Abstract
To evaluate the efficacy and safety of recombinant human epidermal growth factor (rhEGF) in treating diabetic foot ulcers (DFUs), we conducted both database searches (PubMed, MEDLINE, EMBASE, CENTRAL, and Web of Science) and reference searches for randomised controlled trials from the inception of databases to 30 January 2020. Two reviewers independently scrutinised the trials, extracted data, and assessed the quality of trials. The primary outcome was the proportion of complete healing. The secondary outcomes were mean time to complete healing and adverse events. A subgroup analysis was performed by different administration routes. Statistical analyses were performed in RevMan 5.3. The time to complete healing Kaplan‐Meier curves was pooled in the R software. Of the 156 citations, 9 trials (720 participants) met eligibility criteria and were included. The rhEGF achieved a higher complete healing rate than placebo (OR: 2.79, [95% CI: 1.99, 3.99]). The rhEGF also significantly shorten complete healing time (MD: −14.10 days, [95% CI: −18.03, −10.16]). Subgroup analysis showed that topical application was superior to intralesional injection, but that may be because of different ulcer severity they included. No significant difference was shown in adverse events. Results were coherent with sensitivity analyses. Therefore, rhEGF is an effective and safe treatment for DFUs.
Keywords: diabetic foot ulcer, meta‐analysis, recombinant human epidermal growth factor, systematic review, wound closure
1. INTRODUCTION
Diabetic foot ulcers (DFUs) are often an intractable difficulty throughout the world, which are often notoriously difficult to treat. 1 For DFU patients, the quality of life collapses, while their economic burden soars. 2 About 19% to 34% of diabetic patients are likely to develop DFUs worldwide during their lifetime, and 9.1 to 26.1 million diabetic patients are affected by DFUs every year. 3 The mortality rate doubles in 10 years after diabetic patients encounter foot ulcers. 4 For those who survive, about 20% of moderate to severe diabetic foot infections leads to amputation. 5 It is estimated that, every minute, two diabetic patients lose at least a part of their lower limb somewhere in the world. 6 Even worse, DFU treatments account for almost one‐third of the total diabetic care cost. 7 In England, the annual cost of DFUs and amputation was about 1.3 to 1.5 billion dollars, more than the total costs of prostate, lung, and breast cancers. 8
At present, the standard management of DFUs includes surgical debridement, dressings, wound off‐loading, vascular evaluation, glycemic control, and antibiotics if needed. 7 And various agents are currently being researched as adjuvant therapies, such as human growth factors, negative pressure wound therapy, oxygen therapies, non‐surgical debridement agents, acellular bioproducts, skin grafts, and bioengineered skin. 7 This systematic review and meta‐analysis aimed to assess the efficacy and safety of recombinant human epidermal growth factor (rhEGF) for DFUs.
Physiologically, normal ulcer healing has four spatially and temporarily overlapping phases: haemostasis, inflammation, migration/proliferation, and remodelling. 9 Epidermal growth factor (EGF) is a member of the growth factor family that regulates cell migration, proliferation, and differentiation. 10 Previous research studies have illustrated that the decrease of growth factors and their cell‐surface receptors are the main cause of refractory DFUs. 11 Synthesised in Escherichia coli and functioning by binding to receptor kinases on target cells, the rhEGF can work as a mitogen and a differentiation factor for many types of cells. 10 However, the Diabetic Foot Guideline on the prevention and management of foot problems 12 did not recommend this treatment, although it was based on a low‐quality evidence. Therefore, to provide high‐quality evidence distinguishing the true therapeutic effect and safety issues, we conducted this systematic review and meta‐analysis by synthesising all available randomised controlled trials.
2. MATERIALS AND METHODS
This systematic review and meta‐analysis followed the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) 2009 guidelines. 13
2.1. Literature search
We searched the electronic databases including Ovid Medline, Embase, PubMed, the Cochrane Central Register of Controlled Trials (CENTRAL), and Web of Science without language restrictions from the inception of the databases to 30 January 2020. Medical Subject Headings (MeSH) such as “epidermal growth factor,” “diabetic foot,” and “randomized controlled trial” were used, and were combined with various free terms as the retrieval strategy. The search strategies were adjusted to each database. Furthermore, we conducted reference searches to identify additional eligible trials.
