mri of the shoulder in adults - acreditacion-fmc.org€¦ · perspective of shoulder specialists...

© 2015 AAOS Instructional Course Lectures, Volume 64 75

7

This chapter provides a concise over-view of the capacity of MRI to assist in distinguishing pathology and anatomic disruptions adjacent to the shoulder. It focuses on common adult degenerative and posttraumatic conditions, with the intent of further familiarizing the gen-eral orthopaedic surgeon and the sports medicine–oriented care provider to the perspective of shoulder specialists and musculoskeletal radiologists.

Equipment and TechniqueThe fi rst MRI of a live human, per-formed in 1977, required many hours

to acquire a single image.1 Since then, MRI has evolved into a rapid and commonly used evaluation tool; how-ever, its quality is affected by sever-al factors. Higher magnet strength (fi eld strength), de noted by Tesla (T), increases the signal-to-noise ratio; in general, a 3.0-T magnet has a higher signal-to-noise ratio than does a 1.5-T magnet. The “closed magnet,” narrow MRI machine generally produces the highest quality images, whereas the “open magnet” MRI machine allows more room for the patient at the ex-pense of reduced strength (usually 1.0

T or less) and decreased image quality. The receiver coil (the apparatus placed around the tissue of interest) contains readout channels that collect the signal; in general, the greater the number of channels, the better the image quality.

Higher-strength magnets have in-creased metallic artifact susceptibility. Metals cause poor fat suppression on T2-weighted (or fl uid-weighted) se-quences, decreasing the visibility of fl uid; thus, a short tau inversion re-covery T1 sequence (a fl uid-sensitive sequence with more uniform fat sup-pression) may be more useful. Similarly, when metallic implants are a consider-ation, image quality may benefi t from a lower strength magnet.

Routine shoulder MRI proto-cols typically consist of axial, sagittal oblique, and coronal oblique sequences. The sagittal and coronal sequences are oriented along the long axis of the scap-ula to better profi le the rotator cuff. Most protocols employ a combination of fl uid-weighted sequences (T2- or pro-ton density–weighted) and T1-weight-ed sequences. In the former, fl uid and edema are bright (hyperintense); fat is

MRI of the Shoulder in Adults

Erik W. Foss, MDStephen B. Weldon, MD

Dennis C. Crawford, MD, PhD

Dr. Crawford or an immediate family member serves as a paid consultant to or is an employee of Histogenics, Moximed, and Zimmer Biologics; serves as an unpaid consultant to Community Tissue Services; and has received research or institutional support from Community Tissue Services, Histogenics, Moximed, and Zimmer Biologics. Neither of the following authors nor any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this chapter: Dr. Foss and Dr. Weldon.

AbstractMRI can be used to assist in distinguishing patholog y and anatomic disruptions adjacent to shoulder articulations and is a crucial tool in evaluating the symptomatic shoulder. The differentiation of traumatic and degenerative etiologies is important for patient management. A review of common adult degenerative and posttraumatic conditions can help familiarize general orthopaedic surgeons and sports medicine care providers to the perspective of shoulder specialists and musculoskeletal radiologists.

Instr Course Lect 2015;64:75–86.

Orthopaedic Medicine and Practice

76 © 2015 AAOS Instructional Course Lectures, Volume 64

natively bright, although fat suppression is often used to turn fat and bone mar-row dark (hypointense) to increase le-sion visibility. On T1-weighted images, fat and bone marrow are bright, unless fat suppression is used. If arthrographic or intravenous gadolinium contrast is administered, T1-weighted sequences are used, whereas gadolinium is bright.

Normal and Abnormal Appearances of Soft TissuesNormal ligaments, tendons, capsules, and labrum should be hypointense on all sequences (Figure 1). As a result of edema and fl uid, traumatic soft-tissue injuries manifest as increased signal on T2-weighted sequences. Sprained tissues show intrasubstance, inter-mediate-to-hyperintense signal, and ill- defi ned fi bers but not fi ber discon-tinuity. Tear defects demonstrate bright fl uid. Chronic, atraumatic mucoid (myxoid) degeneration may confound a sprain diagnosis but should be amor-phous, nonlinear, and intermediate in signal intensity. Such degeneration is

often seen in the rotator cuff tendons and the labrum.

Acromion and Acromioclavicular JointDegenerative joint disease is common in the adult acromioclavicular (AC) joint and is often reported. However, one study found up to 82% of asymp-tomatic patients had MRI abnormali-ties in the AC joint.2,3 Distinguishing incidental degenerative joint disease from relevant disease requires clin-ical correlation. The joint should be smoothly marginated (Figure 1, A). Abnormal fi ndings include osteo-phytes, bone marrow signal changes, and (rarely) intra-articular abnormal-ities. Capsular distention is suggested as an important factor associated with symptomatic AC joints.4 Degenerative osteophytes at either the distal clavicle or the articulating acromion can result in roof impingement on the supraspi-natus or infraspinatus tendons, leading to tendinopathy and tearing. Static im-pingement is identifi ed on sagittal and

coronal images as inferiorly projecting spurs indenting the superior tendon surface (Figures 2 and 3).

Acromioclavicular (AC) joint and rotator cuff musculature. A, Coronal proton density–weighted image of a normal AC joint (arrow). The articular margins are smooth, without spurring or impingement of the underlying bright fat plane (arrowheads) and the supraspinatus myotendinous junction (black arrow). B, Sagittal T1-weighted image of a type 2 acromion (arrow), which parallels the humeral head contour (arrowheads). Acromial type is assessed on MRI on the two lateralmost sagittal images. C, Sagittal T1-weighted image of normal rotator cuff muscles. The subscapularis (SU), supraspinatus (SS), infraspinatus (IS), and teres minor (TM) muscles all show uniform intermediate-intense signal, without fatty (hyperintense) interdigitation; the SS muscle extends above the tangent line (dashed line). D, Sagittal T1-weighted image of muscle atrophy. The supraspinatus muscle (black arrow) fails to extend above the tangent line, and there also is fatty degeneration of the infraspinatus muscle (white arrow), which appears as interdigitation of bright streaks of fatty tissue, and overall muscle atrophy.

