DOI:10.2214/AJR.04.1684
AJR 2005; 185:1422-1428
© American Roentgen Ray Society
Width of High Signal and Extension Posterior to Biceps Tendon as Signs of Superior Labrum Anterior to Posterior Tears on MRI and MR Arthrography
Michael J. Tuite,
Anthony Rutkowski,
Timothy Enright,
Lee Kaplan,
Jason P. Fine and
John Orwin
Department of Radiology, E3/311, University of Wisconsin Medical School,
600 Highland Ave., Madison, WI 53792.
Received October 28, 2004;
accepted after revision December 16, 2004.
Address correspondence to M. J. Tuite
(mjtuite{at}facstaff.wisc.edu).
Abstract
OBJECTIVE. The purpose of our study was to determine the accuracy of
two signs for superior labrum anterior to posterior (SLAP) tears: increased
width of high signal between the superior labrum and glenoid, and high signal
posterior to the biceps tendon.
MATERIALS AND METHODS. Forty-one patients with SLAP tears and 40
patients without a tear at surgery who had undergone MRI or MR arthrography
were retrospectively evaluated. The MR studies were combined and interpreted
in a blinded manner. The reviewers measured the width of high signal that
extended to the articular surface on oblique coronal images and determined
whether the high signal extended posterior to the biceps. A Student's
t test was used to determine statistical significance between the
means of the signal width.
RESULTS. High-signal width was greater in patients with a SLAP tear
than in the control group on both MRI and MR arthrography (both p =
0.003). The sensitivity and specificity of at least 2.0 mm on MRI are 39%
(11/28) and 89% (24/27) and at least 2.5 mm on MR arthrography are 46% (6/13)
and 85% (11/13). The sensitivity and specificity of high signal posterior to
the biceps are 54% (15/28) and 74% (20/27) on MRI and 69% (9/13) and 54%
(7/13) on MR arthrography.
CONCLUSION. Increased width of high signal has a moderate
specificity but a poor positive predictive value for distinguishing a SLAP
tear from a normal recess. In addition, labral signal posterior to the biceps
tendon is not rare in patients with no SLAP tear.
Introduction
Superior labrum anterior to posterior (SLAP) tears of the glenoid
labrum are a well-recognized cause of shoulder pain
[1,
2]. Several patterns of tears
can be seen in SLAP lesions, and these are often classified as one of 10 types
[3]. Type 1 SLAP lesions
represent degenerative fraying of the superior labrum and typically do not
require surgery [4]. Types 2-10
SLAP tears are present in 4-6% of patients and are usually painful
[1,
5,
6]. Arthroscopic repair or
débridement is often necessary to relieve the pain in these patients
[2,
7].
MRI is useful for diagnosing SLAP tears and therefore identifying patients
who may benefit from surgery. Multiple MRI and MR arthrographic signs of a
SLAP tear have been described in the literature
[3,
5,
8-15].
Several of the more commonly cited MR signs are an irregular high-signal line
in the superior labrum extending to the articular surface, a laterally curved
high-signal line, a detached superior labrum, and two high-signal lines (the
so-called double Oreo sign). All of these MR signs attempt to distinguish a
pathologic tear from the superior sublabral recess, which is considered a
normal variant and is present in about 75% of individuals
[16]. This recess is seen on
MR images as a smooth, medially curved line of high signal between the
superior labrum and the adjacent glenoid rim.
Several studies have found that these MR signs may be absent in some
patients with SLAP tears [17,
18]. Some SLAP tears occur by
partial detachment of the superior labrum from the adjacent glenoid rim, and
these tears, which do not extend laterally into the labrum itself, can be
particularly difficult to distinguish on MR images from a normal superior
recess. For this reason, several authors also include two additional MR signs
of a SLAP tear. The first is abnormal medial-lateral width of the high signal
between the labrum and the superior rim of the glenoid fossa
[8,
12-14].
