Episode 123: The Big IMDC Kidney Cancer episode with Professor Daniel Heng

Simply Oncology podcast - IMDC Dr Heng-20260714_171115-Meeting Recording

ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 0:41 Welcome to this special episode of the Simply Oncology Podcast, Mike. It's very special. Why is it special? Because, John, today we have Johnny Heng all the way from Canada, all the way from the IMDC, all the way from the bosom of kidney cancer. Danny Hing. Welcome to the show, Danny. Daniel Heng 1:03 Thank you very much. It's great to be here across the pond. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 1:08 And we cannot have Professor Daniel Heng, but we're going to call you Danny Heng, right, because I think that's your trademark. If you had, like, if you wore a t-shirt, it would say Danny Heng, TM. We have to talk about the IMDC, the Heng criteria. Danny, would it be okay if we call this episode Hanging with the Heng? Daniel Heng 1:15 Sure, yeah, yeah. Oh, I love that. Okay, yeah, I've had that. I've had hang 10. I've had all sorts of things. So, so hang, yeah, this is great. I love it. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 1:31 Can we do that? Oh damn it, somebody else has already thought of it. Okay. But we can have Danny on without talking about the IMDC and how we prognosticate kidney cancer, how that has become how we segregate our treatment for kidney cancer, certainly here in the UK, and how that's fed into the trials. biomarkers, and of course, how real world evidence helps shape our knowledge from clinical trials and how we'll hopefully shape how we move forward. Is that enough, Danny, do we think? Daniel Heng 2:06 Yeah, it'll be great. I love talking about this stuff. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 2:09 And do you want to just give yourself a little introduction? Daniel Heng 2:12 Sure, I'm Danny Hang. I'm the Division Head of Medical Oncology here at the Arthur Child Cancer Centre in Calgary, Alberta, Canada. I'm also the chair of the IMDC, the International MRCC Database Consortium. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 2:26 And we are delighted to have you on. Now, Danny, we were at a meeting in London a couple of weeks ago, and I heard like a really sad story that you told about how the IMDC started. It means I genuinely felt sorry for you. But from bad moments, great things happen. Well, I've not heard this story. What happened? So how did you get the idea for the IMDC? And just tell us a little bit about that. Daniel Heng 2:43 Maybe you did, maybe you. Yeah. Sounds good. So from tragedy comes innovation like this, this can be a movie. So basically, one boxing day, where you're supposed to go shopping and line up for the best prices for TVs and Nintendo entertainment systems and stuff like that, there's a huge snowstorm. And so I couldn't go shopping. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 2:59 Yeah. The. Daniel Heng 3:17 And so what do people do when they can't go shopping, when they're with their family on Boxing Day after Christmas? They collect data. And so that's how the IMDC started. And so I collected data in my dining room on a computer, laptops, and we didn't have laptops. Did we have laptops? I don't know. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 3:27 Yeah. Daniel Heng 3:37 And it started collecting the first 69 patients with metastatic kidney cancer. And we looked at sunitinib versus interferon and showed a doubling of overall survival. And that led to, and that helped with funding recommendations in our country. And so that led to the IMDC because we developed prognostic factors after that. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 3:47 Mhm. Daniel Heng 4:02 I did other institutions. One of my excellent friends, Tony Schweri, he's an amazing, amazing friend. He corralled people over and got Dana-Farber involved as well. And so with this first six institutions, we had 645 patients and found the IMDC prognostic criteria for today. And so the IMDC prognostic criteria are 6 criteria. KPS less than 80%, diagnosis of treatment interval less than one year, low hemoglobin, high platelet count, high calcium, high neutrophil count. If you have none of those factors, you have favourable risk disease. If you have one or two of those factors, we have intermediate risk disease, if you have three or more of those factors, you have poor risk disease. And we use this to prognosticate metastatic RCC patients all around the world, because it can help in prognostication, patient counselling, and it affects drug reimbursement. So especially for Ipi Nivo, we shouldn't be using Ipi Nivo in very favourable risk patients or payroll risk patients, for example, because it's not funded. And so that's been the IMDC criteria and it's been a great journey so far. