Abstract
Background: The marked loss of glenoid bone
volume or alteration of glenoid version can affect glenoid component
fixation in patients undergoing total shoulder arthroplasty. The
purpose of this study was to evaluate the long-term results associated
with the use of bone-grafting for restoration of glenoid volume
and version at the time of total shoulder arthroplasty.
Methods: Twenty-one shoulders received
an internally fixed, corticocancellous bone graft for the restoration
of peripheral glenoid bone stock at the time of total shoulder arthroplasty
between 1980 and 1989. Grafting was indicated when glenoid bone
stock was insufficient to maintain adequate version or fixation
of the prosthesis. Seventeen shoulders were available for follow-up;
the average duration of follow-up for the thirteen shoulders that
did not have prosthetic failure within the first two years was seventy
months. Total shoulder arthroplasty was performed because of osteoarthritis
in five shoulders, chronic anterior fracture-dislocation in five,
capsulorrhaphy arthropathy in three, inflammatory arthritis in two,
recurrent dislocation in one, and failure of a previous arthroplasty
in one. All patients had some form of anterior or posterior instability
preoperatively. There were five anterior and twelve posterior glenoid
defects. Bone from the resected humeral head was used for grafting
in fifteen shoulders, and bicortical iliac-crest bone was used in two.
Results: The average glenoid version after grafting
was 4° of retroversion, with an average correction of 33°. The graft
failed to maintain the original correction in three shoulders due
to nonunion, dissolution, or shift. Five total shoulder replacements
failed, necessitating glenoid revision at two to ninety-one months
postoperatively. The failures were associated with recurrent massive
cuff tears (one shoulder), persistent instability (two shoulders), improper
component placement (one shoulder), and loss of graft fixation (one
shoulder). There were no humeral component failures. According to
the criteria of Neer et al., the functional result was rated as excellent
in three shoulders, satisfactory in six, and unsatisfactory in eight.
Conclusions: Despite the finding that eight shoulders
had an unsatisfactory functional result at the time of long-term
follow-up, corticocancellous grafting of the glenoid successfully
restored glenoid version and volume in fourteen of the seventeen
shoulders in the present study. Patients with glenoid deficiency often
have associated glenohumeral instability, which may affect the results
of total shoulder arthroplasty. Bone-grafting of the glenoid is
a technically demanding procedure that can restore bone stock in
patients with structural defects.
Total shoulder replacement was introduced by Neer in 1974
for the treatment of arthritic conditions involving the glenoid
and humeral articular surfaces1.
Several authors have reported excellent pain relief and improved
motion in association with this procedure2-12.
The glenoid component has been the object of much scrutiny and
is the most common source of failure of total shoulder arthroplasty.
Asymmetric glenoid wear or loss of bone stock can contribute to common
complications involving the glenoid component. The reported rate
of glenoid component loosening, defined as dislocation, migration,
or a complete progressive radiolucent line measuring 2 mm in width,
is approximately 10% (seventy-eight of 761 reported cases)2-13. Inadequate glenoid bone stock
and a lack of congruent component contact have been associated with
an increased rate of component loosening14-16.
The rate of glenohumeral instability after shoulder arthroplasty
has been reported to be approximately 3% (sixteen of 498
reported cases)3,6-9,11,13. Improper
glenoid component version due to asymmetric bone wear can be a source
of joint instability16-20.
Several conditions result in abnormal glenoid fossa wear16. Rheumatoid arthritis most commonly
results in central glenoid erosion. Primary osteoarthritis is typically
associated with progressive posterior glenoid wear. Arthropathy
related to recurrent or chronic dislocation of the shoulder can
result in excessive anterior or posterior wear, depending on the
direction of the instability. Severe erosion can make proper seating
of the glenoid component difficult. Techniques that can be used
to compensate for lesser degrees of bone loss include changing the humeral
component version, reaming to lower the elevated side of the glenoid
margin, and using an augmented glenoid component. These techniques may
be used alone or in combination, depending on the degree of erosion.
Larger defects may preclude the use of a glenoid component or require
the use of internally fixed bone grafts to restore glenoid fossa volume
and version. Although abnormal glenoid wear is common, the percentage
of patients requiring correction with internally fixed bone grafts
has ranged from only 4% (twenty of 463) to 10% (nine of
eighty-nine)16,21. The purpose
of this study was to evaluate the long-term results associated with
the use of internally fixed bone grafts for the reconstruction of
glenoid deficiencies at the time of total shoulder arthroplasty.
