1 Shoulder

1.1 Introduction

Acute shoulder injuries and chronic shoulder symptoms have become an increasing problem in recent years. Degenerative shoulder symptoms are exceeded only by degenerative back and knee disorders. Causes include sports, leisure activities, and occupational stresses overlaid on age-related degeneration. Decades of repetitive overhead movements in occupational, leisure, or domestic activities can cause stress syndromes. With increasing age, the incidence of degenerative changes of the shoulder approaches 100%.

The age of the patient is a critical factor in the overall evaluation of symptoms. Many disorders of the shoulder girdle occur more frequently within certain age-groups. For example, clavicular fractures and torticollis frequently occur in newborns and infants. Congenital deformities such as Sprengel deformity, Klippel–Feil syndrome, or cleidocranial dysostosis are first observed at this age. Septic arthritides or osteomyelitis can be the cause of fevers in infants (Table 1.1). Acute injuries predominate in adolescents and young adults. These include dislocated shoulders, injuries to the acromioclavicular joint, and clavicular fractures. After the age of 30, degenerative symptoms tend to predominate, primarily impingement syndrome, tears in the rotator cuff, and degenerative changes in the acromioclavicular joint. Impingement syndromes can occur in athletes, even at a young age. In this case, primary impingement syndrome must be carefully distinguished from impingement of instability. In the older population, degenerative changes in the rotator cuff and glenohumeral joint predominate.

When the history reveals dysfunction following prior trauma, it is important to reconstruct the mechanism of injury as precisely as possible. The magnitude, direction, and point of impact of the forces acting on the joint are important in establishing whether a direct or indirect mechanism of injury was involved. This inquiry yields important diagnostic information about the type and extent of injury. Acute trauma with the arm abducted and externally rotated generally results in an anteroinferior dislocation of the shoulder. A typical example is the quarterback who is sacked by the defensive lineman in the midst of throwing. Internal rotation and adduction often causes posterior dislocation of the shoulder. A fall with the arm extended can cause dislocation if the player fails to tense his or her muscles in anticipation of the fall. A direct blow to the lateral aspect of the shoulder suggests acromioclavicular joint separation. A direct blow to the clavicle or a fall on the extended arm can cause a clavicular fracture. If the patient has a history of shoulder instability, then the question of whether the instability was triggered by acute trauma or a unique motion is important. It is also important to ask if reduction was necessary. Did reduction occur spontaneously, or did the patient position the arm to obtain reduction? Was reduction only possible in an emergency-room setting or only under anesthesia? Was reduction painful? In the case of recurrent instability, documenting the initial event is particularly important. Did trauma precede this dislocation? How old was the patient when the first dislocation occurred? All these questions are crucial for classifying the instability and determining appropriate therapy (Table 1.4a, b)

Table 1.3 Associations between sports and shoulder complaints

In the presence of chronic complaints, the type of pain should be quantified:

Aside from the specific questions focusing primarily on the shoulder girdle, inquire about systematic diseases and other disease processes (Table 1.5). For example, angina often radiates into the shoulder and arm; the radiating pain does not always occur on the left side. Depending on variations in myocardial vascular anatomy, pain can radiate into the right upper extremity. Gallbladder and liver disorders can also produce pain in the right shoulder. The patient’s neurologic and endocrine status should be explored. Both rheumatoid arthritis and hyperuricemia manifest themselves in the shoulder. Patients with diabetes mellitus experience an above-average incidence of shoulder complaints.

Even in young patients the possibility of a neo-plastic process cannot be ignored, particularly when the history is inconsistent with known primary shoulder disorders. One of the most frequent neo-plastic causes for shoulder pain is a Pancoast tumor with a typical Horner syndrome. Other causes include primary osteogenic tumors in younger patients, and metastatic disease in older patients. A history of cold or numb fingers suggests a compressive neuropathy or vascular compromise.

To complete the history, document previous interventions. Inquire specifically about the number, type, and site of previous injections. It is also essential to elicit a history of other medications. Any indication of treatment with anabolic steroids is important, particularly in athletes, because these drugs significantly decrease the ability of muscle-tendon units to respond to exertion. Tears can occur in these weakened structures.

As in any physical examination, a systemic approach is the key to clinical examination of the shoulder. Such an approach can help to prevent important findings from being overlooked and can reduce the risk of inadequate documentation.

Every clinical examination consists of three parts: observation, palpation, and tests.

Observation

Observing the patient provides an initial insight into the diagnosis. This begins the moment the patient’s name is called in the waiting room. Note how the patient arises from the chair, his/her gait, and the motion of the contralateral upper extremity. Does the patient undress smoothly and symmetrically? A patient with a frozen shoulder will avoid any shoulder rotation and movements above horizontal when undressing. If acute calcific bursitis is present, the intense pain will cause the patient to anxiously hold the affected arm fixed to the side; often the patient will hold the affected arm with the other hand. Patients with a tear in the rotator cuff will often ask for help when undressing because they lack the strength to abduct the arm. An arm adducted and held in internal rotation (in the “porter’s tip” position) suggests Erb palsy (Fig. 1.1).

Table 1.5 Non-shoulder-related disorders that have components of shoulder complaints

Fig. 1.1 Typical appearance of Erb palsy with the arm adducted and in internal rotation.

Always uncover both shoulders for the examination. Examination gowns that leave the shoulder girdle exposed may be helpful with female patients. Inspection begins anteriorly and proceeds posteriorly. Note any blisters, hematomas, scrapes, or other pathologic skin changes. Asymmetry, especially muscular atrophy, is best revealed by comparing one side with the other. A lateral tangential view is better for comparing the sides. This view will reveal minor differences between the sides in the sternoclavicular joints. These may be caused by subluxation or dislocation, or by degenerative changes in the joint. When observing and comparing the acromioclavicular joint from both sides, be alert to swelling or step-off resulting from acromioclavicular joint separation. The superiorly protruding clavicle with a “piano key” phenomenon is a diagnosis from observation (Fig. 1.2) like a tear of the long head of the biceps tendon with distal migration of the muscular belly (Fig. 1.3). The same applies to many congenital disorders such as Sprengel deformity (Fig. 1.4), Klippel–Feil syndrome, congenital torticollis, or a clavicular fracture frequently occurring in newborns and infants.

Fig. 1.3 Proximal tear of the long head of the biceps showing distal migration of the muscular belly.

Fig. 1.4 Sprengel deformity showing the superiorly protruding scapula.

Isolated atrophy of the supraspinatus muscle suggests a tear in this tendon. Atrophy of the supraspinatus and infraspinatus muscles can be due to a tear in the rotator cuff or a scapular notch syndrome. Isolated atrophy of the infraspinatus muscle can be caused by entrapment of the infraspinatus branch of the suprascapular nerve.

Fig. 1.5 Winged scapula, a sequela of a long thoracic nerve injury.

The drooping dominant shoulder in a tennis player is a frequently seen posture, as is an internally rotated posture with hypertrophy of the pectoralis muscle. The drooping shoulder can result in a secondary thoracic outlet syndrome and secondary impingement syndrome. Scoliotic changes in the spine can also develop. An obvious hematoma in the middle third of the clavicle is a sure sign of a clavicular fracture. Conspicuous protrusion of the scapula (winged scapula) is often the result of paralysis of the muscles by which it is fixed (serratus anterior and trapezius; Fig. 1.5).

Table 1.6 Typical diagnoses of disorders made by observation

Preliminary Examination of the Cervical Spine and Arm

Every examination of the shoulder girdle also includes an examination of the cervical spine (see p. 289). Often, the cause of shoulder symptoms may be found in the cervical spine and vice versa. The examination includes evaluating mobility in flexion and extension, right and left lateral bending, and right and left rotation in neutral head position and with the head flexed and extended. Next, palpate the bony structures, such as the transverse and spinous processes. Often, muscle tension in the paravertebral musculature and the trapezius will cause pain in the shoulder and the back of the neck. The pain is often localized in the descending part of the trapezius muscle. Finally, seek enlarged lymph nodes in the shoulder and neck area.

Look for tenderness to palpation about the elbow that may include medial or lateral epicondylitis. Pain in typical sensory nerve distribution may provide clues about compression neuropathies (cubital tunnel syndrome, carpal tunnel syndrome, or radial tunnel syndrome).

Palpation

The next step involves systematic palpation of the shoulder girdle. This is performed with the patient sitting or standing.

Sternoclavicular joint. Stand behind the patient and begin by examining the Sternoclavicular joint. Grasp the clavicle between the thumb and forefinger and move it back and forth to evaluate instability. Tenderness to palpation without instability is a sign of joint irritation. Move laterally along the clavicle from the Sternoclavicular joint to the acromioclavicular joint. Any irregularities from an old clavicular fracture will be palpable.

Acromioclavicular joint. The joint space is easily palpable when approached from the medial aspect of the clavicle. Particularly in athletes, whose sport involves overhead movements, and in weight lifters or body builders, the joint will often be tender to palpation, indicating irritation of the joint or the onset of degenerative changes. Instability or a loose acromioclavicular joint can be detected by grasping the clavicle between the thumb and forefinger and moving it back and forth. If the lateral end of the clavicle protrudes as the result of acromioclavicular joint separation, it can be reduced by applying vertical pressure. When the pressure is released, it will spring back into the original position (“piano key” phenomenon).

Fig. 1.6 The long tendon of the biceps is palpated in the bicipital groove with the arm in external rotation.

Bicipital groove. The proximal part of the long tendon of the biceps lies in the subacromial space and courses through the narrow bicipital groove. This groove can be palpated deep in the anterior deltoid region. With the patient’s arm flexed at the elbow, passively rotating the arm internally and externally presents the lesser tuberosity, bicipital groove, and greater tuberosity in a more exposed position (Fig. 1.6). If pathology is present, this examination will cause the patient discomfort. If the tendon has a tendency to sublux or dislocate, this test can be used to trigger instability for diagnostic purposes.

Subscapularis tendon. Irritation of the subscapularis tendon in athletes is often the result of overuse in throwing sports. With the arm in external rotation, the tendinous insertion can be palpated near the lesser tuberosity.

Anterior joint capsule. With the arm in external rotation, the anterior joint capsule and labrum complex can be palpated between the coracoid process and the lesser tuberosity. Localized pain in this area due to irritation and lesions of the anterior passive stabilizers, is a sign of anterior instability.

Fig. 1.7 The supraspinatus tendon and the subacromial bursa are palpated with the arm in extension.

Pectoralis major. The tendinous insertion of the pectoralis major is located distal to the lesser tuberosity. Changes in this muscle will often be found in weight lifters. If the history reveals use of anabolic steroids, a steroid-induced pectoralis tear can occur.

Subacromial space. With the shoulder in a neutral position, neither the subacromial bursa nor the supraspinatus tendon are accessible for palpation. Passively extending the shoulder rotates the structures underneath the acromion outward, allowing palpation of the anterior part of the supraspinatus tendon (Fig. 1.7). The parts located further posteriorly are exposed by flexing the arm and palpating immediately behind the posterior margin of the acromion.

Infraspinatus tendon. The infraspinatus tendon in the posterior and lateral portion of the humeral head is accessible with the shoulder in a neutral position. Degenerative changes or overuse injuries are less frequent than in the supraspinatus tendon. Pathologic findings, however, may be present, particularly in athletes with hypermobile or unstable joints as a result of repetitive eccentric stress from the external rotators.

Fig. 1.8 Complex movements such as the Apley “scratch” test where the patient touches the superior medial angle of the contralateral scapula are used for testing combined external rotation and abduction.

Active range of motion. Several tests utilizing complex motions are suitable for initial evaluation of the patient’s active range of motion. These complex motions themselves comprise several separate component motions. One such test is the Apley “scratch” test. To test abduction and external rotation, ask the patient to reach behind his or her head and touch the superior medial angle of the contralateral scapula (Fig. 1.8) Internal rotation and adduction are tested by having the patient grasp the contralateral acromion (Fig. 1.9). Internal rotation and adduction can also be demonstrated by instructing the patient to touch the inferior angle of the contralateral scapula behind his or her back (Fig. 1.10). To test full bilateral abduction, instruct the patient to abduct both arms with the elbows extended and the palms in supination. The patient should be able to touch both hands together in the midline above his or her head (Fig. 1.11). The advantage of these tests is that they quickly demonstrate the patient’s range of motion for both sides simultaneously. Symmetry of motion and even slight limitations on the affected side are easily detected. In the range-of-motion examination, the patient first demonstrates the active range-of-motion of the unaffected side. Then he or she performs the same motions with the affected extremity.

Passive range of motion. These tests are best performed standing behind the patient. The patient may sit or stand. Holding the patient’s arm above the flexed elbow, place the other hand on the shoulder. The hand on the upper arm guides the patient’s arm through the arc of motion as the hand on the shoulder monitors the movements of the scapula and the humeral head. This permits detection of spasm. If motion is limited due to pain, the pendulum test can be used to document passive range of motion while eliminating the need to use rotator cuff muscles and avoiding stress on the subacromial space. To perform this test, instruct the patient to bend forward and allow the arms to hang down loosely from the body.

Fig. 1.11 To evaluate full bilateral abduction, the patient abducts both arms with the elbows extended and the palms in supination.

This test can be useful to document inconsistencies between reported symptoms and physical examination findings (Fig. 1.12).

Limited range of motion after soft-tissue injuries may present as a frozen shoulder syndrome (adhesive capsulitis). Immobilization of the joint in adduction for several weeks can result in shortening of the adductors, which is identifiable by limited abduction and external rotation. Scars in the axilla can contribute to limited motion, as can radiation therapy. Palpable crepitus with audible snapping and grating, particularly during rotation, may have several causes. Crepitus during passive motion is frequently a sign of changes in the subacromial space. These changes may include thickening and irregularities in the subacromial bursa and the rotator cuff, or changes in the greater tuberosity. Crepitus during active motion is suggestive of instability and may indicate a labral lesion. Palpable snapping, particularly with horizontal adduction and abduction, is encountered with acromioclavicular joint pathology. Crepitus with external or internal rotation in abduction may point to instability of the long head of the biceps.

Normal and abnormal range of motion of the joint must be documented. Abnormal motion may be due to labral or capsular pathology or to weakness in the muscles used to actively stabilize the shoulder. Grasp the humeral head between the thumb and fingers and translate it anteriorly and posteriorly with the shoulder muscles relaxed (Fig. 1.13). This will demonstrate laxity in the joint. Excess laxity may be a part of shoulder instability.

Fig. 1.14 Horizontal stretch. Reduction in cross-body adduction indicates posterior capsular contracture.

Reduced horizontal adduction is a sign of shortening of the posterior joint capsule (Fig. 1.14).

The active and passive ranges of motion are measured and documented using the neutral-0 method. The following ranges of motion are regarded by some physicians as normal:

– Adduction and abduction: 75°-0°-180°

– Extension and flexion: 60°-0°-180°

– Horizontal extension and flexion: 45°-0°-135°

– Internal and external rotation in adduction: 80°-0°-65°

– Internal and external rotation at 90° adduction: 70°-0°-90°

These values are not always standardized. Note that any abduction past 90° indicates scapulothoracic, rather than glenohumeral, motion.

