1 Introduction ONE OF the few regenerative capabilities of the human body is the healing of lesions inflicted to soft tissue. The process of repair and regeneration following injury represents one of the most fundamental defence mechanisms of an organism against its environment. Wound healing is often viewed as a sequence of events which automatically take course after the injury. However, closer acquaintance with the problems which may arise during the healing quickly dispels the concept of the inevitability of wound healing and brings the process into proper perspective as one of the keystones of modern surgery (CuoNo, 1985). Indeed, wounds which appear on the body without surgical intervention as well as surgical wounds may not heal. Such 'chronic' wounds are difficult and frustrating to manage. They can last for weeks, months or years and represent a major medical, social, and economic problem for the patients, their relatives and the medical community. Correspondence should be addressed to Renata Karba. First received 2 n d July and in final form 4th November 1991
9 IFMBE: 1992
Medical & Biological Engineering & Computing
The major reason for the appearance of chronic wounds is ischaemia combined with damage to the nervous system and any resultant metabolic abnormalities. There are three major groups of patients who are susceptible to chronic wounds: (a) Spinal cord injured patients who develop pressure sores (decubitus ulcers) due to their inability to move and lack of sensibility in areas subjected to prolonged pressure on the skin, and deeper tissues from an external object while sitting or lying. Typical pressure sore locations are bone tuberocities (trochanter, sacrum, heels). (b) Patients with peripheral vascular disease in whom ischaemic ulcers occur in the lower extremities due to various pathological conditions, the most common being diabetes mellitus. In a large proportion of such patients their condition impairs to such an extent that amputation is required with the usual complication that the wound on the stump resists healing. (c) Geriatric patients in whom ulcers are due to combinations of immobility and reduced blood circulation.
May 1992
257
With these facts in mind it is obvious that any additional treatment modality which might offer hope for accelerated healing of chronic wounds should be amenable to careful investigation. In the following paragraphs we present a survey of several methods which seem to improve wound healing by various applications of electrical and electromagnetic fields. It is our goal to present medical engineers as well as clinicians with adequate evidence about the effects of electricity on wound healing. We hope that additional basic and clinical investigations will be encouraged and that electrical wound healing will soon be generally accepted as a safe, simple and efficient modality for treating chronic wounds. 2 S t i m u l a t i o n modalities A brief survey of the existing literature exposes a fascinating richness in the application of diverse electrical or electromagnetic fields to different parts of the body to heal chronic wounds. Practically all reports are rather exclusive regarding their specific method, and rarely even attempt to compare their treatment techniques with methods published by authors who used a different method. Recently some review papers were published which discuss electrical wound healing (DAYTON and PALLADINO, 1989; KLOTH and FEEDAR, 1990; WEISS et al., 1990). However, none of them addressed all current modalities and electrode locations, nor did they discuss possible mechanisms. To present a fairly complete picture regarding the application of electricity in human wound healing is the goal of the present review. In the second part of this paper (STEFANOVSKAet al., 1993) some of our results as well as problems with the evaluation of the healing process will be discussed. Obviously it is impossible to make a quantitative analysis of the comparative advantages and disadvantages of the different methods because the protocols, quantification and wound aetiologies in the published literature are vastly different. Therefore the aim of the following survey is not to 'grade' and compare the different techniques but only to make the clinician and engineer aware of the fact that there seem to be many truths in electrical wound healing and only time and more research will show if there is something like a single or optimal truth. 2.1 Direct current
It is quite difficult to trace back the history of electrical wound healing. One of its roots was investigations into natural electric fields. DuBoIS-REYMOND (1860) was the first who as early as 1860 described the currents of an injury leaving human skin wounds. The same phenomenon was observed by HERLITZKA(1910). He measured a current of about 1 #A in a skin injury on the hand. Experimental abrasions were made on human skin in research by CUNLIFFE-BARNES (1945) with the purpose of studying the potential difference between intact and injured skin during healing. The abrasions were found to be positive compared with the surrounding intact skin. The potential difference lessened during the healing of the wounds and disappeared when they closed over. The existence of injury currents was confirmed by measuring currents leaving the fingertips in children who had undergone accidental amputation (ILLINGWORTHand BARKER,1980). On average eight days after the wounding a peak average current of 22/~A cm- 2 was obtained. Thereafter it lessened and reached zero value when the wounds healed completely. These reports raised three questions: where is the source of injury currents, what is their physiological background, 258
and what is their role in the healing process ? Only the first of them was elucidated. In 1982 BARKER et al. presented a map of human 'skin battery' voltages, showing the arrangement of potentials on the surface of the intact skin on different parts of the human body with respect to the reference point in the subdermal layer. The surface of the skin was always negative compared with the deeper skin layers. The transcutaneous voltage was up to 40 mV. JAFFE and VANABLE(1984) finally located the skin battery in the epidermis or 'living layer'. Human skin thus possesses endogenous electrical properties that may influence wound healing. But can this natural regenerative process be enhanced by applying external electricity? This question gave rise to numerous studies which this review will deal with. The second historical root of electrical wound healing stems from the seventeenth century when gold leaf was applied to smallpox lesions to prevent scarring (ROBERTSON,1925). Following this route KANOF (1964) and WOLF et al. (1966) applied gold leaf directly to skin ulcers. Encouraged by the results and suspecting that the electrochemical influence of the gold may have been responsible for enhancing tissue healing, they conducted the first studies with low intensity direct electrical current. The rationale for electrical wound healing with direct currents thus originates from two sources: investigations of biopotentials and experiments with gold leaf. Direct currents were consequently applied for wound healing enhancement by several research groups. Initial studies on animal experimental wound models (CAREY and LEPLEY, 1962; ASSIMACOPOULOS, 1968a) demonstrated faster healing, greater wound tensile strength and intensified inflammatory reaction due to direct electrical current. The first report about human application (AsSIMACOPOULOS,1968b) deals with only three patients with chronic leg ulcers due to venous insufficiency which resisted conventional treatment for several years. Following the application of up to 0.1 mA of negativepolarity direct current to the ulcers, they healed completely in six weeks. One year later, the results of an extensive study by WOLCOTT et al. (1969) were published. They comprised 83 ischaemic skin ulcers. Electrical stimulation (ES) treatment consisted of three 