- Email: [email protected]
Nerve stress during reverse total shoulder arthroplasty: a cadaveric study
ARTICLE IN PRESS J Shoulder Elbow Surg (2016) ■■, ■■–■■
www.elsevier.com/locate/ymse
Nerve stress during reverse total shoulder arthroplasty: a cadaveric study Hubert Lenoir, MDa,*, Louis Dagneaux, MDb, François Canovas, PhDb,c, Thomas Waitzenegger, MDd, Thuy Trang Pham, MDe, Michel Chammas, PhDf a
Centre Ostéo-Articulaire des Cèdres, Echirolles, France Hip, Knee and Foot Surgery Unit, Centre Hospitalier Régional Universitaire Montpellier University Hospital, Montpellier, France c Laboratory of Anatomy, Montpellier 1 University, Montpellier, France d Clinique de l’Yvette, Longjumeau, France e Toulouse-Purpan University Hospital Center, Toulouse, France f Hand and Upper Extremity Surgery Unit, Centre Hospitalier Régional Universitaire Montpellier University Hospital, Montpellier, France b
Background: Neurologic lesions are relatively common after total shoulder arthroplasty. These injuries are mostly due to traction. We aimed to identify the arm manipulations and steps during reverse total shoulder arthroplasty (RTSA) that affect nerve stress. Methods: Stress was measured in 10 shoulders of 5 cadavers by use of a tensiometer on each nerve from the brachial plexus, with shoulders in different arm positions and during different surgical steps of RTSA. Results: When we studied shoulder position without prostheses, relative to the neutral position, internal rotation increased stress on the radial and axillary nerves and external rotation increased stress on the musculocutaneous, median, and ulnar nerves. Extension was correlated with increase in stress on all nerves. Abduction was correlated with increase in stress for the radial nerve. We identified 2 high-risk steps during RTSA: humeral exposition, particularly when the shoulder was in a position of more extension, and glenoid exposition. The thickness of polyethylene humeral cups used was associated with increased nerve stress in all but the ulnar nerve. Conclusion: During humeral preparation, the surgeon must be careful to limit shoulder extension. Care must be taken during exposure of the glenoid. Extreme rotation and oversized implants should be avoided to minimize stretch-induced neuropathies. Level of evidence: Basic Science Study; Biomechanics © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. Keywords: Nerve stress; brachial plexus injury; neurologic complications; reverse total shoulder arthroplasty; shoulder position; cadaveric study
This study was carried out with the agreement of the Laboratory of Anatomy of Montpellier, France. *Reprint requests: Hubert Lenoir, MD, Centre Ostéo-Articulaire des Cèdres, 5 Rue des Tropiques, F-38130 Echirolles, France. E-mail address: [email protected] (H. Lenoir).
Neurologic lesions in the context of total shoulder arthroplasty (TSA) are poorly documented.3,11,13,15 The incidence of these lesions was estimated at 1% in a literature review conducted by Bohsali et al.3 However, the vulnerability of the brachial plexus during the operation is probably
1058-2746/$ - see front matter © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. http://dx.doi.org/10.1016/j.jse.2016.07.020
ARTICLE IN PRESS 2 underestimated. In a systematic review, Lynch et al13 reported 18 shoulders with neurologic deficits during 417 shoulder arthroplasties (4%). Using continuous intraoperative nerve monitoring by electromyography, Nagda et al15 recorded nerve dysfunction in 17 of 30 patients. Nerve topography and its anatomic variations have been well specified to prevent iatrogenic risks.1,14 Nevertheless, some authors have emphasized that the nerve topography could change during range of motion of the shoulder and surgical procedures.2,4,5 These changes could increase the risk of nerve injury. Neurologic lesions during shoulder surgery are more likely due to indirect mechanisms.8 Nerves from the brachial plexus can undergo stretching when they reach their mobility limits. These phenomena were illustrated in studies assessing the adverse effects of patient postures under general anesthesia. Using finger palpation, Jackson and Keats9 subjectively evaluated the tension on the nerves of the brachial plexus in a cadaveric study. They highlighted the deleterious effects of different postures in abduction, external rotation, and extension. Coppieters et al7 showed that abduction and external rotation of the shoulder combined with elbow extension induced substantial discomfort for subjects. They concluded that this posture produced increasing strain on the brachial plexus. During TSA, some steps such as the preparation of the glenoid and humerus have been suggested to result in neurologic lesions.13,15 Moreover, reverse TSA (RTSA) is more prone to brachial plexus palsies, probably because of the lengthening effect.11,17 We aimed to evaluate variations in stress on nerves by using a tensiometer in the terminal branches of the brachial plexus in different arm positions of cadavers to highlight the effects of extension, abduction, and rotation. We also aimed to identify high-risk surgical steps that could lead to nerve stress during RTSA and the effect of joint lengthening.
Materials and methods Anatomic preparation We used 10 shoulders in 5 fresh-frozen cadavers (mean age at time of death, 77.6 years; range, 71-88 years). Cadavers were thawed at room temperature for 48 hours before experimentation to avoid softtissue stiffness. None of the upper limbs showed evidence of previous surgical procedures. Cadavers were placed in the beach-chair position, with the trunk at a 45° angle to the horizontal plane. The brachial plexus was exposed via an extensive deltopectoral approach combining a release of the pectoralis major from its humeral insertion and the pectoralis minor from the coracoid process to allow for positioning of a tensiometer (FK 50T; Sauter, Balingen, Germany). The 5 main branches of the brachial plexus to the arm (axillary, radial, median, musculocutaneous, and ulnar nerves) were carefully dissected distal to the brachial plexus cords. These branches were released from the periplexus adipose tissue and from minor adhesive attachments to the muscle to allow for positioning of the tensiometer to measure stress on each nerve (Fig. 1). No further release of the muscle insertion was necessary for positioning this
H. Lenoir et al.
Figure 1 Tensiometer measurement of stress on nerves of cadaveric shoulder undergoing reverse total shoulder arthroplasty.
device. Care was taken to limit dissection to the minimum necessary for placement of the tensiometer.
