The outcome of open and arthroscopic treatment of atraumatic
posteroinferior multidirectional instability of the shoulder is not
consistently
successful1-5,
partly because of the relatively low prevalence of the condition and our poor
understanding of the primary pathologic lesion. Atraumatic posteroinferior
multidirectional instability is seldom associated with labral
changes1,6-8.
Instead, increased volume and redundancy of the capsule have been considered
to be the main pathologic conditions responsible for the
instability1,2,7,9,10.
Therefore, correction of the increased capsular volume has become a
fundamental component of surgical treatment. However, another potential
lesion, retroversion of the glenoid, may also predispose a patient to
posterior instability of the
shoulder11-14.
Previous studies of glenoid version have focused only on the osseous
structures. However, glenohumeral joint stability is probably an integrated
function of bone and the soft-tissue stabilizers. The thickened cartilage at
the periphery15 and
the glenoid
labrum16-18
contribute to the stability of the joint by increasing the concavity of the
glenoid. Accordingly, a more complete measurement of glenoid version can be
obtained when the articular cartilage and the labrum are also considered.
The purpose of this study was to evaluate the chondrolabral containment of
the glenohumeral joint in shoulders with atraumatic posteroinferior
multidirectional instability. We used magnetic resonance imaging-arthrography
to measure the amount of version of the osseous and chondrolabral portions of
the glenoid, the labral height, and the glenoid depth in such shoulders, and
we compared these measurements with those in a normal control group.
Patient Selection
In the eight-year period between January 1995 and December 2002, a
preoperative magnetic resonance imaging-arthrogram was made for eighty-one
patients with recurrent atraumatic posteroinferior multidirectional
instability. The criteria for a diagnosis of such instability were (1) no
history of initiating trauma, (2) a sulcus
sign10 in both
shoulders, (3) a positive jerk
test19, (4) no
anterior apprehension sign (the fulcrum
test20), and (5) no
evidence of a Bankart lesion on the magnetic resonance imaging-arthrogram or
arthroscopic examination. Initiating trauma was defined as a distinctive
traumatic event that was followed by instability of an otherwise asymptomatic
shoulder. The sulcus
sign10 was measured
with use of inferior translation. A downward traction force was applied to the
adducted shoulder, and the inferior translation of the humerus was measured by
estimating the distance between the inferior margin of the lateral aspect of
the acromion and the humeral head. The jerk
test19 was
performed with the patient sitting. With the involved arm held in 90° of
abduction, adduction and a posteriorly directed axial load were simultaneously
applied to the glenohumeral joint. A sudden palpable or audible clunk and pain
were considered to indicate a positive test.
With the patient under anesthesia and in the lateral decubitus position,
anterior and posterior translation was assessed with use of the method of
Altchek et al.7. Any
intra-articular lesions, especially posteroinferior labral and capsular
lesions, were evaluated on arthroscopic examination.
We excluded patients who had traumatic unidirectional posterior instability
(since these patients frequently had a posterior labral tear), those who had
an anterior labral tear (since an anterior labral tear can affect the
measurements), those who were noted to have degenerative arthritic changes
during the arthroscopic examination, and those who had had previous surgery on
the shoulder.
Thirty-three shoulders in thirty-three patients met the inclusion criteria.
There were twenty-eight male and five female patients. The mean age was
twenty-four years, with a range of eighteen to thirty years. All patients had
pain during sports and daily activities such as those requiring a
follow-through motion or carrying a heavy object. The age-matched control
group consisted of twenty-two male and eleven female subjects (mean age,
twenty-three years; age range, seventeen to thirty years) who had volunteered
for the study. They did not have a history of shoulder trauma, and they had no
clinical evidence of instability on any of the tests described above. The
institutional review board of our center approved the study, and each patient
signed a detailed informed consent form.
Magnetic Resonance Imaging-Arthrography and Measurements
All of the patients and control subjects were examined with magnetic
resonance imaging-arthrography. A 1.5-T magnetic resonance imager (Signa;
General Electric Medical Systems, Milwaukee, Wisconsin) was used for all
studies. Just prior to the imaging, 2 to 3 mL of Iomeron 300 contrast medium
(Bracco Imaging, Milan, Italy) was injected to verify the location of the
joint space under fluoroscopy. Afterwards, 40 mL of saline solution mixed with
Magnevist (250:1; Schering, Berlin, Germany) was injected into the shoulder
joint. The magnetic resonance images were made with 4-mm-thick slices with a
1-mm gap between slices, a 16-cm field of view, and a matrix size of 256
× 192. The arm was held at the side of the trunk, and the shoulder was
held in neutral rotation with the thumb pointing upward. As seen on the
oblique midsagittal image, the axial T2-weighted spin-echo transverse axial
images were obtained at 25%, 50%, and 75% points along the long axis of the
glenoid (Fig. 1).