Although the neutral-0 method is used for all measurements of abduction and adduction, flexion and extension, and internal rotation, the extent of internal rotation is also frequently documented as the part of the body that the patient can reach with the thumb (for example, the greater trochanter, sacroiliac joint, lumbar spinous process, or thoracic spinous process). Both the active and passive ranges of motion are documented. Concentric limitation of motion is a sign of adhesive capsulitis. If external rotation, in particular, is limited, then the differential diagnosis must include chronic posterior dislocation or severe osteoarthritis of the shoulder. Concentric limitations of motion with relatively good external rotation while supine is suggestive of a chronic tear in the rotator cuff.

Scapulohumeral rhythm. Irregularities in the scapulohumeral rhythm are best observed from behind the patient. A frozen shoulder, degenerative glenohumeral arthritis, or rheumatoid arthritis of the shoulder will not permit uniform motion. In these conditions, the motion will be segmented and irregular; abduction of more than 90° is rarely possible. Abduction is achieved by scapulothoracic motion rather than by glenohumeral motion. A complete tear of the rotator cuff leads to an even more severe limitation; generally only a few degrees of abduction can be achieved. Again, most abduction will occur as a result of the scapulothoracic motion. With smaller tears, limited abduction will not be so severe; complete physiologic abduction can occasionally be achieved, albeit painfully. Minimally limited abduction may be seen with longitudinal tears. Any weakness detected in abduction is an indication for further diagnostic studies.

Scapula slide test. Physiologic coordinated motion between the scapula and the arm follows certain regular patterns. Abduction of the arm produces a regular lateral sliding motion of the scapula. Observe whether the coordinated motion of the scapula is symmetric. More than 1 cm of asymmetry is regarded as pathologic. For purposes of objective measurement, the distance between the inferior angle of the scapula and the spinous process of the seventh thoracic vertebra is used.

Nerve injuries can produce uncoordinated scapulothoracic motion. Coordinated motion requires stabilizing the scapula on the chest wall. If the stabilizing muscles are weakened or paralyzed, uncoordinated motion results. The main stabilizing muscles are the trapezius and serratus anterior. With neurologic injuries, a winged scapula can be present.

Paralysis of the serratus anterior is encountered most frequently. If this muscle is completely denervated, the diagnosis is made by inspection. Muscular imbalance causes the scapula to protrude, and the inferior angle migrates medially. Maximum scapular winging is demonstrated when the patient elevates his or her arms and presses against a wall (Figs. 1.15a-c). This test will clearly demonstrate slight weakness that will not cause noticeable changes at rest. Paralysis of the trapezius causes an identifiable change in the position of the scapula while the scapula is pulled inferiorly, and the inferior angle migrates laterally and inferiorly from the mid-line.

A lesion of the long thoracic nerve will cause winged scapula similar to a trapezius lesion and is seen in the setting of shoulder complaints that follow a lymph node biopsy. In such situations, the posterior triangle of the neck should be carefully inspected.

A compressive neuropathy of the suprascapular nerve will produce atrophy of the supraspinatus and infraspinatus muscles. If the syndrome involves only the infraspinatus branch, then only that muscle will be affected.

Dislocation of the shoulder can result in damage to the axillary nerve, producing paresis of the deltoid and an area of sensory deficit on the lateral aspect of the shoulder.

Figs. 1.15a-c Different forms of winged scapula. (a) Normal position of the scapula. (b) Paralysis of the serratus anterior; the scapula migrates superiorly and medially. (c) Paralysis of the trapezius; the scapula migrates interiorly and laterally.

Specific Tests (see Table 1.7)

Painful arc. This is a painful segment of passive or active motion. It should be described in flexion, abduction, or external rotation. It is generally accepted to demonstrate subacromial pathology.

Acromioclavicular compression test (cross-body compression). With the patient’s shoulder flexed 90°, maximum internal rotation, and the elbow flexed 90°, press against the patient’s olecranon from lateral to medial. Pain projected into the acromioclavicular joint is a sign of involvement of this joint.

Neer impingement test. Stand behind the patient, immobilizing the scapula with one hand and briskly flexing the patient’s arm with the other (Fig. 1.16). This will cause the greater tuberosity to decrease the volume of the subacromial space. Pain and weakness may indicate an impingement syndrome. To distinguish between rotator cuff pathology and impingement of instability, repeat the test after infiltrating the subacromial space with local anesthetic. If painless weakness persists after infiltration, consider the possibility of a rotator cuff tear.

Differential anesthetic injection in the shoulder examination. Occasionally it is difficult to differentiate between pain due to degenerative changes in the acromioclavicular joint and subacromial impingement on the basis of the history and physical examination. If there is coexisting radiographic evidence of acromioclavicular and subacromial degenerative changes, the diagnostic dilemma is even greater.

In settings such as this, an extension of the Neer impingement test (using local anesthetic injection to isolate the pain generator) is helpful. These techniques are most useful when the possibility of a rotator cuff tear has been eliminated by history, physical examination, or imaging studies.

Table 1.7 Overview of specific tests for various clinical syndromes

Jobe impingement test. Jobe introduced a modification of the impingement test in which the arm is rotated internally in abduction. This is more likely to compress the posterior parts of the supraspinatus tendon in the subacromial space (Fig. 1.17).

Hawkins and Kennedy impingement test. The Hawkins and Kennedy impingement test represents a further modification. Here, 90° of shoulder flexion is coupled with simultaneous forced internal rotation (Fig. 1.18). A positive test suggests the presence of a subcoracoid impingement syndrome.

Fig. 1.17 Jobe impingement test. This compresses the posterior parts of the supraspinatus tendon in the subacromial space.

Drop-arm test. This test can be used to diagnose larger tears in the rotator cuff. First instruct the patient to fully abduct the arm. The patient may be able to do this using compensatory motions. However, when the arm is lowered from about 90° abduction, any tear that interrupts the continuity of the tendons will make it impossible to actively hold the arm, which will drop to the side. Occasionally the patient can maintain 90° abduction with great effort, but a mere light tap on the forearm will be enough to cause the arm to drop. If active abduction is not possible, passively abduct the arm to 90°. When the supporting hand is removed, the patient will be unable to maintain this position (Figs. 1.19a, b).

Fig. 1.18 Hawkins and Kennedy impingement test with horizontal flexion and internal rotation suggests the presence of a subcoracoid compression syndrome.

Figs. 1.19a, b The drop-arm test indicates a lesion in the supraspinatus tendon.

Fig. 1.20 Nabot sign with pain in the subacromial space on compression and rotation.

Nabot sign. Press the arm against the acromion along the longitudinal axis of the humerus while rotating it. Crepitus will often be discernible in the presence of rotator cuff pathology (Fig. 1.20).

Active abduction while relieving stress on the subacromial space. Pain during abduction from neutral can be caused by irritation of the supraspinatus tendon or by subacromial bursitis. To differentiate between the two, the attempt to decompress the subacromial space by applying traction along the longitudinal axis of the humerus with the arm at the patient’s side. This relieves tension on the bursa. If pain persists when the patient now attempts active abduction, this suggests supraspinatus pathology. If abduction is now significantly less painful, this suggests that the bursa is the main cause.

Clinical Signs and Specific Tests for Differentiating Pathology of the Rotator Cuff

Provocative test (isometric rotator tests). If pseudoparalysis is present, further evaluation is needed to precisely determine which component of the rotator cuff is involved. Provocative tests can be very helpful. These evaluate external and internal rotation against resistance with the shoulder in various positions. Weakness is usually attributable to a functional deficit, whereas pain with normal strength is usually attributable to tendinitis or bursitis. Selective testing can evaluate various components of the rotator cuff.

• Supraspinatus
Zero-degree abduction test. Abduction of the arm is initiated by the supraspinatus and deltoid muscles. The 0° abduction test is used to evaluate the abduction initiation role of the supraspinatus. The patient attempts to abduct the hanging arm against resistance (Fig. 1.21).

Jobe test. The 90° supraspinatus test evaluates the holding function of the supraspinatus muscle. In this test, described in the literature as the Jobe test, the patient holds the arms in 90° abduction with the elbows extended. The arms are flexed 30° horizontally in relation to the scapular plane and internally rotated (with the thumbs pointing to the floor). The patient attempts to open the arms further, against resistance (Fig. 1.22). The anterior portions of the rotator cuff are tested with the upper arms in the same position but in external rotation.

Fig. 1.23 Full-can test. Optimal manual muscle-testing position for the isolation of the supraspinatus muscle: elevation at 90° of scapular elevation and 45° of external rotation (full-can position).

Full-can test. This test can also be used to evaluate the supraspinatus muscle in isolation. The arm is abducted 90° in the scapular plane and externally rotated 45°. Attempt to press the arm down (“full-can” position) against the patient’s active resistance (Fig. 1.23).

• Supraspinatus and infraspinatus muscles
Spontaneous internal rotation. The patient stands shirtless or appropriately gowned and as relaxed as possible. Instruct the patient to let the arms hang down at the sides. A tear in the rotator cuff, particularly in the superior and posterior portion, will cause spontaneous internal rotation of the affected arm. This is due to the uncompensated action of the intact internal rotators (Fig. 1.24).

To evaluate the infraspinatus and teres minor muscles, the patient’s arm is adducted, internally rotated 45°, and flexed 90° at the elbow. With the arm in this position, the patient attempts to externally rotate the arm against resistance (Fig. 1.25). To eliminate the deltoid muscle contribution in external rotation, the external rotation can be performed at 90° abduction and 30° flexion. Failure of active external rotation with the arm abducted is indicative of a clinically relevant tear of the infraspinatus tendon (Fig. 1.26).

• Subscapularis
Increased painless passive external rotation with the patient supine and loss of active internal rotation is a sign of a subscapularis tear.

The internal rotators (subscapularis and pectoralis major) are also tested with the arm adducted and flexed at the elbow, this time with resistance against internal rotation (Fig. 1.27). The arm is not placed at the side but held in flexion to eliminate contribution of the deltoid muscle. Painful external rotation against resistance in this position is a sign of involvement of the infraspinatus tendon.

Fig. 1.26 Testing the external rotators while eliminating the effect of the deltoid muscle.

Lift-off test. This test can demonstrate tears in the subscapularis muscle. With the arm in internal rotation, the back of the patient’s hand is placed on his or her back at the belt line, and the patient attempts to lift the hand off the back (Fig. 1.28). A patient with a tear in the subscapularis will be unable to perform this test.

Modified lift-off test. If internal rotation is limited, the patient might not be able to place his or her hand on the back. In such cases the modified lift-off test can be used. The patient places his or her hand flat on the anterior abdomen and attempts to press with force against the abdominal wall. Inability to do so is a sign of a tear in the subscapularis tendon.

Compression and rotation test. This test is suitable for revealing degenerative changes in the glenohumeral joint. The patient lies on the unaffected side. The arm to be examined is flexed and placed at the side. Press on the humeral head from lateral to medial to produce glenohumeral compression, while the patient rotates the arm internally and externally. Cartilaginous lesions will cause audible and palpable crepitus, and the patient will experience subjective symptoms. To exclude possible accompanying subacromial pathology, inject the subacromial bursa with local anesthetic prior to the test (Fig. 1.29).

Specific Tests for the Long Head of the Biceps

Yergason test. This test demonstrates lesions of the long head of the biceps in the bicipital groove, its tendon sheath, or the transverse ligament that anchors it in the bicipital groove. The elbow is flexed 90° with the forearm pronated. The patient now attempts to supinate the forearm and flex the elbow against resistance (Fig. 1.30). Lesions in any of these structures will cause pain in the bicipital groove.

Fig. 1.29 Compression and rotation test. The patient is placed in the lateral position with the arm at the side and the elbow bent to 90°. The glenohumeral joint is compressed by applying downward pressure. The patient rotates the arm externally and internally. Pain and crepitation are signs of chondromalacia and/or degenerative joint diesease.

Speed test. The speed test (palm-up test) is also used to demonstrate lesions of the long head of the biceps. The arm is in 90° forward flexion with the elbow extended and the forearm supinated. Attempt to press the extended arm down against the patient’s active resistance (Fig. 1.31). If the test is positive, the patient will complain of pain in the anterior deltoid region.

Ludington test. Both arms are abducted and the palms placed on the head with the fingers interlocked. Voluntary contraction of the biceps will cause pain in the anterior deltoid region if the test is positive.

Heuter test. Under normal conditions, forceful flexion with the forearm pronated is achieved with the brachialis. The biceps contracts simultaneously, causing supination of the forearm. If this is not observed, then you should look for a lesion of the biceps or its tendons.

The long head of the biceps is directly palpated in the Lippman and De Anquin tests.

Additional tests. These tests may be useful in demonstrating specific pathology. In the Lippman test, the examiner palpates the bicipital groove approximately 3 cm distal to the shoulder with the patient’s elbow flexed at a right angle. If the biceps tendon tends to sublux or dislocate, you can provoke dislocation or subluxation by palpation of the relaxed musculature. This is generally painful for the patient. In the De Anquin test, rotating the upper arm while palpating the biceps tendon in the bicipital groove will cause pain if the tendon is affected. In the Gilcrest test, reducing the subluxed or dislocated long biceps tendon while slowly adducting the arm will cause pain in the anterior deltoid region. The Beru sign demonstrates dislocation of the long head of the biceps tendon. If the long head dislocates, it can be palpated under the anterior deltoid when the biceps is voluntarily contracted. The Duga sign demonstrates an injury of the long head of the biceps tendon. In such an injury, the patient is unable to touch the contralateral shoulder with the hand of the affected arm.

Fig. 1.32 Stretching the long head of the biceps in the bicipital groove.

Stretch test. Passive extension of the shoulder, extension of the elbow, and pronation of the forearm, or the patient’s attempt to actively supinate the forearm, flex the elbow, or flex the shoulder forward from this position will cause pain in the anterior deltoid region (Fig. 1.32).

Compression test. Passive elevation of the arm to the end of the range of motion, and application of posterior pressure, causes pain in the tendon compressed between the acromion and humeral head (Fig. 1.33).

Clinical Tests for Glenoid Labral Lesions

O’Brien test. With the arm flexed 90° and adducted 10°, internally rotate the arm so that the thumb points down. The patient attempts to resist a downward force applied by the examiner. Pain in the acromioclavicular joint is a sign of acromioclavicular arthropathy. Pain projected to the superior part of the glenohumeral joint suggests a SLAP (Superior Labrum from anterior to posterior) lesion. Typically, the pain disappears when the arm is placed in external rotation with the palm up.

Compression and rotation test. With the arm abducted 90° and the elbow flexed 90°, apply compression to the glenohumeral joint by pressing on the humerus while rotating the joint. As in the McMurray knee test, a snapping noise is a sign of damage to the labrum.

Fig. 1.33 Compressing the long head of the biceps against the acromion.

