It remains unclear whether electrical currents can affect biological factors that determine chronic wound healing in humans. Purpose: The aim of this study was to determine whether anodal and cathodal high-voltage monophasic pulsed currents (HVMPC) provided to the area of a pressure injury (PI) change the blood level of cytokines (interleukin [IL]-1β, IL-10, and tumor necrosis factor [TNF]-α) and growth factors (insulin-like growth factor [IGF]-1 and transforming growth factor [TGF]-β1) in patients with neurological injuries and whether the level of circulatory cytokines and growth factors correlates with PI healing progression. Methods: This study was part of a randomized clinical trial on the effects of HVMPC on PI healing. All patients with neurological injuries (spinal cord injury, ischemic stroke, and blunt trauma to the head) and a stage 2, stage 3, or stage 4 PI of at least 4 weeks’ duration hospitalized in one rehabilitation center were eligible to participate if older than 18 years of age and willing to consent to donating blood samples. Exclusion criteria included local contraindications to electrical stimulation (cancer, electronic implants, osteomyelitis, tunneling, necrotic wounds), PIs requiring surgical intervention, patients with poorly controlled diabetes mellitus (HbA1C > 7%), critical wound infection, and/or allergies to standard wound treatment. Participants were randomly assigned to 1 of 3 groups: anodal (AG) or cathodal (CG) HVMPC treatment (154 μs; 100 Hz; 360 µC/sec; 1.08 C/day) or a placebo (PG, sham) applied for 50 minutes a day, 5 days per week, for 8 weeks. TNF-α, IL-1β, IL-10, TGF-β1, and IGF-1 levels in blood serum were assessed using the immunoenzyme method (ELISA) and by chemiluminescence, respectively, at baseline and week 4. Wound surface area measurements were obtained at baseline and week 4 and analyzed using a digitizer connected to a personal computer. Statistical analyses were performed using the maximum-likelihood chi-squared test, the analysis of variance Kruskal-Wallis test, the Kruskal-Wallis post-hoc test, and Spearman’s rank order correlation; the level of significance was set at P ≤.05. Results: Among the 43 participants, 15 were randomized to AG (mean age 53.87 ± 13.30 years), 13 to CG  (mean age 51.08 ± 20.43 years), and 15 to PG treatment (mean age 51.20 ± 14.47 years). Most PIs were located in the sacral region (12, 74.42%) and were stage 3 (11, 67.44%). Wound surface area baseline size ranged from 1.00 cm² to 58.04 cm². At baseline, none of the variables were significantly different. After 4 weeks, the concentration of IL-10 decreased in all groups (AG: 9.8%, CG: 38.54%, PG: 27.42%), but the decrease was smaller in the AG than CG group (P = .0046). The ratio of pro-inflammatory IL-10 to anti-inflammatory TNF-α increased 27.29% in the AG and decreased 26.79% in the CG and 18.56% in the PG groups. Differences between AG and CG and AG and PG were significant (AG compared to CG, P = .0009; AG compared to PG, P = .0054). Other percentage changes in cytokine and growth factor concentration were not statistically significant between groups. In the AG, the decrease of TNF-α and IL-1β concentrations correlated positively with the decrease of PI size (P <.05). Conclusion: Anodal HVMPC elevates IL-10/TNF-α in blood serum. The decrease of TNF-α and IL-1β concentrations in blood serum correlates with a decrease of PI wound area. More research is needed to determine whether the changes induced by anodal HVMPC improve PI healing and to determine whether and how different electrical currents affect the activity of biological agents responsible for specific wound healing phases, both within wounds and in patients’ blood. In clinical practice, anodal HVMPC should be used to increase the ratio of anti-inflammatory IL-10 to pro-inflammatory TNF-α , which may promote healing.

The wound healing process is controlled by cytokines and growth factors that coordinate cellular processes. They can act by autocrine, paracrine, juxtacrine, or endocrine mechanisms and regulate cell migration, proliferation, differentiation, and metabolism. Clinical trial reviews^1,2 note the expression of inflammatory cytokines in chronic wounds is upregulated and prolonged and anti-inflammatory cytokines and growth factors are inhibited, resulting in delayed wound healing.

Interleukin (IL)-1β and tumor necrosis factor (TNF)-α are the primary pro-inflammatory cytokines that are essential in the early phases of wound healing. According to preclinical and clinical study reviews^1,2 and in vitro research by Goldberg et al,³ these cytokines induce neutrophil recruitment and maturation and increase vascular permeability.^1,2 The prolonged expression of IL-1β1,2 and TNF-α^1-3 is believed to increase tissue destruction due to overactivation of immune cells and the production of proteases, along with decreased collagen synthesis and maturation and reduced formation of granulation tissue.^1-3

According to Werner and Grose,² IL-10 plays a major role in limiting and terminating  inflammatory responses. In vivo mouse research⁴ has demonstrated that endogenous IL-10 inhibits infiltration of neutrophils and macrophages toward the site of injury as well as expression of pro-inflammatory agents, mainly TNF-α.  IL-10 also regulates growth and differentiation of keratinocytes and endothelial cells.²

New blood vessel formation, wound granulation, and reepithelialization are stimulated by growth factors such as transforming growth factor-β1 (TGF-β1), insuline-like growth factor (IGF)-1, platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF), and keratinocyte growth factor (KGF).^1,2,4-11 PDGF, FGF, and TGF-β1 also stimulate synthesis of extracellular matrix (particularly type 1 collagen).^1,2,8 TGF-β1 has been shown in in vivo studies with animals to inhibit the activity of matrix metalloproteinases (MMPs),⁴ increase keratocyte chemotaxis,⁸ and induce α-smooth muscle actin (α-SMA) expression in fibroblasts.⁹ Thus, TGF-β1 controls tissue proliferation, contraction, and remodeling.^4,8,9

Pressure injury (PI; previously termed pressure ulcer), one of the most frequent chronic wounds that occur in humans, develops as a result of prolonged pressure to skin and subcutaneous tissues that obstructs arterial blood flow, leading to cell death, tissue necrosis, and inflammation. In an in vivo study, Kurose et al¹² found elevated levels of pro-inflammatory factors and apoptosis-inducing factors, including IL-1β, IL-6, TNF-α, and MMP-3, in the skin of experimental animals exposed to prolonged pressure. Similar results of prolonged pressure also were observed by Jiang et al¹³ in their clinical study in humans; they found increased concentration of pro-inflammatory cytokines (IL-1β, TNF-α) and caspase-3 and decreased concentration of collagen and growth factor (FGF, VEGF) in stage 3 and stage 4 PIs compared with healthy skin.¹³

International clinical guidelines¹⁴ recommend the use of electrical stimulation (ES) in the treatment of stage 2 through stage 4 PIs. The recommendation (strength of evidence = A) is supported by direct scientific evidence from properly designed and implemented controlled trials on PIs in humans that consistently support the recommendation.¹⁴ However, more clinical trials should be conducted to determine how exogenous electrical currents affect the biological processes underlying healing wounds.

