Medicine The American Journal of Sports P

The American Journal of Sports Medicine http://ajs.sagepub.com/

The Effect of a Combined Glenoid and Hill-Sachs Defect on Glenohumeral Stability: A Biomechanical Cadaveric Study Using 3-Dimensional Modeling of 142 Patients Robert A. Arciero, Anthony Parrino, Andrew S. Bernhardson, Vilmaris Diaz-Doran, Elifho Obopilwe, Mark P. Cote, Petr Golijanin, Augustus D. Mazzocca and Matthew T. Provencher Am J Sports Med published online March 20, 2015 DOI: 10.1177/0363546515574677 The online version of this article can be found at: http://ajs.sagepub.com/content/early/2015/03/20/0363546515574677

Published by: http://www.sagepublications.com

On behalf of: American Orthopaedic Society for Sports Medicine

Additional services and information for The American Journal of Sports Medicine can be found at:

P> OnlineFirst Version of Record - Mar 20, 2015 What is This?

Downloaded from ajs.sagepub.com at Univ of Connecticut / Health Center / Library on April 13, 2015

AJSM PreView, published on March 20, 2015 as doi:10.1177/0363546515574677

The Effect of a Combined Glenoid and Hill-Sachs Defect on Glenohumeral Stability A Biomechanical Cadaveric Study Using 3-Dimensional Modeling of 142 Patients Robert A. Arciero,*y MD, Anthony Parrino,y MD, Andrew S. Bernhardson,z MD, LT, MC, USN, Vilmaris Diaz-Doran,y MS, BMe, Elifho Obopilwe,y MS, BMe, Mark P. Cote,y DPT, MSCTR, Petr Golijanin,§ BS, Augustus D. Mazzocca,y MD, MS, and Matthew T. Provencher,§ MD, CDR, MC, USNR Investigation performed at University of Connecticut Health Center, Farmington, Connecticut, USA Background: Bone loss in anterior glenohumeral instability occurs on both the glenoid and the humerus; however, existing biomechanical studies have evaluated glenoid and humeral head defects in isolation. Thus, little is known about the combined effect of these bony lesions in a clinically relevant model on glenohumeral stability. Hypothesis/Purpose: The purpose of this study was to determine the biomechanical efficacy of a Bankart repair in the setting of bipolar (glenoid and humeral head) bone defects determined via computer-generated 3-dimensional (3D) modeling of 142 patients with recurrent anterior shoulder instability. The null hypothesis was that adding a bipolar bone defect will have no effect on glenohumeral stability after soft tissue Bankart repair. Study Design: Controlled laboratory study. Methods: A total of 142 consecutive patients with recurrent anterior instability were analyzed with 3D computed tomography scans. Two Hill-Sachs lesions were selected on the basis of volumetric size representing the 25th percentile (0.87 cm3; small) and 50th percentile (1.47 cm3; medium) and printed in plastic resin with a 3D printer. A total of 21 cadaveric shoulders were evaluated on a custom shoulder-testing device permitting 6 degrees of freedom, and the force required to translate the humeral head anteriorly 10 mm at a rate of 2.0 mm/s with a compressive load of 50 N was determined at 60° of glenohumeral abduction and 60° of external rotation. All Bankart lesions were made sharply from the 2- to 6-o’clock positions for a right shoulder. Subsequent Bankart repair with transosseous tunnels using high-strength suture was performed. Hill-Sachs lesions were made in the cadaver utilizing a plastic mold from the exact replica off the 3D printer. Testing was conducted in the following sequence for each specimen: (1) intact, (2) posterior capsulotomy, (3) Bankart lesion, (4) Bankart repair, (5) Bankart lesion with 2-mm glenoid defect, (6) Bankart repair, (7) Bankart lesion with 2-mm glenoid defect and Hill-Sachs lesion, (8) Bankart repair, (9) Bankart lesion with 4-mm glenoid defect and Hill-Sachs lesion, (10) Bankart repair, (11) Bankart lesion with 6-mm glenoid defect and Hill-Sachs lesion, and (12) Bankart repair. All sequences were used first for a medium Hill-Sachs lesion (10 specimens) and then repeated for a small HillSachs lesion (11 specimens). Three trials were performed in each condition, and the mean value was used for data analysis. Results: A statistically significant and progressive reduction in load to translation was observed after a Bankart lesion was created and with the addition of progressive glenoid defects for each humeral head defect. For medium (50th percentile) Hill-Sachs lesions, there was a 22%, 43%, and 58% reduction in stability with a 2-, 4-, and 6-mm glenoid defect, respectively. For small (25th percentile) Hill-Sachs lesions, there was an 18%, 27%, and 42% reduction in stability with a 2-, 4-, and 6-mm glenoid defect, respectively. With a 2-mm glenoid defect, the medium Hill-Sachs group demonstrated significant reduction in translation force after Bankart repair (P \ .01), and for the small Hill-Sachs group, a 4-mm glenoid defect was required to produce a statistical decrease (P \ .01) in reduction force after repair. Conclusion: Combined glenoid and humeral head defects have an additive and negative effect on glenohumeral stability. As little as a 2-mm glenoid defect with a medium-sized Hill-Sachs lesion demonstrated a compromise in soft tissue Bankart repair, while small-sized Hill-Sachs lesions showed compromise of soft tissue repair with 4-mm glenoid bone loss.