Anterior slide test. The patient is examined either standing or sitting, with the hands on the hips and thumbs pointing posteriorly. Place one hand across the patient’s shoulder with the last segment of the index finger extending over the anterior aspect of the acromion at the glenohumeral joint. Place the other hand behind the patient’s elbow and apply force in an anterosuperior direction. Instruct the patient to push back against this force. Sudden pain in the anterosuperior shoulder, corresponding to the pain typically experienced during exercise, or a palpable snap phenomenon is indicative of a SLAP lesion (Figs. 1.34 and 1.35).

Clunk test. With the patient supine, place your hand on the posterior aspect of the shoulder directly under the humeral head. With the other hand, grasp the distal humerus and elbow condyles. The patient’s affected arm is now brought from the extended position into forward flexion and external rotation. A snap phenomenon or clunk sound indicates damage to the labrum (Fig. 1.36).

Crank test. The crank test is used to demonstrate damage to the labrum and can be performed with the patient sitting or supine. With the patient’s arm abducted 160° in the scapula plane, press along the longitudinal axis of the humerus while rotating the humerus with the other hand. The test is considered to be positive if:

Fig. 1.34 Slide test. Position of the hands and arms for the anterior slide test.

Anterior apprehension test. This test is positive in the presence of anteroinferior instability. The test is performed with the patient sitting or standing and the affected shoulder placed in 90° abduction and 90° external rotation. Continuing the external rotation while applying forward pressure to the humeral head with the thumb is met with muscular resistance as the patient feels that the shoulder is about to dislocate anteriorly (Fig. 1.37). The presence of pain alone should not be regarded as a positive test even though the muscular resistance can often appear simultaneously with the occurrence of pain. Patients will often indicate that “something bad is about to happen” or “it’s starting to go out.” This test can be performed in different degrees of abduction. Performing the test at 45° abduction evaluates the contribution of the medial glenohumeral ligament and the subscapularis tendon. In abduction of 90° and greater, the stabilizing effect of the subscapularis muscle is neutralized and contribution of the inferior glenohumeral ligament is evaluated.

Fig. 1.37 Anterior apprehension test.

Posterior apprehension test This is helpful for diagnosing posterior instability. Place the patient’s affected arm in 90° to 110° abduction at the shoulder and horizontally flex it approximately 20° to 30°. Immobilize the scapula from above with your other hand. The fingers cover the scapular spine and the humeral head while the thumb reaches anteriorly to a point slightly lateral to the coracoid process. As horizontal flexion is slowly increased, the force along the longitudinal axis of the humerus will cause the glenohumeral joint to sublux posteriorly. Both the thumb lateral to the coracoid process and the fingers can palpate the translation of the humeral head. Occasionally, the humeral head will be visible inferior to the acromion as a slight prominence. An extension of 20° to 30° in the same horizontal plane will palpably reduce the humeral head (Fig. 1.38).

Fig. 1.38 Posterior apprehension test.

Evaluation of glenohumeral translation. Glenohumeral translation can be evaluated with the patient supine or sitting. This test provides a semiquantitative method of measuring the extent of laxity. The extent of translation is compared with the nonaffected shoulder.

Passive drawer test. When performing the test on a patient, place your contralateral hand on the patient’s shoulder from posterior so that the index and middle finger lie anteriorly on the humeral head while the ring finger lies on the coracoid process. To test for a passive drawer, grasp the shaft of the humerus with the other hand and move it anteroposteriorly, noting the translation of the humeral head in relation to the coracoid process. With the same landmarks located, you can document an active drawer by having the patient perform active abduction and external rotation of the affected arm (see Fig. 1.13).

AP drawer test. This test can also be used to document AP laxity in the shoulder. The patient is positioned supine with the arm lying comfortably at the side. Attempt to stabilize the scapula while achieving AP translation of the humeral head without encountering muscular resistance form the patient.

In the classic anterior drawer test, stand facing the affected shoulder. To examine the right shoulder, grasp the upper third of the patient’s arm with your right hand while immobilizing the patient’s forearm and hand between your body and upper arm. In this position, flex the patient’s upper arm approximately 20° and place it in slight external rotation. Immobilize the scapula with your other hand so that the fingers reach posteriorly to the scapular spine and the thumb anteriorly to the coracoid process. From this position, carefully translate the humeral head anteriorly. First perform the test at 50° abduction, repeating it at higher degrees of abduction up to 120°. This test can also produce a snapping phenomenon as a result of a torn or degenerative labrum.

Posterior drawer test. Hold the patient’s shoulder in 20° to 30° flexion and 20° to 30° abduction with the elbow flexed 90°. When examining the right shoulder, stand superior to the patient, holding the patient’s arm with your right hand and positioning your left hand so that the thumb is slightly lateral to the coracoid process and the long fingers stabilize the scapular spine posteriorly. Applying pressure with your thumb produces posterior translation.

Leffert test. This test is another way of obtaining quantitative information in the drawer test. Stand above the sitting patient and move the humeral head anteriorly. Anterior motion of your index finger in relation to the middle finger shows the translation of the humeral head (Figs. 1.39a, b).

Fukuda test. This test demonstrates a passive posterior drawer. With the patient sitting, place your thumbs on both scapular spines and your fingers anterior to the humeral head. Pressing posteriorly with the fingers creates a posterior drawer (Fig. 1.40).

Load-and-shift test. This test is performed similarly to the drawer test on the sitting patient. Immobilize the scapula with one hand, grasp the humeral head with the other, apply a medial force to the humerus, and translate the humerus anteroposteriorly.

Rowe test. In the Rowe test, the drawer test is performed with the patient standing and bending forward slightly. This allows the patient to fully relax the shoulder muscles as in the pendulum exercises (Fig. 1.41).

Sulcus sign. If multidirectional instability is present, a positive sulcus sign can be demonstrated. With the patient sitting or standing and the arm of the affected shoulder relaxed along the body, apply traction to the longitudinal axis of the humerus. This will widen the space between the acromion and the humeral head. A sulcus in the skin will appear in this region (Fig. 1.42).

Inferior apprehension test. For this test, place the patient’s arm in 90° abduction. While supporting the arm with one hand, attempt to provoke inferior subluxation by applying pressure to the proximal upper arm with the other hand (Fig. 1.43).

Fulcrum test. The patient is positioned supine on the examining couch with the affected arm externally rotated 90° and abducted 90°. Continuing the external rotation while moving the humeral head anteriorly will provoke muscular resistance by the patient, in combination with pain. Performing the same motion (external rotation) while moving the humeral head posteriorly is painless (Figs. 1.44a, b). This test is helpful in differentiating patients with a simple supraspinatus syndrome from those with tenosynovitis at the insertions of the rotator cuff as a result of hypermobility.

Figs. 1.39a, b Leffert test

Fig. 1.40 Fukuda test

Fig. 1.41 Rowe test. In the standing position, the patient bends slightly forward and the arm is relaxed. A gentle anterior-inferior translation is performed.

Thrower test. In this test, the patient performs a throwing motion against your resistance. This can reveal anterior subluxation during the throwing motion (Fig. 1.45).

Examination under anesthesia. Examining the shoulder under anesthesia permits both clinical and fluoroscopic documentation of joint laxity. The drawer tests described in the previous section may be used.

Occasionally, it is helpful to have an assistant stabilize the scapula. Snapping phenomena resulting from labrum pathology can be readily documented by compressing the humeral head into the glenoid fossa while simultaneously translating the bone (shift-and-load test). The extent of translation with the arm in external rotation should be documented with the arm in various degrees of abduction (60°, 90°, and 135°). Performing a bilateral examination is recommended to allow comparison of both sides. Posterior translation up to half the humeral head diameter can be normal. Anteriorly, the normal translation is significantly less and is usually one-third of the diameter of the humeral head.

Quantifying examination under anesthesia. AP translation is divided into four grades (Fig. 1.46).

• Grade 0: no palpable translation between the humeral head and the glenoid fossa.

• Grade I: the humerus can be moved up to the margin of the glenoid.

• Grade II : Subluxation of the humeral head past the margin of the glenoid with spontaneous reduction. This is the maximum considered normal during examination under anesthesia.

• Grade III: complete dislocation without reduction.

Figs. 1.44a, b Fulcrum test. Moving the humerus anteriorly and rotating it exteriorly produces typical pain (a); external rotation with the head of the humerus centered can be achieved painlessly (b).

Table 1.8 Muscle grading chart

As in all other examinations, the strength of the muscle groups is compared with the opposite side. The chart shown in Table 1.8 is recommended for objective evaluation and documentation.

Forward flexion. Testing evaluates the anterior part of the deltoid muscle (axillary nerve, C5) and the coracobrachialis muscle (musculocutaneous nerve, C5–C6), the secondary flexors, the clavicular head of the pectoralis major, and the biceps. Stand behind the patient and place one hand over the acromion and scapula and the other above the flexed elbow; the fingers are over the anterior segment of the arm and the biceps muscle. Here and in the following tests, instruct the patient to perform the respective motion (in this case flexion). Provide resistance that the patient can barely overcome.

Extension. Testing evaluates the latissimus dorsi (thoracodorsal nerve, C6), teres major (subscapular nerve, C5–C6), and the posterior portion of the deltoid (axillary nerve, C5–C6). The secondary extensors are the teres minor and the long head of the triceps. Extension is tested in a similar position to flexion except that the palm of the hand is now placed on the posterior aspect of the humerus. Instruct the patient to extend the arm posterior, and again provide resistance that the patient can barely overcome.

Abduction. The primary abductors are the middle portion of the deltoid (axillary nerve, C5–C6) and the supraspinatus (suprascapular nerve, C5-C6). The secondary abductors are the anterior and posterior portions of the deltoid and the serratus anterior. Stand behind or beside the patient, placing your hand over the acromion. Place the other hand beside the elbow and press against the lateral epicondyle of the humerus. Instruct the patient to abduct the arm, and determine the amount of resistance that the patient can barely overcome.

Adduction. The primary adductors are the pectoralis major (pectoral nerves, C5–T1) and the latissimus dorsi (thoracodorsal nerve, C6–C8); the secondary adductors are the teres major and the anterior portion of the deltoid. Test adduction in the same position as abduction. Grasp the arm with the resisting hand so that the fingers reach the medial side of the humerus.

External rotation. The primary external rotators are the infraspinatus (suprascapular nerve, C5–C6) and the teres minor (axillary nerve, C5). The posterior portion of the deltoid is the secondary external rotator. Stand beside the patient instructing him or her bend the elbow to 90° with the arm at the side in a neutral position. Place one hand on the elbow and provide resistance on the posterior distal forearm.

Internal rotation. The primary internal rotators are the subscapularis (subscapular nerve, C5–C6), pectoralis major (pectoral nerves, C5–T1), latissimus dorsi (thoracodorsal nerve, C6–C8), and the teres major (subscapular nerve, C5–C6). The anterior portion of the deltoid is the secondary internal rotator. Standing in the same position as for testing external rotation, place your resisting hand close to the radial styloid.

Scapular retraction. The primary retractors of the scapula are the rhomboid major (dorsal scapular nerve, C5) and the rhomboid minor (dorsal scapular nerve, C5). The secondary retractor is the trapezius muscle. Stand in front of the patient and place your hands on the patient’s shoulders. Your palms should lie under the acromion with the fingers reaching around the patient’s shoulders and touching the back. Have the patient retract the scapula while offering as much resistance as the patient can barely overcome.

Scapular protraction. The primary protractor of the scapula is the serratus anterior (long thoracic nerve, C5–C7). Stand behind the patient giving instructions to elevate the arm to 90° so that the humerus is parallel to the floor. Now instruct the patient to attempt to bend the elbow so that the hand touches the shoulder. Place one hand on the patient’s spine; cup the patient’s elbow with your resisting hand.

Scapular elevation. The primary elevators are the trapezius (spinal accessory nerve or cranial nerve XI) and the levator scapulae (C3–C4 and frequently the dorsal scapular nerve, C5). The secondary elevators are the rhomboid major and minor. Stand by the patient and place one hand on each acromion. Normally the elevators of the shoulder girdle are so strong you will hardly be able to overcome them.

Figs. 1.47a, b Dermatomes in the upper extremity.

1.3 Radiology

Fig. 1.49 Anteroinferior dislocation of the shoulder.

True AP view. The true AP radiograph may be taken with the patient standing or supine. The scapula is the reference plane for the glenohumeral joint and is at a 30° to 45° angle to the axis the body. The central ray in the AP radiograph must run parallel to the glenohumeral articular surface. Turn the patient’s torso so that the plane of the shoulder blade lies parallel to the film cassette. The angle between the back and the film cassette will be about 30° to 45° (Figs. 1.48a, b).

The arm is in slight external rotation with the palm facing forward and the elbow extended. Rotation is best evaluated with the elbow flexed 90°. With the arm in external rotation, the greater tuberosity will appear on the lateral aspect of the humerus in profile. The central ray is aimed at the coracoid process and is angled 20° caudally. With the patient and cassette properly positioned, the glenohumeral joint is visualized parallel to the plane of projection, clearly demonstrating the joint space without any overlapping bony structures. For the projection, the humeral head and the glenoid fossa will only overlap if the shoulder is dislocated. The true AP view provides the basic study. These radiographs are suitable for identifying dislocations (Fig. 1.49), tumors, infections (Fig. 1.50), postinfectious conditions, avascular necrosis (Fig. 1.51), osteoarthritis of the shoulder, intra-articular loose bodies (Fig. 1.52), and fractures.

Axial view. The axial radiograph may be taken with the patient supine or sitting. It is most often taken as an inferosuperior projection with the patient supine. Towels or cushions are placed under the patient’s head and affected shoulder to raise them approximately 10 cm. The affected arm should ideally be abducted 90° but pain will often prevent this. The film cassette is placed superior to the shoulder near the neck, with the central ray aimed perpendicularly through the axilla at the acromioclavicular joint (Figs. 1.53a, b).

Figs. 1.53a, b Patient positioning for the axial view.

The axial view is ideal to demonstrate the position of the humeral head in relation to the glenoid fossa. This view is useful in evaluation of a suspected posterior dislocation of the shoulder. In many cases, posterior dislocation cannot be reliably detected in the AP radiograph (see following section). Bony changes to the humeral head, such as a compression fracture or fractures of the greater and lesser tuberos-ities, can be demonstrated in this view, as can avulsion fractures of the glenoid (Fig. 1.54). The axial view can also identify fractures of the coracoid process and acromion, and can document an os acromiale (Fig. 1.55). The os acromiale is classified according to the ossification center as either preacromion, meso-acromion, metaacromion, or basiacromion (Fig. 1.56). The axial view can identify rotator cuff changes not seen on the true AP view.

If limited abduction precludes standard axial views, modified axial or transthoracic projections must be used to obtain images in the second radio-graphic plane.

Combinations of Standard Views for Specific Situations

Particularly in the shoulder, standard projections in two planes should not be deemed satisfactory.

The following views are recommended as a standard series for patients with degenerative shoulder disorders.

— AP view in internal rotation

— AP view in abduction

– Axillary view

A normal radiograph by no means exclude pathologic changes as most pathologic conditions of the shoulder involve soft-tissue injuries that are not visualized on radiographs. Radiographic changes appear only in advanced stages of disorders, for example in chronic tears of the rotator cuff. On the other hand, there is no clear correlation between the extent of bony changes, clinical symptoms, or response to therapy. Calcifications may be conspicuous, but they are not always responsible for symptoms.

