Most occlusions in acute limb ischemia involve the popliteal or more proximal arteries. Therefore, amputation in acute ischemia is frequently at or above the knee level. In addition, significant muscle beds are usually threatened and most vascular surgeons attempt to salvage acutely threatened limbs if the outcome is in doubt; so, many amputations in acute limb ischemia occur in patients suffering significant reperfusion injury with rhabdomyolysis and renal damage. Currently in the United States, many patients with acute limb ischemia also undergo primary endovascular therapy—however, prior incisions can complicate amputation flaps. Acute limb ischemia patients are often receiving anticoagulants for therapy of the underlying thrombotic or embolic event, further increasing the risk of hematoma and wound complications. Finally, increasing cases of amputation due to acute limb ischemia involve iatrogenic vascular injury or delay in diagnosis/referral, increasing the psychological complexity of the care and having medical legal implications.

The vast majority of major lower extremity amputations are due to CLI. Unlike other areas of vascular surgery such as aneurysm and carotid diseases, accurate guidelines on potential salvage of limbs based on revascularization potential and extent of necrosis are lacking. This leads to significant differences in practice patterns based on geography, access to revascularization and vascular imaging, and clinician experience in the use of open or endovascular techniques. Finally, unlike cardiac disease, there is no established route of access for suspected CLI to revascularization in modern health care systems, leading to potential delays in referral and diagnosis, wide differences in practice patterns, and ultimately greater costs. Therefore, eventual amputees with CLI fall into two categories: (a) patients who are not offered revascularization either due to the extent of pedal necrosis on vascular presentation, failure to recognize CLI and futile forefoot amputations, and/or comorbidities (medical and/or functional) that would preclude revascularization, or rarely anatomy not amenable to revascularization; and (b) patients who fail revascularization (either occlusion or ongoing pedal necrosis despite improved circulation). Most centers with aggressive limb salvage practice perform far fewer primary amputations than revascularizations, so that many patients with CLI who require amputation have had prior endo- or open revascularizations that have failed. Although several studies have demonstrated no effect of a failed infrainguinal revascularization on amputation level, below-knee amputations (BKAs) after failed infrainguinal bypasses have greater wound complications and prolonged healing times. With the increase in endovascular techniques for CLI, these wound issues have been reduced greatly.

Younger diabetics with normal circulation who develop overwhelming foot infections are another major amputee group. Typical patterns include first or fifth metatarsal osteomyelitis from a chronic malperforans ulcer in that location. Patients with Charcot foot can have chronic ulceration at the midplantar area due to osteophytes and pressure points at the chronic midfoot fracture. Many of these diabetic amputees have had prior digit or ray amputations that may have healed, only to have another malperforans ulcer occur in the adjacent digit as the initial amputation destabilizes the distribution of pressure on the remaining metatarsal heads. The cause of eventual limb loss in this population is multifactorial. Poor compliance (often demonstrated in significantly elevated glycosylated hemoglobin values), obesity (increasing the pressure on the neuropathic foot when ambulatory), and inadequate footwear are major contributors. A less recognized but equally important cause is patient education. Many poorly educated diabetic patients assume that seeking health care for foot lesions will lead to a major amputation, and therefore delay care until a major amputation is inevitable, which continues to perpetuate this widespread fear. Another issue is access to care as many of these patients lack insurance.

Few vascular surgery procedures have the psychological and functional impact of a major lower extremity amputation. Amputation-free survival is considered the primary end point for any clinical trial for patients with CLI. However, in clinical practice, a number of variables can markedly change this end point that has no relationship to efficacy of therapy. Understandably, patients often adamantly refuse amputation. CLI (the most common indication for amputation worldwide) rarely results in overwhelming infection and imminent patient death if left untreated. Therefore, patient pain tolerance is often the deciding variable. Many patients turn to unproven therapies in desperation. The widespread use/advertising of wound centers gives patients hope that all wounds can ultimately be healed—but the reality is some wounds need to be removed. All of these issues increase amputation-free survival. Therefore, for many patients, the first step in preoperative preparation is acceptance of the procedure—which frequently takes days to weeks and sometimes multiple visits to answer questions and work through the various stages of grief and loss—for the patient and family members. Often, consultation with rehabilitation medicine, physical therapy, prosthetics, and even other amputees is very helpful in preparing the patient and family for the realities of life as an amputee.

