Glenoid Bone Defect in Anterior Shoulder Instability

Abstract

Bipolar bone defect in shoulder instability are main reason for poor results after arthroscopic stabilization for acute or chronic shoulder instability. Precise quantification of those defects and their interaction to each other should be done, when facing patient with multiple shoulder dislocations. Subsequently treatment should be addressed to clinical and imaging findings. CT and MRI are now “gold standards” for decision making regarding bone defects after multiple shoulder dislocations. Bone transfers are established as reliable surgical option for such patients. They provide opportunity to compensate even large bone defects. Laterjet coracoid transfer is one of the best surgical options with easy technique and reliable results.

1. Introduction

Increasing physical activity, especially among young people is reason for respectively, the increase of sport injuries. Shoulder joint, as most mobile joint, is one of the most traumatized joint during sports activities. Shoulder dislocation and shoulder instability are common disorders, that are interrupting normal everyday activities, especially in young athlethes.

Bone defects after shoulder dislocation and recurrent instability are fundamental for diagnostics and treatment of anterior shoulder instability. Precise diagnosis and surgical options are result of proper imaging modalities and the ability of the surgeon to use them. Anterior shoulder dislocation as the reason for shoulder instability affects 40 per 100,000 in men and 15 per 100,000 in women [1]. Results of conservative treatment can lead up to 50% reccurensy [2]. First time dislocators can sustain fracture of anterior glenoid rim or erosion up to 30% of cases [3]. In reccurensy cases described pathologies are increasing up to 86% [4]. Bone defects of anterior glenoid rim (Bankart) and humeral head (Hill-Sachs) are defined as main risk factor for failure of conservative/arthroscopic soft tissue stabilization procedures [5, 6].

Instability of the glenohumeral joint is defined as inability of the humeral head to stay centered into glenoid fossa. This is excessive mobility of joint surfaces to each other, resulting in partial or total dislocation. The difference in partial dislocation is, that articular surfaces still have some contact between and when causative factor is removed, it reduces spontaneous.

In dislocation, translation of the humeral head is so big, that it does not interact with glenoid fossa, respectively it cannot reduce spontaneously.

Definition of shoulder instability comes with proper classification in order to provide proper treatment algorithm. In past decades there are many classifications which are trying to fulfill all requirements – being precise, complete and easy to apply in everyday practice. The closest one, according to literature and our experience, is FEDS classification- it is based on the most important parameters of instability [7] (Table 1).

Direction	Ethiology	Severity	Frequency	Pathoanatomy**
Anterior	Traumatic	Pain*	Single episode	Capsule
Posterior	Requires reduction	Subluxation	2–5 episodes	Labrum
Inferior	No reduction requriment	Dsilocation	>5 episodes	Bone
	Atraumatic; Positional; INvoluntary; Habitual; Microtraumatic*	Incarcerated	Incarcerated

Table 1.

Modified FEDS system.

*

In “overhead” atlethes.

**

Only image detectable.

2. Bone defects in anterior shoulder instability

2.1 Pathomorphology of bone loss in anterior shoulder instability

Understanding shoulder instability and its treatment is impossible without knowledge of the pathomorphological changes occurring in shoulder dislocation. The evolution of the latter into shoulder instability depends on many factors acting synchronously. Changes in the anatomical structures responsible for joint stability are directly related to the specific anatomy of the shoulder joint.

Main topic of this chapter is glenoid bone defect and because of that, soft tissue pathology will not be discussed.

Bone deficiency after sigle or multiple shoulder dislocations is considered to be important factor for reccurensy of instability after soft-tissue stabilization procedures [8, 9]. Thus, it has to be diagnosed and evaluated before any treatment. Multiple studies have reported reccurensy of instability due to glenoid bone defect ranging from 36% to 93.7%, with the rate directly proportional to the number of episodes of instability [10]. Patients with “minimal” glenoid loss (up to 13%) have reported poor functional outcomes after primary soft-tissue stabilization [11]. Similar data have been reported in Hill-Sachs defect, reaching even to 100%, in combine bone defects. Because of that, bone morphology and integrity are very important factors influencing outcomes after surgical treatment of shoulder instability [12].

The amount of bone loss and the recurrence rate often after soft tissue stabilization account for the preference for bony reconstructions in patients with significant bone loss [13, 14]. The frequency of bone defects varies according to literature data, depending on the evaluation methodology used. In patients with anterior instability, the frequency of glenoid bone loss is between 49 and 86%, and of humeral defects - even up to 100% [15].

Bipolar bone defects are almost obligatory finding in patients with anterior glenohumeral instability [13, 14]. They lead to instability because, altering the congruence of the glenohumeral joint and the function of the static stabilizers, like glenoid surface, congruence, etc. In physically active patients, their frequency is particularly high, because of the increased activity, age and potential for injuries of the respective discipline. Contact sports (rugby, martial arts, wrestling, bars, etc.) and overhead sports (basketball, volleyball, handball, etc.) can lead to high-energy traumatic dislocations and are usually accompanied by significant bone defects [16]. In contrast, the rate of bone loss is much less in non-contact sports (track and field), where chondrolabral and pure labral injuries predominate. Other situations associated with significant glenoid bone loss are post-convulsive conditions and bike/motorcycle trauma [17].

Glenoid bone loss is among the causes of glenohumeral instability and one of the most important factors responsible for failure and recurrence after arthroscopic stabilization [12]. A high percentage of patients with chronic anterior glenohumeral instability have some type of glenoid bone loss. Recognizing bone loss and determining its degree is of great importance, with planning individualized treatment algorithm. Detecting chronic bone loss is important for several reasons. Unlike an acute Bankart lesion, a chronic one may be associated with erosive glenoid wear and chronic medialization of the anterior-inferior capsuloligamentous complex. Humeral bone defects also multuply frequency of instability [15].

