Shoulder and Elbow Injuries

Neer Impingement Sign and Impingement Test

Neer first described the impingement test in 1972 ( Fig. 46.2A ). With the patient seated, the examiner raises the affected arm in forced forward elevation while stabilizing the scapula, causing the greater tuberosity to impinge against the acromion. This maneuver produces pain with impingement lesions of all stages. It also produces pain in many other shoulder conditions, such as adhesive capsulitis, osteoarthritis, calcific tendinitis, and bone lesions. Neer also described the impingement test with the use of a subacromial injection of 10 mL of 1% lidocaine (Xylocaine). Pain caused by impingement usually is significantly reduced or eliminated, but pain caused by other conditions (with the exception perhaps of calcific tendinitis) is not relieved. One study revealed sensitivity for the impingement sign of 75% for bursitis and 88% for cuff abnormalities, with specificities of 48% and 51%. The positive predictive values were 36% and 40%, and the negative predictive values were 83% and 89%. The test was positive, however, in 25% of patients with Bankart lesions, 46% of patients with superior labral anterior and posterior lesions, and 69% of patients with acromioclavicular arthritis.

Hawkins-Kennedy Test

Hawkins and Kennedy described their test in 1980 as an alternative to the Neer test, but they did not believe that it was as reliable ( Fig. 46.2B ). The test is performed by forward flexing the humerus to 90 degrees and forcibly internally rotating the shoulder. This maneuver drives the greater tuberosity farther under the coracoacromial ligament, reproducing the impingement pain. Sensitivities of 92% for bursitis and 88% for cuff abnormalities were found in 85 consecutive patients, with specificities of 44% and 43%. The positive predictive values were 39% and 37%, and the negative predictive values were 93.1% and 90%. The test was positive, however, in 31% of patients with Bankart lesions, 69% of patients with superior labral anterior and posterior lesions, and 94% of patients with acromioclavicular joint arthritis.

Jobe Test

Jobe described the “supraspinatus test” in 1983 ( Fig. 46.2C ). The test is performed by placing the shoulder in 90 degrees of abduction and 30 degrees of forward flexion and internally rotated so that the thumb is pointing toward the floor. Muscle testing against resistance shows weakness or insufficiency of the supraspinatus owing to a tear or pain associated with rotator cuff impingement.

Internal Rotation Resistance Stress Test

Zaslav described a test to differentiate between internal and classic outlet impingement ( Fig. 46. 2 D ). The test is performed with the patient seated and the examiner standing behind the patient. The patient’s arm is positioned in 90 degrees of abduction in the coronal plane and approximately 80 degrees of external rotation. A manual isometric muscle test is performed for external rotation and compared with one for internal rotation in the same position. If a patient with a positive impingement sign has good strength in external rotation and weakness in internal rotation, the test is positive. A positive internal rotation resistance stress test suggests internal impingement, and a negative test (more weakness in external rotation) suggests classic outlet impingement. Zaslav reported a sensitivity of 88%, specificity of 96%, positive predictive value of 88%, and negative predictive value of 94%. All of the patients tested had been selected, however, because they had a previous positive Neer impingement test, resulting in a population of “impingers” rather than a population with unconfirmed shoulder abnormalities.

Dynamic Labral Shear Test

O’Driscoll described the dynamic labral shear test to reproduce the shearing mechanism that can cause a superior labrum tear anterior and posterior (SLAP); it is reported to yield reliable results that isolate a SLAP lesion’s causal role in symptoms and impaired function. The test is performed with the patient supine with the affected arm relaxed off the side of the examining table; the scapula is supported by the table, but the humerus is free. To test the right shoulder, the examiner flexes the right elbow 90 degrees and grasps the olecranon and distal humerus. The shoulder is passively rotated externally to its natural limit with the force of gravity alone pulling down on the forearm (approximately 90 degrees); the elbow is then dropped back into its natural limit of horizontal abduction (toward the floor), and the shoulder is passively elevated and depressed while horizontal abduction and external rotation are maintained. During elevation of the shoulder, the magnitude of horizontal abduction will vary and must be permitted to do so without constraint. The degree of external rotation also will vary throughout the arc of elevation and must be unconstrained. While the shoulder is being elevated, the examiner’s right hand is kept on the acromion to stabilize the scapula and to detect any palpable click transmitted through the bony structures ( Fig. 46.3 ). Usually the click will be felt by the left hand on the olecranon, as well as the distal humerus. After full overhead elevation, the entire motion is reversed and the arm is brought down to the side while the natural limits of shoulder external rotation and horizontal abduction are maintained. The test is considered positive if the patient’s deep pain at the posterior glenohumeral joint is reproduced by the test through an arc from approximately 90 degrees to 120 degrees; it is considered negative if it is not painful in this range or it does not reproduce the patient’s pain. The test can also be performed with the patient sitting or standing, and we have found this to be reliable in the clinical setting. Cheung and O’Driscoll reported sensitivity for diagnosis of a type II SLAP lesion of 86%, which was as good as MRI with gadolinium and better than MRI without gadolinium contrast dye. Kibler et al. found a sensitivity of 72%, specificity of 98%, accuracy of 97%, positive predictive value of 97%, and negative predictive value of 77% for detecting labral disease with this test.

