Abstract
Background: The growing frequency of joint arthroplasty has led to
increasing numbers of patients requiring revision surgery. In the treatment of
a failed total elbow arthroplasty not associated with infection, one of the
main issues is poor or absent proximal ulnar bone stock due to osteolysis. We
report our experience with the use of strut allograft reconstruction of the
proximal part of the ulna as an adjunct to revision total elbow arthroplasty
with a noncustom implant. Our aim was to better define the indications,
outcomes, and complications of this technique in a population of patients with
a failed total elbow arthroplasty.
Methods: We reviewed the cases of patients with aseptic failure of a
total elbow replacement and proximal ulnar bone deficiency who were treated
with allograft bone struts. The patients had had an average of 2.5 (range, one
to four) prior open osseous operations addressing the elbow joint. In addition
to revision of the prosthetic components, the deficient bone stock was treated
with allograft strut grafts in one of four ways: (1) discrete cortical defects
were contained, (2) periprosthetic fractures were splinted, (3) deficient
triceps attachments were reconstructed, and (4) expanded segments were
augmented with struts and filled with impaction graft. Twenty-one patients
(twenty-two elbows) were followed for an average of four years (range, two to
eleven years).
Results: The mean Mayo Elbow Performance Score improved from 34
points preoperatively to 79 points at the time of the latest follow-up. The
scores for pain, stability, and activities of daily living improved most;
there was little change in motion. Complications, consisting of four
soft-tissue and four osseous problems, occurred in eight patients. Three
patients had incorporation of 26% to 50% of the graft; five, 51% to 75%; and
fourteen, 76% to 100%.
Conclusions: Most deficiencies of proximal ulnar bone stock and
fractures complicating revision total elbow surgery can be treated with
allograft strut grafting. Although the complication rate is high, this
technique is suitable for discrete cortical lesions, periprosthetic fractures,
and an expanded proximal part of the ulna, which also requires augmentation
with impaction grafting. The technique has been unreliable, however, in
restoring deficient olecranon bone stock.
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.
With the increasing number of total elbow arthroplasties being performed, a
parallel increase in revision surgery has been noted, as it has in the
experience with hip and knee arthroplasties. The literature contains little
information concerning techniques and outcomes of revision surgery following
total elbow
arthroplasty1-6.
Most of the principles regarding stem length, stem diameter, and bypassing
defects in revision elbow surgery were derived from experiences with revision
hip and knee arthroplasty. Although this transfer of principles has had
encouraging results, revision surgery on the elbow poses some unique
problems.
A major issue in revision total elbow arthroplasty is the loss of proximal
ulnar bone stock due to osteolysis, which can occur with loosening of the
ulnar implant. The aim of revision of a failed total elbow replacement due to
aseptic loosening, implant fracture, or periprosthetic fracture in combination
with compromised bone stock is to restore a stable pain-free functional elbow
with reconstitution of proximal ulnar bone. In the past twenty-five years, the
use of allografts to restore skeletal defects has become increasingly popular,
mostly following tumor
resection7-9
but also in lower-limb revision
surgery10. There
has also been increasing use of allograft bone about the elbow for
reconstruction of osseous deficiencies due to trauma and following
arthroplasty7,11-15.
Most reports in the elbow literature address the use of massive osteochondral
allografts or allograft-prosthesis composites. We report our experience with
the use of strut allografts in the reconstruction of a deficient proximal part
of the ulna in association with a failed total elbow arthroplasty. One goal
was to better define the indications and outcomes of this specific technique
for revision total elbow surgery.
We retrospectively reviewed the records on 734 total elbow arthroplasties
performed by the senior author (B.F.M.) over a period of twenty-two years; 189
(26%) were revision procedures. Since 1996, the percentage of revision total
elbow replacements has risen from 17% to >40% at our institution. Among the
revisions, we identified twenty-two, in twenty-one patients, in which a major
proximal ulnar osseous deficiency was treated with a reconstructive procedure;
a revision of the ulnar component was also performed in twenty-one elbows.