Effect of shoulder position in anatomic condition Before the RTSA procedure, we measured nerve stress in different shoulder positions (Table I). Extension was evaluated only with 60° of external rotation with reference to the positions conventionally used during humeral exposition. The other measured positions were those most likely to be used during glenoid exposure, starting with the neutral position of 0° of flexion, 0° of rotation, and 0° of abduction. Nerve tension was separately analyzed for extension, internal rotation, external rotation, and abduction to evaluate the role of each of these positions on nerve stress.
Effect of RTSA procedure A senior surgeon (H.L.) performed the RTSA procedure (Delta Xtend; DePuy Synthes, Raynham, MA, USA). Measurements were performed during 5 potential high-level stress steps (Fig. 2): In step 1, the retractor was placed between the anterior deltoid muscle and conjoined tendon with the shoulder in a neutral position. In step 2, dislocation of the humeral head was performed by a combination of adduction, maximal external rotation, and 45° of extension (humeral shaft vertically). In step 3, dislocation of the humeral head was performed by a combination of adduction, maximal external rotation, and 60° of extension. In step 4, glenoid exposure (after humeral head resection) was performed by using a forked retractor placed on the
ARTICLE IN PRESS Nerve stress and shoulder arthroplasty Table I Shoulder positions used to measure variation in nerve stress in anatomic condition No abduction—60° of external rotation—no extension No abduction—60° of external rotation—30° of extension No abduction—60° of external rotation—45° of extension No abduction—60° of external rotation—60° of extension No abduction—neutral rotation (ie, neutral position) No abduction—30° of external rotation No abduction—60° of external rotation No abduction—90° of external rotation No abduction—30° of internal rotation No abduction—60° of internal rotation No abduction—90° of internal rotation 30° of abduction—neutral rotation 30° of abduction—30° of external rotation 30° of abduction—60° of external rotation 30° of abduction—90° of external rotation 30° of abduction—30° of internal rotation 30° of abduction—60° of internal rotation 30° of abduction—90° of internal rotation 60° of abduction—neutral rotation 60° of abduction—30° of external rotation 60° of abduction—60° of external rotation 60° of abduction—90° of external rotation 60° of abduction—30° of internal rotation 60° of abduction—60° of internal rotation 60° of abduction—90° of internal rotation 90° of abduction—neutral rotation 90° of abduction—30° of external rotation 90° of abduction—60° of external rotation 90° of abduction—90° of external rotation 90° of abduction—30° of internal rotation 90° of abduction—60° of internal rotation 90° of abduction—90° of internal rotation
inferior glenoid rim to move the humeral head posteriorly and inferiorly. In step 5, definitive RTSA insertion was performed with 3 thicknesses of polyethylene humeral cups (+3, +6, and +9 mm) to evaluate the effect of lengthening with the shoulder in a neutral position. The glenosphere was placed centrally on the glenoid such that the inferior border would overhang inferiorly, and the humeral resection level was determined with the cutting assembly positioned with full contact with the superior humeral head.
Statistical analysis Analysis was performed using SPSS software (version 20.0.0; IBM, Armonk, NY, USA). The correlation between nerve stress and arm position in each anatomic plane or humeral cup thickness was analyzed by Spearman rank correlation coefficient analysis. The Wilcoxon signed rank test was used to compare nerve stress at 60° of external rotation with various degrees of extension and with various degrees of internal and external rotation, as well as abduction from the neutral position and during the proposed steps of RTSA versus the neutral position. We also compared step 2 with step 3. P < .05 was considered statistically significant.
3
Results Shoulder position in anatomic condition Extension (with 60° of external rotation) was significantly correlated with stress measurements for all nerves. Spearman correlation coefficients ranged from 0.353 (P = .0255) for the median nerve to 0.5787 (P = 9.177 × 10−5) for the radial nerve. Statistical significance in relation to external rotation without extension was observed for the axillary, radial, and median nerves (Fig. 3). Nerve stress was correlated with internal rotation on the radial and axillary nerves (ρ = 0.2841 [P = .0026] and ρ = 0.5989 [P = 5.81 × 10−11], respectively) and with external rotation on the median nerve (ρ = 0.2928 [P = .0012]). Figures 4 and 5 show the statistical significance of nerve stress depending on external rotation and internal rotation in relation to neutral rotation. Nerve stress was correlated with abduction on the radial nerve (ρ = 0.2343 [P = 7.54 × 10−5]). Figure 6 shows the statistical significance of nerve stress depending on various degrees of abduction in relation to neutral abduction.
RTSA surgical procedure Figure 7 shows nerve stress by surgical step for each nerve in relation to the neutral position. There was no increase in nerve stress with the retractor placed between the anterior deltoid muscle and conjoined tendon (step 1). Humeral exposition with 45° of extension (step 2) was not associated with a significant increase in nerve stress in relation to the neutral position. With an increase from 45° to 60° of extension (step 3), stress increased for all nerves (P < .05) except the axillary nerve (P = .2340) and the ulnar nerve (P = .0592). When we compared steps 2 and 3, stress increased for all nerves (P < .05) except the axillary nerve (P = .0519) and the ulnar nerve (P = .1275). During glenoid exposure (step 4), stress increased for all nerves (P < .05) except the median nerve (P = .0754) and the ulnar nerve (P = .1422). Joint lengthening with increasing polyethylene humeral cup thickness from 3 mm to 9 mm was significantly correlated with increased stress on the axillary nerve (ρ = 0.4574 [P = .0164]). Increased stress was not related for the other nerves using Spearman correlation coefficients. Figure 8 shows the statistical significance of nerve stress by cup thickness for each nerve in relation to the neutral position. With a 9-mm cup, stress increased for all nerves (P < .05) except the ulnar nerve (P = .0592). For the radial and musculocutaneous nerves, all cups increased stress (P < .05).
Discussion Knowledge of situations in which nerve stress increases during TSA is important to limit the risk of traction-induced nerve injuries. We aimed to understand stress variation during RTSA
ARTICLE IN PRESS 4
H. Lenoir et al.
Figure 2
Measurements performed during 5 potentially high-stress steps of reverse total shoulder arthroplasty.