Figure 1

Coronal fat-suppressed T2-weighted image of degenera-tive joint disease (DJD). There is diffuse, severe cartilage thinning and denudation with degenerative subchondral cysts (white arrow), a subchondral bone marrow edema pattern (curved arrow), and osteophytes (arrowheads). A partial-thickness, bursal-sided supraspinatus tendon tear (black arrow) and acromioclavicular DJD also are present.

Figure 2

MRI of the Shoulder in Adults Chapter 7


Four types of acromial shapes have been described and variably associated with rotator cuff pathology.5,6 Type 2, which is the most common, includes a smooth lateral undersurface that par-allels the humeral head (Figure 1, B). Type 3 (hooked) describes downward anterior angulation. These subtypes can be easily evaluated on radio-graphs; however, MRI focused on the lateralmost sagittal acromial images is nearly as reliable.7

Rotator CuffThe rotator cuff tendons should be homogeneously hypointense on all MRI sequences (Figure 4). The tendons are well visualized on sagittal oblique se-quences. The supraspinatus and infra-spinatus tendons also can be evaluated on coronal oblique sequences, whereas

the subscapularis tendon is best seen on axial images. Normal cuff muscu-lature is evaluated on sagittal oblique sequences, with uniform intermedi-ate signal throughout; sagittal oblique T1-weighted images best reveal atrophy and more advanced fatty degeneration.

The spectrum of rotator cuff pathol-ogy ranges from intrinsic tendinopathy to partial- and full-thickness tears that can all appear as areas of intermedi-ate- or fl uid-intense signal within the tendon on fl uid-weighted sequences. Distinguishing these entities requires further evaluation.

Tendinopathy, which is most appar-ent on fl uid-weighted sequences, can represent acute tendinitis, a contusion, or degenerative mucoid tendinosis.8 It appears as ill-defi ned, focal or diffuse intermediate signal intensity in the ten-don (Figure 3). However, the articular and bursal tendon surfaces should re-main smooth and intact.

Conventional MRI is relatively accurate for diagnosing rotator cuff tears, with 80% sensitivity and 95% specifi city for partial-thickness tears and 91% sensitivity and 97% specifi city for full-thickness tears.9 Tendon tears may involve the articular surface or the bursal surface (a partial-thickness tear), both surfaces (a full-thickness tear) or be confi ned to the interstitium (a mid-substance tear; Figures 5 and 6). A tear is diagnosed as a discrete, fl uid-intense abnormality across all or part of a ten-don; it is usually best appreciated on fat-suppressed, T2-weighted sequences. Tears are frequent within 1 cm of the insertion (critical zone) or at the humer-al head insertion. Associated adjacent tendinopathy is common. Partial- thickness tears represent approximately 20% to 25% of cases, with most occur-ring along the articular supraspinatus

footprint.10 Articular-sided footprint tears may be referred to as partial ar-ticular-sided supraspinatus tendon avulsions (PASTAs), which commonly occur in those younger than 40 years.10,11

In the setting of a suspected tear, close inspection at the far-anterior margin (high stress zone) of the tendon is essen-tial;12 the anterior supraspinatus tendon fi bers should extend to the tuberosity at the posterior biceps interval. Distin-guishing small partial-thickness tears from surrounding tendinopathy can be diffi cult. Full-thickness tears may demonstrate only pinpoint extension to one surface and, if oblique in nature, may be mistaken for partial-thickness tears. Larger, full-thickness tears may demonstrate retraction, which is seen as medial and posterior displacement of the tendon and myotendinous junction; the normal junction is located slightly lateral to the AC joint.13

Interstitial tears most commonly oc-cur adjacent to the humeral insertion footprint. A focal, fl uid-intense signal within the tendon, with the articular and bursal surfaces intact, suggests an interstitial-type, partial-thickness tear.

The differentiation of acute trau-matic injuries from chronic degenera-tion and attrition is important for both clinical and surgical management. A de-generated rotator cuff will demonstrate thinning of one or more tendons and may be accompanied by muscle atrophy and fatty degeneration. In contrast, in a traumatic rotator cuff injury, abnor-mal signal will be localized to a discrete portion of the rotator cuff tendon(s), but the rotator cuff will maintain its normal bulk, and the tendon may even be thickened.

Depending on severity, atrophy and fatty degeneration of the rotator cuff muscles suggest long-standing disease

Coronal fat-suppressed T2-weighted image of supraspina-tus tendinopathy, acromioclavicular (AC) degenerative joint disease, and superior labral degenera-tion. The supraspinatus tendon is abnormally thickened and diffusely increased in signal (arrowheads), but no focal, fl uid-intense defect is present. Degenerative osteo-phytes of the AC joint indent the myotendinous junction (arrow). The posterior-superior labrum (curved arrow) demonstrates diffuse intermediate signal compatible with degeneration, without a focal, fl uid-intense tear.

Figure 3



and infl uence clinical management.14,15

Grading is best accomplished on sag-ittal T1-weighted images. Supraspina-tus muscle bulk is assessed by using

the tangent sign: the muscle should extend above a line drawn tangential-ly between the coracoid process and scapular spine16 (Figure 1, C). Fatty

Normal rotator cuff and labrum. A, Coronal post-magnetic resonance arthrogram fat-suppressed T2- weighted image. The superior labrum (curved arrow) is homogeneously dark, triangular, and well defi ned; the sublabral sulcus (arrow), a normal variant, is thin, well defi ned, and parallels the glenoid contour. The supraspinatus tendon (arrowheads) is uniformly hypointense, and the articular and bursal surfaces are well defi ned and smooth all the way to the insertional footprint (dashed line). B, Sagittal fat-suppressed T2-weighted image of the subscapularis (SU), the supraspinatus (SS), the infraspinatus (IS), the teres minor (TM) muscles, and the intra-articular long head of the biceps tendon (LHBT; arrowhead). The tendon tissue is homogeneously dark, with normal surrounding intermediate-intense muscle. The LHBT courses above the anterosuperior humeral head and below the superior glenohumeral ligament–coracohumeral ligament complex, is homogeneously dark, and is normal in thickness. C, Axial fat-suppressed T2-weighted image. The SU and IS tendons are well defi ned and dark, with smooth articular and bursal margins. The anterior labrum (curved arrow) and the posterior labrum (thick arrow) are dark and triangular in shape and blend imper-ceptibly with the adjacent intermediate-intense glenoid hyaline cartilage (arrowheads).