One hundred and thirty-two unconstrained total shoulder arthroplasties
were performed between 1980 and 1989. Twenty-one shoulders
(16%) received a large, internally fixed bone graft for
the reconstruction of a glenoid deficiency at the time of shoulder
replacement. Three patients were lost to follow-up before a minimum
of two years, and one died of unrelated causes. Seventeen shoulders
in sixteen patients (nine men and seven women) were available for
follow-up examination; the average duration of follow-up
for the thirteen shoulders that did not have prosthetic failure
within the first two years was seventy months (range, twenty-four
to 134 months). The average age at the time of the operation was
fifty-six years (range, thirty to eighty-four years). Total
shoulder replacement was performed on the right side in nine patients,
on the left in six, and bilaterally in one. Ten shoulder replacements
were on the dominant side.
Total shoulder replacement was performed because of primary osteoarthritis
in five shoulders, chronic anterior fracture-dislocation in five,
capsulorrhaphy arthropathy in three, inflammatory arthritis in two, arthritis
secondary to recurrent dislocation in one, and failure of a previous
replacement in one. Seven shoulders had undergone previous surgery.
Two of the three patients with capsulorrhaphy arthropathy had been
treated with a Putti-Platt stabilization procedure and
one, with a Magnuson-Stack procedure. Two patients had
had a previous resection arthroplasty of the distal part of the
clavicle. One patient with an anterior fracture-dislocation had
had a prior humeral head replacement for a three-part proximal humeral
fracture. One patient had a failed total shoulder replacement due
to persistent instability (Table I).
Patients were evaluated preoperatively and postoperatively for
pain, motion, strength, stability, and function according to the
system of the American Shoulder and Elbow Surgeons22. All patients were considered to
have some form of instability before the operation (Table I). Five shoulders
had a fixed anterior dislocation. One patient had bilateral recurrent
posterior dislocations. The remaining ten shoulders had a fixed
posterior subluxation. This was represented by the humeral head
being posteriorly displaced due to the lack of glenoid bone support
resulting from the defect. Thus, the center of the humeral head
was displaced posterior to the true central axis of the glenoid
(Fig. 1).
Six shoulders had an associated rotator cuff tear, three of which
were classified as massive (involving two or more tendons). All
cuff tears were diagnosed and repaired at the time of the index procedure.
All shoulders were evaluated preoperatively with an axillary
radiograph, and ten had an adjunctive evaluation with computerized
tomography (Table I). The axillary radiograph was standardized so
that the space between the posterior margin of the coracoid and
the anterior part of the glenoid rim was equal to the space between
the posterior part of the glenoid rim and the scapular spine. Three
factors were used to evaluate the glenoid defect: (1) the version
of the defect, measured relative to the normal glenoid surface version;
(2) the extent of the defect relative to the entire glenoid surface, expressed
as a percentage; and (3) the maximal depth of the defect at the
glenoid margin, measured in millimeters (Fig. 1, Table I). The normal glenoid contour was
extrapolated from radiographs of the uninvolved shoulder (fifteen
shoulders) or from an average glenoid retroversion of 5 (two shoulders
in one patient). All measurements were made on plain radiographs.
A computerized tomographic scan was used, if available, to corroborate
the measurements. In general, the plain radiographic measurements
correlated well with those obtained by computerized tomography.
Five shoulders, in the five patients with chronic anterior dislocation,
had an anterior glenoid defect. These anterior defects had an average
anteversion of 47° (range, 42° to 52°), accounted for an average
of 40% (range, 29% to 57%) of the entire glenoid
surface, and had an average maximal depth of 18 mm (range, 10 to
25 mm). Twelve shoulders, in eleven patients, had a posterior glenoid
defect. Five of these patients had primary osteoarthritis, three
had capsulorrhaphy arthropathy, two had inflammatory arthritis,
and one had bilateral involvement (recurrent dislocation on one
side and failure of a previous arthroplasty on the other). These
posterior defects had an average retroversion of 27° (range, 8°
to 50°), accounted for an average of 93% (range, 61% to
100%) of the glenoid surface, and had an average maximal
depth of 12 mm (range, 4 to 15 mm).
