Rotator Cuff Pathology 

Updated: Feb 24, 2020
Author: R H Bilal, MBBS, MRCS; Chief Editor: S Ashfaq Hasan, MD 


Practice Essentials

Rupture of the rotator cuff tendon was first described by Smith in 1834. Subsequently, degenerative changes of the rotator cuff have been better characterized by Duplay (1872), Von Meyer (1924), Codman (1934), and Neer (1972).[1]  The exact mechanisms leading to the degeneration of the rotator cuff, however, are still being debated.

Shoulder pain is the third most common cause of musculoskeletal disorders (MSDs), after low back and neck pain. Although rotator cuff tendinitis is considered a benign condition, a study of its long-term outcome found that 61% of patients were still symptomatic at 18 months, despite receiving what was considered sufficient conservative treatment, and 26% rated their symptoms as severe. MSDs are the primary disabling conditions of working adults. The prevalence of rotator cuff tendinitis has been found to be as high as 18% in certain workers who performed heavy manual labor.

Webster and Snook estimated that the mean compensation cost per case of upper-extremity work-related MSD was $8070 in 1993; the total US compensable cost for upper extremity, work-related MSDs was $563 million in the 1993 workforce.[2]  The compensable cost is limited to the medical expenses and indemnity costs (lost wages). When other expenses (eg, full lost wages, lost production, cost of recruiting and training replacement workers, cost of rehabilitating the affected workers) are considered, the total cost to the national economy becomes much greater.

The impact of rotator cuff disease on quality of life is even more difficult to assess than its cost. Further studies using valid methods that measure the impact of the disorder on general health (eg, the Medical Outcomes Study [MOS] 36-item short-form health survey [SF-36]) should help in the evaluation of this issue.

Conservative treatment of the degenerative rotator cuff may be effective. (See Treatment.) However, patients with more advanced rotator cuff disease or a more significant injury may not respond to conservative therapies. If the patient believes that his or her quality of life is being significantly impacted by the shoulder dysfunction, then surgical intervention is a reasonable consideration.


The rotator cuff is composed of four muscles—subscapularis, supraspinatus, infraspinatus, and teres minor—and their musculotendinous attachments (see the image below).

Rotator cuff anatomy. Rotator cuff anatomy.

The subscapularis is innervated by the subscapular nerve and originates on the scapula. It inserts on the lesser tuberosity of the humerus. The supraspinatus and the infraspinatus are both innervated by the suprascapular nerve, originate in the scapula, and insert on the greater tuberosity. The teres minor is innervated by the axillary nerve, originates on the scapula, and inserts on the greater tuberosity.

The subacromial space lies underneath the acromion, the coracoid process, the acromioclavicular joint, and the coracoacromial ligament. A bursa in the subacromial space provides lubrication for the rotator cuff.[3, 4]

A solid knowledge of the functional anatomy of the rotator cuff makes it easier to understand the disorder that affect this structure. The rotator cuff is the dynamic stabilizer of the glenohumeral joint; the static stabilizers are the capsule and the labrum complex, including the glenohumeral ligaments. Although the rotator cuff muscles generate torque, they also depress the humeral head. The deltoid abducts the shoulder. Without an intact rotator cuff, particularly during the first 60° of humeral elevation, the unopposed deltoid would cause cephalad migration of the humeral head, with resulting subacromial impingement.


Rotator cuff pathology can result from extrinsic or intrinsic factors. Extrinsic examples include a traumatic tear in tendons from a fall or accident. Overuse injuries from repetitive lifting, pushing, pulling, or throwing are also extrinsic in nature. Intrinsic factors include poor blood supply, normal attrition or degeneration with aging, and calcific invasion of tendons.[3, 5, 6, 7, 8, 9]

Rotator cuff tendinitis is the term used to describe irritation of tendons either from excessive pressure on the acromion or, less commonly, from intrinsic tendon pathology. Irritation of the adjacent bursa is known as subdeltoid or subacromial bursitis. Repetitive overhead activities resulting in irritation of tendons and bursae from repeated contact with the undersurface of the acromion is termed impingement syndrome.

Rotator cuff dysfunction is typically a continuum of pathology ranging from tendinitis and bursitis to partial tearing to complete tearing in one or more of the tendons. Although the earlier stages may resolve with conservative care, actual tearing of the tendon can be more problematic. These tears most commonly occur at the tenoperiosteal (tendon-to-bone) junction. Because this area has a relatively poor blood supply, injury to the tendon at this location is very unlikely to heal well.

Additionally, the constant resting tension in the muscle-tendon unit, or muscle tone, pulls any detached fibers away from the bone, preventing their reattachment. Finally, joint fluid from within the shoulder may seep into the gap created by the tear and prevent the normal healing processes from occurring.


Possible causes of rotator cuff pathology incldue the following:

  • Outlet impingement
  • Subacromial spurs
  • Type 2 and type 3 acromions
  • Osteoarthritic spurs of the acromioclavicular joint (including subacromial spurs)
  • Thickened or calcified coracoacromial ligament
  • Nonoutlet impingement
  • Loss of rotator cuff causing superior migration of the humerus (ie, tear, loss of strength)
  • Secondary impingement from an unstable shoulder
  • Acromial defects (os acromiale)
  • Anterior or posterior capsular contractures (eg, adhesive capsulitis)
  • Thick subacromial bursa

Genetic factors may play a role in the pathogenesis of rotator cuff disease. A systematic review by Longo et al found significant associations between single-nucleotide polymorphisms and rotator cuff disease was found for DEFB1, FGFR1, FGFR3, ESRRB, FGF10, MMP-1, TNC, FCRL3, SASH1, SAP30BP, and rs71404070 located next to cadherin8; results reported for MMP-3 were contradictory.[10]


Shoulder pain is the third most common cause of MSDs, after low back pain and cervical pain. Estimates of the cumulative annual incidence of shoulder disorders range from 7% to 25% in the Western general population. The annual incidence is estimated at 10 cases per 1000 population, peaking at 25 cases per 1000 population in persons aged 42-46 years.

In persons aged 70 years or older, 21% of persons have shoulder symptoms, most of which can be attributed to the rotator cuff. In cadaver studies, the rate of full-thickness tears ranges from 18% to 26%. The rate of partial-thickness tears ranges from 32% to 37% after age 40 years. Before age 40 years, tears are rare. In magnetic resonance imaging (MRI) studies, tears have been observed in 34% of asymptomatic individuals of any age. After age 60 years, 26% of patients have partial-thickness tears, and 28% demonstrate full-thickness tears.

No known racial variation associated with rotator cuff disease is cited in the literature. In one study, a predominance of male patients (66%) seeking consultation for rotator disease was reported, but in other studies, the male-to-female ratio was 1:1. Rotator cuff disease is more common after age 40 years. The average age of onset is estimated at 55 years.


An estimated 4% of cuff ruptures develop a cuff arthropathy. Various authors report the success rate of conservative treatment to be 33-90%, with longer recovery time required in older patients. Surgery results in better function regardless of the patient's age.

Piasecki et al found that arthroscopic revision rotator cuff repair may be a reasonable option even after previous open repair, providing improved pain relief and shoulder function.[11] In the 54 patients studied, the American Shoulder and Elbow Surgeons scores improved from 43.8 ± 5.7 to 68.1 ± 7.2, and the Simple Shoulder Test improved from 3.56 ± 0.8 to 7.5 ± 1.1. Visual Analog Scale (VAS) scores for pain improved from 5.17 ± 0.8 to 2.75 ± 0.8, and forward elevation increased from 121.0º ± 12.3º to 136º ± 11.8º. Female patients and those who had undergone more than one ipsilateral shoulder surgery had poorer results.

In a systematic review of the published literature, Nho et al compared single-row (SR) with double-row (DR) suture anchor fixation in arthroscopic rotator cuff repair.[12] They found no clinical differences between the two approaches. They concluded that the data in the published literature do not support the use of DR suture anchor fixation to improve clinical outcome, though they noted that there are some studies reporting that DR suture anchor fixation may improve tendon healing.

A meta-analysis including three randomized controlled studies and two controlled clinical cohort studies compared outcomes between SR and DR rotator cuff repair.[13] The results showed that whereas the DR technique significantly increased operating time, it provided greater external rotation, improved tendon healing, and decreased recurrence rate. However, there were no significant differences between the two techniques with regard to shoulder function, muscle strength, forward flexion, internal rotation, patient satisfaction, or return to work.

One study analyzed the structural and functional outcomes after arthroscopic rotator cuff repair between SR, DR, and combined DR/suture-bridge techniques.[14] After an average follow-up of 38.5 months, repair with the combined DR/suture-bridge technique resulted in an overall decreased retear rate, especially for large and massive tears. This combined technique proved to be an effective option for arthroscopic rotator cuff repair.

