Costs and benefits of dose-dense chemotherapy scheduling in Hong Kong Chinese patients with primary breast cancer

Abstract

Background: Routine prophylactic administration of recombinant granulocyte colony-stimulating factor (G-CSF, filgrastim) has enabled more frequent scheduling of adjuvant cytotoxic drugs in primary breast cancer. Dose-dense chemotherapy of this kind has become popular in recent years, but few studies have compared the costs and benefits of this treatment approach. Methods: We conducted a retrospective comparison of 150 Hong Kong Chinese primary breast cancer patients treated with either conventional 3-weekly chemotherapy or G-CSF-supported 2-weekly (dose-dense) chemotherapy using similar cytotoxic regimens (four cycles of doxorubicin and cyclophosphamide, followed by four cycles of taxane) between April 2004 and May 2006 in two teaching hospitals. Assessments of treatment tolerance, toxicity, scheduling, and expense were applied to both patient cohorts. Results: Compared to conventionally treated patients (n = 117), dose-dense patients (n = 33) benefited from 52.3% fewer chemotherapy delays (21.2% vs. 44.4%; p = 0.02), 79.3% fewer hospital admissions for febrile neutropenia (3.0% vs. 14.5%; p = 0.07), and 54.8% enhancement of cytotoxic dose intensity (p < 0.001). Dose-dense patients also developed less nausea (p < 0.001) and stomatitis (p = 0.01) but more frequent bone pain (p < 0.001). After subtracting costs saved by fewer hospital admissions, dose-dense patients incurred higher total expenses approximating US$500 per cycle. Conclusion: Routine G-CSF administration to Hong Kong Chinese patients with primary breast cancer is associated with major improvements in cytotoxic drug delivery, which could plausibly translate into greater therapeutic efficacy in some patients. In addition, dose- dense treatment is shorter, safer, and more reliably scheduled, and is associated with unexpected improvements in nausea and stomatitis. The extent to which these gains justify the rise in net costs needs to be further defined in prospective randomized studies, both in different adjuvant patient subsets and in differently resourced socioeconomic contexts.