7

Massive bone defects of the upper limb:

reconstruction by vascularized bone
transfer
Michael B Wood

Introduction

The use of free microvascular bone segment transfer has become an accepted and, in
many instances, the preferred technique for reconstructing massive defects of bone. This
is true in either the upper or the lower limb. The advantages of this technique are
suggested to be more rapid bone union, more rapid bone hypertrophy and fewer instances
of late stress fracture or osteolysis. In addition, some authors suggest that microvascular
autografts may be used under adverse surgical conditions including a septic field or in an
irradiated field. The disadvantages of this technique are that microvascular bone transfer
is a technically demanding procedure and requires a donor site that may result in some
degree of morbidity. This chapter will focus on the indications and technique of
microvascular bone transfers in the upper limb, with illustrative case examples.

History

The suggestion of at least partial sustained viability of cellular elements within bone
segments transferred with an intact soft tissue vascular pedicle dates to at least a century
ago. Huntington in 1905 described the successful healing of a large tibial defect by a
pedicled shift of the ipsilateral fibula (Huntington 1905). Bone segments based on an
intact vascularized soft tissue pedicle lacked widespread clinical application, however,
because they were limited by the arc of rotation of the donor bone segment. It was not
until the clinical feasibility of microvascular anastomosis was demonstrated in the early
1960s that the concept of free vascularized bone grafting emerged. The earliest
experimental work using the rib as the model was by McCulloch in 1973 (McCulloch and
Fredrickson 1973). This was followed by the more comprehensive work of strup and
Fredrickson (1974). Several investigators subsequently confirmed earlier findings of at
least partial preservation of intraosseous cellular elements, a mechanism of bone union
more similar to fracture union than non-vascularized autograft incorporation and more
rapid bone remodeling (Doi et al 1977, Haw et al 1978, Berggren et al 1982, Arata et al
1984, Goldberg et al 1987, DeBoer and Wood 1989, Siegert and Wood 1990). Even
today, however, there remains a controversy about what the most important advantage of
a microvascular bone transfer iswhether it is retained intraosseous cellular viability or
 Severe traumatic defects of the upper limb 86

whether it is the immediate re-establishment of intraosseous blood flow which permits

immediate reseeding of the bone segment with osteoprogenitor cells.
Clinical applications of microvascular bone transfer have been reported over the past
three decades. For long bone reconstruction, the fibula is clearly the preferred donor site.
Taylor is credited with the first report of a successful fibula transfer in 1975 (Taylor et al
1975). However, Ueba reported in 1983 what seems to be the first actual successful
clinical application (Ueba and Fuyikawa 1983). Regardless of who was the first to carry
out free fibular transfer, in the past 20 years, numerous large series have been reported
that have confirmed the value of this technique for reconstructing massive bone defects
(Gilbert 1979, Weiland et al 1979, 1983, Weiland 1981, Taylor 1983, Osterman and Bora
1984, Dell and Sheppard 1984, Wood and Cooney 1984, Wood et al 1984, 1985, Pho et
al 1985, Wood 1986, 1987, Gidumal et al 1987). The authors personal series, reported
by Han et al (1992), resulted in an overall primary union rate of 61% and a secondary
union rate of 81%, with the best results occurring in non-septic reconstructions with a
union rate of 84%.

Upper limb bone defectindications for vascularized bone transfer

In general, most authors suggest that the strongest indications for the use of vascularized
bone graft include situations that are prone to failure or complications with technically
less demanding techniques such as non-vascularized bone autografts or allografts. These
situations in general include massive defects and/or an unfavorable surrounding soft
tissue milieu related to prior bone grafting failure, infection, radiation or other causes of
extensive scarring.

Specific indications

Recipient site considerations

In the upper limb, humerus reconstruction probably represents the most compelling
indication for the use of vascularized fibula transfer. This is because, other than massive
allografts, there are few techniques available to reconstruct a large missing segment of
the humeral shaft. Although large defects of the radius and ulna may also be excellent
indications for reconstruction by vascularized bone transfer, the option of forearm
salvage by a one-bone forearm conversion (Fig. 1) always merits consideration (Peterson
et al 1995). Moreover, more limited defects of the proximal radius or distal ulna may be
consistent with an acceptable level of upper limb function.
 Massive bone defects of the upper limb 87

Figure 1
One-bone forearm reconstruction for en bloc resection of
proximal radius.

