DOI: 10.1148/radiol.2441050029
(Radiology 2007;244:239-248.)
© RSNA, 2007
Cystic Changes at Supraspinatus and Infraspinatus Tendon Insertion Sites: Association with Age and Rotator Cuff Disorders in 238 Patients1
L. Benjamin Fritz, DMD, MD,
Hugue A. Ouellette, MD,
Tara A. O'Hanley, BA,
Ara Kassarjian, MD, and
William E. Palmer, MD
1 From the Department of Musculoskeletal Radiology, Massachusetts General Hospital, 55 Fruit St, YAW 6030, Boston, MA 02114. From the 2004 RSNA Annual Meeting. Received January 12, 2005; revision requested March 31; revision received October 1, 2006; accepted November 1; final version accepted December 22.
Address correspondence to W.E.P. (e-mail: palmer.william{at}mgh.harvard.edu).
 |
ABSTRACT
|
|---|
Purpose: To retrospectively determine the frequency of posterior and anterior cystic abnormalities at rotator cuff insertion site on the greater tuberosity and to determine their relationship to patient age and rotator cuff disorders.
Materials and Methods: Institutional review board approval was given; informed consent was waived. The study was HIPAA compliant. In 238 patients with rotator cuff diagnoses at surgery, preoperative magnetic resonance (MR) imaging studies were reviewed to localize osseous cystic changes as anterior (supraspinatus insertion site) or posterior (infraspinatus insertion site) on the greater tuberosity. If rotator cuff tear was present, tendon retraction and location of partial tear (articular or bursal surface) were recorded. Two radiologists reached conclusions by consensus. Locations of cysts were correlated to surgical cuff diagnoses: no tear, tendinopathy, partial-thickness tear, and complete tear. Prospective interpretations from original MR reports were compared with surgical results. Statistical analyses included one-way analysis of variance, 
2, Fisher exact, and Student t tests, as well as logistic regression and receiver operating characteristic curve comparison.
Results: There were 238 consecutive patients (150 men, 88 women; mean age, 43 years). Cysts were located at or near footprint of cuff tendon and demonstrated fluid or soft-tissue signal intensities. Posterior cysts occurred in 56.7% of shoulders and showed no statistical correlation to age or cuff diagnosis. Anterior cysts occurred in 22.7% of shoulders and were strongly associated with cuff disorders (P < .001). Controlling for cuff disorders, there was no relationship between anterior cysts and age (P > .50). Anterior cysts were more common in partial-thickness articular (48%) than in bursal (13%) tears (P < .001).
Conclusion: Posterior cysts were more common than anterior cysts and showed nearly random distribution among patients, regardless of age and cuff diagnosis. Anterior cysts were closely associated with cuff disorders.
© RSNA, 2007
|
INTRODUCTION
|
|---|
On the basis of necropsy findings, Codman (1) described in 1934 an association between rotator cuff tears and osseous abnormalities involving the greater tuberosity. These osseous abnormalities, including cortical irregularities and cystic changes, were also visible on radiographs of the shoulder. When localized to the rotator cuff attachment site, they became established as important radiographic signs in the diagnosis of cuff disorders, which occurred in two-thirds of patients with tears but were present in only 13%–27% of normal shoulders (2–5).
Magnetic resonance (MR) imaging and ultrasonography (US) are both established as effective imaging tools in the evaluation of the rotator cuff (6–10). Because of soft-tissue discrimination and tomographic section acquisition in multiple planes, these noninvasive imaging modalities are also more accurate than radiography in demonstrating and enabling the localization of osseous abnormalities at the cuff attachment site. Whereas some investigators (10–14) have reported strong associations between rotator cuff tears and osseous cysts or cortical irregularities, other researchers (15–18) have concluded that bone abnormalities of the greater tuberosity simply represent age-related changes that have no relationship to cuff disorders. Several of these studies had enrollment limited to asymptomatic volunteers or patients with known rotator cuff tears. Most of the studies had a small number of subjects or lacked surgical confirmation of cuff diagnoses made by using MR images. The studies rarely distinguished osseous abnormalities on the basis of location on the greater tuberosity. Thus, the purpose of our study was to retrospectively determine the frequency of posterior and anterior cystic abnormalities at the rotator cuff insertion site on the greater tuberosity and to determine their relationship to patient age and rotator cuff disorders.
|
MATERIALS AND METHODS
|
|---|
Patient Selection
Records of patients who underwent arthroscopic or open surgical evaluations between April 2000 and July 2004 performed by one of three orthopedic surgeons who specialize in the shoulder (21, 17, and 14 years of experience in shoulder surgery) were reviewed in chronologic order by two authors (L.B.F., W.E.P.) who determined eligibility. Shoulders were included in our study if preoperative MR images were available in the departmental picture archiving and communication system. Our institutional review board granted study approval and waived informed consent. This study was Health Insurance Portability and Accountability Act compliant.
Shoulders were excluded from the study for the following reasons: previous surgical interventions involved the rotator cuff or greater tuberosity (21 shoulders); surgery was performed more than a year after preoperative MR imaging (nine shoulders); trauma-related fracture (ie, Hill-Sachs defect) of the greater tuberosity (14 shoulders); or inflammatory, crystal-induced, or degenerative arthropathy affected the joint (11 shoulders). The remaining 238 patients were included in our study.
