Hyperbaric Oxygen Therapy and The Diabetic Foot

Abstract

Hyperbaric oxygen therapy is an adjunctive wound-healing modality receiving increasing use for problem wounds, particularly diabetic foot wounds. Nevertheless, few clinicians understand the physiologic basis for this modality; how patients are selected, or the expected results. The author reviews the development of hyperbaric oxygen therapy, selection of patients, and clinical studies of this modality for diabetic patients with foot wounds.

Hyperbaric oxygen therapy has been used in the clinical setting since 1943 when it was adopted by the US Navy for treatment of decompression sickness and air embolism. While these are widely accepted indications for the use of hyperbaric therapy, there have been no double-blind, prospective, randomized, controlled trials showing the efficacy of hyperbaric oxygen therapy for decompression sickness and arterial gas embolism. Its value is based on application of physical laws in a known pathophysiologic process.

In terms of the diabetic patient with a problem foot wound, the known pathophysiologic entity that serves as the primary indication for hyperbaric oxygen therapy is tissue hypoxia. Use of hyperbaric oxygen therapy in a hypoxic foot simply applies the physical laws to reduce tissue hypoxia. However, guidelines for treatment of the diabetic foot with hyperbaric oxygen therapy are not uniform. Lack of uniformity reflects changing opinion as studies clarify the role of this adjunctive wound care modality.

At present, some centers accept all patients with hypoxic foot wounds; others accept only those with nonpressure-based ulcers; others accept patients with a high risk of above-ankle amputation, particularly with limb-threatening infection in the setting of ischemic peripheral vascular disease; and still others accept only chronic diabetic wounds unresponsive to all standard modes of treatment.

As with decompression sickness and air embolism, there have been no double-blind prospectively controlled, randomized trials demonstrating the efficacy of hyperbaric oxygen therapy in the treatment of the diabetic foot. Nevertheless, the physiologic principles for the use of oxygen therapy in general, and the rationale for the use of hyperbaric oxygen therapy in particular, have been developed throughout the years through a variety of studies. The rationale for the use of hyperbaric oxygen therapy is discussed below.

Rationale for the Use of Hyperbaric Oxygen Therapy

The use of hyperbaric oxygen therapy for “problem” wounds (ie, hypoxic wounds) stems from the basic and clinical science research of Silver [1], Niinikoski et al [2], Hunt and Pai [3], Hunt et al [4], Hohn et al [5], Zederfeldt [6], and Knighton et al [7]. The important concepts developed by these and other investigators include:

(1)

oxygen is a controlling factor in wound healing and the control of infection. Oxygen has significant effects with respect to collagen deposition, angiogenesis, and leukocyte function [2,3,4,5]. Normal wound healing alone can probably use more oxygen than is usually delivered [8]. Hypoxic wounds benefit by correction of local tissue oxygen tension;

(2)

the primary method of killing bacteria by phagocytic leukocytes is an oxygen-dependent process [9,10,11]. The common pathogenic species involved in infections in humans are destroyed by oxidative killing after phagocytosis [12]. Hypoxia leads to impaired wound healing and defects in control of infection [5,7,12,13,14,15];

(3)

perfusion is the fundamental determinant of tissue oxygen tension [16,17]; ischemia and hypoxia are not the same thing [18]. Large vessel disease is the most important cause of impaired oxygen delivery to the wound and is the primary defect that must be addressed initially in all hypoxic wounds. It is likely that most, if not all, wounds occurring in ischemia vascular watersheds have some degree of local tissue hypoxia. Tissue hypoxia, however, may also arise as a result of local processes. Infection, ischemia-reperfusion injury, microangiopathic peripheral vascular disease, autonomic neuropathy with altered flow creating arteriovenous shunting, and hemorrheologic abnormalities each contribute to tissue hypoxia. These defects are not bypassable, and in concert with abnormal inflammatory responses and abnormal wound-healing responses, result in delayed wound healing;

(4)

anemia is not inimical to wound healing [6,17,19,20]. Wound healing can take place at normal rates as long as perfusion is maintained at a normal or elevated level [17,20]. Dissolved oxygen can potentially supply or augment the oxyhemoglobin-bound supply of oxygen to wounds;

(5)

tissue oxygen tension is the ultimate factor affecting the progress of wound healing and the control of infection. The supply of oxygen to tissues is diffusion limited [21,22,23,24]. Wounds do not discriminate as to the method of delivery of oxygen. Oxygen supplied to the wound diffuses along a partial pressure gradient into the wound space to supply tissue needs. This is true whether oxygen is bound to hemoglobin or carried in dissolved form in the plasma [16];

(6)

the microcirculation may exert an independent effect on tissue perfusion. Occlusive microangiopathy is not an important issue. However, altered autonomic and microvascular pathophysiologic responses, combined with the effects of activated leukocytes and hemorrheologic abnormalities, may independently affect perfusion, tissue oxygenation, and outcome [25,26,27,28,29,30,31].

