Presentation (CTI)

OBJECTIVES: To assess the acceptability of early endpoints in NICE technology appraisals (TAs) within the (neo)adjuvant and peri-adjuvant oncology settings. This investigation will evaluate approaches undertaken by companies and capture critiques from the Evidence Assessment Group (EAG) and NICE Committee. The significance of early endpoints will be examined alongside the impact of other uncertainties on the final recommendation.
METHODS: A targeted literature review was conducted using the NICE website. NICE TA Committee Papers and Final Draft Guidance (FDG) from 2019 to June 2025 were reviewed to capture the preferences of the EAG and the NICE Committee and to understand the role of early endpoints in NICE’s decision-making. A data extraction grid was developed, applying a grading system (major, moderate, minor) to the impact of each uncertainty discussed in the Committee Papers and FDG, including the use of early endpoints.
RESULTS: Twenty-three published appraisals met the inclusion criteria. The majority (18/23) pertained to adjuvant treatment, while two related to neoadjuvant, and three to peri-adjuvant treatment. In most cases (19/23), treatments were recommended for routine commissioning, with four reimbursed via the Cancer Drugs Fund (CDF). The most recent use of the CDF for (neo)adjuvant indications occurred in 2022, despite eleven appraisals having been published since. The extent to which companies performed correlative analyses of early endpoints with overall survival (OS) varied. Other common areas of uncertainty impacting decision making included the choice of model for survival extrapolation, and treatment effect waning assumptions.
CONCLUSIONS: Due to the relative immaturity of OS data, companies routinely predicted OS using early endpoints in the (neo)adjuvant and peri-adjuvant settings in NICE submissions. Despite the relationship between early endpoints and OS being identified as a major uncertainty in some cases, most therapies were still recommended for routine commissioning. Unmet need and cost-effectiveness remain the main drivers of recommendation.