Management of Diabetic Wounds: Expert Panel Consensus Statement

Macrovascular disease

The global epidemic of diabetes overlaps significantly with the rising prevalence of peripheral artery disease (PAD). Fowkes and colleagues estimated the global incidence of PAD back in the early part of this decade and found that >200 million people are afflicted, and that the increasing prevalence over the last decade is of the order of 25%. ¹⁰⁸ The rise in PAD is not unique to any socioeconomic status. In fact, areas of the greatest growth around the world are lower income countries and areas of Southeast Asia. PAD is a huge and growing global health problem, requiring a lot of resources from health care systems as the population in the world ages and diabetes prevalence continues to rise. The Eurodiale observational studies^109–111 demonstrated that the coprevalence of PAD and a DFU is a very worrisome combination in terms of prognosis for the patient. DFUs precede perhaps up to 85% of all nontraumatic limb amputations, and we know that PAD is highly prevalent in patients with diabetes for 10 years or longer.

In Eurodiale, 49% of the patients with an open DFU had PAD, and the presence of PAD was a very strong negative predictor of healing, particularly when compounded with infection. ¹⁰⁹ Therefore, it is vitally important to diagnose PAD in all patients with DFUs because it is a major driver of outcomes, and also because the treatment needs to be tailored to the presence and severity of the vascular disease. Work from Armstrong and colleagues ⁵⁷ has shown that 5-year mortality in patients with DFU is 25%, a somewhat grim statistic greater than or equal to most types of cancers. Even more sobering is the prognosis for those with chronic limb-threatening ischemia (CLTI), the most advanced stage of PAD, in which inadequate perfusion leads to tissue necrosis. Once individuals are diagnosed with CLTI, their annual mortality rate is generally around 10%, exceeding almost every type of cancer, save perhaps lung cancer. If a patient requires a major limb amputation due to advanced vascular disease, there are substantial data demonstrating that both their life is shortened and their quality of life severely reduced. ⁵⁷

Turning to diagnosis, it is fairly well accepted that patients with diabetes should have a vascular examination, at least annually, and that they should be asked about the classic symptoms of PAD. However, it is also well-known that classical PAD symptoms may be absent or masked in patients with peripheral neuropathy and in those who have a sedentary lifestyle and do not exercise. Furthermore, patients who have a primarily distal pattern of disease (e.g., involving below knee arteries) often will not have exercise-related claudication but may present de novo with frank tissue loss and poor healing. Performing an extensive cardiovascular and particularly vascular history, along with foot pulse examinations, should be a standard of care (SOC) for everyone with diabetes and certainly for those with a prior history of DFU. In patients who have symptoms that are suggestive of PAD with an abnormal vascular examination or an existing foot wound, objective measurement of the severity of vascular disease is critical before surgery or other wound intervention and is a practice guideline measure.

Standard noninvasive arterial studies rely on segmental pressures and Doppler waveforms in the lower leg, ankle, and foot. They include ankle pressure, the ankle brachial index (ABI), toe pressure, toe brachial index (TBI), and the Doppler waveforms, which should be inspected. There are much data concerning this topic, and no single one of these tests is optimal for assessing perfusion in all patients. ¹¹² Moreover, localized perfusion to the wound bed (“angiosome” or “woundosome” ¹¹³ ) is not always reflected in these measures. Current guidelines reflect that one or more of the tests should be done in conjunction, and the visual inspection of the Doppler waveforms associated with the measured pressures is quite useful. The sensitivity and specificity for any one test still leave significant room for improvement. However, significant PAD is quite unlikely if the ankle index is between the range of 0.9 and 1.4 and/or if the toe index is greater than 0.7 in the presence of a multiphasic Doppler waveform. There are special circumstances that can influence these measurements. Most important is the presence of arterial calcification, particularly medial artery calcification (MAC) in the tibial arteries, that can render the ankle pressures either not measurable or falsely elevated. An ABI >1.4 is considered an unreliable measurement for PAD. Digital artery calcification is less common, although not rare in those with diabetes and renal disease. Accordingly, the toe pressure and TBI may be a better measure of vascular disease in the diabetic population and in those with unmeasurable or inappropriately high ABI. It should be noted, however, that MAC can also affect the pedal arteries and thus even digital pressures may be rendered inaccurate^114,115; a plain radiograph of the foot (SOC for most DFUs) demonstrates the presence and severity of pedal MAC, which can help the clinician to interpret the noninvasive study results within this context.

Visual examination of the Doppler waveforms can be helpful to determine if the measured pressure is likely accurate. If the waveform is barely pulsatile, the distal pressure value is likely to be an overestimation. Importantly, these tests are noninvasive, fast, and inexpensive in most office settings. In current practice in most parts of the Western world, patients with DFU should have access to vascular testing to quantify the degree of ischemia more routinely and more consistently. Improvement in this area is critical so that patients obtain an accurate diagnosis and treatment; it also provides better epidemiological data.

An algorithm from the Global Vascular Guidelines (GVG) ¹¹⁶ provides guidance on vascular testing (Fig. 2). For patients with a clinical suspicion of advanced limb ischemia such as pain at rest in the foot that has the characteristics of ischemic rest pain, or any degree of tissue loss in the presence of an abnormal vascular examination, it is suggested that ankle pressures, toe pressures, and Doppler waveforms should be done. When the ankle pressure or waveform is abnormal and if there is any tissue loss, toe pressures are preferred because they have a better predictive value for wound healing than ankle pressures alone. Thus, most patients with DFU benefit from toe pressure/TBI measurement.

OPEN IN VIEWER

Figure 2.

Suggested algorithm for noninvasive vascular testing and evaluation in patient with suspected chronic limb-threatening ischemia (CLTI). ABI, ankle brachial index; PAD, peripheral arterial disease; TBI, toe brachial index; WIfI, Wound, Ischemia, Foot Infection (from Conte et al., 2019 ¹¹⁶ ).

