12.27 Replantation

One of the guiding principles of reconstruction espoused by the father of modern plastic surgery Sir Harold Gillies was to replace like with like. In the successful replantation of an amputated part using microsurgery this principle is taken literally. Since the first reports of an arm replantation and the first thumb replantation, the reattachment of amputated parts has become common place and includes almost any extremity of the body.

Unfortunately successful replantation does not mandate satisfactory function, with the aim of survival alone being useless without a functional appendage. The potential adequacy of bone, tendon, and nerve repairs and joint flexibility must be predicted preoperatively before embarking on an attempt at replantation.

The mechanism greatly influences the degree of damage caused to all anatomical structures and will have significant implications on survival and long-term function. It is prudent to give the patient a guarded outcome as occasionally even in ideal circumstances results can be disappointing.

The industrial machinery which cause amputations are many and varied and the injury can be further influenced by the reflex withdrawal response of the victim. A detailed description of the mechanism of injury combined with examination of the amputated part and radiographs will give clues to the extent of tissue damage but ultimately surgical exploration must determine the technical feasibility of replantation.

Sharp or guillotine injury is the most favourable mechanism of amputation although it is the rarest. Unfortunately it is seen most often at very distal levels where replantation is technically difficult.

Localized or moderate crush injuries caused by power saws and many industrial press-type machines are the most common form of injury. The tissue damage occurs over a wider area and often occurs at multiple levels. Following conservative trimming of damaged structures, this type of injury can usually be converted to the sharp type and whilst prognosis is worse than a sharp injury, in the absence of an avulsional component, a high survival rate can still be achieved.

Diffuse crush injuries caused by blunt squashing or high-velocity cutting machines reveal extensive contusion of skin and soft tissue as well as ragged tendon and nerve ends and often comminution of bones. There may be frank loss of tissue especially bone and skin and this may necessitate either local, regional, or free flap coverage.

Avulsion amputations result from a degloving or longitudinal force and are compounded by the natural withdrawal reflex of the victim. Typically, the vascular structures rupture and protrude from the proximal amputation stump while long tendon and nerve ends protrude from the distal part. Extensive damage occurs to the digital arteries over a variable length, often extending proximally to the next branch point and even into the palmar vessels, and may include skip lesions of thrombus. If the mechanism included a torsional component the arteries can look macroscopically normal in the absence of flow and this is due to intramural spiralling seen clearly with more proximal dissection under the microscope. In contrast, veins rupture rather than avulse. The veins are sheared at the same level of the skin and can be anastomosed primarily whereas the arteries nearly always require vein grafting.

Characteristic of avulsion is the extensive degloving of skin which remains with the amputated part, creating poorly vascularized distally based flaps following replantation. The ultimate fate of the skin depends on the extent of distal resection of the arteries in order to restore blood flow. It is possible to be left with an alive digit distally surrounded by a moat of necrotic skin requiring further reconstruction. In general the long-term functional prognosis is poor given the injury to nerves and tendons, especially if the latter have disrupted at their musculotendinous junction. One caveat to this is the classic isolated ring avulsion which often leaves the skeleton and flexor sheath intact and requires revascularization or replantation of the soft tissues alone thereby permitting early active mobilization and a more favourable outcome.

Physical trauma may be combined with thermal or chemical burns which in the absence of blood supply to the amputated part can be impossible to assess. Incorrect storage can cause frostbite of the amputated part. Gross contamination requires widespread debridement of all tissues including bone and necessitates vein grafting of all vessels.

There are no absolute indications to replant. The decision whether or not to replant isn’t always easy and demands experience both in replantation and hand surgery in order to maximize the chance of a successful functional outcome. The aim is to restore function, growth, and cosmesis. Stiff, cold, or painful fingers may be excluded from everyday use and also interfere with the function of the remaining undamaged hand. There remains a subgroup of patients who for religious or cultural reasons will categorically insist on replantation. Similarly some patients prefer to keep a non-functional digit as it affords a better cosmesis than a bothersome prosthesis. Thus whilst the delayed amputation of a finger at 12 months does not constitute a success, if there is any doubt, it is generally better to err on the side of replantation than primary amputation.

Ischaemia time, level, mechanism of injury, number of digits amputated, associated hand injury, and age will all influence the potential for recovery.

The amount of muscle contained in the amputation is the critical determinant with respect to time to revascularization. Amputations containing minimal muscle generally present well before their maximum permissible ischaemia time has elapsed (warm 12–16h; cold <28h). More proximal amputations, particularly those in the forearm or the lower limb, contain significant portions of muscle and therefore have a much more limited warm ischaemia time than digits (6–8h). With viability unlikely and potentially life threatening consequences these replantations should probably not be attempted if the warm ischaemia time exceeds 8h.