Only patients who had been followed for a minimum of two years from the time
of the index revision procedure were included in this study. The investigation
was approved by the institutional review board.
The study cohort comprised seventeen women and four men, with an average
age at the time of the index ulnar reconstructive surgery of fifty-seven years
(range, thirty-seven to seventy-nine years). The operation was performed in
eight dominant extremities and fourteen nondominant ones (see Appendix). The
original implant had been in situ for an average of 6.7 years (range, eighteen
months to eleven years). The primary diagnoses included rheumatoid arthritis
with joint destruction (ten elbows), juvenile rheumatoid arthritis (three),
and sequelae of trauma (nine) consisting of posttraumatic arthritis (seven),
posttraumatic ankylosis with nonunion (one), or multiple failed attempts at
open reduction and internal fixation (one). One patient had bilateral elbow
fracture (Cases 4 and 10), with a twelve-year interval between fractures.
Eleven patients were known to have a major coexistent pathological condition
that was thought to influence the final outcome. Eight patients; seven of whom
had rheumatoid arthritis, were taking steroid medications; two patients were
diagnosed with osteoporosis; and one had non-insulin-dependent diabetes
mellitus. The patients had had a mean of three (range, one to eight)
operations about the elbow prior to the index reconstructive procedure at our
institution and a mean of 2.5 (range, one to four) open osseous operations
specifically addressing the elbow joint.
The predominant symptom at presentation was pain in twenty-one elbows,
whereas one patient was primarily concerned about instability. The source of
the pain was thought to be the loose components, which were apparent on
radiographs. Additional symptoms included weakness in five patients, reduced
motion in six patients, and deformity and ulnar nerve symptoms in one patient
each.
Six patients had weakness of elbow extension, with two demonstrating
antigravity weakness of the triceps, and one had a flail elbow. Four patients
had an irritable ulnar nerve located in the cubital tunnel, with a positive
Tinel sign and mild paresthesias in the ulnar nerve distribution. The
preoperative workup for infection included aspiration of the elbow in patients
who had pain at rest as well as measurement of the erythrocyte sedimentation
rate and the C-reactive protein level.
The implants in situ at the time of index revision included seven
Coonrad-Morrey, two Coonrad-I, eight Pritchard-Walker, and two London
prostheses as well as one GSB (Gschwend-Scheier-Bahler), one Triaxial, and one
Swanson prosthesis. The radiographic diagnosis was aseptic loosening of the
ulnar component in seventeen elbows. One patient had a well-fixed ulnar
component associated with substantial proximal bone loss, and one patient had
a well-fixed ulnar component with a periprosthetic fracture. Additional
diagnoses included symptomatic loosening of the humeral component in five
elbows, periprosthetic fracture at the component tip in the proximal part of
the ulna in eight (Fig. 1-C),
periprosthetic fracture of the humerus in two, and fracture of the stem of the
ulnar component in two. The Coonrad-Morrey ulnar component that fractured did
so partially through the proximal beaded area, which was a feature designed to
enhance the cement bonding interface. Subsequently, this area was recognized
as an area of increased stress and the implant was
redesigned16,17.
One elbow had been treated with an ulnar component with a precoating of
polymethylmethacrylate (Figs. 1-A,
1-B, and 1-C),
which subsequently has been associated with premature loosening following some
hip
replacements18,19.
The primary indication for reimplantation in this series was aseptic failure.
The primary indications for the allograft strut reconstruction of the proximal
part of the ulna are presented in Table
I.
Although some of the humeral components were not obviously loose on
preoperative radiographs, twelve elbows required revision of both the humeral
and the ulnar component. In one elbow, the ulnar component remained firmly
fixed distally but was associated with substantial proximal ulnar osteolysis.
The component was retained, and the revision consisted of reconstruction of
the proximal part of the ulna alone. Cortical perforations occurred in two
elbows during cement removal with a high-speed burr; they were recognized
intraoperatively and were covered with corticocancellous bone grafts.
