OBJECTIVES: Health-state utility values in oncology economic models are traditionally based on disease progression status, which may not always accurately reflect deterioration in patients' health-related quality-of-life (HRQoL). Alternatively, to account for the decline in HRQoL towards end-of-life, utilities can be based on time-to-death (TTD). We explored the use of the TTD approach and HTA's acceptance or criticism through a review of HTA appraisals in advanced non-small cell lung cancer (aNSCLC).

METHODS: HTA reports from NICE and CADTH regarding first- and second-line treatments for aNSCLC from January 2015 to November 2023 were reviewed. The inclusion criterion was the use of TTD utilities in the economic models. Only NICE technology appraisals (TAs) were described due to limited TTD details in the CADTH reports.

RESULTS: Out of sixty-two TAs identified, only five (8%) used TTD utilities, including assessments of sotorasib, nivolumab+ipilimumab, and three for pembrolizumab.

Statistical models of EQ-5D data collected in relevant clinical trials of aNSCLC, supported by clinical experts, demonstrated that TTD was the strongest predictor of HRQoL.

For the base case, three TAs used TTD utilities, one used progression-based utilities, and one combined both. The incremental cost-effectiveness ratio (ICER) was 10-15% lower with TTD utilities [£29,133-£43,660] compared to progression-based utilities [£32,150- £47,208]. One of the contributory factors could be the greater difference in mean overall survival compared to the mean progression free survival. The extended survival may have contributed to higher QALYs observed in TTD scenarios.

NICE acknowledged the value and uncertainties of both TTD and progression-based methods, requiring robust justification for TTD utilities and emphasizing the importance of both approaches.

CONCLUSIONS: TTD based utilities are not widely used in HTA submissions for aNSCLC. Use of TTD based over progression-based utility can impact ICER, independently from the treatment effect.