DOI:10.2214/AJR.07.2242
AJR 2007; 189:943-946
© American Roentgen Ray Society
Rim-Rent Tear of the Rotator Cuff: A Common and Easily Overlooked Partial Tear
Emily N. Vinson1,
Clyde A. Helms1 and
Laurence D. Higgins2
1 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC
27710.
2 Department of Orthopedic Surgery, Brigham and Women's Hospital, Boston, MA
02115.
Received March 9, 2007;
accepted after revision May 19, 2007.
Address correspondence to E. N. Vinson
(vinso003{at}mc.duke.edu).
Presented at the 2006 annual meeting of the American Roentgen Ray Society,
Vancouver, BC, Canada.
Abstract
OBJECTIVE. The purpose of this study was to determine the incidence
of rim-rent rotator cuff tears in a population of patients referred for
shoulder MRI and to determine the sensitivity of MRI for the detection of
these tears.
CONCLUSION. Rim-rent tears are a common type of partial-thickness
rotator cuff tear, much more commonly present than has been previously
reported. In particular, infraspinatus rim-rent tears are more common than
previously believed. Rim-rent tears of the infraspinatus tendon and those
involving the anterior-most fibers of the supraspinatus tendon are commonly
overlooked on MRI, possibly because of failure to appreciate the high
incidence of these types of tears and failure to inspect the anterior-most
fibers of the rotator cuff.
Keywords: rim-rent tear • rotator cuff tear • shoulder MRI
Introduction
The term "rim-rent" was first used by E. A. Codman
[1] in 1934. He described four
types of incomplete ruptures of the supraspinatus tendon, and one the four he
described was a tear in which "a few of the lower fibers on the joint
side, together with the synovial reflection, may be torn out...," and he
termed these rotator cuff tears rim-rent tears
[1]. To our knowledge, the only
other mention of rim-rent tears in the literature was in a 1998 article by
Tuite et al. [2] that reported
110 consecutive patients who had preoperative shoulder MRI examinations and
either a partial-thickness or small full-thickness tear diagnosed at
arthroscopy. Of these patients, nine had articular surface partial tears that
were less than 2 cm in diameter and adjacent to the rotator cuff insertion
into the greater tuberosity, which Tuite et al. called "rim-rent
tears." Of the 110 patients with partial-thickness or small
full-thickness cuff tears, only one patient's tear was isolated to the
infraspinatus tendon. Although the article did not address whether this one
tear was a rim-rent tear, we had previously believed that infraspinatus
rimrent tears were uncommon.
Through surgical follow-up, we recently discovered that we had overlooked
several rim-rent tears on initial shoulder MRI interpretations. It seemed the
location of the tears and the fact that many of the patients were positioned
in marked internal rotation during the MRI examinations contributed to this.
The purpose of this study was to determine the incidence of rim-rent tears in
patients referred for shoulder MRI and to determine the sensitivity of MRI for
the detection of these tears.
Materials and Methods
The study was performed after approval from our institutional review board.
There were two parts to the study. The first part involved a cohort of
patients with surgically proven rim-rent tears and an available preoperative
shoulder MRI performed at our institution. This population was selected from a
single orthopedic surgeon's database and consisted of 29 patients. There were
22 males and 7 females with a mean age of 41.6 years (age range, 17–77
years). The time interval between MRI and arthroscopy ranged from 7 to 179
days (mean, 51.9 days).
The preoperative MRI examinations were retrospectively reviewed by two
musculoskeletal radiologists. We assessed our ability to see the rotator cuff
tear on MRI, noted the location of the tear with regard to the
anterior-to-posterior position in the rotator cuff, and subjectively assessed
whether the patient was positioned in internal rotation during the MRI
examination on the basis of the location of the bicipital groove on the axial
images. The original MRI interpretation was reviewed to determine how often
the correct diagnosis was made prospectively.
For the second part of the study, 225 consecutive shoulder MRI examinations
performed at our institution between June 2002 and January 2005 were reviewed.
Twenty-five examinations were excluded from the study: Three were excluded due
to extensive artifact from postoperative changes, 10 were excluded because all
of the images were not available for review, nine were excluded because a
nonstandard protocol was performed, two were excluded because extensive
infection obscured visualization of the rotator cuff, and one was excluded
because visualization of the rotator cuff was obscured due to extensive
posttraumatic hematoma. The remaining 200 examinations composed the study
population. There were 109 males and 91 females with a mean age of 46.7 years
(age range, 15–85 years). The examinations were retrospectively reviewed
by two musculoskeletal radiologists for the presence of rotator cuff tears,
and the tears were classified as to location and as full or partial thickness.