Schofer et al, in a prospective, randomized, controlled study, compared high-energy extracorporeal shock-wave therapy (ESWT) with low-energy ESWT in the treatment of rotator cuff tendinopathy.[15] Patients in the high-energy group received 6000 impulses of ED+ 0.78 mJ/mm2 in three sessions, and those in the low-energy group received 6000 impulses of ED+ 0.33 mJ/mm2.

An increase in function and a reduction of pain were found in both groups.[15] Although the improvement in Constant score was greater in the high-energy group, statistical analysis showed no significant difference between the two groups with respect to Constant score, pain reduction, and subjective improvement after 12 weeks and after 1-year follow-up.

Drake et al, in a study reviewing the use of reverse total shoulder arthroplasty (RTSA) in patients with rotator cuff disease,[16]  found that modern RTSA designs restore deltoid tension and a functional fulcrum to the rotator cuff-deficient shoulder, allowing recovery of active shoulder elevation and restoring function. Contraindications for RTSA included the following:

  • Severely impaired deltoid function
  • Isolated supraspinatus tear
  • Presence of full active shoulder elevation with a massive rotator cuff tear and arthritis

Drake et al concluded that for properly selected patients with symptomatic and disabling rotator cuff deficiency, RTSA can yield life-changing improvements in pain, motion, function, and patient satisfaction.[16]

Wellmann et al concluded that for patients with symptomatic and disabling rotator cuff deficiency, RTSA can result in a significant reduction in pain and improvements in motion and function.[17]

Using propensity-matching methods, Oh et al compared the outcomes of patients with pseudoparalytic large-to-massive tears with those of patients with nonpseudoparalytic tears after rotator cuff repair with the aim of determining whether the presence of pseudoparalysis had a negative effect. Evidence of recovery from pseudoparalysis was noted in a large portion of the study group; similar outcomes were noted in postoperative function and cuff healing, whether pseudoparalysis was present or not. Considering the possible complications from treatment with RTSA, the authors suggested that rotator cuff repair should be the first-line treatment option for large-to-massive tears.[18]

Data from a study by Chung et al showed that the failure rate after arthroscopic rotator cuff repair was significantly higher in patients with lower bone mineral density, a higher grade of fatty infiltration of the infraspinatus, and greater amount of retraction.[19]

The T-scale, a measure of the anterolateral translation of the humeral head, has been suggested as a prognostic factor for rotator cuff repair. In a study of 120 consecutive patients with full-thickness rotator cuff tears, Taniguchi et al found that patients who had large-to-massive tears in conjunction with negative values on this scale had poorer clinical outcomes and were at greater risk for repeat tearing.[20] They concluded that a negative T-scale value is a useful prognostic factor for considering reverse shoulder arthroplasty in patients with a higher risk of retear.




Patients with rotator cuff pathology commonly present with an activity-related dull ache in their upper lateral (outer) arm and shoulder. Activity is usually most difficult above shoulder level. Many people have little or no discomfort with below-shoulder-level activities such as golf, bowling, gardening, writing, or typing. Tennis, baseball and softball, basketball, swimming, and painting tend to be more problematic.

A complete medical history should be obtained to direct the physical examination and make the correct diagnosis. Most of the time, the diagnosis can be made following a systematic history. Relevant history findings, treatments, and test results should complement the history of the present injury. Sometimes, relevant social and family histories are necessary.

Patients with degenerative rotator cuff disease are almost always older than 40 years. About 50% of patients have a progressive onset of shoulder pain, whereas the other 50% can identify a specific event responsible for the onset of pain. The evolution of rotator cuff disease is characterized by variable episodes of recurrence following more intensive shoulder activities, followed by remission with rest or treatment.

As the disease progresses, shoulder pain becomes more constant. Overhead and arm-length activities typically increase the pain. Discomfort and night pain can also be present. With time, the individual can notice some weakness during shoulder elevation. Crepitus can also be noted. With evolution of the disease, shoulder pain can be accompanied by cervical and mid-back pain.

The following questions should help the physician assess the patient:

  • What is the patient's age? Shoulder pain in young overhead athletes suggests underlying shoulder instability. In older patients, degenerative rotator cuff disease or frozen shoulder is suggested by shoulder pain.
  • What is the patient's occupation or sport? Repetitive overhead activities and sports predispose to rotator cuff tendinitis.
  • What was the mechanism of injury? A fall on an outstretched arm could indicate a dislocation of the glenohumeral joint or a fracture of the humeral neck. Repetitive overhead motions can cause tendinitis and, in the long run, chronic degenerative changes. A fall or a trauma on the tip of the shoulder can result in an acromioclavicular sprain.
  • What was the onset? Insidious, slow onset may suggest tendinitis or osteoarthritis. Sudden onset is usually due to a trauma causing a fracture, dislocation, or a rotator cuff tear.
  • Where is the pain located? Pain located on the superior or lateral aspect of the shoulder suggests rotator cuff tendinitis. Pain on the anterior aspect of the shoulder may result from bicipital tendinitis, an acromioclavicular sprain, or anterior instability. Neck pain and radicular pain or paresthesias suggest a cervical spine disorder.
  • What is the severity of the pain? An acute burning pain could indicate an acute bursitis. An intermittent dull pain may be due to a degenerative rotator cuff disease.
  • What is the type of pain? Sharp, burning pain suggests a neurologic origin. Bone and tendon pain is deep, boring, and localized. Muscle pain is dull and aching, is not localized, and may be referred to other areas. Vascular pain is aching, cramplike, and poorly localized, and it may be referred to other areas.
  • What is the duration of the symptoms? Frozen shoulder has three stages that can last up to 3-4 years. Acute bursitis has a short-term evolution and responds well to nonsteroidal anti-inflammatory drugs (NSAIDs).
  • What is the timing of the pain? Predominantly night pain suggests frozen shoulder. Morning pain and stiffness improved by activity may be caused by a synovitis. Pain that increases with activity is usually the result of a rotator cuff tendinitis.
  • Which activities/positions increase the pain? Pain increased by overhead activities or arm-length activities suggests rotator cuff tendinitis. Pain increased when throwing is likely to be due to anterior instability. Pain increased by lying on the affected shoulder may be caused by an acromioclavicular sprain.
  • Which activities/positions relieve the pain?
  • Is there any weakness or paresthesia in the upper extremities? Neurologic symptoms are caused by a cervical radiculopathy or peripheral nerve entrapment/lesion.
  • Are the symptoms constant or intermittent? Intermittent symptoms usually result from soft-tissue or joint disorders. Constant symptoms suggest a neurologic lesion.
  • Is joint-motion restriction present? Passive and active joint restriction in all directions of range of motion (ROM) is caused by a frozen shoulder or glenohumeral synovitis. Restriction in internal rotation suggests an impingement syndrome due to rotator cuff tendinitis. The inability to perform active abduction suggests a rotator cuff tear or a frozen shoulder.
  • Is some crepitus noted? Crepitus is the result of degenerative rotator cuff changes. Crepitus is not a normal finding in the shoulder.
  • Have any changes in the color of the arm occurred? Color changes may be due to ischemia secondary to vascular insufficiency. Reflex sympathetic dystrophy (also termed complex regional pain syndrome type 1) can cause skin-color changes.
  • Has the patient had any treatments such as oral medication, injections, or physical therapy to date?
  • Has the patient had any diagnostic tests performed to date?
  • What is the evolution of the symptoms?
  • Has the pain changed?
  • Has the pain spread or moved?
  • Has the pain subsided or increased?

The last three questions help in deciding the appropriate treatment and management.

The importance of obtaining a systematic and detailed history cannot be overemphasized. Any attempt to shortcut the process leads to an unfocused physical examination and an inaccurate diagnosis. In one study assessing the interobserver agreement of a diagnostic classification of shoulder disorders based on history and physical examination, there was only moderate agreement between experienced observers.

Physical Examination

A systematic examination of the shoulder region includes the following:

  • Careful observation
  • Palpation of the bones and soft tissues
  • Assessment of passive and active ROM
  • Impingement and topographic tests, complemented, as needed, by instability tests, labrum tests, and special tests

The examination is completed by a cervical spine examination, along with neurologic and vascular examination.


Observation begins from the moment the patient enters the room. The smoothness and symmetry of the shoulders and the movements of the upper extremities are evaluated, as is the patient's gait. The examiner must be aware of any signs of painful posturing and irregularity of motion of the affected shoulder. Bilateral examination allows comparison of the affected shoulder with the unaffected one.

The patient is then asked to remove the appropriate amount of clothing to facilitate proper assessment of the bone and soft tissues. The shoulder, cervical region, and entire upper extremity must be assessed. The examiner should assess bones and joints for possible asymmetry or deformities and look for soft-tissue changes (eg, swelling, erythema, white shiny skin, loss of hair, or atrophy) suggestive of vasomotor abnormalities. Scars and abrasions also must be noted. Bony contours should be assessed first, then soft tissues. Observation must be completed from the front, the side, and the back.