There are numerous classifications for the level of amputation but from a practical viewpoint there are five distinct types –

Although technically difficult, it has become obvious that distal replantations, only have to survive to be functional. They have the potential for the best results of any replant.

Only sharp type replantations in Type 1 have the potential to restore function, especially when they are multiple and to the ulnar side of the hand. Even with limited distal interphalangeal joint movement or sensation, the intact proximal interphalangeal joint will restore considerable palmar grip function and balance with the outstretched hand. Paradoxically, Type 2 amputations, with more to gain by replantation, are the poorest indication in a single digit because of the complexity of the structures involved. In practice very good function following a single digit non-guillotine injury in Types 1 and 2 levels is rare and replantation is generally contraindicated.

Replants in Types 0–2 have the relative advantage of not requiring any motor neural re-innervation. Amputations proximal to Type 2 in the hand are strong indications for replantation. Replantation of a clean-cut wrist-level amputation is the most rewarding. At the mid forearm level and above the prognosis worsens.

Whilst an amputation through a joint obviates the need for osteosynthesis and bony union and potentially allows for earlier mobilization, the range of motion resulting is universally disappointing. Repair of the unique stabilizing structures of the joint (usually the proximal interphalangeal joint) cannot hope to replicate the pre-injury status and therefore mobility.

The thumb contributes 40% of function to the hand and is the strongest indication for replantation at any level and at any age (Figure 12.27.1). The mobility in three planes of the trapeziometacarpal joint also compensates for the stiffness postreplantation in the interphalangeal or metacarpophalangeal joints without a significant diminution in function. Replantation of a single finger should only be done in special circumstances, but single-digit amputations with a concomitant injury to another finger or part of the hand, magnify the disability and is a stronger indication (Figure 12.27.2).

Multiple-digit amputations are a strong indication for replantation. The barest use of the hand requires a stable, mobile thumb and a pillar against which it can oppose. The radial three digits are concerned with pulp pinch grip, and therefore sensory excellence is of greater importance than mobility. On the ulnar side of the hand however, power grip is the prime function and this requires digital length and mobility; sensation is relatively less important. The potential for joint movement, especially the proximal interphalangeal joint, is of prime importance. Amputations in Types 0 and 1 do not need sophisticated function as simple restoration of length will allow them to contribute to ulnar power grip.

Where multiple digits have been amputated and not all are suitable for replantation, the opportunity to select the least traumatized of them and transfer to the most appropriate stump should always be kept in mind. Most commonly this will involve the thumb.

Young patients carry a better prognosis than the elderly. Due to the remarkable plasticity of the immature brain to create cortical reorientation, replantation of any digit at any level is strongly indicated in children. Beyond the age of 20, nerve regeneration occurs variably and unpredictably, and all but distal amputations should be expected to have significant limitations in their neurological function in terms of both motor and sensory modalities.

The history should include the calculation of the ischaemic time, the nature and mechanism of the injury, and the presence of any contaminating agents. The patient’s age, occupation, hobbies or sporting skills, and hand dominance should be noted. The status of the patient’s tetanus prophylaxis should be checked. Any life-threatening medical comorbidities may make the decision to replant foolish.

The purpose of emergency examination is to evaluate the technical feasibility of restoring circulation and the potential for functional recovery so that an informed discussion with the patient can proceed prior to surgery. The amputated part will reveal the type of injury and its level. By deduction the proximal stump(s) should mirror the pattern of the distal segment, so that it is generally possible to avoid painful unwrapping of the proximal wound. In the more complex major limb and multiple-injury categories, it is important to exclude more proximal multiple-level injury, extensive tissue loss and particularly brachial plexus injury. Even in these situations it is evident that some form of surgery is inevitable, so that meddlesome explorations prior to anaesthesia should be avoided.

Problems often do not appear until intraoperatively and therefore preoperative consent should be as broad as possible and include grafts of veins, skin, nerve, and bone as well as local and free flaps.

Most amputations have undergone some haemorrhage and therefore a haemoglobin/haematocrit and a blood group cross match should be performed routinely. Plain radiographs of both the injured limb and the amputated parts should be obtained and should demonstrate the joint proximal to the injury.

