- Email: [email protected]
Abduction moment arm of transposed subscapularis tendon
Clinical Biomechanics 14 (1999) 265±270
Abduction moment arm of transposed subscapularis tendon Tomotaka Nakajima a, Jain Liu a,1, Richard E. Hughes a, Shawn O'Driscoll b, Kai-Nan An a,* a
Biomechanics Laboratory, Division of Orthopedic Research, Mayo Clinic and Mayo Foundation, 200 First Street S.W., Rochester, MN 55905, USA b Department of Orthopedics, Mayo Clinic and Mayo Foundation, Rochester, MN 55905, USA Received 30 September 1997; accepted 3 August 1998
Abstract Objective. The purpose of this study was to analyze the eects of the procedure of superior transposition of the subscapularis on the biomechanics of glenohumeral abduction. Design. The abduction moment arms of the subscapularis muscle for the normal attachment and transposed tendon were measured on 10 cadaver shoulders and compared to that for the normal supraspinatus tendon for which it is intended to substitute. Background. Superior transposition of the subscapularis tendon has been recommended for surgical repair of massive tears of the rotator cu, but the eect of this procedure on shoulder biomechanics has not been reported. Methods. The moment arm about an instantaneous center of rotation was derived, based on the slope of tendon excursionglenohumeral angle curve. To simulate the insertion of the transposed subscapularis tendon, pseudo-insertion sites were created. Results. Superior transposition of the subscapularis tendon signi®cantly increased its abduction moment arm. The eect was optimal when the simulated insertion site was lateral rather than medial and, to a lesser extent, anterior versus posterior. Conclusions. The results provided a biomechanical rationale for subscapularis tendon transposition in restoring the loss of abduction strength of the shoulder in a massive cu tear. Relevance Superior transposition of the subscapularis tendon has been recommended for surgical repair of massive tears of the rotator cu. The results from this study support the clinical practice of subscapularis tendon transposition in restoring the loss of abduction strength of the shoulder in a massive cu tear. Ó 1999 Elsevier Science Ltd. All rights reserved. Keywords: Rotator cu; Massive tear; Subscapularis tendon transposition; Abduction moment arm; Tendon excursion; Biomechanics; Shoulder joint
1. Introduction Some large rotator cu tears that expose the humeral head completely and following failures of prior repairs the cu tears cannot be repaired back to their original insertions. This is most likely with a `global tear' larger than 5 cm in diameter [1], or a `massive tear', which involves several rotator cu tendons [2,3]. There is no general agreement regarding surgical treatment of massive cu tears. McLaughlin proposed a *
Corresponding author. E-mail: [email protected]. Dr Jain Lui was primarily responsible for conceiving and conducting this research. Tragically, he was struck and killed by lightning in June 1995, soon after completing his fellowship in our laboratory. The co-authors wish to recognize his enthusiasm, intelligence, wit, and hard work. 1
longitudinal suture to a trough on the humeral head with appropriate tension [4]. Since debridement of a torn cu results in a high incidence of superior migration of the humeral head [5,6], patch grafts [3,7±11], pedicle advancement [4,12], and rotational ¯aps [13±17] of adjacent muscles or tendons have been advocated. However, such procedures do not predictably restore elevation or abduction strength when the arm is raised above the horizontal. Superior transposition of the upper 70% of the subscapularis tendon has been performed by many surgeons treating massive tears [2,18], but there are no data on the biomechanics of the transposed subscapularis tendon. The purpose of this study was to measure the abduction moment arm of the subscapularis tendon before and after superior transposition, and compare it to the moment arm of the supraspinatus tendon for which it is
0268-0033/99/$ ± see front matter Ó 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 9 8 ) 0 0 0 7 5 - 8
266
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
substituting. A further objective was to determine the optimum site for attaching the transposed subscapularis tendon. 2. Methods Five right and ®ve left fresh shoulders were harvested from four male and ®ve female cadavers (two shoulders were from the same specimen). The age at death ranged from 40 to 89 years (mean: 67 years). All specimens had full range of motion, no evidence of rotator cu tears, arthritis, or bony deformity. They were kept frozen at ÿ40° C and thawed at room temperature before examination. The acromioclavicular and glenohumeral joints were left intact. With the elbow ¯exed 90°, a ®berglass pin was inserted through a hole drilled in the humerus parallel to the ulna for reference. The humerus was osteotomized just distal to the deltoid insertion. The skin, subcutaneous tissues, and all muscles except the rotator cu were removed. The supraspinatus, infraspinatus, subscapularis, and teres minor muscles were separated from their origins at the scapula, and were transected 1 cm proximal to their musculotendinous junctions. The rotator cu, coracoacromial ligament, and joint capsule were preserved. Fifty-pound test nylon lines were sutured to the midpoints of the musculotendinous junctions of the supra-
spinatus and subscapularis tendons using the modi®ed Mason-Allen (Gerber) method [19]. Eyehooks were ®xed to the scapula at the mid-points of both muscle origins. The nylon line from each tendon passed through a corresponding eyehook to simulate the line of action of each respective muscle, and a 250-g weight hung from each line. The superior capsule was cut open from the superior labral margin to expose the humeral head. To simulate the insertion of the transposed subscapularis tendon, pseudo-insertion sites were created. They were arranged to permit comparison of anterior vs. posterior, and medial vs. lateral insertion sites. Each pseudo-insertion site consisted of a 30-pound test nylon line attached to a custom made acrylic suture anchor (3.25 mm in diameter and 10 mm in length) inserted into the humeral head (holes were predrilled using a 2.5-mm drill bit). The pins were arranged in three pairs from lateral to medial side (pair 1, pair 2 and pair 3). Pair 1, pair 2 and pair 3 were 3, 10 and 17 mm medial to the insertion of the joint capsule, respectively. Each pair had an anterior and posterior pin. The anterior pins constituted a row of three (row A) and posterior pins formed another row (row P). Row A corresponded to the mid-point of the greater tuberosity, and row P to its posterior edge (Fig. 1). A ®berglass rod was inserted into the humeral intramedullary canal and ®xed with acrylic pins and
Fig. 1. Insertion sites on the humeral head simulating superior transposition of the subscapularis tendon. The pins and test nylon lines were arranged at the anterior and posterior pseudo-insertion sites (row A and row P), and at the lateral, intermediate, and medial sites (pair 1, pair 2 and pair 3). Row A corresponded to the mid-point of the greater tuberosity; row P to its posterior edge. Pair 1 was placed 3 mm medial to the insertion of the joint capsule. Pair 2 was 10 mm medial, and pair 3 was 17 mm medial to the insertion. Ó 1999 Mayo Foundation. Published by Elsevier Science Ltd.
