DOI: 10.1148/rg.262045719
RadioGraphics 2006;26:589-604
© RSNA, 2006
From the RSNA Refresher Courses
US of the Rotator Cuff: Pitfalls, Limitations, and Artifacts1
Matthieu J. C. M. Rutten, MD,
Gerrit J. Jager, MD, PhD and
Johan G. Blickman, MD, PhD
1 From the Department of Radiology, Jeroen Bosch Hospital, Nieuwstraat 34, 5211 NL ‘s-Hertogenbosch, the Netherlands (M.J.C.M.R., G.J.J.); and the Department of Radiology, University Medical Center, Nijmegen, the Netherlands (J.G.B.). Presented as a refresher course at several RSNA Annual Meetings. Received July 28, 2004; revision requested September 22; final revision received August 26, 2005; accepted August 29. All authors have no financial relationships to disclose.
Address correspondence to M.J.C.M.R. (e-mail: M.Rutten{at}JBZ.nl).
 |
Abstract
|
|---|
High-resolution ultrasonography (US) has gained increasing popularity as a diagnostic tool for assessment of the soft tissues in shoulder impingement syndrome. US is a powerful and accurate method for diagnosis of rotator cuff tears and other rotator cuff abnormalities, provided the examiner has a detailed knowledge of shoulder anatomy, uses a standardized examination technique, and has a thorough understanding of the potential pitfalls, limitations, and artifacts. False-positive sonographic findings of rotator cuff tears can be caused by the technique (anisotropy, transducer positioning, acoustic shadowing by the deltoid septum), by the anatomy (rotator cuff interval, supraspinatus-infraspinatus interface, musculotendinous junction, fibrocartilaginous insertion), or by disease (criteria for diagnosis of rotator cuff tears, tendon inhomogeneity, acoustic shadowing by scar tissue or calcification, rotator cuff thinning). False-negative sonographic findings of rotator cuff tears can be caused by the technique (transducer frequency, suboptimal focusing, imaging protocol, transducer handling), by the anatomy (nondiastasis of the ruptured tendon fibers, posttraumatic obscuration of landmarks), by disease (tendinosis, calcifications, synovial proliferation, granulation or scar tissue, bursal thickening, massive rotator cuff tears), or by patient factors (obesity, muscularity, limited shoulder motion).
© RSNA, 2006
|
Introduction
|
|---|
Since the 1980s, ultrasonography (US) of the shoulder has been performed for the diagnosis of rotator cuff tears. Technical developments and improvements, increased experience, and detailed knowledge of shoulder anatomy and pathologic conditions have significantly improved sonographic results (1–4).
Owing to the complex shoulder anatomy and various pitfalls, US of the shoulder is susceptible to interobserver variability and has a learning curve. This can be improved by performing US of the shoulder on a regular basis and in a standardized way and by being aware of sonographic pitfalls (5–7).
The purpose of this article is to provide an overview of potential pitfalls, limitations, and artifacts related to US of the shoulder. Causes of false-positive and false-negative misinterpretation of rotator cuff tears can be classified into four different categories: technique-related, anatomy-related, disease-related, and patient-related factors (Tables 1, 2).
|
Causes of False-Positive Diagnoses of Rotator Cuff Tears
|
|---|
Technique-related Causes of Misinterpretation
Anisotropy.—
A common cause of false-positive diagnoses of rotator cuff tears is anisotropy or angle-dependent appearance of tissue structures (6,8). To our knowledge, anisotropy of fibers was first described by Dussik et al (9) in 1958.
The rotator cuff appears echogenic when the ultrasound beam insonates at 90° to the long axis of the tendon fibers because the beam is then reflected maximally (Fig 1a). The more the angle deviates from this angle, the fewer reflected sound waves will be detected by the transducer (Fig 1b). The tendon becomes isoechoic to muscle at angles of 2°–7° (8) and hypoechoic at greater angles (Figs 2, 3). Tendon insertions, where most rotator cuff tears occur, are most vulnerable to the anisotropic artifact due to their curved course (10). If unaware of this artifact, less experienced scanners could erroneously take this for tendinosis or a partial-thickness rotator cuff tear.
View larger version (10K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Anisotropy. (a) Tendon fibers have a parallel arrangement. Emitted sound waves are optimally reflected when they are perpendicular (at 90°) to the long axis of the fibers. (b) Deviation of the insonating beam from this angle causes a decrease in the echogenicity of the fibers because not all of the reflected sound waves will return to the transducer.
|
|
View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Anisotropy. (a) Tendon fibers have a parallel arrangement. Emitted sound waves are optimally reflected when they are perpendicular (at 90°) to the long axis of the fibers. (b) Deviation of the insonating beam from this angle causes a decrease in the echogenicity of the fibers because not all of the reflected sound waves will return to the transducer.
|
|
View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Anisotropy at the insertion of the supraspinatus tendon. GT = greater tuberosity. (a) Long-axis US scan of the supraspinatus tendon (SSP) shows that the fibers parallel to the transducer have a normal hyperechoic linear appearance (arrowheads). However, the fibers at the insertion (arrows) are poorly demonstrated due to anisotropy. (b) Corresponding image obtained with the transducer moved a bit laterally. The fibers at the supraspinatus tendon (SSP) insertion (arrows) are now parallel to the transducer and therefore have a normal hyperechoic appearance.
|
|
View larger version (78K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Anisotropy at the insertion of the supraspinatus tendon. GT = greater tuberosity. (a) Long-axis US scan of the supraspinatus tendon (SSP) shows that the fibers parallel to the transducer have a normal hyperechoic linear appearance (arrowheads). However, the fibers at the insertion (arrows) are poorly demonstrated due to anisotropy. (b) Corresponding image obtained with the transducer moved a bit laterally. The fibers at the supraspinatus tendon (SSP) insertion (arrows) are now parallel to the transducer and therefore have a normal hyperechoic appearance.
