Although the shoulder is no less susceptible than other diarthrodial joints to arthritic conditions, arthritis is not a typical cause of shoulder pain. Probably as few as 5% to 10% of patients with shoulder pain have arthritis as a part of a polyarticular systemic disease, or a monarticular disease process. Painful dysfunction is more often a disorder of one or more of the soft tissue elements of the shoulder.

There exist more than 100 different types of arthritis. They may be defined as inflammatory or degenerative, polyarticular or monarticular, and acute or chronic. Distinction is made from historical factors, physical findings, radiographic analysis, and tissue sampling, including the synovial fluid analysis and serum studies.

The predominant symptom of patients with arthritis is pain, usually intensifying with use and interfering with sleep. The shoulder motion is restricted by synovial membrane inflammation and increased fluid production. The synovium and capsule are richly innervated structures easily irritated by distention from the accumulation of synovial fluid. Exacerbation of discomfort with rotary movement of the elbow at the side is a characteristic physical finding of an inflammatory condition of the glenohumeral joint. Degenerative conditions are more tolerant of passive range of motion. Tenderness over the posterior joint line, as noted by Neer, is notable in glenohumeral osteoarthritis.⁵⁷⁷ Tenderness over the anterior joint line, lateral and inferior to the coracoid process, is very suggestive of an inflammatory disease. It is not commonly elicited in rotator cuff disease or in noninflammatory conditions.

Green and Norris have provided an excellent review of imaging strategies to be employed for patients with glenohumeral arthritis.²⁸⁶ Initial evaluation of the painful shoulder will almost always include plain radiographs.⁴⁰⁹ The views most commonly employed by the orthopedic surgeon are the true anteroposterior (AP) of the glenohumeral joint orthogonal to the plane of the scapula and the transaxillary lateral. Glenohumeral arthritis is easily distinguished from other soft tissue conditions and is classified based on features identified on plain films (Fig. 12-1).⁶⁸⁷ These include alignment and
relation of the humerus to the glenoid and acromion process, the width of the articular cartilage, osseous erosions, productive changes including osteophytes, and the presence of soft tissue swelling and calcification.

Distinction of the articular cartilage space is best made when the X-ray beam is directed tangent to the joint surface. For the shoulder, this requires accurate positioning of the patient, shoulder, and extremity. Once the plane has been determined, the humerus can be externally rotated to approximately 35 degrees (Fig. 12-2). This will profile the load-bearing portion of the articular surface so that the earliest narrowing does not escape detection, and it is also the position favoring visualization of marginal humeral head osteophytes.⁴¹⁰ Apple et al. suggested the weighted abduction Grashey shoulder method as a more sensitive means of detection of loss of articular cartilage.²⁵ When the disease is in more advanced stage, positioning the humerus by external rotation is difficult and cannot be done exclusive of the position of the scapula (Fig. 12-3). For Nelson et al., plain radiography, as opposed to magnetic resonance imaging (MRI), computed tomography (CT) arthrogram, and ultrasonography, was most useful for the diagnosis of osteoarthritis.⁵⁸³ Atlases and classification systems have been proposed for the study and treatment of osteoarthritis.¹⁴,¹⁵ The Samilson-Prieto classification for glenohumeral arthritis offers advantages due to its simplicity and reliability.⁸⁴,⁴⁰⁶,⁷¹⁵

Although shoulder arthrography is employed mostly for the detection of rotator cuff tears, the articular surfaces and contours of the synovial lining of the joint can be easily visualized.⁸⁰² The arthrographic features may be enhanced by the use of poly-tomography or CT.⁴⁵⁶,⁷⁴² Filling defects may be observed in the cartilaginous surfaces, and the thickness of the residual cartilage is estimated. Magnetic resonance arthrography using gadopentetate dimeglumine may offer a more sensitive method for the detection of intra-articular abnormalities.⁴³²,⁶¹⁰ These techniques, for practical purposes, are rarely employed by the author for the routine assessment of glenohumeral arthritis.

The ultrasound study of the glenohumeral articulation has a very limited, yet undefined, role in the evaluation of the articular surfaces. It may prove more useful in the detection of synovitis and periarticular soft tissue disorders.³⁶⁶

Although a CT scan is generally not required for the diagnosis of articular disorders, it can be quite helpful to delineate the extent of bone involvement and to assess glenoid orientation (Fig. 12-4).²⁵¹,⁵⁶¹ Glenohumeral morphology can be detailed, as can the quality and quantity of the glenoid. CT scanning will also help detect subtle cortical erosions, deposits of calcium, and the presence of newly formed bone. A CT scan may more accurately assess glenohumeral relationships when extreme stiffness or deformity does not allow the shoulder to be positioned for optimal plain film study. Combining fluorine-18-fluorodeoxyglucose (FDG) positron emission tomography (PET) with CT proved 100% reliable for the detection of osteoarthritis and RA in patients with shoulder pain.⁵⁵² FDG accumulates diffusely about the shoulder in patients with clinical symptoms and signs of osteoarthritis, although false-positive scans are not infrequent.⁸⁴⁴