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
methylmethacrylate. The scapula was mounted vertically on a Plexiglas apparatus with plastic screws and cement. A Plexiglas guide plane was axed to the test table such that it was coplanar with the scapular plane. A sliding guide was fastened to the distal end of the intramedullary rod to control the abduction angle of the humerus along the Plexiglas guide plane (Fig.2). After mounting the scapula, all nylon lines were passed through their respective eyehooks and through holes in a Plexiglas guide panel to potentiometers (3500S-2-103, Bourns Corp., Riverside, CA; resistance tolerance 3% , linearity tolerance 0.2%). Each line was wound around a cylinder on a potentiometer shaft, and 250-g weights were hung on the ends of the strings to remove slack. The potentiometers were calibrated and checked before use to ensure that they did not reach their end points during testing. The potentiometers were powered by a 10 V power supply. A magnetic tracking device (3Space Isotrack, Polhemus, Colchester, VT) was used to measure the glenohumeral abduction angle. The measurement error of this system has previously been shown to be less than 1 mm displacement and 0.5° [20]. The `source' of the 3Space was attached to the Plexiglas table that did not interfere with the magnetic tracking, and the `sensor' was attached to the humeral shaft. A second `sensor' was used for digitization. The `scapular plane' was de®ned as the plane containing the center of the glenoid and the superior and inferior angles of the scapula. The humerus was abducted from 0° (hanging posture) to maximum abduction in the scapular plane in neutral rotation by weights pulling on the nylon lines and gentle passive assistance. The term `abduction' is used in the current study to describe elevation of the humerus in the scapular plane, and the abduction angle refers to
Fig. 2. Experimental set-up. The scapula was mounted vertically on a Plexiglas frame to which a Plexiglas guide plane was axed coplanar with the scapular plane. A sliding guide was fastened to the distal end of an intramedullary ®berglass rod in the humerus to control abduction angle of the humerus along the Plexiglas guide plane.
267
glenohumeral motion only. Three trials were performed for each test. Each trial was performed within 10 s. 3Space position data and potentiometer rotation data were sampled at 30 Hz. After testing, the joint was disarticulated. The pseudo-insertion sites, the contour of the humeral head, and glenoid were digitized with a second 3Space sensor. Tendon excursions were measured by rotation of the potentiometers, and the glenohumeral abduction angle was determined from the 3Space tracking data. The slope of tendon excursion-glenohumeral angle curve de®ned the moment arm about an instantaneous center of rotation of the humeral head [Fig. 3(a) and (b)]. The theoretical basis for this method has been described by An et al. (1984) [21] using the principle of virtual work: n X dX Q 0; T1 dX i1 where Ti is the tension in tendon i, X is the joint angle, X is the tendon excursion, and Q is the external moment. It is apparent from this equation that the moment arm is dX =dX, which is the slope of the excursion vs. joint
Fig. 3. (a) Comparison of calculated and measured tendon excursion vs. abduction angle. (b) Calculated abduction moment arm (e.g. slope of polynomial ®t of excursion data). The slope of tendon excursionglenohumeral angle curve was the moment arm about an instantaneous center of rotation of the humeral head. This slope was computed analytically by dierentiating a polynomial ®t to the excursion-angle data. The lowest-order polynomial satisfying the condition that the root mean square error be less than 0.5 mm was ®tted to model the relationship between the tendon excursion and abduction angle.
