Sequencing Agents in HCC Based on Patient Subgroups

Transcript:

Richard Finn, MD: Let’s see, right now we have 4 drugs approved in the United States in second-line: regorafenib, cabozantinib, nivolumab, and pembrolizumab. And presumably ramucirumab will be approved at some date. How do we think about this? Do we think about this as frontline TKI [tyrosine-kinase inhibitor], second-line TKI; we’ll just batch TKI versus I-O [immune-oncology] therapy; or do we go TKI, TKI, then I-O? What do you think?

Anthony El-Khoueiry, MD: It’s difficult. Let’s say a patient progresses on first-line therapy, and I have the option today of using regorafenib, cabozantinib, nivolumab, or pembrolizumab. Let’s batch the TKIs together and the PD-1 [programmed cell death protein 1] agents together for a minute. I think the TKI sequence again is a very valid sequence based on phase III evidence, level 1 evidence. And it’s actually a great option for patients who have tolerated TKI therapy. Patients who didn’t tolerate TKIs very well in first-line, I tend to prefer switching them to a different agent with a different safety profile, ie, an anti—PD-1 agent.

Richard Finn, MD: You know I’m going to pin you down on that. Because from the RESORCE analysis that you alluded to where we looked at the survival of the sequence, we also looked at toxicity. Patients who needed a dose reduction with sorafenib, they derive the same benefit with regorafenib. There was similar toxicities for that group of patients, maybe a little higher with full dose regorafenib. But the time on sorafenib didn’t necessarily predict time on regorafenib. The magnitude of the benefit was the same. So, when you say didn’t tolerate frontline TKI, whether it’s regorafenib, lenvatinib, or sorafenib, what does that mean didn’t tolerate it well?

Anthony El-Khoueiry, MD: The easy part of the answer is the patients who would not have qualified for the RESORCE trial with regorafenib, meaning patients who couldn’t even tolerate 400 mg daily of sorafenib. Those patients do exist.

Richard Finn, MD: Yes.

Anthony El-Khoueiry, MD: So those patients, the easy decision is to put them on an anti—PD-1 agent. For the patients who would have met the strict criteria for the RESORCE trial but had repeated grade 2 toxicities, repeated interruptions, I agree with you that the evidence tells us that they could derive the same benefit. But, from a patient perspective and a practitioner perspective, it’s harder. The patient is suffering more, the physician is having to do more interventions, and it may be that in practice those patients tend to be switched to an anti–PD-1 more frequently. But again, this is just practice, not science.

Richard Finn, MD: Of course, yes.

Anthony El-Khoueiry, MD: The last group of patients where we may preferentially want to switch to an anti—PD-1 agent in second-line are the patients who are progressing quickly, where they may not make it to a third-line option. And many physicians feel like we don’t want these patients to lose the opportunity to be on an anti–PD-1 agent in case they are one of those responders. Because we know that the responders tend to do very well, have prolonged duration of response. And that seems to also be associated with really prolonged survival. So, if the patient is not going to make it to third-line, we want to give them the chance of exposure to an anti–PD-1 agent. That’s a clinical judgment.

Richard Finn, MD: How many patients out of 100 are those who get PD-1 and have that great response? What would you estimate that to be?

Anthony El-Khoueiry, MD: Again, I roughly would estimate somewhere between 15% and 20%. But, if we look at CheckMate-040, which has longer follow-up now than the pembrolizumab data, when we look at patients who achieved a response, their median survival has not been even reached.

Richard Finn, MD: Yes, they do very well.

Anthony El-Khoueiry, MD: They do very well.

Richard Finn, MD: And I think all of us who have used these drugs, whether pembrolizumab or nivolumab, if you respond, you do well.

Anthony El-Khoueiry, MD: Yes.

Richard Finn, MD: What about stable disease with them? So let me bring up a few points. We have the sequence of sorafenib to regorafenib, sorafenib to cabozantinib. I’m going to ask you a few questions. You talked about the patient who didn’t tolerate sorafenib so well and maybe you go to regorafenib. Would that influence your choice for cabozantinib? Is it the same logic?

Anthony El-Khoueiry, MD: Probably. Again, because the adverse effect profile tends to be largely overlapping as another TKI.

Richard Finn, MD: Yes. We’ve talked about, and you alluded to the fact, that maybe the RESORCE study was a very selective group of patients. And what I always find interesting is even though they required this minimum period of sorafenib, they had documented disease progression, the control group for RESORCE behaved just the same as the control group in all the other phase III studies. So I think we can say it was selected on some characteristics, but the population in reality behaved very similar.

Anthony El-Khoueiry, MD: Yes, absolutely.

Richard Finn, MD: And so now you say for the patient who might not make it to third-line, you go PD-1 up front because of that 15%, you hope you’re going to be a good responder. So, say you start PD-1 and they have stable disease. How do you monitor them? And what decision do you make to go then to your third-line TKI?

Anthony El-Khoueiry, MD: Right. So like every other cancer therapy we use, as I tell patients, stable disease is considered success in cancer. Now, that’s a large group with anti—PD-1 agents. Hopefully with the phase III data, we’ll understand more what the impact of stable disease is. But, nonetheless, as long as they have stable disease, we monitor them with regular imaging. I tend to not rely heavily on AFP [alpha-fetoprotein], rather on imaging. And as is accepted with I-O agents in other diseases, unless the patient is clinically progressing, we tend not to switch out of the anti–PD-1 agent at the first progression. Meaning if there is a minor radiologic progression, patients doing well clinically, generally I would continue and confirm this progression on a subsequent scan 6 to 8 weeks later.

Richard Finn, MD: So is that something unique to I-O or do you do that with the TKIs as well?

Anthony El-Khoueiry, MD: We’ve done this with sorafenib, as you alluded to earlier. We waited for clinical, not just radiologic, progression. But, again, because of the availability of multiple agents now, we tend to be a bit stricter at calling progression. But with the I-O agents, there is always this concept that especially early in the disease, you may see something that looks like a progression, but the patient subsequently has stabilization or even a response. And, in a nutshell, when we looked at that in the CheckMate-040 study, that was about 19% to 20% of patients who subsequently had restabilization or response again. And they tended to do well.

Transcript Edited for Clarity