|
|
View larger version (113K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Anisotropy. (a, b) Short-axis US scans of the long head of the biceps tendon (BT). (a) When the insonating beam is perpendicular to the tendon fibers, the tendon appears hyperechoic. It is round or oval and lies in the bicipital groove (BG). GT = greater tuberosity, LT = lesser tuberosity. (b) The tendon appears hypoechoic because the transducer is angled relative to the long axis of the tendon. (c, d) Long-axis US scans of the long head of the biceps tendon. (c) When the transducer is parallel to the tendon fibers, the tendon has a normal hyperechoic, linear, fibrillar appearance (arrows). (d) The tendon is not seen (arrows) due to anisotropy. Arrowheads = fibers parallel to the transducer.
|
|
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Anisotropy. (a, b) Short-axis US scans of the long head of the biceps tendon (BT). (a) When the insonating beam is perpendicular to the tendon fibers, the tendon appears hyperechoic. It is round or oval and lies in the bicipital groove (BG). GT = greater tuberosity, LT = lesser tuberosity. (b) The tendon appears hypoechoic because the transducer is angled relative to the long axis of the tendon. (c, d) Long-axis US scans of the long head of the biceps tendon. (c) When the transducer is parallel to the tendon fibers, the tendon has a normal hyperechoic, linear, fibrillar appearance (arrows). (d) The tendon is not seen (arrows) due to anisotropy. Arrowheads = fibers parallel to the transducer.
|
|
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3c. Anisotropy. (a, b) Short-axis US scans of the long head of the biceps tendon (BT). (a) When the insonating beam is perpendicular to the tendon fibers, the tendon appears hyperechoic. It is round or oval and lies in the bicipital groove (BG). GT = greater tuberosity, LT = lesser tuberosity. (b) The tendon appears hypoechoic because the transducer is angled relative to the long axis of the tendon. (c, d) Long-axis US scans of the long head of the biceps tendon. (c) When the transducer is parallel to the tendon fibers, the tendon has a normal hyperechoic, linear, fibrillar appearance (arrows). (d) The tendon is not seen (arrows) due to anisotropy. Arrowheads = fibers parallel to the transducer.
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3d. Anisotropy. (a, b) Short-axis US scans of the long head of the biceps tendon (BT). (a) When the insonating beam is perpendicular to the tendon fibers, the tendon appears hyperechoic. It is round or oval and lies in the bicipital groove (BG). GT = greater tuberosity, LT = lesser tuberosity. (b) The tendon appears hypoechoic because the transducer is angled relative to the long axis of the tendon. (c, d) Long-axis US scans of the long head of the biceps tendon. (c) When the transducer is parallel to the tendon fibers, the tendon has a normal hyperechoic, linear, fibrillar appearance (arrows). (d) The tendon is not seen (arrows) due to anisotropy. Arrowheads = fibers parallel to the transducer.
|
|
Transducer Positioning.—
The position of the transducer is related to the long or short axis of the anatomic structure under examination. When imaging the supraspinatus tendon in the transverse direction (ie, the short axis), the transducer is placed in a sagittal plane with regard to the patient’s shoulder and is moved anterior to posterior following the course of the tendon (Fig 4).
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Transducer position. D = deltoid muscle, H = humeral head. (a) Normal long-axis US scan of the supraspinatus tendon (SSP). GT = greater tuberosity. (b) Short-axis US scan of the supraspinatus tendon obtained too far laterally (at line 4b in a). At this position, the rotator cuff cannot be visualized between the humeral head and deltoid muscle, an appearance suggestive of a full-thickness tear of the supraspinatus tendon. (c) Normal short-axis US scan of the supraspinatus tendon (SSP), obtained at line 4c in a, shows the normal soft-tissue layers around the humeral head. c = hyaline cartilage.
|
|
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Transducer position. D = deltoid muscle, H = humeral head. (a) Normal long-axis US scan of the supraspinatus tendon (SSP). GT = greater tuberosity. (b) Short-axis US scan of the supraspinatus tendon obtained too far laterally (at line 4b in a). At this position, the rotator cuff cannot be visualized between the humeral head and deltoid muscle, an appearance suggestive of a full-thickness tear of the supraspinatus tendon. (c) Normal short-axis US scan of the supraspinatus tendon (SSP), obtained at line 4c in a, shows the normal soft-tissue layers around the humeral head. c = hyaline cartilage.
|
|
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. Transducer position. D = deltoid muscle, H = humeral head. (a) Normal long-axis US scan of the supraspinatus tendon (SSP). GT = greater tuberosity. (b) Short-axis US scan of the supraspinatus tendon obtained too far laterally (at line 4b in a). At this position, the rotator cuff cannot be visualized between the humeral head and deltoid muscle, an appearance suggestive of a full-thickness tear of the supraspinatus tendon. (c) Normal short-axis US scan of the supraspinatus tendon (SSP), obtained at line 4c in a, shows the normal soft-tissue layers around the humeral head. c = hyaline cartilage.
|
|
When the transducer has reached the most lateral part of the supraspinatus tendon insertion at the greater tuberosity, no rotator cuff can be visualized between the deltoid muscle and the humeral head (Fig 4b). This could be misinterpreted as a full-thickness rotator cuff tear. To prevent making such an error, every possible lesion should be verified in two planes.
