OBJECTIVES: To understand factors related to cost-effectiveness modelling that facilitate or hinder positive HTA recommendations for CAR T therapies across Europe, Canada and Australia, using axicabtagene ciloleucel (axi-cel) in second-line treatment of large B-cell lymphoma (2L-LBCL) as a case study.

METHODS: A review of economic assessments published by 7 HTA agencies was conducted. Key factors to successful HTA recommendations and differences in preferred modelling approaches between the manufacturer and HTA agencies were identified. The proportional difference in publicly available submitted versus accepted ICERs was captured to quantify the difference in perceived cost-effectiveness between the manufacturer and HTA agencies.

RESULTS: Three success factors were: 1) the pivotal trial (ZUMA-7) was sufficiently powered and had adequate follow-up (median follow-up 47.2 months) to support stabilisation of cure fractions estimated from mixture cure models; 2) appropriate statistical methods were applied to health-related quality-of-life data (i.e., mixed model for repeated measures) to estimate utilities; and 3) selection of survival extrapolations were supported by external data from previously reimbursed indications (ZUMA-1). Four modelling approaches preferred by HTA agencies were: 1) application of high discount rates and/or shortening of modelled time horizons; 2) inclusion of delivery costs for CAR T that include set-up costs; 3) pessimistic selection of extrapolations despite reduced uncertainty with the introduction of RCT evidence; and 4) no consideration of manufacturing success rate in modelling. On average, a 104% increase was observed between the submitted and accepted ICERs.

CONCLUSIONS: As axi-cel is the only CAR T for 2L-LBCL that has achieved positive HTA recommendations across major cost-effectiveness markets, clear lessons can be learned about building a successful economic evidence package for CAR Ts. However, a potential underappreciation of the evidence required and preferences for different modelling approaches could lead to an imbalance in perceived cost-effectiveness that could impact patient access and future innovation.