2 h segments of 0.2-1 mA direct current applications each day. One copper mesh electrode was placed over the wound and the other proximal to the lesion. The optimum current amplitude was set empirically between visible bleeding (too much current) and copious serious exudate (too little current). Initially the polarity of the electrode in the wound was negative. The polarity was reversed after three days in noninfected wounds, while for infected ulcers the negative electrode was retained in the ulcer until the infection was cleared, plus three additional days. The electrode polarity was changed during the treatment each time the 'growth plateau' was observed. Stimulation was applied for 2 h followed by a 4 h pause. This 2:4 treatment ratio was repeated thrice daily, which totals 6 h of electrotherapy per day. In this study a control group was lacking, but nevertheless the eight participating patients had contralateral wounds of comparable size and aetiology, of which one wound received electrical stimulation and the other served as control. In all such cases significantly faster healing was obtained in stimulated wounds. In addition to accelerated healing, the strong antimicrobial effect of the negative-polarity current was observed in infected wounds. This study served as an example for many succeeding clinical trials. At the same time it is the most frequently cited work in the history of electrical wound healing,
Medical & Biological Engineering & Computing
May 1992
although the rationale for many elements of its study design remained unknown until the present day. The intensity of the stimulation current was adjusted for each wound with regard to its exudate, although it is dependent on the condition of the sore and on the phase of the healing process in which the wound is found, and therefore has no physiologically uniform meaning from which the dosage could be estimated. Electrical stimulation was applied for six hours daily, which represents nearly half of the active part of the patient's day. However, the dependence of the efficacy of ES on daily therapy duration has not been checked and it is possible that the same effect could be achieved with shorter applications, which would add considerably to the patient's comfort. Other questionable features are the alternation of electrode polarity and the daily cycle of current application. It is evident from the abovementioned studies of endogenous electricity (CuNLIEEEBARNES, 1945; ILLINGWORTH and BARKER, 1980) that the wound remains positive compared with the surrounding intact skin throughout the healing process. The potential difference during the healing merely diminishes, the polarity, however, remains the same. Thus if ES is intended to support the natural endogenous electrical events which accompany regenerative processes, there are no arguments for changing the polarity of the stimulation electrodes. To plan and accomplish a correct clinical study of the effect of a particular treatment modality on wound healing is quite difficult because a multitude of intrinsic (the patient's psychophysical condition, age) as well as extrinsic (nutrition, wound care) factors influence the healing process along with the therapy studied, and their effects are not easily separated. From this viewpoint contralateral lesions are acceptable as control wounds, as in this case all the above-mentioned factors are uniform for both groups of sores. Yet we cannot maintain that the effect of the electrical current is merely local. And since its possible systemic action has to be considered as well, intrapatient comparison of treatments does not seem to be the most objective method for evaluating the effects of ES. Moreover, the number of contralateral wounds in the trial discussed is very small, thus not permitting any statistically supported conclusions. Following the protocol of WOLCOTT et al. (1969), a group of 25 patients with venous, arterial and neurotrophic ulcers was treated by EDEL and FREUND (1975). They selected only difficult cases, which had resisted polypragmatic therapies for years. After electrical stimulation, the wounds started to heal in 20 out of 25 patients in treatment regimens of 3 x 1 h or 2 x 2 h of stimulation per day with a current density under the electrodes of about 0.1 mA cm -2, which was just below the sensation level. In 1976 GAULT and GATENS reported on the results of their extensive clinical study dealing with 76 patients having 106 ischaemic skin ulcers of different aetiologies, of whom six patients with contralateral ulcerations served as controls. Exactly the same protocol was used as in the study by WOLCOTT et al. (1969) except that the electrode polarity was reversed only once. The mean healing rate of the control ulcers was 14-7 per cent per week compared with 30 per cent per week of the treated counterparts. The mean healing rate of the 100 ulcers treated with lowintensity direct current was 28.4 per cent per week. Wound asepsis was observed within three to seven days with the use of the negative electrode over the ulcer. This is another study where the undoubted beneficial effects of weak constant direct current were presented, using comparison with a numerically poor control group Medical & Biological Engineering & Computing
of wounds. It is interesting to note that the efficacy of the ES therapy is obviously not reduced if after the first polarity exchange (negative to positive) of the electrode on the wound it remains positive throughout the subsequent treatment. This observation makes the necessity for and the sense of changing the polarity of the stimulation electrodes during the therapy questionable. F r o m among the studies with direct electrical current, due note should be taken of the work by CARLEY and WAINAPEL (1985), which is unique for including a numerically equal (N = 15) control group of randomly assigned patients, who were matched by age, diagnosis, and wound aetiology, location and size with the treated group. Electrical stimulation was applied to the wounds in the treated group twice daily for 2 h. The current level ranged between 0.3 and 0.7 mA, which revealed current densities of 0.03-0.11 mA cm -2, depending on the wound size. Again the electrode placed over the wound was negative for the first three days and positive thereafter. If, during the treatment, a plateau in healing was observed, the protocol of negative polarity at the wound site for a three-day period was restarted. The current amplitude was determined empirically after observing the stimulated tissue response (copious exudate or bleeding). Comparison of the treated wounds with the controls revealed 1.5-2.5 times faster healing of the stimulated wounds, which was statistically significant. Again the antiseptic qualities of the constant negative electric current were reported. Randomised clinical trials seem to be the most adequate way of comparing treatments. Random assignment of subjects to experimental groups abolishes the selection bias which often 'corrupts' the results of the research. In using this study design, attention should be paid to the homogeneity of the population studied with respect to all the remaining patient variables which could affect the research outcome. Nonsymmetrical distribution of patients with regard to age, sex, additional diseases and others can seriously jeopardise the objectivity of the results and blur the effect of the modality studied. A uniform distribution can be achieved with certainty by using a numerically sufficient sample of subjects. The survey of published research reveals that studies performed on a numerically small sample of subjects are extremely common. Such studies might thus be useless from the scientific point of view as they are unable to detect clinically important effects. On the other hand, they are also unethical in their consumption of financial and other resources and above all because of the patients involved (ALTMAN,1985). The nonappearance of the positive effect of the stimulation with a weak constant negative direct current on the healing of venous leg ulcers was reported by KATELARISet al. (1987). They compared the efficacy of four different treatments: standard local application of povidone-iodine and normal saline, both with and without electricity. The current of 0-02 mA was applied using an ulcer dressing which comprised titanium electrodes, the negative electrode overlying the ulcer. No significant differences were found, either between the povidone-iodine and normal salinic treatments, or between normal saline and normal saline with electric stimulation. The combination of povidone-iodine treatment with negative current, however, induced a significant decrease in the rate of healing. The authors attributed the retardation effect to possible intracellular toxic effects of negatively charge iodine ions forced into the tissue by the negative electrode. Although reports about the efficacy of a constantly applied current of negative polarity (AsSIMACOPOULOS, 1968a; b; KONIKOFF, 1976) as well as current of positive polarity (ALVAREZ et al., 1983) can be found in the liter-
