Hepatocellular Carcinoma: Practical Implications of Emerging Data - Episode 14

Safety Profiles of Novel Therapies for HCC

, , , , ,


Richard S. Finn, MD: We’ve talked about a lot of the efficacy discussion about choosing something beyond progression frontline. Katie, you had commented about the use of regorafenib in the inclusion for RESORCE was to tolerate prior sorafenib. Part of that was to mitigate, I think, the adverse-effect profile of regorafenib and sorafenib, given their overlap. Amit, could you give us some idea about how adverse effects of these drugs influence your choice of treatment, including frontline? We talked about bleeding issues, but hypertension, proteinuria—some of these are class effects. But every molecule has some unique aspects.

Amit Singal, MD: This goes back to the question that Tony asked. When you put this all in 1 box, you’re going to have these options you have to choose. As we’ve seen, there are no head-to-head data that suggest that 1 TKI [tyrosine kinase inhibitor] has better efficacy than the other. As you said, there’s a class effect. Some of these AEs are similar, but there are small differences in terms of the proportion of patients who experience any specific AE. For example, comparing sorafenib and lenvatinib, you basically see that sorafenib gives you a higher proportion of hand-foot-skin reaction, whereas with lenvatinib, you’re more likely to have proteinuria, hypertension, and maybe constitutional symptoms like anorexia. Depending on how your patient is doing at that time, when they come to the second-line therapy, some of these AEs may actually help you in terms of distinguishing which therapy you’re going to choose. We’ve already referenced the fact that regorafenib and cabozantinib have similar AEs. But with regorafenib, it’s really with the idea that they were well selected. As Katie mentioned, they had to be on sorafenib for 20 of 28 days to get into regorafenib. It’s unknown if you went directly into regorafenib if you would have had more or worse AEs.

From my standpoint, these subtle differences in terms of AEs are maybe one way you can choose which TKI you’re going to use in the second-line setting and potentially even what AEs they come out of atezolizumab-bevacizumab with. If you use atezolizumab-bevacizumab, and they have some proteinuria, you may be less likely to use lenvatinib in that setting, given that you can have proteinuria with lenvatinib. If they come out and have poorly controlled hypertension, it’s the same kind of decision-making in terms of knowing that lenvatinib is more likely to give you hypertension. Outside of us actually hitting the jackpot and starting to have some biomarkers to help select this, unfortunately, we’re continuing to use these more gross and obtuse measures in terms of trying to decide among these.

As Tony mentioned, I’m hoping, knock on wood, that this is a short-term problem. These other new trials of combo therapies are promising. Some of them have finished enrollment, and we should see data. I do think the future is going to be combos. It’s going to be combo 1 and then sequencing to combo 2. It doesn’t take away how we choose. But that the idea of having to choose between TKI monotherapy 1 and TKI monotherapy 2, I think that discussion goes away.

Richard S. Finn, MD: Katie, what’s your take on that discussion?

R. Kate Kelley, MD: It’s good to have these choices. Just as we practice the art of oncology in other tumor types, like in colorectal cancer, a patient with an ileostomy and a lot of diarrhea will choose FOLFOX [5-fluroracil, leucovorin calcium, oxaliplatin] over FOLFIRI [5-fluroracil, leucovorin calcium, irinotecan] in the first-line. In breast cancer, a patient with neuropathy may guide our first-line choice of therapy. The degree of hypertension and whether a patient has had immediate hand-foot syndrome or prohibitive hand-foot syndrome on their prior therapy should be the kinds of things we use to guide some of our decision-making, and it’s really great to have the choices that allow us to tailor by toxicity and comorbidity. Of course, in addition to the efficacy outcomes, which are paramount, having these choices really does help us individualize therapies.

Richard S. Finn, MD: Anthony, the toxicity from CheckMate040 with CTLA4 inhibition and PD-1 inhibition includes 50% use of steroids. That’s a high number, I think. Many of us think about that and its implication for using that regimen in second-line compared with a TKI. Certainly, we’re not seeing response rates of 32% with TKIs. That’s a very impressive number. You have a fair amount of experience with this combo. What’s your thought about it? Are these protocol-dictated use of steroids? Or are these real toxicities?

Anthony El-Khoueiry, MD: Certainly some of it is protocol dictated. When we quote 50%, this is using 40 mg of prednisone or higher. So 40 mg would be less than the 1 mg/kg standard that we would start with. This is almost all steroid usage. If you look at that, most of it was driven by skin toxicity like dermatitis, about 20% were hepatitis-type events, and then colitis. There was some pneumonitis as well, but the numbers get pretty small for that. What’s important is that over 90% of these events were reversible. Actually, a majority of these patients with immune-mediated events were rechallenged and were able to get treated again. It’s a tough regimen. There needs to be caution. There is a high need for steroid usage, obviously, in some patients. But it’s a manageable toxicity in the sense that if you are careful, you know when to do it—when to stop the regimen, use steroids—you can have most patients recover and even reinitiate therapy. It has to be a regimen that’s reserved for the right patients, so good performance status, Child-Pugh A. As you alluded to, what’s promising there is not just the response rate. It’s this intriguing OS in second-line of over 20 months.

Richard S. Finn, MD:Yeah, wow.

Anthony El-Khoueiry, MD: Right? In the right patient, this could be a beneficial regimen. Again, how does it fit post–atezolizumab-bevacizumab? No one knows. This is something we already discussed.

Richard S. Finn, MD: Reena, from the hepatology standpoint, a lot of the drugs we use cause proteinuria. In colon cancer, with bevacizumab, proteinuria never got us too concerned. Do we need to think about it differently—the proteinuria we see, either from the TKIs or bevacizumab or ramucirumab, for that matter—in patients with underlying cirrhosis, or not really?

Reena Salgia, MD: Not really, Rich, at the Child-Pugh A stage. Certainly as patients start to progress in their level of decompensation, we worry about that more with the renal dysfunction they may experience anyway. But generally, no. As long as we’re monitoring it—many have started to become accustomed to that with lenvatinib in this space, already starting to check for proteinuria. If we apply it to a more decompensated patient or a Child-Pugh B7, for example, going forward, if data allow, then we would need to be monitoring for proteinuria more closely in the cirrhotic patient.

Transcript Edited for Clarity