Anterior shoulder instability is the most common type of shoulder instability. It is typically the result of a traumatic event with dislocation of the glenohumeral joint. However, instability also may present as subluxation, a condition in which the joint symptomatically translates but does not completely dislocate.
Anterior shoulder instability frequently affects patients between the ages of 15 and 40 years, but patients of any age may sustain a shoulder dislocation. It is also more common in athletic persons. Within the general population, the prevalence of glenohumeral dislocation is 1.7%, but in ice hockey players, the incidence rises to 8%. The incidence is estimated to be between 0.08 and 0.24 per 1000 person-years in the civilian population and between 1.69 and 4.35 per 1000 person-years in the military population.
This chapter discusses the anatomy, clinical features, and treatment of anterior shoulder instability. Posterior and multidirectional instability may involve a component of anterior patholaxity; these topics are addressed separately in subsequent chapters. This chapter first reviews the anatomy and pathoanatomy of shoulder instability. The History and Physical Examination sections provide an evidence-based approach and assume a general knowledge of shoulder examination. The Decision-Making Principles and Treatment Options sections discuss acute shoulder reduction and immobilization, timing of surgery, arthroscopic versus open surgery, unique clinical scenarios, and the approach to bone loss. Our preferred technique for treatment of shoulder instability is arthroscopic Bankart repair. For revision situations, we prefer the Latarjet procedure. Methods for both of these scenarios are outlined, along with postoperative management. Lastly, the results of treatment and complications are discussed, along with future directions of research.
The glenohumeral joint is a unique diarthrodial joint. More than any other joint in the body, it must carefully balance function and stability. Because of its role in positioning the arm in space, it has 6 degrees of freedom and is the most mobile of the diarthrodial joints. However, to achieve this range of motion, stability is sacrificed.
When discussing the stabilizing anatomy of the shoulder, it has become common to dichotomize the stabilizers as static or dynamic. We recognize that this approach is an oversimplification and that the entire system functions in a coordinated fashion, but we believe it serves as a useful pedagogic instrument and thus we will retain it for the purposes of this discussion. Static stabilizers of the shoulder may be considered to be the structures that provide a unidirectional limitation to translation. The three principle groups of static stabilizers are bony, ligamentous, and labral. Dynamic stabilizers are primarily musculotendinous and include the rotator cuff, biceps, deltoid, pectoralis major, and latissimus dorsi.
The bony anatomy of the glenohumeral joint has been compared to a golf ball on a tee. The humeral head is significantly larger than the overall size of the glenoid, with only 25% to 30% of the humeral head in contact with the glenoid at any given anatomic position. The bony glenoid concavity is quite shallow, with a depth of only a few millimeters. To confer some degree of stability, the glenoid concavity is deepened by the articular cartilage and the labrum. The articular cartilage is thinner in the center of the glenoid and progressively thickens toward the periphery, thus increasing the functional depth of the glenoid ( Fig. 46-1 ). Loss of the articular cartilage has been found to decrease stability by nearly 50%.
The glenoid labrum is a critical structure for the stability of the glenohumeral joint. It encircles the glenoid to both increase the depth of the glenoid concavity and the overall surface area of the glenoid in contact with the humeral head by approximately 50%. Although it is commonly referred to as a fibrocartilaginous structure, anatomic studies have demonstrated that the glenoid labrum has a structure similar to a tendon, with dense fibrous connective tissue that is devoid of chondrocytes. The labrum has a predominantly triangular cross-sectional shape and, consequently, functions as a “chock block” that prevents translation of the humeral head outside the articular surface of the glenoid. However, the stabilizing mechanism of the labrum is much more complex than this simple mechanical function.
The labrum has a complex anatomic structure and attachment to the glenoid. It is typically more adherent in the inferior half of the glenoid and quite variably attached in the anterior superior quadrant. The labrum serves as an attachment site for the glenohumeral joint capsule and the long head of the biceps tendon, which helps create a suction seal that leads to a negative intraarticular pressure and aids in the creation of the concavity-compression effect of glenohumeral stability.
The glenohumeral capsule that attaches to the glenoid labrum has four distinct thickenings that have been termed the glenohumeral ligaments (GHLs): the superior, middle, and anterior band of the inferior and the posterior band of the inferior ( Fig. 46-2 ). Of these ligaments, the inferior glenohumeral ligament (IGHL) is the most important in providing stability to the shoulder. The GHLs are not as strong as the knee ligaments; for example, the IGHL possesses an ultimate stress capability that is only 15% of that reported for the anterior cruciate ligament. This characteristic underscores the importance of the entire relationship between the static and dynamic components that provide stability to the glenohumeral joint.