The important techniques developed to elucidate the role of oxygen in wound healing and infection include the use of nondamaging microelectrodes to measure oxygen tension at the cellular level [1], the rabbit ear chamber [32], and a variety of methods to measure tissue oxygen tension such as needle electrode [1], tissue tonometer [33], implantable Clark polarographic electrode [34], and the spectrophotometric tissue optode [35].

The adequacy of tissue oxygenation is not exactly the same thing as a normal tissue oxygen tension. In an analogous fashion, a drop of saltwater has the same chemical analysis as an ocean of saltwater, but the quantity clearly differs. Demands for oxygen are usually met with ordinary means. Baseline health status, large vessel status, oxyhemoglobin saturation, and arterial oxygen tension significantly contribute to the ability to deliver oxygen to the wound.

However, the most direct link to satisfying wound demands is tissue oxygen tension. Tissue oxygenation may be able to be met by greater unloading of oxyhemoglobin, or may require augmentation in order to sustain tissues. Tissue metabolic demands determine the quantity of oxygen requisite for control of infection and healing. Methods for measuring wound oxygen demands and quantifying the adequacy of tissue oxygenation do not exist in clinical practice. Measures of tissue oxygen tension similarly exist only in investigational trials [17,36].

The current use of transcutaneous oximetry is an indirect measure of underlying tissue oxygen tension, and does not represent deeper soft tissues or bone. Its limits must be understood. Meier-Hellman et al [37] and Abbot et al [38] have discussed the significance of measuring oxygen supply and demand in wounds with the potential for measuring the efficacy of hyperbaric therapy. Falcone and Caldwell [39] have also addressed the subject of wound metabolism. Currently, no practical, readily available manner exists by which a physician may optimize selection of patients for hyperbaric therapy that improves the limits imposed upon patient selection by the use of transcutaneous oximetry.

The concept of tissue oxygen debt in critical ischemia holds some promise for improving patient selection use of hyperbaric therapy [40]. The use of nitroimidazoles for hypoxia imaging and nuclear magnetic resonance spectroscopy represents attractive methods to assess tissue oxygen debt and image wound metabolism [41,42]. There has not to date been any testing of these techniques for imaging of ischemic limbs or wounds.

Efforts to increase tissue oxygen tension may fail regardless of the modality used to improve conditions at the wound site if a sufficient quantity of oxygen cannot be delivered. Hence, it would seem that there is practical significance in pursuing research directed at elucidating wound supply and demand characteristics.

Physical Laws and Physiology of Hyperbaric Oxygen Therapy

The physical basis for the elevation of oxygen tension resides in Henry’s law: the amount of a gas dissolved in solution is directly proportional to the partial pressure of that gas. As LoGerfo [43] has noted, there is nothing specific about hyperbaric oxygen therapy in the treatment of the diabetic foot. The principles for the use of hyperbaric oxygen therapy are based on application of physical laws in an effort to correct abnormal tissue oxygen tension.

Hyperbaric oxygen therapy is the only widely available modality that is potentially capable of raising the tissue oxygen tension of a hypoxic wound when the oxygen tension cannot be elevated by standard techniques. The increased dissolved oxygen is available for use by the tissues to support wound healing and aid in the control of infection. The body simply does not differentiate between oxygen that has been carried by hemoglobin, and oxygen dissolved in plasma. Providing an adequate oxygen gradient from the capillary to the hypoxic tissue bed is the only requirement that must be fulfilled in order for the tissues to benefit. Whenever impaired tissue oxygen tension is an impediment to the clearance of bacteria or progression of wound healing, correction of tissue hypoxia should result in clinical improvement. Supernormal healing is not an expected benefit, since the role of hyperbaric therapy is only to provide an adequate level of oxygen to the tissues so that healing can take place.

Heimbach has stated that there is probably a range of hyperoxic therapy that can be used to effect adequate tissue oxygenation (personal communication, 1995). The key to effective use of ground-level oxygen or hyperbaric therapy is maximizing perfusion, since no amount of oxygen is effective in nonperfused spaces. Demonstration of elevated tissue oxygen tension is the ideal method of indicating that increasing the oxygen supply results in delivery of oxygen at the wound site. Measuring tissue oxygen tension alone, however, may not always accurately reflect successful use of oxygen therapy, even in the face of obvious clinical improvement. This paradox occurs in infected wounds when all available supply is consumed for metabolic purposes. Other measures of improved responsiveness to hyperbaric therapy are therefore needed.

