Abstract
Background: Treatment of the loss of the distal part of the radius,
including the physis and epiphysis, in a skeletally immature patient requires
both replacement of the osseous defect and restoration of longitudinal growth.
Autologous vascularized epiphyseal transfer is the only possible procedure
that can meet both requirements.
Methods: Between 1993 and 2002, six patients with a mean age of 8.4
years (range, six to eleven years) who had a malignant bone tumor in the
distal part of the radius underwent microsurgical reconstruction of the distal
part of the radius with a vascularized proximal fibular transfer, including
the physis and a variable length of the diaphysis. All of the grafts were
supplied by the anterior tibial vascular network. The rate of survival and
bone union of the graft, the growth rate per year, the ratio between the
lengths of the ulna and the reconstructed radius, and the range of motion of
the wrist were evaluated for five of the six patients who had been followed
for three years or more.
Results: The mean duration of follow-up of the six patients was 4.4
years (range, eight months to nine years). All six transfers survived and
united with the host bone within two months postoperatively. The five patients
who were followed for three years or more had consistent and predictable
longitudinal growth. Serial radiographs revealed remodeling of the articular
surface. The functional result was rated as excellent for all but one patient,
in whom the distal part of the ulna had also been resected because of
neoplastic involvement. No major complication occurred at the recipient site,
whereas a peroneal nerve palsy occurred at the donor site in three patients.
The palsy was transient in two patients, but it persisted in one. No
instability of the knee joint was observed.
Conclusions: After radical resection of the distal part of the
radius because of a neoplasm in children, vascularized proximal fibular
transfer, based on the anterior tibial artery, permits a one-stage skeletal
and joint reconstruction, provides excellent function, and minimizes the
discrepancy between the distal radial and ulnar lengths.
Level of Evidence: Therapeutic study, Level IV (case
series [no, or historical, control group]). See Instructions to Authors for a
complete description of levels of evidence.
Reconstruction of the wrist in children following resection of the distal
part of the radius for oncologic reasons is a challenge. No traditional
procedure can replace the wrist joint and simultaneously restore function and
physiologic growth. In this anatomic region, failure of growth not only means
a cosmetically unacceptable discrepancy between the lengths of the upper
extremities, but also can lead to progressive functional impairment due to the
continued growth of the ulna with subsequent development of ulnocarpal
impingement and radial deviation of the hand resembling a radial clubhand.
It is not possible to prevent such a deformity with non-vascularized
autografts, allografts, or a prosthesis in a growing child. Subsequent
deformity can be treated by lengthening of the radius or shortening of the
ulna, or both, at the conclusion of growth. We hypothesized that
reconstruction following surgical resection of the distal part of the radius
in children may be optimally done with a vascularized proximal fibular graft
that includes its physis and epiphysis. Such a graft has the potential for
growth and contains an articular surface similar to the distal part of the
radius that can provide an effective joint surface.
The aims of this study were to validate the use of the anterior tibial
artery as the vascular pedicle of the proximal part of the fibular graft, to
evaluate the growth rate of the transferred epiphysis, and to assess the
radiographic and functional results of this procedure.
Between 1993 and 2002, six children with a mean age of 8.4 years (range,
six to eleven years) underwent limb salvage surgery for osteogenic or Ewing
sarcoma involving the distal part of the radius
(Table I). All but one of the
patients was followed for longer than two years. The mean duration of
follow-up was 4.4 years (range, eight months to nine years) for all six
patients and 5.2 years (range, three to nine years) for the five patients
followed for three years or more.
The study was approved by our institutional review board, and the parents
of all patients gave their informed consent.
Preoperative and postoperative chemotherapy was administered to all six
patients according to the different protocols for osteogenic
sarcoma1,2
and Ewing sarcoma3
being utilized at the time of diagnosis. All resections involved the distal
part of the radius and a variable amount of the diaphysis. The distal part of
the adjacent ulna was resected in one patient (Case 2). The resulting distal
radial bone defect was reconstructed with a transfer of the proximal part of
the fibula, including its physis and epiphysis, on a vascular pedicle of the
anterior tibial artery. Preoperative angiography was routinely performed in
order to assess the location and the size of the recurrent epiphyseal branch,
which must be preserved in order to provide an adequate blood supply to the
proximal part of the fibula (Fig.
1). This branch is a constant vessel and was identified in all
patients.
