Objective: To compare the cost-effectiveness of interferon-α with that of hydroxyurea as initial therapy for patients with chronic myelogenous leukemia (CML) in the chronic phase.
Design: A decision analysis and Markov model that described the natural history of the therapeutic process. The Markov model contained two treatment arms (interferon-α and hydroxyurea) and eight states of health (complete hematologic remission with cytogenetic response, complete hematologic remission without cytogenetic response, partial hematologic remission, chronic phase without hematologic remission, accelerated phase, blast crisis, bone marrow transplantation, and death). Probabilities, costs, and utilities were obtained from published clinical studies and clinical investigators.
Measurement: Quality-adjusted years of life saved and costs and qualities discounted at 5% per year.
Setting: University medical centers in North America and Europe.
Patients: Meta-analysis of results from patients studied in clinical trials.
Results: The model's predictions of median survival (69 months with interferon-α therapy and 58 months with hydroxyurea therapy) were derived from data in the recent literature. In patients 50 years of age, interferon-α improved life expectancy over hydroxyurea by approximately 18 months. The marginal cost-effectiveness of interferon-α (incremental discounted cost of interferon-α compared with that of conventional therapy) was $34 800 per quality-adjusted year of life saved. The model was sensitive to the monthly cost of interferon-α therapy (if the cost of interferon-α is reduced by one third, the cost-effectiveness becomes $19 300 per quality-adjusted year of life saved) but was not particularly sensitive to the costs associated with blast crisis or bone marrow transplantation. The other significant variable was quality of life during therapy with interferon-α; when this measure was varied from 70% to 100% of the quality of life during hydroxyurea therapy, cost-effectiveness changed from $123 200 to $25 620 per quality-adjusted year of life saved. When the quality of life associated with interferon-α was less than 62% of the quality of life associated with hydroxyurea, the discounted quality-adjusted life expectancy with interferon-α was less than that with hydroxyurea.
Conclusion: Compared with hydroxyurea, interferon-α is, in most clinical scenarios, a cost-effective initial therapy for patients with chronic-phase CML who can tolerate the drug.