268
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
angle relationships. This slope was computed by analytically dierentiating a polynomial ®t to the excursionangle data. The lowest-order polynomial satisfying the condition that the root mean square error be less than 0.5 mm was ®tted to model the relationship between the tendon excursion and abduction angle. A three-way analysis of variance ANOVA model (SAS, SAS Institute, Cary, NC) was used to test the eect of abduction angle (at 10° increments) and pseudo-insertion sites on the abduction moment arms. Specimen was used as a blocking variable to account for interspecimen variability. Average moment arm values were computed across trials and subjects. 3. Results Superior transposition of the subscapularis tendon had a signi®cant eect on its average abduction moment arm (Fig. 4). The mean supraspinatus moment arm is also provided for comparison. The moment arms of the supraspinatus and subscapularis tendons varied continuously with glenohumeral abduction. The mean abduction moment arm of the supraspinatus tendon was greatest (26 mm) at approximately 30° of abduction. The normal subscapularis tendon, prior to transposition, had a small abduction moment arm (8 mm), peaking at 15° of abduction. This was very signi®cantly increased by simulating superior transposition of the subscapularis tendon. The average abduction moment arm of these sites peaked at 20 mm at about 30° of abduction.
Fig. 4. Average abduction moment arms (mm) for the supraspinatus and subscapularis tendons. The supraspinatus abduction moment arm (thin line) was maximal (26 mm) at 30° of abduction. Before superior transposition (dotted line), the subscapularis tendon had a small abduction moment arm, and it reached a peak (8 mm) at 151 of abduction. All pseudo-insertion sites simulating superior transposition of the subscapularis tendon (thick line) increased the abduction moment arm (average peak: 20 mm at approximately 30° of abduction) relative to the normal insertion. The standard deviation of the data is represented by brackets to the right.
The mean increase was signi®cant for all pseudo-insertion sites, although the eect varied signi®cantly from one site to another and at dierent abduction angles [P < 0:0001, Fig. 5(a) and (c)]. The abduction moment arm of the transposed subscapularis tendon was optimized by inserting laterally (vs. medially) and anteriorly (vs. posteriorly). This not only in¯uenced the peak abduction moment arms but also permitted them to occur at higher abduction angles. For example, in contrast to the average maximum value (18 mm) of the posterior insertion (row P) at 20° of abduction, the maximum abduction moment arm of the anterior insertion (row A) was 21 mm at approximately 30° of abduction [Fig. 5(a)]. The mean peak moment arm of the lateral site (pair 1) was 22 mm at 40° of abduction, although the medial site (pair 3) reached a mean maximum of 18 mm at 20° of abduction [Fig. 5(c)]. Our results can be explained by a geometric analysis. Superior transposition of the subscapularis tendon elevates the line of action above the instantaneous center of rotation, which produces a large abduction moment arm. However, the posterior (row P) and medial (pair 3) insertion sites may produce a line of action more oblique to the scapular plane [Fig. 5(b) and (d)]. Such obliquity would decrease the abduction moment arm and increase the extension moment arm, because the line of action would become more aligned with the abduction axis of rotation. 4. Discussion The present study con®rms that superior transposition of the subscapularis tendon increases its abduction moment arm. This result supports the clinical practice of transposing the subscapularis to compensate for a loss of abduction function of the supraspinatus tendon in a cu tear that cannot be repaired. Moreover, anterolateral locations for transposition may produce greater abduction moment arms that postero-medial locations on average. Lateral insertion (vs. medial) was more important than anterior insertion (vs. posterior), based on average moment arms. There are some limitations in the current study. The in¯uence of each part of the cu tendon on the overall excursion could not be identi®ed precisely in muscles with broad origins and insertions. The rotator cu has many oblique or transverse branches combining layers and tendons [22,23]. These morphological factors were neglected in this model because the origin and the line of action of each cu tendon were simulated by small eyehooks and nylon test lines, respectively. The other limitation was that in-vitro kinematics may not simulate in-vivo conditions. However, our results based on the 3Space magnetic tracking device showed that the superior±inferior translation of the geometric center of the
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
269
Fig. 5. (a) Abduction moment arms at anterior and posterior insertion sites (row A and row P). (b) Change in line of action of transposed subscapularis tendon to anterior and posterior insertions. Superior transposition of the subscapularis tendon elevates the line of action above the instantaneous center of rotation, which produces a large abduction moment arm. However, the posterior insertion sites may produce a line of action more oblique to the scapular plane. Such obliquity would decrease the abduction moment arm and increase the extension moment arm because the line of action would become more aligned with the abduction axis of rotation. The line of action is de®ned along the direction tangential to the ®bers at the joint line. (c) Abduction moment arms at lateral and medial insertion sites (pair 1 and pair 3). (d) Change in line of action of transposed subscapularis tendon to medial and lateral insertions. Superior transposition of the subscapularis tendon elevates the line of action above the instantaneous center of rotation, which produces a large abduction moment arm. However, the medial insertion sites may produce a line of action more oblique to the scapular plane. Such obliquity would decrease the abduction moment arm and increase the extension moment arm, because the line of action would become more aligned with the abduction axis of rotation. The line of action is de®ned along the direction tangential to the ®bers at the joint line.