The American Journal of Sports Medicine, Vol. XX, No. X DOI: 10.1177/0363546515574677 Ó 2015 The Author(s)

1 Downloaded from ajs.sagepub.com at Univ of Connecticut / Health Center / Library on April 13, 2015

2

Arciero et al

The American Journal of Sports Medicine

Clinical Relevance: Bipolar bony lesions of the glenoid and humeral head occur frequently together in clinical practice. Surgeons should be aware that the combined defects and glenoid bone loss of 2 to 4 mm or approximately 8% to 15% of the glenoid could compromise Bankart repair and thus may require surgical strategies in addition to traditional Bankart repair alone to optimize stability. Keywords: shoulder instability; biomechanics

In patients with recurrent traumatic anterior instability of the shoulder, glenoid and humeral head defects occur together.3,5,9,12 Glenoid bone loss (GBL) has been found in 49% to 86% of patients with recurrent instability,9 whereas humeral head defects have been found in 70% of first-time dislocations and up to 100% in recurrent instability of the shoulder.3,12 In addition, it has been well established that bone loss is the primary reason for failure after arthroscopic Bankart repair.1,2 Hill-Sachs lesions—which occur when the posterolateral aspect of the anteriorly dislocated humeral head impacts against the glenoid rim—cause an engaging bone lesion that has been shown to possibly compromise soft tissue Bankart repair alone.2 In a cadaveric study, Sekiya et al14 attempted to quantify the size of a Hill-Sachs defect and measure its effect on shoulder stability. Creating an osteotomy of 5/8 radius of the humeral head decreased shoulder stability in external rotation and abduction, while an osteotomy of 7/8 radius further decreased stability at neutral and external rotation. Another study7 showed that defects of 25% of the humeral head may decrease glenohumeral instability, while defects involving \20% of the humeral head are of little clinical significance. In these Hill-Sachs studies, the lesion was based on diameter of the humeral head and did not fully address the exact location, size, and shape of a Hill-Sachs lesion and thus may not have provided data that assess a clinically relevant scenario of bone loss. While there has been extensive work on investigating bone loss problems in the shoulder as a unipolar issue (either humeral head or GBL but not both), there is emerging clinical evidence that bony defects manifest as a bipolar phenomenon. However, little is known about the biomechanical implications of a bipolar bone loss model of the shoulder—one that is clinically manifested in the majority of instability cases. Indeed, in the patient with bipolar bone lesions observed on both the glenoid and humeral head, instability may be potentiated owing to the combined effect of these lesions on glenohumeral translation. Thus, the purpose of this study was to determine the biomechanical efficacy of a Bankart repair in the setting of bipolar bone defects (glenoid and humeral head) utilizing a clinically relevant model of bone loss based on

a 3-dimensional (3D) computed tomography (CT) scan database of 142 patients with recurrent anterior instability. This study sought to assess the biomechanical effects of 2 sizes of Hill-Sachs lesions with increasing amounts of GBL and a soft tissue Bankart repair model. The null hypothesis was that adding a bipolar bone defect (glenoid and humeral bone loss) will have no statistical effect on glenohumeral stability after soft tissue Bankart repair.

