Ghassan K. Abou-Alfa, MD: Amit, checkpoint inhibitors—20%. We just heard that the drug called lenvatinib has a response rate that is close to 41%. Is there any preference as to which patient gets what? How do you explain all of that?

Amit Singal, MD: What’s exciting about immunotherapy and the Checkmate-040 data are the response rates that we see. You see response rates in the RESORCE trial. You’ve seen them in the REFLECT trial. You see response rates in the CheckMate-040 trial. You’ve seen this with TKI therapies, and you see it with immunotherapy. The difference is the durability of that response. With TKI therapies, the response, while you have it, tends to be short-lived. The fascinating and most exciting thing with CheckMate-040, at least in my mind, is that these responses are durable. When you take a look at this, it’s several months—even greater than 12 months in many patients. When you have that durability of response, there’s a higher likelihood that it’s going to translate into meaningful survival differences. And so, I think that’s the biggest thing that’s exciting here.

Ghassan K. Abou-Alfa, MD: You bring up a very important point. Actually, for all of us who are now familiarizing ourselves with different ways of looking at data, we used to see numbers, or response rates. Now we’re seeing waterfall plots that show how much response and nonresponse there is. We’re seeing swimmer plots—how long the response would stay. This is really what you were referring to, in regard to checkpoint inhibitors. Mark, let’s assume that you have patients on nivolumab who are getting a response. They stay on the therapy for a long period of time. Would you stop at some point? What does the data show?

Mark H. O’Hara, MD: That’s the question in I-O therapy, in pretty much every type of cancer—for the I-O therapies, how long is too long? In general, we continue it as long as they’re responding because it’s hard to rationalize or say, “Hey, let’s stop this therapy. It seems to be working.” You don’t want to kick yourself at the end if it ends up growing after that. We don’t know the answer.

If they’re tolerating it, I stick to it. If they’re having side effects that concern me, whether it be immune-mediated or a side effect like fatigue, joint aches, or something that’s really bothersome from a quality of life perspective, I will stop it, pointing to other data and other cancers that say we may still be able to keep a durable response.

Ghassan K. Abou-Alfa, MD: I hear you, and you’re right. Actually, this is being looked at in other diseases. Manish, along that line, what’s your hands-on experience regarding tolerability and adverse events of checkpoint inhibitors?

Manish R. Sharma, MD: We have experience in many diseases, right? My experience with nivolumab and other checkpoint inhibitors was in other disease before it was in HCC, because HCC came later in the game. My experience has been that these are generally very well-tolerated therapies. The number I tell patients, which I think is accurate, is that about 90% of patients will tolerate it with minimal or no side effects. About 10% of patients will have more significant side effects. Those tend to be things like fatigue, or nausea, or diarrhea. We, of course, get excited about immune-related adverse events, and those are very important to recognize. That’s the big thing. With checkpoint inhibitors being in such common use, I think oncologists have become much more comfortable in managing them, stopping the drug, getting high-dose steroids, etc.

In HCC, my experience has been that the tolerability is not fundamentally any different than it is in any other disease. We are always concerned that these are patients with liver disease. Or, they are at higher risk for immune-related hepatitis because the liver is sort of inflamed from the hepatitis or whatever else. We really haven’t seen higher toxicity rates with the checkpoint inhibitors in HCC compared to any other disease, which has been exciting. You’re combining what Amit was saying, about the durability of response, with a really well-tolerated therapy. From a patient’s perspective, that’s very attractive.

Ghassan K. Abou-Alfa, MD: I totally agree. One of the teaching points that I hear every now and then is, like everything, the side effects with the checkpoint inhibitors, until proven otherwise, you can tell how the immune response might be… This is relatively novel to us—in regard to their practice in oncology.

Along that line, we hear about other checkpoint inhibitors, and there’s quite a bit of data coming out. Ruth, can you tell us about pembrolizumab?

A. Ruth He, MD, PhD: Pembrolizumab has been tested in the phase II KEYNOTE-224 study. In comparison with placebo, it has shown a response. The response rate is very similar to nivolumab. I think this is because the mechanism of action is the same. And also, the response can be durable. Because of that data, pembrolizumab has now been tested in the second-line setting. The phase III study has completed enrollment, and we’re waiting for the survival data.

And then, there are additional immune checkpoint inhibitors, such as tremelimumab, durvalumab, and atezolizumab. They’re all in clinical trials, being tested for liver cancer.

Ghassan K. Abou-Alfa, MD: Mark, we have a lot of checkpoint inhibitors, and there is definitely a lot of promise there. What are we going to ultimately do with them? Are we going to mix them together? Are we going to mix them with something else?

Mark W. Karwal, MD: So far, my opinion is that if you mix a PD-L1 drug with a CTLA-4 drug, you get more toxicity. In liver cancer, you don’t get more benefit. Maybe you get more benefit in melanoma, but not in liver cancer.

Ghassan K. Abou-Alfa, MD: Ironically, I might disagree. We have had some experience with durvalumab plus tremelimumab. If anything, we are seeing incredible responses. And, if anything, the toxicity did not necessarily defer from what we know about the single agents. So, I would like to refresh my mind with other perspectives there. We know that there is a huge, ongoing trial in HCC that is looking at durvalumab plus tremelimumab, versus another dose of durvalumab plus tremelimumab, versus durvalumab, versus sorafenib. This is a 1200-patient study. It’s going to be very important in determining the role of anti—PD-1 plus anti–CTLA-4. It’s important to remember the biology—how these things are connected. All of this is about connectivity between the T cell and the tumor, and we are trying to break every bridge we can to prevent that T cell from attacking the tumor. A break of 2 bridges will make it much easier than if it was 1, per se.

There is quite a bit going on, in that regard. Right now, we know that we can potentially use checkpoint inhibitors in HCC, pending further data from the durvalumab versus sorafenib study.

Transcript Edited for Clarity.