Surgical strategies for prevention of amputation of the diabetic foot

Abstract

Prevention of amputation has become a key objective of clinicians providing care to patients with high-risk diabetic foot problems. In this regard, the multidisciplinary diabetic foot team (MDFT) has been embraced as the most effective way to manage patients with foot ulcers, infections, and Charcot feet. Importantly, such specialized teams have also integrated various surgical specialties to enable more expedient management of these often complex conditions. Experienced diabetic foot surgeons over the last three or four decades have contributed much to this discipline, whereby foot-sparing reconstructive procedures or minor amputations have become fundamental strategies for limb preservation teams. Central to limb salvage, of course, is the recognition of underlying vascular insufficiency and the importance of prompt (endo)vascular intervention. Restoration of adequate perfusion is essential to allow the podiatric, orthopaedic, or plastic surgeon to perform indicated functional reconstructive or minor amputation procedures. This evidence-based overview discusses the various indications and surgical principles inherent in modern concepts aimed at preventing amputation in the high-risk diabetic foot.

1. Introduction

One of the great concerns of persons with diabetes mellitus is that of lower limb amputation. Indeed, many diabetic patients have longstanding memories of relatives having undergone amputation or, very commonly, multiple amputations prior to their ultimate death from the disease. In fact, one recent study found that diabetic persons feared major amputation more than death, blindness, and end stage renal disease.¹ Fortunately, our current understanding of the underlying pathophysiology and associated preventive interventions for diabetic foot complications has lead to improvements in major amputation rates in recent years.^{2, 3, 4, 5} While there has been a gradual increase in the number of minor amputations in this population, there has been a commensurate decrease in the number of required major amputations.⁵ Of particular importance is our understanding of the role that peripheral arterial disease (PAD) plays not only in the etiology of foot lesions, but also in management strategies that focus on revascularisation.^{6, 7, 8} Recent multidisciplinary guidelines have therefore highlighted the importance of early diagnosis of PAD in order to re-establish adequate perfusion to the ischemic ulcerated and/or infected foot.^{9, 10, 11} Making this a priority will not only provide for salvage of the at-risk limb, but also to allow for foot sparing or reconstructive surgical procedures.¹²

Diabetic foot ulcers (DFU) remain the most pathognomonic of all diabetic foot complications. Their etiology, pathophysiology, and general approaches to management have been extensively reviewed in recent multidisciplinary guidelines.^6,9,10 There are several validated DFU classification systems.^{13, 14, 15} However, the University of Texas system has been widely adopted for clinical use and is presented in Table 1.¹³

Table 1.

University of Texas diabetic foot ulcer classification system.

Stage		Grade
	0	I	II	III
A	Pre- or postulcerative lesion completely epithelialized	Superficial wound, not involving tendon, capsule, or bone	Wound penetrating to tendon or capsule	Wound penetrating to bone or joint
B	Infected	Infected	Infected	Infected
C	Ischemic	Ischemic	Ischemic	Ischemic
D	Infected and Ischemic	Infected and Ischemic	Infected and Ischemic	Infected and Ischemic

Surgical management for diabetic foot disorders and infections has increasingly become a recognized, if not key, component of care over the last several decades.^16,17 Experienced diabetic foot surgeons over the last three or four decades have contributed much to this discipline, whereby foot-sparing reconstructive procedures or minor amputations have become fundamental strategies for limb preservation teams. It is therefore surprising that several recent diabetic foot guidelines have not specifically addressed the important role of reconstructive surgery for patients suffering from diabetic foot ulcers (DFU) or other complications of the diabetic foot syndrome.^{9, 10, 11}

Several risk-based classes of diabetic foot surgery have been described and are detailed in Table 2.^17,18 Generally performed in the non-ischemic patient, each class of surgery is determined by the presence or absence of a wound, neuropathic status, or the acuity of the wound (i.e. chronic vs. acute necrotizing infection). Many of the procedures commonly used in one category are not mutually exclusive, meaning that a given procedure might be performed to prevent an ulcer or, when present, to effectuate a resolution (“cure”) of the ulcer. For example, a simple hammertoe repair by flexor tenotomy might be used to treat an idiopathic deformity in a sensate diabetic person (Elective, Class I) while also being performed to prevent recurrence of a healed neuropathic ulcer (Prophylactic, Class II). Similarly, a midfoot reconstruction for a Charcot foot deformity is often performed for patients without an active ulcer (Prophylactic, Class II) or in such patients with recalcitrant or recurrent plantar ulcers (Curative, Class III).

