Abstract B094: A combined RAD51 and phosphorylated replication protein assay predict response to platinum chemotherapy in high-grade serous ovarian cancer
Amanda J. Compadre,Carmen Sandoval,Olivia Graham,Jeimmy Rodriguez,Lindsay M. Kuroki,Carolyn K. McCourt,Andrea Hagemann,Premal H. Thaker,L. Stewart Massad,Matthew A. Powell,David G. Mutch,Ian Hagemann,Katherine Fuh,Lulu Sun,Peinan Zhao,Dineo Khabele,Priyanka Verma,Elena Lomonosova,Mary M. Mullen
DOI: https://doi.org/10.1158/1538-7445.ovarian23-b094
IF: 11.2
2024-03-05
Cancer Research
Abstract:Introduction: An effective assay is needed to predict response to platinum chemotherapy in high-grade serous ovarian cancer (HGSOC). We previously showed that patients with RAD51-Low (functionally deficient HR) tumors responded better to platinum chemotherapy than patients with RAD51-High tumors (RR 5.28, P<0.001). This RAD51 foci assay predicted platinum sensitivity with 100% specificity and 100% positive predictive value, but it had 40% sensitivity due to some RAD51-High tumors also being platinum sensitive. To improve the assay, we evaluated whether phosphorylated replication protein A (pRPA) foci, a functional marker of replication stress, could yield a more sensitive assay to predict platinum sensitivity when added to the RAD51 assay. Methods: The pRPA antibody was optimized for immunofluorescence and validated by staining two HGSOC cell lines treated with hydroxyurea in which RPA2 was knocked down. Formalin-fixed paraffin-embedded HGSOC tumor samples (discovery cohort n=31, validation cohort n=51) had all previously been scored as RAD51-High or RAD51-Low using immunofluorescence. Using immunofluorescence, samples were classified as pRPA-Low ( 20% of cells had ≥2 pRPA foci). Associations were calculated between RAD51 and pRPA scores, platinum chemotherapy response, and survival. Results: In the discovery cohort, patients with RAD51-High, pRPA-High tumors had significantly higher rates of pCR (50% vs. 6.7%, P<0.001; RR 11.25, P<0.001) and longer PFS (15.2 vs. 7.8 months, P=0.006) than patients with RAD51-High, pRPA-Low tumors. All RAD51-High, pRPA-High tumors (n=4) were sensitive to platinum chemotherapy. When RAD51-Low (regardless of pRPA score) and RAD51-High, pRPA-High tumors were combined into one group, it predicted platinum sensitivity with 100% PPV, 64% sensitivity, and 100% specificity. In the validation cohort, all RAD51-High, pRPA-High tumors were platinum sensitive (n=9). Tumors that were RAD51-Low or RAD51-High, pRPA-High were more likely to be platinum sensitive (RR ∞, P<0.001) than RAD51-High, pRPA-Low tumors. The combined assay predicted platinum sensitivity with 100%, 54.3% sensitivity, and 100% PPV. Moreover, patients in the combined group of RAD51-Low (regardless of pRPA score) or RAD51-High, pRPA-High had longer PFS (33.0 vs. 10.0 months, P=0.002) and OS (135.2 vs. 85.1 months, P=0.01) than those with RAD51-High, pRPA-Low tumors. Exploratory analysis of patients who received PARP inhibitors (n=14) demonstrated improved PFS (60.0 vs 6.8 months, P=0.001) and OS (134.4 vs 85.1 months, P=0.004) in RAD51-Low or RAD51-High, pRPA-High tumors compared to RAD51-High, pRPA-Low tumors. Conclusions: In this study, RAD51-High, pRPA-High tumors were as responsive to platinum chemotherapy as RAD51-Low tumors. Patients with these tumors had longer PFS and OS than patients with RAD51-High, pRPA-Low tumors. Combined RAD51 and pRPA foci predicted platinum sensitivity with 58% sensitivity, 100% specificity, and 100% PPV. This biomarker should be evaluated in prospective clinical trials. Citation Format: Amanda J. Compadre, Carmen Sandoval, Olivia Graham, Jeimmy Rodriguez, Lindsay M. Kuroki, Carolyn K. McCourt, Andrea Hagemann, Premal H. Thaker, L. Stewart Massad, Matthew A. Powell, David G. Mutch, Ian Hagemann, Katherine Fuh, Lulu Sun, Peinan Zhao, Dineo Khabele, Priyanka Verma, Elena Lomonosova, Mary M. Mullen. A combined RAD51 and phosphorylated replication protein assay predict response to platinum chemotherapy in high-grade serous ovarian cancer [abstract]. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_2) nr B094.
oncology