Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Amputation is defined as the partial or complete removal of a limb or extremity arising from surgery, underlying disease or trauma. Minor amputations are those performed in the foot and major amputations are those above the level of the ankle.
Over 90% of major leg amputations carried out in England are attributed to peripheral arterial disease (PAD) with similar rates in the United States. Therefore the vascular surgeon and surgeon-in-training must be familiar with the pre-, peri- and post-operative management of the amputee patient in terms of not only the amputation itself but also prosthetic management and rehabilitation.
Traditionally, amputation was considered a ‘nadir’ or ‘failure’ in vascular surgery , when revascularisation has failed or the limb is deemed unsalvageable. However, it is increasingly recognised that surgical amputation is the first step in the vascular patient’s rehabilitation process, especially in terms of pain relief, removal of necrotic, ischaemic or infected tissue and improvement in quality of life (QoL), irrespective of whether the patient is a prospective prosthetic limb user or not.
Amputation management should start before the amputation itself and continue into the community. Specialist and holistic assessment of the patient leading to careful pre-operative planning, selection of the amputation level, good surgical technique and optimal post-operative management are all vital in the initial stages of amputee rehabilitation. Subsequent stages involve post-operative early rehabilitation, including wound management, the use of early walking aids, wheelchair and home assessments, prosthetic limb fitting (where appropriate), and continuing long-term follow-up and support to patients and their families in the community.
This chapter focuses on the rehabilitation of adult patients undergoing lower extremity amputation caused by PAD and or diabetes although its general principles can also be applied to non-vascular amputations such as those caused by trauma, infection, frostbite or neoplasm.
In the United States, there are an estimated 2 million amputees with this figure forecasted to double by 2050. The incidence of all amputations is estimated to be over 150 000 per year in the USA.
Major amputations are a common operation in Europe and the numbers are well recorded in routinely collected data. In 2014 16 645 major amputations were performed in Germany and 25 312 amputations over a 6-year period in England.
In higher-income Western countries, the overwhelming majority (>80%) of lower extremity amputations are resulting from PAD, whereas in lower-income countries the most common reason for amputation is trauma.
Amputation carries a significant financial burden with annual costs in the United States alone reaching US$ 4.3 billion.
There is a strong association between PAD and diabetes mellitus in terms of the risk of amputation. Approximately one-third of amputees have diabetes mellitus in the United Kingdom. This figure rises to 50% in Australia and 70% in Germany. In men with PAD, the presence of diabetes mellitus alongside PAD confers a significant amputation risk.
A systematic review in 2011 of 57 studies reported a marked difference in the incidence of amputation globally. Rates were highest in the United States (9600 per 100 000 population in the US state of Louisiana) followed by Northern Europe, and lowest in Taiwan, Japan and Spain (1.5 per 100 000 population). In the United States, there are also significant ethnic and racial variations with African Americans twice as likely to undergo amputation compared to White Americans. However, the causes of this variation are unclear, especially as similar disparities amongst Black and White populations are not found in other countries such as the United Kingdom.
The UK Limbless Statistics report last published in 2012 showed that a total of 5906 new referrals to amputee rehabilitation centres were made for the preceding 1 year, with lower extremity amputations accounting for 91% of referrals and ‘dysvascularity’ (a term that suggests an amputation is the result of inadequate blood supply and usually associated with the complications of PAD and/or diabetes) as the most common cause of amputation. In this lower extremity group, half underwent an amputation at a transtibial level and 34% at transfemoral level. The sex breakdown of new referrals has remained consistent, with female referrals accounting for 30% of all new referrals with an average age of 68 years (65 years for men).
There is increasing evidence that the impact of modern vascular surgery may have resulted in a reduction in the incidence of amputation. A 2018 British study of 103 934 patients with PAD who underwent revascularisation showed that there was a marked reduction in the estimated 1-year risk of major amputation from 5.7–3.9%.
Similar trends have also been observed in the United States, where national billing data have revealed a 45% reduction in the rate of lower extremity amputations during the period 1996 to 2011 associated with a quadrupling in the number of therapeutic endovascular interventions.