Large bone defects

The precise length of a bony defect which would lead one to select a vascularized bone
graft for reconstruction is not particularly well established. Many authors (Taylor et al
1975, Weiland 1981, Osterman and Bora 1984, DeBoer and Wood 1989, Han et al 1992)
have suggested that a 6 cm gap is the point where vascularized bone reconstruction
 Severe traumatic defects of the upper limb 88

should be chosen in place of a non-vascularized autograft. However, it is important to

recognize that with sufficient mechanical protection over several months, and when
dealing with a well vascularized surrounding soft tissue milieu, bone defects exceeding
10 cm may heal with cancellous autograft (Nicoll 1956) or non-vascularized cortical bone
segments (Enneking et al 1980). Moreover massive allografts may be a suitable option
for reconstructing very lengthy defects (Mankin et al 1987). However, it should be
recognized that massive allografts have limited ability to be revascularized and hence a
limited capacity to be replaced by creeping substitution of host osteoprogenitor cells
(Phemister 1914). In general, the author believes that for a defect as short as 6 cm in the
presence of a poor surrounding soft tissue bed and for all defects greater than 10 cm, the
selection of vascularized bone transfer for reconstruction is justifiable.

Prior bone reconstruction failures

Bone defects in the upper limb, without regard to length, which have failed to heal with
nonvascularized autograft may be candidates for a vascularized bone graft. This is
particularly the case when there is no readily apparent explanation for the initial failure
(i.e. inadequate bone graft material, inadequate stabilization, use of allograft or xenograft,
etc.).

Infected bone defects

The use of vascularized bone grafts for reconstructing infected bone defects is
particularly attractive for a number of reasons. Probably the most important fact is that
such bone grafts are inherently a vehicle for local blood supply (Dell and Sheppard 1984,
Wood and Cooney 1984). However, also of importance is the fact that a vascularized
fibula is a generous source of bone length and it makes little difference from the technical
perspective if one transfers a 6 cm or 16 cm graft segment. Thus, a more aggressive
debridement of infected bone ends may be carried out with less concern about creating a
larger bone defect than can be reconstructed.

Non-unions associated with bone radionecrosis

Bone non-union that is associated with radiation osteonecrosis is a particularly
challenging problem that responds poorly to conventional bone grafting techniques
(Duffy 2000). This is because three adverse circumstances exist in the presence of
localized radiation changes: (1) impaired intraosseous blood supply; (2) impaired blood
supply of the surrounding soft tissue; (3) periosteal and intraosseous cell death. These
adverse circumstances are directly addressed by the transfer of vascularized bone
segment obtained from a site well distant to the irradiated field.

Donor sites
The focus of this chapter is repair of large bone defects of the upper limb. For all
practical purposes, for these types of defects, the fibula is the preferred donor bone.
Rarely is the fibula unavailablefor example, if both were previously harvested for bone
 Massive bone defects of the upper limb 89

grafting or in unique patients with osteogenesis imperfecta. The technique of vascularized

fibula isolation has been well described (Gilbert 1979, Wood 1985) and will not be
repeated here. In the rare patient where the fibula is unavailable, one can consider other
vascularized bone donor sites for example, iliac crest, scapula, rib, radius, metatarsal or
any bone segment from a paralyzed or useless limb.

Surgical technique

Scapulo/humeral defects
Scapulohumeral defects most typically result from extraarticular resections of malignant
bone lesions of the proximal humerus. This situation presents significant technical
challenges in
 Severe traumatic defects of the upper limb 90