humeral head averaged 3:7 1:1 mm. This was consistent with in-vivo humeral head translation reported by Poppen and Walker [24]. The subscapularis tendon is anatomically attached to the lesser tuberosity, and its line of action passes anteriorly to the center of rotation of the humeral head. Tendinous bands of this subscapularis muscle are evenly interspersed in the midportion and condensed laterally into a single, large, and ¯at tendon in its superior twothirds [25]. Biomechanically, the upper portion of the
subscapularis tendon has been reported to have a greater abduction moment arm in externally rotated postures of the humerus [26]. This is enhanced by superior transposition of the upper 70% of the subscapularis muscle/tendon complex [2,18], which might therefore be biomechanically bene®cial for massive cu tears. This would be consistent with Co®eld's original clinical report of the subscapularis tendon transposition, which gave not only satisfactory pain relief in more than 80% (22/26 cases) of patients, but also improvement in
270
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
active abduction in the scapular plane up to 120° or 130° [2]. Our results provide a possible explanation for the good postoperative results based on the mechanical advantage of the transposed subscapularis tendon. Acknowledgements This study was partially supported by National Institutes of Health Grant AR41171. References [1] Bateman JE. The diagnosis and treatment of ruptures of the rotator cu. Surg Clin N Am, 1963;43:1523±1530. [2] Co®eld RH. Subscapular muscle transposition for repair of chronic rotator cu tears. Surg Gynec Obstet, 1982;154:667±672. [3] Neviaser JS, Neviaser RJ, Neviaser TJ. The repair of chronic massive ruptures of the rotator cu of the shoulder by use of freeze-dried rotator cu. J Bone Joint Surg, 1978;60(A):681±684. [4] McLaughlin HL. Repair of major cu ruptures. Surg Clin N Am, 1963;431535±1540. [5] Apoil A, Dautry P, Moinet P, Koechlin P. Le syndrome dit `de rupture de la coie des rotateurs de l'`Epaule' ± A propos de 70 observations. Rev Chir Orthop, 1977; 63 (Suppl. II):145. [6] Rockwood CA, Jr. Function following decompression of the shoulder and debridement of the rotator cu. Presented at the 52nd Annual Meeting of the AAOS, Las Vegas, January 27, 1985. [7] Bateman JE. The shoulder and neck, 2nd edn. Philadelphia: W.B. Saunders, 1978. [8] Bush LF. The torn shoulder capsule. J Bone Joint Surg, 1975;57(A):256±259. [9] Heikel HVA. Rupture of the rotator cu of the shoulder: experiences of surgical treatment. Acta Orthop Scand, 1968;39:477±492. [10] Neviaser JS. Ruptures of the rotator cu of the shoulder. New concepts in the diagnosis and operative treatment for chronic ruptures. Arch Surg, 1971;102:483±485. [11] Ozaki J, Fujimoto S, Masuhara K. Repair of chronic massive rotator cu tears with synthetic fabrics. In: Bateman JE, Welsh RP, editors. Surgery of the shoulder. Toronto: B.C. Decker, 1984:185±91.
[12] Debeyre J, Patte D, Elmelik E. Repair of ruptures of the rotator cu of the shoulder: with a note on advancement of the supraspinatus muscle. J Bone Joint Surg, 1965;47B:36±42. [13] Augereau B, Apoil A. La reparation des grandes ruptures de la coie des rotateurs de l'epaule par lambeau de deltoide. Livre des resumes 62e reunion annuelle de la SOFCOT, Paris, 1987; 10 (Nov.):10. [14] Gerber C. Latissimus dorsi transfer for the treatment of irreparable tears of the rotator cu. Clin Orthop, 1992;275:152±160. [15] Mikasa M. Trapezius transfer for global tear of the rotator cu. In: Bateman JE, Welsh RP, editors. Surgery of the Shoulder. Toronto: B.C. Decker, 1984:196±9. [16] Neviaser RJ, Neviaser TJ. Transfer of subscapularis and teres minor for massive defects of rotator cu. In: Bayley JI, Kessel L, editors. Shoulder surgery. Berlin: Springer, 1982:60±3. [17] Takagishi N, Okabe Y, Matsuzaki A. et al. Treatment of the rotator cu tear. J Jpn Orthop Assoc 1975;49:698±9 (in Japanese). [18] Neer CS, II. Cu tears, biceps lesions, and impingement. In: Shoulder reconstruction. Philadelphia PA: W.B. Saunders, 1990:63±77;93±134. [19] Gerber C, Schneeberger AG, Beck M, Schlegel U. Mechanical strength of repairs of the rotator cu. J Bone Joint Surg, 1994;76B:371±380. [20] An KN, Jacobsen MC, Berglund LJ, Chao EYS. Application of a magnetic tracking device to kinesiologic studies. J Biomech, 1988;21:613±620. [21] An KN, Ueba Y, Chao EYS, Cooney WP II, Linscheid RL. Tendon excursion and moment arm of index ®nger muscles. J Biomech, 1983;16:419±425. [22] Clark JM, Harryman DT II. Tendons, ligaments, and capsule of the rotator cu: gross and microscopic anatomy. J Bone Joint Surg, 1992;74A:713±725. [23] Nakajima T, Rokuuma N, Hamada K, Tomatsu T, Fukuda H. Histologic and biomechanical characteristics of the supraspinatus tendon: reference to rotator cu tearing. J Shoulder Elbow Surg, 1994;3:79±87. [24] Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. J Bone Joint Surg, 1976;58A:195±200. [25] Klapper RC, Jobe FW, Matsuura P. The subscapularis muscle and its glenohumeral ligament-like bands: a histomorphological study. Am J Sports Med, 1992;20:307±310. [26] Otis JC, Jiang CC, Wickiewicz TL, Peterson MG, Warren RF, Santner TJ. Changes in the moment arms of the rotator cu and deltoid muscles with abduction and rotation. J Bone Joint Surg, 1994;76(A):667±676.