2.2. Eligibility criteria and exclusion criteria
We used the PICOS strategy to include eligible trials: (a) population (P): type 1 or type 2 diabetic patients with foot ulcers; (b) intervention (I): recombinant human epidermal growth factor (rhEGF); (c) comparison (C): placebo; (d) outcomes (O): not limited; and (e) study design (S): randomised controlled trial. Then, non‐human studies, reviews, protocols, and trials without interesting outcomes were excluded.
2.3. Primary and secondary outcomes
The primary outcome was the proportion of complete healing. Secondary outcomes were mean time to complete healing and adverse events. Complete healing was defined as complete (100%) wound re‐epithelialisation without any drainage or need for dressing.
2.4. Data extraction
Two reviewers screened the titles and abstracts independently, excluding apparently irrelevant manuscripts. The full texts of potentially eligible studies were obtained and double‐screened for eligibility (Figure 1). Disagreements were resolved through discussion.
FIGURE 1.
PRISMA flow diagram of the study selection for the systematic review and meta‐analysis
We collected the following information from each eligible trial: (a) publication data (authors, country, title of the trial, year of publication, study design, and universal trial number); (b) patient data (age, type of DM, duration of DM, HbA1c, and BMI); (c) ulcer baseline condition (duration of ulcers, ulcer size, location, Wagner grade, and ankle‐brachial index); (d) intervention (concentration, route, apply frequency, treatment duration, evaluation time, duration of follow‐up, and sample size); (e) control (sample size); (f) outcomes (proportion of complete healing, mean time to complete healing, and adverse events); and (g) statistical analysis (intent‐to‐treat or per‐protocol analysis). If both per‐protocol and intent‐to‐treat analyses were reported in the trial, we only extracted data from intent‐to‐treat analysis. Furthermore, we sent emails to the authors to acquire important information.
We also used Adobe Photoshop (version: CC 2019) and R software (version 3.6.1) to extract survival analysis data from each Kaplan‐Meier curve.
2.5. Assessment of risk of bias
We evaluated the risks of bias of the included trials according to the following domains of Cochrane risk‐of‐bias tool for randomised controlled trials: (a) random digital series generation; (b) allocation concealment; (c) blindness of patients and personnel; (d) blindness of outcome estimators; (e) incomplete outcome data; (f) selective reporting; and (g) other bias. The risk of bias of each trial was evaluated in the Review Manager (RevMan version 5.3) according to the Cochrane risk‐of‐bias tool for randomised controlled trials. 14 Besides, we drew a funnel plot to identify whether there was any publication bias.
2.6. Statistical analyses
We evaluated the clinical and methodological heterogeneities based on trial characteristics. If the trial characteristics had good homogeneity, we combined data quantitatively in RevMan. The Cochran Q test and the I 2 statistic were used to identify statistical heterogeneity. As the I 2 statistic was lower than 50%, the heterogeneity was regarded as acceptable, and a fixed‐effects model was applied. We also performed a subgroup analysis according to different administration routes.
We used the intent‐to‐treat approach to analyse all data. For binary outcomes, the results were presented as odds ratios (ORs). For continuous outcomes, mean differences (MDs) with corresponding 95% CIs were reported. All statistical tests were two‐sided at a significance level of 5%. Moreover, the time to complete healing Kaplan‐Meier curves were pooled in the R software (version 3.6.1) using Tierney's method. 15
Finally, we performed sensitivity analyses by removing each trial from the pooled analysis and recalculating the rest OR value each time.
This article is based on previously conducted trials, and none of the authors had contact with any human participants or animals.
3. RESULTS
3.1. Study selection and characteristics
A total of 154 citations were identified through systematic literature search and two additional citations were identified through reference search. Of the 156 citations, nine trials16, 17, 18, 19, 20, 21, 22, 23, 24 with 720 participants (404 randomised to the rhEGF group and 317 to the placebo group) met eligibility criteria and were included in the systematic review and meta‐analysis.
The baseline characteristics of patients in each trial were similar (Table 1). All trials were conducted in Asia, except one in Cuba. 22 Most patients included had type 1 or type 2 diabetes mellitus, whereas one trial 17 included only type 2 diabetes mellitus. Patients in most trials16, 17, 21, 22, 23 have a long diabetic history with a mean duration of more than 10 years. Regarding the severity of the ulcers, most trials included Wagner grades 1 and 2, whereas one trial 21 included Wagner grades 1 to 3, and one trial 22 included Wagner grades 3 and 4.
TABLE 1.
