We designed a phase 2-3 seamless statistical design trial looking at uterine leiomyosarcoma. And in this trial, patients were either randomized to olaparib with temozolomide or an investigator’s choice of trabectedin and pazopanib. This was a one-to-one randomization. There were 74 patients randomized equally to each arm. And then the trial began to get complicated...
We designed a phase 2-3 seamless statistical design trial looking at uterine leiomyosarcoma. And in this trial, patients were either randomized to olaparib with temozolomide or an investigator’s choice of trabectedin and pazopanib. This was a one-to-one randomization. There were 74 patients randomized equally to each arm. And then the trial began to get complicated. It would accrue very rapidly; it accrued within about a year, which showed the power of the collaborative efforts in doing rare tumor research. There were a lot of problems with toxicity and dose stoppages and dose delays and dose adjustments within the temozolomide and olaparib arm, and then if you look at the investigator’s choice arm, six patients didn’t start, but this is an intention-to-treat trial. But what we found, unfortunately, was in an all-comer, unbiomarker-selected trial that the investigator’s choice actually did have a progression-free survival of about two months better than the olaparib and temozolomide arm. What’s going on right now is a lot of correlative work to try to determine why this trial was negative. Between case reports, personal experience, and worldwide experience, we know there is a subset of patients that this benefits within the uterine leiomyosarcoma patient population. We now need to actually figure out what to do and how to define these patients. One of the problems is that we needed a CLIA-certified test for RAD51 foci formation that may have helped this trial, but that didn’t exist at the time this trial was done. But we’re working very hard not to just walk away from these findings. But what we’ve learned is an all-comer trial wasn’t going to actually answer this question. We’re going to go back to the drawing board. We’re going to identify who these patients are, and then we will come back probably with a CLIA-stratified test and then do this trial again.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
