ABSTRACT
BACKGROUND AND OBJECTIVES: Atezolizumab is an anti-programmed death ligand 1 (PD-L1) antibody that shows good safety and efficacy for patients with PD-L1-positive triple-negative breast cancer (TNBC). The cost of atezolizumab therapy is expensive, and its economic burden is an important problem. In this study, we evaluated the cost effectiveness of atezolizumab plus nab-paclitaxel therapy (AnP) compared with nab-paclitaxel monotherapy (nP) for PD-L1-positive TNBC under Japanese medical conditions and environments using a Markov model. METHODS: The medical information was collected from data published by the IMpassion130 trial. A Markov model was established to simulate the number of patients in each disease state after AnP or nP therapy during each time period. As indices for cost effectiveness, total cost, quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICER) were calculated. Probabilistic sensitivity analysis (PSA) was used to assess the uncertainty of the model using 10,000 Monte Carlo simulations with difference parameters. RESULTS: The QALYs for AnP treatment were longer than for nP treatment (1.12 vs 0.82 QALYs), but the total cost of AnP treatment was more expensive than that of nP treatment (¥11,070,143 vs ¥2,056,164). The ICER values comparing AnP treatment with nP treatment were ¥30,208,143/QALY. The ICER/QALY was more expensive than the willingness-to-pay (WTP) of ¥15,000,000 per QALY. To achieve a 50% cost-effective probability with a WTP threshold, the price of the atezolizumab should be reduced by 55.1%. CONCLUSIONS: AnP was not cost effective compared to nP for PD-L1-positive inoperable TNBC under the Japanese condition.