Abduction moment arm of transposed subscapularis tendon Tomotaka Nakajima a, Jain Liu a,1, Richard E. Hughes a, Shawn O'Driscoll b, Kai-Nan An a,* a
Biomechanics Laboratory, Division of Orthopedic Research, Mayo Clinic and Mayo Foundation, 200 First Street S.W., Rochester, MN 55905, USA b Department of Orthopedics, Mayo Clinic and Mayo Foundation, Rochester, MN 55905, USA Received 30 September 1997; accepted 3 August 1998
Abstract Objective. The purpose of this study was to analyze the eects of the procedure of superior transposition of the subscapularis on the biomechanics of glenohumeral abduction. Design. The abduction moment arms of the subscapularis muscle for the normal attachment and transposed tendon were measured on 10 cadaver shoulders and compared to that for the normal supraspinatus tendon for which it is intended to substitute. Background. Superior transposition of the subscapularis tendon has been recommended for surgical repair of massive tears of the rotator cu, but the eect of this procedure on shoulder biomechanics has not been reported. Methods. The moment arm about an instantaneous center of rotation was derived, based on the slope of tendon excursionglenohumeral angle curve. To simulate the insertion of the transposed subscapularis tendon, pseudo-insertion sites were created. Results. Superior transposition of the subscapularis tendon signi®cantly increased its abduction moment arm. The eect was optimal when the simulated insertion site was lateral rather than medial and, to a lesser extent, anterior versus posterior. Conclusions. The results provided a biomechanical rationale for subscapularis tendon transposition in restoring the loss of abduction strength of the shoulder in a massive cu tear. Relevance Superior transposition of the subscapularis tendon has been recommended for surgical repair of massive tears of the rotator cu. The results from this study support the clinical practice of subscapularis tendon transposition in restoring the loss of abduction strength of the shoulder in a massive cu tear. Ó 1999 Elsevier Science Ltd. All rights reserved. Keywords: Rotator cu; Massive tear; Subscapularis tendon transposition; Abduction moment arm; Tendon excursion; Biomechanics; Shoulder joint
1. Introduction Some large rotator cu tears that expose the humeral head completely and following failures of prior repairs the cu tears cannot be repaired back to their original insertions. This is most likely with a `global tear' larger than 5 cm in diameter [1], or a `massive tear', which involves several rotator cu tendons [2,3]. There is no general agreement regarding surgical treatment of massive cu tears. McLaughlin proposed a *
Corresponding author. E-mail: [email protected]. Dr Jain Lui was primarily responsible for conceiving and conducting this research. Tragically, he was struck and killed by lightning in June 1995, soon after completing his fellowship in our laboratory. The co-authors wish to recognize his enthusiasm, intelligence, wit, and hard work. 1
longitudinal suture to a trough on the humeral head with appropriate tension [4]. Since debridement of a torn cu results in a high incidence of superior migration of the humeral head [5,6], patch grafts [3,7±11], pedicle advancement [4,12], and rotational ¯aps [13±17] of adjacent muscles or tendons have been advocated. However, such procedures do not predictably restore elevation or abduction strength when the arm is raised above the horizontal. Superior transposition of the upper 70% of the subscapularis tendon has been performed by many surgeons treating massive tears [2,18], but there are no data on the biomechanics of the transposed subscapularis tendon. The purpose of this study was to measure the abduction moment arm of the subscapularis tendon before and after superior transposition, and compare it to the moment arm of the supraspinatus tendon for which it is
0268-0033/99/$ ± see front matter Ó 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 9 8 ) 0 0 0 7 5 - 8
266
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
substituting. A further objective was to determine the optimum site for attaching the transposed subscapularis tendon. 2. Methods Five right and ®ve left fresh shoulders were harvested from four male and ®ve female cadavers (two shoulders were from the same specimen). The age at death ranged from 40 to 89 years (mean: 67 years). All specimens had full range of motion, no evidence of rotator cu tears, arthritis, or bony deformity. They were kept frozen at ÿ40° C and thawed at room temperature before examination. The acromioclavicular and glenohumeral joints were left intact. With the elbow ¯exed 90°, a ®berglass pin was inserted through a hole drilled in the humerus parallel to the ulna for reference. The humerus was osteotomized just distal to the deltoid insertion. The skin, subcutaneous tissues, and all muscles except the rotator cu were removed. The supraspinatus, infraspinatus, subscapularis, and teres minor muscles were separated from their origins at the scapula, and were transected 1 cm proximal to their musculotendinous junctions. The rotator cu, coracoacromial ligament, and joint capsule were preserved. Fifty-pound test nylon lines were sutured to the midpoints of the musculotendinous junctions of the supra-
spinatus and subscapularis tendons using the modi®ed Mason-Allen (Gerber) method [19]. Eyehooks were ®xed to the scapula at the mid-points of both muscle origins. The nylon line from each tendon passed through a corresponding eyehook to simulate the line of action of each respective muscle, and a 250-g weight hung from each line. The superior capsule was cut open from the superior labral margin to expose the humeral head. To simulate the insertion of the transposed subscapularis tendon, pseudo-insertion sites were created. They were arranged to permit comparison of anterior vs. posterior, and medial vs. lateral insertion sites. Each pseudo-insertion site consisted of a 30-pound test nylon line attached to a custom made acrylic suture anchor (3.25 mm in diameter and 10 mm in length) inserted into the humeral head (holes were predrilled using a 2.5-mm drill bit). The pins were arranged in three pairs from lateral to medial side (pair 1, pair 2 and pair 3). Pair 1, pair 2 and pair 3 were 3, 10 and 17 mm medial to the insertion of the joint capsule, respectively. Each pair had an anterior and posterior pin. The anterior pins constituted a row of three (row A) and posterior pins formed another row (row P). Row A corresponded to the mid-point of the greater tuberosity, and row P to its posterior edge (Fig. 1). A ®berglass rod was inserted into the humeral intramedullary canal and ®xed with acrylic pins and
Fig. 1. Insertion sites on the humeral head simulating superior transposition of the subscapularis tendon. The pins and test nylon lines were arranged at the anterior and posterior pseudo-insertion sites (row A and row P), and at the lateral, intermediate, and medial sites (pair 1, pair 2 and pair 3). Row A corresponded to the mid-point of the greater tuberosity; row P to its posterior edge. Pair 1 was placed 3 mm medial to the insertion of the joint capsule. Pair 2 was 10 mm medial, and pair 3 was 17 mm medial to the insertion. Ó 1999 Mayo Foundation. Published by Elsevier Science Ltd.