Limb staging is the next important area to address. There are several systems for vascular staging and wound staging out there, but the only one that combines them both into a single framework is the Society for Vascular Surgery’s WIfI staging system that was published in 2014. ⁶³ This scheme provides a TNM-type approach to grade 3 independently critical factors in patients presenting with tissue loss. To use WIfI, the clinician must characterize the extent and depth of the wound, assess the severity of ischemia noninvasively, and the presence and extent of foot infection. These individual grades are combined into four overall WIfI stages for the limb to predict the 1-year risk for major amputation, and to suggest which patients would benefit most from revascularization. An important concept in WIfI is that rather than applying a simple threshold value of perfusion pressure, there is recognition of a gradient of ischemia that may be limb threatening depending on the clinical circumstance. Previous concepts of “critical limb ischemia” were based on a sharp cutoff value (e.g., ABI <0.4 or a toe pressure <30 mm Hg). However, more moderate degrees of ischemia in the presence of an extensive wound or infection may also in fact be “critical.” This evolving concept also explains the change in terminology recently to CLTI.

When considering the relationship between blood flow and wound healing, a fundamental observation is that regardless of which measure one uses for perfusion, the relationship to wound healing is an S-shaped curve with a large intermediate range of ischemia that may be most relevant clinically in the setting of an advanced wound or infection. In other words, the degree of ischemia that may be clinically important to impair healing is likely to be quite different depending on the complexity of the wound and the presence of infection.

Review of the WIfI grades and stages (Fig. 3) demonstrates how the system is used for prognosis.^63,117 Each of the three primary components is graded on a 0–3 scale, and then these domains are combined into four overall stages of limb threat. An expert consensus panel developed the scheme, informed by evidence review, and suggested the assignment of the overall stages based on the expected risk of major amputation at 1 year. There have been many subsequent publications that have validated the WIfI staging system largely among cohorts with CLTI, demonstrating that it is quite good at predicting 1-year major amputation risk. ¹¹⁸ However, these studies are primarily retrospective in nature, and largely included patients who got revascularization. WIfI is yet to be validated prospectively in all comers presenting in a longitudinal cohort manner and specifically in the DFU population. The available data demonstrate that patients in WIfI stage 1 have a very low to negligible risk of major limb amputation with good basic foot care. Conversely, those presenting in stage 4, even with aggressive revascularization in all of these series, face about a 20% chance of major limb loss. Stages 2 and 3 seem to overlap in intermediate severity. WIfI staging is meant to be used in a manner similar to TNM in cancer, which means it is not a one-time assessment but is used to gauge trajectory over time. It should be reassessed in any patient whose wound or symptoms are not resolving or worsening.

OPEN IN VIEWER

Figure 3.

Society for Vascular Surgery Threatened Limb Classification System based on Wound, Ischemia and Foot Infection (WIfI). (A) Grades shown in top panel. (B) Stages shown in bottom panel correlate with risk of major amputation (adapted from Mills et al., 2014, and Aboyans et al., 2018^117,118).

In practice, one makes an initial assessment about the need for revascularization based on the patient’s overall risk, the severity of the ischemia, and clinical presentation of the limb. This defines an initial treatment approach that should include optimal wound care and off-loading. One should then reassess if the wound is not improving after 4 to 6 weeks, including remeasuring the state of any ischemia and restaging that foot. ¹¹⁹ This provides a systematic, data-driven approach to assess the treatment strategy’s progress and whether a change is needed, particularly in the domain of revascularization.

Regarding limb revascularization in patients with diabetes, there is a long history of using distal extremity revascularization to prevent limb loss dating back to the middle of the 20th century. Pioneering vascular surgeons in the 1960s demonstrated that you could salvage limbs with vein bypass grafts in patients with advanced peripheral vascular disease, including patients with diabetes. It is important to understand that although the absolute certainty of MVD in the foot of patients with diabetes is well recognized, the notion that one could not effectively treat these patients with a macrovascular approach such as bypass surgery was disproven heartily in the 1980s ¹²⁰ by Dr. Frank LoGerfo who is credited with bringing this concept to the forefront and removing the nihilistic approach about MVD that suggested revascularization would not be effective in patients with diabetes.

Contemporary strategies for limb revascularization have evolved rapidly due to the increasing availability and success of catheter-based endovascular techniques in addition to traditional open surgery. This, together with better staging, allows us to consider how to approach these patients in a precision medicine way. First of all, in any patient with DFU in whom PAD has been demonstrated or is suspected, evaluation by a vascular specialist is absolutely indicated to consider the severity of ischemia and the potential benefits and risks of revascularization. The WIfI staging system is helpful to define the potential benefits of revascularization in patients across the spectrum of CLTI. When ischemia is severe, such as a toe pressure less than 30 mm Hg or an ankle pressure less than 50 mm Hg, the consultation should be urgent. When ischemia is mild or moderate, the wound clinician can consider focusing on optimizing wound care management and off-loading with frequent visits to assess progress. If the foot or clinical status is deteriorating, a consultation with a vascular specialist should be indicated.

The GVG ¹¹⁶ provide a patient-oriented framework to define the optimal revascularization strategy based on the PLAN concept of Patient risk, Limb threat severity (WIfI), and the ANatomic pattern of PAD in the limb. For anatomical staging, a new integrated approach called the Global Limb Anatomical Staging System (GLASS) was developed. GLASS integrates the complexity of disease in the limb from the groin to the foot based on the selection of a preferred target artery path (selected tibial vessel) from angiographic imaging including the ankle and foot. Technical success and durable patency of a limb revascularization, just as in coronary disease, are highly dependent on how the anatomical pattern plays off disease. Without delving into the fine details of the GLASS, it involves separately grading the disease severity above the knee (femoral and popliteal) and below the knee (tibial and pedal vessels) in separate components based on high-quality angiographic imaging. These grades are then combined into three overall stages for the limb as a whole, based on the understanding that to achieve effective revascularization you need to achieve inline flow to the foot. The resulting stages range from low-complexity disease, which is very amenable to endovascular treatment with good expected results, to high-complexity disease patterns where both the expected technical success and the downstream patency of an endovascular intervention are markedly reduced. The GLASS is increasingly being validated in published studies, including a recent systematic review and meta-analysis. ¹²¹

The GVG recommendations that an optimal revascularization strategy in CLTI patients should be individualized are based on the defined critical elements of PLAN outlined above. In a patient presenting with signs of advanced limb ischemia or tissue loss, it begins with limb staging. If it is initially a low-risk limb presentation, you may start with wound care and surveillance. If the limb is judged to be in immediate or higher risk, the patient is evaluated as an appropriate candidate for limb salvage based on their overall risk and their functional status. For frail or nonambulatory patients, there is an important role for primary limb amputation and palliative wound care. If the patient is a candidate for limb salvage, you should estimate the operative and cardiovascular risk using one of multiple available risk calculators (e.g.,^122,123) and consider the options for revascularization using WIfI and GLASS.