Figure 2
(a) Parosteal osteogenic sarcoma of the proximal humerus.
(b) Immediate postoperative radiograph following
extraarticular resection of tumor and reconstruction with
massive proximal humerus allograft and side-by-side
vascularized fibular autograft bridging between scapula
and distal humerus. (c) Radiograph 7 years
postoperatively: note allograft, partial resorption and
progressive fibular hypertrophy. Solid scapulohumeral
union.
 Massive bone defects of the upper limb 91

securing adequate fixation between the end of the transferred fibula and the remaining
scapula. The author has successfully used a compression plate and screws bridging from
the remaining scapular spine to the proximal segment of the fibula. However, due to the
cancellous nature of the scapula and small diameter of the fibula, such fixation requires
the additional use of external fixation using a shoulder spica cast. Because of the
problems associated with immobilization of this type of construct, the author prefers to
combine fibula transfer in this location with a proximal humerus allograft (Fig. 2). Such a
construct better ensures scapula to allograft and allograft to the remaining humerus
fixation by the use of larger screws and plates. The fibula

Figure 3
(a) Immediate postoperative radiograph following en bloc
resection of mid-humerus for chondrosarcoma and transfer
of revascularized fibular autograft. (b) Radiograph 13
months postoperatively. Note union and hypertrophy of
fibular segment to nearly the same diameter of normal
humerus.
 Severe traumatic defects of the upper limb 92

is placed parallel to the allograft with contact proximally into the neck of the scapula and
distally to the remaining humerus, using a transosseous screw at each end. The author
prefers to revascularize the fibula in most patients by end-to-side anastomosis of the
donor bone peroneal artery to the recipient site brachial artery. Venous anastomoses are
usually end-to-end between peroneal venae comitantes and either brachial venae
comitantes or the cephalic vein. It is much easier to perform the vascular anastomoses
and to isolate the recipient vessels more distally in the upper limb. Thus, the fibular
segment should be positioned in a retrograde manner in order to position its vascular
pedicle closer to the elbow.

Figure 4
(a) Radiograph of atrophic non-union of humerusprior
gunshot wound with extensive bone loss. (b) Immediate
postoperative radiograph of vascularized fibula transfer.
Note internal fixation with proximal and distal
compression plates.
 Massive bone defects of the upper limb 93

Humeral diaphyseal defect

A large defect of the diaphysis of the humerus is probably the ideal application of
vascularized fibula transfer, because when successful it results in a functioning shoulder
and elbow joint. Typically defects of the humerus shaft result from penetrating trauma,
infected non-unions, tumor resection or radiation necrosis. Fixation of these constructs is
usually easier than with scapulohumeral defects. Preferentially, and when there is an
adequate remaining length of the proximal and distal humerus, a compression plate or
transosseous screws at each end is employed (Figs 35). If a compression plate is used at
each end screw fixation to at least six cortices (three screws, each with two cortex
purchase) should be obtained on either side of the osteosynthesis site. A single plate
spanning the entire construct should be avoided as it may act as a stress-shield and inhibit
appropriate fibular hypertrophy after healing. Even with compression plate fixation, the
additional protection of a shoulder spica cast should be considered for the first 2
postoperative months. As was mentioned for the scapulohumeral reconstructions,
retrograde positioning of the fibula should be done to facilitate access to the vascular
pedicle for anastomoses. Preferentially end-to-side arterial and end-to-end venous
anastomoses are employed as discussed previously.

Radius/ulna diaphyseal defect

Large defects of either radius or ulna most commonly result from penetrating trauma,
infected non-union or tumor resection. Based on size and shape considerations, the fibula
is a near perfect match for the diaphyseal segment of radius or ulna. Because of their
similarity, post-union hypertrophy of the fibular segment is usually not a significant
issue. Fixation may employ a compression plate and screws at each end or a single long
spanning plate may be utilized. If the latter technique is selected, however, it is important
to avoid screw placement in the central portion of the fibula or near the nutrient foramen.
Moreover, six-cortex fixation is required in both the distal and proximal forearm bone
segments whether one or two plates are used. For reconstruction requiring fixation to the
distal metaphyseal flare of the radius, it is preferable to dowel the fibula well into the
metaphysis of the radius (Fig. 6). Internal fixation by any method should be additionally
protected by the use of a long arm cast or splint for 68 weeks postoperatively. The
fibular segment may be placed orthograde or retrograde depending upon the most
convenient vascular access site. Anastomoses usually employ end-to-end coaptation to
either radial or ulnar arteries, provided the second vessel is patent and the superficial
palmar arterial arch permits adequate flow to all digits with the selected recipient artery
occluded. As a final comment, whenever one is considering the use of free vascularized
fibula for reconstructing a defect of either radius or ulna, the possibility of developing a
rather refractory radio-ulnar synostosis should be weighedespecially if the simpler
approach of a one-bone forearm construct is considered a viable option for the patients
functional needs and expectations.
 Severe traumatic defects of the upper limb 94