METHODS Initial CT Scan Modeling and 3D Printing To make a clinically applicable bone loss model of a humeral head defect in a cadaver, 142 patients (mean age, 29.3 years; range, 18-35 years; 138 men and 4 women)— all without prior surgery and the majority (n = 121 of 142) with recurrent anterior shoulder instability—had a 3D CT scan that demonstrated a minimum 5% GBL and bony evidence of a Hill-Sachs lesion. These images were processed with Osirix software, and the volume (cm3), length (cm), width (cm), and depth (cm) of the Hill-Sachs lesion as well as GBL was determined. The patients were then classified according to volume of Hill-Sachs lesion, which was measured according to an automated process utilizing a region-of-interest tool to determine the volume, location, size, dimensions, and shape of the Hill-Sachs lesion for each patient (Table 1). The amount of GBL was also measured and recorded. The total volumetric data were utilized as a baseline to construct a distribution plot volume versus patient number (Figure 1). A 3D plastic model (Makerbot 3D printer) was then created using an actual CT scan, which was representative of a Hill-Sachs lesion representing the 25th- and 50th-percentile means of the 142-patient cohort. The 25th-percentile lesion was identified as a small Hill-Sachs injury and corresponded to a bone deficiency of 0.87 cm3, and the 50th-percentile lesion was identified as a medium Hill-Sachs (Figure 1) based on volumetric assessment (1.47 cm3) (Figure 1). These 2 representative Hill-Sachs lesions from the database were then printed to scale with solid plastic resin on a 3D printer (Makerbot).

*Address correspondence to Robert A. Arciero, MD, Department of Orthopaedic Surgery, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, CT 06034, USA (email: [email protected]). y Department of Orthopaedic Surgery, University of Connecticut Health Center, Farmington, Connecticut, USA. z Department of Orthopaedics, Naval Medical Center, San Diego, California, USA. § Sports Medicine Division, Department of Orthopaedic Surgery, Massachusetts General Hospital, Boston, Massachusetts, USA. Presented at the 40th annual meeting of the AOSSM, Seattle, Washington, July 2014. One or more of the authors has declared the following potential conflict of interest or source of funding: Funding was received from United States Navy Grant N00259-12-P-1866. Implants used in the pilot study were donated by Arthrex Inc.


Vol. XX, No. X, XXXX

Assessment of Combined Bony Glenoid and Hill-Sachs Defects

TABLE 1 Hill-Sachs Lesions, Patients per GBL Group, and Patient Characteristicsa Hill-Sachs lesion characteristics, mean 6 SD Length, cm Depth, cm Height, cm Volume, cm3 Patients according to percentage GBL, No. \10% 10%-20% 10%-20% Patient characteristics (N = 142) Age, y, mean (range) Men, No.

2.5 6 0.55 0.67 6 0.22 1.79 6 0.3 1.22 6 0.13-5.3 36 71 35 29.3 (18-35) 138

a

GBL, glenoid bone loss.

3

there was \60° elevation relative to the scapula, \45° of external rotation, or \20° of internal rotation. Each specimen was thawed overnight at room temperature before preparation. All soft tissues superficial to the rotator cuff muscles were removed. The rotator cuff muscles were elevated from the scapula and the tendinous portions bluntly separated from the capsule, from medial to lateral, to a level that was approximately 1 cm lateral to the glenohumeral joint line. Elevation of the rotator cuff tendons was discontinued at this level to ensure no damage to the capsuloligamentous structures. The muscles of the arm and the periosteum were removed from the humeral shaft, which was then fixed in the center of a polyvinyl chloride (PVC) sleeve with use of polymethylmethacrylate (PMMA). By this step of the dissection procedure, the contour of the glenoid rim was visible from the outside of the joint capsule.

