Problem Statement: Current CAR-T therapies target rare oncologic diseases. Due to clinical trial limitations, manufacturers leverage 2-3 year data from Phase II/III RCTs to support long-term survival claims that drive clinical and economic value propositions. The lack of long-term data increases the perceived data uncertainties on these claims. This detracts value, which translates into access restrictions.

Description: In this study, we identified strategies to support long-term survival claims and, consequently, reimbursement of CAR-T therapies in Germany and the UK. We focused on eight G-BA and NICE reports identified for tisagenlecleucel (ALL and DLBCL), axicabtagene ciloleucel (DLBCL and PMBCL), and brexucabtagene autoleucel (MCL). Across all HTA reports, the long-term clinical value was recognized only when a trend in survival and response was demonstrated through relevant endpoints. Endpoint relevance was assessed based on clinical relevance (i.e., correlation with survival) and/or patient relevance (i.e., patient functioning, feeling, or survival). Moreover, post-launch evidence was requested to validate the initial assumptions on survival. Across NICE assessments, partitioned survival cost-effectiveness models and spline or parametric survival models were preferred, with cure points based on assumptions derived from supporting evidence (i.e., published literature, clinical trial data). Given that long-term survival claims were difficult to substantiate in the absence of long-term data, HTA agencies restricted initial reimbursement (i.e., referral to the CDF by NICE and time-limited resolutions by G-BA) to allow for additional data collection and reassessment based on more mature data.

Lessons Learned: Strategies are needed to support long-term survival claims for CAR-T therapies. Data generation strategies involve using patient-relevant surrogate endpoints that correlate with survival (i.e., show a trend in survival and response benefit) and committing to post-launch evidence generation activities. Data analysis strategies involve the use of partitioned survival cost-effectiveness models and spline or parametric survival models with a validated cure point.

Stakeholder perspective: Industry