Forearm-carpal defects
Forearm-carpal defects most often result from either penetrating trauma or aggressive
tumors, especially recurrent giant cell tumor of the distal radius. In this situation the usual
goal is to obtain a stable wrist arthrodesis (Fig. 7). The technique of vascularized bone
transfer in this area is essentially identical to that of reconstructing forearm defects, with
the exception of distal osteosynthesis fixation. It can be technically challenging to obtain
secure fixation to either the carpal bones or the metacarpals and the selection of the best
form of internal fixation will thus differ with the unique circumstances of each patient.
The author has utilized mini-plate fixation, screws alone, Kirschner wires and cerclage
wiring. In all patients firm cast support is recommended until union is confirmed.

Conclusion

Vascularized bone transfer is increasingly recognized as a very useful and versatile

technique for
 Massive bone defects of the upper limb 95

Figure 5
(a) Radiograph of chronic non-union associated with
radionecrosis of humeral diaphysis for Ewings sarcoma.
(b) Immediate postoperative radiograph following
resection of radionecrotic humerus and transfer of
vascularized fibula autograft. (c) Radiograph of healed
humerus reconstruction 58 months postoperatively. Note
hypertrophy and absence of radionecrosis of humerus.
 Severe traumatic defects of the upper limb 96

Figure 6
(a) Radiograph of radiusrecurrent adamantinoma post
curettage at another institution. (b) Immediate
postoperative radiograph of en bloc resection of portion of
radius and transfer of vascularized fibula autograft. Note
intramedullary placement into metaphysis of radius. (c)
Radiograph 3 months postoperatively demonstrating union
of fibular autograft proximally and distally. (d) Radiograph
38 months post fibular transfer and 30 months post internal
fixation plate removal demonstrating fibula incorporation
identical to normal appearing radius.
 Massive bone defects of the upper limb 97

Figure 7
(a) Radiograph of recurrent aggressive giant cell tumor of
distal radius invading proximal carpal row. (b) Immediate
postoperative radiograph following extraarticular en bloc
resection for aggressive, recurrent giant cell tumor of the
distal radius. Note distal fixation with multiple Kirschner
wires. (c) Radiograph 13 months postoperatively with
union proximally and distally and modest fibular
hypertrophy.
 Severe traumatic defects of the upper limb 98

reconstructing massive bone defects in the upper limb or in patients with especially
challenging conditions, such as infected non-unions and nonunions associated with
radionecrosis of bone. It is especially indicated for the humerus and shoulder region with
more selected applications in the forearm or wrist. Though technically challenging, the
outcomes of this procedure justify consideration along with alternative methods for major
reconstructions of the upper limb.

