Peter Galle, MD: We will, in many countries, at least in those where IMbrave150 results in approval and reimbursement, take over ATEZO-BEVA [atezolizumab-bevacizumab] as first-line therapy, and that requires a new assessment of the other therapeutic options in second and third lines. We will not have the chance to have sufficiently large trials for all these sequences to get an evidence base for the different lines of therapy. One of the rationales to make a decision of what treatment to choose is the specific spectrum of patients who had been treated in the available trial. It can be the adverse-effect profile, and of course it can be the rationale, the mode of action. For example, I would consider it useful to have in 1 line I/O [immuno-oncology]–based therapy and in another line a TKI [tyrosine kinase inhibitor]–based therapy to offer patients the chance to get these different modes of action as an individual patient.

Kate Kelley, MD: Our data for agents in the second line are really predicated on the receipt of first-line sorafenib so far. For example, when we think of regorafenib, that approval was based on the phase 3 RESORCE trial, which required patients to have received sorafenib for 20 of 28 preceding days, at a dosage of at least 400 mg total per day. I just had to understand how to use regorafenib if patients didn’t receive first-line sorafenib. For now, that is a predicament, and it will be interesting to see if there will be future second-line studies on the horizon to better characterize response, toxicity, and overall benefit in a different sequence of receipt.

We know that for antiangiogenic agents in other tumor types, that continuation beyond progression of a first antiangiogenic agent has shown benefit, and in HCC [hepatocellular carcinoma] that has been the case as well. For example, with drugs like cabozantinib, there has been benefit after first-line sorafenib, as well as any second-line other agent. We know that cabozantinib achieved benefit in a population that included third-line patients. I think it is intuitive to continue the use of cabozantinib after a combination regimen based on the way the CELESTIAL second-line trial was conducted in the population it included.

As we start looking at downstream immunotherapies after a first-line PD-L1 inhibitor in the case of atezolizumab plus bevacizumab, for example, that’s where we really don’t understand enough about the mechanism of the primary resistance as well as acquired or secondary resistance to immune checkpoint inhibition in HCC. It will be very important to try to perform studies in the second- and third-line space, along with translational end points, to try to understand which patients are likely to respond to a different immune checkpoint inhibitor approach versus an antiangiogenic approach in the second or third line.

It’s important here to add what a landmark this is, to be discussing later-line therapies as a reality in this disease. A decade ago, we didn’t think second-line therapy was a reality, and now we have not only a handful of second-line agents but also the likelihood of third-line studies on the horizon.

Peter Galle, MD: We do have clearly defined labels for those substances. Sorafenib was first line, and it was very clear regorafenib was approved in the second line following sorafenib. It was actually required that sorafenib had been tolerated for some period of time and approved for patients in the second line following sorafenib. Of course, this does not mean that second-line cabozantinib after ATEZO [atezolizumab] plus BEVA [bevacizumab] does not work, but we have to accept we have neither the evidence nor the label for it. That is an interesting discussion, and it is probably very country specific whether these substances are accepted or not. For example, in Spain, we have very limited second-line options today, even though the evidence is there. You can tell that different countries here come down to different decisions.

Masatoshi Kudo, MD, PhD: The agent used after sorafenib was lenvatinib. After lenvatinib, unfortunately, cabozantinib is not approved in Japan because Japan did not join the phase 3 trial. But currently the phase 2 trial finished and was successful. We will soon see cabozantinib available in Japan as well. But at the present, we have 2 secondary agents, regorafenib and ramucirumab. After regorafenib, we usually introduce sorafenib first. If the patient is tolerable to sorafenib for more than 20 days at 400 mg, we introduce regorafenib. After lenvatinib, if the AFP [alpha-fetoprotein] is greater than 400 mg, we use ramucirumab as a second-line agent after lenvatinib. After sorafenib, of course, we use the regorafenib or the ramucirumab if the dose is greater than 400 mg. In Japan, as a secondary agent after sorafenib, lenvatinib is also effective. As a third line after regular after regorafenib, lenvatinib is also useful. So we can use lenvatinib as a second-line or third-line agent after sorafenib

Minsig Choi, MD: What are some clinical parameters I use in deciding what drug I would use? I have given you so many treatment options. One of the things I factor in is how well did the patient do on their first-line treatment, whether it’s lenvatinib or sorafenib? If the patient did very well on their first-line treatment options, then using another TKI agent would be an effective way. That means regorafenib or cabozantinib. If the patient didn’t do very well on their first-line treatments, they progressed rapidly within 3 or 4 months on their first-line treatment options, then I would personally use immunotherapy like nivolumab or nivo-ipi [nivolumab-ipilimumab] combination or pembrolizumab in my second-line setting. Then you can use both of these options as third-line treatment options in the future.

For patients whose alpha-fetoprotein is above 400 mg, we have another option, ramucirumab, which is a biomarker-based therapy that can be used in HCC patients. I would look at first how did the patients do in their first- and second-line settings, what are their alpha-fetoprotein level, and then what are the toxicities they have. If the patients did very well without many toxicities, then using another TKI might be a reasonable option. If the patient had a lot of adverse effects, like hand-foot syndrome, hypertension, fatigue, weakness, and weight loss in their first-line treatment options, then I would want to change the mechanism of action and go with checkpoint inhibitors in my second-line setting. I would factor in all those things.

It’s a nice problem because now we have so many options. A few years ago I would have told you that you have not many treatment options for liver cancer. We were just using different doses of sorafenib in managing those patients.

Transcript Edited for Clarity