Kim Test for Posteroinferior Labral Tear

This test is performed with the patient sitting with the arm in 90 degrees of abduction. The examiner holds the elbow and lateral aspect of the proximal arm, and applies a strong axial loading force ( Fig. 46.4A ). While the arm is elevated 45 degrees diagonally upward, downward and backward force is applied to the proximal arm ( Fig. 46.4B ). A sudden onset of posterior shoulder pain indicates a positive test result, regardless of accompanying posterior clunk of the humeral head. Because it is important to apply a firm axial compression force to the glenoid surface by the humeral head, having the patient sit against the back of a chair rather than on a stool provides a good countersupport of the axial loading in the involved arm.

In both a study by Kim et al. and a systematic review, the Kim test was found to have a sensitivity of 80% and specificity of 94%. When the jerk test was used in combination with the Kim test, sensitivity increased to 97%.

Jerk Test for Posterior Labral Lesions

The jerk test is highly sensitive and specific for a posterior labral tear. It is performed with the patient sitting. While the examiner holds the scapula with one hand, the patient’s arm is abducted 90 degrees and internally rotated 90 degrees. An axial force is applied with the examiner’s other hand holding the patient’s elbow ( Fig. 46.5A ), and a simultaneous horizontal adduction force is applied ( Fig. 46.5B ). A sharp pain, with or without posterior clunk or click, suggests a positive test result.

Biceps Load II Test for Isolated Slap Lesions

With the patient supine, the examiner places the patient’s shoulder in 120 degrees of abduction, the elbow in 90 degrees of flexion, and the forearm in supination, and then externally rotates the shoulder ( Fig. 46.6 ). The patient then flexes his or her elbow, and any pain while the examiner resists elbow flexion is a positive test. Reported sensitivity of the biceps load II test is 90%, specificity 97%, positive-predictive value 92%, and negative-predictive value 95%.

Gerber Subcoracoid Impingement Test

The Gerber test is designed to identify impingement between the rotator cuff and the coracoid process ( Fig. 46. 2 E ). It is performed in a manner similar to the Hawkins-Kennedy impingement test. The arm is forward flexed 90 degrees and adducted 10 to 20 degrees across the body to bring the lesser tuberosity into contact with the coracoid. Pain with the maneuver indicates coracoid impingement.

Jobe Apprehension-Relocation Test

Jobe described a combination test in 1989 to distinguish between primary impingement and secondary impingement owing to subtle anterior instability ( Fig. 46. 2 F ). With the patient supine, the arm is abducted 90 degrees and externally rotated, which produces pain from impingement. Application of a posteriorly directed force to the humeral head, relocating it in the glenoid, does not change the pain in patients with primary impingement but relieves the pain in patients with instability (subluxation) and secondary impingement, who tolerate maximal external rotation with the humeral head maintained in a reduced position. Evaluation of this test concluded that pain in the position of 90 degrees of abduction and 90 degrees of external rotation could be caused by a variety of disorders and would be diminished by a posteriorly directed force.

Speed Test

In 1966, Crenshaw and Kilgore described a test they attributed to Speed ( Fig. 46.2G ). The Speed test is performed by having the patient forward flex the shoulder to 90 degrees with the elbow extended and the forearm supinated. Resistance is applied to the forearm, and a positive result produces pain localized to the bicipital groove. In an arthroscopic analysis that included biceps tendinitis and superior labral anterior and posterior lesions as positive findings, Bennett found that the Speed test had a sensitivity of 90% and a specificity of 14%. The positive predictive value was 23%, and the negative predictive value was 83%. A study in 2005 showed the following values: sensitivity 38%, specificity 83%, positive predictive value 81%, and negative predictive value 43%.

Yergason Sign

Yergason described the “supination sign” in 1931 ( Fig. 46.2H ). The elbow is flexed to 90 degrees, and the forearm is pronated. The patient attempts to supinate the forearm actively against resistance applied by the examiner at the patient’s wrist. Pain localized to the bicipital groove indicates inflammation of the long head of the biceps. Yergason noted that this test may be negative with partial or complete rupture of the supraspinatus tendon.

Park et al. evaluated eight impingement tests ( Table 46.1 ) in 552 patients and determined that if the Hawkins-Kennedy sign, the painful arc sign, and the infraspinatus muscle (Jobe) test all were positive, the likelihood of a patient having an impingement syndrome of some degree was greater than 95%; if these three tests all were negative, the likelihood of impingement syndrome was less than 24%.