The second sign is extension of the high signal between the labrum and the
superior rim of the glenoid fossa posteriorly to the posterior aspect of the
attachment of the long head of the biceps tendon
[11,
15,
16]. Our purpose was to
determine the accuracy of these two MR signs in a group of surgically proven
SLAP tears.
Materials and Methods
This study was approved and a waiver of consent granted by our
institutional review board.
Forty-one consecutive patients who underwent shoulder MRI followed by
arthroscopy on which a type 2 or greater SLAP tear was diagnosed were included
in the study. During the study period between 1999 and 2003, 28 patients
underwent a conventional MRI examination and 13 underwent direct MR
arthrography with intraarticular gadolinium. At arthroscopy, 31 patients had a
type 2 SLAP tear, three had a type 3, six had a type 4, and one had a type
5.
The patients were 33 males and 8 females, with a mean age of 41 years
(range, 17-78 years). A history of injury was given by 29 of the patients: 7
from throwing, 6 from other sports, 6 after a fall, 4 from weightlifting, and
6 as a result of unspecified trauma. The other 12 patients had chronic pain
with no injury recalled. A SLAP tear was interpreted prospectively in the
original MRI report by one of seven musculoskeletal radiologists in 12/28
conventional MR examinations and 8/13 MR arthrographic examinations. After MRI
and physical examination, the preoperative diagnosis included a SLAP tear in
11 patients (all of which tears were also seen on MRI). The other preoperative
diagnoses were impingement in 19, instability with or without impingement in
7, and biceps tendonosis or glenohumeral arthrosis in 4 patients. After
surgery, a SLAP tear was the only finding seen in 7 patients. Additional
findings in the other 34 patients were a rotator cuff tear in 6; cuff tear and
chondromalacia in 4; cuff tear and biceps tendon tear in 8; cuff tear, biceps
tear, and SLAP tear in 3; cuff tear and labral tear in 5; labral tear in 4;
biceps tear in 1; labral tear and biceps tear in 1; and acromioclavicular
joint arthrosis in 2.
A control group was included in the study and consisted of 40 studies in 37
consecutive patients scanned during a portion of the same time period who
underwent shoulder MRI followed by arthroscopy in whom at surgery the superior
labrum and insertion of the long head of the biceps tendon were normal or had
only mild fraying. Twenty-four patients underwent conventional MRI, 10
underwent MR arthrography with intraarticular gadolinium, and 3 patients
underwent both examinations.
The control group consisted of 30 males and 7 females, with a mean age of
44 years (range, 16-75 years). In the 40 studies, a SLAP tear was incorrectly
interpreted prospectively in the original MRI report in 1/27 conventional MRI
examinations and in 2/13 MR arthrographic examinations. After MRI and physical
examination, the surgeon's preoperative diagnoses were SLAP tear in 1 patient,
impingement in 25, instability in 8, instability and impingement in 1, biceps
tendonosis in 1, and bursitis in 1 patient. After surgery, the findings were a
rotator cuff tear in 6; cuff tear and biceps tendon tear in 5; cuff tear,
biceps tear, and anterior or posterior labral tear in 2; cuff tear and labral
tear in 6; labral tear in 6; biceps tear in 2; and synovitis and labral
fraying in 10.
All MR images were obtained on a 1.5-T scanner using a phased-array
shoulder coil (IGC, Medical Advances). Patients were scanned with the arm
adducted at the side and, whenever possible, in comfortable mild external
rotation. Conventional MR images obtained in 55 patients included oblique
coronal fat-suppressed fast spin-echo T2-weighted images (TR/TE, 2,033/60;
echo-train length, 6; and 3 excitations) with a 4-mm section thickness and a
1-mm interslice gap. The transverse images were fat-suppressed fast spin-echo
intermediate-weighted images (2,000/34; echo-train length, 5; and 3
excitations) with a 3-mm section thickness with 1-mm interslice gap. Oblique
sagittal T2-weighted and axial T1-weighted images were also obtained but were
not used in this study. All images were obtained with a 14-cm field of view
and a 256 x 192 matrix.
