|
Background: The humerus is a frequent involvement site of
benign bone lesions. Various reconstruction methods have been
adopted to restore the defect after excavating the lesion
and/or to treat associated pathological fractures. In this
study, we reviewed the clinical outcomes of using allogenous
cortical struts to the treatment of patients with large humeral
defects resulting from benign bone lesions, and investigated
the mid-term fate of implanted allografts.
Methods: From 1988 through 1997, 29 patients with space-occupying
humeral lesions were treated by eradication of the tumor and
reconstruction with an intramedullary allogenous cortical
strut. No additional internal fixation was needed for support.
Clinical data were recorded, and functional and radiographic
results were evaluated.
Results: The sizes of defects after eradication of the lesions
ranged from 61 to 122 ml (mean, 92 ml). The patients were
followed for a mean of 8.8 years. One local recurrence was
noted and was successfully treated by repeating the procedure.
All patients achieved good to excellent functional results.
Follow-up radiographs showed complete healing of the defects,
with partial to complete incorporation of the allografts into
the host bones. Children had a better chance of complete allograft
incorporation than adults.
Conclusion: Intramedullary allogenous cortical struts act
as internal splint mechanically and bone graft material biologically.
The combined use of intramedullary allogenous cortical struts
and chipped cancellous bone grafts provided good stability
and healing probability for large osseous defects in the humerus
without the need for implant fixation. Allograft incorporation
occurred slowly in adults and might not achieve complete incorporation
in adults.
(Chang Gung Med J 2002;25:656-63)
Key words: allogenous cortical strut, bone tumor, union,
allograft incorporation.
The humerus is the third most common site for benign space-occupying
lesions after the distal area of the femur and the proximal
area of the tibia.(1) Some benign lesions can grow, cause
marked bony destruction, and lead to pathologic fracture of
the shaft or metaphyseal area. Consequently, adequate eradication
of the lesion is necessary, except in the case of simple bone
cysts.(2,3) However, tumor eradication usually creates a defective
and weakened limb. Small defects can be simply managed with
an autograft or may even be left untreated, but extensive
defects carry a high risk of postoperative fracture and can
be slow to heal. Therefore, the treatment of large tumors
or space-occupying tumor-like lesions generally focuses on
restoring structural integrity after eradication of the bone
lesions.
Large osseous defects remain orthopedic challenges. Various
reconstruction options can be considered after eradicating
the lesions, including autografts, allografts, and various
biocompatible bone substitute materials. Autografts, either
vascularized or non-vascularized, are considered the ideal
graft materials.(4-6) However, autografts are not always available
in sufficient quantity, especially in children. Donor site
morbidity is another drawback prohibiting their wide application.
Biocompatible bone substitute materials can only be used in
areas with intrinsic skeletal stability and do not provide
significant structural support.(7-9) Additional supports using
various internal fixation devices or cast protection are often
needed when autogenous cancellous bone grafts or bone substitute
materials are adopted. Torsional and shear forces at the lesion
site make good cast immobilization difficult to achieve, and
the geometry of the humerus, particularly at its proximal
and distal ends, limits implant choice. The site and size
of the bony lesions can make stable intramedullary fixation
difficult even when transfixation screws are used. A plate
and screw fixation relies on the purchase of screws that vary
with the quantity and quality of the remaining bone.
Given the above problems, allogenous cortical struts are attractive
alternatives for treating large bony defects of the humerus.
Allogenous cortical struts have been widely used to treat
spinal conditions for interbody fusions, various nonunions
of long bones, and in revision surgery for total joint arthroplasty.(10-15)
In addition to acting as a bone graft, allogenous cortical
struts provide immediate mechanical strength that can eliminate
the need for any additional intramedullary nails or plates
and screws.(15,16)
Though allografts have been widely used in various reconstructive
fields of orthopedic practice, only a few researchers have
examined the clinical results of using intramedullary allogenous
cortical struts to treat benign humerus lesions. Furthermore,
mid-term results of allogenous cortical struts have also rarely
been discussed. The purpose of this study was to review the
clinical, functional and radiographic results of using intramedullary
allogenous cortical struts to reconstruct large humerus defects
after eradicating the lesions. Mid-term remodeling of the
allogenous cortical struts and incorporation of the allografts
to the host bones were also addressed.