M ay 1992
259
ature, one could conclude that the polarity of the electrode over the wound has to be reversed to obtain optimal effects of low intensity constant direct current. In the article by ROWLEV et al. (1974) the authors report on the effects of direct current on wound healing in rabbits. They claim that the application of the negative electrode to the ulcer suppressed both healing and local infection, while the positive electrode enhanced both. The authors thus suggested a treatment regime in which the negative electrode is applied to the ulcer until aseptic conditions are attained and the positive electrode thereafter to accelerate healing. Such a stimulation regimen likewise proved successful in the majority of the above-mentioned trials, but nevertheless it is hard to maintain that it is optimal because the essence of its action remains unknown. And, in spite of the evident beneficial effects of electrical wound treatment, the lack of knowledge regarding the mechanisms as well as the lack of carefully controlled studies seem to be major arguments used by the medical profession against routine clinical application of direct currents for healing chronic wounds. 2.2 Low-frequency pulsed currents As low-frequency pulsed currents (also known as tetanising currents) are quite popular in physical medicine, it is not surprising that they were tested for enhanced wound healing. However, it is interesting to note that these currents were not only applied locally to the wound or across it, but also to areas quite distant to the lesion. The two major distant locations were the spinal cord and acupuncture points. We shall thus divide this section into three subsections regarding the location of the stimulation electrodes. 2.2.1 Local application: The first report stems from THOMAS et al. (1982). They applied biphasic rectangular pulse trains of 4 0 H z and 0-1 ms pulse width with an amplitude producing slight muscle contraction and zero DC component, close to the wound. The duration of the trains was 5 s followed by a pause of 10 s (Fig. 1). One such session lasted for 30 rain and was repeated every 2 h during the day. With such a regimen several decubiti were healed. The authors report particularly about a quadriplegic patient with a sacral bedsore of 12cm diameter. After application of stimulation the sore healed in five months. Accelerated healing was attributed to an improved supply to the affected area and to decompression due to rhythmical contractions of stimulated muscles. Another successful attempt to heal chronic wounds with low-frequency pulses was reported by WESTERnOF and Bos (1983). They treated a patient with ulcers on the face due to trigeminal nerve transection with pulses of 120Hz and 250#s pulse width. The pulse train was 0.5 s on and 0.5s off for 30 min three times a day. The severe ulceration on the face did not respond to any wound treatment including
E
lttl//ttttltlllltlttt _iI ,/I/,, ,t, t
-5
,-
Fig. 1
260
time
5s
10s
5s
pulse duration :0-1ms frequency :40Hz
Trains of biphasic current stimuli which were applied locally to the wound in the study by THOMAS et al. (1982). (Redrawn from THOMAS et al., 1982)
antibiotic therapy for two years. After application of electrical stimulation the ulcers healed completely after six weeks. It is rather surprising that the above-mentioned case reports represent the only attempts of local application of this type of electrical stimulation for wound healing acceleration. The formation of chronic wounds is principally determined by insufficient tissue supply with oxygen and nutrients due to poor blood flow, irrespective of their diverse aetiologies (pressure sores, wounds of venous and arterial origin, neurotrophic ulcers). Tetanising currents slightly contract muscles in the sore surroundings, thus improving circulation and providing energy for the affected area. The ability of low-frequency pulsed currents to fight the basic cause of problematic healing has been far too little investigated and exploited. Unlike the method for direct currept applications, where one of the stimulation electrodes covers the wound, both electrodes are placed on the healthy skin surrounding the sore when tetanising currents are used. This means a minor risk for possible mechanical damage of newly formed tissue (too strong contractions should be avoided), but also much simpler 'noninvasive' current application. From the engineering point of view the problem concerning the ES dosage is particularly interesting. There have been attempts to build a model of the wound and to assess the distribution of the current in the tissue during electrical treatment with the aim of optimising current amplitude (VALENCIa, 1989). However, it is becoming evident that chronic wounds might be extremely diverse with regard to their location, size, tissue layers involved and state of the wound, which restricts the validity of each particular model because it is difficult to assess all the possible parameters. In setting the current intensity it is therefore best to rely on some decisive, significant physiological indicator (or several of them). This will probably mean using different current intensities for particular wounds, yet the reactions evoked in the tissue should be very similar. In the case of the tetanising current, smooth, scarcely visible contraction of the muscles surrounding the sore offers itself as such an almost ideal indicator. Direct current, however, does not produce any such well established indication and optimal current intensity adjustment will not be possible until the mechanisms of its action are fully understood. 2.2.2 Spinal cord stimulation: Electrical stimulation enhances wound healing even if it is not applied to the wound or its close surroundings. Spinal cord stimulation (SCS) was introduced by SI-mALY et al. (1967) with the main purpose of reducing pain. Its beneficial results were later also observed in multiple sclerosis (Cook and WEINSXEIN, 1973; ILLlS et al., 1980) as well as in improving motor function and reducing spasticity (SIEGFRIED et al., 1978; RICHARDSONet al., 1979). One of the common observations associated with SCS was also an increase in peripheral blood flow (ILLIS, 1982; TALLIS et al., 1983). Because reduced blood flow is one of the major reasons for chronic ulcers, SCS produced healing of wounds as a beneficial side effect. Cook et al. (1976) described the effects of SCS in nine patients with varying degrees of vascular insufficiency in the limbs. They observed pain relief and healing of ulcers and noted increased skin temperature and improved blood flow. Similar results were also reported by RICHARDSON et al. (1979). Out of five cases of patients with spinal cord injuries treated with SCS for autonomic hyperreflexia, three developed pressure sores but all the wounds healed after the application of stimulation to the spinal cord.
