Background: Disparities in access and use of biomarker testing have been reported in patients (pts) with advanced ovarian cancer (AOC) that may result in worse outcomes. This study aimed to characterize socioeconomic status (SES), racial, and US regional differences in the use of biomarker testing and first-line maintenance (1LM) therapy among pts with AOC. Methods: We retrospectively included pts ≥18 years old who were diagnosed with AOC from Jan 2019–Mar 2022 from the nationwide electronic health record-derived Flatiron Health database. Use of biomarker testing (BRCA mutation [BRCAm] onlyand homologous recombination deficiency [HRD] not including BRCAm tests) and 1LM therapy (PARP inhibitor ± bevacizumab, bevacizumab monotherapy, or other 1LM) were evaluated by SES (scored 1 [low]–5 [high]), race, and US region, using descriptive statistics and chi-squared testing. Results: Of 907 pts included, 91% were from community clinics; median age was 68 years. Overall, 87% (n=787/907) of all pts were BRCAm tested (test type: 31% [n=247/787] germline, 52% [n=411/787] somatic, and 16% [n= 129/787] unknown testing type); 28% (n=255/907) were HRD tested. Of pts who were germline BRCAwt, 22% (n=45/201) underwent additional HRD testing. Of all non-BRCAm pts (defined as germline or somatic BRCAwt or BRCAunknown pts), 28% (218/772) underwent HRD testing; they were more likely to have had surgery (91% vs 71%) vs those who were not HRD tested. Pts with SES score 1 and Black pts had numerically lower testing rates vs the overall population and the relationship between BRCAm testing rates and SES was significant (P <0.05) (Table). Of all biomarker-tested pts, 49% (n=387/788) received 1LM therapy (71% [n=96/135] of BRCAm, 68% [n=54/80] of BRCAwt/HRD-positive, and 47% [n=55/118] of HRD-negative pts). Among non-biomarker-tested pts, 82% (n=98/119) did not receive 1LM therapy. 1LM therapy use was comparable across SES, race, and US region. Conclusions: While most pts with AOC underwent BRCAm testing, some differences in biomarker-testing rates by SES and race were observed in our study. Further exploration of these differences is required to better understand potential disparities in testing, treatment, and outcomes.

*P <0.05. ^aSES was missing/unknown for 145 patients.