Hepatocellular Carcinoma: Practical Implications of Emerging Data - Episode 13

Sequencing of Novel Therapies in HCC

, , , , ,

Transcript:

Richard S. Finn, MD: I’m going to challenge all of you because of the rapidly changing landscape. We’re not curing most patients. Atezolizumab-bevacizumab does have, and all the I/O combos seem to have, this 30% plus or minus group that appears to be very long-term survivors. All of us have used these drugs and probably see patients who are cured. Presumably, patients are going to go beyond their second-line agent, even patients who get prolonged stable disease from frontline. All of us hope these drugs are being used earlier in the disease course, as we talked about before using them. Before we see patients who are decompensated, as we get more liver cancers that are earlier in detection, and the whole staging system hopefully will move to the left, to earlier-stage disease and using these drugs earlier. Patients will be going to third- or fourth-line. We can all say what our preferred sequence is, at least for second-line. Obviously, the data get even less supportive as we go to third or fourth line. Tony, do you see a role for sequencing TKIs [tyrosine kinase inhibitors]? Or do patients get frontline I/O combination, second-line TKI, and then we stop because there are no more data? What’s your thought?

Tanios S. Bekaii-Saab, MD, FACP: We haven’t stopped before, so why stop now?

[Laughter]

Richard S. Finn, MD: Tell us how you feel, Tony.

Tanios S. Bekaii-Saab, MD, FACP: I don’t disagree that we’re going to get more patients hopefully to third-line. That’s the natural history: When you do really well in the first-line, you start doing better in the second- and the third-lines. The point is, we’re going to have to sequence. Hopefully, we’ll have more studies to fill these spaces. Let me go back to 2 things. Your second-line has to be a TKI. There’s no doubt about it. You have 4 options you all talked about. I have my preferred 2. I think Katie aligned with that as well. We have 2 agents that have proved value. That doesn’t discount the others. My least preferred remains sorafenib, frankly. I don’t think I’m going to waste much thought about it. I’ll have 3 to work with, 2 favorites in the second-line, and in the third-line, if I do regorafenib, I’m going to look at cabozantinib. If I do cabozantinib, I will do probably regorafenib. You can still argue for lenvatinib as well, perhaps. I’m not sure. I want to go back to the point. There were interesting data from ASCO that looked at nivolumab plus ipilimumab in patients who failed nivolumab. It is not inconceivable that a dual checkpoint inhibitor—adding the CTLA4 inhibitor in patients who failed atezolizumab-bevacizumab, whether in second- or third-line—may actually pull some weight. Of course, it needs to be studied. Theoretically, we’ve already seen that in RCC [renal cell carcinoma]. We see some of it in melanoma. It’s not inconceivable that you can continue to salvage patients with I/O. Just change the partner. If the partner is bevacizumab in the first line, atezolizumab-bevacizumab, when you fail atezolizumab-bevacizumab, did we really fail the bevacizumab or the atezolizumab or both? Can we go to wherever? I/O of choice plus, say, lenvatinib or cabozantinib or regorafenib, which hits multiple other targets, including some that will reverse the VEGF effect on immunosuppression or the VEGF effect on the tumor. I went through the point that we’re not going to make a bigger difference in those patients until we start asking combination questions in second- and third-line, not just limited to first line, where most of the landscape right now is. I hope it’s not going to take us another 10 years to figure out that this is what we need in second and third line.

Richard S. Finn, MD: That’s really excellent commentary, because as we’ve heard before, there are several combinations being developed. I asked Anthony about CTLA4 and I/O, but there are TKI combinations—lenvatinib-pembrolizumab, atezolizumab-cabozantinib—and they’re all being studied frontline. Let’s say they get approval. You can use only 1 frontline. Does that mean there’s no role second-line? Those are very difficult questions to answer, these sequencing-type studies. I don’t know when we would even have that data. It would take time to try to answer that. As we see, the natural history of liver cancer is getting longer. We can likely tell patients who are Child-Pugh A frontline that median survival is probably 2 years, I suspect. Obviously, the IMbrave150 data read out very quickly, and the long-term follow-up from the treatment arm has not matured. We know that there’s a 42% decrease in the risk of death with that combination up front, but it’ll take longer follow-up to see what the true medians are. That median is going to reflect what they get not only front line but also in subsequent lines of therapy.

Transcript Edited for Clarity