Shoulder Impingement Syndromes

Prevalence, Epidemiology, and Definitions

The etiology of rotator cuff tendinosis and tears is likely multifactorial, involving both intrinsic and extrinsic factors. Multiple causal factors have been described in association with rotator cuff pathology, including vascular, degenerative, traumatic, and anatomic/mechanical factors. The diagnosis is based on the patient's history and findings on physical examination, supplemented by imaging findings.

The modern understanding of shoulder impingement owes much to the description and classification described by Neer. He popularized the theory that extrinsic mechanical factors primarily related to the anterior one third of the acromion process, the coracoacromial ligament, and sometimes the acromioclavicular joint are responsible for impingement (predominantly of the supraspinatus tendon and long head of the biceps tendon), which he believed was, in turn, responsible for 95% of rotator cuff tears. This is known as primary mechanical impingement (see eFig. 8-1 ). He postulated that the development of rotator cuff tears was best explained by the shape and slope of the anterior acromion. Patients with a “prominent” anterior edge of the acromion should be more susceptible to impingement and, by implication, rotator cuff tears ( Fig. 8-1 ).

Neer described three stages of impingement lesions. In stage I there is edema and hemorrhage of the rotator cuff. Stage I lesions tend to be seen in patients younger than 25 years old, although they may be seen in patients of any age and may heal with only conservative treatment. Clinical symptoms of these lesions may be the same as those of stage III lesions, making clinical differentiation difficult. With repeated mechanical irritation, there is inflammation and fibrosis of the subacromial/subdeltoid bursa and “tendinitis,” which tends to be exacerbated by activities performed with the arm over the head (stage II). This stage of disease is typically seen in patients aged 25 to 40, according to Neer. Stage III is characterized by partial or full-thickness rotator cuff tears, biceps tendon rupture, and osseous changes.

Ogawa and colleagues studied the relationship between subacromial enthesophyte formation and rotator cuff abnormalities in a series of 1029 shoulders. In their control group, small enthesophyte formation (less than 5 mm) was associated with advancing age; however, the presence of a small enthesophyte had no significant association with rotator cuff tears. This study demonstrated a high association between enthesophytes of 5 mm or greater in size and bursal surface rotator cuff tears, full-thickness supraspinatus tears, and massive rotator cuff tears.

Secondary extrinsic impingement is most commonly seen in throwing athletes younger than 35 years old and is likely a result of glenohumeral or scapular instability. It is thought to be a much less common cause of rotator cuff abnormality related to a decrease in the supraspinatus outlet due to instability of the glenohumeral joint. With repetitive stress, as is often seen in athletes who throw overhand or over the head, there is damage to the gleno­humeral ligaments, leading to mild instability. This instability leads to increased stress on the rotator cuff (dynamic stabilizers) and subsequent fatigue. It allows anterior translation of the humeral head and secondary mechanical impingement of the rotator cuff on the coracoacromial arch. Scapulothoracic muscle weakness or inflexibility (especially of the trapezius, rhomboids, and serratus anterior muscles) may lead to scapular instability, which may, in turn, lead to impingement of the rotator cuff by the coracoacromial arch during throwing.

Intrinsic factors have also been implicated in the development of rotator cuff tears. Multiple studies have shown that abnormalities may occur in the rotator cuff without associated abnormalities of the undersurface of the acromion. Contrary to Neer's hypothesis that mechanical impingement is the leading cause of rotator cuff tears, many authors have postulated that tendon degeneration is the primary cause. Tendon degeneration is thought to be associated with aging, and virtually all ruptured tendons demonstrate evidence of degeneration. In their anatomic and radiologic study of 76 shoulders, Ogata and Uhthoff concluded that the majority of rotator cuff tears are caused by intrinsic degenerative tendinopathy, not by impingement. This view was supported by the finding of increasing incidence and severity of rotator cuff tears with age and by the lack of correlation between aging and degenerative changes of the undersurface of the acromion (with the exception of very advanced degenerative changes). Furthermore, Kjellin and colleagues showed that isolated histologic abnormalities, which may lead to tendon tears, may occur on the articular surface of the supraspinatus tendon. This suggests that impingement, which would be expected to cause predominantly bursal surface abnormalities, may not be the primary or even most common cause of rotator cuff tears.

