Introduction

DR LOVE: Good afternoon, everyone. I’m Neil Love from Research To Practice and welcome to Meet The Professor, as today we talk about the management of hepatobiliary cancers with Dr Katie Kelley from the Helen Diller Family Comprehensive Cancer Center at the University of California, San Francisco. This is the second webinar. The first one we did a few weeks ago with Dr Abou-Alfa. It will be interesting to contrast his views with Dr Kelley.

As always, if you have any questions or cases you’d like to run by us, just type them into the chat room and we’ll talk about as many of these as we have time. We know a lot of people end up listening to these webinars. And if you’re into audio programs, check out our podcast series, Oncology Today, including a recent program on biliary tract cancers with Dr Valle.

Next Monday and Tuesday, we’re going to try a little experiment. We’re actually going to host a world premiere of 2 enduring programs that we’re about to publish. The first on Monday with Drs Bradley and Spigel looking at the management of lung cancer with locally advanced disease that’s not surgically resectable. Of course, Dr Spigel did the classic PACIFIC study. We’ll talk about that. And then on Tuesday, we have Dr Modi who did the plenary presentation at ASCO on T-DXd and HER2-low breast cancer. That’s a really exciting presentation and discussion. And then on Wednesday, we’ll continue our Meet The Professor series in ovarian cancer with Professor Ledermann. Then on Saturday, I’m actually heading out to Las Vegas to the Bellagio. We’re doing a day-long meeting. It’s going to start at 9:00 in the morning, but actually 12:00 Eastern Time. We have some great faculty that’ll all be appearing virtually with the American Oncology Network who will be there with me live. We’ll start out actually at 12:00 noon on Eastern Time with breast cancer with Drs Burstein and O’Shaughnessy. And we’ll just follow it up with other modules on main topics with a great faculty. So come on out on Saturday, a week from Saturday. Then on Thursday, August the 11^th, Dr Sands will be talking about small cell cancer in Meet The Professor series. Dr Strickler on the 17^th to continue our upper GI series.

But today, we’re here to talk about hepatobiliary cancers and actually, some pretty recent developments that we want query Dr Kelley about. But for now, we’re going to also present a bunch of cases from docs in practice who are going to be presenting cases, both of hepatobiliary — HCC as well as biliary tract cancers.

But Katie, I just want to start out with a couple of papers that you wrote and some data that was just presented at the ESMO GI meeting. First, this paper. I wasn’t even aware of this “journal” that does the review articles coming out in Nature, but you were a part of this, it was like an amazing paper, 28 page, everything you ever want to hear about HCC, phenomenal graphics. Anything you want to say about the paper just to kind of let people know about it? There’s so much in there, Katie. What a — I guess this is the new textbook of HCC. Any comments about sort of putting it together and anything in there that you may want to comment on?

DR KELLEY: Thank you. I was honored to be invited to be a part of this by Josep Llovet who led the effort. Nature Review Disease Primers really provide a cross stages, cross molecular profiling, cross disciplined approach to tumor types. And, as you can see from the author list, HCC is highly multidisciplinary with hepatology, medical oncology, interventional radiology, surgeons, molecular medicine folks all represented in the authorship. I guess my — I’ve been using this paper to cite epidemiology updates because it really does explain, delve into why we’re seeing more cases of HCC, more deaths from HCC, particularly in the western world due to NAFLD. So our epi colleagues have gone into that with this paper. And then it also is, I think, a handy review of the seen change in systemic therapy that we’ve seen over the past 5 years. And this really delves into the data underlying the new tyrosine kinase inhibitors and the immunotherapy that we’re starting to use in the clinic.

DR LOVE: Yeah. So it’s July, I guess time for the new fellows who are just on board. So I guess if you’re a new fellow and you want to hear about HCC, if you’ve got a couple of days to read through this paper, I recommend.

Just one other thing that is discussed in the paper, not by you, but one of the other authors, Katie, is the issue of screening. And I hear stories from docs about seeing patients with cirrhosis with cytopenias, get consulted for that. And then them saying, hey, are you screening this patient for HCC? Anything you want to say about that strategy and whether you think moving forward there’s going to be any yield in terms of earlier diagnosis and better long-term outcome?

DR KELLEY: Well I think it is important to remember that anyone with more advanced stages of fibrosis or frank cirrhosis, whatever the cause, whether it’s NAFLD or hepatitis virus or alcohol, should be screened. And our hepatology as well as oncology guidelines, including NCCN, recommend this. So at least twice annually with either ultrasound or CT or MRI and AFP periodically at those same intervals. And then, for patients with higher risk features such as hepatitis B, even at a younger age even without frank HCC — excuse me, even without frank cirrhosis may warrant screening depending on their demographics.

DR LOVE: So here’s another great paper that you did, another review article looking at sort of the biology of HCC. And then, of course, this paper you did in the New England Journal with Tim Greten, which was really a commentary on a paper from Nature looking at NASH and the issue of immunotherapy. Really, I’ve been asking people for a long time now, is there any difference in how people respond to therapy with HCC based on the etiology? And head and neck cancer, I’ve been asking the same thing. You have the same issue, viral, nonviral, and looks like immunotherapy works pretty similarly. But this really gets into another thought that really maybe NASH related HCC is biologically very different, and the question of whether it’s going to respond as well to immunotherapy. Can you kind of — there’s a lot of details and great diagrams. This is, maybe you can use this one to really kind of talk through it. But can you kind of review the thinking now that maybe there is going to be, maybe we’re moving towards different subsets based on etiology?

DR KELLEY: Yeah, definitely. The Pfister Nature paper of which this diagram is a recapitulation of the data, right, the bottom panel there from Pfister et al., really pointed out that the immune microenvironment in NASH itself or nonalcoholic fatty liver disease is different than in viral hepatitis, in particular with different profiles of T-cells, and some which have sort of both an exhausted and an activated phenotype that in animal models seems to have a poorer response or lack of response to immune checkpoint inhibition by comparison to non-NAFLD context. And they followed this paper up with a meta — in the same paper, they presented a metanalysis of several of the contemporary Phase III immunotherapy trials that we’ve had in HCC. And they looked at overall survival as an endpoint by nonviral compared to viral etiologies. And they showed significant benefit for viral, but really no significant difference compared to the control arm for the nonviral. And so, they posited that the immune microenvironment in the NAFLD setting may be different and interfering with response. And so incredibly interesting hypothesis.

