Key points are not available for this paper at this time.
Introduction/Background The predictive value of HRD test for the efficacy of first-line (1L) maintenance therapy with PARPi in newly diagnosed advanced ovarian cancer (OC) patients has been confirmed by randomized controlled trials. However, research on the predictive value of HRD in real-world has been limited. We have developed MEDx HRDetectCDx test and validated its predictive value in a prospective, multicenter real-world study in China (NCT04986371). Methodology The MEDx HRDetectCDx test was performed by capture-based targeted sequencing of ~42,000 genome-wide single nucleotide polymorphisms specific to Chinese population at ±200x coverage and coding exons of 85 genes at ±400x coverage. The HRD score was calculated by analyzing LOH, TAI and LST. A score ≥43 (cut-off) is considered positive. The analytical performance of the HRDetectCDx was evaluated by comparing to the Myriad myChoice (Myriad test). The predictive impact of the HRD score on niraparib efficacy was evaluated in the prospective RENI-1 study, which enrolled OC patients from 22 centers in China who received niraparib as 1L maintenance therapy between 2021 and 2022. The date of PFS data cut-off was 18 October 2023. Results The positive percentage agreement between the HRDetectCDx panel and the Myriad test is 89.5%, the negative percentage agreement is 93.3% and the overall percent agreement is 91.2%. 227 eligible patients were enrolled and 122 patients underwent HRDetectCDx test with 92 evaluable for HRD status. 20 (16.4%) had BRCAm and 53 (43.4%) had HRD. Median PFS was not reached (NR) and 14.4 months in the HRD and HRp groups (HR 0.51; 95% CI 0.27–0.97), respectively. In the HRD group, median PFS was NR for BRCAm and 18.1months for BRCAwt/HRD (HR 0.46; 95% CI 0.15–1.39). Conclusion The MEDx HRDetectCDx demonstrated high consistency with Myriad test and was prospectively validated for predictive value in 1L maintenance therapy with niraparib in the real-world setting. Disclosures All authors have declared no conflicts of interest.
Zuo et al. (Fri,) studied this question.