Diabetic foot ulcers (DFUs) are among the most common complications of diabetes, with an estimated 15–25% risk of a patient with diabetes developing a DFU.1 It is estimated that by 2025, >125 million people will develop DFUs, of which 20 million people will undergo foot amputation.2,3,4 Approximately 50% of patients with infected DFUs who have a foot amputation die within five years.1,5,6 The average length of hospital stay among patients with a DFU was reported to be at least twice as long as that of patients with diabetes but who were ulcer-free.7 DFUs are an economic and social burden, causing high morbidity and mortality, loss of working hours and increased expenditure associated with diabetic foot problems.
The basic treatment process of DFUs includes offloading the affected extremity, treatment of the vascular pathology, debridement of the devitalised tissues, antimicrobial therapy, standard of care (SOC) and appropriate metabolic control. In complex DFUs with neuropathy, vascular failure and/or infection, the rate of amputation remains high despite best treatment. Thus, new products are urgently needed for the treatment of DFUs.
The role of growth factors in the healing mechanisms of DFUs, and their direct effect on the maintenance of tissue integrity and intercellular communication has been known for some time.8 Growth factor therapy is considered to be an adjuvant therapy to SOC in DFUs. Epidermal growth factor (EGF) directly affects wound healing mechanisms. EGF induces mitogenic, motogenic and cytoprotective actions which are crucial for healing. There are studies suggesting that intralesional administration of recombinant EGF is effective in treating DFUs.8,9,10,11,12
This study compares intralesional administration of recombinant EGF treatment (Hebeprot-p, 75μg, produced at the Center for Genetic Engineering, Cuba) with SOC of lower extremity DFUs.
Methods
The data of patients with DFUs were retrospectively analysed. The data of the patients who had had DFUs for at least four weeks and who had been treated in the wound clinic between January 2014 and 2017 were screened and included in the study.
The patients were divided into two groups. The study group consisted of patients in whom intralesional recombinant human EGF, Heberprot-p 75μg, was applied, and the control group consisted of patients in whom EGF was not applied.
Despite the indications for intralesional EGF treatment in many patients with DFUs who had been admitted to the author's clinic, Heberprot-p treatment could not be applied to these patients between 2016 and 2017. All patients with DFUs had to be treated with SOC treatment in the interim. Therefore, the authors retrospectively allocated the patients who were admitted to the clinic and had Heberprot-p treatment before 2016 to the study group and allocated the patients treated between 2016 and 2017 when the reimbursement of the drug had been paused to the control group. The reason for not using Heberprot-p in the patients enrolled in the control group was solely related to the reimbursement status of the drug in Turkey between 2016 and 2017. The Heperprot-p treatment was temporarily removed from the reimbursement list after changes in the reimbursement regulations were implemented in 2016 because of the high treatment cost.
The two groups were compared with respect to: age; haemoglobin A1c level; smoking; history of previous wound; presence of peripheral arterial disease (PAD); wound duration; wound stage according to Wagner classification system; presence of chronic renal failure; history of minor and major amputation before and during the treatment; limb salvage at the end of treatment; history of percutaneous transluminal angioplasty procedures before Heberprot-p treatment; hyperbaric oxygen and negative pressure wound treatment history before Heberprot-p treatment; and time to heal, and were found to be similar in terms of these characteristics.
Full granulation was accepted as the endpoint of the treatment. The wounds were then either closed surgically with split-thickness skin grafts or left to secondary healing and treated with best wound care (Fig 1 and 2).
