Medicine Matters oncology

The DANUBE trial is a large, randomized, phase 3 study. It compares durvalumab alone, durvalumab plus tremelimumab, and chemotherapy, and that's standard chemotherapy, in front line urothelial or bladder cancer.

The trial enrolled over 1,000 patients. It had co-primary endpoints. For the durvalumab alone arm, it was looking at overall survival in the biomarker-positive population. And in the durvalumab plus tremelimumab arm, it was looking at overall survival in the ITT population. The characteristics of the patients were well balanced, and the trial was well conducted and was robust. It actually had the longest follow-up of 41 months, as a follow-up.

When one looks at the durvalumab alone arm, in the biomarker-high population, the hazard ratio was 0.89, which was not statistically significant. The shape of the curve shows a characteristic crossing. However, essentially durvalumab did not outperform chemotherapy in this biomarker-positive setting.

In the durvalumab plus tremelimumab arm, we showed in the ITT population, not the biomarker-driven population, a hazard ratio of 0.85. This was also not statistically significant. Again, there was a characteristic crossing of the curve.

When we look at response rates and progression-free survival, we can see chemotherapy outperforms the combination of durvalumab plus tremelimumab in the ITT population, and durvalumab in the biomarker-positive population. What this represents is chemotherapy performing well in the early part of the disease, and then immune therapy catching up. But essentially, a lot of the ground during the early period is lost.

There was access to subsequent therapy, so 31% of patients in the control arm got subsequent immune therapy. When we looked at the safety profile, we could see that 60% of patients on the chemotherapy arm had grade 3 or 4 adverse events. But this dropped to about 15% in the durvalumab alone arm and 28% in the durvalumab plus tremelimumab arm, showing immune therapy overall being better tolerated than chemotherapy.

Nevertheless, the summary of the trial showed that the durvalumab alone in the biomarker-positives, and durvalumab plus tremelimumab in the ITT population was not significant, and both arms of the trial were negative.

There was some provocative exploratory analysis which looked at durvalumab plus tremelimumab in the biomarker-high population. In this exploratory analysis, we showed a hazard ratio of 0.74. Had that been the focus of the trial, it would have been statistically significant. It wasn't, and it wasn't. That's a shame, because what we have seen, I think, is the efficacy of durvalumab being increased by the addition of a CTLA4 inhibitor. And this requires further evaluation in the future.

We did see increased efficacy with the addition of tremelimumab, but we also showed increased adverse events. This is important when considering drug development for the future.

Overall, what we showed in DANUBE was a large robust study in the frontline setting with the longest follow-up for immune therapy. The trial was negative for its primary endpoint. It does add a great deal to the body of literature, because some of the results of this trial are consistent with other studies in a similar environment, looking at monotherapy or combinations in this setting.

What we've shown is it's quite hard to beat frontline chemotherapy, and other strategies for the future will be needed. We know, for example, chemotherapy followed by maintenance avelumab has a survival advantage.

We look forward to developing CTLA4 inhibitors further in the future.