I think that HCC is so complicated though that when one really delves into the details, it’s probably more than just the immune microenvironment. It is certainly possible that the viral, and likely, that the immune microenvironment of virus associated HCC is different. However, if we look at objective response rates and even PFSs of immune checkpoint inhibitors for viral versus nonviral, they’re not that different. Nonviral and NASH patients still can have good responses and the objective response rates aren’t that different. Where we see the difference is in the overall survival. And that actually ties in a lot of confounding variables like comorbidities, like diabetes and hypertension and things that can be complicated in the NAFLD population more so than, say, a viral population. It also draws into question the impact of region and access to downstream therapy which can contaminate overall survival. So I think it’s not a simple binary yes/no for immune checkpoint inhibitors according to NAFLD. We still use it predominantly in our NAFLD patients as a first line if there’s no contraindications. But I think it’s a really important way to think about HCC, that they’re not homogenous, and viral etiology versus nonviral are different in a lot of ways, not just immune microenvironment.

DR LOVE: So our crack RTP team saw that there was this paper that was going to be presented at ESMO that kind of addresses what I want to ask you about in a second. But first, just to take a look at the updated data from the IMbrave150 study of atezo/bev that was published in a journal. I don’t know that a lot of oncologists read the Journal of Hepatology. But the bottom line is, it looks like the benefit in terms of survival and PFS is continuing to extend now with more follow-up now to 15.5 months. You still see the benefit. But what’s interesting was, I haven’t seen this forest plot before. So if you look at the top part, etiology, hepatitis B, hepatitis C over to the left and they are nonviral, it’s sitting right on the line. FYI, we put the PD-1 stuff in there too if you want to comment on this. But any thoughts about this? Does this dissuade you, make you maybe less interesting in atezo/bev in a patient with nonviral HCC?

DR KELLEY: So I’ll take the bottom part of the question first because it’s easier. So I think we can conclude from the PD-L1 status that while, like a lot of other tumor types, there might be somewhat of a trend towards better outcomes in the higher PD-L1 expressors, the overall and the unknown are very similar to one another and it is not a strong enough biomarker to guide our treatment. And there’s a lot of heterogeneity between tumors. And so I think we can definitively say right now we should not be using PD-L1 to decide on, to predict response or to choose therapy.

Now the harder one is the viral etiology. And as I mentioned in the earlier slide with the Pfister paper from Nature, the OS data is very complicated to dissect out. And it is not simply a response to immunotherapy metric. It is also wrapping up into that metric comorbidity, access to downstream therapy according to region of the world which tracks with viral status and potentially some component of immune microenvironment. But nonviral is not the same as NAFLD, so that nonviral also includes alcohol and other causes of liver disease besides just fatty liver. So for all those reasons, in my practice, and as I mentioned before, objective responses and PFS do not show this identical trend across studies, so it is more complicated than just response to immunotherapy. So for this reason, it doesn’t change my choice of the atezo/bev regimen based on viral status at this time.

DR LOVE: And we’ll get into the other immune option, or what we think hopefully will be an option, durva/tremelimumab and what your thoughts are on that. But just to finish out on this. And this is the paper that was presented at the ESMO meeting, I guess a few days ago. This is a retrospective analysis, I think of 200-some cases, trying to match up all the caveats of retrospective analyses, atezo/bev versus lenvatinib. I was pretty surprised they put hazard rates and p-values and all that stuff, which I don’t think you usually do in a retrospective thing. But, anyhow. Here it is. I guess hypothesis generating. They broke it out into, as you were just saying, NASH and other nonviral causes, and compared outcomes. And they didn’t report, or we didn’t see, response rate. But in terms of survival and PFS, it looked like the lenvatinib patients did better with NASH. And the non-NASH patients also did better. Any thoughts about this, I guess, hypothesis? And, of course, they concluded, this is not level 1 evidence. If you want to test it, you’ve got to do a Phase III trial. I don’t know if Phase III trials are going to look at this. I would imagine they’re going to start at least including in a maybe stratification factor. But, anyhow. You want to sort of sum up what this all means?

DR KELLEY: I think that the key take home messages that we need to be much more careful in are stratification and definitions of viral and nonviral. And nonviral is not good enough anymore. We need to be much more specific about clinical risk factors for NAFLD versus other etiologies in our clinical trials.

As for clinical practice, the data are really interesting in the context of what we already know about the metanalysis data. But there’s also in a retrospective study, there’s patient selection and there could be a lot of differences according to stage of disease where a patient might have been selected for lenvatinib because they’re earlier stage or right after a TACE procedure or something, particularly depending on region of the world. So I think it’s hard to draw any conclusions, but it is an important area for us to be more rigorous in our definitions.

DR LOVE: So one more thing I wanted to get your feedback on before you jump into the cases is this Trial in Progress. I always love the Trial in Progress papers at ASCO and other meetings. I just, when I saw this one, was like, woah, that is interesting. So this is in a bunch of GI cancers. So this is, I think, pancreas, upper GI, biliary tract, colorectal, wild to start with. And then this combination, belzutifan, which, of course, we talk a lot about in our RCC programs, and then pembro and lenvatinib. Any thoughts about the thinking behind this?

DR KELLEY: Right. So the tumor types in this GI basket are tumor types that are known to be sensitive to anti-angiogenic therapies and have a role for hypoxia and tumor signaling. And the belzutifan is a hypoxia inducible factor-2 alpha inhibitor, blocks a particular pathway for proliferation through HIF-2 alpha signaling that may be complimentary to the other anti-angiogenic therapy which is lenvatinib in this combo. So we’re certainly really excited about this multi-basket — multi-tumor basket trial that also includes HCC, I might add. So stay tuned in the next year or 2 for some early results.

DR LOVE: Yeah, that sounds super interesting. And I guess safety-wise, I don’t think belzutifan has any major tolerability problems. I would assume it’d be fairly easy to do this, at least compared to pembro/lenvatinib.

DR KELLEY: There’s some potential for anemia and hemolysis, but I think the risk is low. But, obviously, the study will quantify it.

Case: A man in his mid 80s with newly diagnosed hepatocellular carcinoma (HCC) who has a partial response with atezolizumab/bevacizumab — Minesh Dinubhai Patel, MD

DR LOVE: All right. Well let’s get down to the nitty gritty of where we are today and making clinical decisions today. And we saw a shift in systemic therapy a couple years ago and it’s really reflected in these cases as people really try out atezo/bev in their practice and get various responses. A lot of our HCC cases are like that. So we’re going to start out with an 84-year-old man with newly diagnosed HCC, a patient of Dr Patel who got atezo/bev. Here’s the case and here are his questions.

DR PATEL: 84-year-old with a history of gastric ulcers, osteoarthritis, and hypertension. He met with the interventional radiologist and discussed TACE and Y90. But with his age, he decided against anything too invasive. So he started first line therapy with atezolizumab and bevacizumab, and has tolerated it well. Restaging studies after his 4th cycle actually had showed a partial response, decrease in his primary tumor along with some suspicious lymph nodes as well too. Then he’s been on it for almost a year now and still maintaining a partial response from that.

