Pierre Gholam, MD: Sure. This was 3 or 4 years ago, when it was the main focus of interest and where most of the excitement in the field was. This has changed quite a bit since then. I’ll remind this audience of the pembrolizumab [Keytruda] study, KEYNOTE-240, which compared pembrolizumab to placebo in patients with advanced HCC [hepatocellular carcinoma] who were previously treated with sorafenib [Nexavar]. This was a phase 3 study. It was a much anticipated study that unfortunately didn’t meet a prespecified statistical significant cut-off for OS [overall survival] or PFS [progression-free survival]. However, the median OS was 13.9 months for pembrolizumab and 10.6 months for placebo, for a very respectable hazard ratio of 0.78.

There has since been additional analysis of these data that our colleague Philippe Merle from Lyon, France, and his collaborators performed and reported at the recent American Society of Clinical Oncology Gastrointestinal Cancers Symposium. In this study, they looked at the same patients who were randomized to either pembrolizumab or placebo and followed them over time. They found that the OS was similar, at 13.9 months for pembrolizumab, and 10.6 months for placebo. We talked about the durability of response. They found that overall survival rates at 24 and 36 months were quite respectable for pembrolizumab, at 28.8% at 24 months and 17.7% at 36 months. In comparison, placebo was 20.4% and 11.7%, respectively.

The conclusion of this study, which was much awaited and analyzed, is that there appears to be an improvement in OS and PFS. This appears to be maintained over time compared with placebo. And there appears to be a fairly stable safety profile when one looks at the data over time.

Another study that Dr El-Khoueiry would be much better qualified to summarize is the follow-up study of nivolumab [Opdivo] that led to its FDA [Food and Drug Administration] approval as a subsequent-line therapy for HCC. This approval has since been withdrawn by the company that makes nivolumab. This was a study of different dose escalation and dose expansion cohorts that looked at nivolumab in this setting. The previous use of sorafenib was allowed. Although—correct me if I’m wrong—I don’t think every cohort had sorafenib-experienced patients. There may have also been cohorts that didn’t have sorafenib experience, if I remember correctly. The meat of this paper, which was published in The Lancet almost 4 or 5 years ago, is that in the second line for the sorafenib-experienced cohorts, there was a pretty good objective response rate of about 20% for the nivolumab 3 mg/kg dose, and about 15% in some of the dose expansion and dose escalation cohorts.

A third study looked specifically at the various treatment-related adverse events in the setting of CheckMate 040, the study that looked at the safety and efficacy of nivolumab in the setting of patients who had previously been treated with sorafenib. Overall, this essentially showed that the predictable immune-mediated adverse events and other events included the high prevalence of skin toxicity in about 35%, gastrointestinal and hepatic IMARs [immune-mediated adverse reactions] in about 14%, and the other treatment-related adverse events in 68%. This study was interesting in that it cemented the duration of onset of various adverse events, which we still use in clinical practice. It told us that skin-related IMARs might happen early on, typically at about 3 and a half weeks, whereas hepatic events may not happen until 6 weeks at the median, pulmonary events may not happen until 11 weeks, and endocrine events happen fairly late. Endocrinopathies may not happen until 18 weeks or later.

There’s an option that’s still available for one to consider in second line that includes an immunotherapy option: the combination of NIVO1 [nivolumab 1 mg/kg]/IPI3 [ipilimumab (Yervoy) 3 mg/kg] based on data from the CheckMate 040 study. This study included 2 other cohorts published in JAMA Oncology in November 2020. Basically, in that optimal but increased toxicity arm where the dose of ipilimumab was higher, there was a median overall survival of about 24 months if I remember correctly.

Anthony B. El-Khoueiry, MD: Twenty-three.

Pierre Gholam, MD: Twenty-three months compared with a significant drop-off in efficacy with the other 2 arms, at the expense of significantly more toxicity with the higher dose of the CTLA-4. That summarizes where we stand at this point with immunotherapy in the second line.

Anthony B. El-Khoueiry, MD: Thank you. To make sure the audience got this, all of these trials looking at pembrolizumab single-agent, nivolumab single-agent, or nivolumab/ipilimumab were in patients who were immunotherapy-naïve and were mostly post-sorafenib. That’s 1 important point. The second important point is that KEYNOTE-240, the phase 3 trial Dr Gholam mentioned, didn’t meet statistical significance as it was predefined, but there was a numerical difference in outcome between pembrolizumab vs placebo. Recently, there was an Asian subset of that KEYNOTE-240 study from China that was reported that did reach statistical significance with pembrolizumab vs placebo. Pembrolizumab, unlike nivolumab, kept its accelerated approval in the second line, pending this confirmatory trial. We’ll see what the regulatory outcome will be.

The last thing is that with CheckMate 040, similar to KEYNOTE-240, it’s important to note that the grade 3 and 4 events in the second or third line with these agents were less than 20%. These are well-tolerated agents in subsequent lines of therapy in PD-1 [programmed cell death protein 1] naïve patients. The NIVO1 [nivolumab 1 mg/kg]/IPI3 [ipilimumab 3 mg/kg] combination that you mentioned was given for 4 doses and then nivolumab was continued alone as a single agent after the 4 doses of the combination. All 3 regimens that were compared had a similar response rate in the 30% to 32% range, but the OS was much more favorable with the higher dose of ipilimumab at 3 mg/kg.

Transcript Edited for Clarity