If performed correctly, US allows reliable detection and quantification of rotator cuff tears (4). Both the Crass position (11) and the modified Crass position (12) reflect the true size of supraspinatus tears in the transverse plane. However, in the sagittal plane, the Crass position is more useful to quantify supraspinatus tears, as the modified Crass position leads to overestimation of the size of such tears (13).
Acoustic Shadowing by the Deltoid Septum.—
The deltoid muscle is a large triangular muscle that consists of anterior, intermediate, and posterior parts. The intermediate or central portion of the deltoid muscle, which arises from the acromial process, consists of oblique fibers. These fibers arise from the sides of four tendinous intersections and insert at the sides of three other tendinous intersections. These tendinous intersections pass alternately downward and upward toward one another in the substance of the muscle.
The anterior and posterior parts of the deltoid muscle, which arise from the clavicle and the spine of the scapula, are not arranged in this manner.
The tendinous intersections cause an acoustic shadow (ie, a refractile shadow) when they are relatively thick or scanned tangentially. Acoustic shadowing occurs at an interface between tissues that transmit sound at different velocities. It is characterized by reflection of sound away from the transducer. This causes a hypoechoic area within the tendon, which can simulate a rotator cuff tear (Fig 5). This shadowing decreases or disappears when the transducer is moved to other positions, whereas a true tear remains unchanged in appearance.
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5. Deltoid septum. Short-axis US scan of the supraspinatus tendon (SSP) in a normal volunteer shows hyperechoic lines (arrowheads) in the deltoid muscle (D), which represent septa of connective tissue. A posterior acoustic shadow (arrow) may appear when the insonating beam is perpendicular to the septa; such a shadow could be mistaken for a tear in the underlying tendon.
|
|
Anatomy-related Causes of Misinterpretation
Errors caused by failure to recognize normal anatomy are diverse. They can be overcome by studying anatomy, thorough supervision by an experienced colleague, and by comparison with the contralateral side.
Rotator Cuff Interval.—
The rotator cuff is a continuous tendinous structure around the shoulder joint formed by the tendons of the subscapular, supraspinatus, infraspinatus, and teres minor muscles. There is one single discontinuity, which is named the rotator interval. The rotator interval contains the long head of the biceps tendon, which descends from the glenohumeral joint through the interval into the bicipital groove. The rotator interval has a triangular shape, which is composed of the coracohumeral ligament and the superior glenohumeral ligament and envelops the anterior margin of the supraspinatus tendon and the superior margin of the subscapular tendon (14). The rotator interval varies in size and may not be apparent in some individuals (15).
At sonography, the rotator interval is a hypoechoic area surrounding the cross-sectioned long head of the biceps tendon; this area could be mistaken for a rotator cuff tear by less experienced radiologists (Fig 6). This problem can be overcome by looking for the rounded edge of the anterior part of the supraspinatus tendon and by verifying the biceps tendon by following it into the bicipital groove. The rotator interval is best evaluated with the arm in external rotation or by externally rotating the glenohumeral joint slowly (16).
View larger version (103K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Rotator cuff interval. (a) On a short-axis US scan of the supraspinatus tendon (SSP), the rotator cuff interval (RI) anterior to this tendon can easily be mistaken for a rotator cuff tear. BT = biceps tendon (long head), D = deltoid muscle, H = humeral head. (b) Oblique sagittal T1-weighted magnetic resonance (MR) arthrogram shows the position of the long head of the biceps tendon (BCP) in the rotator cuff interval. A = acromion, C = coracoid process, ISP = infraspinatus tendon, SSC = subscapular tendon, SSP = supraspinatus tendon.
|
|
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Rotator cuff interval. (a) On a short-axis US scan of the supraspinatus tendon (SSP), the rotator cuff interval (RI) anterior to this tendon can easily be mistaken for a rotator cuff tear. BT = biceps tendon (long head), D = deltoid muscle, H = humeral head. (b) Oblique sagittal T1-weighted magnetic resonance (MR) arthrogram shows the position of the long head of the biceps tendon (BCP) in the rotator cuff interval. A = acromion, C = coracoid process, ISP = infraspinatus tendon, SSC = subscapular tendon, SSP = supraspinatus tendon.
|
|
Supraspinatus-Infraspinatus Interface.—
Focal thinning of the rotator cuff is a feature of a rotator cuff tear (17) or can be seen in an atrophic but intact cuff, particularly in patients with rheumatoid arthritis. Bretzke et al (18) showed that thinning of the rotator cuff at the supraspinatus-infraspinatus interface (Fig 7) is a normal finding and should not be mistaken for a partial-thickness tear. One should be aware of this normal anatomic difference in rotator cuff thickness. Comparison with the contralateral shoulder is an additional support for avoiding this pitfall.
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7. Supraspinatus-infraspinatus interface. Short-axis US scan of the supraspinatus tendon (SSP) shows normal thinning of the rotator cuff at the supraspinatus-infraspinatus (ISP) interface (arrows). There is a significant difference between the diameter at the interface and the normal rotator cuff diameter (arrowheads).
|
|
Musculotendinous Junction.—
The junction zone between muscle and tendon is a complex of interdigitating muscle and tendon fibers. The junction of the multipennate subscapular tendon is subject to a varying appearance. The hyperechoic tendon fibers are interposed with hypoechoic muscle fibers, which may be confused with tendinosis by less experienced radiologists (Fig 8).
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8. Musculotendinous junction of the subscapular tendon. Long-axis US scan of the subscapular tendon (SSC) shows varying echogenicity of the interdigitating hyperechoic tendinous fibers and hypoechoic muscle fibers (*), an appearance that mimics tendinosis or a rotator cuff tear.
|
|
The infraspinatus tendon is centrally positioned in the infraspinatus muscle. The surrounding hypoechoic muscle fibers may be confused with a tendon tear, especially when the tendon is scanned obliquely.