So has atezolizumab and bevacizumab replaced sorafenib as the standard of care? Where does lenvatinib and then also tremelimumab and durvalumab fit into treatment algorithms? Is there any preferences for the immunotherapies given a multitude of options with atezolizumab, pembrolizumab, ipilimumab, nivolumab, and durvalumab?

DR LOVE: That’s really interesting. You mentioned tremelimumab and durvalumab. Any questions that you have about that specifically?

DR PATEL: Is there any benefit with that over ipilimumab and nivolumab? I saw a pretty fair amount of enterotoxicity from ipilimumab and nivo. And so is that something similar with the tremelimumab and durvalumab?

DR LOVE: So a lot of questions there. And we’ll take a look at the HIMALAYA data in a second. But any thoughts about this case and some of the questions he brings up?

DR KELLEY: Yeah. So starting from the top. I think the first question is, is sorafenib still our standard of care? And I think that in the fit Child-Pugh A patient without contraindications to bevacizumab or to atezolizumab, atezolizumab plus bevacizumab combination has now effectively replaced sorafenib or lenvatinib as our first line standard of care for patients with advanced or unresectable disease who need systemic therapy. So that’s the first answer.

In terms of the comparison to durvalumab plus tremelimumab in the HIMALAYA trial, I think we’ll look at those slides in a moment.

I think back to the question about sorafenib and lenvatinib. I really reserve those as a first line therapy for patients who aren’t able to get atezo/bev for some reason, either a contraindication to immunotherapy such as a patient with active Crohn's colitis, hypothetically, or in the pandemic, occasionally in patients who aren’t able to travel due to surges in COVID in the past also have used lenvatinib over atezo/bev for pragmatic travel related reasons, or in patients maybe with a really low burden of indolent metastatic disease who say, I’m not ready for infusional therapy. I don’t want to go to an infusion center every 3 weeks. I want to take a pill. I’ve used a TKI, lenvatinib or sorafenib, in those patients. But all of those cases aside, in the average fit systemic therapy patient, right now, atezo/bev is our new first line standard of care.

DR LOVE: So we’ll look, as you said, we’ll look at the HIMALAYA data, but I’m just kind of curious broadly right now at this point, putting aside reimbursement issues, if you could access tremelimumab, you could access this combination, are there situations, of course, you hear a lot about if there’s, obviously, a contraindication to bev, for example. Would you use it in that situation? And what about somebody who doesn’t have a contraindication to atezo/bev? Are there situations where you prefer to use durva/treme?

DR KELLEY: So this actually draws upon our prior conversation about NAFLD to some degree because I think all of us in clinical practice would agree that there are absolute contraindications to bevacizumab or where we feel very uncomfortable using bev, and then there is a spectrum of conditions where we feel increasing degrees of discomfort. Patients who are elderly, very diabetic, brittle blood pressure control that are really low one day and really high the next day. We see them because of when they took their medications. People that we are very anxious using bev in. And I think that’s where having another active second line survival prolonging regimen that doesn’t involve bevacizumab will come into play, particularly as we flush out this viral etiology impact. So in NAFLD patients with extensive vascular disease risk factors, I am going to be increasingly considering the durva/treme regimen because of concern for bev toxicity and potentially lower efficacy for a variety of reasons in the NAFLD setting.

DR LOVE: So we’ll explore that a little bit more in a second after we take a look at the data. But very common question that I hear from docs. They think they see anti-CTLA-4, they think ipi or ipi/nivo. A lot of experiences with all kinds of stuff, hepatic, colitis, et cetera. But the regimen with tremelimumab and durva is a different regimen. You give the treme upfront one time. How would you indirectly compare some of the immune problems you see with treme/durva with that regimen to, say, ipi/nivo?

DR KELLEY: Great question. So the durva/treme regimen in HIMALAYA is so-called STRIDE.

DR LOVE: Right.

DR KELLEY: Which stands for a single priming dose of tremelimumab where tremelimumab is given just once on day 1 at 300 mg and not repeated as a part of the standard regimen. And I think while we can compare specific immune-related adverse events, I think one helpful metric for me is the risk of needing high-dose steroids. And so for the STRIDE regimen, it’s a little less than 25, between 20 and 25% in the Phase II and in the Phase III, both. Whereas the nivolumab 1 mg per kg, ipi 3 mg per kg regimen in both HCC as well as renal cell and other tumor types that have looked at that nivo 1 ipi 3 combo, it’s approaching 50%, or is actually 51% in the Phase II in HCC of patients who required high-dose steroids for immune-related toxicity. And so that’s about a twice as high rate of needing high-dose steroids. And that’s sort of a barometer for me in the clinic of how likely is this person to have a serious problem?

DR LOVE: So a couple of questions in the chat room, one from Dr Kumar. If you think IO doesn’t work as well with NASH HCC, why would you use treme/durva?

DR KELLEY: So I think 1, back to the point, I’m not sure that it isn’t — is less effective from a response rate perspective, recalling again, the response rates aren’t that different if we look across a lot of different studies. It’s the overall survival. And that includes post treatment, comorbidity, toxicity of bev in particular, access to therapy downstream, 2^nd, 3^rd, 4^th line therapy which are more likely in a western population than in an eastern viral hepatitis population. So it’s much more complicated than just activity. But if one looks at the forest plots of the durva/treme study, and let’s just say hypothetically, that there is some immune mediated difference in the microenvironment that’s contributing to these differences which, again, is not — it doesn’t answer all the questions of why the response rates are pretty similar, even if there is — the immune population of T-cells is slightly different. If we look at the durva/treme with a CTLA-4 inhibitor involved, we don’t see that change in the waterfall plot. In fact, the viral/nonviral difference is really erased in that waterfall. And secondly, there’s a biologic hypothesis that the CTLA-4 may actually help recruit different T-cell population and also prime the peripheral — the lymph nodal stations around the tumor. So there may be a different biologic response to immune therapy when you include a CTLA-4 inhibitor. So I think that would be the rationale. But the other rationale is all tied to this question of toxicity. Is the decrement in overall survival related to comorbidity in the NAFLD population that might be exacerbated by a strong anti-angiogenic? If so, that supports a CTLA — a different approach with immune checkpoint combo.

DR LOVE: So I was commenting to Dr Abou-Alfa in our first webinar that it wasn’t that long ago when we would did CME programs on HCC when we were trying to think of what can we talk about other than sorafenib and different ways to use it? Now, it’s like, woah. And here’s another one from the chat room. Where does lenvatinib/pembro fit in?