Surgical Technique
The choice of surgical approach depended, in part, on the previous skin
incision, but the Bryan-Morrey deep
exposure20 was
possible in sixteen elbows and a structurally incompetent triceps or olecranon
allowed a triceps-detaching approach in two. The medullary canal was
débrided of cement and pseudomembrane, and osseous defects were
delineated.
Containment struts (Fig. 2,
A) were used in eight elbows in which the defect was
contained within a minimum of three cortical surfaces (for example, it was
used for a defect of the anterior aspect of the ulna in the region of the
ulnar stem). All grafts were attached with 16 or 18-gauge cerclage wires. The
primary function of containment struts was to provide local bone stock for
incorporation, and a secondary function was to add support to the lesion.
Structural struts were used to directly strengthen an incompetent
ulna-prosthesis unit with a periprosthetic fracture in eight elbows and with
an impending fracture in three (Figs. 2,
B, and
3-A). A secondary function of
the grafts in these elbows with a fracture was to provide local bone stock for
incorporation. The grafts were obtained from AlloSource (Centennial, Colorado)
and were either fresh-frozen or freeze-dried. Eight fibular, six humeral, two
ulnar, five femoral, and three rib grafts were used. Graft selection was
dictated by the structural elements required and the graft availability.
Whereas a large structural defect in a small person could be adequately
addressed with a fibular graft, the same defect in a larger person required a
femoral graft. We attempted to match the graft thickness to the thickness of
the underlying host cortical bone. Hence graft selection was tailored to the
size of the patient, the size of the ulnar bone defect, whether the humeral
side required support as well, and the quality of the surrounding cortical
bone stock. Because the grafts were ordered prior to the surgery, we were not
always able to tailor the graft exactly to the lesion and we erred on the side
of selecting stronger and thicker grafts. With regard to length, the graft was
cut to extend two cortical diameters distal to the lesion. In two patients
with deficient olecranon bone stock (Fig.
1-C), a structural graft was required for the proximal part of the
ulna and was extended proximally in order to secure a triceps attachment site
with a number-5 nonabsorbable suture (Fig.
2, C). Additionally, the graft was intended to increase
the triceps lever arm and decrease the likelihood of impingement on the
prosthesis. One elbow had an old nonunion of the olecranon combined with a
periprosthetic fracture at the tip of the ulnar component. The olecranon
fracture fragment was refreshed at its nonunion site and was securely fixed
with the proximal cerclage wire to the posterior strut graft. In two elbows,
the proximal part of the ulna had expanded, with thinning of the cortices,
around a loose ulnar component. They were treated with a combination of
impaction grafting with morselized allograft and augmented with strut
allografts (Fig. 2,
D).
Intraoperative findings were categorized according to how many surfaces of
the ulna were deficient. Seven specimens had a single-sided deficiency (five
anterior and two lateral), four had a two-sided deficiency (three anteromedial
and one anterolateral), three had a three-sided deficiency (lateral, anterior,
and medial), and three had a four-sided deficiency that involved 4 to 5 cm of
proximal ulnar loss. The four-sided deficiencies were managed with a combined
technique, which consisted of insertion of a fibular allograft-prosthesis
composite in one case and, in addition, anterior and posterior femoral struts
and circumferential cortical strut grafts in the other two cases.
Follow-up
Follow-up evaluation was performed with use of the Mayo Elbow Performance
Score (MEPS)16 and
radiographs. The patients were interviewed over the telephone by an
independent reviewer from our Total Joint Database who is specifically trained
in such communications.