Looking at bony contours, the examiner first makes a general assessment. The dominant side may be lower than the nondominant one; the head and neck should be in the midline; the clavicle should be symmetric without any deformity of the acromioclavicular joint and sternoclavicular joint.

Each of these parts is then examined in more detail. Because of its superficial location, a fracture of the clavicle or a subluxation or dislocation of both ends is easy to identify. A step deformity of the acromioclavicular or sternoclavicular joint, with the clavicle side of the joint migrating superiorly, is due to a dislocation of these joints.

Observation of the soft tissues is directed first at the contours of the deltoid. The mass of the deltoid should be round, with the anterior and posterior aspects symmetric. Flattening of the muscle suggests atrophy of the deltoid and is usually due to a neurologic lesion such as an axillary nerve neuropathy, an upper-trunk brachial plexopathy (Erb palsy), or a C5-6 radiculopathy.

An anterior dislocation of the glenohumeral joint produces flattening of the deltoid with bulging of the anterior aspect of the muscle due to the dislocated head of the humerus, with the patient holding the shoulder in slight adduction and across the trunk. A bulge observed in the middle third of the belly of the biceps when the elbow is flexed suggests rupture of the long head of the biceps tendon.

The side view allows the examiner to assess thoracic spine kyphosis, which is indicated by a protraction of the head or the shoulders. Deltoid atrophy also can be observed.

Looking at bony contours, the examiner seeks evidence of a scoliosis of the thoracolumbar spine and then observes the scapulae. Each scapula extends from the spinous process of T2 (superomedial angle) to the spinous process of T7 (inferomedial angle). The scapulae should be at the same height and at the same distance from the spine.

The examiner should check for a winging of the scapula (ie, a displacement of the medial side of the scapula away from the thorax). When the winging is present with medial displacement of the scapula toward the spine, a serratus anterior muscle palsy is suggested. This palsy usually is due to a long thoracic nerve injury. When the winging is noted with lateral displacement of the scapula, a trapezius muscle palsy or, more rarely, a rhomboid muscle palsy must be suspected.

A trapezius muscle palsy can be due to a spinal accessory nerve (cranial nerve XI) injury, and a rhomboid muscle palsy can be due to a dorsal scapular nerve injury. A prominent spine of the scapula may be due to supraspinatus and infraspinatus muscle atrophy caused by a suprascapular nerve injury in the suprascapular notch or a rotator cuff tear.

Observation of the soft tissues is directed at the posterior aspect of the deltoid muscle. The trapezius muscle is then observed. Atrophy resulting from palsy of the muscle has been discussed previously. Because the rhomboid is overlapped by the trapezius, atrophy of the rhomboids is more difficult to assess.


Like observation, palpation must be performed in an orderly manner, beginning with the anterior structures and finishing with the posterior structures. Palpation must include bony structures and soft tissues. Irregular joint surfaces, swelling, heat, crepitus, pain, tenderness, and muscle tension and spasms must be sought. Palpation can be performed more conveniently with the patient standing. In this position, the examiner can more easily move around the patient. The examiner should stand behind the patient for the palpation.

Beginning with the anterior structures, the examiner palpates the sternoclavicular joint. Superior migration of the medial end of the clavicle is palpated if the joint is dislocated. The examiner must remember that the clavicle is superior to the manubrium. The affected side must always be compared with the contralateral side.

The sternocleidomastoid also must be palpated with the aim of looking for tension and spasms. The muscle contracts to turn the head on the contralateral side. The muscle is easier to identify and palpate in this position. The sternal and clavicular heads of the muscle must be palpated. Hands can be moved medially to palpate the suprasternal notch. The first rib, the costochondral joints, and the sternum also should be assessed.

The clavicle should be palpated along its whole length to check for bumps (suggesting callus formation resulting from fracture), loss of continuity, and crepitus. The acromioclavicular joint is a common site of pain and must be palpated with care. Because the acromioclavicular joint is a superficial joint, swelling, synovial thickening, or crepitus can be palpated. Step deformities with superior migration of the lateral end of the clavicle, seen in dislocation or subluxation, are easily palpable.

The coracoid process can be palpated approximately 2.5 cm (1 in.) inferior and just medial to the acromioclavicular joint. The coracoid process is the site of origin of the short head of the biceps tendon, the coracobrachialis muscle, and the insertion of the pectoralis minor. The pectoralis major and minor also must be palpated. Muscle tension and spasms are commonly associated with shoulder disorders.

The acromion and subacromial space are palpated. The subacromiodeltoid bursa can be palpated indirectly in the subacromial space. Because it is overlapped by the deltoid muscle, the bursa cannot be felt under the fingers; however, the examiner, through pressure on the deltoid muscle, applies indirect pressure on the inflamed bursa, causing pain.

The examiner follows by palpating the greater tuberosity, the long head of the biceps tendon, and the lesser tuberosity. In a lean patient, these structures are easily identified by an experienced examiner; however, identification may be more difficult in an overweight patient or one with abundant muscle mass.

By rotating the shoulder medially (eg, by putting the dorsal aspect of the hand on the buttock), the examiner can feel the greater tuberosity on the anterior aspect of the shoulder, just inferior to the acromion. The supraspinatus, infraspinatus, and teres minor tendons all insert into this structure and, when inflamed, can produce pain upon palpation of the greater tuberosity.

Keeping the fingers on the greater tuberosity, the examiner rotates the shoulder laterally. The fingers feel the bicipital groove where the long head of the biceps tendon can be palpated. Pain or thickening of the tendon sheet indicates an inflamed tendon, whereas its absence suggests a rupture or dislocation. By rotating the shoulder more laterally, the examiner can palpate the lesser tuberosity. The tendon of the subscapularis inserts on that structure, and when it is inflamed, the tendon is painful to palpation.

With the shoulder back to a neutral position, extension of the shoulder allows palpation of the subacromiodeltoid bursae under the anterior edge of the acromion.

All of these structures must be palpated gently because they may be tender. Any painful palpation must be compared with the contralateral shoulder. A positive finding is when pain is more significant on the affected side than on the contralateral shoulder. Any excessive pain caused by a vigorous palpation makes the examination less sensitive.

The biceps muscle should be palpated to check for any bulging that indicates a rupture of the long head of the biceps tendon. The deltoid muscle also must be palpated to look for painful spasm or tension. Tone and atrophy also are assessed.

The examination is continued by palpation of the posterior structures. Bony structures can be rapidly assessed because they are rarely a source of pain. The spine of the scapula is palpated, followed by palpation of the superior medial angle of the scapula. The levator scapulae muscle that inserts on this angle is a common site of pain. The medial border of the scapula is then palpated from the superior to the inferior medial angle. The bony palpation is completed by palpation of the spinous processes of the dorsal and cervical spine.

Because muscle spasm and tension are frequently associated with a rotator cuff disease, the posterior muscles must be palpated with care to identify and treat those muscles. The superior trapezius is commonly tense and painful and must be palpated from its cervical and occipital origin to its insertion on the spine of the scapula and the acromion. Under this muscle, lying in the supraspinatus fossa, the supraspinatus muscle also should be palpated.

The rhomboid muscles, from C7 to T5, run downward to attach on the medial border of the scapula. These muscles, often a source of pain, are difficult to distinguish from the overlying middle trapezius muscle. The rhomboid muscles can be identified by asking the patient to put his or her hand behind the back, with the shoulder internally rotated and the elbow flexed, and to push posteriorly against a resistance. The muscle belly of the rhomboid muscles then becomes palpable.

Muscle palpation is completed by assessing the infraspinatus, teres major and minor, and latissimus dorsi muscles.

Evaluation of range of motion

Both active and passive ROM must be evaluated. Although some authors suggest that an assessment of passive ROM is not necessary if the patient is able to perform complete active ROM without pain, passive ROM must be assessed systematically. Some patients with glenohumeral ROM restrictions have learned to compensate with increased scapulothoracic mobility and seem to have near-normal active ROM.

Movements (with the normal ranges provided) that should be assessed are abduction (70-180°), adduction (30-45°), flexion (160-180°), extension (45-50°), external rotation (80-90°), and internal rotation (90-110°).

Active movements are evaluated first. With the observer behind the patient (who is standing), active abduction is performed.

The scapulohumeral rhythm is observed. If a painful arc (ie, pain or inability to abduct because of pain) is observed at 45-120°, a subacromial impingement syndrome is suggested. If the pain is greater after 120°, when full elevation is reached, an acromioclavicular joint disorder is suggested.

