Engaging Hill-Sachs Lesion: Is There an Association Between This Lesion and Findings on MRI?

Musculoskeletal Imaging Original Research Gyftopoulos et al. MRI of Hill-Sachs Lesions Musculoskeletal Imaging Original Research Soterios Gyftopoulos 1 Avner Yemin Luis Beltran James Babb Jenny Bencardino Gyftopoulos S, Yemin A, Beltran L, Babb J, Bencardino J Keywords: anterior shoulder dislocation, engaging Hill-Sachs lesion, glenoid bone loss DOI:10.2214/AJR.12.10206 Received October 24, 2012; accepted after revision January 24, 2013. 1 All authors: Department of Radiology, NYU Langone Medical Center, 560 First Ave, New York, NY 10016. Address correspondence to S. Gyftopoulos (Soterios20@gmail.com) WEB This is a web exclusive article. AJR 2013; 201:W633 W638 0361 803X/13/2014 W633 American Roentgen Ray Society Engaging Hill-Sachs Lesion: Is There an Association Between This Lesion and Findings on MRI? OBJECTIVE. The objective of our study was to see whether there is an association between engagement on physical examination and the location or size of a Hill-Sachs lesion and the presence and degree of glenoid bone loss as assessed on MRI. MATERIALS AND METHODS. Thirty-three consecutive patients (32 males and one female) with a history of anterior shoulder dislocation who underwent preoperative MRI and arthroscopy at our institution and were tested for engagement on physical examination over a 9-month period were included in the study. Two blinded readers reviewed each study independently and documented the presence and size of the Hill-Sachs lesion, location of the Hill-Sachs lesion with a modified biceps angle, and presence and size of glenoid bone loss. Statistical analysis included the Mann-Whitney, logistic regression, Pearson correlation, and intraclass correlation tests. RESULTS. Eleven patients had evidence of an engaging Hill-Sachs lesion on physical examination and 22 did not. There was no statistically significant difference between any of the dimensions or overall area of the Hill-Sachs lesion when comparing the group with an engaging Hill-Sachs lesion and the group with a nonengaging lesion (surface area, 3.60 vs 3.23 cm 3, respectively; p = 0.272). There was a trend for a larger biceps angle in the engaging group without a statistically significant difference (mean, 154.5 vs 143.9 ; p = 0.069). There was a statistically significant difference in the amount of glenoid bone loss in the engaging group compared with the nonengaging group (mean, 20.2% vs 6.0%; p = 0.001). CONCLUSION. There is a significant association between an engaging Hill-Sachs lesion on physical examination and the degree of glenoid bone loss as well as a trend toward increased engagement with more medially oriented Hill-Sachs lesions. These findings show the importance of considering both the Hill-Sachs lesion and glenoid bone loss when evaluating patients with engagement. T he two most common osseous injuries in the setting of anterior shoulder dislocation are anterior glenoid bone loss and Hill-Sachs lesions [1, 2]. In terms of glenoid bone loss, several studies have discussed the different types, its role in recurrent dislocation, and the importance of accurate quantification for treatment [3 6]. The Hill-Sachs lesion, a posttraumatic impaction injury along the posterolateral aspect of the humeral head, has also been studied extensively to understand and define its significance [3, 7 13]. Although the diagnosis of a Hill-Sachs lesion can be consistently made on imaging, the importance of accurate characterization and the role of a Hill-Sachs lesion in predisposing to recurrent dislocation remain unclear. In several studies, investigators have attempted to define the role of the Hill-Sachs lesion in recurrent shoulder dislocation. Burkhart and De Beer [3] coined the term engaging Hill-Sachs to describe a lesion that predisposes to recurrent dislocation or to the symptoms of recurrent dislocation after Bankart repair [3]. Yamamoto et al. [11] found that the location of the Hill-Sachs lesion was the most important factor for predisposing to engagement and recurrent dislocation using cadavers and 3D CT reconstructions. Cho et al. [13] evaluated Hill-Sachs lesions on 3D CT reconstructions as well and showed that lesion size and orientation were the most important factors predisposing to engagement. The standard apprehension test is an important component of the physical examina- AJR:201, October 2013 W633

