Background
Phase III, randomized, controlled multicenter trial. 123 patients with severe chronic neutropenia (SCN; absolute neutrophil count [ANC] <0.5 × 10⁹/L) and recurrent infections due to idiopathic, cyclic, or congenital (Kostmann) neutropenia. Pediatric and adult patients included. Key entry criterion: history of recurrent serious bacterial infections attributable to neutropenia.
Interventions and follow up
Arm A: Filgrastim (G-CSF) 3.45–11.50 μg/kg/day SC (dose-adjusted to ANC ≥1.5 × 10⁹/L)
Arm B: 4-month observation period (no treatment), followed by crossover to filgrastim
Primary endpoint: ANC response and incidence/duration of infection-related event
mFollow up: Duration of observation period + treatment period
Arm B: 4-month observation period (no treatment), followed by crossover to filgrastim
Primary endpoint: ANC response and incidence/duration of infection-related event
mFollow up: Duration of observation period + treatment period
Results
ANC ≥1.5 × 10⁹/L achieved: 108/120 patients (90%) on filgrastim
Infection-related events: ~50% reduction in incidence and duration vs observation period, P<.05
Antibiotic use: ~70% reduction in duration with filgrastim vs observation
Infection-related events: ~50% reduction in incidence and duration vs observation period, P<.05
Antibiotic use: ~70% reduction in duration with filgrastim vs observation
Adverse events
Main adverse events: Asymptomatic splenomegaly in a majority — frequent but clinically significant in only a few. Bone pain in most patients (managed with analgesics). Headache and rash common. No immediate life-threatening toxicities. Long-term follow-up data (from SCNIR) later showed risk of MDS/AML, particularly in congenital neutropenia — not observed in this trial.
Conclusions
Filgrastim effectively raised ANC and reduced infection-related events in patients with severe chronic neutropenia, regardless of etiology (idiopathic, cyclic, or congenital). This trial provided the pivotal evidence for FDA approval of filgrastim for SCN in 1994 and established G-CSF as the cornerstone of SCN management.
Key Limitations
Key Limitations: Crossover design means the observation period serves as control — not a true parallel placebo-controlled RCT, limiting blinding. Dose variability was high (3.5–11.5 μg/kg/day), making standard dosing difficult to define. Long-term complications including MDS/AML were not assessable within this trial window; registry data later identified transformation risk in congenital (Kostmann) SCN, possibly related to underlying CSF3R mutations rather than G-CSF itself. Cyclic and idiopathic neutropenia patients have better prognosis than congenital forms.
Clinical Context
Filgrastim (G-CSF) is FDA-approved for SCN since 1994 and remains the standard of care. Patients with congenital neutropenia are monitored for MDS/AML transformation (SCNIR registry data: ~8-10% risk at 10 years). CSF3R mutations (which arise somatically in ~25% of congenital SCN) are predictive of progression. Allogeneic HSCT is offered for patients unresponsive to G-CSF or with evidence of transformation. Cyclic and idiopathic forms generally require lower G-CSF doses with a lower transformation risk.
References
References: Dale DC et al, Blood 1993 (filgrastim SCN)