ES with exogenous current applied to the wound surface is believed to stimulate cellular processes and enhance wound healing. Electric currents can alter the expression of pro-inflammatory^15,16 and anti-inflammatory cytokines¹⁷ and growth factors^17,18 according to in vivo studies with animals¹⁸ and humans^16,17 and following the examination of tissue samples from wounds,^15,18 wound fluid,¹⁷ and circulating blood.¹⁶

Gürgen et al¹⁵ estimated the concentration of pro-inflammatory cytokines (IL-1β, TNF-α, IL-6) in vivo in skin wounds in 40 rats divided into 4 groups. In one of the groups, wounds received no treatment. Rats in the other 3 groups were treated for 5 days, receiving, respectively, ES with biphasic currents (2 Hz; 250 μC/sec; 0.89 C/day; 15 minutes), 0.9% sodium chloride solution, and 10% povidone iodine. A fifth group (control) consisted of 8 nonwounded rats. Immediately after incision, the concentration of pro-inflammatory cytokines in skin samples in all 4 wounded groups was greater than in controls (P <.05). After 5 days of treatment, the IL-1β, TNF-α, and IL-6 immunoreaction in the skin was decreased in the ES group, and wound epithelialization was more advanced compared to the other forms of treatment (P <.05 in both cases). The authors concluded decreases in pro-inflammatory cytokines observed in the dermis in the ES group suggested ES shortened the healing process by inhibiting the inflammatory phase.¹⁵

Kim et al¹⁸ evaluated in vivo the effect of high-voltage monophasic pulsed current (HVMPC) on wound healing and expression of wound-healing factors in diabetic rats. The animals were divided into 3 groups (10 animals per group): diabetic rats that were administered ES (ES group), diabetic control rats (sham ES), and nondiabetic control rats (sham ES). The treatment electrode (2 cm x 2 cm) was placed on the dorsal linear wound, and the return electrode was placed distally to the wound area. Rats in the ES group received HVMPC (monophasic, twin-peak pulses of 140 μs; 100 pps) daily for 40 minutes for 1 week. ES was performed such that palpable contractions were barely evident (35 V and 50 V). The polarity of the treatment electrode was negative for the first 3 days of intervention (inflammation phase) and positive for the following 4 days (proliferation phase). In all groups, the results of wound treatment were evaluated after 7 days. The authors observed wound closure was delayed in diabetic control rats compared to the nondiabetic control rats. In diabetic control rats, the expression of levels of collagen-I, α-SMA,  and TGF-β1 mRNAs were reduced. Importantly, compared to diabetic control rats, rats provided HVMPC showed accelerated wound closure and healing (P <.01) and restored expression levels of collagen-I (P = .002), α-SMA (P = .04), and TGF-β1 (P = .01) mRNAs. The authors concluded HVMPC may be beneficial for enhancing diabetic wound healing by restoring the expression levels of TGF-β1, collagen-I, and α -SMA.

In the clinical trial by Karavidas et al,¹⁶ 24 patients with chronic cardiac failure were randomly divided into 2 groups. In both groups, the patients’ quadriceps and triceps surae in both lower extremities were treated with biphasic current (25 pps) 30 minutes per day for 4 weeks. In the functional ES group (FES group), amperage was set to cause tetanic muscle contractions (5-second contraction followed by a 5-second pause), and the control group (8 patients) received current input up to the sensory threshold without muscle contractions. TNF-α, soluble intercellular adhesion molecule-1 (sICAM-1), and soluble vascular adhesion molecule-1 (sVCAM-1) were significantly reduced, and the ratio of IL-10 to TNF-α in blood serum increased, but only in the FES group (P = .007, P = .028, P = .019, and P = .098, respectively).

Chronic wounds in humans are treated using low-voltage monophasic pulsed current^19,20 and HVMPC,^21-31 as well as low-voltage biphasic current,^32,33 applied at a sensory level (>1 mA) below the threshold of skeletal muscle response. Direct and pulsed currents also are used with amperage below 1 mA (so-called microcurrents) and do not induce sensory responses.^34,35   

HVMPC was used in clinical studies to treat PIs,^21,22,27-31  venous leg ulcers,^23,25 and diabetic foot ulcers.²⁴  In all cited clinical studies, patients with PIs received standard wound care (SWC) for ethical reasons regardless of experimental or control group participation.^21,22,26-31 The results of treating experimental groups with HVMPC were compared with the results of control groups that received only SWC^26-28 or SWC combined with sham HVMPC.^21,22,29-31 The authors of all cited studies used double-peaked impulses.^21,22,26-31 In most studies, pulse duration ranged from 100 µs to 154 µs and frequency from 100 pps to 105 pps.^21,22,27-31 Electrical charge delivered to tissue was in the range of 250 μC/s to 500 μC/s.^21,22,29-31 HVMPC mostly was applied once a day for 45 to 60 minutes, 5 to 7 days per week (3.75 to 7 hours per week).^21,22,27-31  In all studies, the treatment electrode was placed on the wound surface and the return electrode on healthy skin at least 15 cm to 20 cm from the treatment electrode.^21,22,26-31

