OBJECTIVES:

Surrogates endpoints (SEP) of overall survival (OS) are increasingly being used as the primary endpoint in oncology. The objective is to evaluate the acceptance of SEP by European Health Technology assessment (HTA) bodies.

METHODS:

A review of the European recommendations for use of SEP (via authorities’ websites) has been performed. A targeted literature research (TLR) on Pubmed and Embase studying correlation of SEP with OS in advanced ovarian cancer (AOC) and non-Hodgkin's lymphoma (NHL) was done. The opinions of the “Haute Autorité de Santé” (HAS) and the National Institute for Health and Care Excellence (NICE) of niraparib in AOC, and axicabtagene ciloleucel in a subtype of NHL, both having used SEP on their submission dossiers, were critically reviewed and compared.

RESULTS:

Variable levels of correlation with OS were found in the TLR: in NHL the linear regression coefficient (r) linking progression free survival (PFS) and PFS at 24 months to OS was superior to 0.85. In AOC an r<0.85 was found with overall response rate (ORR) but an r>0.85 with post progression survival (PPS). The critical analysis discussed the area under the curve of the cost-effectiveness model (CEM) of niraparib and the Mixture Cure Model (MCM) of axicaptagene ciloleucel. While NICE emphasized that approaches to extrapolate long-term survival were not appropriate, HAS rejected the use of PFS as an appropriate SEP of OS to populate the CEM of niraparib. Both HTA bodies accepted the MCM developed in the axicaptagene ciloceucel dossier, although concluding that no model was optimal for constructing long term OS curve.

CONCLUSIONS:

NICE and HAS consider that PFS/ORR/OS SEP relationships are still not proven. OS should remain the main judgement criteria. Acceptance of SEP by NICE looks more opened than HAS. In both CEM, this led to great uncertainty for NICE and invalidation for HAS (major methodologic reserve).