Perhaps, the best imaging study for detecting early chondral injury or damage is MRI, even though the assessment of the articular cartilage can be technically challenging.²⁷⁴ Monographs to which the readers are referred to give the details on MRI of articular cartilage and osteoarthritis.³⁷⁰,⁶⁸⁴ Hayes stated that an accurate MRI scan of articular cartilage requires
good spatial resolution for the detection of small defects, good subject contrast and image contrast for the detection of signal intensity changes in articular cartilage, reliable distinctions for the detection of signal intensity changes in articular cartilage, and reliable distinction between articular cartilage and adjacent subchondral bone and joint fluid.³²³ Broderick’s study drew attention to the tendency for cartilage abnormalities to be underestimated with MRI when compared with arthroscopic observations.⁷⁷ Amin made a similar observation when MRI was compared with radiographic progression.¹⁷ The use of injectable contrast agents to enhance the detection accuracy of articular cartilage is not generally necessary.¹¹⁸,⁴³³,⁵²³

There is a consensus that normal articular cartilage can be imaged by MRI, perhaps to the extent of “zonal” analysis.¹⁴³,⁵⁴⁶,⁶¹⁷,⁷⁰³ There are several means of assessing articular cartilage morphology using MRI techniques with each having advantages and disadvantages in the quest to distinguish articular cartilage from subchondral bone.¹⁹⁵,²⁶⁶,⁴⁶⁹,⁶⁴²,⁶⁷⁹ Three-dimensional imaging data can provide information about joint surface topography utilizing refined techniques.³³,¹¹⁷,⁶²²,⁶²³,⁶⁶⁷,⁷⁹⁷ Quantitative MRI is under investigation as a means of earlier detection and as a tool for monitoring the response of the articular cartilage to certain clinical conditions and to treatment.²²,⁸⁵,¹⁵⁰,²⁰⁵,²⁰⁶,²⁰⁷ and ²⁰⁸,²⁵⁷,³⁰²,⁶⁵⁴,⁸²⁹ Advances have been made in cartilage matrix assessment using MRI quantitative dGEMRIC, diffusion tensor imaging, sodium MRI, T2 and T1rho mapping techniques.⁹⁸,²³⁶,²⁴²,²⁷¹,⁴⁶⁹,⁵³⁴,⁶⁶⁹,⁷⁷⁶,⁸⁵¹,⁸⁵⁹ Presently, the advantages of dGMERIC are recognized and its use is more widespread.⁶¹,⁵⁶³,⁸⁰¹,⁸⁵¹ T2 mapping with dedicated software is more sensitive than routine MRI for the detection of early pathological changes in articular cartilage.⁴⁹⁴

With a few exceptions, the earliest scientific studies of MRI of articular cartilage involved the knee.⁶⁵,¹⁴³,²⁹⁶,³⁰¹,³⁴⁵,⁵²³,⁵²⁴,⁶²²,⁶⁶⁷ The trend continues currently.¹⁹⁵,³²⁶,⁵⁸⁴,⁶⁷¹,⁶⁸⁵,⁶⁸⁶ Huber, in one of the earliest published reports on MRI of the normal shoulder, cited the failure of MRI to effectively define articular cartilage.³⁵⁵ The literature was replete with studies of MRI of the shoulder that had assessed features of the rotator cuff and capsulolabral morphology in impingement and instability disorders, respectively, and rarely mentioned the findings of the glenohumeral articular cartilage. Recent monographs and textbooks concerned with MRI of the shoulder and shoulder imaging inadequately address arthritis imaging by MRI, particularly the articular cartilage changes.⁷⁵³,⁸¹⁴,⁸⁹¹ The preferred timing, sequences, and technical parameters have not been determined with uniform success and are evolving.³⁴⁴ Therefore, MRI applications for glenohumeral articular disorders are not yet as extensive as for the knee. Zlatkin stated that MRI is rarely required to assess patients for glenohumeral osteoarthritis alone.⁸⁹² MRI arthrography utilizing gadolinium compounds may prove to be an effective technique to enhance the glenohumeral articular surfaces.²⁹⁵,⁴³³

A shortcoming of conventional radiographs is failure to detect soft tissue details, although in some cases swelling can be identified. Joint effusions may be suspected on plain films, but are best evaluated by ultrasound, CT, or MRI. The MRI is probably the most useful, accurate, reliable, and expensive method.⁸¹⁴,⁸⁶⁴

In normal shoulders, the MRI findings show only a thin rim of joint fluid around the biceps tendon sheath and narrow bands in the axillary recess and subscapularis bursa; the volume has been observed to be increased with age.⁷⁴³ The accumulation of glenohumeral joint fluid is abnormal; usually it is related to rotator cuff tears or osteoarthritis (Fig. 12-5A,B).⁷⁴³ The intravenous administration of gadopentetate dimeglumine will enhance the MRI of joint fluid, although the response is slower and less vigorous than synovium.⁸⁶³