Normal rotator cuff tendons occasionally contain slight inhomogeneities. Histologically, they consist of a complex of five layers (19), which have a fibrocartilaginous attachment at the humeral tuberosities (20). Vahlensieck et al (21) and Turrin and Cappello (22) reported that the supraspinatus muscle consists of two distinct portions: an anterior fusiform (cylindrical) portion that contains the dominant tendon and a straplike (flat) posterior portion. This causes a fanning out and running in slightly different directions of the fascicles of the supraspinatus tendon. The differently oriented tendon fascicles and complex interdigitation of muscle fibers between the anterior and posterior parts of the tendon result in varying echogenicity of the supraspinatus tendon; this varying echogenicity may be mistaken for tendinosis or a tear.
Fibrocartilaginous Insertion.—
The attachment site of tendons may contain an amount of fibrocartilage (20). This is related to the orientation of the tendon fibers with regard to the bony attachment site. The more the fibers follow a perpendicular course, the higher the content of fibrocartilage in the attachment zone. As with hyaline cartilage, the cartilage in the attachment zone appears hypoechoic. This may result in a thin hypoechoic zone in the tendon insertion near the hyperechoic reflection of the cortical bone. Familiarity with the anatomy will prevent a false-positive diagnosis like tendinosis or rotator cuff tear (Fig 9). The low echogenicity of the fibrocartilage attachment zone is reinforced by the anisotropy of the tendon fibers in this zone, which are curved and parallel with regard to the insonating ultrasound beam.
View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Fibrocartilaginous insertion. (a) Long-axis US scan of the supraspinatus tendon (SSP) shows the hypoechoic appearance of the fibrocartilaginous attachment zone (arrowheads) near the greater tuberosity (GT). C = articular hyaline cartilage, H = humeral head. (b) Macroscopic section from the cadaver of a 78-year-old man shows the fibrocartilaginous insertion (arrowheads) of the supraspinatus tendon.
|
|
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Fibrocartilaginous insertion. (a) Long-axis US scan of the supraspinatus tendon (SSP) shows the hypoechoic appearance of the fibrocartilaginous attachment zone (arrowheads) near the greater tuberosity (GT). C = articular hyaline cartilage, H = humeral head. (b) Macroscopic section from the cadaver of a 78-year-old man shows the fibrocartilaginous insertion (arrowheads) of the supraspinatus tendon.
|
|
Disease-related Causes of Misinterpretation
Definition of and Criteria for Rotator Cuff Tears.—
The US appearances of rotator cuff tears overlap partly with the spectrum of appearances of normal variation and/or other tendon abnormalities. One of the reasons why the results of studies of US detection of rotator cuff tears vary is the fact that different criteria or combinations of criteria for rotator cuff tears (Table 3) were used.
The criteria have been modified (23). A full-thickness rotator cuff tear is a defect in the tendon that reaches from the bursal to the articular margin. A partial-thickness tear is a focal discontinuity at the bursal or articular margin or is located intratendinously.
Brandt et al (23) showed that echogenic foci or bands are not reliable criteria for rotator cuff tears. These hyperechoic foci represent calcification, fibrotic scar tissue, synovitis, or hemorrhage. Partial- and full-thickness rotator cuff tears are visualized as hypoechoic lesions or mixed hyper-and hypoechoic lesions most frequently located in the critical zone of the supraspinatus tendon and should be verified in two orthogonal directions.
Secondary or indirect signs are reliable criteria for the detection of rotator cuff tears (24). Partial-thickness tears are frequently accompanied by cortical outpouchings (pitting) at the insertion of the rotator cuff tendons (1). Fluid in the glenohumeral joint is associated with the presence of rotator cuff tear in 60% of cases (25). When fluid is present in the subacromial-subdeltoid bursa and in the glenohumeral joint, the probability of a rotator cuff tear is 95% (25). In patients with a fluid-filled widened subacromial-subdeltoid bursa, a tear is apparent in 70% to more than 90% of cases (25,26).
Other indirect signs of partial- or full-thickness rotator cuff tears are the ability to compress the deltoid muscle into a cuff defect or against the humeral head (naked tuberosity sign) and a bright aspect of the humeral cartilage (cartilage interface sign or uncovered cartilage sign), which is caused by enhancement of the ultrasound signal due to fluid and loss of cuff tissue above the cartilage.
Tendon Inhomogeneity.—
Inhomogeneities of the tendon are frequently encountered with degenerative changes of the tendon (ie, tendinosis) (Fig 10a) (27,28). In our experience, the combination of tendinosis and anisotropy is the most common cause of a false-positive diagnosis of a partial-thickness rotator cuff tear (Fig 10b). Power Doppler US may be of help by demonstrating low-flow hyperemia associated with tendinosis (29), in contrast to no flow in a full-thickness rotator cuff tear.
