Ghassan K. Abou-Alfa, MD, MBA: Let’s stick to the lenvatinib scenario. Pierre, first-line lenvatinib—it’s not necessarily the only answer. PEMBRO [pembrolizumab] as second-line. What would be your third line?

Pierre Gholam, MD: The plot thickens and the evidence thins here. I believe the only evidence we have in any shape or form for patients receiving “third-line therapy” comes from the CELESTIAL trial of cabozantinib, with fairly generous exclusion and inclusion criteria, which allowed patients to be on sorafenib but also having received a second line of systemic therapy.

I believe about 20% of patients actually fit that scenario, at least 8%—or something like that—may have received some components of the I/O [immuno-oncology] therapy.

We obviously have very small numbers to gauge whether, in that small subset of patients, there was benefit. But at least in the overall cohort of patients who received second-line therapy, there was still some benefit. Therefore, one might surmise that in that particular setting, cabozantinib would be a reasonable option for those patients, assuming they continued to have adequate liver function and functional status that would allow them to pursue therapy. I think that is our collective responsibility: to keep these patients in decent shape from a functional and liver perspective to allow them to pursue potentially more life-extended therapy.

Ghassan K. Abou-Alfa, MD, MBA: Catherine, I like what Peter is saying. Tell us a little more about cabozantinib. Because clearly it is second line, solid, but also suggestive of third-line data as well within the same study. Is it easy to give, and is it well tolerated?

Catherine T. Frenette, MD: It’s relatively easy to give. It’s 1 dose a day, so once a day. The full dose is 60 mg, and you dose reduce to 40 mg and then 20 mg. If you have someone with, say, Child-Pugh B, it hasn’t really been studied in Child-Pugh B cirrhosis, but we do have dosing that those patients would start at 40 mg. You really can adjust the dose if you need to. It has the TKI [tyrosine kinase inhibitor] adverse effects that you expect—you know, diarrhea, hand-food skin reaction, hypertension, poor appetite, some weight loss. But really, with all these adverse effects, they’re manageable. And I’m pretty aggressive with my patients in terms of seeing them back quickly and adjusting the dose, treating their adverse effects, keeping their adverse effects under control to be able to keep them on therapy so that they can have that clinical benefit.

Ghassan K. Abou-Alfa, MD, MBA: There’s a lot of information we’ve learned from all our colleagues here. No. 1, I would like to start where Dr Frenette started the discussion, that there was a default that sorafenib was the only available drug at that time, and that’s why all those studies were kind of 1 way or the other compared and considered after prior exposure to sorafenib.

No. 2 is—and I very much also like what Catherine ended up with—that she wants to make sure that the patients are monitored very carefully and very closely. Because these TKIs can have certain complexities to them, let alone the I/Os. I’m sure these patients are under close observation at all times.

In regard to the choices, we heard them a little 1 way or the other, and it’s not as if I have any better answer than what our colleagues have mentioned. The first choice at the moment is regorafenib. But as we heard from Dr El-Khoueiry and Dr Gholam, it has a very robust requirement of prior exposure to sorafenib, tolerance of sorafenib, and progression on sorafenib, so patients get regorafenib. It is already within the FDA approval that prior sorafenib would be required.

No. 2 is cabozantinib, which has very robust data. As we heard, it can apply not only for second line, where it’s a primary end point of the study—and by all means it’s a priority—bus also a third-line opportunity.

We heard from Dr Frenette that the tolerance of the drug is acceptable. I have to admit, I have a lot of experience with cabozantinib, and we have seen that the drug is acceptably well tolerated. It might require certain dose adjustments because of the fatigue component. As we heard, 60 mg to 40 mg to 20 mg. The data that we published in the New England Journal of Medicine really supported the activity of the drug to patients who are equal to all of us who we see in clinic.

The third option we heard about, which is very important, was ramucirumab, which is anti-VEGF—a great tolerance. But there was a requirement in that study, which is the REACH-2 study, the majority of patients had an alpha-fetoprotein more than 400 ng/mL to get the drug.

This remained a little bit of a debate. What’s the value of the alpha-fetoprotein? By all means we have seen incredible data, especially with the prior REACH study, which allowed all patients to get on. It looked specifically at the patients with the high alpha-fetoprotein. When they joined this from the REACH-2 study, the outcome definitely was more apparent, and it’s an option.