A minimum of 2 electrodes is required for delivery of current into tissues. The cathode (the negative electrode) attracts positive ions (cations), and the anode (positive electrode) attracts negative ions (anions) in the tissues. Preclinical and clinical studies show the anode and cathode may have different effects on wound healing processes. For example, in vitro^36,37 and in vivo³⁸ studies have shown anodal stimulation facilitates electrotaxis of macrophages³⁶ and neutrophils^37,38 for autolysis and reactivation of the inflammatory phase of healing. In vitro studies^39-42 have demonstrated that fibroblasts,^39,40 keratinocytes,⁴¹ and epithelial cells⁴² migrate toward the cathode, suggesting that cathodal ES may promote cellular proliferation. In vitro study⁴³ has shown both the anode and cathode can stimulate cellular processes that enhance growth of blood vessels; vascular endothelial cells migrate toward the cathode, whereas vascular fibroblasts and smooth muscle cells migrate toward the anode. In a clinical study by Asadi et al,⁴⁴ cathodal ES using direct current increased the concentration of VEGF in fluid obtained from diabetic foot ulcers. In their in vitro study, Daeschlein et al⁴⁵ showed low-voltage monophasic pulsed current has an antibacterial effect that is greater with anode than cathode stimulation and that bacterial reduction differed significantly between anode and cathode control, with the highest log 10 reduction factor achieved with the anode. In the current authors’ previous clinical study,³¹ both anodal and cathodal stimulation was observed to increase blood flow at the edges of the PIs. After 2 weeks of treatment with both anodal and cathodal HVMPC, periwound skin blood flow was significantly greater than in the control group (anodal ES group increased 109.52% from baseline versus 131.54% in the cathodal ES group and by 35.83% in the control group). The differences between both ES groups and the control group were significant (P = .047 and P = .0152, respectively).

In clinical studies, the polarity of the treatment electrode varied significantly from study to study. In some trials, PIs were treated using only anode^21,31 or only cathode stimulation.^22,29,31 In others, the polarity of the treatment electrode was reversed during treatment.^26-28,30 Some authors changed polarity weekly starting with cathodal stimulation26; others employed cathodal stimulation in the first 1 to 2 weeks^27,29,30 then switched to anodal stimulation for the next 4 to 5 weeks.^27,28,30  

In the authors’ previous clinical study, cathodal and anodal HVMPC were found to increase blood flow in the wound edges and to cause a similar surface area reduction in stage 2 through stage 4 PIs in patients with neurological impairment.³¹ The purpose of this study was to investigate how anodal and cathodal HVMPC influence the concentration of pro- and anti-inflammatory cytokines and growth factors in blood serum in humans. Clinical studies of similar scope have not been conducted to date.

Study design. The study was designed to compare the concentrations of cytokines (TNF-α, IL-1β, IL-10) and growth factors (IGF-1, TGF-β1) in blood serum and their correlations with changes in the surface area of stage 2 through stage 4 PIs after 4 weeks of treatment among 3 parallel groups of patients receiving SWC plus anodal HVMPC (AG), SWC plus cathodal HVMPC (CG), and SWC plus sham HVMPC (PG), respectively.
The results of this study are part of a prospective, randomized controlled clinical trial designed to gain insight into the effects of anodal and cathodal electrical currents on the PI healing process. The authors’ previous study³¹ examined the clinical effects of healing stage 2 through stage 4 PIs and changes in blood flow at the edge of PI after treatment with anodal, cathodal, and sham ES among 61 patients. In the current study, changes in blood levels of cytokines and growth factors were assessed in 43 patients, 40 of whom (65.57%) were part of the previous research³¹ and had agreed to donate blood samples. The study was conducted from December 1, 2015, to January 30, 2017; 3 additional patients were included in the more recent study from January 30, 2017, to February 27, 2017.

Study enrollment and criteria. Patients screened for the study were treated as inpatients at one rehabilitation center between December 1, 2015, and February 27, 2017. Their eligibility to participate was assessed by their physician using the following main criteria: neurological injuries (spinal cord injury [SCI], ischemic stroke, blunt trauma to the head), 18+ years old, at high risk of PI development (>15 points on the Waterlow scale),⁴⁶ and having a stage 2, stage 3, or stage 4 PI of at least 0.5 cm² in size and at least 4 weeks’ duration, located on the pelvic girdle or lower extremities.

An additional criterion for inclusion in the study was consent to donating blood samples. Patients with local ES contraindications (cancer, electronic implants, osteomyelitis, tunneling, necrotic wounds) were excluded from the study, as well as patients with PI requiring surgical intervention, poorly controlled diabetes mellitus (HbA1c >7%), critical wound infection, allergies to standard wound treatment, and/or no consent for donating blood samples. Patients were informed in writing by the research manager about the aim and course of the study and that they could withdraw from the study at any time without any consequences on their further treatment.

Randomization. Patients were randomly allocated to 3 groups using sealed envelopes containing special codes (A for the AG, B for the CG, and C for PG). The letters were inserted into envelopes that were sealed and then numbered randomly in computer-generated order. The envelopes were delivered to the main investigator, who opened them and directed the patient to the appropriate group according to the enclosed symbol.

Blinding. All patients, medical personnel, and researchers were blinded as to what type of ES was being applied to patients (anodal, cathodal, or sham ES). The exceptions were the main investigator and the principal physiotherapist who set the ES device to apply active or sham ES. The persons measuring the area of the wound surface, determining the levels of cytokines and growth factors in the blood, and performing statistical analysis of the results of the study also were blinded.

Study variables. Demographic information on the patients enrolled in the study was obtained from standardized participant interviews, physical examinations, the results of additional examination, and history of concomitant diseases. Study variables for assessing wounds, PI risk, and patients’ nutritional status were collected with paper-and-pencil instruments and transferred to a data sheet. Patient case history, results of blood laboratory tests, and previous diagnostic and treatment results were obtained from the electronic hospital database. All obtained information then was entered to a computer database to facilitate analysis.

Standard wound care administered to all groups. All patients in the AG, CG, and PG received SWC (scheduled repositioning, wound dressing changes, and physiotherapy treatment) under the supervision of the physician and the principle investigator and following best practices.^14,47,48 To protect the trial participants from developing more PIs, pressure-redistribution surfaces, foam devices, and pillows were provided. Patients who were immobile were repositioned by a nurse or physiotherapist at least every 2 hours, and persons who could move were instructed to change position as often as they could.