The proximal fibular graft had a mean length of 10.2 cm (range, 7 to 13.5
cm) and was fixed to the host bone by a plate and screws in five patients and
with a step-cut osteotomy and screws in the sixth patient. The vascularity of
the transferred fibula was confirmed by the rapid bone-healing that occurred
within two months postoperatively.
The patients were followed periodically with standard anteroposterior and
lateral radiographs to evaluate the growth rate and the morphological changes
of the articular surface of the proximal fibular epiphysis in its new
anatomical location. The progressive growth of the transferred fibula was
evaluated by measuring the distance between two fixed landmarks: the tip of
the epiphysis and the distal end of the metal plate, or the most distal screw
in the one patient in whom only screws were used. Three-dimensional computed
tomography was used in patients with more than four years of follow-up to
demonstrate the progressive remodeling of the articular surface of the
proximal fibular epiphysis.
The functional outcome was evaluated by assessing the range of motion of
the wrist joint, the sensation of the hand, and the grip strength.
Postoperative clinical examination of the donor extremity was performed to
assess knee stability and the function of the muscles supplied by the peroneal
nerve. Radiographs of the donor knee and ankle were not made because both
joints were asymptomatic in all patients.
Operative Technique
Harvesting of the Proximal Part of the Fibula
Because of its anatomical similarities with the distal part of the radius,
the proximal part of the contralateral fibula is preferred for reconstruction
following distal radial loss. An anterolateral approach is used to isolate the
proximal part of the fibula based on the anterior tibial arterial network. The
dissection is carried out in the intermuscular plane between the tibialis
anterior and extensor digitorum longus muscles
(Fig. 2). The latter, together
with the peroneus longus muscle, is sharply detached from its proximal
insertion at the level of the emergence of the peroneal nerve into the
anterior compartment of the leg. The proximal muscular cuff must be left
attached to the fibular head since it contains the recurrent epiphyseal branch
of the anterior tibial artery on which this transfer is based. During the
diaphyseal dissection, as many periosteal branches as possible are preserved.
For this reason, it is recommended that the interosseous membrane and a
longitudinal strip of muscle be harvested as well in order to protect the
small branches from the main artery to the diaphyseal periosteum of the
proximal part of the fibula.
In the proximal one-half of the leg, the peroneal nerve must be dissected
from the anterior tibial vascular bundle. The nerve surrounds this bundle in
an intricate three-dimensional pattern and sends many branches to the muscles
of the anterior compartment. Some of these motor branches may perforate the
space between the vascular bundle and the bone and therefore cannot be
dissected (Fig. 3). In this
case, the motor branch is divided and then repaired with use of microsurgical
techniques.
The fibula is resected and is separated from the surrounding muscles and
the peroneal artery, which is located in close proximity to the posteromedial
aspect of the middle and distal parts of the fibula. The segment of diaphysis
should not exceed the proximal two-thirds in order to preserve an adequate
vascular supply to the periosteum.
The proximal tibiofibular joint is then opened, with care taken to preserve
as much of the lateral collateral ligament of the knee as possible. A strip of
biceps femoris tendon is incorporated in the graft in order to reinforce the
soft-tissue repair at the recipient site. Finally, the proximal dissection of
the pedicle is carried out until the origin of the anterior tibial artery is
exposed and ligated.
Care is taken when repairing the lateral structures that stabilize the knee
joint. The lateral collateral ligament, enhanced by the residual of the biceps
femoris tendon, is fixed to the lateral aspect of the tibia with nonabsorbable
sutures into the periosteum, and stability is evaluated.
Reconstruction of the Distal Part of the Radius
Fixation of the proximal part of the fibula to the distal part of the
radius can be achieved either with a plate and screws or with lag screws if a
step-cut osteotomy is performed. Either procedure is facilitated by the
similarity between the diameters of the donor and recipient bones. The wrist
joint is temporarily stabilized with a 1.2-mm Kirschner wire, which is removed
one month postoperatively. The strip of biceps femoris tendon is used for
soft-tissue repair and is anchored to the remaining distal radiocarpal capsule
and ligaments (Fig. 4). In
contrast, the distal radioulnar joint is usually left slightly lax in order to
prevent any possible impingement during pronation and supination.
A reverse-flow arterial end-to-end anastomosis is then performed with the
recipient vessel, which may be either the radial artery or the common
interosseous artery. The recipient vein is usually the cephalic vein. At the
end of the vascular repair, bleeding should be observed from the muscular cuff
surrounding the transferred proximal part of the fibula. An above-the-elbow
cast is worn during the first two months postoperatively and is then replaced
with a wrist splint.
