3000 CORRESPONDENCE Randomized Comparison of Interferon-a, Hydroxyurea, and Busulfan in Chronic Myeloid Leukemia:

and Baccarani regarding our report.’ Kantarjian and Talpaz compare their retrospective study on 274 interferon-a (R%)-treated chronic myeloid leukemia (CML) pa-tients ’ with the IFN arm of our randomized prospective study that compares I F N with hydroxyurea and busulfan. Although the quality of the group’s studies is acknowledged, it cannot be expected that a randomized study with defined inclusion criteria that follows the intention-to-treat principle will yield identical results. The goal of our study was to analyze the impact of IFN on the duration of the chronic phase and on survival in newly diagnosed chronic-phase CML requiring treatment as compared with standard busulfan or hydroxyurea. There were almost no entry restrictions, particularly no age limits (our oldest patient was 85 years at diagnosis), to obtain information on the true benefit of the respective therapies in a nonpreselected patient population representative of general hema-tologic practice. In addition, it should be noted that our protocol required monotherapies. In consideration of these differences it is not surprising that Kantarjian and Talpaz ’ patient population contains almost twice as many low-risk patients as our patient population and much fewer intermediate- and high-risk patients (Table 1). This difference correlates with a median age of 41 years in the Houston patients as compared with 47.4 years in our patients. Kantarjian and Talpaz note that only 3 1 % of our patients attained complete hematologic responses as compared with 80 % in their group. However, the rates of complete and partial hematologic re-sponses together are similar in both studies (83 % in our patients and 87 % in the Houston patients) despite the differences in patient grou