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Tendon inhomogeneity. D = deltoid muscle, H = humeral head. (a) Short-axis US scan of the supraspinatus tendon (SSP) shows a hypoechoic appearance of the anterior part of the tendon (arrows) due to tendinosis. (b) Short-axis US scan of the supraspinatus tendon (SSP), obtained with a minor change in the angle of the insonating beam, shows near invisibility of the tendon due to a combination of tendinosis and anisotropy.
|
|
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Tendon inhomogeneity. D = deltoid muscle, H = humeral head. (a) Short-axis US scan of the supraspinatus tendon (SSP) shows a hypoechoic appearance of the anterior part of the tendon (arrows) due to tendinosis. (b) Short-axis US scan of the supraspinatus tendon (SSP), obtained with a minor change in the angle of the insonating beam, shows near invisibility of the tendon due to a combination of tendinosis and anisotropy.
|
|
Tendon inhomogeneities may be due to postoperative changes. Crass et al (30) and Mack et al (31) showed that in the postoperative shoulder, the rotator cuff is more echogenic than in the normal shoulder due to fibrosis or granulation tissue and that the soft-tissue planes are distorted or absent. Prickett et al (32) showed that despite these changes, the diagnostic accuracy of US in surgically treated shoulders appears to be comparable with that previously reported in shoulders that had not been operated on.
Acoustic Shadowing by Scar Tissue or Calcification.—
Trauma or surgery may also cause fibrosis or scarring of the soft tissues around the shoulder. Behind these lesions, acoustic shadows may occur (Fig 11). Intratendinous or intrabursal calcium deposits manifest as focal echogenicity and acoustic shadowing. Because of its structure (eg, milk of calcium) or size, calcification may not always cause well-defined acoustic shadowing. Correlation with plain radiographs is necessary to recognize these deposits and prevent misinterpretation of the hypoechoic shadowing zone as a rotator cuff tear (Fig 12b). Acoustic shadowing is also proportionate to transducer frequency.
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11. Acoustic shadowing. Long-axis US scan of the subscapular tendon (SSC) shows scar tissue (arrows) in the deltoid muscle (D). The scar tissue produces an acoustic shadow (arrowheads) at the insertion of the tendon, an appearance that mimics tendinosis or a tear. LT = lesser tuberosity.
|
|
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12a. Acoustic shadowing. (a) Long-axis US scan of the supraspinatus tendon (SSP) shows calcification (arrows) in the subacromial-subdeltoid bursa. The calcification produces an acoustic shadow, which obscures visualization of the tendon insertion. D = deltoid muscle, H = humeral head. (b) Radiograph shows the calcification (arrows).
|
|
View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12b. Acoustic shadowing. (a) Long-axis US scan of the supraspinatus tendon (SSP) shows calcification (arrows) in the subacromial-subdeltoid bursa. The calcification produces an acoustic shadow, which obscures visualization of the tendon insertion. D = deltoid muscle, H = humeral head. (b) Radiograph shows the calcification (arrows).
|
|
Rotator Cuff Thinning.—
Focal thinning of the rotator cuff is a feature of a partial-thickness rotator cuff tear, but it is also anatomy related (see the section on the supraspinatus-infraspinatus junction and Fig 7) and related to atrophy due to factors such as rheumatoid arthritis, disuse, nerve impingement, or surgery (33). Comparison with the contralateral side may be of help in differentiation from a rotator cuff tear. However, one should keep in mind that in a symptomatic shoulder without a rotator cuff tear or with a partial-thickness rotator cuff tear, there is a 0.5%–4.3% prevalence of a contralateral tear; with a full-thickness tear, there is a 35.5% prevalence of a full-thickness tear on the asymptomatic side (34).
The average thickness of an intact rotator cuff is approximately 4.7 mm (34). There is a slight but not significant difference in rotator cuff thickness between the dominant limb (range, 3.6–7.0 mm; mean, 5.3 mm) and nondominant limb (range, 3.2–7.0 mm; mean, 5.1 mm) (35). The rotator cuff thickness is not related to age, gender, or symptoms (34,36–38).
|
Causes of False-Negative Diagnoses of Rotator Cuff Tears
|
|---|
Technique-related Causes of Misinterpretation
Transducer Frequency.—
US of the shoulder should be performed with a high-frequency transducer of at least 7.5 MHz. Examinations performed with a 5-MHz transducer show disappointing results (Table 4). Nowadays, high-resolution transducers (multifrequency broadband, 7.5–15 MHz) with adequate tissue penetration are available. Linear-array transducers are preferred because of their high resolution and because curved-array and mechanical sector transducers are more vulnerable to anisotropic artifacts (Table 4).
Focusing.—
Small or partial-thickness rotator cuff tears may be missed due to suboptimal focusing, which diminishes spatial resolution. Therefore, the near field should constantly be adjusted to the depth of the structure under examination.
Imaging Protocol.—
The rotator cuff tendons and especially the supraspinatus tendon are, in the neutral position of the arm, generally hidden underneath the acromion. However, for reliable sonographic evaluation, the tendons need to be totally exposed. This can be achieved with a standardized imaging protocol (Table 5) (17,53–55).
Recent technical developments such as transmit compounding, extended-field acquisition processes, and three-dimensional US acquisition and four-dimensional imaging may improve performance. However, to date no such studies are available.
Despite these advances, rotator cuff tears may be missed due to limited movements of the shoulder. This is especially the case for the supraspinatus tendon, where about 90% of rotator cuff disease is located (the anterior part) (10).
Dynamic evaluation of the rotator cuff may be helpful for identifying nonretracted full-thickness rotator cuff tears by looking for separation of the margins of a tear. It can also be helpful in overcoming anisotropic artifacts (eg, the subscapular tendon). In addition, subluxation or dislocation of the long head of the biceps tendon (56) can be diagnosed by means of external rotation of the forearm. Instability of the shoulder (57–59) may be diagnosed with dynamic examinations; however, in these cases MR imaging is frequently the imaging modality of first choice.
