Arndt Vogel, MD: Other novel approaches in HCC are really difficult at the moment. We had great hopes for the tivantinib trial, which was recently published. The trial was based on a phase II trial that indicated that c-MET in HCC could be a negative prognostic marker, and that patients who have c-MET-high tumors would derive a benefit from treatment with tivantinib. Based on the phase II data, a phase III study has been initiated and has now been reported. Unfortunately, the study was completely negative and, interestingly, in 2 respects.

First of all, it has the longest median overall survival that has been ever reported in a second-line trial, which indicated that c-MET is not such a negative prognostic marker since all these patients had a really long median overall survival. And unfortunately, it’s not only a negative prognostic marker, it’s also not a predictive marker because the patients did not derive any benefit from tivantinib. So, biomarker-driven trials are really difficult at the moment in HCC. We have 1 biomarker, AFP, which we really do not understand. But based on the REACH-1 study, we have the impression that patients would derive a benefit from ramucirumab if they have tumors with a high AFP level. And based on this first phase III study, and new phase III study that has been started, the REACH-2 study only includes patients with high AFP. And we have to see whether AFP is really a reasonable biomarker in HCC.

We have a better understanding of genetic lesions in HCC, but so far, there are no really promising candidates that would lead to systemic therapies in the very near future. So, I think, at the moment, the most likely next steps will include combination therapies. And here, the combination of immunotherapies with local therapies or other systemic therapies, such as lenvatinib or regorafenib, are most likely the most promising candidates.

Oliver Waidmann, MD: Biomarkers are really a good question in HCC. We have just learned from the tivantinib trial that if you have a biomarker-driven study, it can also be negative. And with all the drugs that have been shown to be effective in HCC, namely sorafenib, regorafenib, lenvatinib, and even nivolumab, we didn’t have really good biomarkers to predict the treatment. All 3 drugs—sorafenib, regorafenib, and lenvatinib—all the TKIs, which have a broad mode of action that’s not specific, and also the immunotherapeutics, such as nivolumab, we don’t have a good biomarker because we found that PD-L1 expression—which is quite often used to predict treatment efficacy in patients with immunotherapy—was not active. There was no big difference between the patients who get PD-1 high or PD-L1 low tumors. So, I think it’s really a good question, but a difficult question to answer. I think there are some interesting data if you could target stemness—so the BBI608 in colorectal cancer and other diseases, which we also tested in HCC. So, stemness is a way and I also think Wnt signaling will be interesting. But I think at this time, we don’t really know what type of therapy we will see.

Transcript Edited for Clarity Brought to you in part by Eisai