Table 2.

Classification of diabetic foot surgery.

Class I	Elective	Reconstructive procedures on patients without neuropathy Examples: Hammertoe, bunion, osteotomy, Achilles lengthening (TAL), etc.
Class II	Prophylactic	Reconstructive procedures performed to reduce the risk of ulceration or re-ulceration in neuropathic patients who do not have a wound present Examples: Keller arthroplasty, TAL, Exostectomy, Charcot reconstruction, etc.
Class III	Curative	Procedures performed to assist in healing of open wounds Examples: Metatarsal head resection, Keller arthroplasty, toe amputation, etc.
Class IV	Emergent	Procedures performed to arrest or limit progression of infection Examples: Incision & drainage, Guillotine/open amputation, fasciotomy, etc.

Emergent procedures (Class IV) are performed to control the progression of severe or ascending infections (wet gangrene, necrotizing fasciitis, etc.). This category of foot (or leg) surgery frequently requires initial open amputation or fasciotomies, followed by definitive reconstructive or wound coverage procedures. Obviously, procedures performed in the presence of open wounds are at higher risk for non-healing or infection than are those performed in patients without wounds. Successful surgical outcomes are therefore predicated on thorough assessment of both the patient and the specific foot complication under consideration.

It is recognized that diabetic foot surgical care in a region is best provided as a ‘hub and spoke’ model, where the hub has a multidisciplinary diabetic foot team (MDFT) that includes a physician (endocrinologist), podiatrist, wound care specialist, orthotist, orthopaedic/podiatric surgeon, vascular surgeon, plastic surgeon, and a microbiologist. All surgical competencies needed for functional limb salvage of complex diabetic foot presentations should be offered at the MDFT. Minor diabetic foot presentations can be managed by a small diabetic foot team at the spoke unit, but complex and limb threatening conditions such as critical limb ischemia or infected Charcot deformity are best referred to the MDFT. Diabetic foot surgical management often includes correction of bone, soft tissue and vascular components.

Over recent decades there has been a progressive inclination of diabetic foot specialists, MDFTs, and society guidelines to promote limb salvage procedures as alternatives to major limb amputations.^2,3,6,10 While there are no formal direct comparative studies, cross-sectional studies consistently indicate the trend for such teams to reduce the necessity for major amputation by aggressive vascular intervention coupled with early surgical management for deep, limb threatening infections.^3,8,19 Nonetheless, there are times when major amputation can be a preferred option to lengthy (or failed) limb salvage attempts. The authors have had numerous patients who had struggled with unsuccessfully treated chronic/recurrent wounds for many months or years only to be relieved upon undergoing definitive major amputation. Of course, such decisions must be made after thoughtful discussions with the patient, family, and treating clinicians. This article therefore covers the spectrum of typical presentations of increasing complexity and their treatment by the modern diabetic foot team.

2. Osteomyelitis and Charcot foot

Osteomyelitis and deformity together contribute to a significant risk of limb loss in a neuropathic diabetic foot. The deformity can result from muscle imbalance and musculotendinous units’ contractures, tissue loss and contracture due to infection, bone, and joint affections from Charcot neuroarthropathy (CN), or a combination of these. The surgical ladder for prevention of a major amputation in an infected or Charcot foot can be described for the following presentations:

• 1.Acute Infection: Diabetic foot attack surgery
• 2.Osteomyelitis surgery
• 3.Acute Charcot foot reconstruction
• 4.Charcot deformity correction
• 4aNon-infected Charcot deformity correction
• 4bInfected Charcot deformity correction