The IGHL should be considered to be a complex rather than an individual structure. Within the IGHL structure, an anterior band originates at the 3-o’clock position of the glenoid, a posterior band originates at the 8-o’clock position of the glenoid, and an intervening capsule exists that is termed the “axillary pouch.” This complex is likened to a hammock. When the arm is brought into the “apprehension position” of abduction and external rotation, the axillary pouch becomes more taut as the anterior band is placed under tension from being pulled superiorly and anteriorly to span the midportion of the glenohumeral joint ( Fig. 46-3 ). In this position, the anterior band of the IGHL functions to prevent anterior translation of the humeral head while the taut axillary pouch prevents anteroinferior translation.
If the arm is brought into 45 degrees of abduction, the middle GHL becomes taut and prevents anterior displacement. In 0 degrees of abduction, the superior GHL has a minor function in preventing anterior displacement and an accessory role in preventing inferior displacement along with the supraspinatus, deltoid, and coracohumeral ligament.
The dynamic stabilizers of the shoulder provide stability by providing compression of the humeral head into the glenoid and coracoacromial arch concavities, which is termed the “concavity-compression effect.” The deltoid and the rotator cuff function to dynamically compress and center the humeral head within these concavities during shoulder motion, thereby enhancing (or, perhaps, providing primary restraint to) glenohumeral stability.
In a review of anatomy relating to anterior shoulder instability, it is important to note two relatively common anatomic variants that should not be mistaken for pathology. The anterosuperior aspect of the capsulolabral complex is the most variable in the shoulder. A sublabral foramen (or sublabral hole) is the complete separation or absence of the labrum from the glenoid. It is the most common variant, with an incidence between 3% and 18%. A Buford complex is the combination of an absent anterosuperior labrum with an associated “cordlike” middle glenohumeral ligament that attaches to the superior labrum near the base of the biceps tendon. The incidence varies between 1.5% and 6%. Other less common variants exist as well. Overall, recognition of variant anatomy both on magnetic resonance imaging (MRI) and at arthroscopy is vital to avoid inappropriate diagnosis or treatment.
When A.S. Blundell Bankart described the pathology and treatment of recurrent dislocation of the shoulder, he emphasized the importance of understanding the pathoanatomy to guide treatment. To him, the “essential lesion” occurred when the “[humeral] head shears off the fibrous or fibrocartilaginous glenoid ligament from its attachment to the bone.” His original description was based on his work with only four patients, and his subsequent description 15 years later was based on work with a further 23 patients. Bankart believed that the essential lesion occurred in 100% of cases and that “no one who has ever seen this typical lesion exposed at operation could possibly doubt that the only rational treatment is to reattach the glenoid ligament (or the capsule) to the bone from which it has been torn.”
Since Bankart provided this description, a more nuanced understanding of the pathoanatomy of recurrent anterior shoulder instability has been developed. Fundamentally, the pathoanatomy can be capsulolabral, osseous, or both. The capsulolabral lesions are typically related to the glenoid labrum and the anterior band of the IGHL. The nomenclature of these lesions has evolved into a mixture of eponymous terms and acronyms that can confuse even the most experienced clinician. The nomenclature includes the Perthes lesion, Bankart lesion (or “anterior labral tear”), bony Bankart lesion, anterior labroligamentous periosteal sleeve avulsion (ALPSA) lesion, and humeral avulsion of the glenohumeral ligaments (HAGL) lesion ( Table 46-1 ).
|Associated with Anterior Instability|
|Perthes||Avulsion of the anterior-inferior glenolabral complex with preservation of the medial scapular neck periosteum|
|Bankart||Complete avulsion of the anterior-inferior glenolabral complex along with a piece of scapular neck periosteum|
|Bony Bankart||Osseous avulsion fracture of the anterior-inferior glenolabral complex|
|ALSPA||Avulsion of the anterior-inferior glenolabral complex with stripping of the medial scapular neck periosteum but preservation of a medial hinge; the loose fragment subsequently scars medially down the scapular neck|
|HAGL||Avulsion of the glenohumeral ligaments from their humeral-sided attachment|
|Not Associated with Instability|
|Glenolabral articular disruption||A superficial tear of the anterior-inferior labrum with associated cartilage injury but preservation of the anterior-inferior glenolabral complex; presents with a painful shoulder but is not a cause of shoulder instability|
|SLAP||Disruption of the superior labrum, originally described to stop at the midglenoid notch; recent descriptions have associated SLAP tears with Bankart lesions, but SLAP lesions alone are not a cause of shoulder instability|
Despite this confusing jargon, the Bankart lesion remains an “essential lesion” and has been found to occur in approximately 90% of patients with recurrent anterior instability. A Bankart lesion can be arbitrarily defined as an avulsion of the anterior-inferior capsulolabral complex with extension into the scapular periosteum and rupture of the periosteal tissue ( Fig. 46-4, A ). Acutely, the avulsed tissue is free to move about the shoulder. As such, the stabilizing function of the labrum is lost and the anterior band of the IGHL can no longer resist anterior translation in abduction and external rotation. A bony Bankart lesion occurs when the capsulolabral complex is avulsed along with a variably sized fragment of bone ( Fig. 46-4, B ), which can create a significant osseous injury that can contribute to the pathogenesis of shoulder instability in a manner more severe than that of a soft-tissue Bankart lesion (discussed later).
