Opinion|Videos|November 13, 2026

ROSELLA and KEYNOTE-B96: Two New Standards, Head-to-Head on Design and Data

The ROSELLA trial is particularly exciting, in my opinion.

Vue: Yes, definitely. The ROSELLA trial is particularly exciting, in my opinion. ROSELLA investigated a new agent called relacorilant — a selective glucocorticoid receptor modulator — in the platinum-resistant ovarian cancer population. Investigators found that patients treated with relacorilant in combination with nab-paclitaxel had delayed disease progression and lived an average of approximately 4 months longer than patients who received standard therapy. At 2 years on treatment with relacorilant, approximately 50% of patients were still alive, compared with approximately 25% in the control arm. That is very compelling, and importantly, this benefit was seen without any new toxicity signals. Is there another trial in the space you’re excited about?

Miller: Yes. The ROSELLA trial showed us progression-free and overall survival benefit with the addition of relacorilant to a taxane backbone. The KEYNOTE-B96 trial looked at a different question — the addition of immunotherapy to a taxane backbone. KEYNOTE-B96 was a 1:1 randomized trial of weekly paclitaxel with or without pembrolizumab, with investigator’s choice to add bevacizumab. Similar to ROSELLA, it showed both progression-free and overall survival benefit with the addition of immunotherapy. We obviously know that immunotherapy carries its own toxicities, but when pembrolizumab was combined with paclitaxel we did not identify any new safety signals with the combined regimen.

Miller: It is a unique moment in platinum-resistant disease to see several trials, in short succession, each improve overall survival for this patient population. We have mirvetuximab soravtansine from the MIRASOL trial, pembrolizumab plus paclitaxel from KEYNOTE-B96, and relacorilant plus nab-paclitaxel from ROSELLA — all demonstrating an overall survival benefit in patients who have historically been very challenging to treat. That is really significant progress. With all of these options on the table, are there patient-related factors that would prompt you to select one regimen over another?


Related to this article