Cost-Effectiveness Analysis (CEA) of Sequential Treatment with Tyrosine Kinase Inhibitors (TKIs) for Chronic Myelogenous Leukemia (CML)

Background: Imatinib has been approved for treatment of newly diagnosed CML since 2001. For patients failing imatinib, other treatment options such as dasatinib, nilotinib, bosutinib, and high-dose imatinib are recommended. Previous economic analyses assumed that patients are treated until progression, despite guidelines recommending changing treatments for non-responders. The objective was to perform a CEA of sequential treatment with 2^nd line TKIs, from a commercial payer perspective in the United States (US).

Methods: A Markov-cohort model was used to simulate lifetime treatment costs and health outcomes (discounted 3.0% per annum) for patients resistant or intolerant to 1^st line imatinib. It compared six treatment sequences, starting from 2^nd line (shown in the results table). The model included five health states: chronic phase 2^nd line TKI, chronic phase 3^rd line TKI, chronic phase no TKI, post-progression, and death. After 12 months of TKI treatment, patients without a major molecular or complete cytogenetic response (MMR or CCyR) moved to the next line of therapy (per NCCN guidelines). Patients could also move to the next treatment line if they lost MMR or CCyR, or discontinued due to drug-related toxicity. Data for response achievement, risks of progression, death, adverse events, and discontinuation were primarily based on data in published trials. Due to the lack of head-to-head studies for dasatinib vs. nilotinib in 2^nd line, and a lack of time-to-response data for all three 2^nd line treatments, MMR and CCyR rates were interpolated from available data points in 2^nd line, while dasatinib and nilotinib rates were assumed to be equal in 3^rd line. Dasatinib and imatinib progression, loss of response, and survival rates for 2^nd line responders were assumed equal, due to data limitations. In each health state, patients accrued drug costs, resource use (related to monitoring, AE management, and disease management) costs and quality-adjusted life years (QALYs). Resource use, cost, and utility estimates were based on FDA labels, RedBook, Medicare and AHRQ Healthcare Cost Utilization Project data, and published economic analyses. Multi-way uncertainty analyses evaluated key contributors to uncertainty in the results, by testing various assumptions for probabilities of discontinuation, response, loss of response, progression, and survival.

Results: The model predicts that 2^nd line dasatinib provides increased survival (ΔLYs = ~0.4-2.6 years) and QALYs (ΔQALYs = ~0.4-2.8 years) in all patient groups when compared with 2^nd line high-dose imatinib or nilotinib sequences. Also, 2^nd line dasatinib was more costly (ΔLifetime Costs = ~$65,000 - $225,000) than high-dose imatinib and nilotinib, primarily due to longer survival and corresponding longer time on TKI treatment. In ±20% univariate sensitivity analyses, the model was most sensitive to 2^nd line progression and survival estimates.

Conclusions: This analysis suggests that dasatinib may be associated with increased life expectancy and quality of life when compared with high dose imatinib or nilotinib, among patients who are resistant or intolerant to 1^st line imatinib, primarily based on higher cytogenetic response rates observed in studies of dasatinib. Other studies have shown improved quality of life for responders, and landmark analyses have shown improved survival for patients achieving cytogenetic response, but head-to-head clinical studies of sequential use of dasatinib and nilotinib are needed to confirm the model result. Based on the threshold of $150,000/QALY, dasatinib can be considered cost-effective in the US.

BOS=bosutinib; DAS=dasatinib; HDI= high-dose imatinib; NIL=nilotinib; ICER=incremental cost-effectiveness ratio