Glenoid insufficiency was defined as glenoid bone loss that met
one of three criteria: (1) cortical penetration of the glenoid neck
by the glenoid component keel or peg, (2) incomplete peripheral
contact of the glenoid component flange, or (3) >20° of
retroversion or anteversion of the glenoid component surface with
complete seating. Thus, the indication for glenoid bone-grafting
was (1) insufficient bone stock for appropriate component fixation
or (2) peripheral wear that was severe enough to result in component
malpositioning that could not be corrected by glenoid reaming or
a change in humeral component version. These situations typically could
be predicted preoperatively by templating a true axillary radiograph
of the glenohumeral joint. If the keel of the component clearly
penetrated the cortex as a result of either excessive version or decreased
volume, grafting was anticipated.
All shoulders were treated with the Neer-II prosthesis
(Kirschner Medical, Fair Lawn, New Jersey). The humeral component
was fixed with cement in three shoulders and was press-fit
in fourteen. The humeral component was placed in the standard 30° to
40° of retroversion in fifteen shoulders, and it was placed in neutral
version in two (Cases 3 and 10). An all-polyethylene glenoid component
was used in five shoulders, and a metal-backed component
was used in twelve. All glenoid components were fixed with cement.
The glenoid defect was prepared and grafted as described by Neer
and Morrison16. There were five
anterior and twelve posterior glenoid bone grafts. Bone from the
resected humeral head was used for grafting in fifteen shoulders,
and bicortical iliac-crest bone was used in two (Cases 3 and 12).
No allograft bone was used. The graft was transfixed with screws
in sixteen shoulders; in the seventeenth shoulder, an attempt was
made to insert the graft without fixation by wedging it in a fissure created
by an osteotomy along the glenoid neck.
Fourteen shoulders required one or more supplemental procedures
in addition to the index operation. Six shoulders required a rotator
cuff repair, six required lengthening of the subscapularis tendon, one
required an anterior capsular release because of an internal rotation
contracture, and two required a posterior capsulorrhaphy.
The result at the most recent follow-up examination was
graded according to the criteria described by Neer et al.11. A standard radiographic evaluation
was performed with use of anteroposterior and lateral radiographs of
the glenohumeral joint relative to the scapular plane as well as
an axillary radiograph. Radiographs were made preoperatively, immediately postoperatively,
at three and six months, and yearly thereafter. Postoperative radiographs
were analyzed with the method described by Amstutz et al.13. Fixation of the glenoid component
was classified according to the criteria described by Franklin et
al.23 (Table I).
The arthroplasty failed in five shoulders (Cases 2, 3, 13, 14,
and 16) (Table I). All failures were associated with symptomatic
loosening of the glenoid component; there were no humeral component
failures. The patient with bilateral involvement (Cases 2 and 3)
had bilateral failure. The right shoulder (Case 2) had had a previous
total shoulder arthroplasty for the treatment of arthritis secondary
to recurrent dislocation. The revision, which involved insertion
of a posterior bone graft to correct glenoid version, failed twenty-one
months postoperatively due to excessive polyethylene wear in the
glenoid component secondary to persistent instability and development
of a rotator cuff tear. The procedure in the left shoulder (Case
3) failed ninety-one months postoperatively due to persistent
instability. The third failure (Case 13) occurred eight months postoperatively
due to persistent instability that resulted in nonunion of the graft
and shifting of the component. The fourth failure (Case 14) occurred
because the glenoid component had been placed too inferiorly, resulting
in superior subluxation of the humerus and shifting of the glenoid
component two months postoperatively. The fifth failure (Case 16), which
involved the only shoulder in which the graft was not fixed with
a screw, occurred when the glenoid component and bone graft became
dislodged two months postoperatively. All failures were definitively
treated with removal of the glenoid component. Glenoid version and
peripheral bone volume were noted to be well restored after removal
of the component from two (Cases 2 and 3) of the five shoulders.
Functional Evaluation
The twelve patients who did not have prosthetic failure had marked
relief from pain at the time of the most recent follow-up. All twelve
had had marked or disabling pain preoperatively. At the time of
the most recent follow-up examination, ten patients had
slight or no pain, one had moderate pain with activity, and one
had moderate pain at rest. The average total elevation at the time
of the latest evaluation was 107° (range, 30° to 165°), with an
average postoperative improvement of 19° (range, —40° to
90°). Three patients had less than 90° of elevation postoperatively.
The average external rotation was 28° (range, —15° to 90°),
and the average internal rotation was to the twelfth thoracic level
(range, the sacrum to the sixth thoracic vertebra); the average
improvement in these values was 28° and four spinal levels, respectively.