The use of MRI allows direct visualization of inflammatory synovitis and synovial proliferation within the joint, as well as its penetration into the adjacent bone and periarticular soft tissues (Fig. 12-5C).⁶⁷³,⁸⁶⁵ Contrast enhancement with the use of gadolinium provides a more specific assessment of synovitis or pannus.⁴³⁰,⁶⁷³ The rate and intensity of synovial enhancement may vary, depending on the activity of the synovium as a reflection of the activity of the disease process, as well as the image timing.⁸⁶³

Radionuclide isotope uptake in tissue reflects the rate and extent of blood flow through the tissue and, in bone, is a measure of bone metabolism, especially osteoblastic activity. Scintigraphy, in particular triple-phase bone scan, may prove useful in distinguishing soft tissue inflammation from osteoarticular changes.²⁷⁰,⁴⁹⁵,⁸¹¹ It is very sensitive, but it does not distinguish acute inflammatory disease from bone and joint sepsis.⁸¹⁰ Technetium scans have been used to confirm the presence of RA before the appearance of radiographic abnormalities. 189 Indium-111 chloride has been used for the detection of RA as well as for following the course of disease activity.⁷³²,⁷⁵² Immune complex scintigraphy holds promise for monitoring disease activity and response to treatment.¹⁷¹,⁴¹⁵ Uptake defects can exist in osteonecrosis and can be enhanced with single photon emission computed tomography images.⁸¹²,⁸⁴⁹ White cell-labeled scans are useful when an infectious disease process is being considered.⁸¹¹,⁸³³

Serum studies may be helpful, but only rarely can a diagnosis be obtained from this information alone. Complete blood count, autoantibodies (rheumatoid factor [RF] and antinuclear antibody), uric acid and acute-phase reactant measurements, erythrocyte sedimentation rates (ESRs), and C-reactive protein (CRP) are most often performed as an initial screening battery and are most helpful when the diagnosis is in question.²⁵⁸,⁷³⁶ Acute-phase reactants are serum proteins formed in the liver and include coagulation, transport, complement, and miscellaneous reactive proteins. Their production is accelerated in the presence of inflammatory states and tissue necrosis.²⁵⁹,⁴⁴¹,⁷³⁷ The ESR is a rough, indirect quantification of their serum levels. The CRP level has been used more recently. Both are serum markers that reflect the extent or degree of inflammation and can be used for the assessment of disease activity or response to treatment over time.¹⁴²,⁴⁴¹ The CRP levels (normally less than 1 mg/dL) may exhibit moderate elevation (1 to 10 mg/dL) in most of the connective tissue diseases; marked elevation (more than 10 mg/dL) may signify an acute bacterial infection such as septic arthritis.⁵⁵³

Clinically detectable effusions are uncommon in most arthritides. When present or suspected, however, a fluid sample should be obtained and submitted for analysis.⁷⁸⁶ It should also be obtained any time the patient’s history, physical examination, and radiographs support the diagnosis of arthritis but a diagnosis has yet to be determined.²⁶¹ The findings of joint fluid analysis have been shown to change a clinically suspected diagnosis, and often treatment, in 20% of samplings.²¹⁵ Particularly in acute arthritis, synovial fluid analysis is of major diagnostic value.⁷⁸⁶

Arthrocentesis of the shoulder can be performed from the anterior or posterior approach. Anteriorly, the tip of the coracoid process is palpated. After local anesthetic instillation, an 18- or 21-gauge spinal needle is passed through the deltoid muscle at a point approximately 1.5 cm inferior and 1.5 cm lateral to the tip of the coracoid process. The cephalic vein is nearby and may be inadvertently punctured. Entry into the joint is gained with passage through the subscapularis muscle and the joint capsule. Posteriorly, the point of entry is two to three fingerbreadths inferior and one to two fingerbreadths medial to the posterolateral corner of the acromion process.
Directed toward the coracoid process, the needle will penetrate the deltoid and infraspinatus muscles to enter through the capsule into the joint.

As much fluid as possible is withdrawn and its volume is determined. It is characterized as shown in Table 12-1.¹⁴⁶ The gross characteristics are noted. These include viscosity and color, which is normally clear and yellowish tinged. The fluid is typically not bloody, although blood tingeing may suggest a traumatic tap or the existence of a pathologic lesion.

Fluid in normal joints is present in quantities that may preclude sampling by arthrocentesis. Recht detected fluid in 14 of 20 shoulders by MRI in 12 asymptomatic young volunteers; in none was more than 2 mL evident.⁶⁶⁸ If the joint fluid is obtained, the white count is less than 200 and predominantly monocytic as opposed to polymorphonuclear, and there are no red cells.²⁶¹

Additional joint fluid types have been noted.²⁶¹ Noninflammatory fluid has a white cell count usually less than 2,000.¹⁴⁶ The fluid is transparent or nearly so. This type of fluid can be seen in trauma, osteoarthritis, systemic lupus erythematosus (SLE), sarcoid, and hypothyroidism.