Figure 4

Coronal fat-suppressed T2-weighted image of a distal articular-sided partial-thickness tear of the supraspinatus tendon (a partial articular-sided supraspinatus tendon avulsion). The undersurface of the distal supraspinatus tendon shows a partial-thickness, fl uid- intense defect involving 50% of the tendon thickness (arrow). The bursal surface fi bers are intact. Subtle intermediate-intense signal in the intact distal tendon (curved arrow) suggests mild tendinopathy.

Figure 5

A, Coronal fat-suppressed T2-weighted image of a full-thickness supraspinatus tendon tear. There is a fl uid-intense gap in the distal tendon that extends through both the bursal and articular surfaces (arrow). Mild thick-ening and increased signal within the residual fi bers (arrowheads) is compat-ible with tendinopathy and possible partial tearing. B, Coronal fat-suppressed T2-weighted image of a chronic, retracted, full-thickness supraspinatus tendon tear. The humeral head is uncovered and high riding, nearly contacting the acromion. The supraspinatus tendon is retracted to the level of the glenoid (arrow). Fluid is seen within the subacromial/subdeltoid bursa (arrowheads).

Figure 6



degeneration of the muscles is evaluated on sagittal T1-weighted images with the Goutallier grading system, which has been adapted for MRI.17 Normal muscle has no fatty (bright) streaks through the muscle, whereas fatty degeneration will show streaks of fat (Figure 1, D).

Assessing a repaired rotator cuff for a retear is challenging because shaver debris, anchor artifact, and progressive tissue remodeling may all obscure the certainty of diagnosis. Granulation tissue and fi brovascular scars have intermediate and/or bright signal on fl uid-weighted sequences, which are often diffi cult to differentiate from a retear. Spielmann et al18 reported that only 10% of postoperative patients had normal (dark) signal intensity in the re-paired rotator cuff, whereas 53% had mildly increased signal intensity, and 20% had foci of frank fl uid intensity. Postoperative MRIs of the rotator cuff should therefore be interpreted with caution, and a retear diagnosis should be limited to a symptomatic patient with a discrete focus of fl uid-intense signal. MRI arthrography may be useful in the setting of prior rotator cuff repair or periarticular hardware.

Long Head of the Biceps TendonThe long head of the biceps tendon (LHBT) is normally homogeneously dark, and T2-weighted sequences provide the best visualizations. The biceps-labral anchor is most appar-ent on coronal sequences, whereas an intra-articular tendon is seen best with sagittal imaging (Figure 7). The extra-articular tendon and sheath are best assessed axially (Figure 8).

At the upper margin of the bicipi-tal groove, the subscapularis tendon, the coracohumeral ligament, and the

superior glenohumeral ligament bor-der and secure the biceps in the groove. Tendon displacement suggests injury to one or all of these structures.19,20 LHBT position is assessed with axial imaging. The tendon should be located lateral to the lesser tuberosity within the bic-ipital groove. When it is superfi cial to the subscapularis tendon, the coracohu-meral ligament and the superior gleno-humeral ligament are torn; when it is in the glenohumeral joint, a full-thickness subscapularis tear is present (Figure 9).

A complete tendon avulsion from the labral anchor can be diagnosed on coronal imaging via nonvisualization of the tendon at the superior labrum. Often, the tendon will retract into the upper arm and also may be absent from the bicipital groove.

Tendinopathy typically shows ten-don thickening and intermediate sig-nal intensity, often accompanied by an increased amount of fl uid in the ten-don sheath, commonly near the biceps interval where adjacent rotator cuff

tendon tears and associated bicipital groove instability may irritate the ten-don (Figure 10).

LabrumThe labrum should be homogeneously dark on all sequences (Figure 4, A and C and Figure 8). It is normally triangu-lar in shape, although a slightly blunted free edge can be a normal fi nding.21 The labrum should blend imperceptibly into adjacent hypointense cortical bone and intermediate-intensity hyaline cartilage. The joint capsule and the inferior gleno-humeral ligament may attach directly to the labrum, the superfi cial labral base, or the paralabral glenoid.

The labrum has several variants that can be mistaken for disease or in-jury.21,22 The anterosuperior labrum may be diminutive in size to absent (with a thickened middle glenohumeral liga-ment known as the Buford complex; Figure 11) but should reconstitute at

Sagittal, post-mag-netic resonance arthrography, fat- suppressed T1-weighted image shows the normal intra-articular long head of the biceps tendon (LHBT). The LHBT (arrow) is oval, dark, and normal in size, and it courses along the anterosuperior margin of the humeral head. Con-trast (bright signal) outlines the joint capsule.

Figure 7 Axial, post-magnetic resonance arthrography, fat-sup-pressed T1-weighted image of the normal anterior and posterior- inferior labrum. Contrast material (bright signal) outlines a smooth, homogeneously dark, and triangular anterior (white arrow) and posterior (black arrow) labrum, which blend imperceptibly with the adjacent intermediate-intense hyaline car-tilage. A normal long head of the biceps tendon lies in the bicipital groove (arrowhead).