Figure 3 Relation between shoulder extension and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 0° of extension (with 60° external rotation) (E0), 30° of extension (E30), 45° of extension (E45), and 60° of extension (E60). With 60° of extension, nerves stress increased for the axillary, radial, and median nerves. One asterisk indicates P < .05 and two asterisks indicate P < .01 in relation to 60° of external rotation without extension. Statistically significant results are shown in bold.
because the stretching mechanism is most often implicated in neurologic complications after TSA.13,15 Stress was measured in 10 shoulders of 5 cadavers by use of a tensiometer on each nerve from the brachial plexus, with the shoulders in different arm positions and during different surgical steps of RTSA. Extension was correlated with increased nerve stress on all nerves. In relation to neutral rotation, internal rotation statistically increased stress on the radial and axillary nerves, and external rotation statistically increased stress on the musculocutaneous, median, and ulnar nerves. Abduction increased stress on the radial nerve. During RTSA, stress was increased during humeral exposition when the shoulder was in a position of more extension, as well as during glenoid
exposition. Definitive RTSA insertion with a 9-mm cup was significantly associated with increased stress in relation to the neutral position on all nerves except the ulnar nerve. According to our data, the radial and musculocutaneous nerves appear to be most susceptible. In 1965, Jackson and Keats9 highlighted tension on the plexus in various positions. Tension was arbitrarily estimated by observation and palpation. In our study, we used a tensiometer for objective quantitative measurement. We prefer this device over miniature linear displacement transducers6 because it can be used to evaluate nerve lengthening, which is not perfectly correlated with nerve damage because of nerve elasticity.10,16
ARTICLE IN PRESS Nerve stress and shoulder arthroplasty
5
Figure 4 Relation between external rotation and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 30° of external rotation (ER 30), 60° of external rotation (ER 60), and 90° of external rotation (ER 90). Nerve stress increased only for the musculocutaneous, median, and ulnar nerves (with anterior course relative to humerus). One asterisk indicates P < .05, two asterisks indicate P < .01, and three asterisks indicate P < .005 in relation to neutral rotation (NR). Statistically significant results are shown in bold.
Figure 5 Relation between internal rotation and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 30° of internal rotation (IR 30), 60° of internal rotation (IR 60), and 90° of internal rotation (IR 90). Nerve stress increased for the axillary and radial nerves (with posterior course relative to humerus). Two asterisks indicate P < .01 and three asterisks indicate P < .005 in relation to neutral rotation (NR). Statistically significant results are shown in bold.
We found the axillary and radial nerves were often stretched. Previous in vivo studies have had similar findings.3 Along with the musculocutaneous nerve, these nerves are closer to surgical exposure. These 3 nerves were also particularly sensitive to lengthening. Compared with the median and ulnar nerves, their vulnerability is probably a result of different functional lengthening: Compared with the median and ulnar nerves, these 3 nerves are shorter (axillary nerve) or have a fixed point (radial nerve at the posterior aspect of the humerus, musculocutaneous nerve at the coracobrachial muscle) near the shoulder. Rotational shoulder motions affected stress on the brachial plexus branches. The axillary and radial nerves are sensitive to internal rotation, whereas the musculocutaneous, median, and ulnar nerves are sensitive to external rotation.
This difference by direction of rotation is probably explained by the difference in the course of these nerves: either anterior or posterior relative to the humerus. We found no significant increase in stress related to abduction except for the radial nerve. This finding is not consistent with results of Jackson and Keats9 or Coppieters et al.7 These authors emphasized the adverse effect of the combination of abduction, external rotation, and extension. Their results are probably not comparable to ours because they studied the effect of combined positions. We found a strong negative effect of extension with 60° of external rotation for all nerves. In our study, extension was always evaluated with external rotation. The protective effect of external rotation on the axillary and radial nerves was not substantial enough to protect against extension.
ARTICLE IN PRESS 6
H. Lenoir et al.
Figure 6 Relation between shoulder abduction and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 0° of abduction (Abd 0), 30° of abduction (Abd 30), 60° of abduction (Abd 60), and 90° of abduction (Abd 90). Nerve stress increased significantly for the radial nerve from 30° to 90° of abduction. There was also an increase in nerve stress for the musculocutaneous nerve with 60° of abduction and a decrease in nerve stress for the axillary nerve with 90° of abduction. One asterisk indicates P < .05, two asterisks indicate P < .01, and three asterisks indicate P < .005 in relation to 0° of abduction. Statistically significant results are shown in bold.
Figure 7 Nerve stress (in newtons) by surgical step. Nerves stress increased during humeral exposition but only with 60° of extension and during glenoid exposition. One asterisk indicates P < .05 and two asterisks indicate P < .01 in relation to the neutral position.
This ascertainment of the deleterious effect of extension is supported by the increase in nerve stress observed in comparing steps 2 and 3 of RTSA. Because the humeral shaft is often positioned in a vertical plane by the assistant to present it to the surgeon when reaming, placing patients in a position that is too supine could lead to excessive shoulder extension. We recommend limiting shoulder extension during this step, particularly when patients are in a more supine position (Fig. 9). Surprisingly, the axillary and ulnar nerves
were not stretched during humeral preparation. Nevertheless, when the effects of shoulder positioning were studied, extension was significantly correlated with deleterious effects for these nerves by use of Spearman correlation coefficients. In vivo, this may not necessarily be true because during the RTSA procedure, while performing the humeral reaming, the assistant surgeon naturally pushed up the humerus to correctly expose the humerus, which partially released the nerves.
ARTICLE IN PRESS Nerve stress and shoulder arthroplasty
7
Figure 8 Nerve stress (in newtons) by cup thickness. One asterisk indicates P < .05, two asterisks indicate P < .01, and three asterisks indicate P < .005 in relation to the neutral position (NP). Statistically significant results are shown in bold.
Figure 9 (A, B) Shoulder extension during humeral preparation. The surgeon should limit the shoulder extension as it tends to increase when the patient is in a more supine position.
Another deleterious step for the brachial plexus branches during RTSA was the glenoid exposure (step 4). Lynch et al13 presented results of 18 patients with neurologic damage after TSA. We also assumed, as they stated, that posterior translation of the humerus for glenoid preparation may stretch the plexus. We used a combination of posterior and inferior translation as necessary to achieve appropriate surgical exposure. The shoulder position was variable to provide the best visibility on the glenoid. The type of arthroplasty used is also an important factor regarding neurologic complications. Lädermann et al11 showed a higher incidence of subclinical lesions after RTSA as compared with anatomic TSA. The arm is lengthened by more than 2 cm with these implants.12 Lädermann et al concluded that this arm lengthening may be responsible for most nerve injuries. In performing a computed tomography scan before and after RTSA procedures on 2 cadavers, Van Hoof et al17 gave a good illustration of brachial plexus lengthening. All these reports agree with our results because we found correlations between polyethylene humeral cup thickness and stress. We used the Delta Xtend RTSA in this study, and we point out that differences in RTSA design could make our results not applicable to other reverse prostheses.
