Management of Bone Loss in Glenohumeral Instability




This review discusses the evaluation and management of bone loss in glenohumeral instability. The glenohumeral joint may experience a dislocation or subluxation associated with traumatic injury or through repetitive atraumatic events. Nearly 62% of cases with recurrent dislocation have both Hill-Sachs and bony Bankart defects. Treatment of unstable bone defects may require soft-tissue repair, bone grafting, or both, depending on the size and nature of the defects. The most common treatment is isolated soft-tissue repair, leaving the bone defects untreated, although emerging evidence supports directly addressing these bony defects.


Key points








  • With anterior dislocations, bony defects of the anterior glenoid and posterosuperior aspect of the humeral head occur with relative frequency.



  • In shoulders sustaining a Hill-Sachs lesion at the initial dislocation, there exists a statistically significant association with recurrent dislocation.



  • When a patient has symptomatic anterior instability associated with an engaging Hill-Sachs lesion with an articular arc deficit, treatment must be directed at both repairing the Bankart lesion, if present, and preventing the Hill-Sachs lesion from engaging the anterior glenoid.



  • Glenoid bone loss often requires bone-block transfers using the coracoid (Bristow/Latarjet) or iliac crest autograft.



  • Humeral bone loss can be addressed through a variety of surgical options, including humeroplasty, remplissage, partial resurfacing, allograft transfers, and total shoulder arthroplasty.






Introduction


The human shoulder is the most mobile joint in the body and consequently the glenohumeral joint is one of the most commonly dislocated joints in the body. Glenohumeral instability affects approximately 2% of the general population, with anterior dislocations occurring 95% to 98% of the time. With anterior dislocations, bony defects of the anterior glenoid and posterosuperior aspect of the humeral head occur with relative frequency ( Fig. 1 ). These osseous injuries directly affect recurrent instability by altering joint-contact area, congruency, and function of the static restraints. Thus, restoration of normal articular geometry should be considered when critical bone loss exists, especially in cases of failed soft-tissue stabilization procedures.




Fig. 1


Mechanism of traumatic anterior shoulder dislocation. ( A ) Combined forces in external rotation and anterior translation overcome internal restraints, resulting in anterior dislocation. ( B ) This process results in compression of the posterolateral aspect of the humeral head onto the anterior glenoid rim.

( Courtesy of Cleveland Clinic Foundation, Cleveland, OH.)


This review presents the epidemiology and pathophysiology of bone loss relevant to anterior shoulder instability, and summarizes the evaluation and management of this problem.




Introduction


The human shoulder is the most mobile joint in the body and consequently the glenohumeral joint is one of the most commonly dislocated joints in the body. Glenohumeral instability affects approximately 2% of the general population, with anterior dislocations occurring 95% to 98% of the time. With anterior dislocations, bony defects of the anterior glenoid and posterosuperior aspect of the humeral head occur with relative frequency ( Fig. 1 ). These osseous injuries directly affect recurrent instability by altering joint-contact area, congruency, and function of the static restraints. Thus, restoration of normal articular geometry should be considered when critical bone loss exists, especially in cases of failed soft-tissue stabilization procedures.




Fig. 1


Mechanism of traumatic anterior shoulder dislocation. ( A ) Combined forces in external rotation and anterior translation overcome internal restraints, resulting in anterior dislocation. ( B ) This process results in compression of the posterolateral aspect of the humeral head onto the anterior glenoid rim.

( Courtesy of Cleveland Clinic Foundation, Cleveland, OH.)


This review presents the epidemiology and pathophysiology of bone loss relevant to anterior shoulder instability, and summarizes the evaluation and management of this problem.




Humeral bone loss


One of the first descriptions of the lesions found on the humeral head was by Flower in 1861, with many subsequent investigators reporting on these bony defects. In 1940, 2 radiologists, Hill and Sachs, reported that these defects were actually compression fractures produced when the posterolateral humeral head impinged against the anterior rim of the glenoid. In their series of recurrent anterior glenohumeral instability, these lesions were found in 74% of patients. The true incidence of Hill-Sachs lesions is unknown; however, they are associated with approximately 40% to 90% of initial anterior glenohumeral dislocations. The incidence in recurrent instability can vary from 70% up to 100%, with arthroscopy often identifying lesions not appreciated on imaging. The management of Hill-Sachs lesions depends mainly on the size of the lesion and whether it is engaging. Most lesions are small and clinically insignificant. Often, lesions that are clinically relevant may be indirectly managed with procedures aimed at addressing primary instability at the glenoid (ie, Bankart repair, glenoid reconstruction, and so forth).