Biomechanical Testing The scapula was trimmed and potted in a custom box, with the glenoid parallel to the top of the scapular box and the superior-inferior and anterior-posterior axes aligned with the width and height of the box. A custom shoulder-testing system permitting 6 degrees of freedom positioning of the glenohumeral joint was used.8 The scapular box was mounted onto a vertical linear bearing translator and lever arm system on top of 2 translation plates that allowed anterior-posterior and superior-inferior translation. The humerus was attached to an arc at the top of the testing system that allowed humeral rotation, abduction, and horizontal adduction. A similar testing protocol was utilized as described by Itoi et al6 and Yamamoto et al.15-18 A compressive load of 50 N was applied perpendicular to the glenoid. This compressive load allowed the glenohumeral joint to find its anatomic neutral position, which was defined as the starting reference point for the specimen. Weight was applied through the vertical linear bearing translator and lever arm system. In both directions, the least lateral position of the humeral head on the glenoid surface was defined as the reference-neutral position. First, the neutral position was identified by measuring the position where the humeral head was seated at the bottom of the glenoid, as previously reported.6,15,16 This reference-neutral position was used for the subsequent displacement-control study. Figure 1. Bell-shaped curve generated by the volumetric analysis of 142 consecutive patients with recurrent anterior glenohumeral instability. The 25th-percentile lesion measured 0.87 cm3, and the 50th-percentile lesion measured 1.47 cm3.

Cadaveric Application of the 3D Model A total of 21 fresh-frozen cadaveric shoulders were used in the final analysis. Cadavers were visually inspected and excluded if they had a defect. Specimens included left and right shoulders of men and women aged 42 to 65 years. All specimens were free of moderate or severe glenohumeral arthritis or joint contractures confirmed by direct visualization. Joint contractures were considered present if

Testing Positions and Translation Conditions A MicroScribe G2 (Immersion Corp) identified the starting reference point once the humeral head was centered in the glenoid. Once the starting reference point was recorded with the MicroScribe, the humeral head was translated in the anterior direction for 10 mm at a rate of 2.0 mm/s via a cable attached to the actuator of a servohydraulic test system (MTS 858; MTS Systems). This rate and amplitude of displacement has been used in prior studies of unipolar bone loss assessment.6,15-18 The MTS machine load cell recorded the peak translational force required to move the humeral head 10 mm and served as the primary biomechanical outcome measure. The final anterior displacement of the center of the humeral head was recorded


4

Arciero et al


using the MicroScribe G2. Inter- and intrareliability testing was performed using this device. Ten predetermined lengths were measured with the electronic calipers and then the MicroScribe. The percentage error between the 2 devices was calculated and found to be 2.4% 6 1.02%. Three trials were performed in each condition, and the mean value was used for data analysis. The test positions were 60° of shoulder abduction and 60° of external rotation as described by Sekiya et al,14 which corresponded to 90° of glenohumeral abduction and 90° of external rotation. Rotational range of motion was quantified using a 360° goniometer with a precision of 0.5°, attached to the top of the shoulder-testing jig using the long head of the biceps as a consistent marker in rotational measurements.8

Creation of Glenoid and Humeral Head Bony Defects A posterior capsulotomy was performed to permit exposure to create all the bony defects and repairs. Care was taken to not violate the anteroinferior capsule. In the pilot study to determine the optimal access to the glenohumeral joint with preservation of soft tissue, the specimens were tested intact, after posterior capsulotomy and then with the posterior capsule repaired. The load to anterior translation was not statistically different with posterior capsulotomy and with the posterior capsule repaired even in the presence of the humeral head lesion. Therefore, based on the pilot testing data, the posterior capsulotomy was utilized for access to bony defect creation, for the creation of the Bankart labral tear, and also for Bankart repair as described below. Furthermore, this was prepared in the same testing condition as prior studies that did not have the capsule intact.6,15-18 All Bankart lesions were made sharply from the 2- to 6-o’clock positions for a right shoulder and 10- to 6-o’clock positions for a left shoulder. Bankart repair was made with 2.0-mm transosseous tunnels using No. 2 Fiberwire (Arthrex). Transosseous tunnels were placed at the edge of the glenoid to reestablish the glenoid concavity (Figure 2). These tunnels exited the posterior scapular neck approximately 1 cm medial to the posterior glenoid margin. Sutures were tied over the bone posteriorly in a mattress fashion. This repair obviated the need for repeated drilling and reinsertion of suture anchors and the risk of compromising or removing bone in addition to the planned glenoid osteotomy and, based on pilot data, allowed for a reproducible and stable repair. All initial and subsequent repairs were performed with the arm in 0° of abduction and 30° of external rotation relative to the scapular plane, which is equivalent to adduction and neutral rotation relative to the trunk, and utilized same tissue planes for reproducibility of repair. A digital micrometer was used to measure and create glenoid defects with a saw parallel to the anterior glenoid as previously described by Yamamoto et al16 (Figure 3). The glenoid defects represented a clinically relevant model of GBL along a line that parallels the long axis of the glenoid and were made in 2-mm increments with a template based on corresponding glenoid bone resin-printed models. The 2 plastic resin Hill-Sachs lesions created with the 3D printer were then converted to a cadaveric model, and the exact Hill-Sachs lesion was transcribed to each cadaver