At the time of the last follow-up, one patient (Case 4) had died 3.5 years
postoperatively as a result of lung metastasis and the remaining five patients
were alive and free of disease. All six patients had satisfactory clinical and
radiographic results, with no major complications such as fracture or
infection. The four surviving patients who had been followed for three years
or more used the reconstructed extremity for activities of daily living with
minimal if any restriction. In spite of mild articular incongruity observed on
serial radiographs, the functional outcome was very good. Postoperatively, the
mean flexion of the wrist was 66° (range, 30° to 80°), the mean
extension was 54° (range, 20° to 65°), the mean supination was
80° (range, 75° to 85°), and the mean pronation was 72°
(range, 45° to 85°). The overall range of wrist joint motion was
approximately 70% of that of the contralateral extremity in the four patients
in whom the ulna had been spared in the initial resection. The wrist flexion
and extension of the patient with the one-bone forearm were 30° and
20°, respectively, and the overall range of motion of the wrist was
limited to 29% of that of the contralateral extremity, with neither pronation
nor supination possible. All patients had normal sensation, satisfactory grip
strength, and full motion of the digits.
Follow-up evaluation of the donor extremity revealed a full range of
motion, no evidence of laxity of the knee joint, and no long-term functional
impairment.
The results concerning the quantity and quality of growth in the five
patients who were followed for three years or more are presented in
Table II. This group includes
the patient who died 3.5 years postoperatively. All of the vascularized
proximal fibular transfers demonstrated continual longitudinal growth
following cessation of postoperative adjuvant chemotherapy, a factor
identified as a retardant of longitudinal growth. The mean growth rate per
year was 0.8 cm (range, 0.7 to 1.1 cm) (Figs.
5-A and 5-B6-A,6-B,6-C).
There was less of an increase in the diameter of the transferred fibula,
probably because of the absence of weight-bearing. The distal radioulnar joint
was left quite lax in order to prevent impingement during pronation and
supination. The transferred proximal part of the fibula lacks a sigmoid notch
and does not provide an articular surface for the ulna. Although no specific
ligament reconstruction had been done, no distal ulnar instability had yet
been observed at the time of writing. Computed tomography showed excellent
adaptive changes of the articulation of the transferred fibula with the
carpus.
Complications
The only complication was a peroneal nerve palsy in three patients. All
three patients had had division of motor branches of the peroneal nerve during
the harvest of the proximal part of the fibula, and the branches had been
repaired with microsurgical techniques. Two of the palsies completely resolved
within one year postoperatively, but one patient had a permanent paralysis of
the tibialis anterior muscle. A transfer of the tibialis posterior muscle was
suggested to this patient in order to eliminate need for the ankle-foot
orthosis that he used when walking.
Until the potential for vascularized bone allotransplantation becomes a
reality, the free vascularized epiphyseal transfer with a variable amount of
adjoining diaphysis represents the most effective one-stage procedure for
reconstruction of large epiphyseal bone defects in children while maintaining
the potential for longitudinal
growth4-6.
Experimental applications of the procedure were described by Bowen et
al.7-10,
Nettelblad et
al.11,12,
Manfrini et al.13,
and Donski et
al.14,15.
The first clinical vascularized epiphyseal transfers were performed nearly two
decades ago by a few surgeons who reconstructed either the distal part of the
radius or the proximal part of the humerus in children by means of a
vascularized fibular graft that included its proximal
epiphysis16-19.
For distal radial reconstruction in children, the proximal part of the
fibula offers many advantages over other donor sites such as the iliac crest
or the distal portion of the
scapula20. Those
structures are actually apophyses, not true epiphyses, and therefore lack an
organized articular surface. Furthermore, neither the scapula nor the iliac
crest provides tubular bone for simultaneous diaphyseal reconstruction.
In contrast, the proximal part of an immature fibula provides a true physis
and epiphysis and a variable amount of diaphysis, which allows long diaphyseal
reconstruction and facilitates bone fixation. Because of its anatomical
similarity to the distal part of the radius, the proximal part of the
contralateral fibula is the ideal donor bone. Another advantage is the plastic
property of the fibular head, which is revealed by the remodeling that occurs
at the articular surface over time. Lastly, in this series of patients, the
growth rate of the transplanted fibula resembled that of the distal part of
the radius, limiting the potential for asymmetric growth of the radius and the
ulna.
The need to provide an adequate blood supply both to the proximal part of
the fibula and to the diaphysis is a crucial technical problem. Initially this
required harvesting a double pedicled graft consisting of the descending
genicular or the anterior tibial artery to supply the fibular head and the
peroneal artery to supply the diaphyseal
segment19,20.