Medical & Biological Engineering & Computing
May 1992
Another successful case of wound healing with SCS was reported by MEGLIO et al. (1981). A patient with excruciating pain and trophic ulcers on his right foot due to obstruction of the femoral artery had epidural electrodes implanted at the vertebral levels T-7 and T-8. After stimulation immediate pain relief was obtained and after two months a large reduction of the ulcers was observed concomitant with improved blood flow. The stimulation frequencies in these studies were between 30 and 100 Hz and the pulse width ranged between 100 and 500 #s. The amplitude was adjusted to a level where the patient felt a tingling sensation in the extremities. In the study by AUGUSTINSSON et al. (1985) 34 patients with severe limb ischaemia were treated by epidural spinal electrical stimulation. Complete healing was observed in 50 per cent of previously non-healing ulcers. In 70 per cent of the patients, improved skin temperature recordings were obtained. The stimulation was given in pulses of 200/~s with a frequency of 100 Hz. The stimulus amplitude was high enough to give comfortable paraesthesia in the affected limb. Stimulation with the same parameters was also used in the work by JIVEGARD et al. (1987) and by GRABER and LIFSON (1987). The first study included 32 patients with impending or already established distal arteriosclerotic or diabetic lower limb gangrene. SCS had no effect on the progression of established gangrene. Nevertheless, pain relief and improved skin ulcer healing were observed. Nine diabetic patients with non-healing ischaemic ulcers took part in the study by GRABER and LIFSON (1987). Of these, seven experienced pain relief and improved ulcer healing. Regretfully, in the reports dealing with SCS neither the quantification of the effect (healing rates of the stimulated wounds) nor the comparison with a control group of untreated ulcers was given. Therefore the evaluation of SCS effectiveness with regard to other stimulation modalities is unfeasible and it is hard to estimate whether it is justified to insert stimulation electrodes directly above the patients' spinal cord. The authors claim that the implantation is simple; however, it is an invasive intervention in the patient's body and if the implant is not complete (electrodes penetrating through the skin on the back) the patient is exposed to a considerable degree of discomfort. 2.2.3 Stimulation of acupuncture points: The painrelieving effect of transcutaneous electrical nerve stimulation (TENS) and acupuncture has been reported (LONG and HAGFORS, 1975; FOX and MELZACK, 1976). However, little attention has been paid to the peripheral circulatory effects of such stimulation. KAADA (1982) was among the first who tested changes in vasodilation due to the stimulation of acupuncture points. He performed TENS with pulses of 0.2ms duration one to three times daily for 3045 min at a low frequency (2-5 Hz). The low rate stimuli comprised bursts of pulses with an internal frequency of 100Hz (Fig. 2a). The current intensity was adjustable up to 50mA and increased until local contraction of adjacent muscles was achieved without pain. Various electrode positions were tried but the best responses were obtained when stimulating the web between the I and II metacarpal bones. This location is known as the 'ho-ku' point of traditional Chinese acupuncture. The negative electrode was attached to this point. The positive electrode was placed at the ulnar edge of the same hand (Fig. 2b). In four patients with Raynaud's disease and in two with diabetic polyneuropathy, low-frequency TENS of remote sites caused a dramatic peripheral vasodilation associated with relief of ischaemic pain and a rise in skin temperature. Medical & Biological Engineering & Computing
In another study KAADA (1983) promoted healing of chronic ulcerations of various aetiologies in ten patients with peripheral vascular insufficiency. He succeeded in healing wounds due to neuropathy, arteriosclerosis, varicles, thrombophlebitis, decubiti and scleroderma. In several later publications he reconfirmed his initial success and developed a theory about the effect of electrical stimulation on endogenous mediators causing vasoconstriction and vasodilation (KAADA, 1987). 50( 25(
kk
kk
cJ
time
>o -250 -500
Fig. 2
pulse duration frequency a v e r a g e current peak c u r r e n t
= 45-100~s = 1-125Hz = 0.5- 2.5 mA = 2 - 2.5A
b TENS stimulation of acupuncture points: (a) electric stimulation parameters; (b) electrode placement. (Redrawn from KAADA 1987)
In all his studies the patients served as their own controls. Therefore only ulcers which had been treated conventionally for several months without success were included. Interestingly, the beneficial effects of the stimulation lasted for several hours after the end of each stimulation session. Despite the excellent results Kaada reported, his work with surface stimulation of acupuncture points remained solitary, without successors. In many reports about trials without a control group of patients the authors refer to the duration of the wounds' existence prior to the beginning of ES treatment and to the resistance of sores to conventional treatment during that time. Yet, simultaneously with the beginning of ES therapy, often more attentive conventional care of the wounds is provided as well (debridement, daily wound cleaning, footbaths, antibiotics administration) so that eventual beneficial effects actually cannot be attributed solely to ES and the results obtained are not objective. Again, comparison with other electrical stimulation techniques used for wound healing acceleration is impossible due to the lack of objective data (absence of a control group), but it should be emphasised that stimulating acupuncture points represents by far the most elegant modality of electricity application. Owing to the simple electrode placement and current intensity adjustment, it is also particularly convenient for home use in non-hospitalised patients, The carryover effect observed eliminates the necessity for lengthy daily involvement by the patient. Nevertheless, extensive comparative studies will have to be performed before we become impressed about this or any other treatment modalities. 2.3 Hiyh-voltage pulsed currents Recently several companies have started to produce stimulators with high output voltage (100-500 V) and very short pulse durations, some of them using so-called 'twin
May 1992
261
pulses' (Fig. 3). The rationale for producing such pulses is not quite clear. Strength/duration curves of excitable tissue predict action potentials or muscle contraction at low pulse amplitudes and longer pulse durations as well as at large amplitudes and small pulse duration. For the use of twin pulses no physiological explanation could be found. The literature on the use of high-voltage stimulation for wound healing is scanty and out of the studies performed only one was controlled.