Standard evaluation with a thorough history and physical examination should be performed in all patients. Active congestive heart failure, unstable angina, or concurrent myocardial infarction must be addressed. In the absence of active cardiac symptoms, evidence of congestive heart failure on history or physical examination, or acute changes on electrocardiogram, thorough cardiac evaluation is not routinely performed. It is important to recognize that most patients presenting for amputation have had limited activity and may not manifest covert cardiac disease due to the lack of physiologic stress, so even minor symptoms should be very concerning.

Beta-blockade has been shown to be beneficial in perioperative management of patients undergoing revascularization for aortic and lower extremity disease. However, due to the shared risk factors, the positive impact of beta-blockers can likely be extrapolated to the amputation population. Management of associated hypertension, diabetes, and renal failure should be optimized. An aggressive approach to normalize glucose levels is essential to ensure proper wound healing. The timing of hemodialysis in relation to operation is important in managing fluids and electrolytes in the perioperative period. Perioperative nutrition is critical, as many patients are depressed, manifested by severe emaciation. Depressed patients not eating or participating in their postoperative rehabilitation programs are at risk for a host of complications including further nutritional depletion, incisional wound complications, decubitus ulcers, contractures, and deep venous thrombosis (DVT).

Finally, due to relative inactivity and frequent hypercoagulable states in the amputation population, venous thromboembolism
is a major source of postoperative morbidity/mortality. Recent data would indicate that the incidence is 15%. In addition to the risk of fatal pulmonary embolism, postoperative DVT may cause severe stump swelling and impaired primary healing of the amputation. Routine DVT prophylaxis and/or careful perioperative screening for DVT with duplex scanning in amputation patients appears prudent.

The goals of major lower extremity amputation are several: to eliminate the nonviable tissue, to provide a stump that has the best chance to heal, and to provide a stump with the best chance of long-term function—ambulation with a prosthesis. However, which of these factors is most important often depends on the age, baseline function, and comorbidities of the patient. On balance, it is important to avoid additional operative procedures in this compromised group of patients. In all reported series, healing of an above-knee amputation (AKA) is more rapid and involves less incisional complications than healing a BKA. However, any chance at rehabilitation and prosthetic use in older vascular patients requires the knee joint. Finally, physicians and patients wish to preserve as much limb length as possible. More recently, the through-knee amputation has been advocated. With the use of a computerized C-leg prosthesis, patients have a much better return to ambulation than with an AKA due to the longer lever arm for mechanical advantage and balance. In addition, a through knee amputation has minimal muscle transaction and is considered to be faster and with less blood loss.

Worldwide, the ratio of AKAs to BKAs performed is 1. Many AKAs are performed in patients with limited rehabilitation potential and significant comorbidities, as the primary healing is better and there is a risk of knee contracture in BKAs done in patients who are unable to participate in postoperative physical therapy. Despite this, most patients are considered for BKA due to a variety of reasons, and the surgeon needs to determine if a BKA will heal primarily. A number of studies in the literature describe various circulatory measures to determine amputation level healing. As a general point, many of these techniques are not practical in clinical practice. A palpable pulse at a level immediately proximal to a proposed amputation has been associated with a 100% rate of healing. Most surgeons agree that a BKA without a femoral pulse is unlikely to do well. However, a palpable femoral pulse does not guarantee success. The clinical judgment of the operating surgeon (skin temperature, hair growth, tissue bleeding, viable muscle, and wounds without tension) has been demonstrated to be reasonably accurate (75% to 85%) in several large clinical series (combined, around 1,000 patients). Finally, despite the plethora of research regarding the technical aspects of salvaging the knee joint, the sobering fact is that many vascular patients will take minimal advantage of this in rehabilitation.