Survival rates after an amputation depend upon the cause of the amputation rather than the amputation itself. Patients who have amputations following trauma tend to have good long-term outcomes in terms of QoL and survival. However, those who undergo amputation because of dysvascularity face a significantly worse prognosis. A 2020 systematic review and meta-analysis of 61 studies comprising a total of 36 037 patients from 17 countries and regions, and showed that patients who had amputations because of dysvascularity (including associated diabetes mellitus) faced pooled mortality rates of 33.7% at 1-year post-amputation, and rising steadily to 80% at 10-years follow-up.
Following an initial major leg amputation caused by dysvascularity, amputees who also had diabetes were 2.8 times more likely to undergo an additional major amputation of the contralateral leg, compared to those without diabetes.
The clinical decision to perform amputation can appear straightforward when there is extensive tissue loss resulting in non-functioning limb. In other situations, the decision to amputate may not be so straightforward and the role of amputation in the management of chronic limb-threatening limb ischaemia (CLTI) remains controversial.
The Trans-Atlantic Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II) recommends that if revascularisation is unlikely to be successful and the patient suffers intolerable pain or has a spreading infection, then primary amputation should be considered. This approach is supported by the literature in that it appears to be the most viable outcome in terms of QoL and cost-effectiveness.
TASC II also recommended that non-ambulatory elderly patients presenting with CLTI should be considered for primary amputation, especially if they have flexion contractures. The reasoning is that where patients with CLTI also have co-existing disabilities or co-morbidities that would render them unable to make use of a salvageable limb, then primary amputation with appropriate rehabilitation offers the best option (even if the patient is unlikely to be a prosthetic limb user). Further examples of such patients include those with severe dementia, dense hemiplegia or spinal paralysis, severe arthritis and severe cardiorespiratory disease. Fig. 6.1 is an algorithm on the TASC II recommendations on the management of CLTI.
Selection of the anatomical level of amputation depends on the potential for surgical wound healing and rehabilitation as well as prosthetic considerations. The potential for rehabilitation and setting of likely goals can only be determined through a holistic assessment of the patient by the multidisciplinary specialist amputee rehabilitation team. This assessment should include co-morbidities and disabilities, cognitive state and motivation, likely discharge destination, patient’s occupation, lifestyle and support network, as well as their aspirations and wishes.
In general terms, the more proximal the level of amputation, the more difficult it will be for the patient to achieve independent ambulation with a prosthetic limb, especially if functional native joints such as the knee are lost.
The most common amputation levels in CLTI are transtibial, transfemoral, knee disarticulation and Gritti–Stokes amputation. Using an anatomical approach, this section will cover the main amputation levels along the lower extremity from distal to proximal, before summarising on amputation level selection.
Minor amputations that are distal to the ankle joint require an abundant blood supply to heal. Unless the foot can be successfully revascularised, wounds following local amputation of single or multiple digits are generally difficult to heal. ,
Minor amputations should generally be avoided unless the foot can be successfully revascularised. ,
Ray amputation, which involves removal of the toe and partial removal of the metatarsal can be considered in forefoot ischaemia and can ensure a more adequate surgical debridement of the infected or necrotic margins. In 2014 a single centre study, carried out in Singapore, of 150 patients with diabetes with infection or gangrene of the forefoot who underwent ray amputation found that 71% had healed without complication at 1-year follow-up. However, a systematic review carried out in 2012, of 435 partial first ray amputations revealed that a fifth of these patients required more proximal re-amputation within 26 months.
Transmetatarsal amputations (TMA), if technically feasible with or without revascularisation, also produce satisfactory results in forefoot ischaemia.
Intra-operatively, soft tissue should be preserved as much as possible, and consideration is given to leaving the flaps open with delayed closure, as infection and gangrene of the flaps may occur if primary closure is attempted.
A single centre study in Italy from 2016, explored the effects of TMA with revascularisation in 206 patients with diabetes with forefoot infection and gangrene. The authors found that on discharge from hospital, 38% of these patients were ambulating independently, 43% with assistance and 60% were directly discharged home. At 1-year follow-up, the percentage of patients ambulating independently had risen to 77% and 81% of patients were living at home.
