Presentation (CTI)

OBJECTIVES: Immature survival data (median overall survival [OS] not reached) is prevalent in oncology health technology assessments. Tai et al. (2021) found 41% of NICE cancer appraisals relied on immature data and showed the incremental cost-effectiveness ratio (ICER) halved with mature evidence for olaparib in ovarian cancer. This study evaluates data maturity impact on nivolumab versus everolimus in previously treated advanced renal cell carcinoma (NICE TA417).
METHODS: TA417 was replicated using a three-state partitioned survival model (progression-free survival [PFS], progressed disease, death). OS and PFS from CheckMate 025 were available as an immature interim cut (2015) and mature five-year follow-up (2020). OS used employed dependent generalised gamma (immature; replicating TA417) and independent log-logistic (nivolumab)/generalised gamma (everolimus) (mature); PFS used independent 2-knot spline odds throughout. A lifetime horizon, NHS perspective, and 3.5% discount rate were applied.
RESULTS: Immature results aligned with the replicated TA417 submission. Comparing immature with mature results, incremental quality-adjusted life years (QALYs) rose from 0.63 to 0.99 (+57%) and life years from 0.74 to 1.25 (+69%); costs rose from £55,255 to £68,957 (+24%). The ICER fell 21%, from £88,335 to £69,925 per QALY gained, driven by greater long-term nivolumab OS benefits. Probabilistic sensitivity analyses showed reduced uncertainty: all mature-data iterations indicated nivolumab as more effective, whereas immature-data iterations included everolimus-dominated scenarios. No mature-data one-way sensitivity scenario reached the immature ICER.
CONCLUSIONS: Immature survival data underestimated long-term nivolumab benefits, inflated cost-effectiveness estimates, and increased ICER uncertainty. As a single-indication case study using list-price drug costs, the magnitude and direction of impact may differ elsewhere. The reduction in decision uncertainty with more mature data highlights the importance of embedding routine reassessment into NICE appraisals based on immature evidence. While managed access arrangements (e.g. Cancer Drugs Fund) help mitigate uncertainty and maintain timely patient access, more consistent reassessment could further reduce suboptimal reimbursement decision risks.