If a reverse scapulohumeral rhythm (ie, an abduction initiated by the scapulothoracic joint rather than by the glenohumeral joint) is observed, a frozen shoulder is suggested. Look for a winging of the scapula caused by trapezius or rhomboid muscle weakness. Active flexion is also evaluated. In the presence of a subacromial impingement syndrome, this movement can also be painful. Active flexion may elicit winging of the scapula caused by weakness of the serratus anterior.

Other motions can be evaluated through a combination of active movements. The Apley scratch test is probably the best known of these maneuvers. This test combines internal rotation and adduction of one shoulder with external rotation and abduction of the other.

Evaluation of passive ROM can be performed with the patient standing, sitting, or lying down. For practical purposes, the examination is more often performed with the patient standing. Passive abduction is assessed with the observer behind the patient. Full abduction is performed first to evaluate the combination of scapulothoracic and glenohumeral motion. Then, the scapulothoracic joint is locked by putting one hand over the scapula and the clavicle to resist any motion of this joint. This maneuver enables more selective evaluation of the glenohumeral joint (90-120°).

The same procedure can be used to evaluate full flexion that combines scapulothoracic and glenohumeral motion and flexion performed selectively by the glenohumeral joint. This maneuver is followed by an evaluation of adduction. The external rotation is tested with the elbow held close to the waist and flexed at 90°. Then, the arm is rotated externally. The examination is followed by evaluation of the extension and assessment of the internal rotation. The full range of internal rotation is achieved with the forearm passing behind the trunk with the shoulder slightly extended.

Impingement testing

Positive impingement tests result from the reproduction of impingement of the rotator cuff tendon by different provocative maneuvers.[21] With anterosuperior impingement syndrome, the impingement occurs underneath the coracoacromial arch. With posterosuperior impingement syndrome, the impingement is on the posterosuperior border of the glenoid cavity. Finally, with anterointernal impingement syndrome, the impingement occurs in the subcoracoid space or in the coracohumeral interval.

Impingement tests confirm the presence of an impingement syndrome; however, they do not determine the location of the rotator cuff lesion.

A study of cadaveric shoulders showed that some provocative impingement tests—namely, the Neer and Hawkins-Kennedy tests—appear to elicit contact consistent with impingement.

The Neer impingement test is carried out as follows:

  • With the examiner standing behind the patient, the shoulder is passively flexed. The author positions the shoulder in internal rotation (though this positioning was not part of Neer's original description).
  • When the result is positive, this test produces pain caused by contact of the bursal side of the rotator cuff on the anterior third of the undersurface of the acromion and the coracoacromial ligament, as well as by contact of the articular side of the tendon with the anterosuperior glenoid rim.
  • A positive test result suggests an anterosuperior impingement syndrome. The sensitivity of this test, assessed based on operatively observed anatomic lesions, is 89%.

The Hawkins-Kennedy test is performed as follows:

  • With the examiner standing behind the patient, the shoulder is flexed passively to 90°, followed by repeated internal rotation.
  • When the result is positive, this test produces pain caused by contact of the bursal side of the rotator cuff on the coracoacromial ligament and by contact between the articular surface of the tendon and the anterosuperior glenoid rim. Contact between the subscapularis tendon and the coracoid process is also observed.
  • A positive test result suggests an anterosuperior or an anterointernal impingement test.
  • The author uses a modified version of this test with the shoulder positioned initially at 90° of abduction and 30° of flexion in the plane of the scapula. Along with repeated internal rotation motion, the shoulder is brought progressively to 90° of flexion. If pain is present when the shoulder is flexed at 30°, it is caused by an anterosuperior impingement syndrome. If the pain is present only when the shoulder is brought to 90° of flexion, reducing the coracohumeral interval, an anterointernal impingement syndrome is suggested.
  • The sensitivity of this test is 87%.

The Yocum test is carried out as follows:

  • With the examiner standing behind the patient, the hand on the ipsilateral side of the examined shoulder is placed on the contralateral shoulder. The elevation of the elbow is resisted by the examiner.
  • When the result is positive, this test produces pain caused by contact of the bursal side of the cuff tendon with the coracoacromial ligament and possibly the undersurface of the acromioclavicular joint.
  • A positive test suggests an anterosuperior or an anterointernal impingement syndrome. The sensitivity of this test is only 78%; however, the sensitivity of these three tests together is 100%, which supports the view that the three tests should be systematically performed together.

The posterior impingement test is performed as follows:

  • With the patient lying down, the shoulder is positioned at 90-100° of abduction and maximally externally rotated.
  • When the result is positive, this test produces pain in the posterior aspect of the shoulder caused by impingement of the articular side of the cuff tendon between the greater tuberosity and the posterosuperior glenoid rim and labrum. Relocation of the humeral head, performed by applying a posteriorly directed force to the humeral head, causes a reduction in pain.
  • The sensitivity of this test is 90%.

Topographic testing

Using resisted isometric contraction of specific muscles of the rotator cuff, the location of the tendon lesion causing the impingement can be identified.

To identify the supraspinatus tendon, use the Jobe test or the full-can test.

In the Jobe test, the shoulder is placed at 90° of abduction and 30° of flexion in the plane of the scapula. Shoulder elevation is resisted. The test result is considered positive if pain is noted. When compared with surgical observations, the sensitivity of this test is 86% and the specificity is 50%. The positive predictive value (the ratio of true positive tests on all the positive tests) of the Jobe test is 96%, and its negative predictive value (the ratio of all the true negative tests on all the negative tests) is 22%.

In the full-can test, the shoulder is placed at 90° of flexion and 45° of external humeral rotation (with the thumb pointing upward, as if someone is holding a full can right side up). Shoulder elevation is resisted. The test result is considered positive if this produces pain. Electromyography (EMG) studies show that this test results in the greatest supraspinatus activation with the least activation from the infraspinatus.

To identify the infraspinatus tendon, use the infraspinatus isolation test or, less optimally, the Patte test.

In the infraspinatus isolation test, the shoulder is positioned at 0° of elevation (elbows against the waist flexed at 90°) and 45° of internal rotation. Shoulder external rotation is resisted. The test result is considered positive if it produces pain. EMG shows this to be the optimal infraspinatus isolation test.

In the Patte test, the shoulder is placed at 90° of abduction, in neutral rotation, and in the plane of the scapula. The examiner holds the patient's elbow, and external rotation is resisted. The test result is considered positive if this produces pain. The test has a sensitivity of 92% but a specificity of only 30%. It has a positive predictive value of 29% and a negative predictive value of 93%. A palsy of the external rotator also can be tested. With the elbow held against the waist, the shoulder is positioned passively in external rotation. A positive test result is signaled by inability to maintain the shoulder in external rotation, suggesting a full tear of the external rotators.

To identify the teres minor tendon, use the same tests used for the infraspinatus tendon. No specific teres minor isolation tests have been developed.

To identify the subscapular tendon, use the Gerber liftoff test or the Gerber push-with-force test.

In the Gerber liftoff test, the shoulder is placed passively in internal rotation and slight extension by placing the hand 5-10 cm from the back with the palm facing outward and the elbow flexed at 90°. The test result is positive when the patient cannot hold this position, with the back of the hand hitting the patient's back. The sensitivity and specificity of this test are 100% when the subscapularis is fully torn.

In the Gerber push-with-force test, the shoulder is placed in the same position as the liftoff test; however, the patient must keep the hand away from the back and must resist a push in the palm of the hand. EMG shows that this is the optimal subscapularis isolation test, with minimal activation of the pectoralis and latissimus dorsi muscles.

To identify the long head of the biceps tendon, use the Speed palm-up test.

In the speed palm-up test, the shoulder is placed at 90° of flexion with the elbow in extension and the forearm in supination, bringing the palm of the hand up. The flexion of the shoulder is resisted. The test result is positive if the maneuver produces pain. The sensitivity of this test is 63%, but its specificity is only 35%. Its positive predictive value is 43%, and its negative predictive value is 55%.

The Yergason test, in the author's opinion, is technically difficult and ineffective; therefore, it is not described.

Generally, the topographic tests are sensitive but not specific, except for the Gerber liftoff test. The combination of impingement tests and topographic tests helps determine whether a patient's symptoms are caused by rotator cuff disease.

As mentioned, the examination must be completed by instability and labrum tests, special tests (eg, thoracic outlet syndrome tests), a cervicothoracic spine examination, and a neurologic and vascular examination, but it is not the purpose of this section to describe them all.





Approach Considerations

A wide variety of imaging examinations are offered to image the rotator cuff, including plain radiography, arthrography, computed tomography (CT), magnetic resonance imaging (MRI), and ultrasonography (US). Each of them has advantages and limitations. Bone scintigraphy is not used routinely in imaging for rotator cuff disease.