Although most amputations are isolated injuries the emergency management of the patient should be along trauma guidelines. As there are no absolute indications for replantation, other life-threatening injuries take priority. The patient should be adequately stabilized and resuscitated. There are numerous reports in the literature of digital replants being done hours or days later when the patient is stable.

Preparation of the amputated part in a saline-soaked sterile gauze, placed in a sealed plastic bag, and packed in an ice-water slurry should be done immediately. Despite intensive research, no perfusate has been shown to be convincingly beneficial in cytoprotection of limb tissue.

Haemorrhage can mostly be arrested by a compression bandage, with a tourniquet only used as a last resort. In cases of incomplete amputation with devascularization, the appendage should be splinted in an anatomical position to maximize any residual circulation.

Ideally replantation should be performed as soon as possible by a two-team approach, with one operating on the amputated part and the other on the stump. Much time can be wasted unless a disciplined sequence of repair is followed shown in Table 12.27.1

General anaesthesia is favoured but if time permits an axillary block and catheter may be performed preoperatively which has the dual advantage of assisting with vasodilation due to a sympathetic blockade and providing postoperative analgesia. Although there is now evolving data in the anaesthetic literature that regional anaesthesia may cause a steal phenomenon to an already sympathectomized digit or free flap and may be counterproductive. Distal replants (Type 0) may be done under local anaesthesia only.

Elements of the anaesthesia are vital to optimize success and unfortunately replantations often occur after hours when experienced anaesthetists may not be available. It is incumbent upon the surgeon to ensure perfusion requirements to the digit are achieved. This includes a gentle induction and extubation, maintenance of a core body temperature of 37°C, a urine output of 1mL/kg/h, avoidance of vasopressors and diuretics, normovolaemia, and a haemoglobin of approximately 10. In our experience packed red blood cells followed by concentrated albumin are the best colloids to maintain intravascular volume, and with the use of depth of anaesthesia brainwave monitors, often volatile anaesthetic gases can be reduced with a subsequent improvement in systemic blood pressure.

Contaminating foreign material should be removed and conservative debridement of damaged tissue carried out. All vital structures should be systematically identified and tagged. Proximal damage and retraction of the digital artery and/or nerves will mandate a skin incision for exposure. A Bruner type incision is preferable, so that flap closure is possible over the microvascular anastomosis even if the incision itself cannot be fully closed because of swelling. The ends of the flexor tendons are located and a core suture is inserted into each.

Bone shortening is an essential and frequently underrated initial step in replantation. Guillotine injuries and distal replants are the exception to the rule. The prime reasons for shortening are to allow good quality tension-free wound closure for primary healing, extensor tendon apposition, and nerve approximation. Although skeletal shortening does assist in vascular end to end repair, this is the least valid reason and in fact can lure the operator into inadequate resection of damaged vessels. In cases of disarticulation where shortening is not possible this may necessitate vein graft lengthening.

Rigid fixation to allow early movement is a vital operative step and the most difficult to achieve. Too often the functional benefit of successful revascularization is reduced by malunion although non-union is unusual. The dilemma resides in the balance between rigid low-profile plate fixation with its ensuing widespread periosteal stripping and the less invasive pin fixation favouring maximization of vascularity to the fracture but a consequent decrease in rigidity. Properly placed crossed K-wires carry a number of significant advantages as a method of skeletal fixation in replantation providing stable but not rigid fixation, minimal periosteal stripping and easy removal (Figure 12.27.3).

The extensor tendon, including the lateral bands and intrinsics, should be repaired with a continuous running technique. It is preferable to perform the venous anastomoses prior to the arteries to avoid unnecessary and prolonged bleeding if the reverse sequence is followed. In distal replants, some severe crush injuries and those with prolonged ischaemia where revascularization is doubtful then it is advisable to perform the arterial repair first. The dorsal digital veins are usually easily located in the subcutaneous tissue by the presence of small blood clots and the location of the veins on the distal amputation stump will mirror their position on the proximal stump. Occasionally adequate sized veins can be seen on the volar aspect. Radical resection is rarely necessary so direct anastomosis is common. The dorsal skin is closed cautiously over the venous anastomoses. Inadequate skin closure due to insufficient bone shortening or degloving risks venous exposure, desiccation, infection, secondary haemorrhage and late thrombosis after 4–5 days.

