6.12 Tumours and hand reconstruction

A multidisciplinary approach which employs improvements in diagnosis, staging, adjuvant therapy, surgical resection, and particularly reconstruction has made limb salvage a viable option in the treatment of most upper-extremity malignancies. The adequacy of surgical resection should be paramount. However, reconstructive considerations should be contemplated during the phase of preoperative planning in order that the best possible functional and aesthetic outcome is achieved. Emphasis should be placed on a thorough evaluation by all members of the treatment team prior to the initiation of treatment. A broad range of techniques are available for both salvage and restoration of function.

In principle, the management of tumours of the bone and soft tissues in the hand and wrist differ little from those encountered elsewhere, except that vital neurovascular structures are more commonly juxtaposed and there is an inherent desire to preserve upper-extremity function. Certain tumours occur more frequently, e.g. ganglion cysts, epidermoid inclusion cysts, and giant cell tumours of the tendon sheath. Conversely, tumours common in other locations present rarely, e.g. metastatic carcinoma, lymphoma, and myeloma. The following guidelines may prove useful in the approach to the patient presenting with a mass in the hand:

The approach to tumours of the hand and wrist incorporates the following principles:

A multidisciplinary approach to management is mandatory for malignant tumours and some, typically larger, benign tumours. The initial goals of management include diagnosis and staging of the tumour prior to biopsy. The biopsy should be planned with the reconstruction in mind.

Patients presenting with a bony mass may be investigated by means of plain radiographs, computed tomography (CT), and magnetic resonance imaging (MRI). The usual criteria for the radiographic analysis of bone tumours include size, shape, location, gross architectural alterations, disturbance of the bony cortex, and the radiological appearance of the matrices of these tumours.

Most tumour surgeons employ the modified Enneking staging system for malignant neoplasms of bone (Table 6.12.1). This is generally also used for staging soft tissue tumours. This staging scheme depends on three primary determinants: the grade of the tumour (G), the anatomical location of the lesion (T), and the presence or absence of distant metastases (M). Benign tumours are usually designated G0, with low-grade malignant lesions classified as G1 and high-grade malignant lesions as G2.

The anatomical location is assessed by imaging studies including CT and MRI. The essential definition is whether there is preservation (T1) or violation (T2) of the involved compartment. For bony tumours, the involved bone fulfils the definition of a compartment, such that extraosseous extension is considered a T2 lesion. Soft tissue tumours that extend beyond the fascial compartment or involve the neurovascular bundle are considered to be T2 tumours.

For giant cell tumours of bone and chondrosarcoma, Campanacci (1976) has devised a grading system (Table 6.12.2). Giant cell tumours are graded into three radiographic types (calm, active, aggressive) and three histological types (typical, aggressive, sarcoma). Chondrosarcomas are also graded into three radiographic and three histological types (grades I, II, III). Cutaneous malignancies such as melanoma have their own unique staging scheme.

The American Joint Cancer Committee Staging System is usually employed in staging melanoma. This system is based primarily on tumour thickness and the presence of lymphatic metastasis. Tumour thickness and ulceration are the major determinants of prognosis. Subungual melanoma represents a specific clinical entity comprising 1–3% of all presenting melanomas. In these patients, tumour thickness is often difficult to determine due to fixation and biopsy techniques.

The approach to patients presenting with either bone or soft tissue tumours includes assessment of the grade of the tumour and the anatomical extent of the lesion, and determination of metastatic spread. The tumour is graded on the histopathology including special stains and even electron micrography. A working clinical diagnosis can usually be formulated on the basis of the physical examination, plain radiographs, and blood tests. This should enable the clinician to distinguish between benign tumours, primary malignant tumours, metastatic tumours, and lymphomatous deposits.

If the lesion is thought to be benign, an excisional biopsy without further workup is appropriate. Frozen section should be employed intraoperatively if there is any doubt. If the lesion is thought to be a primary malignant tumour, further imaging studies (CT or MRI) should be performed to stage the tumour accurately. MRI of the hand and wrist is accurate in the detection of mass lesions and can distinguish benign from malignant tumours in the most cases. In the case of suspected metastatic tumour, appropriate imaging of the probable primary source should be ordered. The focus of this search may be appropriately guided by the results of the biopsy.