METHODS
From 1988 through 1997, 189 patients were admitted to the
orthopedic department of the Chang Gang Memorial Hospital
with benign lesions in the humerus. Most patients were admitted
for corticosteroid injections and multiple drilling for simple
bone cysts, osteochondroma excision, or curettage and autogenous
or combined allogenous and autogenous cancellous bone grafting
for small lesions. Large humerus lesions were defined as those
larger than 6 cm, expansile, associated with cortical destruction,
and having a cavity exceeding 60 ml in volume. Twenty-nine
patients were diagnosed with large humeral bone lesions, and
treated using curettage and reconstruction of the humerus
with an allogenous cortical strut, with or without supplementary
cancellous bone graft. Clinical data regarding gender, age
at surgery, diagnosis, location, estimated volume of the lesions,
the presence or a history of pathological fracture, previous
treatment, operation time, length of hospital stay, and complications
were collected.
Allograft preparation
The allografts were harvested and stored according to the
guidelines set by the American Association of Tissue Banks.(17-19)
Potential donors with malignancy, hepatitis, septicemia, death
from unknown causes and those known to be at high risk (human
immunodeficiency virus (HIV)-positive patients, acquired immune
deficiency syndrome (AIDS) patients, drug abusers, homosexuals,
prostitutes, and hemophiliacs) were excluded. All donors were
screened for hepatitis, HIV, and venereal diseases. Blood
cultures for aerobic and anaerobic organisms were undertaken
and multiple bacteriological cultures of soft tissues and
bone marrow were obtained during the recovery phase of bone
banking. The bones were harvested within 24 hours of death
and stored at -70oC. Small bones, such as the radius, ulna,
or fibula, were used as cortical struts.
Operative technique
The lesion was approached directly through a window in the
cortex. Tissues were taken for the frozen section first. Definite
surgery was performed after the diagnosis of the frozen section.
Curettage was first performed and a dental burr was used to
extend the margin to beyond the reactive zone of the tumor
to ensure adequate eradication of the lesions. Local adjuvant
chemocauterization with phenol and alcohol was performed in
cases of fibrous dysplasia, aneurysmal bone cysts, and giant
cell tumors.(20) The defect was then reconstructed with one
or two appropriately contoured fresh-frozen allogenous cortical
struts depending on the size of the defect. Most of the cortical
allografts could be inserted and firmly anchored due to the
geographic irregularities of the remaining cortex of the humerus.
In some cases, when the allogenous cortical strut was not
securely fixed, two additional screws were passed through
the cortical allograft and the humeral cortices to achieve
a so-called quadricortical fixation. Supplementary autogenous
or allogenous cancellous bone grafts were used to completely
fill the bony defect in 22 patients. No internal fixation
implants such as intramedullary nails or plates and screws
were used as the cylinder cortical graft acted as an internal
splint and provided sufficient mechanical strength. Antibiotics
were given once before operation and continued for 1 day postoperatively.
Postoperative management
The humerus was initially immobilized in a sling and swathe.
The patients were encouraged to perform gentle, protective,
passive range of motion exercises from the third day after
surgery, then gradually shifted to active motion exercise
at 6 weeks after surgery. Patients were followed every 2 months
for the 6 months, every 3 months for the next 6 months, and
every 6 months thereafter.
Functional evaluation
Functional results were evaluated at the last follow up, and
were classified using the Musculo-Skeletal Tumor Society rating
score of Limb Salvage for the shoulder and elbow.(21)
Radiographic evaluation
Plain radiographs, including antero-posterior and lateral
views, were taken preoperatively, immediately after surgery,
and at every follow-up examination. All radiographs were independently
reviewed by the authors and radiologists. The radiographic
analysis was comprised of two aspects: the first aspect involved
estimating the volume of the lesion in cubic centimeters using
the method described by Glancy et al.,(4) while the second
aspect involved determining the healing of the lesion and
the extent of incorporation of the allogenous cortical struts
into the host bones. Each radiograph was examined for trabeculation,
internal callus formation, bone density, and borders between
the cortical struts and the cavity. A lesion was considered
healed if the preoperative cavity was completely obliterated.
The lesion was considered partially or incompletely healed
when residual lytic areas remained. The union was considered
a failure if the cavity was not obliterated, no evidence of
trabecular formation existed, or the graft was resorbed. Allograft
incorporation into the host bone was considered complete if
the graft was completely obliterated. Incorporation was considered
partial if the graft was still visible but its border was
blunted, and no incorporation if the contour of the allograft
was unchanged from that of the initial postoperative radiograph.