If revascularization appears feasible in a high-risk patient, one may choose to use a catheter-based approach as much as possible. However, in a standard-risk patient with CLTI, choosing between open bypass and endovascular intervention should consider the anatomical complexity and the ability to undergo a vein bypass. The GVG suggested a rubric for average surgical risk patients that is based on WIfI and GLASS, as well as the availability of great saphenous vein for a bypass conduit (Fig. 4). In this scheme, many patients are approached with an endovascular-first approach highlighted in yellow. Those patients who have very complex anatomy and high-risk limbs probably do better with an open bypass (Fig. 4, green). There are many that fall into an indeterminate zone based on current data, and in general a stronger evidence base is needed to guide recommendations.

OPEN IN VIEWER

Figure 4.

Suggested initial revascularization strategy for average surgical risk CLTI (chronic limb-threatening ischemia) patients with an available great saphenous vein is based on limb severity and anatomical complexity (from Conte et al., 2019 ¹¹⁶ ).

Very recently, level 1 data have been published comparing the effectiveness of revascularization strategies in CLTI patients. The recently published BEST-CLI randomized trial was conducted primarily in the United States and was funded by the National Institutes of Health over an 8-year period. This is a landmark trial that was published in the New England Journal of Medicine in 2022 ⁷² ; its results are reflected in the most current International Working Group guidelines on PAD management ¹²⁴ and likely will be reflected in other upcoming societal guidelines related to the management of CLTI. Investigators in BEST-CLI trial randomized patients with CLTI, who were deemed to be a reasonable surgical risk and who could be treated with either a catheter-based approach or with open bypass surgery. Patients were screened for the presence of an adequate great saphenous vein and placed into two different trial cohorts. Cohort 1 included patients who were deemed to be likely to get a saphenous vein bypass with a single segment of vein. Cohort 2 included subjects likely to require an alternative conduit for bypass based on their preoperative vein mapping. In cohort 1, over 1,400 patients were randomized and followed for nearly 3 years. The majority of these patients had diabetes, more than 70%. The primary endpoint of the trial was the occurrence of major adverse limb events, which included conversion to open bypass surgery, thrombectomy/thrombolysis, and major amputation, or all-cause death. Among patients in cohort 1 after a median of 2.7 years of follow-up, there was a 32% relative risk reduction in the primary endpoint among the patients randomized to open bypass first. This comprised a 65% relative risk reduction in major reinterventions, a 27% relative risk reduction in major limb amputation, and no difference in all-cause mortality or major adverse cardiovascular events. This finding in cohort 1 was robust across nearly every patient subgroup, including those with diabetes, tissue loss, and infrapopliteal disease. There was no significant difference in the primary endpoint in cohort 2 (those lacking an adequate great saphenous vein).

It is important to recognize that similar to any clinical trial, generalizability of the BEST-CLI results is an important question. The trial was designed to focus on CLTI patients who were assessed to be of reasonable surgical risk and had an acceptable anatomy for either type of revascularization. It is unclear at present what percentage of the overall CLTI population this may represent, but it is likely to be a minority. Overall results demonstrated excellent limb salvage and improved quality of life in both treatment arms of the trial. Nonetheless, the results of BEST-CLI strongly suggest that some patients with CLTI should be treated initially with an open bypass for a more effective outcome. There is much more to come from subsequent analyses in this trial. It is also important to note that the BASIL-2 trial from the United Kingdom was also recently reported and presented a contrasting result from a different (and smaller) population. ¹²⁵ Comparison of these trial populations and outcomes is beyond the scope of this review. However, the much larger scale and inclusive scope of the BEST-CLI trial, combined with the robustness of its findings, have led to renewed enthusiasm for the clinical effectiveness of vein bypass surgery in appropriate patients with CLTI.

In conclusion, it should be clear that the early and accurate diagnosis of PAD in patients with diabetic foot wounds is paramount. Limb staging and restaging are important for prognosis and to define treatment success. More validation studies of the WIfI system are needed with possible modifications going forward based on well-done outcome studies. Revascularization should be considered in any DFU patients with PAD, based on their clinical staging, their clinical trajectory, and on their overall patient risk and goals of care. For some frailer patients, primary amputation and palliation have important roles to play.

Microvascular disease

While macrovascular disease is well known to be associated with the development of and the poor healing-associated DFU, MVD association with DFU remains controversial. MVD is also known as microangiopathy or small-vessel disease and has been associated with a range of conditions that impact the arterioles, venules, and/or capillaries, in general vessels with diameters of 100 um. ¹²⁶ Disease that affects these small vessels manifest locally in organ-specific diseases. MVD is associated with T2DM, particularly those patients where uncontrolled blood glucose levels cause chronic cellular stress include retinopathy, nephropathy, coronary microangiopathy, and neuropathy in the peripheral nervous system.^127–130

As a review, central regulation of skin blood flow balances sympathetic-mediated vasoconstriction and vasodilatation, responsible for most of the blood flow to the skin capillary network, including the foot. ¹³¹ Numerous arteriovenous anastomoses allow direct communication between meta-arterioles and venous plexus shunting large volumes of blood to the skin surface for heat dissipation but at the same time bypassing nutrient capillaries. Local regulation of blood flow is mediated mainly by local vascular responses and by neurogenic C-fibers via the nerve axon reflex (Lewis triple response) stimulated by noxious stimuli, including mechanical, chemical, and thermal injury. Thus, direct injury (e.g., thermal injury above 42°C) increases blood flow specifically at the site of injury by decreasing the precapillary sphincter tone, a direct vascular mechanism. In contrast, in the surrounding skin there is also an increase in blood flow, the “flare response.” Nociceptive C fibers responsible for neural conduction to the thalamus to mediate the pain sensation also simultaneously conduct to the local neural network, resulting in the release of vasoactive substances, including substance P, histamine, and calcitonin gene-related peptide, into the surrounding skin at the site of injury to cause vasodilation.^132–135

With diabetes, structural changes occur, such as thickening of the capillary basement membrane, diminished capillary luminal size, and pericyte degeneration. These are more pronounced in lower extremities due to gravity-induced hydrostatic pressure and loss of postural vasoconstriction in people with diabetes. ¹³⁶ Increased capillary pressure increases fluid filtration and probably induces inflammatory responses in the microvascular endothelium, contributing to basement membrane thickening and hyalinosis of the arterioles. These microvascular changes have also been linked to increased glycosylation and formation of nonenzymatic AGEs. ¹³¹ As a result, basement membrane thickening can impair normal transport across the capillary wall.