2.1. Diabetic foot attack surgery

Diabetic foot infection can occasionally progress rapidly, spreading along the tissue planes to the adjacent tissues causing necrosis and generating a systemic response.²⁰ Such presentation is labelled as a diabetic foot attack and is potentially limb and life threatening without emergent intervention. Diabetic foot attack presentation warrants rapid diagnosis with appropriate clinical assessment and urgent investigations. Harvesting deep tissue samples for microbiological studies prior to the commencement of empirical systemic antibiotics helps provide targeted antibiotic treatment once the culture sensitivities become available. The definitive treatment is, however, aggressive surgical debridement performed emergently. The debridement often necessitates the need for large surgical incisions to reach and excise the necrotic tissue completely to the healthy normal tissue. The usage of drains/loops/strings to connect infected regions is not recommended; always connect with an incision as there probably is necrotic and infected tissue underneath. Negative pressure wound therapy (NWPT) and or a plastic surgical procedure can help to achieve soft tissue coverage for the open wound. The role for aggressive and expedient surgical intervention for deep infection was highlighted by Faglia et al. who found that those patients with delayed referral or surgical intervention fared worse than those who received urgent surgical debridement and revascularisation.²¹ Higher amputation levels and longer lengths of hospital stay were associated with the group of patients in whom surgical care was delayed.

2.2. Osteomyelitis surgery

Diabetic foot osteomyelitis (DFO) as compared to osteomyelitis of long bones lacks a clear definition as well as a classification system as does osteomyelitis of long bones.²² Reports on drug treatment as an alternative to surgery have to be interpreted with caution when they are confined to infections of phalanges and metatarsal bones. Lazaro-Martinez published no differences between conservative and surgical treatment, but the surgery group included more severe cases with metatarsal instead of phalangeal involvement.²³

Surgical treatment of DFO must consider the quality of the soft tissue envelope. Infected or necrotic skin may interfere with primary wound closure. Open wound treatment and/or NPWT facilitate delayed closure and help to preserve foot length in case of partial foot amputations. Dead bone (sequestrum or involucrum) usually calls for surgical intervention since uptake of antibiotics to that area is limited. Penetration of the infection into an adjacent joint is another reason to prefer surgery over antibiotic treatment. The aim of surgery is to resect as much as needed albeit as little as possible of viable tissue adjacent to the necrotic tissue.

Curettage is the least invasive procedure, sometimes followed by application of local antibiotic carriers which also could shorten the duration of postoperative antibiotic treatment.^24,25 Resection of single bones must take into account possible biomechanical consequences. Load transfer, skeletal collapse and progressive foot deformity may happen after those procedures. Internal pedal amputations are more radical, but may help to obtain a plantigrade and shoeable foot where polyneuropathy allows for pain-free ambulation.²⁶

The last step on this surgical ladder of osteomyelitis surgery before it comes to major limb loss are partial foot amputations. They also can be performed as staged procedures to preserve skeletal architecture. One crucial part of the operative procedure is tendon balancing to prevent postoperative foot deformity which in turn could lead to pressure ulcers.²⁷ All in all, there are a variety of conservative surgical techniques to prevent limb loss in case of DFO.

2.3. Acute Charcot foot reconstruction

CN can be staged using the Eichenholtz system, based on changes noted on plain radiographs, into stage 1- bone fragmentation, 2- coalescence and 3- consolidation. Stage 0, a pre-radiographic stage where the abnormality is noted only on MRI or isotope scanning, was later added to the classification. CN of foot and ankle is classified based on the anatomical location of bone involvement, and the two classifications commonly used are by Brodsky and Saunders & Frykberg. The mainstay of treatment of acute CN is immediate offloading in a total contact cast or brace and protected weight-bearing that is continued until bone consolidation is achieved.