Patients with more than three dislocations or with persistence of more than 4 hours in a dislocated state are suspected of significant posttraumatic bone loss [16]. A history of frequent recurrent dislocations, especially during daily activities, is suspicious for a bony defect. It is recommended, that bone loss be differentiated as acute or chronic. This is done with a careful history and X-ray examination. Although acute large bony Bankart lesions cause clinically significant bone loss, they can be treated successfully with open or arthroscopic techniques of internal fixation. Bigliani et al. reported good results after anatomic repositioning and fixation of glenoid rim fragments in accute setting. Thus, the resistance of the glenoid against axial load and, respectively, the arch of movement is restored. Missing that moment, bone loss is tramsfering from acute to chronic and the function of the shoulder is reversibly affected by significant bone loss.

Bony defects of the anterior glenoid are creating mismatch between the articular surfaces, leading to increased anterior translation of the humeral head with minimal force [5]. The difference in the length of the articular surfaces has a negative effect on the ability of the glenoid to resist axial loading. Bone deficiency also decreases the concavity of the glenoid fossa, increasing the inability to hold the humeral head in translation [5]. In the presence of bipolar bone defects, the instability is further aggravated.

In 1998, Bigliani et al. [17] were the first to introduce the term bony Bankart when defining an anteroinferior bony fragment of the glenoid with a preserved insertion of the inferior glenohumeral ligament (Biliani type I). Biliani type II is a fragment with poor consolidation and firm connection to the ligamentous complex. The third type of defect is the erosive one with two subtypes: III A - with defect size below 25% and III B - above 25% of the total glenoid surface.

Radiologically, chronic unstable glenohumeral joints, have about 90% bone loss [18]. Based on arthroscopic and CT findings, Sugaya et al. found that, only 10% have intact glenohumeral joints, 40% have erosive-compression defects of the glenoid and 50% - with bony- Bankart [14]. Itoi et al. stated that a lesion of 21% of the glenoid length is reducing the stability provided by the glenoid by 50%, corresponding to an 18% deficit on West Point radiographic projection and a 50% loss of glenoid depth determined by CT [19]. In a cadaver study, autors defined that the position of the defect was usually at 4:30 o’clock (in right joints) and was more anterior than anteroinferior [20]. Bone defects greater than 21% of the glenoid width significantly reduce stability, compromise Bankart reconstruction, and therefore are indicated for bone augmentation [21]. The localization of this “critical” defect is predominantly anterior rather than anteroinferior [20]. More recent studies have shown a significant decrease in stability after defect formation beyond 26% of the glenoid width (20% of the length) [21]. In addition, stability after soft tissue reconstruction (Bankart) is significantly reduced in the presence of such a defect (Figure 1) [21].

Figure 1.

Typical location of the glenoid defect in anterior shoulder instability and creating “ inverted pear” glenoid.

Bony defects between 0 and 15% of the glenoid width (< 3–4 mm) are usually imperceptible to regular patients (excluding high-level athletes), while defects greater than 20% (>6-10 mm) are clinically significant.

The concept of the “inverted pear” proposed by Burkhart [5] provides an explanation of what is clinically significant bone loss. This is when a large bony Bankart lesion turns the normal shape of the glenoid (pear-shaped, with a larger transverse dimension at its bottom) into a so-called “inverted pear” (Figure 1). This form is recognizable either arthroscopically or with imaging studies (CT, MRI).

The geometric center of the inferior glenoid it is defined as “bearing point” and this is the anatomical landmark for accurate arthroscopic determination of the radius of the glenoid. This is done through an anterior superior-lateral arthroscopic portal, with the introduction of a calibrated probe through the posterior portal. Burkhart and De Beer [5] used the “bearing point” as a reference point for arthroscopic measurement of glenoid loss. Sugaya et al. [14] demonstrated that the inferior part of the glenoid is almost circular in shape (about 24 mm in diameter) and determined the size of the defect along the missing portion of this circle. These two methods determine the degree of bone loss without information about the native shape of the glenoid in question.

Anatomical studies of the glenoid are extensive, with two of them referring to its inferior part [22]. De Wilde et al. confirm again the round shape of this part of the glenoid, the center of which is Asaki’s tubercle - the underlying bone of the so-called “bearing point”/“bare spot” [23].

Aigner et al. [24] measured the distances from this point to the cartilaginous edge of the glenoid and found that most of the “bearing spots” were eccentrically located and therefore not a reliable guide in determining the amount of bone loss. Conversely, Burkhart et al. [5] define precisely the “bare spot” for the geometric center of the inferior glenoid. The measurements of Aigner et al. differ in millimeters, which makes the differences negligible and confirms the reliability of the concept of the “bare spot”. Huysmans et al. explain the differences in the lengths of the distance from the center to the glenoid rim with the triangular shape of the labrum and the uneven density of the underlying bone. Therefore, when the distance from the “bare spot” to the posterior labrum, in patients with a detached anterior labrum, is equal to the distance from the “bare spot” to the anterior glenoid rim, a small bone defect is already present [25]. It is for this reason that the concept of the “bare spot” can be used to determine the degree of glenoid loss, especially in combination with CT or MRI.

When discussing bone deficiency in anterior shoulder instability it is obligatory to include defects of the humeral head, because both defects interact dynamically in shoulder motion. Thus, when treating glenoid deficiency, we need to consider humeral head defects either.

The Hill-Sachs lesion is an impression fracture of the posterosuperior, lateral part of the humeral head, first described by Malgen in 1855 and studied radiographically in detail by Hill and Sachs in 1940. It can occur with any dislocation because the impaction of the soft humeral head on the hard bone of the glenoid (Figure 2).

Figure 2.

Hill-Sachs defect.

The frequency of the defect is between 65% and 67% after a primary dislocation and 84–93% after a recurrent one. This frequency is actually highly dependent on radiographic projections. On anteroposterior views, it is reported in only 7% of primary dislocations, but on MRI it is 93% [26]. The frequency of lesions in young patients with traumatic dislocations in several arthroscopic series ranges between 47% and 100% [14]. Hill-Sachs defects are graded depending on the percentage of involvement of the humeral head: less than 20% (small defects); between 20% and 45% (medium) and above 45% (major defects). The critical threshold of the humeral defect is still a subject of debate - 4 cm in length, 20–25% of the surface of the humeral head and about 250–1000 mm3 volume of the defect [27].