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
methylmethacrylate. The scapula was mounted vertically on a Plexiglas apparatus with plastic screws and cement. A Plexiglas guide plane was axed to the test table such that it was coplanar with the scapular plane. A sliding guide was fastened to the distal end of the intramedullary rod to control the abduction angle of the humerus along the Plexiglas guide plane (Fig.2). After mounting the scapula, all nylon lines were passed through their respective eyehooks and through holes in a Plexiglas guide panel to potentiometers (3500S-2-103, Bourns Corp., Riverside, CA; resistance tolerance 3% , linearity tolerance 0.2%). Each line was wound around a cylinder on a potentiometer shaft, and 250-g weights were hung on the ends of the strings to remove slack. The potentiometers were calibrated and checked before use to ensure that they did not reach their end points during testing. The potentiometers were powered by a 10 V power supply. A magnetic tracking device (3Space Isotrack, Polhemus, Colchester, VT) was used to measure the glenohumeral abduction angle. The measurement error of this system has previously been shown to be less than 1 mm displacement and 0.5° [20]. The `source' of the 3Space was attached to the Plexiglas table that did not interfere with the magnetic tracking, and the `sensor' was attached to the humeral shaft. A second `sensor' was used for digitization. The `scapular plane' was de®ned as the plane containing the center of the glenoid and the superior and inferior angles of the scapula. The humerus was abducted from 0° (hanging posture) to maximum abduction in the scapular plane in neutral rotation by weights pulling on the nylon lines and gentle passive assistance. The term `abduction' is used in the current study to describe elevation of the humerus in the scapular plane, and the abduction angle refers to
Fig. 2. Experimental set-up. The scapula was mounted vertically on a Plexiglas frame to which a Plexiglas guide plane was axed coplanar with the scapular plane. A sliding guide was fastened to the distal end of an intramedullary ®berglass rod in the humerus to control abduction angle of the humerus along the Plexiglas guide plane.
267
glenohumeral motion only. Three trials were performed for each test. Each trial was performed within 10 s. 3Space position data and potentiometer rotation data were sampled at 30 Hz. After testing, the joint was disarticulated. The pseudo-insertion sites, the contour of the humeral head, and glenoid were digitized with a second 3Space sensor. Tendon excursions were measured by rotation of the potentiometers, and the glenohumeral abduction angle was determined from the 3Space tracking data. The slope of tendon excursion-glenohumeral angle curve de®ned the moment arm about an instantaneous center of rotation of the humeral head [Fig. 3(a) and (b)]. The theoretical basis for this method has been described by An et al. (1984) [21] using the principle of virtual work: n X dX Q 0; T1 dX i1 where Ti is the tension in tendon i, X is the joint angle, X is the tendon excursion, and Q is the external moment. It is apparent from this equation that the moment arm is dX =dX, which is the slope of the excursion vs. joint
Fig. 3. (a) Comparison of calculated and measured tendon excursion vs. abduction angle. (b) Calculated abduction moment arm (e.g. slope of polynomial ®t of excursion data). The slope of tendon excursionglenohumeral angle curve was the moment arm about an instantaneous center of rotation of the humeral head. This slope was computed analytically by dierentiating a polynomial ®t to the excursion-angle data. The lowest-order polynomial satisfying the condition that the root mean square error be less than 0.5 mm was ®tted to model the relationship between the tendon excursion and abduction angle.
268
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
angle relationships. This slope was computed by analytically dierentiating a polynomial ®t to the excursionangle data. The lowest-order polynomial satisfying the condition that the root mean square error be less than 0.5 mm was ®tted to model the relationship between the tendon excursion and abduction angle. A three-way analysis of variance ANOVA model (SAS, SAS Institute, Cary, NC) was used to test the eect of abduction angle (at 10° increments) and pseudo-insertion sites on the abduction moment arms. Specimen was used as a blocking variable to account for interspecimen variability. Average moment arm values were computed across trials and subjects. 3. Results Superior transposition of the subscapularis tendon had a signi®cant eect on its average abduction moment arm (Fig. 4). The mean supraspinatus moment arm is also provided for comparison. The moment arms of the supraspinatus and subscapularis tendons varied continuously with glenohumeral abduction. The mean abduction moment arm of the supraspinatus tendon was greatest (26 mm) at approximately 30° of abduction. The normal subscapularis tendon, prior to transposition, had a small abduction moment arm (8 mm), peaking at 15° of abduction. This was very signi®cantly increased by simulating superior transposition of the subscapularis tendon. The average abduction moment arm of these sites peaked at 20 mm at about 30° of abduction.
