The Same but Different: Inconsistencies in Key Cost-Effectiveness Assumptions for Curative Therapies across Clinically Similar NICE Appraisals

BACKGROUND: NICE appraisals in similar therapeutic areas and/or evaluating similar products (e.g. CAR-Ts) are frequently routed to different external assessment groups (EAGs) and can be appraised by different committees. This can potentially lead to different preferences regarding key modelling assumptions, particularly those related to cure.

OBJECTIVES: To compare the preferred model cure assumptions adopted by NICE committees appraising CAR-Ts and/or their comparator treatments bridging to stem-cell transplant (SCT) in relapsed/refractory leukaemia and lymphoma.

METHODS: The following data were extracted from technology appraisals (TAs) published on the NICE website: the preferred cure timepoint/fraction assumptions; the post-cure health-related quality of life (HRQoL) assumption and the post-cure mortality assumption.

RESULTS: 8 TAs were analysed, of which 2 were targeted leukaemia therapies bridging to SCT, 2 were CAR-T therapies for leukaemia and 4 were CAR-T therapies for lymphoma. Among the leukaemia TAs the cure timepoint ranged from 3-4 years, with 2 TAs instead assuming cure fractions. In lymphoma, the cure timepoint was 5 years, with 2 TAs assuming cure fractions. Standardised mortality ratios (SMRs) of cured patients ranged from 1 to 4, with wide ranging SMRs within the same indication. Half of TAs assumed post-cure HRQoL equal to the general population with the other half assuming a decrement, without consistency across indications.

CONCLUSIONS: There is inconsistency in key modelling assumptions preferred by EAGs and/or NICE committees regarding curative therapies. Differences in cure assumptions will logically lead to differences in cost-effectiveness results and the price achievable for a new therapy. A more consistent, structured approach should be considered for centralised application across haematological cancers that considers (1) the burden of the curative treatment; CAR-T vs. SCT (2) the relative contribution of disease- vs. treatment-specific post-cure mortality and its temporality (3) the age of the patient and ability to recover from the disease and treatment.