Dr El-Khoueiry invited us to really look into the checkpoint inhibitors in regard to second-line therapy. I would say that disappointingly, but it is what it is. The data in the first line of nivolumab versus sorafenib are negative. I like very much how our colleague at ESMO [European Society for Medical Oncology], Dr Arndt Vogel, said it. He said it 3 times: it’s a negative study.

In addition, with regard to the PEMBRO [pembrolizumab] versus placebo in second line, we like to really see the “clinical benefit,” as Dr El-Khoueiry referred to it. Statistically it was negative, so it’s a negative study. Nonetheless, to give credit, those 2 drugs are available and already approved in the second-line setting in HCC [hepatocellular carcinoma]. Why is that? Actually the FDA probably was mostly clear on the duration of response. In other words, patients can stay on therapy for a long time, and that’s why we like them very much. As such, it is nothing but appropriate and correct to give pembrolizumab as an option in the second-line setting. Of course, nivolumab could be another option in second-line setting.

As you noticed, I played the scenario with the patients with prior exposure to lenvatinib specific. Why did I choose lenvatinib? Because, No. 1, we’re well aware of the conditional thing between sorafenib—REGO [regorafenib], and I don’t want all the answers to come, but it’s straightforward. I wanted to ensure we built a little more complexity to the story.

And No. 2, to be fair, lenvatinib has its presence in the market because of the appropriate tolerance, the great management, and of course the excellent progression-free survival. As we heard from Dr Gholam, there was a 3-times increase in PFS [progression-free survival] compared with sorafenib, even though it was an inferiority study, and an impressive response rate by imaging close to 24% by RECIST 1.1, and close to 41% by the modified RECIST trial.

Ghassan K. Abou-Alfa, MD, MBA: I can definitely see the 2 lines. By all means, I can see the 2 lines. Interesting enough, Catherine, what we heard from Peter a second ago is that cabozantinib has “second- and third-line applicability”—ie, if a patient gets ATEZO-BEV [atezolizumab-bevacizumab]. Let’s say we’re going to go through the scenario. We’ll go to second line automatically, and we’ll do CABO [cabozantinib]. But let’s say now we would like to revisit the first-line approaches, as was also suggested by Anthony, and we’ll go to sorafenib or lenvatinib. Do you agree that it’s nice that we have third-line data on cabozantinib anyway?

Catherine T. Frenette, MD: It is nice to have that third-line data on CABO [cabozantinib]. I don’t know that that would make me want to save it back for third line, because clinically, very few patients are going to make it to third line. Usually by the time you know you’re progressing through first line, you know half of them can’t get second-line because of progression of their liver disease or poor performance status. It’s really a very small percentage that will make it to third line, I think. And I don’t want to lose the opportunity to give them a very good TKI in cabozantinib, if that is where I think the patient would be most appropriately treated.

Ghassan K. Abou-Alfa, MD, MBA: Well, that’s quite fascinating. We are hearing different opinions on that, and it’s not as if we are—any of us—working on very thin ice in regard to the choices of therapy. Nonetheless, the schools of thought are, ironically, all turning around, ultimately around cabozantinib, because this is the only 1 that really had a second- or third-line therapy. If we think of that approach, as Catherine just mentioned, we want to make sure we’re using the best available data for the patient, and start moving forward, rather than go backward. As such, ATEZO-BEV [atezolizumab-bevacizumab] followed by cabozantinib.

On the other hand, we also heard from Pierre that maybe we are keen on giving every opportunity because—I probably can understand Pierre’s view—after all, the data that we spend tons of time on, and a lot of patients contributed too, are not a waste in regard to sorafenib, lenvatinib, etc. The idea of going with ATEZO-BEV [atezolizumab-bevacizumab] and back to, for example, lenvatinib, and moving forward to cabozantinib based on the third-line data is also valuable.

Time with further data availability and, of course, the real-world experience are going to be very important. I think the new advantage we have is, as it used to be called in the old days, the “phase IV studies.” But now we have certain entities that can help us—real-world evidence and real-world data that are going to be probably guiding a lot of our decision-making in regard to dose treatment. Again, this is yet to be discovered.

Transcript Edited for Clarity