General characteristics of included trials
| Author/year | Country | Study design | Mean age (y) | Type of DM | Duration of DM (y) | Duration of ulcers (wk) | Ulcer size (cm2) | Ulcer location | Wagner grade | ABI | Total cases |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Viswanathan et al. 2019 | India | Stratified open‐labelled RCT |
I = 57.9 ± 9.6 C = 55.0 ± 6.8 |
Type 1 and 2 |
I = 10.7 ± 8.2 C = 12.9 ± 5.5 |
NA |
I = 9.1 ± 9.5 C = 8.4 ± 7.9 |
Toe (>50%) | Grades 1 and 2 | ABI >0.8 | 50 |
| Xu et al. 2018 | China | RCT |
I = 65 ± 3.65 C = 63 ± 4.56 |
Type 2 |
I = 13 ± 4.88 C = 12 ± 4.26 |
I = 16 ± 0.62 C = 13 ± 0.35 |
I = 4.7 ± 0.3 C = 4.2 ± 0.4 |
NA | Grade 2 | NA | 99 |
| Park et al. 2018 | South Korea | Double‐blind multicenter RCT |
I = 56.52 ± 12.71 C = 59.31 ± 12.64 |
Type 1 and 2 | NA | NA |
I = 2.80 ± 3.72 C = 2.35 ± 2.69 |
Toe, sole, heel, interphalangeal area, malleolus | Grades 1 and 2 | NA | 167 |
| Prabakar et al. 2016 | India | RCT |
I = 47.8 ± 2.35 C = 47.9 ± 2.16 |
NA | NA | NA | NA | Dorsum, plantar, thigh | Grade 2 | NA | 60 |
| Singla et al. 2014 | India | RCT |
I = 58.8 C = 55.84 |
Type 1 and 2 | NA | NA |
I = 19.56 C = 21.2 |
NA | Grades 1 and 2 | ABI > 0.75 | 50 |
| Gomez‐Villa et al. 2014 | Mexico | Double‐blinded multicenter RCT |
I = 62.1 ± 12.8 C = 55.1 ± 10.6 |
Type 1 and 2 |
I = 17.3 ± 10.0 C = 15.3 ± 8.4 |
I = 25.8 ± 44.0 C = 36.5 ± 75.8 |
I = 19.2 ± 15.7 C = 11.9 ± 11.8 |
Dorsum, sole, heel, others |
Grades 1, 2, and 3 | ABI > 0.6 | 34 |
| Fernandez‐Montequin et al. 2009 | Cuban | Double‐blind multicenter RCT |
I1 = 63 ± 25.4 I2 = 65·5 ± 27.5 C = 64·0 ± 32.7 |
Type 1 and 2 |
I1 = 19·5 ± 21.7 I2 = 15·0 ± 11.0 C = 15·0 ± 17.2 |
I1 = 4·3 ± 11.6 I2 = 4·3 ± 9.8 C = 4·9 ± 16.5 |
I1 = 28·5 ± 58.7 I2 = 20·1 ± 39.0 C = 21·8 ± 44.4 |
Toes, internal edge, external edge, dorsum sole, trans metatarsal heel | Grades 3 and 4 | NA | 149 |
| Afshari et al. 2005 | Iran | Single‐blind RCT |
I = 56.9 ± 12.7 C = 59.7 ± 12.3 |
Type 1 and 2 |
I = 12.6 ± 7.5 C = 14.9 ± 7.1 |
I = 6.13 ± 5.49 C = 8.53 ± 7.93 |
I = 0.88 ± 1.03 C = 1.03 ± 1.47 |
NA | Grades 1 and 2 | NA | 50 |
| Tsang et al. 2003 | China | Double‐blind RCT |
I1 = 68.76 ± 10.45 I2 = 62.24 ± 13.68 C = 64.37 ± 11.67 |
Type 1 and 2 |
I1 = 9.85 ± 7.79 I2 = 9.05 ± 6.19 C = 10.11 ± 8.29 |
I1 = 8.2 ± 5.6 I2 = 11.5 ± 14.7 C = 12.0 ± 15.47 |
I1 = 2.78 ± 0.82 I2 = 3.40 ± 1.1 C = 3.48 ± 0.82 |
Sole, toe, ankle, others | Grades 1 and 2 | ABI ≥ 0.7 | 61 |
Abbreviations: ABI, ankle‐brachial index; C, control group; I, intervention group; rhEGF, recombinant human epidermal growth factor.
There were two types of administration routes: topical application and intralesional injection (Table 2). The apply frequencies and treatment durations differed among trials (Table 2). For severity of the ulcers, trials using topical application enrolled Wagner grades 1 and 2, whereas intralesional injection enrolled Wagner grades 1 to 3 21 or Wagner grades 3 and 4. 22
TABLE 2.