There is controversy regarding the significance of structural changes in the small vessels of the diabetic foot. One study of 152 amputation specimens, of which 92 had diabetes, described endothelial proliferation sufficient to almost occlude the lumen of digital arteries and smaller vessels. ¹³⁷ Subsequent studies using light microscopy, vascular casting, and physiological studies did not confirm the presence of occlusion. ¹²⁰ However, a recent study on T2DM found that capillary microangiopathy was present in neuro-ischemic and neuropathic diabetic foot skin, with a predominance of arteriolar occlusions in the neuro-ischemic foot. ¹³⁸

Unfortunately, standard vascular tests (ABI and TBI) do not correlate with known clinical diagnoses that support the presence and severity of MVD. These data agree with the notion that ABI and TBI are generally used for evaluation of large-vessel arterial flow, rather than diseases of the small vessels. In practice, most physicians do not directly measure perfusion in the small vessels. More recently, significant improvements to evaluate skin microcirculation have emerged, they are still not used by most clinicians caring for DFU. Among them is capillaroscopy, which is a sensitive and exacting method for estimating the skin’s microcirculation at a microscopic level by in vivo visualization of the density, morphology, and blood flow in capillaries using a real-time video technique. ¹³⁹ Laser Doppler flowmetry and imaging noninvasively measure blood flow in the superficial 1 to 1.5 mm of skin and as such include capillaries, superficial meta-arterioles, venous plexuses, and acral areas. Using monochromatic laser light, back-scattered by moving red blood cells, blood flow and velocity can be assessed. ¹⁴⁰ Hyperspectral imaging is a technique assessing skin blood flow noninvasively based on the principle that oxy- and deoxy-hemoglobin have different absorption spectra resulting in differences in reflectance of light emitted in specific wavelengths. With the aid of complex imaging algorithms, information regarding the perfusion of tissues.^141,142 Transcutaneous oxygen pressure, often used for patients undergoing hyperbaric oxygen, noninvasively measures the partial pressure of oxygen molecules, diffusing through the skin from the superficial microvasculature. Heating greatly increases perfusion such that the transcutaneous oxygen levels more closely correlate with local perfusion, and thus, transcutaneous oxygen pressure in heated skin has been considered a measure of skin perfusion and included in some vascular units, a measure that will influence vascular intervention.^143–146

Often MVD in patients with diabetes is clinically defined and the presence of clinically defined MVD increased peripheral amputation risk independently of PAD (3.7 times higher than no MVD) and worked synergistically with PAD (amputation risk 13.9 times higher than baseline) to yield a 22.7 times higher amputation risk in patients who had both MVD and PAD. MVD has also been shown to correlate with increased rates of DFU development. ¹⁴⁷

Limited high-quality evidence exists to support specific intervention to address or prevent MVD. Glucose control is paramount, and recently, a study evaluating a healthy lifestyle consisting of nonsmoking, having a healthy body weight, vigorous exercise, high-quality diet, and moderate alcohol consumption was published. ¹⁴⁸ After multivariable adjustment, a healthy lifestyle before and after the diagnosis of T2DM was associated with a lower risk of developing microvascular complications.

Key Consensus Points:

Early and accurate diagnosis of PAD in patients with diabetic foot wounds is paramount, and revascularization should be considered in any DFU patient with PAD within the context of overall patient risk and goals of care.

Limb salvage versus amputation

A successful limb salvage procedure is defined by maximized anatomical limb preservation and the ability to remain ambulatory at a functional level. For patients with diabetic wounds of the lower extremity, limb salvage is generally preferable to amputation. However, regardless of whether a limb can be salvaged, the critical factor for long-term survival is the level of activity where ambulation equals more life years especially with the ability of independent function. If a patient with diabetes exercises three times a week, their chances of increasing their life years increases by 2.4 to 3.6 years and if they exercise daily, then their life years can increase by as much as 5 years.^151–153 The decision between limb salvage and major LEA (MLEA; e.g., below-knee amputation (BKA), above-knee amputation (AKA)) hinges on various factors, including the patient’s overall health, wound location and healing potential, and patient goals and preferences. If anatomical limb salvage does not lead to improvement in ambulation and level of activity, then a more functional amputation may be preferable if that will result in increased ambulation through the use of prosthetics. For less active patients who need to manage daily activities, foot level amputations such as Chopart’s or Lisfranc’s may suffice. ¹⁵⁴ These procedures help prevent further soft tissue loss, preserve limb length, and minimize progression toward MLEAs.

However, if the foot cannot be functionally reconstructed to meet a patient’s goals, a functional and carefully performed BKA is often recommended. BKAs are particularly suitable for active patients seeking high-functioning limbs, offering quicker recovery and effective functional rehabilitation. Generally, MLEAs are considered a last resort due to their association with depression, reduced quality of life, and high 5-year mortality. ¹⁵⁵ BKA ambulation rates, ranging from 16% to 77%, highlight the importance of careful patient selection based on overall health, surgical technique, rehabilitation, and social support. ¹⁵⁶

MLEA 5-year mortality rates exceed 60%, with BKA mortality surpassing 68.0%, largely due to worsening comorbidities. ⁵³ Diabetes independently increases 1- and 5-year mortality rates by up to 1.4 times. The pooled 5-year mortality rate for all cancers is 31%, similar to 30.5% for DFU alone. The 5-year mortality of minor and major amputations was 46.2% and 56.5%, respectively. ⁵⁷ The highest mortality occurs in patients who are nonambulatory, have renal failure, and have had an AKA. Patients with AKAs are up to four times less likely to regain ambulatory function compared with those with BKAs. This reduced likelihood of walking contributes to the higher mortality rates associated with AKAs. ¹⁵⁷ There is a noted bias toward AKAs, despite a recommended BKA:AKA ratio of 2.5:1. This is likely skewed by wound healing complications associated with BKAs. About 90% of AKAs heal, compared with lower healing rates for BKAs, with 50% of unhealed BKAs eventually requiring AKAs.^158,159

The 5-year mortality rates for BKA and AKA patients at one institution were noted to be 36.3% and 44.0%, respectively, the lowest reported in the literature. ¹⁶⁰ This may be attributed to a multidisciplinary approach and careful patient selection aimed at maximizing postoperative ambulatory function, regardless of limb salvage or amputation. The ultimate goal is that ambulation and daily exercise can extend life by up to 5 years. ⁵⁷ Consequently, a function-based approach is recommended that emphasizes ambulation as crucial to extending life years.