9 IFMBE: 1992
Medical & Biological Engineering & Computing
The major reason for the appearance of chronic wounds is ischaemia combined with damage to the nervous system and any resultant metabolic abnormalities. There are three major groups of patients who are susceptible to chronic wounds: (a) Spinal cord injured patients who develop pressure sores (decubitus ulcers) due to their inability to move and lack of sensibility in areas subjected to prolonged pressure on the skin, and deeper tissues from an external object while sitting or lying. Typical pressure sore locations are bone tuberocities (trochanter, sacrum, heels). (b) Patients with peripheral vascular disease in whom ischaemic ulcers occur in the lower extremities due to various pathological conditions, the most common being diabetes mellitus. In a large proportion of such patients their condition impairs to such an extent that amputation is required with the usual complication that the wound on the stump resists healing. (c) Geriatric patients in whom ulcers are due to combinations of immobility and reduced blood circulation.
May 1992
257
With these facts in mind it is obvious that any additional treatment modality which might offer hope for accelerated healing of chronic wounds should be amenable to careful investigation. In the following paragraphs we present a survey of several methods which seem to improve wound healing by various applications of electrical and electromagnetic fields. It is our goal to present medical engineers as well as clinicians with adequate evidence about the effects of electricity on wound healing. We hope that additional basic and clinical investigations will be encouraged and that electrical wound healing will soon be generally accepted as a safe, simple and efficient modality for treating chronic wounds. 2 S t i m u l a t i o n modalities A brief survey of the existing literature exposes a fascinating richness in the application of diverse electrical or electromagnetic fields to different parts of the body to heal chronic wounds. Practically all reports are rather exclusive regarding their specific method, and rarely even attempt to compare their treatment techniques with methods published by authors who used a different method. Recently some review papers were published which discuss electrical wound healing (DAYTON and PALLADINO, 1989; KLOTH and FEEDAR, 1990; WEISS et al., 1990). However, none of them addressed all current modalities and electrode locations, nor did they discuss possible mechanisms. To present a fairly complete picture regarding the application of electricity in human wound healing is the goal of the present review. In the second part of this paper (STEFANOVSKAet al., 1993) some of our results as well as problems with the evaluation of the healing process will be discussed. Obviously it is impossible to make a quantitative analysis of the comparative advantages and disadvantages of the different methods because the protocols, quantification and wound aetiologies in the published literature are vastly different. Therefore the aim of the following survey is not to 'grade' and compare the different techniques but only to make the clinician and engineer aware of the fact that there seem to be many truths in electrical wound healing and only time and more research will show if there is something like a single or optimal truth. 2.1 Direct current
It is quite difficult to trace back the history of electrical wound healing. One of its roots was investigations into natural electric fields. DuBoIS-REYMOND (1860) was the first who as early as 1860 described the currents of an injury leaving human skin wounds. The same phenomenon was observed by HERLITZKA(1910). He measured a current of about 1 #A in a skin injury on the hand. Experimental abrasions were made on human skin in research by CUNLIFFE-BARNES (1945) with the purpose of studying the potential difference between intact and injured skin during healing. The abrasions were found to be positive compared with the surrounding intact skin. The potential difference lessened during the healing of the wounds and disappeared when they closed over. The existence of injury currents was confirmed by measuring currents leaving the fingertips in children who had undergone accidental amputation (ILLINGWORTHand BARKER,1980). On average eight days after the wounding a peak average current of 22/~A cm- 2 was obtained. Thereafter it lessened and reached zero value when the wounds healed completely. These reports raised three questions: where is the source of injury currents, what is their physiological background, 258
and what is their role in the healing process ? Only the first of them was elucidated. In 1982 BARKER et al. presented a map of human 'skin battery' voltages, showing the arrangement of potentials on the surface of the intact skin on different parts of the human body with respect to the reference point in the subdermal layer. The surface of the skin was always negative compared with the deeper skin layers. The transcutaneous voltage was up to 40 mV. JAFFE and VANABLE(1984) finally located the skin battery in the epidermis or 'living layer'. Human skin thus possesses endogenous electrical properties that may influence wound healing. But can this natural regenerative process be enhanced by applying external electricity? This question gave rise to numerous studies which this review will deal with. The second historical root of electrical wound healing stems from the seventeenth century when gold leaf was applied to smallpox lesions to prevent scarring (ROBERTSON,1925). Following this route KANOF (1964) and WOLF et al. (1966) applied gold leaf directly to skin ulcers. Encouraged by the results and suspecting that the electrochemical influence of the gold may have been responsible for enhancing tissue healing, they conducted the first studies with low intensity direct electrical current. The rationale for electrical wound healing with direct currents thus originates from two sources: investigations of biopotentials and experiments with gold leaf. Direct currents were consequently applied for wound healing enhancement by several research groups. Initial studies on animal experimental wound models (CAREY and LEPLEY, 1962; ASSIMACOPOULOS, 1968a) demonstrated faster healing, greater wound tensile strength and intensified inflammatory reaction due to direct electrical current. The first report about human application (AsSIMACOPOULOS,1968b) deals with only three patients with chronic leg ulcers due to venous insufficiency which resisted conventional treatment for several years. Following the application of up to 0.1 mA of negativepolarity direct current to the ulcers, they healed completely in six weeks. One year later, the results of an extensive study by WOLCOTT et al. (1969) were published. They comprised 83 ischaemic skin ulcers. Electrical stimulation (ES) treatment consisted of three 2 h segments of 0.2-1 mA direct current applications each day. One copper mesh electrode was placed over the wound and the other proximal to the lesion. The optimum current amplitude was set empirically between visible bleeding (too much current) and copious serious exudate (too little current). Initially the polarity of the electrode in the wound was negative. The polarity was reversed after three days in noninfected wounds, while for infected ulcers the negative electrode was retained in the ulcer until the infection was cleared, plus three additional days. The electrode polarity was changed during the treatment each time the 'growth plateau' was observed. Stimulation was applied for 2 h followed by a 4 h pause. This 2:4 treatment ratio was repeated thrice daily, which totals 6 h of electrotherapy per day. In this study a control group was lacking, but nevertheless the eight participating patients had contralateral wounds of comparable size and aetiology, of which one wound received electrical stimulation and the other served as control. In all such cases significantly faster healing was obtained in stimulated wounds. In addition to accelerated healing, the strong antimicrobial effect of the negative-polarity current was observed in infected wounds. This study served as an example for many succeeding clinical trials. At the same time it is the most frequently cited work in the history of electrical wound healing,