After periosteal repair the flexor tendons should be repaired by tying the core sutures previously placed and in so doing restoring the natural posture of the finger thus reducing tension on the neurovascular repairs. Digital arteries are identified by their constant dorsal relationship to the nerves in the fingers. Towards their termination in the distal phalanx they converge towards the midline to form an arcade with often a single midline artery deeply placed on the bone. The proximal vessels must be resected until pulsatile flow is encountered. If the vessel spasms, manoeuvres to restore flow include distraction of the vessel to its physiological length, a bolus of intravenous heparin, dependency of the limb, topical antispasmodics, and warm saline irrigation. Experience dictates an adequate distal artery seen under the microscope. Vein grafts should be harvested from the flexor aspect of the forearm and should be placed under mild tension in order to avoid redundancy. Where access is difficult, it is helpful to anastomose the distal end of the vein graft into the amputated part prior to skeletal fixation. Increased survival is directly proportional to the number of anastomosed vessels.

If no useful arteries and/or veins can be located for revascularization, afferent arteriovenous anastomosis may provide adequate perfusion or efferent arteriovenous anastomosis may provide outflow.

Both digital nerves should be repaired using an epineural suture technique. The medial cutaneous nerve of the forearm, the posterior interosseous nerve, or an interpositional vein graft are convenient donor sites if no rejected amputation segments are available. Skin closure should be tension free and designed so that vascular pedicles are covered. This may necessitate flap transpositions with skin grafts to the secondary defects (Figure 12.27.4). Circumferential dressings should be avoided when placing a volar plaster.

The capacity of replanted tissues to develop sensory reinnervation gives limb replantation a unique advantage over any contemporary prosthesis. The mixed nature of the proximal nerves involved, the large bulk of muscle, the local tissue destruction, and sometimes the avulsive mechanism of injury all weigh heavily against an ideal outcome. Larger vessels make vascular repair relatively simple, but permissible ischaemia times are severely reduced. The prognosis for significant recovery following repaired brachial plexus or proximal peripheral nerve injuries, even in ideal circumstances, is poor. In general, above-elbow replantations are contraindicated except in children. Double-level injuries, especially an associated brachial plexus avulsion are a contraindication.

In the lower limb the replanted foot or leg is likely to remain insensate and will cause difficulties for the patient during ambulation and with footwear. Patients tolerate below-knee prostheses well and rehabilitate rapidly. Therefore, except in children, replantation of lower limbs should be reserved for incomplete amputations where the posterior tibial nerve is in continuity.

There are several principles in major limb replantation that are not pertinent to the replantation of digits. Major limb replantation is a true surgical emergency. Contrary to traditional earlier credos, all major limb replantations should be revascularized prior to any bony fixation or other soft tissue repair. Temporary endarterectomy shunts can be used to rapidly restore arterial and venous continuity. Venous repair should always be delayed to allow toxins in the venous effluent to dissipate prior to reconnection into the circulation. Intraoperative cooling is also important in these large-muscle-containing parts.

The objectives of bone shortening in major limb replantation are different from those in digital replantation. The priority is the permission of end-to-end nerve and musculotendinous repair and this may necessitate quite radical resections of 10cm or more. Fixation depends on level and site of the fracture and should be performed along AO guidelines. Despite the shortening, free flaps are frequently necessary to obtain wound closure.

Musculotendinous divisions are repaired in the same way as injuries in a non-replant situation. Side to side tenorrhaphies, secondary tendon transfers, or even secondary free muscle transplants may be indicated.

Adherence to microvascular technical principles is important. Only those vessels with demonstrable flow should be anastomosed. Grafts will almost always be necessary for arterial reconstruction. Because of high flow rates and high pressure demands over many years, the long-term fate of vein grafts in these sites is uncertain and consideration may be given to primary arterial grafts, for example from the thoracodorsal or opposite radial artery.

If any doubt exists about the compartment pressures, prophylactic fasciotomies should be performed. These may include the interossei and a carpal tunnel release in the hand. Systemic reperfusion toxicity and myoglobinuria are theoretical grave risks after revascularization and are proportionate to the length of ischaemia of the part.

The patient should be kept warm, well hydrated, and comfortable with the limb elevated on pillows. Broad-spectrum systemic antibiotics are indicated for prophylaxis. Smoking is totally prohibited given its proven association with vascular spasm and increased risk of thrombosis. Visible access to the region of the amputation site is particularly important to detect early signs of local necrosis, sepsis, or vessel exposure. Given the propensity for myonecrosis in major replants it may be prudent to perform a ‘second-look’ operation after 24–48h in order to ascertain muscle viability.