Figure 8



the midanterior glenoid. It is either attached to the underlying hyaline cartilage or separated by a sublabral foramen; however, the anterior labrum should be fi rmly attached below the 3-o’clock position.23 Superiorly, the

labrum may attach to the underlying glenoid or hyaline cartilage (cartilage undercutting) or be separated from both by a thin, fl uid-fi lled sulcus that parallels the glenoid contour (known as the sublabral sulcus; Figure 4, A). Distinguishing a normal sulcus from

chondrolabral separation can be diffi -cult, but a sulcus is usually thin (< 2 to 3 mm wide) and parallels the glenoid.23-25

The pathology of the superior la-brum usually falls into two categories: traumatic superior labrum anterior to posterior (SLAP) tears and degener-ation. Traumatic tears appear as hy-perintense signal on fl uid-weighted sequences; are morphologically lin-ear; and extend to the labral surface, coursing away from the glenoid. Labral degeneration is amorphous and inter-mediate in signal intensity (Figure 3).

Given labral variability, it is recom-mended to initiate a search for superior labral tears at the biceps-labral anchor and the posterior-superior labrum on coronal images; axial images comple-ment this search. With tear suspicion, the morphology should be determined to differentiate a stable lesion from an unstable lesion.

Currently, 10 types of SLAP tears have been described.26 Type I is char-acterized by irregularity and fraying at the biceps-labral complex and is of-ten an incidental fi nding or related to degeneration (Figure 12, A). Type II tears are most common and demon-strate linear hyperintense signal at the biceps-labral anchor extending into the posterior- superior labrum (Fig-ure 12, B). Type III tears are buck-et-handle tears of the superior labrum in which the central portion of the la-brum can displace into the joint, but the biceps tendon is spared. These tears are most apparent on coronal images as linear hyperintensities coursing across the labrum from one portion of the articular surface to another. Type IV injuries extend through the anchor into the biceps tendon and are seen on coro-nal and sagittal images (Figure 12, C). The remaining types are uncommon,

Sagittal, fat-suppressed T2-weighted image of long head of the biceps tendon (LHBT) tendinopathy. The tendon (arrowheads) is abnormally thickened and increased in signal. Tearing of the subscapularis (arrow) and supraspinatus tendons (curved arrow) likely traverses the biceps interval; LHBT pathology is often seen with such tears.

Figure 10

Axial, post-magnetic resonance arthrography, fat-suppressed T2-weighted image shows an anterosuperior labral variant. The anterosuperior labrum (thin arrow) is small and separated from the underlying glenoid by a sublabral foramen, as evidenced by intervening bright fl uid. There is associated thickening of the middle glenohumeral ligament (arrowhead).

Figure 11

A, Axial fat-suppressed T2-weighted image of a long head of the biceps tendon (LHBT) dislocation. The LHBT (black arrow) lies medial to the bicipital groove and slightly anterior to the lesser tuberosity, with overlying soft-tissue edema. There is a full-thickness tear of the subscapularis tendon (curved arrow), along with tendon thickening. Tears of the superior gleno-humeral ligament and the coracohumeral ligament are likely present, given the anterior displacement of the LHBT relative to the lesser tuberosity. B, Axial fat-suppressed T2-weighted image of intra-articular dislocation of the LHBT related to a full-thickness subscapularis tendon tear. The subscapularis tendon (curved arrow) is retracted proximally. The LHBT (black arrow) lies anterior to the glenohumeral joint. The bicipital groove (arrowhead) is empty.

Figure 9



unstable variants of type II injuries ex-tending into adjacent contiguous struc-tures (for example, the adjacent labrum or the middle glenohumeral ligament).

Lesions of the anterior-inferior la-brum are usually divided into bony Ban-kart (osteochondral) fractures, Bankart lesions, and Bankart-like lesions. Axial MRIs are most helpful for visualizing linear, fl uid-intense signal abnormali-ties coursing into the labrum itself or through the chondrolabral junction.26

Bankart fractures (Figure 13, A) al-most universally include an articular component and should therefore be considered osteochondral fractures; axial T1-weighted images optimize cortex-medullary bone differentiation and help demonstrate the osseous com-ponent, whereas fl uid-weighted images help reveal cartilage involvement. A Bankart lesion is diagnosed if a labral tear extends anteromedially beyond the margin of the scapula through the scapular periosteum, whereas a Bankart variant-type tear spares the periosteum,

which is normally an uninterrupted, thin, linear, hypointense structure extending from the anterior scapular margin to the labral base.

A Hill-Sachs impaction fracture of the humeral head may be a helpful sec-ondary sign of dislocation, raising sen-sitivity for detecting a Bankart injury. In

an anterior dislocation, it is found on ax-ial images through the superior margin of the humeral head as a posterolateral articular surface concavity (Figure 13, B). Engaging lesions are associated with a more horizontally oriented defect that may require multiple sequences to ap-preciate.27,28 Posterior dislocation may

Superior labrum anterior to posterior (SLAP) tears. A, Coronal fat-suppressed T2-weighted image of superior labral degeneration (type I SLAP tear). The superior labrum (arrow) shows amorphous intermediate signal without linear extension to the surface. B, Coronal fat-suppressed T2-weighted image of a type II SLAP tear. Linear fl uid extends into the base of the posterior-superior labrum (arrow); no extension into the biceps tendon or the remainder of the labrum was seen on other images. Although this tear parallels the glenoid, it should not be confused with a sublabral sulcus because dark labral tissue is seen beneath this tear. Note also the intermediate-intense cartilage undercutting (arrowhead), a normal variant, interposed between the glenoid and the labral base. C, Coronal fat-suppressed T2-weighted image of a type IV SLAP tear. The biceps-labral anchor (arrow) demonstrates abnormal increased signal that was contiguous with a posterior-superior labral tear on sequential images; the proximal long head of the biceps tendon (arrowhead) shows linear hyperintensity that was contiguous with the labral tear on other images.

Figure 12

A, Axial fat-suppressed T2-weighted image of a Bankart fracture. There is a focal defect of the anterior-inferior glenoid (arrow) with an adjacent triangular osseous fragment (arrowhead); fl uid-intense signal separates the two and extends anteriorly into the soft tissues, confi rming a tear of the scapular periosteum. B, Axial T1-weighted image of a Hill-Sachs impaction fracture. The concavity along the posterolateral aspect of the upper humeral head represents the impaction deformity (arrow); underlying marrow hypointensity corresponds to marrow edema, which is dark on T1-weighted images.