Transducer Handling.—
Increasing the pressure of the transducer facilitates the identification of nonretracted full-thickness rotator cuff tears. On the other hand, pressure should be eased to detect tiny fluid collections in the bicipital tendon sheath and subacromial-subdeltoid bursa (17,25). The detection of fluid in the subacromial-subdeltoid bursa or in both the joint and the subacromial-subdeltoid bursa is highly specific (96% and 99%, respectively) and has a high positive predictive value (70% and 95%, respectively) for the diagnosis of associated rotator cuff tears in symptomatic patients (25).
Anatomy-related Causes of Misinterpretation
Nondiastasis of the Ruptured Tendon Fibers.—
A recent partial- or full-thickness tear is accompanied by fluid (ie, hematoma). The surrounding fluid enhances the ultrasound signal, which is favorable for the depiction of rotator cuff tears.
In a long-standing tear, fluid may be absorbed. Partial- and full-thickness rotator cuff tears in which the ruptured tendon fibers do not recede may then be more difficult to depict (Fig 13).
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13. Nondiastasis of ruptured tendon fibers. Long-axis US scan of the supraspinatus tendon (SSP) shows a full-thickness tear (arrows) in the anterior part of the tendon insertion. Note that the torn ends of the long-standing tear lie close together. There is hardly any fluid in the tendon defect to facilitate depiction.
|
|
Posttraumatic Obscuration of Landmarks.—
Fractures can alter the bony landmarks used for orientation. Soft tissues may become hypoechoic due to contusion, edema, and tendon strain or rupture.
Disease-related Causes of Misinterpretation
Tendinosis.—
Neer (60) described three stages in the pathogenesis of the impingement syndrome. A spectrum of sonographic abnormalities of the rotator cuff tendon are likely to correlate with edema, hemorrhage, tendinosis, fibrosis, and rotator cuff tear caused by the impingement syndrome.
In tendinosis, a tendon can appear hypoechoic due to an increased amount of fluid and/or amyloid deposits in and between the tendon fibers (see the section on tendon inhomogeneity) (61). The hypoechoic appearance decreases when one scans with too much gain and increases when one uses too little gain or in combination with anisotropy (Fig 10). Tendinosis is often coexistent with partial-thickness rotator cuff tears. These may be difficult to detect when they are located in an area of tendinosis.
Calcifications.—
With long-standing impingement (ie, chronic tendinosis), calcium may be deposited in the rotator cuff tendons and/or the subacromial-subdeltoid bursa. Most patients with calcifying tendonitis are 30–50 years old, a much younger age group than those who develop rotator cuff tears.
These deposits may be uni- or multilocular and have a varying hyperechoic aspect and/or acoustic shadow (Fig 12). These calcifications appear in the critical zone of the tendon, where most tears tend to occur. Tears may be obscured by shadowing from these calcifications. Therefore, plain radiographs should be available prior to sonographic examination.
Synovial Proliferation, Granulation or Scar Tissue.—
Synovial, granulation, and scar tissue can have varying echogenicities. Bretzke et al (18) stated that focal areas of increased echogenicity are a feature of rotator cuff tear. However, most focal hyperechoic areas in the rotator cuff are due to degenerative changes of tendon fibers (fibrosis in chronic tendinosis), scar tissue, or calcium deposits (see the section on tendon inhomogeneity).
Granulation or scar tissue and intraarticular or bursal synovial tissue may fill in partial- or full-thickness rotator cuff tears, thereby impeding sonographic visualization (Fig 14).
View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14a. Proliferation of synovial tissue. (a) Arthroscopic image of the subacromial-subdeltoid bursa shows extensive proliferation of synovial tissue (arrows). (b) Short-axis US scan of the supraspinatus tendon (SSP) shows proliferating synovial tissue in the subacromial-subdeltoid bursa (*). The synovial tissue fills in several tendon defects (arrowheads), including a full-thickness rotator cuff tear (arrows).
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14b. Proliferation of synovial tissue. (a) Arthroscopic image of the subacromial-subdeltoid bursa shows extensive proliferation of synovial tissue (arrows). (b) Short-axis US scan of the supraspinatus tendon (SSP) shows proliferating synovial tissue in the subacromial-subdeltoid bursa (*). The synovial tissue fills in several tendon defects (arrowheads), including a full-thickness rotator cuff tear (arrows).
|
|
Thickened Bursa Mimicking the Rotator Cuff.—
The synovial tissue of the subacromial-subdeltoid bursa can thicken substantially with (chronic) bursitis or due to synovitis in patients with rheumatoid arthritis. These thickened bursal layers may mimic the rotator cuff (Fig 15) or fill in a partial- or full-thickness rotator cuff tear (Fig 14b). As in conventional arthrography, this is one of the causes of a false-negative finding. Compression with a transducer would be useful in making a diagnosis of a full-thickness rotator cuff tear in these cases.
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15. Extensive chronic bursitis. Short-axis US scan of the supraspinatus tendon (SSP) shows thickened synovial bursal layers (B), which mimic a thickened rotator cuff tendon (ie, tendinosis).
|
|
Massive Rotator Cuff Tear.—
Most rotator cuff tears are located in the supraspinatus tendon and may extend to the infraspinatus or subscapular tendon. With massive rotator cuff tears, the cuff cannot be visualized at US because it has completely avulsed off the greater tuberosity and retracted under the acromion. When the rotator cuff is missing, the deltoid muscle lies directly on the humeral head (Fig 16) or is separated from the humeral head by a fluid layer (Fig 17). Missing the diagnosis can be overcome by counting the layers around the humeral head (Fig 18). Normally, three layers can be recognized: the subcutaneous tissue, deltoid muscle, and rotator cuff (Fig 18). The thickness of these layers may vary due to several factors, for instance, obesity or muscle and tendon atrophy.