Glenoid bone loss


The characteristic anteroinferior capsulolabral injury (ie, Bankart lesion) associated with an acute anterior shoulder dislocation has been termed the essential lesion. Rowe and colleagues first described glenoid bone loss as a “rim fracture” following anterior instability. Rowe’s key finding was the importance of the anterior glenoid rim in providing anterior shoulder stability, by creating a deepened concave surface of the glenoid and increased articular coverage. The importance of the rim fracture is shown in multiple studies by analyzing the relationship of the glenoid and humerus, especially in external rotation and abduction. Bigliani and colleagues provided the first detailed description of osseous glenoid rim injuries, which included rim fractures and erosions ( Fig. 2 ). In a radiographic study of patients with recurrent instability, 87% of shoulders involved the presence of either a glenoid rim fracture or erosion. Griffith and colleagues used 2-dimensional (2D) computed tomography (CT) to find glenoid bone loss in 41% of 66 patients with a first-time dislocation and 86% of 137 patients with recurrent instability. The predominant pattern of injury was attritional bone loss, with glenoid rim fractures reported in only 21% of 233 dislocated shoulders. However, using 3-dimensional (3D) CT to assess glenoid bone loss in 100 patients with recurrent instability, Sugaya and colleagues found that only 40% of patients had erosive or attritional bone loss.




Fig. 2


Mechanism of glenoid fossa fractures. The force vector at the time of impact between the humeral head and glenoid fossa determines the morphology of the glenoid fracture. ( A ) Small rim-type fracture. ( B ) Larger fracture extending into the glenoid vault.

( Courtesy of Cleveland Clinic Foundation, Cleveland, OH.)




Bipolar bone loss


The literature investigating osseous lesions of both the glenoid and humeral head is limited. The prevalence of combined bone defects is reported to be 64% to 70% in first-time anterior dislocations and 79% to 84% in recurrent anterior glenohumeral instability.




Pathophysiology


Knowledge of the pathoanatomy and biomechanics of glenohumeral bone loss and instability is crucial in the appropriate management to prevent recurrent instability. The most common mechanism of traumatic anterior shoulder dislocation occurs with an indirect force on the abducted and externally rotated arm. The humeral head externally rotates relative to the glenoid while translating anteriorly. The static glenohumeral restraints (ie, capsule, ligaments, labrum) are stretched or torn with further anterior translation, and dislocation, of the humeral head. The posterosuperolateral aspect of the humeral head then impacts on the anterior aspect of the glenoid rim and can create a Hill-Sachs lesion and/or a bony Bankart lesion.


Richards and colleagues investigated the location and depth of 28 arthroscopically confirmed Hill-Sachs lesions. On an axial view with 0° representing direct anterior, the typical Hill-Sachs lesion lies between 170° and 260° with a midpoint at 209°. Saito and colleagues used axial CT imaging to demonstrate that the normal bare area of the humeral head was located deeper than a typical posterolateral humeral head Hill-Sachs defect, allowing differentiation between the two.


Palmer and Widen and Burkhart, Danaceau, and De Beer described an “engaging” Hill-Sachs lesion as one that encounters the anterior glenoid rim with the arm in the “active” position of abduction (90°) and external rotation (0°–135°) and can lever the humerus from the glenoid concavity ( Fig. 3 ). These humeral head defects are parallel to the surface of the anterior glenoid when the arm is abducted and externally rotated. This defect has been termed an articular arc deficit, as there is disruption of the glenohumeral articulation on motion. Lesions that are not parallel to the glenoid rim in the active or athletic position do not engage, and are termed nonengaging lesions. The Hill-Sachs defect passes diagonally across the anterior glenoid with external rotation; therefore, there is continual contact of the articulating surfaces and no engagement of the Hill-Sachs lesion by the anterior glenoid.




Fig. 3


Engaging and nonengaging Hill-Sachs lesions. ( A ) An engaging lesion is parallel to the anterior glenoid rim when the shoulder is in a functional position. ( B ) The “engagement point” of a nonengaging lesion occurs with the arm in a nonfunctional position. ( C ) In a functional position, a nonengaging lesion is diagonal and nonparallel to the anterior glenoid rim.