Inflammatory fluid may show varying degrees of clarity and color. The white cell count is more than 2,000, but usually not more than 50,000.¹⁴⁶ This is commonly seen in RA, gout, and possibly some infectious disorders. It is also associated with Reiter’s syndrome, ankylosing spondylitis, psoriatic arthritis, and juvenile RA.

Pyogenic fluid is opaque or grossly purulent. The white cell count exceeds 50,000, often higher, and is predominantly polymorphonuclear cells.¹⁴⁶ This is typical of infectious arthritis and, in some cases, gout or other very inflammatory arthritides.

Nonpyogenic fluid has variable characteristics. The white blood cell count is usually less than 20,000.146 Additionally, joint fluid can be characterized as hemorrhagic. This will occur in cases of trauma, hemophilia, and other bleeding disorders.

Microscopic identification of crystals is a laboratory test with notable problems in sensitivity, specificity, and interobserver differences.¹²³,⁶⁹³,⁷⁴⁷ Despite this observation, the fluid should routinely be examined for calcium pyrophosphate (CPP) and uric acid crystals that are further characterized by polarized light microscopy.⁶¹⁴ Gout crystals are pointed and negatively birefringent.⁶³⁰ CPP crystals are rhomboid and positively birefringent.²⁶²,³⁹⁵

Synovial fluid glucose levels are normally approximately 20 mg/dL less than the serum. Lower concentrations may be
observed in joint sepsis. Synovial fluid protein is often increased in inflammatory disease. The utility of synovial fluid glucose and protein levels is questionable.⁷⁸⁶ Appropriate organism stains and cultures should be obtained when there is even the most remote suspicion of an infectious process.

The occurrence of osteoarthritis at the glenohumeral joint is much less common than at the hip or knee.⁵⁶⁶ It is more likely to be found in women than men and in patients over 60 years of age.⁵⁶⁶ Its incidence increases with age, and it is the site with the oldest average age of onset.¹⁶³,³⁸⁷ The difficulty in determining the early diagnosis, the timing of its onset, and the absence of longitudinal data make estimates of the prevalence of osteoarthritis and its incidence imprecise.

DePalma had an early interest in the aging shoulder and carefully studied its morphologic aspects.¹⁷⁶,¹⁷⁸ On the glenoid side, from his study of cadavers ranging in age from 14 to 87, he determined that the degenerative process reached its maximum by the sixth decade. The most significant changes seemed to be located at the superior aspect of the glenoid. At the glenoid site of labral attachment, particularly in the anterior and anteroinferior regions, there was a marginal proliferation of bone and cartilage. This was believed to be secondary to traction forces applied to the soft tissues in these areas, with a resultant functional response to stress loading. Over time, a generalized thickening of the synovial membrane was observed. This occurred secondary to proliferation of the fibrous stromal elements in the synovial areas, resulting in an increased number of villous projections. On the humeral side, articular surface changes were never profound. The changes were certainly never equal to those seen on the glenoid side and were felt to be attributable to the mismatch of the humeral to glenoid surface area (4:1).

Cadaver studies by others have confirmed the relative absence of humeral cartilage thinning with increasing age.⁵³¹,⁶²⁶ Petersson and Raedlulnd-Johnell radiographically reviewed the glenohumeral joint spaces in normal persons and concluded that, with age alone, the joint space does not decrease.⁶²⁵ A positive correlation with age and joint space narrowing was shown by Kircher.⁴¹⁶

Age is the greatest risk factor for the development of primary (idiopathic) osteoarthritis in all joints.⁷³,⁶²⁹ Additional systemic factors (gender, hormonal status, nutrition, race, and ethnicity), vascular pathology, and intrinsic joint vulnerabilities (previous damage, bridging muscle weakness, misshaped joint, malalignment, and proprioceptive deficiencies) may increase one’s susceptibility to osteoarthritis, and under the influence of loading conditions (obesity, trauma, and physical activities), osteoarthritis may evolve or progress.⁹²,²³⁷,⁴⁰⁷,⁵⁰⁰,⁶²⁹ Some of these factors may influence to some degree by genetic variation.⁸¹⁹,⁸²⁰ and ⁸²¹ Bone density may play a role, as noted by the apparent inverse relation between osteoarthritis and osteoporosis.⁴²² Various crystals are present in osteoarthritic joints, yet whether they are incidental or whether they contribute to the initiation and acceleration of joint deterioration is uncertain.²⁵³,⁴⁷³,⁴⁹¹,⁵¹⁶,⁵⁹⁰,⁶⁴⁷,⁶⁴⁸,⁶⁹⁴ There are no special associations in glenohumeral arthritis, except a coexistence of degenerative changes, including tearing of the rotator cuff.¹¹⁸,⁴¹⁰ Feeney identified a strong correlation between tears of the rotator cuff and degenerative changes of the articular cartilage that was independent of the factor of age.²²⁹ However, the area of cartilage damage does not necessarily increase as the size of the tear increases.³⁵³ Moreover, the incidence of full-thickness rotator cuff tears in osteoarthritis is low. Superior labrum anterior posterior lesion-associated chondral lesions have been observed and may contribute to early-onset glenohumeral osteoarthritis.⁶¹⁶

