Ghassan K. Abou-Alfa, MD, MBA: I’d like to go back to our discussion about the etiology-specific or etiology-nonspecific therapy. You are hearing different opinions. Already I’m going to start to summarize it, but let me hear first further thoughts. Ahmed, a patient of yours comes in and you say, “Patient, you have hepatitis C—related HCC [hepatocellular carcinoma]. I definitely would like to commit you to sorafenib. Of course, I’m going to do that after regorafenib.”

And the patient looks at you and says, “I heard about this drug lenvatinib,” and they really insist they want to take lenvatinib. Would you be adamant and say, “No way I’ll give you this”? In other words, how solid are your data in regard to hepatitis C—related HCC should be treated with sorafenib?

Ahmed Kaseb, MD: Yeah. Up front we really don’t have any randomized data, and that’s what we all at last will be able to discuss with our patients. In general, you just have to know your patient inside and out. By that I mean upper endoscopies—you look at the scan yourself. You evaluate the varices. If this patient has got grade 3 varices, and the risk of bleeding on lenvatinib is higher than sorafenib, you really have to weigh the risk-benefit ratio and discuss it up front with the patient. A lot of patients love to be able to discuss the pros and cons. We all would love to have nice randomized data in every risk factor—hepatitis C—related HCC versus hepatitis B. But we all know we’re never going to see those studies. They’re very hard to conduct and accrue and complete. Until then I think it’s going to have to be up to the clinician and the patient weighing the risk-benefit ratio, the risk of very strong anti-VEGF drug like lenvatinib on their blood pressure, maybe their blood pressure is very hard to control with 4 antihypertensive medications. Their varices are at severe risk for bleeding. These factors are going to have to determine what to start with.

Ghassan K. Abou-Alfa, MD, MBA: Anthony, to carry on the same question—I like what you said, but I would like to dissect it further. And if you recall from your nice data in regard to the nivolumab, which I remember many years ago, even when you presented it at the ASCO [American Society of Clinical Oncology] Annual Meeting, I also was asked to comment on how much clarity or nonclarity we have in regard to etiology-based responses to checkpoint inhibitors?

Anthony, I hear that with the checkpoint inhibitors, which I remember you nicely presented more than once on the CheckMate040—and I remember I was asked to comment on it at ASCO when he presented this. Is there any kind of particularity or interest in regard to hepatitis C versus hepatitis B response, or is it the same?

Anthony B. El-Khoueiry, MD: The initial data we have again with both NIVO [nivolumab] and PEMBRO [pembrolizumab] are that responses occurred across all groups independent of etiology. Again, in the cohorts, for example, CheckMate040 had separate expansion cohorts: hepatitis B, hepatitis C, noninfected, sorafenib exposed or nonexposed. None of these was really powered to be compared with one another, but numerically you saw objective responses across all the cohorts, and they were within the same range when you look at the confidence intervals. At this point these are the data we have.

When we start looking at the larger phase III studies—the CheckMate 459 in first line, KEYNOTE-240—we’re still relying again on subgroup analyses, and that is a noninfected group. For example, in CheckMate 459 that’s called out, if you look in the subgroup analysis maybe that group doesn’t benefit as much, but who knows what that means. It’s a subgroup analysis—the numbers get much smaller. So we’re going to need in the future, as we now have more available therapies, to really refine how we do clinical trials and dig deeper into the etiology story and maybe marry the etiology and the carcinogenesis mechanism to the therapy a bit better. But my short answer is, right now responses seem to occur across all etiologies similarly.

Ghassan K. Abou-Alfa, MD, MBA: Catherine, back to you. Regarding hepatitis B and C, you mentioned already and you described it very clear—they are not friends, they are not brothers, they are not even cousins. They are really separate from each other. How much do we know? If you recall from your part of the field a long time ago from the NCI [National Cancer Institute], describe to us how that pathway and the cancer evolve in regard to hepatitis B, being it’s totally different from how it evolves in regard to hepatitis C. For example, the RAF, MEK, ERK. Dr Kaseb is really pressing the interest in sorafenib for hepatitis C. How much do we make out of this?