Silver [1] demonstrated the characteristics of oxygen gradient extending from the nearest intact blood vessel to the central avascular portion of a wound. He also demonstrated that changes in arterial oxygen tension using ground-level oxygen resulted in a steep gradient, but with relatively little increase in the diffusion distance from the wound. Diffusion distance varies directly with the oxygen tension. Theoretic work by Krogh [44] estimated that a tenfold increase in arterial oxygen tension under hyperbaric conditions at 3 atmospheres absolute produces a potential threefold increase in diffusion distance from the capillary. Overcoming diffusion barriers has been cited as one potential benefit of hyperbaric therapy [45].

A series of elaborate experiments performed over a number of years at multiple centers provided the physiologic rationale for hyperbaric therapy. The most direct determinant of wound healing and control of infection is tissue oxygen tension. Perfusion, however, is the fundamental determinant of delivery of oxygen to the wound. At the wound site, diffusion of oxygen occurs along a partial pressure gradient. The partial pressure of oxygen is therefore important. For some wounds, the only reasonable method of elevation of tissue oxygen tension is with the use of hyperbaric oxygen therapy.

Zederfeldt [6] demonstrated that anemia does not impair wound healing as long as perfusion is maintained at a high level and hematocrit is not <15%. Wound oxygen tension may be low despite normal or elevated cardiac index, arterial O₂ content, and arterial O₂ tension. Heppenstall et al [24] demonstrated in a series of trauma patients that wound oxygen tension might be disastrously low despite normal or elevated cardiac index (increased as much as 50%), normal arterial oxygen content, and elevated arterial oxygen tension. The authors postulated that local events in the microcirculation could independently lead to tissue hypoxia, and result in increased risk of infection and impaired or delayed healing. This early physiologic research was consistent with and complemented other research showing that changes in the microcirculation such as the degree of leukocyte activation, or hemorrheologic properties of blood, could have an independent effect leading to hypoperfusion, hypoxia, and microvascular shutdown. The work of these investigators was done primarily on acute wounds.

Perfusion is the fundamental determinant of tissue oxygen supply [16]. However, the supply of oxygen to the wound’s healing edge is diffusion limited along a partial pressure gradient. Hence, the partial pressure of oxygen at the wound must be considered. The importance of the partial pressure of oxygen and the partial pressure gradient has been repeatedly appreciated by acclimated mountaineers on expedition to Mount Everest and residents of alpine terrain who experience poor wound healing and infection control of minor injuries despite having physiologically adapted to the reduced parital pressures of oxygen at high altitude.

Oxygen diffuses from red cells into the tissues of wound space along a gradient and is competitively consumed by leukocytes, fibroblasts, and bacteria, thus creating an oxygen demand. Metabolic demands are a multifactorial product as depicted by Wipke-Tevis [46] (Fig. 1). Oxygen unloading from hemoglobin under normal conditions easily supplies the metabolic demands of cells nearest the blood vessel. However, multiple factors may lead to the metabolic demand outstripping oxygen supply. Clinically, this is most commonly seen when infection is superimposed upon ischemic peripheral vascular disease. When peripheral vascular disease nears the threshold of critical ischemia, infection may profoundly affect tissue oxygenation. Increasing the partial pressure of oxygen with hyperbaric oxygen therapy increases the partial pressure gradient and can result in increasing the tissue oxygen tension to normal levels. The differences between the normoxic and hyperbaric environment are seen in the figures adapted from Boykin [47] (Figs. 2A and B).

Sheffield [48,49] and Sheffield and Heimbach [50] used implantable polarographic electrodes to demonstrate that hyperbaric oxygen therapy resulted in the elevation of tissue oxygen tension in diabetic patients with chronic wounds. In a series of 13 diabetic patients with chronic foot wounds, Sheffield demonstrated the hypoxic nature of these wounds, their response to 100% oxygen at normal atmospheric pressure, and their response under hyperbaric pressures. The response of these wounds over time to intermittent hyperbaric oxygen therapy was also demonstrated. Multiple intermittent episodes of hyperbaric therapy elevated the wound oxygen tensions for some chronically hypoxic wounds: some, but not all, of these wounds went on to heal. His work essentially extended and expanded the work of previous investigators’ research with acute wounds.

The most hypoxic region along a blood vessel is located at the venous end, the so-called hypoxic corner. This region can be thought of as having a “normal” hypoxia under ordinary wound-healing circumstances owing to physiologic processes that consume oxygen. Despite a “normal” hypoxia related to ordinary circumstances, this region remains susceptible to development of infection.

Infection exacerbates underlying hypoxia caused by ischemic peripheral vascular disease. Patients with diabetes are known to have abnormal wound-healing responses. These abnormal responses simply compound the problem of wound healing and control of infection. Patients with impaired immune responses are similarly at risk of abnormal wound healing and control of infection, but for differing reasons than the patient with diabetes. In either case, infection in patients with underlying ischemia or immunologic impairment increases the risk of infectious necrosis. Hohn et al [5] noted that as tissue oxygen tension falls toward zero, leukocytes lost approximately one half of their maximal killing capacity. The greatest drop in killing capacity occurred when local oxygen tensions fell below 30 mmHg.