According to Saito, the defect is located between 0 mm and 24 mm from the tip of the humeral head. Spatschil presents a relationship between the number of dislocations and the frequency of defects. In patients with only one dislocation, the incidence of Hill-Sachs defects is around 67%, whereas in patients with more than one dislocation, the incidence increase up to 84% [4]. That’s why there is direct proportional dependence between the number of dislocations and the size of the defect.

Accurate assessment of the size of the Hill-Sachs lesion and its interaction with glenoid bone loss improved with Itoi’s introduction of the glenoid track concept [21]. Itoi [28] was the first to determine the “critical size” of the glenoid biomechanically in cadavers. With a fixed scapula, movement of the humeral head relative to the glenoid is reproduced in a position of elevation from 0° to 60°, which corresponds to a position of 90° to the body and maximum external rotation. “Glenoid track” represents the contact area between the humeral head and the glenoid fossa in the various phases of motion—the varying degrees of elevation at maximum external rotation and maximum extension. This end position is critical for instability due to tension on the capsuloligamentous complex and prevention against anterior translation of the humeral head. It is in this position that patients feel a positive apprehension test. During the movement from inferomedial to superolateral, an “imprint” is formed on the humeral head from the glenoid fossa - the so-called “glenoid track”. As long as the Hill-Sachs defect is located within the limits of glenoid track, there is no risk that the anterior edge of the glenoid will “engage” the defect and result in a lever mechanism, resp. dislocation. It is classified as an “on track” lesion. However, if the above mentioned defect exceeds the size of glenoid track, there is a risk of “engagement” with the anterior edge of the glenoid and resulting in instability. This type of pathology is classified as an “off track” lesion.

The distance from the medial edge of the contact zone to the medial edge of the rotator cuff is approximately 18 mm, or 84% of the glenoid width at 60° abduction [21].

Itoi et al. [29] measured glenoid track also in vivo by MRI. The position of the hand was put in seven different degrees of elevation with maximal external rotation of the limb maintained. Computer analysis showed that localization of the medial edge of the glenoid track was 93%, 85%, 82%, and 82% at 60°, 90°, 120°, and 150° elevation, respectively. Values at 90° elevation are closest to those measured on cadavers. Burkhart and De Beer’s [5] concept of an engaging/non-engaging defect is fully consistent with Itoi’s [21] concept of the “glenoid track”. Together, both concepts explain and evaluate the interaction of bipolar bone defects on dynamic shoulder function. Because ot that, these two concepts are “milestones” of understanding bone deficiency in shoulder instability.

Assessment of bone loss by 3D CT is an essential tool when evaluating patient with instability. There are numerous biomechanical studies on the mechanisms of bipolar bone defects. However, there is a lack of data demonstrating the pathomorphological evolution of these defects in a specific patient.

Nakagawa performed such a study on 204 shoulders evaluated by CT preoperatively, postoperatively, and postrecurrence. The size of the defect is undoubtedly a determinant, but the development of these defects over time also has therapeutic importance for predicting instability. The author reported a mean glenoid defect value of 2.2% with a fragmentary type of bone loss in half of the cases. The high frequency of Hill-Sachs defect (66% of cases) against such a small glenoid defect is cofusing. The overall incidence of bipolar bone loss is approximately 30%. This is somewhat at odds with other studies, that postulate a predominantly glenoid defect and much less Hill-Sachs lesions. Regarding the influence of the degree of dislocation/subluxation, Nakagawa reported a significant difference in the presence and size of Hill-Sachs lesions in dislocated and subluxated joints. This is not absolute for glenoid bone loss. CT findings after a first episode of instability show a logical 100% increase in the amount of bone loss measured (in the study from 0–10% to 6–17%). Regarding the occurrence of Hill-Sachs lesions, no significant difference was found, although the size of the loss increased.

Multiple studies based on 3D CT have shown a directly proportional relationship between the size of bipolar bone defects, age, and the number of dislocations [30]. The frequency of occurrence of these defects increases with increasing episodes of instability. Nakagawa studied bipolar bone loss in 153 shoulders and reported a population prevalence of over 56%. Including the isolated defects, the total frequency of bone defects becomes more than 85%. On average, the presence of bipolar bone loss in patients with a primary episode of instability is 34%, while in a recurrent episode it is already 62% and increases proportionally with the number of episodes [30].

The question arises as to the timing of the stabilization procedure. Is such an intervention necessary immediately after the first dislocation, when the defect is CT evaluated, or to wait for the first epizode of recurrence and the formation of already “critical” bone loss, preventing further enlargement of the defect and increasing the risk of instability after already done stabilization. Large Hill-Sachs defects are also more common after instability recurrence. Nakagawa reported that 12% of male professional athletes with recurrent instability had off-track Hill-Sachs [31]. This suggests that these defects develop from several episodes of instability, rather than a single one.

When we define glenoid bone loss as “fragmentary” or “erosive” type, it should be kept in mind that very often small bone fragments undergo resorption and thus loss after recurrence is defined as “erosive” type. In Nakagawa’s studies, a bone fragment was not always differentiated in the first dislocation, and the anterior glenoid rim often showed signs of a compression fracture without a bone fragment present [32]. This is may be because of the different timing in appearance of the patient. That’s why it is very important, patients with shoulder dislocations to be closely followed-up and evaluated for glenoid deficiency.

Randelli [33] did an extensive analysis of risk factors for instability, particularly after arthroscopic soft-tissue stabilization. The age of the patient is the most common reason for recurrence. Many autors reported a significant negative correlation between postoperative instability and age at first operation: 25% in patients younger than 20 years; 20% at 20 years - 30 years; 7% in patients between 30 and 40 years. Taking the 20th year as borderline, the frequency of recurrence below this age is 33.6%, while above it - 11.8%. Porcellini [34] reported a 10% risk of recurrence in males, compared to 2.8% in females. The overall recurrence rate in men is about 15%, and in women 8.7%. The number of dislocations is also an important prognostic marker for the risk of recurrence. According to Imhoff [35] the risk is directly proportional to the number of preoperative dislocations, reaching a 44% with more than 10 dislocations. Other series give similar data at baseline of 5 dislocations [36]. There is no significant relationship between the type and number of anchors used in arthroscopic stabilization and the risk of recurrence [37].