Figure 13



produce a similar morphology (reverse Hill-Sachs lesion) along the anterior surface of the humeral head.

Diagnosis of an anterior-inferior labral tear may be assisted by the ABER (abduction-external rotation) sequence (Figure 14). To obtain the ABER se-quence, the ipsilateral palm is placed onto the occiput, forward fl exing the humerus and placing stress on the anteroinferior labroligamentous junc-tion. The shoulder is then imaged in an oblique plane. Although the plane of orientation can be confusing, one lo-cates the anterior aspect of the joint (ref-erencing the coracoid process) where the capsule appears most taut; this is the anteroinferior labroligamentous junction29 (Figure 15).

Magnetic Resonance ArthrographyMagnetic resonance arthrography (MRA) can be accomplished in two ways. Direct MRA (dMRA) is the tra-ditional and common technique, where diluted gadolinium contrast material is injected directly into the glenohumeral joint. Bright gadolinium signal outlines the labrum, the capsule, and the articu-lar surface of the rotator cuff, revealing any tears. Indirect MRA (iMRA) is a newer technique, where gadolinium is injected intravenously. After about 15 minutes, there is evidence of gadolin-ium passively diffusing into the joint, providing arthrographic contrast; passive or active exercise for 10 to 15 minutes may improve diagnostic accu-racy.30,31 The former technique (dMRA) is dependable, minimally invasive, and well-studied, and it has a minimal risk of postprocedure synovitis. The latter technique (iMRA) is well tolerated but shows enhancing vascularity in the la-brum and the rotator cuff, which can be

mistaken for a tear.32 Of note, although commonly used in practice and general-ly accepted as routine, the FDA has not approved gadolinium contrast material for MRA, so its use must be considered off-label for such applications.

For assessing rotator cuff pathology in patients with no history of rotator cuff surgery, conventional MRI is the usual test of choice. Although dMRA is more sensitive and slightly more specifi c for detecting partial-thickness tears (86% sensitive and 96% specifi c compared with 64% and 92% for con-ventional MRI), it has little advantage over conventional MRI for diagnosing all rotator cuff tears.33 In postoperative patients, dMRA is up to 87% accurate for all rotator cuff retears and may be benefi cial in diagnosing full-thick-ness retears, but it is only moderate-ly diagnostic for partial-thickness tears.34,35 Contrast extravasation into the subacromial/subdeltoid bursa may

suggest a full-thickness cuff but also can result from an incidental capsular defect or an injection error.

MRA is most commonly used to evaluate labral tears. Contrast material will extend into most tears, being vi-sualized as a bright linear signal in the labrum. However, 3.0-T conventional MRI is up to 86% to 90% sensitive and 100% specifi c for labral tears, similar to dMRA.36-38 Thus, in patients with no history of labral surgery, conventional MRI may suffi ce for diagnosis. For pa-tients with prior labral surgery, dMRA may still be benefi cial. The repaired labrum tends to be heterogeneous in signal as a result of proton density– and T2-hyperintense granulation and fi brovascular tissue, potentially being confused with a tear, and anchor and suture artifacts may obscure small re-tears. dMRA has shown up to 92% ac-curacy for diagnosing retears compared with 83% for conventional MRI.35,39

Abduction-external rotation, fat-suppressed T1-weighted MRI of a posttraumatic cartilage lesion associated with a labral tear. There is a full-thickness, sharply marginated cartilage defect of the anterior-inferior glenoid (arrowhead) communicating with a tear of the anterior-inferior labral base (arrow). This is often termed a glenolabral articular disruption.

Figure 14

Abduction-external rotation, post-magnetic resonance arthrography, fat-suppressed T1-weighted image shows an anterior-inferior labral tear. There is linear, gadolinium-intense (bright) signal extending into the labrum (arrow). No extension into the anterior soft tissues is seen to suggest a tear of the scapular periosteum. The anterior-inferior portion of the labrum can be deduced from the taut joint capsule (arrowheads).

Figure 15



Periarticular CystsCysts around the glenohumeral joint are common and may be incidental or secondary signs of pathology. Differ-entiating between the two may assist in fi nding additional pathology and/or associated injury.

Subcortical cysts are frequently seen in the humeral head. Cysts in the poste-rior half of the greater tuberosity show no statistical correlation with rotator cuff disease.40 However, cysts of the anterolateral humeral head (deep to the supraspinatus insertion) are highly specifi c for rotator cuff tears.40 Cysts within and immediately superior to the lesser tuberosity are highly correlated with supraspinatus and subscapularis tendon tears.41

Cysts deep to the glenoid and humer-al head articular cartilage are important to recognize. These are almost invari-ably secondary to degenerative joint dis-ease, presumably refl ecting a ball-valve phenomenon related to intra-articular fl uid, and assist the viewer in recogniz-ing and characterizing other degener-ative fi ndings.

Cysts around the labrum are use-ful clues for diagnosing labral tears. A nearly 100% incidence of adjacent labral tears in the setting of a paralabral cyst has been reported42 (Figure 16, A). Paralabral cysts vary in size but can be large enough to cause mass effect on adjacent structures and remodeling of adjacent bone. When located in the spinoglenoid notch, they can impinge the traversing suprascapular nerve and produce infraspinatus muscle denerva-tion edema and/or atrophy.

Capsule, Synovium, and BursaMultidirectional instability in the shoulder can be associated with a pat-ulous capsule. Several studies have

investigated MRI diagnosis, with lim-ited success.43,44 Most use complex, time-consuming measurements that lack practical application. MRA is usu-ally needed to demonstrate the potential degree of joint distention. However, it should be emphasized that the concept of a patulous capsule is neither specifi c nor diagnostic.