A Perthes lesion and an ALPSA lesion are variants of the Bankart lesion ( Fig. 46-5 ). A Perthes lesion can be thought of as representing the initial stages of a Bankart lesion. The capsulolabral complex is avulsed from the anterior-inferior aspect of the glenoid, but the medial scapular periosteum remains intact. In essence, it is a nondisplaced Bankart lesion. An ALPSA lesion occurs when the capsulolabral complex is avulsed and the medial scapular periosteum is stripped (but not detached as in a Bankart lesion) and subsequently displaced down the denuded anterior glenoid neck. In essence, it can be conceptualized as a medialized Bankart lesion. In either lesion, the capsulolabral function is lost and recurrent instability ensues.
Although the most common site of injury relating to anterior instability is at the glenoid, rupture of the glenohumeral ligaments can also occur on the humeral side. This rupture has been termed a HAGL lesion and has an incidence between 1% and 9%. Similar to that which occurs on the glenoid, the rupture of the humeral insertion along with a bony avulsion is coined a bony HAGL or B-HAGL (pronounced “bagel” lesion). Although loss of the GHL tension mechanism results in either case, it is important to recognize this lesion and thus perform the correct surgery.
In addition to capsulolabral injury, some degree of bony injury occurs in virtually every patient with anterior shoulder instability. Osseous glenoid lesions are thought to occur as the humeral head either passes over the glenoid rim or as the posterior superior aspect of the humeral head has an impact on the anterior glenoid rim upon dislocation. Conversely, osseous injury to the posterior superior humeral head occurs via an impression-impaction fracture of the soft humeral head bone against the less giving, sclerotic anterior rim of the glenoid when subluxation or dislocation occurs. This lesion was first described by two radiologists, Hill and Sachs, in 1940 and has since been termed a Hill-Sachs lesion.
Glenoid bone injury is common and predominately occurs in two configurations. A visible bone fragment with its attached capsulolabral structures may fracture from the anterior glenoid rim and is termed a bony Bankart lesion. Alternatively, the anterior glenoid rim may be impacted or eroded from the force of the humeral head during subluxation or dislocation. In a three-dimensional (3D) computed tomography (CT) scan of 100 consecutive shoulders with recurrent anterior instability, only 10% were found to have normal glenoid morphology ; 50% of patients had some degree of bony fragment, and 40% presented with erosion or a compression fracture of the glenoid. A subsequent CT study found similar results, with 40% of persons with first-time dislocations and 85% of persons with recurrent dislocations sustaining a degree of glenoid bone loss. These noninvasive studies corroborate Rowe’s observations that 73% of his patients undergoing open Bankart surgery had glenoid rim damage.
Hill-Sachs lesions are common, but until recently, they were less regarded in the pathogenesis of recurrent instability. These lesions occur in approximately 40% of patients with recurrent subluxation but no dislocation, in 90% of patients with a single dislocation, and in virtually 100% of patients with recurrent dislocations. The majority of these lesions are small, and in general, they are clinically insignificant. However, a minority have been termed “engaging,” which means the Hill-Sachs lesion is oriented in such a manner that placing the shoulder in abduction and external rotation results in the humeral head losing contact with the glenoid and subsequent subluxation or dislocation of the glenohumeral joint.
If we return to the analogy of the glenohumeral joint as a golf ball on a tee, it becomes apparent that if either the tee (glenoid) or golf ball (humeral head) is damaged, the stability is altered or even lost. But what degree of injury will result in instability? It is easiest to first consider the glenoid.
In a classic biomechanical study by Itoi and colleagues, a series of glenoid osteotomies were performed to determine the effect of glenoid bone loss on the glenoid concavity mechanism of shoulder stability. Using the stability ratio, these investigators found a significant loss of stability after creation of a 6-mm defect, which is roughly equivalent to 28% of the total glenoid width. Practically speaking, however, it is difficult to measure the total glenoid width because this bone may be lost due to impaction or erosion. The glenoid length, defined as the superior to inferior distance of the glenoid (i.e., from 12 o’clock to 6 o’clock), can be used as a surrogate if some geometric assumptions and transformations are used. Accordingly, this critical defect size corresponded to 20% of the glenoid length. When the study was repeated with the soft tissue structures retained, the authors confirmed their original findings but also demonstrated that repair of the Bankart lesion failed to confer added stability. Stability was restored only after a coracoid bone graft was placed.