2.2 Imaging and quantification of glenoid defects

Accurate identification and quantification of both glenoid bone loss and the presence of a Hill-Sachs defect is essential in the diagnosis and therapeutic approach of anterior shoulder instability, as it is directly related to functional treatment outcomes. The ability to correctly diagnose and treat these potential bony defects of the glenohumeral joint is based on multiple factors—the patient’s history, physical findings, and the various imaging modalities in a clinical setting that are fundamental to a successful outcome. There are multiple methods for quantifying bone loss based on imaging or arthroscopy, and anyone involved in this pathology should consider all bone loss imaging options, their specificity, and sensitivity.

Standard radiographs are the initial imaging of the patient with shoulder instability. While some studies recommend standard radiographs, diagnostic sensitivity, specificity, and accuracy can be greatly influenced by patient positioning during the examination.

In studies reporting the accuracy and reliability of specific radiographic projections, the axillary view, West Point View, true AP view, and Bernageau projection are those that would be useful. Of these, the Bernageau view is reported to be the most accurate and reliable when compared with 3D CT images (Figure 3) [19, 38]. A disadvantage is the ignorance of the projection itself by the X-ray laboratory technicians and the poor imaging of the lower bony lesions of the glenoid face [13]. Edwards et al. [13] propose a classification of glenoid loss in this projection into 3 groups: fracture, “cliff” and “blunt angle”, but they do not determine its size. Bernageau view represents radiographic projection of the glenoid, which is compared to the healthy contralateral side and serves to calculate glenoid bone loss. This technique is easy to apply and has a sensitivity and specificity of over 90% [13]. With the developing of 3D-CT and MRI, the calculation of bone loss has become more accurate, and many authors have proposed protocols for its determination [39]. Due to popularity of CT imageing, Bernageau view it is not standard procedure, but it should be kept in concideration.

Figure 3.

Bernageau view. Uninjured side (left) and unstable side (wright).

In general, standard radiographs are unreliable relative to modalities such as 3D CT in quantifying glenoid deficiency, which may affect the therapeutic algorithm. Therefore, it is not recommended that, standard radiographs to be used as the sole method for diagnosing glenoid bone loss.

Several methodologies exist in the literature to quantify the extent of glenoid bone loss. These techniques are based on surface area measurements or based on glenoid diameter. The general concept of measuring glenoid bone loss is the understanding that, the inferior glenoid in the “en face” view resembles a true circle and that the degree of bone loss can be calculated using the geometric properties of a circle or by comparing the ratio of measurements in the healthy, contralateral shoulder [40]. Although there is a lack of consensus and heterogeneity in the literature on the method of measurement, seventeen different methods are available in practice [41].

All methods calculate the ratio of glenoid surface and missing bone on 2D images. Thirteen methods were based on unilateral glenoid measurement and four were based on bilateral measurement. Sixteen methods performed measurements using the glenoid surface (sagittal “en face” view of the glenoid) and one method used the glenoid rim (sagittal view). Eleven of the methods used a reference distance that was divided, 5 methods calculated the defect that was divided by a reference surface, and 1 method used the position of the defect to quantify glenoid bone loss [41].

All methodologies have similar characteristics focused over the glenoid surface in “en face “view. One of the most popular and reliable method is the “ideal circle” (Figure 4).

Figure 4.

Method of the “ideal” circle - bare spot method, one-sided. On a sagittal projection, the longitudinal axis from the supraglenoid tubercle to the inferior glenoid rim is determined. It is drawn perpendicularly through the widest part of the inferior glenoid. The intersection of the two lines is the so-called “bare spot” and it is the center of a circle. Bone loss was defined as the ratio of anterior (AD) to posterior distance (PD) using the formula [(PD-AD)/(2xPD)] x 100 [41].

Other reproducible technique is “diameter” method (Figure 5).

Figure 5.

Method of the “ideal” circle (diameter method), one-sided. On a sagittal projection, the transverse axis of the glenoid is drawn and a circle is placed that most accurately fits between 3′ and 9′ o’clock of the inferior glenoid. A line perpendicular to the transverse axis is drawn until it reaches the bone defect (A). The diameter of the circle (B) is also determined. Bone loss is calculated using the formula (B-A/B) x 100 [41].

2.2.1 Surface-based methods

The amount of bone loss of the glenoid face is estimated by surface area methodology using so called “ideal” circle, which is not covered by best-fitting circle Sugaya et al. promote a “circle method” for determining bone loss, which was calculated by dividing the area not filled by the glenoid by the surface area of the best-fit circle [40]. Baudi upgrade this technique to method known as the PICO: calculation of the area of the glenoid loss by the area of the healthy contralateral surface of the glenoid, defined by two identical best fit circles [42]. The PICO method has been reported to be highly accurate and reliable in assessing bone loss and has also been shown to be clinically more useable than linear techniques when performed on 3D CT images [42, 43]. PICO technique needs bilateral CT, and some studies reported equal result to MRI, considered as more favorable in clinical setting [44]. It should be noted that calculations using the surface method require more sophisticated software and are more complex than linear methods.

Surface-based measurements are considered as reliable and precise in terms of glenoid bone loss.

2.2.2 Diameter (linear-based) techniques

They are convenient, because they do not require sophisticated image processing software and can be used instantly in clinical practice [45]. These methods, even can be used intraoperatively using arthroscopically assisted measurements. Several variants of linearly based models have been described in the literature. Estimation of bone loss can be done using the contralateral arm or using the “ideal” circle method. The first option can be useful, as no major differences between the two arms have been reported in the morphology of the glenoid region. This does not apply to cases with bilateral involvement [25]. In comparison, the “ideal” circle methodology, with an emphasis on absent bone loss of width, is applied only to the pathological shoulder by estimating the width of the glenoid face and overlapping with the “ideal circle” [45].

