Clinical Study On Treating Burn Wounds With Recombinant Human Granulocyte-Macrophage Colony-Stimulating Factor Hydrogel

Granulocyte-macrophage colony-stimulating factor (GM-CSF) is a cytokine with pleiotropic functions. Found in lung-conditioned medium by Burgess et al., it was named for its ability to stimulate the formation of a granulocyte and macrophage colony. With progress in the basic and clinical research of GM-CSF, recombinant human GM-CSF (rhGM-CSF) has been widely used in the clinical treatment of many conditions. As the most attractive cytokine of research in the last 10 years, the application of rhGM-CSF in the treatment of wounds has become popular increasingly.GM-CSF has multifarious biologic effects such as stimulating proliferation and differentiation of hematopoietic progenitor cells in the myeloid and myeloblasts into neutrophils, eosinophils and macrophages. The topical application of GM-CSF on wounds starts a cascade reaction of trauma repair to improve wound healing. GM-CSF promotes the proliferation and activation of mature macrophages, neutrophils, endothelial cells, epithelial cells and fibroblasts, which are important in the reparative process after trauma. The underlying mechanisms of GM-CSF also include attracting inflammatory cells and endothelial cells with subsequent prevention from escaping the wound, inducing the proliferation and differentiation of epidermal keratinocytes, and attracting Langerhans cells into dermal layers. Locally administered GM-CSF can enhance the wound healing in diabetic mice, immuno-suppressed or adriamycin-treated rats. Other animal experiments also reported beneficial effects including burns, full-thickness wounds, chronic granulating wounds, doxorubicin induced tissue necrosis, radiation-impaired wound, as well as surgical incisions. Clinically, locally applied rhGM-CSF assists wound healing of diverse etiologies including venous ulcers, diabetic ulcers, pressure ulcers, sickle cell leg ulcers, hydroxycarbamide-related leg ulcers, hemoglobinopathy-related ulcers and so on.The topical application of rhGM-CSF on burn wounds has been only reported for wounds early after deep second-degree burns. Until now, whether rhGM-CSF can accelerate healing of residual burn wounds is not clear. Residual burn wounds are difference from wounds early postburn, which are non-healing chronic wounds with particular forming causation and mechanism. So it is necessary to perform a clinical trial to evaluate the efficacy of rhGM-CSF in the topical treatment of residual burn wounds.Although many animal experiments and clinical studies have reported the beneficial efficacy and safety of GM-CSF in the topical treatment of various wounds, there are some studies with negative results. A randomized, double-blind, placebo-controlled study has showed GM-CSF did not significantly accelerate the healing rate in murine excisional wounds model. Another animal experiment also found GM-CSF did not significantly decrease the incidence of incisional hernias in rats. Both in vitro and in vivo study demonstrated GM-CSF treatment suppressed specifically and dose-dependently collagen deposition and may impair tissue repair processes during surgery. Except that, rhGM-CSF was shown to have no effect on the healing of the superficial wounds in healthy volunteers, and have no effect on pressure ulcers if used without other growth factors.Owing to above inconsistent conclusions, it is necessary to conduct a systematic review of randomized controlled trials (RCT) of locally administered rhGM-CSF on wounds and adequately evaluate its efficacy and safety. The first relevant systematic review searched various databases up to January 2010, of which the Meta analysis about locally administered rhGM-CSF on burn wounds is incorrect. In addition, we have obtained the observation data of locally applied rhGM-CSF on residual burn wounds one year later. It is necessary to conduct a new systematic review of RCT for local rhGM-CSF administration on burn wounds and evaluate its efficacy and safety more adequately.Based on above-mentioned two aims, this study consists of two parts mainly: PartⅠ:A clinical trial of rhGM-CSF hydrogel for the treatment of residual burn wounds. PartⅡ:Systematic review for efficacy and safety of rhGM-CSF hydrogel for the treatment of burn wounds. In addition, the relevant studies on accelerating wound healing effect of GM-CSF were introduced for reference (PartⅢ).PartⅠ:A clinical trial of rhGM-CSF hydrogel for the treatment of residual burn woundsObjective:Aim to evaluate the efficacy and safety of topical rhGM-CSF hydrogel treatment for residual burn wounds and explore the potential mechanisms. Our study will help find a new way for treating residual burn wounds and reveal new pharmacological basis for drug exploiture.Methods:The study was a randomized, double-blind, placebo-controlled and self-controlled clinical trial. One hundred and thirty-eight residual burn wounds were divided into group A(n= 69, with treatment of rhGM-CSF hydrogel [State Drug Approval Document No:S20080003]) and group C (n= 69, with treatment of placebo), which were treated for 21 days totally. Local