ELYPSE 2: A prospective randomized trial comparing filgrastim (G-CSF) in primary and secondary prophylaxis in patients (pts) at high risk for febrile neutropenia (FN)

Abstract

8614 Background: FN is a frequent life-threatening consequence of cytotoxic chemotherapy (CT). G-CSF reduces the risk of FN, but primary (1ry) prophylaxis using G-CSF may be cost efficient only if FN incidence is ≥20%. The identification of pts at high risk for FN with simple criterias would be useful in clinical practice. Here we report a randomized phase II trial comparing G-CSF in 1ry vs 2ry prophylaxis in a high risk group of pts (based on our risk model JCO 1996;14:737, Br J Cancer 2003;88:181). Methods: Pts ≥18 years with solid tumors or NHL at high risk for FN after CT were incluable: these were pts with day 1 or day 5 lymphocytes (Ly) count ≤700/μL AND “high risk CT” (HRCT) In pts with d1 and d5 Ly ≤700/μL, the observed incidence of FN was 40% and 66% respectively G-CSF (300 to 480μg/d from d6–12) was randomized either as in 1ry prophylaxis (Arm 1), or as 2ry prophylaxis after FN (Arm 2). Primary endpoint was the rate of grade 4 FN. Hypothesis was a reduction of 40% of the risk of FN with 1ry prophylaxis. Results: Between 03/97 and 12/04, 137 pts were included in 7 centers. The median age was 53 years (range 18–80) with 54% males. Most frequent tumors were sarcomas (36%), breast carcinomas (18%), lymphomas (15%), head and neck carcinomas (10%), and lung carcinomas (6%). 23% patients had PS>1 at the first line of chemotherapy. No difference was observed in terms of duration of hospitalization or antibiotherapy. Median number of days of G-CSF administration was 14 days (0–24) vs 0 (0–17) days (p<0.0000). After the 1st course, grade 4 FN was 38% in Arm 2 (2ry prophylaxis) and 25% in Arm 1 (1ry prophylaxis), showing a 34% reduction of FN in arm 1 (p=0.14): 1ry prophylaxis was associated with a significant reduction of FN using logistic regression (p=0.04). Incidence of FN after the 2 course in pts receiving 2ry prophylaxis was 22%. Among the subgroup of pts with PS>2 and Ly ≤700/μL, (a group with a reported 20% risk for early death, Br J Cancer 2001;85:816), 2 of 8 patients (25%) died after the 1st course, vs 0/13 in the G-CSF group (p=0.05). This difference was not significant in the whole group. Conclusions: This study confirms that lymphopenic pts receiving HRCT are a high risk group of pts for FN for whom 1ry prophylaxis with G-CSF reduces the incidence of FN. [Table: see text]