Traumatic injuries to the capsule may include a tear or an avulsion of the inferior glenohumeral ligament (Figure 16, B). The normal inferior glenohumeral ligament is seen in all three imaging planes as a hypointense, U-shaped band connecting the inferior labrum or the paralabral glenoid to the humerus; the anterior and posterior

A, Axial fat-suppressed T2-weighted image of a paralabral cyst secondary to a labral tear. A linear, fl uid-intense tear traverses the posterior labrum (arrow). A fl uid-intense paralabral cyst (arrowhead) extends from the tear into the spinoglenoid notch. B, Coronal, post-magnetic resonance arthrography, fat-suppressed T1-weighted image of the normal anterior band of the inferior glenohumeral ligament. The inferior glenohumeral ligament is a focal thickening of the axillary recess capsule and can be inferred at the anterior and posterior margins of the recess. The inferior glenohumeral ligament (arrow) should form a U shape, extending from the labrum/perilabral glenoid to the humerus. C, Coronal fat-suppressed T2-weighted image of a humeral avulsion of the glenohumeral ligament lesion. The anterior band of the inferior glenohumeral ligament is torn, failing to extend to the humerus, forming the J sign (arrow). There is a large amount of posttraumatic soft-tissue edema. There also is a large full-thickness tear of the supraspinatus tendon (arrowheads). D, Coronal fat-suppressed T2-weighted image of axillary recess synovitis. The capsule is thickened and shows a small amount of adjacent soft-tissue edema (arrowhead). There is mildly heterogeneous signal in the axillary recess compatible with thickened synovium (arrow).

Figure 16



bands of the ligament lie at the ante-rior and posterior margins of the ax-illary capsule. A humeral avulsion of the glenohumeral ligament, known as a HAGL lesion (Figure 16, C), is di-agnosed when the anterior band of the inferior glenohumeral ligament fails to insert on the humeral head, forming the appearance of the letter J instead of a U (the J sign), which is best visualized on coronal sequences.45 Acutely, it will be surrounded by soft-tissue edema. A variant of this injury includes bony humeral avulsion of the glenohumeral ligament, which will demonstrate an osseous avulsion from the humerus. A posterior humeral avulsion of the gleno-humeral ligament, known as a PHAGL lesion, is a bony humeral avulsion of the glenohumeral ligament lesion of the posterior band of the inferior gleno-humeral ligament.46

The glenohumeral synovium is normally thin and imperceptible on MRI. When infl amed, the thickened synovium may be visualized. Synovitis (Figure 16, D) is relatively hyperintense on fl uid-weighted sequences and may

be diffi cult to distinguish from a joint effusion but tends to be more hetero-geneous in signal. Intravenous contrast may help in the diagnosis because the thickened synovium will enhance.

The subacromial/subdeltoid bursa is normally decompressed, but when distended by fl uid or synovitis, it will appear as a curvilinear, fl uid-intense structure interposed between the ro-tator cuff, the acromion, and the del-toid muscle. Trace physiologic fl uid may be present but usually lies lateral to the AC joint and is less than 2 mm in thickness.47 Bursal fl uid is often seen in rotator cuff tears (Figure 6, B) but can be seen in other processes, such as tendinitis, rheumatoid arthritis, gout, and infection.

Bursitis can be diffi cult to differ-entiate from bursal fl uid because both are hyperintense on fl uid-weighted se-quences; however, bursitis typically ap-pears slightly more heterogeneous and less bright than fl uid. If intervenous gadolinium contrast is given, the walls of an infl amed bursa will show thicker wall enhancement (usually greater than

1 mm), whereas a noninfl amed bursa will show only very thin uniform wall enhancement.

Cartilage and Degenerative Joint DiseaseVisualization of glenohumeral articular cartilage disease is challenging because of the relatively thin cartilage depth, the limited spatial resolution of routine MRI sequences, and blurring between the cartilage base and the subchondral bone plate. Fat-suppressed T2- weighted sequences are usually used for evalu-ation (Figure 17). The normally in-termediate-intense cartilage signal is contrasted superfi cially with the hyper-intense joint fl uid; fi ssures, delaminat-ing tears, and ulcerations fi ll with fl uid and appear as bright defects. Studies of conventional MRI show variable di-agnostic accuracy ranging from 65% to 96%.48,49 One study found MRA to have only a slightly increased diagnostic accuracy over conventional MRI (86% versus 84%).50

MRI evaluation of cartilage has made tremendous progress in the past decade, but glenohumeral cartilage MRI research is still in its early stages. The use of specialized fl uid-weighted, three-dimensional gradient-recalled echo sequences can increase spatial resolution and potentially raise the sen-sitivity for detecting partial- thickness lesions.49 However, the drawbacks include additional scan time and in-creased metallic susceptibility artifact.

Molecular imaging of hyaline car-tilage is a rapidly evolving area of re-search focused on identifying early degeneration and characterizing dis-ease progression. It seeks to reveal regional variations in the cartilage molecular matrix (such as glycosami-noglycans and water), which may be

A, Coronal, post-magnetic resonance arthrography, fat-suppressed T2-weighted image of normal glenohumeral cartilage. Normal cartilage is intermediate in signal intensity with a smooth surface (arrowheads). B, Coronal, fat-suppressed T2-weighted image of focal full-thickness cartilage disease and underlying marrow change. There is a cartilage defect (thick arrow) that is full-thickness (extending to the dark subchondral bone plate). An underlying subchondral cyst (thin arrow) suggests a chronic process.

Figure 17



altered in degeneration. T2-mapping and delayed gadolinium-enhanced MRI of cartilage are two techniques that have been applied to the gleno-humeral joint; although still in early investigations, normal index values are being elucidated to help differen-tiate normal cartilage from early de-generation.51 Other advanced imaging methods, such as sodium and T1-rho imaging, have not yet been extensively investigated in the shoulder.

SummaryMRI is a crucial tool in evaluating the symptomatic shoulder. Recent advances in MRI technology have enhanced di-agnostic accuracy. Differentiation of traumatic and degenerative etiologies is important for patient management and a thorough understanding of optimal sequences and imaging planes for iden-tifying individual shoulder pathologies is crucial to an accurate diagnosis. Al-though an off-label use of arthrography, the judicious use of gadolinium contrast can enhance visualization for selected pathologies.

