Prognostic factors in adult chronic myelomonocytic leukemia: an analysis of 107 cases.

Adult chronic myelomonocytic leukemia (CMML), as defined by the French-American-British (FAB) group, is associated with a variable survival, ranging from a few weeks to several years. From 1971 to 1986, we made the diagnosis of CMML in 107 cases, according to FAB criteria (except for patients with 20% to 30% bone marrow [BM] blasts who were also included). Median survival was 30 months (range 1 to 81 months) and life expectancy did not seem to be influenced by treatment modalities other than supportive care. Eighteen patients (17%) progressed to acute nonlymphoblastic leukemia (ANLL). In a Cox regression model, main factors associated with short survival were: an excess of marrow blasts (P = 10(-6], anemia (P = .17 x 10(-5], high peripheral blood (PB) monocytosis (P = .26 x 10(-5], presence of PB blasts (P = .49 x 10(-4], and to a lesser extent hyperleukocytosis (P = .001), presence of PB immature granulocytes, thrombopenia, and splenomegaly. Survival (less than 1 year v greater than 1 year) could be predicted at diagnosis in a multivariate stepwise discriminant analysis using two parameters only (percentage of BM blasts and hemoglobin level), with 82% accuracy. Among patients surviving greater than 1 year, initial PB leukocyte count was higher in patients with intermediate survival (12 to 42 months) than in long survivors (greater than 42 months) and was the only discriminating factor between these two subgroups in multivariate analysis. Abnormal cytogenetic findings and increased lysozymuria were also poor prognostic factors, but could not be analyzed in the multivariate models, as they were determined in a minority of patients. Parameters associated with subsequent progression to ANLL included younger age at diagnosis, thrombopenia, increased BM blasts, and splenomegaly. Our study allows for the identification of subgroups with different prognoses in CMML, on the basis of a small number of hematologic parameters, particularly initial percentage of BM blasts, hemoglobin level, and leukocytosis. These subgroups probably require different therapeutic approaches.