Predictors of febrile neutropenia among Medicare patients with breast, lung and colorectal cancer

Abstract

9120 Background: Neutropenia and associated impaired immunity is the major dose limiting toxicity for systemic chemotherapy administration. Current guidelines recommend the prophylactic use of G-CSF when the risk of febrile neutropenia (FN) is approximately 20% because it decreases the duration and risk of FN. Advanced age has been identified as a risk factor for FN; however, little is known about the impact of other factors on the incidence of FN in the elderly. Methods: We analyzed SEER-Medicare data on patients diagnosed between 1995 and 2000 with breast, colorectal, and lung cancer, who received chemotherapy within 11 months of diagnosis. We examined a number of factors that might potentially be associated with FN in the first cycle of chemotherapy, including patient demographics, cancer stage at diagnosis, number of myelosupressive chemotherapeutics, frequency of chemotherapy administration, and comorbid conditions, including previous history of anemia and other blood disorders. Outcome was defined as admission to hospital within 30 days of initial chemotherapy administration with neutropenia (ICD9 288.0) coded as an admission diagnosis. Results: Of 18,637 patients who met inclusion criteria 751 (4.0%) had an episode of FN within 30 days of initial chemotherapy administration. In multivariate logistic regression analysis, independent predictors of FN included cancer type (lung cancer, OR 3.77, CI: 1.02–13.9, as compared to breast cancer) and history of blood disorder (OR 1.49, CI: 1.10–2.0), with a trend towards history of COPD (OR 1.19, CI: 0.98–1.44) and three or more myelosupressive agents (OR 4.01, CI: 0.76–21.2, as compared to non-myelosupressive chemotherapy). Initiating chemotherapy less than 1 month after cancer diagnosis was also associated with a higher odds of developing FN (OR 1.88, CI:1.42–2.49 compared to 3 months or more after diagnosis). Discussion: By utilizing the SEER-Medicare dataset we were able to examine predictors of febrile neutropenia in a large population of elderly patients with common malignancies. Knowledge of predisposing factors for neutropenia will allow more tailored prophylactic use of G-CSF in this at-risk population. No significant financial relationships to disclose.