Characteristic changes that occur after shoulder instability can be detected in suitable radiographs (instability series). If typical findings are visualized, anterior instability can be clearly distinguished from posterior instability. These projections include:

— AP view in internal rotation

— West Point axillary view

— Modified axillary view (such as the Stryker notch view)

A trauma series may be prepared for patients presenting with intense pain and suspected fractures. this series includes three projections at right angles to each other that can be obtained with minimal manipulation of the injured extremity.

Fig. 1.54 Axial view showing a posterolateral Hill-Sachs lesion on the humeral head and bony Bankart lesion of the glenoid.

AP projection. Radiographs may be obtained in external rotation, internal rotation, or neutral position. The patient may be positioned sitting or supine but it is important for the plane of the shoulder blade to be parallel to the film cassette. In applicable cases, AP views in varying rotation can help to localize calcifications in the specific portions of the rotator cuff (Fig. 1.57). Cystic or sclerotic changes in the greater tuberosity and acromion are identified. Superior migration of the humeral head indicates a tear of the rotator cuff (Fig. 1.58). Measuring the acromiohumeral interval (AHI) allows one to infer the condition of the rotator cuff. This interval is normally between 12 and 15 mm. In a complete tear of the rotator cuff, this distance is 6 to 7 mm. If the interval is less than 3 mm, one can diagnose a chronic massive tear where the edges of the rotator cuff tendons have retracted (Fig. 1.59). Only rarely will the inferior glenoid lesions associated with instability be seen on the true AP radiograph.

Fig. 1.58 Superior migration of the humeral head in a chronic tear of the rotator cuff with severe degenerative joint disease.

Fig. 1.62 Positioning for superoinferior view of the bicipital groove.

Abduction view. This can document the contribution of glenohumeral motion to total abduction. It can also show an unobstructed view of the acromioclavicular joint. The patient stands at a right angle to the film cassette with the affected shoulder abducted 90° and the elbow flexed. In the AP projection, the central ray is aimed at the coracoid process (Fig. 1.60).

Leclercq view. Normally the humeral head will be centered in the glenoid fossa in an AP view. If there is a tear in the rotator cuff, the humeral head will be displaced superiorly when the patient actively abducts the arm against resistance (Fig. 1.61).

Fig. 1.63 Positioning for inferosuperior view of the bicipital groove.

Bicipital groove (superoinferior projection). The patient stands and bends over the X-ray table with the elbow flexed and the supinated forearm placed on the table. The film cassette is placed on the forearm and held in the patient’s supinated hand, or supported by sandbags or towels in a horizontal position. The patient then bends so that the longitudinal axis of the humerus is 10° to 15° to the axis of projection. The bicipital groove is palpated and marked. The central ray enters the marked area vertically, producing an unobstructed projection of the bicipital groove (Fig. 1.62).

Bicipital groove (inferosuperior projection). A simpler, more practical method is the inferosuperior projection. The patient is positioned supine, the affected arm slightly abducted, and the upper arm supported in external rotation. The central ray parallels the axis of the humerus and is centered on the anterior margin of the humeral head. The film cassette is placed above the affected shoulder and is held by the patient (Fig. 1.63).
Both projections of the bicipital groove only provide reliable information if the central ray enters precisely tangential to the groove at its deepest point (Fig. 1.64). In actual clinical practise this is often not achieved, resulting in poorly projected images that are difficult to evaluate. If the radiographic visualization of the bicipital groove is vital, we recommend spot views via fluoroscopy. Ultrasound in the hands of an experienced examiner can provide the desired information.

Visualization of the subacromial space. The patient is positioned for the AP view in external rotation except that the central ray is angled 50° caudally. This view readily demonstrates calcifications of the cora-coacromial ligament.

Coracoid process position. An inferosuperior projection shows the contour of the coracoid process. The patient is positioned supine with the arms at the sides, and the film cassette is placed under the patient’s shoulder approximately 3–5 cm proximal to the coracoid process. The arm is slightly abducted with the hand supinated. The central ray is aimed at the coracoid process angled 15° to 30° cranially. In round-shouldered patients, this angle is greater than in flat-shouldered patients.

Fig. 1.64 radiographic image of the bicipital groove.

Supraspinatus outlet view. This demonstrates the shape of the acromion in the sagittal plane. The patient stands so that the body of the scapula is perpendicular to the film cassette. The scapular spine is positioned parallel to the floor, and the humerus is in neutral. The central ray is angled 15° caudally and aimed at the acromioclavicular joint (Fig. 1.65). This technique images the coracoacromial arch with the supraspinatus tendon beneath it and may be used to demonstrate changes in the anterior margin of the acromion. Bigliani has identified three acromial shapes (Figs. 1.66 a–c):

• Type I: flat acromion

• Type II: curved acromion

• Type III: anterior-to-inferior hook

Fig. 1.65 radiographic positioning for demonstrating the supraspinatus tunnel.

However, the angle of inclination of the acromion appears to be more important in subacromial stenosis (Fig. 1.67).

Rockwood view. Routine AP views will not generally show calcifications of the coracoacromial ligament or inferior osteophytes on the acromion. Rockwood uses the 30° caudal projection. With the patient standing, an AP radiograph is taken with the central ray aimed 30° caudally (Fig. 1.68). This view detects changes in the subacromial area.

Special Views for Instability

Anterior dislocation of the shoulder can produce bony changes in the anteroinferior glenoid fossa (Bankart lesions or capsular calcifications) and on the postero-lateral humeral head (Hill-Sachs lesion). Standard radiographs will often fail to detect both these lesions; special views are necessary to document them. However, standard radiographs can provide important information in the presence of shoulder instability. True AP radiographs can provide information about the direction of dislocation and about accompanying injuries. An axial radiograph should always be obtained since this is the only way to positively exclude posterior dislocation (Figs. 1.69a, b). Even chronic dislocation with its typical associated injuries can be clearly documented in the axial projection (Fig. 1.70).

Figs. 1.66a–c The three types of acromion in the supraspinatus outlet view.

Fig. 1.67 The increase in acromial inclination decreases the supraspinatus outlet.

After reduction, or in the presence of, recurring instability, the AP view provides information about associated bony injuries (Hill-Sachs defects or Bankart lesions), possible muscular deficits (Figs. 1.71 a, b), and secondary osteoarthritis of the shoulder.

Figs. 1.69a, b Chronic posterior dislocation in the AP (a) and axial (b) views.

Fig. 1.70 Chronic anterior dislocation in the axial projection.

West Point view. This technique demonstrates the anteroinferior bony margin of the glenoid in a PA projection. The patient is positioned prone with the arm abducted 90°; the elbow is flexed to allow the forearm to hang down over the side of the table. The affected shoulder is raised approximately 8 cm, and the head and neck are inclined toward the 3contralateral side. The film cassette is placed above the shoulder and the central ray is aimed directly at the axilla, angled 25° down from horizontal and 25° medially (Figs. 1.72a, b). Properly positioned, this will provide a tangential view of the anterior inferior glenoid fossa (Fig. 1.73). Over 90% of patients with recurrent anterior shoulder dislocation will show erosion, irregularities, or capsular calcifications in this region.

Figs. 1.71 a, b Inferior subluxation of the humeral head (b) compared to a normal shoulder (a).

Apical oblique projection. This view is helpful in diagnosing shoulder instability. The patient is positioned sitting, with the injured shoulder in contact with the film cassette. This may be done using a sling to immobilize the shoulder. The patient’s torso is rotated approximately 45° to align the plane of the shoulder blade parallel to the film cassette. The central ray is also angled 45° caudally and is aimed at the base of the coracoid process; it runs parallel to the body of the scapula which is held at this angle to the chest with the extremity at rest or adducted (Figs. 1.74a, b).

When evaluating radiographs obtained in this projection, it is helpful to use the base of the coracoid process as an anatomic landmark. This structure lies between the anterior and posterior margins of the glenoid. Structures demonstrated in this projection include the anterior and posterior margins of the bony glenoid as well as intra-articular fractures.

Figs. 1.72a, b West Point axillary shoulder position.

Fig. 1.73 Bony Bankart defect in the West Point projection.

AP view at 60° internal rotation. Probably the simplest radiograph for demonstrating a Hill-Sachs is the AP projection in maximum internal rotation since the posterolateral compression fracture of the humeral head only appears in profile with the arm in internal rotation. This projection is not intended to visualize the glenohumeral joint parallel to the plane of projection, but rather demonstrates the posterolateral contour of the humeral head. The patient stands with his or her back to the film cassette. The central ray is aimed at the coracoid process, at a right angle to the coronal plane. With the elbow flexed 90°, the 60° internal rotation is easily verified (Fig. 1.75).

Figs. 1.74a, b Patient positioning for the apical oblique projection.

Fig. 1.75 Positioning for the 60° internal rotation view for visualizing a Hill-Sachs lesion.

Fig. 1.76 Positioning for the Stryker notch view.

Stryker notch view. This technique is used to image the Hill-Sachs lesion longitudinally. With the patient supine, the upper arm is raised, the elbow flexed, and the hand placed on the head, with the heel of the hand resting at the top of the back of the head. The shaft of the humerus is parallel to the sagittal axis of the body. The upper arm is flexed over 90° and slightly internally rotated. The central ray is angled 10° cranially and aimed at the coracoid process, with the film cassette placed under the patient’s shoulder (Fig. 1.76).

Didiée shoulder position. The Didiée view visualizes the anteroinferior margin of the glenoid and can demonstrate the presence of a Hill-Sachs lesion. The patient is prone, with the arm and back of the hand placed on the posterior iliac crest. The elbow is flexed. The film cassette is under the injured shoulder, and the central ray is angled 45° lateromedially with respect to the floor and aimed at the humeral head (Fig. 1.77). In this view a Hill-Sachs lesion may appear distorted, making it difficult to determine its actual size.

Hermodsson view. This view can demonstrate a posterolateral compression fracture of the humeral head. The patient is supine or standing with the affected arm in internal rotation and the elbow flexed 90°. The hand is placed on the patient’s back, or between the back and table, with the palm touching the table. The central ray is aimed at the humeral head at an angle of 30° to the longitudinal axis of the humerus (Fig. 1.78).

Saxer and Johner posterior tangential view. This technique can demonstrate the width and depth of Hill-Sachs lesions. The patient is supine and the elbow flexed 90°; the forearm is placed on the chest so that the fingertips touch the contralateral shoulder. The film cassette is placed superior to the affected shoulder. The central ray is aimed at the glenohume-raljoint in a caudocranial projection, angled 20° laterally in the coronal plane and 20° posteriorly in the sagittal plane.

Views for Subluxation of the Glenohumeral Joint

Cephaloscapular projection. The cephaloscapular projection can be helpful in diagnosing subluxation of the glenohumeral joint. This projection will demonstrate the acromion, glenoid, humeral head, and coracoid process (overlapped and only slightly obscured by the clavicle). The patient is positioned standing or sitting and bending forward 45° so that the coracoid process and acromion form a vertical plane. The film cassette is placed behind the patient parallel to this plane, with the central ray aimed at the glenohumeral joint and entering horizontally. If radiographs at rest are to be obtained, the elbows is flexed and the pronated forearm is placed on the table in a relaxed position (Figs. 1.79a, b). If posterior stress is desired to demonstrate posterior subluxation, this is achieved by using the patients body weight. Anterior stress is induced by allowing the forearm to hang down and by loading it with weights.

Figs. 1.79a, b Positioning for the cephaloscapular view.

Transthoracic lateral view. With the patient sitting or standing, the lateral side of the affected shoulder is placed against the film cassette, and the contralateral arm is raised so that the supinated forearm is lying on top of the patient’s head. The upper torso is rotated posteriorly. The central ray is perpendicularto the film cassette, aimed between the spine and sternum at a point slightly inferior to the coracoid process (Fig. 1.80). This radiographic projection is difficult to interpret due to overlapping structures, although the scapulohumeral arc (Moloney’s line) is helpful in determining the position of the humeral head in relation to the glenoid fossa. In normal anatomy Moloney’s line, formed by the shaft of the humerus and the axillary margin of the scapula, appears as a smooth continuous curve. In the posterior dislocation of the humeral head, it will be at an acute angle; in anterior dislocation, the angle will be very obtuse (Figs. 1.81 a–c).

Fig. 1.80 Positioning for the transthoracic lateral view.

True lateral (scapular Y) shoulder position. Another technique for obtaining radiographs in the second imaging plane with a painful shoulder is to use the scapular Y position. The painful shoulder may be in internal rotation in a sling. The patient stands with the affected shoulder against the film cassette, with the body and the affected shoulder forming an angle of approximately 60°. The cassette is against the ante-rolateral region of the affected shoulder. The central ray is directed tangentially along the posterolateral margin of the rib cage in line with the scapular spine and perpendicular to the cassette (Fig. 1.82). This positioning produces a true lateral image of the scapula and the glenohumeral joint, demonstrating the position of the humeral head relative to the glenoid fossa. The scapula itself appears as a Y with the scapular body forming the vertical portion. The two upper limbs are formed anteriorly by the coracoid process and anteriorly by the scapular spine and acromion. The glenoid fossa lies at the center of these three lines and the humeral head will normally appear centered within it. In an anterior dislocation, the humeral head will appear inferior to the coracoid process in front of the glenoid fossa; in a posterior dislocation it will be displaced posteriorly relative to the glenoid fossa (Fig. 1.83a–c). Fracture dislocations involving the greater tuberosity can be seen in this view, although fractures of the anterior or posterior glenoid margin will not be detected.

Figs. 1.81 a–c Moloney lines for evaluating the transthoracic radiograph, (a) Normal glenohumeral articulation. (b) Posterior dislocation showing an acute angle. (c) Anterior dislocation showing an obtuse angle.

Fig. 1.82 Positioning for the scapular Y view.

Figs. 1.83a–c Evaluation of the scapular Y radiograph. Normal articulation (a), posterior dislocation (b), anterior dislocation (c).

Velpeau view. This view is a further modification of the axillary projection. It can be used to obtain an axillary view in a patient with a shoulder immobilized in a sling, without abducting the arm. With the arm in a Velpeau bandage or sling, the patient stands with his or her back to the X-ray table and leans backward approximately 30° so that the affected shoulder lies above the film cassette placed on the table. The central ray is centered on the glenohumeral joint in a craniocaudal projection (Fig. 1.84). This technique clearly demonstrates the position of the humeral head relative to the glenoid fossa. Interpretation of the radiograph requires experience as it significantly distorts the size of the humeral head and glenoid fossa. This distortion does not diminish the value of this technique to demonstrate a dislocation.

Cuillo axillary view. This technique allows the supine positioning of a trauma patient without removing the arm from a sling. The patient is supine with the arm internally rotated and the elbow resting on radiolucent cushion material so that the shoulder is flexed approximately 20°. The central ray is aimed at the acromioclavicular joint, entering through the axilla (Fig. 1.85). This view demonstrates the anterolateral margin of the glenoid particularly well since it is unobstructed by interposed tissue.

