The use of anthracycline-based combination chemotherapy to treat DLBCL in older pts is complicated by the more common occurrence of therapy-related myelosuppression, which may lead to dose reductions that might impair efficacy. The preliminary results of initial therapy with cyclophosphamide, doxorubicin, vincristine, prednisone alone (CHOP) or with rituximab (R-CHOP) as induction therapy in patients age 60 years (yrs) or greater with DLBCL have been reported (

, Abstr 8
). Pts were randomized at induction to receive 6–8 cycles of CHOP or R-CHOP. Responding patients were then randomized to receive either no treatment or maintenance rituximab. Per protocol, GF was not to be given with the 1st cycle of therapy, and thereafter ASCO guidelines were to be followed regarding use of GF. The objectives of this study were to determine the frequency of and indications for GF usage and the incidence of FN. GF usage was defined as the use of either GCSF or GMCSF for the purpose of preventing dose reduction/dose delay or as prophylaxis in pts with a prior hospitalization for FN. Of the 546 eligible pts, adequate data was available to report on 528 patients. Among these pts, 75% were >=65 yrs, 41% hemoglobin <12 g/dl, 50% male, 59% elevated LDH and 59% HI/High IPI. 264 of 528 pts (50%) used GF in at least 1 cycle of therapy (GF used in 307 cycles to treat neutropenia). 181 of the 528 pts (34%) used GF per our definition with the 93% using GCSF only in at least 1 cycle of therapy. For these 181 pts, the median number of cycles of GF use was 5 (range 1–8) and the median duration of GF use was 9 d (SD=3). GF was used in 552 of 3256 cycles (17%). Significantly more patients used GF at later cycles of therapy (65/513,13% at cycle 2 vs 98/429, 23% at cycle 6, p<0.001). The use of GF was evaluated by study entry characteristics (age, sex, performance status, IPI score). The only significant difference was that pts age 60–64 yrs were less likely to have received GF than pts >=65 yrs (25% vs 37%,p=0.01). 217 pts (41%, 95% CI [37,45%]) had at least one FN event. FN occurred in 267/3247 (8%) cycles of therapy with 103 (39%, 95% CI [33,45%]) FN events occurring in cycle 1. The median time to FN was d 11 (SD=3) of the cycle. In this older pt population treated with CHOP+/− R therapy for DLBCL, GF was used in 17% of the cycles, with usage more common in later cycles of therapy. FN was most common in the first cycle of therapy occurring mostly within the first two weeks. This early occurrence of FN supports consideration of GF use to prevent FN from the 1st cycle of CHOP-based therapy in older pts (>=60 yrs), in addition to any other usage per ASCO guidelines. Results will be presented on the impact of delivered dose intensity on GF usage and FN.

Author notes

Corresponding author