Of the partial-thickness tears, those of the rim-rent variety were noted. We
assessed subjectively whether the patients were positioned in internal
rotation during the examination. The original MRI interpretations and, when
available, corresponding operative reports were reviewed.
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —25-year-old man with rim-rent tear of supraspinatus tendon.
Fat-suppressed fast spin-echo T2-weighted coronal oblique (TR/TE, 3,967/65)
(A) and sagittal oblique (3,850/75) (B) MR arthrography images
show rim-rent tear (arrows) at insertion of supraspinatus tendon onto
greater tuberosity.
|
|
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —25-year-old man with rim-rent tear of supraspinatus tendon.
Fat-suppressed fast spin-echo T2-weighted coronal oblique (TR/TE, 3,967/65)
(A) and sagittal oblique (3,850/75) (B) MR arthrography images
show rim-rent tear (arrows) at insertion of supraspinatus tendon onto
greater tuberosity.
|
|
View larger version (165K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —38-year-old man with rim-rent tear of infraspinatus tendon.
Fat-suppressed fast spin-echo T2-weighted coronal oblique (TR/TE, 4,000/65)
(A) and sagittal oblique (3,000/74) (B) MR arthrography images
show rim-rent tear (arrows) at insertion of infraspinatus tendon onto
greater tuberosity. This tear was not seen prospectively but was diagnosed at
time of surgery.
|
|
View larger version (154K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —38-year-old man with rim-rent tear of infraspinatus tendon.
Fat-suppressed fast spin-echo T2-weighted coronal oblique (TR/TE, 4,000/65)
(A) and sagittal oblique (3,000/74) (B) MR arthrography images
show rim-rent tear (arrows) at insertion of infraspinatus tendon onto
greater tuberosity. This tear was not seen prospectively but was diagnosed at
time of surgery.
|
|
All of the patients included in the study were scanned on a 1.5-T MRI
scanner (Signa, GE Healthcare). In the cohort of 29 patients with surgically
proven rim-rent tears, 27 had preoperative shoulder MR arthrograms and two had
preoperative conventional shoulder MR examinations. In the cohort of 200
patients who underwent shoulder MRI at our institution, 86 had shoulder MR
arthrograms and 114 had conventional shoulder MR examinations.
Our conventional shoulder MRI protocol consists of axial and oblique
coronal fat-suppressed fast spin-echo T2-weighted images (TR/TE range,
4,000/65–75) and fat-suppressed fast spin-echo proton density images (TR
range/TE range, 3,500–4,000/20–35), oblique sagittal T1-weighted
images (500–620/14–15), and oblique sagittal fatsuppressed fast
spin-echo T2-weighted images (4,000/65–75). The MR arthrogram images
were obtained after the intraarticular instillation under fluoroscopic
guidance via an anterior approach of 12 mL of dilute gadolinium (gadoteridol
[Prohance, Bracco Diagnostics]) diluted 1:200 with normal saline). These
patients were subsequently imaged using our shoulder MR arthrography protocol:
axial, oblique coronal, and oblique sagittal fat-suppressed fast spin-echo
T2-weighted images (3,000–4,000/65–75) and fat-suppressed
T1-weighted images (500–650/14–15), and oblique sagittal
T1-weighted images without fat suppression (500–617/14–15). The
number of excitations was 2 for fast spin-echo sequences and 1 for T1-weighted
sequences, and the echo-train length was 8 for all fast spin-echo sequences. A
slice thickness of 4 mm with a 0.4-mm interslice gap was used with a field of
view of 14 cm and matrix of 256 x 192 for all studies.
Full-thickness tears were diagnosed when there was discontinuity of the
tendon with fluid signal intensity traversing the gap on the fast spin-echo
T2-weighted images or dilute gadolinium traversing the gap on the
postarthrography T1-weighted images. Partial-thickness tears were diagnosed
when there was fluid signal intensity involving a portion of the tendon on the
fast spin-echo T2-weighted images, without a visible gap. Rim-rent tears were
defined as those partial-thickness tears occurring on the articular side of
the rotator cuff at the tendon insertion on the greater tuberosity. Critical
zone partial-thickness tears were defined as those partial-thickness tears
occurring approximately 1–2 cm proximal to the cuff tendon insertion on
the greater tuberosity.
