistance is a signal post PD1 is it enough to get your regulatory approval and a phase three I don't know I don't know I don't know and you know I think I mean that comment from me is about looking for a signal at the end of the day not always but it's likely you may have to do a randomized trial and then thanks for something I mean that's another subject in this own right what the control on is and so on and so forth and and this is is where I just want to put a word of caution is that sometimes what we're seeing is we're seeing signals in this space that are in the 25 30 percent, but they haven't carefully curated or defined the population that entered. It's just a PD1 experienced population. So that could be PD1 experience in the adjuvant setting and you finished your adjuvant course and sometime later your melanoma comes back. Technically that's, well that's a secondary resistance. If you are, I mean, oh, I don't even know where we're placing that. Although this paper house and adjuvant exposure table. So that what a, you know, PD1 experience in the adjuvant setting is not the same thing as taking in a mean primary or secondary resistance. So then whatever signal that, you know, a company might be seeing in their phase one to then be very careful that going straight to a phase three. Do some phase two, but clean up the population. Go looking for that signal in this population or this population or this. And you know, I mean, I like you would be happy to think this through with pharma as far as specimen samples. And then looking for a signal, which one of these do we think is actually happening? Are you reviving an old effect? Are you bringing an effect or T cells where they weren't there before? So helping with a primary resistance issue, etc, etc. Anyway, yeah. So I was just going to say I think that goes back yet again to the importance of doing randomised phase two trials, correlated biology, I think goes with that saying. And then in terms of trial design, you know, do you then need to do a separate phase three or can you use the randomised phase two as a starting point for a phase three? I mean, that would be to me a smart way to do it, I think. You know, a relatively for seven was sort of done that way. It was a phase two slash three with an endpoint that they used for phase two as go no go once they hit it. They picked a clean endpoint for phase three that the phase two portion of patients could still be evaluated. Yeah, just to say something about control arms and randomised trials as well, though. So this whole business about doing randomised trials, you know, to some extent assumes that there is going to be a satisfactory control on. And they actually say this in the ZC paper, it's quite an important point. In other words, let's say you were to give single agent, anti PD one again for the sake of discussion. What is the expectation of efficacy in the control on? Because again, ethically, if you don't have any significant expectation of efficacy in the control on, then there's a big question mark. And then to broaden that out, you could end up with that side to talk to chemotherapy as a control on. Well, and we've had this conversation with companies that are in this space. At that point, maybe arguably you could just say investigators choice because there is a no real front runner. The no one therapy is going to do necessarily so much better than another. So just let the investigators pick whatever they would do for control minus perhaps local therapy. Because all of the medesthetic disease, all of the progressive disease, you know, localised therapy does have a role for that kind of stuff. So I think the caveat being whatever system of therapy you might do. Yeah, and I think to your point as well, then I think you then have the investigators with the patient in front of them and they're thinking about patient selection for the trial. Yeah. And then that means that they have to consider what the control on is and whether they in their own mind knowing the patient really well can have that conversation with the patient or not. And so then one would hope that would mean that localised therapies, kinds of patients actually weren't going in the first place because that's just not going to help with anything. Do you prefer a trial that says investigators choice of this menu of options or do you prefer a trial that says investigators choice is the control, whatever you can come up with systemic? I somewhat prefer a menu if I'm honest. I just think it's cleaner for one of the better word. I'm actually thinking about the immune control as well, which we know quite well. I have answered the immunocrowin. To enter fast. Okay. Okay. Investing at UNO in that one quite well. And you know, you could say, well, what if it's not available? So I think that's part of the feasibility process that you're going to talk to people or inevitably or mainly globally and you say, okay, what have you got access to? And then you actually have discussions with investigators about what's available and what they are comfortable with as a control on, whereas I think if it's completely open, then it's sort of kind of messy, I think. I think I don't know if it's a strong view, but I think I prefer that. I mean, it's messy clean as we just said if all the controls are roughly missing the mark to the same degree. I mean, you bring up you feel so I feel obligated to say the issue with the tabby study at some site, certainly in the US, as they didn't feel comfortable with the control menu of options, Pembroda, Carbazine and Temizola, maybe just really felt uninteresting and they didn't feel they could comfortably randomize. In fact, as you and I know, there were some study contact issues with German sites that patients were getting randomized and to the control and then getting taken off study, which has an OSN point, then you're never going to read out. And in fact, this happened with Hepsano with the delcath treatment. It was randomized to best supportive care. You mean Hepsano? You mean Hepsano? I'm calling it my official name, Hepsano. So the control arm there, basically it was best alternative care. I think I said best supportive, but it was best alternative care. And I mean, it was just so uninteresting that people were dropping out. So that might have been in the absence of a menu, you know, the Tubby study had a menu. So there's probably no perfect way to do it, but your point is pretty much the same. Listen, I think it's a lot of different trade-offs here. That's the reality of this. It's not perfect, clearly. And actually, it's not just going to be about the sponsor and the clinicians, there's regulatory