MR Imaging Protocol
All MR examinations were performed with 1.5-T units (GE Medical Systems, Milwaukee, Wis; Siemens Medical Solutions, Erlangen, Germany) by using routine clinical protocols and dedicated shoulder coils. Pulse sequences in 135 patients included (a) oblique coronal (repetition time msec/echo time msec, 2700/17) and fat-suppressed (2900/54) intermediate weighted; (b) oblique sagittal T1 weighted (600/14) and fat-suppressed intermediate weighted (2900/54); and (c) transverse gradient echo (600/15, 20° flip angle). In 103 patients, shoulder arthrography was combined with MR imaging (MR arthrography). Gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) was injected in a concentration of 2 mmol/L. To obtain this concentration, 2 mL of gadopentetate dimeglumine was added to 250 mL of normal saline. Ten milliliters of this solution was mixed with 5 mL of iodinated contrast material (Isovue-M 200; Bracco Diagnostics, Princeton, NJ) and 5 mL of lidocaine 1%. By using the single-contrast technique, 10–15 mL of this contrast solution was injected into the glenohumeral joint with fluoroscopic guidance, followed by MR imaging within 30 minutes. MR pulse sequences included (a) oblique coronal T1 weighted (660/18), fat-suppressed T1 weighted (800/18), and fat-suppressed intermediate weighted (2900/54); (b) oblique sagittal T1 weighted (600/14); and (c) transverse T1 weighted (550/18). An abduction and external rotation sequence was performed at 18 MR arthrographic examinations to obtain fat-suppressed T1-weighted (660/18) images. In all sequences, field of view was 16 cm, section thickness was 3–4 mm, and matrix was 192–384 x 256.
Data Collection
Demographic data, medical history, MR imaging reports, and surgical reports were obtained though the hospital information system and were recorded by a single author (L.B.F.). Information included patient age, sex, date of MR imaging examination, date of surgery, and pertinent surgical findings relating to the supraspinatus and infraspinatus tendons. Rotator cuff diagnoses were taken from the surgical reports and organized according to a modification of the Snyder classification: no tear, tendinopathy, partial-thickness articular-sided tear, partial-thickness bursal-sided tear, and full-thickness or complete tear (19). Tendinopathy was diagnosed surgically whenever the cuff tendon was abnormal (eg, thinned or thickened, frayed without focal defect, or softened) but not torn. Although there were a few cases that showed discrepancies between imaging and surgical findings (ie, normal cuff at surgery but apparent tear on MR images), the surgical diagnosis was always used as the reference standard. We also recorded the degree of tendon retraction when it was described by the surgeon. If the surgical report did not provide sufficient information about a surgically proved rotator cuff tear (eg, the degree of tendon retraction in a full-thickness tear or the involvement of the articular vs the bursal surface in a partial-thickness tear), this information was obtained retrospectively from the MR images.
Analysis of MR Images
Unaware of the surgical diagnoses and of the original interpretations at imaging, two radiologists (H.A.O., 9 years of experience in musculoskeletal MR imaging; W.E.P., 15 years of experience) inspected retrospectively the greater tuberosity and the anatomic neck along the rotator cuff insertion site for intraosseous cystic changes on MR images and reached conclusions by consensus. An independent investigator organized the MR images into groups of 15–20 and presented them in random order to the reviewers. Cysts could be solitary or multiple in number and were defined as discrete, subcortical, or cortically based lesions measuring at least 2 mm in dimension. Cysts could be rounded, elliptic, or multicentric in contour. They demonstrated sharp margins and low signal intensity on T1-weighted images, except when they appeared filled with contrast material on MR arthrograms. T2-weighted images showed either high signal intensity, which indicated fluid, or mixed signal intensity, which suggested the ingrowth of synovium or granulation tissue. Because cysts could have an identical appearance to vascular channels on single images, they were excluded if they connected to tubular structures on sequential images. It was necessary for cysts to be visible with at least two different pulse sequences and in two different imaging planes.
Cysts on the greater tuberosity were defined as anterior or posterior in location relative to the insertion sites of the supraspinatus and infraspinatus tendons. In anatomic studies, these insertion sites are composed of superior and middle facets, which together measure about 5 cm in length (1,2,20,21). The supraspinatus inserts on the superior facet of the greater tuberosity and the anterior aspect of the middle facet, whereas the infraspinatus inserts entirely on the middle facet. The precise transition point between the supraspinatus and infraspinatus tendons may be difficult to identify on MR images because of overlapping and blending of their fibers at the anterior aspect of the middle facet. Because the combined attachment of the supraspinatus and infraspinatus measures approximately 5 cm in length, we defined the supraspinatus insertion site as the anterior 2.5 cm of the greater tuberosity and the infraspinatus insertion site as the posterior 2.5 cm (20,21) (Fig 1).
