Introduction: Neutropenia and its complications are major dose-limiting toxicities of systemic cancer chemotherapy. Febrile neutropenia (FN) generally prompts immediate hospitalization for evaluation and empiric broad-spectrum antibiotics adding to the cost of cancer care. Randomized controlled trials (RCTs) have demonstrated that prophylactic recombinant granulocyte colony-stimulating factor (rG-CSF; filgrastim) is capable of reducing the risk of FN, infection related mortality (IRM) and duration of hospitalization associated with FN. In the absence of compelling clinical indications, current guidelines support the use of prophylactic G-CSF only when using regimens associated with FN risk of 40% or higher. While recent RCTs have demonstrated that a long acting, pegylated form (pegfilgrastim) is at least as effective and safe as filgrastim, its economic impact has received little study.

Methods: An economic analysis was conducted comparing prophylactic pegfilgrastim to no prophylactic growth factor in patients receiving systemic chemotherapy. Direct medical cost estimates (US$ 2002) were obtained from the University HealthSystem Consortium and Marketscan databases. Pegfilgrastim costs were based on the average wholesale Red Book© prices. Risk and efficacy estimates were based on recent RCTs and meta-analysis of prophylactic rG-CSF. Sensitivity analyses and estimation of cost neutral thresholds for the range of feasible values of all variables were conducted based on cost-minimization.

Results: Under baseline conditions (FN= 17%; RRR=0.941), an incremental cost savings with pegfilgrastim of $428 was estimated. At baseline, a cost neutral threshold for FN risk of 14.6% was estimated with an expected cost for either strategy of $2721. Sensitivity analysis demonstrates a baseline threshold for RRR of 0.74 with lower rates of RRR supported by increasing hospital cost or length of stay, increasing infection-related mortality or incorporation of indirect costs. Threshold estimates for FN risk were robust over all possible values assumed for RRR of infection related mortality. The FN risk threshold falls below 10% when hospitalization cost per day exceeds $3000. Further results, including a Monte-Carlo simulation will be presented.

Conclusions: Incorporation of recent RCTs and cost data into clinically relevant models demonstrates that pegfilgrastim is cost saving at levels of FN risk associated with common chemotherapy regimens (< 20%). Primary prophylaxis with pegfilgrastim should be considered in patients being treated for curative intent with moderately myelosuppressive regimens.

Author notes

Corresponding author

Sign in via your Institution