Gyftopoulos et al. were used for imaging. The patient s shoulder was placed in a neutral position with the thumb up for each examination. Appropriate positioning was confirmed during scanning and repositioning was done when necessary. Fig. 1 22-year-old man with history of recurrent shoulder dislocation and engagement on physical examination. A and B, Axial fat-suppressed proton density weighted (A) and sagittal T1-weighted (B) images show anteroposterior (line A, blue line, A), depth (line B, green line, A), and craniocaudal (red line, B) measurements of Hill-Sachs lesion. In A, dotted line shows humeral head. tion of the patient with a history of anterior shoulder dislocation [14, 15]. Although this tool can be used to detect an unstable joint, the root of this instability, such as osseous injuries like an engaging Hill-Sachs lesion, can be difficult to discern [16]. Engagement is also typically tested for during the physical examination component of arthroscopic surgery at our institution; however, having this information before surgery would be helpful, especially if it can be found on standard preoperative imaging such as MRI. Having this information would result in more complete surgical planning and allow the surgeon to better explain to the patient the extent of his or her injuries as well as the benefits and risks of the treatments available. The purpose of this study was to see whether there is an association between engagement on physical examination and the location or size of the Hill-Sachs lesion as well as an association with the presence and degree of glenoid bone loss as assessed on MRI. Materials and Methods Institutional review board approval was obtained, and informed consent was waived for the retrospective HIPAA-compliant study. Patients We conducted a retrospective review of 60 consecutive patients who underwent MRI over a 9-month period. The patients were chosen irrespective of age, sex, and ethnicity. The inclusion criteria were the following: Patients had to have a history of anterior shoulder dislocation, to have undergone preoperative MRI (1.5 or 3 T) at our institution, to have undergone arthroscopic surgery performed by one of three fellowship-trained sports medicine orthopedic surgeons, and to have documentation of testing for engagement of the humeral head onto the glenoid in the operative reports. The exclusion criteria were no stated history of shoulder dislocation, MRI not performed at our institution, and a lack of documentation for engagement testing in the operative report. This search yielded 33 patients (32 males, one female; age range, 16 59 years; mean age, 29.8 years) who were included in the study. The physical examination was conducted while the patient was under anesthesia before arthroscopic surgery and consisted of placing the shoulder in 90 of abduction and a range of external rotation between 0 and 135 to test for a catching or popping sensation that would indicate engagement. MRI Technique Thirteen of the MR studies were conducted on a 1.5-T scanner and 20 MR examinations were performed using a 3-T scanner. The sequences for the unenhanced 1.5- and 3-T MR examinations included the following: coronal oblique turbo spin-echo proton density weighted (slice thickness, 3 mm; TR range/te range, 2100 2500/25 35) and fatsuppressed T2-weighted (slice thickness, 3 mm; TR range/te range, 3500 4000/55 60), sagittal oblique T1-weighted (slice thickness, 3 mm; TR range/te range, 450 600/12 15) and fat-suppressed T2- weighted (slice thickness, 3 mm; TR range/te range, 3500 4000/55 65), and axial fat-suppressed proton density weighted (slice thickness, 3 mm; TR range/te range, 2100 2500/25 37). The FOV was 140 mm, and the matrix was 256 256 for all the studies. Dedicated 16-channel shoulder array coils Imaging Evaluation Two musculoskeletal radiologists (each with 3 years of experience) reviewed these studies independently and documented the presence and, if present, the size of the Hill-Sachs lesions (craniocaudal anteroposterior depth; volume) (Fig. 1). The location of the Hill-Sachs lesion was documented using a revised biceps angle measurement, which consisted of the angle between the