Some acute CN presentations develop progressive bone fragmentation and foot deformation despite optimal offloading and carry a high risk of developing skin breakdown, infection, and potential limb loss without timely intervention. Functional limb salvage of such presentation warrants early surgical stabilisation during the bone fragmentation phase of the disease. The procedure is considered once the foot swelling and redness resolve after a period of strict leg elevation and offloading in a total contact cast. Temporary stabilisation of the foot and ankle in an external fixation device can also provide excellent offloading. The definitive reconstruction can be achieved by using external or internal fixation methods (Fig. 1a–d). It is critical that the patient’s care is delivered by a multidisciplinary diabetic foot team. There is paucity of literature on this subject and it overall shows satisfactory limb salvage rates, but variable functional outcomes. Simon et al. in their series of 14 patients in Eichenholtz stage 1 managed with extensive debridement, and open reduction and internal fixation with autologous bone graft reported good outcomes at a mean follow up of 4 months.²⁸ Caravaggi et al., in 2012, reported on a cohort of 45 patients presenting with early chronic stage CN with deformity and no deep ulceration. They achieved an 86% limb salvage rate and no ulcer recurrence.²⁹

Fig. 1.

1a, 1b. Dorsoplantar and lateral views of foot showing bone changes due to Acute CN triggered by a minor twisting injury. Fig. 1c, d: dorsoplantar and lateral views of foot taken following Internal fixation of the medial and lateral columns.

2.4. Charcot deformity correction

CN is a devastating complication of diabetic peripheral neuropathy and is known to decrease the life expectance by 14 years.³⁰ The common surgical interventions in patients with CN include ulcer debridement, exostectomy, and deformity correction.

Charcot deformity correction requires a dedicated MDFT for perioperative assessment and delivery of care. It is recognized that the prevalence of PAD in Charcot feet is high ³¹ and a detailed vascular assessment and, if necessary, revascularisation should be performed prior to the deformity correction.

2.4.1. Non-infected Charcot deformity correction

Presence of a non-infected ulcer is not usually a contra-indication for surgical reconstruction. The deformity correction is achieved by wedge or rhomboid bone resections on the convex side, along with soft tissue balancing through release/lengthening of contracted tendons. The skeletal stabilisation is achieved by using an internal fixation or external fixation method. The established principle of internal fixation is ’long-segment and rigid fixation with optimal bone opposition’. (Fig. 2a and b). A recently published systematic review reported a fusion rate of 81% with the internal fixation technique.³² One recent study has reported a hardware failure rate of 24% with internal fixation at a mean follow up of 31.8 months, but the functional outcomes in this cohort were satisfactory.³³

Fig. 2.

2a: Weight bearing AP and lateral view of foot and ankle showing CN changes involving the midfoot and hindfoot regions with marked deformity. This patient presented with additional instability but no infection. Fig. 2b: Weight bearing AP and lateral view of foot and ankle taken showing hindfoot and midfoot reconstruction using long segment and rigid internal fixation technique with optimal bone. The medial column locking plate has been extended to distal tibial to achieve additional rotational rigidity of the hindfoot nail construct.

2.4.2. Infected Charcot deformity correction

An infected and deformed Charcot foot is one of the most challenging clinical presentations in patients with diabetic neuropathy and is known carry a high rate of limb loss.³⁴ Reconstruction of actively infected Charcot foot carries a risk of infection recurrence and failure of bone union and therefore is often performed as a staged procedure.^33,35 Microbiological cultures of infected bone and soft tissue is critical for successful outcomes. The first stage consists of radical debridement of the ulcer and infected bone along with temporary fixation of foot and ankle using threaded wires if there any instability due to bone resections, and application of local antibiotic eluting preparation to eradicate any residual infection. Culture and sensitivity specific intravenous antibiotics are continued until clinical and serological evidence of infection clearance is noted. Any associated open wounds are managed with NPWT. Definitive reconstruction using an external fixator ³⁶ or internal fixation devices is performed after achieving infection eradication (Fig. 3 a-f).³⁷ The clinical outcomes of infected Charcot deformity reconstruction using this protocol is similar to one stage Charcot foot corrections.³³

Fig. 3.

3a: Charcot foot with osteomyelitis and plantar ulcer. Fig. 3b: Dorsoplantar X-rays at first presentation. Fig. 3c: Lateral X-ray. Fig. 3d: after internal tarsal resection and external fixation. Fig. 3e: final outcome 21 months postoperatively. f) Clinical image after removal of the frame.