A 61-year-old male patient with a diabetic foot ulcer on the left heel of six months' duration (a, b) and which was treated with 36 applications of Heberprot-p. The patient has peripheric arterial disease, cardiac insufficiency and chronic renal failure. After debridement and at commencement of Heberprot-p treatment (c). Progress of the wound during Heberprot-p treatment (d–g). At the end of treatment (h)
A 68-year-old male patient with an infected diabetic foot ulcer (DFU) of four months' duration. At 20 days before hospital admission, the wound became infected, and four toes were amputated in another clinic before being referred to our clinic. Heberprot-p treatment was initiated after surgical debridement of all infected soft tissues and bones. Full granulation was achieved after 12 application of Heberprot-p. The wound was closed with split-thickness skin graft after the wound bed is totally granulated. Infected DFU, second, third, fourth and fifth toes are amputated but the surgical site remains infected (a, b). After debridement of the foot and before initiation of Heberprot-p treatment (c, d). During treatment with Heberprot-p, growth of granulation tissue filling the wound cavity is observed (e, f). The foot has completely healed, following skin grafting (g–i)
All patients were screened for PAD at the time of outpatient clinic admittance. The patients whose ankle–brachial pressure index (ABPI) was <0.8 or >1.3 were sent for Doppler examination. If any sign of arterial flow obstruction and stenosis were found, then the patients underwent either peripheral arterial computerised angiography or conventional angiography, and were scheduled for revascularisation procedures. The result of percutaneous transluminal angioplasty procedure is generally accepted as successful if:
- There is inline flow from the groin to the ankle with <50% of residual stenosis in the target (treated) vessel
- There is an increase in ABPI of ≥0.15
- There is an increase in Rutherford Classification.
In the present study, the result of percutaneous transluminal angioplasty procedure was accepted as successful if there was inline flow from the groin to the ankle with <50% of residual stenosis in the target vessel.
Amputations above the ankle were accepted as major amputations. The Wagner classification was used to define the wound stage:12
- Grade 1: superficial ulcer
- Grade 2: deep ulcer extending to tendon, capsule or bone
- Grade 3: deep ulcer involving joint or bone
- Grade 4: localised gangrene of forefoot or heel
- Grade 5: gangrene of the entire foot.
The selection criteria for the Heberprot-p treatment were patients with DFUs at Wagner Grades 2 and 3. Patients with DFUs at Wagner grades 4 and 5 were treated with Heberprot-p following the debridement of all devitalised tissues and amputations, where the wounds were left to heal by secondary intention because of infection and/or insufficient arterial perfusion. The treatment was initiated after the wound stage became Wagner Grade 3 following amputation or debridement without any sign of infection. Broad-spectrum antibiotics were used to treat infections, determined according to tissue cultures from the wound and the blood analysis.
The patients were screened for osteomyelitis. All infected bones were amputated and debrided before Heberprot-p treatment. Heberprot-p treatment was initiated after the infection was treated and there was no sign of infection in the wound.
Exclusion criteria for Heberprot-p treatment was: presence of and/or known history of cancer; acute infected ulcer; cardiac insufficiency under treatment; and pregnancy. These are the standard exclusion criteria for Heberprot-p treatment, which the author routinely uses for every patient with a DFU admitted to their clinic.
The two groups received SOC during treatment. In the study group, Heberprot-p therapy was also administered. In both groups, DFUs were cleaned with sodium hypocloride solution (crystalin) at each dressing change. Wound dressings, including alginates, foams and hydrocolloids, were chosen according to the wound exudate and stage.
Negative pressure wound therapy (NPWT) and hyperbaric oxygen treatment (HBOT) were used in both groups whenever possible. Sharp debridement was performed whenever necessary to remove devitalised tissues. Offloading of the affected extremity was performed whenever possible. Laboratory tests were performed at admission and whenever required during treatment. Blood glucose was closely monitored and treated to maintain optimal metabolic control. Tissue cultures were routinely obtained at the time of admission and whenever there was suspicion of new wound infection.
The Heperprot-p injections were given three times a week, every other day. The injections were applied into the deep layers of the wound with a standard 26-gauge needle injector of 1cm in length, starting from the perimeter of the wound. At each patient visit, 5cc Heberprot-p was injected, equally divided across the whole wound area. Adverse events (AEs) were recorded during the treatment. The number of injections were determined according to the wound size, the healing rate, the patient's adherence with the treatment and any AEs.