Function-based approach to BKAs

The amputee’s ability to ambulate depends on the patient’s comorbidities, baseline functional status, social support, and access to prosthetics, physical therapy, and other rehabilitation services. From the surgeon’s perspective, efforts to provide a well-padded, pain-free stump in collaboration with prosthetists, physical therapists, and other rehabilitation providers are crucial to optimizing long-term outcomes. Recent studies have found that prosthetic fitting after amputation is associated with improved survival after 3 years ¹⁶¹ and that failure to achieve greater ambulation and mobility through successful prosthetic fitting in these patients is linked to reduced survival in a manner that is not explained by other patient characteristics and comorbidities. ¹⁶²

For BKAs, technical considerations include maintaining a minimum clearance of eight inches from the ground to fit a prosthesis and its artificial ankle, with ideal measurements ranging from 12 to 18 cm from the joint line depending on the patient’s anatomy. A posterior flap design may be used, which requires an additional 8 to 10 cm of soft tissue distal to the bone cut for the posterior flap. In terms of blood supply, a patent popliteal and tibial peroneal trunk are preferred due to their association with higher healing rates and lower complication rates, although successful outcomes were found with a closed popliteal artery if the sural arteries remain open. ¹⁶³ At one institution, it was found that 60% of those patients were ambulatory at 9.5 months, and the conversion rate of BKA to AKA was 4%. ¹⁶³

Pain is a major factor in limiting ambulation and the level of functional activity after amputation. Notably, residual limb pain after MLEA affects 50% to 80% of patients, with 20% to 50% requiring narcotics, and it has been noted that 14.6% of patients develop symptomatic neuroma, primarily affecting the superficial peroneal and saphenous nerves; however, diabetes and obesity were protective against symptomatic neuroma formation. ¹⁶⁴ Preventing neuroma formation is a key strategy to avoid pain, and this has been facilitated by the development of targeted muscle reinnervation (TMR) where a cut sensory nerve such as the superficial peroneal nerve, posterior tibial nerve, or saphenous nerve is connected to a motor nerve branch entering a muscle so that the cut sensory nerve then grows down into the muscle itself and dissipates, preventing formation of a neuroma. This technique has shown a significant reduction in overall pain (41.5% vs. 67.2%; p = 0.010). ¹⁶⁵

In addition, myodesis where the peroneal, anterior, and deep posterior muscles are attached directly to the residual tibia can help prevent muscle wasting and atrophy. This surgical approach also helps mitigate the instability that can make it more difficult for the patient to wear a prosthesis and ultimately ambulate. ¹⁶⁶ The posterior flap consisting of the superficial posterior compartment muscle and overlying skin is inset by tenodesing the soleus, gastrocnemius muscles, and Achilles tendon to the anterior tibia. ¹⁶⁷ Furthermore, shortening the fibula, beveling the tibia, and getting rid of dogears along the incision prepare the limb for immediate postoperative prosthetic fitting. This requires a multidisciplinary collaboration among the surgeon, the prosthetist, and the physical therapist. In cases where infection is present, a two-stage procedure can be used with initial drainage and amputation followed by compression therapy of lymphedema to optimize the leg for amputation and minimize complications. ¹⁶⁸

For the athletic and other active patients with few comorbidities, the Ertl technique, which fuses the tibia and fibula distally using a free or vascularized bone graft to transfer torque directly to the artificial ankle when running, can be utilized to prevent the fibula from moving past the tibia.^167,169,170 It was found that patients who underwent this surgery began using prostheses sooner (2.5 vs. 3.5 months; p = 0.008) and began ambulating sooner (2.3 vs. 3.7 months; p = 0.001) than their non-Ertl counterparts, with no differences in postoperative complication rates. ¹⁷¹ Furthermore, the use of the Ertl technique with TMR has been found to result in 92% rate of patients ambulating compared with just 71% in patients who had neither technique performed, with 70% of patients experiencing no pain. ¹⁷¹

Optimizing the choice of limb salvage options is crucial to maximize the patient’s chances of ambulation after amputation. Knowledge of the patient’s preoperative medical and surgical status is a key component for success. In the context of the traditional paradigm of LEAs, advances in surgical technique, TMR, Ertl procedure, and prosthetics have evolved BKAs into a functionally well-regarded option. Collaborating with a true multidisciplinary team enables the provision of all necessary resources to heal the wound and prevent recurrence of the wound or a more proximal amputation, allowing a focus to be placed on achieving and optimizing function rather than simply anatomical limb salvage for the sake of limb salvage.

Key Consensus Point:

Diabetic limb salvage should focus on achieving and optimizing function rather than optimizing tissue preservation, requiring a multidisciplinary approach and collaborative care with prosthetists and physical therapists.

Biomechanical surgery

Surgical treatment includes wound bed preparation, direct closure methods including flaps or grafts, and biomechanical surgery. Biomechanical surgery takes into account the long-term functional durability of the diabetic foot and may be appropriate when other measures, including orthotic devices, are not effective. Altered biomechanics and foot deformity often precede the development or chronicity of a DFU.^179,180 Repetitive or acute trauma in an insensate or vascularly compromised foot leads to ulcer formation that can lead to infection and/or limb loss. The foot is a high-demand end organ specifically designed to be a mobile adaptor as well as a rigid construct depending on the phase of gait. The patient with diabetes is at high risk due to underlying changes in nerves, arteries, muscles, ligaments, tendon, and bone. The triad that leads to limb loss is peripheral neuropathy, peripheral vascular disease, and bone deformity in the environment of a compromised host. Thus, the surgical approach to DFU management must include a working knowledge of the implications of foot biomechanics.

