Abstract Background: PARP inhibitors (PARPi) improve outcomes in early and advanced HER2-negative (HER2-) breast cancer patients (pts) harboring a germline BRCA1/2 (gBRCA) pathogenic variants (PVs). However, gBRCA alone does not capture all tumors with homologous recombination deficiency (HRD), such as those with epigenetic silencing or other homologous recombination repair (HRR) gene alterations. We previously developed a functional HRD biomarker based on tumor RAD51 nuclear foci, validated retrospectively in early-stage BC (EBC) pts (GeparSixto and GeparOla trials). To date, no prospective study has assessed the predictive value of the RAD51 test for PARPi sensitivity. Methods: SOLTI-RADIOLA trial (NCT0534041) is a multicenter, open-label, two cohorts phase II study in pts with HER2- metastatic BC (mBC) with ≤ 2 prior lines of chemotherapy. Cohort 1(C1) included pts with known germline or somatic PVs in BRCA1/2, PALB2, or RAD51C/D. Cohort 2 (C2) included pts without/unknown HRR PVs but classified as HRD based on RAD51. RAD51 test was performed on pre-treatment tumor samples. HRD was defined as ≤10% RAD51-positive cells (RAD51-low). All pts received olaparib 300 mg BID until progression or unacceptable toxicity. Primary endpoint was overall response rate (ORR) by RECIST v1.1 in C1. Secondary endpoints included progression-free survival (PFS), clinical benefit rate (CBR) and safety. The study was designed to have 90% power to detect differences in the ORR between RAD51-low and -high tumors in C1. Results: From April 2022 to June 2024, 86 pts were screened in C1 and 278 in C2 (RAD51-low: C1- 45/67 67.2%; C2 - 17/201 8.5%), and 65 were enrolled (C1: n=55; C2: n=10). The median follow-up time was 6.8 months (mo). At the data cut-off, 6 pts (9.2%) remained on treatment. Most pts were female (96.9%), post-menopausal (74.6%), had HR-positive disease (58.5%), visceral metastases (84.3%) and received previous systemic treatment for mBC (89.2%). In C1, 96.3% had germline PVs (BRCA2: 43.6%, BRCA1: 40.0%, PALB2: 14.6%, RAD51D: 1.8%) and 3.7% had somatic PVs. Among C1 pts, 74.5% (n=41) were RAD51-low and 25.5% (n=14) RAD51-high. In C1, the ORR was significantly higher in RAD51-low vs RAD51-high tumors (68.3% vs 21.4%; OR = 7.90, 95%CI 2.07-39.54; p=0.002). PFS was longer in the RAD51-low group (7.1 vs 5.6 mo; hazard ratio = 0.51, 95% CI 0.26-0.98). In C2, the ORR was 10.0%, the 12-mo PFS rate was 30.0% (95%CI 11.6 - 77.3). Treatment-related toxicity led to dose reduction in 12.3% of pts. Conclusion: Our results demonstrate that the RAD51 assay enriches for PARPi response in pts with HRR PVs. These findings are especially relevant in EBC, where HRD is more frequent, supporting broader and more precise use of PARPi-based therapies. Citation Format: I. Pimentel, T. Pascual, L. Lema, P. Tolosa, B. Adamo, M. T. Martínez, I. Blancas, J. Salvador Bofill, J. Ponce, G. Viñas, M. Perez-Lopez, I. Teruel, M. González, M. Tapia Céspedes, M. Legerén, M. Cejuela Solís, I. Lozano Cubo, X. González Farré, J. M. Ferrero-Cafiero, A. Llop-Guevara, G. Villacampa, M. Paes Dias, A. Prat, V. Serra, J. Balmaña. Predicting olaparib sensitivity in patients with metastatic HER2-negative breast cancer with BRCA1/2, PALB2, RAD51C/D mutations according to their homologous recombination status by the RAD51 test: Primary analysis of the RADIOLA phase II trial abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr RF5-05.
Pimentel et al. (Tue,) studied this question.