Understanding How Best to Sequence in Advanced HCC

Transcript:

Ghassan K. Abou-Alfa, MD: I’ll go back to the point that we kind of, we spoke about and referred to, and really based on those 7 drugs that we have, plus the data for. And as you see, the field is getting very busy in regard to many other drugs per se, and it’s not necessarily a point of preference, but I would like to ask the question from the perspective of patients [who] come to you, Rich, and they say, “You know what, I heard about this immunotherapy. I want it now.” And how would you really now, especially with the approval as we said, of nivo [nivolumab] and pembrolizumab in second line, would you really go to that? And short of clinical trial, of course, go for checkpoint inhibitors first line? Or is it a TKI [tyrosine kinase inhibitor] per se?

Richard S. Finn, MD: You know, there [are] no data yet that [show that] using these drugs improves survival in any setting. And while we do have very provocative response, duration of disease control in second-line, their approval is currently limited to second-line therapy. So in the frontline, really the go-to, outside a clinical trial, is either sorafenib or lenvatinib.

With that being said, and this is for unique cases, now that we have all these drugs, and I’m sure we’re going to talk about how we sequence all these drugs, there are patients who we see who still fall into a gap, and this is outside of guidelines. These are patients who, in my opinion, are very sick, they have advanced liver cancer, and they need some big response.

And there [are] some of these patients—and we can talk about what they look like, but they have a big tumor burden—and if they aren’t 1 of these people, [who] has a dramatic response, they’re probably going to die pretty quickly. They’re not going to live long enough to get a TKI, then an I/O [immuno-oncology] second line, or a TKI in sequence 3 or 4 drugs. So I would be dishonest if I said I’ve never given an I/O drug frontline, but it’s certainly the minority of patients, they’re not proven, they’re not labeled there. But we do it with the hope that they’re 1 of those 15% to 20% of patients who have a significant response. Because if they don’t, they’re going to probably succumb to their disease pretty quickly. And before this, these were maybe the patients at 1 point I wanted given chemotherapy instead of NEXAVAR, because they needed a response.

Ghassan K. Abou-Alfa, MD: Andrew, same question but in a different setting. Of course, we all get excited about our data, and the question would be, you have a patient on sorafenib first line. It happens, you know? Would you give pembro [pembrolizumab] as second line, or would you do rego [regorafenib] first and then pembro; ie, do you position the checkpoint inhibitors in any way, shape, or form?

Andrew X. Zhu, MD, PhD: So again, in 2019, I think when we make this type of decision we still don’t have a very clear guideline—I have to confess. But having said that, we have to take into consideration a few things. I think Richard mentioned about the tempo, the tempo of the disease progression, the tumor burden, which I think probably would drive the clinical decision to a certain extent. So if I have a patient with explosive disease [who] actually requires immediate intervention, something that’s really active to control the disease quickly, I probably will be more inclined to offer checkpoint inhibitor as a second line right away.

But if I have a patient doing incredibly well on sorafenib for a prolonged period of time, just a few pulmonary nodules that are changing in size by 0.5 cm, I think it’s very reasonable to switch to the regorafenib, in my opinion. So again, we have to take into consideration the individual patient, how they respond to sorafenib, and was there tumor burden? I know so obviously there are other variables that we’ll take into consideration, including the alpha-fetoprotein level, things like that.

Ghassan K. Abou-Alfa, MD: Fair, fair. Katie, along the same line, if you recall on the correlative studies that were done on the SHARP [Sorafenib Hepatocellular Carcinoma Assessment Randomized Protocol] trial of sorafenib versus placebo that were reported 7 years after, it was intriguing that patients who did not respond had a high AGF, which, as we know, could be related to c-MET per se. From the cabozantinib standpoint, would you make an argument [that] if a patient progressed on sorafenib, you’ll go for the cabo based on this and some other facts? Or would you rush directly to checkpoint inhibitors?

R. Kate Kelley, MD: I think that’s a great point about [how] we don’t fully understand how to sequence these drugs. I think I also would interject the data we have from the tivantinib study, which tells us that from the biomarker analyses of tivantinib, not the study itself but the data from the METIV-HCC tivantinib study, which showed that patients who had had prior sorafenib had higher tumor c-MET levels than prior to sorafenib therapy. Meaning that sorafenib exposure increased MET expression because it’s a very plastic marker and can upregulate in response to certain other therapies. In this case such as antiangiogenic therapies.

And so the rationale that you’re alluding to is the question of might a MET-targeted antiangiogenic agent do better in succession directly after a TKI such as sorafenib rather than interrupted by a PD-1 [programmed cell death protein 1] inhibitor, based on the rationale that you mentioned of the HGF, overexpression as a mechanism of resistance, maybe primary resistance, but also the potential for secondary resistance, upregulated on the TKI therapy.

So I think that is exactly the kind of study we need to be doing right now to understand, rather than flip-flopping just to try something different. I think our next generation of clinical trials needs to take—now that we have the luxury of 7 drugs to work with—to take rigorous samples, particularly noninvasive samples, as much as we can to try to understand how best to sequence drugs. Because I do agree that there may be a mechanism of efficacy that depends on prior sequencing right after an antiangiogenic agent like sorafenib, particularly for cabo.

Ghassan K. Abou-Alfa, MD: No, by all means, I totally agree. If anything, the sequence that we are seeing, and clearly it’s evident and proven, is, for example, sorafenib-rego. The other one [that] probably is inviting or at least intriguing to think of is, for example, sorafenib followed by ....; correction, sorafenib followed by cabozantinib. And interestingly, that this cabozantinib can be even effective as a third line, per se. And interestingly, despite those sequences that you are seeing, ironically when it comes to the checkpoint inhibitors, we really have no understanding yet where do they fit best. And this is something that hopefully from the studies that are currently ongoing, being the nivo versus sorafenib first line, as well as ramucirumab versus placebo in second line, will hopefully shed some light in regard to that setting.

Transcript Edited for Clarity