Risk of neutropenia-related hospitalization (NRH) related to duration of short-acting granulocyte colony stimulating factor (sG-CSF) for primary prophylaxis.

e18181 Background: In clinical trials, efficacy of 10–11 days of primary prophylactic (PP) sG-CSF is similar to a single dose of pegfilgrastim for preventing febrile neutropenia (FN). However, most patients receive < 10 days of PP sG-CSF in clinical practice. This study assessed the effect of PP sG-CSF duration on the risk of NRH. Methods: Using Medicare 20% sample data, we conducted a nested case-control study within a cohort of patients aged ≥66 y with breast, colorectal, lung, ovarian, or prostate cancer or NHL who initiated first cycle of chemotherapy 1/1/2008–9/30/2016, had ≥1 y continuous coverage in Medicare parts A and B before cycle day 1 (baseline), and received PP sG-CSF. We identified NRH cases (ICD-9, 288.0X; ICD-10, D70.X) from cycle day 5 to end of cycle 1. We matched each case to up to 4 controls based on age (± 1 y), tumor type, regimen risk for FN (intermediate/high [ > 10%], other), and year using incidence density sampling. Duration of sG-CSF (days of use from cycle day 1 to the day before case date) was categorized as < 5 and ≥5 days. We used conditional logistic regression adjusted for race, sex, and Charlson comorbidity index (CCI) at baseline to estimate relative risk of NRH related to duration of sG-CSF. Results: Of 1431 patients receiving PP sG-CSF, 68 cases matched 231 controls. Cases were similar to controls in age (76 vs 75 y), tumor type (NHL, 62% vs 62%; lung cancer, 25% vs 25%), intermediate-/high-risk regimen (65% vs 67%), and male sex (50% vs 48%) but had slightly higher CCI (3.9 vs 3.3). The percentage of patients with ≥5 days of PP sG-CSF use was 26% in cases and 41% in controls. The adjusted OR (95% CI) for NRH was 0.48 (0.25–0.93) for ≥5 vs < 5 days of PP sG-CSF; results were consistent across sensitivity analyses (Table). Conclusions: Among elderly cancer patients receiving PP sG-CSF, ≥5 days of sG-CSF use was associated with substantial reduction in the risk of NRH.[Table: see text]

Duke Scholars

Author Gary Herbert Lyman Medicine, Medical Oncology