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1: Three-dimensional reconstruction MR image of shoulder depicts insertion sites of supraspinatus and infraspinatus tendons. Greater tuberosity is composed of superior (S), middle (M), and inferior facets. Supraspinatus tendon attaches to superior facet and anterior aspect of middle facet, whereas infraspinatus tendon attaches entirely to middle facet. Therefore, supraspinatus and infraspinatus tendons overlap in middle portion of greater tuberosity (crosshatched region). We defined a cyst as anterior if located at anterior 2.5 cm of greater tuberosity and as posterior if located at posterior 2.5 cm of greater tuberosity.
|
|
To determine the transition point between the supraspinatus and infraspinatus tendons, we identified the anterior margin of the greater tuberosity on oblique sagittal or transverse MR images and measured 2.5 cm posteriorly along the cuff insertion by using a calibrated on-screen caliper provided with our institutional picture archiving and communication system. The cross-reference tool was helpful in correlating the transition point in all imaging planes. For each shoulder, we recorded the location and number of anterior and posterior cysts and measured the diameters of the largest anterior and posterior cysts by using the digital caliper. Cyst locations were not differentiated on the basis of rotator cuff insertion site versus the adjacent bare area of bone at the anatomic neck of the humeral head.
If surgical and MR imaging reports did not provide sufficient information about a surgically proved rotator cuff tear to complete our database, this information was obtained retrospectively from MR images by two reviewers (A.K., 5 years of experience; W.E.P.) who reached conclusions by consensus. There were two instances when this was necessary: (a) to determine whether a partial-thickness tear involved the bursal or articular surface and (b) to measure the degree of tendon retraction in a full-thickness tear. Retraction of 10 mm or less was recorded as grade 1, 11–20 mm was recorded as grade 2, 21–30 mm was recorded as grade 3, 31–40 mm was recorded as grade 4, and more than 40 mm was recorded as grade 5.
Statistical Analysis
Contingency table analysis and one-way analysis of variance tests were used to compare cystic changes with rotator cuff status and patient age. 2 Statistics were used to determine the significance of results. The Student t test or Fisher exact test (in the case of two-level variables) was used to correlate findings and test the significance of differences. Logistic regression and receiver operating characteristic curve comparison were employed to calculate sensitivity, specificity, positive and negative predictive values, and accuracy of MR imaging in the diagnosis of rotator cuff disorders. Statistical significance of results was set at 
= .05. A software package (JMP 5.0.1.2; SAS Institute, Cary, NC) was used for statistical calculations.
|
RESULTS
|
|---|
Demographic Data and Surgical Findings
Of 238 consecutive patients, 150 (63.0%) were men and 88 (37.0%) were women. The mean age of patients was 43 years (median age, 42 years), and the age range was 18–79 years. Mean age of women was 44 years and that of men was 42 years; the difference was not statistically significant. On average, 89–108 days (mean, 98 days; median, 80 days; range, 10–310 days) passed between MR imaging and surgical intervention.
MR arthroscopic or open surgical findings in 238 shoulders showed normal rotator cuffs in 121 (50.8%) patients; tendinopathy without tear in 13 (5.5%) patients; and torn cuffs in 104 (43.7%) patients, including partial-thickness tears in 48 (20.2%) and full-thickness tears in 56 (23.5%) patients. In only one case, a full-thickness infraspinatus tear was present without supraspinatus involvement. Otherwise, all full-thickness tears involved the supraspinatus tendon, and 31 (55%) of these extended posteriorly into the infraspinatus. In shoulders with partial-thickness tears, 33 (69%) were articular sided and 15 (31%) were bursal sided (Table 1).
View this table:
[in this window]
[in a new window]
|
Table 1. Relationship between Cystic Change and Rotator Cuff Diagnosis in 238 Patients on Basis of Retrospective MR Anthroscopic or Open Surgical Findings
|
|
Prospective MR imaging interpretations and surgical findings agreed in 117 of 121 normal cuffs (97% were true-negative findings) and 114 of 117 abnormal cuffs (97% were true-positive findings). Four MR imaging findings were not confirmed at surgery (false-positive findings), and three abnormalities identified at surgery were overlooked on MR images (false-negative findings). Thus, overall sensitivity and specificity of the diagnosis of cuff lesions were both 97%. In the diagnosis of full-thickness tears, sensitivity of MR imaging was 93% and specificity was 98%, with a positive predictive value of 95% and a negative predictive value of 98%. In the diagnosis of partial-thickness tears, sensitivity and specificity, respectively, were 82% and 98% for articular-sided tears and 73% and 98% for bursal-sided tears.
Surgical and preoperative MR imaging reports did not provide sufficient information to complete our database for 12 shoulders with surgically proved rotator cuff tears (5.0% of all shoulders, 10.2% of shoulders with cuff disorders). At retrospective review of MR images, we determined that partial-thickness tears were articular sided in three shoulders and bursal sided in one shoulder. It was necessary to measure the degree of tendon retraction in eight shoulders with full-thickness tears.