Codman initially described a region of relative hypovascularity in the rotator cuff 5 to 20 mm from its insertion on the greater tuberosity, the critical portion or zone. Rathbun and Macnab's cadaveric study also showed a consistent zone of relative avascularity in the supraspinatus tendon varying in location between 1 cm from the point of tendon insertion up to the point of tendon insertion. They also postulated that pressure exerted by the head of the humerus on the supraspinatus tendon with the arm adducted may “wring out” the vessels in this area of relative avascularity ( eFig. 8-2 ). They demonstrated a similar area of avascularity in the intracapsular portion of the biceps tendon where it is stretched out over the humeral head near its point of insertion. These authors also showed that the changes of tendinitis, calcification, and tearing of the supraspinatus tendon occur in this region of relative avascularity and that the changes of avascularity precede the degenerative changes in the tendon. Lohr and Uhthoff demonstrated that whereas the bursal surface of the supraspinatus tendon in the critical zone does show abundant blood supply, the articular surface is deficient in blood supply, likely making it more susceptible to degeneration and tearing than the bursal surface. Although many authors have suggested that there is decreased vascularity to the critical zone of the rotator cuff, it must be noted that other authors disagree. Whereas cadaveric studies have suggested relative avascularity in the critical zone, other studies in symptomatic patients have revealed hyperemia or neovascularization in the critical zone in symptomatic patients. This is supported by the fact that hyperemia and marginal bleeding are commonly observed at the margins of rotator cuff tears at the time of surgery. Some have suggested that symptomatic lesions may be due to mechanical impingement, leading to hypervascularity in the critical zone, and that the tears seen in association with hypovascularity in the cadaveric studies likely represent predominantly asymptomatic lesions. Furthermore, after their study of 32 cadaveric shoulders, Clark and Harryman concluded that while the vessels in the deeper portion of the supraspinatus tendon are relatively small compared with those in the more superficial layers, the blood supply to this portion of the tendon is sufficient for the metabolic needs of the tissue.

A prospective study of 66 patients with partial-thickness tears of the rotator cuff by Ko and colleagues suggests that articular surface tears are predominantly associated with intrinsic pathology and degeneration of the rotator cuff and that bursal surface tears are predominantly associated with subacromial impingement superimposed on less severe degenerative changes of the rotator cuff. Their results also indicate that the acromial insertion of the coracoacromial ligament is the primary area where impingement of the rotator cuff occurs, although the acromioclavicular joint can become the area of impingement in some circumstances.

Trauma may also play a role in the development of rotator cuff tears. The presence of a displaced fracture of the greater tuberosity implies a rupture of the rotator cuff by definition ( eFig. 8-3 ). Acute trauma to the shoulder with or without associated dislocation may be responsible for tears of the rotator cuff, especially in older individuals (who likely have underlying degeneration) ( eFig. 8-4 ). Recurrent multidirectional and anterior instability as well as sports that require repetitive motions in an abducted, extended, and externally rotated position (throwing or serving motion) have also been implicated as traumatic causes of rotator cuff tears.

Clinically, patients usually present with a history of slowly developing anterior shoulder pain increasing over a period of weeks to months that may radiate to the lateral humerus. The pain is often related to activities performed with the arm over the head. Patients typically experience pain with abduction and external rotation or elevation and internal rotation but usually retain a full range of shoulder motion.