Twenty patients (twenty-one elbows) were available specifically for this
final review. One patient, who was in a federal prison, could not be
contacted, but the last clinical follow-up of that patient had been performed
at five years after the revision. The final in-person assessment at our
institution was conducted at an average of 3.5 years (range, two to six
years). The final radiographic assessment of six elbows was carried out
specifically for this study, radiographs of five elbows were sent to us by the
local physicians, and the last radiographs made at our institution were used
for the final follow-up of eleven elbows. The radiographs were assessed with
respect to the bone-cement interface, proximal bone stock, and amount of graft
surface area incorporated into the underlying host bone. The bone-cement
interface was classified as one of five types, as previously
described16. Type 0
indicated no radiolucency of any part of the cement-bone interface; type 1, a
nonprogressive radiolucent line involving <50% of the interface; type 2, a
nonprogressive radiolucent line involving >50% of the interface; type 3, a
progressive radiolucent line involving <50% of the interface; type 4, a
progressive radiolucent line involving >50% of the interface; and type 5,
gross loosening of the implant. Graft incorporation was graded according to
the degree of host-graft distinction on both anteroposterior and lateral
radiographs. Grade 0 indicated that the whole graft-host interface was clearly
distinct and separate without any crossing trabeculae; grade 1, that =25%
of the interface was indistinct with crossing trabeculae; grade 2, that 26% to
50% of the interface was indistinct; grade 3, that 51% to 75% of the interface
was indistinct; and grade 4, that 76% to 100% of the interface was indistinct
(Fig. 3-B).
Clinical
The final follow-up of the twenty-one patients was performed at a mean of
four years (range, two to eleven years). The interview for determination of
the Mayo Elbow Performance Score (MEPS) was completed for twenty of the
twenty-one patients, whereas information on the incarcerated patient was
derived from the medical records at a five-year examination. All twenty-one
patients had radiographs of the elbow made at least two years after the
strut-graft procedure.
At the time of the final follow-up, the average MEPS had improved from 34
points (range, 15 to 55 points) to 79 points (range, 30 to 100 points). The
prerevision functional score had been poor for all twenty-two elbows, whereas
at the final review the score was poor for two elbows, fair for three, good
for five, and excellent for seven. Of the twenty-one patients, only one had a
lower functional level at the final review than before the revision, and this
patient had had multiple operations and complications. The pain scores
improved the most, from an average of 9 points (range, 0 to 30 points) to an
average of 32 points (range, 15 to 45 points). No patient had pain that was as
severe as it had been before the revision, although four patients continued to
have moderate pain and ten had mild pain. Seven patients were completely
pain-free. The average stability score increased from 0.48 point (range, 0 to
5 points) to 7 points (range, 5 to 10 points). The average score for
activities of daily living also improved considerably, from 8 points (range, 0
to 20 points) to 21 points (range, 5 to 25 points). The least improvement was
seen in the flexion arc, which consisted of a 17° improvement (range,
5° to 20° improvement) in flexion and an 18° improvement (range,
5° to 20° improvement) in extension.
Complications
Eight (36%) of the twenty-two elbows underwent additional surgery after the
index reconstructive revision operation for reasons related to the original
pathological condition or the index revision surgery
(Table II). Of these, six
elbows had a single complication and two had two complications.
Two ulnar components loosened following the index revision surgery, at
three and six years. One patient (Case 1), a prison inmate, had a history of
narcotics abuse, a long history of rheumatoid arthritis with steroid use,
osteoporosis, multiple documented falls, and insertion of a pre-coat ulnar
component (Coonrad-Morrey) at the original revision procedure. This patient
was treated with impaction cancellous allograft and strut allograft as well as
reinsertion of an ulnar component. Two years later, she sustained a
periprosthetic fracture about the tip of the ulnar stem in a fall, and this
was successfully treated in a plaster cast. The other patient with aseptic
loosening of the ulnar component, with an erosion and an impending fracture at
the tip of the ulnar stem, had had excellent relief of symptoms and had
returned to manual labor despite postoperative instructions to limit usage of
the elbow. The ulnar component was successfully revised with a longer stem and
strut grafts.
A third patient sustained a second periprosthetic fracture at the tip of
the ulnar stem, just distal to the original periprosthetic fracture, two years
after the revision procedure. Poor bone quality, a history of long-term
steroid use, and heavy cigarette smoking were possibly contributing factors. A
revision procedure with use of a long-stem ulnar component bypassing the
fracture and a posteromedial fibular strut to splint the fracture site was
performed. The patient had a well-fixed implant and a healed fracture at the
most recent assessment, at two years.