Medical & Biological Engineering & Computing
May 1992
although the rationale for many elements of its study design remained unknown until the present day. The intensity of the stimulation current was adjusted for each wound with regard to its exudate, although it is dependent on the condition of the sore and on the phase of the healing process in which the wound is found, and therefore has no physiologically uniform meaning from which the dosage could be estimated. Electrical stimulation was applied for six hours daily, which represents nearly half of the active part of the patient's day. However, the dependence of the efficacy of ES on daily therapy duration has not been checked and it is possible that the same effect could be achieved with shorter applications, which would add considerably to the patient's comfort. Other questionable features are the alternation of electrode polarity and the daily cycle of current application. It is evident from the abovementioned studies of endogenous electricity (CuNLIEEEBARNES, 1945; ILLINGWORTH and BARKER, 1980) that the wound remains positive compared with the surrounding intact skin throughout the healing process. The potential difference during the healing merely diminishes, the polarity, however, remains the same. Thus if ES is intended to support the natural endogenous electrical events which accompany regenerative processes, there are no arguments for changing the polarity of the stimulation electrodes. To plan and accomplish a correct clinical study of the effect of a particular treatment modality on wound healing is quite difficult because a multitude of intrinsic (the patient's psychophysical condition, age) as well as extrinsic (nutrition, wound care) factors influence the healing process along with the therapy studied, and their effects are not easily separated. From this viewpoint contralateral lesions are acceptable as control wounds, as in this case all the above-mentioned factors are uniform for both groups of sores. Yet we cannot maintain that the effect of the electrical current is merely local. And since its possible systemic action has to be considered as well, intrapatient comparison of treatments does not seem to be the most objective method for evaluating the effects of ES. Moreover, the number of contralateral wounds in the trial discussed is very small, thus not permitting any statistically supported conclusions. Following the protocol of WOLCOTT et al. (1969), a group of 25 patients with venous, arterial and neurotrophic ulcers was treated by EDEL and FREUND (1975). They selected only difficult cases, which had resisted polypragmatic therapies for years. After electrical stimulation, the wounds started to heal in 20 out of 25 patients in treatment regimens of 3 x 1 h or 2 x 2 h of stimulation per day with a current density under the electrodes of about 0.1 mA cm -2, which was just below the sensation level. In 1976 GAULT and GATENS reported on the results of their extensive clinical study dealing with 76 patients having 106 ischaemic skin ulcers of different aetiologies, of whom six patients with contralateral ulcerations served as controls. Exactly the same protocol was used as in the study by WOLCOTT et al. (1969) except that the electrode polarity was reversed only once. The mean healing rate of the control ulcers was 14-7 per cent per week compared with 30 per cent per week of the treated counterparts. The mean healing rate of the 100 ulcers treated with lowintensity direct current was 28.4 per cent per week. Wound asepsis was observed within three to seven days with the use of the negative electrode over the ulcer. This is another study where the undoubted beneficial effects of weak constant direct current were presented, using comparison with a numerically poor control group Medical & Biological Engineering & Computing
of wounds. It is interesting to note that the efficacy of the ES therapy is obviously not reduced if after the first polarity exchange (negative to positive) of the electrode on the wound it remains positive throughout the subsequent treatment. This observation makes the necessity for and the sense of changing the polarity of the stimulation electrodes during the therapy questionable. F r o m among the studies with direct electrical current, due note should be taken of the work by CARLEY and WAINAPEL (1985), which is unique for including a numerically equal (N = 15) control group of randomly assigned patients, who were matched by age, diagnosis, and wound aetiology, location and size with the treated group. Electrical stimulation was applied to the wounds in the treated group twice daily for 2 h. The current level ranged between 0.3 and 0.7 mA, which revealed current densities of 0.03-0.11 mA cm -2, depending on the wound size. Again the electrode placed over the wound was negative for the first three days and positive thereafter. If, during the treatment, a plateau in healing was observed, the protocol of negative polarity at the wound site for a three-day period was restarted. The current amplitude was determined empirically after observing the stimulated tissue response (copious exudate or bleeding). Comparison of the treated wounds with the controls revealed 1.5-2.5 times faster healing of the stimulated wounds, which was statistically significant. Again the antiseptic qualities of the constant negative electric current were reported. Randomised clinical trials seem to be the most adequate way of comparing treatments. Random assignment of subjects to experimental groups abolishes the selection bias which often 'corrupts' the results of the research. In using this study design, attention should be paid to the homogeneity of the population studied with respect to all the remaining patient variables which could affect the research outcome. Nonsymmetrical distribution of patients with regard to age, sex, additional diseases and others can seriously jeopardise the objectivity of the results and blur the effect of the modality studied. A uniform distribution can be achieved with certainty by using a numerically sufficient sample of subjects. The survey of published research reveals that studies performed on a numerically small sample of subjects are extremely common. Such studies might thus be useless from the scientific point of view as they are unable to detect clinically important effects. On the other hand, they are also unethical in their consumption of financial and other resources and above all because of the patients involved (ALTMAN,1985). The nonappearance of the positive effect of the stimulation with a weak constant negative direct current on the healing of venous leg ulcers was reported by KATELARISet al. (1987). They compared the efficacy of four different treatments: standard local application of povidone-iodine and normal saline, both with and without electricity. The current of 0-02 mA was applied using an ulcer dressing which comprised titanium electrodes, the negative electrode overlying the ulcer. No significant differences were found, either between the povidone-iodine and normal salinic treatments, or between normal saline and normal saline with electric stimulation. The combination of povidone-iodine treatment with negative current, however, induced a significant decrease in the rate of healing. The authors attributed the retardation effect to possible intracellular toxic effects of negatively charge iodine ions forced into the tissue by the negative electrode. Although reports about the efficacy of a constantly applied current of negative polarity (AsSIMACOPOULOS, 1968a; b; KONIKOFF, 1976) as well as current of positive polarity (ALVAREZ et al., 1983) can be found in the liter-