To prescribe the most useful examination, one must start with a good clinical history and physical examination (see Presentation). Imaging should be used to confirm the anomaly and to describe its extension and the associated findings. The following sections briefly explain the indications, the technique, and the findings for each modality available for imaging the rotator cuff.[22, 23]

Plain Radiography

Plain radiography is not very specific or sensitive for rotator cuff disease, but it remains the first examination to perform.

Radiographs are used for gross evaluation of the mineralization of the bone, the alignment, posttraumatic changes, the normal variant of the acromion shape, the presence of degenerative changes, and the presence of fine soft-tissue calcifications that could be missed by other modalities. This is the most useful test in trauma situations or in patients with complete chronic tears. In the last stage of complete chronic rotator cuff tear, radiography could be the only imaging modality needed to confirm the diagnosis (see the image below).

In this patient's shoulder radiography, the humera In this patient's shoulder radiography, the humeral head no longer matches up with the glenoid because the rotator cuff is torn and the strong deltoid muscle is pulling the head superiorly toward the acromion. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

Plain films are acquired routinely in three planes (ie, neutral, internal rotation, and external rotation).


The main indication for arthrography is to identify complete rotator cuff tears and intra-articular infiltration of the corticoid. As a diagnostic tool, it is generally combined with CT arthrography.[24]

Arthrography is performed by injecting iodine contrast medium, air, or both into the glenohumeral joint. Either 8-12 mL of contrast or 3-4 mL of contrast with 10-12 mL of air is injected to distend the joint capsule. If both air and contrast material are injected, the term double-contrast study is used. Then, plain films are taken in different positions, such as external rotation, internal rotation, and subacromial views, before and after motions. (See the image below.)

This image depicts the channel between the articul This image depicts the channel between the articular capsule and the subacromial-subdeltoid bursa in a complete rotator cuff tear. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

In the presence of a complete tear, the contrast floods from the glenohumeral joint into the subacromial-subdeltoid bursa (see the images below).

Even if the channel cannot be always identified, t Even if the channel cannot be always identified, the presence of contrast medium in the subdeltoid-subacromial bursa signals the presence of a complete rotator cuff tear. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
CT-arthrography of the shoulder in the axial plane CT-arthrography of the shoulder in the axial plane. Note the presence of air and contrast in the subacromial-subdeltoid bursa. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

With a partial tear, the contrast is seen as a line or a small filled cavity within the tendon, but without contrast in the subacromial-subdeltoid bursa. This finding is more difficult to demonstrate in a complete tear. Intratendon tears and tears on the superior aspect of the tendon (bursal side) are not visualized with this technique. Arthrography can also provide some information about the long portion of the biceps tendon, loose bodies, and synovial disorders (eg, inflammatory synovitis, osteochondromatosis, and villonodular pigmented synovitis).

CT Arthrography

CT arthrography, though highly accurate (100% sensitivity, 100% specificity) in depicting complete rotator cuff tears, is limited in the evaluation of tendinitis and partial tears, for which its sensitivity drops to 17-43%. On the other hand, this test yields more information than conventional arthrography regarding the joint itself and the soft tissues around it. The ability to evaluate the labrum, the glenohumeral ligaments, the long head of the biceps tendon, and the bony structures, as well as the presence of loose bodies, makes this a useful study.

CT arthrography is performed in exactly the same way as double-contrast (air and iodine contrast) arthrography but is followed by tomodensitometry imaging (CT scanning). For this examination, the shoulder is imaged in the axial plan in internal and external rotation. Thin slices as small as 2-3 mm are acquired throughout the entire joint. With new CT technology, it has become easy to reformat images in multiple planes.

The semiologic signs of rotator cuff tears are essentially the same as seen with conventional arthrography. The presence of contrast in the subacromial-subdeltoid space confirms the diagnosis of complete rotator cuff tears (see the images below). The contrast can also facilitate determination of the size and location of the tear and thereby help the surgeon plan operative treatment. Degenerative findings such as osteophytes, geodes, sclerosis, and articular space narrowing are also well depicted.

CT-arthrography of the shoulder in the axial plane CT-arthrography of the shoulder in the axial plane. Note the presence of air and contrast in the subacromial-subdeltoid bursa. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
CT-arthrography of the shoulder in the axial plane CT-arthrography of the shoulder in the axial plane. Note the presence of air and contrast in the subacromial-subdeltoid bursa. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

In conjunction with conventional arthrography, CT arthrography can identify labral and glenohumeral ligament tears. The presence of contrast between the labrum and the articular space indicates the presence of a tear. The axial views also permit good visualization of the long head of the biceps tendon in its groove. Therefore, subluxation of this tendon, or rupture, can also be diagnosed. Finally, the shape of the acromion can be evaluated on the oblique sagittal reformatted study, which requires a special acquisition.

MRI and MR Arthrography

MRI is the state-of-the-art diagnostic tool for a full evaluation of the shoulder. It allows fine evaluation of the bone marrow, tendons, muscles, ligaments, capsules, bursae, and labrum. It combines the advantage of visualization of the bony structures and of all the soft tissues about the shoulder and in any plane desirable.

With this imaging modality, the full continuum of rotator cuff disease, from simple tendinosis to complete tears, can be diagnosed. MRI is much more powerful than the previous modalities when used to identify partial tears, and it also can identify intratendon tears or tears on the bursal aspect of the tendon. As with CT and plain radiography, the bone structures resulting or contributing to the impingement syndrome can be evaluated.[24, 25, 26]

MRI can also yield information about retraction of the muscle, atrophy, bursitis, and bone marrow abnormalities (eg, edema, contusion), which all are associated findings of rotator cuff disease.

MRI is somewhat limited in the evaluation of the labrum and glenohumeral ligaments. Magnetic resonance (MR) arthrography is the study of choice for the evaluation of these structures.

Because this technique takes advantage of the properties of hydrogen protons submitted to a magnetic field and radiofrequency waves, the patient is not subjected to radiation exposure. Multiple sequences are available to highlight different substances, such as water, fat, blood, or solid structures. Mainly, spin-echo T1, spin-echo T2, and gradient-echo sequences, in axial, sagittal, and coronal oblique plans, are acquired in different combinations. Inversion recovery, fat saturation, and injection of gadolinium (intravenous or intra-articular) can be added if necessary.

MRI shows great detail of the anatomy in multiple planes. MRI also allows better visualization of the nature of a structure or an anomaly (ie, according to its intrinsic property). Therefore, the examiner should know some characteristics of the MRI signals for the most common structures.

Fat, methemoglobin, melamine, gadolinium, and some forms of calcium all are hyperintense in T1-weighted images, whereas water is hypointense. In T2-weighted images or in gradient echo, the liquids are hyperintense, as are most lesions, meaning that edema, inflammatory processes, tumors, tendinitis, and tendon tears are hyperintense in T2-weighted images and hypointense in T1-weighted images. Therefore, the presence of fluid in a bursa or articular joint is hyperintense in T2 or gradient echo and indicates inflammatory or posttraumatic fluid. A full-thickness tear of the tendon is demonstrated by a signal hyperintensity in T2 that extends throughout the tendon (see the image below).

Full-thickness tear of the supraspinatus as seen a Full-thickness tear of the supraspinatus as seen as a hyperintensity line through the full thickness of the tendon in a flash 2-dimensional MRI sequence in coronal oblique plane. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

Tendinitis is recognized as a gray signal in the tendon. Finally, calcification and cortical bone appear hypointense in all sequences (see the image below).

Calcifications are seen as hypointense foci in fla Calcifications are seen as hypointense foci in flash 2-dimensional images. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

MR arthrography is the gold standard for diagnostic imaging of a rotator cuff tear. It follows the same principle as CT arthrography. This modality can help identify labral tears (see the image below) and glenohumeral tears.

MRI arthrography can help to identify labral tears MRI arthrography can help to identify labral tears, as seen in this image. The contrast medium penetrates between the labrum and the articular surface. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

The size and morphologic features of rotator cuff tears may influence treatment selection and affect final outcomes. MR arthrography allows observation of these features and other intra-articular structures.

In one series, Toyoda et al[27] compared MRI with MR arthrography. To assess the utility of MRI in assessing size and morphologic features, 41 shoulders were retrospectively analyzed in 37 consecutive surgically treated patients (mean age, 63.2 years) who had MRI followed by MR arthrography. The maximum rotator cuff defect size in the anteroposterior direction defined transverse size, and the maximum defect size in the mediolateral direction defined longitudinal size. Sensitivity for detecting full-thickness rotator cuff tears was 90.2% for MRI vs 100% for MR arthrography.

MR arthrography also allowed morphologic classification of the torn tendon as blunt end, tapering end, indistinct end, horizontal tear, and global tear.[27] There was good agreement in classifying torn edges: the imaging findings agreed with findings at surgery. MR arthrography was more accurate in evaluating both rotator cuff tear size and morphologic features than MRI.