Twenty-six patients underwent MR arthrography as requested by the referring
physician. After informed written consent was obtained, a 22-gauge needle was
placed in the joint under fluoroscopic control. Twelve to 15 mL of a 1:200
dilution of gadolinium (Omniscan [gadodiamide], Amersham Health) was instilled
into the joint. The patient was escorted immediately to the MR scanner.
MR arthrographic images included the same oblique coronal fat-suppressed
fast spin-echo T2-weighted images as for the conventional MRI and oblique
coronal fat-suppressed T1-weighted images (633/18, 1 excitation) with a 4-mm
section thickness and 1-mm interslice gap. The transverse images were
fat-suppressed T1-weighted images (667/17, 1.5 excitations) with a 3-mm
section thickness and 1-mm interslice gap. Oblique sagittal and
abduction-external rotation oblique axial T1-weighted images were also
obtained but were not used in this study.
The mean interval between MRI and arthroscopy was 100 days (range, 1-427
days). Most of the arthroscopies were performed by a single orthopedic
shoulder surgeon with 14 years of postfellowship experience; five were
performed by a second orthopedic shoulder surgeon at our institution with 1
year of postfellowship experience. The surgeons diagnosed a SLAP tear if there
was hemorrhage or evidence of fiber tearing either in the superior labrum or
where the superior labrum was detached from the superior glenoid. If these
findings were equivocal, they also diagnosed a SLAP tear if the biceps anchor
was unstable and it was consistent with the patient's mechanism of injury,
symptoms, and clinical examination.
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —65-year-old man with normal superior recess and minimal
degenerative fraying of superior labrum at surgery. Maximum width of high
signal between superior labrum (arrow) and glenoid rim occurred at
articular surface and measured 2.0 mm (arrowheads) on this oblique
coronal fat-suppressed fast spin-echo T2-weighted conventional MR image.
Patient had full-thickness rotator cuff tear confirmed at surgery.
|
|
The oblique coronal and axial MR images were reviewed on a PACS
workstation. The examinations of the patients with SLAP tears and the control
patients were combined, and the studies were evaluated in alphabetical order
by last name with the reviewer blinded to the patient's medical history and
arthroscopic results. The measurements were made and the studies scored in
consensus by a fellowship-trained musculoskeletal radiologist with 12 years'
experience and by a fourth-year medical student.
The reviewers first determined whether high signal (brighter than and
distinct from the hyaline cartilage) in the superior labrum extended to the
articular (inferior) surface on an oblique coronal image. The MR images were
magnified 11 times, and the reviewers measured the maximum width of the
high-signal line down to 0.1 mL using the linear distance tool on the PACS
workstation (Figs. 1,
2A, and
2B). If no high-signal line was
seen, the width was recorded as 0. For the MR arthrographic images, either the
T1- or T2-weighted image was used, whichever gave the widest line (Figs.
3 and
4). Care was taken not to
measure the line on the most anterior oblique coronal image through the
anterior glenoid, where a sublabral foramen might be present. We also did not
obtain any measurements on far anterior or posterior oblique coronal images on
which the labrum is curving in the acquired image section and the margins of
the high signal were indistinct because of partial averaging. Eight weeks
later, the high-signal width was measured again by the musculoskeletal
radiologist. A different fourth-year medical student also separately measured
the width while blinded to the patient's medical history and arthroscopic
results to determine interobserver variability. The three measurements were
averaged, and these values were used to calculate the mean, sensitivity, and
specificity.
