- Email: [email protected]
Functional morphology of the supraspinatus tendon
Journal of Orthopaedic Research 20 (2002) 920–926 www.elsevier.com/locate/orthres
Functional morphology of the supraspinatus tendon Jonathan Fallon a, Field T. Blevins a
b,*
, Kathryn Vogel c, John Trotter
d
Department of Orthopaedics and Rehabilitation, University of New Mexico School of Medicine, Albuquerque, NM, USA b Animas Orthopaedic Associates, 1199 Main Ave. Suite #242, Durango, CO 81301, USA c Department of Anatomy and Cell Biology, University of New Mexico School of Medicine, Albuquerque, NM, USA d Department of Biology, University of New Mexico, Albuquerque, NM, USA Received 8 August 2000; accepted 23 January 2002
Abstract Grossly normal supraspinatus tendons were analyzed by stereomicroscope dissection and three-dimensional serial-section reconstruction. Four structurally independent subunits were identified: the tendon proper extended from the musculotendinous junction to approximately 2.0 cm medial to the greater tuberosity. It was composed of parallel collagen fascicles oriented along the tensional axis and separated by a prominent endotenon region. There was no interdigitation of fascicles, and an 18% incidence of fascicle convergence as the fascicles coursed from muscle toward greater tuberosity. The attachment fibrocartilage extended from the tendon proper to the greater tuberosity, consisted of a complex basket-weave of collagen fibers, and stained diffusely with alcian blue. The densely packed unidirectional collagen fibers of the rotator cable extended from the coracohumeral (CH) ligament posteriorly to the infraspinatus, coursing both superficial and deep to the tendon proper. The capsule was composed of thin collagen sheets each with uniform fiber alignment that differed slightly between sheets. These data describe a specialized tendon capable of internally compensating for changing joint angles through fascicles which are structurally independent and can slide past one another. The tendon attachment exhibits a structure adapted to tensional load dispersion and resistance to compression. Ó 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. Keywords: Rotator cuff; Supraspinatus; Tendon anotomy; Tendon function
Introduction Rotator cuff pathology is a significant cause of morbidity and disability particularly in the middle-aged and elderly population [1,14–17,23,26]. The etiology of cuff pathology remains poorly understood due at least in part to an incomplete understanding of the structure and functional anatomy of the normal supraspinatus tendon. A recent biochemical and histologic investigation done in our laboratory noted increased proteoglycan content and localized alcian blue stained ‘‘seams’’ within the supraspinatus tendon as compared with the long head of the biceps tendon [4]. Although the biochemical profile of the supraspinatus tendon was consistent with those observed in regions of tendons
*
Corresponding author. Tel.: +1-970-259-3020; fax: +1-970-2599766. E-mail address: [email protected] (F.T. Blevins).
undergoing compression, the histologic characteristics were not. It was hypothesized that proteoglycan in the supraspinatus tendon functioned to separate and lubricate collagen bundles minimizing shear stress as fascicles moved relative to each other [4]. A natural extension of this hypothesis was to perform a detailed examination of the microanatomy of the supraspinatus tendon with the goal of identifying whether or not it was composed of structurally independent subunits. Thus, the present investigation employed stereomicroscopic dissection and three-dimensional computer reconstruction to study the anatomy of the supraspinatus tendon and test the hypothesis that the collagen fascicles are structurally independent parallel units, separated by a lubricating proteoglycan rich matrix and thus capable of movement relative to one another. Methods This investigation consisted of two parts: (1) characterization of supraspinatus tendon anatomy utilizing gross and stereomicroscopic
0736-0266/02/$ - see front matter Ó 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 2 ) 0 0 0 2 3 - 2