Nine patients obtained full strength. The strength deficit in two
patients (Cases 4 and 12) was due to the recurrence of a massive
rotator cuff tear after repair. Glenohumeral stability returned
to normal in nine of the twelve patients (Table I). In three patients (Cases
4, 9, and 12), the original anterior-posterior instability was corrected
but superior migration of the humeral head subsequently developed.
The average preoperative functional score was 23 points (range,
4 to 39 points), and the average postoperative improvement was 20
points (range, 2 to 41 points). All twelve patients were satisfied
with the procedure, with eight patients stating that they were much
better and four stating that they were better.
When the twelve patients who did not have prosthetic failure
were evaluated according to the criteria of Neer et al.11, the functional result was rated
as excellent in three shoulders, satisfactory in six, and unsatisfactory
in three. The unsatisfactory results were related to recurrence
of a rotator cuff tear (Cases 4 and 12) and to limited postoperative
motion (Case 17).
Radiographic Evaluation
Fourteen of the original seventeen bone grafts, including two
grafts in shoulders in which the glenoid component failed, healed
in proper position. Graft-healing was determined during the revision operation
in the shoulders with component failure. Three grafts failed in
association with failure of the glenoid component: two (Cases 13
and 14) failed as a result of nonunion caused by persistent glenohumeral
instability or improper placement of the glenoid component, and
one (Case 16) failed as a result of loss of graft fixation (Table
I). The average glenoid version after grafting was 4° of retroversion,
with a range of 22° of anteversion to 30° of retroversion. The average
correction in version of the glenoid surface, excluding the three
failures, was 33°.
The initial postoperative radiographs demonstrated periprosthetic
radiolucency around nine of the seventeen glenoid components. The
final radiographic evaluation of the twelve components that did
not fail revealed no radiolucency around two components, radiolucency
limited to the area of the flange of two components, incomplete
radiolucency at the keel of three components, and a complete radiolucent
line with a width of <2 mm around five components. The
final radiographic evaluation of the five components that failed
demonstrated that three had grossly shifted and one had dislocated
(Table I). The radiographic evaluation of the fifth component that
failed showed radiolucency limited to the area of the flange; although
this component did not demonstrate severe loosening, it was revised because
of excessive polyethylene wear.
The final radiographic evaluation demonstrated no radiolucent
lines around twelve of the seventeen humeral components. A radiolucent
line with a thickness of £1 mm was observed in one or more zones
around five humeral components. Two of these five components were
associated with global radiolucency, two were associated with radiolucency
in the medial and lateral metaphyseal regions, and one was associated
with radiolucency that was isolated to the lateral metaphyseal region. No
humeral component was associated with progressive radiolucency on
serial radiographs.
Total shoulder replacement has provided good long-term results,
with a satisfactory outcome reported after 72% (263) of
366 reported procedures2,7,8,10,11.
Revision of the glenoid component because of loosening has been
reported after approximately 2% to 3% of primary
shoulder replacements14,23,24.
Glenoid bone loss can be a contraindication to glenoid resurfacing.
Bone-grafting can restore glenoid volume and version. Neer and Morrison16 reported that only 4% (twenty)
of 463 shoulders required a large, internally fixed bone graft for
the treatment of glenoid deficiency at the time of total shoulder
arthroplasty.
The indications for bone-grafting in a patient with glenoid deficiency
include (1) uneven wear that cannot be accommodated by small changes
in glenoid or humeral component version and (2) insufficient volume
to support the glenoid component16.
Friedman et al.25 recommended
bone-grafting if the retroversion of the glenoid surface exceeded
15° as determined by computerized tomography. Lesser degrees of
bone loss can be compensated for by changing the humeral component
version, lowering the high side of the glenoid with reaming, or
using an augmented glenoid component. Alternatively, the surgeon
may decide not to resurface the glenoid but to perform a humeral
hemiarthroplasty. It is our opinion that the need for grafting can
be predicted preoperatively by evaluating glenoid volume and version
on a true axillary radiograph or with computerized tomography. Grafting
is required if the component would penetrate the cortex of the glenoid
neck after corrections in glenoid version are made.