View larger version (159K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —34-year-old man with type 2 superior labrum anterior to
posterior (SLAP) tear at surgery. Two adjacent oblique coronal fat-suppressed
fast spin-echo T2-weighted conventional MR images show high signal extending
into superior labrum (arrows). Greatest width, 3.1 mm
(arrowheads, A), was obtained at focal widening of high signal
on more posterior image (A).
|
|
View larger version (158K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —34-year-old man with type 2 superior labrum anterior to
posterior (SLAP) tear at surgery. Two adjacent oblique coronal fat-suppressed
fast spin-echo T2-weighted conventional MR images show high signal extending
into superior labrum (arrows). Greatest width, 3.1 mm
(arrowheads, A), was obtained at focal widening of high signal
on more posterior image (A).
|
|
View larger version (103K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3 —20-year-old man with normal superior recess at surgery. Mild
focal widening of increased signal between labrum and glenoid rim several
millimeters above articular surface is seen on this oblique coronal
fat-suppressed T1-weighted MR arthrogram. Area of increased signal was thought
to be distinct from hyaline cartilage (arrow). Width measured 2.4 mm
(arrowheads).
|
|
View larger version (100K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —29-year-old woman with type 2 superior labrum anterior to
posterior (SLAP) tear at surgery. Width of high-signal line measured 3.2 mm
(arrowheads) on this oblique coronal fat-suppressed T1-weighted MR
arthrogram. Portion of high signal has a laterally curved branch (lateral
arrowhead); a SLAP tear was interpreted on original MR image.
|
|
The most posterior oblique coronal image that showed a high-signal line was
then localized on the transverse image that showed the insertion of the long
head of the biceps tendon into the superior labrum. The reviewers recorded
whether this oblique coronal image was posterior to the posterior aspect of
the tendon at its insertion (Figs.
5A,
5B,
5C,
5D,
6A, and
6B). If the biceps tendon at
its insertion could not be identified, the finding was recorded as
"biceps not seen."
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —39-year-old man with normal superior recess and minimal
degenerative fraying of superior labrum at surgery. Two consecutive oblique
coronal fat-suppressed fast spin-echo T2-weighted conventional MR images show
smooth medially curved high signal (arrows) between superior labrum
and cartilage at top of glenoid rim (arrowheads). Image A is
one section posterior to image B.
|
|
View larger version (138K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —39-year-old man with normal superior recess and minimal
degenerative fraying of superior labrum at surgery. Two consecutive oblique
coronal fat-suppressed fast spin-echo T2-weighted conventional MR images show
smooth medially curved high signal (arrows) between superior labrum
and cartilage at top of glenoid rim (arrowheads). Image A is
one section posterior to image B.
|
|
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C —39-year-old man with normal superior recess and minimal
degenerative fraying of superior labrum at surgery. Screen capture from PACS
workstation shows localization (curved arrows) of image A
(posterior) on an axial image through upper glenoid fossa. Oblique coronal
section is posterior to portion of superior glenoid rim included on axial
image.
|
|
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5D —39-year-old man with normal superior recess and minimal
degenerative fraying of superior labrum at surgery. Screen capture from
adjacent more superior axial image shows localization (curved arrows)
of image B ("Mid") and a portion of long head of biceps
tendon (straight arrow). Back of long head of biceps tendon at
insertion onto labrum is indicated with arrowhead. Full-thickness rotator cuff
tear was confirmed at surgery.
|
|
View larger version (97K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —20-year-old man with normal superior recess at surgery.
Oblique coronal fat-suppressed T1-weighted MR arthrogram shows medially curved
high signal (arrow) between labrum and glenoid rim.
|
|
View larger version (101K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —20-year-old man with normal superior recess at surgery.
Screen capture from PACS workstation shows localization (curved
arrow) of image A on axial image through top of glenoid rim.
Anterior portion of long head of biceps tendon (solid straight arrow)
and posterior aspect of tendon at its insertion onto labrum
(arrowhead) are shown. High-signal diluted gadolinium contrast
material is also seen in recess on axial image (broken arrow).
|
|
The MRI findings were compared with the arthroscopic results, which were
used as the gold standard. The Student's two-sample t test was used
to determine the difference between the means for the high-signal widths. The
differences in width measurements between the consensus interpretation and the
subsequent interpretation made by the musculoskeletal radiologist alone, and
the interpretations made by the separate observers, were determined using the
interclass correlation coefficient for continuous data. The Fisher's exact
test was used to determine statistical significance for high signal posterior
to the biceps tendon. A p value of less than 0.05 was considered
significant.