With the aid of fat-suppressed imaging, full thickness and partial cuff tears can be identified with 100% sensitivity and specificity. Fat-suppressed images also showed intratendinous contrast material imbibition in three torn cuffs with frayed, friable tendon margins. Fat suppression in MR arthrography is valuable in the differentiation between partial and full-thickness cuff tears and in the detection of small partial tears of the inferior tendon surface.

De Jesus et al performed a meta-analysis to compare the diagnostic accuracy of MRI, MR arthrography, and US in diagnosing rotator cuff tears.[23] Their analysis showed that MR arthrography is the most sensitive and specific technique for diagnosing both full- and partial-thickness rotator cuff tears and that US and MRI are comparable in both sensitivity and specificity.

In this study, summary ROC (receiving operating characteristic) curves for MR arthrography, MRI, and US for all tears showed that the area under the ROC curve was greatest for MR arthrography (0.935), followed by US (0.889) and then MRI (0.878); however, pairwise comparisons of these curves showed no significant differences between MRI and US.[23]


The advantages of US in the evaluation of the rotator cuff lie in its low cost, high availability, and high resolution.[25, 28, 29, 30] US is a dynamic study for demonstrating impingement syndrome. Its main disadvantages are that it is time-consuming for the radiologist and is operator-dependent. In addition, it cannot study bone structures, because sound does not penetrate bone very well.

With US, the normal tendon is an echoic structure, whereas the cartilage and fluids are hypoechoic. All of the tendons, bony landmarks (eg, humerusm and greater tuberosity), and intra-articular or intrabursal effusions are easily recognized. Tendinitis is diagnosed when the tendon loses its echogenicity and becomes diffusely hypoechoic. Calcifications appear as bright foci within the tendon, accompanied by a posterior shadowing, because the sound cannot pass through the calcium.

The main, and most sensitive, sign of a complete rotator cuff tear is an interruption in the tendon that fills with fluid, producing a hypoechogenic focus extending from the cartilage surface to the subdeltoid-subacromial bursa (see image below). The secondary signs include the uncovered cartilage (cartilage appears hyperechoic at the site of the tear), bursa herniation, loss of convexity of the tendon and bursa, and effusion within the glenohumeral articulation and the subacromial-subdeltoid bursa.

Ultrasound is another modality to demonstrate a co Ultrasound is another modality to demonstrate a complete rotator cuff tear, as seen here with a gap of more than 2 cm between both extremities of the torn tendon. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

The diagnosis of a partial rotator cuff tear is made when the hypoechoic or bursal herniation does not cross the full width of the tendon. US also allows the operator to demonstrate, in real time, the impingement of the supraspinatus tendon on the acromion when the arm is positioned in internal rotation and moved in abduction or flexion.



Approach Considerations

Conservative treatment of the degenerative rotator cuff involves the following:

  • Pain relief
  • Avoidance of painful motions and activities
  • Simple analgesics and nonsteroidal anti-inflammatory drugs (NSAIDs)
  • Manual physical therapy for the glenohumeral, scapulothoracic, acromioclavicular, and sternoclavicular joints and the parascapular and scapula-stabilizer muscles
  • Subacromial corticosteroid injection
  • Bupivacaine suprascapular nerve block
  • Restoration of motion and normal scapulohumeral rhythm
  • Stretching of the glenohumeral capsule and muscles
  • Manual therapy of the cervicodorsal spine (often necessary because of its close relation with the shoulder)

Patients with more advanced rotator cuff disease or a more significant injury may not respond to conservative therapies. If the patient believes that his or her quality of life is being significantly impacted by the shoulder dysfunction, then surgical intervention is a reasonable consideration.

In some cases, simple debridement of a frayed or partially torn cuff tendon, along with smoothing of the undersurface of the acromion (acromioplasty) above the tendon, may be all that is needed. More significant partial tearing (>50% of tendon thickness) and complete tears require reattachment of the tendon ends back to the humeral head.[31, 32, 33]

Conservative Therapies

Physical therapy

Physical therapy can be a useful adjunct in the conservative treatment of rotator cuff degeneration. Although numerous studies have been performed on conservative treatment of and surgical approaches to the painful shoulder and, more specifically, the rotator cuff, the conclusion of a review of randomized, controlled trials of common interventions for painful shoulder was that little evidence supports or refutes their efficacy.[34, 35, 36, 37]  Drawing firm conclusions about the efficacy of any of these interventions remains difficult for the following reasons (among others):

  • Lack of definition and strict diagnostic criteria for the different painful shoulder conditions
  • Uncertain validity of randomization procedures
  • Absence of blinding
  • Unavailability of valid scales for outcome measurement
  • Heterogeneous study populations

In their approach to conservative patient treatment, clinicians must be critical and try to use an evidence-based medicine approach as much as possible.[38] The clinician must also use a combination of experience and intuition to compensate for the lack of scientific evidence supporting the different therapeutic modalities.

Pharmacologic pain relief


Acetaminophen is recommended as initial treatment because of the toxicity associated with NSAIDs, the need for an analgesic rather than anti-inflammatory effect, the lower cost of a simple analgesic, and the chronicity of degenerative rotator cuff disease.

NSAIDs are known to be effective in reducing pain and improving function and range of motion (ROM), but they may exert their effect through their analgesic rather than their anti-inflammatory properties. One study with poor methodologic quality showed no short-term superiority of NSAIDs over acetaminophen in the treatment of painful shoulder syndrome.

Long- and short-term studies comparing the efficacy of NSAIDs with that of acetaminophen for osteoarthritis of the knee have shown similar efficacy for the two. Moreover, the finding that even the presence of inflammatory signs did not predict a better response to treatment with NSAIDs suggests that improvements are not necessarily dependent on an anti-inflammatory effect.

The analgesic effect of acetaminophen is mediated by prostaglandin inhibition. Recommendations for its use are as follows:

  • Adult dose - 325-650 mg PO q4-6hr or 1000 mg tid/qid; not to exceed 4 g/day
  • Pediatric dose - For those younger than 12 years, 10-15 mg/kg PO q4-6hr prn, not to exceed 2.6 g/day; for those older than 12 years, 325-650 mg PO q4hr, not to exceed five doses in 24 hours
  • Contraindications - Documented hypersensitivity; known glucose-6-phosphate dehydrogenase deficiency
  • Interactions - Rifampin can reduce analgesic effects of acetaminophen; coadministration with barbiturates, carbamazepine, hydantoins, and isoniazid may increase hepatotoxicity
  • Pregnancy category B - Usually safe, but benefits must outweigh the risks
  • Precautions - Hepatotoxicity possible in chronic alcoholism following various dose levels; severe or recurrent pain or high or continued fever may indicate a serious illness; acetaminophen is contained in many over-the-counter products, and combined use with these products may result in cumulative doses exceeding recommended daily dose

Nonsteroidal anti-inflammatory drugs

Numerous studies on the efficacy of NSAIDs for different shoulder conditions have been published, but because most of them have poor methodologic quality, no conclusions can be drawn about the efficacy of these agents.

Review articles, using strict inclusion criteria based on the quality of the methodology, concluded that the trials with the best methodology show a superior short-term (2 weeks) efficacy for NSAIDs as compared with placebo; however, at 4 weeks, there were no statistically significant differences. Therefore, a short course (10-14 days) of NSAIDs is indicated as a second-line treatment. No evidence supports longer use.

If pain persists, other therapeutic modalities should be considered. A comparison between different types of NSAIDs did not evidence that any given NSAID was superior to others in terms of efficacy. Therefore, an NSAID with the fewest adverse effects, such as a cyclooxygenase (COX)-2 selective agent or an NSAID combined with a prostaglandin E1 analogue (diclofenac-misoprostol), should be the drug of choice.

In an aging population taking additional medication that may interact with NSAIDs, drug interactions must be avoided. Some 40-60% of drugs consumed are over-the-counter medications—most often analgesics and NSAIDs, which increase the risk of potential adverse gastrointestinal (GI) side effects. The patient should be asked whether he or she is taking any medications concomitantly, such as the following:

  • Anticoagulants (hemorrhage)
  • Corticosteroids (peptic ulcer)
  • Diuretics and antihypertensives (decreased blood pressure control)
  • Angiotensin-converting enzyme (ACE) inhibitors (acute renal failure [acute kidney injury])
  • High-dose methotrexate (increased methotrexate toxicity)
  • Lithium, digoxin, aminoglycosides (decreased renal clearance)
  • Phenytoin (decreased albumin binding)
  • Antacids (decreased NSAID levels)

NSAIDs should be avoided, if possible, in elderly patients who have congestive heart failure or renal or hepatic dysfunction and who are taking other medications.