References 1. Geva T: Magnetic resonance imaging:

Historical perspective. J Cardiovasc Magn Reson 2006;8(4):573-580.

2. Shubin Stein BE, Ahmad CS, Pfaff CH, Bigliani LU, Levine WN: A comparison of magnetic resonance imaging fi ndings of the acromiocla-vicular joint in symptomatic versus asymptomatic patients. J Shoulder Elbow Surg 2006;15(1):56-59.

3. Stein BE, Wiater JM, Pfaff HC, Bigliani LU, Levine WN: Detection of acromioclavicular joint pathology in asymptomatic shoulders with magnet-ic resonance imaging. J Shoulder Elbow Surg 2001;10(3):204-208.

4. Choo HJ, Lee SJ, Kim JH, et al: Can symptomatic acromioclavicular joints be differentiated from

asymptomatic acromioclavicular joints on 3-T MR imaging? Eur J Radiol 2013;82(4):e184-e191.

5. Bigliani LU, Morrison DS, April EW: The morphology of the acromion and its relationship to rotator cuff tears. Orthop Trans 1986;10: 228.

6. Vanarthos WJ, Monu JU: Type 4 ac-romion: A new classifi cation. Contemp Orthop 1995;30(3):227-229.

7. Mayerhoefer ME, Breitenseher MJ, Roposch A, Treitl C, Wurnig C: Comparison of MRI and conven-tional radiography for assessment of acromial shape. AJR Am J Roentgenol 2005;184(2):671-675.

8. Opsha O, Malik A, Baltazar R, et al: MRI of the rotator cuff and internal derangement. Eur J Radiol 2008;68(1):36-56.

9. Smith TO, Daniell H, Geere JA, Toms AP, Hing CB: The diagnostic accuracy of MRI for the detection of partial- and full-thickness rotator cuff tears in adults. Magn Reson Imaging 2012;30(3):336-346.

10. Schaeffeler C, Mueller D, Kirchhoff C, Wolf P, Rummeny EJ, Woertler K: Tears at the rotator cuff footprint: Prevalence and imaging characteristics in 305 MR arthrograms of the shoul-der. Eur Radiol 2011;21(7):1477-1484.

11. Tuite MJ, Turnbull JR, Orwin JF: Anterior versus posterior, and rim-rent rotator cuff tears: Prevalence and MR sensitivity. Skeletal Radiol 1998;27(5):237-243.

12. Mesiha MM, Derwin KA, Sibole SC, Erdemir A, McCarron JA: The biomechanical relevance of anterior rotator cuff cable tears in a cadaveric shoulder model. J Bone Joint Surg Am 2013;95(20):1817-1824.

13. Kaplan PA, Dussault R, Helms CA, Anderson MW: Musculoskeletal MRI. Philadelphia, PA, WB Saunders Com-pany, 2001, p 186.

14. Melis B, DeFranco MJ, Chuinard C, Walch G: Natural history of fatty infi ltration and atrophy of the supraspinatus muscle in rota-tor cuff tears. Clin Orthop Relat Res 2010;468(6):1498-1505.

15. Melis B, Wall B, Walch G: Natu-ral history of infraspinatus fatty

infi ltration in rotator cuff tears. J Shoulder Elbow Surg 2010;19(5):757-763.

16. Zanetti M, Gerber C, Hodler J: Quan-titative assessment of the muscles of the rotator cuff with magnetic resonance imaging. Invest Radiol 1998;33(3):163-170.

17. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC: Fatty muscle de-generation in cuff ruptures: Pre- and postoperative evaluation by CT scan. Clin Orthop Relat Res 1994;304:78-83.

18. Spielmann AL, Forster BB, Kokan P, Hawkins RH, Janzen DL: Shoulder after rotator cuff repair: MR imaging fi ndings in asymptomatic individ-uals: Initial experience. Radiolog y 1999;213(3):705-708.

19. Beall DP, Williamson EE, Ly JQ, et al: Association of biceps tendon tears with rotator cuff abnormalities: Degree of correlation with tears of the anterior and superior portions of the rotator cuff. AJR Am J Roentgenol 2003;180(3):633-639.

20. Khazzam M, George MS, Churchill RS, Kuhn JE: Disorders of the long head of biceps tendon. J Shoulder Elbow Surg 2012;21(1):136-145.

21. De Maeseneer M, Van Roy F, Lenchik L, et al: CT and MR arthrography of the normal and pathologic antero-superior labrum and labral-bicipital complex. Radiographics 2000;20(Spec No):S67-S81.

22. Rudez J, Zanetti M: Normal anatomy, variants and pitfalls on shoulder MRI. Eur J Radiol 2008;68(1):25-35.

23. Tuite MJ, Rutkowski A, Enright T, Kaplan L, Fine JP, Orwin J: Width of high signal and extension posterior to biceps tendon as signs of superior labrum anterior to posterior tears on MRI and MR arthrography. AJR Am J Roentgenol 2005;185(6):1422-1428.

24. Beltran J, Bencardino J, Mellado J, Rosenberg ZS, Irish RD: MR arthrography of the shoulder: Variants and pitfalls. Radiographics 1997;17(6):1403-1415.

25. Chandnani VP, Yeager TD, DeBe-rardino T, et al: Glenoid labral tears: Prospective evaluation with MRI imaging, MR arthrography, and CT



arthrography. AJR Am J Roentgenol 1993;161(6):1229-1235.

26. Mohana-Borges AV, Chung CB, Res-nick D: Superior labral anteroposteri-or tear: Classifi cation and diagnosis on MRI and MR arthrography. AJR Am J Roentgenol 2003;181(6):1449-1462.

27. Burkhart SS, De Beer JF: Traumatic glenohumeral bone defects and their relationship to failure of arthroscopic Bankart repairs: Signifi cance of the inverted-pear glenoid and the humeral engaging Hill-Sachs lesion. Arthroscopy 2000;16(7):677-694.

28. Cho SH, Cho NS, Rhee YG: Preoperative analysis of the Hill-Sachs lesion in anterior shoulder instability: How to predict engage-ment of the lesion. Am J Sports Med 2011;39(11):2389-2395.

