For metastatic urothelial cancer platinum-based chemotherapy has been the standard of care for many years. In recent years, approvals for checkpoint inhibition have been added to the treatment landscape. Several trials have already tried to replace chemotherapy with monotherapy checkpoint inhibition but have largely failed.
In this study we wanted to investigate the combination of ipilimumab plus nivolumab, so anti-CTLA-4 plus anti-PD-1, which in earlier studies was found to add benefit to monotherapy checkpoint blockade.
In this study, the CheckMate-901 study, which was a multi-arm study with multiple comparisons, we looked at the comparison of ipilimumab plus nivolumab against chemotherapy for previously untreated advanced urothelial cancer. The study had two primary endpoints, one was the cisplatin ineligible population, which is the main topic of the presentation at ASCO, the other population was the PD-L1 positive on tumour cells. So in previous studies we had seen better results with patients who were PD-L1 positive so the idea was that these patients would benefit better and we would have a better outcome for this group of patients.
The main comparison of the study was ipilimumab/nivolumab versus carboplatin/gemcitabine, so in the cisplatin ineligible sub-population. Unfortunately, the result was negative. There was a hazard ratio of 0.79 with a p-value of 0.0245 where the boundary was 0.017. So we barely missed the endpoint in that population.
As I said previously, we thought that the PD-L1 positive subgroup would do better but unfortunately that was not the case. The hazard ratio for that subgroup was actually a bit worse than for the overall population.
In the end we also looked at the overall population, so there were also some patients in the control arm who could receive cisplatin. When those were added it was about 700 patients, and in that full population the hazard ratio was 0.78. However, we could not formally analyse that statistical endpoint because of the hierarchical design of the study.
Some of the important features of the study were that we saw a very durable response in a subset of patients. So if we look at the durability of response it was actually very good and especially for patients who had a complete response. So a subset of those had long-term durability without any subsequent therapy. Also when you look at landmark overall survival, so the median study follow-up was over 5½ years, the landmark overall survival was also favourable for ipilimumab plus nivolumab. But, as we evaluate these studies for the full extent of the curve, we didn’t make the primary endpoint unfortunately.
What is next for this study?
That’s a very good question. For ipilimumab/nivolumab I think this the last study that has been done with this combination. Although we see in non-metastatic cancer also some very good efficacy in smaller trials, at this point with enfortumab vedotin/pembrolizumab now taking over the treatment landscape, there are no new studies with this combination in the near future.
There is, however, one study still ongoing with tremelimumab, which is also anti-CTLA-4, which is added to enfortumab vedotin plus durvalumab in the neoadjuvant setting. We expect the results of that randomised phase III trial very soon, hopefully end of this year, next year, depending, of course, on the events taking place in that study.