Results
Width of High Signal
The width of the high-signal line on conventional MR images varied between
0 and 2.6 mm for the 27 control patients, and between 0 and 3.5 mm for the 28
patients with a SLAP tear (Table
1). The mean width for the control patients was 0.9 mm (±
0.8 mm [SD]), and for the SLAP tear group was 1.5 mm (± 1.0 mm)
(p = 0.003). A width of at least 2.0 mm had a sensitivity for
diagnosing a SLAP tear of 39% (11/28) and a specificity of 89% (24/27). We
also computed from our data the threshold width for various fixed
specificities (Table 2). By
this calculation, a specificity of 90% has a threshold width of 2.1 mm and a
sensitivity of 36%. Using a threshold of 2.1 mm in a population with a SLAP
tear prevalence of 5% would give a calculated positive predictive value (PPV)
of 14%.
For MR arthrography, the width of the high-signal line varied between 0 and
3.0 mm for the 13 control patients, and between 0.8 and 4.4 mm for the 13 SLAP
patients (Table 3). The mean
width for the control patients was 1.1 mm (± 1.2 mm) and for the SLAP
tear group was 2.3 mm (± 1.0 mm) (p = 0.003). A width of at
least 2.5 mm on MR arthrography had a sensitivity for a SLAP tear of 46%
(6/13) and a specificity of 85% (11/13). All patients with a SLAP tear (13/13)
had at least some high signal between the labrum and the glenoid rim, whereas
46% (6/13) of the control group had a 0-mm width. We again computed from our
data the threshold width for various fixed specificities
(Table 4). By this calculation,
a specificity of 90% has a threshold width of 3.0 mm and a sensitivity of 31%.
Using a threshold width of 3.0 mm in a population with a SLAP tear prevalence
of 5% would again give a calculated PPV of only 14%.
The correlation coefficient between the initial consensus interpretation
and that done later by the musculoskeletal radiologist alone was 0.87, with an
average absolute difference of 0.3 mm (± 0.5 mm). For MR arthrography,
this correlation coefficient was 0.9, with an average difference of 0.4 mm
(± 0.5 mm). The interobserver correlation coefficient between the
separate observers on the MRI images was 0.86, with an average difference of
0.3 mm (± 0.6 mm). For MR arthrography, the interobserver correlation
coefficient was 0.83, with a difference of 0.5 mm (± 0.6 mm).
Posterior to Biceps Tendon
Of the 55 patients who underwent conventional MRI, the insertion of the
long head of the biceps tendon was seen in 46 patients
(Table 5). The tendon was not
identified in three of the control patients and in six patients with a SLAP
tear. The SLAP tear type in these six patients was a type 2 in five patients
and a type 3 in one patient. High signal extending to the surface of the
superior labrum on an oblique coronal section posterior to the back of the
long head of the biceps tendon had a sensitivity for a SLAP tear of 54%
(15/28) and a specificity of 74% (20/27) (p = 0.02).
For the 26 patients who underwent MR arthrography, the insertion of the
long head of the biceps tendon was seen in all 13 patients with a SLAP tear
and in 11 of the 13 control patients (Table
6). Although neither of these control patients had a true labral
or biceps tendon tear diagnosed at arthroscopy, one had a glenohumeral
dislocation at the time of MR arthrography and the other had some synovitis
and minimal fraying of the tendon lateral to the insertion. High signal
extending to the surface of the labrum posterior to the biceps tendon was seen
in 9 of 13 patients with a SLAP tear (sensitivity, 69%) but also in 6 of 13
control patients (specificity, 54%) (p = 0.7).
Discussion
Although several articles have found MRI and MR arthrography to be accurate
for diagnosing SLAP tears
[8-11,
15], other authors have
recently reported accuracies for SLAP tears to be as low as 63% on
conventional MR images [18]
and 74% on MR arthrography [5].
We believed it would be useful to determine whether two signs that have been
proposed in the literature, but to our knowledge have not been specifically
evaluated, help improve the accuracy of MRI for SLAP tears.