In Turkey, healthcare reimbursement covers up to 36 injections. Despite the incomplete healing of the wound at the end of dose 36, the injections had to be stopped in some patients because of regulations and coverage of the reimbursement.
The primary endpoint was a complete granulation response, which was either complete healing or major amputation of the affected extremity. Time to complete granulation response, time to complete closure, duration of treatment, duration of follow-up, minor and major amputations, and mortality were recorded.
Ethical approval and patient consent
Institutional review board and local ethics committee approval was obtained (approval number: 07.01.2023-43473).
All patients were treated in the outpatient wound clinic. All patients were informed about the treatment and written consent was obtained before treatment, including for the use of photographs.
Statistical analysis
Statistical analysis was performed using the Statistical Package for the Social Sciences software (SPSS, version 20, IBM Corp., US). Continuous variables were reported as the mean±standard deviation, and categorical variables were presented as frequencies and percentages. Univariate comparisons of continuous variables were made using independent t-test. Categoric variables were compared using the Chi-squared test. Binary logistic regression analysis was used to examine the relationship between potential risk factors and treatment of Heberprot-p. Separate logistic regression analyses were used to first assess the individual impact of each variable. Subsequently, a logistic regression analysis, using all variables in a stepwise method, was performed to identify independent risk factors. Cox proportional hazard models were performed to examine predictors for reoperation. A two-sided p-value of <0.05 was considered statistically significant.
Results
A total of 51 patients were enrolled in the study. The study and control groups were composed of 29 patients and 22 patients, respectively. Both groups were found to be statistically similar with respect to age (p=0.63), sex (p=0.96), HbA1c (p=0.94), smoking (p=0.42), chronic renal failure (p=0.65), dialysis (p=0.51), history of previous wound (p=0.58) and duration of the wound (p=0.32). The two groups' demographics and study data are given in Table 1.
Table 1.
Table showing the demographics and the data of study and control groups
| Control group n (%) | Study group n (%) | p-value | |
|---|---|---|---|
| Number of patients | 22 | 29 | |
| Male | 17 (77.3) | 22 (75.9) | |
| Female | 5 (22.7) | 7 (24/1) | |
| Age, years, mean±SD, (range) | 64.5±8.6 (48–78) | 65.8±10.3 (47–92) | 0.63 |
| Mean HBA1c, %, mean±SD | 7.7±1.7 | 7.7±1.9 | 0.94 |
| Chronic renal failure | 7 (31.8) | 11 (38) | 0.65 |
| Need for dialysis | 5 (22.7) | 9 (31) | 0.51 |
| History of smoking | 12 (54.5) | 19 (65.6) | 0.42 |
| Mean wound duration, days, mean±SD, (range) | 95.2±115.3 (15–500) | 354.8±1211.9 (15–6570) | 0.32 |
| Wagner wound classification | 0.104 | ||
| W2 | 1 (4.5) | 0 | |
| W3 | 6 (27.3) | 4 (13.8) | |
| W4 | 15 (68.2) | 18 (62) | |
| W5 | 0 | 7 (24) | |
| History of previous wound | 6 (27.3) | 10 (34.5) | 0.58 |
| Mean treatment time, weeks, mean±SD | 6.9±4.2 | 3.6±1.3 | 0.000 |
| Mean follow-up time weeks, mean±SD | 16.1±8.7 | 12.6±8.2 | 0.358 |
| Exitus | 2 (9.1) | 5 (17.2) | 0.402 |
| Limb salvage at the end of treatment | 18 (82.0) | 25 (86.2) | 0.67 |
| Wound healing at the end of treatment | 9 (41.0) | 22 (76.0) | 0.011 |
| Minor amputation | 10 (45.5) | 8 (27.6) | 0.86 |
| Major amputation | 4 (18.2) | 3 (10.3) | 0.343 |
| Patients who underwent percutaneous transluminal angioplasty | 14 (63.6) | 24 (82.8) | 0.121 |
SD—standard deviation
The mean treatment time in the control group and study group were 6.97±4.18 weeks and 3.65±1.34 weeks, respectively. The treatment time was statistically shorter in the study group (p=0.000). The wound healing rates were 40.9% in the control group and 75.8% in the study group. The healing rate was significantly higher in the study group (p=0.011).