Thermography, skin perfusion via radioisotope injection, and segmental pulse volume recordings (PVRs)/limb pressures have been used to try to improve the success of clinical judgment in predicting BKA healing. Each has limitations ranging from clinically impractical (thermography and skin perfusion) to not accurate due to calcified vessels (PVRs and pressures) in many amputees. Arguably, the most reproducible noninvasive test used to determine BKA healing is transcutaneous measurement of the partial pressure of oxygen (TcPO₂). Measurements are obtained in the supine position by placing an oximetry probe on the skin in question and heating the area to 44ºC. The generation of heat results in local skin hyperemia, a decrease in flow resistance, and subsequent arterialization of capillary blood. The probe then records skin oxygen tension reflecting the true arterial oxygen pressure in the area in millimeters of mercury (mm Hg). Absolute measurements, measurements compared to a control location such as the chest, and measurements compared to transcutaneous carbon dioxide values have been reported. Various studies have shown that TcPO₂ measurements of greater than or equal to 30 mm Hg have an accuracy of 87% to 100% in predicting BKA healing. In addition, this technique can be used in patients with arterial calcification. However, supplemental oxygen can falsely elevate TcPO₂, and edema can falsely reduce TcPO₂.

The authors follow several general principles in amputation surgery. When at all possible, Esmarch bandages and tourniquets are used to minimize blood loss despite liberal use of the knife for dissection—supported by a single small randomized trial in the UK. The exceptions are diabetic patients with significant calcification where all are accomplished in reducing venous outflow and actually increasing venous bleeding. All major vessels are suture ligated. We do not use bone wax for control of any bone marrow bleeding, for concern of leaving a foreign body and potential infection. We transect both the tibia and the fibula with a saw, to avoid any spiral fracture of the fibula. Major nerves are transected short under tension, with minimal use of chromic ligatures to avoid excessive neuroma formation later and accepting some bleeding as a consequence. Care is taken to debride as much soleus from the BKA muscle flap to avoid a bulbous stump, as this is a difficult prosthetic fit. Some have recommended myodesis, with muscle flaps sutured to the bone, but currently that has not been our practice. Skin is closed with nylon suture in a vertical mattress with long ends, as we often leave the sutures in for several weeks (due to edema and concern for wound separation in patients who are often malnourished) and it helps with later removal in the clinic. We plan plaster casting of BKAs, usually within a few days. We have not done this with AKAs. We have also used subcutaneous and subfascial fenestrated catheters to deliver local analgesia via a computerized pump for 48 hours postoperatively with success.

If there is any question regarding uncontrolled pedal sepsis, a two-staged amputation should be considered. These are compromised patients with significant perioperative morbidity/mortality—much of which can be septic. Guillotine amputations offer the ability to remove the septic focus relatively quickly and also provide direct examination of the calf muscle compartments for septic extension. If discovered, a longitudinal incision either medially or laterally in the leg (similar to a fasciotomy) drains the affected compartment and allows for debridement of nonviable muscle. The patient can be returned to the operating room for the BKA once the edema, fever, and leukocytosis have been resolved (usually 3 to 5 days).

The guillotine amputation is relatively straightforward. The limb is prepped to the knee level in case an incision in the calf is needed to drain purulence/necrosis proximally (most commonly the anterior compartment). The authors do not find tourniquets for guillotine necessary, as the assistant can provide some level of arterial control with a firm grip at the distal calf. In addition, Esmarch exsanguination in a septic foot would be contraindicated. The knife is used to make a circular incision just above the malleolus down to the tibia and fibula (Fig. 1). Electrocautery is often not useful due to significant edema. Either a power saw or a Gigli saw is used to
transect the distal tibia and fibula. The three major vascular bundles are oversewn with silk (Fig. 2). Of the three, the peroneal artery is the most difficult to control as it is between the fibula and the tibia, making exposure a problem. The calf is compressed to check for purulence in the compartments. Finally, the authors place a large piece of thrombin-soaked Gelfoam over the stump to assist in hemostasis postoperatively. A bulky dressing is applied over this and then an ace wrap. It is our practice to not examine the stump for a day or 2 if there was no sign of proximal infection, as the patient’s clinical course is often adequate to gauge therapy, and early dressing changes in a patient with relatively normal circulation at the ankle (as is often the case) risks hemorrhage.