References

Arata MA, Wood MB, Cooney WP (1984) Revascularized segmental bone transfers in
the canine. An analysis of viability, J Reconstruct Microsurg 1: 1119.
Berggren A, Weiland AJ, Dorfman H (1982) The effect of prolonged ischemia time on
osteocyte and osteoblast survival in composite bone grafts revascularized by
microvascular anastomoses, Plast Reconstr Surg 69:2908.
DeBoer NH, Wood MB (1989) Bone changes in the vascularized fibula graft, J Bone
Joint Surg 71B:3748.
Dell PC, Sheppard JE (1984) Vascularized bone grafts in the treatment of infected
forearm nonunions, J Hand Surg 9A:6538.
Doi K, Tominaga S, Shubata T (1977) Bone grafts with microvascular anastomoses of
vascular pedicles: an experimental study in dogs, J Bone Joint Surg 59: 80915.
Duffy GP, Wood MB, Rock MG, Sim FH (2000) Vascularized free fibular transfer
combined with autografting for the management of fracture nonunions associated with
radiation therapy, J Bone Joint Surg 82A:54454.
Enneking WF, Gady GL, Burchardt H (1980) Autogenous cortical bone grafts in the
reconstruction of segmental skeletal defects, J Bone Joint Surg 62A: 103958.
Gidumal R, Wood MB, Sim FH et al (1987) Vascularized bone transfer for limb salvage
and reconstruction after resection of aggressive bone lesions, J Reconstr Microsurg
3:1838.
Gilbert A (1979) Vascularized transfer of the fibular shaft, Int J Microsurg 1:1002.
Goldberg VM, Shaffer JW, Field G, Davy DT (1987) Biology of vascularized bone
grafts, Orthop Clin North Am 18:197205.
Han CS, Wood, MB, Bishop AT, Cooney WP (1992) Vascularized bone transfer, J Bone
Joint Surg 74A: 14419.
Haw CS, OBrien B Mc, Kurata T (1978) The microsurgical revascularization of resected
segments of tibia in the dog, J Bone Joint Surg 60:2669.
Huntington TW (1905) Case of bone transference. Use of a segment of fibula to supply a
defect in the tibia, Ann Surg 41:249.
Mankin HJ, Gebhardt MC, Tomford WW (1987) Use of frozen cadaveric allografts in the
management of patients with bone tumors of the extremities, Orthop Clin North Am
18:27589.
McCulloch DW, Fredrickson JM (1973) Neovascularized rib grafts to reconstruct
mandibular defects, Can J Otolaryngol 2:96100.
Nicoll GA (1956) The treatment of gaps in long bones by cancellous insert grafts, J Bone
Joint Surg 38B:7082.
 Massive bone defects of the upper limb 99

Osterman AL, Bora FW (1984) Free vascularized bone grafting for large-gap nonunion
of long bones. Orthop Clin North Am 15:13142.
strup LT, Fredrickson JM (1974) Distant transfer of a free living bone graft by
microvascular anastomosis: an experimental study, Plast Reconstr Surg 54:27485.
Peterson CA, Maki S, Wood MB (1995) Clinical results of the one-bone forearm, J Hand
Surg 20A:60918.
Phemister DB (1914) The fate of transplanted bone and regenerative powers of its
various constituents, Surg Gynecol Obstet 19:30333.
Pho RW, Levack B, Satku K, Patradul A (1985) Free vascularized fibula graft in the
treatment of congenital pseudarthrosis of the tibia, J Bone Joint Surg 67B:6470.
Siegert JJ, Wood MB (1990) Blood flow evaluation of vascularized bone transfers in a
canine model, J Orthop Res 8:2916.
Taylor GI (1983) The current status of free vascularized bone grafts, Clin Plast Surg
10:185209.
Taylor GI, Miller GDH, Ham FJ (1975) The free vascularized bone graft: a clinical
extension of microvascular techniques, Plast Reconstr Surg 55:53344.
Ueba Y, Fuyikawa S (1983) Nine years follow-up of a vascularized fibular graft in
neurofibromatosis: a case report and literature review, Orthop Trauma Surg 26:595.
Weiland AJ (1981) Current concepts review. Vascularized free bone transplants, J Bone
Joint Surg 63A:1669.
Weiland AJ, Kleinert HE, Kutz JE, Daniel RK (1979) Free vascularized bone grafts in
surgery of the upper extremity, J Hand Surg 4:12944.
Weiland AJ, Moore JR, Daniel RK (1983) Vascularized bone autografts: experience with
41 cases, Clin Orthop 174:8795.
Wood MB (1986) Free vascularized bone transfers for nonunion, segmental gaps, and
following tumor reconstruction, Orthopedics 9:81016.
Wood MB (1987) Upper extremity reconstruction by vascularized bone transfers: results
and complications, J Hand Surg 12A:4227.
Wood MB (1990) Free osseous tissue transfer. In: Atlas of Reconstructive Microsurgery.
Aspen Publications: Rockville, 6591.
Wood MB, Cooney WP (1984) Vascularized bone segment transfers for management of
chronic osteomyelitis, Orthop Clin North Am 15:46172.
Wood MB, Cooney WP, Irons GB (1984) Posttraumatic lower extremity reconstruction
by vascularized bone graft transfer, Orthopedics 7:25562.
Wood MB, Cooney WP, Irons GB (1985) Skeletal reconstruction by vascularized bone
transfer. Indications and results, Mayo Clin Proc 60:72934.