Churchill et al. [12] reported that the normal glenoid is retroverted a mean of 1.2°. Their study reported a range between 9.5° of anteversion to 10.5° of retroversion in 344 human scapulae with a mean age of 25.6 years, indicating that a high degree of anatomic variability exists across the general population. Although it is unknown whether these shoulders were afflicted with recurrent instability, most evidence suggests that excessive glenoid version (anterior, posterior, superior, or inferior) or humeral torsion may be associated with decreased glenohumeral stability [12–17] and rotator cuff tears (Fig. 6.5) [18].

The coracoacromial arch is situated anterosuperiorly above the humeral head and is composed of the anterior acromion and the coracoid with the coracoacromial ligament spanning between these structures. This arch is known to prevent excessive anterosuperior migration of the humeral head (Fig. 6.6). However, contact of the humeral head with the undersurface of the acromion can be both a cause and effect of significant rotator cuff disease (see Chap. 4). In general, clinical instability as a result of superior humeral head migration in the absence of a large rotator cuff tear is an extremely rare entity.

The glenoid labrum is a triangular, fibrocartilaginous structure that adheres to the circumference of the glenoid rim (see Fig. 6.1). Its primary function is to provide an extension of the bony glenoid by increasing both its depth and surface area (Fig. 6.7)—factors that have been shown to contribute to approximately 10 % of glenohumeral stability [19, 20]. Recently, Park et al. [21] studied the effect of labral height on subjective outcomes in 40 patients who underwent arthroscopic repair of soft-tissue lesions of the anteroinferior glenoid (i.e., Bankart lesions). Patients with decreased labral height after repair demonstrated inferior clinical outcomes 1 year postoperatively (via Rowe scores) when compared to those with higher labral height. In addition to improving glenoid depth and contact surface area, the glenoid labrum also serves as an attachment site for the joint capsule and the glenohumeral ligaments. Approximately 40–60 % of the LHB substance originates from the superior labrum (the remaining percentage originates from the supraglenoid tubercle) (Fig. 6.8) [22].

With the humerus in a neutral position (such as the “loose pack position” described in Chap. 2), stability is achieved through dynamic muscle contraction since the glenohumeral joint capsule and associated ligamentous structures are somewhat lax in this position [23]. However, these structures become variably taught with both active and passive shoulder motion which both maximizes articular surface contact and prevents abnormal humeral head translation (Fig. 6.9). The joint capsule itself, the coracohumeral ligament (CHL) and the superior, middle, and inferior glenohumeral ligaments (SGHL, MGHL, and IGHL, respectively) make up this important capsuloligamentous complex (Fig. 6.10).

The glenohumeral joint capsule is one of the primary static restraints involved in maintaining shoulder stability. It is primarily composed of collagen fiber bundles with varying degrees of thickness and fiber orientation [24, 25]. The CHL, SGHL, MGHL and IGHL make up distinct bands or areas of capsular thickening that are important for the maintenance of glenohumeral stability in any plane of shoulder motion. These structures become variably taught with both active and passive motion, thus maximizing articular surface contact and preventing abnormal humeral head translation. The shape and function of each individual ligament is determined by the position of the humerus in space.

The CHL has two distinct bands (anterior and posterior) that originate from the lateral base of the coracoid and travel between the supraspinatus and subscapularis tendons. The anterior band merges with the insertional fibers of the subscapularis tendon and the joint capsule near the lesser tuberosity whereas the posterior band inserts over the anterior aspect of the greater tuberosity. With the arm at the side, the anterior band primarily resists excessive external rotation and the posterior band primarily functions to prevent excessive internal rotation. Both the anterior and posterior bands of the CHL also provide resistance to inferior humeral head translation with the arm at the side and posterior humeral head translation when the arm is horizontally adducted [26–29].

The SGHL originates from the superior rim of the glenoid near the biceps-labral complex, travels parallel to the much larger CHL, and inserts on the lesser tuberosity, blending with the fibers of the subscapularis tendon. Its usual functions are similar to that of the CHL, preventing excessive external rotation [30] and inferior translation [31] when the arm is at the side and preventing posterior translation when the arm is horizontally adducted. However, its diameter, strength, and relative contribution to shoulder stability are highly variable across the population.

The anatomy of the MGHL is also highly variable. It can originate from the scapular neck, the anterosuperior glenoid rim or the supraglenoid tubercle with the biceps-labral complex. Similar to the CHL and the SGHL, the distal insertion of the MGHL blends with the fibers of the subscapularis tendon. The morphologic phenotype of the MGHL can also range in appearance from a round, cord-like ligament to a flat, sheet-like structure that blends with the IGHL inferiorly. Biomechanical ligament sectioning studies have shown that the MGHL primarily functions as a restraint to anterior translation when the humerus is between 0° and 45° of abduction and externally rotated [23]. In addition, the MGHL may also be important in limiting external rotation when the humerus is abducted greater than 60°.