center of the biceps groove and medial margin of the lesion (Fig. 2). This measurement accounted for the medial extent of the Hill-Sachs lesion, which is the margin that would first engage onto the glenoid in the setting of abduction and external rotation. The readers documented the presence and size of glenoid bone loss as the percentage of articular surface bone loss along the width of the glenoid using a modified circle method (Fig. 3) [17 19]. Each reader also noted the presence of an anterior capsulolabral injury. Statistical Analysis Mann-Whitney tests were performed to compare the measurements between the engaging and nonengaging groups. Logistic regression and a receiver operating characteristic (ROC) curve (area under the curve [AUC]) analysis were performed to assess the utility of each type of measurement for differentiating patients with engagement from those without engagement. Pearson correlation tests were used to Fig. 2 34-year-old man with history of multiple prior shoulder dislocations and engagement on physical examination. Axial fat-suppressed proton density weighted image shows revised biceps angle measurement (dotted line) between center of biceps groove and medial margin of Hill-Sachs lesion. W634 AJR:201, October 2013

MRI of Hill-Sachs Lesions Fig. 3 24-year-old man with history of recurrent dislocation and engagement on physical examination. Circle method of measuring glenoid bone loss is shown on this sagittal T1-weighted image. Vertical line (blue) was drawn between superior glenoid tubercle and inferior margin of glenoid to mark center of glenoid. Best-fit circle (red) was drawn around inferior aspect of glenoid with center of circle placed on vertical line and its borders along intact posterior and inferior glenoid margins. Horizontal line (yellow) was drawn through center of circle to signify estimated width of intact glenoid. Second line (green) was drawn between anterior aspect of remnant glenoid and anterior margin of circle to represent amount of missing bone. This measurement was then divided by estimate of intact glenoid to produce percentage of glenoid bone loss. This defect measured approximately 30%. assess the association between each of the measurements and the presence of an anterior capsulolabral injury. The estimates of glenoid bone loss made on MRI and in the operating room were also compared using an exact Wilcoxon test. The intraclass correlation coefficient was used for calculation of interobserver agreement of the measurements. Results Eleven patients had evidence of engagement on physical examination and 22 did not have engagement on physical examination; these examinations were performed while the patient was under anesthesia before arthroscopy. Nine patients (one in the engaging group and eight in the nonengaging group) had a reported acute anterior shoulder dislocation and 24 (10 in the engaging group and 14 in the nonengaging group) had reported chronic, recurrent shoulder instability. Hill-Sachs lesions were found in all of the patients on MRI and arthroscopy. For the engaging group, the mean measurements for the Hill-Sachs lesions were as follows: 2.15 ± 0.63 (SD) cm (anteroposterior) 2.13 ± 0.54 cm (craniocaudal) 0.69 ± 0.23 cm (depth); and 3.60 ± 2.83 cm 3 (volume) (Table 1). The mean biceps angle measurement for this group was 154.5 ± 15.8. For the nonengaging group, the mean measurements for the Hill-Sachs lesions were as follows: 2.10 ± 0.66 cm (anteroposterior) 2.03 ± 0.76 cm (craniocaudal) 0.63 ± 0.27 cm (depth); and 3.23 ± 2.69 cm 3 (volume) (Table 1). The mean biceps angle measurement for this group was 143.91 ± 17.55. There was no statistically significant difference between any of the Hill-Sachs measurements when comparing the two groups of TABLE 1: Measures of the Engaging and Nonengaging Groups Nonengaging Hill-Sachs Lesion (n = 22) patients (range of p values, 0.27 0.72). There was also no statistically significant difference between the mean biceps angle measurements, but there was a trend toward the angle being larger in the engaging group (p = 0.07). Glenoid bone loss was found in all but one of the patients in the engaging group (10/11) and in 10 of the patients in the nonengaging group (10/22). Eight patients, five in the engaging group and three in the nonengaging group, had evidence