Factors that may have influenced the incorporation of the
allogenous cortical struts were evaluated, such as patient
age, defect size, allograft length, and supplementary autogenous
or allogenous cancellous bone graft. Contingency tables were
created for each factor as related to incorporation, and subjected
to Fisher's exact test for nonparametric data and t test for
continuous values. A statistical significance was defined
as p<0.05 for each test.
RESULTS
A total of 29 patients were included in this study. There
were 17 men and 12 women. The ages of the patients ranged
from 11 to 52 years, with a mean of 21 years. Bone defects
were located in the proximal end of the humerus in four patients,
in the midshaft in 13 patients, and in the distal third of
the humeral shaft in four patients. Four patients had bone
involvement extending from the upper to the middle third of
the humerus, two had bone involvement extending from the middle
to the lower third of the humerus, and two had lesions involving
the whole humerus. Four (14%) patients had experienced one
or more pathologic fractures, and three were treated with
casting and local prednisolone injections, while one was treated
with surgery. Twenty-five patients sought medical help because
of pathological fracture after minor traumas. Four patients
were referred for further treatment because of local recurrence
following previous surgical intervention or failed repeated
corticosteroid injection. The pathologic diagnoses were simple
bone cyst in nine patients, fibrous dysplasia in 10 patients,
aneurysmal bone cyst in five patients, and giant cell tumor
in five patients. No immediate postoperative complications
occurred. The mean hospital stay was 6 days.
Oncological results
A minimum follow-up of 5 years was necessary to ensure no
signs of local recurrence and to assess the long-term results
of the allografts. After a mean follow-up period of 8.8 years
(range, 5 to 14 years), one patient died of an unrelated cause
and the remainders were alive at the last follow up. One patient
suffered from local recurrence 1.5 years after surgery and
was treated successfully with another curettage and allogenous
cortical strut reconstruction. Another patient who had polyostotic
fibrous dysplasia with whole humerus involvement also had
residual tumor over the metaphyseal region, and was regularly
followed without further surgical intervention. All pathologic
fractures healed.
Functional results
The functional evaluation revealed that 25 (86%) patients
had no pain or discomfort, and enjoyed normal use of the affected
limb, while four (14%) patients experienced occasional soreness
of the affected limb after extended use. The second group
of patients were still able to perform daily life activities
without limitations and were satisfied with the results. At
the last follow up, the average active shoulder elevation
was 165 degrees, total shoulder abduction was 170 degrees,
and arc of rotation exceeded 160 degrees. The average range
of active elbow flexion was 145 degrees, pronation was 90
degrees, and supination was 100 degrees. The overall results
were rated as good in four patients and excellent in 25 patients.
Radiographic results
The sizes of the defects ranged from 61 to 122 ml with a mean
of 92 ml after eradicating the lesions. Radiographs at 2 months
postoperatively showed signs of new bone formation in the
cavitary defect, which indicated healing. All radiographs
except for the two above-mentioned cases showed complete union
without any residual tumors 6 months after surgery.
Consolidation and remodeling of the implanted allograft struts
were observed in radiographs taken between 9 months and 1
year after operation. Incorporation of the allograft struts
increased with time and reached a plateau at 2 to 3 years
after surgery except in some children. Bone remodeling continued
after 2 to 3 years and complete obliteration of allograft
strut border was noted in six children. At the last follow
up, the radiographic evaluation revealed blunting of the border
between the allograft struts and the host bone bed in most
patients (23/29, 79%) (Fig. 1). The allogenous cortical struts
remained visible in the intramedullary canal but were incorporated
into the endosteal surfaces of the humerus. Six patients displayed
complete obliteration of the allogenous cortical struts, indicating
complete incorporation of the allograft to the host bone (Fig.
2). Younger patients (under 16 years old) had better chances
for complete allograft incorporation than the adults (6/18
versus 0/11, p=0.039). However, incorporation of the allograft
to the host bone was not influenced by lesion size or location.
There were also no differences in the incorporation time of
patients with or without supplementary cancellous bone grafts.
DISCUSSION
Benign space-occupying lesions in the humerus are often small
and silent. Only less than 10% of lesions are large enough
to mandate surgery. Simple bone cysts, though large in size,
can usually be treated by drilling and repeated steroid injections
in most cases.(2,3) Small lesions that do not compromise the
structural integrity of the bones can be treated with curettage
and cancellous bone grafts only. The use of intramedullary
allogenous cortical strut reconstruction techniques for benign
space-occupying lesions in the humerus is thus limited to
a few specially selected patients. This was reflected by the
fact that only 29 of 189 patients admitted for treatment of
benign lesions in the humerus received this procedure. Indications
for this technique include lesions that are large, have thin
cortex, display cortical destruction, fail to heal, or fail
to respond to non-surgical treatment.
