Mod Pathol. 2024 Feb 8:100445. doi: 10.1016/j.modpat.2024.100445. Online ahead of print.

ABSTRACT

HRD assays are an important element of personalized oncology in ovarian carcinomas, but the optimal tissue requirements for these complex molecular assays remain unclear. As a result, a considerable percentage of assays are not successful, leading to sub-optimal diagnoses for these patients. In this study, we have systematically analyzed tumor and tissue parameters for HRD analysis in a large cohort of real-world cancer samples. The aim of this study is to give recommendations for pathologists and gynecological oncologists for selection of tissue samples to maximize the success rate of HRD analyses. Tumor samples from 2702 patients were sent to the Institute of Pathology of the Philipps-University Marburg between October 2020 and September 2022, 2654 were analyzed using the Myriad MyChoice CDx HRD+ assay. 2396 (90.3% of 2654) were successfully tested, of which 984 (41.1% of 2396) were HRD positive and 1412 (58.9%) HRD negative. 363 (15.2%) of 2396 samples were BRCA1/2-mutated, 27 samples had a BRCA mutation and a GIS<42. 22 (0.9%) failed GIS measurement, but displayed a BRCA1/2 mutation. BRCA1/2-mutated samples showed significantly (p<0.0001) higher GIS values than those with a wild type BRCA1/2-status. Tumor cell content, tumor area, and histology significantly (p<0.0001) affected the probability of successfully analyzing a sample. Based on a systematic analysis of tumor cell content and tumor area, we recommend selecting patient high-grade serous ovarian cancer samples that display a tumor cell content ≥30% and a tumor area ≥0.5 cm² (on their H&E) for HRD testing, to allow for optimal chances of a successful analysis and conclusive results. Considering histological and sample conditions, success rates of up to 98% can be achieved. Our comprehensive evaluation contributes to further standardization of recommendations on HRD testing in ovarian cancer, which will have a large impact on personalized therapeutic strategies in this highly aggressive tumor type.

PMID:38341130 | DOI:10.1016/j.modpat.2024.100445