Aims/hypothesis. To re-evaluate the use of Granulocyte-Colony Stimulating Factor (G-CSF) in the treatment of infected diabetic foot ulcers.

Methods. Thirty-seven diabetic subjects were randomised to Granulocyte-Colony Stimulating Factor (G-CSF) (n=20) or placebo (n=17). The primary endpoint was resolution of cellulitis, which was evaluated clinically and with an infection summary score. Patients were hospitalised for 10 days and received subcutaneously either 5 µg/kg G-CSF or placebo daily. Ulcers were treated with a standard wound protocol and the patients were instructed to stay in bed. All subjects received antibiotics (clindamycin and ciprofloxacin) intravenously until the inflammation had subsided.

Results. Patients who received G-CSF did not have an earlier resolution of clinically defined cellulitis (p=0.57). The infection summary score declined, but comparably, in both groups (G-CSF: 29.5±18.4 to 6.7±6.3 p<0.001, placebo: 24.2±16.9 to 8.9±7.2 p<0.001). The ulcer volume, which was not greater among placebo patients, was reduced by 59% in G-CSF and by 35% in placebo patients.

Conclusion/interpretation. We conclude that antibiotic and non weight-bearing therapy (bed rest) accelerated the resolution of cellulitis in infected foot ulcers. Additional treatment with G-CSF had no further beneficial effect.