Limitations of this study include a large dissection and release of the brachial plexus from the periplexus adipose tissue and from minor adhesive attachments to the muscle that could underestimate the measurements. In addition, because of the tensiometer’s size, its introduction could slightly increase the amount of nerve stress. Because we performed a cadaveric study, the function of physiological muscle tension in protecting nerves was not evaluated. For all these reasons, determining an objective threshold for nerve stress is difficult to establish with mobilization of the shoulder or distalization of the humerus. Nagda et al15 performed an in vivo study using continuous intraoperative nerve monitoring by electromyography and found increased nerve dysfunction particularly during humeral and glenoid preparation. Our study is complementary because we defined which nerves are vulnerable and which surgical steps seem particularly vulnerable to lesions. However, intraoperative nerve monitoring during any step in the procedure seems difficult in current practice for several reasons: This monitoring requires a neurophysiologist, these measurements complicate patient setup, and curarization to provide muscular relaxation is not compatible with this monitoring. Finally, it would have been interesting to assess the influence of different
ARTICLE IN PRESS 8
H. Lenoir et al.
implant designs especially with a lateralized glenosphere; in the same way, we did not evaluate nerve stress with the superolateral approach.
Conclusions This study helps to understand the mechanisms involved in cases of nerve injuries during RTSA and highlights some recommendations to avoid traction neurapraxia during RTSA. During humeral preparation, arm extension should be limited. Care must be taken during exposure of the glenoid. Extreme rotation and oversized polyethylene implants should be avoided to minimize stretch-induced neuropathies.
Acknowledgments We are indebted to technicians from the University Laboratory of Anatomy at Medicine Faculty of Montpellier for their support in this study.
Disclaimer The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
References 1. Apaydin N, Uz A, Bozkurt M, Elhan A. The anatomic relationships of the axillary nerve and surgical landmarks for its localization from the anterior aspect of the shoulder. Clin Anat 2007;20:273-7. http:// dx.doi.org/10.1002/ca.20361 2. Bailie DS, Moseley B, Lowe WR. Surgical anatomy of the posterior shoulder: effects of arm position and anterior-inferior capsular shift. J Shoulder Elbow Surg 1999;8:307-13.
3. Bohsali KI, Wirth MA, Rockwood CA Jr. Complications of total shoulder arthroplasty. J Bone Joint Surg Am 2006;88:2279-92. http://dx.doi.org/ 10.2106/JBJS.F.00125 4. Burkhead WZ Jr, Scheinberg RR, Box G. Surgical anatomy of the axillary nerve. J Shoulder Elbow Surg 1992;1:31-6. 5. Cheung S, Fitzpatrick M, Lee TQ. Effects of shoulder position on axillary nerve positions during the split lateral deltoid approach. J Shoulder Elbow Surg 2009;18:748-55. http://dx.doi.org/10.1016/j.jse.2008.12.001 6. Coppieters MW. Shoulder restraints as a potential cause for stretch neuropathies: biomechanical support for the impact of shoulder girdle depression and arm abduction on nerve strain. Anesthesiology 2006;104:1351-2. 7. Coppieters MW, Van de Velde M, Stappaerts KH. Positioning in anesthesiology: toward a better understanding of stretch-induced perioperative neuropathies. Anesthesiology 2002;97:75-81. 8. Ho E, Cofield RH, Balm MR, Hattrup SJ, Rowland CM. Neurologic complications of surgery for anterior shoulder instability. J Shoulder Elbow Surg 1999;8:266-70. 9. Jackson L, Keats AS. Mechanism of brachial plexus palsy following anesthesia. Anesthesiology 1965;26:190-4. 10. Kwan MK, Wall EJ, Massie J, Garfin SR. Strain, stress and stretch of peripheral nerve. Rabbit experiments in vitro and in vivo. Acta Orthop Scand 1992;63:267-72. 11. Lädermann A, Lübbeke A, Mélis B, Stern R, Christofilopoulos P, Bacle G, et al. Prevalence of neurologic lesions after total shoulder arthroplasty. J Bone Joint Surg Am 2011;93:1288-93. http://dx.doi.org/10.2106/ JBJS.J.00369 12. Lädermann A, Williams MD, Melis B, Hoffmeyer P, Walch G. Objective evaluation of lengthening in reverse shoulder arthroplasty. J Shoulder Elbow Surg 2009;18:588-95. http://dx.doi.org/10.1016/j.jse.2009.03.012 13. Lynch NM, Cofield RH, Silbert PL, Hermann RC. Neurologic complications after total shoulder arthroplasty. J Shoulder Elbow Surg 1996;5:53-61. 14. McFarland EG, Caicedo JC, Guitterez MI, Sherbondy PS, Kim TK. The anatomic relationship of the brachial plexus and axillary artery to the glenoid. Implications for anterior shoulder surgery. Am J Sports Med 2001;29:729-33. 15. Nagda SH, Rogers KJ, Sestokas AK, Getz CL, Ramsey ML, Glaser DL, et al. Neer Award 2005: peripheral nerve function during shoulder arthroplasty using intraoperative nerve monitoring. J Shoulder Elbow Surg 2007;16(Suppl 3):S2-8. http://dx.doi.org/10.1016/j.jse.2006.01.016 16. Sinclair EB, Andarawis-Puri N, Ros SJ, Laudier DM, Jepsen KJ, Hausman MR. Relating applied strain to the type and severity of structural damage in the rat median nerve using second harmonic generation microscopy. Muscle Nerve 2012;46:899-907. http:// dx.doi.org/10.1002/mus.23443 17. Van Hoof T, Gomes GT, Audenaert E, Verstraete K, Kerckaert I, D’Herde K. 3D computerized model for measuring strain and displacement of the brachial plexus following placement of reverse shoulder prosthesis. Anat Rec (Hoboken) 2008;291:1173-85. http:// dx.doi.org/10.1002/ar.20735
www.elsevier.com/locate/ymse
Nerve stress during reverse total shoulder arthroplasty: a cadaveric study Hubert Lenoir, MDa,*, Louis Dagneaux, MDb, François Canovas, PhDb,c, Thomas Waitzenegger, MDd, Thuy Trang Pham, MDe, Michel Chammas, PhDf a
Centre Ostéo-Articulaire des Cèdres, Echirolles, France Hip, Knee and Foot Surgery Unit, Centre Hospitalier Régional Universitaire Montpellier University Hospital, Montpellier, France c Laboratory of Anatomy, Montpellier 1 University, Montpellier, France d Clinique de l’Yvette, Longjumeau, France e Toulouse-Purpan University Hospital Center, Toulouse, France f Hand and Upper Extremity Surgery Unit, Centre Hospitalier Régional Universitaire Montpellier University Hospital, Montpellier, France b