Vasoconstriction does occur in response to hyperbaric oxygen therapy, but oxygenation of the tissues can nevertheless take place. Wells et al [51] demonstrated elevation in muscle and subcutaneous tissues during hyperbaric therapy. Post-treatment levels fell but remained above baseline values for 3 to 4 hr after exposure. The required ambient pressure necessary for elevating tissue oxygen tension to normal undoubtedly varies with local conditions. Similarly, the duration of elevation after treatment may not be the same for all wounds.

Oxygen radicals have been implicated in the pathophysiology of peripheral vascular disease, retinopathy stroke, and myocardial infarction [52,53,54,55,56]. These radicals are derived from leukocytes and can arise spontaneously, or as a result of ischemia, hypoxia, inflammation, infection, and other conditions. Hyperbaric oxygen therapy has frequently been feared as a potential contributor to further generation of toxic oxyradicals. Hence, many clinicians have reservations in referring ischemic patients for hyperbaric therapy despite recognized therapeutic indications. There is evidence that this either does not occur or is of only marginal concern with intermittent therapy. Thom [57] showed that hyperbaric therapy blocks leukocyte activation and subsequent oxyradical damage. Visonà et al [58] followed malondialdehyde levels as a marker of oxyradical-induced damage in patients with peripheral vascular disease undergoing hyperbaric therapy. The levels were actually shown to decrease throughout the duration of therapy (30 days). However, the patients in another study were treated with the rheologic agent pentoxifylline, which has been reported to inhibit oxyradical generation [59]. Finally, studies of ischemia-reperfusion injury indicate that hyperbaric oxygen therapy mitigates the damaging effects of toxic oxygen radicals [60]. In summary, there seems to be no reason to fear that hyperbaric therapy might increase tissue damage as a result of generation of oxygen radicals.

Pathophysiology of the Diabetic Foot

Physiologic and rheologic abnormalities in the vasculature and microvasculature of patients with diabetes have been described by Edmonds [60], Edmonds et al [61], Watkins and Edmonds [62], Tooke [63], Boulton et al [64], Rendell et al [65], Ditzel [28], and Schmid-Schönbein and Volger [29]. The derangements identified in the microvascular system of diabetics leading to altered local hemodynamics suggest a physiologic role for microvascular disease in the pathogenesis of foot ulceration. Excellent reviews of diabetic microvascular disease have been produced by Tooke [63], Jaap and Tooke [26,66], Flynn and Tooke [67], and McMillan [68].

Pecoraro et al [69] summarized the local abnormalities found by various investigators as potentially contributory to decreased skin oxygenation in the diabetic foot (

Table 1. Contributors to Impaired Oxygenation in the Diabetic Foot.

).

It is probably impossible to define or catalog the full extent of contribution of any of the aforementioned physiologic and rheologic abnormalities that affect an individual patient’s status. In the infected patient with multiple underlying abnormalities, the ability to draw upon reserves to supply hypoxic tissues with sufficient oxygen for current metabolic needs may be critically tested.

Confusion regarding the use of hyperbaric oxygen arises from early accepted ideas of the occlusive nature of diabetic microvascular disease and current knowledge of this entity. The effects of hyperbaric therapy depend solely on the physical laws and not the changing medical opinion regarding the nature of microvascular disease.

Investigations in the 1960s and 1970s into the physiology of oxygen tension in the wound space were paralleled by investigations into the anatomical basis of diabetic microvascular disease. Anatomical studies by Goldenberg et al [74] had demonstrated what was then thought to be the sine qua non of the diabetic foot: basement membrane disease. It was thought that occlusive microcirculatory disease in patients with diabetes resulted in tissue hypoxia. Basement membrane disease supposedly led to a diffusion defect in the microcirculation of the diabetic foot [45].

The widely held view that microcirculatory disease produced an occlusive lesion creating a hypoxic foot thwarted the development of surgical bypass procedures of the limb and foot. In 1984, LoGerfo and Coffman [80], and in 1990, Morain and Colen [81] described a series of studies that demonstrated the lack of specificity of basement membrane disease as an anatomical sine qua non for diabetic microvascular disease. They also questioned the significance of occlusive microangiopathy in creating a diffusion defect for oxygen at the tissue level. Strandness et al [82] repeated the study of Goldenberg et al in blinded fashion and was unable to confirm changes in the basement membrane as a sine qua non of diabetes. Conrad [83] made anatomical casts from amputated specimens of diabetic and nondiabetic limbs and found no difference in small vessel disease, no difference in transcutaneous values in diabetics versus nondiabetics suggesting the lack of a diffusion defect. Based on these studies, vascular surgeons began to reject the importance if not the overall concept of microcirculatory disease in patients with diabetes and began to perform bypass procedures below the ankle into the foot at rates never before seen.

