HTA38 Characterizing Evidentiary Criteria in HTA Assessments of CAR-T Therapies

CAR-T therapies are one-time administration cell-based therapies with high upfront costs, resulting in growing payer scrutiny of associated evidence packages. In this research, we aimed to characterize the clinical and economic evidence of CAR-T therapies that maximally impact reimbursement decisions across key HTA agencies. HTA websites for NICE, CADTH, HAS, and the G-BA were searched for CAR-T HTA reports published up to 31 December 2021. Key outcomes were extracted and mapped for clinical and cost effectiveness markets. A total of 20 reports were extracted for 4 drugs (Kymriah, Yescarta, Tecartus and Abecma) across 5 oncology indications (DLBCL, PMBCL, ALL, MCL and MM). Only 20% received positive recommendations (all from HAS), with the remaining 80% receiving restricted reimbursement or were rejected. All of the drugs reported only single-arm phase I/II trials, often accompanied by Real World Evidence (RWE) including external comparator studies and Indirect Treatment Comparisons, typically with a sample size of N<200. Primary endpoint was overall response rate and remission rate for 75% and 20% of all reviewed trials, respectively. Median follow-up time was in 11.3 – 28.6 months range. For cost-effectiveness markets, limited data resulted in uncertainty of the exact size of the benefit of each drug compared with salvage chemotherapy. Where conclusions could be drawn, treatments were mostly not considered cost-effective. Despite suboptimal evidence, most HTAs provided restricted reimbursement for CAR-T therapies due to lack of standard of care and on the condition of additional evidence generation, such as longer follow-up trial results, generation of comparative evidence through RWE observational studies and the use of appropriate indirect comparators and endpoints. Overall, for novel CAR-T products with limited clinical data at launch, payers are increasingly employing RWE to minimize uncertainty in evidence and cost consequence estimates as reimbursement decision drivers, even as their evidentiary thresholds continue to evolve.