Neer described the “impingement test” to help distinguish shoulder impingement from other causes of shoulder pain. To perform this test, the scapula is stabilized with one hand while the examiner's other hand raises the patient's arm in forced forward elevation to impinge against the acromion. If the shoulder pain produced by this maneuver is relieved by a subacromial injection of 10 mL of 1.0% lidocaine (Xylocaine), then according to Neer, the cause of the shoulder pain can be attributed to (primary mechanical extrinsic) impingement. Several uncommon impingement syndromes deserve special mention. These include subcoracoid impingement and internal (posterosuperior) impingement. Subcoracoid impingement is an uncommon type of shoulder impingement in which there is narrowing of the coracohumeral space and resultant impingement of the subscapularis tendon. Patients complain of pain in the anterior aspect of the shoulder that is caused by adduction, internal rotation, and forward flexion, because this position decreases the coracohumeral space (space between the tip of the coracoid process and the lesser tuberosity). Internal impingement is classically, but not exclusively, associated with throwing overhand or over the head and occurs when the shoulder is placed in marked abduction and external rotation, at which time there may be impingement of the posterior supraspinatus and anterior infraspinatus tendons between the humeral head and glenoid. Findings in this type of impingement include posterosuperior labral lesions, superior labral tears (SLAP tears), humeral head articular cartilage lesions, humeral head cystic changes, and rotator cuff abnormalities of the infraspinatus and, less commonly, supraspinatus tendons. Anterosuperior impingement is a more recently described form of impingement thought to be a cause of anterior shoulder pain in some patients. Findings associated with this form of impingement include lesions of the biceps pulley, medial subluxation of the long head of the biceps tendon, and partial tears of the cranial fibers of the subscapularis tendon.

Definitions

Several definitions should be established before proceeding to a review of rotator cuff pathology. Partial-thickness rotator cuff tears do not produce a communication between the glenohumeral joint and the subacromial/subdeltoid bursa, and full-thickness tears do produce such a communication. The former can be subdivided into articular and bursal surface tears depending on which side of the rotator cuff is involved. Rim-rent tears, initially described by Codman, are partial insertional articular surface tears of the anterior leading margin of the supraspinatus tendon ( Fig. 8-2 ). Interstitial or intrasubstance tears are partial-thickness tears that are confined to the substance of the tendon and do not contact either the bursal or articular surface of the tendon. Small full-thickness tears are defined as those that measure less than 1 cm in longest diameter, medium-sized tears measure less than 3 cm, large tears measure less than 5 cm, and massive tears measure greater than 5 cm in longest diameter.

Classification

Patte has proposed a classification system for rotator cuff tears based on the following criteria: extent of tear, topography of the tear in the sagittal plane, topography of the tear in the frontal plane, quality of the muscle, and state of the long head of the biceps.

He defined the extent of the tear in centimeters measured at the level of the osseous insertion and divided it into four possible categories: group I lesions are defined as partial tears or full-thickness tears measuring 1 cm in sagittal diameter at the level of the osseous detachment, group II lesions represent full-thickness tears of the entire supraspinatus tendon, group III lesions are full-thickness tears involving more than one tendon, and group IV lesions are massive tears with secondary osteoarthritis. In the sagittal plane, lesions are divided into six segments: segment 1 represents an isolated subscapularis tear; segment 2 represents an isolated coracohumeral ligament tear (usually traumatic in etiology); segment 3 includes isolated supraspinatus tears; segment 4 includes tears involving the supraspinatus and upper half of the infraspinatus tendon; segment 5 lesions involve the supraspinatus and entire infraspinatus tendons; and segment 6 lesions involve the entire cuff, including the subscapularis, supraspinatus, and infraspinatus tendons ( eFig. 8-5 ).

This system does not address the teres minor, because tears of the teres minor tendon are extremely uncommon.

In the coronal plane, tears are divided into three stages: stage 1 tears show little retraction of the proximal tendon stump, stage 2 lesions demonstrate retraction of the stump to the level of the humeral head, and stage 3 lesions demonstrate retraction of the tendon to the level of the glenoid ( eFig. 8-6 ). Fatty atrophy in Patte's system is divided into four stages and is based on CT assessment. Currently, however, MRI is predominantly used for determination of the degree of muscle atrophy (see later discussion). Determination of muscle atrophy is important because functional outcome in rotator cuff repair is strongly associated with the recovery of muscle strength. The status of the biceps tendon, including dislocation and tearing, is also important when assessing the rotator cuff. In the setting of extensive tears, abnormalities of the biceps tendon contribute to elevation of the humeral head and may adversely affect the results of rotator cuff repair.