The tests described next ( Fig. 46.7 ) are designed to assess rotator cuff integrity and fall into two types: tests that determine whether a movement can be undertaken actively and tests that determine whether a passive position can be maintained (the lag signs). The main finding on physical examination of patients with full-thickness tears depends on which tendons are torn, what percentage of the tendon width is torn, and the duration for which the tendons have been torn. During an examination of a patient with pain, it may be difficult to determine if weakness is caused by the pain or to a torn tendon. Weakness with pain should be interpreted with caution and not be assumed to indicate a full-thickness tear.

Lift-Off Test

In 1991, Gerber and Krushell described the lift-off test for detection of an isolated rupture of the subscapularis tendon ( Fig. 46.7A ). With the patient seated or standing, the arm is internally rotated, and the dorsum of the hand is placed against the lower back. If the patient is unable to lift the dorsum of the hand off the back, the test is positive. Electromyography has confirmed that the subscapularis muscle is maximally active with the hand in the midlumbar position and with resistance applied.

Belly Press Test

Gerber et al. described the belly press test for patients who have decreased internal rotation ( Fig. 46.7B ). In this test, the patient presses the abdomen with the flat of the hand and attempts to keep the arm in maximal internal rotation. If active internal rotation is strong, the elbow does not drop backward, meaning it remains in front of the trunk. If the strength of the subscapularis is impaired, maximal internal rotation cannot be maintained, the patient feels weakness, and the elbow drops back behind the trunk. The patient exerts pressure on the abdomen by extending the shoulder, rather than by internally rotating it. Other investigators have noted that when the subscapularis tendon is torn, patients tend to flex the wrist to press against the abdomen and are unable to hold the elbow forward.

Bear-Hug Test for Subscapularis Tear

The bear-hug test is performed with the patient’s palm of the involved side placed on the opposite shoulder with fingers extended ( Fig. 46.8A ) and the elbow positioned anterior to the body. The patient is asked to hold that position (resisted internal rotation), and the examiner tries to pull the patient’s hand from the shoulder with an external rotation force applied perpendicular to the forearm ( Fig. 46.8B ). The test is considered positive if the patient is unable to hold the hand against the shoulder or if he or she shows weakness of resisted internal rotation of more than 20% compared with the opposite side. If strength is comparable to that of the opposite side, without any pain, the test is considered negative. A painful bear-hug test without weakness is considered negative.

External Rotation Stress Test

The external rotation stress test is intended to test the integrity of the external rotators of the shoulder, specifically the infraspinatus and the teres minor ( Fig. 46.7C ). With the patient’s arms by his or her side in neutral flexion and abduction, the shoulders are externally rotated 45 to 60 degrees. The examiner applies force against the dorsum of the hands, attempting to rotate the shoulders internally back to neutral while the patient is asked to resist. Pain and weakness suggest inflammation or tearing of the infraspinatus or the teres minor or both.

External Rotation Lag Sign

The external rotation lag sign test is designed to test the integrity of the supraspinatus and infraspinatus tendons ( Fig. 46.7D ). The patient is seated with his or her back to the examiner. The elbow is passively flexed to 90 degrees, and the shoulder is held at 20 degrees of elevation and near maximal external rotation (maximal external rotation –5 degrees to avoid elastic recoil in the shoulder) by the examiner. The patient is asked to maintain the position of external rotation actively as the examiner releases the wrist, while maintaining support of the arm at the elbow. The sign is positive when a lag, or angular drop, occurs. As with all tests, performance and interpretation are complicated by pathologic changes in the passive range of motion. The external rotation lag sign of more than 40 degrees has been shown to have a sensitivity of 100% and a specificity of 92%.

Patte (Hornblower) Sign

The Patte sign is used to determine the strength of the teres minor. With the patient standing, the examiner elevates the patient’s arm to 90 degrees in the scapular plane and flexes the elbow to 90 degrees ( Fig. 46.9 ). The patient is then asked to laterally rotate the shoulder. Weakness and/or pain constitutes a positive test. This sign was reported to have a sensitivity of 93% to 100% and a specificity of 72% to 93%.

Drop Sign

The drop sign is intended to test the integrity of the infraspinatus ( Fig. 46.7E ). The patient is seated with his or her back to the examiner. The affected arm is held at 90 degrees of elevation in the scapular plane and at almost full external rotation with the elbow flexed at 90 degrees. The patient is asked to maintain this position actively as the examiner releases the wrist while supporting the elbow, which is mainly a function of the infraspinatus. The sign is positive if a lag or “drop” occurs. The drop sign has a reported sensitivity of 87% and specificity of 88%.

Internal Rotation Lag Sign

The internal rotation lag sign test is designed to test the integrity of the subscapularis tendon ( Fig. 46.7F ). The patient is seated with his or her back to the examiner. The affected arm is held by the examiner in almost maximal internal rotation. The elbow is flexed to 90 degrees, and the shoulder is held at 20 degrees of elevation and 20 degrees of extension. The dorsum of the hand is passively lifted away from the lumbar region until almost full internal rotation is reached. The patient is asked to maintain this position actively as the examiner releases the wrist while maintaining support at the elbow. The sign is positive when a lag occurs.