Patients who were diagnosed with malnutrition based on weight observation and blood tests received individual nutritional support.¹⁴ Patients’ wounds were examined regularly by a physician to determine the appropriate type of topical treatments. Tissue debridement was combined with infection and inflammation control, maintaining moisture balance in the wounds, and monitoring epithelial wound edges and epithelialization. Before ES was applied, necrotic tissue was removed from PIs with surgical/sharp, conservative sharp, or enzymatic debridement. Patients with elevated leukocyte levels received antibiotics selected according to the results of microbiological culture and sensitivity tests. All immobile patients received low-molecular-weight heparin. SWC, nutritional support, and general treatment of patients have been described in detail in the authors’ previous publication.³¹

Anode group. Patients in the AG were administered SWC and anodal HVMPC energy. The device used to deliver HVMPC was the Intelect Advanced Combo (DJO Global, Vista, CA); the device has 2 independent electrical circuits, but only one was active. The device generated a twin-peak monophasic pulse (154 µs) consisting of 2 77-µs exponential pulses in rapid succession. Pulse frequency was set at 100 pps and voltage above 100 V for amperage of 0.36 so as not to elicit motor reactions. The electrodes delivered a charge of 360 µC per second and 1.08 C per day. Patients participated in 50-minute sessions every week day (Monday through Friday). All patients had a personal set of conductive carbon-rubber electrodes. During the procedure, the treatment electrode (50 cm²) was placed on the wound and the return electrode (100 cm²) was applied to healthy periwound skin at least 20 cm from the PI. Both electrodes were separated from the tissue by aseptic gauze pads saturated with physiological saline. Anodal stimulation was applied to PIs once a day.

Cathode group. HVMPC protocol in the CG was identical, except cathodal instead of anodal stimulation was used.

Sham group. This group received SWC and sham HVMPC. The arrangement of electrodes during the procedure was the same as in the ES groups. The monitor of the ES unit displayed all parameters, but because the electrodes were connected to the inactive electrical circuit, current energy was not delivered to wounds.

General protocol. For all treatments, the lead physiotherapist connected the electrodes and selected the polarity of the treatment electrode. The procedure was performed in an inconspicuous manner so neither the patient nor the members of the medical team could see whether real or sham ES was applied.

In the active ES groups, amperage was set to 0.36 A (the same value was displayed on the monitor for patients receiving sham ES), which did not cause muscle contractions but only weak tactile sensations. Because most patients in the groups had tactile sensory problems and did not perceive the current, patients in the sham ES group did not know that they were not being treated. All treatment sessions had the same duration and frequency and followed the same protocol whether sham or active ES was applied.

The electrodes were sterilized before and after each session in approved disinfectant solution. As soon as the procedure ended, patients’ wounds were washed thoroughly with a 0.9% sodium chloride solution and covered with SWC dressings. In patients with more than 1 PI, all wounds were treated, but only the most severe were analyzed statistically.

Data collection.
Cytokine and growth factor concentrations. To determine cytokine and growth factor concentration in blood serum, peripheral blood samples (5 mL) were taken from the patients twice, immediately before treatment and at week 4. Blood serum was centrifuged away and stored at -80˚ C until the levels of cytokines and growth factors were determined. TNF-α, IL-1β, IL-10, and TGF-β1 levels were assessed using the immunoenzyme method (ELISA) using the following kits: Human TNF-α ELISA kit (Diaclone, Besancon Cedex, France), Human IL-1β ELISA kit (Diaclone), Human IL-10 ELISA kit (Diaclone), and TGF-β1 kit (Cloud-Clone Corp, Katy, TX). The concentration of IGF-1 was measured by chemiluminescence and a set of Human IGF-1 kit reagents (DiaSorin, Saluggia, Italy).

Wound measurements. Wound surface area (WSA, calculated in cm²) measurements were taken at baseline and after each week of therapy. The WSA was determined by tracing wound shapes onto acetate sheets and from the sheets onto rigid, transparent film for measurement with a planimeter. Measurements were processed by a digitizer (Mutoh Kurta XGT, Altek, Phoenix, AZ) connected to a personal computer with C-GEO software (version 4.0, Nadowski SoftLine, PL) that also was used for making computations and storing the results.

Primary outcomes. The primary outcomes of the study included:
1.    Percentage changes in blood serum concentration of pro-inflammatory cytokines (TNF-α and IL-1β) at week 4 of treatment compared with pretreatment levels;
2.    Percentage changes in blood serum concentration of anti-inflammatory cytokine IL-10 at week 4 of treatment compared with pretreatment levels;
3.    The percentage change in the concentration ratio of pro-inflammatory cytokine IL-10 to inflammatory cytokine TNF-α (IL-10/TNF) in blood serum at treatment week 4 compared with pretreatment values; and
4.    The percentage change in the concentration of growth factors IGF-1 and TGF-β1 in blood serum at treatment week 4 compared with pretreatment values.

The formulas for calculating the percentage changes listed above are presented in Table 1.

Secondary outcomes. Secondary outcomes included correlations between the concentrations of cytokines and growth factors in blood serum and the decrease in the PI size.

Statistical analysis.
Intention-to-treat analysis. To retain data of all randomly allocated participants, an intention-to-treat analysis was performed. Data that were not available were approximated using an exponential regression function written as WSA = b exp(-at), where WSA is wound surface area; b and a are the regression constant and the exponential regression coefficient calculated for each patient using WSA (cm²) obtained over the period of treatment, respectively; exp is the exponential regression function with a base of e≈2.718282 (the Euler’s number); and t is the week of treatment.⁴⁹ The function allows WSA decreases to be described and can be calculated with data from at least 3 weeks.⁴⁹ The exponential correlation coefficient proved negative for each patient and higher than 0.9 for the absolute WSA.

To minimize the risk of baseline interpatient differences biasing the results of the study, the relative values were used to determine the changes of cytokine and growth factor concentration in blood serum at week 4 and to calculate correlations between cytokine and growth factor concentration in blood serum and percentage reductions in wound surface area.

Patient characteristics were tested for normal distribution using the Shapiro-Wilk W-test, which showed that their distribution was not normal. Levene's test revealed heterogeneity of variance. Despite the absence of normal distribution, because of low absolute values of skewness and kurtosis (<2.5), a mean was used as the central value and a standard deviation as a measure of dispersion.

The baseline homogeneity of patient characteristics between groups was assessed using the maximum-likelihood chi-squared test, the analysis of variance (ANOVA) Kruskal-Wallis test, and the Kruskal-Wallis post-hoc test. To compare mean percent changes in blood levels of cytokines and growth factors between groups, the ANOVA/ Kruskal-Wallis test and Kruskal-Wallis post-hoc test were employed.

The correlations of cytokine and growth factor with percentage wound area reduction (PAR) were tested with Spearman’s rank order test. The following correlations were calculated:
1.    Between percentage changes of cytokines (TNF-α, IL-1β and IL-10) at week 4 and the PAR noted at week 4 (%PAR 4) and at week 8 (% PAR 8) (see Table 2); and
2.    Between percentage changes of growth factors (IGF-1 and TGF-1β) at week 4 and the PAR noted at week 4 and at week 8.  