adverse reaction, wound healing time and wound healing rate, total effective rate at the 7th and 14th day after treatment were investigated. The wound tissue samples (n=6) around the residual burn wounds at different time points were obtained for taking count of blood capillaries and fibroblasts through optical microscope and evaluating the expression change of proliferating cell nuclear antigen (PCNA), transforming growth factor-β1 (TGF-β1) and vascular endothelial growth factor (VEGF) through EliVisionTM plus method. The statistical analysis was done with SPSS13.0 software. Intention-to-treat (ITT) analysis was performed. A two-sided P<0.05 was considered significant. Results:Sixty-nine patients with residual burn wounds were enrolled in this study. Nine patients dropped out of the trial. The data of sixty patients were analyzed. The wound diameter in group A was (3.70±0.69) cm, which was (3.63±0.72) cm in group C. There were no differences between two groups (P> 0.05). ITT analysis results:The wound diameter in group A was (3.76±0.72) cm, which was (3.71±0.75) cm in group C. There were no differences between two groups (P>0.1).The median wound healing time was 12 (95%CI:11-13) days in group A, which was significantly shorter than that in group C [18 (95%CI:17-19) days, P<0.001]. The wound healing rate in group A was (62±13)% and (95±10)% respectively at the 7th and 14th day after remedy, which was significantly higher than that in group C [(46±11)% and (83±12)%, P<0.001]. The total effective rates in group A (30.00% and 96.67%) were also higher than that in group C [3.33% and 83.33%, P<0.05]. The therapeutic efficacy of group A was significantly better than group C(P<0.001). The main adverse event after remedy was local irritating. Local slight red and swollen with mild pain was found in the wound of a few cases, which disappeared without management.ITT analysis results:The median wound healing time was 13 (95%CI:12-14) days in group A, which was significantly shorter than that in group C [18 (95%CI: 17-19) days, P<0.001]. The wound healing rate in group A was (61±14)% and (94±11)% respectively at the 7th and 14th day after remedy, which was significantly higher than that in group C [(45±11)% and (80±13)%, P<0.001]. The total effective rates in group A (29.00% and 95.65%) were also higher than that in group C [2.90% and 76.81%, P< 0.01]. The therapeutic efficacy of group A was significantly better than group C(P< 0.001).At the 7th and 14th day after remedy, histopathology showed the number of blood capillaries (10.30±0.58 and 14.47±0.93 respectively) and fibroblasts (143.1±10.3 and 137.8±6.9 respectively) in group A was significantly increased compared with group C (blood capillaries:7.23±0.70 and 10.37±0.77; fibroblasts 110.2±11.7 and 126.4±7.7) respectively (P<0.01 in all). It was found the positive expression of PCNA in the cell nucleus of fibroblasts and endothelial cells, TGF-β1 in the cytoplasm and surface of fibroblasts and endothelial cells, VEGF in the cytoplasm of endothelial cells, fibroblasts and some inflammatory cells. The expression intensity of PCNA, TGF-β1 and VEGF in wound tissue of group A was also significantly higher than that in group C at the 7th and 14th day after remedy.Conclusions:Topical rhGM-CSF hydrogel treatment can accelerate healing of residual burn wounds with superior safety. Its mechanism may be in association with promoting blood capillaries formation, inducing the proliferation or transferring of fibroblasts and increasing expression intensity of PCNA, TGF-β1 and VEGF in wound tissue.Part II:Systematic review for efficacy and safety of rhGM-CSF hydrogel for the treatment of burn woundsObjective:To systematic evaluate the efficacy and safety of topical rhGM-CSF hydrogel treatment of burn wounds and provide evidence-based medicine study information for clinical practice.Methods:According to research purposes, the appropriate detailed inclusion and exclusion criteria (including study characteristics, interventions and measurement of outcome) were constituted by PICO program (Participant, Intervention, Comparator, Outcome) and a systematic and comprehensive search strategy was formulated according to above-mentioned standards. We searched PubMed (1950 through Jan.2011.); EMBASE (1966 through Jan.2011.); Cochrane Central Register of Controlled Trials (up to Jan.2011.); WANFANG Database (1982 through Jan. 2011.); The Chinese Journal Full-text databases (1994 through Jan.2011.); The Chinese Biomedical Database (CBM)(1978 through Jan.2011.). The review used a subject and text word strategy with "burns; scald; electric injuries; chemical injuries; radiation injuries; granulocyte-macrophage colony-stimulating factor; molgramostim; leucomax; mielogen" and relevant Chinese terms as the primary search terms, handsearched 10 Chinese journals related to drugs for external use in burn wounds (Zhonghua Shao Shang Za Zhi, Chinese Journal of Trauma, Acta Academiae Medicinae Militaris Tertiae, Chinese Journal of Surgery, Chinese Journal of Reparative and Reconstructive Surgery, Chinese Journal of Practical Surgery, Chinese Journal of New Drugs, Chinese Journal of New Drugs and