Pathophysiology

Anatomy

Clark and Harryman studied the structure of the rotator cuff tendons in 32 cadavers and demonstrated that all 4 of the tendons of the rotator cuff fuse to form a common insertion on the humeral tuberosities. There is merging between the fibers of the subscapularis and supraspinatus anteriorly and between the fibers of the infraspinatus and supraspinatus posteriorly ( eFig. 8-7 ). The superficial fibers of the tendons appear to pass along lines parallel to the orientation of the individual muscles, and the coalescence of the tendon fibers appears to occur primarily in the deep layers (see later discussion). Tendinous slips extend from the supraspinatus and subscapularis tendons to form a sheath around the biceps tendon (see later discussion). The deep portion of the sheath is formed by fibers from the subscapularis (primarily) and supraspinatus tendons, and the roof of the sheath over the biceps tendon is formed by fibers from the supraspinatus tendon ( eFig. 8-8 ).

Histologic examination demonstrated that the cuff-capsule complex of the supraspinatus and infraspinatus tendons is composed of five layers: layer 1 (most superficial) is a thin layer composed of obliquely oriented fibers that represent an extension of the coracohumeral ligament and that pass to the greater tuberosity in the rotator interval between the supraspinatus and subscapularis tendons; layer 2 is formed by tightly packed tendon fibers in large bundles that extend from their respective muscles to the humerus; layer 3 is composed of tendon bundles that are smaller and less tightly packed than those in layer 2 and run at a 45-degree orientation to one another; layer 4 is composed of loose connective tissue as well as thick collagen bands (which merge with fibers from the coracohumeral ligament and form a ligamentous sleeve around the anterior portion of the supraspinatus tendon); and layer 5 is composed of a contiguous layer of interwoven collagen fibrils and forms the capsule of the glenohumeral joint ( eFig. 8-9 ). The subscapularis tendon is formed by four to six thick parallel collagen fascicles that pass from the muscle of the subscapularis to the lesser tuberosity where they fan out before inserting on the bone. The collagen bundles are tightly packed in the superficial portion of the tendon and are more loosely packed in the deeper portion of the tendon where the tendon fibers are separated by loose connective tissue.

Several anatomic variants have been implicated in the etiology of shoulder impingement and rotator cuff tears. Among primary extrinsic factors responsible for impingement, acromial shape has received much attention. Bigliani and colleagues described three different acromial morphologies: type I flat (17%), type II curved (43%), and type III hooked (39%). In their study, 70% of rotator cuff tears were seen in patients with a type III acromion, whereas only 3% of tears were seen in patients with a type I acromion. In addition, 70% of the patients with an anterior subacromial enthesophyte also demonstrated rotator cuff tears. These data suggest a high degree of correlation between anterior acromial morphology and rotator cuff pathology.

A type IV acromion, which is thought to be relatively uncommon, has also been described. This type of acromion demonstrates a convex distal undersurface and has not been associated with rotator cuff tears. Although sometimes implicated as a cause of impingement, at least one study has suggested that lateral downsloping of the acromion is not significantly associated with extrinsic impingement.

The presence of an os acromiale has also been implicated by some as a potential factor leading to impingement.

The acromial apophysis forms from four individual ossification centers (preacromion, mesacromion, meta-acromion, and basiacromion). Failure of fusion between the acromial apophysis and the spine of the scapula, which should occur by age 22 to 25, results in an os acromiale, the type depending on which ossification centers fail to fuse ( eFig. 8-10 ). The os acromiale may lead to shoulder pain due to instability at the site of nonunion or may be displaced inferiorly into the rotator cuff by the pull of the deltoid muscle, leading to impingement ( eFig. 8-11 ).