Subcoracoid Impingement

In 1909, Goldthwait first described pain in the shoulder caused by contact between the rotator cuff and the coracoid process. Gerber et al. suggested that this painful contact might be caused by a prominent coracoid, for which there may be numerous reasons, including idiopathic and iatrogenic conditions. The iatrogenic form was most common in their series, and it was found in patients who had undergone a Trillat osteotomy of the coracoid for the treatment of anterior instability. In their series of 475 patients with rotator cuff tears, Park et al. identified subcoracoid impingement in 13%; of 110 with subscapularis tears, 56% had subcoracoid impingement. Among patients with subacromial impingement but no rotator cuff tears, 41% had subcoracoid impingement. This entity has not been studied extensively, and it remains a diagnosis of exclusion.

Physical findings attributed to this condition include tenderness over the coracoid and a positive coracoid impingement test (see Fig. 46.2E ). An injection of lidocaine into the subcoracoid region similar to the Neer impingement test (see Fig. 46.2A ) has been used to evaluate patients for coracoid impingement. Relief of pain suggests the diagnosis, but the proximity of multiple structures in the subcoracoid region, including the glenohumeral joint itself, makes the accuracy of these injections questionable. CT has been used in the diagnosis of coracoid impingement; a suggested distance of 6.8 mm between the coracoid tip and the closest portion of the proximal humerus indicates impingement. For suspected impingement, open or arthroscopic coracoplasty has been recommended. A comparison of outcomes between patients with arthroscopic coracoplasty and those without found a significant increase in internal rotation in the treated group, especially in those with large to massive rotator cuff tears. We have no experience with this procedure.

Internal Impingement

In this condition, internal contact of the rotator cuff occurs with the posterosuperior aspect of the glenoid when the arm is abducted, extended, and externally rotated as in the cocked position of the throwing motion. This contact probably is a normal phenomenon but becomes pathologic in certain patients. It often occurs in throwers who have lost internal rotation of the shoulder. This loss causes the center of rotation of the humeral head to move upward so that the contact between the rotator cuff and the biceps tendon attachments increases. Several cadaver studies have attempted to identify factors related to the development of internal impingement. One such study determined that increased capsular laxity significantly increased horizontal abduction and contact pressure in the glenohumeral joint, resulting in impingement of the supraspinatus and infraspinatus tendons and posterosuperior labrum between the greater tuberosity and glenoid. In contrast, another cadaver study found that excessive posteroinferior capsular tightness caused forceful internal impingement of the shoulder at maximal external rotation, whereas another identified increased internal scapular rotation and decreased upward scapular rotation as significantly increasing glenohumeral contact pressure and the area of impingement of the rotator cuff. Arthroscopic findings include partial rotator cuff tears, posterior and superior labral tears, and anterior shoulder laxity. Early in the course of the condition, aggressive physical therapy with attention to regaining internal rotation and rotator cuff strengthening is often successful. Arthroscopic management of this problem is discussed in Chapter 52 .

Primary (External) Impingement

The initial treatment of a patient with tendinopathy caused by classic primary extrinsic impingement is a well-planned and well-executed nonoperative regimen, including anti-inflammatory medications and one or at most two subacromial cortisone injections. Hyaluronic acid injections have been suggested to improve results, but a comparison of hyaluronic acid injections (51 patients) with corticosteroid (53 patients) or placebo injections (55 patients) found no benefit from hyaluronic acid injections; corticosteroid injections produced a significant reduction in pain in the short term (3 to 12 weeks), but in the long term the placebo injection produced the best results. Platelet-rich plasma (PRP) has been offered as a treatment for subacromial impingement, but a single-blinded randomized controlled trial comparing PRP to exercise therapy found that both were effective in reducing pain and disability. Medical treatment is followed by a physical therapy program focusing on stretching for full shoulder motion and strengthening the rotator cuff. A scapular motor control retraining protocol was reported to be successful in relieving symptoms in young patients. If the patient fails to respond after 3 to 4 months of conservative therapy, operative intervention may be indicated and should be directed to the specific lesion. Significantly greater improvements in Constant scores have been reported for patients with a positive Hawkins-Kennedy sign in the neutral position, and positive Neer and Jobe tests compared with those with negative signs. Patients with four positive tests out of the five studied (Neer, Hawkins-Kennedy in neutral and in abduction, Jobe empty can, and painful arc) had greater improvement than those with three or fewer positive test results.

Arthroscopic or open acromioplasty when indicated is the surgical treatment of choice for external impingement syndrome. The reported results of open anterior acromioplasty vary widely. In more than 20 large series of open acromioplasties, the overall success rate was approximately 85%. Failures were related to incorrect diagnosis, technical inadequacy, and other complications. For acromioplasty to be successful, impingement must be the cause of the pain and a thorough history and physical examination are necessary. Acromioclavicular arthritis, glenohumeral arthritis, subtle shoulder instability in throwing athletes, early adhesive capsulitis, and fibromyalgia all can present diagnostic dilemmas. Cervical spondylosis with nerve root irritation and suprascapular nerve injury also can mimic the symptoms associated with an impingement syndrome.