3. Peripheral arterial disease

In the treatment of a diabetic foot ulcer with PAD one must first understand the specifics of the disease. This helps in choosing and interpreting the diagnostic tests and making a customised treatment plan with your multidisciplinary team and the patient. PAD in diabetes mellitus (DM) is somewhat different compared to the general population; it is more common, starts earlier, is multi-segmental and bilateral, more distal and shows more medial calcification. Furthermore, there is impaired formation of collaterals and faster progression with higher risk of amputation. PAD and DFU results in a 5-year mortality of 50%, which is worse than many common cancers. When not revascularised the limb salvage rate is around 50% at 1 year which increases to 80–85% after revascularisation.^38,39 The infected DFU with PAD should be considered as a medical emergency where time = tissue.

Urgent vascular imaging and revascularisation should be considered when the ankle pressure is < 50 mmHg, ankle-brachial index (ABI) < 0.5, toe pressure <30 mmHg, or where transcutaneous oxygen pressure (TcPO2) <25 mmHg. Revascularisation might also be indicated when an ulcer does not improve after 6 weeks of optimal care, in the presence of massive infection, a large ulcer surface area or when it is located at the heel (which is notoriously poorly vascularized).^40,41 Diabetic microangiopathy should not automatically assumed to be the cause of poor healing.

The goal of a revascularisation is to improve wound or ulcer healing, treat incapacitating intermittent claudication or rest pain, prevent amputation, or improve wound healing after amputation. In those patients who are a high operative risk and/or non-ambulatory, a revascularisation should be avoided.⁴²

When revascularisation is indicated a color duplex ultrasound (CDUS) assesses the aorto-iliac and femoropopliteal vessels. Occlusions or hemodynamically significant stenoses should be treated. If there are none, crural artery pathology is logically the cause and crural duplex imaging can be skipped. Angiography should be performed with concurrent direct crural angioplasty. A CT-angiography or contrast-enhanced magnetic resonance angiography (CE-MRA) evaluates the entire lower extremity arterial circulation directly and have both advantages and disadvantages over CDUS.

Evidence is inadequate in favour for endovascular, open, or hybrid revascularisation techniques.³⁹ Therefore, decision-making is based on individual factors such as morphological distribution of the pathology, availability of autogenous vein for bypass surgery, patient co-morbidities and, finally, local expertise. In this new decade where pandemics and an aging population lead to scarce availability of resources, hospital beds, operating rooms and intensive care units, the goal to get the patient out of the hospital as soon as possible becomes even more essential. This means opting for minimally invasive techniques where possible and the ability to avoid general anesthesia. Sometimes this results in a less durable intervention with lower patency rate but can at least allow for initial healing of limb threatening foot lesions. It takes more oxygen to heal tissue than to keep it healed.

From top to bottom for aorto-iliac pathology, percutaneous transluminal angioplasty (PTA) is the first choice. Even in total occlusive disease a percutaneous Covered Endovascular Repair of the Aortic Bifurcation (CERAB) can be performed with excellent long-term results.⁴³ If the common femoral artery (CFA) is diseased an open surgical thromboendarterectomy (TEA) with patch-plasty is a simple, durable procedure. However, there are recent reports showing acceptable short-term results of PTA with all the benefits of avoiding a surgical cutdown.⁴⁴ For selected patients with long occlusions in the above the knee femoropopliteal location, the patency rates of an endovascular endoluminal bypass with a covered stent are the same as surgical bypass but with faster recovery and less morbidity.⁴⁵ For shorter pathology (<10 cms) a PTA with stenting on indication seems best. For the popliteal artery, a PTA is first choice and primary stent placement should be avoided in this flexion/extension region. For long occlusions, a bypass should be performed. In some patients a hybrid procedure is needed performing a TEA of the CFA followed directly or later (2-staged) by a PTA. Below the knee arteries are the most affected in DM. A PTA is the treatment of choice for these small caliber arteries. Because of the heavily calcified medial atherosclerosis an antegrade passage is sometimes impossible. A retrograde passage can be achieved through a tibio-pedal access with a micropuncture set. Drug coated balloons have no role in infrapopliteal PTA for improving limb salvage although the suggested higher mortality rates are not confirmed in a recent meta-analysis.^46,47 Due to the poor collateral formation in DM restoration of flow to an artery directly supplying the affected area (angiosome) is associated with better outcomes than indirect angiosome revascularisation.^8,48,49 If conventional techniques are not possible a (percutaneous) deep venous arterialization can be considered.⁵⁰

After a revascularisation procedure a multidisciplinary team together with the patient are part of a comprehensive care plan including extensive cardiovascular risk management (CVRM) such as smoking cessation, hypertension and glycaemia control, statin drug and a low dose platelet inhibitor.