Manifestations of the Disease

Subacromial Impingement and Rotator Cuff Tears

Radiography

Radiographs are essential in evaluation of the rotator cuff. They are important to evaluate for other pathologic processes that may simulate rotator cuff tears, such as fractures or calcific tendinitis ( Fig. 8-3 ). Calcific tendinitis may coexist with underlying rotator cuff tears, especially in older patients and in those with small calcium deposits.

Subacromial enthesophyte formation, as discussed earlier, is significantly associated with shoulder impingement and has been shown to be presumptive evidence for it. The anterior portion of the acromion is often difficult to see on conventional views of the shoulder because of superimposition on the body of the acromion. An impingement view has been described in which the x-ray beam is angled 22 to 25 degrees caudad from the anteroposterior projection in an attempt to see the anterior portion of the acromion, the area that Neer believed to be responsible for the vast majority of impingement lesions, to better advantage.

In a study of 523 patients with chronic shoulder pain, the impingement view demonstrated 100 subacromial enthesophytes, whereas only 18 were visible on the routine views. The supraspinatus outlet view, originally described by Neer and Poppen, is a lateral scapular view with caudal angulation of the X-ray beam that, when properly performed, allows evaluation of the shape and slope of the acromion, prominence of the acromioclavicular joint, thickness of the acromion, presence of subacromial enthesophyte formation, and adequacy of subacromial decompression ( eFig. 8-12 ).

Osteophyte formation on the undersurface of the acromioclavicular joint has also been associated with impingement. Impingement due to large acromioclavicular osteophytes likely occurs with greater degrees of shoulder abduction than subacromial impingement.

Cone and colleagues did not observe impingement by acromioclavicular joint osteophytes as an isolated form of impingement but noted that large acromioclavicular osteophytes commonly coexisted with subacromial enthesophytes.

An acromiohumeral interval of greater than 6 to 7 mm is considered to be normal, and narrowing of this space to less than this is highly suggestive of a rotator cuff tear ( Figs. 8-4 and 8-5 ). Concavity of the undersurface of the acromion (acetabulization) due to articulation between the humeral head and acromion is an additional radiographic sign of a large chronic rotator cuff tear ( Fig. 8-6 ).

Obliteration of the peribursal fat plane around the subacromial bursa, which is normally best seen on the internal rotation view of the shoulder, is a sensitive but not a specific sign of rotator cuff tears and may be seen in association with other inflammatory processes, such as calcific tendinitis and rheumatoid arthritis. Other radiographic findings associated with rotator cuff tears include sclerosis and flattening of the greater tuberosity, subchondral cyst formation and periosteal reaction, enthesophyte formation involving the greater tuberosity, narrowing of the glenohumeral joint space, and anterior subluxation of the humeral head ( Fig. 8-7 ).

Magnetic Resonance Imaging

Technical Aspects

Magnetic resonance imaging has become the imaging study of choice for evaluation of the rotator cuff because of its ability to reliably detect full-thickness tears of this structure. There is no ideal imaging protocol for evaluation of the rotator cuff, and imaging protocols should be tailored to the available equipment and to the reader's experience. It is generally agreed that imaging should be performed in multiple planes, provide good spatial and contrast resolution, and require only a reasonable amount of time. Coronal oblique and sagittal oblique imaging planes plotted relative to the long axis of the supraspinatus tendon have proved to be the most useful for evaluation of the rotator cuff. Several studies have shown that fast spin-echo techniques are equivalent to conventional spin-echo techniques in evaluation of the rotator cuff.

Sonin and colleagues demonstrated 100% correlation between T2-weighted spin-echo and turbo spin-echo sequences in evaluation of the integrity of the rotator cuff and an improved signal-to-noise ratio on the turbo spin-echo images. Carrino and colleagues reported that T2-weighted fast spin-echo sequences yielded diagnostic results similar to those for conventional spin-echo sequences. Reviewers in this study also subjectively preferred the fast spin-echo images to the conventional spin-echo images because of the smaller slice thickness that was possible with the fast spin-echo sequences, and also likely because of decreased motion artifact on the fast spin-echo sequences.