Results
Of the 29 patients with surgically proven rim-rent tears, the tear was
visible on MRI in retrospect in 28 cases (96.6%). In 17 (60.7%) of these 28
cases, the tear involved the anterior-most fibers of the supraspinatus tendon
(Fig. 1A,
1B), and the tear was
prospectively diagnosed on the initial MRI interpretation in 12 (70.6%) of
these cases. In nine (32.1%) of these 28 cases, the tear involved the
infraspinatus tendon (Fig. 2A,
2B), and the tear was
prospectively diagnosed on the initial MRI interpretation in four (44.4%) of
these cases. Overall, the presence of a rim-rent tear was prospectively
diagnosed in 19 of the 29 (65.5%) surgically proven cases. Of the 10 cases in
which the tear was missed prospectively, five tears involved the anterior-most
fibers of the supraspinatus tendon and four tears involved the infraspinatus
tendon; in one of the cases, the tear was not visible in retrospect. In 20 of
the 29 cases, the patients were subjectively assessed to have been positioned
in excessive internal rotation during imaging. Of the nine cases in which the
tear was missed prospectively but visible in retrospect, seven were assessed
to have been positioned in excessive internal rotation.
Of the cohort of 200 consecutive shoulder MRI examinations, a total of 117
tears were identified in retrospect in 101 of the examinations. Seventy
(59.8%) of the tears were partial-thickness tears, and of these, 49 (70.0%)
were of the rim-rent variety. The incidence of rimrent tears in this
population was 49 in 200, or 24.5%. Twenty-one of the 70 partial-thickness
tears were not rim-rent tears: there were nine (12.9%) tears in the critical
zone, 10 (14.3%) interstitial tears, and two (2.9%) bursal-sided tears. Of the
49 rim-rent tears, 24 (49.0%) involved the anterior-most fibers of the
supraspinatus tendon, one of which extended to involve the infraspinatus
tendon. Twenty-three (46.9%) involved the infraspinatus tendon alone. Two
tears were isolated to the more-posterior aspect of the supraspinatus tendon.
Of the 24 tears involving the anterior-most fibers of the supraspinatus
tendon, 16 (66.7%) were prospectively diagnosed; of the eight cases not
prospectively diagnosed, seven were imaged in excessive internal rotation
(Fig. 3A,
3B). Of the 24 rimrent tears
involving the infraspinatus tendon, nine (37.5%) were prospectively diagnosed.
Of all 49 rim-rent tears, the correct diagnosis was made prospectively in 27
(55.1%) cases.
View larger version (167K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —56-year-old man with rim-rent tear of anterior fibers of
supraspinatus tendon; patient positioned in internal rotation. Fat-suppressed
fast spin-echo T2-weighted coronal oblique (TR/TE, 4,000/67) (A) and
sagittal oblique (4,000/67) (B) MR images show rim-rent tear at
insertion of supraspinatus tendon onto greater tuberosity involving insertion
of far anterior fibers of supraspinatus tendon (arrows) adjacent to
biceps tendon (arrowhead, A). This tear was not diagnosed at
time of initial MRI interpretation.
|
|
View larger version (160K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —56-year-old man with rim-rent tear of anterior fibers of
supraspinatus tendon; patient positioned in internal rotation. Fat-suppressed
fast spin-echo T2-weighted coronal oblique (TR/TE, 4,000/67) (A) and
sagittal oblique (4,000/67) (B) MR images show rim-rent tear at
insertion of supraspinatus tendon onto greater tuberosity involving insertion
of far anterior fibers of supraspinatus tendon (arrows) adjacent to
biceps tendon (arrowhead, A). This tear was not diagnosed at
time of initial MRI interpretation.
|
|
Of the cohort of 200 patients, 62 subsequently underwent shoulder surgery
and had available operative reports. Fifty-two of these patients had rotator
cuff tears diagnosed surgically, and 27 of these were full-thickness tears, of
which 26 (96.3%) were diagnosed by MRI prospectively. Twenty-five of the tears
were either partial-thickness tears or subscapularis tears, of which 15
(60.0%) were diagnosed by MRI prospectively. Of the 10 cases not diagnosed by
MRI prospectively, one was interpreted as a full-thickness tear prospectively,
and nine were not reported. Six of these nine were visible on MRI in
retrospect, and all six of these were rim-rent tears. Of the 62 patients who
underwent shoulder surgery after MRI, a total of 16 rim-rent tears were found.
Of these, only seven (43.8%) were diagnosed prospectively by MRI. In all nine
of the rim-rent tears not reported prospectively, the patients were imaged in
internal rotation. Seven of these tears were visible in retrospect; four
involved the anterior fibers of the supraspinatus tendon, and three involved
the infraspinatus tendon.