Overall, 14 (63.6%) patients in the control group and 22 (75.7%) patients in the study group underwent a revascularisation procedure before Heberprot-p treatment.
The two groups were similar in terms of wound stage according to the Wagner classification system (p=0.104); however, wound stage was found to be a significant risk factor for wound healing (p=0.014). It was also shown that wound stage was a risk factor for Heberprot-p treatment (p=0.044), which meant that Heberprot-p treatment was applied to patients with higher Wagner stages.
The demographics with regards to wounds, including wound localisation, wound size, presence of infection before treatment and ABPI of the affected extremity before percutaneous transluminal angioplasty procedure, is given in Table 2.
Table 2.
Table showing wound localisation, wound size, presence of infection before treatment and ankle-brachial index of the affected extremity before percutaneous transluminal angioplasty procedure
| Control group (n=22) | Study group (n=29) | p-value | |
|---|---|---|---|
| Wound size, cm2, mean±standard deviation (SD) | 20.23±14.60 | 27.72±20.30 | 0.30 |
| Ankle–brachial pressure index, mean±SD | 0.72±0.13 | 0.75±0.14 | 0.404 |
| Patients with infected ulcer before treatment, n (%) | 15 (68.2) | 19 (65.5) | 0.842 |
| Result of treatment | n (%) | n (%) | |
| Full granulation | 5 (27.7) | 5 (17.2) | |
| Complete healing | 4 (18.2) | 17 (58.6) | |
| No healing | 13 (59.1) | 7 (24.1) | |
| Death during treatment | 2 (9.1) | 5 (17.2) | 0.402 |
| Location of wound (number of patients) | |||
| Sole | 9 (40.9) | 6 (20.7) | |
| Heel | 2 (9.1) | 4 (13.8) | |
| Foot dorsum | 1 (4.5) | 1 (3.5) | |
| Inner/outer side of foot | 0 (0.0) | 2 (6.9) | |
| Forefoot amputation stump | 2 (9.1) | 3 (10.3) | |
| Below knee amputation stump | 0 (0.0) | 1 (3.5) | |
| Toe amputation stump | 8 (36.4) | 10 (34.5) | |
| Toe amputation stump and foot dorsum | 0 (0.0) | 1 (3.5) | |
| Toe amputation stump and sole | 0 (0.0) | 1 (3.5) | |
There were 10 (45.5%) minor and four (18.2%) major amputations in the control group, and eight (27.6%) minor and three (10.3%) major amputations in the study group during the treatment period. When the two groups were compared in terms of minor and major amputations, no significant difference was detected (p=0.186 and p=0.343, respectively). When the two groups were compared in terms of limb salvage rate, no significant difference was observed between the groups (82% and 86%, respectively; p=0.67).
During the treatment period, two (9.1%) patients died in the control group and five (17.2%) patients died in the study group. All deaths were related to cardiac causes. There was no statistical difference between the two groups (p=0.402).
The mean follow-up times were 16.09±8.72 weeks and 12.55±8.18 weeks for control and study groups, respectively. There was also no statistical difference between the follow up times in both groups (p=0.358).
There were 14 (63.6%) patients in the control group and 22 (75.9%) patients in the study group who received a percutaneous transluminal angioplasty procedure during their treatment period. The difference in the two groups were not statistically significant (p=0.121). When the relationship between percutaneous transluminal angioplasty procedure and the need for amputation and wound healing were investigated, no relationship was found to either parameters (p=0.811 and p=0.553, respectively).
