Joseph Llovet, MD, PhD: We can talk a bit about personal experience. Do you want to talk about, for instance, this personal experience with cabozantinib, ramucirumab, or whether these trials were run in your institution? What’s your experience with regorafenib and so on?

R. Kate Kelley, MD: I think you’re getting at this tricky decision now where we have to identify which agents to use in second- versus third- and, in some cases, even fourth-line therapy now with so many drugs available. We haven’t used ramucirumab much. It does not yet have an FDA label in the United States, but from my experience in clinical trials, it is generally a well-tolerated drug. Though one notes in the trial data that there is a small number of patients who can have encephalopathy that needs to be carefully monitored, as well as ascites or fluid retention. I think right now the data are limited to Child-Pugh A and fit patients with liver function. Is that your experience as well?

Joseph Llovet, MD, PhD: Yes. So in fact, if I recall well, in the REACH trial at the beginning, Child A and B patients were accepted, but after a first watching of adverse events related to particularly Child-Pugh B patients in terms of encephalopathy, this indication or inclusion criteria were removed. And the adverse event of encephalopathy now is managed in Child A patients. But still you are seeing that 3.5% or 4.0% of the patients may experience that, which is acceptable for these types of drugs.

Arndt Vogel, MD, PhD: First of all, we have no more options for systemic therapies. When we are sitting in the tumor board, it’s very good because we have a lot of arguments to say that we should stop local therapies and switch to systemic therapies. And in the past, I think we really have pushed the limit of local therapies to an extent that was not really good for the patient. And most of these patients during this treatment develop progressive liver disease, and there are no more candidates for systemic therapies. And at that time, we already had 1 drug available.

Now we have more drugs available, and what we are seeing from our clinical practice is that we sequence these patients. And we look at the clinical trials, for example, the RESORCE study, which was a second-line trial with a select population of patients. But nevertheless, when we look at the overall survival from beginning of first-line, the median overall survival was 28 months, which is really impressive, or 26 months, which is really impressive. And today we have seen data from the lenvatinib trial and also in this trial patient either with lenvatinib or sorafenib that received second-line therapy that show a median overall survival of 20 and 17 months, which indicates if you sequence the treatment, survival of our patients will be better.

And, therefore, in the future, we really need to see that we stop local therapies as early as possible. We should, of course, use them if they are candidates for local therapies. But if they are not candidates for local therapies, or if they do not really respond well to these local therapies, we should switch to systemic therapies. And now we have options. We can decide a little bit depending on the patient. We can see whether the patient tolerates either drug and then switch to the other drug, which makes it really easier in the field. What is the best sequence? We don’t know yet. And a lot will really depend on the label, so which drug we can use at which time point.

Masatoshi Kudo, MD, PhD: We have 2 checkpoint inhibitors and 5 targeted agents. So, a first choice of treatment will be checkpoint inhibitors because response rate is not so high, up to 20%. But once response is obtained, the response is durable, long-lasting, and always better; but in the remaining 80%, no response is observed after the 3 months. Response is observed by 3 months. So at the time of 3 months, if there is no response, we switch to lenvatinib because lenvatinib shows high response. And response correlates with a longer overall survival. So we choose lenvatinib. And then in third-line, maybe we choose ramucirumab, then maybe cabozantinib or other agents.

Transcript Edited for Clarity