Fig. 1.85 Positioning for the Cuillo view.

Neer classification (Fig. 1.86) has become an established method for distinguishing fractures of the proximal humerus. This classification uses fragment identification and considers anatomic and biomechanical forces.

Fig. 1.86 Neer’s four-segment classification Group I (minimal or no displacement not shown).

– Fracture of the anatomic neck of the humerus

– Fracture of the lesser tuberosity

– Fracture of the greater tuberosity

– Fracture of the surgical neck of the humerus

Injuries in which the fragments are either separated more than 2 cm or are at an angle exceeding 45° are referred to as displaced fractures; injuries involving lesser degrees of displacement are referred to as non-displaced or one-segment fractures. There are fracture dislocations when the head is displaced outside the joint space, not merely rotated.

Clavicular Fractures

Neer divides clavicular fractures into three groups:

– Fracture in the middle third (80% of all fractures)

– Fracture in the lateral third (15% of all fractures)

– Fracture in the medial third (85% of all fractures)

Fig. 1.87 Classification of fractures according to Ideberg.

Ideberg subdivides glenoid fractures themselves into five groups (Fig. 1.87).

Special Views of the Acromioclavicular Joint

To best image the acromioclavicular joint, remind the X-ray technologist to reduce peak kilovoltage by about 50%. This will significantly improve the image quality because the acromioclavicular joint is overexposed at standard X-ray unit settings.

Zanca view of the acromioclavicular joint. Often the acromioclavicular joint will be obscured by the scapular spine in the AP view. Possible solutions include an abduction or Zanca view. Aim the central ray at the acromioclavicular joint and angle it 10° cranially (Fig. 1.88). This technique is often the only way to demonstrate changes in the lateral clavicle, acromion, or coracoclavicular ligaments.

Acromioclavicular stress view. This is used to confirm separation of the acromioclavicular joint that presents with superior displacement of the lateral end of the clavicle. The patient sits with the back against an upright film cassette. The cassette should be wide enough to image both acromioclavicular joints simultaneously. Both arms are weighted with about 10 kg. The weights are suspended by wrist straps, not held by the patient. This eliminates any stabilizing effect of the shoulder muscles on the acromioclavicular joints. If the acromioclavicular joint is separated, the clavicle on the affected side will be superiorly displaced relative to the acromion and coracoid process (Figs. 1.89a, b). This study is very painful in the case of an acutely injured acromioclavicular joint.

Alexander lateral projection of the scapula. Patient positioning, image size, and film cassette are the same as for the lateral scapula Y projection. The patient attempts to thrust both shoulders as far forward as possible. Films are taken ofthe acromioclavicular joint in a relaxed and forward thrust position. In a separated acromioclavicular joint, the acromion will be visible anteriorly and inferiorly under the distal end of the clavicle when compared with the relaxed position view (Fig. 1.90). The uninjured side can be used as a comparison view.

Fig. 1.88 Positioning for the Zanca view.

The Rockwood classification of dislocations in the acromioclavicular joint is as follows (Fig. 1.91).

• Type I: strain of the acromioclavicular and coracoclavicular ligaments with pain and localized swelling. The ligaments are intact; the acromioclavicular joint is stable.

• Type II:tear of the acromioclavicular ligament and strain of the coracoclavicular ligaments. Instability in the AP plane. Minimal upward displacement of the distal clavicle.

• Type III:tear of the acromioclavicular and coracoclavicular ligaments with significant superior displacement of the lateral end of the clavicle.

• Type IV:same as type III but with posterior displacement of the lateral end of the clavicle.

• Type V:exceeds type III with massive dislocation.

• Type VI:rare form with dislocation of the clavicle under the acromion accompanied by intense swelling.

Fig. 1.91 Rockwood classification of AC joint injury

Rockwood sternoclavicular view: This is also called the serendipity view, as Dr. Rockwood incidentally noted the value of this projection. The patient is supine on the X-ray table, with the arms at the sides and palms facing the table. A 28x35 cm film cassette is placed beneath the patient’s neck and shoulders. The X-ray is angled 40° cranially from vertical, with the central ray entering at the sternal angle. Attempt to project at least the medial halves of both clavicles onto the middle of the film. The distance between the X-ray tube and film cassette should be 100 cm in children and 140 cm in adults (Figs. 1.92a-d). Normally, both clavicles will be located in the same horizontal plane. Anterior deviation of a clavicular axis is a sign of an anterior dislocation of the joint, posterior deviation of posterior dislocation. Degenerative or other processes can also be imaged in this projection, although tomography will provide more information in many situations (Fig. 1.93).

Figs. 1.92a-d (a) Rockwood sternoclavicular projection, (b) Normal sternoclavicular articulation, (c) Anterior dislocation, (d) Posterior dislocation.

Hobbs sternoclavicular view. In this technique, the patient stands at the end of the X-ray table, bending forward so that the cervical spine is almost parallel to the table. The central ray is aimed at the sternoclavicular joint from posterior (Fig. 1.94).

Dislocation most frequently occurs anteriorly; posterior dislocation is rare. The injuries are roughly classified as follows:

– Mild sprain

– Moderate sprain (subluxation)

– Severe sprain (dislocation)

1.4 Ultrasound

Indications, Diagnostic Value, and Clinical Relevance

Ultrasound extends the range of shoulder imaging modalities and has demonstrated its value in diagnosing rotator-cuff lesions and shoulder instability. Ultrasound and plain radiography are primary imaging modalities that complement each other and they can often confirm tentative clinical diagnoses. Ultrasound studies alone are often able to confirm a complete tear of the rotator cuff. They can be used to verify and quantify fragment dislocation in avulsion fractures of the tuberosities.

Normal Findings

Certain transducer positions have been defined in the interest of standardizing the examination technique. These positions are named according to the anatomic orientation relative to the musculoskeletal system. A complete examination of the shoulder should be performed using the following standard projections:

– Anterior transverse plane

– Coracoacromial window

– Lateral coronal plane

– Posterior transverse plane

Fig. 1.95 Anterior transverse imaging plane.

Anterior transverse plane. The anterior transverse plane is located anterior to the shoulder as a transverse section taken through the bicipital groove. This projection demonstrates the medial portions of the subscapularis tendon as far as its insertion at the lesser tuberosity and, laterally, the long biceps tendon. The long biceps tendon is evaluated in the bicipital groove that is bounded medially by the lesser tuberosity and laterally by the greater tuberosity (Fig. 1.95). The long biceps tendon appears as an echogenic structure; usually the transducer must be inclined slightly to obtain hyperechoic image. Within the bicipital groove, the tendon courses not quite parallel to the surface of the skin. These portions of the tendon can only be imaged perpendicular to the direction of sound propagation by tilting the transducer laterally to vary the angle of incidence until the tendon is visualized as an echogenic structure. The dynamic action of the subscapularis tendon should be evaluated in external and internal rotation.

The coracoacromial window. The coracoacromial window is located by placing the transducer on the contour of the shoulder lateral and parallel to the axis of the coracoacromial ligament. In this imaging plane the rotator cuff can be evaluated initially with the arm in a neutral position, then in internal rotation with extension, and finally with the arm in external rotation. The coracoid process and the convex arc of the humeral head with its typical posterior acoustic shadow appear as echogenic structures along the boundaries of the coracoacromial window; these structures serve as bony landmarks. Above the echogenic contour of the humeral head lies echogenic articular cartilage; above this is the rotator cuff, appearing as a moderately hyperechoic region. The upper boundary layer of the cuff cannot be distinguished from the lower lamina of the subdeltoid bursa; these two structures appear as a single echogenic arc of reflection. A narrow hypoechoic halo is discernible above this combined boundary layer, corresponding to the actual subdeltoid bursa. The upper lamina of the bursa combines with the lower fascia of the deltoid muscle to form a sharp, echogenic arc of reflection. This typical sequence of layers is referred to as a wheel pattern, with the contour of the humeral head forming the rim, the rotator cuff the tire, and the line of the bursa representing the tread (Fig. 1.96).

Fig. 1.96 Coracoacromial window.

Different portions of the rotator cuff can be examined through the coracoacromial window by changing the rotation of the arm. The long tendon of the biceps is defined as the border structure between the supraspinatus and subscapularis tendon. The subscapularis tendon lies medial to the biceps tendon, and the supraspinatus tendon lateral to it. The examination in internal rotation with extension brings the insertion of the tendon into the imaging plane in front of the bony margin of the acromion. In normal ultrasound anatomy, the supraspinatus tendon will appear as a band-shaped, moderately echogenic structure above the narrow, hypoechoic layer of articular tissue and the hyperechoic contour of the humeral head.

Lateral coronal plane. In the lateral coronal plane, the transducer lies medially on the acromion and laterally on the deltoid. Particularly in abduction movements, this position permits evaluation of the gliding action of the supraspinatus tendon as well as imaging of the lateral portion of the subdeltoid bursa. It is also used for documenting craniocaudal hypermobility. The typical hypoechoic boundary layers of the bony margins of the acromion and humeral head with its posterior acoustic shadow serve as anatomic landmarks. The moving supraspinatus tendon lies laterally on the humeral head, extending to its insertion on the greater tuberosity. The halo of articular cartilage is visible as a hypoechoic region beneath the tendon. The echo structure formed by the subdeltoid bursa and the lower deltoid fascia is clearly discernible above it (Fig. 1.97).

Fig. 1.97 Coronal lateral imaging plane.

Posterior transverse plane. In the posterior transverse plane, palpate the scapular spine and place the transducer immediately inferior to it, aiming the transducer horizontally. The hyperechoic bony boundary layer of the posterior glenoid and the contour of the humerus should be sharply distinguished in the image. The triangular hyperechoic glenoid labrum rests on the margin of the glenoid. As in the wheel pattern in the anterior imaging planes, the musculotendinous portion of the infraspinatus muscle lies on the contour of the humerus in the posterior transverse plane. Moving the arm into external rotation permits static and dynamic imaging of this tendon as far as its insertion on the greater tuberosity (Fig. 1.98).

Pathologic Findings in the Presence of Degenerative Changes

Changes in the Subacromial Bursa

Irritation of the subdeltoid bursa and subacromial bursa accompanies many changes in the subacromial space. This irritation will appear in the sonogram as a hypoechoic region correlating with an effusion. Extensive accumulations of fluid cause what is known as relative enhancement behind the affected area due to the reduced attenuation in the fluid. Acute bursitis usually leads to massive effusion with clearly discernible separation of the bursa laminae (Fig. 1.99).

Tear of the Rotator Cuff

Tears in the rotator cuff are divided into two categories. Safe geometric tear signs are differentiated from unsafe structural tear signs that specifically include changes in the echogenicity of the tendons (Table 1.9).

• Safe geometric tear sign
– Missing rotator cuff. It has become common to refer to a “bald head” in describing a complete tear of the rotator cuff. In the coracoacromial window sonogram, the typical wheel pattern of the cuff will be missing above the contour of the humeral head; the deltoid is in direct contact with the contour of the humeral head (Fig. 1.100). These findings are particularly striking when the opposite shoulder is imaged. This is a sign of a complete rotator-cuff tear. In the anterior transverse plane, positive diagnosis of a complete tear of the subscapularis tendon is also possible. The supraspinatus tendon tear is characterized in the lateral imaging plane by direct contact between the deltoid muscle and the contour of the humeral head; the contour of the cuff will be missing completely.

Table 1.9 Safe and unsafe signs of rotator cuff pathology in ultrasound

Fig. 1.100 “Bald” humeral head with massive effusion beneath the deltoid. The contour of the rotator cuff is missing completely.

– Segmental narrowing with reversed contour. Segmental narrowing and flattening of the rotator cuff by over 50% compared with the contralateral side, combined with a dip in the normally convex contour, is also to be regarded as a sign of a tear. However, this sign is not as clear as the presence of a completely “bald head.” Segmental narrowing of the cuff is referred to when the tendinous plate medial and lateral to the affected area is of normal width. The decisive criterion is the reversed contour of the cuff boundary, i. e., the region with a concave depression in the tendon structure (Fig. 1.101).

– Abrupt change in diameter. Narrowing may take the form of an abrupt change in the diameter of the supraspinatus tendon, extending laterally to the point of insertion.

– Step sign. Less frequently, a step sign is encountered in which the hyperechoic boundary of the bursa is interrupted. This is a sign of a tear in the rotator cuff. Such a sign is seen in either a partial or complete tear of the rotator cuff and is primarily observed with traumatic tendon defects.

• Inconclusive structural signs of a tear and changes in echogenicity
Changes in the echogenicity of the tendon structure that are not accompanied by changes in geometry are inconclusive signs that may indicate a tear in the rotator cuff. Positive diagnosis of a tear in the rotator tendons requires a change in tendon geometry; a change in echogenicity alone is not sufficient.

Hyperechoic zones in the supraspinatus tendon are best detected in the coracoacromial window in internal rotation with extension (Fig. 1.102). It is essential to distinguish such a hyperechoic zone from a central echo that frequently occurs in the normal tendon as it is imaged horizontally, and from the long head of the biceps tendon coursing beneath it. Dynamic examination is helpful here. The hyperechoic zone should always be demonstrated in a second imaging plane. For example, after the initial examination of the coracoacromial window, imaging findings should be confirmed in the longitudinal section of the supraspinatus. A hyperechoic zone may be a sign of degenerative changes in the rotator cuff. Other possible causes can include inflammatory processes such as calcific tendonitis, tendon fibrosis, or scarring from a tear in the rotator cuff accompanied by formation of granulation tissue.

When evaluating changes in echogenicity be aware that ultrasonography does not permit histologic differentiation. Ensure that the description of image data is strictly separated from the final evaluation of these findings.

A combination of a hyperechoic and hyperechoic zone is encountered far more frequently than an isolated hyperechoic area. This phenomenon can occasionally occur in calcific tendinitis with accompanying inflammation of portions of the rotator cuff. Hypoechoic zones in the rotator cuff can occur on the bursa side, joint side, or within the tendons (Fig. 1.103). They are frequently encountered together with geometric changes.

Fig. 1.103 Echogenic zone within the rotator cuff.

Tendinitis of the biceps tendon. Tendinitis of the biceps tendon usually results in increased accumulations of fluid in the tendon sheath. The resulting widening of the tendon sheath is detectable in the anterior transverse sonogram as a hypoechoic halo around the hyperechoic sheath. This phenomenon is referred to as a biceps halo (Fig. 1.104). A biceps halo can also occur without any irritation of the tendon when effusion into the shoulder joint continues into the tendon sheath. In this case the condition is attributable to pathology of the glenohumeral joint itself.

Pathologic Findings in the Presence of Instability

Ultrasound studies of bony surfaces for diagnosing instability require different settings than for soft tissue, where gray-scale values are important. Bony margins produce maximum reflection, resulting in a hyperechoic image. Sonograms for diagnosing the stability of the glenohumeral joint are obtained in two imaging planes. Anterior and posterior can be examined in the posterior transverse plane. Inferior instability can be verified by a stress test in the lateral coronal plane. Systematic examination of active and passive joint motion can provide information on the direction of instability.