Damage to the joint and reaction to it may result in an early, subclinical stage of arthritis that may exist for years or decades. The degeneration takes the form of thinning and softening of the articular cartilage. Surface fibrillation and, in the deeper layers, fissures and vertical clefts develop as the cartilage succumbs, exposing subchondral bone.⁶⁴⁹ The classification of Outerbridge, initially proposed for chondromalacia of the knee, has been modified for the staging of articular cartilage lesions of the glenohumeral joint (viewed arthroscopically).⁶⁰⁶,⁸⁵⁰ Stage I is softening or blistering of the cartilage. Stage II is fissuring and fibrillation. Stage III is deep ulceration, and stage IV is exposed subchondral bone. Its limitation is that stages II and III do not take into account accurately the depth of the lesion, although it is implied. There may be evidence of fibrocartilaginous repair.

The synovium thickens in response to the joint debris produced by the reactive changes in the articular cartilage and bone. Villous hypertrophy with random synovial cell hyperplasia may characterize the membrane. It may be filled with cartilage fragments and foreign body giant cells.⁹¹ Loose or sequestered osteocartilaginous bodies may be seen.⁹¹ The subsynovial region stroma may be filled with a mild chronic inflammatory reaction.²⁷⁵

The gross pathologic findings of glenohumeral osteoarthritis are fairly consistent and have been characterized by Neer.⁵⁷²,⁵⁷³,⁵⁷⁷ Thinning or absence of the cartilage of the humeral head is most pronounced in a position corresponding to 60 to 90 degrees of abduction, the area of maximum joint reaction force.³⁶¹ The normally convex humeral head flattens. The exposed bone becomes eburneous and sclerotic, acquiring a marble-like appearance, often stippled with small reddish-brown vascular proliferations and fibrocartilaginous plugs (Fig. 12-7). Patterns of surface wear may be influenced by the variable mineral distribution in the subchondral plate of the humeral head that is felt to reflect mechanical loading of the joint.⁸⁹⁴ Osteophytes appear circumferentially at the margin of the articular surface of the humeral head, resulting in its apparent enlargement (Fig. 12-6). The large inferomedial osteophytes, which have been termed “the goat’s beard” by Matsen, may envelope and tension the adjacent capsule, contributing to limitations of external rotation (Fig. 12-7).⁴¹⁶,⁵⁰⁸ Subchondral cysts are often present on the glenoid as well as on the humeral head.

The rotator cuff and the biceps tendon long head are intact in 87% to 95% of cases.¹³⁸,¹³⁹,⁵⁷²,⁵⁷⁸,⁸⁴¹ Neer believed that their integrity and capability to generate essential glenohumeral
compressive forces were a prerequisite for the development of osteoarthritis.⁵⁷⁸ He further believed that an enlarged osteoarthritic head helps prevent upward migration, resulting in fewer impingement-type rotator cuff tears.⁵⁷⁸ Rupture of the biceps tendon long head can rarely occur, but the mechanism is in response to the presence of spurs in the intertubercular groove and not by the process of impingement against the coracoacromial arch.⁵⁷⁸ The subacromial bursa has been noted to be thickened in the absence of acromial pathology in a high percentage of patients.⁸⁵⁰ During the arthroscopic evaluation and treatment of glenohumeral arthritis, the incidence of a concomitant lesion requiring treatment was 47%.¹⁰⁴

Although primary glenohumeral osteoarthritis usually begins on the glenoid, the glenoid cartilage is typically spared anteriorly (Fig. 12-8).⁶⁶⁴ The wear is more pronounced posteriorly, and a true crista (Fig. 12-9) may be formed as a demarcation between intact articular cartilage and exposed subchondral bone. The disease progression favors excessive glenoid bone erosion posteriorly with resultant posterior humeral head subluxation. This stretches and attenuates the posterior capsule while the anterior capsule significantly thickens and contracts. Peripheral osteophytes may “enlarge” the glenoid (Figs. 12-8B and 12-10). Osteocartilaginous bodies, either loose or attached, seek the recesses of the synovial cavity, especially the subscapularis bursa (Fig. 12-11). They may enlarge and become intimately attached to the adjacent bony structures, significantly distorting the normal anatomy. The soft tissue envelope tightens further as it drapes over the osteocartilaginous bodies and osteophytes. A validated morphologic study of the glenoid in osteoarthritis has enabled the classification of a well-centered humeral head, a posteriorly subluxed humeral head, and a retroverted, primarily dysplastic glenoid.²¹³,⁵⁸⁹,⁸⁴⁰,⁸⁴¹ In addition, four types of eccentric inferior glenoid wear patterns have been observed frequently and described.²⁹⁷

The same findings are observed in secondary osteoarthritis. In addition, in cases of ochronosis, the characteristic blackening of the articular cartilage caused by homogentisic acid deposition is noted.