Interventional details of included trials
| Author/year | Intervention | Control | Route | Apply frequency | Treatment duration | Evaluation time | Duration of follow‐up | Complete healing rate | Mean time to complete healing (d) |
|---|---|---|---|---|---|---|---|---|---|
| Viswanathan et al. 2019 | hEGH gel (Regen‐D) (n = 27) | Placebo (n = 23) | Topical application | NA | 30 d | NA | 30 d |
I = 78% (21/27) C = 52% (12/23) |
I = 45 ± 12 C = 72 ± 18 |
| Xu et al. 2018 | rhEGF 40 IU/cm2 (n = 50) | placebo (saline) (n = 49) | Topical application | Once a day | 60 d | NA | NA | NA |
I = 38.51 ± 1.46 C = 47.52 ± 1.82 |
| Park et al. 2018 | rhEGF 0.005% (50 lg/ml) (n = 82) | placebo (saline) (n = 85) | Topical application | Twice a day | Until ulcer healing or up to 12 wk | Weekly | 12 wk |
I = 73.2% (60/82) C = 50.6% (43/85) |
I = 56.00 C = 84.00 |
| Prabakar et al. 2016 | rhEGF (n = 30) | placebo (saline) (n = 30) | Topical application | NA | 14 wk | NA | 18 mo |
I = 83.3% (25/30) C = 66.67% (20/30) |
Reported patient‐level data |
| Singla et al. 2014 | rhEGF (Urogastrone) gel 15 g (n = 25) | betadine dressing (n = 25) | Topical application | NA | 8 wk | Every 2 wk | 8 wk |
I = 92.0% (23/25) C = 44.0% (11/25) |
Illustrated with Kaplan‐Meier curve |
| Gomez‐Villa et al. 2014 | rhEGF 75 mg/ml (n = 17) | placebo (n = 17) | Intralesional injection | Thrice per week | 8 wk | Every 2 wk | 8 wk |
I = 23.5% (4/17) C = 0% (0/17) |
NA |
| Fernandez‐Montequin et al. 2009 |
I1 = rhEGF 75 μg (n = 53) I2 = rhEGF 25 μg (n = 48) |
placebo (n = 48) | Intralesional injection | Thrice per week | 8 wk | Weekly | 12 mo |
I1 = 77·4% (41/53) I2 = 52·1% (25/48) C = 56·2% (27/48) |
I1 = 98 ± 37.8 I2 = 84 ± 60.2 C = 140 ± 172.3 |
| Afshari et al. 2005 | rhEGF 1 mg/1000 mg (n = 30) | Placebo (n = 20) | Topical application | Once a day, every day | 4 wk | Weekly | 4 wk |
I = 23.3% (7/30) C = 10% (2 /20) |
NA |
| Tsang et al. 2003 |
I1: rhEGF 0.02% (n = 21) I2: rhEGF 0.04% (n = 21) |
Actovegin 5% cream (n = 19) | Topical application | NA | 12 wk | Every other week | 24 wk |
I1 = 57.1% (12/21) I2 = 95.0% (20/21) C = 42.1% (8/19) |
Illustrated with Kaplan‐Meier curve |
Abbreviations: C, control group; I, intervention group; rhEGF, recombinant human epidermal growth factor.
3.2. Risk of bias
The risk of biases was a bit high (Figure 2). Only three trials (33.3%) reported how they generated random sequences, four trials (44.4%) reported how they concealed allocation, three trials (33.3%) blinded patients and personnel, and two trials (22.2%) blinded evaluators. One trial 19 did not report how they performed the RCT and barely did statistical analysis. Fortunately, it reported raw data, so we could calculate them.
FIGURE 2.
Risk‐of‐bias quality summary
The funnel plot showed an obvious asymmetry (Figure 3). After excluding two literature studies on intralesional injection, the funnel plot switched to be symmetric (Figure 4).
FIGURE 3.
Funnel plot
FIGURE 4.
Funnel plot (excluding intralesional injection trials)
3.3. Primary outcomes
The proportion of complete ulcer healing with rhEGF was significantly higher than that of placebo (OR: 2.79, [95% CI: 1.99, 3.99]) (Figure 5). Subgroup analyses based on different routes of administration illustrated that rhEGF by topical application had a much better therapeutic effect than by intralesional injection (Figure 5). The OR for topical application was 3.37 ([95% CI: 2.19, 5.20]), whereas that for intralesional injection was only 1.76 ([95% CI: 0.91, 3.41]).