Procedure selection is determined by the patient’s functional capacity and the plane of deformity. The patient’s functional capacity has a broad range and needs to be determined to meet the ambulatory demands of the patient. The activity level of a patient with diabetes may be limited to a wheelchair or working daily with frequent and prolonged standing or walking. In the case of a patient resigned to a wheelchair, typically the ulcer location may be where the foot platform of the wheelchair is in constant contact with the foot (e.g., hindfoot). Whereas the ambulatory patient wound location is typically in the midfoot or forefoot. Surgery selection must take into account these factors. The plane of deformity (i.e., sagittal, coronal, transverse) is identified though a biomechanical examination and weightbearing radiographs. The biomechanical examination includes identifying restricted range of motion of joints, which can cause uneven distribution of pressures experienced in the foot.^181,182 The weightbearing radiographs will identify abnormal bone relationships and articulations. The goal of biomechanical surgery is to better balance the foot on the weightbearing surface to redistribute pressure more evenly.

Biomechanical surgical correction of the diabetic foot begins with discriminating between soft tissue and bone limitations, or the combination of both. A tendon or capsule procedure cannot correct a deformity in which the bone or joint is impacted due to their immobility through processes such as end-stage osteoarthritis. If the major deforming force is tendon driven, then tendon release or lengthening is indicated. In patients with long-standing diabetes and poor blood glucose control, there is atrophy of the intrinsic muscles of the foot as well as structural changes to the tendon and capsules.^183,184 This tendon substance thickens and loses its elastic properties leading to joint contracture. ¹⁸⁴ This translates to excess forces experienced in isolated areas of the plantar foot rather than being broadly distributed.^182,185

Tendon procedures include flexor or extensor tenotomy, Achilles tendon lengthening (TAL), tendon transfer of the tibialis anterior or peroneal tendons. Bone reconstruction includes exostectomy (bony prominence removal), arthroplasty (joint resection) and arthrodesis/osteotomy (joint realignment and fusion). A few examples of bone reconstruction include hammertoe arthroplasty, hallux abductor valgus reduction, pes planus correction, and ankle fusion. These types of procedures include the use of internal or external hardware with sometimes prolonged recovery. An extreme example of a bone procedure is for diabetic Charcot neuropathy where there is significant bone destruction, which necessitates the use of multiple robust hardware constructs across multiple joints. Again, the goal for soft tissue and bone procedures is to redistribute pressure experienced on the plantar aspect of the foot. In general, soft tissue corrections are less complex surgeries with quicker recovery.

There is limited peer-reviewed evidence for biomechanical surgery of the diabetic foot. Largely the evidence is relegated to small case series representing a single surgeon or group experience. The most robust evidence for the success of a tendon procedure is for the TAL. Muller et al. published a prospective randomized clinical study comparing the TAL plus total contact versus a total contact alone for the treatment of recurrent diabetic forefoot ulcers. ¹⁸² They report a 38% recurrence rate at 2 years for the TAL group compared with an 81% recurrence rate for the other group. Andersen et al. in a multicenter, randomized controlled trial in diabetic digital preulcers and ulcers report 100% ulcer healing and higher ulcer prevention (1 vs. 7) in the group assigned to flexor tendon tenotomies. ¹⁸⁶ A meta-analysis performed by Yammine et al. reports the results of metatarsal osteotomies for improving the off-loading of DFUs of 119 patients with a mean follow-up of 10.9 months. ¹⁸⁷ This meta-analysis only included four studies with a reported healing rate of 98.7%. A systematic review conducted by Ha et al. compiled studies of 1,089 patients with diabetic Charcot neuropathy reconstruction and report no benefit of one technique over another and also report a 36% complication rate and a 5.5% amputation rate. ¹⁸⁸ As mentioned above, there is a lack of robust comparative studies examining the outcomes of biomechanical surgery; results are largely driven by surgeon experience.

Partial foot amputation is a viable surgical treatment option. Again, the goal for diabetic foot surgery is to preserve a functional and durable foot. A properly performed partial foot amputation may be the best option in the right patient. Amputation-level selection is often guided by the extent of soft tissue and bone loss due to infection or ischemia. However, the surgeon must also consider the biomechanical implications of the amputation level selected. As the amputation level progresses proximally (toe, ray, tarsometatarsal, midfoot, hindfoot), the lever arm of the foot reduces. Furthermore, the extensor mechanism responsible for dorsiflexion is lost, thereby increasing flexor dominance. The propulsive ability decreases with more proximal foot amputations. There may also be greater plantar pressures experienced at the distal plantar stump potentially leading to reulceration and reamputation. This initiates the cascade of repeat surgery with more of the foot requiring amputation. Isolated digital or ray amputations have repercussion as this leads to redistribution of pressure to other plantar foot locations, which again will make these areas prone to ulceration and reamputation of 34% (digital) and 10% for ray amputations. ⁵² After a trans-metatarsal amputation, the reamputation rate is reported to be approximately 30%. ¹⁸⁹ The goal of partial foot amputation is to maintain balance to the pressures experienced to the plantar aspect of the remaining foot.

The surgical approach to the DFU includes a thoughtful approach to the biomechanical implications. Closure or coverage of the DFU is not enough. Alterations in the underlying biomechanics of the diabetic foot must be addressed for long-term success. Unlike follow-up of other type of surgical procedures, long-term vigilance and close monitoring are absolutely necessary. This includes the use and exchange of appropriate shoes, orthotics, braces, and prosthetics as well as close collaboration with a multidisciplinary team including physical therapists to optimize long-term functional outcomes.

Wound bed preparation

Wound bed preparation is a fundamental principle to advance wound closure or coverage.^190,191 This includes optimizing the macroenvironment by enhancing host factors such as tightening glucose control and optimizing nutritional support.^192,193 Furthermore, more local factors such as maximizing perfusion through vascular interventions to the targeted angiosome where the wound is located are preferred. ¹⁹⁴ Other local factors include eliminating infection and decreasing colonized bacterial load in the wound through appropriate and adequate wound hygiene and topical antimicrobials. Periwound tissue stabilization, as well as joint immobilization, is also important through compression wraps, splinting, casting, orthotics, and other off-loading devices, and fixation using internal and external hardware.