Clinically, these results have been supported by numerous studies that have found significant bone loss to be a risk factor for failure after a Bankart repair. Burkhart and De Beer described their clinical experience with bony glenoid loss. If the glenoid is viewed en face, it has an appearance similar to that of a pear. When anterior glenoid bone loss occurs, Burkhart and De Beer indicated that the glenoid has an “inverted pear” appearance and hypothesized that this arthroscopically viewable appearance is an indicator of clinically significant bone loss. When patients were dichotomized into a normal glenoid or inverted pear glenoid group, the recurrent dislocation rate after arthroscopic Bankart repair was 4% for the normal glenoid group but 61% for the inverted pear glenoid group. Lo, Parten, and Burkhart then reexamined these data and compared them with a cadaveric model, demonstrating that approximately 6.5 mm of glenoid bone loss (or 29% of the glenoid width) was necessary to achieve an inverted pear glenoid configuration.
The significance of humeral bone loss via Hill-Sachs lesions has been more difficult to characterize. Numerous authors have proposed classification schemes in an attempt to guide clinical decision making. They have variably defined the size according to length and depth, percentage of humeral head involvement at arthroscopy, percentage of humeral head involvement on a Notch View radiograph, and circular degree involvement and location on axillary MRI. Regardless of classification, it has been recognized that persons with large Hill-Sachs lesions are at risk for recurrent dislocation. Small lesions have conversely been viewed as clinically insignificant. Intermediate lesions have represented something of a management quandary. Of the classification schemes that have emerged in the literature, none is particularly helpful or valid, likely because of the multiple variables associated with a Hill-Sachs lesion.
Recently, interest in how both glenoid and humeral osseous defects can contribute to the pathogenesis of instability has increased and has resulted in the concept of the “glenoid track.” The glenoid track is defined as the zone of contact between the glenoid and the humeral head when the humerus is in maximal external rotation and horizontal extension and the arm is elevated and depressed ( Fig. 46-6, A ). If a Hill-Sachs lesion is entirely contained by the glenoid track, regardless of its length, depth, or percentage of the humeral head, then it never has the ability to engage the glenoid and dislocate ( Fig. 46-6, B ). Conversely, a small and shallow Hill-Sachs lesion can become symptomatic if its location is medial to the glenoid track ( Fig. 46-6, C ). It is also vital to understand that, by definition, any glenoid-side bone loss will decrease the width of the glenoid track ( Fig. 46-6, D ).
Although the glenoid track is a recently proposed concept and has yet to be conclusively demonstrated in a clinical setting, it provides an excellent framework for considering the importance of both glenoid and humeral bone loss in the pathogenesis of shoulder instability. Given the high incidence of osseous injury in persons with either acute or recurrent anterior instability, it is vital for the clinical evaluation to account for potential osseous causes of instability to avoid risk of treatment failure.
The purpose of the history is both to establish the diagnosis of shoulder instability and to obtain information that will guide treatment. Often the diagnosis of shoulder instability is not in question after a traumatic dislocation and a physician-assisted reduction. Conversely, some persons present with more subtle forms of instability and require differentiation from patients with posterior or multidirectional instability. Information to guide treatment must include data that will allow the surgeon to identify the risk of recurrence, the risk of failure of arthroscopic surgery, and the risk of associated pathology.
When first approaching a patient who reports shoulder instability, it is useful to classify the patient as either having sustained or not having sustained a dislocation. If the patient has experienced a dislocation, the details should be ascertained, including the mechanism of injury, the use of radiographs, the need for reduction (as well as who performed the required reduction), and the length of disability resulting from the dislocation. What arm position aggravates the sense of instability? If the patient has experienced multiple dislocations, it is useful to proceed in chronologic order to evaluate how the mechanism of injury may have evolved. Are the dislocations becoming more frequent? Are they occurring with less traumatic force? Lastly, it is vital to document the age of the patient when the first dislocation occurred and his or her age at the time of subsequent dislocations, because age is a significant prognostic indicator for recurrence.
Symptoms of shoulder subluxation without dislocation are vague, difficult to evaluate, and difficult to distinguish from the symptoms of patients with multidirectional instability. Typically, the symptoms have an insidious onset and include a sense of looseness, shoulder achiness, and possibly transient neurologic deficits. The activities causing symptoms are often random and, in contrast to the cases of patients with traumatic instability, may occur with activities of daily living.
To distinguish patients with anterior instability from patients with posterior and multidirectional instability, identifying the arm position that aggravates symptoms is occasionally useful. Classically, anterior instability presents with discomfort when the shoulder is abducted and externally rotated (such as when throwing a ball). Posterior instability presents with discomfort when the shoulder is internally rotated, adducted, and forward elevated, such as when pushing a door open. Multidirectional instability presents with symptoms in a variety of positions, but by definition, symptomatic inferior translation is part of the spectrum of complaints.