The IGHL complex circumferentially attaches to the inferior aspect of the glenoid labrum anteriorly, inferiorly, and posteriorly and runs laterally to widely insert over an area extending between the lesser tuberosity anteriorly and the triceps tendon posteriorly. The IGHL is composed of thick anterior and posterior bands with an interposed “hammock-like” pouch that loosely cradles the inferior aspect of the humeral head (see Fig. 6.10). Its function is to resist both anterior and posterior humeral head translation when the humerus is abducted more than 60° [30]. Specifically, as the humerus is abducted and externally rotated, the anterior band of the IGHL complex along with the anteroinferior capsule becomes taut and prevents anterior humeral head translation. When the humerus is flexed, adducted, and internally rotated, the posterior band of the IGHL complex and the posterior capsule become taut and prevent posterior humeral head translation (see Fig. 6.9).

The rotator interval is a triangular space over the anterosuperior aspect of the joint capsule. The supraspinatus forms the superior border, the subscapularis forms the inferior border, and the coracoid process forms the medial base (Fig. 6.11). The CHL, the SGHL, the MGHL, the LHB tendon and the anterosuperior joint capsule all reside within this triangular space. Jost et al. [32] performed one of the more detailed cadaveric studies in which the rotator interval was described as being composed of several layers. However, the precise anatomy of the rotator interval is still under investigation and is beyond the scope of this chapter.

The exact function of the rotator interval is also the subject of numerous biomechanical studies [26, 32–39]. However, many of their reported results have been conflicting. Harryman et al. [36] performed one of the first comprehensive and descriptive studies that examined the function of the structures within the rotator interval. After dividing the capsule and ligamentous structures within the rotator interval in a series of 80 cadaveric shoulders, the investigators noted an increase in passive glenohumeral flexion, extension, external rotation, and adduction capacity. Medial–lateral imbrication of the same structures resulted in the opposite effect, thus decreasing these motions. The authors concluded that the rotator interval provided resistance against excessive motion while also functioning to limit posteroinferior glenohumeral translation. Nobuhara and Ikeda [39] also showed that tightening of the rotator interval decreased the propensity for humeral head translation in the posteroinferior direction. As a result of these studies, most surgeons believe that the rotator interval does provide some degree of stability, especially inferiorly when the humerus is externally rotated.

As mentioned in Chap. 5, the rotator interval also contributes to stability of the LHB tendon as it travels through the bicipital groove towards the superior labrum and supraglenoid tubercle. Specifically, the SGHL, CHL, and subscapularis tendon together form a structure known as the biceps reflection pulley which supports the tendon as it enters the glenohumeral joint (Fig. 6.12) [2, 26, 27, 40].

Within the closed joint, a negative intra-articular pressure is produced by either a “piston-in-valve” mechanism, an adhesion-cohesion effect of the viscous synovial fluid, or both, which exist to provide some degree of stability to the glenohumeral joint [41]. Any perforation of the joint capsule can “vent” the joint, eliminating this nascent pressure gradient (Table 6.1) [41–43]. However, although this mechanism may provide some joint stability, capsular venting alone is not an apparent cause of clinical instability.

Many of the structures described above have several anatomic variations that are important to recognize from a management perspective (Fig. 6.15). However, because these variations are not discernible by physical examination, an in-depth discussion of each variation would not be helpful in the context of this chapter. Rather, we have provided a summary of this information with their corresponding references in Table 6.2 in an effort to direct the reader towards the most relevant published evidence related to these anatomic variations. In addition, we recommend reviewing the article published by Tischer et al. [53] which provides a detailed discussion of the relevance of each anatomic variation as they relate to the arthroscopic management of instability.

Instability is typically described according to severity (microinstability, subluxation, or dislocation), direction (anterior, posterior, inferior, or multidirectional), and chronicity (acute, chronic, or acute on chronic). In other cases, instability can also be described as being voluntary or involuntary [54] and hereditary or acquired [55]. Although several classification systems have been proposed, none of these have been comprehensive nor have they been proven to adequately facilitate communication between physicians, guide treatment decisions or predict outcomes. In 1989, Thomas and Matsen [56] categorically divided those with instability into two distinct groups according to the distinctive characteristics of their condition. In one group, the acronym TUBS was used to describe individuals with Traumatic, Unilateral instability with a Bankart lesion that generally requires Surgical repair. The other group was described using the acronym AMBRI which included patients with Atraumatic, Multidirectional instability which was typically Bilateral, Responded to physical therapy and sometimes required Inferior capsular plication to prevent recurrence. In this group, many individuals are afflicted with underlying multiligamentous laxity who gradually develop instability as they age. However, although this classification system has some academic merit, its usefulness in clinical practice is very limited since most patients present with pathologic traits that overlap between the TUBS and AMBRI groups. For example, many patients with generalized hyperlaxity present with uni- or bi-directional instability [57] whereas up to one-fourth of patients with traumatic instability display evidence of contralateral involvement and increased capsular elastin content which suggests the possibility of familial inheritance [55]. This overlap indicates that the term “instability” most likely encompasses a continuous spectrum of pathologic features where the TUBS and AMBRI groups represent the terminal ends.

