Risk of Febrile Neutropenia (FN) in Select Myelosuppressive Chemotherapy Regimens

Introduction

Febrile neutropenia (FN) is a serious adverse effect of myelosuppressive chemotherapy, which often results in hospitalization and chemotherapy dose modification. FN risk depends on patient characteristics and chemotherapy regimen risk. Understanding the FN risk associated with individual chemotherapy regimens can help guide the use of prophylactic granulocyte colony-stimulating factor (G-CSF) and patient monitoring. To this end, the NCCN has classified regimens into high (≥20%), intermediate (10%-20%), or low (<10%) FN risk based primarily on clinical trial data. However, even for the same regimen, the FN risk is often higher in clinical practice than in clinical trials. In this study, we assessed the FN risk associated with several regimens for which FN risk has not been determined or has shown substantial variability outside of a clinical trial setting, using data from Kaiser Permanente Southern California (KPSC), a large, community-based practice.

Methods

Included were patients diagnosed with incident non-Hodgkin's lymphoma (NHL), breast cancer (BC), or multiple myeloma (MM) between 2008 and 2013 at KPSC who initiated the following chemotherapy regimens: bendamustine ± rituximab for NHL; docetaxel, carboplatin, and trastuzumab (TCH) or docetaxel and cyclophosphamide (TC) for BC; or Q4W lenalidomide 25 mg/dexamethasone for MM. Bendamustine ± rituximab, TCH, and lenalidomide are not classified by NCCN; TC is classified as intermediate FN risk but has shown considerable variability of FN incidence when used in clinical practice. Data on cancer diagnosis, chemotherapy use, G-CSF use, neutrophil count, and infections were obtained from KPSC's electronic medical records to estimate the incidence proportions of FN and grade 3 and 4 neutropenia. FN was defined as (1) hospitalization with absolute neutrophil count (ANC) <1000/µL or (2) hospitalization with primary or secondary diagnosis codes of neutropenia (ICD-9 288.0x) and fever (ICD-9 780.6), diagnosis code for bacterial/fungal infection, or antibiotic use. Grade 3 neutropenia was defined as ANC ≥500/µL to <1000/µL; grade 4 neutropenia as ANC <500/µL. Patients who received prophylactic G-CSF within 5 days of chemotherapy initiation were excluded from analysis.

Results

Overall, 40 (12%) NHL patients; 149 (24%) and 340 (28%) BC patients who received TCH and TC, respectively; and 0 (0%) MM patients were excluded due to prophylactic G-CSF. Over the first 6 cycles of bendamustine (median 338.4 mg/m²) ± rituximab for NHL patients (n = 307), 7.2% experienced FN, 4.2% grade 3 neutropenia, and 17.6% grade 4 neutropenia. Over the first 6 cycles of TCH for BC patients (n = 462), 24.2% experienced FN, 10.6% grade 3 neutropenia, and 44.6% grade 4 neutropenia. Over the first 6 cycles of TC for BC patients (n = 859), 20.5% experienced FN, 9.5% grade 3 neutropenia, and 37.5% grade 4 neutropenia. Over the first 4 cycles of lenalidomide/dexamethasone for MM patients (n = 186), 3.8% experienced FN, 5.9% grade 3 neutropenia, and 18.3% grade 4 neutropenia (Table 1).

Conclusions

Using NCCN criteria, bendamustine ± rituximab for NHL and lenalidomide/dexamethasone for MM would be classified as low-FN-risk regimens (<10%). By contrast, BC regimens TCH and TC would be classified as high-FN-risk regimens (>20%) based on our data. These results could help inform prophylactic G-CSF use for the selected regimens in clinical practice.