Selecting a Frontline TKI for Advanced HCC

Transcript: Andrew Zhu, MD, PhD: You mentioned the REFLECT trial, which I think is a very interesting study, because after 12 years, clearly this was the first drug where we were able to bring a second agent to the first-line setting. The study design, as you pointed out, was actually noninferiority design. Tell us a little about the safety profile of these 2 agents and how the physicians can actually make the right decision to use sorafenib versus lenvatinib.

Nikolaos Pyrsopoulos, MD, PhD: Very valid point. And there is a different profile in regard to adverse events. For example, we know that palmar-plantar erythrodysesthesia—in other words, the hand-foot syndrome—is much more prominent with a utilization of sorafenib. On the other hand, another profile that is more prominent with lenvatinib can be hypertension, or thyroid abnormalities that we’ve seen in 12% of the population.

With sorafenib we can see gastrointestinal symptomatology like diarrhea and also weight loss. And with lenvatinib, something we need to pay attention to is proteinuria, especially for patients with cirrhosis who might experience malnutrition and decreased protein levels. This is something we need to pay attention.

Andrew Zhu, MD, PhD: Very good.

Nikolaos Pyrsopoulos, MD, PhD: Alopecia is very important as well, because if we look at the registration trial, the body image that was reflected in the quality questionnaire was not that good with a sorafenib. In other words, the weight loss and the alopecia make patients feel not very well.

Andrew Zhu, MD, PhD: Very good point. You mentioned the safety profile between these 2 agents. How about the secondary end points of these 2 drugs? Was that actually affecting your decision choosing 1 versus the other?

Nikolaos Pyrsopoulos, MD, PhD: Well, the secondary end points were statistically significant actually for lenvatinib. Because if we look at the time to progression, the progression-free survival, it was 7.9 and 8.9 months versus 3.7 months. So it looks as though there is a very strong home run. But at the end, both of them are finishing at the same time in the finish line.

Andrew Zhu, MD, PhD: For the overall survival.

Nikolaos Pyrsopoulos, MD, PhD: Exactly. The overall survival remains the same. Now, how can we incorporate that in our daily practice? Do we need to pretreat patients? And before we exhaust the result, do we introduce a second agent? Do we need to start from the get-go with a combination therapy? There are some trials, as we mentioned earlier, with a utilization of lenvatinib plus pembrolizumab on the way. I’m really curious to see the results of these trials.

Andrew Zhu, MD, PhD: Very good. Yes, I definitely agree with you. I think with a selection of sorafenib versus lenvatinib, it’s fair to say we do not have a predictive biomarker to guide us to choose 1 versus the other. It sounds like we still have to make a decision based on perhaps the secondary end point—ie, lenvatinib perhaps will have a better chance to induce radiological responses—and also the safety profile of these 2 agents. They differ dramatically.

Nikolaos Pyrsopoulos, MD, PhD: I agree because the mRECIST [Modified RECIST] response was almost triple for lenvatinib, 24% versus 9%. But it was the mRECIST, and the mRECIST is looking at the enhancement. It is looking as if RECIST 1.1 that is investigating the shrinkage of the tumor. But something else that people might think about is that once you ask the question “What to choose?”, it is as if you’re in the candy store. You see different candies, but you don’t know which 1 tastes differently.

For example, with sorafenib, we have some long-term experience. And the GIDEON trial revealed that it is relatively safe for patients with Child-Pugh score of B and C for chronic liver disease, something that we need to see with new data for lenvatinib. In the registration trial with lenvatinib, they were investigating branch portal vein thrombosis. We know that portal vein thrombosis has been investigated with sorafenib. Something else that we might take under consideration is the tumor burden. Supposedly, very large tumors have a good response with sorafenib. But I believe that along the way more will be revealed.

Something else we need to look at is the precision medicine, the way we’re moving nowadays. In other words, how will we be able to match the individual characteristics of the patients with the profile of the medication we try to introduce? For example, somebody with skin disorders. I will be skeptical to utilize sorafenib. On the other hand, somebody with significant uncontrolled hypertension, I don’t know how safe it would be or how much we would like to utilize the combination of medications where hypertension is in the downside.

Transcript Edited for Clarity