OBJECTIVES

The aim of this analysis was to assess the cost-effectiveness of nivolumab as an adjuvant treatment for adult patients with melanoma with involvement of lymph nodes or metastatic disease, from a healthcare system perspective in Switzerland.

METHODS

: A Markov de novo global cost-effectiveness model was adapted to the Swiss setting. The analysis took a health-care payer perspective, over a 60-year time horizon. Nivolumab (240mg, every 2 weeks) was compared to observation, standard of care in Switzerland.

Recurrence-free survival (RFS) for nivolumab and observation was estimated through an indirect treatment comparison based on patient-level data from CheckMate 238 and CA184-029 trials, since both include ipilimumab as comparator. Overall survival (OS) was modelled using a RFS/OS correlation equation supported by published literature.

Drug acquisition, administration, monitoring, subsequent therapy and adverse events costs were sourced from published prices and clinical expert input. Utility values based on French EQ-5D tariffs were obtained from CheckMate 238 collected data. An annual discount rate of 3% was applied for both costs and effects. Input parameters’ uncertainty was tested in sensitivity analyses. Incremental cost-utility ratio (ICUR) was compared to a willingness-to-pay (WTP) of 100,000 CHF (Swiss Francs) per QALY gained.

RESULTS : In the base case, an incremental effect of 3.63 QALYs and incremental costs of CHF 64,114 were estimated with nivolumab, resulting in an ICUR of CHF 17,667 per QALY gained compared to observation. The parameters with the highest impact on the results were the effects’ discount rate and the HR of OS vs Post-recurrence survival (PRS) in the placebo arm. Nivolumab has approximately 53.2%, 74.8% and 99.0% probability of being cost-effective at a WTP threshold of CHF 18,000, 20,000 and 30,000 respectively.

CONCLUSIONS

: Compared to observation, nivolumab was cost-effective as adjuvant treatment of melanoma in Switzerland, considering a WTP threshold of CHF 100,000 per QALY.