Discussion
Partial-thickness tears of the rotator cuff are a significant and common
cause of shoulder pain. However, both the cause and management of these
lesions remain elusive and without consensus. Codman
[1] believed that some
combination of tendinitis and trauma occurring in a degenerated tendon led to
incomplete ruptures of the supraspinatus tendon. Neer
[3] emphasized the importance
of impingement of the rotator cuff beneath the coracoacromial arch in the
development of incomplete rotator cuff tears and espoused anterior
acromioplasty, in addition to rotator cuff repair, to relieve the subacromial
impingement in patients with complete and incomplete supraspinatus tears and
in patients with irreparable tears. Chang et al.
[4], using 3D shape analysis of
the morphologic features of the acromial undersurface in a cohort of patients,
concluded that osseous impingement by any portion of the acromion is not a
primary cause of impingement syndrome or rotator cuff tears. Most now embrace
a multifactorial pathogenesis of rotator cuff tears, with intrinsic
degeneration, microtrauma, trauma, hypovascularity, and subacromial
impingement all having a potential role
[5,
6]. Although Codman believed
that partial-thickness rotator cuff tears could heal spontaneously without
surgery, this assumption has not been substantiated histologically, and many
believe it to be the rare exception rather than the rule
[5,
7].
It has long been thought that the majority of rotator cuff tears occur in
the so-called critical zone—that portion of the tendon approximately 1
cm proximal to the cuff insertion onto the greater tuberosity—due to the
relative hypovascularity of the tendon in that region
[8,
9]
(Fig. 4). However, feedback
from our surgical colleagues contradicts this long-held belief. Their
experience and the results derived in our study strongly suggest that rim-rent
tears involving the insertional fibers of the rotator cuff are the most common
type of partial-thickness rotator cuff tear. Since making this observation, we
have also noted that many full-thickness rotator cuff tears appear on imaging
as if they originated as rim-rent tears and progressed to involve the full
thickness of the tendon (Fig.
5). The results of our study also suggest that bursal-sided
partial-thickness tears are less common than previously believed, constituting
only 2.9% of all partial-thickness tears in this series.
View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —58-year-old woman with critical zone partial-thickness tear.
Coronal oblique fat-suppressed fast spin-echo T2-weighted (TR/TE, 4,000/67) MR
image shows partial-thickness, articular-sided tear of supraspinatus tendon
1–2 cm proximal to its insertion, in so-called critical zone
(arrow).
|
|
View larger version (159K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —46-year-old man with full-thickness rotator cuff tear.
Coronal oblique fat-suppressed fast spin-echo T2-weighted (TR/TE, 4,000/66) MR
arthrography image shows full-thickness tear of rotator cuff from insertion on
greater tuberosity (arrow).
|
|
MRI has proven to be reliable in the diagnosis of full-thickness rotator
cuff tears, with sensitivities reported as high as 100%
[10]. Reported sensitivities
for the detection of partial-thickness rotator cuff tears with MRI have been
more variable, ranging from 35% to 92%
[10–13].
During our routine quality assurance conferences, it was noted that we had
missed several partial-thickness and full-thickness rotator cuff tears
localized to the anterior-most fibers of the supraspinatus tendon, sometimes
referred to as the leading edge of the rotator cuff.
We noted that many of these tears were of the rim-rent variety and that the
tears were difficult to visualize in many instances because the patients were
positioned in marked internal rotation during imaging. This internal rotation
made the anterior cuff difficult to visualize on MRI, particularly on the
sagittal oblique images. We learned to routinely inspect the anterior-most
fibers of the cuff on the coronal oblique images by finding the image with the
long head of the biceps tendon passing over the humeral head and inspecting
the cuff from that point posteriorly (Fig.
6A,
6B). Through experience, we
have also learned that partial-thickness tears in general—and rim-rent
tears in particular—are not rare in the infraspinatus tendon, and we
carefully inspect the infraspinatus insertion on both the coronal oblique and
sagittal oblique images for this type of tear.
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —36-year-old man with normal supraspinatus anterior fibers.
Consecutive far anterior coronal oblique fat-suppressed fast spin-echo
T2-weighted (TR/TE, 4,000/67) MR arthrography images show normal appearance of
anterior fibers of rotator cuff insertion (thick arrows) adjacent to
long head of biceps tendon (thin arrows).
|
|
View larger version (143K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —36-year-old man with normal supraspinatus anterior fibers.