In all tests, the level of significance was P <.05. All statistical procedures were blinded and utilized Statistica software, version 13.1 (StatSoft Polska Sp. z o.o.).

Ethical approval. Ethical approval was granted by the Academy Bioethics Commission.
The trial was registered with the Australian-New Zealand Clinical Trials Registry: ACTRN12616001709437; and funded by the J. Kukuczka Academy of Physical Education (Katowice, Poland).

Of the 73 persons screened for the study from December 1, 2015, to January 30, 2017, 12 persons failed to meet the inclusion criteria, leaving 61 persons in the study. The results of the treatment of these 61 patients were presented in the authors’ previous study,³¹ which focused on changes in PI surface areas and periwound blood flow after anodal, cathodal, and placebo ES. Of those, 40 patients (65.57%) also agreed to donate blood and were included in the current study (14 patients were randomized to AG, 12 to CG, and 14 to PG).

From January 30, 2017, to February 27, 2017, an additional 10 persons volunteered for the study, 5 of whom qualified and were randomized to the groups: 1 person to AG, 3 to CG, and 1 to PG. The purpose of additional randomization was to obtain 15 people in each group.

Between December 1, 2015, and February 27, 2017, 45 persons were randomized into the study groups (15 patients to each group). Two (2) patients randomized to CG were not treated with ES (1 person because of sudden death and 1 person because of sudden deterioration of health). The remaining 43 people were treated for 4 weeks. Between 4 and 8 weeks of therapy, 12 people dropped out (4 from AG, 3 from CG, and 5 from PG). The flow of participants through the trial is illustrated in the Figure.
 

Baseline patient and wound characteristics.
Sample characteristics. Of the 43 patients enrolled in the trial (age range 26–78 years), 11 were women (25.58%) and 32 were men (74.42%). All patients were found to be at risk of PI development (Waterlow scale score >15). In addition, 28 patients (65.12%) were malnourished and were administered nutrition therapy, 33 (76.74%) had diabetes (HbA1C <7%), 21 (48.84%) had SCI, 20 (46.51%) had experienced a cerebral stroke, and 3 (6.98%) had a head injury. Tetra- or quadriplegia was diagnosed in 6 patients (13.95%); 19 (44.19%) patients had paraplegia and 13 (30.23%) had hemiparesis. The patients had a total of 43 PIs that ranged in size from 1.00 cm² to 58.04 cm². Most PIs were stage 3 (29; 67.44%), 5 PIs were stage 2 (11.63%), and 9 stage were 4 (20.93%). Most PIs were located in the sacral region (32; 74.42%), 4 (9.30%) were on the ischial tuberosity or the trochanter, and 7 (16.28%) were on the lower extremities (lower leg and foot). The duration of the PIs ranged from 4 to 47 weeks.
 

Anode group characteristics. The AG included 15 patients (4 women, 11 men), average age 58.87 ± 13.30 (range 23–75) years. The mean risk of PI development (Waterlow scale score) was 31.3 ± 6.96 (range 18–46) points; 10 patients (66.67%) were malnourished and were administered nutrition therapy, 12 (80%) had diabetes (HbA1C <7%), 6 (40%) had SCI, 8 (53.33%) had experienced a cerebral stroke, and 1 (6.67%) had a head injury. One (1) (6.67%) patient had quadriplegia, 6 (13.95%) had paraplegia, and 5 (33.33%) had hemiparesis. The patients had a total of 15 PIs, with mean size of 17.23 ± 18.02 cm² (range 1.69–55.95 cm²). One PI was stage 2 (6.67%), 11 were stage 3 (73.33%), and 3 were stage 4 (20%). Most PIs were located in the sacral region (12; 80%), 1 (6.67%) was on the ischial tuberosity, and 2 (13.33%) were on the lower extremities (lower leg and foot). The mean duration of the PIs was 14.53 ± 12.91 (range 6–48) weeks. Four (4 , 26.67%) PIs were covered with slough, 7 (46.67%) started to granulate, and reepithelialization occurred in 4 (26.67%). Between ES treatments, SWC consisting of hydrogel dressings was applied in 9 patients (60%), hydrocolloid dressings in 3 (20%), and alginate in 3 (20%) patients. The mean blood concentration of TNF-α was 11.36 pg/mL, IL-1β was 10.46 pg/mlL, IL-10 was 5.57 pg/mL, IGF-1 was 99.07, and TGF-1β was 1.85 pg/mL. The mean ratio of IL-10/TNF-α was 0.62.
 

Cathode group characteristics. The CG included 13 patients (4 women, 9 men), average age 51.08 ± 20.43 (range 22–78) years. The mean risk of PI development (Waterlow scale score) was 29.1 ± 6.06 (range 23–43) points, 8 patients (61.54%) were malnourished and were administered nutrition therapy, 12 (92.31%) had diabetes (HbA1C <7%), 6 (46.15%) had SCI, 7 (53.85%) had experienced a cerebral stroke, and 1 (7.69%) suffered a head injury. Tetra- or quadriplegia was diagnosed in 3 patients (23.08%), 5 (38.46%) had paraplegia, and 3 (23.08%) had hemiparesis. The patients had a total of 13 PIs, mean size 26.13 ± 16.72 cm² (range 2.62–54.93 cm²). One (1) PI was stage 2 (7.69%), 8 were stage 3 (61.54%), and 4 were stage 4 (30.77%). Most PIs were located in the sacral region (10; 76.92%), 1 (7.69%) was on the trochanter, and 2 (15.38%) were on the lower extremities (lower leg and foot). The mean duration of the PIs was 13.69 ± 10.35 (range 6–40) weeks. Five wounds (5; 38.46%) were covered with slough, 5 (38.46%) started to granulate, and reepithelialization occurred in 3 (23.08%). Between ES treatments, SWC involving hydrogel dressings was applied in 5 patients (38.46%), hydrocolloid dressings in 5 (38.46%), and alginate in 3 (23.08%). The mean blood concentration of TNF-α was 8.05 pg/mL, IL-1β was 9.43 pg/mL, IL-10 was 5.65 pg/mL, TGF-1β was 1.61 pg/mL, and IGF-1 was 128.12. The mean ratio of IL-10/TNF-α was 0.7.
 

