Real-world effectiveness of prophylactic granulocyte colony-stimulating factor (G-CSF) early (week 1) and late (weeks 2-3) in the cycle for the prevention of febrile neutropenia (FN) among patients (pts) with breast cancer (BC) after high FN–risk chemotherapy (chemo).

Abstract

599 Background: G-CSF mitigates chemotherapy-induced neutropenia (CIN) and reduces FN risk. G-CSF moves the nadir of absolute neutrophil count (ANC) earlier (to week 1) in the chemo cycle and shortens nadir duration (Crawford, NEJM 1991), suggesting the potential for suboptimal CIN protection early (week 1) in the chemo cycle. The relative FN risk in week 1 vs. weeks 2-3 of the cycle with G-CSF is unknown and was analyzed compared with no G-CSF in the real-world setting with high FN risk chemo. Methods: Using a database of administrative claims representing 100% of fee-for-service Medicare, we analyzed BC pts who initiated docetaxel (T), doxorubicin (A), or cyclophosphamide (C) monotherapy or combination therapy between 01/01/2015 – 12/31/2019. Sample pts included adults aged ≥ 65 years with continuous coverage in Medicare Parts A, B, and D for 6 months before and 20 days after chemo initiation. Pts were categorized as receiving vs. not receiving G-CSF therapy within 3 days after chemo. Rate of FN events starting in week 1 vs. weeks 2-3 in cycle 1 was calculated. We defined FN as an inpatient admission with a primary or secondary diagnosis of neutropenia and measured the interval between chemo initiation and FN admission. Results: Among 18,788 Medicare beneficiaries with BC treated with T, A, and/or C, 72% received G-CSF therapy. More pts receiving G-CSF were treated with ≥ 2 of T, A, and/or C compared to pts who did not receive G-CSF (71% vs. 51%). Overall FN incidence in cycle 1 was significantly lower among pts receiving G-CSF (4.0%; n=546) compared to pts not receiving G-CSF (8.8%; n=462) (p<0.0001). In pts with G-CSF, 81% (440/546) of all 1st-cycle FN events started in week 1 vs. 19% (106/546) in weeks 2-3. In pts not receiving G-CSF, the start of 1st-cycle FN events was more equally distributed: 41% (190/462) started in week 1 vs. 59% (272/462) in weeks 2-3. Results were robust to sensitivity analyses restricted to pts receiving ≥ 2 of T, A, and/or C. The rates of 1st-cycle FN events starting in weeks 1 and 2-3 with and without G-CSF following chemo initiation is shown below. Conclusions: Prophylactic G-CSF was highly effective for the prevention of FN in weeks 2-3, but relatively ineffective in week 1 of cycle 1 in the real-world setting, leaving pts largely unprotected during the first week. This represents an unmet medical need in week 1 of the cycle, despite use of G-CSF. Clinical trial information: NCT03294577. [Table: see text]