( Adapted from Burkhart SS, De Beer JF. Traumatic glenohumeral bone defects and their relationship to failure of arthroscopic Bankart repairs: significance of the inverted-pear glenoid and the humeral engaging Hill-Sachs lesion. Arthroscopy 2000;16:677–94; Courtesy of Cleveland Clinic Foundation, Cleveland, OH.)


Cho and colleagues looked at 3D CT scans of 107 shoulders undergoing surgery for recurrent anterior instability to preoperatively predict engagement of a Hill-Sachs lesion. The mean width was 52% (range, 27%–66%) and depth 14% (range, 8%–20%) of the humeral head diameter on axial images. The magnitude of bone loss that coincides with a Hill-Sachs lesion depends on multiple factors including dislocation frequency, chronicity, and force. Cetik and colleagues found an increasing percentage of articular surface involvement with increasing frequency of dislocations. Intraoperative assessment revealed an average involvement of 26% of the articular head in patients with greater than 20 dislocations. The size of the humeral head defect is also directly related to dislocations of longer duration, as seen with neglected and locked shoulder dislocations.


The critical limit or threshold of humeral bone loss on glenohumeral stability has been investigated in many ways. In a cadaveric study, Sekiya and colleagues found that defects that were 25% of the humeral head diameter or larger revealed significantly ( P <.05) less anterior translation before dislocation, and decreased stability ratios (displacing force divided by compressive load) when compared with the intact specimens. Furthermore, Kaar and colleagues noted that defects that were greater or equal to five-eighths of the humeral head radius lead to decreased glenohumeral stability when tested in the functional position of abduction and external rotation.


Hill-Sachs lesions typically are accompanied with other abnormality including soft-tissue and/or bony Bankart lesions and anterior glenohumeral ligament disruption. In the clinical setting, there are essentially 2 types of anterior glenoid defects that occur after an instability event: rim fracture or avulsion, and compression fracture or erosive bone loss. The angle of the humeral head and shaft relative to the glenoid fossa, along with the energy, determine the extent of the resulting glenoid rim fracture (see Fig. 2 ). Recurrent or repetitive subluxations may have more shear and less axial load, leading to attritional bone loss rather than a large rim fracture that may accompany a high-energy axial load. When viewing the glenoid en face, the area of bone defect is nearly parallel to the long axis of the glenoid fossa. Saito and colleagues found the average osseous glenoid injury to range from 12:08 to 6:32 on the clock-face scheme with the midpoint in line with 3:01. However, clinical bone loss can still occur in more anterior-inferior locations.


Glenoid defects are typically classified with large lesions accounting for greater than 20% of the glenoid fossa, medium lesions ranging from 5% to 15%, and small lesions usually less than 5% of the glenoid fossa. Itoi and colleagues performed a biomechanical analysis on amount of glenoid defect and force required to translate the humeral head to dislocation. The investigators made sequentially larger glenoid defects in the anterior-inferior glenoid (45° from the longitudinal axis) and found that stability progressively decreased as the size of the glenoid defect increased. Specifically, defects at least 21% of the glenoid length led to instability and limited the range of motion of the shoulder. Similarly, Yamamoto and colleagues looked at anterior glenoid rim defects and found that the stability ratio significantly decreased with defects that were 20% or greater of the glenoid length. Optimal surgical management requires addressing these lesions and management of the clinically significant Hill-Sachs lesion.


The concept of the glenoid track was proposed by Yamamoto and colleagues in 2007 ( Fig. 4 ), and serves to illustrate the dynamic of glenohumeral instability in cases of combined defects. The glenoid track represents the pattern of articular contact between the humeral head and the glenoid with the arm in a position of vulnerability for anterior dislocation. The width of the glenoid track was found to be 84% of the inferior glenoid surface. When a glenoid defect exists, the resulting glenoid width is multiplied by 0.84 to calculate the new glenoid track width. If a humeral head defect exists and remains within the glenoid track, there will be no engagement with the anterior glenoid rim. However, if even a margin of the humeral head defect extends beyond the glenoid track, there is a risk that it will engage the glenoid rim. This concept has proved valuable in understanding the clinical significance of bipolar bone loss.




Fig. 4


Glenoid track concept. ( A ) In extremes of external rotation and abduction, the glenoid displaces the cuff tendon close to its footprint, creating a glenoid track that is close to 84% of the glenoid width. ( B ) When a glenoid defect exists, the defect width is subtracted from the 84% width obtained from the normal glenoid to calculate the true glenoid track width.

