Naiyer A. Rizvi, MD: These drugs received accelerated approval in the third-line setting, and then we had 2 negative phase III trials with either NIVO [nivolumab] or NIVO-IPI [nivolumab-ipilimumab]. Taofeek, can you walk us through that and what your thoughts are about that?

Taofeek K. Owonikoko, MD, PhD: Yeah. The use of this agent in the relapsed setting got very strong traction based on the CheckMate 032 in a small subset of patients but where that benefit is quite durable. We’re all looking for the tail, and the tail was there. I’ve used both NIVO [nivolumab] and NIVO/IPI [nivolumab-ipilimumab] as well as PEMBRO [pembrolizumab] now in the relapsed setting, initially based on the NCCN [National Comprehensive Cancer Network Guidelines] recommendation that allowed us to get insurance to approve it for those who are not able to get on a clinical trial.

The disappointment came when we had the head-to-head comparison of nivolumab to topotecan in the CheckMate331 trial with what we saw in the CheckMate 032, which was somewhat less of a controlled study, even though we had multiple cohorts. Those are now being compared with each other or any standard agent but then going head-to-head against topotecan, which we all don’t use, we all don’t like. But we’re stuck with it because it’s the only approved agent. And for that we have come up short. It has made everybody start to rethink whether the initial observation in this CheckMate 032 trial was real or just a statistical fluke.

But when we see these patients in clinic, and we’ve all seen them, we know that this agent actually does work for patients. It’s just that knowing who those patients are is key. So the CheckMate331 trial has really not dissuaded me from the use of nivolumab and nivolumab-ipilimumab in appropriate patients. I tend to use it now as third line, based on FDA approval. In the second line I would go with a cytotoxic agent that is not topotecan for the most part. And then for PEMBRO [pembrolizumab], actually the 1 thing I don’t use is the CPS [Cognitive Performance Scale] score to select patients, but if we have to go by what the data show, that’s actually where you are likely to see the outstanding benefit that you see in that small subset of patients with this agent in the relapsed setting.

And then talking about the other trial of NIVO-IPI [nivolumab-ipilimumab] and NIVO [nivolumab] in the maintenance setting, which is the CheckMate 451 study, disappointingly, that’s probably the expectation that when we maximally cytoreduce the patient with induction chemotherapy and then come afterward with maintenance immunotherapy, that we’re going to see survival benefit. That study actually failed to show any survival advantage with the use of nivolumab and ipilimumab.

Part of that could be that the regimen that was tested was a little too toxic for patients to tolerate and stay on for a long period of time. And the median number of cycles delivered in that arm of the study was only 1 cycle, and that was the 1 mg/kg nivolumab versus 3 mg/kg of ipilimumab. Because of the design, although there was a trend toward improvement with nivolumab, you could not test that statistically because the primary arm of the study was the nivolumab-ipilimumab arm, and it was negative. There was actually nothing that you could statistically conclude in terms of the nivolumab arm of the study. Now we’re left with this situation where the attempt to use immunotherapy in the frontline or to use it in the relapsed setting or as a maintenance has failed, at least based on the CheckMate 451 trial. I do not see any foreseeable opportunity for us to retest that concept, given where the field has moved.

Transcript Edited for Clarity