Fast spin-echo sequences provide a significant decrease in imaging time over conventional spin-echo sequences as well. They are also useful for minimizing artifact in the postoperative shoulder.

A recent meta-analysis of 65 articles reviewed accuracy of MRI, MR arthrography, and ultrasound in diagnosis of rotator cuff tears. The analysis found MRI to have an overall sensitivity of 85.5% and specificity of 90.4%, with a higher percentage for full-thickness tears than for partial tears. Results were similar for the ultrasound. However, MR arthrogram was found to have a higher sensitivity and specificity.

Another meta-analysis demonstrated sensitivity and specificity of 80% and 95% for partial-thickness tears, and 91% and 97% for full-thickness tears, respectively. Interestingly, there was no significant difference in diagnostic accuracy between MRIs reviewed by general radiologists and musculoskeletal radiologists. However, higher field strength (3 T) MRIs showed greater accuracy than lower-field-strength systems.

Fat suppression is useful in improving soft tissue contrast because it allows an expansion in the dynamic range of image display, eliminates chemical shift misregistration artifacts that occur at fat-water interfaces, and reduces artifacts from respiratory motion. However, because of the limitations of the fat-suppressed sequence, such as lower signal-to-noise ratio, some authors have advocated a modified inversion recovery sequence (modified short tau inversion recovery [STIR] sequence with an inversion time decreased from 150 to 110 msec) to provide more homogeneous fat suppression.

Singson and colleagues demonstrated that T2-weighted fast spin-echo images with and without fat suppression were both excellent for the diagnosis of full-thickness tears but that partial-thickness tears were better demonstrated with the use of fat suppression because of increased lesion conspicuity. Reinus and colleagues demonstrated improved detection of both full-thickness and partial-thickness tears with the use of fat suppression, although the overall detection of partial-thickness tears in their study was poor (35% detection rate with fat suppression and 15% detection rate without fat suppression). Fat-suppression techniques with fast spin-echo T2-weighted sequences employ chemical fat saturation techniques, which rely on the differences in the different resonance frequencies between the protons in fat and water and are subject to several limitations. These limitations include lower signal-to-noise ratios than non–fat-suppressed images, uneven fat suppression because of inhomogeneity in the static and radiofrequency magnetic fields, complete failure of fat suppression, and limited use on systems with low magnetic field strength due to the closer resonance frequencies of fat and water. The modified inversion recovery sequence advocated by some authors (STIR sequence with an inversion time decreased from 150 to 110 msec) improves the relatively low signal-to-noise ratios seen with typical STIR sequences. Kijowski and colleagues showed relatively similar results for this modified STIR sequence relative to fat-suppressed T2-weighted fast spin-echo sequence regarding integrity of the rotator cuff.

Rotator Cuff Tendinosis

Tendinosis is diagnosed on MRI by increased intratendinous signal intensity on proton density or T2-weighted images without tendon disruption. The increased signal intensity seen in the setting of tendinosis may be homogeneous or heterogeneous and may be focal or diffuse. Other MRI findings associated with tendinosis include enlargement of the tendon, a longitudinal intratendinous band of increased signal intensity, or a focal intratendinous zone of increased signal proximal to the tendon insertion on the greater tuberosity ( Fig. 8-8 ).

Total or partial loss of the peribursal fat plane was present in over half of the patients with tendinosis in a study by Rafii and colleagues ( eFig. 8-13 ). A small amount of fluid is also not uncommon in the subacromial/subdeltoid bursa in the setting of tendinosis.

One should be careful not to overcall tendinosis on short TE sequences because of the magic angle phenomenon, which occurs in the downsloping lateral portion of the supraspinatus tendon, where the tendon fibers are oriented at about 55 degrees to the static magnetic field. The falsely increased signal on short TE sequences fades on T2-weighted sequences.

Most rotator cuff tears are the result of an ongoing attritional process. Once present, a tear is likely to gradually increase in size. This process is difficult to see on MRI, though it may sometimes be appreciated in the presence of joint effusion or on MR arthrogram ( Fig. 8-9 ).

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here