Key Takeaways in the Treatment of HCC - Episode 4

Key Takeaway 4: Second-Line Regorafenib After TKI in HCC


Richard Finn, MD: After the approval of sorafenib in 2007, there were 2 large unmet needs; 1) another drug in frontline, but even more, what do we do for patients who progress on sorafenib or get no benefit from sorafenib? For many years, this was felt to be an easy spot for drug development, because the control arm was placebo. There was nothing approved. However, as we saw in frontline, it wasn’t so easy, and there were numerous failures initially with ramucirumab, everolimus, and brivanib, until the RESORCE study, which was the first positive study in liver cancer in about a decade.

This study evaluated regorafenib, which is a multikinase inhibitor, similar to sorafenib, but with some differences. It has a slightly broader kinase profile and is a little more potent. This study accrued patients who were on sorafenib for a minimum period, so they were not sorafenib-intolerant. They had tolerated a minimal dose of sorafenib for a minimum period of time. Then, progression was carefully collected as a mRECIST [modified response evaluation criteria in solid tumors] type of progression, so either a new lesion or increase in tumor burden by more than 20%. At progression, these patients were randomized in a blinded fashion to 160 mg regorafenib, 3 weeks on and 1 week off, versus placebo.

There was not a lot of optimism for this study, because we had 3 other failures before it, as well as the feeling that regorafenib was somewhat similar to sorafenib, even though in a single-arm phase II study before RESORCE, we did see single-agent activity in this drug. We know this drug is a little different than sorafenib. REGO [regorafenib] was approved in colon cancer where sorafenib did not have activity, and we also know that regorafenib is also approved in GI [gastrointestinal] stromal tumors. Sorafenib doesn’t have activity there.

Needless to say, this was a positive study. It improved survival by about 3 months over placebo. Regorafenib translated to a hazard ratio in the 0.65 range, so it was more than a 30% decrease in the risk of death with regorafenib in this population. There were also concerns about the tolerability of regorafenib, given that in the colon cancer population, regorafenib has toxicities that can make it challenging. In the liver cancer population, partially because these patients had tolerated prior sorafenib, the toxicity with regorafenib was not too bad. Patients tolerated it. Many patients needed dose delays or dose reduction, but the toxicity was not as bad as expected. In my experience, it seems even better-tolerated than in the colorectal population.

I think that comes because of our experience with sorafenib and TKIs [tyrosine kinase inhibitors], and the fact that the colorectal population patients probably have had a lot more prior treatments and chemotherapy over a long period of time. Whereas, in the liver cancer population, it’s sorafenib and regorafenib. The sequence we’re seeing is improving survival. We had published an analysis of survival for patients on the RESORCE trial from the start of sorafenib to death, and we showed that that survival is about 24 months for those patients. Again, this is a subset of patients from clinical trial data, but 24 months is not a number that we’ve seen in frontline or second-line liver cancer. This relates to the challenges now in drug development in frontline. We’re extending survival using the sequence of these new drugs. Regorafenib also met other endpoints for improving progression-free survival and time to progression. Response rate is a little higher than with sorafenib, but not significantly, and most of the benefit is probably from slowing progression.

Another drug approved in the second-line space is cabozantinib. Cabozantinib is a multikinase inhibitor, although it’s a little different from the others we’ve talked about in the liver cancer space in that it not only hits the VEGF [vascular endothelial growth factor] receptor, but 2 other kinases, AXL and c-MET. C-MET is the hepatocyte growth factor receptor and has been pursued as a target in liver cancer in and of itself. It’s a receptor tyrosine kinase. That’s very important in the development of hepatocytes, and this drug hits c-MET and AXL, another oncogenic protein. This drug was looked at in a second-line setting versus placebo. It was launched after the RESORCE study and subsequently read out sometime afterward, but showed a very similar result.

What was somewhat different about the design of CELESTIAL was that it also took patients who were sorafenib-intolerant. That was different than the RESORCE study, where they had to tolerate sorafenib. About 25% of patients in that study also had 2 agents, so they had a third-line population. That population had been treated with sorafenib, another agent—and there were various drugs used—and then cabozantinib. In the intent-to-treat population, which was the entire true second-line and third-line population, survival was improved by about 2 to 3 months, which was not too dissimilar from the regorafenib data. The hazard ratio was around 0.78, which is a 22% decrease in the risk of death. Again, like the other kinase inhibitors, it did this without necessarily improving response rates. It slowed progression.

If we look at the survival for the truly second-line population and exclude those third-line patients, the hazard ratio drops down to the low 0.7 range. Toxicity with this drug is also similar to sorafenib and regorafenib. It does have fatigue, GI toxicity, some hand-foot skin reaction, but generally, it’s tolerable and is another option. Now, we’re faced with choices in liver cancer, which is a good thing. It’s a good thing for patients and for us.

Patients progress on frontline therapy, whether it’s sorafenib or lenvatinib, and then they come to second-line. All of these drugs have been studied after prior sorafenib, but I think we make the assumption that they’re just active in second-line liver cancer. Whether they have lenvatinib frontline, we would still consider either regorafenib or cabozantinib for this population of patients.

I think it’s a toss-up which drug you use. For patients who have tolerated sorafenib and did well, it’s very logical and evidence-based to go to regorafenib in that sequence, and that 2-year survival data are very provocative. For patients who might have more problems with hypertension or had trouble with sorafenib, the data would support using cabozantinib over that. Ultimately our goal is to sequence things as much as possible, and that means following patients. There are data using cabozantinib in the third-line setting, so conceivably, a sequence of sorafenib, regorafenib, and cabozantinib is evidence-based. The other thing to keep in mind is, if patients have an elevated AFP [alpha-fetoprotein], ramucirumab was also approved in the second-line setting, which we’ll probably talk about shortly.

Transcript Edited for Clarity