The goals of treatment for benign bone lesions are lesion
eradication to prevent local recurrence, and restoration of
functional integrity to avoid subsequent fracture or deformity.
Bone grafts are often needed to fill bone defects resulting
from tumor evacuation. Large bone defects also weaken the
bony structure and metallic internal fixation devices are
often needed. Conventional internal fixation devices such
as intramedullary nails or plates and screws are associated
with morbidity and are rarely performed in children. Consequently,
a more sophisticated surgical procedure that simultaneously
provides bone grafts and stability is desired.
Allogenous cortical struts have been used both to provide
mechanical support and supplement bone healing in various
areas.(10-15) The present series successfully extended this
technique to treat large bone defects caused by the extraction
of space-occupying lesions in the humerus. The mechanical
loading of a cortical strut graft is primarily in compression.
However, the loading of the strut grafts may also include
some bending or torsion in some settings. Previous studies
have shown that deep-freezing had only a minimal influence
on the biomechanical properties of bone allografts.(4,16,22)
The mechanical demands of the upper limbs are much less than
those of the lower limbs. Healing of the allograft struts
to the adjacent host bone further reduced the mechanical demands
on the struts.(23,24) The intramedullary cortical struts may
thus provide sufficient stability to the lesion site by acting
as an internal splint, while simultaneously functioning as
a bone graft material. This technique may not be as successful
in treating long bone defects of the lower limbs, where the
mechanical demands of the long bones are relatively high.
Conventional massive allograft reconstruction demonstrated
union at the cortico-cortical junction of the allograft and
the host bone. This union occurred very slowly (taking approximately
12 months) and was achieved by host derived external callus
bridging the junction and filling the gap between abutting
cortices.(23-25) Placing the allogenous cortical struts intramedullaryly
facilitated union because of the large contact area between
the cortical strut and the endosteal surface of the humerus.
The preservation of periosteal blood supply further enhanced
the defect healing. In the present series, all bony defects
showed good union at less than 6 months after surgery. This
added an additional advantage of the novel technique in reconstructing
large humeral bone defects.
Another advantage of the novel technique is its relatively
unlimited availability. Repeated grafting with another allograft
strut remains possible in cases of incomplete healing or failure
of union. In the present series, one patient was noted to
have local recurrence and was successfully treated with repeated
curettage and an additional intramedullary allogenous cortical
strut. Disadvantages of using allogenous cortical struts include
a small risk of transmission of infectious organisms, such
as human immunodeficiency and hepatitis B and C viruses. However,
these risks are significantly reduced by using modern screening,
preservation, and sterilization techniques.
The fate of allogenous cortical struts has seldom been discussed.
Allograft incorporation is a complex, multifaceted process,
and multiple variables influence the rate, pattern, and completeness
of healing. Although union with the host bone can be achieved
with large allografts, incorporation occurs slowly if at all.
A large series of clinicopathological studies of retrieved
human allografts showed that the extent and distribution of
cortical revascularization rarely exceeded 2 mm during the
first year. Beyond two years, osteoclastic resorption was
markedly decreased and rarely penetrated more than 10 mm into
the cortex.(25)
Radiographical assessment of the remodeling or incorporation
of allograft struts is very difficult. Since no retrieved
specimens were available in the present series, we can simply
use radiographic signs, such as blunting or complete obliteration
of allograft borders, as signs of allograft remodeling. This
method was quite subjective, but reasonably qualitative. Based
on our observation, the children appeared to have a better
chance of allograft strut incorporation than the adults. However,
allograft strut incorporation was not influenced by diagnosis,
lesion size or location. Incorporation time was not influenced
by whether or not patients had supplementary cancellous bone
grafts. However, since the host bone takes over the mechanical
load after bone defect unions, the degree of allograft incorporation
did not appear to affect the final results of the patients.
In conclusion, an intramedullary allogenous cortical strut
acts as an internal splint mechanically and as a bone graft
material biologically. The combined use of intramedullary
allogenous cortical struts and chipped cancellous bone grafts
provided good stability and healing rate for large osseous
defects of the humerus without the need for implant fixation.
Remodeling of the allograft and incorporation to the host
bone occurred slowly and might not be complete in adults.
|