Once the history of the present illness is concluded, a comprehensive social history is obtained. In which sport (or sports) does the patient participate? Are these sports competitive or recreational? Is the activity a contact sport? Which way does the patient typically shoot, throw, or swing? What is the patient’s occupation? Does the patient’s occupation require forced overhead activity? This vital information will guide treatment.
A small number of patients present with the ability to voluntarily dislocate their shoulder. A classic study by Rowe and colleagues found that a substantial proportion of patients who could voluntarily dislocate their shoulder had an associated psychiatric condition and did very poorly after surgical stabilization. However, these investigators acknowledged that the majority of persons who could voluntarily dislocate their shoulder had no cognitive disturbance. Nevertheless, the association between persons who can voluntarily dislocate their shoulder and psychopathology has persisted. The astute clinician should screen for psychopathology but recall that the majority of these patients warrant more than cognitive and physical therapy. It is important to differentiate “voluntary” from positional instability. “Voluntary” suggests a muscular contraction and a volitional component, whereas patients with positional instability (commonly posterior) can reproduce the instability by positioning their hand in space. It is very common for patients with posterior instability to demonstrate their instability in forward elevation.
From the history alone, it is possible to estimate the risk for the presence and size of a glenoid bony defect. Milano and colleagues demonstrated that the presence of a defect was significantly associated with multiple dislocations, male gender, and type of sport. The size of the defect was significantly associated with recurrent dislocation, an increasing number of dislocations, timing from first dislocation, and manual labor. Presence of a critical defect, defined as 20% of the glenoid width, was significantly associated with the number of dislocations and age at first dislocation.
As with the history, the physical examination must be goal directed. Goals of the examination are to:
Either narrow the differential diagnosis or confirm the suspected diagnosis after the history is taken
Rule out possible associated pathology
Obtain information that will influence management
The standard systematic orthopaedic approach of “look, feel, move,” which was popularized by A. Graham Apley, should be used on a routine basis. First, a general impression of the patient is obtained. For example, unlike a 15-year-old underweight swimmer, a 22-year-old muscular football player is likely to have anterior instability. If the patient is older than 40 years, a rotator cuff tear should be considered, whereas if the patient is older than 60 years, a rotator cuff tear, axillary nerve injury, or brachial plexus palsy should be considered.
After inspection, palpation, and range of motion testing, the examination proceeds to strength testing (see Chapter 43 ). All components of the rotator cuff musculature should be evaluated with great care. Massive subscapularis ruptures may manifest as anterior instability, whereas associated supraspinatus tears are not uncommon in older patients after shoulder dislocations.
Once a generalized appreciation of both shoulders has been obtained, special tests are performed to confirm the diagnosis of anterior shoulder instability. It is important to note that instability is a subjective complaint of the patient and not a physical examination finding. By definition, instability is “symptomatic translation,” not merely the presence of larger magnitudes of translation upon physical examination.
A multitude of tests have been described for instability, but within the literature, the most accurate test is the apprehension-relocation test. To perform this test, the patient is placed supine at the edge of the examination table and the affected shoulder is brought into an abducted/externally rotated position. This arm position is considered to be the apprehension position because it maximizes the tension on the IGHL. If the patient has anterior instability with an incompetent IGHL, the restraints to anterior translation are lost and the patient feels a sense of apprehension that the shoulder will dislocate. The patient may alternatively experience pain, but pain is considered to be a less accurate criterion for a positive apprehension sign. The relocation test, or Fowler’s sign, is a continuation of the apprehension test that entails placing a posterior force on the arm to relieve the symptoms of apprehension ( Fig. 46-7 ). Without warning, the examiner may remove his or her hand, thus reproducing the symptoms of the apprehension test. This maneuver is termed the surprise test or anterior release test. The sensitivity of these tests is moderate, but they have high associated specificities, and the positive likelihood ratios for these examinations vary between 6 and 20.
After the apprehension tests are used to diagnose anterior shoulder instability, the examination turns to ruling out associated pathology. The most important concomitant pathology is a rotator cuff tear, which should have been detected with muscle strength testing. The belly-press test and lift-off test can be used to diagnose subscapularis insufficiency, the Jobe empty can test evaluates supraspinatus integrity, and massive cuff tears can be identified with the dropping sign and horn blower’s sign ( signe de Clarion ). Other commonly associated lesions include superior labral anterior to posterior (SLAP) tears, posterior or circumferential labral tears, and incompetence of the rotator interval. SLAP tears are notoriously difficult to diagnose with physical examination. Anecdotally, we have found that O’Brien’s test works best to diagnose SLAP tears when combined pathology is present, but the significance of concomitant SLAP lesions is unclear. Posterior labral pathology can be detected with the posterior apprehension test and jerk test. Competency of the rotator interval can be evaluated by performing the sulcus test in 30 degrees of external rotation.