DR KELLEY: Great question. So the LEAP-002 trial is a Phase III registrational trial looking at lenva/pembro compared to lenva in the Phase III context building upon the Phase Ib study which showed really provocative high response rates for that combination, but also some toxicity and, again, it was a Phase Ib study of only about 100 patients. So that LEAP-002 trial will really determine, both its efficacy and its safety metrics, will really determine its place in the standing of first line HCC therapy. I think just from a pragmatic clinician perspective, there are going to be patients that are not anti-angiogenic candidates or on the spectrum towards the end where we’d rather not use an anti-angiogenic like the brittle blood pressure patients, for example, where we might favor durva/treme. I think that would be a hard sell to use pembro/lenva because lenva does have a lot of hypertension. But, again, I think it’ll depend on how the data pan out in the Phase III.

DR LOVE: So I want to get your thoughts on a second comment that Dr Patel made to me. And I can reflect back. So many oncologists have mentioned to me, come up to me after meetings, frustrated that by the time they see the patient, the patient’s already had multiple ablative procedures to the liver, they have poor liver function, they can’t receive systemic therapy and they’re very frustrated. And then when atezo/bev came out, we really started to see some response rates. So first thing I thought about is, I wonder if that’s going to change? I wonder if people are going to start thinking about systemic therapy rather than intrahepatic therapy? He brings up this question, also related another, an interesting anecdote to me. So here’s the comment.

DR PATEL: A lot of times in the community, we’ll oftentimes inherit patients after they’ve had several ablations and they already had a cirrhotic liver and it’s at the point where it’s a Child-Pugh B or C at that point.

DR LOVE: The interventional radiologist that you were working with this on that case, was that person kind of tuned in to what’s going on with atezo/bev and that you see responses?

DR PATEL: So one of the first patients that I actually had started on them had a remarkable response. So he actually called me afterwards and asked me, what is this patient on? And I told him. And he was like, this is remarkable. I’ve never seen it before. They’re being more in tune with it now. So more and more part of the conversation upfront rather than seeing patients on the back end.

DR LOVE: So any thoughts about the interdisciplinary dynamic of HCC? You really see the hepatologists, the surgeons, the interventional radiologists very, very actively involved. Any thoughts about, are we seeing a movement maybe away from quickly going to intrahepatic therapy to systemic treatment?

DR KELLEY: I completely agree with Dr Patel’s comment there where our tumor board I think has taken a year and a half or so since atezo/bev for the interventional radiologists to start seeing with their own 2 eyes these examples. And in both our institutional academic tumor board, but also a community outreach tumor board, I’m starting to hear the IR docs or the surgeons say, hey, this is probably a good atezo/bev patient, Katie, or whoever the oncologist is there, you want to see them? And it’s happening a lot earlier. Instead of doing a Hail Mary where people feel just obliged to try something and don’t have a great confidence in any of the choices, people are saying let’s try this earlier and then we can always go back to liver-directed therapy if it didn’t work.

DR LOVE: So we’ll see if the interventional radiologists learn anything about durva/treme as well. Here is the HIMALAYA study. We’ve talked about it before. It kind of looks a lot like atezo/bev in terms of the advantage over sorafenib, at least indirectly. They had this other arm that was durva alone that didn’t seem to do as well. Any comments? Any other things you want to say about this trial? There was clearly an advantage seen with the STRIDE regimen. And you were talking about tolerability. We saw this data that was just presented, again, at the ESMO GI meeting looking at the HIMALAYA study based on liver function, which I think is really interesting. And the first thing, of course, is you all used the different type of grading for cirrhosis. You have the ALBI system. But the bottom line, this was just presented. It doesn’t look like there’s that — I mean, it looks like they benefitted regardless whether with ALBI 1 or 2 or 3. Any thoughts about this and also how you convert ALBI to Child-Pugh?

DR KELLEY: Right. So the ALBI score is a calculation, purely mathematical, of the albumin and bilirubin which is really nice for clinical trial purpose because it can just be extracted from a database and doesn’t require the subjectivity and heterogeneity of clinicians calling mild, moderate, or severe ascites or encephalopathy, which are quite subjective. So it’s a little purer of a metric, particularly in clinical trials. Roughly, ALBI Grade 1 is going to be a Child-Pugh A5, as good as it can be. An ALBI Grade 2 is either somewhere between A5 or A6 in a Child-Pugh. And Grade 3 is more likely to be a B7 although, again, they don’t convert perfectly. Otherwise, we’d just use one or the other. But in any case, I think we always know that a higher Child-Pugh score, a higher ALBI grade, meaning worse liver function, patients have greater toxicity, shorter time to event endpoint like PFS or OS, and usually lower response rates because they have less time to respond. They have more competing comorbidity. So it’s really reassuring to see quite similar numbers for this objective response rate and median duration of response for both subgroups. And likewise, this table you’re showing here —

DR LOVE: What about the tolerability? Yeah.

DR KELLEY: Yeah. The toxicity here, it’s really — I think this is highly encouraging for a CTLA-4 based combination to see if you look, say, at the treatment related adverse events leading to discontinuation, that’s sort of a metric that I think is important. And we see that the discontinuations due to adverse events are quite similar for Grade 1 and Grade 2/3 ALBIs. And conversely, if you look at sorafenib, we see more patients in the higher ALBI, twice as many almost, discontinued. And that fits with what I see in clinical practice is that we usually tend to stop a TKI earlier for toxicity in the poor liver function. And this paper —

DR LOVE: This is kind of interesting too, yeah.

DR KELLEY: Yeah, yeah, totally. This also reinforces that same principle that liver function is preserved or not harmed by the HIMALAYA durva/treme regimen and that the green and blue lines are nice and static, no major change in liver function where the sorafenib, probably correlating with progression more so than toxicity I would guess in this case. But it reassures that the CTLA-4 in the HIMALAYA regimen is not causing undue hepatotoxicity.

Case: A man in his mid 60s with HCC and lung metastases (AFP = 30,000) who receives front-line atezolizumab/bevacizumab — Warren S Brenner, MD

DR LOVE: So I’m just going to flip through. And people can check out some of these papers we put in there if you want to learn more about durva and treme and read about a lot of the basis for it. But let’s get back to some cases. This is a patient of Dr Brenner’s, a patient with HCC and lung mets who gets frontline atezo/bev and doesn’t really respond. Here’s his questions.

DR BRENNER: I have a patient right now that just failed atezo/bev. So we have this whole armamentarium of second line drugs, but they’re all post-sorafenib. So the question for investigators is, is there signs or rationale behind choosing one agent over another in the second line setting?

He had significant pulmonary metastases that clearly progressed on therapy. Alpha-fetoprotein was over 30,000. So I’m going to treat him with dose reduced cabozantinib 40 mg.

And then, what do we do about patients who have Child-Pugh B8? I’ve been treating those patients. You could have a 1 point difference in INR or bilirubin and it makes a difference between a Child-Pugh A/B. That’s important for clinical trial, but I think in real-world practice, it doesn’t make any sense to deny patients potentially much more effective therapy based on very small differences.