There are few reports on the results of glenoid bone-grafting
in patients undergoing total shoulder replacement. Neer and Morrison16 evaluated nineteen shoulders after
an average duration of follow-up of fifty-two
months. All nineteen shoulders had a satisfactory result, and all
grafts healed. Six of the nineteen glenoid components were associated
with an incomplete radiolucent line that measured <1 mm
in width. Two screws broke, and one screw was worn because of contact
with the humeral component. Hulsey and Norris26 reported
on the first fourteen patients in the present study at an average
duration of follow-up of thirty-six months. There
were three reported failures. In the present study, there was one
late failure that occurred since the time of the initial evaluation and
another failure that occurred in a patient who was not involved
in the initial evaluation. Thus, the failure rate in the present
study (five of seventeen) was higher than that in the study by Neer
and Morrison (zero of nineteen)16.
The major difference between the two studies was in the use of metal-backed
glenoid components. A metal-backed component was used in
five of the nineteen shoulders in the study by Neer and Morrison16 as opposed to twelve of the seventeen
shoulders in the present study.
Fourteen of the seventeen glenoid bone grafts in the present
study healed in a proper position, which was maintained during an
average follow-up period of 5.8 years despite failure of
the glenoid component in two of the fourteen shoulders in which
the graft healed. All shoulders had some form of preoperative instability,
and nine of the seventeen had a return to normal stability with
the aid of grafting. Prosthetic failure occurred in five of the
seventeen shoulders. All five failures involved the glenoid component,
and all were treated with removal of the component. The failure
was related to persistent instability in two shoulders and to recurrent
rotator cuff tear, improper inferior placement of the glenoid component,
and inadequate fixation of the graft in one shoulder each. A satisfactory
functional result was obtained in nine of the seventeen shoulders.
The eight unsatisfactory functional results were related to prosthesis
failure (five shoulders), recurrence of a large rotator cuff tear
(two), and postoperative stiffness (one). Three of six shoulders had
a recurrence of a rotator cuff tear that had been repaired at the
time of the arthroplasty; the recurrence was associated with component
failure in one of the shoulders and with a poor functional outcome
in two.
There were differences in the morphology of the anterior and
posterior defects. All anterior defects were related to fracture-dislocation,
chronic dislocation, or recurrent anterior dislocation. The anterior
defects involved a smaller proportion of the glenoid surface than
did the posterior defects (average, 40% compared with 93%).
However, the average version and depth of the anterior defects (47° and
18 mm, respectively) were greater than those of the posterior defects
(27° and 12 mm, respectively). The grafts that were used for posterior
defects were technically more difficult to place and transfix than were
those used for anterior defects. One of the five components that
had been inserted with an anterior graft failed, compared with four
of the twelve components that had been inserted with a posterior graft.
One anterior graft and two posterior grafts failed to maintain the
original correction.
In conclusion, bone-grafting for patients with glenoid deficiency
is a technically demanding procedure. Patients who require glenoid
bone-grafting at the time of primary total shoulder replacement
have a tenfold higher rate of glenoid component failure than those
who have adequate glenoid version and volume. Patients with glenoid
deficiency frequently have preoperative glenohumeral instability.
The present long-term study demonstrated that glenoid bone-grafting
has the potential to restore and maintain volume and version in
patients undergoing total shoulder arthroplasty. When early failure
is avoided, there does not appear to be any tendency for the graft
to resorb or to predispose the glenoid component to early loosening.
Glenoid bone-grafting can facilitate the restoration of normal glenohumeral
stability when combined with appropriate soft-tissue balancing procedures.
In the present study, failure of this procedure was related to recurrence
of glenohumeral instability, rotator cuff tears, loss of graft fixation,
or improper component placement. These same factors also have been
shown to adversely affect the results of primary total shoulder
replacement.
Neer CS 2nd: Replacement arthroplasty for glenohumeral osteoarthritis. J Bone Joint Surg Am,1974.56: 1-13, 561
1974
[PubMed]
Bade HA 3rd, Warren RF, Ranawat CS,
Inglis AE. Long-term results of Neer total shoulder replacement.
In: Bateman JE, Welsh RP, editors. Surgery of the shoulder. St.