Recently, studies have questioned the superiority of acromioplasty over physical rehabilitation. Ketola et al. randomized 140 patients into a structured exercise program or an arthroscopic acromioplasty group that also did the therapy postoperatively. At follow-up, there were no statistically significant differences in either the amount of perforating ruptures of the supraspinatus tendon or in the changes in muscle volume at 5 years. Kolk et al. in a double-blind randomized clinical trial compared arthroscopic bursectomy alone with bursectomy combined with acromioplasty. Forty-three patients were examined at a median of 12 years postoperatively. Based on Constant score, Simple Shoulder test score, visual analog scale (VAS) pain score, and VAS shoulder function score, there were no relevant additional effects of arthroscopic acromioplasty on bursectomy alone with respect to clinical outcomes or rotator cuff integrity at 12 years follow-up. Paavola et al. compared arthroscopic acromioplasty, diagnostic arthroscopy, and physical therapy. There were no differences between acromioplasty and diagnostic arthroscopy at 2 years, and the differences between acromioplasty and therapy did not reach a minimum clinically important difference.

Technical inadequacy has been implicated as a cause of failed acromioplasties. Adequate bone must be removed to alleviate outlet stenosis. Inadequate bone removal seems to occur more often in arthroscopic than open acromioplasties. In addition to the anterior lip, the portion of the acromion anterior to the anterior clavicular border must be removed to obtain optimal results. Original technical descriptions called for resecting and removing a portion of the coracoacromial ligament to prevent the cut edge from scarring back to the acromion. Our current practice is to release the ligament. We believe that the ligament can be part of the pathologic process and anticipate that it would heal back to the acromion, restoring the coracoacromial arch and preventing anterosuperior subluxation of the humeral head.

We use arthroscopic and occasionally open techniques. We believe that either open or arthroscopic acromioplasty is satisfactory if the main principles of the original procedure as described by Neer are kept in mind, as follows:

• ▪

  Release (but not resection) of the coracoacromial ligament

• ▪

  Removal of the anterior lip and lateral edge of the acromion

• ▪

  Removal of part of the acromion anterior to the anterior border of the clavicle

• ▪

  Removal of the distal 1 to 1.5 cm of clavicle if significant degenerative changes are found

Complications

Complications after acromioplasty include, but are not limited to, infection, seroma formation, hematoma, synovial fistula, biceps rupture, pulmonary embolus, acromial fracture, and complex regional pain syndrome. Poor patient motivation, poor rehabilitation compliance, or a poorly designed rehabilitation program also can lead to failure because of continued pain and stiffness. Bouchard et al. cited co-planing and workers’ compensation claims as poor prognostic indicators.

Without question, the worst common complication is loss of anterior deltoid function, which is caused by either axillary nerve injury or detachment of the deltoid from the acromion. Loss of anterior deltoid function produces a poor outcome despite technically adequate bone work and ligament resection. Very little can be done to restore function to a detached, retracted deltoid. Because deltoid detachment, retraction, and scarring are much more common after the “deltoid on” approach, we recommend suturing to the acromion with heavy sutures whenever tissue is unavailable for direct side-to-side repair.

Partial-Thickness and Full-Thickness Tears

Partial-thickness tears may be articular-sided, bursal-sided, or intratendinous. The true incidence of partial-thickness tears is unknown. Most of the information is from cadaver studies, which reflect an older population; the true incidence in young overhead-throwing athletes is unknown. Among partial-thickness tears, cadaver studies indicate that intratendinous tears are more common than articular-sided or bursal-sided tears, whereas a clinical study found that articular-sided tears constituted 91% of all partial-thickness tears in a population of young athletes. This discrepancy between cadaver and clinical studies may result because intratendinous tears are more difficult to diagnose with arthroscopy, MRI, or ultrasound than are bursal-sided or articular-sided tears. The true prevalence of partial-thickness tears is likely to be greater than that currently documented in the literature. In a systematic review of the literature, Lazarides et al. determined that among young patients with rotator cuff tears, most had full-thickness traumatic tears, but in a subgroup of elite throwers, most tears were partial-thickness tears resulting from chronic overuse.

The natural history of partial-thickness tears is not fully known. Imaging and clinical studies have suggested that partial-thickness tears progress in as many as 80% of patients.

For partial-thickness rotator cuff tears, a nonoperative program that includes activity modification, stretching and strengthening exercises, and antiinflammatory medication is appropriate as initial treatment. Operative management is indicated if conservative management fails. Arthroscopic evaluation is required to determine the extent of the lesion, and subacromial decompression is indicated when outlet impingement is present. The causes of the tear should be treated at the time of surgery. Debridement or repair of partial-thickness rotator cuff tears depends on the degree of the tear and the activity level and age of the patient. We currently perform arthroscopic debridement of partial-thickness tears when they are found on inspection during arthroscopic acromioplasty. If a lesion involves less than 50% of cuff thickness, acromioplasty and debridement are sufficient treatment. If a tear is longer or thicker, elliptical excision of the diseased tendon and repair are indicated. Good results have been reported with arthroscopic repair of bursal-side partial-thickness tears.

