Use of External Control Arms in Reimbursement Submissions: A Review of CAR-T Appraisals by NICE and CDA

OBJECTIVES: Chimeric antigen receptor (CAR)-T cell therapies, targeting rare diseases with high unmet need, often cannot be ethically or practically evaluated through randomized controlled trials (RCTs). In these instances, an external control arm (ECA) can provide context for single-arm trial evidence. This study investigated the use of ECAs in health technology appraisals (HTAs) of CAR-T cell therapies by the National Institute for Health and Care Excellence (NICE) in the UK and Canada’s Drug Agency (CDA).

METHODS: We reviewed all CAR-T cell therapy appraisals by NICE and CDA, published as of June 2024 and based on single-arm trial and ECA evidence, focusing on ECA construction methods, economic modelling approaches, agency critiques, and reimbursement decisions.

RESULTS: Twelve appraisals were reviewed (NICE: 5; CDA: 7). Most evidence submissions used individual patient data from historical clinical trials, retrospective real-world studies, or prospective observational studies to build ECAs. Statistical methods included matching-adjusted indirect comparison (n=7), propensity score weighting (n=4), and propensity score matching (n=4), with some submissions using multiple methods. All submissions employed partitioned survival economic models. Incremental cost-effectiveness ratios (ICERs) per quality-adjusted life year (QALY) gained ranged from ~£20,000 to £58,223 and C$127,679 to C$1,276,217 across NICE and CDA appraisals, respectively. Agency critiques of evidence submissions concentrated on exclusion of prognostic factors and treatment effect modifiers, small effective sample sizes and/or low data quality, clinical heterogeneity, lack of face validity, and uncertainty in comparative efficacy assessments. Despite methodological issues, 10 out of 12 appraisals led to positive reimbursement decisions.

CONCLUSIONS: HTA submissions for CAR-T cell therapies based on single-arm trials with ECAs are generally accepted by NICE and CDA. Future research endeavors should prioritize exploring the incorporation of prognostic factors and effect modifiers in ECA construction to enhance the robustness and reliability of HTA evaluations.