The first sign that we evaluated was the medial-lateral width of the
high-signal line extending into the superior labrum or between the labrum and
the adjacent glenoid rim. Various authors have reported that a SLAP tear is
present on MR images if there is "displacement" of the superior
labrum [12,
13], "accumulation of
contrast between the labrum and glenoid"
[13], or "wide
separation of the superior labrum from the glenoid rim"
[8].
In our study, a width of at least 2 mm on conventional MR images and 2.5 mm
on MR arthrography has a specificity of 85-89% for SLAP tears. The prevalence
of SLAP tears in our study was higher than in our clinical population and that
of most other MR centers. When we extrapolated our data to a hypothetical
patient population with a more typical 5% prevalence of SLAP tears, the
positive predictive value for these two thresholds calculated to about 12-14%.
In addition, the sensitivity of this sign is low, so the superior labrum
should also be carefully inspected for the other more common MRI signs of a
SLAP tear, such as irregular, branching, or laterally curved high signal or a
completely detached labrum.
Another weakness of this sign is that the width of the high signal is small
relative to the spatial resolution of the image. The transverse width of each
pixel is about 0.6 mm in the oblique coronal plane on which we made our
measurements. Interpolation software built into the MR scanner smoothes the
signal differences at the interface between pixels but contributes to some
subjectivity in determining where the exact margins are of the high signal.
The interobserver correlation coefficient and that between the consensus
interpretation and the one by the musculoskeletal radiologist alone were good,
however, with a range of 0.83-0.89.
The second sign that we studied was high signal posterior to the long head
of the biceps tendon, including high signal that was smooth and curved
medially. Smith et al. [16]
evaluated 26 cadavers in their study and found that "no recess was
identified in the posterior third of the superior labrum" and, agreeing
with a similar finding by Cooper et al.
[19], concluded that the
recess "does not extend posterior to the long head biceps tendon."
Tuite et al. [17] looked at
this "posterior third rule" and reported a specificity of 81-94%
for this sign on conventional MRI images. Waldt et al.
[15] also used extension of
contrast medium into the posterior third of the superior labrum as one of
their signs of a SLAP tear and reported an overall specificity of 99% on MR
arthrographic images. Bencardino et al.
[11] specifically included
contrast material posterior to the biceps tendon anchor as one of their MR
arthrogram signs of a SLAP tear and reported a 90% specificity. We found this
sign to have limited usefulness as a stand-alone sign, with a specificity of
74% on conventional MRI and 54% on MR arthrography.
We are not certain why so many of our control patients had high signal
posterior to what we identified as the back of the biceps tendon. One reason
may be the limitations in the technique that we used to determine the back of
the biceps tendon. Some have reported that traction on the arm at MR
arthrography more accurately shows the back of the insertion of the biceps
tendon, but this was not used in our study
[20].
Another possible explanation is that the normal superior recess may extend
that far back in some people. Kreitner et al.
[21] reported that the
superior recess "extended throughout the entire base of the superior
labrum" in one of the 12 cadavers in their study. Tuite et al.
[17] reported that 6-19% of
their patients without a SLAP tear had abnormal high signal in the posterior
third of the superior labrum. Jee et al.
[5] reported that no difference
was seen between false-positives and true-positives for SLAP tears in the
number of oblique coronal sections behind the biceps anchor that had contrast
material beneath the labrum. Both groups had contrast material beneath the
labrum an average of 4.5 ± 0.4 sections (3-mm thick, 0.3-mm interslice
gap) behind the biceps anchor, indicating that a normal recess extended fairly
far posteriorly in some of their patients. Finally, DePalma et al.
[22] found that the recess is
more common in people as they age and therefore may represent a physiologic
partial separation of the labrum from the adjacent glenoid. If true, it is not
entirely clear that the recess would have to end always at the posterior
aspect of the biceps tendon, especially given that many of the collagen fibers
of the biceps tendon continue as the circumferential fibers in the
posterosuperior labrum
[23].