Cystic Changes Involving Greater Tuberosity
One or more cysts were identified in 157 (66.0%) of 238 shoulders. Posterior cysts occurred in 135 (56.7%) shoulders; anterior cysts, in 54 (22.7%) shoulders; and both posterior and anterior cysts, in 32 (13.4%) shoulders. More than 95% of the posterior cysts were localized to the posterior centimeter of the greater tuberosity (posterior margin of the infraspinatus tendon). Anterior cysts were localized to the anterior centimeter of the greater tuberosity (anterior margin of the supraspinatus tendon). Thus, posterior and anterior cysts were geographically distinct and easily distinguished (Fig 2). In only three shoulders, cysts involved the middle region where supraspinatus and infraspinatus tendons overlapped. Both posterior and anterior cysts were usually solitary (73% of posterior cysts and 78% of anterior cysts). When multiple, they rarely numbered more than two in either location (four shoulders with posterior occurrence of multiple cysts, one shoulder with anterior occurrence of multiple cysts). Mean diameters of anterior and posterior cysts were 3 mm ± 1.3 (standard deviation) (range, 2–8 mm) and 4 mm ± 1.8 (range, 2–11 mm), respectively.
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2: T1-weighted (600/14) oblique sagittal MR arthrographic image of anterior and posterior cystic changes at greater tuberosity in 41-year-old patient with surgically proved full-thickness tear of supraspinatus tendon. Subcortical cystic change (open arrow) involves anterior margin of greater tuberosity. Overlying high signal intensity contrast material fills the gap created by retracted tear of supraspinatus tendon (*). Another subcortical cyst (white arrow) involves posterior margin of greater tuberosity at insertion site of infraspinatus tendon (arrowhead), which remains intact.
|
|
Mean ages of patients (Table 2) varied significantly depending on the presence or absence of rotator cuff disorders and anterior cysts, whereas age showed no significant variation with the presence or absence of posterior cysts. After controlling for rotator cuff diagnosis, we found that mean ages of patients showed no statistical variation, regardless of any cystic changes at the tendon insertion site (Table 3).
View this table:
[in this window]
[in a new window]
|
Table 2. Mean Age of Patients according to Rotator Cuff Diagnosis or Cystic Change at Rotator Cuff Tendon Insertion
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Mean Age of Patients according to Cystic Change at Rotator Cuff Insertion and Presence or Absence of Rotator Cuff Disorder
|
|
Therefore, posterior cysts occurred with similar probability across all age groups (Fig 3), and there was no significant difference between the mean age of patients with posterior cysts (44 years) and that of patients without posterior cysts (41 years) (P > .07). Sixty-one (45.2%) of 135 shoulders with posterior cysts had intact rotator cuffs (Figs 4 and 5). The prevalence of posterior cysts in shoulders with intact cuffs (50.4%) was not significantly different from that in shoulders with cuff disorders (63.2%) (P > .05). Posterior cysts showed no relation to age in shoulders with normal cuffs (P > .20) or in those with cuff disorders (P > .90). Thus, receiver operator characteristic analysis revealed a poor sensitivity of 63% and specificity of 50% for posterior cysts in the prediction of cuff disorders (positive predictive value, 55%; negative predictive value, 58%; odds ratio, 1.7).
View larger version (7K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a: Logistic probability plots show relationship of patient age to presence of anterior cysts, posterior cysts, and rotator cuff tears. (a) Plot shows increasing likelihood of rotator cuff disease (RCD) (solid line) in patients between 40 and 60 years old (P < .001). Anterior cysts (CA) (dotted line) also begin increasing in likelihood at approximately 40 years (P < .001). Posterior cysts (CP) (dashed line) show no significant correlation with patient age (P > .05). (b) In 121 patients with normal rotator cuffs at surgery, logistic probability curves remain relatively flat for both anterior and posterior cysts; this indicates no significant relationship between age and presence of either anterior (P > .50) or posterior cysts (P > .20). Probability of posterior cysts is intermediate (.4–.6), whereas probability of anterior cysts is low (less than .05). (c) In 117 patients with rotator cuff disorders at surgery, probabilities of anterior and posterior cysts remain constant with patient age. Therefore, neither posterior (P > .9) nor anterior cysts (P > .7) show correlation with age. Posterior cysts show same probability in patients with rotator cuff disorders as that in patients with normal cuffs, whereas probability of anterior cysts increases to approximately .4 from .05 in shoulders with cuff disorders (P < .001).
|
|
View larger version (6K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b: Logistic probability plots show relationship of patient age to presence of anterior cysts, posterior cysts, and rotator cuff tears. (a) Plot shows increasing likelihood of rotator cuff disease (RCD) (solid line) in patients between 40 and 60 years old (P < .001). Anterior cysts (CA) (dotted line) also begin increasing in likelihood at approximately 40 years (P < .001). Posterior cysts (CP) (dashed line) show no significant correlation with patient age (P > .05). (b) In 121 patients with normal rotator cuffs at surgery, logistic probability curves remain relatively flat for both anterior and posterior cysts; this indicates no significant relationship between age and presence of either anterior (P > .50) or posterior cysts (P > .20). Probability of posterior cysts is intermediate (.4–.6), whereas probability of anterior cysts is low (less than .05). (c) In 117 patients with rotator cuff disorders at surgery, probabilities of anterior and posterior cysts remain constant with patient age. Therefore, neither posterior (P > .9) nor anterior cysts (P > .7) show correlation with age. Posterior cysts show same probability in patients with rotator cuff disorders as that in patients with normal cuffs, whereas probability of anterior cysts increases to approximately .4 from .05 in shoulders with cuff disorders (P < .001).