Clinical Remedies, Chinese Journal of Hospital Pharmacy, Chinese Journal of Modern Applied Pharmacy) and searched correlative websites, such as "google". The studies covered the references of eligible studies were additionally searched. Some other expand search methods were also used. At least two reviewers independently screen the studies for eligibility, evaluated the quality and extracted the data from the eligible studies, with confirmation of cross-check. Different opinions would be solved by the third party. The methodological quality of included studies such as randomization, allocation concealment, blinding and attrition was evaluated. The lacking data was obtained through contacting with author at full steam. Meta-analyses were conducted using Revman Analyses software (Revman 5.0.17) of the Cochrane Collaboration. We would use odds ratio (OR) for binary data, standardized mean difference (SMD) for continuous data with 95% confidence intervals (CI). The clinical heterogeneity would be analyzed before pooling data. Subgroup analysis would be performed if there was clinical heterogeneity. We would use Chi-square test to evaluate heterogeneity among various study with the test level of a=0.10. Heterogeneity was tested using the I2-statistic with values>50% and Chi-square test with P<0.1 indicating strong heterogeneity. Data without heterogeneity could be pooled using fixed effect model, and those with heterogeneity could be solved by sensitivity analysis, subgroup analysis or randomized effect model. A two-sided P<0.05 was considered significant. We will use non-quantitative synthesis for outcome that cannot be pooled or with a very low incidence.Results:Three studies (422 Patients with burn wounds) were included. There were 260 cases in the treatment group (with treatment of rhGM-CSF hydrogel) and 162 cases in the control group (with treatment of placebo). All included studies were graded in term of randomization, allocation concealment, blinding and attrition. Two studies were graded B and the other one was graded A. Subgroup analyses (two subgroups according to wound category) were performed because of the existed clinical heterogeneity.Meta analysis results:(1) Effects on the 8th day wound healing rate of rhGM-CSF hydrogel vs. placebo:Meta analysis within subgroups showed that SMD of the two subgroups were 0.51 (95%CI:0.27-0.75) and 1.44 (95%CI:1.04-1.84), P< 0.0001. Heterogeneity existed between subgroups (Chi2=14.89, P=0.0001, I2= 93%) and the random effect model was selected for pooling the effects. The pooled SMD was 0.96 (95%CI:0.05-1.87), P=0.04. Sensitivity analysis results indicated that the stability of statistical inference was good. (2) Effects on the 14th day wound healing rate of rhGM-CSF hydrogel vs. placebo:Meta analysis within subgroups showed that SMD of the two subgroups were 0.63 (95%CI:0.38-0.87) and 1.09 (95%CI:0.70-1.47), P<0.00001. Heterogeneity existed between subgroups (Chi2= 3.95, P=0.05, I2=75%) and the random effect model was selected for pooling the effects. The pooled SMD was 0.83 (95%CI:0.38-1.28), P=0.0003. Sensitivity analysis results indicated that the stability of statistical inference was good. (3) Effects on the wound healing time of rhGM-CSF hydrogel vs. placebo:Heterogeneity between subgroups (Chi2=83.36, P<0.00001,I2=99%) was too strong and pooling the effects was given up. Meta analysis within subgroups showed that SMD of the two subgroups were-7.79 (95%CI:-9.09--6.50) and-1.44 (95%CI:-1.86-1.03), P< 0.00001. (4) Effects on the 8th day total effective rate of rhGM-CSF hydrogel vs. placebo:Meta analysis within subgroups showed that OR of the two subgroups were 2.55 (95%CI:1.55-4.19) and 11.81 (95%CI:3.31-42.15), P<0.001. Heterogeneity existed between subgroups (Chi2=4.93, P=0.03,I2=80%) and the random effect model was selected for pooling the effects. The pooled OR was 4.89 (95%CI: 1.09-21.91), P=0.04. Sensitivity analysis results indicated that the stability of statistical inference was good. (5) Effects on the 14th day total effective rate of rhGM-CSF hydrogel vs. placebo:Meta analysis within subgroups showed that OR of the two subgroups were 3.01 (95%CI:1.47-6.17) and 5.80 (95%CI:1.21-27.73), P< 0.05. Homogeneity existed between subgroups (Chi2=0.57, P=0.45,I2=0%) and the fixed effect model was selected for pooling the effects. The pooled OR was 3.48 (95%CI:1.82-6.65), P=0.0002. Sensitivity analysis results indicated that the stability of statistical inference was good.Evaluation results of safety:The main adverse event of locally administered rhGM-CSF on burn wounds was local irritating. Local slight red and swollen with mild pain was found in the wound of a few cases, which disappeared without management. Any systemic adverse reaction related to rhGM-CSF hydrogel was not observed.Conclusions:Locally applied rhGM-CSF hydrogel has a positive effect on burn wounds with certain safety. Owing to possibility of selection bias in included studies, there may be a negative impact on evidence intensity of our results. We expect that future systematic review could involve more high quality evidences provided by high quality double-blind randomized control trials.