of a fracture fragment or fragments along the anterior glenoid, whereas the remaining patients had a flattened anterior glenoid. The mean percentage of glenoid loss in the engaging group was 20.2% ± 10.2% and was 6.0% ± 6.9% in the nonengaging group (Table 1). This difference was statistically sig- Engaging Hill-Sachs Lesion (n = 11) Measure Mean SD Mean SD p a Biceps angle ( ) 143.91 17.55 154.46 15.81 0.0687 Glenoid bone loss (%) 6.00 6.92 20.18 10.23 0.0006 Hill-Sachs lesion Anteroposterior (cm) 2.10 0.66 2.15 0.63 0.7273 Volume (cm 3 ) 3.23 2.69 3.60 2.83 0.2718 Craniocaudal (cm) 2.03 0.76 2.13 0.54 0.2881 Depth (cm) 0.63 0.27 0.69 0.23 0.4215 a Each p value is the exact two-sided significance level from a Mann-Whitney test to compare subject groups. TABLE 2: Areas Under the Receiver Operating Characteristic Curve (AUCs) and p Values From Logistic Regression to Assess the Utility of Each Measure for Discriminating Between Patients With Engaging Hill- Sachs Lesions and Those With Nonengaging Hill-Sachs Lesions Measure AUC p Biceps angle ( ) 0.698 0.1102 Glenoid bone loss (%) 0.847 < 0.0001 Hill-Sachs lesion Anteroposterior (cm) 0.539 0.7174 Volume (cm 3 ) 0.570 0.7077 Craniocaudal (cm) 0.564 0.6899 Depth (cm) 0.587 0.5018 TABLE 3: Glenoid Bone Loss as Measured on MRI and at Arthroscopy Measure Mean SD p Glenoid bone loss (%) Measured on MRI 16.42 12.53 Measured during arthroscopy 16.25 12.48 Within-subject difference between MRI and arthroscopy measures a 0.167 1.528 0.893 a Calculated as follows: (glenoid bone loss measured on MRI) (glenoid bone loss measured at arthroscopy). An exact Wilcoxon test was conducted to compare the measurements. AJR:201, October 2013 W635

Gyftopoulos et al. nificant (p = 0.001). ROC curve analysis using the AUC showed that glenoid bone loss was the only measurement that could be used to reliably discriminate between the two groups (AUC = 0.847, p < 0.0001) (Table 2). The diagnostic accuracy of MRI for engagement was maximized when using a threshold of 23% glenoid bone loss, which achieved 63.6% (7/11) sensitivity and 100% (22/22) specificity. Logistic regression and Pearson correlation analysis showed that no set of two or more measurements was a significant independent predictor of engagement. For 12 of the patients, seven in the engaging group and five in the nonengaging group, estimates of glenoid bone loss were performed using the bare area method during surgery and reported in the operative report: For this subset of patients, the mean percentage of glenoid loss was 16.4% ± 12.5% (SD) on MRI and 16.25% ± 12.5% in the operating room, without a statistically significant difference (p = 0.893) (Table 3). The intraclass correlation coefficient for the two readers was 0.98 for glenoid bone loss and the biceps angle, 0.97 for Hill-Sachs craniocaudal length, 0.96 for Hill-Sachs anteroposterior length, 0.28 for Hill-Sachs depth, and 0.75 for Hill-Sachs area. Each of the patients in the engaging and nonengaging Hill-Sachs groups had an anterior capsulolabral tear on MRI and arthroscopy. Thus, the presence of an anterior capsulolabral tear was omitted from analysis because it was constant and cannot show an association with any other measure. Discussion Burkhart and De Beer [3] originally described the term engaging Hill-Sachs lesion as an osseous injury that occurs in the setting of anterior shoulder dislocation and predisposes to failure of arthroscopic repair. Engagement was defined as the reproduction of symptoms of instability, such as increased apprehension or a catching or popping sensation, when the shoulder was placed in a functional position of abduction and external rotation, which was defined as 90 of abduction and between 0 and 135 of extension. Symptoms of instability were theorized to be a result of the geometry of Hill-Sachs lesions with horizontal defects relative to the humeral shaft in the nonabducted position that become parallel to the anterior margin of the glenoid in the abducted and externally rotated (ABER) position, which predisposes to engagement (Fig. 4). Alternatively, Hill- Sachs lesions oriented more vertically relative to the humerus in the nonabducted position and extending diagonally to the anterior glenoid in the ABER position result in continuous, uninterrupted contact between these