Proper wound bed preparation involves the removal of all nonviable tissue, including necrotic, fibrotic, liquified, and indurated tissue. After sufficient wound bed preparation is performed, the remaining tissue should be bleeding, pliable, and without malodor. ¹⁹⁰ The wound perimeter should also be excised because bacteria can harbor in these areas, and in chronic DFUs, the wound edge is often rolled (epibole), or callused, which inhibits epithelization across the wound surface if secondary healing is the goal. ¹⁹⁵

The surgical approach to wound bed preparation includes the use of sharp instrumentation, including scalpel, scissors, rongeur, curette, and/or hydro-surgical scalpel. Surgical-based excisional debridement is different from nonoperating room-based sharp debridement. In the operating room setting, the surgeon can be more precise, more aggressive with the ability to control pain through anesthesia, and control bleeding through multiple methods. Furthermore, particularly in DFUs, there is always a concern for bacteria. DFUs contain not only planktonic bacteria but also biofilm. ¹⁹⁶ Thus, full surgical excision of the wound is necessary with the use of adjuncts such as antiseptics and preservatives. The mechanisms of action of antiseptics and preservatives are different than the mechanism of action of antibiotics, and thus can impact biofilm. ¹⁹⁷ There are multiple methods of delivering antiseptics and preservatives to the wound bed, including wet-moist dressing changes in preparation for wound closure or coverage, as an irrigation solution during the operation, or as the instillation solution used in conjunction with NPWT with instillation (NPWTi).

Serial or staged excisional debridement may be necessary. This allows for demarcation of infection and nonviable tissue as well as an opportunity to optimize the blood flow and the patient. A patient with a DFU frequently presents with a deep-seated abscess. Thus, the goal of the initial excisional debridement decompresses the infection and reduces the overall bacterial bioburden before definitive closure or coverage.^52,198 The use of NPWT can be an effective method to bridge between operations. The goal of NPWT is to remove exudate, promote granulation tissue, increase local perfusion, and to decrease bacterial contamination.^199,200 Armstrong et al. report that in a prospective, multicenter, randomized clinical trial, the use of NPWT in patients hospitalized for a diabetic partial foot amputation results in higher rates of complete healing. ²⁰¹ NPWTi is a modification of traditional NPWT, which has been reported to be superior in the acute-care setting. NPWTi combines the benefits of traditional NPWT with continuous dwelling of topical solution. A meta-analysis performed by Gabriel et al. report that NPWTi results in a decreased number of operations, faster time to closure or coverage, decreased bacterial counts, and overall decreased length of therapy compared with other wound care modalities, including traditional NPWT. ²⁰² Further modifications of NPWTi include a large perforated foam design, which has been reported to detach nonviable tissue on the wound surface, essentially hydromechanically debriding the surface of the wound. ²⁰³ This innovation can potentially expedite wound bed preparedness for wound closure or coverage.

Wound bed preparation precedes any next step in wound management, whether the goal is wound closure, coverage, or secondary healing. There are clear benefits to surgical excisional debridement. However, factors including surgeon or operating room availability as well as the patient’s medical stability and consent may limit this approach. Furthermore, while surgical treatment for DFU is well accepted, the literature supporting common clinical care is not at the highest levels of evidence. Reasons for this are complex and include the difficulty in blinding surgical procedures. In randomized clinical studies, provider and patient bias toward active treatment arms can be substantial. Nevertheless, the best evidence we have is based on randomized controlled clinical trials and meta-analysis and what follows is a review of the current high-level evidence for each of these surgical interventions.

Free tissue transfer

Free tissue transfer (FTT) takes a segment of tissue along with its blood supply and moves this to a debrided ulcer site. In addition to suturing the tissue in place, microvascular anastomosis is a technique used with at least an artery and a vein. Sometimes structures such as a sensory nerve are included. These are highly technical procedures in patients who have substantial comorbid disease. These operations can take several hours to perform and can have a 5% to 10% total failure rate in lower extremity wound cases. As such, these are most often performed in specialized centers with the surgical expertise and the postoperative ability to closely monitor these flaps. Bhat et al. recently performed a meta-analysis which suggests that for carefully selected cases, this technique can be valuable, albeit with a substantial risk profile. ²⁰⁴ Kim et al. ²⁰⁵ recently performed a registry study of 21 free flaps that were compared with patients treated with a skin substitute product, NPWT, or local flaps. They found comparable outcomes using all of these methods, but presumably the FTT patients had more severe disease. ²⁰⁵ In conclusion, FTT appears to be a reasonable option in a specialized center, although the level of evidence to support this is low.

Local flaps and skin grafts

Local flaps and skin grafts are less technically demanding than FTT but are generally used in smaller areas. In DFU, the clinician needs to make sure there is adequate perfusion to the flap in the setting of a high percentage of patients with vascular disease. Although the amount of tissue available to use as a flap in the foot and ankle is limited, when these are used in conjunction with a skin graft, critical structures such as bone, joint, tendon, and large neurovascular pedicles are optimally closed with a local flap, while the rest of the wound can be treated with a skin graft. This is nicely reviewed by Clemens et al. ²⁰⁶ Santema et al. ²⁰⁷ performed a 2016 Cochrane review on skin grafting and tissue replacements for the treatment of DFU. They found that these treatments slightly lowered the chance of amputation compared with the SOC. The data available were insufficient to draw conclusions of the effectiveness of different types of skin grafts or tissue replacement therapies. ²⁰⁷

Products to treat DFUs

There are a large number of skin substitutes available in the market today for clinicians to use. Because many of these require little clinical data before going to market, the number of studies justifying their effectiveness is small. In most cases, the best studies available compare a specific construct to the SOC, which is most often some sort of moist wound healing dressing along with off-loading of the foot. What follows is a review of the highest level of studies for skin substitutes available in the literature, which has also been reviewed by the Centers for Medicare and Medicaid Services. ²⁰⁸

Placental constructs

Mohammed et al. ²⁰⁹ performed a systematic review and meta-analysis for patients with DFUs treated with dehydrated human amnion and chorion allograft (DHACA) compared with SOC. They pooled 11 randomized clinical trials of 655 patients with DFUs, where 328 patients were treated with DHACA and compared this with 327 patients treated with SOC alone. They concluded that DHACA, in addition to SOC enhanced wound healing, reduced the mean time to heal, and diminished adverse events.