The other common question I have is, in patients who have disease confined to the liver, how do we determine which patients should be considered for surgical resection versus interventional procedure such as Y90 or even when do we consider referral for potential transplantation?

DR LOVE: Do you feel like you’re walking down the hall and you’re making rounds and they’re throwing questions at you right and left? Anyhow. You want to tackle a few of those?

DR KELLEY: Yeah. I think the first one is the question on everybody’s mind of what to do in the second line after atezo/bev because we don’t have prospective randomized trial data to tell us yet. I will highlight a trial that’s ongoing called IMbrave251 which looks at patients who had progressed on atezo/bev and get randomized to either lenvatinib or sorafenib according to physician’s choice or the combination of lenvatinib plus atezolizumab continuation or sorafenib plus atezolizumab continuation according to the physician’s choice of TKI. But the randomization is either to continue the atezo or not and then TKI is used second. So it doesn’t really address the question of which TKI to use in the second line, but it does address the question of, is there a benefit to continuing the PD-1 or PD-L1 inhibitor after first line? So we don’t have any results from that trial yet, but that is, I think, one of the key questions on the table right now.

The second question which is what was posed, was which TKI to use. And I’m curious to hear, Neil, what Ghassan answered in his response. I think, I’ll put my money down first, which I think either cabozantinib or lenvatinib are both very reasonable choices. Sorafenib, if a patient has comorbidities and really compromised liver function. I might use sorafenib in a patient that I’m more worried about the compensation and just because it has a longer track record for safety and more deteriorated liver function. But cabozantinib has the advantage of it was studied in the second or third line. And so some patients in the CELESTIAL trial actually had received a checkpoint inhibitor prior to cabozantinib so that in a purest clinical trial perspective, cabo has the best supporting data for using in the second line with respect to after a checkpoint inhibitor combination from a prospective data perspective. I think the rationale also works for either lenvatinib or, again, in certain patients, sorafenib.

DR LOVE: The idea of continuing the IO is really interesting. I don’t remember hearing that kind of strategy in other solid tumors at all. Maybe they’re doing it, but I haven’t heard, like the trastuzumab kind of idea, keep it going and switch the partner, I guess. What about ramucirumab? This patient had an AFP over 30,000.

DR KELLEY: So great question. I think the data for ramucirumab after bevacizumab in HCC has not been studied. And, obviously, there is some data for anti-angiogenics post-progression on bev and ram post-progression on bev in colorectal, for example, but we don’t have such data yet in HCC. I infer that this patient might be more decompensated with respect to liver function as a Child-Pugh B, and so I would hesitate to use ram in a Child-Pugh B patient because there was more toxicity with higher degree of liver dysfunction in the REACH trial and the preceding trials.

DR LOVE: That’s really interesting.

Case: A man in his mid 60s with metastatic HCC and rapid disease progression on first-line atezolizumab/bevacizumab — Shaachi Gupta, MD, MPH

DR LOVE: So I want you to take a look at this next case. Incidentally, this patient had a background of hepatitis C related cirrhosis. And tell me biologically, why is this tumor different than the first case with the patient who responded for a year to atezo/bev? Here’s Dr Gupta. She was all excited that that was what was going to happen, but it didn’t.

DR GUPTA: This is a 66-year-old male who, when he presented to me, he was in a fairly good condition. He was noted to have several liver lesions which were noted on the CT scan. He had bilobar disease. He had portal vein thrombosis. I decided to give atezolizumab and bevacizumab combination. I thought he was a really good candidate for that, and I really hoped he would respond to it. He started that therapy and his symptoms only got worse. And 2 cycles later, he displayed disease progression, but not just increase in the size of lesions, but with new lesions. And I felt like, was there something else that could have been done? Was starting with atezolizumab/bevacizumab the best choice? Should I have started with lenvatinib? The patient had a lot of faith in me and I had a lot of faith in atezo/bev. And it didn’t work out.

One question that comes up, nobody wants to biopsy. They say it’s a diagnosis made on imaging. So there’s no biopsy and we cannot do any further testing genomic analysis. Should that be something that should be looked at?

DR LOVE: And what’s the role of liquid biopsy? I would add to that one. Any thoughts?

DR KELLEY: Yes. So great questions. I guess, quickly, I would’ve definitely used atezo/bev as well as a first line in a fit patient eligible for it. There is data presented in an ASCO abstract from first author Breder et al. in 2021 looking at the main portal vein tumor thrombus subgroup of atezo/bev IMbrave150 trial. And while those patients still showed a benefit compared to the sorafenib arm, their outcomes are much poorer. And so it is very fair to say that patients with, I didn’t catch if this patient had main portal vein tumor thrombus, but with extensive vascular invasion and main portal vein tumor thrombus or VP4, either bilateral or main, the outcomes are just worse. And that’s why those patients have been excluded from some of the other programs including the LEAP-002 and including from HIMALAYA too, for that matter. So this patient had some poor prognostic factors going into it. So I would’ve used atezo/bev but, again, looking at the disease control rate, 20 to 30% of patients will have primary progression even on atezo/bev, even though it’s so successful overall. So this sounds like just one of those patients. And maybe because of his extensive vascular disease, vascular invasion by tumor that he was predisposed.

But I think it brings up, I didn’t get a good chance to peruse the scan on this, but the question about biopsy is a really important one. And while historically, I think a decade ago, that was the party line in a hepatocellular carcinoma, particularly with an eye to the possibility of someday maybe transplanting patients. I think our perspective in the medical oncology realm has completely changed in that, if there’s ever any doubt, we strongly vote for getting a biopsy because missing an intrahepatic cholangiocarcinoma can significantly change a patient’s outcome with other, more active therapies for cholangio, for example. There’s an interesting study of the UK system that showed that there is potentially up to about a 10% failure rate of just using LI-RADS 5 type radiographic criteria in patients with — calling them HCC. There are going to be either mixed HCC cholangio or cholangio in up to around 10% of the time. That was published as a national audit earlier this year, Childs et al., not to be confused with Child-Pugh, but Childs. So I favor biopsy when anytime in doubt. And if there’s any question at all, I advocate for getting a biopsy. I don’t think a liquid biopsy helps most of the time, but if you catch an FGFR2 fusion or an IDH1 mutation, it can definitely, or KRAS for that matter, something that’s more likely in cholangio, it can bias you to say, oh, I better really biopsy this because this doesn’t fit.

DR LOVE: Interesting, KRAS. We’ll get to that in a second.

Case: A man in his early 80s with NASH liver cirrhosis and advanced HCC who develops proteinuria on atezolizumab/bevacizumab — Liudmila N Schafer, MD

DR LOVE: One more HCC case. 82-year-old man with NASH liver cirrhosis, incidentally. This is sort of life with atezo/bev and how you deal with it. Let’s hear what Dr Schafer has to say about this case.