If the patient has no contraindications to the use of ibuprofen, it is usually the drug of choice for the treatment of mild-to-moderate pain. It inhibits inflammatory reactions and pain by decreasing prostaglandin synthesis. Recommendations for the use of ibuprofen are as follows:

  • Adult dose - 400 mg PO q4-6hr, 600 mg PO q6hr, or 800 mg PO q8hr while symptoms persist; not to exceed 3.2 g/day
  • Pediatric dose - For age 6 months to 12 years, 10-70 mg/kg/day PO divided tid/qid; begin at lower end of the dosing range and titrate upward; not to exceed 2.4 g/day; for older than 12 years, administer as in adults
  • Contraindications - Documented hypersensitivity; peptic ulcer disease, recent GI bleeding or perforation, renal insufficiency, or high risk of bleeding
  • Interactions - Coadministration with aspirin increases risk of inducing serious NSAID-related adverse effects; probenecid may increase concentrations and, possibly, toxicity of NSAIDs; may decrease effect of hydralazine, captopril, and beta blockers; may decrease diuretic effects of furosemide and thiazides; may increase risk of methotrexate toxicity; phenytoin levels may be increased when administered concurrently; monitor prothrombin time closely (instruct patients to watch for signs of bleeding)
  • Pregnancy category B - Usually safe but benefits must outweigh the risks.
  • Precautions - Category D in the third trimester of pregnancy; to be used with caution in congestive heart failure, hypertension, and decreased renal and hepatic function; to be used with caution in anticoagulation abnormalities or during anticoagulant therapy

Celecoxib primarily inhibits COX-2, which is considered an inducible isoenzyme—that is, it is induced during pain and inflammatory stimuli. Inhibition of COX-1 may contribute to NSAID GI toxicity, but at therapeutic concentrations, celecoxib does not inhibit the COX isoenzyme; therefore, GI toxicity may be decreased. The lowest dose of celecoxib should be sought for each patient. Recommendations for its use are as follows:

  • Adult dose - 200 mg/day PO qd; alternatively, 100 mg PO bid
  • Pediatric dose - Not established
  • Contraindications - Documented hypersensitivity
  • Interactions - Coadministration with fluconazole may cause an increase in celecoxib plasma concentrations because of inhibition of celecoxib metabolism; coadministration of celecoxib with rifampin may decrease plasma concentrations
  • Pregnancy category C - Safety for use during pregnancy has not been established.
  • Precautions - Use with caution in patients with compromised cardiac function, hypertension, and conditions predisposing to fluid retention, because NSAIDs may cause fluid retention and peripheral edema; use with caution in patients with severe heart failure and hyponatremia, because NSAIDs may deteriorate circulatory hemodynamics; use with caution in the presence of existing controlled infections, because NSAIDs may mask the usual signs of infection; evaluate symptoms and signs suggesting liver dysfunction, cardiac dysfunction or renal dysfunction

Ketoprofen is used for relief of mild-to-moderate pain and inflammation. Small initial dosages are indicated in small and elderly patients and in persons with renal or liver disease. Doses of more than 75 mg do not increase therapeutic effects. Administer high doses with caution and closely observe patients for response. Recommendations for the use of ketoprofen are as follows:

  • Adult dose - 25-50 mg PO q6-8hr prn; not to exceed 300 mg/day
  • Pediatric dose - For those aged 3 months to 14 years, 0.1-1 PO mg/kg q6-8hr; for those older than 14 years, administer as in adults
  • Contraindications - Documented hypersensitivity; GI disease
  • Interactions - Coadministration with aspirin increases risk of inducing serious NSAID-related adverse effects; probenecid may increase concentrations and, possibly, toxicity of NSAIDs; may decrease effect of hydralazine, captopril, and beta blockers; may decrease diuretic effects of furosemide and thiazides; may increase risk of methotrexate toxicity; phenytoin levels may be increased when administered concurrently; monitor prothrombin time closely (instruct patients to watch for signs of bleeding)
  • Pregnancy category B - Usually safe, but benefits must outweigh the risks.
  • Precautions - Category D in third trimester of pregnancy; use with caution in patients with congestive heart failure, hypertension, and decreased renal and hepatic function; use with caution in patients with anticoagulation abnormalities or during anticoagulant therapy

Ultrasound therapy

Ebenbichler showed in a randomized, doubled-blind, placebo-controlled study that the use of pulsed ultrasound 5 times a week for 15 minutes at a time (0.89 MHz frequency, 2.5 W/cm2, pulsed mode 1:4) significantly resolves calcification of the shoulder, decreases pain, and improves the short-term quality of life.[39, 40]

Long-term follow-up did not reveal significant differences; however, in the long term, the symptoms of calcifying tendinitis may be self-limiting and may improve independently from the resolution of the calcium deposit. This theory may explain why the use of ultrasound is only significantly effective in the short term. The short-term efficacy of ultrasound therapy has been demonstrated only in calcifying tendinitis. Its efficacy in other shoulder disorders has not been shown.

Extracorporeal shockwave therapy

Shock waves were used first for the treatment of delayed union and nonunion of fractures by stimulating osteogenesis.

In an uncontrolled study, extracorporeal shockwave therapy (ESWT) with 1500 impulses of 0.28 mJ/mm2 reportedly disintegrated calcium deposits partially or completely in 62% of patients, and 75% had significant improvement in pain, power, ROM, and shoulder function. The authors of the study concluded that a larger-scale, placebo-controlled trial should be conducted to analyze the benefits of this modality.

A prospective, randomized, controlled study using a valid functional shoulder scale showed the efficacy of ESWT. At 3-6 months, significant improvement occurred in pain and function. At 6 months, calcium deposits disappeared or disintegrated in up to 77% of patients' radiographs. Comparing different regimens of shockwaves, the authors concluded that the improvement in pain and function and the radiologic disintegration of calcification were dose-dependent.[41, 42, 43, 44, 15]

ESWT appears to be a promising treatment for calcifying tendinitis. Like ultrasound therapy, it has not been definitively shown to be efficacious in other shoulder conditions.


Randomized, controlled studies have shown the efficacy of topical steroids, NSAIDs, and acetic acid iontophoresis, as compared with placebo, in different musculoskeletal disorders (MSDs); however, the studies were not specifically on rotator cuff disease. Moreover, a subsequent trial did not show any difference in outcomes between no treatment and treatment with acetic acid iontophoresis followed immediately by 9 sessions of ultrasound therapy in a constant mode (0.8 W/cm2 at a frequency of 1 MHz for 5 min) over a period of 3 weeks.

Some authors could not show any effect of iontophoresis on steroid migration through in-vivo and in-vitro studies, whereas others did. Thus, no conclusions can be made regarding the efficacy of iontophoresis in the treatment of rotator cuff disease.

Subacromial corticosteroid injection

As with NSAID therapy, many of the studies on the efficacy of corticosteroid injection for various shoulder conditions are of poor methodologic quality. Green, van Der Heijden, and Sibilia performed a systematic review of all the randomized clinical trials on corticosteroid injection.[45, 46, 47, 48] Although these studies selected essentially the same trials, their conclusions differ because of the different assessment methods. Two of these articles suggested that corticosteroid injection may be superior to placebo in the short-term treatment of rotator cuff tendinitis, whereas one suggested that no conclusive evidence was found regarding the efficacy of corticosteroid injection.

Subacromial corticosteroid and local anesthetic agent injection also appear to be more effective than an injection of a local anesthetic alone, though some authors disagree. Corticosteroid injection also appears to be significantly more effective than NSAIDs. Therefore, subacromial corticosteroid injection appears indicated when pain persists after simple analgesics and NSAIDs have been used.[49]

Because some authors have reported poorer surgical outcome in patients who have received three or more corticosteroid injections, the recommendation is that no more than two injections be given. No trials have compared the different routes of corticosteroid injection; thus, the physician should select his or her preferred route. Additionally, no trial has compared the efficacy of different corticosteroids. Triamcinolone acetonide is the agent most frequently studied.

The action mechanism is inhibition of prostaglandin formation by selective COX-2 activity. The optimal dose has not been evaluated. Recommended doses range from 20 to 80 mg in the different trials. The author recommends 20-40 mg of triamcinolone acetonide. Adverse effects can be local or systemic. Although systemic adverse effects can occur following a subacromial injection, only local adverse effects are discussed here. Possible adverse effects include the following:

  • Dermal atrophy
  • Necrosis and loss of pigmentation
  • Synovitis
  • Septic arthritis
  • Hemarthroses
  • Cartilage damage and degeneration
  • Tendon rupture
  • Charcot arthropathy

Bupivacaine suprascapular nerve block

The bupivacaine suprascapular nerve block is a relatively unknown, though effective, method to treat different painful shoulder disorders. A few randomized controlled trials have demonstrated its efficacy for painful shoulder associated with rheumatoid arthritis, for chronic rotator cuff disease, and for frozen shoulder.