Dislocation

The extent and direction of instability can be determined by changing the normal relation between the humeral head and the glenoid. In acute or chronic dislocation, the position of the humeral head can be imaged in relation to the glenoid fossa. Since no special positioning is required, the patient is not only spared radiation exposure but also the painful positioning needed to obtain radiographs. Pregnant patients can be examined perfectly safely. In posterior dislocations, the humeral head will lie significantly posterior to the posterior margin of the glenoid fossa (Fig. 1.105). Imaging the contralateral shoulder will confirm the diagnosis, even for an inexperienced examiner. In an anterior dislocation, the posterior margin of the humeral head will lie anterior to the posterior margin of the glenoid. Voluntary dislocations can be documented under ultrasonographic control (Fig. 1.106).

Fig. 1.106 Anterior dislocation of the shoulder in the posterior horizontal imaging plane (right) compared with the normal contralateral side (left).

Inferior instability. Inferior instability is evaluated in the lateral coronal plane; a groove test is performed under ultrasonographic control. The hyperechoic contour of the acromion and the contour of the humeral head, which disappears medially in the acoustic shadow of the acromion, serve as bony anatomic landmarks. The distance from the superior margin of the acromion to the contour of the humerus is initially measured with the arm in a neutral position. A distal stress is then applied to the relaxed arm as in clinical testing for the sulcus sign and the results are documented in a second sonogram (Fig.1.107). Occasionally, the humeral head can be moved more than 2 cm out of the glenoid fossa. The sulcus test under ultrasonographic control is more sensitive and objective than the clinical test which, when positive, is itself characteristic proof of multidirectional joint instability. In demonstrating inferior instability and permitting quantitative documentation of findings, the ultrasound test is important in the differential diagnosis of instability because it can distinguish multidirectional instability from unidirectional forms.

Hill-Sachs Lesions

Ultrasound can detect the compression fracture in Hill-Sachs Lesions. The shoulder is first examined in the posterior transverse plane with the arm in a neutral position, and the convexity of the humerus is examined for irregularities and interruptions in its contour. Here, pay particular attention to the layer of cartilage. Continue the ultrasound examination by instructing the patient to slowly lift the arm to 90° flexion while you guide it. This allows examination of an entire posterosuperior humeral convexity. Typically, a Hill-Sachs lesion can be recognized as an approximately triangular depression in the contour of the humerus with an echo at its base (Fig. 1.108). Compression in the bone should not be confused with the physiologic transition between the greater tuberosity and the humeral head. Ultrasonography permits exact measurement of the Hill-Sachs lesion.

Other Pathologic Findings
Fractures of the Humeral Head

Ultrasound can be used as a supplementary modality in support of radiographic studies in fractures of the humeral head. These studies concentrate primarily on the rotator cuff when secondary injuries to the cuff are suspected. However, ultrasound can also be used in evaluating the fracture itself, particularly in the case of isolated avulsion fractures of the greater tuberosity. Traction of the infraspinatus and supraspinatus causes the greater tuberosity to migrate super-oposteriorly. Because imaging can be freely directed in ultrasound studies, the course of the fracture cleavage can be followed during the imaging processes. This makes it possible to precisely evaluate the relationship between the fracture fragments. Step-offs in the surface contour of the humerus are more readily discernible on the sonogram than on the radiograph. Frequently the fragment of the greater tuberosity will be seen to have migrated superiorly; comparison with the opposite shoulder will reveal significant superior displacement and a step-off in the contour of the humeral head. The narrow band of soft tissue contracted toward the acromion will be discernible even with the arm in a neutral position. In abduction, the fragment of the greater tuberosity presses the supraspinatus tendon against the inferior aspect of the acromion. Here, ultrasound studies are a valuable supplement to standard radiographs.

Fig. 1.109 Intra-articular effusion in the posterior horizontal imaging plane.

Intra-articular and extra-articular effusion. Intra-articular effusion of the glenohumeral joint is characterized by a hypoechoic expansion of the capsular space, primarily in the posterior transverse plane (Fig. 1.109). The fluid is often conducted to the sheath of the biceps tendon, causing the familiar biceps halo. In the presence of shoulder irritation of uncertain cause, ultrasound studies can differentiate between intra-articular and extra-articular fluid accumulations.

Intra-articular effusion with hypoechoic expansion of the intra-articular space can result in the rotator cuff being lifted off the contour of the humeral head. This superior migration can be identified in the lateral coronal plane. Extra-articular fluid accumulations lie above the rotator cuff near the transducer and can be distinguished from the intra-articular space (Fig. 1.110).

Fig. 1.110 Suppurative arthritis in the coracoacromial window.

1.5 Arthrography

Indications, Diagnostic Value, and Clinical Relevance

For a long time, arthrography was the only diagnostic modality available for imaging the intra-articular space and the surrounding soft-tissue structures of the shoulder. The method was first described in 1933 and most examiners injected air as a contrast medium. During the 1950s a single-contrast process using a water-soluble contrast medium became popular. The first double-contrast arthrographic studies of the shoulder were described in 1942. However, this technique was slow in becoming established. While the double-contrast method makes it possible to visualize more structures than in single-contrast arthrography, the numerous overlapping structures often make evaluation quite difficult. Occasionally, the double-contrast technique can be used to obtain more detailed information on such phenomena as intra-articular loose bodies and to evaluate articular surfaces or the underside of the rotator cuff. Bursography of the subacromial bursa is also performed. Arthrography may be indicated for both traumatic and degenerative changes (Table 1.10). Invasive arthrography should come at the end of the diagnostic imaging sequence.

Serious complications are rare in arthrography. A known history of allergy to contrast media can be a contraindication to arthrography, as can the presence of an acute infection in the shoulder or surrounding soft tissues. Genuine contrast-medium allergies are extremely rare in shoulder arthrography. One author found only 318 incidents in a retrospective evaluation of 126000 arthrographic studies. No deaths were observed. Serious complications included one air embolism, four cases of severe hypotention, one case of laryngeal edema, and three instances of infection. The most frequent complications included chemical synovitis (0.15%), vasovagal reactions (<0.1%), and urticaria (<0.01%).

Table 1.10 Indications for arthrography

In the AP projection with the arm in external rotation, the contrast medium surrounds the articular surface of the humeral head, extending as far as the insertion of the capsule at the anatomic neck. Since the subscapularis tendon is taut in external rotation, the subscapular bursa will only partially fill. The sheath of the long head of the biceps tendon is located on the lateral portion of the humeral head and is also filled with contrast medium. The biceps tendon itself causes a filling defect in the tendon sheath. The image in internal rotation shows the humeral head surrounded by contrast medium, and its hemispheric shape is clearly visible. The inferior capsular pocket is filled with contrast medium. The biceps tendon and the filled tendon sheath now lie further medially, and their course can be followed distally (Fig. 1.111). In abduction, the contrast medium is again pressed out of the subscapular bursa and in particular the inferior recess that is now under tension. Both structures are less discernible. The axial view clearly shows the filled subscapular recess at the base of the coracoid process. The filled sheath of the biceps is visualized, as is the posterior recess of the joint that is visible in this position because the joint capsule is flaccid. The contrast medium flows around the articular surfaces of the glenoid fossa and the humeral head. The inferior recess is compressed and is not visible because the joint capsule is under tension. The anterior and posterior portions of the glenoid labrum appear as filling defects.

Fig. 1.112 Contrast medium escapes from the glenohumeral joint into the subacromial bursa in a complete rotator cuff tear.

Leakage of contrast medium from the glenohumeral joint into the subacromial bursa is a clear sign of a complete tear of the rotator cuff. Arthrography has 95–100% accuracy in diagnosing complete tears of the rotator cuff. In addition to visualizing the joint cavity, it also visualizes the subacromial bursa. Soft-tissue technique is particularly suitable for imaging contrast-medium leakage (Fig. 1.112). However, neither the rotator cuff itself nor the tear can be clearly imaged. Double-contrast techniques may be helpful here.

If the bursa fills only as the shoulder is moved, it may be presumed that the tear is small. Occasionally, there will be a space between the bursa and the joint space that is free of contrast medium. This shows that rotator-cuff tissue is still interposed between the head and the acromion, meaning that the tear is not yet severe (“bald” humeral head). More extensive chronic rotator-cuff tears can lead to degenerative processes in the subacromial space. This will be seen on the plain radiographs as superior displacement of the humeral head. Occasionally, the joint space of the acromioclavicular joint will communicate with the subacromial bursa. This will lead to what is known as the “geyser sign” in arthrography (Fig. 1.113).

Fig. 1.113 Contrast medium escapes into the subacromial bursa, spreading as far as the acromioclavicular joint (geyser sign).

The diagnostic value of arthrography in evaluating complete tears of the rotator cuff depends on various factors. These include the time that the image documentation was obtained, radiographic parameters such as screen-film system, image definition, clarity of detail, contrast density, and radiographic projections. After the successful repair of a rotator-cuff tear, contrast-medium leakage can be demonstrated. For this reason, an arthrogram after rotator cuff surgery with persisting symptoms will not be conclusive.

Partial Tear of the Rotator Cuff

Partial tears of the rotator cuff that are limited to the bursal side or an intratendinous tear cannot be demonstrated with arthrography. Since the defect does not go through the entire thickness of the tendon, there is no communication with the subacromial bursa and no leakage of contrast medium into the bursa. However, most partial tears lie on the under-surface of the tendon and can be detected in an arthrogram. Irregularities of the underside will appear. If severe degeneration is present, contrast medium will also leak into the substance of the rotator cuff, and small depots of contrast medium will appear in the tendon itself. In such cases, it is important to move the shoulder through a range of motion after injecting contrast medium to press the contrast medium into the defects in the tendon (Fig. 1.114). These ulcerated or crater-shaped contrast-medium depots in the tendon generally appear laterally at the greater tuberosity and are referred to as a “rooster’s comb” phenomenon.

Fig. 1.114 Partial inferior tear of the rotator cuff showing contrast medium escape into the tendinous tissue.

Intra-articular placement of the needle is difficult in adhesive capsulitis. The normal capacity of the joint space is reduced to 5–6 mL. Attempting to expand the fibrotic joint capsule any further will often cause pain. In its early stages, the disorder is visualized as a furrowed capsular margin with near normal filling volume. In the later stages, the joint space becomes progressively smaller, and the subscapular recess and axillary recess merge. The AP radiographs show tight capsular ligaments. In external rotation, the inferior recess will appear very small and will occasionally be missing entirely (Fig. 1.115). In internal rotation, the subscapular bursa and the extremely tight sheath of the long head of the biceps tendon will fill slightly with contrast medium. The biceps tendon will appear lateral and slightly superior to the humeral head and is easily confused with a rotator cuff tear.

Therapeutic arthrography. In adhesive capsulitis, one may attempt therapeutic arthrography. The goal of this procedure is to improve the range of motion of the shoulder by distension arthrography with precisely targeted injection of local anesthesia (possibly in combination with steroids), and to improve this range of motion with a subsequent program of specific mobilization. However, this technique has not proven very valuable since capsular rupture almost always occurs at the subscapular bursa. While this region represents the weak point of the capsule, it is not the reason for limited motion in adhesive capsulitis.

Plain-film radiography, CT, and MRI usually focus on a single area. Nuclear medicine imaging offers significant advantages because it can visualize the entire body simply, quickly, and cost-effectively with comparatively little radiation exposure. However, despite significant improvements in nuclear medicine imaging, it cannot approach the detailed anatomical visualization provided by CT or MRI. The skeleton and joints are not rigid or static systems, rather they are in dynamic equilibrium where metabolic changes can frequently be visualized before the onset of structural changes.

With inflammatory disorders, different radio-pharmaceuticals can be used to provide a wide range of information about joint function. A broad range of radiopharmaceuticals with widely varying uptake is available for imaging inflammatory and infectious processes. Examples from the group of tracers with unspecific uptake (for visualizing exudates or phagocytosis) include gallium-67 citrate, technetium-99m Nanocoll, or technetium-99m-labeled immunoglobulins. When infection is suspected, more specific metabolic visualization is possible using leukocyte scans, for example with autologous separated granulocytes, labeled with indium-111 or technetium-99m-Hexa-methyl propyleneamine oxime (HMPAO). In place of this relatively complicated multi-step procedure, a comparatively simple method can now be used. This involves the use of a technetium-99m-labeled monoclonal murine antibody against the granulocyte surface antigen NCA 95 (Fig. 1.116). This method requires only intravenous injection of the labeled antibody. It also permits specific identification of granulocyte infection foci. In osteomyelitis, the sensitivity and specificity are equivalent to the multi-step procedure. Other radiopharmaceuticals are beyond the scope of this discussion. Detailed consultation with the nuclear medicine department is always appropriate.

Fig. 1.116 Chronic osteomyelitis of the left tibial head. Two different radiopharmaceuticals are used. The weak, unspecific uptake pattern in the technetium-99m Nanocoll image (left) tends to exaggerate the extent of the infection. The intensive uptake in the image obtained with technetium-99m-labeled monoclonal antigranulocyte antibodies (right) clearly demarcates the focus of the infection.

Degenerative joint changes or osteoarthritis are a frequent clinical indication for radiologic examination. Degenerative changes can be observed in varying severity in almost every older patient and are usually an incidental finding during nuclear medicine studies. Degenerative processes lead to osteoblastic activation and reactive bone formation, which can be detected early in nuclear medicine studies (Fig. 1.117).

Although these changes are usually incidental findings, they must be distinguished from other causes of bone metabolism such as metastases, fractures, or inflammatory changes. Parameters such as the phase of uptake, the intensity of uptake, the topographic location of increased metabolism in detailed images (pinhole or SPECT), and correlation with radiographic findings help clarify the etiology of these changes.

Occasionally, nuclear medicine studies are performed to document the metabolic activity of lesions identified on radiographs where degenerative causes must be distinguished from other causes of arthropathy.

The most intense increase in bone metabolism often occurs in the region of subchondral cysts where pressure is high, in areas of bone remodeling, and less so in matured osteophytes. Therefore, it is not surprising that the intensity of increased uptake does not always correlate with the most prominent site of degenerative changes in the radiograph. However, the intensity of metabolic activity in bone scans may correlate with the clinical pattern of symptoms (Merrick, 1992).

Nuclear medicine can provide important information in the evaluation of osteoarthritis in the temporo-mandibular joint (Fig. 1.118) and the small vertebral joints, for example in the evaluation of facet syndrome and back pain of uncertain cause (see also chapter 8, p. 380, Spine) (Holder, 1990; Holder et al, 1995.)

Inflammatory Joint Disorders and Joint Infection

Diagnostic imaging studies play a special role in the evaluation of rheumatic joint disorders. Individual imaging modalities are important in answering the following questions:

1. Is an inflammatory joint disorder present?

2. Which joint disorder is present?

3. How big is the affected area and what is the pattern of involvement?

4. How intense is the inflammatory activity?

5. Are complications present?

6. How is the disorder progressing, particularly under medication?

7. Is surgical intervention indicated?

8. Are postoperative complications present?

Nuclear medicine studies can provide important diagnostic information with regard to several of these questions (Kaye, 1990).

