OBJECTIVES: The value for money associated with pembrolizumab versus nivolumab has not been evaluated head-to-head. The objective of this analysis is to assess the cost-effectiveness of pembrolizumab versus nivolumab in advanced or metastatic NSCLC patients from a third-party payer’s perspective.

METHODS: We developed a three-state transition model [progression-free survival(PFS), post-progression survival, and death] based on the phase III KEYNOTE-042 trial. Cycle length was 3 weeks with a 10-year time horizon. Transition probabilities for pembrolizumab were obtained following individual patient data (IPD) reconstruction of the survival curves of the KEYNOTE-042 trial and parametric model fitting. The hazard ratios comparing pembrolizumab and nivolumab were obtained from on network meta-analysis and used as factors to derived transition probabilities for nivolumab. Utilities of NSCLC states were obtained from a real-world utility survey study. Direct medical costs were obtained from publicly available sources. A 3% annual discount rate for cost and outcomes was used for adjustment. The incremental cost-effectiveness ratio (ICER) and net monetary benefit (NMB)was estimated for pembrolizumab versus nivolumab. Sensitivity analyses were conducted to assess the model robustness.

RESULTS: Patients accrued 0.97 QALY at a cost of $263,480 with pembrolizumab while accruing 0.96 QALY at a cost of $259,835, resulting in an ICER of $358,127/QALY. NMB was estimated to be $-2,118 at a willingness to pay(WTP) threshold of $150,000. One-way sensitivity analysis has shown that the model is most sensitive to the acquisition costs for both medications and PFS hazard ratio with the ICER ranging from $-1,402,502 to $2,118,756 per QALY.

CONCLUSIONS: The results of this analysis show that pembrolizumab is not cost-effective compared to nivolumab, at a WTP of $150,000 in advanced or metastatic NSCLC patients.