Estimated Acquisition Cost Differences between Fixed and Simulated Body Weight-Based Nivolumab Dosing Using Real-World Claims Data in Japan

BACKGROUND: Nivolumab was initially approved with body weight-based dosing; however, fixed-dose regimens were later adopted based on pharmacokinetic evidence supporting comparable benefit-risk profiles. Nevertheless, the acquisition cost impact of fixed dosing relative to a body weight-based counterfactual remains unclear in Japanese real-world data. OBJECTIVES: The purpose of this study was to estimate drug acquisition cost differences at the index administration by comparing observed fixed-dose quantities with a simulated body weight-based counterfactual using nationwide Japanese claims data. METHODS: We retrospectively analyzed nationwide Japanese claims data. Fixed-dose regimens of 80, 240, 360, and 480 mg were mapped to corresponding body weight-based regimens of 1, 3, 4.5, and 6 mg/kg, respectively, and patient-specific body weight-based doses were calculated using recorded body weight. The index administration was the first nivolumab administration observed for each patient during the database observation period. Costs were estimated at the index administration using official Japanese drug prices. Sensitivity analyses examined alternative body weight measurement windows and simulated measurement error. Bootstrap resampling was used to estimate uncertainty in the median cost difference. RESULTS: Among 3243 patients, estimated drug acquisition costs based on observed fixed-dose quantities were higher than those calculated for the simulated body weight-based counterfactual at the index administration. The median cost difference was 89,540 Japanese yen (bootstrap 95% confidence interval 86,945-92,135 Japanese yen). Median costs were 311,443 and 251,750 Japanese yen for observed fixed-dose quantities and the simulated body-weight-based counterfactual, respectively. The direction of the estimated cost difference remained consistent across sensitivity analyses, although its magnitude varied across simulated body weight measurement-error scenarios. CONCLUSIONS: Using real-world Japanese claims data and official drug prices, this study estimated higher acquisition costs based on observed fixed-dose quantities than those calculated for a simulated body-weight-based counterfactual at the index administration. The body-weight-based comparator was simulated for the same patients and did not represent an observed treatment group.