New studies report significant reliability of the glenoid index (SI/AP) for assessment of glenoid bone loss. This method is based on the ratio of the measured widths of the pathological and healthy glenoid [16]. Some autors did not find statistically significant differences between the calculations with this methodology and the measurement technique based on surface area in patients with bone loss greater than 6%. They noted that, although not significantly, when the rate of bone loss increased, the differences in measurements also increased [45]. The position of the bone defect affects accuracy of quantifying bone loss in linear measurements, due to presentation only in frontal plane. Bone loss in lower glenoid can be underestimated if defect is located oblique to the frontal plane of the glenoid [46].

Diameter-based measurements can overestimate glenoid bone loss [45]. The “perfect” circle technique has been argued to be inaccuracy, due to an incorrect geometric formula, that would more accurately be applied to calculate the area of a square rather than a circle [47].

In some series, these inaccuracies ranged between 15 and 25%. Because of that, unnecessary bone augmentation may occur due to an exaggeration of bone loss that exceeds the critical threshold of 20–21% for recurrent instability [47]. CT imaging is the most extensively studied modality for assessing bone loss in shoulder instability [46]. Although CT has some well known disadvantages as increased ionizing radiation than other modalities, few studies support its replacement [14, 16]. Many studies show superiority of the CT to other modalities. This often “enforces” 3D CT methodology as the “gold” standard [48]. 3D CT has almost perfect specificity in estimating bone loss [14, 49].

3D CT is reported to be a more accurate and reliable modality than two dimension techniques [48]. The “en face” view of the glenoid fossa, provided by 3D, gives more precise estimation, using surface measurement methods. In contrast, 2D CT images are affected by scapular tilt, which alters the “en face” projection of the glenoid and affects the accuracy of the calculations [48].

Any inclination of the en-face view results in a change of the measured surface and the diameter of the “ideal” circle. Although superior, inferior, and posterior scapular tilts tend to reduce the actual defect, anterior tilts have no such relationship, even the opposite. This leads to underestimation in the first case and overestimation of the defect in the second case. Vertical tilt, in particular, has a significant effect even at 5° malrotation. From a clinical point of view, these changes are not always significant. Naturally, changes in defect measurement are directly proportional to scapular inclination [48].

2.2.3 Arthroscopic evaluation

The term “bare spot” as anatomical center of the glenoid was introduced by Burkhart [50]. This is done during shoulder arthroscopy and it relies to the fact that posterior rim of the glenoid face is intact and can be used as a reverence point measuring the center of the glenoid fossa. Using calibrated probe, author has validated the accuracy of the measurement, compared to 3D CT [51, 52].

However, some studies have shown that, this point could be inconsistent and even absent in some patients up to 48% [53]. Other authors noted that, “bare spot” can be eccentrically located, thus leading to overestimation of the glenoid bone loss. Because of that literature does not recommend to use this method as a single tool.

There is controversy in the literature what is the ideal method for assessing the bone deficiency in glenohumeral instability. In a recent systematic review, Walter et al. compared the efficacy of different imaging methods for quantifying glenoid bone loss. A review of 1425 shoulders concluded that CT (2D/3D) and MRI provide accurate measurement of bone loss in shoulder instability. There is no complete consensus regarding the so-called “gold standard” ac its determination [41, 54].

Regarding the best way to quantify glenoid bone loss, consensus of leading experts, stated that “en face” view of the glenoid face using 3D CT provides the most accurate assessment of the glenoid bone defect. Regarding methods for calculation the size of the defect, 86% of the experts concluded that all the different options available, including the surface area method, the superimposed circle method, the PICO method, etc. are sufficiently reliable [12, 41].

2.2.4 ISIS (instability severity index score) total

In a prospective study, Balg and Boileau proposed the ISIS-score as a way to predict outcomes after arthroscopic stabilization of anterior shoulder instability [55]. In a 10-point preoperative scoring system, risk factors leading to an increased risk of recurrence (also after arthroscopic Bankart) are: patient age less than 20 years at the time of surgery; professional or contact sports; hyperlaxity; Hill-Sachs lesion visible on anteroposterior view of the shoulder ac external rotation and/or loss of the sclerotic contour of the inferior glenoid (Table 2).

Age	under 20 yrs. – 2 pts	above 20 yrs. – 0 pts
Sport	Professional- 2 pts	Recreative- 0 pts
Type of sport	Martial / over-head - 1 pts.	Other - 0 pts
Laxity	Hyperlaxity −1 pts	Normal −0 pts
HS defect	positive in ER - 2 pts	Negative in ER – 0 pts
Glenoid	Loss of anteroinferior contour - 2 pts.	No loss – 0 pts

Table 2.

ISIS score.

Since the introduction of the ISIS-score as a predictor of recurrence, multiple studies to validate this score have shown inconclusive data and a lack of consensus. A total ISIS score of less than 4 points is significantly more accurate, than the defined base-limit of 6 points [44, 56].

Dekker performed an extensive review of 217 patients over a 5-year period with arthroscopic stabilization to determine the reliability of the instability severity index score (ISIS) for assessing the risk of recurrent instability. The analysis of patients is based on the main points of the ISIS-score:

1. number of instability episodes,

2. general period of instability,

3. presence and degree of glenoid bone loss,

4. presence and degree of Hill-Sachs defect.

The evaluation systems used in the study are Western Ontario Shoulder Instability Index (WOSI), Single Assessment Numerical Evaluation (SANE) and American Shoulder and Elbow Society scores (ASES). The results of the study indicated no correlation between treatment outcomes and the measured ISIS total. The mean glenoid loss measured in the series was less than 14.5%, which was expected to be lower in patients with good functional outcome, mean Hill-Sachs defect volume greater than 1.3 cm3, and duration of symptoms greater than 3 months. This led the study authors to recommend a revision of the ISIS score, incorporating factors that have increasing weight in deciding on a therapeutic approach. Of particular importance are the degree of glenoid loss and the volume of the Hill-Sachs defect. Loppini did a large controlled 5-year study showing the protective role of a fixed baseline of 3 points, with success rates on this basis of 93.7% [56].