Such a procedure requires two recipient vessels or complex vascular
assemblage, which increases the operating time and the risk of microvascular
failure. According to the anatomical studies by Taylor et
al.21 and Bonnel et
al.22, the anterior
tibial artery can supply both the epiphysis and the diaphysis and there is no
need to provide a double pedicle to the graft. The harvest of the fibula must
preserve the epiphyseal artery, which arises from the main vessel just after
its entrance into the anterior compartment of the leg, as well as the small
musculoperiosteal branches that constitute the diaphyseal periosteal network.
After appropriate dissection, bleeding from both the muscular cuff surrounding
the fibular head and the diaphysis had been observed prior to division of the
fibula in all patients. The choice of this pedicle greatly simplified the
harvesting technique and the subsequent vascular anastomosis. The dissection
of the nerve from the vascular bundle is the critical point of the procedure:
a very intricate network of small motor branches surrounds the vessel, and it
usually is not possible to salvage all of them. When a major branch is
sacrificed, either a neurorrhaphy or a direct neurotization should be
performed in order to reduce the risk of a permanent palsy.
Unlike conventional microvascular skeletal reconstruction, in which
hypertrophy is the only parameter to be studied, a vascularized epiphysis is a
much more complex unit, the growth of which not only produces quantitative
changes but also induces adaptive modifications governed by new functional
requirements. Three-dimensional computed tomography was found to be
particularly useful in the study of the plasticity of the transferred bone and
revealed details that were not fully appreciated on standard radiographs. The
major morphological modifications occurred at the level of the epiphysis,
where progressive remodeling made the articular surface more congruent with
the proximal carpal row (Figs.
6-B and
6-C), thus increasing the range
of motion and the stability of the wrist joint. From a functional point of
view, our patients had improvement in the performance of the reconstructed
joint for up to several years after the surgery, taking advantage of the
articular remodeling that made the fibular head more and more similar to a
radial epiphysis.
The transferred fibula seems to grow at a rate matching that of the host
bone. This remarkable finding was based not only on the comparison with the
contralateral side but also on the comparison with the growth rate of the
ipsilateral ulna (Fig. 5-B).
Assessment of symmetrical growth of the two bones of the forearm is actually a
more accurate index, and it provides very reliable information about the
growth of the transferred fibula in its new location. At this time, we do not
have enough data to suggest that the growth rate depends on local factors
rather than on an intrinsic property of the bone; however, in the near future,
the study of the behavior of fibulae transferred to different locations will
hopefully provide an answer to this question.
In conclusion, following a surgical resection of the distal part of the
radius in children, a biological reconstruction should simultaneously replace
the bone defect and prevent future upper-extremity length discrepancy. The
vascularized transfer of the proximal part of the fibula, including its physis
and epiphysis, produced consistent and predictable longitudinal growth, thus
preventing wrist deformity and subsequent functional impairment.
Note: The authors thank Dr. Jesse Jupiter and Dr. Scott Levin
for their advice and encouragement and for the patient review, and they also
thank Francesco Di Nardo for his artistic drawings.
Ferrari S, Mercuri M, Picci P,
Bertoni F, Brach del Prever A, Tienghi A, Mancini A, Longhi A, Rimondini S,
Donati D, Manfrini M, Ruggieri P, Biagini R, Bacci G. Nonmetastatic
osteosarcoma of the extremity: results of a neoadjuvant chemotherapy protocol
(IOR/OS-3) with high-dose methotrexate, intraarterial or intravenous
cisplatin, doxorubicin, and salvage chemotherapy based on histologic tumor
response. Tumori.1999;85:
458-64.85458
1999
[PubMed]
Bacci G, Ferrari S, Mercuri M, Longhi
A, Capanna R, Tienghi A, Brach del Prever A, Comandone A, Cesari M, Bernini G,
Picci P. Neoadjuvant chemotherapy for extremity
osteosarcoma—preliminary results of the Rizzoli's 4th study.
Acta Oncol.1998;37:
41-8.3741
1998
[PubMed][CrossRef]
Rosito P, Mancini AF, Rondelli R,
Abate ME, Pession A, Bedei L, Bacci G, Picci P, Mercuri M, Ruggieri P, Frezza
G, Campanacci M, Paolucci G. Italian Cooperative Study for the treatment
of children and young adults with localized Ewing sarcoma of bone: a
preliminary report of 6 years of experience. Cancer.1999;86:
421-8.86421
1999
[PubMed][CrossRef]
Innocenti M, Ceruso M, Lauri G,
Angeloni R, Manfrini M, Capanna R, Camnasio F. Free vascularised growth
plate transplantation in a growing child. In: Limb salvage: current
trends. Proceedings of the Seventh International Symposium.