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 5:20 And Danny, so many people will say to you, and even I've seen people walking up to you and challenging you on the IMDC, essentially it's a full blood count, a calcium, how fit you are and how long it is before you start treatment. I want to ask you a question because I want to hear it straight from the source. We as clinicians dictate one of these scores because if we say I want to start treatment or not survey, I can make somebody an intermediate risk patient by starting treatment within 12 months. Can you just clarify that point? Daniel Heng 5:58 Yeah, yeah. So you're right. The diagnosis treatment interval, it can be a bit subjective based on what you choose. And so, and the idea of it in the world of research, it makes sense, but clinically, sometimes it could be fudged a little bit. So it's from any diagnosis. So if you had kidney cancer five years ago and resected five years ago, ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 6:03 Mm. Daniel Heng 6:20 and then developed metastatic disease today and then started kidney cancer treatment, systemic therapy today, then your interval is five years. And so that interval is meant to give us an idea of how active and how aggressive the tumour is. So if your treatment is quite quick, if we relapse really quickly, that time period is very short, less than one year, ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 6:36 No. Daniel Heng 6:42 then that portends a worse prognosis. If it's greater, then it portends a better prognosis. The other thing that could be fungible, I suppose, is also the KPS. Like you can make someone 80%, you can make somebody 70%. But really, it's just a tool for us to help prognosticate as ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 6:54 Yeah. Daniel Heng 7:02 best we can. These prognostic criteria are not the are not 100% solid. They're not 100% accurate. They're the best thing we have so far. We try to improve on it every six months. So every six months we get a data dump from all the centres around the world and we try to improve on the IMDC criteria. We press this statistics program, the little running man. starts this programme and it spits out all of these statistics. And every time we try to add something, we only get ahead by maybe 1% in terms of C index accuracy and the discrimination ability. And so it actually doesn't add much to add more features into the IMDC criteria. And so that's why we haven't changed it. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 7:30 Yeah. So you essentially, at the get-go, got the perfect prognostic score. It cannot get any better. Is that what you're saying? Whoa, that's a stupid. Let's stop trying. It's perfect. No. Daniel Heng 7:55 Whoa. That's a high bar. That's a high bar. So it's not perfect. It's very imperfect. And in fact, it's imperfect for a couple of reasons. The intermediate risk group is really large. Over 50% of people are intermediate risk. And it's imperfect because sometimes it can predict wrongly ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 8:02 Hmm. Hmm. Yeah. Daniel Heng 8:16 And so I think how to improve on it, we've reached the ceiling for clinical factors. So adding more clinical factors like NLR, like, like, you know, other clinical things, race, gender, age, stuff like that, adding those things ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 8:23 Mmh. Mhm. Daniel Heng 8:35 don't seem to help, but I think one day we'll have some biomarkers that we'll be able to add to the IMDC score to make that accuracy better. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 8:44 No, we're going to hold our fire on biomarkers, okay, because no kidney cancer talk is complete without the biomarker chit chat. But why do you think full blood count has become the cornerstone of the IMDC? We like to say in meetings and Daniel Heng 8:52 You got it. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 9:03 and talks, it's a downstream inflammatory effect because of raised platelet count, raised white cell count and anemia. These are all inflammatory markers. And is that is that your description of or how do you sell it one to yourself and two to, you know, a clinical fellow coming over to work with you? Daniel Heng 9:23 Absolutely. The complete blood count is so simple, hey? It does have markers for inflammation, so platelet count rising, that's an acute phase reactant. Same with neutrophils rising, that's a marker of the immune system. Anaemia is also can be from chronic inflammation or from blood loss if you're peeing out, you're peeing out blood, for example. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 9:27 Mm. Daniel Heng 9:47 Um... How do I, I don't know if we need to sell it or anything. I think these markers are simple enough to be able to explain the variability in the database in prognosis better than CRP, better than BNP, better than neutrophil lymphocyte ratio. And so. The best part about it is that everybody has a CBC, so you don't get any missing data. There's no such thing as missing data from a CBC because everyone has a complete blood count. Whereas if you look for CRP, not everyone has a CRP. LDH, not everyone has an LDH. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 10:16 Yeah. Mhm. Daniel Heng 10:28 And these more expensive tests, like biomarkers, genetic tests, genomic tests, those are definitely not readily available unless you pay a lot of money for them. And so those are things that we talk about parsimony, how parsimonious is the model, how, you know, how do you get a model that's simple enough to remember that's easy enough to get, that explains as much variability as possible in all of the patients in your subset to be able to make do and prognosticate well enough compared to other things. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 11:04 I think it's, you know, having, you know, slightly jokingly mentioned that you got the perfect test right up front. The point is that no prognostic marker is perfect. And actually, the fact that we even have something that we can use is better than most tumour sites. We don't have it in bladder cancer. We don't have it in, or we can pretend we have it in prostate cancer with high and low volume, but not. not the level of granularity of favourable risk 42 months, you know, intermediate risk X months, you know, we just don't have that. So it's all we've already made a huge thing advanced from that point of view, but the counter to it being so good and so easy to use, and because it's become so prevalent and embedded in clinical trials and therefore Reimbursement regulations. One of my sort of points about it is people have become very rigid with it and then particularly regulators. You know, definitely in the UK, we can't use certain drugs in favourable, we can use it intimately, we can use them poor, and therefore you're more driven towards that fudging and then moving away from the evidence because you want to try and access the drug. So I suppose that will then come into the conversation about emerging biomarkers, about how we can try and make that a bit more useful and less and more evidence based rather than us just trying to get the right thing for the right patient. And I'm sure that's something you've heard before, Danny, like people saying to you, this is a prognostic tool. Daniel Heng 12:11 Mhm. Yeah, you got it. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 12:31 designed in the TKI era, but you have recently confirmed in the IO era is holding true. So, what was a favourable,an intermediate and a poor risk looking at now. Daniel Heng 43:05 Yeah, so it's interesting, like often the medians haven't even been reached yet, but they've quadrupled from before. So for example, in the interferon era, an overall survival of an RCC patient, medicine RCC patient would be 12 months. And then in the TKI era, it'd be 24 months. Now with our combination therapies, we're at 48, 50, 55, pushing that. So So we've really more than quadrupled our overall survival. It was never designed as a predictive stratification tool for combination therapies versus single agent TKI, but it has become that. Daniel Heng 12:48 Yeah, no, you're absolutely right. This is a prognostic marker and we accidentally stepped into the predictive field because it was included as a stratification factor in Checkmate 214 Ipi Nivo versus Sunitinib. And so because of that, a lot of regulators have used it. So many regulators around the world don't allow Ipi Nivo in the favourable risk world. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 13:01 Mhm. Daniel Heng 13:12 Are there some favourable risk patients that could benefit from Ipi Nivo? Absolutely. But I guess the problem is we don't know who they are in the favourable risk group. And the problem is there are actually more deaths in the favourable risk group for people getting Ipi Nivo for the first four years before the curves actually start to cross. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 13:20 Uh-huh. Daniel Heng 13:32 And so you have to be comfortable with that. Okay, if I give you ipi nivo and your favourable risk, there's actually a higher chance of you dying for the first four years, which is a very long time. And then the curves start to cross. And I think there's more emerging data about long-term durable responses in other types of combination therapies. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 13:54 Yeah. Daniel Heng 13:54 And so with IO-VEGF therapies, you know, they exist as well, these long-term durable responses, and the complete responses, they exist as well. And the relapse rate from the complete responses is similar between IO-IO and IO-VEGF. It's about 10% or so. And so I think we have many choices. I think If it be nivo were the only available option for our patients, absolutely I would feel bad about excluding favourable risk patients. But now we've got so many different options we can use in our patients. I think we're a lot more tailored than we could have been maybe five years ago because of the options that we have. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 14:24 Mhm. And you mentioned at the start, very favourable risk. Has this got your seal of approval? Is there going to be a fourth section or is there going to be four groups with very favourable risk? Daniel Heng 14:51 That's a really good question. So, John, you're welcome. You're welcome. You know, like. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 14:52 Oh, thank you. Danny, you've ruined the episode. Danny Heng said that was a good question. Don't give him any kind of validation, Danny. He's just... Don't listen to him. Danny's over there in Canada being positive and you're being negative right here in the UK. Daniel Heng 15:04 It is a very good question, John. Ohh. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 15:14 He only does positive. Daniel Heng 15:15 Um... Too funny. So the very favourable risk group are people from the favourable risk group that don't have bone, liver, or brain metastases, that have a KPS that's really good, 90 or 100%. And their diagnosis or treatment interval is really long, like 3 years. And so this is the cream of the cream of the cream of the crop. It's not a very common patient. But Martin Zarba, who was an IMDC fellow, did this analysis and looked at the different types of combination therapies and the different types of treatments we can use in very favourable risk patients. And the very favourable risk patients did not really respond very well to IO-IO ipi nivo. They responded really well to VEGF. based therapies, be it a VEGF inhibitor alone or a VEGF IO therapy. And this was true in terms of response rate, time to next treatment and overall survival. In fact, there was an inferior overall survival when we used Ipi nivo. And I think it just suggests that the very favourable risk group is just very angiogenic driven. And so if you don't have an anti-angiogenic agent in your cocktail, then you're probably not treating the patient if they need treatment at all. And so I think that's a story that we need to continue to validate. I don't think it's ready for regulatory approvals or anything like that. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 16:38 Yeah. Daniel Heng 16:41 But I think it provides that story that the more favourable you are, the more towards VEGF inhibitors your treatment is. And I think the more poor risk you are, the less towards VEGF inhibitors and more toward immunotherapy or doublet immunotherapy you are. And so there's that spectrum of people. And, you know, we'll get better at this spectrum because the spectrum is not perfect. But when we do things like Kim 1, when we do things like the emotion 151 clusters, for example, to identify angiogenic versus immune, immunophile like tumors, I think we'll get better at sorting out who should get what. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 17:22 You've opened Pandora's box. He said it. He couldn't help himself. Do you know what? I think Danny was getting... No, I think Danny was getting bored of defending the IMCC. You don't need to defend it. We are fully, you know, in all... I'm sold. Yeah, yeah, no, we're all in. But if at any stage you want to make a big like... Daniel Heng 17:26 Ha ha ha! No, I just love it. I love it. Haha. Okay. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 17:42 statement about the IMDC where you're going to change everything and you want to do an exclusive, feel free. Daniel Heng 17:48 Will do. I'll call you guys immediately. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 17:50 Yes. But you've said it, biomarkers. You've said Kim 1, you've talked about Reeney's clusters. We're talking about Brighton Reeney's new trial, the optic trial. We now have to talk, and we're about to open Care 1 here with PDL 1. We are entering the era of biomarkers, we are moving away from the simple full blood count. Which of these things... Into the into the murky land of biomarkers. I mean, you know, PDL one, don't even get me started. What does that mean? Well, what excites you, Danny? Say we've... Daniel Heng 18:24 Mm. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 18:29 picked up a gambling habit. Please don't gamble anybody who's listening. No, gamble responsibly. No, just don't. Daniel Heng 18:36 TSA, that's a great public service announcement. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 18:40 Just don't, I think. But if we were and we're backing horses, which horses would we be putting money on? Daniel Heng 18:41 But. Oh, I'm banking, I'm banking on the horses that involve composite biomarkers. So things like the emotion 151 score with Brian Reedy, Bob Mozer, just because I don't think one thing will do it. Like you guys said earlier, I don't think it'll just be one thing. I think it'll be a combination of Kim 1 and some sort of genetic transcriptomic map. You know, I think that PD-L1, we've tried it many, many, many, many times. All the clinical trials, we've done PD-L1 testing, and we found that it's not very helpful, but it does enrich for a population of responders. So if you're PD-L1 positive, I think the response rate is higher. It doesn't mean that it doesn't work at all immunotherapy. It just means that it might work a little bit better in PD-1 positive patients. However, there's lots of problems with testing for PD-L1. So, you know, the difference between us and melanoma, for example, is that our tissue is often very old. So ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 19:46 Yeah. Daniel Heng 19:50 So if you get a specimen or an effect me, it might be five years old, it might be 10 years old even. And then you're testing PD-L1 on that. And so is that okay to do? Are we