A more recent Chinese study from 2019 explored the effects of TMA on 97 patients with diabetes with forefoot gangrene who were unsuitable for revascularisation. The authors found that 64.9% of these patients had their wound completely healed after a median of 8 months.
Amputations of the midfoot include Lisfranc and Chopart amputations. A Lisfranc separates all five metatarsals from the cuboid and their respective cuneiforms, whereas a Chopart amputation removes more of the midfoot and separates the talonavicular and calcaneocuboid joints.
Midfoot amputations have a significant risk of post-operative equinovarus deformity caused by detaching the insertions of peroneal longus, brevis and tibialis anterior as part of the amputation surgery, which in turn leads to unopposed plantarflexion by the gastrocsoleus complex. If midfoot amputations are absolutely necessary, the surgeon should consider inter-operative rebalancing techniques such as Achilles tenotomy, tibialis anterior or extensor digitorum transfer to the talus or even ankle arthrodesis.
Prosthetic options with these amputations are limited and midfoot amputations also have high rates of subsequent proximal revision amputations (i.e., ‘salami-slicing surgery’).
Amputations of the midfoot are not usually recommended in CLTI.34,35
A Syme amputation involves disarticulation of the ankle joint, removal of both malleoli and securing a long posterior flap containing the heel pad to the anterior tibia to facilitate intermittent weight-bearing. It can be used to successfully treat forefoot gangrene, especially in diabetic patients with dysvascularity. Viability of the posterior flap and heel pad is heavily dependent on the patency of the posterior tibial artery. Thus an ankle-brachial index of less than 0.5 is generally considered to be a contraindication to a Syme amputation.
As the heel pad allows for intermittent weight-bearing, the residual limb in a Syme amputation can be walked on for very short distances without a prosthetic limb (such as when ambulating from the bedroom to the toilet).
Compared to a transtibial amputation (TTA), a Syme amputation offers reduced energy expenditure when ambulating, greater gait stability and thus requires less physiotherapy and gait retraining. However, breakdown and migration of the posterior flap and heal pad may occur if a prosthesis is used regularly. Prosthetic options are also limited with a Syme amputation and there are considerable difficulties in prosthetic fitting as the prosthesis would inevitability be bulky to encompass the bulbous residual limb, which may also impede the contralateral limb when walking.
Syme amputations are not usually recommended in CLTI, especially if the patient is a potential regular prosthetic limb user.
Approximately half of all referrals to amputee rehabilitation centres in the United Kingdom involve TTAs making it the most common amputation level. The second most common level was transfemoral amputation (TFA) at 34%.
Preservation of the native functioning knee joint has tremendous advantages in terms of mobility with a prosthetic limb.
Compared to independently mobile non-amputees, the energy cost of walking with a prosthetic limb in dysvascular amputees who have undergone a unilateral TTA is increased by 40%, whilst in unilateral transfemoral amputees by 120%. This energy cost can increase to 280% in bilateral transfemoral amputees when mobilising with two prosthetic limbs. The increased energy demands invariably affect an amputee’s ability to use a prosthetic limb with approximately 74% of bilateral amputees not being able to walk with prosthetic limbs 5 years following surgery.
In 2013 a Dutch longitudinal study of 82 patients exploring QoL following lower extremity amputation and found that amputation level was a key factor in determining QoL. Using the RAND-36 questionnaire, the authors noted that transtibial amputees reported better QoL scores compared to transfemoral amputees within the 18 months following an amputation. This was because of a better walking ability leading to greater social participation amongst patients who had a TTA.
A previous British observational study from 2003, of 281 unilateral lower extremity amputees, found that amongst older (>50 years old) amputees, approximately 50% and 60% of transtibial amputees were ambulating independently in the community or at home, respectively 1-year following surgery. Whereas for transfemoral amputees, the respective figures were much lower at 25% and 50%.
The native knee joint should be preserved where possible to facilitate rehabilitation and ambulation with a prosthetic limb. ,
Traditionally, through-knee amputation was only considered in those patients who were deemed incapable of walking. The often-quoted Lower Extremity Assessment Project (LEAP) study showed worse outcomes in trauma patients who had knee disarticulation compared to transtibial and TFAs, slower walking speeds and greater difficulty walking on uneven ground or outdoors in bad weather.
