Manish A. Shah, MD: Actually, there are 2 cohorts in KEYNOTE-059—one was presented by Dr. Fuchs, and the other one was presented the day earlier—which combined a PD-1, pembrolizumab, with chemotherapy. So Dr. Fuchs’ study was a single-arm study, third-line or above of pembrolizumab by itself. To Yelena’s point, and she’s absolutely right, the response rates were very low—less than 10%. And interestingly, and Kohei will talk to this I think, it’s actually the same response rate that was seen with nivolumab in the ONO study. And in that study, the hazard ratio was 0.65, and we’ve not had a randomized study in gastric cancer for metastatic disease with that hazard ratio. So even though the response rate is low, there is benefit to the PD-1 inhibitor. I think it’s a fair bet that it will likely get approved at some point.

Johanna C. Bendell, MD: When Dr. Fuchs finished the presentation, a lot of people asked, “Well, is this just by microsatellite instability status?” So I will tell you a little story. The first time I met Yelena was not at one of these, but it was because she e-mailed me about 5 times before I responded. Sorry.

Yelena Y. Janjigian, MD: It’s OK.

Johanna C. Bendell, MD: And then she called me and she said, “I need to know the MSI status of your patients who were put on my study with immunotherapy.”

Yelena Y. Janjigian, MD: I think it was because of this type of analysis, the collaborations, having a report, and being able to reach out, and you were the closest investigator I called. This trial, as I showed on the map when I presented, was an international study, so there were folks involved in Europe and across the United States. This is the CheckMate-032 study, with a cohort of patients in esophageal and gastric adenocarcinoma, and, mirroring the ATTRACTION-2 study, looking at heavily pretreated patients with metastatic disease but in the West. Because up to this point across different subtypes of biomarker analysis, everyone always believed that gastric cancer in the United States should not be approached the same as gastric cancer in Asia. And, in fact, if you have a positive study in Asia, you cannot use it for regulatory purposes for any drug approval in the United States, because it’s just a completely different ball game. So the CheckMate-032 study was a United States and Europe study. It was a smaller study, nonrandomized—or rather it was randomized but not designed to compare and contrast against placebo or other regimens.

We looked at nivolumab monotherapy and nivolumab in combination with 2 different doses of ipilimumab. It’s funny, Charlie and I presented at the same meeting, and it looked like we had copied the conclusions from each other. That’s a good sign when 2 experts in the field agree to that extent. Word for word, the conclusions were very similar. Again, there’s a subset of patients who benefit, and they clearly benefit. When you treat these patients in your clinic, you’re a believer when you see 1 or 2 of these responses. But when you look at the median progression-free or median overall survivals, they’re not impressive. So the curves are pretty much similar, but what we’re looking for is the tail on the curve, and the tail on the curve is a complete game changer.

And whether or not a combination therapy or monotherapy is the right answer, it’s hard to say. Combination of nivolumab/ipilimumab certainly improves response rate. And again, this is a recurring theme in gastric cancer. One drug is probably enough. Addition of a second or third drug is maybe important in a subset of patients, similar to what we talked about with cytotoxic agents but at a cost of toxicity. And improvement in response rate would have marginal, if any, improvement in survival. And so, in a younger patient who is fit and has a certain tumor biology, which we can talk about later, perhaps combination therapy may be explored. These are questions that are all hypothesis-generating at best and are being explored in a phase III setting now.

Johanna C. Bendell, MD: Dr. Shitara, tell us about ONO, the nivolumab study. It was the first randomized phase III data we saw of immunotherapy for patients with gastric cancer.

Kohei Shitara, MD: This is our Asian trial that enrolled patients from Japan, Korea, and Taiwan and targeted the patients who are refractory to at least 2 standard therapies. Most of the patients from Japan were enrolled as a first-line therapy, so all available treatments were conducted, and then for the clinical trial, it clearly showed a survival benefit. Hazard ratio is 0.63, and tolerability is very good. Already, companies have made these data available for the Japanese regulatory authority, last December, so I hope in the future, this nivolumab will be approved. After this, we can use nivolumab based on this trial, either third-line or fourth-line treatment.

At the same time, we also joined the KEYNOTE-059 trial, and they enrolled some patients. Important message from these 2 studies: There is no clear difference. In cross-trial comparison, there are some mutations, but the response rate of PFS is similar to some small Japanese patient populations and adjuvant trials with nivolumab. So this is a very important message. And also, we joined a pembrolizumab trial. It’s a phase III in second-line and third-line, which only targeted PD-L1—positive tumor, especially in first-line. We also joined CheckMate-649—which was developed by Yelena’s data—which compared ipilimumab/nivolumab, chemotherapy, and chemotherapy/nivolumab. So these trials are all very important to give us another step for immune therapies.

Transcript Edited for Clarity