Patients may present for evaluation and treatment as early as age 40 to 50 years. Their lives are clearly altered by the disease process manifesting primary complaints of shoulder pain and loss of range of motion.²⁶⁰,⁵⁰⁷ The source of pain may be local, peripheral, or central in origin.¹⁸⁴,⁵²⁵ It does not appear to be related to radiological parameters.⁴¹⁶ Joints effusions, bone marrow lesions (edema), synovial hypertrophy, tendonitis, and bursitis are nocioreceptive structures shown to be involved in osteoarthritis.²³² More than 50% of patients are unable to sleep on the affected side and cannot perform common functional tasks.⁵¹³ Unremitting progression of the disease process will lead to more intense symptoms, failure of conservative management measures, and consideration for surgical treatment. The gender distribution is equal for this active and generally healthy subset of patients.⁵⁷⁸ The dominant arm is involved to a greater extent than the nondominant arm.

Generalized atrophy of the shoulder is often notable, especially when the disorder is unilateral. Early in the disease, motion changes are minimal and difficult to detect, unless the most sensitive testing position—supine—is used. As the glenohumeral changes become more advanced, motion of the soft tissues is restricted, with a significant diminution of external rotation (Fig. 12-12). This is a more sensitive test of intraarticular activity than is the loss of elevation. Attempts at active and passive motion are painful and may produce catching and squeaking sounds from the rough glenohumeral articulation. Shoulder proprioception is significantly altered with respect to an uninvolved contralateral shoulder.¹⁵⁹ A specific point of localized tenderness is over the posterior joint line, more easily elicited as the humeral head becomes subluxed with advancing disease. Synovial thickening and a large joint effusion may be palpable in thin individuals.

Plain films are most helpful for making the diagnosis of osteoarthritis, but probably underestimate the extent of pathologic osteoarticular changes.³⁹³,⁶⁶⁴ The AP view of the humeral articulation to the scapular plane and a high-quality axillary lateral view are all that is needed. They enable one to ascertain joint orientation, the amount of erosion of the glenoid, humeral head position, and the extent of disease activity. It appears that joint space narrowing and the development of osteophytes are independent radiological parameters.⁴¹⁶ Weighted views may prove useful to help identify reduction of the glenohumeral cartilage space by as little as 1 mm.⁸¹⁶ Because of soft
tissue contractures or pain, optimum films are sometimes not possible to obtain. In those instances, a CT scan will prove extremely useful in assessing glenoid morphology, glenohumeral relations, the presence of osteocartilaginous bodies, and the estimation of the anatomic version (Fig. 12-13).³⁹ The accuracy of version determination, wear orientation, and estimation of extent of glenoid bone loss is enhanced by proper orientation of the scapula, three-dimensional reconstruction, and a sophisticated method of computer-generated modeling of the glenoid vault.⁸⁷,²⁵⁶,³⁴²,⁷²³,⁷²⁴ and ⁷²⁵ Other findings of osteoarthritis are densification of subchondral bone, subchondral cysts, and peripheral osteophytes along the glenoid margins and adjacent to the articular surface of the humeral head, especially inferiorly. The humeral head may be flattened and enlarged. A triple-phase bone scan may prove beneficial for confirmation of disease localization when the plain films show no or early changes of glenohumeral arthritis (Fig. 12-14). The role of ultrasonography is uncertain although its utility for the detection of synovitis is recognized.³²²,⁴⁰⁰ The routine use of MRI for disease detectable by plain films would be considered exotic. However, it may have a role as a screening tool in pre-radiographic or early disease, to detect synovitis, to quantify articular cartilage volume, thickness and/or to monitor response to nonsurgical therapeutic interventions.⁴⁴,⁹⁷,³²²,⁴⁹⁶

Laboratory studies are rarely helpful in the evaluation of primary osteoarthritis. However, investigators are in pursuit of promising identifiable molecular markers of disease in body fluids, such as COMP, antigenic keratin sulphate, hyaluronic acid, YKL-40, type III collagen N-propeptide, catabolic and anabolic epitopes, and urinary glucosyl-galactosyl pyridinoline.¹²⁰,¹⁷²,²⁷⁸,⁶⁵³,⁷⁰⁰,⁸⁶⁰ Blood studies and synovial fluid analysis may help identify an underlying cause of secondary osteoarthritis. Synovial fluid levels of biochemical markers (aggrecan-aggregates) of catabolic activity within the articular cartilage enable the detection of all stages of glenohumeral osteoarthritis, with a high degree of accuracy.⁶⁶⁴ The identification of the disorder before radiographic changes are present may prove extremely useful to expand the treatment options and to monitor the disease progression and response to treatment.⁶⁶⁴

Arthroscopy of the glenohumeral joint can improve the diagnostic accuracy when the clinical diagnosis is suspected and the radiographs do not show advanced stages.²¹⁷,²¹⁹,⁶⁶⁴