really capturing what's happening in the tumour microenvironment? And I think there's a lot of heterogeneity as well between the primary and the metastasis. So if you are staining PD-L1 in the primary, that's great, but what's happening in the metastasis? And in Brian Reaney's optic trial, they found that the transcriptomic map of the metastasis was much more representative of what was going on as opposed to the transcriptomic map of the kidney, the primary itself. And so they actually changed the protocol. So you actually had to have a biopsy of your metastasis instead of just using the transcriptomic map of the kidney. And so, you know, I think that's what I'd bet my money on. PD-L1 might be in the mix there somewhere, but as a test on its own, it's not good enough because we've tested it so many times, like you've said. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 20:55 And CT DNA, you know, we got really excited about CT DNA in urothelial cancers, but kidney cancer is not really jumping on that bandwagon. Daniel Heng 21:05 Yeah, it's not great. Hey, John. So Tony Shwary presented the keynote 564 analysis at ASCO looking at CT DNA use in the adjuvant setting. So if you took out your kidney, resected it, and then it did CT DNA, what happens? ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 21:20 Yeah. Thank you. Daniel Heng 21:24 It turns out that very few people are ctDNA positive, like 8%, 9%. And so the sensitivity is quite poor. The people that were ctDNA positive did worse. Okay, so it is prognostic that way, but it doesn't capture enough people. 8% is not enough. And so Maybe it's the assay we're using, maybe different assays are better. I know Signatarid has newer assays that might be deeper in terms of what they can see with ctDNA. And so I think that field still remains to be developed a little bit more with more sensitive assays. I think ctDNA. is in there somewhere, but we're just not sensitive enough at this point. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 22:08 And Kim One, like we have to talk about Kim One exciting data. And I think as I spoke to somebody, a urologist who's a friend of mine, about Kim One and how it was transforming things. And he was saying, oh, I was looking into that like 15 years ago in benign. Daniel Heng 22:08 Yeah. I love Kim One, love it, yes. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 22:27 Yeah, and so it's an old test that's got a new shiny suit and it's been brought to the ball, isn't it? Are you using it or do you have access to it? Daniel Heng 22:36 Yeah. So we don't have access to it. We're just doing experimentally. So all of our adjuvant patients will get Chim 1 measurement testing, but on an experimental research basis only. The reason why it's experimental is because we don't know what a level of 20 means or 300, for example. Is that high? Is that low? What's the median? So we still have to figure that out. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 22:43 Yeah. Mm. Daniel Heng 23:02 And my Kim 1 test is probably a little bit different from your Kim 1 test. There are two specific types of tests. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 23:07 Well, it's definitely different to Arc in one test because we don't have one. Daniel Heng 23:11 But there are two different assays, like there's an ELLIS assay, and so there are two different assays, and so, so I think we require more standardisation and we and we just need to look at this prospectively a little bit more. The data is pretty amazing for metastatic disease in IpiNivo, for example. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 23:13 Yeah, yeah, yeah. Mm. Daniel Heng 23:29 Vincent Zhu from Dana-Farber presented that data where he showed that with Ipi Nivo, you know, as short as week three, if your Kim 1 level falls, it means you'll do better with the Ipi Nivo. If it rises, it means you're not doing well. If you started off with a high Kim 1, it means you don't do as well. If you started off with a low Kim 1, it means you do better. And so that's very... very interesting, but interestingly, it didn't work with Sutent. So with Sunitinib, it didn't show those. And so will it work with the IO-VEGF therapies? And how important is the VEGF component to that? That remains to be seen. And in the adjuvant space, Laurence Albigier presented that as well. showing that if you did have a high baseline Kim 1 level, then you may stand to benefit from adjuvant therapy a little bit more. So I think this is a quickly evolving story and all of us are jumping on the bandwagon to do sort of research protocols for Kim 1. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 24:19 Ohh. Daniel Heng 24:32 And so I look forward to those prospective evaluations of Kim 1 testing, and they're coming. And I think it'll be very helpful to have something that tells us about MRD, minimal residual disease, and just the bulk of disease in metastatic disease. ROWE, Michael (ROYAL CORNWALL HOSPITALS NHS TRUST) 24:45 Yeah. All of that is absolutely incredible, like the amount of change that is happening scientifically driving all of that, those changes to biomarkers.