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926 dissection; and (2) three-dimensional computer montage reconstruction of supraspinatus tendon cross-sections to examine the course and relationship of individual collagen fascicles from musculotendinous junction to insertion on the greater tuberosity. Detailed dissections were carried out on six fresh frozen and four embalmed grossly normal human shoulders (mean age 68 yr, range 44– 89 yr) with particular attention to the rotator cuff, rotator interval, coracohumeral (CH) ligament, tendon of the long head of the biceps, and cuff insertion sites. The fresh supraspinatus tendon and muscle preparations were fixed in embalming fluid for seven days, sectioned, and reincubated in the embalming solution for 24 h. Tissue was then stored in 7% acetic acid. The embalmed preparations were bathed in 7% acetic acid for at least two weeks. Fiducials were required for accurate three-dimensional reconstruction of tissue architecture. 2-0 Ethibond polyester sutures coated with polybutilate (meniscal repair sutures) (Ethicon, Inc.) were used to create continual fiducials throughout the length of the supraspinatus tendon. Sutures were immersed in India ink prior to introduction, to aid in visualization. The 25.4 cm, straight suture needle was run through the paratenon and capsule, oriented parallel to the fiber bundle axis. The suture was cut, leaving an approximately 6 cm section within the tendon. Four to five fiducials were utilized per tendon. The tissue samples were then embedded in a dental alginate, Kromopan 100 (LASCOND SpA—Laboratori Scientifici Odontoiatria Firenze, Italy), to provide stability for slicing. The embedded tissue was then sectioned transversely with a Chefmate Deli Slicer (Globe Food Equipment Co., Dayton, OH). Slices were approximately 2 mm in thickness and were immediately stored in 7% acetic acid in preparation for histological evaluation. Alginate was discarded post-sectioning and prior to storage in acetic acid. Alcian blue staining of the endotenon and peripheral tendon tissue was used to visualize collagen fiber/fascicle organization and to qualitatively analyze glycosaminoglycan (GAG) content in tendon tissue. GAGs stain dark blue with this stain. Sectioned tissue from embalmed cadavers was stained with 0.5% alcian blue (in 7% acetic acid, pH 2.5) for 20 min, removed from the alcian blue and washed in 7% acetic acid for 45 min, then bathed in deionized water overnight. Stained tissue was removed from the water and placed in 7% acetic acid for long term storage [25]. Fixed, fresh frozen tissue was stained in the 0.5% alcian blue solution for 10 min, removed and washed in 7% acetic acid for 45 min. Tissue was then placed in 7% acetic acid for long term storage. Sections were then examined under a dissection stereomicroscope (magnification 7.5) while in a liquid media. Microdissection instruments were utilized to manually manipulate fascicles while noting areas of motion between adjacent fascicles. Specifically, microforceps were used to hold a fascicle or small group of fascicles while a microprobe was used to push longitudinally on adjacent fascicles in a direction parrallel to their orientation. Thus a shear force between adjacent fascicles was produced. For the purposes of systematic analysis of the fascicular architecture, we digitally reconstructed fascicles from or near the myotendinous junction to the attachment fibrocartilage. This was done through a series of computer programs developed for three-dimensional reconstruction from serial sections at the University of Pennsylvania [21,22]. Reconstruction of the supraspinatus tendon required photographing sections in series, digitizing a number of fascicles, and digitally reconstructing these series. A photographic montage was created at a magnification of 7.5 on a stereomicroscope with a camera mount. The transilluminated tissue was photographed using AGFAPAN APX25 film (Agfa-Gevaert AG, Germany) with a shutter speed of 1/4 of a second and an aperture of 25ð1Þ. A montage was then created of each section within the series and acetate sheets were overlaid upon each montage. Representative fascicles (6–9 per tissue) were followed through the section via stereomicroscope dissection and documented as outlines of the respective fascicles on the acetate. Acetate series were then digitized with a Summagraphics MM series digitizer bit pad (Summagraphics Corp., Fairfield, CT), using the Montage computer program. The data acquired in Montage were then analyzed by a computer program which accumulated and reconstructed the serial sections of each supraspinatus sample. All computer work was done with Red Hat Linux 4.2 operating system (Red Hat Software, Inc., Durham, NC) on a Pentium Pro SCSI system built by SW Technology (Richardson, TX).