Sham group characteristics. The PG included 15 patients (3 women and 12 men), average age 51.2 ± 14.47 (range 27–68) years. The mean risk of PI development on the Waterlow scale was 32.55 ± 8.72 (range 19–46) points; 10 patients (66.67%) were malnourished and were administered nutrition therapy, 13 (86.67%) had diabetes (HbA1C<7%), 9 (60.00%) had SCI, 5 (33.33%) had experienced a cerebral stroke, and 1 (6.67%) had a head injury. Quadriplegia was diagnosed in 2 patients (13.33%), 8 (53.33%) had paraplegia, and 5 (33.33%) had hemiparesis. The patients had a total of 15 PIs, mean size 21.28 ± 17.89 cm² (range 3.59 – 59.57 cm²). Three (3) PIs were stage 2 (20%), 10 were stage 3 (66.67%), and 2 were stage 4 (13.33%). Most PIs were located in the sacral region (10; 66.67%), 2 (13.33%) were on the ischial tuberosity and the trochanter, and 3 (20%) were on the lower extremities (lower leg and foot). The mean duration of the PIs was 11.93 ± 9.62 (range 4–36) weeks. Four (4) PIs (26.67%) were covered with slough, 9 (60%) started to granulate, and in 2 PIs (13.33%) reepithelialization occurred. Between ES treatments, SWC with hydrogel dressings was provided in 8 patients (53.3%), hydrocolloid dressings in 2 (13.3%), and alginate in 5 (33.33%). The mean blood concentration of TNF-α was 9.58 pg/mL, IL-1β was 10.96 pg/mL, IL-10 was 6.64 pg/mL, TGF-1β was 1.56 pg/mL, and IGF-1 was 126.07. The mean ratio of IL-10/TNF-α was 0.75.

The groups were not statistically different for any of the baseline demographic and wound characteristics considered (P >.05). Main baseline patient characteristics are presented in Table 3.
 

Primary outcomes.
Changes in blood serum concentration of pro- and anti-inflammatory cytokines. At week 4, pro-inflammatory TNF-α concentration decreased from 11.36 pg/mL at baseline to 7.71 pg/mL at week 4 (24.89% ± 16.52%), from 7.48 to 6.12 pg/mL (22.66% ± 12.03%), and from 9.58 to 7.56 pg/mL (17.35% ± 51.24%) in the AG, CG, and PG, respectively. The between-group differences were not statistically significant.

Pro-inflammatory IL-1β concentration also was found to be lower in all groups but not significantly — from 10.46 to 4.77 pg/mL (22.41% ± 31.05%) in AG, from 9.43 to 4.80 pg/mL (17.74% ± 30.89%) in CG, and from 10.96 to 6.15 pg/mL (12.31% ± 30.05%) in PG (see Table 4). Anti-inflammatory IL-10 concentration at week 4 decreased in all 3 groups, from 5.57 to 4.85 pg/mL (9.8% ± 10.78%), from 5.65 to 3.29 pg/mL (38.54% ± 22.88%), and from 6.64 to 4.20 pg/mL (27.42% ± 23.47%) in the AG, CG, and PG, respectively. However, the decrease was significantly smaller in the AG than in the CG. Neither AG and PG nor CG and PG concentrations were found to be statistically different (see Table 4).   

The ratio of pro-inflammatory IL-10 to anti-inflammatory TNF-α. The ratio of IL-10/TNF-α increased from 0.62 to 0.72 (27.29% ± 28.76%) in the AG at week 4 versus baseline; in the CG, it changed from 0.71 to 0.54 ( 26.79% ± 21.80%) and in the PG from 0.75 to 0.59 (18.56% ± 21.34%). Differences between AG and CG and between AG and PG were significant ( P = .00009 and  P = .0054, respectively), unlike differences between CG and PG (P = .9180) (see Table 4).

Changes in blood serum concentration of growth factors. Between baseline and week 4 of treatment TGF-1β concentration increased from 1.85 to 1.91 pg/mL (7.27% ± 32.21%) in the AG and decreased from 1.61 to 1.56 pg/mL (2.52% ± 27.28%) and from 1.56 to 1.52 pg/mL (1.24% ± 20.75%) in the CG and PG, respectively. Intergroup differences in percent changes in TG-1 after treatment versus pretreatment were not statistically significant (P >.05) (see Table 4).

IGF-1 increased in all groups: from 99.07 to 122.78 pg/mL (9.80% ± 40.3) in the AG, from 128.12 to 133.69 pg/mL (2.06% ± 27.36) in the CG, and from 126.07 to 127.21 pg/mL (3.01% ± 21.50) in the PG. Intergroup differences in percent changes in IGF-1 after treatment versus pretreatment were not statistically significant (P >.05) (see Table 4).

Secondary outcomes.
Cytokine concentration at week 4 and wound area reduction. In all groups, the surface area of PIs after 4 and 8 weeks of treatment was significantly smaller than before treatment. In the AG, it decreased from 17.23 cm² ± 18.02 cm² at baseline to 11.30 cm² ± 15.71 cm² at week 4 and to 8.55 cm² ± 13.09 cm² at week 8 (P = .0006 and P = .0006, respectively; Wilcoxon test); in the CG from 26.13 cm² ± 16.72 cm² to 14.88 cm² ± 14.30 cm² at week 4, and to 8.52 cm² ± 10.61 cm² at week 8 (P = .0015 and P = .0015, respectively; Wilcoxon test); in the PG, it decreased from 21.28 cm² ± 17.90 cm² to 16.04 cm² ± 15.62 cm² at week 4 and to 12.75 cm² ± 13.15 cm² at week 8 (P = .0009 and P = .0008, respectively; Wilcoxon test). After 4 weeks of treatment, the surface area of PIs in the AG decreased by 48.75% ± 30.95%, in the CG by 51.33% ± 21.55%, and in the PG by 29.57% ± 25.57%. The differences between the groups were not statistically significant (P = .0559; ANOVA Kruskal-Wallis test). After 8 weeks of treatment, the surface area of PIs in the AG decreased by 67.98% ± 26.53%, in the CG by 73.78% ± 21.78%, and in the PG by 46.10% ± 30.02%. The differences between AG and PG and CG and PG were statistically significant (P = .0441 and P = .0393, respectively; ANOVA Kruskal-Wallis test).  

