Factors associated with transfusion requirements during treatment for acute myelogenous leukemia

Supportive care is a prerequisite for intensive chemotherapy in leukemic patients. Little has been published about quantitative aspects of red blood cell and platelet transfusions. We evaluated transfusion requirements and factors associated with observed differences in 206 patients undergoing initial induction consolidation chemotherapy for newly diagnosed acute myelogenous leukemia. All patients were treated during a 5-year period in 12 hospitals on a common protocol of the Swiss Study Group for Clinical Cancer Research (SAKK). Protocol 30/85 comprises a double induction and one course of consolidation. Of 206 registered patients, 199 were evaluable; 118 of 199 (59%) patients entered completed all three cycles of chemotherapy. These 118 patients received a median (range) of 18 (3-44) units of red blood cells and 12 (2-61) platelet transfusions during 112 (70-129) days of hospitalization. Patients with a hemoglobin > 10 g/l, platelets > 100 x 10(9)/l, and white blood cell counts < 5 x 10(9)/l at diagnosis received fewer transfusions than patients with less favorable blood counts during the first cycle of chemotherapy (p < 0.05). Patients with FAB subtype M3 received more platelet transfusions during the first cycle. Female patients received more platelet transfusions than male patients. In multivariate analyses the participating center was the most important single factor associated with the number of red cell and platelet concentrates given per patient and cycle (p < 0.05), the number of days in hospital (p < 0.05), and the risk of premature withdrawal from the study. These data define factors associated with transfusion requirements in patients treated for newly diagnosed AML. They include severity and subtype of disease at diagnosis, age and sex of the patients, and the participating institution. Results suggest that medical decision-making varies from center to center. The participating institution is strongly associated with differences in transfusion requirements, hospitalization time, and premature withdrawal from study. Leukemia trials tend to focus on the prospective evaluation of chemotherapy or growth factors. Our results suggest that other variables, such as management strategies, should be included for prospective analysis.