Prior to the review of LoGerfo and Coffman [80], the extreme view of the importance of occlusive microcirculatory disease in diabetes was that peripheral bypass procedures would have little benefit in changing the overall course of events in patients with diabetic foot problems. An additional extreme view existed as well: hyperbaric oxygen therapy might overcome this problem, perhaps even without the need for peripheral vascular bypass procedures. Hyperbaric oxygen therapy appeared to be able to overcome the diffusion defect created by thickened basement membranes and other abnormalities in the diabetic patient that led to a hypoxic foot.

Faulty reasoning regarding the importance and significance of occlusive microcirculatory and basement membrane disease, coupled with the success of surgical revascularization procedures, led some to question the importance of microvascular disease. The potential physiologic importance of Sheffield’s [48,49,50] work with hyperbaric oxygen seems to have become caught up in the medical prejudices of the day. The lack of importance of microvascular disease as an occlusive problem seems to have been well demonstrated by the success of revascularization reported by Gibbons [84], Gibbons et al [85], and LoGerfo et al [86].

In terms of satisfying tissue oxygen needs in patients with diabetes, hyperbaric oxygen therapy may well be the right modality in the right circumstance, even when used with faulty reasoning. No study has examined expected differences in outcome as a result of aggressive revascularization versus hyperbaric therapy during the era of “occlusive” microvascular disease. Such a retrospective study would be of historic interest only, since the role of hyperbaric therapy has never been to supplant revascularization. Hyperbaric oxygen therapy has always been intended to be used adjunctively.

Gibbons et al [85] have reviewed the success of revascularization of the lower limb and foot and have rejected the notion of occlusive microvascular disease of the foot. LoGerfo [43] has taken the surgical pendulum almost to the opposite extreme, stating that there is nothing specific about hyperbaric oxygen therapy with respect to the diabetic foot. He is correct. But that is not the same as stating that there is no role for hyperbaric oxygen therapy in the treatment of the diabetic foot. As noted earlier, the effects of hyperbaric therapy depend solely on the physical laws, and not changing medical opinion regarding the nature of microvascular disease. Conclusive proof of the efficacy of hyperbaric oxygen therapy in the diabetic foot lesion awaits double-blind, prospectively controlled trials. LoGerfo et al [86] view aggressive surgical revascularization as preferable to amputation, noting that the morbidity and mortality are less, while the costs are comparable to amputation. In the author’s opinion, appropriate use of hyperbaric therapy may significantly aide in determining the choice between revascularization for limb salvage, and amputation.

Mechanisms of Action

The studies by Hunt and Pai [3] and Knighton et al [7] regarding the role of oxygen in wound healing have formed the basis for the conclusion that hyperbaric oxygen therapy stimulates angiogenesis, development of granulation tissue formation through the enhancement of fibroblast function, and control of infection with enhancement of leukocyte function by elevating the partial pressure of oxygen in tissues. These studies explain the role of oxygen as a metabolite involved in many basic functions.

Other authors have cited additional actions and possible mechanisms by which hyperbaric oxygen therapy may exert beneficial effects in wound healing. The studies of Thom [87] have already been mentioned. However, it should be noted that the blocking effect of hyperbaric therapy on leukocytes lasts up to 8 hr, outstripping the duration of hyperoxic elevation of hyperbaric oxygen therapy. Mathieu et al [88] found that red blood cell deformability was improved with hyperbaric oxygen therapy. Nemiroff [89] found that hyperbaric oxygen therapy and the rheologic agent pentoxifylline were synergistic in improving red blood cell flow. In a rabbit ischemic ear model, Zhao et al [90] found that hyperbaric oxygen therapy and growth factors acted synergistically in enhancing wound healing.

Siddiqui et al [91] also used a rabbit ischemic ear model. The group has suggested that hyperbaric oxygen therapy results in increased tissue oxygen capacitance. This increased capacitance is seen after 14 treatments in hyperbaric-exposed animals and is manifested by an increased tissue responsiveness to 100% oxygen inhalation at normal atmospheric pressure. The baseline values were not different between the hyperbaric-treated group and the control group, but the responsiveness to oxygen inhalation was significantly different. The effect is not seen in nonhyperbaric-treated animals. Animals exposed to only five treatments demonstrated in an intermediate capacity to respond to oxygen challenge. The authors have proposed that hyperbaric therapy may act not only as a simple metabolite available for wound healing, but also as a stimulus to signal transduction of specific genetic pathways important for wound healing. It was suggested that following increased tissue oxygen capacitance by following responsiveness to 100% O₂ inhalation in the clinical setting might serve as a useful predictor of healing. The investigators did not believe that tissue oxygen capacitance was explicable by neovascularization.

