Expert perspectives on how PARP inhibition will fit into the real-world management of patients with metastatic castration-resistant prostate cancer.
Evan Y. Yu, MD: I want to press on that point that you just brought up. That's the big discussion most provocative point right now. I want to know what everybody thinks about it here, on selected patients. The data. We saw it in the initial study. We saw it again in Propel. It’s 2 studies now. What do you all think of it? How does it work? Does it work? The data seems to look like it's doing something.
Tanya B. Dorff, MD: We know PARP inhibitors have other functions and they do impact downstream expression of antigen regulated genes. There's a scientific rationale. The data aren't strong enough and even more importantly, most of our patients who have CRPC will have been exposed to an AR-targeted agent upfront. And then do they fit into the dataset? It's going to be hard to apply those data and I more studies are needed.
Evan Y. Yu, MD: Let me just ask this practical question for the group then. When you say more studies are needed. I heard Pedro say more studies are needed. Is there anything from the current study, from the Propel that we're waiting for that might come out that we'll say I'm good. I don't need another study?
Andrew J. Armstrong, MD, MSc: Let me just point out that the Propel was published on Friday in the Journal medicine evidence. You can go to that. It's an online journal and it's going to have all the breakdowns of the HRD positive and negative.
Evan Y. Yu, MD: They're that at ASCO [American Society of Clinical Oncology] we don't have time to dissect data.
Andrew J. Armstrong, MD, MSc: There's going to be updates for Propel for survival in the next 6 months. I am curious to know if that 5 month improvement in HRD negative subset translates to a survival benefit eventually. We would all like to know that because we're not sure that RPFS is a surrogate yet in this setting.
Evan Y. Yu, MD: Let's say it does. Is it a done deal?
Andrew J. Armstrong, MD, MSc: There are some patients that do quite well with a monotherapy. With an AR inhibitor. Five-year survival, 5 year progression-free survival even. We have models. Prognostic models, monograms that can identify men that do well with monotherapy. I don't want to give a PARP inhibitor to every single patient. But there's an incremental increasing benefit going from BRCA-2 to BRCA-1 to then the other HRD genes and then to everybody else. It would be important to identify in that everybody else group, who's got the genomic scar. Who's got HRD. Who's got some AR co-factor that might make PARP work better and get biomarkers to figure out who needs the combination.
Evan Y. Yu, MD: The thing that none of these studies prove is that combination is better than sequencing.
Pedro C. Barata, MD, MSc: That's the kind of study that ultimately to your point, if we see survival in that group we also have survival for the sequencing monotherapy in the biomarker positive patients. The biomarker negative maybe that would help answer. But even the biomarker positive we might need the study about sequencing. Or combination?
Mary-Ellen Taplin, MD: Andy, you're an expert on blood biopsies. I believe in Propel they allow liquid biopsies. Could the non-mutated group, could they have just missed a significant proportion of the league of mutated patients because of deletion?
Andrew J. Armstrong, MD, MSc: I'm glad you asked that. In Propel, everybody who had an attempt at a foundation solid biopsy and a foundation plasma liquid biopsy. There are patients where the archival tumor tissue didn't yield a valuable tumor. About 30% didn't have valuable tumor tissue but 70% did. Of the plasma, you have some patients that have a low tumor fraction. They're not considered valuable, but a smaller fraction is. When you combine the 2, you have a much smaller number that are not valuable by either one. Despite that, there could have been a few BRCA patients that have homologous homozygous deficiency like deletions that are missed because there's so much wild type DNA there. You just can't pick up a copy law very well. But that's going to be probably less than 1 or 2%. I don't think that's driving the wild type outcomes.
Transcript edited for clarity.