( Adapted from Yamamoto N, Itoi E, Hidekazu A, et al. Contact between the glenoid and the humeral head in abduction, external rotation, and horizontal extension: a new concept of glenoid track. J Shoulder Elbow Surg 2007;16:649–56; Courtesy of Cleveland Clinic Foundation, Cleveland, OH.)




Natural history


Hovelius and colleagues prospectively followed 229 shoulder dislocations for 25 years. All patients were treated nonoperatively initially and prognostic factors, recurrence, and surgical intervention were monitored. At 10 years, 99 of 185 (53.5%) shoulders that were evaluated with radiographs had evidence of a Hill-Sachs lesion; of these 99 shoulders, 60 redislocated at least once and 51 redislocated at least twice during the 10-year follow-up, compared with 38 (44%) of the 86 shoulders that did not have such a lesion documented ( P <.04). However, at 25 years, the investigators concluded that a small humeral impression fracture at the time of initial dislocation did not influence the recurrence rate. Rowe and colleagues analyzed the long-term results of Bankart repairs for recurrent instability, and found an overall recurrence rate of 3.4% (5 of 145); the recurrence rates were 4.7% and 6% for patients with moderately severe and severe Hill-Sachs lesions, respectively. Whereas Rowe and colleagues used depths of 3 mm, 5 mm, and greater than 10 mm to differentiate their size of Hill-Sachs lesions, various other methods of determining size and/or volume of the humeral head defect have been proposed without consensus; these include the Hill-Sachs quotient, articular arc circumference, and Hill-Sachs angle.


Lo and colleagues noted that bone loss of 25% or greater of the diameter of the inferior glenoid will create an “inverted pear” appearance, and recommended coracoid transfer when glenoid deficiency reached this magnitude. The inverted-pear glenoid had a poor prognosis in the study by Burkhart and De Beer evaluating a series of 194 patients who underwent primary soft-tissue repair for anterior instability. Of the 21 patients with recurrent instability, 14 had either an engaging Hill-Sachs lesion (n = 3) or an inverted pear glenoid shape (n = 11).




History and physical examination


A thorough orthopedic history must be obtained from the patient regarding shoulder instability. Specifics of the history that must be elucidated include the mechanism of instability and timing of initial symptoms. Arm position and amount of force required for instability may be an evolving process with progressively less rotation or force required for subsequent dislocations. Need and method of reduction of the glenohumeral joint may indicate the extent of laxity present. Presenting symptoms should be noted, including pain, frequency, instability, and level of function. Though infrequent, pertinent medical history including collagen disorders or epilepsy should be noted. Many patients will report a history of recurrent dislocations or multiple surgical attempts to correct the instability. All previous surgical procedures performed on the shoulder should be considered and, if possible, operative reports and photos should be obtained.


Physical examination should focus on inspection for previous scars, gross asymmetry, a thorough comparison of active and passive range of motion, strength testing, particularly evaluation of the integrity and strength of the rotator cuff, and axillary nerve function. The clinician should perform a detailed examination for glenohumeral laxity in the anterior, posterior, and inferior directions. Examination for apprehension should be performed in multiple positions (ie, sitting, standing, supine), as patients with large Hill-Sachs lesions usually exhibit apprehension that often occurs with the arm in significantly less than 90° abduction and 90° external rotation. A positive anterior apprehension will be associated with anterior labral injuries. Moreover, apprehension with fewer degrees of abduction may indicate a significant and symptomatic bony contribution to the instability.




Imaging and other diagnostic studies


The ideal imaging technique is easy to reproduce, with excellent reliability among physicians, and able to predict clinically significant bone defects. In addition to standard radiographs of the shoulder, specialized views allow for evaluation of bony defects of the glenoid and humeral head. Preoperative imaging includes a comprehensive radiographic evaluation with anteroposterior (AP), true AP, axillary, West Point axillary, and Stryker notch views of the involved shoulder ( Fig. 5 ).




Fig. 5


Axillary view of the right shoulder shows congruency of the glenohumeral joint. This view also allows for evaluation of bone loss and glenoid version.


The Stryker notch view, in addition to AP internal rotation views, has been found to be most sensitive in detecting humeral head lesions on plain radiographs. Based on these views, various quantification methods have been described ( Fig. 6 ). However, Bois and colleagues note that no method has been universally accepted because of the learning curve required to obtain these radiographic projections and the inconsistency often seen in special radiographic views.