FIGURE 5.
Proportion of complete healing
3.4. Secondary outcomes
The rhEGF group had a significantly shorter time to complete ulcer closure compared with the placebo group (MD: −14.10 days, [95% CI: −18.03, −10.16]). The pooled Kaplan‐Meier curves of time to complete healing with placebo (Figure 6) are remarkably lower and flatter than that with rhEGF (Figure 7). The slope in the rhEGF group increased significantly around 20 days after treatment initiation and eventually reached an increased height.
FIGURE 6.
Kaplan‐Meier curves of time to complete healing (placebo group)
FIGURE 7.
Kaplan‐Meier curves of time to complete healing (rhEGF group)
For serious adverse events (SAEs), two trials18, 22 reported wound‐related infection, three trials20, 22, 24 noticed amputation, and one trial 22 had one case of death after healing (Table 3). Fortunately, in most trials, the incidence rates were low and there was no significant difference between the rhEGF group and the placebo group. And a trial 20 found that rhEGF therapy significantly decreased the incidence of cellulitis. All the SAEs were not considered to be product‐related.18, 22 Other adverse events reported were mild and easily manageable, such as pain, burning sensation, chills, tremors, dizziness, nausea, and vomiting. Subgroup analysis illustrated that dizziness, 21 tremors, and chills 22 appeared more frequently in the rhEGF group through intralesional injection.
TABLE 3.
Adverse events
| Author/year | Proportion of adverse events | Serious adverse events (rhEGF group) | Serious adverse events (placebo group) | Wound‐related infections | Possible product‐related adverse Events | Amputation |
|---|---|---|---|---|---|---|
| Viswanathan et al. 2019 | NA | NA | NA | NA | NA | NA |
| Xu et al. 2018 17 | NA | NA | NA | NA | NA | NA |
| Park et al. 2018 |
I = 7.3% (6/82); C = 8.2% (7/85) |
I: 6 cases (7.3%) 1 superficial infection, 1 cellulitis, 1 acute myocardial infarction, 1 diarrhoea, 1 gastrointestinal haemorrhage, 1 acute renal failure |
C: 7 cases (8.2%) 3 superficial infections, 1 cellulitis, 1 diarrhoea, 1 idiopathic thrombocytopenic purpura, 1 coronary artery disease |
I = 2.4% (2/82); C = 4.7% (4/85) |
I = 0%; C = 0% |
I = 0%; C = 0% |
| Prabakar et al. 2016 | NA | NA | NA | NA | NA | NA |
| Singla et al. 2014 | NA | NA | NA |
I = 60% (15/25); C = 80% (20/25) |
NA |
I = 4% (1/25); C = 0% (0/25) |
| Gomez‐Villa et al. 2014 | NA | 0 | 0 | I = 0%; C = 0% |
I = 0%; C = 0% |
I = 0%; C = 0% |
| Fernandez‐Montequin et al. 2009 |
I1 = 69.8% (37/53); I2 = 58.3% (28/48); C = 64.5% (31/48) |
I1: renal failure, cellulitis I2: renal failure (lethal), myocardial infarct, pneumonia |
C: acute pulmonary, 2 oedema (1 lethal), cellulitis, knee abscess |
I1 = 7.5% (4/53); I2 = 8.3% (4/48); C = 4.2% (2/48) |
I1 = 0%; I2 = 0%; C = 0% |
I1 = 13.2% (7/53); I2 = 20.8% (10/48); C = 25% (12/48) |
| Afshari et al. 2005 |
I = 0%; C = 0% |
I = 0%; C = 0% |
I = 0%; C = 0% |
I = 0%; C = 0% |
I = 0%; C = 0% |
I = 0%; C = 0% |
| Tsang et al. 2003 | NA | NA | NA | NA | NA |
I1 = 9.5% (2/21); I2 = 0% (0/21); C = 10.5% (2/19) |
Abbreviations: C, control group; I, intervention group; rhEGF, recombinant human epidermal growth factor.
3.5. Sensitivity analyses
By removing one trial from the pooled analysis each time, we found that all the OR values for the proportion of complete wound healing did not significantly change. They were always higher than 2.4 (Figure 5). Moreover, the result barely changed when removing Prabakar's trial, 19 which provided us patient‐level data. Therefore, our result was quite robust.