Lastly, but very certainly not of least importance, is the evaluation of laxity. Several tests are used to gain comprehensive knowledge of the various ligamentous structures. Anterior shoulder laxity is determined by performing external rotation with the arm at the side ( Fig. 46-8, A ). More than 85 degrees of external rotation is considered to be lax. Inferior hyperlaxity is evaluated with the Gagey hyperabduction test, which is performed with the patient in the seated position and the examiner standing behind the patient. The examiner’s arm is forcefully placed on the shoulder to prevent scapular movement ( Fig. 46-8, B ). The shoulder is passively abducted and the amount of glenohumeral abduction, prior to the initiation of scapulothoracic movement, is noted. More than 105 degrees of abduction is consistent with inferior laxity. The examination is concluded with an assessment of generalized ligamentous laxity, using the Beighton criteria ( Fig. 46-8, C ).
Imaging is vital in the diagnostic process of anterior shoulder instability. Imaging is obtained for several purposes; the foremost reason is to ensure that the joint is not currently dislocated. Imaging also allows for determination of glenoid or humeral bone loss, identification of the pathoanatomy, and detection of associated pathology. Plain radiographs, CT, and MRI all have a role in diagnosing anterior shoulder instability.
Plain radiographs are obtained initially and should include anteroposterior (AP) shoulder views in internal and external rotation, a true AP Grashey view, and an axillary view. The AP view in internal rotation allows for a generalized assessment of the joint and surrounding structures. Unexpected pathology, such as abnormal calcifications and tumors, can be detected with radiographs. Radiographs also should be carefully inspected for the presence of a Hill-Sachs lesion. The external rotation AP view is also examined for presence of a Hill-Sachs lesion, because only large lesions are typically visualized with this view. The Grashey view is used to assess the glenohumeral joint and the presence of subtle subluxation or joint space narrowing consistent with a cartilage defect or dislocation arthropathy. The axillary view is key in detecting dislocation of the joint and can also be used to visualize a Hill-Sachs lesion. All patients should have an axillary radiograph performed before leaving an emergency department setting to confirm glenohumeral reduction.
Standard views are extremely helpful, but because of the anatomy of the shoulder, overlap of the humeral head and the glenoid on other structures is often significant. As such, a specialized view should be obtained to examine for the presence and severity of glenoid and humeral bone loss. Glenoid bone loss can be detected using the West Point view or Bernageau view. Hill-Sachs lesions are easily identified with use of the Stryker Notch view or Didiée view. We prefer using the apical oblique view (Garth view), because it adequately visualizes both glenoid bone loss and Hill-Sachs lesions with minimal patient positioning ( Fig. 46-9 ).
Advanced imaging is dictated by the history, physical examination, and radiographic findings. CT is used for the determination of bony anatomy, whereas MRI can be used to detect occult soft tissue pathology. Either scan can be performed in association with arthrography to increase overall sensitivity of the examination. The choice of CT versus MRI is somewhat controversial, but the current trend is toward a CT scan because of the importance of detecting bony pathology before performing standard arthroscopic stabilization. Indications for obtaining a CT scan include multiple dislocations, increasing ease of dislocations and/or reductions, apprehension on physical examination with the arm in less than 75 degrees of abduction, and radiographic evidence of glenoid bone loss.
Quantitative evaluation of glenoid bone loss is best performed using 3D CT. Use of this imaging modality permits the humerus to be subtracted and the glenoid to be viewed en face. Numerous methods have been described to quantify bone loss, and currently, no standard has been agreed upon. Perhaps the most straightforward technique is based on the observation that the inferior aspect of the glenoid is a true circle, with the bare spot located in the exact center. As such, the simple mathematical principle that the diameter of the circle is twice the radius can be used to determine the percentage of bone loss.
This quantification method has been termed the AP distance from bare area method in the literature ( Fig. 46-10 ). To quantify bone loss, the 3D CT scan is used with the humerus subtracted. The first goal is to estimate the bare area. To do so, a vertical line from the supraglenoid tubercle at the most superior aspect is drawn. Next, a horizontal line at the widest AP distance is drawn. At the intersection of these two lines is the approximated bare area. The next goal is to draw a best-fit circle, which is centered about the bare area. The distance from the bare area to the posterior glenoid rim is measured. The distance from the bare area to the remaining anterior glenoid rim is also measured. The percentage of bone loss can be calculated as [(posterior distance)−(anterior distance)]/(2 × posterior distance). Significant instability that is not amenable to standard arthroscopic techniques is thought to occur at more than 30% of glenoid width loss.
