LAUSR

OBJECTIVE Nivolumab is the first drug to demonstrate a survival benefit for platinum-refractory recurrent or metastatic head and neck cancer. We performed a cost-utility analysis to assess the economic value of nivolumab as compared to alternative standard agents in this context. MATERIALS AND METHODS Using data from the CheckMate 141 trial, we constructed a Markov simulation model from the US payer's perspective to evaluate the cost-effectiveness of nivolumab compared to physician choice of either cetuximab, methotrexate or docetaxel. Alternative strategies considered included: single-agent cetuximab, methotrexate or docetaxel, or first testing for PD-L1 to select for nivolumab. Costs were extracted from Medicare and utilities from the literature and CheckMate. Probabilistic sensitivity analysis (PSA) was used to evaluate parameter uncertainty. $100,000/QALY was the primary threshold for cost-effectiveness. RESULTS When comparing nivolumab to the standard arm of CheckMate, nivolumab demonstrated an incremental cost-effectiveness ratio (ICER) of $140,672/QALY. When comparing standard therapies, methotrexate was the most cost-effective with similar results for docetaxel. Nivolumab was cost-effective compared to single-agent cetuximab (ICER $89,786/QALY). Treatment selection by PD-L1 immunohistochemistry did not markedly improve the cost-effectiveness of nivolumab. Factors likely to positively impact the cost-effectiveness of nivolumab include better baseline quality-of-life, poor tolerability of standard treatments and/or a lower cost of nivolumab. CONCLUSIONS Nivolumab is preferred to single-agent cetuximab but requires a willingness-to-pay of at least $150,000/QALY to be considered cost-effective when compared to docetaxel or methotrexate. Selection by PD-L1 does not markedly improve the cost-effectiveness of nivolumab. This informs patient selection and clinical care-path development.