Background: Neurologic lesions are relatively common after total shoulder arthroplasty. These injuries are mostly due to traction. We aimed to identify the arm manipulations and steps during reverse total shoulder arthroplasty (RTSA) that affect nerve stress. Methods: Stress was measured in 10 shoulders of 5 cadavers by use of a tensiometer on each nerve from the brachial plexus, with shoulders in different arm positions and during different surgical steps of RTSA. Results: When we studied shoulder position without prostheses, relative to the neutral position, internal rotation increased stress on the radial and axillary nerves and external rotation increased stress on the musculocutaneous, median, and ulnar nerves. Extension was correlated with increase in stress on all nerves. Abduction was correlated with increase in stress for the radial nerve. We identified 2 high-risk steps during RTSA: humeral exposition, particularly when the shoulder was in a position of more extension, and glenoid exposition. The thickness of polyethylene humeral cups used was associated with increased nerve stress in all but the ulnar nerve. Conclusion: During humeral preparation, the surgeon must be careful to limit shoulder extension. Care must be taken during exposure of the glenoid. Extreme rotation and oversized implants should be avoided to minimize stretch-induced neuropathies. Level of evidence: Basic Science Study; Biomechanics © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. Keywords: Nerve stress; brachial plexus injury; neurologic complications; reverse total shoulder arthroplasty; shoulder position; cadaveric study
This study was carried out with the agreement of the Laboratory of Anatomy of Montpellier, France. *Reprint requests: Hubert Lenoir, MD, Centre Ostéo-Articulaire des Cèdres, 5 Rue des Tropiques, F-38130 Echirolles, France. E-mail address: [email protected] (H. Lenoir).
Neurologic lesions in the context of total shoulder arthroplasty (TSA) are poorly documented.3,11,13,15 The incidence of these lesions was estimated at 1% in a literature review conducted by Bohsali et al.3 However, the vulnerability of the brachial plexus during the operation is probably
1058-2746/$ - see front matter © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. http://dx.doi.org/10.1016/j.jse.2016.07.020
ARTICLE IN PRESS 2 underestimated. In a systematic review, Lynch et al13 reported 18 shoulders with neurologic deficits during 417 shoulder arthroplasties (4%). Using continuous intraoperative nerve monitoring by electromyography, Nagda et al15 recorded nerve dysfunction in 17 of 30 patients. Nerve topography and its anatomic variations have been well specified to prevent iatrogenic risks.1,14 Nevertheless, some authors have emphasized that the nerve topography could change during range of motion of the shoulder and surgical procedures.2,4,5 These changes could increase the risk of nerve injury. Neurologic lesions during shoulder surgery are more likely due to indirect mechanisms.8 Nerves from the brachial plexus can undergo stretching when they reach their mobility limits. These phenomena were illustrated in studies assessing the adverse effects of patient postures under general anesthesia. Using finger palpation, Jackson and Keats9 subjectively evaluated the tension on the nerves of the brachial plexus in a cadaveric study. They highlighted the deleterious effects of different postures in abduction, external rotation, and extension. Coppieters et al7 showed that abduction and external rotation of the shoulder combined with elbow extension induced substantial discomfort for subjects. They concluded that this posture produced increasing strain on the brachial plexus. During TSA, some steps such as the preparation of the glenoid and humerus have been suggested to result in neurologic lesions.13,15 Moreover, reverse TSA (RTSA) is more prone to brachial plexus palsies, probably because of the lengthening effect.11,17 We aimed to evaluate variations in stress on nerves by using a tensiometer in the terminal branches of the brachial plexus in different arm positions of cadavers to highlight the effects of extension, abduction, and rotation. We also aimed to identify high-risk surgical steps that could lead to nerve stress during RTSA and the effect of joint lengthening.
Materials and methods Anatomic preparation We used 10 shoulders in 5 fresh-frozen cadavers (mean age at time of death, 77.6 years; range, 71-88 years). Cadavers were thawed at room temperature for 48 hours before experimentation to avoid softtissue stiffness. None of the upper limbs showed evidence of previous surgical procedures. Cadavers were placed in the beach-chair position, with the trunk at a 45° angle to the horizontal plane. The brachial plexus was exposed via an extensive deltopectoral approach combining a release of the pectoralis major from its humeral insertion and the pectoralis minor from the coracoid process to allow for positioning of a tensiometer (FK 50T; Sauter, Balingen, Germany). The 5 main branches of the brachial plexus to the arm (axillary, radial, median, musculocutaneous, and ulnar nerves) were carefully dissected distal to the brachial plexus cords. These branches were released from the periplexus adipose tissue and from minor adhesive attachments to the muscle to allow for positioning of the tensiometer to measure stress on each nerve (Fig. 1). No further release of the muscle insertion was necessary for positioning this
H. Lenoir et al.
Figure 1 Tensiometer measurement of stress on nerves of cadaveric shoulder undergoing reverse total shoulder arthroplasty.
device. Care was taken to limit dissection to the minimum necessary for placement of the tensiometer.
Effect of shoulder position in anatomic condition Before the RTSA procedure, we measured nerve stress in different shoulder positions (Table I). Extension was evaluated only with 60° of external rotation with reference to the positions conventionally used during humeral exposition. The other measured positions were those most likely to be used during glenoid exposure, starting with the neutral position of 0° of flexion, 0° of rotation, and 0° of abduction. Nerve tension was separately analyzed for extension, internal rotation, external rotation, and abduction to evaluate the role of each of these positions on nerve stress.