Accurate identification and quantification of both glenoid surface loss and a potential Hill-Sachs lesion is essential in the treatment of anterior shoulder instability, as it directly affects patients outcomes. In a so called first-time dislocator or patient with a history of multiple subluxations, the suspected bony defect should be evaluated comprehensively including soft tissues either.

There are multiple methods for quantifying bone loss, and the surgeon must be aware of both the advantages and disadvantages of each imaging modality when it comes to the accuracy of identifying glenoid bone loss and accurately quantifying the size of bony lesions.

3. Operative treatment for shoulder instability with bone deficiency

Surgical treatment of bone loss in anterior shoulder instability aims to restore the deficit and achieve joint stability. There are several basic operative techniques that can be summarized as: “bone pathology - bone surgery”. Once the degree of bone loss has been precisely determined, the method of bone augmentation is a matter of choice for the surgeon, depending on his capabilities and available technical support, as well as the patient’s profile.

Treatment of anterior shoulder instability is generally based on two types of operative techniques: anatomic and nonanatomic reconstruction. The anatomical reconstruction aims to restore the glenoid labrum to its anatomical position and achieve optimal tension of the capsulolabral complex. These techniques currently are performed arthroscopically.

Nonanatomic techniques are stabilizing the joint through bone augmentation and increasing the glenoid articular surface, preventing of excessive humeral translation, and stabilization of the joint through isolation of the humeral bone defect. The current understanding of bone loss in shoulder instability, recommends operative techniques to be addressed to the “bipolar bone loss” and not to each one separately. This is because, both bone defects interact dynamically and it would not be appropriate to isolate the treatment for each of the types. This is the reason why this chapter examines the individual operative techniques and their effect on the bone damage of the glenoid-humeral head pair.

3.1 Coracoid transfer: Laterjet

Coracoid transfer - Laterjet was first described in 1958 in the treatment of anterior shoulder instability and involves trasfering the coracoid process over the formed glenoid defect along with its adjacent musculature. The main effects achieved with this technique are:

1. bone replacement of the existing defect and increase of the contact surface of the glenoid with the humeral head. This is how “isolation” of the humeral defect is achieved and its engaging is prevented (“glenoid track concept “);

2. the movement of the coracoid muscles over the lower edge of the subscapularis muscle causes the muscle to “stretch” and thus to stabilize the joint;

3. the so-called “sling effect” - the tendon of the coracoid muscles achieves a “retaining” effect at 90° abduction and 90° external rotation, providing additional stability of the joint by preventing the engaging of the humeral bone defect at this “most unstable” position.

Gradually over time, two main variants of coracoid transfer have been established: open and arthroscopic techniques. Open techniques have been popularized due to consistent postoperative results, relatively less need for surgical expertise, and predictable postoperative complications. The advantages of the arthroscopic technique are less surgical trauma, better visualization and the ability to address concomitant soft tissue injuries. The most common disadvantages are the need for a “learning curve” for the surgeon and more sophisticated technical support.

A number of authors recommend the following treatment algorithm for bipolar bone defects. With glenoid loss ≤13% and the presence of a small “on track” Hill–Sachs, arthroscopic stabilization is recommended. Glenoid loss ≤13% and medium-sized “off track” Hill–Sachs is an indication for arthroscopic stabilization and remplissage. Surgeons experienced in arthroscopy advocate the recommendation for arthroscopic variant of Laterjet transfer. With glenoid loss over 13% and a small “on track” Hill–Sachs – arthroscopic Laterjet technique. The combination of glenoid loss greater than 13% and moderate/large “off track” Hill–Sachs lesions are indications for an open Laterjet technique or free bone graft and replissage [33, 57].

The indications for arthroscopic Laterjet stabilization are:

• Isolated anterior glenoid loss >13%

• Anterior glenoid loss <13% associated with a Bankart lesion

• Anterior glenoid loss >10% and < 20% with an ISIS score of 3–6 points

• First episode of shoulder dislocation no more than 3 years ago

• No more than 5 dislocations

Nonanatomic bone techniques are the subject of extensive debate in the literature. A systematic review of 46 series involving 3211 shoulders showed low rates of recurrence of instability after bone augmentation compared with arthroscopic Bankart [58]. A biomechanical study of the effectiveness of variations of coracoid transfer (Bristow or Laterjet) demonstrated the superiority of Laterjet in restoring stability and reducing the risk of recurrence [59].

The “congruent-arc” modification, which rotates the graft by 90°, has several advantages over the classical one: it restores the radius of curvature of the glenoid face, thus reducing the superior translation of the head and allowing compensation of larger defects. Similar comparisons have been made between the remplisage technique and Laterjet in engaging Hill-Sachs defects.

Ali compared the overall functional outcome after open and arthroscopic Latarjet stabilization by assessing range of motion, strength, and clinical outcomes using the Rowe, WOSI, and visual analog scale (VAS) scoring systems. The author makes an additional assessment radiographically. The most important finding of this study is that both open and arthroscopic techniques produce similar overall functional and radiographic results [60].

In a comparative meta-analysis, Hurley found no significant difference in functional outcomes and complications between arthroscopic and open coracoid transfer (Latarjet) [60]. Data from the systematic review show the same parameters for both techniques: low recurrence and complication rates. Given the greater technical complexity, the arthroscopic version of the Latarjet has been shown to be safe and suitable as an alternative to the open one. It is important to emphasize the significant learning curve for the use of the arthroscopic Laterjet, which suggests its use in large centers with experienced arthroscopists available. With the open technique, the author found a lower frequency of persistent apprehension test, which he explained by the higher frequency of patients returning to sports, especially contact one. Potential benefits of the arthroscopic technique for coracoid transfer lie in less postoperative pain. However, this does not always reflect in less use of analgetics, especially with an opioid component [61, 62].