In the AG, the percentage change in the concentration of pro-inflammatory cytokine TNF-α between weeks 0 and 4 negatively correlated with the PAR in the same period (R = -0.667, P = .066) and with the PAR between weeks 0 and 8 (R= -0.735, P = .0027). Thus, a greater decrease in TNF-α concentration between weeks 0 and 4 was associated with a larger PAR in the same period and with PAR between weeks 0 and 8 (see Table 5).

A negative correlation in the AG also was found between the percentage change in IL-1β concentration and PAR between weeks 0 and 4 (R = -470, P = .047). Thus, the greater reduction in IL-1β concentration from week 0 to week 4 was associated with a greater PAR in that time frame (see Table 5).

Other correlations between percentage change in cytokine concentration between weeks 0 and 4 and PAR were not statistically significant in the AG (P >.05). Similarly,  correlations between changes in TNF-α and IL-1β concentrations and PAR were not significant in the CG and PG (see Table 5).

Growth factor concentration at week 4 and wound area reduction. No significant correlations were found in any of the 3 groups regarding percentage change in growth factor concentration in weeks 0 to 4 and PAR in weeks 0 to 4 and 0 to 8 (see Table 5). 

Principle findings. Four (4) weeks of HVMPC treatment significantly reduced the concentration of pro-inflammatory cytokines (TNF-α and IL-1β) in blood serum in all 3 groups, with no significant differences among groups (P >.05). The level of anti-inflammatory cytokine IL-10 also decreased significantly in all groups but less in the AG than in the CG group (P = .0046). An increase was noted in the AG in the ratio of anti-inflammatory IL-10 to pro-inflammatory TNF-α at week 4, an increase that was significantly greater than in both the CG and the PG (P = .00009 and P = .0054, respectively). ES did not affect changes in the concentration of circulating growth factors (TGF-β1 and IGF-1).    

Other research has noted that SCI patients have PIs associated with high levels of inflammatory agents in the blood compared to healthy persons. In a cross-sectional study, Segal et al⁵⁰ noted that plasma concentrations of bioactive molecules IL-6, IL-2R (the soluble interleukin-2 receptor) and ICAM-I were numerically or significantly elevated in 70 patients (19 with PIs) with SCI as compared to 20 healthy, able-bodied  individuals. The greatest increase in concentration was seen in patients with PIs who demonstrated slow wound healing. Similar results were obtained in a cross-sectional study by Davis et al⁵¹ that involved 56 people with SCI and 35 healthy persons in control group. In their study, serum levels of Il-6, TNF-α, and IL-1RA (the IL-1 receptor antagonist) were higher in patients with SCI (P <.005) than in the control group. Elevated cytokine levels were not associated with high white blood cell count, severity, and level of SCI, but they were obvious in patients with SCI who were asymptomatic due to medical complications and were even more elevated in patients with neuropathic pain, urinary tract infection, or PIs. The authors concluded these results may indicate protective autoimmunity, which is a consequence of an existing infection, and that high levels of circulating adhesion molecules and pro-inflammatory cytokines in patients with SCI may impede PI healing.

The results of the current study show anodal HVMPC used to treat PIs in people with SCI can increase the ratio of anti-inflammatory IL-10 to pro-inflammatory TNF-α, which, in the authors’ opinion, can contribute to reducing inflammation in the wound area.

It is noteworthy that in the group receiving anodal HVMPC, the decrease of TNF-α and IL-1β concentrations during the first 4 weeks of treatment positively and  significantly correlated with the percentage of wound area reduction; statistically significant correlations between the amount of cytokines and growth factors were not noted in the CG and the PG. The results of this study do not allow one to conclude that decreased concentrations of TNF-α and IL-1β, induced by anodal HVMPC, caused enhanced PI healing. Nevertheless, these results are interesting, and it is worth continuing the research in a larger group of patients to answer the question of whether anodal HVMPC can accelerate the healing of PIs by affecting blood levels of cytokines and growth factors.

This study did not find that HVMPC applied in the PI area influenced the concentration of pro-inflammatory cytokines (IL-1β, TNF-α) or growth factors (TGF-1β, IGF-1) in blood serum. However, the ratio of anti-inflammatory interleukin IL-10 to pro-inflammatory interleukin TNF-α in the AG increased and reduced concentrations of IL-1β and TNF-α at week 4 correlated with progress in wound area closure.

Human clinical studies investigating the effect of ES on cytokines and growth factors are few. The current researchers failed to find reports evaluating changes induced by ES in blood levels of cytokines and growth factors in patients with PIs. Only 3 human studies that evaluated the concentration of prohealing factors in the wound area were found^7,44,52; 2 studies involved healthy persons with acute experimental wounds^17,52 and 1 was a clinical study on patients with diabetic foot.⁴⁴

In a study with healthy humans by Sebastian et al,¹⁷ a punch biopsy was taken from the upper arm of 20 volunteers on day 0, the wounds were treated with ES, and the biopsy was repeated on day 14 (experimental wounds). The contralateral upper arm served as a control; a biopsy was taken on day 0 and repeated on day 14. ES utilizing biphasic current was applied 4 times, every second day, for 30 minutes. At day 14, a near-completed inflammatory stage of healing in ES wounds was demonstrated by upregulation of anti-inflammatory agents  (IL-10 and macrophage migration inhibitory factor; P <.05 compared to control wounds). ES wounds demonstrated an increased expression of angiogenic gene factor (connective tissue growth factor, TGF-β1, and MMP-2; P <.05). In addition, factors responsible for wound proliferation and remodeling (type IV collagen, fibronectin, and interferon-γ; P <.05) also were noted. The authors concluded cutaneous wounds receiving ES display accelerated healing in reduced inflammation, enhanced angiogenesis, and an advanced remodeling phase.

In a cohort study conducted among 40 healthy volunteers by Ud-din et al,⁵² skin punch biopsies were performed on both upper inner arms of the participants. The wounds on one arm acted as the controls and were not treated. The ES protocol was the same as in the study by Sebastian et al.¹⁷ On days 10 and 14, the surface area of ES wounds was significantly smaller than the control wounds (P = .001 and P <.001, respectively). Expression of angiogenetic factors increased in ES-treated wounds, including VEGF-A (P <.05) and PDGF (P <.05) compared to the control wounds. The authors concluded ES enhanced wound healing by reducing wound dimensions and increased VEGF-A and PLGF expression in acute cutaneous wounds.