View larger version (75K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16. Massive rotator cuff tear. Short-axis US scan shows a full-thickness tear of the supraspinatus tendon. The deltoid muscle (D) lies directly on the humeral head (H), thus mimicking the rotator cuff. * = long head of the biceps tendon in cross section.
|
|
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17. Massive rotator cuff tear. Short-axis US scan shows a massive full-thickness tear of the supraspinatus tendon. The hypoechoic layer between the deltoid muscle (D) and humeral head (H) is intraarticular fluid (F). The deltoid muscle should not be mistaken for the rotator cuff. ST = subcutaneous tissue, * = long head of the biceps tendon in cross section.
|
|
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18a. Soft-tissue layers around the humeral head. (a) Short-axis US scan of the normal supraspinatus tendon (SSP) shows three soft-tissue layers: the subcutaneous tissue (ST), deltoid muscle (D), and rotator cuff tendon. In each sonographic section of the shoulder, these layers can be visualized around the humeral head. (b) Diagram of a short-axis view of the normal supraspinatus tendon (SSP) shows the subcutaneous tissue (ST), deltoid muscle (D), and rotator cuff tendon around the humeral head (H). B = subacromial-subdeltoid bursa, BT = biceps tendon (long head), C = cutis, HC = hyaline cartilage.
|
|
View larger version (68K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18b. Soft-tissue layers around the humeral head. (a) Short-axis US scan of the normal supraspinatus tendon (SSP) shows three soft-tissue layers: the subcutaneous tissue (ST), deltoid muscle (D), and rotator cuff tendon. In each sonographic section of the shoulder, these layers can be visualized around the humeral head. (b) Diagram of a short-axis view of the normal supraspinatus tendon (SSP) shows the subcutaneous tissue (ST), deltoid muscle (D), and rotator cuff tendon around the humeral head (H). B = subacromial-subdeltoid bursa, BT = biceps tendon (long head), C = cutis, HC = hyaline cartilage.
|
|
The outside of the subcutaneous tissue is bordered by a small hyperechoic line representing the cutis. The deltoid muscle and the rotator cuff are separated by a small hyperechoic layer representing the peribursal fat line and the subacromial-subdeltoid bursa. The small hypoechoic line on the inside of the rotator cuff following the contour of the humeral head represents the hyaline cartilage of the humeral head (Fig 18). The deltoid muscle and rotator cuff can be easily differentiated because the deltoid muscle extends over the humeral head and greater tuberosity to insert on the proximal lateral humeral shaft, whereas the normal rotator cuff tapers and inserts on the greater tuberosity.
In some cases, the space left by the missing rotator cuff may be filled in by fluid (echogenic due to debris) and/or proliferating synovial tissue mimicking the rotator cuff (Fig 19).
View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 19a. Massive full-thickness rotator cuff tears in two patients with rheumatoid arthritis. H = humeral head. (a) Short-axis US scan shows a tear of the supraspinatus tendon. Three layers are seen; the hypoechoic inner layer consists of fluid and pannus (large *). The deltoid muscle (D) could be mistaken for the rotator cuff. The hypoechoic zone in the deltoid muscle represents pannus (small *). (b) Long-axis US scan of the infraspinatus tendon (ISP). The intraarticular (*) and bursal (B) pannus could be mistaken for a rotator cuff. D = deltoid muscle, G = glenoid.
|
|
View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 19b. Massive full-thickness rotator cuff tears in two patients with rheumatoid arthritis. H = humeral head. (a) Short-axis US scan shows a tear of the supraspinatus tendon. Three layers are seen; the hypoechoic inner layer consists of fluid and pannus (large *). The deltoid muscle (D) could be mistaken for the rotator cuff. The hypoechoic zone in the deltoid muscle represents pannus (small *). (b) Long-axis US scan of the infraspinatus tendon (ISP). The intraarticular (*) and bursal (B) pannus could be mistaken for a rotator cuff. D = deltoid muscle, G = glenoid.
|
|
Patient-related Causes of Misinterpretation
Obesity or Muscularity.—
The depiction of rotator cuff tears in obese or muscular patients may be limited or insufficient. Despite the development of high-frequency transducers and especially the improvement of resolution and penetration depth, it may still be difficult to evaluate the rotator cuff of an obese or muscular patient sonographically. The fatty tissue layer or large deltoid muscle absorbs too much of the emitted high-frequency (ie, low-energy) sound waves. However, large tears can readily be diagnosed. To overcome this problem, it rarely may be necessary to use a lower-frequency transducer. Inherent to lower frequencies is a decrease in spatial resolution, which limits the reliable depiction of rotator cuff tears and therefore the accuracy of the examination (Table 4).
Limited Shoulder Motion.—
Full assessment of the rotator cuff is difficult in patients with shoulder pain and/or disability. In our experience, this problem can be overcome by physical support of the arm movements by the examiner and by moving the arm slowly and performing the examination quickly. This especially applies to the external rotation (ie, subscapular tendon) and hyper-extension–internal rotation (ie, supraspinatus tendon) positions. Alternatively, the rotator cuff can be evaluated with the arm hanging beside the body, preferably with maximal hyperextension.
|
Sonographic-Pathologic Correlation
|
|---|
The accuracy of US also depends on the accuracy of the reference standard. Terms such as fraying, fibrillation, scuffing, fringe, degeneration, or synovitis are ill-defined (by arthroscopists), frequently used terms to describe a cuff that has an irregular margin but that has no tear. There is a floating scale between these terms and the description of a small tear. Furthermore, some rotator cuff injuries may be created by the arthroscopic procedure itself, and intrasubstance tears or tears covered by thickened synovium may escape arthroscopic visualization, as synovium thickened by synovitis in the glenohumeral joint or by bursitis in the sub-acromial space inhibits inspection of the surface of the rotator cuff.
