OBJECTIVES: Entrectinib is a potent agent used for targeted therapy of patients with NTRK fusion-positive cancers irrelevant of their origin or histology. Not surprisingly, such tumor diversity represents major challenge for its pharmacoeconomic evaluation. Here, we tried to overcome these issues and estimate its ICER in comparison with standard of care (SoC) from the payer´s perspective.

METHODS: We used national data as input for a three-health states partitioned survival model. Progression-free survival (PFS) and overall survival (OS) curves determined the patient proportions in each health state. We derived PFS and OS in entrectinib arm from parametric regression of empirical data from pooled trials. In absence of H2H studies, we used mean PFS and OS values for SoC and derived median values from literature search with exponential function. Mean PFS and OS for individual tumor types were weighted according to the frequencies observed in the pooled entrectinib NTRK+ subgroups. We used utilities for PFS state from STARTRK-2 EQ5D-3L data, mapped them with UK tariffs, whereas for post-progression state we used average value based on literature. We also used previously published data, including adverse events management, pre-/post-progression costs, as main economical inputs. Deterministic and probabilistic analyses were performed to identify uncertainty.

RESULTS: Over a timeframe of 30 years with 3% discount rates entrectinib generates 0.84 and 1.05 incremental QALY and LYs, respectively, with 2.2 mio CZK of additional costs compared to SoC. Based on these results, we determined ICER of entrectinib therapy as 2.6 mio CZK per QALY and 2.1 mio CZK per LY gained.

CONCLUSIONS: In patients with NTRK fusion-positive tumors, entrectinib therapy has clinical benefits with acceptable incremental costs. Nevertheless, due to the poor clinical evidence and naïve-indirect comparison, local HTA agency considered these results insufficient, exposing limitations of conservative HTA approaches in face of novel innovative therapies.