921
Results Supraspinatus tendon gross and microscopic anatomy Based on gross and microscopic dissections, four structural subunits within the supraspinatus tendon and the associated structures were identified: tendon proper, attachment fibrocartilage, rotator cable, and capsule. The tendon proper extended from the supraspinatus musclo-tendinous junction (approximately 5 cm medial to the greater tuberosity) to the attachment fibrocartilage (2 cm medial to the greater tuberosity. The tendon proper enlarged and broadened towards the distal attachment manifesting as a thick, ‘rope-like’ structure in the anterior tendon and a thin ‘strap-like’ region that spread posteriorly, creating a broad point of attachment. The internal structure of the tendon proper consisted of collagen bundles, grouped as fibers and fascicles, running parallel to the axis of tension. The fascicles were separated by a thick endotenon region that stained prominently with alcian blue (Fig. 1), indicating the presence of negatively charged GAG [22]. The loose connective tissue of the endotenon was a soft meshwork that allowed the fascicles to easily be separated and maneuvered. When fascicles in cross-sections of the tendon proper were manipulated with microdissection instruments under the stereomicroscope, the stout, solid fascicles slid past one another, and the motion occurred in the region of the alcian blue staining endotenon. As the collagen network converted from the fascicular structure of the tendon proper to the basket-weave of the attachment fibrocartilage, it appeared to compress and reorganize. This process originated in the thick anterior portion of the tendon proper, occurring closer to the bony attachment in the thinner, posterior regions. The attachment fibrocartilage of the supraspinatus extended from the tendon proper to the greater tuberosity (Fig. 2A). The average fibrocartilage length was 1:8 0:5 cm. This region encompassed the ‘‘critical zone’’ described by Codman where tears are often seen [10,11]. The collagenous ultrastructure of the fibrocartilage was a basket-weave of undetectable pattern with diffuse alcian blue staining (Fig. 2B), histologically resembling fibrocartilage subject to compression. Rotator Cable. Our observations correlated well with those of Burkhart et al. [6,7] who originally described this structure. We observed the rotator cable to be the thicker, deeper one of two extensions of the CH ligament into the supraspinatus structure. The first projection of the CH ligament was a thin layer of tissue enveloping the superior border the tendon proper. The second contribution of the CH ligament (the rotator cable) to the supraspinatus extended perpendicular to the axis of the tendon proper, deep to the tendon and superficial to the joint capsule (Fig. 2). This thick collagenous strap traversed the tendon proper and
922
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
Fig. 1. Cross-sectional photomicrograph of alcian stained tendon proper (magnification 7.5). (A) Note the parallel fascicles of the tendon proper (TP), the dense rotator cable (RC) inferior to the tendon proper, and the thin capsule on the articular (inferior) surface. (B) Cross-sectional photomicrograph of tendon proper fascicle and GAG rich endotenon, (magnification 60). Localization of the GAG containing matrix is primarily around the funcional unit of the fascicle.
Fig. 2. Longitudinal photomicrograph of sectioned supraspinatus tendon and its attachment to the greater tuberosity. (A) Attachment fibrocartilage (AF), rotator cable (arrow), greater tuberosity (GT), tendon proper (TP). (B) Cross-section photomicrograph of fibrocartilage in the plane of the dashed line (magnification 10), alcian blue stain. Note the diffuse GAG stain and basketweave collagen orientation (GAG darkly stained).
extended into the infraspinatus tendon. It surrounded the fibrocartilage and the area correlating with the
critical zone. Manual tension applied to the medial extreme of the CH ligament demonstrated a structural connection to the rotator cable in all except one shoulder. There was no distinguishable CH ligament in one of ten shoulders. The joint capsule was a thin collagenous structure that lined the articular surface of the rotator cuff. The capsule was a composite of thin collagen sheets; each individual sheet having a uniform fiber alignment, which differed slightly between sheets (Fig. 3), combining to form a tough structure of varying fiber orientation. The capsule became inseparable from the attachment fibrocartilage just medial to the point of attachment to the greater tuberosity. The capsule and the infraspinatus tendon remained distinct structural entities until close to the point of attachment, where the entire rotator cuff became a solid structure, in the vicinity of the supraspinatus tendon. An additional area of interest was the rotator interval (the region between the supraspinatus and subscapularis). A thick, soft, collagenous tissue that was well vascularized, connected the supraspinatus and subscapularis tendons and stained heavily with alcian blue. A solid, porous structure was observed in two of the five (5) fresh shoulders in which this region was examined and sectioned. The deposit was 4–6 mm in thickness with no consistent shape, and was confined to the area abutting the tendon proper of the supraspinatus.
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
923
Fig. 3. Polarized photomicrograph of joint capsule (magnification 40). Collagen fibril direction alternating with different layers (arrowheads).
cant difference was observed between embalmed cadaver tissue and fixed fresh frozen tissue.