Consecutive far anterior coronal oblique fat-suppressed fast spin-echo
T2-weighted (TR/TE, 4,000/67) MR arthrography images show normal appearance of
anterior fibers of rotator cuff insertion (thick arrows) adjacent to
long head of biceps tendon (thin arrows).
|
|
Limitations of this study include the fact that this was a retrospective,
nonblinded study. Also, although the cohort of 29 patients was chosen because
they had surgically proven rim-rent tears, no surgical data were available for
the majority of the cohort of 200 patients. However, the sensitivity of MRI
for the prospective detection of rim-rent tears was similar across the cohorts
of patients studied. In the cohort of 29 surgically proven rim-rent tears, 19
(65.5%) were diagnosed correctly by MRI prospectively. In the cohort of 200
consecutive shoulder MRI examinations, there were 49 rim-rent tears, 27
(55.1%) of which were diagnosed correctly by MRI prospectively. Within that
group, there were 16 patients who underwent shoulder surgery and were
diagnosed with rim-rent tears, of which seven (43.8%) were diagnosed
prospectively. This low sensitivity for rim-rent tears may be rectified by
learning to inspect the leading edge of the supraspinatus, just lateral to the
bicipital groove on the oblique coronal images, and to look closely at the
infraspinatus insertion because tears were initially overlooked in both of
these locations.
In conclusion, rim-rent tears are a common type of rotator cuff tear,
occurring in 24.5% of 200 consecutive shoulder MRI examinations in this series
and accounting for 70% of all partial-thickness rotator cuff tears. Critical
zone and bursal-sided partial-thickness tears were much less common,
accounting for 12.9% and 2.9% of partial-thickness tears, respectively.
Contrary to previous beliefs, the infraspinatus is a relatively common
location of rim-rent tears, with almost half of all rimrent tears involving
the infraspinatus tendon in this series. Rim-rent tears, especially
infraspinatus rim-rent tears, are often over-looked on MRI, possibly in part
because we did not previously recognize this to be a common type of tear.We
believe that excessive internal rotation and a failure to inspect the
anterior-most fibers of the rotator cuff contributed to prospectively missing
many supraspinatus rim-rent tears because they commonly involve the far
anterior fibers of the rotator cuff. We also believe that, rather than
healing, many rim-rent tears will progress to full-thickness tears if not
diagnosed and treated appropriately. Being more aware of these tears and more
comfortable looking for them on the far anterior coronal oblique images and on
the sagittal oblique images should improve our sensitivity in detecting these
common tears on MRI.
References
- Codman EA. The shoulder. Boston, MA: Thomas
Todd, 1934: 123–177
- Tuite MJ, Turnbull JR, Orwin JF. Anterior versus posterior, and
rim-rent rotator cuff tears: prevalence and MR sensitivity.
Skeletal Radiol 1998;27
: 237–243[CrossRef][Medline]
- Neer CS 2nd. Anterior acromioplasty for the chronic impingement
syndrome in the shoulder: a preliminary report. J Bone Joint Surg
(Am) 1972; 54:41
–50[Abstract/Free Full Text]
- Chang EY, Moses DA, Babb JS, Schweitzer ME. Shoulder impingement:
objective 3D shape analysis of acromial morphologic features.
Radiology 2006;239
: 497–505[Abstract/Free Full Text]
- Fukuda H. Partial-thickness rotator cuff tears: a modern view on
Codman's classic. J Shoulder Elbow Surg2000; 9:163
–168[CrossRef][Medline]
- Ko J-Y, Huang CC, Chen W-J, Chen C-E, Chen S-H, Wang C-J.
Pathogenesis of partial tear of the rotator cuff: a clinical and pathologic
study. J Shoulder Elbow Surg 2006;15
: 271–278[CrossRef][Medline]
- Fukuda H. The management of partial-thickness tears of the rotator
cuff. J Bone Joint Surg (Br) 2003;85
: 3–11[CrossRef][Medline]
- Neviaser RJ, Neviaser TJ. Observations on impingement.
Clin Orthop Rel Res 1990;254
: 60–63[Medline]
- Bergman AG. Rotator cuff impingement. Magn Reson Imaging
Clin N Am 1997; 5:705
–719[Medline]
- Singson RD, Hoang T, Dan S, Friedman M. MR evaluation of rotator
cuff pathology using T2-weighted fast spin-echo technique with and without fat
suppression. AJR 1996;166
:1061
–1065[Abstract/Free Full Text]
- Balich SM, Sheley RC, Brown TR, Sauser DD, Quinn SF. MR imaging of
the rotator cuff tendon: interobserver agreement and analysis of interpretive
errors. Radiology 1997;204
: 191–194[Abstract/Free Full Text]
- Magee T, Williams D. 3.0-T MRI of the supraspinatus tendon.
AJR 2006; 187:881
–886[Abstract/Free Full Text]
- Rafii M, Firooznia H, Sherman O, et al. Rotator cuff lesions:
signal patterns at MR imaging. Radiology1990; 177:817
–823[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
Top
Abstract
Introduction
