Abstract / Summary
ABSTRACT Background Homologous recombination repair (HRR) testing is increasingly relevant in prostate cancer because it identifies patients who may benefit from biomarker‐guided treatment strategies, including poly(ADP‐ribose) polymerase inhibitors (PARPi). However, real‐world implementation of HRR testing remains variable across countries and clinical settings, and the extent to which current testing pathways capture actionable alterations is not well defined. Methods ARON‐3 is a multinational, multicentre, retrospective real‐world study of patients with prostate cancer. In this subanalysis, we assessed genomic testing uptake within the overall ARON‐3 cohort and characterised HRR testing pathways and detected pathogenic alterations among patients with available genomic results. Genomic testing included germline, somatic tumour, and/or circulating tumour DNA (ctDNA) assays performed in routine practice across participating centres. Multivariable logistic regression was used to identify clinicopathologic factors associated with undergoing genomic HRR testing. Results Among 2,192 patients included in ARON‐3, genomic testing results were available for 719 (33%). Among tested patients, 660 (92%) underwent germline testing and 610 (85%) somatic testing; 549 (77%) underwent both, whereas 109 (15%) had germline‐only and 61 (8%) somatic‐only testing. In multivariable analysis, metastatic castration‐resistant prostate cancer, high‐volume disease, and Gleason score > 8 were independently associated with higher odds of testing, while age ≥ 70 years was associated with lower odds of testing. Overall, 238 tested patients (33%) harboured pathogenic HRR alterations, corresponding to 11% of the entire ARON‐3 cohort. Somatic pathogenic alterations were more frequent than germline alterations. BRCA2 was the most frequently altered gene, while non‐BRCA1/2 alterations were also detected, particularly in centres using broader multigene panels. Most centres relied on BRCA1/2 ‐focused assays, and the use of liquid biopsy remained limited. Conclusions This multinational real‐world analysis demonstrates substantial heterogeneity in HRR testing uptake and testing pathways across participating ARON‐3 centres. Testing was more commonly performed in patients with more advanced disease features, suggesting that clinically relevant subgroups may still be differentially selected for genomic assessment. Broader and more standardised implementation of comprehensive HRR profiling may improve the identification of actionable alterations and strengthen precision medicine in prostate cancer.