Current issues of ACP Journal Club are published in Annals of Internal Medicine


Granulocyte-colony stimulating factor hastened healing of severe foot infections in patients with diabetes mellitus

ACP J Club. 1998 Mar-April; 128:32. doi:10.7326/ACPJC-1998-128-2-032

Source Citation

Gough A, Clapperton M, Rolando N, et al. Randomised placebo-controlled trial of granulocyte-colony stimulating factor in diabetic foot infection. Lancet. 1997 Sep 20;350:855-9. [PubMed ID: 9310604]



To determine whether granulocyte-colony stimulating factor (G-CSF) is an effective adjuvant therapy for severe foot infections in patients with diabetes mellitus.


7-day, randomized, double-blind, placebo-controlled trial.


A U.K. diabetes foot clinic and associated hospital.


40 patients > 18 years of age (mean age 66 y, 73% men, 83% white) with diabetes and severe foot infections. Infections were defined as extensive cellulitis (acute spreading infections of the skin with involvement of the subcutaneous tissues clinically characterized by erythema > 2 cm in diameter, in association with purulent discharge). Exclusion criteria were an absolute neutrophil count < 1.0 or > 50 × 109/L, history of malignant conditions, HIV infection, renal insufficiency, hepatic disease, previous organ transplantation, immunosuppressive therapy, corticosteroid therapy, pregnancy, lactation, multiple organ failure secondary to septicemia, or critical leg ischemia.


All patients received antibiotics until cellulitis and ulcer discharge resolved. 20 patients received subcutaneous G-CSF for 7 days (5 µg/kg of body weight daily with lower subsequent doses on the basis of increased absolute neutrophil counts); 20 patients were allocated to placebo. Follow-up was 100%.

Main outcome measures

Antibiotic requirements and time to resolution of infection and hospital discharge. Secondary outcomes were needed for surgery, time to elimination of wound pathogens, and generation of neutrophil superoxide.

Main results

Patients in the G-CSF group did better than patients in the placebo group for all main outcome measures: proportion of patients with resolution of cellulitis at day 7 (55% vs 20%, P = 0.05) (Table); median time to hospital discharge (10 vs 17.5 d, P = 0.02), resolution of cellulitis (7 vs 12 d, P = 0.03), withdrawal of intravenous antibiotics (8.5 vs 14.5 d, P = 0.02), and negative results on swab culture (4 vs 8 d, P = 0.02); and increased production of neutrophil superoxide (P < 0.001). The groups did not differ for need for surgery (P = 0.1) or proportion of patients with healed ulcers at day 7 (P = 0.09).


Granulocyte-colony stimulating factor as adjuvant therapy for severe foot infections in patients with diabetes mellitus increased the proportion of patients whose infections healed and reduced the median time to hospital discharge, resolution of cellulitis, withdrawal of antibiotics, and negative results on swab culture.

Source of funding: Amgen Ltd.

For article reprint: Dr. A. Gough, Diabetes Centre, Birmingham Heartlands Hospital, Bordesley Green East, Birmingham B9 5SS, England, UK. FAX 44-121-685-5539.

Table. Granulocyte-colony stimulating factor (G-CSF) vs placebo*

Outcome at 7 d G-CSF EER Placebo CER RBI (95% CI) ABI |EER-CER| NNT (CI)
Resolution of cellulitis 55% 20% 175% (14 to 126) 35% 3 (2 to 20)

*Abbreviations defined in Glossary; RBI, ABI, NNT, and CI calculated from data in article.


The morbidity associated with diabetic foot infection persists as a major burden to patients, a therapeutic challenge for clinicians, and a reminder to all of the need for more preventive efforts. The study by Gough and colleagues evaluates an adjunctive measure, G-CSF, in the treatment of this serious and costly condition. The methods are clearly described and carefully done; the results are analyzed and presented in a comprehensible manner. A plausible and attractive case is made for using G-CSF along with antibiotics in the treatment of patients with cellulitis in diabetic foot disease. That said, the clinician must also be aware that the results are constrained by small sample size, imprecise description of the associated foot ulcers, focus on ulcers of short duration, and concurrent use of multiple antibiotics. Although cellulitis had to exceed 2 cm for inclusion in the study, the reader is left without a description of the relative severity of each patient's condition.

Further study of G-CSF is indicated, particularly with a larger sample that includes patients with ulcers of longer duration. However, the data presented certainly support the value of this adjunctive treatment. Because G-CSF has been used in infections in patients with neutropenia, physicians will be more familiar with this treatment than with a previously unknown treatment. Thus, given the positive differences noted in the group that received G-CSF—faster hospital discharge, quicker resolution of infection, earlier discontinuation of intravenous antibiotics, faster time to negative results on swab culture, and larger neutrophil superoxide production—clinicians should consider this treatment for their patients with diabetic foot infection.

Marvin H. Waldman, DPM, MPH, MS
Department of Veterans Affairs Medical CenterDetroit, Michigan, USA

Marvin H. Waldman, DPM, MPH, MS
Department of Veterans Affairs Medical Center
Detroit, Michigan, USA