A limitation of any MR study of SLAP tears that uses arthroscopy as the
gold standard is that, because of the superior recess, diagnosing a SLAP tear
at surgery can be somewhat subjective. Snyder and Wuh
[24], who were among the first
to describe SLAP tears in 1990, recognized this in an article published 1 year
later in which they stated, "The initial tendency is clearly one of
over-diagnosis of these lesions." Like many surgeons, our shoulder
orthopedists use several arthroscopic and clinical criteria to diagnose a SLAP
tear. Our shoulder surgeons state that some debate exists among orthopedists
as to whether a partial lack of attachment of the superior labrum to the
glenoid rim that extends posterior to the biceps tendon, if otherwise normal,
is by itself an absolute criterion for a SLAP tear.
Our study had several additional limitations. The surgeons knew the results
of the MRI examinations and so may have been biased in their arthroscopic
findings. Our study included both conventional MRI and MR arthrography, and
there were three control group patients who had both imaging tests. The width
of the high signal on MR arthrography was usually similar on the
fat-suppressed T1-weighted images and the fat-suppressed T2-weighted images,
but in a few patients in whom we used the T2-weighted images it may have not
been wider because of a slight difference in slice placement but because it
may have included adjacent intralabral high signal that was not actually part
of the tear. In addition, we did not obtain an oblique coronal proton
density-weighted sequences, which might have shown small tears not seen on
T2-weighted images. Finally, 10 patients had a delay longer than 6 months
between MRI and surgery, which may have lowered the sensitivity of these two
signs.
In summary, we found that a width of the high-signal line between the
superior labrum and glenoid rim of at least 2.0 mm on conventional MRI and 2.5
mm on MR arthrography is a moderately specific sign of a SLAP tear. These
threshold widths are less useful in a clinical setting with a more typical 5%
prevalence of SLAP tears, where the calculated PPV falls to about 12-14%.
Extension of the high signal posterior to the biceps tendon has poor
specificity for a SLAP tear, and smooth medially curved signal in this region
is a weak stand-alone sign of a SLAP tear, especially on MR arthrography.
Although we used the arthroscopic findings as the gold standard, some debate
exists among surgeons about what distinguishes some large stable normal
variant superior recesses from a SLAP tear requiring treatment, and this may
have affected the accuracy that we report for these MR signs.
References
- Handelberg F, Willems S, Shahabpour M, Huskin J, Kuta J. SLAP
lesions: a retrospective multicenter study.
Arthroscopy 1998;14
: 856-862[Medline]
- Snyder S, Banas M, Karzel R. An analysis of 140 injuries to the
superior glenoid labrum. J Shoulder Elbow Surg1995; 4:243
-248[CrossRef][Medline]
- Mohana-Borges A, Chung C, Resnick D. Superior labral
anteroposterior tear: classification and diagnosis on MRI and MR arthrography.
AJR 2003; 181:1449
-1462[Free Full Text]
- Maffet MW, Gartsman GM, Moseley B. Superior labrum-biceps tendon
complex lesions of the shoulder. Am J Sports Med1995; 23:93
-98[Abstract/Free Full Text]
- Jee WH, McCauley TR, Katz LD, Matheny JM, Ruwe PA, Daigneault JP.
Superior labral anterior posterior (SLAP) lesions of the glenoid labrum:
reliability and accuracy of MR arthrography for diagnosis.
Radiology 2001;218
: 127-132[Abstract/Free Full Text]
- Kim TK, Queale WS, Cosgarea AJ, McFarland EG. Clinical features of
the different types of SLAP lesions. J Bone Joint Surg
Am 2003; 85:66
-71[Abstract/Free Full Text]
- Snyder S, Karzel R, Del Pizzo W, Ferkel R, Friedman M. SLAP lesions
of the shoulder. Arthroscopy 1990;6
: 274-279[Medline]
- Connell D, Potter H, Wickiewicz T, Altcheck D, Warren R.