two surfaces without engagement (Fig. 5). A limitation of this engaging Hill-Sachs study by Burkhart and De Beer was that diagnostic imaging was not used. Yamamoto et al. [11] used cadavers and 3D CT to conduct a biomechanical and anatomic study of Hill-Sachs lesions [11]. This study showed that the position of the Hill-Sachs lesion and not the size of the lesion was the most important factor in terms of predisposing to engagement using what they termed the concept of the glenoid track. The authors defined the glenoid track as the contact area between the humerus and glenoid during abduction and external rotation and measured it to be approximately 84% of the glenoid width (Fig. 6A). A Hill-Sachs lesion that remained in this track would have constant contact without risk for engagement (Fig. 6B). A Hill-Sachs lesion that extended past the medial margin of the glenoid track would result in overriding and engagement (Fig. 6C). Although this Fig. 4 Engaging Hill Sachs lesion. A, Hill-Sachs lesion (blue) is obliquely oriented on neutral view. B, When shoulder is in abducted and externally rotated position, margins of Hill-Sachs lesion (blue) and anterior glenoid are parallel, which predisposes to engagement. Fig. 5 Nonengaging Hill Sachs lesion. A, Hill-Sachs lesion (blue) is vertically oriented on neutral view. B, When shoulder is in abducted and externally rotated position, margins of Hill-Sachs lesion (blue) are diagonal to anterior glenoid, which decreases chance of engagement. W636 AJR:201, October 2013

MRI of Hill-Sachs Lesions study was not the first to examine the importance of the relationship between glenoid loss and a Hill-Sachs lesion in predisposing to instability [20 22], it was the first to provide a biomechanical explanation. Several studies have examined the role that the size of the Hill-Sachs lesion plays in engagement and predisposing to instability and have used estimates between 20% and 40% of humeral head involvement as thresholds for treatment [7 10, 23 25]. A recent study by Cho et al. [13] used 3D CT to evaluate the role of size, location, and orientation of Hill-Sachs lesions in engagement. Their study found that engaging Hill-Sachs lesions were typically larger and more horizontally oriented than those that did not engage, which was consistent with prior studies. Our study correlated the findings on physical examination with MR characteristics of Hill-Sachs lesions and glenoid bone loss. The average size of the Hill-Sachs lesion in the engagement group was, overall, larger than the average lesion in the nonengaging group; however, the difference was not statistically significant. Thus, the size of the Hill-Sachs lesion did not appear to be associated with engagement. These results differ from previously published studies [10, 12, 13]. We were A Fig. 6 Glenoid track theory [11]. A, Contact area between glenoid and humeral head during abduction and external rotation, referred to as glenoid track (blue), measures approximately 84% of glenoid width, whereas remainder of glenoid (16%, green) contacts medial margin of rotator cuff footprint. B, Hill-Sachs lesion remains within glenoid track (arrow), which provides consistent contact between two surfaces and limits risk for engagement. Green = rotator cuff. C, Hill-Sachs lesion extends past medial margin of glenoid (arrow), which leads to overriding of margins and predisposes to engagement. Green = rotator cuff. B surprised by this finding and are unsure of the reason for this discrepancy. We assessed the location of the Hill-Sachs lesions using a revised biceps angle. As opposed to prior studies that used the biceps groove as a reference point for location of the Hill-Sachs lesion [26] or an angle between the groove and the central aspect of the Hill-Sachs lesion to assess location [13], our biceps angle accounted for the medial extent of the lesion because this portion would be the first portion to engage the anterior glenoid with the shoulder in abduction and external rotation. There was a trend for this angle to be larger in the engaging group, but the biceps angle of the engaging group was not significantly different from that of the nonengaging group (p = 0.07). Given this trend in a relatively small sample of patients in the engaging group, the presence of medially oriented Hill-Sachs