M ay 1992
259
ature, one could conclude that the polarity of the electrode over the wound has to be reversed to obtain optimal effects of low intensity constant direct current. In the article by ROWLEV et al. (1974) the authors report on the effects of direct current on wound healing in rabbits. They claim that the application of the negative electrode to the ulcer suppressed both healing and local infection, while the positive electrode enhanced both. The authors thus suggested a treatment regime in which the negative electrode is applied to the ulcer until aseptic conditions are attained and the positive electrode thereafter to accelerate healing. Such a stimulation regimen likewise proved successful in the majority of the above-mentioned trials, but nevertheless it is hard to maintain that it is optimal because the essence of its action remains unknown. And, in spite of the evident beneficial effects of electrical wound treatment, the lack of knowledge regarding the mechanisms as well as the lack of carefully controlled studies seem to be major arguments used by the medical profession against routine clinical application of direct currents for healing chronic wounds. 2.2 Low-frequency pulsed currents As low-frequency pulsed currents (also known as tetanising currents) are quite popular in physical medicine, it is not surprising that they were tested for enhanced wound healing. However, it is interesting to note that these currents were not only applied locally to the wound or across it, but also to areas quite distant to the lesion. The two major distant locations were the spinal cord and acupuncture points. We shall thus divide this section into three subsections regarding the location of the stimulation electrodes. 2.2.1 Local application: The first report stems from THOMAS et al. (1982). They applied biphasic rectangular pulse trains of 4 0 H z and 0-1 ms pulse width with an amplitude producing slight muscle contraction and zero DC component, close to the wound. The duration of the trains was 5 s followed by a pause of 10 s (Fig. 1). One such session lasted for 30 rain and was repeated every 2 h during the day. With such a regimen several decubiti were healed. The authors report particularly about a quadriplegic patient with a sacral bedsore of 12cm diameter. After application of stimulation the sore healed in five months. Accelerated healing was attributed to an improved supply to the affected area and to decompression due to rhythmical contractions of stimulated muscles. Another successful attempt to heal chronic wounds with low-frequency pulses was reported by WESTERnOF and Bos (1983). They treated a patient with ulcers on the face due to trigeminal nerve transection with pulses of 120Hz and 250#s pulse width. The pulse train was 0.5 s on and 0.5s off for 30 min three times a day. The severe ulceration on the face did not respond to any wound treatment including
E
lttl//ttttltlllltlttt _iI ,/I/,, ,t, t
-5
,-
Fig. 1
260
time
5s
10s
5s
pulse duration :0-1ms frequency :40Hz
Trains of biphasic current stimuli which were applied locally to the wound in the study by THOMAS et al. (1982). (Redrawn from THOMAS et al., 1982)
antibiotic therapy for two years. After application of electrical stimulation the ulcers healed completely after six weeks. It is rather surprising that the above-mentioned case reports represent the only attempts of local application of this type of electrical stimulation for wound healing acceleration. The formation of chronic wounds is principally determined by insufficient tissue supply with oxygen and nutrients due to poor blood flow, irrespective of their diverse aetiologies (pressure sores, wounds of venous and arterial origin, neurotrophic ulcers). Tetanising currents slightly contract muscles in the sore surroundings, thus improving circulation and providing energy for the affected area. The ability of low-frequency pulsed currents to fight the basic cause of problematic healing has been far too little investigated and exploited. Unlike the method for direct currept applications, where one of the stimulation electrodes covers the wound, both electrodes are placed on the healthy skin surrounding the sore when tetanising currents are used. This means a minor risk for possible mechanical damage of newly formed tissue (too strong contractions should be avoided), but also much simpler 'noninvasive' current application. From the engineering point of view the problem concerning the ES dosage is particularly interesting. There have been attempts to build a model of the wound and to assess the distribution of the current in the tissue during electrical treatment with the aim of optimising current amplitude (VALENCIa, 1989). However, it is becoming evident that chronic wounds might be extremely diverse with regard to their location, size, tissue layers involved and state of the wound, which restricts the validity of each particular model because it is difficult to assess all the possible parameters. In setting the current intensity it is therefore best to rely on some decisive, significant physiological indicator (or several of them). This will probably mean using different current intensities for particular wounds, yet the reactions evoked in the tissue should be very similar. In the case of the tetanising current, smooth, scarcely visible contraction of the muscles surrounding the sore offers itself as such an almost ideal indicator. Direct current, however, does not produce any such well established indication and optimal current intensity adjustment will not be possible until the mechanisms of its action are fully understood. 2.2.2 Spinal cord stimulation: Electrical stimulation enhances wound healing even if it is not applied to the wound or its close surroundings. Spinal cord stimulation (SCS) was introduced by SI-mALY et al. (1967) with the main purpose of reducing pain. Its beneficial results were later also observed in multiple sclerosis (Cook and WEINSXEIN, 1973; ILLlS et al., 1980) as well as in improving motor function and reducing spasticity (SIEGFRIED et al., 1978; RICHARDSONet al., 1979). One of the common observations associated with SCS was also an increase in peripheral blood flow (ILLIS, 1982; TALLIS et al., 1983). Because reduced blood flow is one of the major reasons for chronic ulcers, SCS produced healing of wounds as a beneficial side effect. Cook et al. (1976) described the effects of SCS in nine patients with varying degrees of vascular insufficiency in the limbs. They observed pain relief and healing of ulcers and noted increased skin temperature and improved blood flow. Similar results were also reported by RICHARDSON et al. (1979). Out of five cases of patients with spinal cord injuries treated with SCS for autonomic hyperreflexia, three developed pressure sores but all the wounds healed after the application of stimulation to the spinal cord.
