Emerging Treatment Patterns for Hepatocellular Carcinoma - Episode 13

Opportunities for Checkpoint Inhibitors in Advanced HCC


Ghassan K. Abou-Alfa, MD: Amit, back to you. I would like to dwell on this further. Let’s talk about another drug and then, maybe, put this in perspective. There’s also pembrolizumab, which is, if anything, quite evolving in HCC [hepatocellular carcinoma]. Tell us about KEYNOTE-224, and then we’ll talk more about nivolumab.

Amit Singal, MD: KEYNOTE-224 also evaluates immunotherapy but with pembrolizumab in the second-line setting. This is a phase II study, just like CheckMate-040. Similar to CheckMate-040, it shows very promising durable response rates that are right around the same mark—around 15%. So if anything, with 2 phase II studies, it reinforces what you can see in uncontrolled settings, in terms of immunotherapy having very good response rates. Like CheckMate-040, where we’re waiting for data. With KEYNOTE-224, we’re also waiting for the phase III data from KEYNOTE-240. So I think that we see very promising phase II data in uncontrolled settings and we’re waiting for larger phase III data in a comparative fashion.

Ghassan K. Abou-Alfa, MD: Absolutely. If anything, KEYNOTE-224 is going to be more pertinent to the second-line setting, whereas the CheckMate-459 data is more useful in the first-line setting. So it’s good to ask this question even though some people in the beginning are like, “Wait, you can’t ask that question.” Theoretically, for argument’s sake, let’s say CheckMate-459 does not really show an improvement in outcome for nivolumab versus sorafenib. What does that mean? How do you interpret that?

Peter Galle, MD, PhD: It would add another level of complexity to our discussion. The expectation is clear. We believe from what we have seen that it’s a positive trial, but your point, Ghassan, is extremely well taken. There’s a lot happening after the patient is off the drug, and that might heavily influence whether it’s crossover or other trials. So the second-line and even third-line treatment options, which are there and haven’t been there in the past, will certainly impact overall survival. Postprogression pattern is extremely relevant. It might be required in order to understand what has been happening, whether it’s a positive or negative trial.

Nevertheless, I’m very optimistic. When we reconsidered the situation at the ASCO [American Society of Clinical Oncology] Annual Meeting this year in Chicago, it had been and continues to be such an active field. You have been approached by everybody about not only doublets but triplets. And so the combinations are already there. You know that the HIMALAYA trial and the IMbrave150 trial are both testing different combinations. Whether or not single-agent matures into an effective therapy, there will be more around, and this field is already moving very, very fast.

Ghassan K. Abou-Alfa, MD: I totally agree. If anything, again, we bring up the question only for argument’s sake. I don’t think any of us are questioning the value and the efficacy of the checkpoint inhibitors. But I think what you are bringing up is exactly what Riad brought up earlier. What’s happened to the patients during their whole sequence of therapy? Could it be that they received checkpoint inhibitors after sorafenib on that trial of sorafenib versus nivolumab anyway? It is possible. If anything, it raises the question of: What is the timing of the therapies, and will it matter? It appears to me, based on our conclusion from the discussion on TKIs [tyrosine kinase inhibitors], that it might matter.

On the other hand, with checkpoint inhibitors it’s just a matter of where they come into play. This is fascinating because there are a lot of opportunities after nivolumab or sorafenib that patients might get. One is for pembrolizumab versus placebo. They already received something in the first line, and there’s more second line…. It will be fascinating to see. We’re not really aware of any data for any of those studies, but it will be nice to bring those discussions into play, to ultimately see what the endpoints will be, so we can figure out how to interpret those endpoint questions that we brought up a bit earlier.

Transcript Edited for Clarity