Where clinical and laboratory findings are inconclusive, nuclear medicine studies can detect an inflammatory joint disorder in its early stages before any changes appear in radiographs. Bone scans should always be performed as three-phase studies in this setting since the early phase in articular images shows inflammatory infiltration and synovial activity (Fig. 1.119). Positive findings in such studies are an important sign of the inflammatory nature of the disorder.

Fig. 1.118 Severe arthritis in the temporomandibular joint.

The examination should be designed to provide an overview of the state of inflammatory activity in as many joints as possible in the early phase. Inaccurate information could be obtained if clinicians rely only on the bone phase of a study. The bone phase visualizes the reactive bone metabolism, which however will respond to synovial inflammation at an early stage. This is important e.g. in the early diagnosis of sacroileitis (see chaps 8 and 9, Spine and Pelvis).

The early phases can provide important information in the presence of arthropathy of uncertain cause by differentiating between inflammatory and noninflammatory disorders. However, this distinction is not completely accurate as active chronic osteoarthritis can simulate acute arthritis. Since the field of view is limited in the first two phases (or at least in the first phase), the clinically more significant joint is imaged. If no one joint is primarily affected, the hands are imaged for documentation purposes since they are a frequent site of the disorder.

Aside from the high sensitivity of nuclear medicine studies, visualization of the entire skeletal system offers a major advantage. Radiologic studies focus on one area or joint; nuclear medicine studies can document all areas or joints with relatively short examination times. Often significantly more areas or joints are affected than clinical findings would indicate.

Identification of specific disease entities is not possible. Only the pattern of involvement may provide information on the underlying arthritic condition. Usually the cause of the increased uptake can be identified with greater certainty by other means.

Fig. 1.119 Florid rheumatoid arthritis. The vascular phase (5 minutes after injection) visualizes the inflammatory activity in the individual joints. In the bone phase (3 hours after injection), changes in bone metabolism are visualized as well.

Detection and quantification of the activity of the disorder are crucial to therapeutic management. For this reason, substances have been sought that permit specific quantitative information in addition to the unspecific visualization of synovial inflammation. Blood-pool markers such as technetium-99m albumin, technetium-99m-DPTA, ortechnetium-99m per-technetate have failed to provide significant advantages. Labeled and separated autologous leukocytes provide more specific information that indicates the degree of cellular infiltration in the pannus (Fig. 1.120). Although the uptake mechanism is not known (presumably unspecific bonding to Fc receptors on polymorphic leukocytes, lymphocytes, and macrophages and via increased vascular permeability), the uptake intensity of technetium-99m-labeled polyclonal human immunoglobins correlates well with the degree of inflammation found in histopathologic studies. This method has utility in the detection, evaluation, and monitoring therapy of chronic inflammatory joint disorders (DeBois et al, 1995).

Fig. 1.120 Rheumatoid arthritis in the knee. The unspecific uptake in the bone scan in the upper series of images exaggerates the inflammatory process. In the labeld leukocyte study in the lower series of images, uptake is primarily seen where inflammation has infiltrated the synovia.

Antibodies such as technetium-99m-CD3 anti-T-lymphocytes or technetium-99m-CD4 anti-T-helper lymphocytes detect very specific cellular immune reactions. These substances are currently in the process of further clinical testing.

Rheumatoid joint disorders must be distinguished from septic arthritis and osteomyelitis close to the joint. Nuclear medicine studies can visualize both situations with high sensitivity. However, unspecific activated bone metabolism is seen in both infectious and inflammatory settings. Reactive bone metabolism can also remain once infections have healed. Excluding other causes of increased bone metabolism can also be difficult. For these reasons, specific leukocyte studies are often indicated (Fig. 1.121). This method can achieve accuracy of about 90% outside the axial skeleton. Usually these studies identify and specifically visualize the infected area. Nuclear medicine studies are usually combined with MRI where more precise anatomical correlation of the extent of infection is required, for example in determining joint involvement in epiphyseal and metaphyseal osteomyelitis.

In some settings it may not be possible for nuclear medicine studies to distinguish non-bacterial and rheumatoid inflammation from bacterial infection.

Fig. 1.121 Osteomyelitis of the patella. The bone scan in the right image shows uptake in all bony structures, whereas the labeled leukocyte study in the left image correctly limits the area of increased focal uptake to the infected patella. This excludes arthritis.

Extensive studies have demonstrated the high sensitivity of bone scans in the diagnosis and evaluation of fractures: this method diagnosed 95% of all fractures in patients less than 65 years old within 24 hours of injury. Seventy-two hours after injury it was able to demonstrate 100% of all fractures in patients less than 65 years old and 95% of all fractures in patients over 65 years old (Matin, 1988). More recent data show that these results vary according to location; the speed and probability of diagnosis decrease from the periphery to the axial skeleton. A well-known indication is early confirmation of the diagnosis of a scaphoid fracture in the presence of normal radiographic findings (Fig. 1.122). Fractures of the spine, cranial vault, and hips are difficult to evaluate in older patients with significant osteoporosis.

Bone scans are important both for diagnostic and forensic purposes in excluding or confirming fractures in cases of child abuse. Similarly, the method can be easily used in adults with multiple trauma to detect occult fractures in the presence of clinical findings with negative radiographic findings.

It can be important to determine the age of a fracture, even in a non-forensic setting. Vascular images will return to normal within 3–4 weeks of an acute fracture; soft-tissue activity will be normal after about 8–12 weeks. Bone metabolism changes its configuration and intensity over time. Follow-up examinations at short intervals within 10 days can be used to document quantifiable changes in intensity in acute fractures that will be absent in chronic fractures. After 2 years, 90% of all fractures will have returned to normal; moderately increased metabolism will persist throughout the patient’s lifetime in the remaining fractures. Nuclear medicine studies play an important diagnostic role in evaluating occult fractures, periostitis, stress fractures, and in enthesopathy (Fig. 1.123). These studies may also be indicated to evaluate posttraumatic complications such as delayed fracture healing or malunion, pseudarthrosis, reflex sympathetic dystrophy, avascular necrosis, or infection.

Fig. 1.122 The images shows an acute navicular fracture that did not show up in the radiographs. The soft-tissue phase (upper image) and bone phase (lower image) reveal abnormal findings.

Fig. 1.123 Stress fracture of the tibia.

The pathophysiologic process begins as the result of disturbed circulation, which leads to necrosis of the bone marrow and matrix cells. Repair processes result in revascularization via collateral circulation and reactive bone remodeling or apposition. Disintegration of the articular surfaces due to cartilage destruction may precipitate secondary degenerative changes (Tumeh, 1996). As a result, findings in nuclear medicine studies will vary depending on the stage of the disorder. The early stage is characterized by photopenia. This is often observed in traumatic avascular necrosis of the femoral head. In the revascularization stage, hyperemia can mask the defect and produce false negative findings. However, there will typically be a zone of increased perfusion, soft-tissue activity, and mineralization activity at the boundaries of the necrotic defect that appears as a cup-shaped sign. The repair stage is usually characterized by a circumscribed increase in bone metabolism. Since these changes can be very discreet, studies should be obtained using a multi-phase acquisition technique. Perfectly symmetric patient positioning is needed to permit comparison with the activity of the contralateral joint. Carefully prepared detailed images are important. These are best obtained as pin-hole magnification images and, in large joints, as SPECT images (Fig. 1.124).

Fig. 1.124 SPECT image showing avascular necrosis in the left femoral head. There is a defect in the femoral head and reactive bone metabolism in the growth plate.

Fig. 1.125 Aseptic implant loosening in the left hip. Intensely increased metabolic activity is seen in the entire implant bed (right image), whereas the labeled leukocyte study (left image) reveals no abnormal findings.

Leukocyte studies are a more specific method with significantly higher accuracy in detecting septic implant loosening (Fig. 1.125). This method primarily utilizes autologous granulocytes, which are labeled with indium-111 oxinate or Tc-99m-HMPAO after separation (Palestro et al, 1990). Comparative studies with gallium-67 citrate, which can also detect infection, have shown leukocyte imaging to be the superior method (Schauwecker et al, 1984).

Instead of these quite complicated methods, one can achieve comparable results (i.e., about 85% accuracy) using technetium-99m-labeled monoclonal antibodies. This method is always available, does not require cell separation, and can be performed with a single intravenous injection. Where the bone scan is evaluated simultaneously, this method achieves a sensitivity of about 89% and a specificity of about 84% (Sciuk et al, 1992). Findings should be compared to the bone or marrow scintigram to reduce the rate of false positive findings.

Where aseptic granulocyte accumulation is present in the periprosthetic tissue, one should consider a body rejection reaction, granuloma, hematoma, unspecific reaction to abraded particles from the implant surface, or reaction to metal. Comparison with the bone scan is of little help in these cases because increased bone metabolism may also be expected. Another diagnostic pitfall that can produce false positive findings is periprosthetic islands of bone marrow. Due to their physiologic uptake mechanism, the antibodies will mark these islands of marrow as they bond to the precursor cells of granulo-poiesis. This may be misinterpreted as increased uptake due to infection. Comparing the leukocyte scan with a technetium-99m-sulfur colloid image has been suggested to avoid this type of misinterpretation. However, incongruent visualization may be expected in the presence of infection and congruent visualization in the presence of a marrow island because this colloid will mark bone marrow but not the site of an infection (Palestro et al, 1990). Comparison with the bone scan is also helpful in distinguishing a marrow island from an infection. The marrow island will show normal perfusion, normal soft-tissue activity, and a degree of mineralization identical to that of the rest of the skeleton, whereas bone metabolism will be increased in infected parts of the skeleton. Sites of infection along the femoral components of a total hip implant appear as areas of increased uptake in the leukocyte scan. Infections, and particularly chronic infections in the acetabular component lead to decreased uptake (cold lesions). The cold areas are an unspecific sign of bone marrow depression. They can be regarded as both a sequela of infection and as postoperative destruction of the bone marrow in the acetabulum.

Indications for plain CT are bony lesions that are difficult to evaluate on plain radiographs (Table 1.11). Due to the overlapping structures in the shoulder, determining the exact location and orientation of fragments and possible articular surface involvement is not always possible with radiographs. Plain CT can provide information for planning therapy by precisely visualizing fractures of the proximal humerus and glenoid and by visualizing bony defects after dislocations.

If soft-tissue injuries are suspected, CT arthrography is indicated. The primary indication is a defect in the glenoid labrum. CT arthrography can be used to evaluate the joint capsule. Together with MRI, CT arthrography is the method of choice for visualizing labral lesions. In deciding between imaging modalities, the availability of CT scanning systems compared with MRI scanning systems should be weighed against the disadvantage of CT arthrography as an invasive procedure.

Visualization of rotator cuff pathology is another possible indication. However, modalities such as ultrasound and arthrography are superior due to their lower cost.

Technique, Instrumentation, and Examination Procedure

Plain CT. The patient is positioned supine. To image the joint in a neutral position, instruct the patient to hold the hand flat on the hip. The patient must not place his or her arm on the abdomen because respiratory excursion will produce motion artifacts. For images in internal rotation, the patient places the hand and forearm beneath the buttocks. After obtaining a topographic image, tomograms are obtained from the acromioclavicular joint to the inferior recess. The images are evaluated with both the soft-tissue windows and bone windows.

Table 1.11 Indications for CT

Plain CT or CT arthrography is rarely indicated for degenerative changes in the subacromial space. Calcific tendinitis, rotator-cuff tears, and muscle atrophy are mentioned as indications. The high resolution of radiodense structures with CT allows imaging and localization of calcifications. This may permit differentiation between degenerative and reactive calcifications and formation and resorption phases. It will sharply delineate a deposit and determine its size, location, and density. This information is important for aspiration, but the same information can be obtained from plain radiographs, with the joint in different rotational positions, and from ultrasound studies. CT diagnosis of rotator cuff lesions requires injection of contrast medium (Fig. 1.126). Studies describe discontinuity, asymmetry, and triangular or drawstring configurations as signs of severe damage to the rotator cuff. In this system, exact classification of the cause of pathology is contingent on demonstrated atrophy of the supraspinatus.

Despite the capabilities discussed here, CT is not very helpful in routine diagnostic imaging of rotator cuff disorders. Although leakage of contrast medium into the subacromial bursa can demonstrate a tear in the rotator cuff and provide some preoperative information about the extent of the tear, equivalent diagnostic information can be obtained with less costly examination methods that are more suitable for routine use.

Fig. 1.127 V-shaped posterolateral Hill–Sachs lesion.

Hill–Sachs lesions are reliably demonstrated in CT scans (Fig. 1.127). Even minor Hill–Sachs lesions are discernible. Plain radiographs will not detect changes in the cartilaginous region of the humeral head, whereas CT clearly demonstrates even minor defects. This modality can also be used to determine preoperatively the extent of the compression fracture in three planes so that the surgeon can plan the reconstruction in applicable cases. However, since CT is complicated, costly, and may require invasive testing, it is not commonly used to document Hill–Sachs lesions.

Bankart Lesion

Bony Bankart lesions and capsular calcifications can also be more clearly visualized than in plain radiographs (Fig. 1.128). However, it must be ensured that both shoulders are imaged in the same position. Otherwise it will not be possible to compare the injured shoulder with the normal shoulder. Soft-tissue changes in the anterior capsular ligaments such as a patulous capsule can be demonstrated with contrast medium in CT arthrography. The crucial question of whether the labrum is damaged can only be answered by a CT arthrogram (Fig. 1.129). Bankart lesions can be visualized in this manner, as can sub-periosteal separation of the capsule along the anterior glenoid. The wide range of normal anatomical variants of the glenoid labrum make conclusive evaluation difficult. Failure to visualize the labrum and tears in the labrum at typical locations (anteroinferior glenoid) are relatively conclusive signs. Changes in the superior portions of the labrum are frequently the cause of false diagnoses.

Fig. 1.130 Chronic posterior dislocation.

In a chronic dislocation, the CT will clearly demonstrate the position of the humeral head. The head usually appears locked on the glenoid with an obvious Hill–Sachs lesion. Aside from documenting the dislocation, CT can visualize associated injuries to the articular surfaces (Fig. 1.130).

Other Pathologic Findings

Fractures of the Proximal Humerus

In fractures of the proximal humerus, CT clearly demonstrates the fractures. In this type of injury, the number of fragments is less important in the prognosis than the degree of displacement and rotation. Fragments that are displaced less than 1 cm and rotated less than 45 °minimally influence the prognosis. Fragments widely separated from their surrounding structures increase the risk of avascular necrosis of the humeral head. The number and position of the bone fragments with respect to each other and their position relative to the glenoid can be demonstrated in CT scans (Figs. 1.131 and 1.132). These can provide additional information for prognosis, the clinical course of the disorder, and the treatment.

Fractures of the Glenoid Fossa

Here, CT is superior to conventional radiographs (Fig. 1.133). In addition to showing the number of fragments, it is possible to image the intra-articular course of the fracture line. This information can determine further treatment, the clinical course of the injury, and the prognosis.