|
|
View larger version (6K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3c: Logistic probability plots show relationship of patient age to presence of anterior cysts, posterior cysts, and rotator cuff tears. (a) Plot shows increasing likelihood of rotator cuff disease (RCD) (solid line) in patients between 40 and 60 years old (P < .001). Anterior cysts (CA) (dotted line) also begin increasing in likelihood at approximately 40 years (P < .001). Posterior cysts (CP) (dashed line) show no significant correlation with patient age (P > .05). (b) In 121 patients with normal rotator cuffs at surgery, logistic probability curves remain relatively flat for both anterior and posterior cysts; this indicates no significant relationship between age and presence of either anterior (P > .50) or posterior cysts (P > .20). Probability of posterior cysts is intermediate (.4–.6), whereas probability of anterior cysts is low (less than .05). (c) In 117 patients with rotator cuff disorders at surgery, probabilities of anterior and posterior cysts remain constant with patient age. Therefore, neither posterior (P > .9) nor anterior cysts (P > .7) show correlation with age. Posterior cysts show same probability in patients with rotator cuff disorders as that in patients with normal cuffs, whereas probability of anterior cysts increases to approximately .4 from .05 in shoulders with cuff disorders (P < .001).
|
|
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a: MR images of posterior cystic change in 27-year-old patient with intact rotator cuff at MR arthroscopy. No anterior cyst was present. (a) On intermediate-weighted (2700/17) oblique coronal image, cystic abnormality (arrow) is located posteriorly on greater tuberosity at attachment site of infraspinatus tendon (white arrowhead). Cyst extends to cortical bone (black arrowhead) in bare area of anatomic neck. (b) On T1-weighted (600/14) oblique sagittal image, cyst (arrow) shows multicentric configuration and cortical extension (black arrowhead). Typical for posterior cysts, overlying infraspinatus tendon (white arrowhead) is intact.
|
|
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b: MR images of posterior cystic change in 27-year-old patient with intact rotator cuff at MR arthroscopy. No anterior cyst was present. (a) On intermediate-weighted (2700/17) oblique coronal image, cystic abnormality (arrow) is located posteriorly on greater tuberosity at attachment site of infraspinatus tendon (white arrowhead). Cyst extends to cortical bone (black arrowhead) in bare area of anatomic neck. (b) On T1-weighted (600/14) oblique sagittal image, cyst (arrow) shows multicentric configuration and cortical extension (black arrowhead). Typical for posterior cysts, overlying infraspinatus tendon (white arrowhead) is intact.
|
|
View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a: After MR arthrography, posterior cyst fills with contrast material in 45-year-old patient with intact rotator cuff. No anterior cyst was present. (a) On fat-suppressed T1-weighted (800/18) oblique coronal MR arthrographic image, high signal intensity of contrast material fills posterior cyst (arrow) at attachment site of infraspinatus tendon (arrowhead). (b) Fat-suppressed T1-weighted (800/18) oblique sagittal MR arthrographic image confirms posterior location of cyst (arrow). Presence of high signal intensity of contrast material reflects communication between cyst and joint space at bare area of cortical bone (arrowhead) underlying infraspinatus tendon attachment site.
|
|
View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b: After MR arthrography, posterior cyst fills with contrast material in 45-year-old patient with intact rotator cuff. No anterior cyst was present. (a) On fat-suppressed T1-weighted (800/18) oblique coronal MR arthrographic image, high signal intensity of contrast material fills posterior cyst (arrow) at attachment site of infraspinatus tendon (arrowhead). (b) Fat-suppressed T1-weighted (800/18) oblique sagittal MR arthrographic image confirms posterior location of cyst (arrow). Presence of high signal intensity of contrast material reflects communication between cyst and joint space at bare area of cortical bone (arrowhead) underlying infraspinatus tendon attachment site.
|
|
Patients with anterior cysts were significantly older (mean age, 52 years) than patients without anterior cysts (mean age, 40 years) (Table 2, Fig 3a) (P < .001), which is analogous to the age difference in patients with rotator cuff disorders (mean age, 53 years) and that of those without cuff disorders (mean age, 33 years) (Table 2, Fig 3a) (P < .001). Anterior cysts showed correlation with cuff disorders independent of age (Table 1) (P < .001), but no correlation with age independent of cuff disorders (Table 3, Fig 3b, 3c) (P > .50 for no rotator cuff disorder and P > .70 for rotator cuff disorder). Anterior cysts were present in 43.6% of shoulders with cuff disorders, including six of 13 shoulders with surgically proved tendinopathy, whereas anterior cysts were present in only 2.5% of shoulders with normal cuffs (Table 1). Ninety-four percent of shoulders with anterior cysts had a cuff disorder. Receiver operator characteristic curves showed sensitivity of 44%, specificity of 98%, and accuracy of 71% for anterior cysts in the prediction of cuff disorders (positive predictive value, 94%; negative predictive value, 64%; odds ratio, 30.4) (P < .05).
