Variation in chemotherapy delivery among patients with advanced epithelial ovarian cancer in the Netherlands and its association with survival: a nationwide cohort study.

OBJECTIVE: Population-based information on variation in the delivery of first-line chemotherapy for epithelial ovarian cancer is scarce. This study aimed to evaluate chemotherapy delivery, reasons for chemotherapy modifications, and their associations with overall survival. METHODS: Patients with advanced-stage epithelial ovarian cancer diagnosed from January 2015 through December 2021 were identified in the Netherlands Cancer Registry. We included patients who underwent cytoreductive surgery with platinum- and taxane-based chemotherapy and categorized them as receiving chemotherapy without or with modifications (dose reduction, interruption, and/or fewer cycles). We assessed reasons for modifications and used Kaplan-Meier curves and Cox proportional hazards models to analyze overall survival. RESULTS: Among 3687 patients, 54% underwent chemotherapy modifications. Dose reduction (38%) was most common, followed by interruption (24%) and fewer cycles (9%). Patients with modifications more often had poorer performance scores, higher comorbidity indices, and primary cytoreductive surgery. Neurotoxicity and hematologic toxicity were the primary reasons for platinum (33% and 37%) and taxane (47% and 35%) modifications. Dose reduction and interruption were not associated with survival. Fewer cycles were associated with lower 5-year overall survival (32% [95% confidence interval 26% to 38%] vs 36% [95% confidence interval 34% to 38%]) after multi-variable adjustment (hazard ratio 1.36, 95% confidence interval 1.17 to 1.59). CONCLUSIONS: A substantial proportion of Dutch patients with advanced-stage epithelial ovarian cancer undergo chemotherapy modifications. Dose reduction and chemotherapy interruption did not appear to be associated with overall survival, warranting prospective studies. A reduction in the number of chemotherapy cycles was negatively associated with overall survival, possibly reflecting underlying treatment ineffectiveness. However, this association could also be explained by residual confounding (eg, underlying disease severity or treatment response not fully captured in our data) and warrants further confirmation.