with scaling based on percentage difference in humeral head articular diameter. The humeral head defect was measured and the volume determined, and a near size-matched humeral head from a cadaver was utilized. If there was a difference in the 3D model humeral head diameter versus the specimen, the cadaveric volume of the defect was increased or decreased by the same percentage. The creation of the Hill-Sachs lesion was performed according to measurements from the bicipital groove, the center of the humeral head, and the junction of the articular margin and neck of the humerus. A positive mold of PMMA was created from the Hill-Sachs defect of the 3D model and used to precisely reproduce the corresponding size, location, and volume in the cadaveric shoulders (Figures 4 and 5). The defect was created using a combination of oscillating saw, bur, and rasp to achieve the precise defect size and shape based on the plastic positive mold of the defect. The volume was confirmed with the positive mold made from the printed 3D plastic model, and the final size, depth and volume confirmed via this mold of defect form from the 3D printer. To validate the technique and ensure that the humeral head defects were proportional to the defect in the 3D models, the humeral head and the humeral head defects of the 3D models and all cadaveric specimens were digitized using a MicroScribe. These data were interfaced with Rhinoceros 3D modeling software (McNeel and Associates) to create a 3D image. This permitted the volume and surface area of the defect to be compared for each specimen to the size of the humeral head and to the 3D models. The percentage defect of the humeral heads of the specimens with the small Hill-Sachs defect were within 1.5% of the volume of the defect relative to the entire humeral head of the 3D model. For the medium-sized defect, the percentage defect created in the cadaveric specimen was within 2% of the volume of the defect of the 3D model of the medium-sized defect. Testing was then conducted in the following sequence for each specimen: (1) intact, (2) posterior capsulotomy, (3) Bankart lesion, (4) Bankart repair, (5) Bankart lesion with 2-mm glenoid defect, (6) Bankart repair, (7) Bankart lesion with 2-mm glenoid defect and Hill-Sachs lesion, (8) Bankart repair, (9) Bankart lesion with 4-mm glenoid defect and Hill-Sachs lesion, (10) Bankart repair, (11) Bankart lesion with 6-mm glenoid defect and Hill-Sachs lesion, and (12) Bankart repair.

Statistical Methods Data analysis was based on the work of Yamamoto et al,16,18 who employed a similar study design to examine the critical size of anterior glenoid defects. Peak translational forces with the various humeral head and glenoid defects, with and without Bankart lesion repair, were compared using a 1-way repeated-measures analysis of variance. Significant effects on the analysis of variance were analyzed further with the Dunnett test. This was performed for 2 testing conditions: progressive defects with and without a Bankart repair. For the conditions without a Bankart repair, the force needed to translate the humeral head 10 mm after the creation of the Bankart lesions was used as the reference group. For the conditions with a Bankart




Figure 2. Superior view of right shoulder demonstrating 3 mattress transosseous sutures to repair the Bankart lesion. Note that the anteroinferior capsule is left intact.

Figure 3. Cadaveric right shoulder with 2 mm of glenoid bone being resected, Bankart lesion, and Hill-Sachs defect. repair, the force needed to translate after the Bankart repair in the absence of any bone loss condition was used as the reference group, as this was felt to represent the ‘‘best soft tissue repair state.’’ This analysis was performed for both the 0.87-cm3 and the 1.47-cm3 Hills-Sachs lesion. The level of significance was set at P \ .05.

RESULTS Small Hill-Sachs Defect (0.87 cm3) For intact specimens, the mean force to translate 10 mm was 68 6 15 N. After the creation of a Bankart lesion, the force needed decreased to 55 6 15 N. Compared with

5

Figure 4. A ‘‘positive’’ mold of polymethylmethacrylate created for a medium (1.47 cm3) Hill-Sachs defect generated by the 3-dimensional printer model.

