Background: Biomarker-based treatments have revolutionized treatment in ovarian cancer. Our objective was to examine the association of biomarker testing in ovarian cancer with patient and health system factors. Methods: We conducted a retrospective cohort study of ovarian cancer patients diagnosed from 2011-2021using the nationwide de-identified, electronic health record-derived Flatiron Health database. We used multi-level regression modeling to analyze the association of patient, sociodemographic, health system, and cancer factors with receipt of biomarker testing. We analyzed results overall and by biomarker: BRCA, homologous recombination deficiency (HRD) or genomic instability score (GIS). Results: Of 8,519 patients with ovarian cancer, 55.4% (95%CI 54.3-56.4) had undergone any biomarker testing. 55.3% (95%CI 54.3-56.5) had undergone BRCA testing and 6.1% HRD/GIS (95%CI 5.6-6.6). Medicare insurance, older age, ECOG of 2+, and living in Puerto Rico was associated with lower likelihood of biomarker testing, while recurrence or a later year of diagnosis was associated with greater likelihood of biomarker testing. Findings were similar for BRCA and HRD/GIS. Conclusions: Only 55% of patients underwent biomarker testing. Insurance type, ECOG status, region, and recurrence were predictors of ever undergoing biomarker testing. Interventions to improve Medicare coverage and care of medically complex patients may improve biomarker testing and use of directed therapies in ovarian cancer.

Analyses also adjusted for stage at diagnosis, BMI, and histology. Risk ratios are for associations at diagnosis (e.g., insurance at diagnosis).N/A = Not applicable* P-value < 0.05 ** P-value < 0.001 (Bonferroni correction applied)