As a result, other classification systems were proposed to account for this continuum of pathologies associated with glenohumeral instability. Particularly noteworthy is the Stanmore classification developed by Lewis et al. [58] in 2004 in which the three points of a triangle represent the polar pathologic characteristics associated with instability (Fig. 6.16). Type I represented traumatic instability, type II represented atraumatic instability, and type III represented neurologic dysfunction and muscle patterning (i.e., voluntary instability). The line between types I and II corresponded with the spectrum of disease between traumatic (TUBS) and atraumatic (AMBRI) etiologies. The line connecting types II and III corresponded with the spectrum of disease in those with atraumatic instability and neurologic dysfunction. As one moves towards the apex of the triangle (i.e., towards the polar type I pathology), the involved pathology increasingly resembles the clinical picture of traumatic instability. The opposite is true when one moves towards the base of the triangle. While this system adequately represents the variability in clinical presentation, it fails to guide the clinician towards a specific treatment option for individual clinical scenarios.

In 2008, Kuhn et al. [59] systematically reviewed the literature to determine the frequency of various criteria that have been used to classify glenohumeral instability. The authors observed that four of these criteria were used in more than 50 % of the proposed classification systems. They also noted that a survey conducted by the American Shoulder and Elbow Surgeons (ASES) echoed these same results. As a result, a classification system that accounts for Frequency, Etiology, Direction and Severity of instability (FEDS system) was developed since these features were found to be the most important factors involved in treatment decision-making. Subsequent analysis revealed that the FEDS classification had high inter- and intra-rater agreement [60]. However, the major drawback of the FEDS system is its inability to categorize all patients who present with instability. Therefore, further studies that correlate this classification system with treatment outcomes are needed. With this information, clinical results could be predicted at the time of initial clinical evaluation.

Due to the lack of a single validated classification system for glenohumeral instability, the clinician must utilize the concepts derived from several different classification schemes and published literature to make individual treatment decisions based on the etiology and underlying pathologic lesions associated with the injury.

It is often difficult for clinicians to define which patients are afflicted with atraumatic instability because the effects of activities of daily living and/or sporting activities may lead to undetectable glenohumeral joint damage. The cumulative effects of this damage may lead to unilateral or bilateral instability without an apparent cause. However, some degree of genetic predisposition is implied when patients present with atraumatic bilateral shoulder instability [83]. For example, multidirectional instability (MDI) is defined as atraumatic anterior or posterior instability with a component of increased inferior translation. Studies have demonstrated increased elastin content in both the skin and capsular tissue of many patients with MDI [84] in addition to increased capsular volume [57, 85, 86] and, in some cases, laxity of the rotator interval [87]. These findings suggest that undiagnosed Ehlers–Danlos syndrome or multiligamentous laxity may be a substantial contributing factor involved in the development of instability in many of these patients. It should be also noted that traumatic and atraumatic instability can occur simultaneously in the same patient and thus should not be considered entirely independent from one another. For example, a patient diagnosed with MDI can also present to the clinic after a traumatic dislocation, potentially resulting in any of the pathologic lesions associated with traumatic instability (i.e., bony Bankart lesion, Hill–Sachs lesion, HAGL lesion, etc.).

The semantic relationship between laxity and instability should be recognized and understood by all clinicians who evaluate patients with various shoulder pathologies. The term “laxity” refers to the normal physiologic motion allowed as a result of the position and tension of the ligaments that maintains stability of a joint [88–90]. Without this physiologic laxity, joint motion would not be possible. Because the shoulder requires a large range of motion, its physiologic laxity has a greater magnitude than the other joints within the body. Therefore, laxity testing in the shoulder requires that the clinician understands the difference between “normal” and “abnormal” joint motion as they relate to the entire clinical picture. Along the same lines, the clinician should also understand that increased joint laxity does not necessarily equate pathologic instability, even if this finding occurs unilaterally. As mentioned above, these conditions lie along a spectrum of disease that is most frequently and conveniently labeled as “instability.”

Currently, there are three basic methods by which humeral head motion is quantified: (1) translation in millimeters, (2) translation as a percentage of the humeral head diameter, and (3) the sensations felt when the humeral head is translated. A fourth modality includes the use of instrumentation or imaging; however, these methods are currently under development.

Patient guarding or apprehension is one of the main challenges associated with laxity testing of the shoulder in the office setting. With these maneuvers, it is required that the patient remains relaxed to allow the humeral head to translate appropriately during testing. While many of these tests can be performed in the sitting or supine position, several authors have noted that laxity testing with the patient in the supine position may produce the best results because the patient is generally more relaxed [111, 112]. Another challenge associated with laxity testing is the interpretation of the clinical findings. Although the end feel classification system derived by Cyriax and Cyriax [113] in 1947 has been used in the past (see Chap. 2), defining the quality of the end point in shoulder laxity testing is not practical since none of these qualities can be associated with any specific pathology, treatment or outcome. However, in the clinical setting, the reproduction of symptoms is often a strong indicator of the underlying diagnosis and may also direct the use of other provocative maneuvers.