A review of 37 unpublished cases followed in the author’s institution with serial intermittent transcutaneous oximetry performed at 14 and 21 days, along with serial wound photos, would support Siddiqui’s contention of an increased tissue capacity to respond to oxygen inhalation. However, the author disagrees with the conclusion that capacitance is caused by initiation of signal transduction. Myers and Wolf [92] demonstrated that neovascularization in rabbits under normal wound-healing conditions was demonstrable as early as 48 hr in muscle tissue, 72 hr in skin, and 5 days in fascia and bone. Neovascularization was greatest at 7 days. These data seem to be within the time frame in Siddiqui’s study. Uhl et al [93] performed a similar study in an ischemic mouse ear model and followed blood-flow changes with a laser Doppler. The study did not find intergroup differences in tissue blood flow, but noted that the hyperemic response was less in the ischemic model. The investigators concluded that changes in re-epithelialization were caused by increases in arteriolar oxygen content and oxygen diffusion. Finally, wound photos from the author’s patient population clearly show a pronounced granulation tissue formation that parallels oxygen responsiveness and appears to be consistent with the time frame reported by Siddiqui. From the author’s viewpoint, improved responsiveness to oxygen challenge is the result of simple neovascularization. Regardless of the mechanism, it is clear that repeated treatments result in improved responsiveness to oxygen challenge.

What are the indications for hyperbaric oxygen therapy in the treatment of the diabetic foot; how are patients selected; what studies support the use of hyperbaric oxygen therapy in the treatment of diabetic patients with problem foot wounds; how often should patients be treated; and what is the appropriate duration of therapy?

Indications and Patient Selection for Hyperbaric Oxygen Therapy

The basic and clinical science investigations described in the previous sections have demonstrated the role of oxygen in wound healing and infection, and the importance of perfusion in oxygen supply. The important contribution of the partial pressure of oxygen and oxygen gradients in establishing an adequate tissue oxygen tension has been discussed in the preceding sections. Rabkin and Hunt [94] and La Van and Hunt [95] have described the physiologic principles for the therapeutic use of oxygen in wound healing and the control of infection, and the potential uses of hyperbaric oxygen therapy in clinical situations.

The clinical application regarding which wounds are suitable for consideration of hyperbaric therapy includes partially ischemic wounds wherein elevation of tissue oxygen tension can be demonstrated, infected wounds, skin grafts, and revascularized wounds or flaps with secondary failure demanding re-exploration. The Undersea and Hyperbaric Medical Society has provided guidelines for the use of hyperbaric oxygen therapy in the management of problem wounds [96]. These guidelines will undoubtedly change with continuing critical review of applicable research.

The selection of patients with diabetes for hyperbaric oxygen therapy may be based on chronicity, wound grade, and transcutaneous oximetry data. Selection of diabetic patients with foot wounds for treatment with hyperbaric oxygen therapy was initially based on chronicity and the demonstration of tissue hypoxia in wounds that had lapsed into a state of nonhealing [48,49]. Strictly speaking, the definition of a chronic wound had never been defined on an absolute temporal basis [97].

Currently, patient selection is based on wound grade and transcutaneous oximetry. Some clinicians prefer to select patients based on wound grade and use transcutaneous oximetry data primarily for following trends throughout a course of treatment [98]. The Wagner Scale is the most commonly used wound-grading system for the diabetic foot [99]. Matos and Nuñez [100] have advocated a modified Wagner Scale that accounts for the presence of critical ischemia at all levels of involvement as an added risk factor for nonhealing. Brakora and Sheffield [101] used the scale by Knighton et al [102] in their algorithm previously published in the clinics.

The initial studies of chronic wounds in patients with diabetes were accomplished using implantable electrodes. Unfortunately, routine clinical use of implantable tissue oxygen tension devices is not currently done. Transcutaneous oximetry has supplanted the use of implantable tissue oxygen tension monitoring. The use of transcutaneous oximetry for patient selection is based on studies by Orenstein et al [103], Dowd et al [104], White and Klein [105], Hauser [106], and Harward et al [107].

Transcutaneous oxygen tension is the most commonly used objective criterion for patient selection. The data by Wyss et al [108], demonstrated an increasing probability of wound healing failure with decreasing transcutaneous oxygen values. However, even at a tension of zero, the probability of failure was not 100%. Magnant et al [109] state that if the wound PO₂ cannot be raised to > 30 mmHg by normobaric 100%O₂, then hyperbaric oxygen therapy should be considered.