As a rule, neoplasms of the connective tissues spread locally by invasion of the adjacent vascular and connective tissues rather than by invasion of lymphatics. The adjacent normal tissues become compressed and may demonstrate a fibroproliferative response. This compressed fibroproliferative layer is known as the ‘reactive zone’. Histological analysis of the reactive zone will frequently demonstrate microscopic tumour invasion. Discrete nodular metastases formed within the same compartment are known as skip lesions. These are more frequent with high-grade neoplasms such as osteosarcoma. Several studies have demonstrated the correlation of survival and local recurrence with the adequacy of local control. Based on the proximity of the surgical margin to the tumour and the pattern of tumour spread within the compartment, Wolf and Enneking have devised four types of surgical procedure:

There are no set guidelines for deciding when each of these four procedures should be applied. Table 6.12.3 gives a reasonable approach based on the tumour grade.

Factors that may affect this decision include the age and wishes of the patient, functional demands on the limb, and adjunctive chemotherapy or irradiation. At operation, it is vitally important to limit the dissection, and to observe meticulous haemostasis, to limit potential seeding of tumour cells. The biopsy and subcutaneous tract should be included in the specimen. A bone weakened by resection requires protection or definitive reconstruction. If the reconstructive procedure requires the harvest of autogenous tissue, the two surgical sites should be kept separated until such time as the specimen has been removed.

Finally, the entire ablative procedure should be performed only after careful planning and preoperative consultation with the oncologist and the pathologist. Only after an oncologically sound plan has been formulated should consideration be given to the restoration of limb function and form.

Primary bone tumours of the hand and wrist are uncommon. Of 1046 primary bone tumours in the Netherlands Bone Registry, only 64 (5.8%) occurred in the hand and 3 (0.2%) in the wrist (Table 6.12.4). Of the 420 benign lesions, 53 (12.6%) were in the hand and 3 (0.7%) were in the wrist. No malignancies involved the carpus, whilst eight (1.2%) were in the hand.

The classification of tumours is given in Table 6.12.5.

Postoperative adjuvant radiation therapy is usually employed in the management of malignant tumours when the surgical margin is narrow or for high-grade malignancy. Adjuvant radiation therapy should be considered in sarcomas larger than 6cm and where there is vascular invasion because of the poorer prognosis in these patients. Local recurrence of sarcoma with marginal resection and radiation is greater than with resection alone. The use of adjuvant radiation therapy diminishes local recurrence in extremity sarcomas. The use of preoperative adjuvant radiation therapy can be useful for diminishing tumour size prior to excision. In some patients presenting with sarcomas in the hand and forearm, adjuvant radiation therapy may facilitate limb-salvage surgery as an effective alternative to amputation.

More proximal lesions requiring division of branches of the brachial plexus should warrant a frank discussion with the patient as to the merits of amputation. Slow-growing soft tissue tumours seldom require additional skin coverage. The concomitant expansion of the adjacent tissues will usually ensure that skin closure can be achieved without difficulty. Reconstruction is more often required for rapidly growing tumours or where the provision of an adequate margin of resection dictates a greater need for skin coverage.

Flap coverage following resection of small tumours in the digits is mandatory if such resection results in exposure of neurovascular structures or tendons. However, where the paratenon covering a particular tendon is intact, a full-thickness skin graft will often provide ideal coverage with minimal donor site morbidity. Where skin grafting is not an option, a local digital flap will provide good coverage of small defects. In principle, these are either harvested from the same finger (homodigital) or an adjacent finger (heterodigital). The use of heterodigital tissue for the reconstruction of a skin defect is only recommended following excision of benign tumours.

Amputation of the thumb constitutes a significant functional loss, rendering reconstruction a priority. The major goal of reconstruction is to provide a stable thumb of sufficient strength and length in the correct position. Lesser priorities for the reconstructed thumb are motion, sensibility, and appearance. Reconstructive alternatives following amputation of the thumb depend primarily on the length of the thumb remnant and the presence of the basal joint.