The incidence of this disorder has not been accurately established. Hovelius performed a 10-year prospective study of patients younger than the age of 40 to assess the outcome of a primary anterior dislocation of the shoulder treated with closed reduction, both with and without immobilization.³⁴⁸ The incidence of glenohumeral arthrosis was 20%. The degree of involvement was mild in 11% and severe in 9%. Cameron determined the overall prevalence of osteoarthritis after either acute or chronic glenohumeral instability to be small, if not rare.¹⁰⁵ It was not influenced by the direction of the instability. They noted that the risk of osteoarthritis increased with the time from injury. In other series of patients treated surgically for anterior glenohumeral instability, the incidence of preoperative arthritis ranged from 0% to 20%.¹¹,⁹⁹,⁶⁵⁵,⁸²⁵

In the large series (570 patients) reviewed by Buscayret as well as in earlier series, the factors of age at the onset of instability, osseous glenoid rim lesions, humeral head impaction fractures, and interval between the onset of instability
and surgical treatment correlated with the preoperative development of arthritis.¹¹,⁹⁹,³⁴⁹,⁷¹⁵,⁷⁵⁹ Marx noted that the glenohumeral dislocation requiring reduction was found to be associated with a 10- to 20-fold increase in the risk for the subsequent development of severe arthrosis sufficiently severe to warrant shoulder arthroplasty.⁵⁰⁵

By far, the most common cause of arthritis of dislocation is iatrogenic. It has been termed “capsulorrhaphy arthropathy” by Matsen.⁵⁰⁹ With rare exception, its existence is acknowledged by most experts in the field of shoulder surgery, although its incidence cannot be accurately determined.⁷⁸,⁵¹⁴,⁶¹³ In series with greater than 10 years follow-up, the incidence of significant arthritis ranges as high as 72%.⁵⁶,¹²⁴,²²⁵,³⁵¹,³⁵²,⁴²⁸,⁵⁸⁵,⁶²⁰,⁶⁹⁰,⁷⁰¹,⁷³⁴,⁷⁵⁹,⁸²⁵,⁸⁵⁷,⁸⁸⁷ Similarly, after 7 to 11 years following arthroscopic Bankart suture-anchor repair, the incidence of glenohumeral arthrosis was up to 39%.¹¹⁴,³⁹⁶ Interestingly, in patients with no preoperative arthritis, at a mean 6.5 years after surgical treatment, postoperative arthritis occurred in
nearly 20% of patients.⁹⁹ Once again, age at the time of the initial instability episode seemed to be most important for the subsequent development of arthritis. Other important factors were age at the time of surgery and a longer interval of follow-up. In contrast to Rachbauer, the presence of arthritis was influenced by the number of instability episodes prior to surgery.⁶⁵⁵ The findings regarding decreased external rotation were inconclusive. Considering these observations, Buscayret suggested that surgery does not influence the risk factors nor prevent the development of arthritis in shoulders that undergo attempts at surgical stabilization.⁹⁹ Ogawa et al. stated more emphatically, “…unquestionably apparent that no operative procedure prevents development of OA.”⁵⁹⁹ Additionally, the arthritis they recognized postoperatively had developed before surgical stabilization.

Hovelius’ data suggested no relation between dislocation arthropathy and the number of recurrent dislocations or the treatment rendered either operative or nonoperative.³⁴⁸ The lack of association with the number of dislocations was confirmed by others.⁵³,³⁰⁶ Hovelius’ study pointed out further that advanced glenohumeral arthritis may occur even after a single anterior glenohumeral dislocation treated without surgery.³⁴⁸ Traumatic shoulder dislocations are the result of a significant force applied to the glenohumeral joint and generally involve damage to soft tissue, including the capsule, labrum, and rotator cuff, as well as bone and cartilage.⁷⁹⁴

Morrey and Janes prudently observed that looseness of the joint played a role in the development of surgical failures after the performance of a unidirectional repair.⁵⁵⁴ Factors shared by these patients included a positive family history, occurrence of bilateral glenohumeral instability, and the coexistence of posterior instability in the shoulder that received operative treatment. Neer introduced the term “multidirectional instability,” perhaps for the same subset of patients.⁵⁷¹ He was also the first to draw the association between the disorder and arthritis of dislocation.⁵⁷² It was his belief that multidirectional instability existed on the basis of lax ligaments, repeated minor injuries, or a combination of both. The most frequent cause of multidirectional instability was felt to be an acquired laxity, explained on a heritable basis. The initial instability episode in this group of patients most often occurred without high forces, major injury, or significant associated injuries. Typically, if a dislocation was present, it reduced spontaneously or required minimal, usually self-manipulative, effort. Generalized ligamentous laxity was often recognized. The examination of the contralateral shoulder revealed excessive anterior and posterior translations, as well as inferior translation, as evidenced by the “sulcus sign.” The presence of this so-called sulcus sign was pathognomonic for multidirectional laxity. Neer concluded that “by far the most frequent etiologic mechanism for the development of arthritis of recurrent dislocation is a ‘standard’ operative procedure intended to remedy recurrent unidirectional dislocations that is unsuspectingly performed on a loose, multidirectional shoulder.”⁵⁷⁹ He observed that the “procedures displace(d) the humeral head in a loose shoulder away from the side of the repair, creating a fixed subluxation. The subluxed head wears unevenly on the glenoid, and arthritic changes can develop surprisingly fast.”⁵⁷⁹