The financial aspect of the arthroscopic technique continues to be a consideration. Randelli’s analysis shows a 3 times higher cost of the arthroscopic versus the open technique, which is an obstacle to the increasing popularity of this technique even in the USA [63]. In both techniques, the accurate positioning of the coracoid process is considered one of the critical steps of the Latarjet technique. Medialized offset of the graft can lead to permanent instability, while lateralization of the graft potentially speed-up degenerative changes,because of increased contact with the humeral head [64, 65]. Optimal graft position is below the glenoid equator, neither too medial nor too lateral, less than 10 mm from the articular cartilage. Visualization of the glenoid for positioning may be more difficult with the open technique, however, many of the technical steps are more easily accomplished with an open exposure. The advantage of the arthroscopic technique is the visualization of the entire glenoid articular surface, which reduces the risk of graft malposition [66].

Graft resorption is one of the most common complications after coracoid transfer and is reported as high as 63.9% in the literature. This consequence has a controversial clinical significance for recurrence of instability and appears to be greater in patients with greater glenoid bone loss and after arthroscopic technique [67]. Conversely, graft resorption in some series is seen more often after open reconstruction compared to arthroscopic. A significant correlation was found between the rate of graft resorption and the positive apprehension test postoperatively, but it did not correlate with the other functional results [68].

3.2 Free tricortical iliac graft

In case of significant glenoid bone loss, one of the preferred stabilization techniques is the open or arthroscopic Eden-Hybinette. This procedure is done with a tricortical iliac graft and has been shown to produce good results, with low rates of recurrent instability. Indications for this technique include severe glenoid loss (> 40%), recurrent instability after Latarjet or distal tibial allotransplantation, patients with abnormal coracoid morphology [69, 70].

Eden-Hybinette offers the possibility of tailoring the size of the bone augment to match the corresponding glenoid bone loss. Furthermore, this technique exculdes the risk of potential allogeneic contamination and is less expensive, despite the obvious disadvantage of donor site morbidity. Arthroscopic reconstruction with a tricortical iliac graft was introduced by Sugaya. Soon after, other authors such as Taverna and Kraus reported excellent short-term results with this technique [71]. Both versions of the Eden-Hybinette technique – arthroscopic or open technique – are equally effective [72].

A study by Malahias on different techniques for processing and fixation of a tricortical iliac graft showed that, regardless of the type of fixation and shaping of the graft, this technique gave excellent functional results and low rates (6%) of reoperation [73]. In addition, all important aspects of shoulder stabilization techniques such as: recurrences (4.8%), positive apprehension test (4.8%), nonunion (2.2%), osteolysis (0.4%), etc. are very low. Based on the meta-analysis, the author recommends the technique as safe and predictable in terms of functional results and complications. Bone augmentation of the glenoid with a tricortical graft did not differ with respect to the source – allograft or autograft. Although the iliac crest is a reliable source of bone, this technique is associated with developing degenerative changes at the recipient site and sometimes pain at the donor site [74]. Allogeneic grafts are designed to reduce these disadvantages [75]. Although the concern of early resorption and inadequate osseointegration of the allograft arises, clinical data show excellent functional results, low recurrence rates, and high osseointegration.

Ernstbrunner analyzed the position of a free tricortical graft in glenoid defect reconstruction, comparing arthroscopic and open surgical technique in two groups of 20 patients. In the arthroscopic group, a steeper angle of impact of the graft and a significantly greater medial offset of the graft were found compared to the open technique. The advantages of the arthroscopic technique are not directly related to the position of the graft and the reconstruction of the glenoid defect. They are rather the general advantages of arthroscopy: lower risk of infection, faster rehabilitation, cosmetics and last but not least - less trauma to the subscapularis muscle. This technique prevents the hypotrophy and fatty degeneration seen with the open technique [76].

A comparative detailed analysis by Moroder on the most used techniques for bone augmentation, namely coracoid transfer (Latarjet) and free tricortical iliac graft (Eden-Hybinette/ (ICBGT)) shows the lack of a significant advantage of each of the mentioned techniques [77]. At one-year follow-up, no statistically significant difference was observed in the functional scores - WOSI, SSV, ASOSS and Rowe. Regarding the range of motion (at 0° and 90° abduction), again no significant difference was reported except for internal rotation in favor of the free tricortical graft. A probable reason for this is the permanent structural splitting of the subscapularis muscle from the common tendon of the coracoid (the so-called conjoined tendon). Biomechanical studies of the sling effect with the Latarjet technique show a significant loss of total rotation at 90° abduction, compared with the free tricortical graft technique.

CT analysis shows another important feature of the free bone graft technique. This is the improved bony augmentation of the glenoid defect in the sense of increased total surface area, reduced volume of the defect, increased diameter and depth of the glenoid fossa. This is partly due to the purely anatomical limitation of the native coracoid process. Conversely, the size of the free graft is determined by the extent of the defect and the surgeon’s technical performance [77].

Another interesting fact from Moroder’s study is the absence of a statistically significant difference in the isokinetics of the patients after one of these techniques. The difference in this parameter was maintained in favor of the free tricortical graft within 1 year after surgery. At the end of the follow-up, the results were identical.

3.3 Tibial auto/allograft

Wong and Urquhart published an arthroscopic technique for glenoid augmentation using a distal tibial allograft that avoids damage to the subscapularis muscle and therefore allows for faster recovery [78]. The technique is based on the classic Bankart arthroscopic technique. Only one additional medial portal (Halifax portal) is required, which is done using an inside-out technique to avoid injury to the neuro-vascular structures. This is done through the posterior working portal, directed superiorly over the subscapularis muscle and laterally from the conjoined tendon of the coracoid - through the deltopectoral interval. Because the portal is inside-out, it is parallel to the glenoid surface and reproducible for safe passage of the graft.