Radiographic Findings
Cement-Bone Interface
The cement-bone interface was type 0 in four elbows, type 1 in eight, type
2 in five, type 3 in two, type 4 in one, and type 5 in one. The two patients
with a type-4 or 5 interface required component revision. Interestingly, the
third patient who required a component revision had a type-1 interface.
Graft Incorporation
No elbow had grade-0 or 1 graft incorporation. Three had 26% to 50%
incorporation (grade 2); five, 51% to 75% incorporation (grade 3); and
fourteen, 76% to 100% incorporation (grade 4)
(Fig. 3-B).
Over the past two decades, the functional outcomes and rates of prosthetic
survival following primary total elbow arthroplasty have greatly
improved6,16,21,22.
Because the candidates for elbow replacement include many steroid-dependent
patients at one end of the spectrum and younger, more active patients at the
other end, an increasing number of patients have required revision surgery.
Revision of an ulnar component in a patient who has associated bone loss is a
complex undertaking. Many factors, including joint instability, infection,
periprosthetic fracture, olecranon nonunion, and difficulty with implant or
cement removal, make revision of an ulnar component challenging. A concern
with any revision arthroplasty is the quality and quantity of the remaining
host bone and thus the mechanical integrity of the limb segment.
There are two opposing approaches to the management of a failed
arthroplasty associated with an osseous deficiency: (1) replace the deficient
segment of host tissue and the loose implant with a custom-made device that
replaces the
deficiency1,23,24,
or (2) restore bone stock in conjunction with reimplantation of a standard
prosthesis. While several reports on the use of custom-made devices have been
published1,23,24,
we are unaware of any reports on the technique of restoring bone stock. Our
experience has been exclusively with the latter technique for appropriately
selected patients.
Bone stock can be restored with autologous bone graft, which is the most
effective grafting material since it provides all three elements required for
bone regeneration: osteoinduction, osteoconduction, and osteogenic cells.
Allograft material is attractive as it is readily available, eliminates the
morbidity associated with harvesting, and shortens the surgical procedure.
Concerns about transmission of infection (which occurs in up to 15% of cases),
nonunion (which occurs in nearly 10% of cases), and fracture (which occurs in
18% of cases) have been
documented25.
Although there were no graft fractures in the present study, one of our
patients, not in this series, who had a high level of activity had a fracture
of the strut graft at six months postoperatively and, at the time of writing,
was awaiting revision surgery.
Cortical onlay plate grafts have been used for as long as forty
years26. Most of
the current knowledge of the behavior of onlay grafts was derived from
experience with revision hip
surgery27. The use
of onlay allografts in the upper limb has been less extensive.
Gresham28 reported
using such grafts to treat fracture nonunions in the forearm, with an 85%
success rate in eighty patients, in whom host replacement of the allograft was
observed. Structural bone grafts serve two functions: first, they provide
strength to an otherwise structurally compromised construct, and, second, they
restore bone mass. An advantage of the strut graft over a traditional
allograft-prosthesis composite is that the former has a substantially larger
surface area for incorporation with the host. Our experience corroborates this
finding, as the bone-allograft junction was completely indistinct in thirteen
of the twenty-two elbows.
There are four distinct patterns of proximal ulnar bone loss related to
revision total elbow arthroplasty. These include discrete cortical defects,
which were contained in this series; periprosthetic fractures, which were
splinted; deficiency of the triceps attachment, which was reconstructed; and
expanded segments, which were augmented with struts and filled with impaction
graft. Granuloma formation, due to the host response to particulate debris,
can lead to endosteal cortical erosion progressing to discrete, multiple, or
multiply interconnected cortical perforations. When the cortical integrity is
sufficiently compromised, periprosthetic fractures can occur, usually in the
region about the tip of the ulnar stem or the olecranon process. The surgical
goal is always the same: to restore bone stock and regain stability with
function.