We used these images to evaluate four measurements thought to represent
glenohumeral containment: osseous and chondrolabral glenoid version, labral
height, and glenoid depth. A reference line drawn through the midpoint of the
transverse glenoid diameter at the level of the articular surface and the
medial rim of the scapular blade defined the axis of the scapula. The angle
between a line perpendicular to this reference line and a line drawn through
the anterior and posterior aspects of the glenoid labrum defined the version
of the chondrolabral portion of the glenoid. The angle between a line
perpendicular to the reference line and a line drawn through the anterior and
posterior aspects of the glenoid rim defined the version of the osseous
portion of the glenoid (Fig.
2). Labral height was defined as the shortest distance from a
tangential line drawn through the deepest point of the glenoid articular
surface to the tip of the anterior or posterior aspect of the labrum. Glenoid
depth was defined as the distance from the deepest point of the glenoid
articular surface to a line drawn through the tips of the anterior and
posterior aspects of the labrum (Fig.
3).
Positive angles represented retroversion, and negative angles represented
anteversion. All measurements were performed three times by three authors
(K.-C.N., J.-S.P., and B.-D.R.) who were blinded with regard to the shoulder's
group (control or instability). To assess intraobserver reliability, the
measurements were repeated twice for all patients after one week. The mean of
the values of the three observers was considered to be the final value. All
measurements were performed on the PACS (Picture Archiving and Communications
System) monitor (General Electric, Chicago, Illinois) with use of a mouse
pointer (cursor) and automated computer calculation of the angle and
length.
Statistical Analysis
Patient Sample Size
Since the aim of the study was to compare two different groups, the null
hypothesis (Ho) was that there is no difference in the version of the
chondrolabral portion of the glenoid between two groups (Ho: Rµ1 = Rµ2).
To determine the appropriate sample size, we utilized a statistical program
(nQuery Advisor 3.0; Statistical Solutions, Cork, Ireland), which incorporated
difference means. With a 0.05 two-sided significance level and 80% power, it
was postulated that the expected difference in mean glenoid version was 5°
between the two groups and the common standard deviation was 6.05° (the
largest standard deviation among the angles between the two groups). With the
assumption that a t test would be used, with p < 0.05 indicating
significance, the sample size was determined to be thirty-three patients in
each group.
Outcome Analyses
Intraobserver reliability of repeated measurements was assessed with use of
the correlation coefficient between the measurements at two different times.
An independent-sample t test was used for the comparisons between the two
groups. One-way analysis of variance was used for the comparisons between the
three planes within each group. The paired-sample t test was used to compare
any difference in version between the osseous and the chondrolabral portion of
the glenoid. Statistical analyses were performed with the alpha value set at
0.05 with use of the SPSS software package (SPSS for Windows, Release 11.0;
SPSS, Chicago, Illinois).
Accuracy of Measurements
The reliability of the repeated measurements by one observer was high, with
the intraobserver reliability coefficient ranging from 0.84 to 0.98. The
interobserver reliability coefficient ranged from 0.80 to 0.92.
Glenoid Version
Compared with the control group, the shoulders with posteroinferior
instability had more retroversion of the osseous and chondrolabral portions of
the glenoid in the middle and inferior planes (p < 0.05)
(Table I). There was no
significant difference in the version in the superior plane between the two
groups.
The chondrolabral portion of the glenoid was more retroverted than the
osseous portion of the glenoid (7.0° ± 4.9° compared with
4.6° ± 4.5°) in the inferior plane (p = 0.008) in the group
with instability (Fig. 4).
There was no significant difference between chondrolabral and osseous
retroversion in the other planes in the unstable shoulders or in any of the
three planes in the control group (p > 0.05).
Labral Height
The height of the posterior aspect of the labrum in the inferior plane in
the unstable shoulders was significantly less than the height in the control
group; it was also significantly less than the heights in the other two planes
in the unstable shoulders (p < 0.001). No other significant differences
with regard to the height of the anterior and posterior aspects of the labrum
were noted (Table II).
Glenoid Depth
The glenoid depth in the middle and inferior planes was significantly
shallower in the unstable shoulders than it was in the control group (p <
0.05). There was no significant difference in glenoid depth in the superior
plane between the two groups (Table
III).