However, more recent studies have suggested no significant differences in post-operative complications or functional outcomes in terms of mobility, activities of daily living, QoL or pain between knee disarticulation and transfemoral amputees in both vascular and trauma patients.
With through-knee amputations, the longer lever-arm of the residual limb assists transfers, sitting balance and maintains hip adductor muscle attachment and distal proprioception. ,
The conventional through-knee amputation (i.e., knee disarticulation) may cause problems related to the leakage of synovial fluid, bulbous femoral condyles and the retained patella, which could potentially cause difficulties with prosthetic socket fitting if the patient becomes ambulatory later on. The Gritti–Stokes amputation theoretically avoids these problems by resecting the bone at the supra-condylar femoral level, enucleating the patella and anchoring it to the distal femur as an end-cap ( Fig. 6.2 ). However excision of femoral condyles could adversely affect the suspension and rotational stability of the applied prosthesis. Enucleating the patella and anchoring it to the distal femur has been associated with femora-patellar non-union and mal-union. In a Gritti-Stokes amputation, the load bearing axis lies posterior to the residual limb. This could to lead to greater mechanical stress on the femur, which in turn could adversely effect ambulation with a prosthetic limb. , , Therefore, caution is strongly advised when considering Gritti-Stokes amputation in ambulatory patients.
Due to a longer femoral component and lowered prosthetic knee centre, as well as limited availability of stance and swing phase control mechanisms in the prosthetic knee joint at this level, prosthetic limb fitting can be technically challenging in through-knee amputees.
Hip disarticulation and hemipelvectomy are rare procedures in patients with PAD. They comprise approximately 1% of all lower extremity amputees presenting to UK amputee rehabilitation centres. The majority of these procedures are carried out for neoplasm (usually sarcoma) followed by infection and major trauma, rather than dysvascularity.
The complex neurovascular and visceral anatomy of the pelvic region and relative rarity of these procedures present significant technical, surgical and rehabilitative challenges. With intra-operative blood loss in hemipelvectomy averaging 3 litres and peri-operative mortality as high as 44%, both these operations should be considered as amputations of last resort. Of the survivors, with an integrated multidisciplinary rehabilitative approach, approximately 43% can eventually use a prosthetic limb following either surgery.
The selection of the amputation level in major lower extremity amputation can be multifactorial and complicated. Thus, it should be decided on after careful multidisciplinary assessment and discussion between the surgical and rehabilitation teams. However, there are general rehabilitation principles that should be taken into consideration by the vascular surgeon when making this decision, as outlined below. Fig. 6.3 illustrates this decision-making process via an algorithm.
Preserve the native functional knee joint if possible
Preserving the native knee joint is the key factor in reducing the energy and metabolic cost when using a prosthetic limb. This should be considered whenever possible, especially if it is anticipated that the patient has the potential to transfer (e.g., from bed to chair) or walk with a prosthetic limb.
In non-ambulatory patients, consider knee disarticulation or Gritti–Stokes amputation
In patients who are bed- or chair-bound, a TTA residual limb risks non-healing and may also lead to flexion contractures of the hip and knee joints and thus impeding patient transfers. A knee disarticulation or Gritti–Stokes amputation can avoid these aforementioned complications. These two amputations are also preferable to TFA as they provide longer lever-arms and larger residual limb surface areas, which are more conducive to transferring and promote wheelchair mobility by providing a more stable counterbalance. ,
TFAs should be considered in fixed knee flexion of ≥30 degrees and/or severe knee arthropathy.
A fixed knee flexion deformity of more than 15% can reduce foot clearance during the swing phase, which may impede walking and can lead to quadriceps fatigue and anterior knee pain. Usually, a flexion contracture of 30 degrees is considered the maximum limit for transtibial prosthetic limb fitting. In addition, severe knee arthropathy can also impede walking. Therefore in these two circumstances, a more proximal amputation should be considered.
In ambulatory patients who cannot have a TTA, a TFA is the next best option.