If the diagnosis of shoulder instability is in question or an associated rotator cuff tear is suspected, an MR arthrogram is obtained. MRI is able to accurately detect a variety of soft tissue pathologies, but sensitivity is approximately 70%. Intraarticular injection of gadolinium contrast dye results in distention of the joint, thus enabling better anatomic resolution, as well as permitting T1 imaging with a higher signal-to-noise ratio compared with T2 imaging. This use of contrast dye improves the sensitivity into the mid-90% range. Additionally, placing the arm into the abducted and externally rotated position results in increased tension on the IGHL and can accentuate subtle pathology, thus further improving sensitivity. The various soft tissue pathologies are demonstrated in Figure 46-11 . MRI is also accurate in identifying the HAGL lesion ( Fig. 46-12 ), which has traditionally been treated in an open fashion and, as such, is key to identifying preoperatively if arthroscopic repair is typically performed. Recent literature has suggested that CT arthrography is equal, if not superior, to MR arthrography in identifying HAGL lesions. Given the superior performance of multidetector CT scans in identifying bony lesions, CT arthrography may become the reference standard in the future.
Fundamentally, the most important decision in patients with anterior shoulder instability is operative versus nonoperative management. If surgical treatment is elected, the type of procedure must be determined. Options include open or arthroscopic stabilization or the Latarjet (coracoid transfer) procedure. The goals of treatment, regardless of method, are to provide the patient with a stable, durable shoulder that permits the patient to perform his or her desired activities without a loss in quality of life. To achieve these goals, a variety of factors are important to consider when making this decision. These factors include age at initial dislocation, first-time versus recurrent instability, in-season versus off-season with regard to sports activity, sporting activity level, and associated injuries.
Age at Initial Dislocation
Age at the time of the initial dislocation is the single most important factor for predicting recurrent instability. McLaughlin and Cavallaro first discovered that the incidence of recurrence is approximately 90% in patients younger than 20 years and that the incidence of recurrence decreases to only 10% in persons older than 40 years. In a study spanning 20 years of admissions to the Massachusetts General Hospital, Rowe observed an 83% recurrence rate in persons younger than 20 years. The largest prospective study with a 25-year follow-up was published in 2008 by Hovelius and colleagues. They also found that age was the most important predictive factor, with a 72% recurrence rate in patients younger than 20 years. Recurrence decreased with age, with a 56% rate of recurrence in persons aged 23 to 29 years and a 27% rate in persons between 30 and 40 years of age.
Investigators have also examined recurrence rates in groups of patients on either end of the age spectrum. Marans and colleagues found a 100% recurrence rate in children with open physes at the time of initial dislocation. The mean time to redislocation was 8 months. The rate of recurrence does not decrease to a negligible level in elderly patients, however. Gumina and Postacchini found a recurrence rate of 22% in patients older than 60 years.
Many other studies have been published that have indirectly found associations between age and recurrent instability. The results of these studies can be broadly summarized as observing a high rate in young patients and a lower rate in older patients. Recurrence occurs in 70% to 95% of persons younger than 20 years, 60% to 80% in persons aged 20 to 30 years, and 15% to 20% in persons older than 40 years at the time of the initial dislocation.
First-Time Versus Recurrent Instability
The decision to proceed with surgical stabilization after a first-time shoulder dislocation is controversial. Four randomized controlled trials and two metaanalyses have been published to guide this decision. Three of the four randomized trials compare arthroscopic Bankart repair with conservative care, whereas the fourth compared open Bankart repair with use of a sling.
The metaanalyses of these studies demonstrate that Bankart repair results in a significant reduction in the risk of recurrent instability over a 2- to 10-year period compared with use of a sling or arthroscopic lavage. Furthermore, with use of the Western Ontario Shoulder Instability Index as a disease-specific quality of life measurement tool, Bankart repair is associated with improved quality of life over a 2- to 5-year period. However, global patient satisfaction was no different between conservative treatment and operative groups.
These trials and metaanalyses have significant limitations that render the conclusions somewhat speculative. Two of the four studies use operative techniques that are no longer performed. Activity levels were poorly defined, different rehabilitation protocols were used, and the length of follow-up was highly variable.
The present literature supports use of the Bankart repair in young persons with a first-time dislocation as a method of reducing recurrent dislocation and improving quality of life. Further research is required to better substantiate these claims.
Anterior shoulder instability frequently occurs in athletes, with most traumatic events occurring during competition. Approximately half of instability events lead to more than 10 days lost to sport, despite the typical pressure on the athlete to return as quickly as possible. As such, the management of anterior shoulder instability during the sporting season is difficult. Ultimately, the goals of treatment remain the same, but in the competitive athlete, additional goals include minimizing time away from sport, the prevention of additional injury, and ensuring safe return to full activity.
In an study of 30 athletes who sustained an in-season dislocation across a variety of sports, 90% were able to return to their sport in the same season using a supervised physical therapy program and no period of immobilization. Seventy-five percent of patients returned to sport wearing either a Duke Wyre or Sully brace, and 37% had recurrent in-season instability episodes. Ultimately, 53% of all patients underwent surgical stabilization.