Discussion
DFUs cause substantial morbidity and impaired QoL, and result in high treatment costs. A DFU is the most important risk factor for lower-extremity amputation.1 Apart from repetitive biomechanical stress and impaired tissue perfusion, healing in DFUs is intrinsically impaired. Complex molecular interactions are damaged as a result of prolonged systemic and local inflammation. There is excessive production of advanced glycation end-products; these end-products are toxic to the vascular system.13 This functional impairment is also shown to be related to the reduced levels of active growth factors in the wound environment.14
EGF is a 53–amino acid polypeptide that exerts potent mitogenic and motogenic activity through binding to its specific cell membrane receptor.15,16 It specifically stimulates fibroblast migration to the ulcer area, induces formation of the granulation tissue through de novo angiogenesis, induces wound contraction through myofibroblast formation and activation, and helps with the re-epithelialisation of the ulcer by stimulating keratinocyte proliferation and migration.17
Heberprot-p contains EGF, which is produced using recombinant DNA. Its biological mechanism is similar to endogenous EGF in terms of stimulating the migration and proliferation of fibroblasts, keratinocytes, endothelial and other cells.18 Recombinant EGF was first produced by the Cuban Center for Genetic Engineering in 1988 and licensed in 2006 to be used as an adjuvant therapy to standard wound care in DFUs. It is licensed in 26 countries worldwide and has been reimbursed and used in Turkey since 2012. In 2018, it was estimated that more than 159, 000 patients around the world were treated with this therapy.19
The primary clinical outcome of Heberprot-p treatment is the increase in vascularity of the wound bed, detected as an increase in granulation tissue. A significantly increased rate in the granulation tissue after eight weeks of treatment (with three injections of Heberprot-p 75μg per week) was reported by Fernandez-Montequ et al.20 and Ertugrul et al.21 Fernandez-Montequ et al. reported a 77% increase in granulation after Heberprot-p treatment versus a 56% increase with standard care.20 Ertugrul et al. reported 75% healing in the wound bed (referred to as ‘complete granulation’ in the study). In the present study, the healing rate of the study group was 75.8% compared with 40.9% in the control group which correlates with similar studies in the literature.20,21
Ertugrul et al.22 reported that the presence of osteomyelitis and a DFU duration of ≥75 days had a statistically significant negative effect on wound healing. The presence of osteomyelitis was not recorded in the present study, but Heberprot-p treatment was initiated after all infected and devitalised tissue and toes were debrided and amputated. However, DFU duration was not found to be a significant risk factor for delayed wound healing. None of the other variables investigated in Ertugrul et al.'s22 study, such as age, diabetes type, renal failure, renal dialysis, wound location and size, presence of PAD or severity of infection, that were found to have a significant effect on healing were found to be significant in our study.
Garcia–Herrera et al.11 found healing time to be 2.1 times quicker with use of Heberprot-p treatment. Unlike the present study, the authors detected that a shorter duration of diabetes, a lower duration of ulceration, female sex and older age were all indicators for early ulcer healing.18
Presence of chronic renal failure or dialysis is another concern in the healing of DFUs. Some researchers preferred to exclude these patients from their studies. Presence of chronic renal disease was not shown to have any effect on wound healing in the present study and the two groups were found to be similar with respect to the presence of renal disease. Ertugrul et al.21 also reported that renal failure did not have any negative effect on Heberprot-p treatment.
In contrast to the literature,18,21,24 Heberprot-p treatment in the present study showed no significant decrease in amputation rates; however, when the study and the control group demographics were evaluated, it was observed that the study group was composed of patients with higher morbidity and risk factors. The mean Wagner classification, the mean HbA1c level, the rate of chronic renal failure and number of dialysis patients were statistically higher in the study group; in other words, Heberprot-p treatment was applied to patients whose DFUs were more difficult to manage and less expected to heal with standard wound care. Despite this higher rate of morbidity and risk factors in the control group, the rate of wound healing was statistically better after Heberprot-p treatment. The reported rates of lower extremity amputation for patients with diabetes with a severe foot ulcer are generally 14–25%.22,23,25
Ertugrul et al.21 reported the lowest major amputation rate at 2.9%, Yera-Alos et al.10 reported it as 9.3%10, Velazquez et al.26 as 9.38% and Guillermo et al.27 as 10.7%. In a comparative study, Gonzalez–Acosta et al.24 found that, compared with standard wound therapy, Heberprot-p treatment added to SOC therapy was associated with a lower rate of major amputation (26.7% versus 8.3%, respectively).