Anterior cysts were present in 27 (48%) of 56 full-thickness tears, 16 (48%) of 33 partial-thickness articular-sided tears, two (13%) of 15 partial-thickness bursal-sided tears, and six (46%) of 13 surgically reported cases of tendinopathy (Table 1). For prediction of full-thickness tears (odds ratio, 5.3), sensitivity was 48%, specificity was 85%, positive predictive value was 50%, and negative predictive value was 84% (P < .001). For prediction of partial-thickness articular-sided tears (odds ratio, 4.1), sensitivity was 48%, specificity was 81%, positive predictive value was 30%, and negative predictive value was 91% (P < .001). For prediction of partial-thickness bursal-sided tears, anterior cysts showed no statistical significance (P > .30). Therefore, anterior cysts were far more likely to occur in shoulders with partial-thickness articular-sided tears (Fig 6) than in those with bursal-sided tears (P < .001). Compared with anterior cysts alone, the presence of both anterior and posterior cysts did not improve the prediction of cuff disorders.
View larger version (163K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a: MR images of anterior cystic lesion underlying supraspinatus tendon in 68-year-old patient with surgically proved partial-thickness, articular-sided rotator cuff tear. No posterior cyst was present. (a) Intermediate-weighted (2700/17) oblique coronal image shows sharply marginated subcortical lesion (arrow) at anterior aspect of greater tuberosity. Overlying supraspinatus tendon is diffusely thickened, with abnormal signal intensity (arrowhead) involving articular-sided fibers. (b) On fat-suppressed intermediate-weighted (2900/54) oblique coronal image, partial-thickness, articular-sided rotator cuff tear is filled with fluid, whereas bursal-sided fibers remain intact on greater tuberosity. Anterior cyst shows surrounding marrow edema (arrows).
|
|
View larger version (173K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b: MR images of anterior cystic lesion underlying supraspinatus tendon in 68-year-old patient with surgically proved partial-thickness, articular-sided rotator cuff tear. No posterior cyst was present. (a) Intermediate-weighted (2700/17) oblique coronal image shows sharply marginated subcortical lesion (arrow) at anterior aspect of greater tuberosity. Overlying supraspinatus tendon is diffusely thickened, with abnormal signal intensity (arrowhead) involving articular-sided fibers. (b) On fat-suppressed intermediate-weighted (2900/54) oblique coronal image, partial-thickness, articular-sided rotator cuff tear is filled with fluid, whereas bursal-sided fibers remain intact on greater tuberosity. Anterior cyst shows surrounding marrow edema (arrows).
|
|
Most full-thickness tears showed tendon retraction (55 of 56). Fourteen (25%) of 55 were retracted more than 40 mm (grade 5), five (9%) were retracted between 31 and 40 mm (grade 4), 13 (24%) were retracted between 21 and 30 mm (grade 3), 17 (31%) were retracted between 11 and 20 mm (grade 2), and six (11%) were retracted up to 10 mm (grade 1). There was no significant correlation between anterior or posterior cysts and the degree of tendon retraction. Marrow edema involved the greater tuberosity in only 12 (5.0%) of 238 patients and was always associated with cystic changes (10 anterior, two posterior).
|
DISCUSSION
|
|---|
Our results indicate that anterior and posterior cysts are distinct both in their geographic locations along the greater tuberosity and in their relationships to rotator cuff disorders. In our patients, posterior cysts demonstrated slightly increased prevalence in older patients and patients with cuff disease, but the trends were not statistically significant (P > .07 and P > .05, respectively). Anterior cysts were closely related with cuff disorders (P < .001) but showed no significant correlation with age after we controlled for cuff disorders (P > .50).
Posterior and anterior cysts share several MR imaging appearances. Both are located at the rotator cuff insertion site on the greater tuberosity or the adjacent bare area of bone at the anatomic neck of the humeral head (20). They can be cortically based and similar to erosions seen in inflammatory arthritis or can be subcortical and similar to cysts seen in degenerative joint disease. Although posterior and anterior cysts commonly occur simultaneously in the same shoulder, in each location they are solitary or total only two or three in number. They are usually small (3–4 mm) but sometimes measure more than 1 cm.
Despite these similarities on MR images, our results suggest that posterior and anterior cysts may have different origins. Several authors (12,15–18) have proposed a degenerative cause for posterior cysts, supporting this conclusion with data indicating an increased incidence of posterior cysts with age. Given the high prevalence (56.7%) in both younger and older patients, it is difficult to explain the presence of posterior cysts on the basis of degenerative change or any specific shoulder lesion. Several other explanations are possible. Because posterior cysts occur at the bare area of the humeral head where cortical bone is not protected by overlying articular cartilage (1,20), they may represent intraosseous synovial cysts similar to the herniation pits that commonly involve the femoral neck (22). Cysts can be lined with synovium that merges with the synovial lining of the glenohumeral joint (1). Because the posterior aspect of the greater tuberosity impacts against the superior glenoid during abduction and external rotation, posterior cysts could result from posterosuperior impingement. They could also represent dilated vascular channels or intraosseous varices, although we excluded any cystic changes that connected to tubular structures.