Medical & Biological Engineering & Computing
May 1992
Another successful case of wound healing with SCS was reported by MEGLIO et al. (1981). A patient with excruciating pain and trophic ulcers on his right foot due to obstruction of the femoral artery had epidural electrodes implanted at the vertebral levels T-7 and T-8. After stimulation immediate pain relief was obtained and after two months a large reduction of the ulcers was observed concomitant with improved blood flow. The stimulation frequencies in these studies were between 30 and 100 Hz and the pulse width ranged between 100 and 500 #s. The amplitude was adjusted to a level where the patient felt a tingling sensation in the extremities. In the study by AUGUSTINSSON et al. (1985) 34 patients with severe limb ischaemia were treated by epidural spinal electrical stimulation. Complete healing was observed in 50 per cent of previously non-healing ulcers. In 70 per cent of the patients, improved skin temperature recordings were obtained. The stimulation was given in pulses of 200/~s with a frequency of 100 Hz. The stimulus amplitude was high enough to give comfortable paraesthesia in the affected limb. Stimulation with the same parameters was also used in the work by JIVEGARD et al. (1987) and by GRABER and LIFSON (1987). The first study included 32 patients with impending or already established distal arteriosclerotic or diabetic lower limb gangrene. SCS had no effect on the progression of established gangrene. Nevertheless, pain relief and improved skin ulcer healing were observed. Nine diabetic patients with non-healing ischaemic ulcers took part in the study by GRABER and LIFSON (1987). Of these, seven experienced pain relief and improved ulcer healing. Regretfully, in the reports dealing with SCS neither the quantification of the effect (healing rates of the stimulated wounds) nor the comparison with a control group of untreated ulcers was given. Therefore the evaluation of SCS effectiveness with regard to other stimulation modalities is unfeasible and it is hard to estimate whether it is justified to insert stimulation electrodes directly above the patients' spinal cord. The authors claim that the implantation is simple; however, it is an invasive intervention in the patient's body and if the implant is not complete (electrodes penetrating through the skin on the back) the patient is exposed to a considerable degree of discomfort. 2.2.3 Stimulation of acupuncture points: The painrelieving effect of transcutaneous electrical nerve stimulation (TENS) and acupuncture has been reported (LONG and HAGFORS, 1975; FOX and MELZACK, 1976). However, little attention has been paid to the peripheral circulatory effects of such stimulation. KAADA (1982) was among the first who tested changes in vasodilation due to the stimulation of acupuncture points. He performed TENS with pulses of 0.2ms duration one to three times daily for 3045 min at a low frequency (2-5 Hz). The low rate stimuli comprised bursts of pulses with an internal frequency of 100Hz (Fig. 2a). The current intensity was adjustable up to 50mA and increased until local contraction of adjacent muscles was achieved without pain. Various electrode positions were tried but the best responses were obtained when stimulating the web between the I and II metacarpal bones. This location is known as the 'ho-ku' point of traditional Chinese acupuncture. The negative electrode was attached to this point. The positive electrode was placed at the ulnar edge of the same hand (Fig. 2b). In four patients with Raynaud's disease and in two with diabetic polyneuropathy, low-frequency TENS of remote sites caused a dramatic peripheral vasodilation associated with relief of ischaemic pain and a rise in skin temperature. Medical & Biological Engineering & Computing
In another study KAADA (1983) promoted healing of chronic ulcerations of various aetiologies in ten patients with peripheral vascular insufficiency. He succeeded in healing wounds due to neuropathy, arteriosclerosis, varicles, thrombophlebitis, decubiti and scleroderma. In several later publications he reconfirmed his initial success and developed a theory about the effect of electrical stimulation on endogenous mediators causing vasoconstriction and vasodilation (KAADA, 1987). 50( 25(
kk
kk
cJ
time
>o -250 -500
Fig. 2
pulse duration frequency a v e r a g e current peak c u r r e n t
= 45-100~s = 1-125Hz = 0.5- 2.5 mA = 2 - 2.5A
b TENS stimulation of acupuncture points: (a) electric stimulation parameters; (b) electrode placement. (Redrawn from KAADA 1987)
In all his studies the patients served as their own controls. Therefore only ulcers which had been treated conventionally for several months without success were included. Interestingly, the beneficial effects of the stimulation lasted for several hours after the end of each stimulation session. Despite the excellent results Kaada reported, his work with surface stimulation of acupuncture points remained solitary, without successors. In many reports about trials without a control group of patients the authors refer to the duration of the wounds' existence prior to the beginning of ES treatment and to the resistance of sores to conventional treatment during that time. Yet, simultaneously with the beginning of ES therapy, often more attentive conventional care of the wounds is provided as well (debridement, daily wound cleaning, footbaths, antibiotics administration) so that eventual beneficial effects actually cannot be attributed solely to ES and the results obtained are not objective. Again, comparison with other electrical stimulation techniques used for wound healing acceleration is impossible due to the lack of objective data (absence of a control group), but it should be emphasised that stimulating acupuncture points represents by far the most elegant modality of electricity application. Owing to the simple electrode placement and current intensity adjustment, it is also particularly convenient for home use in non-hospitalised patients, The carryover effect observed eliminates the necessity for lengthy daily involvement by the patient. Nevertheless, extensive comparative studies will have to be performed before we become impressed about this or any other treatment modalities. 2.3 Hiyh-voltage pulsed currents Recently several companies have started to produce stimulators with high output voltage (100-500 V) and very short pulse durations, some of them using so-called 'twin
May 1992
261
pulses' (Fig. 3). The rationale for producing such pulses is not quite clear. Strength/duration curves of excitable tissue predict action potentials or muscle contraction at low pulse amplitudes and longer pulse durations as well as at large amplitudes and small pulse duration. For the use of twin pulses no physiological explanation could be found. The literature on the use of high-voltage stimulation for wound healing is scanty and out of the studies performed only one was controlled.