Asadi et al⁴⁴ conducted a clinical randomized, single-blind, placebo-controlled trial on 30 patients with type 2 diabetes and ischemic foot ulceration. The authors randomly assigned 15 patients to receive either low-intensity cathodal direct current (0.56 mA/cm²) at the sensory threshold or placebo ES (15 patients) for 1 hour/day, 3 days/week, for 4 weeks. After debridement during the first and twelfth active or placebo ES, wound fluid was collected to determine the levels of hypoxia inducible factor (HIF)-1α, nitrous oxide, VEGF, and sVEGFR-2. After the first session, wound fluid levels of HIF-1α were significantly increased in the ES group compared with the control group (P = .01). After the first and twelfth sessions, wound fluid levels of VEGF were also significantly increased in the ES group compared to the control group (P = .007 and P = .019, respectively). The authors concluded that applying low-intensity cathodal direct current to ischemic diabetic foot ulcers can be a promising way to promote angiogenesis and wound healing.          

The cited preclinical studies in animals^15,18 and healthy people^17,52 confirm different electrical currents (low-voltage biphasic pulsed  currents,15 degenerative biphasic currents,^17,52 cathodal HVMPC¹⁸) have an effect on cytokines, growth factors, and other agents modifying wound healing.
Few clinical studies have explored the effects of ES on biological agents involved in wound healing, but the study by Asadi et al⁴⁴ provides evidence that cathodal direct current can promote the secretion of angogenetic factors in the diabetic foot (HIF-1 and VEGF). The current study shows anodal HVMPC can increase the ratio of pro-inflammatory IL-10 to anti-inflammatory TNF-α in the blood of patients after SCI, which can inhibit inflammation and accelerate PI healing

One of the most important strengths of the current study was the blinding of the research team (physicians, nurses, physiotherapists), the person in charge of measuring WSA, and the statistician. Moreover, participants were hospitalized at the same rehabilitation center, making it possible for the medical staff to supervise the uniform application of PI prevention measures and treatments and to ensure the ES protocol was observed at all times. All patients completed at least 4 weeks of treatment, so the concentrations of cytokines and growth factors were calculated and compared for all of them. In the intention-to-treat analysis, the exponential regression function, which allows WSA decreases to be precisely represented, was used to approximate the likely treatment effects between weeks 4 and 8.

The major limitation of the study was the relatively small number of participants. The blinding rate of patients and assessors was not assessed. Although the general PI prevention and treatment program was developed for all 3 groups using the same best practice recommendations,¹⁴ it also contained some specific solutions addressing the needs of individual patients. No adverse effects of applying HVMPC were observed in this study, and they have not been reported by other researchers.

Further research should be conducted to determine whether and how different electrical currents affect the activity of biological agents responsible for specific wound healing phases, both within wounds and in a patients’ blood. It is also important to examine whether electric currents with different waveforms affecting the concentration of biological agents in wounds and in blood promote the healing of chronic wounds of various etiologies.

A randomized clinical study involving patients with SCI and stage 2 through stage 4 PIs has shown HVMPC (154 µs; 100 pps; 100 V; 250 µC/sec) applied 50 minutes a day, 5 times a week, with the anode as a treatment electrode, favorably modifies the serum concentration of anti-inflammatory IL-10 cytokine to pro-inflammatory TNF-α cytokine. The study also indicates that in patients whose wounds have been subjected to anodal HVMPC, a decrease in proinflammatory cytokines (IL-1β and TNF-α) correlates with greater progression of PI healing. However, more clinical trials are necessary to elucidate the nature of the relationship between anodal stimulation of cytokine concentrations and changes in wound area. Based on the results of this and the authors’ previous research,³¹ in clinical practice a goal of provisional care should be to increase the ratio of anti-inflammatory IL-10 to pro-inflammatory TNF-α and to stimulate PI healing, anodic HVMPC (154µs; 100 pps; 250-360 µC/s) should be applied for 50 minutes a day, 5 times a week for 4 weeks. Further research to support these protocols is warranted.

The authors are grateful to the physicians, physical therapists, and nurses who assisted in performing this study, particularly Prof. Andrzej Franek, MSc, PhD; Tomasz Ickowicz, PT; Ewa Hordynska, MD; Jaroslaw Szczygiel, MD; and Krystian Oleszczyk, MD, PhD.

Dr. Polak is an Assistant Professor, Department of Physical Therapy, Academy of Physical Education, Katowice, Poland; and a physical therapist, Rehabilitation Center “Technomex,” Gliwice, Poland. Dr. Kloth is an Emeritus Professor of Physical Therapy, Marquette University, Milwaukee, WI. Dr. Paczula is a surgeon and a Medical Rehabilitation Specialist, Rehabilitation Center “Repty,” Tarnowskie Gory, Poland; and a lecturer, Department of Physical Therapy, Academy of Physical Education. Dr. Nawrat-Szoltysik is an Adjunct, Department of Physical Therapy, Academy of Physical Education; and a physical therapist, St. Elzbieta Caritas Skilled Nursing Facility, Ruda Slaska, Poland. Dr. E. Kucio is a Clinical Pharmacist, Multispecialty Hospital, Siemianowice Slaskie, Poland; and an assistant, Department of Physical Therapy, Academy of Physical Education. Dr. Manasar is a Specialist in Medical Analytics, Silesian Medical Laboratories, Katowice. Dr. Blaszczak is an Associate Professor, Department of Medical Biophysics, Medical University of Silesia, Katowice. Dr. Janikowska is an Adjunct, Department of Analytical Chemistry, Medical University of Silesia. Dr. Mazurek is a Professor of Medical Sciences, Medical Biology; and the Head of Department of Molecular Biology, Medical University of Silesia. Dr. Malecki is a Professor of Medical Sciences, Neurology and Pharmacology Specialist, Department of Physical Therapy, Academy of Physical Education. Dr. C. Kucio is an Associate Professor of Medical Sciences, Department of Physical Therapy, Academy of Physical Education; and Internal Medicine Specialist and Head of Internal Medicine Department, Multispecialty Hospital, Jaworzno, Poland. Please address correspondence to: Prof. Anna Polak, PT, PhD, Department of Physical Therapy, Academy of Physical Education, Katowice, Poland; email: a.polak@awf.katowice.pl.

The study was funded by the Academy of Physical Education, Katowice, Poland.