Between 1980 and 1997, six patients (seven shoulders) with glenoid
dysplasia and osteoarthritis were treated with shoulder arthroplasty
at our institution because of moderate or severe shoulder pain that
failed to respond satisfactorily to nonoperative management. Four
hemiarthroplasties and three total shoulder arthroplasties were
performed. One patient underwent bilateral total shoulder arthroplasty.
All seven shoulders with complete preoperative evaluation and operative
records that had been followed for a minimum of two years or until
the time of revision surgery were included in the study. The average
duration of follow-up was 7.3 years (range, 1.3 to sixteen years).
Four shoulders required revision surgery: one, at sixteen months;
one, at twenty months; one, at thirty-four months; and one, at 5.8
years. Each of the remaining three shoulders was followed for at
least ten years.
The diagnosis was based on preoperative radiographic findings
that were consistent with glenoid dysplasia
1,4,5
. Additionally, the diagnosis was confirmed at the time of surgery.
All of the patients were men, and their ages ranged from thirty-nine
to sixty years (average, fifty-one years). One patient underwent
arthroscopic d�bridement of a labral tear three years prior to the
shoulder arthroplasty. The remaining six shoulders had no history
of surgery.
Clinical Review
Pain was graded on a scale of 1 through 5, as previously described
by Neer et al.
6
and by Cofield
7
: 1 point was assigned when there was no pain; 2 points were assigned
for slight pain; 3 points, for pain after unusual activities; 4
points, for moderate pain; and 5 points, for severe pain. Active
elevation and external rotation were recorded in degrees, and internal
rotation was graded according to the posterior spinal region that
the thumb could reach actively.
Radiographic Review
The radiographs of the patients who had had a hemiarthroplasty
were reviewed to determine the presence or absence of glenohumeral
subluxation, periprosthetic radiolucency, subsidence or a shift
in the position of the humeral component, loss of glenoid cartilage,
and erosion of the glenoid. The amount of glenoid erosion, which
was determined in relation to the anterior subchondral plate, was
graded as mild (<5 mm), moderate (5 to 10 mm), or severe (>10
mm). The radiographs of the patients who had had a total shoulder
arthroplasty were reviewed to determine the presence or absence
of glenohumeral subluxation, periprosthetic radiolucency, subsidence
or a shift in the position of the humeral component, and migration
or tilt of the glenoid component.
Periprosthetic lucency was classified as grade 0 when there was
no radiolucent line, grade 1 when the line was ≤1 mm wide
and incomplete, grade 2 when the line was 1 mm wide and complete,
grade 3 when the line was 1.5 mm wide and incomplete, grade 4 when
the line was 1.5 mm wide and complete, and grade 5 when the line
was 2 mm wide and complete. Glenohumeral subluxation was evaluated
with regard to its direction and the amount of translation of the
center of the prosthetic head relative to the center of the glenoid
or the glenoid component. It was recorded as none, mild when there was <25%
translation, moderate when there was 25% to 50% translation, and
severe when there was >50% translation. Migration or tilt of the
glenoid component was recorded as either present or absent, as was
subsidence or shift of the humeral component.
Operative Technique
Following implantation of the trial humeral component, the glenoid
was inspected. Three shoulders with degenerative changes of the
glenoid articular surface underwent total shoulder arthroplasty.
Two of the glenoid components-an all-polyethylene Neer component
and a metal-backed Neer component-were implanted with cement. A
Cofield tissue-ingrowth glenoid component was placed in the third
shoulder. One shoulder underwent bone-grafting of the posterior
aspect of the glenoid with bone obtained from the humeral head. Another
shoulder underwent z-lengthening of the subscapularis.
Four shoulders underwent hemiarthr7oplasty, and, in one of them,
a glenoid osteotomy was performed from a posterior approach. In
all four shoulders, at least one-half of the glenoid articular cartilage
was noted to be completely worn, exposing subchondral bone.
Revisions
Four shoulders, three originally treated with hemiarthroplasty and
one originally treated with total shoulder arthroplasty, underwent
revision surgery. All three revisions of the hemiarthroplasties
were performed because of pain associated with glenoid arthrosis.
Of the shoulders originally treated with hemiarthroplasty, one underwent
revision to total shoulder arthroplasty with glenoid bone-grafting
and placement of a glenoid component twenty months following the
hemiarthroplasty (
Figs. 1-A
,
1-B
,
1-C
,
1-D
,
1-E
, and
1-F
). Another underwent revision to total shoulder arthroplasty with
placement of a glenoid component and exchange of the humeral head
sixteen months following the hemiarthroplasty. The third shoulder
underwent revision to bipolar arthroplasty at another institution
thirty-four months following the hemiarthroplasty. The patient originally
treated with total shoulder arthroplasty underwent component removal
because of a suspected Pseudomonas infection (positive preoperative
aspirate) and a loose glenoid component 5.8 years following the
total shoulder arthroplasty. All cultures of intraoperative specimens were
negative. The patient underwent reimplantation of a hemiarthroplasty
component two months following resection.
Pain
All shoulders had moderate or severe pain prior to surgery. Three
shoulders with glenoid arthrosis had moderate or severe pain after
hemiarthroplasty. Each of these shoulders underwent revision surgery.
The remaining four shoulders had no or slight pain.