An area in which strut grafts failed to reconstruct the defect effectively
was the territory of the olecranon process. The purpose of using strut grafts
in this region, in elbows with circumferential proximal ulnar bone loss (two)
or with nonunion of the olecranon (one), was to provide a secure attachment
for the triceps and to increase its lever arm for function. None of these
three attempts succeeded in preserving this segment of allograft, which lacked
a stable and vascularized host tissue bed.
Technical features of graft application have been previously noted to
affect graft behavior. Close contact between the cortical graft and the host
bone has been shown to be of utmost importance in canine
studies27. With our
technique of graft preparation, the deep surface of the graft is sculpted,
with a high-speed burr, to a shape reciprocal with that of the host, and the
superficial surface is smoothed of sharp edges, while as much graft thickness
as possible is maintained. However, because of the subcutaneous location of
the ulna, the strut grafts were palpable by eight patients. Hence it is
important to balance the reconstructive requirement with the capacity of the
soft tissues. The grafts were secured with cerclage wires. We observed that
cementation was improved when defects were digitally occluded during insertion
of the cement and implant, with subsequent final graft placement and wire
tightening.
To treat these elbows successfully, the surgeon needs to be familiar with a
wide range of reconstructive techniques. Strut graft reconstruction appears to
stabilize and reconstitute the proximal part of an ulna with cortical defects
and a periprosthetic fracture. In conjunction with impaction grafting, strut
grafts increase host cortical bone stock. However, osseous deficiencies in the
region of the olecranon process and those that are not supported by stable and
vascularized host bone do not appear to be amenable to this form of
reconstruction. This type of complex reconstructive revision more than doubles
the patients' functional ability, especially in those with periprosthetic
fracture. However, the considerable complication rate, even when there is
successful incorporation of the graft, should be kept in mind.
A table showing demographic data on all twenty-one patients (twenty-two
elbows) is available with the electronic versions of this article, on our web
site at
(go to the article citation and click on "Supplementary Material")
and on our quarterly CD-ROM (call our subscription department, at
781-449-9780, to order the CD-ROM).
Dent CM, Hoy G, Stanley JK.
Revision of failed total elbow arthroplasty. JBone Joint Surg
Br.1995;77:
691-5.77691
1995
Ferlic DC, Clayton ML. Salvage of
failed total elbow arthroplasty. J Shoulder Elbow
Surg.1995;4:
290-7.4290
1995
[CrossRef]
Figgie HE 3rd, Inglis AE, Mow C.
Total elbow arthroplasty in the face of significant bone stock or soft tissue
losses. Preliminary results of custom-fit arthroplasty. J
Arthroplasty.1986;1:
71-81.171
1986
[CrossRef]
Figgie MP, Inglis AE, Mow CS, Wolfe
SW, Sculco TP, Figgie HE 3rd. Results of reconstruction for failed total
elbow arthroplasty. Clin Orthop.1990;253:
123-32.253123
1990
[PubMed]
Inglis AE. Revision surgery
following a failed total elbow arthroplasty. Clin
Orthop.1982;170:
213-8.170213
1982
Morrey BF, Bryan RS, Dobyns JH,
Linscheid RL. Total elbow arthroplasty. A five-year experience at the Mayo
Clinic. J Bone Joint Surg Am.1981;63:
1050-63.631050
1981
[PubMed]
Dean GS, Holliger EH 4th, Urbaniak
JR. Elbow allograft for reconstruction of the elbow with massive bone
loss. Long term results. Clin Orthop.1997; 341:
12-22.34112
1997
[PubMed]
Jaffe KA, Morris SG, Sorrell RG,
Gebhardt MC, Mankin HJ. Massive bone allografts for traumatic skeletal
defects. South Med J.1991;84:
975-82.84975
1991
[PubMed][CrossRef]
Mnaymneh W, Malinin T. Massive
allografts in surgery of bone tumors. Orthop Clin North
Am.1989;20:
455-67.20455
1989
Brady OH, Garbuz DS, Masri BA, Duncan
CP. The treatment of periprosthetic fractures of the femur using cortical
onlay allograft struts. Orthop Clin North Am.1999;30:
249-57.30249
1999
[PubMed][CrossRef]
Kay RM, Eckardt JJ. Total elbow
allograft for twice-failed total elbow arthroplasty. A case report.