Catherine T. Frenette, MD: It’s very interesting about the molecular mechanisms with the different viruses as well as with the metabolic-related liver disease, such as NASH [nonalcoholic steatohepatitis]. There are also different mechanistic developments in that case as well. We’ve also been able to molecularly subtype HCC into about 6 different subtypes depending on which pathway you choose. I’m not sure at this point that we have enough data to say hepatitis B patients should get X, and hepatitis C patients should get Y. We clearly know that if somebody has virally active disease, they benefit from control of the virus, right?

For hepatitis C patients, if you get the virus under control, you treat them, you cure them, and their liver generally does better. For hepatitis B patients, you suppress the virus during their cancer treatment. It’s clear that they respond better to the cancer treatment and their liver function stays stable. That’s about as much as I would make of the underlying etiology at this point, other than saying it’s really interesting, and I think there is something there. We just need more data.

Ghassan K. Abou-Alfa, MD, MBA: I like it, and I’ll go back to Pierre. Let’s take an example of cabozantinib. Admittedly I know the data, and as I recall, at least as best as I can say, it’s not as if there is any amazing, beyond-impressive difference between hepatitis B versus hepatitis C. Would you say that cabozantinib should be used more preferentially for 1 group versus the other, or it’s really available for all?

Pierre Gholam, MD: I would not. I think it remains. The data on etiology of liver disease dictating differentially 1 agent or the other are certainly not there. Perhaps, as Dr Kaseb pointed out, the best longitudinal evidence of some sort of signal for what agent is more effective in HCV [chronic hepatitis C] patients versus others is that for sorafenib, interestingly not only from the positive studies in SHARP and from the Asian Pacific, but even more so in failed trials, including those for brivanib, erlotinib, etc.

If one makes a rationale in first line choose sorafenib in those patients, it might not be entirely based on just opinion. It might actually have some evidence to support it. Beyond that, I do not see that there is anything beyond maybe nuances that would guide us in terms of our treatment decision.

Ghassan K. Abou-Alfa, MD, MBA: I like the fact that we were able to put everybody on the same page. We kind of started a little bit with different views in regard to the etiology and the treatment choices, but it seems we settled on recognizing the limitation of any data that really existed to us in 1 way or another. And I’m saying that after all, if a patient needs cabozantinib, for example, irrespective of the etiology, they need the cabozantinib. Interestingly, to go back to the point that actually also Catherine was bringing up, despite that we had those data published more than 20 years ago that tell us that the etiology would drive a certain pathway in regard to the development of the disease. Data that were published recently is suggesting differently. If anything, from our program at Memorial Sloan Kettering Cancer Center, we published that in CCR [Clinical Cancer Research], and we have shown actually that I think a little bit of a Manhattan subway map. Hepatitis B is like right on the west side subway line, Nos. 1, 2, 3, the red line. Hepatitis C is on the green line, on the east side, Nos. 4, 5, 6. It’s as if they are totally separate. They evolve with the HCC along those 2 lines, so it’s not really the case.

We all know there is what’s called the S Line—or the shuttle line between the west side and the east side, that’s for 42nd Street—and very well the hepatitis C, and hepatitis B and hepatitis C can mix up together. It’s really only a suggestion that, after all, patients who really have a certain etiology of disease, the idea that they are only on 1 specific pathway of development of their disease might not necessarily be true, and others under certain thresholds might actually evolve to another line of the evolution of the genetic makeup of the tumor. This is very important to go back to where we started. Biopsy is critical, as we heard, and funny enough. We might not even leave you with only biopsy, but we might leave you in the future. We might say to you that biopsy is critical on more than 1 occasion; ie, at every progression, maybe we need a biopsy. But time will tell of course.

Transcript Edited for Clarity