Amar presents in his series the positives of arthroscopic reconstruction with tibial allograft and short-term CT results [79]. The technique appears to reduce musculocutaneous and axillary nerve injuries because the insertion of the subscapularis muscle remains intact. Placement of the additional Halifax portal using an inside-out technique allows identification of the aforementioned two structures, thereby protecting them from injury. With this technique, the Halifax portal was found to be an average of 4 cm from the above neural structures. In the Amar study, 18 patients (58%) had either less than 50% or more than 50% graft resorption. However, none of these patients with documented bone resorption reported residual shoulder instability [79]. Most importantly, even in the case of greater than 50% resorption, there was an increase in sagittal glenoid diameter of an average of 5.1 mm, which has been shown to be sufficient for clinical stability of the shoulder. This is probably due to the larger size of the graft that was used compared to the physiological size of the defect.

3.4 Graft remodeling

Coracoid transfer (Latarjet), free tricortical iliac graft (ICBGT), and distal tibial transfer are considered the most common operative techniques for bone augmentation [71, 77, 80].

The biomechanical effect of these techniques has been repeatedly demonstrated on cadavers [81, 82]. Over time, remodeling processes lead to changes in the articulating glenoid surface diameter, version, and depth after bony reconstruction. These processes vary between individual patients, depending on the degree of exercise, and are directly related to Wolff’s law [83].

Sigrist analyzed the effectiveness of free tricortical graft (ICBGT) in restoring the glenoid surface and bone remodeling after this technique - both autograft and allograft. According to the author, the degree of remodeling also affects the degree of stability (stability ratio/SR), and increased retroversion leads to increased stability (SR) [84]. Proportional to Wolf’s law, areas with less mechanical stimulation show greater bone resorption and remodeling. Conversely, at the positions where the bone is subjected to loading - 2′-4′ o’clock, there is a lack of bone resorption, increased glenoid diameter and depth [85]. This also supports the benefits of autologous bone transfer from a biomechanical and morphological perspective. In his study, Sigrist reported statistically greater residual bone defects with the use of allograft. Increased bone resorption in allografts has also been reported in other series. The likely cause of the excessive bone resorption is the lack of cellular recolonization and vascularization of the allograft. The use of fresh-frozen allografts instead of lyophilized ones improves the results of allograft bone augmentation. For example, distal tibial allografts show excellent resistance to bone resorption [86]. Autologous bone transfer (auto-ICBGT) increases the biomechanical stability of the shoulder not only immediately postoperatively, but also in the long-term aspect. Due to their rapid osteolysis, all allogeneic bone grafts suffer from the lack of improvement in stability and increase in bone parameters of the glenoid - width, depth, version, etc.

In 2020 Bois published a consensus of members of the three largest orthopedic associations in the United States regarding the approach and therapeutic options for bony defects in anterior shoulder instability [87]. With regard to known risk factors associated with critical bone loss, consensus among subspecialties has been reached that the total number of shoulder dislocations directly influences the magnitude of bone loss. In 2019, Dickens et al. conducted a study of athletes over a 4-year period to determine the amount of glenoid bone loss associated with first dislocation and recurrent instability. After a first dislocation, the mean loss of glenoid width was 6.8%, increasing to 22.8% after a second episode of instability [88]. In Gottschalk’s 2017 systematic review on studies reporting percentage loss of glenoid width, 23.6% of cases had glenoid width loss between 10% and 25% [89].

Furthermore, in a prospective multicenter cohort study, Rugg showed that first-time dislocator patients were less likely to have bone loss or biceps pathology and the treatment procedure of choice was arthroscopic capsulolabral repair. Recurrent dislocations are more likely to require an open Latarjet to address critical bone loss [90]. These studies support a link between recurrent instability and the creation of critical bone loss. Some autors demonstrated in a prospective cohort study that male patients <25 years of age had a 78% chance of recurrent instability within 2 years of injury, which increased to 85% within 5 years of the initial injury when patients were treated nonoperatively [90].

Regarding the treatment of combined critical bipolar bone loss and previous failed Bankart soft tissue stabilization, Latarjet transfer is unanimous. For revision caces in which glenoid bone loss is maintained constant, and humerus loss is corrected, again the recommendation is for Latarjet stabilization.

Although other bone augmentation techniques were indicated as potential options for study participants, Latarjet stabilization was the most commonly preferred bone procedure for correction of bone deficiency. In a recent systematic review (13 studies, 845 shoulders) analyzing long-term outcomes after Latarjet stabilization, Hurley found an overall high rate of return to sport (84.9%) and a low rate of recurrence of instability (8.5%) [91]. The return-to-sport and recurrent instability rates found by Hurley remain significantly better than those reported after arthroscopic Bankart [92].

The general understanding based on the consensus is that, the choice for bone augmentation in the presence of bipolar bone loss is the augmentation of the glenoid defect, at the expense of the size of the Hill-Sachs defect, which appears to be secondary [87].

The cut-off value for determining the optimal surgical treatment in the presence of bipolar bone loss remains controversial. In Rossi’s study, 90.1% of the experts agreed that in cases where the glenoid bone deficit was greater than 20%, the reconstruction of the glenoid bone loss should be with bone augmentation; in the presence of an acute bony Bankart lesion, fixation of the fragment can be performed arthroscopically using anchors [93].

Regarding graft type, the general statement of the panel is that any of the available options (coracoid, tricortical iliac graft, distal clavicle or distal tibia) is appropriate. However, in cases where the defect is very severe, the best option is to use a tricortical iliac graft. Some recent studies have reported that the critical cutoff value of glenoid bone loss should be lower than 20%. Shaha evaluated the clinical outcomes and recurrence rates among 72 patients with anterior shoulder instability divided into groups according to the size of the bony defects. It suggests that a glenoid bone defect greater than 13.5% will result in a clinically significant reduction in functional instability scores (WOSI), even in patients without postoperative recurrence [11].

4. Conclusion

Glenoid bone deficiency is major problem after shoulder dislocation and main reason for shoulder instability. Precise diagnosis and measurement ot this defect is essential for proper treatment. Bone augmetation techniques are treatment options for gainig stable joint and preventing reccurency. Open teshniques are still more popular because their simplicity, reliability and effectiveness. Atrhroscopic variants of those open techiques are becoming more popular, but still require technical support, surgeon experience and bigger expenses.