Proof is in the PARP.
Top Line: Do PARP inhibitors improve survival in men with metastatic prostate cancer and DNA repair defects?
The Study: Initial results of the PROfound trial showed that men with metastatic castration resistant prostate cancer (mCRPC) and associated defects in DNA repair genes benefited from olaparib (PARP inhibitor) after progressing on second-line androgen axis inhibition. Overall survival results were immature at that time, but they aren’t anymore. As a reminder, patients in PROfound were put in two different cohorts: cohort A included those with a BRCA1, BRCA1, or ATM alteration while cohort B included alterations in a broader list of DNA repair genes. In cohort A, olaparib significantly prolonged median overall survival 14.7→ 19.1 months (HR 0.69). The difference in the overall population was encouraging (14→ 17.3 months), but it was driven by the BRCA/ATM group as there was no significant improvement among cohort B. Two-thirds of patients in the control arm crossed-over to receive olaparib. Accounting for that cross-over demonstrated an even bigger magnitude of benefit in cohort A (HR 0.42).
TBL: PARP inhibition with olaparib prolongs survival in men with mCRPC and BRCA/ATM mutations after progression on abiraterone or enzalutamide. | Hussain, N Engl J Med 2020 (ESMO 2020)
We should not applaud this trial. It has many shortcomings, as brought up by Dr. Prasad.
ReplyDeletehttps://soundcloud.com/plenarysession/ep252#t=3:04
[Plenary Session 2.52]: Open-label design. Prior taxanes are not mandated, and only 2/3 received it (mostly docetaxel, not cabazitaxel). Crossover to olaparib is allowed, even though taxane is likely the best bet. This is Crossover Errors 101. Radiographic PFS is the primary endpoint, which is inappropriate. At best, one in ten patients will be eligible for this in the clinic, but will they benefit from this? Around 40-45% received enza, 40-45% received abi, while 20% received both drugs before enrollment. It is inappropriate to give more enza or more abi despite progression on BOTH drugs before! Prasad acknowledges that it is likely best to sequence abi and enza in alphabetical order. PSA response rate was only 8% in the control arm, suggesting many patients likely switched from enza to abi. Classic sub-par control arm, as seen in [POLO]. We do not know about the response in the DDR aberrations cohort, as it is combined with BRCA1/2 and ATM mutations. This is a very cheeky way to report. The DDR aberration cohort should be reported independently of the BRCA1/2 and ATMmt cohort. Why does cohort B have an equivalent PFS and a slightly better OS than the control arm (Fig S4)? Over 80% of patients on the control arm crossed over to olaparib, while the experimental arm went on to receive a taxane! Salvage taxanes or platinum agents are accessible earlier on in the experimental arm, which is likely responsible for the difference in OS.