Range of Motion
Result Rating
The total shoulder arthroplasties were followed by two excellent
results and one unsatisfactory result. The hemiarthroplasties were
followed by one excellent result and three unsatisfactory results.
Radiographic Analysis
Total Shoulder Arthroplasty
Preoperative radiographs were available for two of the three shoulders
that underwent total shoulder arthroplasty. Both shoulders had moderate
glenoid hypoplasia according to the scale of Wirth et al.
4
, and one of them had mild posterior and superior subluxation preoperatively.
The most recent follow-up radiographs showed no periprosthetic
lucency adjacent to any of the three humeral components, but there
was a radiolucent line adjacent to each of the three glenoid components.
The metal-backed Neer glenoid component was surrounded by a 1.5-mm
complete lucent line, and there was a shift in component position
consistent with loosening of the component. There was a complete
lucent line of >2 mm around the all-polyethylene Neer component, which
had also shifted in a manner consistent with loosening. There was
a 1-mm incomplete lucent line adjacent to the Cofield tissue-ingrowth
component, which had not shifted in position. At the most recent
follow-up evaluation, one shoulder had mild posterior and superior
subluxation.
Hemiarthroplasty
Of the four shoulders that underwent hemiarthroplasty, one had
had moderate preoperative hypoplasia and three had had severe preoperative
hypoplasia according to the scale of Wirth et al.
4
. Preoperatively, three shoulders had had subluxation: moderate
posterior and mild superior subluxation had been noted in two of
these shoulders, and mild posterior and superior subluxation had
been noted in one. There was wear of the glenoid cartilage in all
four shoulders.
At the most recent follow-up evaluation, glenohumeral subluxation
was present in all four shoulders: mild posterior and superior subluxation
was seen in two shoulders; moderate posterior subluxation, in one;
and moderate posterior and mild superior subluxation, in one. There
was a 1.5-mm incomplete radiolucent line adjacent to two humeral
components, without shift or subsidence of either component. There was
glenoid bone erosion in three shoulders: it was mild in two and
moderate in one.
Glenoid dysplasia is a developmental anomaly of the scapula that
is characterized by incomplete ossification of the inferior aspect
of the glenoid and the scapular neck
9
. Glenoid dysplasia has also been termed "glenoid hypoplasia" and
"dentated glenoid."
1
Ossification of the glenoid fossa occurs through two separate ossification
centers in normal shoulder development10. The first ossification
center appears around the tenth year and closes by the fifteenth
year, and it originates in the superior aspect of the glenoid fossa,
at the base of the coracoid
10
. The ossification center of the inferior aspect of the glenoid fossa
appears around puberty10. Glenoid dysplasia is thought to result
from a failure of development of the inferior aspect of the apophysis1.
Typically, radiographic findings include dysplasia of the neck of
the scapula with an irregular articular surface of the glenoid
1,5
. Associated findings may include dysplasia of the humeral head,
dysplasia of the coracoid, and hooking of the lateral part of the
clavicle3. Additionally, the acromion may be elongated and inferiorly
directed
4
. Patients with glenoid dysplasia typically have symptoms in adolescence
or in the fifth and sixth decades of life, and this condition has
been reported more frequently in men11.
We are not aware of any prior reports addressing the management
of glenoid dysplasia in the setting of shoulder arthroplasty. To
our knowledge, the largest report on glenoid hypoplasia in the orthopaedic
literature is by Wirth et al.4. In their study of sixteen patients,
two had glenohumeral degenerative joint disease. Neither of these
patients underwent surgery. The authors noted that hemiarthroplasty
"may play a role in the rare patient who has unremitting symptoms
and moderate or severe degenerative changes of the shoulder joint."
4
The management of these patients requires a thorough preoperative
examination that includes careful evaluation of the preoperative
radiographs. Determination of the severity of the glenoid dysplasia
preoperatively will facilitate adequate preparation for surgery.
Preoperative computed tomography scanning with or without three-dimensional
reconstruction, although not used for the patients in this study,
may improve the ability to determine the degree of glenoid bone
deficiency.
In the current study, posterior deficiency occurred in conjunction
with inferior deficiency. The radiographs demonstrated frank deficiency
of bone formation in the posterior region rather than the usual
pattern of posterior wear seen in osteoarthritis. Additionally,
at the time of surgery, the deficiency of inferior and posterior
bone was in contrast to the usual appearance of posterior wear alone
that is typically seen in osteoarthritis.
The data from this study suggest that glenoid deficiency and cartilage
wear should be addressed at the time of shoulder arthroplasty in
patients with glenoid dysplasia. Currently, there are three options
for the treatment of the inferior and posterior bone deficiency.
First, one can consider bone-grafting the deficient region in conjunction
with placement of a glenoid component
12
. Second, one can consider bone-grafting or osteotomy alone in conjunction
with a hemiarthroplasty. Finally, one can accommodate the deficient
inferior-posterior region with a glenoid component that substitutes
metal for the deficient bone.
Hemiarthroplasty alone appears to be an unsatisfactory option for
the treatment of this disease process. Three of the four hemiarthroplasties
in our series had to be revised because of glenoid arthrosis. Radiographic
review demonstrated that these three shoulders had substantial inferior
and posterior glenoid deficiency that had not been corrected at
the time of the primary arthroplasty. The one shoulder that did
not require revision surgery after hemiarthroplasty had a glenoid
osteotomy performed at the time of the hemiarthroplasty.