Singapore; 1993. p 427-30.427
1993
Innocenti M, Ceruso M, Manfrini M,
Angeloni R, Lauri G, Capanna R, Bufalini C. Free vascularised growth-plate
transfer after bone tumor resection in children. J Reconstr
Microsurg.1998;14:
137-43.14137
1998
[CrossRef]
Innocenti M, Ceruso M, Delcroix
L. Vascularized epiphyseal transfer in upper limb skeletal reconstruction
in children. Indications and operative technique. In: Schuind F, de Fontaine
S, Van Geertruyden J, Soucacos PN, editors. Advances in upper and
lower extremity microvascular reconstructions. Singapore: World
Scientific; 2002. p 90-105.90
2002
Bowen CVA, O'Brien BM.
Experimental study of the microsurgical transfer of growth plates.
Can J Surg.1984;27:
446-52.27446
1984
Bowen CVA. Experimental free
vascularized epiphyseal transplants. Orthopedics.1986;9:
893-8.9893
1986
[PubMed]
Bowen CVA, Ethridge CP, O'Brien BM,
Frykman GK, Gumley GJ. Experimental microvascular growth plate transfers.
Part 1—investigation of vascularity. J Bone Joint Surg
Br.1988;70:
305-10.70305
1988
Bowen CVA, O'Brien BM, Gumley GJ.
Experimental microvascular growth plate transfers. Part 2—investigation
of feasibility. J Bone Joint Surg Br.1988;70:
311-4.70311
1988
[PubMed]
Nettelblad H, Randolph MA, Weiland
AJ. Physiologic isolation of the canine proximal fibular epiphysis on a
vascular pedicle. Microsurgery.1984;5:
98-101.598
1984
[PubMed][CrossRef]
Nettelblad H, Randolph MA, Weiland
AJ. Free microvascular epiphyseal-plate transplantation. An experimental
study in dogs. J Bone Joint Surg Am.1984;66:
1421-30.661421
1984
[PubMed]
Manfrini M, Randolph MA, Andrisano A,
Weiland AJ. Orthotopic knee grafts in rats: a model for growth plate
transplantation. Microsurgery.1988;9:
242-5.9242
1988
[PubMed][CrossRef]
Donski PK, Carwell GR, Sharzer
LA. Growth in revascularized bone grafts in young puppies.
Plast Reconstr Surg.1979;64:
239-43.64239
1979
[PubMed][CrossRef]
Donski PK, O'Brien BM. Free
microvascular epiphyseal transplantation: an experimental study in dogs.
Br J Plast Surg.1980;33:
169-78.33169
1980
[PubMed][CrossRef]
Pho RW, Patterson MH, Kour AK, Kumar
VP. Free vascularised epiphyseal transplantation in upper extremity
reconstruction. J Hand Surg [Br].1988;13:
440-7.13440
1988
[PubMed]
Weiland AJ, Daniel RK.
Microvascular anastomoses for bone grafts in treatment of massive defects in
bone. J Bone Joint Surg Am.1979;61:
98-104.6198
1979
[PubMed]
Zhong-Wei C, Guang-Jian Z.
Epiphyseal transplantation. In: Pho RWH, editor. Microsurgical
technique in orthopaedics. Boston: Butterworths;
1988. p 121-7.121
1988
Tsai TM, Ludwig L, Tonkin M.
Vascularized fibular epiphyseal transfer. A clinical study. Clin
Orthop.1986;210:
228-34.210228
1986
Wood MB, Gilbert A, editors.Microvascular bone reconstruction. St. Louis: Mosby;
1997. p 85-9.85
1997
Taylor GI, Wilson KR, Rees MD,
Corlett RJ, Cole WG. The anterior tibial vessels and their role in
epiphyseal and diaphyseal transfer of the fibula: experimental study and
clinical applications. Br J Plast Surg.1988;41:
451-69.41451
1988
[PubMed][CrossRef]
Bonnel F, Lesire M, Gomis R, Allieu
Y, Rabischong P. Arterial vascularization of the fibula microsurgical
transplant techniques. Anatomia Clinica.1981;3:
13-22.313
1981
[CrossRef]