The results of nonoperative treatment of partial-thickness tears are unknown because no long-term follow-up studies using a standardized treatment protocol on a well-defined uniform patient population exist in the literature. Excellent and good results after arthroscopic debridement have been reported in from 80% to 90% of patients, with improvements in pain, function, active forward flexion, and strength. Studies also have shown no significant difference in outcome between patients with full-thickness and patients with partial-thickness tears or between patients with partial-thickness tears less than 50% of tendon thickness and patients without any tears.

The primary goal of operative management of rotator cuff tears is pain relief, and this is accomplished with predictable results. Improvement of function is a secondary but important consideration. Functional improvement is not as predictable as pain relief and depends on the age of the patient, the age and size of the tear (which suggests the quality of the tissue and the condition of the muscle), and the postoperative rehabilitation program. In elderly patients or patients with low activity, we attempt a course of conservative treatment (8 to 12 weeks). If there is a positive response, the nonoperative approach may be continued, but if there is no improvement, we proceed to surgery to minimize the atrophy of the rotator cuff musculature. Surgery is appropriate for an acute rotator cuff injury in a young patient or in an older patient (60 to 70 years old) with a defined injury who suddenly is unable to rotate the arm externally against resistance. In our experience, these patients usually have an excellent return of strength and function. Surgery is contraindicated in patients with rotator cuff tears and concomitant stiffness (secondary to adhesive capsulitis). Any significant preoperative stiffness must be corrected before rotator cuff repair to avoid severe postoperative stiffness. It is imperative that nearly full motion be regained before surgical intervention to prevent severe postoperative stiffness.

The clinical results of rotator cuff repair in symptomatic patients who have been followed for 10 years are good to excellent in a high percentage of cases, even though rerupture of the cuff occurs in 20% to 65%. In four large series (Hawkins et al., Neer et al., Ellman et al., and Cofield et al.) that included 476 patients, success rates ranged from 78% to 86%, with excellent or good results reported in 383 (80%) of the 476 patients. In a review of several series of rotator cuff repairs, overall pain relief was obtained in 87% of patients, with a 77% patient satisfaction rate. Most repair failures have been found to occur within the first 2 years after surgery; if the repair survives this initial period, its 10-year survival is likely. Our results with rotator cuff repair are similar, achieving pain control and return of function in approximately 80% of patients. Outcomes of repair generally are better in patients younger than 60 years of age, although clinical results often are good despite poor imaging results. Factors consistently found to be associated with failed repair include age of 65 years or older, large and massive tears (>3 cm), moderate to severe muscle atrophy, more than 50% fatty infiltration of the involved cuff, tear retraction of more than 2.5 cm, and diabetes. Some newer studies, however, have challenged these traditional risk factors. In a comparison of 40 patients older than 70 years and 40 patients younger than 50 years, Moraiti et al. found that functional gain was similar, even though healing was more frequent in younger patients. Although increasing age may predict a diminished healing environment, many studies have demonstrated excellent outcomes in older patients. Chung et al. reported that tendinosis severity assessed by preoperative MRI was the only factor associated with failure to heal in their 55 patients (mean age, 58 years), and Inderhaug et al. identified preoperative use of NSAIDs, long-standing symptoms before surgery, and nonacute onset of symptoms as predictors of inferior long-term outcomes in 147 patients. Other studies have noted that preoperative range of motion, obesity, fatty infiltration, or cuff retraction were not supported as prognostic factors for quality of life after arthroscopic rotator cuff repair. These conflicting reports emphasize the importance of careful preoperative evaluation and specific treatment plans tailored for the individual patient.

Several investigators have compared the results of decompression alone with repair and found much better results with repair. Satisfactory outcomes after decompression have ranged from 8% to 59%, and results have been reported to deteriorate over time. Most rotator cuff tears are now approached arthroscopically. The tear and size can be confirmed, and other intraarticular pathology can be treated. A decision can be made to treat the tear arthroscopically, or arthroscopically assisted (mini-open), or to convert to an open procedure. We currently believe that arthroscopic-assisted or arthroscopic repair is appropriate for partial-thickness and small to medium and some large full-thickness rotator cuff tears. Proposed advantages of arthroscopic repair include access for glenohumeral inspection and treatment of intraarticular lesions, no deltoid detachment, less soft-tissue dissection, and smaller incisions. Arthroscopic techniques can reliably assess rotator cuff tear size, tendon quality, tendon mobility, and suture anchor placement. A recent meta-analysis of randomized controlled trials comparing arthroscopic to mini-open repairs found no differences in surgery time, functional outcome scores, VAS pain scores, or functional outcomes. For the arthroscopic repair technique, see Chapter 52 .

