We read with great interest the excellent report by Schofield and colleagues [1] on the efficacy of low-dose interferon- in patients with chronic myeloid leukemia. In this study, 41 patients with Philadelphia chromosome (Ph)-positive chronic myeloid leukemia received interferon- maintenance therapy, 2 x 10⁶ U/m² body surface area three times a week. The complete hematologic response rate was 70%, the major cytogenetic response rate was 22%, and the median survival from the time of diagnosis was 84 months; these figures compared favorably with the historical experience with hydroxyurea therapy.

Although these results confirm our original data, we disagree with the conclusion that low-dose interferon- is as effective as higher doses. The conclusions were based on comparison of the small sample of 27 patients with early-phase chronic myeloid leukemia.

We completed our analysis of 274 patients with early-phase chronic myeloid leukemia treated between 1982 and 1990 with interferon-, 5 x 10⁶ U/m², or the lower maximally tolerated individual dose. The median patient age was 47 years, approximately 10 years higher than the age of the patients treated by Schofield and associates. Comparative results show higher hematologic (85% compared with 71%), cytogenetic (58% compared with 34%), major (36% compared with 22%), and complete (26% compared with 7%) cytogenetic response rates in our patients. We also found significant differences in response and prognosis when data were evaluated for various patient risk groups [2, 3]. The favorable results observed by Schofield and colleagues may have been due in part to a higher percentage of patients with more favorable risk profiles in their study. Among such patients, the expected major cytogenetic response rate would be 50%, a figure that is higher than the 22% reported in their study.

Our concern is that this study, which used a small patient sample without an in-depth analysis of prognostic determinants, may result in practices and investigations unfavorable to the general population of patients with chronic myeloid leukemia. It will encourage the use of lower doses of interferon-, thus producing lower hematologic and cytogenetic response rates. Because survival prolongation has been associated with induction of cytogenetic response [3, 4], patient survival may worsen. The conclusions of Schofield and coworkers will also argue for initiation of randomized studies to compare lower and higher doses of interferon-, despite sufficient data showing that higher doses are more effective in inducing cytogenetic responses associated with increased survival. Such studies may divert investigational resources at a time when only about 25% to 30% of patients (those with major durable cytogenetic responses to interferon-) truly benefit from such therapy. Instead, we believe that a series of pilot trials of interferon- combined with other agents may be more appropriate to increase the percentage of the patients who achieve major durable cytogenetic responses to a reasonable range (40% to 50%). Only then will randomized trials be needed to answer questions of reduced toxicity, cost, and other quality-of-life measures.