Figure 5. Cadaveric reproduction of humeral head defect created from 3-dimensional printer model. the Bankart lesion, the mean force to translate after the creation of the Hill-Sachs lesion was 47 6 15 N (P = .305) and decreased to 42 6 18 N (P = .111) with the addition of a 2-mm glenoid defect. There was a statistically significant decrease in translation force with a 4-mm glenoid defect (34 6 15 N; P = .007) and with a 6-mm glenoid defect (27 6 15 N; P = .001). After repair of the Bankart lesion, the mean force to translate the humeral head 10 mm was 63 6 15 N. Compared with the Bankart repair, the mean force after repair with a Hill-Sachs lesion was 56 6 14 N (P = .383) and 55 6 14 N for the repair with a 2-mm glenoid defect (P = .305). With the repair and a 4-mm glenoid defect, there was a statistically significant decrease (46 6 15 N; P = .031), which continued with a 6-mm glenoid defect (41 6 16 N; P = .005). This pattern of reduction is depicted in Figure 6.


6

Arciero et al


Figure 6. Mean force to dislocate with the small (0.87 cm3) defect and subsequent Bankart repair. Note progressive reduced load to translation with progressive with glenoid bone loss. Significant reduction in translation began at 4 mm of glenoid bone loss. Bk, Bankart; HS, Hill-Sachs; PC, posterior capsulotomy; rep, repair.

Medium Hill-Sachs Defect (1.47 cm3) For intact specimens, the mean force to translate 10 mm was 72 6 18 N. After the creation of a Bankart lesion, the mean force needed decreased to 53 6 17 N. Compared with the Bankart lesion, the mean force with a 2-mm glenoid defect was 41 6 21 N (P = .155) and decreased to 39 6 14 N (P = .089) with the addition of the Hill-Sachs lesion. There was a statistically significant decrease with a 4-mm glenoid defect (30 6 13 N; P = .003) and with a 6-mm glenoid defect (22 6 12 N; P \ .001). After repair of the Bankart lesion, the mean force to translate the humeral head 10 mm was 58 6 16 N. Compared with the Bankart repair, the mean force decreased with a 2-mm glenoid defect to 54 6 13 N (P = .446). With the addition of the Hills-Sachs lesion to the 2-mm glenoid defect, there was a statistically significant decrease (45 6 17 N; P = .048). This decrease continued with the repair and 4-mm glenoid defect (44 6 12 N; P = .037) and repair with a 6-mm glenoid defect (34 6 12 N; P = .001). This pattern of reduction is depicted in Figure 7. Engagement occurred at the end of translation testing and is reflected in the translation force data, which was especially pronounced at .4 mm of glenoid loss.

DISCUSSION The principal findings of this study demonstrate the potentiation effect of a combined bony glenoid deficiency and Hill-Sachs injury, even at relatively small amounts of GBL. Our results show that in patients with a moderate (medium-sized) Hill-Sachs defect (50th percentile), a glenoid defect of only 2 mm will significantly compromise a soft tissue Bankart repair alone. Even in a small Hill-Sachs lesion

Figure 7. Mean force to dislocate with the medium defect (1.47 cm3) and Bankart repair. Note progressive reduced load to translation with progressive with glenoid bone loss. Significant reduction in translation began at 2 mm of glenoid bone loss. Bk, Bankart; HS, Hill-Sachs; PC, posterior capsulotomy; rep, repair. (25th percentile), soft tissue Bankart repair was compromised with progressively larger glenoid defects. As the aim of this cadaveric study was to determine the effect of the combination of humeral head and GBL on glenohumeral joint translation in a bipolar bone loss model, these findings suggest that the combined effect on glenohumeral stability of a glenoid defect and Hill-Sachs injury are significant and so emphasize the importance of considering bony defects as a bipolar problem in the shoulder. Glenoid bone defects have been evaluated with the orientation of the defect obliquely located at the 3-o’clock position on a right shoulder.6 It is now well understood that in the majority of patients with GBL, the defect occurs at an angle that is parallel to the long axis of the glenoid. A large glenoid defect after dislocation results in an inverted pearshaped glenoid, which alone has been shown to compromise the soft tissue repair necessitating bony procedures to improve outcomes. In a cadaveric study, Itoi et al6 determined that a glenoid osseous defect with a width .20% of the glenoid remains unstable after Bankart lesion repair and suggested bone grafting to restore the glenoid concavity to restore stability. The issue of bone loss has significant clinical implications, as it is the number one reason for surgical failure.1,2 The lack of guidelines to improve surgical decision making and to consistently predict which bone defects (ie, those that have high rates of recurrent instability despite arthroscopic or open stabilization addressing only the soft tissue) is critical to improved patient outcomes. It should be recognized that the issue of bone loss is complex and the relationship of the glenoid defect to the humeral head lesion in a bipolar presentation is poorly understood.10,11 Yamamoto et al15 introduced the concept of the ‘‘glenoid track’’ to evaluate and model the effect of simultaneous-occurring glenoid and humeral head defects in an attempt to