Monitoring of capillary return, colour, and temperature should be hourly for the first 24h and every 2h for the next 24h. Healthy revascularized tissue should have a capillary return of 2s. Venous congestion is detected by a violet-blue colour and a rapid capillary return. When pricked the part will bleed briskly with dark blue blood. Absolute arterial obstruction is obvious. The digit is pale white and cold with no capillary return, the pulp is empty, and there is absence of bleeding on needle puncture. Partial arterial insufficiency is more difficult to recognize and may be misinterpreted as a venous problem. In this situation the finger may appear blue; however, the capillary return will be very slow or absent, and bleeding on needle puncture will be either absent or a delayed slow ooze of dark blue deoxygenated blood.

The paucity of human studies and conflicting animal studies make decisions regarding anticoagulation arbitrary. Anticoagulation regimens differ from centre to centre and range from no anticoagulation to full therapeutic heparinization varying on a case by case basis. We have experienced numerous cases of digital replantation that have clearly been salvaged by heparin. It is particularly valuable as a static dose intraoperatively in cases where reperfusion is protracted.

The goal of early active movement of replanted fingers, although theoretically tempting, is difficult in practice because of repairs on both extensor and flexor surfaces. Complete immobilization for approximately 3 weeks is the best option in most cases.

Results in the literature are numerous and relatively consistent

Based on a meta-analysis the following results can be expected.

Results of digital replantations can be summarized as follows.

A number of reports of major replantation have been published. In an eloquent analysis the functional results of major replantation were superior to prosthetic reconstruction.

Arterial insufficiency requires prompt return to the operating room and results from inadequate shortening of damaged vessels thus revision is usually with vein grafts. If revision is unlikely to be successful, full systemic heparinization, warming, and placing the hand below heart level may be beneficial, but expectations should be guarded. Late vascular occlusion is rare but may occur in proximal amputation due to intimal hyperplasia.

This is rapidly lethal and early detection is vital for salvage. Relative congestion may be relieved by elevation above heart level, but if this is not successful then exploration and revision should be undertaken. If this is not possible then either medicinal or ‘chemical’ leeching may be used.

Leech bleeding is due to release of hirudin which is a potent anticoagulant. Leeches should be applied as required when the digit takes on a bluish hue. Tetracycline or ceftriaxone should be administered to prevent Aeromonas hydrophila infection which is resident in the gut of H. medicinalis

This technique involves making a chevron-shaped pulp incision and the injection of 500IU heparin into the subcutaneous tissue. Regular milking (every 30min) is required to ensure continued bleeding. Injections need to be repeated several times daily.

Despite the insalubrious manner in which many amputations occur, frank wound or bone infection is surprisingly rare. Haematoma and necrotic skin margins are the harbingers of infection which will threaten vascular patency. Thrombosis occurring after 4–5 days is ominously infective in origin and is generally unsalvageable, highlighting the essential need to obtain primary healing of skin. Early intervention with flap coverage is the only hope.

It is very important to make detailed notes of any variation in neurovascular anatomy that has been created by the initial surgery to minimize risk during essential secondary surgery. Revision surgery on replanted parts is the most technically challenging of all hand surgery and may be threatening to the survival of the part. Arthrolyses and tenolyses are frequently unrewarding. Therefore it is essential to perform optimal repairs at the time of the initial operation, even to the extent of primary nerve grafting if gaps are present. Silicone joints and tendon rods may also be considered to avoid reoperation through the replanted area, even though this risks infection.

The incidence of cold intolerance is initially high but improves with time. In our practice only a very small number of patients have sought reamputation although it has been a major problem in colder climates. It is highly likely to be due to the injury and not the amputation pre se.

In replantation opportunity strikes but once and it is of little use to realize later that some part of a non-replantable limb could have been used for an imaginative reconstruction. The durable sensate weight-bearing skin from the sole of the foot is ideal for resurfacing denuded upper tibial skeletal structures (Figure 12.27.6) and thereby avoiding above-knee amputations

The ultimate dream for the replantationist is the transfer of human limb allografts. Surgical enthusiasm is currently in a state of flux following recent events. The first transplant was rapidly rejected. A more recent French attempt led to a second wave of hand transplants around the world.

There is precise documentation of the outcomes of at least 25 cases over several years. Most have had complications from the immunosuppression including transient rejection episodes, diabetes, and avascular necrosis of the hip; however, this has necessitated only two secondary amputations. The spectacular functional results from bilateral cases as well as a face transplant have tempered the chorus of objections from early hardliners. The achievement of lymphohaematopoietic chimerism is the Holy Grail as it will allow composite tissue transplantation without the inherent risks of immunosuppression.

In the longer term xenotransplantation of joints, muscles, tendons, and nerves from transgenic pigs may have a role in secondary reconstruction.