Fig. 4. Average abduction moment arms (mm) for the supraspinatus and subscapularis tendons. The supraspinatus abduction moment arm (thin line) was maximal (26 mm) at 30° of abduction. Before superior transposition (dotted line), the subscapularis tendon had a small abduction moment arm, and it reached a peak (8 mm) at 151 of abduction. All pseudo-insertion sites simulating superior transposition of the subscapularis tendon (thick line) increased the abduction moment arm (average peak: 20 mm at approximately 30° of abduction) relative to the normal insertion. The standard deviation of the data is represented by brackets to the right.
The mean increase was signi®cant for all pseudo-insertion sites, although the eect varied signi®cantly from one site to another and at dierent abduction angles [P < 0:0001, Fig. 5(a) and (c)]. The abduction moment arm of the transposed subscapularis tendon was optimized by inserting laterally (vs. medially) and anteriorly (vs. posteriorly). This not only in¯uenced the peak abduction moment arms but also permitted them to occur at higher abduction angles. For example, in contrast to the average maximum value (18 mm) of the posterior insertion (row P) at 20° of abduction, the maximum abduction moment arm of the anterior insertion (row A) was 21 mm at approximately 30° of abduction [Fig. 5(a)]. The mean peak moment arm of the lateral site (pair 1) was 22 mm at 40° of abduction, although the medial site (pair 3) reached a mean maximum of 18 mm at 20° of abduction [Fig. 5(c)]. Our results can be explained by a geometric analysis. Superior transposition of the subscapularis tendon elevates the line of action above the instantaneous center of rotation, which produces a large abduction moment arm. However, the posterior (row P) and medial (pair 3) insertion sites may produce a line of action more oblique to the scapular plane [Fig. 5(b) and (d)]. Such obliquity would decrease the abduction moment arm and increase the extension moment arm, because the line of action would become more aligned with the abduction axis of rotation. 4. Discussion The present study con®rms that superior transposition of the subscapularis tendon increases its abduction moment arm. This result supports the clinical practice of transposing the subscapularis to compensate for a loss of abduction function of the supraspinatus tendon in a cu tear that cannot be repaired. Moreover, anterolateral locations for transposition may produce greater abduction moment arms that postero-medial locations on average. Lateral insertion (vs. medial) was more important than anterior insertion (vs. posterior), based on average moment arms. There are some limitations in the current study. The in¯uence of each part of the cu tendon on the overall excursion could not be identi®ed precisely in muscles with broad origins and insertions. The rotator cu has many oblique or transverse branches combining layers and tendons [22,23]. These morphological factors were neglected in this model because the origin and the line of action of each cu tendon were simulated by small eyehooks and nylon test lines, respectively. The other limitation was that in-vitro kinematics may not simulate in-vivo conditions. However, our results based on the 3Space magnetic tracking device showed that the superior±inferior translation of the geometric center of the
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
269
Fig. 5. (a) Abduction moment arms at anterior and posterior insertion sites (row A and row P). (b) Change in line of action of transposed subscapularis tendon to anterior and posterior insertions. Superior transposition of the subscapularis tendon elevates the line of action above the instantaneous center of rotation, which produces a large abduction moment arm. However, the posterior insertion sites may produce a line of action more oblique to the scapular plane. Such obliquity would decrease the abduction moment arm and increase the extension moment arm because the line of action would become more aligned with the abduction axis of rotation. The line of action is de®ned along the direction tangential to the ®bers at the joint line. (c) Abduction moment arms at lateral and medial insertion sites (pair 1 and pair 3). (d) Change in line of action of transposed subscapularis tendon to medial and lateral insertions. Superior transposition of the subscapularis tendon elevates the line of action above the instantaneous center of rotation, which produces a large abduction moment arm. However, the medial insertion sites may produce a line of action more oblique to the scapular plane. Such obliquity would decrease the abduction moment arm and increase the extension moment arm, because the line of action would become more aligned with the abduction axis of rotation. The line of action is de®ned along the direction tangential to the ®bers at the joint line.