Effect of RTSA procedure A senior surgeon (H.L.) performed the RTSA procedure (Delta Xtend; DePuy Synthes, Raynham, MA, USA). Measurements were performed during 5 potential high-level stress steps (Fig. 2): In step 1, the retractor was placed between the anterior deltoid muscle and conjoined tendon with the shoulder in a neutral position. In step 2, dislocation of the humeral head was performed by a combination of adduction, maximal external rotation, and 45° of extension (humeral shaft vertically). In step 3, dislocation of the humeral head was performed by a combination of adduction, maximal external rotation, and 60° of extension. In step 4, glenoid exposure (after humeral head resection) was performed by using a forked retractor placed on the
ARTICLE IN PRESS Nerve stress and shoulder arthroplasty Table I Shoulder positions used to measure variation in nerve stress in anatomic condition No abduction—60° of external rotation—no extension No abduction—60° of external rotation—30° of extension No abduction—60° of external rotation—45° of extension No abduction—60° of external rotation—60° of extension No abduction—neutral rotation (ie, neutral position) No abduction—30° of external rotation No abduction—60° of external rotation No abduction—90° of external rotation No abduction—30° of internal rotation No abduction—60° of internal rotation No abduction—90° of internal rotation 30° of abduction—neutral rotation 30° of abduction—30° of external rotation 30° of abduction—60° of external rotation 30° of abduction—90° of external rotation 30° of abduction—30° of internal rotation 30° of abduction—60° of internal rotation 30° of abduction—90° of internal rotation 60° of abduction—neutral rotation 60° of abduction—30° of external rotation 60° of abduction—60° of external rotation 60° of abduction—90° of external rotation 60° of abduction—30° of internal rotation 60° of abduction—60° of internal rotation 60° of abduction—90° of internal rotation 90° of abduction—neutral rotation 90° of abduction—30° of external rotation 90° of abduction—60° of external rotation 90° of abduction—90° of external rotation 90° of abduction—30° of internal rotation 90° of abduction—60° of internal rotation 90° of abduction—90° of internal rotation
inferior glenoid rim to move the humeral head posteriorly and inferiorly. In step 5, definitive RTSA insertion was performed with 3 thicknesses of polyethylene humeral cups (+3, +6, and +9 mm) to evaluate the effect of lengthening with the shoulder in a neutral position. The glenosphere was placed centrally on the glenoid such that the inferior border would overhang inferiorly, and the humeral resection level was determined with the cutting assembly positioned with full contact with the superior humeral head.
Statistical analysis Analysis was performed using SPSS software (version 20.0.0; IBM, Armonk, NY, USA). The correlation between nerve stress and arm position in each anatomic plane or humeral cup thickness was analyzed by Spearman rank correlation coefficient analysis. The Wilcoxon signed rank test was used to compare nerve stress at 60° of external rotation with various degrees of extension and with various degrees of internal and external rotation, as well as abduction from the neutral position and during the proposed steps of RTSA versus the neutral position. We also compared step 2 with step 3. P < .05 was considered statistically significant.
3
Results Shoulder position in anatomic condition Extension (with 60° of external rotation) was significantly correlated with stress measurements for all nerves. Spearman correlation coefficients ranged from 0.353 (P = .0255) for the median nerve to 0.5787 (P = 9.177 × 10−5) for the radial nerve. Statistical significance in relation to external rotation without extension was observed for the axillary, radial, and median nerves (Fig. 3). Nerve stress was correlated with internal rotation on the radial and axillary nerves (ρ = 0.2841 [P = .0026] and ρ = 0.5989 [P = 5.81 × 10−11], respectively) and with external rotation on the median nerve (ρ = 0.2928 [P = .0012]). Figures 4 and 5 show the statistical significance of nerve stress depending on external rotation and internal rotation in relation to neutral rotation. Nerve stress was correlated with abduction on the radial nerve (ρ = 0.2343 [P = 7.54 × 10−5]). Figure 6 shows the statistical significance of nerve stress depending on various degrees of abduction in relation to neutral abduction.
RTSA surgical procedure Figure 7 shows nerve stress by surgical step for each nerve in relation to the neutral position. There was no increase in nerve stress with the retractor placed between the anterior deltoid muscle and conjoined tendon (step 1). Humeral exposition with 45° of extension (step 2) was not associated with a significant increase in nerve stress in relation to the neutral position. With an increase from 45° to 60° of extension (step 3), stress increased for all nerves (P < .05) except the axillary nerve (P = .2340) and the ulnar nerve (P = .0592). When we compared steps 2 and 3, stress increased for all nerves (P < .05) except the axillary nerve (P = .0519) and the ulnar nerve (P = .1275). During glenoid exposure (step 4), stress increased for all nerves (P < .05) except the median nerve (P = .0754) and the ulnar nerve (P = .1422). Joint lengthening with increasing polyethylene humeral cup thickness from 3 mm to 9 mm was significantly correlated with increased stress on the axillary nerve (ρ = 0.4574 [P = .0164]). Increased stress was not related for the other nerves using Spearman correlation coefficients. Figure 8 shows the statistical significance of nerve stress by cup thickness for each nerve in relation to the neutral position. With a 9-mm cup, stress increased for all nerves (P < .05) except the ulnar nerve (P = .0592). For the radial and musculocutaneous nerves, all cups increased stress (P < .05).
Discussion Knowledge of situations in which nerve stress increases during TSA is important to limit the risk of traction-induced nerve injuries. We aimed to understand stress variation during RTSA
ARTICLE IN PRESS 4
H. Lenoir et al.
Figure 2
Measurements performed during 5 potentially high-stress steps of reverse total shoulder arthroplasty.