Preliminary data from a study on chronic impingement syndrome conducted at the Montreal Rehabilitation Institute showed its efficacy as compared with placebo. At 3 months, a significant improvement in pain and function, measured by a valid functional shoulder scale, was observed. The efficacy of this procedure is supported by randomized controlled studies, and it appears to be a promising approach to the treatment of rotator cuff disease.[50]

The technique for nerve block is very inexpensive, simple, and safe. It consists of injecting 10 mL of bupivacaine 0.5% in the supraspinatus fossa of the scapula to produce an indirect suprascapular nerve block. In rotator cuff disease, two injections are administered 4 weeks apart.

Surgical Intervention

Open repair

Rotator cuff repair is commonly accomplished by performing an open surgical procedure, which typically requires a 5- to 10-cm incision at the top of the shoulder. The deltoid muscle is split, and the undersurface of the acromion is smoothed. Strong stitches are placed in the torn ends of the rotator cuff tendons, and they are attached back to the bone of the humerus through specially created tunnels or commercially available suture anchors.

Because the entire shoulder cannot be visualized through the open approach, many surgeons perform an initial diagnostic arthroscopy of the shoulder at the time of the repair to be sure that no other coexisting problems are present within the shoulder that could also be addressed at the open procedure. This technique may be performed in an inpatient setting or in an outpatient surgery facility, providing that the patient is comfortable enough to return home the same day.

Standard tendon repair techniques combined with anterior acromioplasty, postoperative limb protection, and monitored physiotherapy can produce consistent and lasting pain relief and improvement in ROM. Open rotator cuff repair has been known to have excellent outcomes and patient satisfaction since the early 1980s. Romeo et al reported 94% patient satisfaction 4 years after open rotator cuff repair, with lasting relief of pain and improved function.[51]

In another series, Baysal reported that 96% of patients were satisfied or very satisfied with the results of their repair; 78% of patients who were working before surgery returned to work without modification by 1 year postoperatively. For the most part, patient age and size of tear did not influence postoperative ROM or health-related quality of life.[52]


Arthroscopic surgery, first used to treat conditions of the knee, has become quite common for treating many knee, shoulder, elbow, wrist, hip, ankle, and foot problems.[53, 54, 55, 56, 57]

Arthroscopic treatment of rotator cuff disease initially consisted of rotator cuff inspection and debridement and arthroscopic acromioplasty. If a repairable rotator cuff tear was discovered, an open or miniopen repair of the tendon was then performed. As surgeons' skills improved and more specialized instrumentation was developed, it became possible to fix relatively small tears by using arthroscopic techniques to insert anchors, pass sutures, and tie knots. In current practice, surgeons can use these arthroscopic techniques in the shoulder to repair even large rotator cuff tears.

Arthroscopic rotator cuff repair (see the images below) is a technically challenging procedure that requires advanced arthroscopic surgical skills, careful preoperative planning, and a stepwise, systematic approach.[58, 59] The procedure may be performed with the patient in a "beach chair" (sitting) or in a lateral decubitus position. Usually, the patient is under general anesthesia.

View of large tear from posterior (behind). Socket View of large tear from posterior (behind). Socket is to the right. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
Visualizing torn rotator cuff from within the join Visualizing torn rotator cuff from within the joint. The biceps tendon is running vertically on the left. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
Motorized burr removing under-surface of acromion. Motorized burr removing under-surface of acromion. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
View of large tear from the "50 yard line." Courte View of large tear from the "50 yard line." Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
The side-to-side stitches begin to close the large The side-to-side stitches begin to close the large tear defect. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
An arthroscopic knot-tying instrument is used to p An arthroscopic knot-tying instrument is used to pass tie knots in the suture to secure the repair. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
Small metallic anchors (5 mm) with sutures attache Small metallic anchors (5 mm) with sutures attached are then inserted into the humerus at the site desired for tendon reattachment. The anchors are recessed below the surface, so only the suture is visible. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.
Sutures are anchored with the metallic anchors. Co Sutures are anchored with the metallic anchors. Courtesy of Dr Thomas Murray, Orthopaedic Associates of Portland.

Small (5-mm) incisions are created in the back, side, and front of the shoulder, and the arthroscope and instruments may be switched between each of these positions as necessary.

A complete diagnostic arthroscopy with bursoscopy (inspection of bursa) is initially performed. Care is taken to inspect the biceps tendon within the shoulder, the fibrous ring or labrum that surrounds the glenoid, the capsule and ligaments, the cartilage surfaces of the head and glenoid, and the rotator cuff tendons. Any pathology is addressed only after a complete inspection, so as not to miss any significant findings.

Careful preoperative radiographic evaluation of the shape and size of the acromion, along with a notation of any spurs, serves as a guide for the extent of any acromioplasty (undersurface smoothing) necessary. Because the arthroscope magnifies the structures seen, irregularities in the surface of less than 1 mm can be seen and are removed. The goal is to smooth and flatten the undersurface of the acromion to provide more room for the repair and to relieve pressure from the healing tendon.

An overly aggressive acromioplasty must be avoided because excessive removal of the anterior acromion can result in the humeral head sliding forward, up, and out of the socket (anterosuperior subluxation).

The rotator cuff tear is then visualized through the lateral (side) portal from the "50-yard-line view." The size and pattern of the tear are assessed. Any thin or fragmented portions are removed, and the area where the tendon will be reattached to the bone is lightly debrided to encourage new blood vessel ingrowth for healing.

The sutures are once again passed through the tendon and systematically tied. The sutures pull the tendon down to the prepared bone surface, closing the defect. This completes the repair.

At the completion of the procedure, the shoulder is injected with a long-acting local anesthetic to assist with postoperative pain management. Each portal incision is closed with a single nylon stitch and covered with a sterile bandage tape, followed by a dry, sterile dressing. A cryotherapeutic shoulder pad (eg, Cryocuff) is applied to provide postoperative cold therapy. This assists in management of pain and swelling. Finally, a sling (eg, Don Joy UltraSling II) is applied for immobilization and protection. The patient is then taken to the recovery room.

Arthroscopic rotator cuff repair has achieved good-to-excellent results in a large percentage of patients (95% in one series), with the results being independent of tear size. U-shaped tears repaired by margin convergence have been shown to have results comparable to those of crescent-shaped tears repaired directly by a tendon-to-bone technique. There is a rapid return to full overhead function after arthroscopic rotator cuff repair (average, 4 months for all tear sizes). A delay between the time of injury and the time of diagnosis, even of several years, is not a contraindication for arthroscopic rotator cuff repair.

The results from one study suggest that patients who underwent arthroscopic rotator cuff repair with or without acromioplasty experienced no difference in function or quality of life.[60]

In a prospective study of 88 patients, Castricini et al showed that augmentation of a double-row arthroscopic surgical repair of a small to medium-sized tear of the rotator cuff with autologous platelet-rich fibrin matrix (PRFM) did not improve the healing.[61]  

In most studies, injection of platelet-rich plasma (PRP) for rotator cuff tendinopathy has not demonstrated significant clinical benefit as compared with other nonoperative treatments; however, PRP injection appears to improve rotator cuff tear healing and reduce early postoperative pain when used to augment surgical repair, though it does not significantly enhance postoperative shoulder function.[62]

Prosthetic implants (tissue engineering)

Tissue-engineering techniques are being used to develop therapies for tendon reconstruction. Biologic and synthetic scaffolds can both repair tendon defects and improve healing by allowing for the regeneration of the tendon's natural biologic composition to restore its mechanical capacity. This process can be further enhanced through augmentation methods such as cell seeding, growth factor implantation, and gene therapy.

Many engineered prosthetic materials are currently in use, but treatment for massive irreparable rotator cuff tears remains challenging. Interposition grafting and superior capsule reconstruction are among the surgical options suggested for such tears.[63, 64]

In a prospective multicenter study of 33 patients with chronic degenerative partial-thickness tears of the supraspinatus tendon, Schlegel et al evaluated clinical and radiologic outcomes of placement of a bioabsorbable collagen implant after arthroscopic subacromial decompression without repair.[65]  At 1 year, clinical scores were significantly improved, and tendon thickness increased by 2.0 mm. MRI showed complete healing in eight patients and a considerably reduced defect size in 23; one lesion remained stable. No serious implant-related adverse events were reported.


Patients should be instructed to limit their activities so as to ensure rest of the affected shoulder. Patients should be referred to a physical therapist for conservative treatment and postoperative therapy.

Long-Term Monitoring

Outpatient follow-up care should be arranged with an orthopedic surgeon and rehabilitation services to continue conservative therapy. A follow-up reassessment examination 6 weeks after beginning conservative therapy is essential to determine if treatment is successful or if further surgical treatment is needed.