OBJECTIVES : The quality of life of patients with ES-SCLC is known to decline over time, but limited literature is available on EQ-5D derived utility values. The phase III CASPIAN trial collected EQ-5D-5L data directly from participants.

The objective is to derive time to death (TTD) utility values from EQ-5D-5L results by progression health state and to estimate EQ-5D-3L based utilities from CASPIAN EQ-5D-5L and EORTC QLQ-C30 data using mapping algorithms.

METHODS : United Kingdom (UK) and United States (US) tariffs were applied to EQ-5D-5L data derived from the combined durvalumab + chemotherapy and chemotherapy arms in CASPIAN to obtain utility estimates for the UK and the US. Two algorithms were used to derive EQ-5D-3L and EORTC utilities: the ‘3L crosswalk’ approach to map EQ-5D-5L to 3L values, and the ‘Longworth’ approach to map EORTC QLQ-C30 to EQ-5D-3L values. Mixed effects/hierarchical regression models were fitted to investigate the effect of treatment, progression state, time in study and time to death.

RESULTS : Utility declined with TTD. Patients with a life expectancy of at least 60 days had a mean utility of 0.85, 0.79, and 0.77 for the UK 5L Direct, 3L Crosswalk, and Longworth approaches respectively, and 0.84 for the US 3L crosswalk. When life expectancy was between 30 and 59 days, mean utilities were 0.74, 0.68, and 0.67 for the UK 5L Direct, 3L Crosswalk, and Longworth approaches, respectively, and 0.75 for the 3L US crosswalk. Treatment effect was not significant in the fitted univariate and multivariate regression models explaining utility. However, time to death categories were associated with significant effects on estimated utilities.

CONCLUSIONS : This study provides utility estimates to inform economic evaluations in ES-SCLC. Declines of utility estimates as TTD decreased were noted across all algorithms. EQ-5D-5L tended to be associated with higher utility estimates.