OBJECTIVES: This study assessed the CE profile of nivolumab versus surveillance with respect to duration of follow-up in CheckMate-274 trial’s efficacy and EQ-5D-3L data for the treatment of high-risk MIUC patients from a UK payer perspective.

METHODS: Patient level data corresponding to two successive database-locks (DBLs) from the trial with 11.0- and 31.6-months of minimum follow-up were used to populate a semi-Markov model consisting of disease-free (DF), local recurrence (LR), distant recurrence (DR), and death states. Expected costs and quality-adjusted life-years (QALY) were calculated for the intention-to-treat population over a 30-year time horizon. Costs were sourced from UK public sources and the utility scores were calculated from trial EQ-5D data. For the estimation of transitions from the DF state until year 5, upon which DF patients were assumed to be cured, reported DFS data from the CheckMate-274 trial (years 0-3) and the control arm from the EORTC-30994 trial (years 3-5) are used. Transitions from LR and DR states were informed by the CheckMate-274 data and literature in metastatic treatment of UC, respectively. Between the two DBLs, all trial-specific model inputs were updated without changing modeling assumptions.

RESULTS: The incremental cost-utility ratio (ICUR) of nivolumab versus surveillance marginally improved from £48,407/QALY to £45,200/QALY (Δ=7%) when using CheckMate-274 data from the latest DBL. The difference between incremental QALYs and costs across the arms was modest-to-negligible (Δ<1% in costs, Δ<10% in QALYs) between the two DBLs. Total (and DF) QALYs increased by 0.07 (and 0.08) for nivolumab and by 0.01 (and 0.01) for surveillance. The total costs decreased from £74,310 to £73,764 for nivolumab, and from £35,049 to £34,281 for surveillance.

CONCLUSIONS: The longer follow-up data from CheckMate-274 study had a marginal impact on the original nivolumab’s CE versus surveillance, confirming the robustness of its economic value for the adjuvant treatment of MIUC.