A study of novel febrile neutropenia risk factors related to bone marrow or immune suppression, barrier function, and bacterial flora

Abstract

Background: Previously identified patient-level risk factors for chemotherapy-induced febrile neutropenia (FN) indicate several potential underlying pathogenic mechanisms, including bone marrow suppression, impaired neutrophil function, or disturbances of barrier function. This study evaluated whether additional clinical characteristics related to these pathogenic mechanisms were risk factors for FN. Patients and Methods: The study population included patients diagnosed with non-Hodgkin's lymphoma or breast, lung, colorectal, ovarian, or gastric cancer between 2000 and 2009 at Kaiser Permanente Southern California and treated with myelosuppressive chemotherapy. Those who received prophylactic granulocyte colony-stimulating factor or antibiotics were excluded. Potential risk factors of interest included surgery, radiation therapy, selected dermatologic/mucosal conditions, and use of antibiotics and corticosteroids. All data were collected using electronic medical records. Multivariable Cox models were used to evaluate associations between these factors and risk of FN in the first chemotherapy cycle, and adjusted using propensity score-based functions. Results: A total of 15,971 patients were included. Of these, 4.3% developed FN in the first chemotherapy cycle. Use of corticosteroids was significantly associated with increased risk of FN (adjusted hazard ratio [aHR], 1.53; 95% CI, 1.17-1.98). Selected dermatologic/mucosal conditions and intravenous antibiotic use were marginally associated with increased risk of FN (aHR, 1.40; 95% CI, 0.98-1.93, and 1.35; 95% CI, 0.97-1.87, respectively). Surgery, radiation therapy, and oral antibiotic use were not statistically significantly associated with FN. Conclusions: Dermatologic or mucosal conditions that might affect barrier integrity and use of corticosteroids and intravenous antibiotics prior to chemotherapy may increase risk of FN and should be considered in prophylaxis use and FN prediction modeling.