When evaluating fractures and the displacement of fracture fragments, it is always useful to obtain scans of the uninvolved side. Three-dimensional reconstruction is possible to image avulsed fragments of the glenoid (Fig. 1.134). However, since these images do not offer any additional advantages for determining treatment, three-dimensional reconstruction has not gained general acceptance.

The most frequent indications of an MR examination of the shoulder are, as for all other joints, the visualization and documentation of traumatic lesions and degenerative changes (Table 1.12). For tumorous conditions, MRI primarily serves to determine the extent of the tumors and to visualize the affected tissue. Determining the type of tumor is rather limited, as is true for all other imaging methods. Except for the usual contraindications, there are no further contraindications.

MRI is equivalent to CT arthrography in evaluating the rotator cuff but is superior for assessing extra-capsular changes. Osseous changes are equally well diagnosed with either modality. Position and size of osseous fragments, however, are better determined by CT, while so-called occult osseous lesions are better detected by MRI.

Technique, Instrumentation, and Examination Protocol

Care should be taken that the patient is positioned comfortably. Normally, the patient lies supine with the arm slightly internally rotated and adducted. Vacuum cushions or foam wedges can be used for stabilization. The resonance signal is generally acquired with surface coils adapted to the shoulder.

The planes are selected to suit the structures to be examined. The shoulder is generally examined in three planes with the glenoid fossa used as the structure of reference. The oblique sagittal sections are parallel to the glenoid fossa and should include the supraspinous fossa for evaluation of the supraspinatus muscle. The transverse and coronal sections are placed perpendicular to this plane. The transverse sections should encompass the acromioclavicular joint. To improve the delineation of the tendon of the supraspintus muscle, the coronal sections must be aligned to the orientation of the tendon, which is determined by the rotation of the humerus and noted on the transverse image. Analogous to the visualization of the other joints, proton density weighted sequences are used for general orientation. The evaluation of pathologic changes requires T2 and T2*-weighted sequences as well as fat suppression. Enhancement after intravenous injection of a gadolinium-based contrast medium is best appreciated on fast GRE sequences or on fat-suppressed T1-weighted sequences. Intra-articular instillation of contrast medium generally is not necessary, though it is favored by some investigators.

In craniocaudad direction, the transverse sections delineate the trapezius muscle and its insertion at the acromion and the clavicle. The acromioclavicular joint, together with its capsule and cartilage, is well visualized on slightly T1-weighted GRE sequences. Using high-resolution techniques, the small intra-articular disk can be identified. The stabilizing ligaments, comprising coracoclavicular, acromioclavicular, and coracoacromial ligaments, are difficult to identify in this plane and lesions can be assigned to the ligaments on the basis of their topography only. Further caudally, the transverse sections delineate the glenohumeral articulation (Fig. 1.135). While the humeral head and the glenoid fossa are covered with hyaline cartilage, the ring-like enforcement of the glenoid fossa – the glenoid labrum – consists of fibrocartilage, which usually blends with the joint capsule. The labrum is of variable thickness. In the middle anterior portion, it appears to sit like at a triangle on the glenoid labrum, separated from it by a layer of hyaline cartilage. Owing to its fibrous structure, it has a lower signal intensity than the hyaline cartilage. In about 20% of cases the capsule inserts medially to the labrum, and in about 10% of cases even more medially at the scapular neck. Between the anterior labrum and the capsule, the capsule is reinforced by the glenohumeral ligament, with its middle portion generally recognized on the transverse sections. The long biceps tendon is seen on cross-section as a thick, almost signal-void structure in the intertubercular sulcus. Likewise, the nearly signal-void structures of the tendinous extension of the infraspinatus and sub-scapular muscles can be followed to their insertion at the major and minor tuberosities. T1-weighted sequences display the fat beneath the deltoid muscle as a fine, high-signal band. T2-weighted images visualize the subdeltoid bursa as a homogenously bright band, especially if a joint effusion is present.

The sections perpendicular to the plane (Fig. 1.136) of the glenoid fossa show in posteroanterior order the infraspinatus muscle and teres minor with their tendon extending to the major tuberosity. Since the glenoid fossa is not always strictly perpendicular to the scapula, the oblique coronal sections must follow the course of the muscle and tendon as seen on the transverse sections for the evaluation of the supraspinatus muscle. The tendon of the supraspinatus muscle can display a medium signal intensity on the T1-weighted images. This increase in signal intensity is imparted by interference of the tendinous structure with the main magnetic field, whenever tendon and field are at a certain angle to each other (“magic angle”). The subacromial fat is seen as a bright band above the tendon. On heavily T2-weighted sequences, the subacromial and subdeltoid bursae are delineated, especially in the presence of an effusion. This plane also visualizes the long biceps tendon, but usually only its distal intra-articular segment. To visualize the tendon in its entirety, the sections must accommodate the intra-articular course, as seen on the transverse sections. To avoid partial volume effects, the section thickness should not exceed 3 mm. The superior and inferior labrum appear rounded in comparison with the anterior labrum. The inferior portion of the capsule is folded when the humerus is in typical adduction.

Fig. 1.135 Transverse image (SE 500/20) showing normal glenohumeral joint.

The third, sagittal, plane, which is parallel to the glenoid fossa, shows, in mediolateral order, the characteristic-shape of the scapula with the supraspinatus fossa, and the subscapular muscle along the anterior surface of the y-shaped scapula (Fig. 1.137). The trapezius is dissected superior to the supraspinatus muscle and separated from it by a thick layer of fat. Inferior and lateral to the infraspinatus muscle, the teres minor extends from the infraspinous fossa and lateral margin of the scapula to the major tuberosity. The teres major originates inferiorly along the lateral scapular margin and attaches to the minor tuberosity. Anteriorly, the tendon of the pectoralis minor is seen converging to the coracoid process, and the pectoralis major converging to the minor humeral tuberosity. Diagnostically important is the section that traverses the acromioclavicular joint, showing the cross section of the subacromial supraspinatus muscle and its tendon. Between the anteriorly located coracoid process and the acromion stretches the coracoacromial ligament. The origin of the long biceps tendon is identified at the anterosuperior circumference of the glenoid (Fig. 1.138). The transition where the tendons of the subscapular, supraspinatus, and infraspinatus muscles form the rotator cuff is further lateral. It retains the humeral head (Fig. 1.139).

Fig. 1.136 Coronal oblique image (SE 500/20) showing normal glenohumeral and acromioclavicular joints.

Impingement is the major cause of degenerative changes involving the rotator cuff. Impingement is associated with morphologic variations of the acromion, especially an inferiorly curved process, by osteophytes that arise anteroinferiorly from the acromion, and by hypertrophy of the acromioclavicular joint. Three stages are distinguished. Stage 1: mild intratendinous edema and possible hemorrhage. Stage 2: tendinitis of the rotator cuff, fibrosis, and thickening of the tendon. An effusion is seen in the acromial and subdeltoid bursa. Stage 3: partial and complete tears of the rotator cuff. These stages correspond to the three grades that categorize the MR findings of the rotator cuff. Grade I: diffuse increase in the signal intensity on the T1-weighted and proton-density image. An increased signal intensity is often better appreciated on the T1-weighted fast SE sequence or on a T2*-weighted GRE sequence. Grade II: increased signal intensity in all sequences, best seen on the T2*-weighted and fast SE images. The supraspinatus tendon is thinned and irregular in outline (Fig. 1.140). Grade III: partial or complete defect in the tendon, with increased signal intensity in all sequences. The tears are also well detected on the T2-weighted images due to intrusion of synovia into the created clefts. Inflammatory tissue is characterized by locally expanded extracellular space. After intravenous injection of gadolinium-based contrast medium, the damaged soft-tissue structures enhance on the T1-weighted images, even more so on ΊΊ-weighted fast GRE sequences. Moreover, the irritated synovial membrane enhances due to hyperemia and intensified secretory activity (Fig. 1.141).

Fig. 1.140 Coronal oblique image (GRE 500/11 out of phase) showing grade II degeneration of the supraspinatus tendon.

Grade I is frequently associated with a decrease in the subacromial fat plane, best seen on the T1-weighted images. On the T2-weighted or STIR image, a small amount of fluid is occasionally observed in the subacromial-subdeltoid bursa. Grade II shows more often a bursal fluid accumulation and an obliterated fat plane. Grade III is, in the case of a completely torn rotator cuff (Fig. 1.142), always accompanied by an effusion in the glenohumeral joint and bursa through the concomitant communication between joint capsule and bursa. With progression of the disease, the supraspinatus muscle retracts, and atrophy and fatty degeneration develop, best diagnosed on the T1-weighted image. Massive tears can lead to inactivity atrophy of the remaining shoulder muscles with resultant instability and upward migration of the humeral head. Vascular changes of the humeral head, contributed by the loss of joint fluid, induce subchondral osseous necrosis (Fig. 1.143), cartilage degeneration, and loss of cartilage. Cartilaginous changes are best recognized on T2*-weighted sequences or on T1-weighted fast SE sequences.

Enthesopathy of the Rotator Cuff

In athletes whose activities include overhead motions, changes at the osseous insertion of the rotator cuff can be occasionally observed. The tendon shows a few areas of increased signal intensities or even tears as found in the more proximally located degenerative tendon changes. This is accompanied by small cortical to subcortical lesions that are of increased signal intensity on the T2-weighted images, and centrally increased signal intensity surrounded by a rim of decreased signal intensity on the T1-weighted sequences. The T1-weighted image delineates these lesions as areas of decreased signal intensity (Fig. 1.144). After intravenous injection of gadolinium-based contrast medium, a distinct central enhancement is usually observed (Fig. 1.145). In addition, the distal segment of the tendon frequently enhances. These changes are summarized under the term “overuse enthesopathy.”

Fig. 1.144 Coronal oblique image (GRE 500/11 out of phase) showing stress necrosis at the insertion of the rotator cuff and partial tear of the supraspinatus tendon.

In severe degenerative changes of the rotator cuff, minor trauma can cause tears of the long biceps tendon (Fig. 1.146). The tear commonly occurs where the tendon leaves the sheath. In most labral lesions with superior anterior to posterior extension (SLAP lesions), the long biceps tendon is detached from the glenoid. The intracapsular segment of the torn tendon can displace the subscapular muscle. The tendon sheath is filled with fluid and no tendon components are found in the intertubercular sulcus. The tear itself generally is often difficult to visualize owing to the curved course of the tendon. An empty sulcus is also observed with medial dislocations of the long biceps tendon.

Calcifying Tendinitis

The calcifications of chronic inflammatory arthritis cannot be directly visualized. Areas with a signal void in all weightings and with artifacts on GRE images, together with increased signal intensities on T2 or STIR sequences as seen in inflammatory tissues, allow the diagnosis of calcifying tendinitis (Figs. 1.147 and 1.148). Postoperative artifacts caused by metallic particles compromise the evaluation considerably.

Fig. 1.148 Coronal oblique image (GRE 500/11 out of phase after application of intravenous gadolinium contrast medium) showing calcification at the insertion of the rotator cuff. Contrast-medium enhancement is seen in the synovial membrane with subdeltoid bursitis.

Severe degenerative changes of the glenohumeral joint are often associated with degenerative changes of the acromioclavicular joint. These changes consist of a thickened capsule with high signal intensity on T2-weighted images, and marked enhancement (Fig. 1.149). Furthermore, an irregularly outlined joint surface and, in long-standing arthropathy, subchondral sclerosis are found in the clavicle and acromion, visualized especially well on GRE sequences. The enhancement observed in these subchondral lesions indicates increased bone turnover (Fig. 1.150).

Pathologic Findings in Instabilities

Changes of the Labrum (Bankart Lesion)

In shoulder instabilities, lesions of the labrum are invariably detectable. They are classified as follows:

– Type I, constituting degenerative changes with increased signal intensity centrally on the proton-density and T1-weighted phase-contrast images.

– Type II (Fig. 1.151), consisting of tears, primarily at the base, again recognized by the increased signal intensity, but also on the T2-weighted image.

– Type III (Fig. 1.152), consisting of tears with dislocation of the labral fragment.

– Type IV (Fig. 1.153), consisting of tears with subperiosteal elevation of the capsule and possible fissure even in the glenoid. Fissures are clearly delineated in fat-suppressed sequences (e. g., STIR).

In athletic activities with overhead motions, the labral lesion occurs commonly in the anterosuperior region and extends posteriorly (SLAP lesion).

Intravenous administration of gadolinium-based contrast medium does not improve sensitivity and specificity of the MR examination for the detection of labral lesions. Enhancement at the base where the labrum rests on the glenoid and in the region of the detached capsule indicates an acute injury.

Fig. 1.152 Transverse image (STIR 1900/125) showing separation of the glenoid labrum with dislocation (type III Bankart lesion).

Tears in the joint capsule are indirectly detected through the leak of synovia, primarily on the T2-weighted image. The irritated synovial membrane, either by trauma or by primary inflammatory processes, always enhances after intravenous administration of gadolinium-based contrast medium (Fig. 1.154).

Hill–Sachs Lesion

This is defined as a compression fracture of the dorsolateral humeral head, generally following an anterior dislocation. The transverse sections show the corresponding impression of the humeral head, with peri-focally decreased signal intensity on the T1-weighted and increased signal intensity on the T2-weighted images. These signal changes are due to bone marrow edema and can even be found at this site without impression, presumably representing trabecular fractures caused by the same pathomechanism (Fig. 1.155 and 1.156).

Fig. 1.157 Sagittal image (SE 600/20 after application of intravenous gadolinium contrast medium) showing an impacted fracture.

Fractures of the humeral head are almost always diagnosed on conventional radiographic examination. So-called occult fractures are disclosed on MRI as linear-to-patchy areas of increased signal intensity, best seen on fat-suppression sequences. After intravenous contrast medium, enhancement is seen along the fracture line or as patchy areas in fractures confined to the trabecular bone. In addition, the accompanying soft-tissue injuries can be visualized by MRI (Figs. 1.157 and 1.158).

Avascular Necrosis

Avascular necrosis is the most frequent complication after fractures of the humeral head (Figs. 1.159 and 1.160). Corticosteroid medication, alcoholism, sickle cell anemia, and caisson disease are nontraumatic causes. Analogous to other joints, the area of necrosis is marked by a well-demarcated decreased signal on the T1-weighted image that extends to the articular surface. The T2-weighted image shows increased signal intensity centrally. Enhancement depends on the age of the necrosis and is peripheral for old and central for acute necrotic lesions.

Tumors

Tumors in the region of the shoulder are well delineated and localized on MRI (Figs. 1.161 and 1.162). A specific diagnosis can only be made with considerable reservation. For instance, a small enchondroma cannot by distinguished with certainty from a fibro-sarcoma, unless the cortex is destroyed. Even with cortical involvement, the differential diagnosis can be difficult if it represents a secondary fracture. Furthermore, the enhancement pattern does not permit a definitive differentiation between malignant and benign tumorous osseous lesions.

Fig. 1.162 Coronal image (SE 500/20) showing a Ewing’s sarcoma of the proximal humerus.