OBJECTIVES:

Many therapies have been approved as adjuvant or neoadjuvant treatments for early-stage cancer. However, economic evaluation of these therapies faces unique challenges such as overall survival (OS) immaturity from clinical trials. We aimed to review the Technology Appraisals (TAs) of adjuvant or neoadjuvant cancer therapies recently published by National Institute for Health and Care Excellence (NICE).

METHODS:

We searched TAs in the NICE website with search terms of “adjuvant” or “neoadjuvant”, status of “published”, and the published year of 2019-2022. The initial TA and the updated ones were counted as one TA. Published documents were reviewed and summarized with a focus on model structure, OS surrogacy, long-term treatment effect, and final recommendations.

RESULTS:

Thirteen TAs were identified: pembrolizumab (N=4) and nivolumab (N=3) as the most frequently evaluated therapies, breast cancer (N=5) and melanoma (N=3) as the top relevant diseases, and one in the neoadjuvant setting. Model structures varied regarding the number of health states, with four health states (recurrence-free, local recurrence, distant metastasis, death) being commonly considered. OS were immature in all TAs, including five where OS was not available. Evidence for the use of surrogate endpoints and validations of modeled long-term survivals were provided and examined in most TAs. In modeling long-term treatment effect beyond the follow-up period of the trial, assumptions of “Waning” (N=7) and “Cure” (N=7, in more recent TAs) were explored in the base-case and scenario analyses. All TAs received positive recommendations: nine for routine use, two initially entering Cancer Drugs Fund (CDF) then for routine use, and two currently available via CDF.

CONCLUSIONS:

Challenges such as OS immaturity persist in the economic evaluation of early-stage oncology. Modeling approaches justified by scientific evidence, efforts to examine uncertainty, and plausible cost effectiveness would help justify the value of these cancer therapies for patient access.