determine which patient will require additional adjunctive stabilization techniques in addition to Bankart repair alone. Recently, Di Giacomo et al4 further described this concept and presented an algorithm for management based on the glenoid track measured from preoperative imaging or at arthroscopy. Although this concept provides a working model to suggest which patients require procedures such as coracoid transfer, bone grafting, and remplissage, it has not been validated by a biomechanical or clinical study. To our knowledge, there is no prior biomechanical study that has evaluated the combined effect of these lesions. Previous work by Itoi et al6 and Yamamoto et al16 suggested that a 20% to 25% glenoid lesion (~6 mm GBL) evaluated in isolation compromises soft tissue repair in a cadaveric anterior instability model. The findings of our study however, suggest that if a small- or medium-sized Hill-Sachs lesion is present, the amount of bone loss on the glenoid side required to compromise glenohumeral translation can be as little as 2 mm. In addition, we utilized a clinically relevant model with a large database of Hill-Sachs injuries modeled with 3D printer for exact cadaveric matching and applicability. The strength of this current study is the creation of clinically relevant Hill-Sachs lesions from actual patients and the evaluation of small- and medium-sized defects with known GBL. We aimed to make the location, size, and characteristics of these defects as clinically applicable as possible. Previous studies created humeral head defects by making wedge defects of various size off the radius of the humeral head; we, however, sought to model our defects from an actual clinical database. This database consisted of 142 recurrent anterior instability patients who underwent CT scans with 3D reconstructions. From those scans, the volume of the bony defects was calculated to facilitate the creation of a bell curve from which the 25th- and 50th-percentile volumes were determined. Using the 3D CT imaging of these patients, we created 3D plastic models that correlated to our defect sizes of interest, and we used the models to make our Hill-Sachs lesions on the cadavers. Given that prior studies showed consistency and clinical applicability in the creation of glenoid defects, we chose to model our defects from previous reported methods. To our knowledge, both the method of simulating the Hill-Sachs lesions and the size of our 3D CT database are unique to the literature. As with any biomechanical study, there are limitations. It is difficult to say with full certainty how these findings would translate into actual patients given the multitude of static, dynamic, and patient characteristics that factor in shoulder instability and its successful treatment. Second, the majority of our cadaver specimens consisted of elderly donors. While we ensured that they were free of arthritis, contractures, prior labrum, or cuff damage, we acknowledge that this is not the target population being treated for shoulder instability. Third, although the creations of the Hill-Sachs defects were more clinically oriented, we acknowledge the variability in creating these lesions. Because we did not create the defects in straight osteotomies in standard increments, such as the glenoid defects, we took measurements from a reproducible location to aid in consistency. To minimize variability, we

7

utilized a positive mold of PMMA to precisely model the defect to improve the ability to match the size and depth in each cadaveric specimen. In addition, this study demonstrated a novel application of a 3D printer to define bony injuries in a precise fashion, as opposed to prior studies that extrapolated data. In summary, our study investigated the combination of humeral head and GBL on glenohumeral joint translation in a novel bipolar bone loss model with a Bankart lesion and after Bankart repair. Our results support the notion that in cases of instability where there is combined glenoid and humeral head defects, the percentage of bone loss required for each before soft tissue Bankart repair is less than what has been reported in the literature for isolated defects. Itoi et al6 reported that a 6-mm glenoid defect, representing approximately 20% anterior bone loss, compromises a soft tissue repair. This amount is often quoted as the critical defect for the point at which surgeons should consider more than the standard soft tissue Bankart repair for satisfactory outcomes. The findings of this study suggest that in the presence of combined lesions, the actual amount of glenoid and humeral head bone loss may be much smaller, as little as 2 mm or an average of 8% GBL. The negative effect of these combined lesions for glenohumeral stability is additive. Combined lesions, even when small, may require surgical strategies that address bone defects to optimize outcomes.