Skin substitutes

Santema et al. performed a 2016 systematic review and meta-analysis of a variety of skin substitutes, where they looked at 17 trials with 1,655 randomized patients. ²⁰⁷ One of the small trials included an amniotic product, but there were a wide variety of other products, including processed allogeneic and xenogeneic scaffolds and bioengineered skin substitutes. They concluded that in addition to SOC, skin substitutes increase the likelihood of achieving complete wound closure compared with SOC alone. The noted a lack of long-term and cost-effectiveness data. Tchero et al. performed a systematic review and network meta-analysis of five regeneration matrices (Integra, Nevelia, Matriderm, Pelnac, and Renoskin), where they looked at 13 studies. ²¹⁰ They concluded that these products have low failure rates and low ulcer recurrence rates. They called for more studies looking at safety, efficacy, and failure rates.

Regulatory review and studies needed to justify reimbursement

Many advanced products that are novel require complex approval processes from the Food and Drug Administration (FDA), including class 3 medical devices, drug approval, or biological approval. These studies generally require large relatively homogeneous patient populations with carefully designed inclusion and exclusion criteria. Because of the expense of these products, there often is a run-in period where the control dressing is used for some time to make sure that the wound neither greatly expands nor quickly heals. Studies tend to last 6 to 12 weeks and complete healing is very often the primary outcome.

Studies for reimbursement, generally require fewer patients, but need to be done in a comparative manner. To avoid the complex FDA approval studies, many have focused on simpler mechanisms to clear the FDA, including the 510(k) process for medical devices (substantially similar to a predicate product) and 21 CFR 1271 for tissues and biologics. An orphan drug status is a simplified application for drug approval. The 510(k) process and 21 CFR 1271 are much simpler than a typical approval process as little clinical data need to be applied. Clinical data are often required for insurance approval, but this can often be done with a simpler study. As a result, most new products used in wound healing today go through the 510(k) or 21 CFR 1271 process.

DFUs are complex with a heterogenicity of wound sizes, locations, and patient comorbidities. Both surgical technologies and new products to treat these wounds bring new hope to these patients and in a large range of studies have good healing rates that presumably reduce amputations. Because of the heterogenicity of the ulcers and patients involved, doing good studies is challenging. Randomized controlled trials have been regarded as the best method to study these patients, but give only a limited view on how an actual patient is treated in our medical systems. These trials typically have a complex set of inclusion and exclusion criteria and only a fraction of patients with DFUs are included in a trial. Some have advocated for registry studies to further elucidate the effectiveness of these products and surgical techniques. Comparative effectiveness studies would be very helpful but would require large numbers of patients. Because of the expense of these studies, most are funded by industry, and they are often designed to show the safety and efficacy of a specific product. Independent funding and asking more hypothesis-based questions could potentially help move new therapies forward.

Emerging treatments and hyperbaric oxygen therapy

The roles of emerging modalities including adipose-derived stem cells (ASCs) in the treatment of diabetic wounds are still being investigated and currently not approved or recommended for general use in the management of diabetic wounds. Clinical data for the use of ASCs in DFUs have been limited to a handful of studies done outside of the United States.^211–215 A pilot study performed in 2010 in South Korea ²¹¹ relied on topical application as did a more recent study from Poland, ²¹² while other recent studies in Nicaragua, Turkey, and Italy used injections of ASCs directly into the wound bed.^213–215 While these studies had variability in patient selection, cell dose, and the methods by which ASCs were isolated and applied, they generally document the safety of the technique and provide data suggesting that healing rates can be improved using ASCs compared with the current SOC. Other modalities being investigated for use to treat diabetic wounds include miR146a, a regulatory microRNA that targets upstream regulators of inflammatory and oxidative stress pathways, which has been found to be suppressed in diabetic mouse wounds.^216,217 A form conjugated to cerium oxide nanoparticles is currently under development with clinical studies in humans planned.^218,219

TOT has been used to enhance the healing of DFUs as an adjunct therapy following treatment with hyperbaric oxygen. Recent evidence for the use of TOT in the treatment of DFUs was reviewed in a meta-analysis by Carter et al. ²²⁰ Their systematic review identified four randomized controlled trials published from 2017 through 2021.^221–224 The authors concluded that these studies provided a moderate level of evidence to support the use of TOT to treat noninfected, nonischemic DFUs of at least 4-week duration that have not adequately responded to SOC measures such as optimal off-loading and wound care. Further studies are needed to clarify the indications for its use and efficacy in other wound types.

The use of hyperbaric oxygen therapy (HBOT) for the treatment of chronic wounds was recently reviewed by a previous consensus panel. ¹ As noted by the consensus panel with regard to diabetic wounds:

“In a meta-analysis of the relevant literature, two separate Cochrane reviews conducted 10 years apart came to significantly different conclusions regarding the usefulness of HBOT in DFUs. The 2004 meta-analysis concluded that HBOT decreased the incidence of major amputations and improved healing at 1 year. ²²⁵ However, an updated 2015 metanalysis contradicted this conclusion and found that HBOT improved healing in the short term (6 weeks), but not at 1 year. ²²⁶ Since this last Cochrane review, numerous new studies have been conducted, including both randomized controlled studies and newer ‘big data’ studies based on large data bases derived from the electronic medical record. These newer techniques again yielded conflicting results with one study showing no impact on amputation outcomes ²²⁷ while a later study on the same data base demonstrated efficacy in a subset of more severe DFUs (Wagner 3 and 4) in terms of wound closure.^228,229 Clearly, how the analysis is performed and how the questions are asked can skew results in different directions. Unfortunately, since none of these studies is perfect or definitive, this allows the controversy to continue ad infinitum. However, when considering the totality of the existing data, among the potential indications for HBOT in the field of chronic wounds, the strongest evidence exists for ischemic, infected DFUs, that is, Wagner Grade 3 or worse. ²³⁰ ”

Key Consensus Points:

Surgical treatment includes a thoughtful approach to the biomechanical implications where closure or coverage of the DFU is not enough. The alterations in the underlying biomechanics of the diabetic foot must be addressed for long-term success where long-term vigilance and close monitoring are absolutely necessary. This includes the use and exchange of appropriate shoes, orthotics, braces, and prosthetics, as well as close collaboration with a multidisciplinary team.