DR SCHAFER: 82-year-old male, good performance status. He had Y90 and then subsequently progressed with retroperitoneal lymphadenopathy, gastrohepatic, periaortic, and gastrohepatic ligament. We started him on atezolizumab and bevacizumab. He has a good partial response on the first line with atezolizumab and bevacizumab per IMbrave150. And, unfortunately, he started developing proteinuria. We evaluated with nephrologist and he did not meet right criteria for nephrotic syndrome. Unfortunately, he developed acute small lacunar infarct. So now, he’s still in good performance status, good response.

And my question, would faculty use in the second line considering that this patient developed lacunar infarct as well as hypertension? Another question is challenge. This patient has high alpha-fetoprotein. And another regimen is ramucirumab. However, we don’t really know what is the role of second VEGF inhibitor?

DR LOVE: So any thoughts about the case? This patient actually is responding, so when she brings up second line, it’s like maybe what’ll happen in the future. But also, whether to continue it now that you had proteinuria, you have this lacunar infarct. Any thoughts?

DR KELLEY: So I think this harkens back to the theme we’ve talked about a lot already which is, what’s different about NAFLD patients? Is it immune microenvironment or is it downstream issues on anti-angiogenic therapy? Is it the stroke, the hypertension, the nephrotic syndrome that happens on bev, or just independent of bev even in a NAFLD population? And so just to underscore that not all of it’s the lack of response to immune checkpoint inhibition, but it’s what happens on anti-angiogenic therapy or independent of all of it. And so this kind of patient, the older, NASH, fatty liver patient, I don’t know if he has diabetes, but those patients are more likely to get proteinuria or hypertension, other complications on bev. It does make us quite interested at having the option of a durva/treme that doesn’t have anti-angiogenic therapy. Proteinuria can be very difficult. I’ve had a lot of patients just like this who happen to be older men, diabetes, hypertensive, and it’s hard to reverse. And nephrotic syndrome doesn’t predispose even other complications like clots.

So I would say first, if the patient’s still responding, I would continue atezo monotherapy as you alluded to, Neil, rather than necessarily stopping the regimen altogether. If he progresses on atezo monotherapy, I think that all of the tyrosine kinase inhibitors do have some degree of risk for proteinuria. I think lenvatinib is higher proportionally in its clinical trials than cabo, so I’d probably favor cabo as a second line. And I’d favor it over ramucirumab given the relative contraindications from a vascular perspective for the pure VEGFR-2 inhibitor.

Journal Club with Robin K (Katie) Kelley, MD

DR LOVE: So speaking of cabo, any comments on this recently published, you’re a part of this study, COSMIC-312 Phase III study, cabo/atezo? I don’t know why you see TKI and IO doing so great in renal, but you kind of don’t see it in HCC. I know they’re different diseases, but a similar concept. Anyhow. Any thoughts about this study and why it was negative?

DR KELLEY: Yeah. So this was a study — I was, obviously, very involved in running this study. The survival showed really no difference between the 2 arms, as you’ve shown here. Yet, the dual primary endpoint of progression free survival did favor the combination arm and showed a benefit. But exactly why the overall survival didn’t mirror that, I think is perhaps related to population. This occurred during COVID and so the proportion of patients with hepatitis B were much lower which, again, harkens back to how predictive of a marker is viral infection. And it may be wrapped up into the populations being studied, but it’s hard to fully understand.

DR LOVE: So here’s another paper people might want to check out. I want to get to biliary tract, but I’ve got to ask you one other question because you were at a symposium we did right after the HIMALAYA data was presented. And when we asked Dr Abou-Alfa what he thinks about first line therapy in our webinar, he gave the same answer he gave then which is, of course, he was the PI, but he’s into the HIMALAYA strategy, the STRIDE regimen. So my question to you is, if you had a patient who was presenting for first line therapy, no contraindications to bev, but had seen Dr Abou-Alfa first and he, let’s assume you could get durva/treme, and that’s what he recommended, would you say, I agree with that? Would you say, I wouldn’t do that, I’d use atezo/bev, but I think it’s a reasonable option? Or would you say, I don’t think that’s a good option?

DR KELLEY: So I think first, I would couch my answer in a, we can’t compare outcomes across trials because there are too many differences in eligibility criteria and timing and all the population differences. So we really can’t compare apples to oranges. I think right now, I’m making the decision between those two regimens based on comorbidity profile until we learn more about NASH and NAFLD microenvironment. I think it’s really about comorbidity and risk factors for anti-angiogenic therapy, and immunotherapy too for that matter. In a vacuum, I myself have been choosing atezo/bev for the fit, non-comorbid patient because the response rate I think is, the absolute numbers comparing between trials, which I just said you shouldn’t do, is higher and we see an 8% rate of complete responses. That said, I think longer term follow-up will be very important as we look at, say, the HIMALAYA data, we see the 3-year survival rate for the STRIDE regimen is 30%. I’d like to see the long-term landmarks for atezo/bev as well and compare those again with the caveats of cross trial. I think to walk back what I said, I think right now, we can’t draw direct comparisons across trials. And so it’s really a patient by patient comorbidity driven decision.

DR LOVE: All right. Let’s talk a little bit about biliary tract cancers and, of course, the TOPAZ data that was presented at the same meeting, the GI ASCO meeting. Two gigantic papers this year, at least in terms of these topics. So it’s pretty much straightforward, chemo +/- IO. We’ve seen that story before. It’s standard in lung cancer. Or actually, IO alone also is a possibility in lung. Any thoughts about what was seen? It looks like a modest improvement. Not a huge downside. I hear a lot of people saying it wasn’t maybe a homerun, but I’d like to do it. Is that the way you see this?

DR KELLEY: Yeah. I think, again, we’re not seeing a huge synergy signal that we’re converting a ton of patients to immunotherapy responders, but it looks like across all the endpoints from objective response to progression free survival to OS, the addition of durva improves all of those metrics quite symmetrically, and across subgroups on that forest plot you just showed, they’re pretty similar. And thankfully, without a safety signal. So while I would caveat the data saying we’re not seeing a dramatic increase of converting immune non-responders to responders, there appear to be a subset that are because the tails of the curve actually grows with time. The proportion with ongoing response in survival is slightly greater with landmarks passing. So, to me, those are all metrics that show it’s a positive study and it helps. It doesn’t help quite the percentage of patients I would have wanted it to, but it’s consistent and statistically significant.

DR LOVE: Any thoughts about this data that was just, again, presented at the ESMO GI meeting now looking at tumor location and type of tumor? It kind of looks like, when you look at the forest plot, it looks like everybody benefits, intrahepatic, extrahepatic, gallbladder. Any other comments about this sort of subset analysis of TOPAZ-1?