Noncontrast magnetic resonance imaging of superior labral lesions: 102 cases
confirmed at arthroscopic surgery. Am J Sports Med1999; 27:208
-213[Abstract/Free Full Text]
- Legan JM, Burkhard TK, Goff WB, et al. Tears of the glenoid labrum:
MR imaging of 88 arthroscopically confirmed cases.
Radiology 1991;179
: 241-246[Abstract/Free Full Text]
- Gusmer PB, Potter HG, Schatz JA, et al. Labral injuries: accuracy
of detection with unenhanced MR imaging of the shoulder.
Radiology 1996;200
: 519-524[Abstract/Free Full Text]
- Bencardino JT, Beltran J, Rosenberg ZS, et al. Superior labrum
anterior-posterior lesions: diagnosis with MR arthrography of the shoulder.
Radiology 2000;214
: 267-271[Abstract/Free Full Text]
- Palmer WE, Brown JH, Rosenthal DI. Labral-ligamentous complex of
the shoulder: evaluation with MR arthrography.
Radiology 1994;190
: 645-651[Abstract/Free Full Text]
- Beltran J, Rosenberg ZS, Chandnani VP, Cuomo F, Beltran S, Rokito
A. Glenohumeral instability: evaluation with MR arthrography.
RadioGraphics 1997;17
: 657-673[Abstract]
- Smith AM, McCauley TR, Jokl P. SLAP lesions of the glenoid labrum
diagnosed with MR imaging. Skeletal Radiol1993; 22:507
-510[Medline]
- Waldt S, Burkart A, Lange P, Imhoff AB, Rummeny EJ, Woertler K.
Diagnostic performance of MR arthrography in the assessment of superior labral
anteroposterior lesions of the shoulder. AJR2004; 182:1271
-1278[Abstract/Free Full Text]
- Smith DK, Chopp TM, Aufdemorte TB, Witkowski EG, Jones RC.
Sublabral recess of the superior glenoid labrum: study of cadavers with
conventional nonenhanced MR imaging, MR arthrography, anatomic dissection, and
limited histologic examination. Radiology1996; 201:251
-256[Abstract/Free Full Text]
- Tuite MJ, Cirillo RL, De Smet AA, Orwin JF. Superior labrum
anterior-posterior (SLAP) tears: evaluation of three MR signs on T2-weighted
images. Radiology 2000;215
: 841-845[Abstract/Free Full Text]
- Yoneda M, Izawa K, Hirooka A, Hayashida K, Wakitani S. Indicators
of superior glenoid labral detachment on magnetic resonance imaging and
computed tomography arthrography. J Shoulder Elbow
Surg 1998; 7:2
-12[CrossRef][Medline]
- Cooper D, Arnoczky S, O'Brien S, et al. Anatomy, histology, and
vascularity of the glenoid labrum. J Bone Joint Surg
Am 1992; 74:46
-52[Abstract/Free Full Text]
- Chan K, Muldoon K, Yeh L, et al. Superior labral anteroposterior
lesions: MR arthrography with arm traction. AJR1999; 173:1117
-1122[Abstract/Free Full Text]
- Kreitner K, Botchen K, Rude J, Bittinger F, Krummenauer F, Thelen
M. Superior labrum and labral-bicipital complex: MR imaging with
pathologic-anatomic and histologic correlation. AJR1998; 170:599
-605[Abstract/Free Full Text]
- DePalma AF, Gallery G, Bennett G. Variational anatomy and
degenerative lesions of the shoulder joint. In: Edwards J, ed.
Instructional course lectures: the American Academy of Orthopedic
Surgeons. St. Louis, MO: Mosby, 1949:225
-281
- Vangsness CT, Jorgenson SS, Watson T, Johnson DL. The origin of the
long head of the biceps from the scapula and glenoid labrum. J Bone
Joint Surg Br 1994; 76:951
-954
- Snyder S, Wuh H. Arthroscopic evaluation and treatment of the
rotator cuff and superior labrum anterior posterior lesion.
Operative Techniques in Orthopedics 1991;1
: 207-220
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
