Ghassan K. Abou-Alfa, MD: Dr Salem, you and I were heavily involved in regard to the development. I remember one thing that you always were concerned in regard to was the potential adverse events, especially when you brought in the sorafenib. Tell us about your experience and what fears patients have about sorafenib.

Riad Salem, MD, MBA: From an interventional radiology standpoint, after now a decade or 15 years of seeing sorafenib, there are patients who suffer from hand-foot syndrome. We have seen, on rare occasion, in the interventional suite, some nonspecific vascular effects that theoretically aren’t supposed to happen, but some level of vasal constriction happens. This is a real-life tumor board, so real-life scenarios exist. We have patients, for example, who are stable or maybe slightly progressing on sorafenib. They have 1 lesion that for some reason is not responding or progresses as an isolated area of progression. Everything else is stable. We do some angiograms or we do an ablation and we see the vascularity change, and we’ve seen some of that reversed relatively quickly. This is anecdotal, and it’s a rare scenario, but it’s certainly something that we start to look at and we start to investigate. I think now, after 10 years, we’re starting to have more experience with real-world settings of patients on sorafenib or on lenvatinib, where we see some sort of minor abnormality. It’s not supposed to happen, but some weird or strange effects have been observed. What we’ve been able to do in all scenarios, however, is ultimately get the patients treated with whatever we’re going to do in embolization, ablation, etc. That’s why I bring it back to this multidisciplinary discussion. Here’s what we see. We discuss it with our hepatologist, with our oncologist and we proceed.

Ghassan K. Abou-Alfa, MD: They listened to Dr Yarchoan’s recommendation. The patient was given 12 mg of lenvatinib and, understandably, the potential adverse effects did occur. There was hypertension that led to a dose reduction to 8 mg, which is very easy and straightforward to do. The patient was referred to therapy, which I’m surprised was done, as we’re hearing that some subtle weight loss and some poor appetite can occur. Nonetheless there was imaging after 16 weeks, a short partial response, which does not surprise me; we’ve seen that with lenvatinib. After 8 months of initially taking therapy, however, there was some progression of disease. That’s what we have so far with regard to that patient. To just go over a little bit of the data that we just brought in, the first-ever study of a TKI [tyrosine kinase inhibitor] vs placebo, sorafenib vs placebo, and the outcome was looking toward overall survival. We must remember that this study was very positive, and we were delighted with 10.7-month median survival for sorafenib vs 7.9-month median survival for placebo. This was clinically and statistically significant. We have to remember, though, a very important component. This impartiality that sorafenib had will never happen again because, if anything, there was nothing to be given for a patient before sorafenib and nothing to be given after sorafenib, and as such, this survival that we are seeing here is purely sorafenib. It’s important because as we evolve for other therapies—and we have now second- and third-line treatment—the contribution of the first, second, and third line is adding onto the survival, and this why this 10-month median survival will never be seen again. However, we’re going to see a longer survival because of the added contribution of other therapies. For that reason, when the lenvatinib study did occur, the REFLECT study [NCT01761266], it was randomizing patients by eligibility criteria. They are the standard, as you heard from all of us; this includes unresectable disease, advanced disease, spread out disease, good performance status, appropriate liver functionality. And if anything, patients were randomized to lenvatinib vs sorafenib and looking at overall survival as a primary end point. Interestingly, this study was designed within noninferiority, and the reason is because the investigators thought, if this is how much the TKI can do, we are not going to change it anymore but we’re going to see that this is still noninferior to sorafenib, and there’s a specific design for that. At the same time, we’ll look into other competency as we just heard from Dr Yarchoan in regard to the different end points. Dr Singal, is this crazy what I just spoke about: having a noninferiority study as if it’s the cheat sheet? What are your thoughts on that?

Amit Singal, MD: No, I think, Ghassan, a noninferiority study is a well-accepted trial design. It has to be done a priori. I think the difference is the REFLECT study did this coming in, vs the quote unquote spin that you would see on some negative trials. When you take a look at this this, this is a gigantic trial. You have nearly 1000 patients in this trial, so you have to adequately power that study for noninferiority to prove that it’s similar. Once again. I understand now we’ve done better than sorafenib, but when the REFLECT trial came out, we joked sorafenib was the Muhammad Ali of hepatocellular carcinoma systemic therapies. It was 10 years in the running where nothing beat it, so it was a very good therapy and is a good therapy in select patients. I think a priori designing, and this as a noninferiority study, is not a cheat. This is an accepted trial design where we can then have alternatives, and by doing so, it met its primary end point in noninferiority as you mentioned, and then it allows us to see significant improvements in some of the secondary end points, as I’m sure you’ll discuss.

Ghassan K. Abou-Alfa, MD: Perfect, that’s great. Thanks so much, Amit, for explaining what noninferiority is, which is not just going through the back door. This is very well designed. It has its own requirements statistically and I consider it to be a smart approach on the hand of the investigators for that purpose, but interestingly, Dr Yarchoan you really hit on very important things. First, let me show the data: 13.6 months for Lenvatinib, 12.3 months for sorafenib, so no difference at all. It did meet the primary end point of noninferiority. As we said, we don’t expect the 10 months, but rather we’re going to expect more because of other therapies that could have happened. However, Mark, you mentioned other secondary end points and where they could be a benefit, and I’m showing you the first one which is PFS [progression-free survival]. Please describe to us what you are seeing here and your reflection on it.

Mark Yarchoan, MD: Lenvatinib clearly improved PFS vs sorafenib; patients were able to stay on the drug about twice as long. Obviously, that didn’t translate as discussed into longer overall survival, but I do think this matters for patients, and it’s a reason that I think lenvatinib can be an important option to include in our discussions.

Ghassan K. Abou-Alfa, MD: Thanks so much. I’ll revise an important point. What we just heard is that the PFS for lenvatinib compared with sorafenib. It is almost 2.7 times longer for lenvatinib. It has some biologic perspective to it because of the anti-FGF [fibroblast growth factor] component of the lenvatinib where the IC-50 already is way more robust than this for sorafenib. It’s like having a race track, as we just heard from Dr Singal, for the survival which is exactly a defined length. The investigators knew what the race track length is but it does not mean, as Dr Yarchoan is telling us, you can have a faster car on the race track, which is PFS. This is exactly what we are seeing over here.

Transcript edited for clarity.