29. Saleem AM, Lee JK, Novak LM: Usefulness of the abduction and external rotation views in shoulder MR arthrography. AJR Am J Roentgenol 2008;191(4):1024-1030.

30. Steinbach LS, Palmer WE, Sch-weitzer ME: Special focus session: MR arthrography. Radiographics 2002;22(5):1223-1246.

31. Fallahi F, Green N, Gadde S, Jeavons L, Armstrong P, Jonker L: Indirect magnetic resonance arthrography of the shoulder: A reliable diagnostic tool for investigation of suspect-ed labral pathology. Skeletal Radiol 2013;42(9):1225-1233.

32. Herold T, Bachthaler M, Hamer OW, et al: Indirect MR arthrogra-phy of the shoulder: Use of abduc-tion and external rotation to detect full- and partial-thickness tears of the supraspinatus tendon. Radiolog y 2006;240(1):152-160.

33. de Jesus JO, Parker L, Frangos AJ, Nazarian LN: Accuracy of MRI, MR arthrography, and ultrasound in the diagnosis of rotator cuff tears: A meta-analysis. AJR Am J Roentgenol 2009;192(6):1701-1707.

34. Duc SR, Mengiardi B, Pfi rrmann CW, Jost B, Hodler J, Zanetti M: Diagnos-tic performance of MR arthrography after rotator cuff repair. AJR Am J Roentgenol 2006;186(1):237-241.

35. Probyn LJ, White LM, Salonen DC, Tomlinson G, Boynton EL: Recurrent symptoms after shoulder instability repair: Direct MR arthrographic assessment. Correlation with sec-ond-look surgical evaluation. Radiolog y 2007;245(3):814-823.

36. Magee TH, Williams D: Sensi-tivity and specifi city in detection of labral tears with 3.0-T MRI of the shoulder. AJR Am J Roentgenol 2006;187(6):1448-1452.

37. Waldt S, Burkart A, Imhoff AB, Brue-gel M, Rummeny EJ, Woertler K: An-terior shoulder instability: Accuracy of MR arthrography in the classifi cation of anteroinferior labroligamentous in-juries. Radiolog y 2005;237(2):578-583.

38. Waldt S, Burkart A, Lange P, Imhoff AB, Rummeny EJ, Woertler K: Diagnostic performance of MR arthrography in the assessment of superior labral anteroposterior lesions of the shoulder. AJR Am J Roentgenol 2004;182(5):1271-1278.

39. Wagner SC, Schweitzer ME, Morrison WB, Fenlin JM Jr, Bartolozzi AR: Shoulder instability: Accuracy of MR imaging performed after surgery in depicting recurrent injury. Initial fi nd-ings. Radiolog y 2002;222(1):196-203.

40. Fritz LB, Ouellette HA, O’Han-ley TA, Kassarjian A, Palmer WE: Cystic changes at supraspinatus and infraspinatus tendon insertion sites: Association with age and rotator cuff disorders in 238 patients. Radiolog y 2007;244(1):239-248.

41. Wissman RD, Ingalls J, Hendry D, Gorman D, Kenter K: Cysts within and adjacent to the lesser tuberosity: Correlation with shoul-der arthroscopy. Skeletal Radiol 2012;41(9):1105-1110.

42. Tirman PF, Feller JF, Janzen DL, Peterfy CG, Bergman AG: Associa-tion of glenoid labral cysts with labral tears and glenohumeral instability: Radiologic fi ndings and clinical signif-icance. Radiolog y 1994;190(3):653-658.

43. Provencher MT, Dewing CB, Bell SJ, et al: An analysis of the rotator interval in patients with anterior, posterior, and

multidirectional shoulder instability. Arthroscopy 2008;24(8):921-929.

44. Lee HJ, Kim NR, Moon SG, Ko SM, Park JY: Multidirectional instability of the shoulder: Rotator interval dimen-sion and capsular laxity evaluation using MR arthrography. Skeletal Radiol 2013;42(2):231-238.

45. Bui-Mansfi eld LT, Taylor DC, Uhorchak JM, Tenuta JJ: Humeral avulsions of the glenohumeral liga-ment: Imaging features and a review of the literature. AJR Am J Roentgenol 2002;179(3):649-655.

46. Chung CB, Sorenson S, Dwek JR, Resnick D: Humeral avulsion of the posterior band of the inferior glenohumeral ligament: MR ar-thrography and clinical correlation in 17 patients. AJR Am J Roentgenol 2004;183(2):355-359.

47. White EA, Schweitzer ME, Haims AH: Range of normal and ab-normal subacromial/subdeltoid bursa fl uid. J Comput Assist Tomogr 2006;30(2):316-320.

48. Guntern DV, Pfi rrmann CW, Schmid MR, et al: Articular cartilage lesions of the glenohumeral joint: Diagnostic effectiveness of MR arthrography and prevalence in patients with sub-acromial impingement syndrome. Radiolog y 2003;226(1):165-170.

49. Dietrich TJ, Zanetti M, Saupe N, Pfi rrmann CW, Fucentese SF, Hodler J: Articular cartilage and labral lesions of the glenohumeral joint: Diagnostic performance of 3D water-excitation true FISP MR arthrography. Skeletal Radiol 2010;39(5):473-480.

50. Hayes ML, Collins MS, Morgan JA, Wenger DE, Dahm DL: Effi cacy of diagnostic magnetic resonance imaging for articular cartilage lesions of the glenohumeral joint in pa-tients with instability. Skeletal Radiol 2010;39(12):1199-1204.

51. Bittersohl B, Miese FR, Dekkers C, et al: T2* mapping and delayed gado-linium-enhanced magnetic resonance imaging in cartilage (dGEMRIC) of glenohumeral cartilage in asymp-tomatic volunteers at 3 T. Eur Radiol 2013;23(5):1367-1374.

mri of the shoulder in adults - acreditacion-fmc.org€¦ · perspective of shoulder specialists...

Documents