lesions may be of clinical importance. The only statistically significant difference between the measurements of the two groups of patients was in the amount of glenoid bone loss: The engaging group had an average defect of 20%, whereas the nonengaging group had an average of 6% (p = 0.001). The diagnostic accuracy of MRI for engagement was maximized when using a glenoid bone loss threshold of 23%, which achieved 63.6% (7/11) sensitivity and 100% (22/22) specificity. Thus, the presence of a glenoid defect that involves at least 23% of the glenoid width on MRI may be predictive of engagement on physical examination. Our results showing there was no statistically significant difference between glenoid bone loss measurements produced on MRI and in arthroscopy (p = 0.893) validated the findings of a previous cadaveric study that showed the ability to accurately quantify glenoid bone loss on conventional MRI [19]. The findings of our study show the importance of considering both the Hill-Sachs lesion and glenoid bone loss when evaluating patients with engagement on physical examination and is supportive of the glenoid track theory [11]. Greater glenoid loss results in a narrower glenoid track, which, in turn, allows the Hill-Sachs lesion needed to engage along the anterior glenoid to be smaller. Although only a trend, the increasing medial extent of a Hill-Sachs lesion may also predispose to engagement especially in a patient with underlying anterior glenoid loss. Several treatment options have been described for engaging Hill-Sachs lesions including Latarjet and Bristow coracoid transfer procedures; remplissage, which consists C AJR:201, October 2013 W637

Gyftopoulos et al. of transfer of the infraspinatus tendon into the Hill-Sachs defect; capsular transfer in the form of a Putti-Platt procedure; rotational humeral osteotomy; and bone graft placement [27 30]. Given our findings, glenoid loss augmentation with procedures such as the Latarjet may need to be considered as a firstline treatment in patients with engagement. On the other hand, Hill-Sachs augmentation procedures and the associated risks and morbidities may not be necessary for this subset of patients. There were several limitations for our study including its retrospective nature and small sample size of patients in both the engaging and nonengaging groups. We were also unable to reliably assess the orientation of the Hill-Sachs lesion, a previously described predictor of engagement, because of the 2D nature of the MRI used in the study. There was poor correlation between the two readers in terms of the measurements of Hill-Sachs lesion depth. There is no clear reason for this difference. The fact that depth was typically the smallest dimension of the lesion (< 1 cm in the majority of the patients) may have resulted in a significant discrepancy during analysis even if the measurements differed by a small amount (2 3 mm). Conclusion Our study results indicate that there is a significant association between an engaging Hill- Sachs lesion on physical examination and the degree of glenoid bone loss as assessed on MRI. There was also a trend toward increased engagement with more medially oriented Hill-Sachs lesions. These findings show the importance of considering both the Hill- Sachs lesion and glenoid bone loss when evaluating patients with engagement and are supportive of the glenoid track theory [11]. References 1. Kim DS, Yoon YS, Ho Yi C. Prevalence comparison of accompanying lesions between primary and recurrent anterior dislocation in the shoulder. Am J Sports Med 2010; 38:2071 2076 2. Yiannakopoulos CK, Mataragas E, Antonogiannakis E. A comparison of the spectrum of intra-articular lesions in acute and chronic anterior shoulder instability. Arthroscopy 2007; 23:985 990 3. 