Targeting resistant prostate cancer, with or without DNA repair defects, using the combination of ceralasertib (ATR inhibitor) and olaparib (the TRAP trial)

Abstract

Background: Men with metastatic, castration resistant prostate cancer (mCRPC) harboring DNA repair defects (̃20%) achieve a radiographic progression free survival of 7.4 months with PARP inhibitors (PARPi). Preclinical studies combining a PARPi (olaparib) and DNA damage checkpoint inhibitor (ATR inhibitor, ceralasertib) show synergy, providing the rationale to test this combination in men with mCRPC, including where single agent olaparib has been shown to be active. Methods: Two cohorts were accrued to a trial combining ceralasertib with olaparib in men a) with or b) without DNA repair defects. All patients progressed on ≥1 prior mCRPC therapy with no prior PARPi or platinum chemotherapy. The primary endpoint was disease response (confirmed PSA decline ≥50% and/or RECIST response), while disease progression was defined per Prostate Cancer Working Group 3 definition. Each cohort is analyzed independently for disease endpoints, while both groups were combined for toxicity assessments. Results: The 12 person DNA repair-deficient (DRDef) cohort allowed patients with germline BRCA2 loss (n = 4), somatic BRCA2 loss (n = 1) and ATM loss (n = 1 germline, n = 5 somatic and n = 1 somatic with unknown germline). 35 men without BRCA2/BRCA1 or ATM genomic loss were accrued to the DNA repair-proficient (DRPro) cohort. These men had next-generation sequencing (NGS) on contemporary biopsies (prior to enrolment without intervening therapy, 12), prior NGS on metastatic tissue (10), prior NGS on primary prostatic tissue (n = 3), or cell-free analyses (5). Five patients have incomplete cell-free analyses. At data cutoff (October 2021), in the DRDef cohort, the response rate by confirmed ≥50% PSA decline was 4/10 (40%) including 3 of 4 BRCA2 patients, and another is awaiting sufficient follow up; 1 of 6 ATM-deficient patients responded and another is awaiting sufficient follow up. All 4 DRDef responders remain on therapy (median of 8 months). For patients in the DRPro cohort who have completed therapy and response assessment (n = 21), 3 responded, one with a duration of 12 months, two with 6 months. An updated analysis will be presented. Conclusions: This analysis suggests potential activity of the doublet for DRDef (BRCA2 mainly) and DRPro mCRPC. Ongoing biomarker analysis (e.g. ATM IHC, contemporaneous cell-free DNA analysis rather than archived tissue) may help guide selection of patients most likely to benefit.