In subtotal amputations, metacarpal lengthening by distraction may provide adequate length. In total amputations, the options include osteoplastic reconstruction, pollicization, or toe-to-hand transfer. Which of these options is chosen depends on patient factors, such as compliance, suitability for microvascular surgery, and the level of the amputation, as well as surgical preference. Reconstruction of the thumb with tissue derived from the great toe gives the best results in patients suited to this procedure. This is by far the superior technique for thumb reconstruction. The tissue derived from the foot can be used in various ways. These include resurfacing of the skeletal thumb by the wrap-around technique, the trimmed great toe transfer, which minimizes the bulk of the great toe, and transfer of the second toe. Pollicization can yield excellent results, but in adults it is usually best reserved for the patient with pre-existent partial index finger amputation. Osteoplastic reconstruction, using a groin flap combined with bone grafting, is suitable for patients who are not candidates for microsurgery.

In general the larger defects created by tumour ablation over the dorsum of the hand require flap coverage that cannot be satisfied by local tissue alone. One exception is where a fillet flap can be harvested from a digit to resurface a dorsal or palmar skin defect. The options include pedicled flaps from the groin or chest wall, regional flaps from the ipsilateral forearm, and free tissue transfers.

Based on the superficial circumflex iliac vessels, the groin flap has proved reliable for the resurfacing of dorsal hand defects. However, the use of the pedicled groin flap is limited to reconstruction following benign tumour resection. Although reliable, it commits the patient to a period of inactivity whilst the pedicle is attached, with the potential for stiffness in the small joints of the hand. Its use as a free flap obviates the need for a pedicle.

Flaps derived from the forearm include those based on fascioseptocutaneous perforating branches from the radial artery, the ulnar artery, and the posterior interosseous artery. They are referred to as the distally based radial forearm, ulnar forearm, and posterior interosseous flaps respectively, and are indicated for reconstruction following removal of benign tumours. Free tissue transfer is a better option for malignant tumours. Inherent in the use of local tissues for resurfacing skin defects is the concern about tumour seeding in the flap donor site. In the case of the radial and ulnar forearm flaps, there is the additional concern of deprivation of a major source of arterial inflow in the extremity. Donor site morbidity in the form of delayed healing is common, particularly where primary closure of the donor site is not possible. The posterior interosseous flap, although probably the least reliable of the group, has significantly less donor site morbidity.

The ideal flap would provide durable coverage with the potential for a gliding environment without being too bulky. Broadly, the options include free muscle, skin, or fascial transfers. The success rate for free tissue transfer in large series now approaches, or equals, 100%. Cost analysis has demonstrated a significant saving with free tissue transfers compared with pedicled flaps.

There are many options with their own particular merits and limitations. They include: the gracilis and serratus anterior as small to medium free muscle flaps; the latissimus and rectus abdominis as large muscle flaps; the contralateral ulnar flap, the free groin flap, and the scapular flap as skin flaps; and the serratus anterior fascia, the superficial temporal fascia, the lateral arm fascia, and the scapular fascia as fascial flaps.

The uniquely sensate glabrous skin of the palm is ideally suited to prehension. The particular aim of reconstruction in the palm should be sensate skin which is of similar volume, texture, and colour. For smaller defects, a large variety of local pedicled or island flaps can be applied. The reconstruction of large palmar defects of the hand remains a difficult problem because of the specificity of the anatomic structures and the highly sophisticated function of the palm. Free tissue transfer remains the preferred alternative for larger defects with exposed tendons, nerves, or other essential structures. The medial plantar flap provides the capability of transposing glabrous skin from the non-weight-bearing instep of the foot to the palm of the hand. Occasionally, owing to the relative abundance of subcutaneous tissue in the palm, vital structures may be protected such that a full-thickness or thick split-thickness skin graft from the sole of the foot will suffice.