In our study, anterior cysts were related to rotator cuff disorders and occurred most often in older patients. In the absence of any cuff disorder, the intact supraspinatus tendon may protect against anterior cyst formation by attaching firmly to bone and forming a barrier to synovial fluid. At its osseous insertion site, or "footprint," the supraspinatus tendon measures approximately 12 mm in thickness (width) and is separated from the adjacent articular cartilage of the humeral head by only 1.5–1.7 mm (20). This distance increases as the supraspinatus tears and detaches from bone (20), thereby exposing the osseous footprint to synovial fluid. If the tendon creates cortical infractions as it continues to tear, fluid may leak into the greater tuberosity and facilitate the ingrowth of synovium and granulation tissue, which excavates bone to form a cyst. This pathogenesis may explain why anterior cysts are more common in partial-thickness articular-sided tears than in bursal-sided tears. Fluid may also gain access to the cuff insertion site as the tendon degenerates and its fibers delaminate.
Results of prior imaging studies (10–14) have revealed an association between rotator cuff disorders and osseous abnormalities, including cystic changes involving the greater tuberosity. Authors of one of these studies (12) differentiated anterior from posterior cysts on MR images and analyzed their relationships to cuff disorders. The authors related posterior cysts to aging and anterior cysts to supraspinatus tendon tears, but they did not have surgical confirmation of cuff diagnoses. In a human cadaveric study (14), US findings were correlated with trabecular and cortical changes in the greater tuberosity. After the cuff was characterized as intact, torn, or partially torn, the humeral head was sliced into 3-mm slices for microradiography. Abnormalities of the greater tuberosity included pitting, irregular contour, sclerosis, cortical thickening, fragmentation, and erosion formation. These changes were mild in shoulders with intact cuffs and prominent in shoulders with cuff tears, especially full-thickness tears.
US can depict cortical irregularities of the greater tuberosity that may be difficult or impossible to identify on MR images. Results of a previous study (10) with US revealed cortical irregularities at the supraspinatus footprint in 86% of shoulders with full-thickness tears, 40% of shoulders with partial-thickness bursal-sided tears, and 50% of shoulders with partial-thickness articular-sided tears. The authors concluded that cortical surface irregularities represented an important secondary sign of cuff disorder. Compared with other US signs, it showed the highest sensitivity and negative predictive value in the diagnosis of tendon tear. US may be less sensitive than MR in the demonstration of subcortical cysts.
Our study had limitations. Cysts were retrospectively identified on MR images and analyzed without histopathologic confirmation. In a prior study (23) of the capability of MR imaging in monitoring the progression of bone abnormalities in degenerative joint disease in a guinea pig model, MR image analysis was found to be a valid method in the detection of osseous cysts. Thus, all subchondral cysts identified on MR images were histologically confirmed on 1-mm slices (23). Although we standardized our identification of cysts by using strict criteria, bias was possible because our reviewers were able to see the rotator cuff. In our population of surgical patients, patient selection bias (24) may have resulted in the overestimation of the prevalence of cystic changes and, therefore, may have resulted in the demonstration of more cysts than are actually present in the general population.
We did not subcategorize cysts as located at either (a) the osseous insertion site, or footprint, of the cuff tendon; or (b) the bare area, or anatomic neck, of the humeral head adjacent to the osseous insertion site. Because it was difficult to make this anatomic distinction on the basis of MR imaging criteria, we only categorized cysts as either anterior or posterior along the greater tuberosity. Therefore, we were more limited in drawing conclusions about the causes of cysts. Although surgery was always the reference standard for the diagnosis of normal cuff, tendinopathy, partial-thickness tear, and full-thickness tear, bias may have been introduced because in four patients the surgical report did not indicate whether the partial-thickness tear was bursal sided or articular sided. Therefore, this information was determined by the reviewers.
Our study results show that cystic changes often involve the greater tuberosity and fall into two distinct geographic locations that are easily distinguished on MR images. Posterior cysts at the infraspinatus insertion site were common abnormalities that showed no significant correlation with either age or cuff disorders, whereas anterior cysts at the supraspinatus insertion site were closely related to rotator cuff disorders, including partial-thickness tears that may be difficult to identify on MR images. In the presence of equivocal cuff tears, anterior cysts should increase diagnostic confidence or lead to closer scrutiny of the supraspinatus tendon.
|
ADVANCES IN KNOWLEDGE
|
|---|
- Cystic changes demonstrated two distinct geographic locations along the greater tuberosity of the humerus and had different origins.
- Anterior cystic change at the supraspinatus attachment site was closely related to adjacent rotator cuff abnormality, including partial tendon tear.
- Posterior cystic change near the infraspinatus attachment site was a common MR imaging finding that was not associated with patient age or rotator cuff disorder.
|
IMPLICATION FOR PATIENT CARE
|
|---|
- Anterior cystic change may increase diagnostic confidence in the presence of cuff disorder or may lead to closer scrutiny of the supraspinatus tendon for subtle tear.
|
ACKNOWLEDGMENTS
|
|---|
The authors acknowledge Martin P. Torriani, MD, Elkan F. Halpern, PhD, and Mareike Braun, MS, for their substantial contribution to the material discussed in this article.
|
FOOTNOTES
|
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantor of integrity of entire study, W.E.P.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, L.B.F., T.A.O., A.K., W.E.P.; clinical studies, L.B.F., H.A.O., A.K., W.E.P.; statistical analysis, L.B.F., W.E.P.; and manuscript editing, all authors
|
References
|
|---|
- Codman EA. The shoulder. Boston, Mass: Thomas Todd, 1934.
