A new model for predicting infectious complications during fludarabine-based combination chemotherapy among patients with indolent lymphoid malignancies.

Abstract

Fludarabine-containing combination chemotherapy regimens are increasingly used in the treatment of indolent lymphoid malignancies, with the associated risk of infection being the major toxicity. Predictors of infection during fludarabine-containing combination therapy are poorly defined and optimal strategies for infection prophylaxis are not known. The authors analyzed their experience with patients treated with the fludarabine-mitoxantrone (FM) or fludarabine-cyclophosphamide (FC) regimens to develop a predictive model for infections. Ninety-two patients with indolent lymphoid malignancies were treated with FM (n = 29) or FC (n = 63). Baseline variables including age, gender, regimen, disease histology, previous therapy, time from diagnosis to current treatment, performance status, renal function, absolute neutrophil count (ANC), lymphocyte count, and immunoglobulin G levels were examined retrospectively for their association with risk of infectious complications during or within 4 weeks of therapy. Six risk factors were associated with infectious complications: age > 60 years, > or = 3 previous therapies, previous fludarabine exposure, time from diagnosis to current treatment of > 3 years, performance status > or = 2, and baseline ANC < 2.0 x 10(9)/L. Compared with patients with 0-2 risk factors, patients with > or = 3 risk factors had higher infection rates (26% vs. 7% per cycle, P < 0.0001), more Grade 4 neutropenia (41% vs. 8% per cycle, P < 0.0001), and more neutropenic sepsis (15% vs. 1% per cycle, P < 0.0001). Infection risk during fludarabine-containing combination chemotherapy was predicted with a model comprising six baseline risk factors. Patients predicted to be at high risk of infection were an appropriate group for consideration of prophylactic strategies.