4. Plastic surgery in diabetic limb salvage

Once it has been established that a biomechanically functional foot can be reconstructed, soft tissue reconstruction can be achieved by simple techniques 90% of the time and complex flap reconstruction in 10% of cases. Optimizing the blood flow to the wound using the angiosome principle is critical to successful reconstruction.⁴⁸ A doppler and selective occlusion assesses the patency of these connections and direction of flow. This is critical in designing local flaps, pedicled flaps, and amputations. If there are any questions about the quality of the blood flow, arteriograms with magnified views of the foot are critical in planning the reconstruction.

Assuming there is adequate blood flow, the first step is to establish a clean and healthy wound base.⁵¹ If the wound is acute, a clean base can be established with excisional debridement and either immediate closure or covering the wound with a NPWT+/-instillation in preparation for subsequent closure.⁵² If it is a chronic wound, it must be converted into an acute one through serial debridement until a clean wound base is established.

The reconstructive ladder is then applied to choose the most functional soft tissue reconstruction option. That may involve amputating part of the foot or removing bone. The soft tissue options include healing by secondary intention, delayed primary closure (DPC), split thickness skin graft (STSG), STSG ± neo-dermis (S&N), and local,⁵³ pedicled⁵⁴ or microsurgical free flap (FF).⁵⁵ If DPC is not possible, the following anatomic reconstruction options exist.

For the plantar foot, one needs a well-padded surface over the bone to prevent recurrence. If there is adequate underlying tissue, STSG or S&N provides good outcome. For smaller plantar deep defects, local flaps or pedicle muscle flaps are another option. For larger plantar wounds, a FF (muscle + STSG or fascia-cutaneous flap) works well.

For the posterior heel, a modified calcaneal resection and primary closure is adequate for a braced functional foot.⁵⁶ For medial or lateral heel smaller wounds, local muscle flaps with STSG are a good option. For medial or lateral ankle wounds, transposition flaps based on perforators can be rotated to cover exposed malleolus. For larger defects, FF are the best option. For Achilles tendon wounds, STSG or S&N works well if the Achilles tendon is well vascularized. Otherwise, a perforator flap or FF⁵⁷ provides good soft tissue cover.

For the distal forefoot, fillet of toe(s) can be used although the amount of tissue obtained is often disappointing. For the dorsum of the foot with intact tendons, S&N is good solution. For deeper wounds, a thin FF is ideal. For the superficial anterior ankle, S&N suffices and for deeper wounds, a pedicled or FF should be used.

Avoiding tension during the closure and protection of soft tissue reconstruction through immobilization and offloading are the keys to optimizing the soft tissue closure. This can take up to 4–6 weeks. Subsequent appropriate orthotics or assisted foot orthotic will help prevent recidivism.

5. Conclusion

Management of minor diabetic foot surgical conditions and most infection clearance procedures require core surgical competencies and can be optimally managed by small teams. However, complex and limb threatening diabetic foot presentations often have a combination of infection, deformity, and vascular components. Accordingly, successful functional limb salvage is best delivered by an MDFT that includes a team of surgeons. Diabetic foot surgery is a rapidly evolving subspeciality and the collective surgical expertise of the surgical team in the MDFT should include bony (infection clearance, deformity correction), soft tissue (infection clearance, plastic surgery) and vascular corrective treatments.

Author contributions

V.K. contributed to conceptualization. R.G.F. supervised the manuscript writing. R.G.F., V.K., A. K., L.S., and C.A. wrote and edited the manuscript. A.B. edited the manuscript, All authors have read and approved the final manuscript.

Disclosure

None.

Funding of the study

No funding was involved in this study.

Declaration of competing interest

All authors report no conflicts of interest in the writing of this manuscript.