Results from the CASPIAN Trial for Extensive-Stage Small Cell Lung Cancer

Kathryn Gold, MD: Let’s move on to the second major trial that was presented, the CASPIAN trial, with durvalumab, etoposide, and platinum therapy plus or minus tremelimumab in extensive-stage small cell lung cancer. Would you mind discussing the design of that trial?

Taofeek K. Owonikoko, MD, PhD: The CASPIAN trial is similar to the IMpower133 trial in terms of concept and the end point of interest. This was a 3-arm trial trying to test the combination of checkpoint blockade using durvalumab or checkpoint blockade using durvalumab and tremelimumab, whether targeting both PD-L1 and CTLA4 to chemotherapy would be better than just giving chemotherapy alone. The initial plan and interim analysis of this study showed significant improvement in terms of the primary end point of overall survival and progression-free survival for the addition of durvalumab to chemotherapy. At that point, we knew 1 arm of the 3-arm trial was significantly better and improved the outcome when compared with chemotherapy. The data were not mature enough for the durvalumab/tremelimumab-plus-chemotherapy arm to be compared with the placebo arm.

Recently at the last ASCO [American Society of Clinical Oncology] Annual Meeting, we had the opportunity to see the full data set, and this helped us crystallize how important it is to add the immune checkpoint blockade to chemotherapy and, more importantly, how best to go about it—that maybe more is not going to be better. This was a randomized trial, 1:1:1 to the 3 arms of the study. Approximately 800 patients were enrolled, with about 260 patients in each arm of the trial. At the time of the data presented at ASCO, with follow-up of around 25 months, more than 80% of the data had matured. The initial report of durvalumab added to chemotherapy did better than chemotherapy was reconfirmed at this more mature data point. Hazard ratio was 0.75, indicating 25% reduction in risk of death with the addition of durvalumab to chemotherapy.

The median overall survival, which is generally not a good estimate of impact, was very favorable overall to the durvalumab-plus-chemotherapy arm—about 12.9 months vs 10.5 months. Looking at the long-term survival of the study is more important. You’re more likely to be a long-term survivor; 22% of those patients have been alive for about 2 years compared with those who had only chemotherapy, which was about 15% of the patients. These consistent data sets are similar to what we knew before about the addition of durvalumab.

What was new was the result of adding tremelimumab to the mix, so durvalumab plus tremelimumab plus chemotherapy. Three things stood out from this trial: One, by adding the tremelimumab to durvalumab and chemotherapy, there was increased toxicity; second, the increased toxicity was not insignificant because it also led to discontinuing treatment at a higher rate than when you only added durvalumab to chemotherapy; and finally, if the toxicity was improved or increased but still able to show better efficacy, they might say, “Maybe that’s a trade-off from increased toxicity and improved efficacy.” Unfortunately, that was not the result. The efficacy was not statistically significant. There was a trend toward improved overall survival for those patients who got durvalumab, tremelimumab, and chemotherapy compared with chemotherapy alone.

When we look at what the impact of the addition of immunotherapy to chemotherapy is—because we are able to deliver durvalumab along with chemotherapy more safely during that induction phase—you are able to realize the benefit. By adding tremelimumab into the mix, if you look at the median number of cycles of induction chemotherapy-immunotherapy that were delivered, it was about half of what you would do with durvalumab and chemotherapy alone. My own interpretation of the data is that what you ended up doing is not giving the patient the immunotherapy you’re trying to give; you also compromised the delivery of the chemotherapy itself, which, as we know it, is the more effective component of the regimen. That is why we did not see that early separation of the curves when we looked at the 4-drug regimen vs chemotherapy. With the 3-drug regimen of durvalumab plus chemotherapy vs chemotherapy alone, that curve separated much earlier at around 4 to 6 months. For durvalumab, tremelimumab, and chemotherapy, it did not separate until around 9 to 12 months. At this point in time, this would be our new standard of care for patients—that is, the addition of durvalumab to chemotherapy—just as we are already doing for atezolizumab with chemotherapy.

One minor point to make about the design of the CASPIAN trial compared with IMpower133 is that this allowed for the use of any platinum agent, carboplatin or cisplatin; for the IMpower133 trial, you are limited to just carboplatin. I don’t use cisplatin with my extensive-stage patients. For those practitioners who, for whatever reason—maybe clinical or personal preference—are still enamored with cisplatin for extensive-stage disease, this gives you the data set to support that the addition of immunotherapy to a cisplatin-containing regimen will be just as good as a carboplatin-containing regimen.

Transcript Edited for Clarity