OBJECTIVES:  To evaluate the cost-effectiveness of dasatinib versus nilotinib and high-dose imatinib (600 or 800 mg) for people with chronic myeloid leukemia (CML) in chronic, accelerated and blast phase (CP, AP and BP), which are resistant or intolerant to normal-dose imatinib, from the perspective of the Chinese healthcare system. METHODS: A Markov cohort model was adapted to Chinese treatment practice. Clinical and utility data were taken mostly from the literature by a MEDLINE search. Costs were based on local charges. The primary output was reported in terms of incremental cost-effectiveness ratios (ICERs). Univariate and probabilistic sensitivity analyses were performed to examine the robustness of the model output. The impact of patient assistance program (PAP, buy three- month and get nine-month free) was assessed.

RESULTS: In chronic phase, the base case analysis showed the dominance of dasatinib versus nilotinib and imatinib 800 mg strategies because dasatinib gained more quality-adjusted life years (QALYs) with lower costs. For CML-CP patients, dasatinib is expected to provide more QALYs with more costs in comparison with imatinib 600 strategy, yielding an ICER of ￥369,767 /QALY. When PAP was available, the ICER of dasatinib versus 600 mg imatinib was ￥101,453, and nilotinib and 800 mg imatinib were dominated in patients with CP. When 3 times per Capita GDP of China (￥125,414 /QALY) according to WHO recommendations was used as the threshold and PAP was provided, dasatinib achieved about 90% probabilities of cost-effectiveness in chronic phase. The uncertainties could be contributed to the costs of drugs. CONCLUSIONS: The results indicate that PAP could notably improve the cost-effectiveness of dasatinib versus nilotinib and high-dose imatinib among imatinib-resistant or intolerant CML patients in China.  Dasatinib is likely to be cost-saving versus nilotinib or imatinib 800 mg strategies in CML-CP patients.