Matos and Nuñez [100] and Brakora and Sheffield [101] also advocate the use of the transcutaneous oximetry to aid patient selection. Matos and Nuñez recommended hyperbaric therapy for patients with TcPO₂ values < 40 mmHg, and regional perfusion index values (the ratio of wound value to chest reference value) of < 0.6. Brakora and Sheffield recommended acceptance of patients for hyperbaric therapy if transcutaneous values on room air were less than 40 mmHg and a 50% rise could be demonstrated by oxygen inhalation challenge (Fig. 1). These authors also reported that wounds with room-air transcutaneous values of < 10 mmHg had little chance of healing even when hyperbaric oxygen therapy is used.

Kindwall and Goldmann [110] have cautioned against the use of transcutaneous criteria as the sole indicator for treatment. This caution is a result of recognition of the technical limits and interpretive differences with this technique. As an example, it is well known that a value of 0 mmHg may result from no blood flow, or when supply equals demand. Oishi et al [111] reported that primary wound healing without hyperbaric therapy could be acheived regardless of the initial value if a 10 mmHg rise in transcutaneous values could be acheived with oxygen-inhalation testing. Harward et al [107] reported that primary wound healing without hyperbaric therapy could be achieved if the initial value on room air was 10 mmHg, or an increase of > 10 mmHg was achieved with oxygen-inhalation testing.

In the author’s experience, neither the absolute value on room air, nor the values obtained with oxygen-challenge testing are sufficiently reliable to eliminate clinical experience in the selection of patients.

Oxygen-inhalation testing seems to be of use primarily as an indication that certain wounds are incapable of developing an elevated transcutaneous oxygen tension, no matter what partial pressure of oxygen is provided when administered at ground level (ie, without hyperbaric pressure). Failure to elevate transcutaneous oxygen tension with ground-level oxygen is an indication only to repeat the examination under hyperbaric exposure conditions. Finally, low values on room air are not a priori hopeless. Experience must be used to augment the absolute numerical values obtained and decide the meaning of such low values. Low values have a differential diagnosis and may indicate poor perfusion and poor oxygenation, overwhelming local demands, or a combination of events. Scheffler and Rieger [112] have reviewed the information content of transcutaneous oximetry. Clinical expertise is still required when evaluating patients with abnormal transcutaneous oximetry studies.

The physician is sometimes faced with a decision regarding the wisdom of revascularization surgery for limb salvage versus primary amputation. Such limbs are often infected and transcutaneous values will be abnormally low as a result of infection alone, thus affecting any decision regarding selection of the site of amputation. Gibbons [113] has advocated deferment of the primary operative procedure until after control of systemic and local infection. In cases with preexisting critical limb ischemia, superimposed infection will further shift the balance toward amputation. Hyperbaric oxygen therapy should be strongly considered as an adjunct in controlling infection prior to committing the patient to amputation. In the postoperative revascularized case, anatomical revascularization is not immediately synonymous with physiologic return to normal. Reperfusion injury may complicate the case, and is another indication for hyperbaric oxygen therapy.

Hyperbaric oxygen therapy has also been reported to be of use in preparing wound beds for skin grafting and accepting flaps for coverage of large defects [114]. Improved percentage of “take” has been reported for skin grafts and flaps. Flaps rotated into previously irradiated recipient beds are ideal candidates for hyperbaric therapy. The flap itself usually contains an adequate blood supply. However, the recipient bed being compromised, may fail to hold the transposed flap and benefits from perioperative hyperbaric therapy.

Hyperbaric Studies

Although hyperbaric oxygen therapy has been in clinical use since 1943, meticulously detailed studies examining its use for specific problems have been slow to be published. Despite a progression of studies of the diabetic foot, the cost and difficulty of development of prospectively controlled, double-blind, randomized trials have been impediments to the scientific standards of medicine today. Davis et al [18] responded to critics that patients with chronic wounds unresponsive to standard techniques were appropriate historic controls demonstrating the efficacy of hyperbaric therapy. It has been pointed out that no surgical technique for management of diabetic foot problems has been subjected to prospectively controlled trials [18,115,116] Feldmeier et al [117] have stated that double-blind studies do not exist in the entire field of radiation therapy. For technical and ethical reasons, such studies are impossible.

The expense of double-blinding hyperbaric studies has long been recognized, as has the necessity for performing such studies. Outcome based on reduction of amputation rates, and “healing” defined as the lack of need for amputation above the ankle are commonly used study endpoints in hyperbaric studies. Absolute measures of healing such as measurement of granulation tissue are not present in existing studies.

Finally, Cianci et al [118] and Cianci [119] have frequently described the paradox of medical practice that compensates more for limb amputation than for limb salvage. It is clear from the growing problem of management of chronic wounds, especially in an aging society, that novel techniques based on the physiology of the wound microenvironment are needed. Hyperbaric oxygen therapy and growth factor research are approaches that specifically address identified pathophysiologic processes. Of the two modalities, hyperbaric oxygen therapy is the most specific treatment for the identified abnormality.