These results suggest that that a highly competitive athlete who sustains an in-season shoulder instability event has the potential to be treated nonoperatively. The algorithm developed by Owens and colleagues is particularly helpful in identifying patients who are appropriate for a trial of nonoperative management and early return to sport. After a complete physical examination and radiographic evaluation, players with an initial instability episode who do not have osseous defects of the glenoid or humeral head more than 25% and are at the beginning of the season with time to appropriately rehabilitate are most suitable for physical therapy and sport-specific training. Criteria for return to play include full range of motion and normal strength. A brace should be used upon return to play when possible.
Athletes in Contact Sports
Athletes who compete in contact sports are particularly prone to anterior shoulder instability. Achieving return to sport is much more difficult for them, as evidenced by a study of professional rugby players whose mean absence from competition after a shoulder dislocation was 81 days. In this cohort of athletes, the decision to proceed with surgical stabilization is typically not difficult. However, deciding between an open or arthroscopic technique can be problematic.
Athletes in contact sports have typically been treated with an open Bankart stabilization, based on its historical gold standard status and early reports of increased failure rates after arthroscopic stabilization. Pagnani and Dome reported no postoperative dislocations and 3% postoperative subluxation in American football players after open Bankart stabilization, whereas Uhorchak and colleagues reported a 3% recurrent dislocation rate and a 19% subluxation rate in contact athletes after open Bankart stabilization and anterior capsulorrhaphy. Despite the low rates of recurrence after open Bankart stabilization, more sophisticated arthroscopic techniques have recently been developed, and arthroscopic stabilization is being increasingly recommended to avoid the morbidity associated with open Bankart stabilization. Recently, open stabilization using the Latarjet technique has been reported in rugby players. Neyton and colleagues reported no postoperative dislocations or subluxations with a mean 12-year follow-up.
Given the current status of the literature, at our clinic we have begun to recommend the Latarjet procedure or arthroscopic Bankart repair for athletes participating in contact sports.
A variety of associated injuries may occur after an anterior shoulder dislocation, including rotator cuff tear, greater tuberosity fracture, brachial plexus palsy, axillary nerve palsy, and axillary artery injury. Associated injuries occur predominately in the older population, with 40% sustaining some form of injury after dislocation. Dislocation with associated greater tuberosity fracture is most common and occurs in approximately 15% of older patients. Rotator cuff tears, which are the second most common asssociated injury, occur in 10% of older patients. Neurologic deficit, primarily involving the axillary nerve, occurs in 5% of older patients.
Identifying these associated injuries is critically important because they can significantly influence management. Patients with an associated greater tuberosity fracture are generally viewed as having a better prognosis than patients who have an associated rotator cuff tear. Hovelius and colleagues found a 32% rate of redislocation in patients with isolated anterior dislocation but no redislocations in patients with an associated greater tuberosity fracture. Few studies have been performed to guide parameters for operative fixation of associated greater tuberosity fractures, but given the biomechanical disadvantage of small amounts of displacement of the greater tuberosity, the current consensus in the literature is fixation of fractures with more than 5 mm of displacement.
Rotator cuff function must be carefully evaluated in the older patient. It is not uncommon for rotator cuff dysfunction to be confused with axillary nerve palsy, and any patient with suspected axillary nerve palsy after anterior dislocation should undergo ultrasonographic or MRI evaluation to rule out a rotator cuff tear. Furthermore, given the prevalence of rotator cuff tears in the older population, any history of preexisting shoulder dysfunction should be obtained. As with greater tuberosity fractures, few studies are available in the literature to support management. Older patients who sustain an anterior shoulder dislocation should be carefully monitored. Patients who experience significant dysfunction or pain should be considered for operative management.
When diagnostic imaging shows that patients with suspected axillary nerve palsies do not have rotator cuff tears, electrodiagnostic studies may be obtained. The prognosis for recovery is excellent, with one study demonstrating a 100% recovery rate.
Should surgery be elected, the aforementioned considerations must be synthesized into a comprehensive view of the patient. Balg and Boileau have developed the Instability Severity Index Score as a way to determine which patients would ultimately benefit most from arthroscopic Bankart repair or the Latarjet procedure. In a prospective case-control study, these investigators identified six risk factors that, when combined as a scoring system, resulted in unacceptably high rates of failure after arthroscopic stabilization ( Table 46-2 ). Patients with more than 6 points had a recurrence risk of 70%, whereas patients with a score of 6 or less had a recurrence risk of 10%. Accordingly, patients with an Instability Severity Index Score greater than 6 should undergo an open Latarjet procedure, whereas patients with 6 points or fewer are acceptable candidates for arthroscopic Bankart repair. Currently, at our clinic, competitive athletes younger than 20 years who are involved in a contact or forced overhead sport are counseled regarding the risks, benefits, and alternatives regarding Latarjet and arthroscopic Bankart reconstruction.