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; 16:677 694 4. Itoi E, Lee SB, Berglund LJ, Berge LL, An KN. The effect of a glenoid defect on anteroinferior stability of the shoulder after Bankart repair: a cadaveric study. J Bone Joint Surg Am 2000; 82:35 46 5. Lo IK, Perten PM, Burkhart SS. The inverted pear glenoid: an indicator of significant glenoid bone loss. Arthroscopy 2004; 20:169 174 6. Greis PE, Scuderi MG, Mohr A, Bachus KN, Burks RT. Glenohumeral articular contact areas and pressures following labral and osseous injury to the anteroinferior quadrant of the glenoid. J Shoulder Elbow Surg 2002; 11:442 451 7. Bock P, Kluger R, Hintermann B. Anatomical reconstruction for reverse Hill-Sachs lesions after posterior locked shoulder dislocation fracture: a case series of six patients. Arch Orthop Trauma Surg 2007; 127:543 548 8. Chen AL, Hunt SA, Hawkins RJ, Zuckerman JD. Management of bone loss associated with recurrent anterior glenohumeral instability. Am J Sports Med 2005; 33:912 925 9. Miniaci A, Gish MW. Management of anterior glenohumeral instability associated with large Hill-Sachs defects. Tech Shoulder Elbow Surg 2004; 5:170 175 10. Sekiya JK, Wickwire AC, Stehle JH, Debski RE. Hill-Sachs defects and repair using osteoarticular allograft transplantation: biomechanical analysis using a joint compression model. Am J Sports Med 2009; 37:2459 2466 11. Yamamoto N, Ito E, Abe H, 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; 16:649 656 12. Kaar SG, Fening SD, Jones MH, Colbrunn RW, Miniaci A. Effect of humeral head defect size on glenohumeral stability: a cadaveric study of simulated Hill-Sachs defects. Am J Sports Med 2010; 38:594 599 13. Cho SH, Cho NS, Rhee YG. Preoperative analysis of the Hill-Sachs lesion in anterior shoulder instability: how to predict engagement of the lesion. Am J Sports Med 2011; 39:2389 2395 14. Pollock RG, Flatow EL. Classification and evaluation. In: Bigliani LU, ed. AAOS monograph series: the unstable shoulder. Rosemont, IL: American Academy of Orthopaedic Surgeons, 1996:25 36 15. Farber AJ, Castillo R, Clough M, Bahk M, Mc- Farland EG. Clinical assessment of three common tests for traumatic anterior shoulder instability. J Bone Joint Surg Am 2006; 88:1467 1474 16. Bushnell BD, Creighton RA, Herring MM. The bony apprehension test for instability of the shoulder: a prospective pilot analysis. Arthroscopy 2008; 24:974 982 17. Sugaya H, Moriishi J, Dohi M, Kon Y, Tsuchiya A. Glenoid rim morphology in recurrent anterior glenohumeral instability. J Bone Joint Surg Am 2003; 85:878 884 18. Huysmans PE, Haen PS, Kidd M, Wouter JD, Willems JW. The shape of the inferior part of the glenoid: a cadaveric study. J Shoulder Elbow Surg 2006; 15:759 763 19. Gyftopoulos S, Hasan S, Bencardino J, et al. Diagnostic accuracy of MRI in the measurement of glenoid bone loss. AJR 2012; 199:873 878 20. Weber BG, Simpson LA, Hardegger F. Rotational humeral osteotomy for recurrent anterior dislocation of the shoulder associated with a large Hill-Sachs lesion. J Bone Joint Surg Am 1984; 66:1443 1450 21. Lazarus MD, Harryman DT 2nd. Open repairs for anterior instability. In: Warner JJP, Ianotti JP, Gerber C, eds. Shoulder surgery. Philadelphia, PA: Lippincott-Raven, 1997:47 63 22. Burkhart SS, Danaceau SM. Articular arc length mismatch as a cause of failed Bankart repair. Arthroscopy 2000; 16:740 744 23. Rowe CR, Zarins B, Ciullo JV. Recurrent anterior dislocation of the shoulder after surgical repair: apparent causes of failure and treatment. J Bone Joint Surg Am 1984; 66:159 168 24. Bühler M, Gerber C. Shoulder instability related to epileptic seizures. J Shoulder Elbow Surg 2002; 11:339 344 25. Gerber C, Lambert SM. Allograft reconstruction of segmental defects of the humeral head for the treatment of chronic locked posterior dislocation of the shoulder. J Bone Joint Surg Am 1996; 78:376 382 26. Workman TL, Burkhard TK, Resnick D, et al. Hill-Sachs lesion: comparison of detection with MR imaging, radiography, and arthroscopy. Radiology 1992; 185:847 852 27. Helfet AJ. Coracoid transplantation for recurring dislocation of the shoulder. J Bone Joint Surg Br 1958; 40:198 202 28. Hovelius L, Sandström B, Sundgren K, Saebö M. One hundred eighteen Bristow-Latarjet repairs for recurrent anterior dislocation of the shoulder prospectively followed for fifteen years. Part I. Clinical results. J Shoulder Elbow Surg 2004; 13:509 516 29. Purchase RJ, Wolf EM, Hobgood ER, Pollock ME, Smalley CC. Hill- Sachs remplissage : an arthroscopic solution for the engaging Hill- Sachs lesion. Arthroscopy 2008; 24:723 726 30. Kazel MD, Sekiya JK, Greene JA, Bruker CT. Percutaneous correction (humeroplasty) of humeral head defects (Hill-Sachs) associated with anterior shoulder instability: a cadaveric study. Arthroscopy 2005; 21:1473 1478 W638 AJR:201, October 2013