Louis: CV Mosby; 1984. p 294-302
Barrett WP; Franklin JL; Jackins SE; Wyss CR; and Matsen FA 3rd: Total shoulder arthroplasty. J Bone Joint Surg Am,1987.69: 865-72, 69865
1987
[PubMed]
Boyd AD; Thomas WH; Scott RD; Sledge CB; and Thornhill TS: Total shoulder arthroplasty versus hemiarthroplasty. Indications
for glenoid resurfacing. J Arthroplasty,1990.5: 329-36, 5329
1990
[PubMed]
Brenner BC; Ferlic DC; Clayton ML; and Dennis DA: Survivorship of unconstrained total shoulder arthroplasty. J Bone Joint Surg Am,1989.71: 1289-96, 711289
1989
[PubMed]
Clayton ML; Ferlic DC; and Jeffers PD: Prosthetic arthroplasties of the shoulder. Clin Orthop,1982.164: 184-91, 164184
1982
[PubMed]
Cofield RH: Total shoulder arthroplasty with the Neer prosthesis. J Bone Joint Surg Am,1984.66: 899-906, 66899
1984
[PubMed]
Figgie HE 3rd; Inglis AE; Goldberg VM; Ranawat CS; Figgie MP; and Wile JM: An analysis of factors affecting the long-term results
of total shoulder arthroplasty in inflammatory arthritis. J Arthroplasty,1988.3: 123-30, 3123
1988
[PubMed]
Hawkins RJ; Bell RH; and Jallay B: Total shoulder arthroplasty. Clin Orthop,1989.242: 188-94, 242188
1989
[PubMed]
McCoy SR; Warren RF; Bade HA 3rd; Ranawat CS; and Inglis AE: Total shoulder arthroplasty in rheumatoid arthritis. J Arthroplasty,1989.4: 105-13, 4105
1989
[PubMed]
Neer CS 2nd; Watson KC; and Stanton FJ: Recent experience in total shoulder replacement. J Bone Joint Surg Am,1982.64: 319-37, 64319
1982
[PubMed]
Wilde AH, Borden LS, Brems JJ. Experience
with the Neer total shoulder replacement. In: Bateman JE, Welsh
RP, editors. Surgery of the shoulder. St. Louis:
CV Mosby; 1984. p 224-8
Amstutz HC; Sew Hoy AL; and Clarke IC: UCLA anatomic total shoulder arthroplasty. Clin Orthop,1981.155: 7-20, 1557
1981
[PubMed]
Brems J: The glenoid component in total shoulder arthroplasty. J Shoulder Elbow Surg,1993.2: 47-54, 247
1993
Fenlin JM Jr; Vaccaro A; Andreychik D; and Lin S: Modular total shoulder: early experience and impressions. Semin Arthroplasty,1990.1: 102-11, 1102
1990
[PubMed]
Neer CS 2nd, and Morrison DS: Glenoid bone-grafting in total shoulder arthroplasty. J Bone Joint Surg Am,1988.70: 1154-62, 701154
1988
[PubMed]
Cofield RH: Integral surgical maneuvers in prosthetic shoulder arthroplasty. Semin Arthroplasty,1990.1: 112-23, 1112
1990
[PubMed]
Craig EV: Complications in shoulder arthroplasty. Semin Arthroplasty,1990.1: 160-71, 1160
1990
Moeckel BH; Altchek DW; Warren RF; Wickiewicz TL; and Dines DM: Instability of the shoulder after arthroplasty. J Bone Joint Surg Am,1993.75: 492-7, 75492
1993
[PubMed]
Wirth MA, and Rockwood CA Jr: Complications of shoulder arthroplasty. Clin Orthop,1994.307: 47-69, 30747
1994
[PubMed]
Norris TR. Bone grafts for
glenoid deficiency in total shoulder replacements. Proceedings of
the Third International Conference on Surgery of the Shoulder. Tokyo:
Professional Postgraduate Services; 1987. p 373-6
Richards RR; An KN; Bigliani LU; Friedman RJ; Gartsman GM; Gristina AG; Iannotti JP; Mow VC; Sidles J; and Zuckerman JD: A standardized method for the assessment of shoulder function. J Shoulder Elbow Surg,1994.3: 347-52, 3347
1994
Franklin JL; Barrett WP; Jackins SE; and Matsen FA 3rd: Glenoid loosening in total shoulder arthroplasty. Association
with rotator cuff deficiency. J Arthroplasty,1988.3: 39-46, 339
1988
[PubMed]
Cofield RH, and Edgerton BC: Total shoulder arthroplasty: complications and revision
surgery. Instr Course Lect,1990.39: 449-62, 39449
1990
[PubMed]
Friedman RJ; Hawthorne KB; and Genez B: Glenoid augmentation for total shoulder arthroplasty. Orthop Trans,1992.16: 66, 1666
1992
Hulsey RE, and Norris TR: Bone grafts for glenoid deficiency in total shoulder replacements. Orthop Trans,1990.14: 254, 14254
1990