Selecting the Appropriate Treatment in Second-Line and Beyond in Advanced HCC

Anthony B. El-Khoueiry, MD: Dr Li, in the second line and beyond, what are the factors that you think about to determine whether somebody is a candidate and how you’re going to treat them?

Daneng Li, MD: That’s the million-dollar question as our treatment options expand much more in the first-line setting. I approach this by looking at 3 or 4 main factors. One is looking at the potential mechanisms action of the second-line agents that I’m going to use relative to the first-line agent that a patient might have been previously treated with. For example, many of these tyrosine kinase inhibitors [TKIs] target various pathways that are involved with HCC [hepatocellular carcinoma] growth, not necessarily just VEGF. The type of agent a patient received in the frontline setting might help inform me in terms of which agent I might select in the second-line or beyond setting.

In addition to that, there are 2 other major factors. One is how the patient looks overall in terms of comorbidities and fitness. With any of these different types of agents or even additional immunotherapy, how they tolerate treatment matters. Because as we go further along in their HCC journey, patients tend to sometimes get a little weaker from their prior treatments.

Finally, it depends on the overall burden of disease. There are potential differences in the potency of various different types of TKIs. Lenvatinib tends to be a much more potent agent and maybe has more of an ability to shrink bulkier tumors, whereas cabozantinib is more of a stabilizing agent and gets good disease control in terms of overall stability. It depends on the patient population where it’s coming in. We certainly have more data with cabozantinib for patients who have Vp4 [main portal vein thrombosis] disease because CELESTIAL [NCT01908426] allowed for all comers in terms of Vp4 inclusion within the trial itself. Meanwhile, in REFLECT [NCT01761266], where we’re drawing that data from, lenvatinib excluded those patients. That also helps me decide. Those 3 factors help me in clinical practice outside of predictive biomarkers, as you mentioned, to decide which ideal choice I go with in the second-line or beyond setting.

Anthony B. El-Khoueiry, MD: Great. Dr Rich, Dr Li said he would look at mechanism of action, differentiators, performance status, tumor burden, and goals related to that. As a hepatologist, is there anything else you’d be concerned about in thinking about second-line and beyond?

Nicole Rich, MD: Like Dr Li mentioned, this is a million-dollar question that everyone is struggling with right now. From the pure hepatology standpoint, I’m concerned about what’s going on with the liver function. And if someone progressed on immunotherapy from Child-Pugh A to a Child-Pugh B, you may be hesitant to consider a second immunotherapy option. As you previously mentioned, TKIs have been shown to be safe in Child-Pugh B. That would be a consideration.

A point to raise would be that all of the second-line treatments that we’ve discussed have been studied after sorafenib first-line therapy because it’s been the standard of care when these studies were planned. Future studies will compare to atezolizumab-bevacizumab if that’s the standard, but we don’t have a lot of data on what to do second line, especially after atezolizumab-bevacizumab, as you mentioned.

There’s a small study that showed that some patients still have a benefit after treatment with nivolumab-ipilimumab after immunotherapy. But in general, if you look at the atezolizumab-bevacizumab study, a lot of patients went on to treatment with a TKI afterward. That suggests that a TKI would also be appropriate in that setting. We need more data on that. Another consideration would be what happened to the patient in the first line. If the patient’s HCC seemed to not respond at all to immunotherapy, we need to be patient and wait. There are concerns for pseudoprogression. If you feel the patient progressed on first-line atezolizumab-bevacizumab, you’d want to consider a TKI.

Anthony B. El-Khoueiry, MD: Great. You partially answered my next question, or gave a great segue to it. Let’s walk through some scenarios. Let’s say a patient received atezolizumab in the first line, which is the current standard of care for the majority of Child-Pugh A patients. What’s your go-to second-line therapy? You said probably a TKI. Are you starting with sorafenib or lenvatinib or are you jumping to a second-line agent, like cabozantinib or regorafenib? What are you doing in your clinic?

Nicole Rich, MD: I say this as a hematologist working in conjunction with my medical oncology colleagues. We throw this around. In our practice, we generally jump to sorafenib or lenvatinib in that case. Although, we don’t have the data to guide us. The other thing would be whether a patient is a candidate for trial after atezolizumab and bevacizumab. I’d be interested to see what the rest of the panel is doing in that situation, but in general, we’ve been going to either sorafenib or lenvatinib. The choice there is based on the adverse effect profile and individualized to the patient and their other comorbidities, such as if they work with their hands and you’re really worried about hand-foot syndrome, etc.

Anthony B. El-Khoueiry, MD: Very good. Dr Li, any comment? What’s your go-to regimen post–atezolizumab-bevacizumab?

Daneng Li, MD: To add to what Dr Rich said and my prior comments, if someone has significant burden and I’m looking for tumor shrinkage, I’m going to choose the agent that has potentially the best chance for maximum response in terms of shrinkage of the tumor to possibly provide symptomatic relief. In that setting, I’d probably go to lenvatinib. For other patients, especially patients who had Vp4 disease, I’d go to cabozantinib in the second-line setting.

Dr Rich mentioned that study from a couple of institutions in Asia looking at nivolumab-ipilimumab post–atezolizumab-bevacizumab. That’s reasonable to try; especially in patients who had really good responses on frontline atezolizumab-bevacizumab. When they looked at this subset of the population, the patients who fared well were patients who initially derived tumor control and had a response to treatment; not patients who essentially had initial resistance to immune checkpoint inhibitors. Patients who potentially had more acquired resistance would benefit from challenge with nivolumab-ipilimumab in that second-line setting post atezolizumab-bevacizumab.

Anthony B. El-Khoueiry, MD: To be fair, as you said, those data were from 2 institutions. It was more like a phase 2 study with no control arm, for full disclosure. But it raises intriguing data and an intriguing option that maybe we could study further. Dr Gholam, do you have any dissenting opinions or anything to add?

Pierre Gholam, MD: Not so much a dissenting opinion, but my guiding principle, at least for the time being, is to remain loyal to the sequencing as outlined in FDA-approved therapies. Once a patient fails or progresses beyond first-line agents, it has been my practice to commit them to an FDA-approved second-line agent. That would be cabozantinib in the majority of these patients and regorafenib selectively in some patients, especially if they’ve had a trial of tolerance of sorafenib in the past and have tolerated it reasonably well, which is what the RESOURCE trial [NCT01774344] was all about. I will very sparingly use ramucirumab in patients who have an AFP [⍺-fetoprotein] greater than 400 ng/mL.

I haven’t ventured to treat someone who received IO [immuno-oncology] in the first line with subsequent IO. A medical oncologist would be much better versed in this, but my limited review of this topic is that that strategy hasn’t historically bared fruit. I’d love to hear both your thoughts about it, and certainly Dr Rich’s if she has any experience with it.

Anthony B. El-Khoueiry, MD: You’re absolutely right. The data are very limited at this point. There are some ongoing phase 2 studies looking at either continuation of PD-1 inhibition and bringing on a TKI on top post progression on a PD-1 or looking at PD-1/CTLA-4 combos. The data we have are very limited, basically a single institutional series. Maybe there’s a hint of response rate, but it’s certainly not the 30% we saw with nivolumab-ipilimumab post sorafenib. You can correct me if I’m wrong, but in that study that was reported, the response rate was lower, in the 10% range.

Daneng Li, MD: Yes. That’s absolutely correct.

Anthony B. El-Khoueiry, MD: That’s where we stand.

Transcript Edited for Clarity