Figure 3 Relation between shoulder extension and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 0° of extension (with 60° external rotation) (E0), 30° of extension (E30), 45° of extension (E45), and 60° of extension (E60). With 60° of extension, nerves stress increased for the axillary, radial, and median nerves. One asterisk indicates P < .05 and two asterisks indicate P < .01 in relation to 60° of external rotation without extension. Statistically significant results are shown in bold.
because the stretching mechanism is most often implicated in neurologic complications after TSA.13,15 Stress was measured in 10 shoulders of 5 cadavers by use of a tensiometer on each nerve from the brachial plexus, with the shoulders in different arm positions and during different surgical steps of RTSA. Extension was correlated with increased nerve stress on all nerves. In relation to neutral rotation, internal rotation statistically increased stress on the radial and axillary nerves, and external rotation statistically increased stress on the musculocutaneous, median, and ulnar nerves. Abduction increased stress on the radial nerve. During RTSA, stress was increased during humeral exposition when the shoulder was in a position of more extension, as well as during glenoid
exposition. Definitive RTSA insertion with a 9-mm cup was significantly associated with increased stress in relation to the neutral position on all nerves except the ulnar nerve. According to our data, the radial and musculocutaneous nerves appear to be most susceptible. In 1965, Jackson and Keats9 highlighted tension on the plexus in various positions. Tension was arbitrarily estimated by observation and palpation. In our study, we used a tensiometer for objective quantitative measurement. We prefer this device over miniature linear displacement transducers6 because it can be used to evaluate nerve lengthening, which is not perfectly correlated with nerve damage because of nerve elasticity.10,16
ARTICLE IN PRESS Nerve stress and shoulder arthroplasty
5
Figure 4 Relation between external rotation and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 30° of external rotation (ER 30), 60° of external rotation (ER 60), and 90° of external rotation (ER 90). Nerve stress increased only for the musculocutaneous, median, and ulnar nerves (with anterior course relative to humerus). One asterisk indicates P < .05, two asterisks indicate P < .01, and three asterisks indicate P < .005 in relation to neutral rotation (NR). Statistically significant results are shown in bold.
Figure 5 Relation between internal rotation and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 30° of internal rotation (IR 30), 60° of internal rotation (IR 60), and 90° of internal rotation (IR 90). Nerve stress increased for the axillary and radial nerves (with posterior course relative to humerus). Two asterisks indicate P < .01 and three asterisks indicate P < .005 in relation to neutral rotation (NR). Statistically significant results are shown in bold.
We found the axillary and radial nerves were often stretched. Previous in vivo studies have had similar findings.3 Along with the musculocutaneous nerve, these nerves are closer to surgical exposure. These 3 nerves were also particularly sensitive to lengthening. Compared with the median and ulnar nerves, their vulnerability is probably a result of different functional lengthening: Compared with the median and ulnar nerves, these 3 nerves are shorter (axillary nerve) or have a fixed point (radial nerve at the posterior aspect of the humerus, musculocutaneous nerve at the coracobrachial muscle) near the shoulder. Rotational shoulder motions affected stress on the brachial plexus branches. The axillary and radial nerves are sensitive to internal rotation, whereas the musculocutaneous, median, and ulnar nerves are sensitive to external rotation.
This difference by direction of rotation is probably explained by the difference in the course of these nerves: either anterior or posterior relative to the humerus. We found no significant increase in stress related to abduction except for the radial nerve. This finding is not consistent with results of Jackson and Keats9 or Coppieters et al.7 These authors emphasized the adverse effect of the combination of abduction, external rotation, and extension. Their results are probably not comparable to ours because they studied the effect of combined positions. We found a strong negative effect of extension with 60° of external rotation for all nerves. In our study, extension was always evaluated with external rotation. The protective effect of external rotation on the axillary and radial nerves was not substantial enough to protect against extension.
ARTICLE IN PRESS 6
H. Lenoir et al.
Figure 6 Relation between shoulder abduction and nerve stress on the 5 terminal branches of the brachial plexus. Nerve stress (in newtons) was assessed at 0° of abduction (Abd 0), 30° of abduction (Abd 30), 60° of abduction (Abd 60), and 90° of abduction (Abd 90). Nerve stress increased significantly for the radial nerve from 30° to 90° of abduction. There was also an increase in nerve stress for the musculocutaneous nerve with 60° of abduction and a decrease in nerve stress for the axillary nerve with 90° of abduction. One asterisk indicates P < .05, two asterisks indicate P < .01, and three asterisks indicate P < .005 in relation to 0° of abduction. Statistically significant results are shown in bold.
Figure 7 Nerve stress (in newtons) by surgical step. Nerves stress increased during humeral exposition but only with 60° of extension and during glenoid exposition. One asterisk indicates P < .05 and two asterisks indicate P < .01 in relation to the neutral position.
This ascertainment of the deleterious effect of extension is supported by the increase in nerve stress observed in comparing steps 2 and 3 of RTSA. Because the humeral shaft is often positioned in a vertical plane by the assistant to present it to the surgeon when reaming, placing patients in a position that is too supine could lead to excessive shoulder extension. We recommend limiting shoulder extension during this step, particularly when patients are in a more supine position (Fig. 9). Surprisingly, the axillary and ulnar nerves
were not stretched during humeral preparation. Nevertheless, when the effects of shoulder positioning were studied, extension was significantly correlated with deleterious effects for these nerves by use of Spearman correlation coefficients. In vivo, this may not necessarily be true because during the RTSA procedure, while performing the humeral reaming, the assistant surgeon naturally pushed up the humerus to correctly expose the humerus, which partially released the nerves.
ARTICLE IN PRESS Nerve stress and shoulder arthroplasty
7
Figure 8 Nerve stress (in newtons) by cup thickness. One asterisk indicates P < .05, two asterisks indicate P < .01, and three asterisks indicate P < .005 in relation to the neutral position (NP). Statistically significant results are shown in bold.
Figure 9 (A, B) Shoulder extension during humeral preparation. The surgeon should limit the shoulder extension as it tends to increase when the patient is in a more supine position.
Another deleterious step for the brachial plexus branches during RTSA was the glenoid exposure (step 4). Lynch et al13 presented results of 18 patients with neurologic damage after TSA. We also assumed, as they stated, that posterior translation of the humerus for glenoid preparation may stretch the plexus. We used a combination of posterior and inferior translation as necessary to achieve appropriate surgical exposure. The shoulder position was variable to provide the best visibility on the glenoid. The type of arthroplasty used is also an important factor regarding neurologic complications. Lädermann et al11 showed a higher incidence of subclinical lesions after RTSA as compared with anatomic TSA. The arm is lengthened by more than 2 cm with these implants.12 Lädermann et al concluded that this arm lengthening may be responsible for most nerve injuries. In performing a computed tomography scan before and after RTSA procedures on 2 cadavers, Van Hoof et al17 gave a good illustration of brachial plexus lengthening. All these reports agree with our results because we found correlations between polyethylene humeral cup thickness and stress. We used the Delta Xtend RTSA in this study, and we point out that differences in RTSA design could make our results not applicable to other reverse prostheses.