DR KELLEY: Yeah. Going back to our theme of viral versus nonviral disease and immunotherapy, the groups that are interesting here I think are Asia versus other, bearing in mind that intrahepatic cholangiocarcinoma also tends to be a disease with a risk factor of viral hepatitis and Asia has a higher proportion of patients with B positivity as the cause of their liver disease, whether it leads to cholangio or HCC. But it’s reassuring to see a very similar trend for all the forest plot subgroups whether Asia or Europe. There was another poster that looked at viral versus nonviral which, again, showed a more prominent benefit for the viral group, but still a benefit for nonviral that mirrored it. So I think overall, we are seeing more pronounced benefit in Asia in nonviral, both in the TOPAZ-1 as well as, say, the IMbrave150 trial. But it’s not 0 and there’s still trend towards benefit in the other subgroups.

Case: A woman in her early 70s with metastatic pancreaticobiliary cancer thought to be intrahepatic cholangiocarcinoma — Priya Rudolph, MD, PhD

DR LOVE: All right. Well let’s jump into sort of the clinical day-to-day real world of biliary tract cancers. We’re going to start out with a patient, a 71-year-old woman. Actually, it wasn’t really clear what the primary was as you’re going to hear. There wasn’t a lot of tissue. It was thought to be a pancreatic biliary tract cancer, thought to be an intrahepatic cholangio. They weren’t 100% sure, and I’m not sure it matters. But in any event, here’s the case from Dr Rudolph.

DR RUDOLPH: The liver lesion was biopsied and was moderately differentiated adenocarcinoma with pancreatic or biliary primary based upon immunostains which was CK7 positive and negative for CK20, CDX2, TTF1, and GATA3.

My questions are, her liver lesions are very small. Is blood based NGS testing sufficient for detection of FGFR alterations or IDH1 mutation? Is there really any reason to pursue tissue biopsy or can we work with blood? The other question that always lingers in the back of my mind is, I’m treating her as cholangiocarcinoma. How do investigators confirm that this is biliary primary when you kind of have this situation to work this? I would love to know if I could extrapolate the TOPAZ data for this patient and use chemotherapy along with immunotherapy. It’s not on the NCCN guidelines and it’s not FDA approved. However, I would love to know how investigators are doing it either upfront or is that in an advanced setting after progression?

DR LOVE: Any thoughts?

DR KELLEY: Yeah. So to the question of diagnosis, cholangiocarcinoma has historically been to some degree a diagnosis of exclusion. If there’s a biopsy showing adenocarcinoma occurring in the liver, how do we know if it started there or traveled there? And so there are some cases where the radiology can really answer that for people who look at a lot of liver tumors. If there’s a dominant, very large tumor in one lobe, for example, and satellite lesions around it, and then some scattered lesions in the contralateral lobe, that really supports the hypothesis it started in the big lesion and then traveled. If there is signs of progressive scarring and mass effect on a bile duct, for example, like biliary dilatation distal to the tumor or puckering like the radiologist will describe an atrophic or puckered appearance of the ipsilateral lobe, that suggests that the tumor is sort of arising from the bile duct and pulling, tugging on the capsule of the liver so it creates sort of a puckered appearance. If you have a tumor that you’re describing with a lot of little small tumors scattered throughout the liver, those are ones where we really want to do a metastatic workup and make sure the patient’s had colonoscopy, upper endoscopy, probably I’d get a PET in those patients in addition to regular scans, make sure that we’re not missing a small bowel or other primary. But if, again, I think to the question about next generation sequencing, it can rule in or nearly rule in a cholangio if you get lucky and get an IDH1 or an FGFR2 fusion. If you get something like KRAS or get a lot of nonspecific p53, for example, it doesn’t help you.

Now to your — the other part of the question was, does cell-free DNA, is it adequate if you have scant tissue which is often the problem? I do get it if I’m worried that the biopsy was just an FNA or an ERCP biopsy where it may not be enough tissue to get tumor based next gen sequencing. But it is important to know that some platforms have, most platforms are not able to detect all of the potential FGFR2 fusions. If you catch an FGFR2 fusion, it’s real and it’s there. If you don’t catch it, it may have been missed because there are so many different potential intronic partners. It is — cell free assays are very good for coding regions of hot spot mutations like in BRAF or KRAS or IDH1. So you can feel very confident that you didn’t miss it with a cell-free assay in those settings. But FGFR is tricky because it’s intronic.

DR LOVE: You mentioned earlier these sort of combined tumors where it’s like HCC and cholangio and it seems like it’s somehow mixed or 2 tumors. What do you do in that situation?

DR KELLEY: We don’t really know. They have a common progenitor cell and there can be transdifferentiation or plasticity in the evolution of these therapies — of these cells. I almost always will try to get next gen sequencing of a mixed tumor in hopes of catching a targetable mutation in the cholangio component. And if it has IDH1 or FGFR2, I treat it accordingly. Otherwise, it’s really a best guess scenario. There’s some data out of colleagues of mine in pathology at UCSF that hep C related mixed HCC cholangios tend to look potentially more like HCC, but, and so we might try to — I basically talk to the pathologists and use all the clues that we have and start with, pick a path and go down it. If it’s truly mixed where you just have some cholangio, some HCC, and they’re both there, the TOPAZ-1 regimen makes a lot of sense because you’ve got durvalumab which showed benefit in the HIMALAYA trial in just HCC, and then you have the gem/cis which we know works in cholangio. So I think TOPAZ-1 is a nice regimen for the really true, you’re convinced it’s a real mixed tumor. So that’s the gem/cis plus durvalumab.

DR LOVE: So somebody just, and the chat room typically will read my mind. I don’t know. After a while, these people know what I’m thinking. But I was just about to show this slide here and Hassan in the chat room asks the question of, can you see patients who receive a first line FGFR and then benefit second line? This is a slide from Dr Valle’s program that I was just mentioning earlier. And I was surprised to hear that, we were talking about covalent BTK inhibitors, pirtobrutinib. So exciting. In CLL, you see responses in people who don’t respond to noncovalent BTK. And it kind of sounds like there’s a similar story here with futibatinib. They all have a typical mutation kind of waterfall plot. But my understanding is that futibatinib, you can see responses to people who’ve had one of the other 3. Have you actually seen that? Is that the case and what’s the reason? I guess it’s covalent. Does that have something to do with it?