An arthroscopic-plus-open technique has been described for combined tears of the subscapularis, supraspinatus, and infraspinatus tendons. Repair of the posterosuperior rotator cuff is done arthroscopically, followed by open repair of the subscapularis tendon. Cited advantages of this method include an ability to treat concomitant pathology, relative ease of repair, and creation of a strong, reliable construct.

Open rotator cuff repairs also can be done through a miniarthrotomy deltoid-splitting approach (see Technique 1.95.

Massive and Irreparable Tears

Cordasco and Bigliani identified five factors that improved results of operative treatment of large and massive rotator cuff tears:

• 1.

  Adequate subacromial decompression

• 2.

  Maintaining the integrity of the deltoid origin

• 3.

  Mobilizing torn tendons and performing an interval slide when indicated

• 4.

  Repairing tendons to bone

• 5.

  Carefully supervising and staging postoperative rehabilitation

In the ideal repair, the arm can be brought down to the patient’s side without tension on the repair. Occasionally, despite the most diligent efforts to mobilize the tendons, tension remains on the repaired tendons. For these difficult problems, there are few options. If the tendon can be brought to the bony trough with the arm in abduction, the repair is completed and the shoulder is immobilized in an abduction orthosis for 6 weeks to permit tendon-to-bone healing. There are many problems with this technique, but we still believe it to be an option for certain patients.

Excessive tension at the repair that results in suture cutting through the tendon is believed to be the most common mechanism of failure of rotator cuff repair. The choice of suture type and technique can improve repair strength, as can decreasing the postoperative activity level. The latter option, however, has a significant problem with patient compliance. It is difficult to enforce the use of an abduction splint for 6 weeks because of problems with hygiene, comfort, and driving. Without early movement, adhesive capsulitis is likely and may produce a poorer result than the original tear and loss of collagen strength caused by immobilization. Overall results are consistent with improved pain relief and function; however, a decrease in range of motion and muscle strength may remain.

In an effort to decrease tension and increase strength and thus improve the rate and quality of biologic healing of rotator cuff repairs, a number of augmentation methods have been developed, including autografts (biceps, subscapularis, teres minor, latissimus dorsi, coracoacromial ligament), allografts (tendon, ligament, freeze-dried rotator cuff graft), xenografts, and synthetic grafts. The most frequently used augmentation methods involve scaffold devices, which have been developed from polylactic acid, poly(lactide- co -glycolide), and polytetrafluoroethylene; extracellular matrix (ECM) from human, porcine, bovine, and equine sources ( Table 46.2 ); chitin; and chitosan-hyaluronan. Most of the published studies describe animal or biomechanical research involving ECM, with only a limited number of follow-up studies in human patients; these studies have reported mixed results in surgical outcomes and complication rates. Several studies, including one prospective comparative study, have reported improvements in outcomes after repairs of massive and recurrent rotator cuff tears with ECM augmentation, whereas others have found no significant improvement in outcomes. Other series have shown better results with synthetic patch augmentation than with biologic patch augmentation (retear rates of 17% and 51%, respectively). Studies of the use of a porcine xenograft scaffold have found not only worse outcomes with the scaffold but also severe postoperative inflammatory reactions requiring open debridement.

A number of factors influence the extent to which a scaffold device can augment the mechanical properties of a tendon repair, including the mechanical and suture retention properties and the surgical methods of scaffold application (e.g., the number, type, and location of fixation sutures; pretensioning of the scaffold at the time of repair). A study using an analytical model for rotator cuff repairs determined that 70% to 80% of the load is distributed to the tendon after repair, with 20% to 30% of the load carried by the augmentation device. The host response and remodeling of biologic scaffolds also are affected by the species and tissue of origin and the processing and sterilization methods used in preparing the scaffold. Because of the scarcity of clinical data on which to base indications for the use of biologic scaffolds, Derwin et al. developed a grading system that correlates tear size, geometry, and ability to be repaired to the appropriate use of ECM scaffolds ( Table 46.3 ). We have limited experience with the use of ECM augmentation.

Molecular and cellular studies have targeted the tendon-bone interfaces, researching the use of growth factors and cell-coated scaffolds to improve healing. The delivery of transforming growth factor-β3 with an injectable calcium-phosphate matrix was reported to improve healing in a rat model, whereas other studies in a rat model found that application of mesenchymal stem cells genetically modified to overexpress bone morphogenetic protein-13 (BMP-13) did not improve healing, but that stem cells modified to overexpress the developmental gene MT1-MMP produced more fibrocartilage at the interface and improved biomechanical strength. Heringou et al. compared outcomes in rotator cuff tears with and without the use of mesenchymal stem cells and found significant improvements in healing time and substantial improvement in tendon integrity at 10 years after surgery.