Three-dimensional reconstruction A systematic three-dimensional investigation of the fascicular interrelationships demonstrated that fascicles within the supraspinatus tendon proper do not interdigitate (Table 1). Fascicles were functionally defined as fibrillar structures surrounded by endotenon that retain their integrity during manual manipulation under stereomicroscope dissection. The ability to manipulate fascicles allowed structural integrity to be observed through the section of tissue. This technique, in addition to gross and stereomicroscopic observations, was used to investigate possible collagenous interaction between adjacent fascicles. Six to nine fascicles were digitized in each tendon proper along a length of approximately 1:6 0:4 cm. In all tendons examined, fascicles remained distinct parallel structures, demonstrating no interdigitation. Convergence of smaller fascicles into large fascicles (proximal to distal) was seen in approximately 18% of all fascicles reconstructed (Table 1). No signifi-
Table 1 Fascicle convergence and interdigitation Sample
Fascicles
Convergence
Interdigitation
% of convergence (%)
S5 S8 S13 S14 S15 S16 S18 S19
6 7 8 8 6 9 7 6
2 0 2 3 1 1 0 1
0 0 0 0 0 0 0 0
33 0 25 38 17 11 0 17
Average
18
Discussion This study was designed as a descriptive investigation of supraspinatus anatomy and fascicular architecture. When correlated with data from recent histologic and biochemical investigations; structure–function relationships may be proposed [4,5,7,9,20]. A potential limitation of the stereomicroscopic portion of this study is that fresh tissue could not be used, and embalming may have affected the ability of the fascicles to move in relation to one another. The sectioning methods in this study required the use of formalin treated tissues. Formalin increases protein cross-links tending to stiffen collagenous tissue. If there was an effect of formalin on the movement between fascicles of the supraspinatus tendon, it would likely be to decrease the ability of the fascicles to move in relation to each other (due to increased cross-links), leading to an underestimation of the true in vivo motion which occurs. The preservation of tissue samples was done in a manner to prevent dehydration, which could have caused distortion, affecting the endotenon regions. The tendon proper acts in series with the attachment fibrocartilage to transmit tensional force from the muscle to the bone. It has a distinct fascicular structure with independent, parallel tendons oriented along the line of tension. The parallel fascicles observed in this region correspond well with Clark and Harryman’s layer #2 [9]. Fascicles are separated by a prominent endotenon tissue comprised of loose connective tissue that stains heavily for negatively charged GAGs. The loose endotenon
924
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
Fig. 4. Differential movement of different regions of the supraspinatus tendon with various movements. (A) During horizontal adduction, the posterior region (X) of the supraspinatus tendon has a greater excursion than the anterior region (Y). (B) During abduction, the superior region (X) has a greater excursion then the inferior (articular) surface (Y).
allowed for easy manual separation of individual fascicles, and the presence of such a distinct endotenon suggests an important function within tendon physiology. GAGs have been implicated in such mechanisms as assisting nutrient flow, simple occupation of space, protection and resistance from compressional forces, and lubrication against shear and frictional forces [4,26,27]. Although any of these hypotheses are possible functions for this GAG containing medium such as the endotenon, the latter appears especially plausible. The kinematics of the shoulder joint and shape of the supraspinatus tendon dictate that different regions of the supraspinatus tendon move independently in relation to each other. In Fig. 4, the tendon structure in area X would have a different excursion than that of in area Y during horizontal adduction (Fig. 4A), and abduction (Fig. 4B). Because of differing degrees of excursion the muscle–tendon structure would have to compensate at some level in order to maintain tension throughout the tendon during active motion. The architectural data provide indirect evidence of a tendon that allows fascicles to slide relative to each other providing a mechanism of compensation. Data from digital reconstruction of the supraspinatus tendon further strengthen this theory. Both fresh and cadaver tissue demonstrated no interdigitation in the six to nine fascicles followed throughout the length of each tendon proper, indicating a structural independence of fascicles. Convergence of smaller fascicles into larger fascicles did occur in a significant number of tendons (Table 1, Fig. 5). The convergence observed in the tendon proper demonstrates the formation of fascicles on a larger scale, with subfibrils grouping to form a large fascicle. Although these converging fascicles may be
Fig. 5. Representation of tendon fascicles converging into one functional unit (18% incidence) and interdigitation between fascicles (not seen).
