A risk model for severe anemia to select cancer patients for primary prophylaxis with epoetin α: A prospective randomized controlled trial of the ELYPSE study group

Background: Epoetin (EPO) administration reduces the need for transfusion. Identifying patients at high risk of anemia requiring red blood cell (RBC) transfusion is needed. This multicentric phase III trial tested epoetin α (EPOα) administration according to our risk model on the basis of three clinical parameters: hemoglobin (Hb) <12 g/dl, lymphocytes ≤700/μl, and/or performance status (PS) >1. Patients and methods: Patients ≥18 years with chemotherapy-treated solid or hematologic tumors were randomized to 150 UI/kg/TIW s.c. EPOα (arm 1) or no EPOα (arm 2) and stratified on Hb level at day 0, lymphocyte count, and PS. The primary end point was transfusion rate; secondary end points included overall survival (OS), safety, and quality of life. Results: From September 2000 to January 2005, 218 patients (median age 64 years, 42.7% males) with principally breast cancer, sarcoma, or lung carcinoma were included. In total, 93% patients had PS >1 and 35% had ≤700/μl lymphocytes. Baseline Hb levels were 10.1 g/dl (range 6.9-11.9). Two hundred and thirteen patients were assessable for the primary end point: 36% received RBC in arm 1 and 58% in arm 2 (P = 0.0012). Median OS was 7.6 [95% confidence interval (CI): 5-12] and 6 (95% CI: 5-8) months in arms 1 and 2, respectively. Median OS was significantly worse for patients with three prognostic factors (3.6 months) compared with two factors (8.3 months) (P < 0.001). No difference in toxicity (47% versus 41%) or thrombovascular events (4.5% versus 3.7%) was observed. Conclusion: Patients at high risk for RBC transfusion according to the ELYPSE model could be given prophylactic EPO with significantly reduced RBC transfusions and no significant impact on side-effects, progression-free survival, and OS.