It was actually Hindmarsh, however, who first implicated the role of stabilization surgery in the production of moderate to severe glenohumeral arthritis.³³⁵ A variety of surgical procedures have been performed in an attempt to stabilize the glenohumeral joint and to prevent recurrences that led to glenohumeral arthritis.²⁴³,³¹⁸,³¹⁹,³⁵⁷,⁴⁸⁶,⁴⁸⁹,⁵⁷²,⁵⁸⁸,⁵⁹⁶,⁷⁷⁹,⁸⁵⁷ When a unidirectional repair, such as the Putti-Platt,³¹⁹,³⁵⁰,⁴¹⁹,⁴²⁸,⁴²⁹,⁶⁵⁵,⁸²⁵,⁸⁸⁶ Eden-Hybbinette-Lange,⁸³,⁹⁹,³³⁵,⁴²⁹,⁵⁸⁵,⁶⁵⁵,⁶⁵⁹,⁸⁵⁷ Max Lange,⁸⁷⁵ Trillat,²⁶⁵ Magnuson-Stack,⁷¹⁵ Bristow-Latarjet,¹¹,¹¹²,¹⁹¹,³⁴⁷,³⁵¹,³⁵²,⁴⁷⁷,⁵⁸⁸,⁷³⁴,⁷⁵⁹,⁷⁷⁴ DuToit,²⁰²,⁵⁹⁶,⁷¹⁵,⁷⁶¹,⁸⁸⁷,⁸⁹³ Bankart⁵⁶,³⁵⁰,³⁵¹,⁵⁸⁸,⁶⁹⁰,⁷⁰¹,⁷¹⁵ or rotation osteotomy243, was performed to eliminate instability recurrences, the stage was set for further glenohumeral joint deterioration. The incidence of osteoarthritis was notably less with anatomic repairs (Bankart) when compared with nonanatomic repairs.⁵⁶,⁶⁹⁰,⁷⁰¹

The pathology encountered is often dependent on the index stabilization operation. In general, excessive scarring and adhesions are the predominant finding.⁵⁹ Operations involving greater manipulation of the soft tissues, such as the Bristow procedure, often result in encasement of all anterior soft tissue structures, sometimes including the axillary nerve.⁵⁹ The subcoracoid, subacromial, and subdeltoid planes of motion are usually scarred. The subscapularis musculotendinous unit will be contracted, as is the anterior joint capsule. Occasionally, the subscapularis integrity will be preserved, such as in a Magnuson-Stack procedure.⁴⁹² Transferred bony blocks may overhang the lateral margin of the anterior inferior glenoid.¹¹

The articular surface contours may appear normal with simply posterior subluxation and instability of the humeral head. More often, the glenoid articular surface involvement is quite significant (Fig. 12-22). The anterior glenoid cartilage may be reasonably preserved, whereas the more posterior cartilage demonstrates thinning and erosion to subchondral bone. In advanced cases, posterior glenoid bone erosion and loss is observed. The humeral head shows concomitant wear changes. The penetration of metallic fixation devices into the intraarticular space may be observed in open as well as arthroscopic procedures, with associated articular surface changes, including arthropathy (Fig. 12-23).³⁸,²¹⁷,³⁸⁸,⁵⁹⁶,⁶⁷⁶,⁷⁵⁶,⁸⁸⁵,⁸⁹³ Severe arthrosis is not usually present. It is these and other pathoanatomical features that may complicate all aspects of subsequent definitive treatment with prosthetic arthroplasty.⁴⁶⁰,⁷¹³ When operative care has not been rendered, the findings are essentially those of glenohumeral osteoarthritis. Preferential glenoid wear or humeral head subluxation may be observed in the direction of the instability. Variable soft tissue changes may exist.

The patient, not infrequently younger than 40 years of age, usually presents to the surgeon for evaluation many years after the glenohumeral stabilization operation or the index dislocation.⁵⁹,⁷⁶,³¹⁹,⁵⁷² Within 2 to 3 years of presentation, shoulder symptoms have begun to evolve. At the time of the evaluation, the chief complaints are related to intense, often disabling pain. Almost all patients have had long-standing limitation of range of motion, especially external rotation.⁷⁷¹ Many develop an internal rotation contracture. Moderate to severe functional disability is present secondary to pain and altered range
of motion.⁶¹² Some have had multiple surgical procedures. Despite mild to moderate radiographic changes of arthrosis, the clinical impact, in some instances, is negligible.¹¹³