humeral head averaged 3:7 1:1 mm. This was consistent with in-vivo humeral head translation reported by Poppen and Walker [24]. The subscapularis tendon is anatomically attached to the lesser tuberosity, and its line of action passes anteriorly to the center of rotation of the humeral head. Tendinous bands of this subscapularis muscle are evenly interspersed in the midportion and condensed laterally into a single, large, and ¯at tendon in its superior twothirds [25]. Biomechanically, the upper portion of the
subscapularis tendon has been reported to have a greater abduction moment arm in externally rotated postures of the humerus [26]. This is enhanced by superior transposition of the upper 70% of the subscapularis muscle/tendon complex [2,18], which might therefore be biomechanically bene®cial for massive cu tears. This would be consistent with Co®eld's original clinical report of the subscapularis tendon transposition, which gave not only satisfactory pain relief in more than 80% (22/26 cases) of patients, but also improvement in
270
T. Nakajima et al. / Clinical Biomechanics 14 (1999) 265±270
active abduction in the scapular plane up to 120° or 130° [2]. Our results provide a possible explanation for the good postoperative results based on the mechanical advantage of the transposed subscapularis tendon. Acknowledgements This study was partially supported by National Institutes of Health Grant AR41171. References [1] Bateman JE. The diagnosis and treatment of ruptures of the rotator cu. Surg Clin N Am, 1963;43:1523±1530. [2] Co®eld RH. Subscapular muscle transposition for repair of chronic rotator cu tears. Surg Gynec Obstet, 1982;154:667±672. [3] Neviaser JS, Neviaser RJ, Neviaser TJ. The repair of chronic massive ruptures of the rotator cu of the shoulder by use of freeze-dried rotator cu. J Bone Joint Surg, 1978;60(A):681±684. [4] McLaughlin HL. Repair of major cu ruptures. Surg Clin N Am, 1963;431535±1540. [5] Apoil A, Dautry P, Moinet P, Koechlin P. Le syndrome dit `de rupture de la coie des rotateurs de l'`Epaule' ± A propos de 70 observations. Rev Chir Orthop, 1977; 63 (Suppl. II):145. [6] Rockwood CA, Jr. Function following decompression of the shoulder and debridement of the rotator cu. Presented at the 52nd Annual Meeting of the AAOS, Las Vegas, January 27, 1985. [7] Bateman JE. The shoulder and neck, 2nd edn. Philadelphia: W.B. Saunders, 1978. [8] Bush LF. The torn shoulder capsule. J Bone Joint Surg, 1975;57(A):256±259. [9] Heikel HVA. Rupture of the rotator cu of the shoulder: experiences of surgical treatment. Acta Orthop Scand, 1968;39:477±492. [10] Neviaser JS. Ruptures of the rotator cu of the shoulder. New concepts in the diagnosis and operative treatment for chronic ruptures. Arch Surg, 1971;102:483±485. [11] Ozaki J, Fujimoto S, Masuhara K. Repair of chronic massive rotator cu tears with synthetic fabrics. In: Bateman JE, Welsh RP, editors. Surgery of the shoulder. Toronto: B.C. Decker, 1984:185±91.
[12] Debeyre J, Patte D, Elmelik E. Repair of ruptures of the rotator cu of the shoulder: with a note on advancement of the supraspinatus muscle. J Bone Joint Surg, 1965;47B:36±42. [13] Augereau B, Apoil A. La reparation des grandes ruptures de la coie des rotateurs de l'epaule par lambeau de deltoide. Livre des resumes 62e reunion annuelle de la SOFCOT, Paris, 1987; 10 (Nov.):10. [14] Gerber C. Latissimus dorsi transfer for the treatment of irreparable tears of the rotator cu. Clin Orthop, 1992;275:152±160. [15] Mikasa M. Trapezius transfer for global tear of the rotator cu. In: Bateman JE, Welsh RP, editors. Surgery of the Shoulder. Toronto: B.C. Decker, 1984:196±9. [16] Neviaser RJ, Neviaser TJ. Transfer of subscapularis and teres minor for massive defects of rotator cu. In: Bayley JI, Kessel L, editors. Shoulder surgery. Berlin: Springer, 1982:60±3. [17] Takagishi N, Okabe Y, Matsuzaki A. et al. Treatment of the rotator cu tear. J Jpn Orthop Assoc 1975;49:698±9 (in Japanese). [18] Neer CS, II. Cu tears, biceps lesions, and impingement. In: Shoulder reconstruction. Philadelphia PA: W.B. Saunders, 1990:63±77;93±134. [19] Gerber C, Schneeberger AG, Beck M, Schlegel U. Mechanical strength of repairs of the rotator cu. J Bone Joint Surg, 1994;76B:371±380. [20] An KN, Jacobsen MC, Berglund LJ, Chao EYS. Application of a magnetic tracking device to kinesiologic studies. J Biomech, 1988;21:613±620. [21] An KN, Ueba Y, Chao EYS, Cooney WP II, Linscheid RL. Tendon excursion and moment arm of index ®nger muscles. J Biomech, 1983;16:419±425. [22] Clark JM, Harryman DT II. Tendons, ligaments, and capsule of the rotator cu: gross and microscopic anatomy. J Bone Joint Surg, 1992;74A:713±725. [23] Nakajima T, Rokuuma N, Hamada K, Tomatsu T, Fukuda H. Histologic and biomechanical characteristics of the supraspinatus tendon: reference to rotator cu tearing. J Shoulder Elbow Surg, 1994;3:79±87. [24] Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. J Bone Joint Surg, 1976;58A:195±200. [25] Klapper RC, Jobe FW, Matsuura P. The subscapularis muscle and its glenohumeral ligament-like bands: a histomorphological study. Am J Sports Med, 1992;20:307±310. [26] Otis JC, Jiang CC, Wickiewicz TL, Peterson MG, Warren RF, Santner TJ. Changes in the moment arms of the rotator cu and deltoid muscles with abduction and rotation. J Bone Joint Surg, 1994;76(A):667±676.