To reduce these limitations, we emphasize how important it is that one develop specific definitions and terminology with his or her own (orthopedic) surgeon(s).
|
Conclusions
|
|---|
The diagnostic accuracy of US is good and comparable with that of MR imaging in regard to identifying and measuring the size of partial- and full-thickness rotator cuff tears (4). US and MR imaging can be used as a primary modality for evaluating the rotator cuff. US is a reliable, fast, inexpensive, and for the patient easily tolerable diagnostic modality provided the examiner has a detailed knowledge of shoulder anatomy, uses a standardized examination technique, and has a thorough understanding of the potential pitfalls, limitations, and artifacts.
|
Acknowledgments
|
|---|
The authors thank James M. P. Collins, MD, for his assistance in manuscript preparation.
|
References
|
|---|
- van Holsbeeck MT, Kolowich PA, Eyler WR, et al. US depiction of partial-thickness tear of the rotator cuff. Radiology 1995;197:443–446.[Abstract/Free Full Text]
- Teefey SA, Hasan SA, Middleton WD, Patel M, Wright RW, Yamaguchi K. Ultrasonography of the rotator cuff: a comparison of ultrasonographic and arthroscopic findings in one hundred consecutive cases. J Bone Joint Surg Am 2000;82:498–504.[Abstract/Free Full Text]
- Zehetgruber H, Lang T, Wurnig C. Distinction between supraspinatus, infraspinatus and sub-scapularis tendon tears with ultrasound in 332 surgically confirmed cases. Ultrasound Med Biol 2002;28:711–717.[CrossRef][Medline]
- Teefey SA, Rubin DA, Middleton WD, Hildebolt CF, Leibold RA, Yamaguchi K. Detection and quantification of rotator cuff tears: comparison of ultrasonographic, magnetic resonance imaging, and arthroscopic findings in seventy-one consecutive cases. J Bone Joint Surg Am 2004;86-A:708–716.[Abstract/Free Full Text]
- Middleton WD, Reinus WR, Melson GL, Totty WG, Murphy WA. Pitfalls of rotator cuff sonography. AJR Am J Roentgenol 1986;146:555–560.[Abstract/Free Full Text]
- Fornage BD. The hypoechoic normal tendon: a pitfall. J Ultrasound Med 1987;6:19–22.[Abstract]
- Friedman L, Finlay K, Popowich T, Jurriaans E. Ultrasonography of the shoulder: pitfalls and variants. Can Assoc Radiol J 2002;53:22–32.[Medline]
- Crass JR, van de Vegte GL, Harkavy LA. Tendon echogenicity: ex vivo study. Radiology 1988;167: 499–501.[Abstract/Free Full Text]
- Dussik KT, Fritch DJ, Kyriazidou M, Sear RS. Measurements of articular tissues with ultrasound. Am J Phys Med 1958;37:160–165.[Medline]
- Rathbun JB, Macnab I. The microvascular pattern of the rotator cuff. J Bone Joint Surg Br 1970;52: 540–553.
- Crass JR, Craig EV, Feinberg SB. The hyperextended internal rotation view in rotator cuff ultrasound. J Clin Ultrasound 1987;15:416–420.[Medline]
- Crass JR, Craig EV, Feinberg SB. Ultrasonography of rotator cuff tears: a review of 500 diagnostic studies. J Clin Ultrasound 1988;16:313–327.[Medline]
- Ferri M, Finlay K, Popowich T, Stamp G, Schuringa P, Friedman L. Sonography of full-thickness supraspinatus tears: comparison of patient positioning technique with surgical correlation. AJR Am J Roentgenol 2005;184:180–184.[Abstract/Free Full Text]
- Harryman DT 2nd, Sidles JA, Harris SL, Matsen FA 3rd. The role of the rotator interval capsule in passive motion and stability of the shoulder. J Bone Joint Surg Am 1992;74:53–66.[Abstract/Free Full Text]
- 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]
- van Holsbeeck MT, Introcasso J. Sonography of the shoulder. In: Musculoskeletal ultrasound. 2nd ed. St Louis, Mo: Mosby, 2001; 463–516.
- Middleton WD, Reinus WR, Totty WG, Melson GL, Murphy WA. Ultrasonographic evaluation of the rotator cuff and biceps tendon. J Bone Joint Surg Am 1986;68:440–450.[Abstract/Free Full Text]
- Bretzke CA, Crass JR, Craig EV, Feinberg SB. Ultrasonography of the rotator cuff: normal and pathologic anatomy. Invest Radiol 1985;20:311–315.[CrossRef][Medline]
- Clark JM, Harryman DT 2nd. Tendons, ligaments, and capsule of the rotator cuff: gross and microscopic anatomy. J Bone Joint Surg Am 1992; 74:713–725.[Abstract/Free Full Text]
- Benjamin M, Evans EJ, Copp L. The histology of tendon attachments to bone in man. J Anat 1986; 149:89–100.[Medline]
- Vahlensieck M, Pollack M, Lang P, Grampp S, Genant HK. Two segments of the supraspinous muscle: cause of high signal intensity at MR imaging? Radiology 1993;186:449–454.[Abstract/Free Full Text]
- Turrin A, Cappello A. Sonographic anatomy of the supraspinatus tendon and adjacent structures. Skeletal Radiol 1997;26:89–93.[CrossRef][Medline]
- Brandt TD, Cardone BW, Grant TH, Post M, Weiss CA. Rotator cuff sonography: a reassessment. Radiology 1989;173:323–327.[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
Top
Abstract
Introduction