In this study of patients with atraumatic posteroinferior
multidirectional instability of the shoulder, the posteroinferior aspect of
the labrum was consistently abnormal in its configuration. Decreased
containment resulting from the loss of height and retroversion of the
posteroinferior aspect of the labrum was a consistent finding in these
patients.
A large number of factors influence the stability of the glenohumeral
joint. In addition to the dynamic stabilizers such as a compressive load
throughout the range of glenohumeral
motion21,
intra-articular negative
pressure22,23,
and positioning of the glenoid, a number of static stabilizers, such as the
capsule and its
ligaments24-26,
play an important role. Although the posterior aspect of the labrum has been
implicated in posterior instability of the shoulder, and previous reports have
described various posteroinferior labral lesions in shoulders with atraumatic
posteroinferior multidirectional
instability6,27,
pathologic changes of the labrum and the important stabilizing role of the
labrum have not been fully
elucidated28-31.
Since Neer and
Foster10 described
the inferior capsular shift for the treatment of inferior and multidirectional
instability, the concept of a redundant capsular ligament as a primary cause
of multidirectional instability has been popularized. The outcomes of the
inferior capsular shift have been
reasonable32,
although less satisfactory results have been reported in some
series3.
Excessive retroversion of the glenoid has been considered a developmental
deformity and a primary cause of posterior
instability12,14,33;
however, treatment of this problem with posterior glenoid osteotomy has had
variable success rates and
complications12-14,34,35.
Other
investigators11-14
have measured version of the osseous portion of the glenoid on computed
tomography images, magnetic resonance images, or plain radiographs. Anatomical
studies12 have
established that the normal retroversion angle of the osseous portion of the
glenoid is about 4°. Weishaupt et
al.33 confirmed
this value with measurements in a healthy control group of fifty volunteers,
in whom the angle averaged 4.4°. Also, all of their patients with
posterior recurrent instability of the shoulder had a retroverted osseous
portion of the glenoid (mean, 7.8°; range, 3° to 21.4°). However,
the static stability of the glenohumeral joint is an integral function of both
the osseous and the ligamentous structures. The labrum and the
capsule-ligamentous complex are considered to be the main sources of anterior
stability16-18,24-26.
Lippitt and
Matsen21
demonstrated a 20% decrease in mechanical stability associated with removal of
the glenoid labrum. Lazarus et
al.17 found a 65%
decrease in the mechanical stability ratio associated with the creation of an
anteroinferior chondrolabral defect that resulted in an 80% reduction in the
height of the glenoid. Version of the glenohumeral joint is measured more
completely when the contribution of both the articular cartilage and the
labrum is considered.
Cooper et al.36
demonstrated that the inferior part of the glenoid labrum is usually a
rounded, elevated fibrous structure that is firmly continuous with the
articular cartilage. There are few differences between the morphology of the
anteroinferior part of the labrum and that of the posteroinferior part.
Although the anterior part may be slightly thicker, the dimensions of the two
parts do not differ substantially. Our study demonstrated that the height of
the posteroinferior part of the labrum was similar to that of the
anteroinferior part in the normal population.
It is unclear whether the retroversion of the posteroinferior portion of
the labrum noted in our patients with atraumatic posteroinferior
multidirectional instability was a cause or a consequence of the instability.
We hypothesized that the posteroinferior aspect of the capsule may pull the
posteroinferior part of the labrum away when the humeral head subluxates
posteroinferiorly and repeated subluxation of the humeral head over the
glenoid rim may compress the triangular shape of the labrum. Regardless of
whether abnormal containment is a cause or a consequence, its correction may
improve stability in these patients. A simple inferior capsular shift or
arthroscopic capsular plication can address the capsular redundancy
successfully. However, those procedures do not correct the abnormal
containment. Capsulolabroplasty, which we recently
described37,
restores the abnormal containment by creating a full-thickness tear around the
rim of the posteroinferior aspect of the labrum and then reattaching the
labrum to the surface of the glenoid with capsular plication. This procedure
was successful in thirty-three of thirty-four patients in whom we performed it
to treat posteroinferior multidirectional
instability37.
In conclusion, excessive retroversion and flattening of the chondrolabral
portion of the glenoid resulting in the loss of containment of the humeral
head were observed in all of our patients with atraumatic posteroinferior
multidirectional instability. We think that this is an important contributor
to posteroinferior shoulder instability and that it is a factor that should be
considered during surgical repair. ?