Fig. 6


Methods used to quantify Hill-Sachs lesions. Such defects may be quantified using ( A ) depth or width measurements, ( B ) percentage of humeral head involvement [(X/Y) × 100], and/or ( C ) measurement of the Hill-Sachs angle.

( Courtesy of Cleveland Clinic Foundation, Cleveland, OH.)


Multiple variations of the axillary lateral view have been proposed to evaluate for glenoid defects on plain radiographs (ie, West Point, Bernageau, Garth, and Didiee). The West Point view has been found to be the most accurate for demonstrating glenoid bone loss.


Despite the vast array of radiographic views available, bone loss may often go undetected on plain radiographs. Preoperative advanced imaging study (CT and/or magnetic resonance imaging [MRI]) are obtained to better define the bony architecture of the glenoid and humeral head ( Fig. 7 ). CT can offer the added value of providing better bony detail, with 1-mm slices and 3D reconstructions improving the accuracy of determining the true location and size of the defect. Kodali and colleagues investigated the reliability and accuracy of making width and depth measurements of different-sized Hill-Sachs lesions using axial, sagittal, and coronal 2D CT images. Measurements by 5 physicians were compared with measurements from a 3D laser scanner, and were found to be reproducible and most accurate in the sagittal and axial planes. However, on the glenoid side, in a study performed by Bois and colleagues comparing various 2D and 3D methods of measurement of glenoid bone loss, there was variable agreement and inaccuracy for all 6 observers using 2D CT in measuring defect length and calculating the width/length ratio. Rather, 3D reconstruction is the most reliable and accurate imaging modality for the assessment of glenoid bone loss, and can be a useful tool to more clearly define the size and location of the defect and to estimate the amount of articular surface involved. Quantification of glenoid bone loss can be performed via either a linear method or measurement of surface area. Again, various methods of quantification for both the glenoid and humeral head have been described as being useful in preoperative planning, without universal acceptance.




Fig. 7


Axial computed tomography image of the left shoulder showing anterior glenoid bone loss in addition to a large Hill-Sachs defect on the humeral head.


Despite the known advantages of 3D CT imaging techniques, the disadvantages include the financial burden to the institution and possibly the patient, the need for specialized computer software to quantify bone loss, and the lack of awareness in the orthopedic and radiology communities of the multiple measurement methods available and their general validity.


Dynamic arthroscopy remains the gold standard for evaluation of bone loss, and can be useful for the preoperative planning of patients undergoing open osteoarticular allograft reconstruction to address bony deficiency.




Classification of bone loss


Ideally classification schemes incorporate clinical, radiographic, and prognostic factors. In bone loss with anterior glenohumeral instability, few studies have validated classification schemes ( Table 1 ). On the humeral side, Burkhart and De Beer differentiated between engaging and nonengaging defects; clinically, patients with engaging lesions had a higher failure rate with soft-tissue stabilization procedures. This diagnostic sign has since been adopted by most investigators and surgeons as the method of classifying Hill-Sachs lesions. Glenoid defects were classified by Bigliani and colleagues into 3 main types based on the nature of the rim fracture. This classification was later modified after the work of Boileau and colleagues demonstrated a 75% recurrence rate in patients with a glenoid compression fracture and a stretched inferior glenohumeral ligament ( Fig. 8 ).



Table 1

Classification schemes in bone loss associated with anterior glenohumeral instability






























Authors, Ref. Year Basis Classification
Humeral Head
Rowe et al, 1984 Size (length and depth)


  • Mild




    • 2 × 0.3 cm





  • Moderate




    • 4 × 0.5 cm





  • Severe




    • ≥4 × 1.0 cm


Bigliani et al, 1996 Percentage of head involvement


  • Mild




    • <20%





  • Moderate




    • 20%–45%





  • Severe




    • >45%


Glenoid
Bigliani et al, 1998 Size of rim


  • Type I




    • A displaced avulsion fracture with attached capsule





  • Type II




    • A medially displaced fragment malunited to the glenoid rim





  • Type III




    • Erosion of the glenoid rim with <25% (type IIIA) or >25% (type IIIB) bone loss


Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Oct 6, 2017 | Posted by in ORTHOPEDIC | Comments Off on Management of Bone Loss in Glenohumeral Instability

Full access? Get Clinical Tree

Get Clinical Tree app for offline access