The treatment of benign tumours of the phalanges is usually intralesional treatment involving curettage only or curettage and bone grafting. Amputation is usually indicated for benign tumours with extraperiosteal spread or for malignant tumours. The rate of metastasis of chondrosarcomas presenting in the phalanges is very low, unlike those occurring elsewhere in the body. Treatment of this locally aggressive lesion may be by curettage rather than amputation if the latter may result in significant functional loss.

Reconstruction is generally required for tumours of the metacarpals, the distal radius, and the distal ulna that fall into one of the following three categories:

This applies to tumours in stages 2 or 3 of the Enneking grading system, or grades 2 or 3 of the Campanacci radiological grade.

The distal radius is a relatively common site for skeletal neoplasms and is the third most common site of giant cell tumours. Resection of the distal radius is indicated for malignant neoplasms, as well as for locally recurrent or invasive benign tumours. Many patients who fall into this category are relatively young and have high functional demands. The limited soft tissue coverage in this area and the proximity of neurovascular structures and tendons add to the difficulty in achieving clear marginal clearance of tumours of the distal radius. The pronator quadratus is a useful watershed on the volar aspect of the radius. Curettage remains the mainstay of management in cases where the tumour is confined within the radius. The resultant intramedullary bone defect can be filled with autogenous cancellous bone or methylmethacrylate cement, or it can be left unfilled. Cryosurgery or phenol instillation is used in an attempt to reduce the incidence of local recurrence. Giant cell tumours are particularly prone to recurrence after curettage, especially in the distal radius where recurrence rates of 50% have been reported. Malignant change in recurrent giant cell tumour may be associated with irradiation. For these reasons, en bloc resection and reconstruction of the distal radius may be necessary.

In the reconstruction of the distal radius, the surgeon needs first to decide whether or not to retain wrist motion. Factors contributing to this decision include the proximity of the tumour to the joint, the tumour grade, and patient preference. The reconstructive options that retain wrist motion include excision arthroplasty, prosthetic replacement, ulnar translocation of the carpus, allograft replacement, and the use of a non-vascularized fibular bone graft or a vascularized fibular bone graft. Depending on the nature of the bony defect, and if a decision has been made to sacrifice wrist motion, a vascularized or non-vascularized bone graft will usually be required at the time of the arthrodesis. However, recent reports have focused on the use of either osteoarticular allograft or vascularized bone grafts in the management of this difficult reconstructive problem.

The use of cadaveric osteoarticular allograft employs a size-matched fresh-frozen distal radial allograft. The advantages of this technique include the ability to provide an exact three-dimensional reconstruction and to reconstruct larger defects, including ligamentous reconstruction, and the absence of a donor site. Despite the encouraging results with this technique, non-vascularized autogenous bone grafts are predisposed to delayed union, stress fracture, resorption, and collapse.

The primary advantage of vascularized autogenous bone grafts is a lower fracture rate compared with large non-vascularized bone grafts or cadaveric bone grafts. Vascularized bone has a greater propensity to hypertrophy as well as a diminished tendency to non-union. Osteotomy sites heal in a fashion more akin to fracture healing than bone-graft incorporation. Furthermore, bony healing occurs irrespective of the dimensions of the vascularized bone segment. An additional advantage in the use of vascularized bone is the facility to provide skin as part of a composite skin–muscle reconstruction. In order to reduce the fracture rate further, the fibula can be double-barrelled for additional strength and bulk. When harvested correctly, the donor site morbidity should not be significant. When used for upper-extremity reconstruction, the vascularized fibula may not be capable of the hypertrophy seen in the lower extremity, particularly where the fixation is rigid.

As for the distal radius, T1 lesions of the distal ulna may be treated by curettage and bone grafting. Tumours that are not amenable to curettage may be resected en bloc with good functional results. Where the tumour extent allows subperiosteal resection, the ulnar collateral ligament and triangular fibrocartilage complex are amenable to reconstruction. This will significantly reduce the incidence of instability following resection. An ulna-stabilizing procedure may be necessary for the patient with demonstrable ulna instability following resection.

Standard systems have been derived for the evaluation of outcome following reconstructive procedures after tumour excision. It is recommended that these are used to facilitate comparison of reconstructive techniques.