Background: The combination of docetaxel and cyclophosphamide (TC) for adjuvant treatment of early stage breast cancer improves overall survival compared with doxorubicin and cyclophosphamide (AC) (Jones et al., 2006). Although cardiotoxicity is avoided with TC, the risk of febrile neutropenia (FN) is higher. For TC, reported rates of FN without prophylactic granulocyte colony-stimulating factor (G-CSF) range from 5% in the phase III trial to as high as 46% in retrospective chart reviews. G-CSF is not covered by our provincial cancer funding agency for primary prophylaxis of FN with TC chemotherapy, however it is often prescribed for patients with private insurance. Our aims were twofold: i) to determine the incidence of FN with TC chemotherapy with and without prophylactic G-CSF or antibiotics in two Ontario comprehensive cancer centres, and ii) to evaluate the cost-effectiveness of primary prophylaxis with G-CSF vs. antibiotics.

Methods: Patients who received adjuvant TC chemotherapy between January 1, 2008 and December 31, 2012 were identified through pharmacy databases. Electronic charts were retrospectively reviewed to extract patient characteristics, treatment details including G-CSF and antibiotic use, as well as incidence of FN and duration of hospitalization. A Markov model comparing primary G-CSF prophylaxis, primary antibiotic prophylaxis and secondary G-CSF prophylaxis was constructed to compare the cost-effectiveness of these strategies over a four cycle time horizon. Costs were based on resource utilization from this retrospective cohort and supplemented by the published literature, adjusted to 2012 Canadian dollars. The model took the perspective of the third party payer. Both one-way and probabilistic sensitivity analyses were performed.

Results: 340 patients were treated with TC over the study period. Of the 73 (21%) who did not receive any primary prophylaxis with G-CSF or antibiotics, 23 (32%) developed FN requiring hospitalization and treatment with intravenous antibiotics. However, only 2 of the 192 patients (1%; P <0.0001) who received primary G-CSF prophylaxis (funded by the patient or a third party payer), and 6 of the 53 patients (11%; P <0.01) who received primary antibiotic prophylaxis (97% receiving ciprofloxacin) developed FN. Age ≥65 was a significant risk factor for FN in the absence of G-CSF (56% vs. 25%, P = 0.02). The results of the cost-effectiveness analysis will be presented at the meeting.

Conclusions: The FN rate associated with TC chemotherapy without primary prophylaxis exceeds 30% but may be reduced with prophylactic antibiotics or G-CSF. Unless prophylactic antibiotics are substantially more cost-effective than prophylactic G-CSF for TC chemotherapy in a particular region or country, primary prophylactic G-CSF should be funded, given its greater effectiveness than antibiotics and the global need to minimize the emergence of antibiotic resistance.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P6-07-06.