Limitations of this study include a large dissection and release of the brachial plexus from the periplexus adipose tissue and from minor adhesive attachments to the muscle that could underestimate the measurements. In addition, because of the tensiometer’s size, its introduction could slightly increase the amount of nerve stress. Because we performed a cadaveric study, the function of physiological muscle tension in protecting nerves was not evaluated. For all these reasons, determining an objective threshold for nerve stress is difficult to establish with mobilization of the shoulder or distalization of the humerus. Nagda et al15 performed an in vivo study using continuous intraoperative nerve monitoring by electromyography and found increased nerve dysfunction particularly during humeral and glenoid preparation. Our study is complementary because we defined which nerves are vulnerable and which surgical steps seem particularly vulnerable to lesions. However, intraoperative nerve monitoring during any step in the procedure seems difficult in current practice for several reasons: This monitoring requires a neurophysiologist, these measurements complicate patient setup, and curarization to provide muscular relaxation is not compatible with this monitoring. Finally, it would have been interesting to assess the influence of different
ARTICLE IN PRESS 8
H. Lenoir et al.
implant designs especially with a lateralized glenosphere; in the same way, we did not evaluate nerve stress with the superolateral approach.
Conclusions This study helps to understand the mechanisms involved in cases of nerve injuries during RTSA and highlights some recommendations to avoid traction neurapraxia during RTSA. During humeral preparation, arm extension should be limited. Care must be taken during exposure of the glenoid. Extreme rotation and oversized polyethylene implants should be avoided to minimize stretch-induced neuropathies.
Acknowledgments We are indebted to technicians from the University Laboratory of Anatomy at Medicine Faculty of Montpellier for their support in this study.
Disclaimer The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
References 1. Apaydin N, Uz A, Bozkurt M, Elhan A. The anatomic relationships of the axillary nerve and surgical landmarks for its localization from the anterior aspect of the shoulder. Clin Anat 2007;20:273-7. http:// dx.doi.org/10.1002/ca.20361 2. Bailie DS, Moseley B, Lowe WR. Surgical anatomy of the posterior shoulder: effects of arm position and anterior-inferior capsular shift. J Shoulder Elbow Surg 1999;8:307-13.
3. Bohsali KI, Wirth MA, Rockwood CA Jr. Complications of total shoulder arthroplasty. J Bone Joint Surg Am 2006;88:2279-92. http://dx.doi.org/ 10.2106/JBJS.F.00125 4. Burkhead WZ Jr, Scheinberg RR, Box G. Surgical anatomy of the axillary nerve. J Shoulder Elbow Surg 1992;1:31-6. 5. Cheung S, Fitzpatrick M, Lee TQ. Effects of shoulder position on axillary nerve positions during the split lateral deltoid approach. J Shoulder Elbow Surg 2009;18:748-55. http://dx.doi.org/10.1016/j.jse.2008.12.001 6. Coppieters MW. Shoulder restraints as a potential cause for stretch neuropathies: biomechanical support for the impact of shoulder girdle depression and arm abduction on nerve strain. Anesthesiology 2006;104:1351-2. 7. Coppieters MW, Van de Velde M, Stappaerts KH. Positioning in anesthesiology: toward a better understanding of stretch-induced perioperative neuropathies. Anesthesiology 2002;97:75-81. 8. Ho E, Cofield RH, Balm MR, Hattrup SJ, Rowland CM. Neurologic complications of surgery for anterior shoulder instability. J Shoulder Elbow Surg 1999;8:266-70. 9. Jackson L, Keats AS. Mechanism of brachial plexus palsy following anesthesia. Anesthesiology 1965;26:190-4. 10. Kwan MK, Wall EJ, Massie J, Garfin SR. Strain, stress and stretch of peripheral nerve. Rabbit experiments in vitro and in vivo. Acta Orthop Scand 1992;63:267-72. 11. Lädermann A, Lübbeke A, Mélis B, Stern R, Christofilopoulos P, Bacle G, et al. Prevalence of neurologic lesions after total shoulder arthroplasty. J Bone Joint Surg Am 2011;93:1288-93. http://dx.doi.org/10.2106/ JBJS.J.00369 12. Lädermann A, Williams MD, Melis B, Hoffmeyer P, Walch G. Objective evaluation of lengthening in reverse shoulder arthroplasty. J Shoulder Elbow Surg 2009;18:588-95. http://dx.doi.org/10.1016/j.jse.2009.03.012 13. Lynch NM, Cofield RH, Silbert PL, Hermann RC. Neurologic complications after total shoulder arthroplasty. J Shoulder Elbow Surg 1996;5:53-61. 14. McFarland EG, Caicedo JC, Guitterez MI, Sherbondy PS, Kim TK. The anatomic relationship of the brachial plexus and axillary artery to the glenoid. Implications for anterior shoulder surgery. Am J Sports Med 2001;29:729-33. 15. Nagda SH, Rogers KJ, Sestokas AK, Getz CL, Ramsey ML, Glaser DL, et al. Neer Award 2005: peripheral nerve function during shoulder arthroplasty using intraoperative nerve monitoring. J Shoulder Elbow Surg 2007;16(Suppl 3):S2-8. http://dx.doi.org/10.1016/j.jse.2006.01.016 16. Sinclair EB, Andarawis-Puri N, Ros SJ, Laudier DM, Jepsen KJ, Hausman MR. Relating applied strain to the type and severity of structural damage in the rat median nerve using second harmonic generation microscopy. Muscle Nerve 2012;46:899-907. http:// dx.doi.org/10.1002/mus.23443 17. Van Hoof T, Gomes GT, Audenaert E, Verstraete K, Kerckaert I, D’Herde K. 3D computerized model for measuring strain and displacement of the brachial plexus following placement of reverse shoulder prosthesis. Anat Rec (Hoboken) 2008;291:1173-85. http:// dx.doi.org/10.1002/ar.20735