Our result was most sensitive to Fernandez‐Montequin's trial, 22 which was a multi‐central RCT with intralesional injection. When removing this trial away, the OR value accelerated from 2.79 ([95% CI: 1.95, 3.99]) to 3.51 ([95% CI: 2.29, 5.37]) (Figure 8).
FIGURE 8.
Proportion of complete healing (sensitivity analyses)
4. DISCUSSION
This systematic review of RCTs assessed the efficacy and safety of rhEGF in patients with DFUs. A total of 720 diabetic patients with foot ulcers (404 randomised to the rhEGF group and 316 to the placebo group) were included in the meta‐analysis. Most patients had a long diabetic history and most trials were conducted in Asia.
There were already some meta‐analyses published.25, 26, 27 However, there are some common limitations in previous research. For example, the risk‐of‐bias quality was not valid enough. All the research studies classified the trials without blinding patients and personnel to have unclear performance bias, which might overestimate the quality of that study. Another flaw was that they did not include all the available relevant literature. All of them did not include an article published in 2016, 19 which provided us with patient‐level data, and that might produce selection bias. Moreover, previous research 26 used per‐protocol analysis, which might break the baseline character balance and overestimate the therapeutic effect. Therefore, we reconducted a systematic review and meta‐analysis by comprehensively searching the databases, including as many trials as we could, evaluating all the articles cautiously, and using intention‐to‐treat analyses.
The most encouraging finding of our research was that rhEGF could enhance ulcer healing. Besides a higher probability of wound recovery, rhEGF reduces healing time by about 2 weeks. The pooled Kaplan‐Meier curves of time to complete healing with rhEGF (Figure 7) are remarkably higher and steeper than that with placebo (Figure 6). The sensitivity analyses showed that our result was quite robust (Figure 8). A possible explanation for the great therapeutic effect might be the role growth factors plays in wound recovery. EGF could stimulate glycolysis, mitosis, and protein synthesis, so as to promote the proliferation of epidermal cells. 11 EGF could also improve the wound microenvironment and tissue's nutrition by inducing the migration of inflammatory cells away from the ulcer. 11
There were two types of administration routes in our meta‐analysis. Subgroup analyses showed that the topical application trials had quite low heterogeneity (I 2: 0%) and a better therapeutic effect. It seems the intralesional injection was less promising than topical application (I 2 for subgroup differences: 61.5%) (Figure 5). The OR value for topical application was 3.37 (95% CI: 2.19, 5.20), whereas that for intralesional injection was only 1.76 (95% CI: 0.91, 3.41) (Figure 5). Sensitivity analyses also showed the OR value significantly accelerated when removing a multi‐central RCT with intralesional injection 22 (Figure 8). Furthermore, dizziness, 21 tremors, and chills 22 appeared more frequently in the rhEGF group by intralesional injection. Therefore, it seems that intralesional injection is less effective and has more side effects than topical application. This is contrary to our previous intuition. Theoretically, the intralesional injection could directly shoot the rhEGF to the desired region and achieve a better therapeutic effect than topical application. The accessibility of growth factor in the ulcer's deeper layer is critical to acquire sufficient efficacy, 22 for active‐agent diffusion would be affected by necrotic tissue, wound proteases, and inflammation. 28 A possible explanation for this discrepancy may be that patients enrolled in these two kinds of research studies had different severity ulcers. When using topical applications, patients were Wagner grades 1 or 2. When using intralesional injections, patients also included Wagner grades 3 and 4 (Table 1), which were severer with infection. Therefore, to compare the true therapeutic effect, more trials with patents of the same Wagner grade are needed.
Although there were some adverse events reported in these trials and some were serious, including wound‐related infection,18, 22 amputation,20, 22, 24 and death after healing, 22 the incidence rates were low, and there was no significant difference between the rhEGF group and the placebo group (Table 3). All the SAEs were not considered to be product‐related.18, 22 Other adverse events reported were mild and easily manageable, such as pain, burning sensation, chills, tremors, dizziness, nausea, and vomiting. Therefore, based on the currently available evidence, rhEGF is a safe administration to DFUs.
The funnel plot showed an obvious asymmetry (Figure 3). Intuitionally, this indicates a potential publication bias, which might overestimate the therapeutic effect. However, subgroup analyses illustrated that the asymmetry mainly came from intralesional injection trials, which had more severe ulcers. After excluding them, the funnel plot switched to be symmetric (Figure 4). Therefore, the result is credible.