considerations. There's also, so you know, to be honest, at the end of the day, the sponsor needs to come up with a study that's going to do the job, I think, taking to account all of these different perspectives, but you're not going to keep everybody happy. I think that's the reality. Can I put you on the spot to go through table three and table four, and then we can wrap? Yeah, it's a table three adjuvant setting. Primary resistance in the adjuvant setting less than or equal to 12 weeks after finishing therapy or recurrence on therapy. Undeterminable is the other category, which is greater than 12 weeks. So basically, using a three month cutoff there. So I know. But Tubby clear, James, that also means less than or 12 weeks. Okay, maybe I'm thinking over thinking it. I was thinking, like, what happens when you discontinue your therapy due to talks? But yes, then less than or 12 weeks from therapy discontinuation, if you have a recurrence, it doesn't matter that you stopped it for talks. They're saying, as long as you've had two doses under your belt, they're going to call that primary resistance. Yeah, I just bet they spent a lot of time hammering this stuff out. And do you know what I like? I'm all for simplicity anyway, and I think that this is simple, it's clear. Yeah, and for it to be usable and workable, I think that's the right balance. Figure one gets a little complicated, but I get the point of it. How about table four? Yeah, so this is called resistance scenarios after holding therapy in their metastatic setting. So here, they're making the distinction between induction therapy. So that might be a beneath, I say, and then we'll keep original cancer for that matter when you have the two drugs together. As induction and then versus continuous, an example of continuous would be Niva Rella, actually, where you're getting both drugs the whole time. And I think there's one or two, I think it's in maybe Mesa, Thelio, with it, but where your epi is continuously given one milligram per kilogram every six weeks, as opposed to what we used to in 2021. Yeah, so just for the record, I guess. Anyway, so take into account those two scenarios. One is the continuous part of the treatment, one is the induction part of the treatment. You see what the young gestures here. You did this and I started to others. Yeah, and you know, I was doing it as well in the visit last week with the Princess of Wales. That was like a picture of me just call myself. I don't know, maybe I'd need to cut down. Or maybe that's another new year's resolution. That's it. It's because I'm enthusiastic, I don't think it's an egg. I don't think it's a negative. Anyway, so you've got these two scenarios, that's the first thing. And then the second thing is the second column on the table is timing of tumour regrowth. That's what they're calling in that. And so then that can be either less and more than 12 weeks after the last dose of treatment, so similar sort of cut off. And then the final column of the table, the third column is called resistance phenotype. So that is either primary or secondary resistance to the combination in the continuous setting. And then in the induction setting, it can be either primary resistance to the drug that was continued. And then the final one, which is greater than 12 weeks after the last dose and the induction setting. It says must be assessed based on rechallenge. So I think the point for that very final one is that in that situation, actually, there probably is gonna be some potential sensitivity to rechallenge. And if you're thinking about Ibnivo and Malano Manau, we all know in practice, again, and we talked about this before, since the days of Ibn mono, can I say, that these patients who get Ibn mono, they get some benefit wears off. We're not doing this so much anymore. We know that there's gonna be some residual sensitivity there to be rechallenge. - So can I ask you in table four, the second line, continuous administration, let's say of combination, 'cause that's what this really means. More than 12 weeks after the last dose, can we really call that secondary resistance? Because wouldn't you rechallenge that group as well and then assess response or resistance based on rechallenge? What we're saying here is that it's more than 12 weeks after the last dose. - Yeah, I mean, they've had to come up with a number here. And I take your point, and I think I would certainly think about that in clinical practice. But again, I'd come back to simplicity and try to resist them here. - But that's my point. Why wouldn't they make it simple to match the induction greater than 12 weeks? Anything greater than 12 weeks after the last dose, you've walked away from the therapy and now we're putting it back in. And much to that earlier point, is there not some residual sensitivity? Because they're recurring right at the four half-life mark of the drug being washed out. Are we sure that secondary resistance? Anyway. - No, I agree with it. I agree with it. I think it is likely in that group of patients that will be some residual sensitivity there. And that's what I would be thinking about in the clinic, I think. - So for simplicity's sake, I could have collapsed us into a two-line table. Whether it's continuous or induction dosing, it's less than or 12 weeks or greater than 12 weeks after the last dose and it's primary resistance and then it's indeterminable based on rechallenge is what I would say. I guess, or obviously tumor biopsies is the need of the paper. - Yeah, so we would need to get the transcript, I think, of the meeting. 'Cause I suspect this was discussed. - Yeah. - Maybe just to finish, we'll just say something about the different groups here, the working groups, which have just listed out rather the end of this paper. So Harriet Cleager, chair at the check point inhibitor combinations working group. And then there's another group for check points inhibitors and chemo, chair but run, Sullivan. And then finally, check points inhibitors, it's called Schumertargetes and anti-angiogenic therapy. And then, who said, "Tawi was the chair of the last one and the names are in there." So there's quite a lot of FDA people and quite a fair amount of people for the industry as well as well as academics and clinicians. So anyway, so what I think about this is it's great. I think it's important. It's never gonna be perfect, but I think Kudos to everybody involved actually established in these definitions. We really needed them. Great framework for us to be able to lay our data on top of. So yeah. - I'll make podcasts about. - Yeah. All right, thank you. Thank you. - Thank you, take care. (upbeat music)