DR KELLEY: Well actually, I think the reason is more that futibatinib, because of its binding location, and its covalent properties, but it has activity against the common resistance mutations. So you’ll see an author on a lot of our FGFR data is Lipika Goyal from Harvard and she and colleagues at Harvard have really looked at the kinase domain resistance mutations that arise in the setting of selection pressure from an ATP competitive inhibitor like infigratinib or pemigatinib. And there are a host of different places in the coding region, not where the intronic fusion happens, but in the coding region where the tumor escapes the inhibitor, gets a new mutation in the kinase domain, and then becomes resistant to the ATP competitive inhibitor. Futibatinib, because it’s not binding at the ATP binding pocket, actually inhibits some of those resistance mutations, not all of them. The exciting part is that — so futibatinib is an option for some of the resistance mutations, there are new drugs that are being developed that are going to hit more of those resistance mutations. So it’s going to become like an ALK or an EGFR story where we’ll have serial FGFR inhibitors and be testing for resistance mutations to inform which one to use.

DR LOVE: So I wanted you to hear —

DR KELLEY: It’s really exciting.

DR LOVE: Super exciting. I call biliary tract cancers the new non-small cell lung cancer. Just kidding. But there, they’ve got like 10 targetable lesions that they focus on. A question in the chat room. Of course, T-DXd is on a lot of people’s mind. We are talking about the HER2-low presentation in our program next week which is really incredible, incidentally, with Dr Modi. But there was also a paper at ASCO on T-DXd in biliary tract cancer, I can’t say it was totally shocking, that looks like it works, at least in some patients. It seems like it works. It works in lung, upper GI, colon, of course, breast. Any thoughts about this? Have you seen HER2-positive biliary tract? Have you ever used T-DXd in these patients?

DR KELLEY: I have just started my first patient. It was a gallbladder as it turned out. But I think that this data plus the MyPathway cholangiocarcinoma cohort, biliary cohort I should say, of pertuzumab/trastuzumab which also showed an intriguing durable response rate for that combination, both of those really suggest that we should be looking at HER2 amplification or overexpression in our biliary tract cancer patients. The prevalence is low. For gallbladder, it’s the highest, it’s in the 15 to 20% range. For cholangios, it’s lower depending intra versus extra, up to about 5% intra, 10, 15 in the extra. So it’s not as common, and with such a small tumor type. This is not breast cancer or even not gastric cancer either. So it’s going to be incredibly difficult to validate exactly what the cut points are for this location pathologically and to do randomized Phase III trials. But I think that these results being symmetric to what we see in other bigger population tumor types really reinforces benefit. And I am using HER2 targeted therapy. And actually, the biggest practice change is to make sure that our next gen sequencing panel includes HER2 or that we check HER2 IHC in addition if it doesn’t.

DR LOVE: Speaking of HER2 IHC, what about HER2-low? Want to use it in HER2-low biliary tract cancer?

DR KELLEY: That, it’s much smaller numbers, so I don’t know about that yet.

DR LOVE: And, of course, you all have to get into the — oh, you’re already in that from the other tumors, the whole issue of ILD and how much of an issue that’s going to be. Let’s see if we can maybe —

DR KELLEY: Well I was going to raise that because I guess just a quick aside, Neil, is just that biliary does have a predilection in Asian populations. And if we think there’s a pharmacogenomic or other reason that it’s a higher risk in Asian populations, that’s particularly important for, just like in lung cancer, it’s important in biliary to be aware of.

Case: A woman in her late 40s with intrahepatic cholangiocarcinoma and micrometastases in portal lymph nodes — Pavel A Levin, MD, PhD

DR LOVE: All right. Let’s see if we can squeeze out one more case here. This is a 49-year-old woman, a patient of Dr Pavel Levin. Here’s the case.

DR LEVIN: So according to NCCN guidelines, somebody with intrahepatic cholangiocarcinoma with micrometastatic disease to portal lymph nodes, the recommendation is to try to avoid surgical resection because oftentimes, at that point, it’s already high likelihood for recurrence. So what are their criteria that they are using to decide on who would and would not be a good candidate for surgical resection if the surgeon feels that he would be able to resect the disease completely? In this kind of situation, what would be the preferred neoadjuvant regimen in somebody without targetable mutations? What agents would they use in somebody with perfectly good renal function and in somebody with, let’s say, suboptimal renal function where cisplatin would not be an option?

DR LOVE: Any thoughts?

DR KELLEY: So to the question of surgical resection, I think there was a very interesting paper, I believe 2021, from Juan Valle was one of the authors. He’s one of your speakers in a couple weeks, I saw. But the first author was Angela Lamarca and it looked at SEER data as well as ENSCCA European data on the impact of multi-focality in intrahepatic cholangiocarcinoma. What they showed is that intrahepatic cholangiocarcinomas that are not a single lesion, but multi-focal have pretty similar prognosis to metastatic disease and node-positive disease also. And so what it tells us is that we have to be very careful as a clinician to send someone to a big surgery when there are, say, big conglomerate tumor with a lot of satellites around it. You take out half of their liver and they already have some biliary obstruction. That may really cripple the patient’s options for downstream therapies. And they have a very, very high risk of behaving like an M1 metastatic disease. So it’s based on 5,000+ patients from a combined SEER and European data set, Lamarca et al. 2021, I think it was Journal of Hepatology. So that’s a caveat to upfront surgery in patients like this.

I think there are very — to the question of neoadjuvant therapy, it’s small numbers of patients, very hard to do prospective randomized trials. But I think in a patient where you think they’re likely to be advanced, I would definitely give them a good 6 months or at least 3 months of systemic therapy before re-entertaining the idea of surgery. Which regimen? I think there’s 2 front runner options. Of course, for unresectable disease, there’s the TOPAZ-1 regimen which is probably my first choice. But there’s another important regimen to have on the radar which also has a higher response rate which is the triplet of gemcitabine, cisplatin, and nab-paclitaxel, the SWOG 1815 trial, which has not yet reported its Phase III data, but in a Phase II, showed a much higher response rate, close to around 40% range. And so that regimen has attractive appeal as a short-term neoadjuvant strategy for cytoreduction. For renal dysfunction, I think the platinum and cisplatin, gem/cisplatin is low. The cisplatin dose is only 25 mg per m². So we can be a little more permissive than, say, a full dose cisplatin that we use in esophageal or other cancer types. That said, if the patient has really significant renal disease, gem/nab-paclitaxel, so gem/nab-paclitaxel without platinum, has also shown response, good response and improvement over gem alone in a randomized Phase II.

DR LOVE: Awesome. Well, Katie, thank you so much for working with us today. Audience, thank you for attending. Come on back on Monday. We’ll see what Dr Bradley and Spigel have to say about what’s new in management of locally advanced disease. So many things happening in lung cancer, neoadjuvant, IO plus chemo. We’ll see what they have to say about that as well as what’s happening in terms of the use of durva. Be safe, stay well, and have a great night. Thanks so much, Katie.

DR KELLEY: Thank you.