Our research has some strengths. First, to our knowledge, this is the most comprehensive and systematic meta‐analysis on rhEGF for DFUs. Besides systematic literature searching, we searched potential references. We also sent emails to the authors for missing data. Second, we used intent‐to‐treat analysis, and the baseline comparability was well preserved. Third, we pooled the time to complete healing Kaplan‐Meier curves. And how the therapeutic effects change as time passes by was illustrated visually. Fourth, we conducted subgroup analyses to compare two types of administration routes carefully. Finally, we performed sensitivity analyses to evaluate how robust our result was by omitting each trial and recalculate the OR values.
Our research also has some limitations. First, all included trials were performed in Asia, except one in Cuba. Whether it is as effective in other ethnic population is still uncertain. Second, there are still some potential biases. Only half trials reported how they avoid performance biases and detection biases during the blinding implementation periods. And more than half trials did not report how they avoid the selection biases. Third, economic analysis is important for patients with long diabetic history, especially in developing countries. However, none of the trials provided cost‐effectiveness data. Fourth, although we found different therapeutic effects between two types of administration routes, we failed to tell which is more superior, because the patients enrolled in these two types of trials had different levels of wound severity.
5. CONCLUSIONS
This systematic review and meta‐analysis demonstrated that rhEGF is effective and safe in treating DFUs. The rhEGF could both enhance ulcer recovery opportunities and accelerate wound healing. However, more studies are needed in different populations. And head‐to‐head trials are needed to assess the therapeutic effect between topical application and intralesional injection.
CONFLICT OF INTEREST
The authors declare no conflicts of interest.
FUNDING INFORMATION
No funding or sponsorship was received for this research or article publication.
Zhao D‐Y, Su Y‐N, Li Y‐H, Yu T‐Q, Li J, Tu C‐Q. Efficacy and safety of recombinant human epidermal growth factor for diabetic foot ulcers: A systematic review and meta‐analysis of randomised controlled trials. Int Wound J. 2020;17:1062–1073. 10.1111/iwj.13377
Ding‐Yun Zhao and Ya‐Na Su contributed equally to this study, and are co‐first authors.
Contributor Information
Jing Li, Email: lijing68@hotmail.com.
Chong‐Qi Tu, Email: chongqitu@yeah.net.
REFERENCES
- 1. Grennan D. Diabetic foot ulcers. JAMA. 2019;321(1):114. [DOI] [PubMed] [Google Scholar]
- 2. Boulton AJM, Vileikyte L, Ragnarson‐Tennvall G, Apelqvist J. The global burden of diabetic foot disease. Lancet. 2005;366(9498):1719‐1724. [DOI] [PubMed] [Google Scholar]
- 3. Armstrong DG, Boulton AJM, Bus SA. Diabetic foot ulcers and their recurrence. New England Journal of Medicine. 2017;376(24):2367‐2375. [DOI] [PubMed] [Google Scholar]
- 4. Walsh JW, Hoffstad OJ, Sullivan MO, Margolis DJ. Association of diabetic foot ulcer and death in a population‐based cohort from the United Kingdom. Diabet Med. 2016;33(11):1493‐1498. [DOI] [PubMed] [Google Scholar]
- 5. Lavery LA, Armstrong DG, Wunderlich RP, Tredwell J, Boulton AJM. Diabetic foot syndrome—evaluating the prevalence and incidence of foot pathology in Mexican Americans and non‐Hispanic whites from a diabetes disease management cohort. Diabetes Care. 2003;26(5):1435‐1438. [DOI] [PubMed] [Google Scholar]
- 6. IDF (International Diabetes Federation) . Diabetes Atlas. 8th ed. Brussels, Belgium: International Diabetes Federation; 2017. http://www.diabetesatlas.org/resources/2017-atlas.html. [Google Scholar]
- 7. Everett E, Mathioudakis N. Update on management of diabetic foot ulcers. Ann N Y Acad Sci. 2018;1411(1):153‐165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Kerr M, Barron E, Chadwick P, et al. The cost of diabetic foot ulcers and amputations to the National Health Service in England. Diabet Med. 2019;36(8):995‐1002. [DOI] [PubMed] [Google Scholar]
- 9. Marti‐Carvajal AJ, Gluud C, Nicola S, et al. Growth factors for treating diabetic foot ulcers. Cochrane Database Syst Rev. 2015;10:1–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Mohan VK. Recombinant human epidermal growth factor (REGEN‐D 150): effect on healing of diabetic foot ulcers. Diabetes Res Clin Pract. 2007;78(3):405‐411. [DOI] [PubMed] [Google Scholar]