In patients where a TTA is not possible but the patient has the potential to walk, a TFA is preferred rather than knee disarticulation or Gritti–Stokes amputation. This is because of problems of prosthetic fitting, which may compromise cosmetic appearance and function. Cosmesis is affected because the prosthetic knee joint extends beyond the length of the femur. Therefore for patients with a through-knee amputation, the femoral component is longer. Functionally, this may mean that sitting in small spaces is more cumbersome or restricted because of this increased length. Conversely, some studies have demonstrated better functional outcomes in patients with a knee disarticulation. This is explained by the longer lever-arm and the fact that the adductor muscles of the hip are preserved in a knee disarticulation. Both of these factors improve the thigh motion for ambulation, decreases energy expenditure and help maintain distal proprioception, which helps facilitate walking with a prosthetic limb. ,
A UK National Confidential Enquiry into Patient Outcome and Deaths (NCEPOD) in 2014 revealed that only 44% of patients undergoing amputations for dysvascularity were assessed as having received a good level of care. The enquiry also reported that half of the patients lacked multidisciplinary involvement in their care as well as a clear patient pathway. The enquiry suggested that its findings may explain the marked discrepancy between mortality rates for amputation in patients with dysvascularity between the United Kingdom and the United States (12.4% vs. 9.6%, respectively).
Therefore a multidisciplinary team (MDT) should be involved in the early stages of planning care for amputees with access to other medical and surgical specialities and allied health professionals, both pre- and post-amputation. In light of the significant risks of morbidities and mortalities, all amputations should be performed by experienced surgeons at consultant level and should not be delegated to unsupervised and inexperienced trainees.
Specific surgical considerations for transfemoral, transtibial and foot amputations will be covered later on in this section.
To enhance post-operative rehabilitation and facilitate limb sitting, the following general surgical principles should be followed:
Tissues must be handled meticulously with care and strict haemostatic control
A thigh tourniquet should be considered to reduce significant blood loss and transfusion requirements as well as ensure a bloodless operative field for the surgeon ,
Bony prominences around cut bone should be removed, the edges should be smoothed off and bevelled
To reduce painful neuroma formation, nerves should be cut under tension, proximal to the bone edge in a scar-free environment and large nerves should be ligated when near an associated vessel
To reduce arteriovenous fistula or aneurysm formation, larger arteries and veins should be individually dissected and ligated
Skin and muscle flaps should be oversized initially and then shaped and trimmed as appropriate
There should be good-quality sensate skin coverage of stump without any tension
Muscle flaps should be anchored over the bone edge by myodesis, a long posterior flap sutured anteriorly or a balanced myoplasty
Skin and muscle flaps should be trimmed, fashioned and shaped to prevent ‘dog-ears’, redundant tissue or a bulbous stump
Closure should be in layered with meticulous suturing technique. The skin should be closed with sutures rather than clips to reduce post-operative wound infection
The ideal transection of the femur is 12 cm (10–15 cm acceptable) above the native knee joint, which is measured from the medial joint line. This length allows enough clearance for incorporating a prosthetic limb, and avoids using a lowered prosthetic knee centre, which gives an unacceptable cosmetic and functional disability when walking. The exact level of the bone cut will depend on the thickness of the thigh muscles and subcutaneous tissue. In morbidly obese patients with excessive soft tissue coverage, a through-knee amputation should be considered instead.
It is important to ensure adequate myoplastic coverage over the cut femur to reduce pain and discomfort as well as to allow muscle balancing of the hip adductors to abductors and rectus femoris to hamstrings.
If muscle quality is adequate, adductor myodesis should be considered. The procedure, which involves anchoring of muscle to bone (i.e., the femur) via drill holes, reduces the risk of muscle slippage and overhang, creates a dynamic muscle balance, removes tension from the anterior myocutaneous flap, and provides a soft tissue envelope to facilitate limb fitting.
Similarly, a quadriceps myodesis should be considered, which can either be secured to the adductor flap or the posterior femur.
The myoplasty and myodeses should be performed with the hip joint in neutral (in the sagittal plane) and the naturally adducted position of 5-10°. This prevents both hip flexion and abduction contractures and is ideal for prosthetic function.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here