- Cotton RE, Rideout DF. Tears of the humeral rotator cuff: a radiological and pathological necropsy survey. J Bone Joint Surg Br 1964;46:314–328.[Medline]
- Harrison SH. The painful shoulder: significance of radiographic changes in the upper end of the humerus. J Bone Joint Surg Br 1949;31:418–422.
- Golding FC. The shoulder: the forgotten joint. Br J Radiol 1962;35:149–158.[Medline]
- Kernwein GA. Roentgenographic diagnosis of shoulder dysfunction. JAMA 1965;194:1081–1085.[CrossRef][Medline]
- Kieft GJ, Bloem JL, Rozing PM, Obermann WR. Rotator cuff impingement syndrome: MR imaging. Radiology 1988;166:211–214.[Abstract/Free Full Text]
- Iannotti JP, Zlatkin MB, Esterhai JL, Kressel HY, Dalinka MK, Spindler KP. Magnetic resonance imaging of the shoulder: sensitivity, specificity, and predictive value. J Bone Joint Surg Am 1991;73:17–29.[Abstract/Free Full Text]
- Palmer WE. MR arthrography of the rotator cuff and labral-ligamentous complex. Semin Ultrasound CT MR 1997;18:278–290.[Medline]
- Seibold CJ, Mallisee TA, Erickson SJ, Boynton MD, Raasch WG, Timins ME. Rotator cuff: evaluation with US and MR imaging. RadioGraphics 1999;19:685–705.[Abstract/Free Full Text]
- Jacobson JA, Lancaster S, Prasad A, van Holsbeeck MT, Craig JG, Kolowich P. Full-thickness and partial-thickness supraspinatus tendon tears: value of US signs in diagnosis. Radiology 2004;230:234–242.[Abstract/Free Full Text]
- Wohlwend JR, van Holsbeeck M, Craig J, et al. The association between irregular greater tuberosities and rotator cuff tears: a sonographic study. AJR Am J Roentgenol 1998;171:229–233.[Abstract/Free Full Text]
- Sano A, Itoi E, Konno N, Kido T, Urayama M, Sato K. Cystic changes of the humeral head on MR imaging: relation to age and cuff-tears. Acta Orthop Scand 1998;69:397–400.[Medline]
- Burk DL Jr, Karasick D, Mitchell DG, Rifkin MD. MR imaging of the shoulder: correlation with plain radiography. AJR Am J Roentgenol 1990;154:549–553.[Free Full Text]
- Jiang Y, Zhao J, van Holsbeeck MT, Flynn MJ, Ouyang X, Genant HK. Trabecular microstructure and surface changes in the greater tuberosity in rotator cuff tears. Skeletal Radiol 2002;31:522–528.[CrossRef][Medline]
- Mirowitz SA. Normal rotator cuff: MR imaging with conventional and fat-suppression techniques. Radiology 1991;180:735–740.[Abstract/Free Full Text]
- Needell SD, Zlatkin MB, Sher JS, Murphy BJ, Uribe JW. MR imaging of the rotator cuff: peritendinous and bone abnormalities in an asymptomatic population. AJR Am J Roentgenol 1996;166:863–867.[Abstract/Free Full Text]
- Huang LF, Rubin DA, Britton CA. Greater tuberosity changes as revealed by radiography: lack of clinical usefulness in patients with rotator cuff disease. AJR Am J Roentgenol 1999;172:1381–1388.[Abstract/Free Full Text]
- Wright T, Yoon C, Schmit BP. Shoulder MRI refinements: differentiation of rotator cuff tear from artifacts and tendonosis, and reassessment of normal findings. Semin Ultrasound CT MR 2001;22:383–395.[CrossRef][Medline]
- Snyder SJ, Pachelli AF, Del Pizzo W, Friedman MJ, Ferkel RD, Pattee G. Partial thickness rotator cuff tears: results of arthroscopic treatment. Arthroscopy 1991;7:1–7.[Medline]
- Ruotolo C, Fow JE, Nottage WM. The supraspinatus footprint: an anatomic study of the supraspinatus insertion. Arthroscopy 2004;20:246–249.[Medline]
- Minagawa H, Itoi E, Konno N, et al. Humeral attachment of the supraspinatus and infraspinatus tendons: an anatomic study. Arthroscopy 1998;14:302–306.[Medline]
- Pitt MJ, Graham AR, Shipman JH, Birkby W. Herniation pit of the femoral neck. AJR Am J Roentgenol 1982;138:1115–1121.[Abstract/Free Full Text]
- Watson PJ, Hall LD, Malcolm A, Tyler JA. Degenerative joint disease in the guinea pig: use of magnetic resonance imaging to monitor progression of bone pathology. Arthritis Rheum 1996;39:1327–1337.[Medline]
- Colton T. Statistics in medicine. Boston, Mass: Little, Brown, 1974.
This article has been cited by other articles:
|
|
|
|
 
U. Studler, C. W. A. Pfirrmann, B. Jost, V. Rousson, J. Hodler, and M. Zanetti
Abnormalities of the Lesser Tuberosity on Radiography and MRI: Association with Subscapularis Tendon Lesions
Am. J. Roentgenol.,
July 1, 2008;
191(1):
100 - 106.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