In the late 1990s and early 2000s, platelet-rich plasma (PRP), which had been used successfully for many years in other medical specialties, became a popular treatment modality for a variety of orthopaedic conditions, including acute soft-tissue injuries and chronic tendinopathy. PRP is defined as a “volume of plasma that has a platelet count above the baseline of whole blood”; however, PRP preparations can vary markedly according to the amount of blood used and the efficacy of platelet recovery, the presence or absence of white or red blood cells, the activation of platelets with thrombin, and the level of fibrin production. The effect of PRP on healing also differs with different musculoskeletal structures, adding to the difficulty of determining its efficacy. Most studies of the efficacy of PRP in rotator cuff healing have found no benefit regarding retear rates or clinical outcomes. Five meta-analyses and two randomized controlled trials all reached the same conclusion: PRP does not improve early tendon-bone healing or functional recovery. Two comparative studies of leukocyte-plate-rich plasma (L-PRP) found no improvement in the quality of tendon healing or clinical outcomes with the use of L-PRP. A randomized comparison, however, found that PRP significantly decreased the rate of retears and increased the cross-sectional area of the supraspinatus in repair of large to massive rotator cuff repairs; there was no significant difference in clinical outcomes at 1 year.

Occasionally, despite the surgeon’s best efforts and the use of all techniques of mobilization, some tears are so large or retracted, or both, that an anatomic repair is impossible. In this situation, several options are available, none of which is ideal. The two repair options are nonanatomic repair or partial repair. Muscle transfers or slides are another option. The final option is simple debridement.

McLaughlin described suturing the tendon to a trough in bone at whatever point it could be advanced onto the humeral head ( Fig. 46.16 ). This may be more proximal (approximately 2 cm) through the anterior neck area. Although this repair allows a watertight closure, the mechanical advantage of the muscle-tendon unit is lost with this much proximal advancement. Partial repair of massive rotator cuff tears has been proposed to assist in closing large defects and as an alternative to debridement only or tendon transfers. The initial step is a side-to-side tendon repair that results in “marginal convergence” toward the greater tuberosity, which decreases the strain at the free margin of the rotator cuff tear, enhancing the mechanics of the construct. A combination of the tendon-to-tendon repair with tendon-to-bone repair can result in a functional rotator cuff. Partial repair has been shown to be superior to debridement, tendon transfers, and tendon augmentation procedures for the treatment of massive irreparable rotator cuff tears.

Tendon transfers for the treatment of irreparable rotator cuff tears may involve transfer of rotator cuff tendons or other muscle-tendon units. Cofield described subscapularis tendon transposition to fill large gaps in the supraspinatus insertion ( Fig. 46.17 ). The flap is created by separating the outer portion of the subscapularis from the inner capsular portion. It is detached from the lesser tuberosity and mobilized superiorly to cover the humeral head. Other surgeons prefer to use the upper half of the subscapularis tendon by separating it from the anterior capsule and transferring it superiorly. This repair results in great tension in abduction and external rotation and disrupts the subscapularis force couple, which could prove detrimental to shoulder function.

For anterosuperior tears involving the subscapularis and the supraspinatus, transfer of the pectoralis major has been described. The coracoacromial arch should be intact. The technique involves sharply releasing the sternocostal portion of the pectoralis major from its common insertion site on the humerus and bluntly dissecting it from the more superficial clavicular head. Care should be taken not to injure its nerve supply. The sternocostal head is passed deep to the clavicular head and underneath the conjoined tendon to the lesser tuberosity. The musculocutaneous nerve should be identified and protected; the tendon is passed superficial to the nerve. Passing the tendon beneath the conjoined tendon improves the posterior and inferior vectors of the transferred tendon. This transfer also is indicated for treatment of anterior soft-tissue deficiencies and instability after shoulder arthroplasty.

For posterosuperior tears involving the infraspinatus and supraspinatus, the latissimus dorsi has been transferred. Clinically small but statistically significant gains can be expected in motion and strength. Factors reported to be associated with better clinical results include better preoperative function in active forward flexion and external rotation and synchronous in-phase contraction of the transferred latissimus dorsi by electromyography; poor shoulder function and generalized muscle weakness before surgery have been correlated with a poor clinical result. Patients with unsatisfactory results after this procedure may be clinically worse than they were preoperatively.

Transfer of the subcoracoid pectoralis major has been reported for patients with anterosuperior subluxation associated with massive rotator cuff tears, with approximately 80% satisfactory results. Other muscles used for transfer include the teres minor, deltoid, and trapezius, but these are used infrequently and are associated with compromised function. We have no experience with this technique.

Others have used free grafts (autologous or autogenous), such as the intrinsic portion of the biceps, the coracoacromial ligament, and fascia lata, or synthetics to augment or replace deficient rotator cuff tendon. There are few published reports of results, and most are not encouraging; however, successful closure of 14 massive cuff defects was reported with the use of a “tendon patch” fashioned from the long head of the biceps tendon. The disadvantages of the synthetic material are the potential for foreign body reaction to synthetics and tissue rejection. Such materials do not replace the atrophic and weakened rotator cuff musculature present with chronic massive tears.