Clin Orthop.1994;303:
135-9.303135
1994
[PubMed]
Urbaniak JR, Aitken M. Clinical
use of bone allografts in the elbow. Orthop Clin North
Am.1987;18:
311-21.18311
1987
Urbaniak JR, Black KE Jr.
Cadaveric elbow allografts. A six-year experience. Clin
Orthop.1985;197:
131-40.197131
1985
Breen T, Gelberman RH, Leffert R,
Botte M. Massive allograft replacement of hemiarticular traumatic defects
of the elbow. J Hand Surg [Am].1988;
13: 900-7. Erratum in: J
Hand Surg [Am]. 1989;14:582.13900
1988
[PubMed][CrossRef]
Foulkes GD, Mitsunaga MM.
Allograft salvage of failed total elbow arthroplasty. A report of two cases.
Clin Orthop.1993;296:
113-7.296113
1993
[PubMed]
Morrey BF, Adams RA.
Semiconstrained arthroplasty for the treatment of rheumatoid arthritis of the
elbow. J Bone Joint Surg Am.1992;74:
479-90.74479
1992
[PubMed]
Schneeberger AG, Adams R, Morrey
BF. Semiconstrained total elbow replacement for the treatment of
post-traumatic osteoarthrosis. J Bone Joint Surg Am.1997;79:
1211-22.791211
1997
[PubMed]
Collis DK, Mohler CG. Loosening
rates and bone lysis with rough finished and polished stems. Clin
Orthop.1998;355:
113-22.355113
1998
[CrossRef]
Mann KA, Bhashyam S. Mixed-mode
fracture toughness of the cobalt-chromium alloy/polymethylmethacrylate cement
interface. J Orthop Res.1999;17:
321-8.17321
1999
[PubMed][CrossRef]
Bryan R, Morrey BF. Extensive
posterior exposure of the elbow. A triceps-sparing approach. Clin
Orthop.1982;166:
188-92.166188
1982
Inglis AE, Pellicci PM. Total
elbow replacement. J Bone Joint Surg Am.1980; 62:
1252-8.621252
1980
[PubMed]
Kudo H, Iwano K, Nishino J. Total
elbow arthroplasty with use of a nonconstrained humeral component inserted
without cement in patients who have rheumatoid arthritis. J Bone
Joint Surg Am.1999;81:
1268-80.811268
1999
Ross AC, Sneath RS, Scales JT.
Endoprosthetic replacement of the humerus and elbow joint. J Bone
Joint Surg Br.1987;69:
652-5.69652
1987
Figgie HE 3rd, Inglis AE, Ranawat CS,
Rosenberg GM. Results of total elbow arthroplasty as a salvage procedure
for failed elbow reconstructive operations. Clin
Orthop.1987;219:
185-93.219185
1987
Clarke HD, Berry DJ, Sim FH.
Salvage of failed femoral megaprostheses with allograft prosthesis composites.
Clin Orthop.1998;356:
222-9.356222
1998
[PubMed][CrossRef]
Head WC, Malinin TI. Results of
onlay allografts. Clin Orthop.2000;371:
108-12.371108
2000
[PubMed][CrossRef]
Malinin T, Latta LL, Wagner JL, Brown
MD. Healing of fractures with freeze-dried cortical bone plates.
Comparison with compression plating. Clin Orthop.1984;190:
281-6.190281
1984
[PubMed]
Gresham RB. The freeze-dried
cortical bone homograft: a roentgenographic and histologic evaluation.
Clin Orthop.1964;37:
194-201.37194
1964
[PubMed]