- 11. Zhang J, Hu W, Diao Q, et al. Therapeutic effect of the epidermal growth factor on diabetic foot ulcer and the underlying mechanisms. Exp Ther Med. 2019;17(3):1643‐1648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Jones NJ, Harding K. 2015 international working group on the diabetic foot guidance on the prevention and management of foot problems in diabetes. Int Wound J. 2015;12(4):373‐374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Moher D, Liberati A, Tetzlaff J, Altman DG, Grp P. Preferred reporting items for systematic reviews and meta‐analyses: the PRISMA statement. BMJ—British Medical Journal. 2009;339:b2535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. Cochrane Collaboration: London, UK; 2011. [Google Scholar]
- 15. Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. Practical methods for incorporating summary time‐to‐event data into meta‐analysis. Trials. 2007;8:16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Viswanathan V, Juttada U, Babu M. Efficacy of recombinant human epidermal growth factor (Regen‐D 150) in healing diabetic foot ulcers: a hospital‐based randomized controlled trial. Int J Low Extrem Wounds. 2019;1534734619892791. [DOI] [PubMed] [Google Scholar]
- 17. Xu J, Min D, Guo G, Liao X, Fu Z. Experimental study of epidermal growth factor and acidic fibroblast growth factor in the treatment of diabetic foot wounds. Exp Ther Med. 2018;15(6):5365‐5370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Park KH, Han SH, Hong JP, et al. Topical epidermal growth factor spray for the treatment of chronic diabetic foot ulcers: a phase III multicenter, double‐blind, randomized, placebo‐controlled trial. Diabetes Res Clin Pract. 2018;142:335‐344. [DOI] [PubMed] [Google Scholar]
- 19. Prabakar A, Raj RS. Efficacy of recombinant epidermal growth factor in the healing process of diabetic ulcer. J Evol Med Dent Sci. 2016;5(31):1642‐1648. [Google Scholar]
- 20. Singla S, Garg R, Kumar A, Gill C. Efficacy of topical application of beta urogastrone (recombinant human epidermal growth factor) in Wagner's grade 1 and 2 diabetic foot ulcers: comparative analysis of 50 patients. J Nat Sci Biol Med. 2014;5(2):273‐277. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Gomez‐Villa R, Aguilar‐Rebolledo F, Lozano‐Platonoff A, et al. Efficacy of intralesional recombinant human epidermal growth factor in diabetic foot ulcers in Mexican patients: a randomized double‐blinded controlled trial. Wound Repair Regen. 2014;22(4):497‐503. [DOI] [PubMed] [Google Scholar]
- 22. Fernandez‐Montequin JI, Valenzuela‐Silva CM, Gonzalez Diaz O, et al. Intra‐lesional injections of recombinant human epidermal growth factor promote granulation and healing in advanced diabetic foot ulcers: multicenter, randomised, placebo‐controlled, double‐blind study. Int Wound J. 2009;6(6):432‐443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Afshari M, Larijani B, Fadayee M. Efficacy of topical epidermal growth factor in healing diabetic foot ulcers. Therapy. 2005;2(5):759‐765. [Google Scholar]
- 24. Tsang MW, Wong WKR, Hung CS, et al. Human epidermal growth factor enhances healing of diabetic foot ulcers. Diabetes Care. 2003;26(6):1856‐1861. [DOI] [PubMed] [Google Scholar]
- 25. Yang S, Geng Z, Ma K, Sun X, Fu X. Efficacy of topical recombinant human epidermal growth factor for treatment of diabetic foot ulcer: a systematic review and meta‐analysis. Int J Low Extrem Wounds. 2016;15(2):120‐125. [DOI] [PubMed] [Google Scholar]
- 26. Thien Quoc B, Quoc Van Phu B, Nemeth D, et al. Epidermal growth factor is effective in the treatment of diabetic foot ulcers: meta‐analysis and systematic review. Int J Environ Res Public Health. 2019;16(14):1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Yang Q, Zhang Y, Yin H, Lu Y. Topical recombinant human epidermal growth factor for diabetic foot ulcers: a meta‐analysis of randomized controlled clinical trials. Ann Vasc Surg. 2020;62:442‐451. [DOI] [PubMed] [Google Scholar]
- 28. Acosta JB, Savigne W, Valdez C, et al. Epidermal growth factor intralesional infiltrations can prevent amputation in patients with advanced diabetic foot wounds. Int Wound J. 2006;3(3):232‐239. [DOI] [PMC free article] [PubMed] [Google Scholar]