Davis et al [18] authored a retrospective review of the group’s experience in 1987. They reported a 70% success and 30% failure rate of healing. Healing was considered successful if the patients did not require amputation above the ankle. Most failures occurred in patients with large vessel disease, which limits perfusion distally.

Additional retrospective reports [120,121,122,123] have been published, but the most provocative data have been provided by Stone et al [124] and Stone and Scott [125] in two separate reviews. In the first review, Stone et al noted that patients referred for hyperbaric therapy had greater wound volumes, more wounds per patient, and had a greater percentage recommended for amputation. Despite having more serious wounds, the limb-salvage rate was greater in the hyperbaric treated group (72% versus 53%; p < 0.002).

In the second review, Stone and Scott measured wound volumes and divided patients into three groups: standard care, standard care with adjuvant hyperbaric therapy, and both combined with growth factors. While there was no significant difference in healing rates for patients treated by standard care (60% healing) versus standard care with hyperbaric therapy (63% healing), the average wound volume (2,750 mm³; n = 38) in the hyperbaric-treated group was double that of the standard care group (1,144 mm³; n = 250). The group treated with the combination of standard care, hyperbaric therapy, and growth factors had the greatest healing rate (80%; 2,414 mm³; n = 49). The study suggested that combination therapy might be synergistic, as had been reported in animal studies.

Baroni et al [126] produced the earliest prospective, controlled trial of hyperbaric therapy in the treatment of Wagner classes 3 and 4. The study involved 28 patients with 18 in the hyberbaric oxygen group and ten in the control group. The groups were similar in terms of age, duration of diabetes, diabetic complications, control of diabetes, and lesion size by surface area measurement. Treatment was similar for both groups; however, the treating surgeons were blinded as to which patients received hyperbaric oxygen therapy. In the hyperbaric-treated group, 16 out of 18 healed (88%) and two required above-ankle amputation. In the control group, one out of ten healed, four required amputation, and five remain unchanged. Although statistical significance was not reached because of overlap, the hyperbaric-treated group averaged 62 hospital days versus 82 days in the control group.

Stone and Cianci [127] are currently pursuing a prospectively controlled, randomized double-blind trial of hyperbaric oxygen therapy in the treatment of diabetic foot wounds. Despite the effort required to pursue such a study, criticism has already surfaced regarding the study design, which includes growth factors in both the treatment and control groups. In addition, the study is unfortunately not stratified to reveal subgroups that may be better candidates for hyperbaric therapy. The Herculean effort to produce this trial is nevertheless to be applauded. The only existing double-blind trial of the benefit of hyperbaric therapy for wound healing was performed by Hammarlund and Sundberg [128]. While the study showed a significant benefit on ulcer healing, the design specifically excluded diabetics and patients with atherosclerotic peripheral vascular disease.

Complications of Hyperbaric Oxygen Therapy

The risks involved in undergoing hyperbaric therapy are no greater than for many other procedures in medicine today. Direct complications usually involve relatively minor problems such as ear block, barotitis, barosinusitis, and a usually spontaneously reversible hyperoxic myopia. Oxygen toxicity seizures are stated to occur at a rate of one in 10 to 12 thousand. This information was based on retrospective review of the experience of multiplace units and was not controlled for seizures from other causes. The true rate of oxygen toxicity seizures may be half the generally reported value. Pulmonary oxygen toxicity remains a theoretic risk that has not been reported as a result of protocols in therapeutic use today. Pulmonary barotrauma with resultant pneumothorax or gas embolism is reported as a potential complication in one in 50 to 60 thousand treatments.

Fire in the hyperbaric unit is the most feared risk. Multiplace units are equipped with sprinkler systems that provide a deluge if triggered. Monoplace units cannot be outfitted with a sprinkler system. Hyperbaric units place a premium on prevention. Fires generally occur at rate of one per year worldwide and are accorded the attention of an airline crash. Hospital fires in general, and operating room fires in particular occur at much greater rates, but with much less attention.

Conclusion

La Van and Hunt [95] have long promoted the importance of oxygen and the role of perfusion in wound healing, and have recognized the role of adjunctive hyperbaric oxygen therapy.

Hyperbaric exposure is the extreme upper end of the safe use of oxygen therapy in an attempt to meet physiologic needs of hypoxic wounds by increasing the partial pressure of oxygen. Present techniques do not allow for precise selection of patients, but the current clinical use in patients with hypoxic problem wounds has been formally approved by the Undersea and Hyperbaric Medical Society. The ongoing study of Stone should clarify some questions regarding the usefulness of hyperbaric therapy in the treatment of the diabetic foot. And it will undoubtedly raise other questions. Careful, well designed studies are required to stratify patients for referral to hyperbaric therapy; to address issues of amputation, and treatment endpoints including functional outcome, costs of providing hyperbaric therapy for indigent and disabled patients versus overall charges, and combination therapy with other wound-healing modalities undergoing current research.