CONCLUSION Combined glenoid and humeral head defects have an additive and negative effect on glenohumeral stability. As little as a 2-mm glenoid defect with a medium-sized Hill-Sachs lesion demonstrated a compromise in soft tissue Bankart repair, while a small-sized Hill-Sachs lesion showed compromise of soft tissue repair with 4 mm or more GBL. This has notable clinical implications, as bipolar bony lesions of the glenoid and humeral head occur frequently together in clinical practice. Surgeons should be aware that smaller combined defects and GBL of 2 to 4 mm, or approximately 8% to 15% of the glenoid, could compromise Bankart repair and thus may require surgical strategies in addition to traditional Bankart repair alone to optimize stability.

ACKNOWLEDGMENT The computed tomography scans and 3-dimensional modeling were obtained from 142 patients from the practice of M.T.P. REFERENCES 1. Balg F, Boileau P. The instability severity index score: a simple preoperative score to select patients for arthroscopic or open shoulder stabilisation. J Bone Joint Surg Br. 2007;89:1470-1477. 2. Burkhart S, 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 HillSachs lesion. Arthroscopy. 2000;16:677-694.


8

Arciero et al


3. Calandra JJ, Baker CL, Uribe J. The incidence of Hill-Sachs lesions in initial anterior shoulder dislocations. Arthroscopy. 1989;5(4): 254-257. 4. Di Giacomo G, Itoi EJ, Burkhart S. Evolving concept of bipolar bone loss and the Hill-Sachs lesion: from ‘‘engaging/non-engaging’’ lesion to’’ on-track/off-track’’ lesion. Arthroscopy. 2014;30(1):90-98. 5. Edwards TB, Boulahia A, Walch G. Radiographic analysis of bone defects in chronic anterior shoulder instability. Arthroscopy. 2003;19(7): 732-739. 6. 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. 7. Kaar SG, Fening SD, Jones MH. Effect of humeral head defect size on glenohumeral stability. Am J Sports Med. 2010;38(3):594-599. 8. Mihata T, Lee Y, McGarry MH, Abe M, Lee TQ. Excessive humeral external rotation results in increased shoulder laxity. Am J Sports Med. 2004;32(5):1278-1285. 9. Piasecki DR, Verma NN, Romeo AA, et al. Glenoid bone deficiency in recurrent anterior shoulder instability: diagnosis and management. J Am Acad Orthop Surg. 2009;17:482-493. 10. Provencher MT, Bhatia S, Ghodadra NS, et al. Recurrent shoulder instability: current concepts for evaluation and management of glenoid bone loss. J Bone Joint Surg Am. 2010;92(suppl 2): 133-151.

11. Provencher MT, Frank MR, LeClere LE, et al. The Hill-Sachs lesion: diagnosis, classification, and management. J Am Acad Orthop Surg. 2012;20:242-252. 12. Rowe CR, Zarins B. Recurrent transient subluxation of the shoulder. J Bone Joint Surg Am. 1981;63(6):665-668. 13. Saito H, Itoi E, Sugaya H, et al. Location of the glenoid defect in shoulders with recurrent anterior dislocation. Am J Sports Med. 2005;33:889-893. 14. 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. 1997;25(3):306-311. 15. 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. 16. Yamamoto N, Itoi EJ, Abe H, et al. Effect of an anterior glenoid defect on anterior shoulder stability. Am J Sports Med. 2009;37(5):949-954. 17. Yamamoto N, Muraki T, Sperling JW, et al. Does the bumper created during Bankart repair contribute to shoulder stability? J Shoulder Elbow Surg. 2013;22:828-834. 18. Yamamoto N, Muraki T, Sperling JW, et al. Stabilizing mechanism in bone-grafting of a large glenoid defect. J Bone Joint Surg Am. 2010;92:2059-2066.

For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav
