Predicting the Risk of Neutropenic Complications and Reduced Dose Intensity in Patients with Malignant Lymphoma: Results from a Prospective Study.

Background: Chemotherapy-induced neutropenia is frequently associated with neutropenic complications resulting in reduced relative dose intensity (RDI) potentially compromising clinical outcomes. Retrospective analyses have identified several factors but are limited by variable reporting and missing data.

Methods: Results from an ongoing prospective study of cancer chemotherapy patients based at more than 100 randomly selected US centers are reported. Demographic, clinical and treatment-related variables were obtained on 345 patients with malignant lymphoma (ML). Data are presented on the first 265 patients completing at least one cycle of chemotherapy. Planned and unplanned reductions in relative dose intensity (RDI) were estimated along with the risk of neutropenia and febrile neutropenia (FN) over the first four cycles of treatment are presented.

Results: Patients with Hodgkin’s disease (HD; N=43) received five different regimens including ABVD (84%) while patients with Non-Hodgkin’s lymphoma (NHL; N=222) received 33 different regimens the most common of which was CHOP (53%). Neutrophil nadirs <1000 were reported in 54% of subjects, two-thirds of which occurred during the 1^st cycle of treatment. One or more episode of FN occurred in 18% (29% HD; 16% NHL) with 60% in cycle 1. Overall average RDI was 90% with 25% receiving <85%. Differences in delivered RDI were found based on practice setting (P=.07), geographic location (P=.007), age (P=.007), ECOG performance status (P=.065) and peripheral vascular disease (P=.021). Significant reductions in planned RDI but not unplanned RDI were observed for increasing age (P=.008). Over the first three cycles of treatment, patients with HD experienced a greater risk of FN (29%) than patients with NHL (16%) (P=.042). Correspondingly, HD patients received an average RDI of 93% with two-thirds of reductions unplanned compared to NHL patients who received an RDI of 90% with two-thirds were planned reductions (P=.03). In multivariate logistic regression analysis, significant predictors of reduced RDI were age, chemotherapy regimen, peripheral vascular disease and cerebrovascular disease (P<.001; R²=.435).

Conclusions: This prospective study confirms previous retrospective surveys that showed a substantial proportion of patients with ML receive reduced RDI potentially compromising clinical outcomes. Predictive models based on the risk factors identified should enable clinicians to target supportive care efforts in a more cost-effective manner.