The rates of both minor and major amputations in both groups in this present study were in accordance with the literature. When the two groups were compared in terms of minor and major amputations, no statistically significant difference was detected. However, the Heberprot-p treatment decreased the amputation rate by nearly half compared with standard treatment.
The treatment duration was not in correlation with the time to wound healing; even though wound treatment was prolonged in hard-to-heal wounds, applying Heberprot-p did not change the overall outcome. On the other hand, the treatment time was statistically shorter in the study group.
The cost-effectiveness of any treatment modality is of major concern in every reimbursement system. At the time of writing, a single flacon of Heberprot-p cost approximately $1100 USD, excluding the application fee and the cost of standard wound care treatment. The clinician must demonstrate evidence for choosing to use more sophisticated, expensive treatment modalities than SOC. In the absence of double-blinded randomised controlled trials to show the clinical benefit of Heberprot-p treatment, the author suggests, based on their clinical experience, a minimum of 12 Heberprot-p injections, which is in concordance with other studies in literature, and to look for measurable macroscopic changes in the wound, terminating the treatment if no favourable outcome is documented.
In a study by Eggert et al.,28 the total healthcare expenditure of a patient with lower extremity amputation due to DFU infection was reported to range between $66,300–73,000 USD.28 When compared with the socioeconomical profit in saving a patient's limb along with the cost of 12 flacons of Heberprot-p treatment (12×$1100=$13,200), Heberprot-p treatment can be accepted as a cost-effective DFU therapy.
Limitations
The study has some limitations. All patient records were collected from outpatient wound clinic patient data. For practical purposes, every wound was recorded using the Wagner classification system. There are some drawbacks to using the Wagner classification system as it is based on dimension (width and depth) but does not include two very basic factors—infection and perfusion. Therefore, the author tried to exclude these factors by stating that the treatment was only applied to patients after infections were treated and by stating the number of patients in whom PAD were treated (14 patients in control group; 24 patients in the study group).
In the present study, the result of percutaneous transluminal angioplasty procedure was accepted as successful if there was an inline flow from the groin to the ankle, with residual stenosis <50% in the target vessel. Questions such as: ‘To what extent were patients ischaemic and did percutaneous transluminal angioplasty interventions provide enough perfusion?’ still remain unanswered.
Another drawback of the study is the use of NPWT and HBOT in standard wound care. These two treatment modalities are known to increase the healing rate in wounds and it may be hard to attribute the significant healing in the Heberprot-p group solely to the drug in the presence of these treatment modalities. However, the two groups both received these two treatments before Heberprot-p treatment and were found to be statistically similar with respect to these treatments.
Conclusion
Heberprot-p is an effective, new generation of drug to heal recalcitrant DFUs. In the era of increasing prevalence of diabetes and its complications, Heberprot-p, with its unique mechanism of action, should be present in every wound caregiver's armamentarium. As far as the author is concerned, this is the first study in Turkey comparing SOC with Heberprot-p treatment in patients with DFUs. If the claims in the literature that Heberprot-p lowers amputation rate are correct, then the low amputation rate and the shorter treatment time may outweigh its treatment costs.
Reflective questions
- What is the role of growth factors in wound healing?
- How does epidermal growth factor function in diabetic foot ulcer (DFU)?
- What are the benefit(s) of using an intralesional application of EGF as part of standard of care or as an adjuvant therapy in the treatment of DFUs?