Ghassan K. Abou-Alfa, MD, MBA: Stay tuned, because it’s not like in the second line it’s any better. It’s as complex and interesting, and that’s great news that now we’re talk about second-line therapy. There have been lots of events in regard to second-line therapy, and probably almost all of them already have been FDA approved. Let me start again with Catherine. We’re going to play it in 2 scenarios, and all of us will be involved in this. For a patient who got lenvatinib as a first-line therapy, why continue in second line?

Catherine T. Frenette, MD: We don’t really have any data at this point of what’s due in second line after lenvatinib. All the second-line studies were approved after sorafenib because that was the standard of care at the time. I think I would choose my second line depending on how the patient tolerated the lenvatinib. Did they have any response, and how good of a response?

You can make an argument for the TKIs [tyrosine kinase inhibitors], multiple other TKIs like cabozantinib, like regorafenib, that they are being used after sorafenib tolerably. You could also make an argument for I/O [immuno-oncology]. You could also go and say, “What about using ATEZO-BEV [atezolizumab-bevacizumab] as second line now in this patient.” We don’t have a lot of data to guide us, unfortunately. It’s going to be a little bit like the Wild West until some of the data come out.

Ghassan K. Abou-Alfa, MD, MBA: That’s a good introduction, and you probably got the taste of what Dr Frenette said in regard to what will be the second line, even though there could be some notions that would take us 1 way or the other. I’ll carry on with the same question, and I’ll ask Dr El-Khoueiry. If lenvatinib is first line, what will be your second line?

Anthony B. El-Khoueiry, MD: Extrapolating that, you know lenvatinib and sorafenib are both multitargeted tyrosine kinase inhibitors. So we extrapolate a lot of medicines. Even though the studies were post sorafenib, we extrapolate and use those agents post lenvatinib. As Catherine mentioned, we could use cabozantinib. Regorafenib—that study was particularly for patients who tolerated sorafenib. It may be used less post lenvatinib. We have ramucirumab for patients with an AFP [alpha-fetoprotein] above 400 ng/mL. Now with the I/O agents, I’ll spend a minute on that.

Ghassan K. Abou-Alfa, MD, MBA: I heard about something called molecular genetics. But I also heard about something called immunotherapy.

Anthony B. El-Khoueiry, MD: Yeah. In that space, meaning the second-line space, we have at least 2 sets of data from phase I/II studies with nivolumab and pembrolizumab that resulted in the accelerated approval of those agents in that space. Now that accelerated approval was based on a response rate ranging between 15% and 20% for both studies, plus another third agent called pembrolizumab, showing consistent data.

In addition to the response rate that was at that time novel and promising, it was the durability of the responses. We’re talking the responses being durable above, usually beyond 1 year. Actually, the CheckMate040 has the longest follow-up median durability of response at 17 to 19 months. And others showed that patients who responded did very well as far as survival. This promise led to these agents being used, both sorafenib and post lenvatinib, in second line.

The challenge came when KEYNOTE-240, which is a phase III trial, was conducted in the second-line space post sorafenib, and randomized patients to pembrolizumab versus placebo.

When we first look at this trial, we see a median OS [overall survival] of almost 14 months with pembrolizumab, a median OS of 10, 11 months with placebo. And a P value of .02-something. It looks positive on the surface, right? And with a clinical benefit. However, because that study had coprimary end points and multiple interim looks at the data, that P value needed to be less than .017.

Statistically, that study was not positive. But we do believe that there’s a clinical benefit with a clean separation of the curve and a difference in the median OS in that study. It poses a bit of a challenge where the single agent, PD-1 [programmed cell death protein 1] agents will fit. Of course, if we’re going to use PD-1 combinations or PD-L1 [programmed death-ligand 1] combinations in first line, then using single agent PD-1 or PD-L1 in second line would not make sense. But for those patients who got TKIs, it remains a reasonable option.

Ghassan K. Abou-Alfa, MD, MBA: I have to admit, you’re confusing me even further because—and Ahmed, please help me here—I agree with both thoughts. I’m hearing the same thing, that really we don’t know what to do. On top of that I’m hearing about a statistically positive study but clinically not positive.

Ahmed Kaseb, MD: The other way.

Ghassan K. Abou-Alfa, MD, MBA: Or the other way around. But as a positive study, do you depend on the use of pembrolizumab, for example, based on the phase II data, which actually led to the conditional approval? Or do you make an argument based on the phase III trial that was rather negative? And this is really what you’re going to call it, a negative study. How do you put all this together?

Ahmed Kaseb, MD: This is a question we face every day in clinic. The simple way of discussing this with patients how the TKIs, performed—if they had horrible adverse effects, the TKIs had hand-foot syndrome, and so on. Do you want to take a break in which case we can try some immunotherapy and see if they can handle it? In the meantime, you also see those patients who did very well with TKIs—you know, a year and a half, tolerating it well. Especially with sorafenib, as you know, as Dr Frenette said. Then there is a case to make here for regorafenib, for example.

It really comes back to our initial thought about personalizing therapy approaches for those patients, which one would carry more risk than the other. Until we get solid data from clinical trials about sequencing, we’re not going to know. As we all know, we’re never going to see those studies because the field is moving so fast. Immunotherapy now moved up to frontline. Now the challenge is going to be, as Dr El-Khoueiry was saying, being exposed to anti-VEGF and anti—PD-1, what to do next.

Transcript Edited for Clarity