OBJECTIVES: Due to the extended survival times in chronic cancers, there is an interest in using surrogate outcomes for approval and reimbursement decisions in the absence of mature survival data from trials. However, there are examples where the mature survival endpoints fail to demonstrate the same efficacy as surrogate markers. We aimed to investigate the use of surrogate or extrapolated endpoints in cost-effectiveness analyses from recent HTAs on chronic cancers by CADTH, NICE, IQWiG, and SMC between 2019 and May 2022 to determine whether these guidelines are being followed.

METHODS: The International Network of Agencies for Health Technology Assessments was searched in May 2022 to identify English HTAs for chronic cancers including chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), chronic myeloproliferative disease (CMD), chronic gastrointestinal tumors (cGI), or indolent non-Hodgkin’s lymphoma (iNHL).

RESULTS: Among the 21 HTAs identified (6 from CADTH, 8 from IQWiG, 4 from SMC, and 3 from NICE), 13 included an approximated outcome (either a surrogate or an extrapolated endpoint; CLL:11; CML: 0; CMD: 0; cGI:1; iNHL:1). Two HTAs from CADTH used progression-free survival (PFS) in iNHL or durable response (cGI) as surrogate endpoints for overall survival (OS). Both reports noted there was a high degree of uncertainty in the measure; however, only the HTA on iNHL was accepted. The remaining 11 HTAs on CLL for venetoclax, venetoclax plus rituximab, venetoclax plus obinutuzumab, or acalabrutinib used extrapolated OS or PFS. In 9 of these HTAs, the extrapolation method was considered acceptable, however, only 5 validated the extrapolation method. Validation methods included fitting parametric or covariate models.

CONCLUSIONS: Despite surrogate and extrapolated endpoints being used and accepted by HTA agencies, inconsistencies were observed in the validation of approximated markers. Our research suggests a need for validation of approximated markers to appropriately assess the validity of these outcomes.