thought to correspond to Clark and Harryman’s layer #3, their location with respect to the parallel fascicles is different. Clark and Harryman’s layer #3 is deep to layer #2; the converging fascicles in the present investigation were observed throughout the tendon proper, intermixed with the parallel fascicles [9]. Convergence of small fascicles into a larger fascicle may suggest multiple muscle fibers acting in concert to affect a particular joint movement. It may also be a necessary mechanism by which the volume of the tendon is reduced as it courses from the musculotendinous junction through the supraspinatus outlet to the greater tuberosity. Biochemical data from previous studies also are consistent with this theory. Histological assays in the tendon
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
proper region of the supraspinatus have localized GAG to the endotenon, between fascicular structures. The isolation of this high quantity of GAG to the tissue separating fascicles indicates its functional significance. While the precise function has not been determined, hyaluronic acid and chondroitin sulfate have been implicated in protection against compression and shear forces throughout the body [4,26,27]. However, the histological profile of the tendon proper does not fit that of tendon under compression. The placement, composition, and quantity of the GAG does make it ideal for the lubrication of moving structures. The synthesis of these results yields a tendon proper region with structurally independent units, separated by a thick loose connective tissue rich in lubricant. This schematic fits into a tissue model that allows fascicles to move independently in order to compensate for differential joint angles. The tendon proper attaches to the greater tuberosity through an extended attachment fibrocartilage. The morphology of the attachment fibrocartilage seen in the supraspinatus tendon is unlike that found in most epiphysial tendon attachments. This fibrocartilagenous region extends well beyond the normal 500–700 lm [3,12] to approximately 2 cm. The histological characteristics are similar to those described in compressional fibrocartilage [4]; a basket-weave collagenous structure with diffuse GAG staining. Fibrocartilage within a tendon is thought to be an adaptation to compression. The tensional strength of such a disorganized collagen structure is inferior to that of tensional tendon, but is better able to resist compression [2]. For this reason, it is to the advantage of the tendon to adapt this morphology only when compressional strain is present. Studies have implicated this region as being compressed between the humeral head and acromion [8,18,19]. Thus, the morphology and composition of the attachment fibrocartilage region may function to resist compression. A possible second function for the attachment fibrocartilage pertains to disbursement of stress in the region of the tendon insertion into the greater tuberosity. An extended attachment fibrocartilage may allow for increased load disbursement at the point of attachment. Anatomical dissections have determined the existence of the rotator cable as a collagenous extension of the CH ligament [6,7]. Our findings regarding this structure are in agreement with those Clarke and Harryman [9], Burkhart et al. [7], and Gohlke et al. [13]. Our findings of the joint capsule agree with those of Gohlke et al. [13], and Steiner and Hermann [24].
Conclusion Four functional subunits were observed within the supraspinatus tendon. Tendon fascicles surrounded by
925
proteoglycan rich material appear to have the ability to move independent of one another.
Acknowledgements Funding for this project was received from a University of New Mexico School of Medicine Research Allocations Committee Grant; and from a University of New Mexico School of Medicine, Department of Orthopaedics and Rehabilitation George E. Omer, MD Grant. No authors have professional or financial affiliations, which would be perceived to have biased the presentation. References [1] Beach WR, Caspari RB. Arthroscopic management of rotator cuff disease. Orthopaedics 1993;16:1007–15. [2] Benjamin M, Evans EJ. Fibrocartilage. J Anat 1990;171:1–15. [3] Benjamin M, Evans EJ, Copp L. The histology of tendon attachments to bone in man. J Anat 1986;149:89–100. [4] Berenson MC, Blevins FT, Plaas AH, Vogel KG. Proteoglycans of human rotator cuff tendons. J Orthop Res 1996;14:518–25. [5] Blevins FT, Djurasovic M, Flatow EL, Vogel KG. Biology of the rotator cuff tendon. Orthop Clin North Am 1997;28:1–16. [6] Burkhart SS. Fluoroscopic comparison of kinematic patterns in massive rotator cuff tears. Clin Orthop Rel Res 1992;284:144–52. [7] Burkhart SS, Esch JC, Jolson RS. The rotator crescent and rotator cable: an anatomic description of the shoulder’s ‘‘Suspension Bridge’’. J Arth Rel Surg 1993;9:611–6. [8] Chansky HA, Iannotti JP. The vascularity of the rotator cuff. Clin Sports Med 1991;10:807–22. [9] Clark J, Harryman. Tendons, ligaments, and capsule of the rotator cuff. J Bone Joint Surg A 1992;74:713–25. [10] Codman EA. The shoulder; rupture of the supraspinatus tendon and other lesions in or about the subcromial bursa. Boston: T. Todd Company; 1934. [11] Codman EA, Akerson IB. The pathology associated with rupture of the supraspinatus tendon. Ann Surg 1931;94:348–59. [12] Cooper RR, Misol S. Tendon and ligament insertion. A light and electron microscopic study. J Bone Joint Surg B 1970;52:1– 20. [13] Gohlke F, Essigkrug B, Schmitz F. The pattern of the collagen fiber bundles of the capsule of the glenohumeral joint. J Shoulder Elbow Surg 1994;3:111–28. [14] Hagberg M. Neck and shoulder disorders. In: Rosenstock L, Cullen MR, editors. Textbook of occupational and environmental medicine. Philadelphia: WB Saunders; 1994 (Chapter 14.2). [15] Hagberg M. Neck and arm disorders. Br Med J (BMJ) 1996;313:419–22. [16] Hagberg M, Wegman DH. Prevalence rates and odds ratios of shoulder–neck diseases in different occupational groups. Br J Ind Med 1987;44:602–10. [17] Kvitne RS, Jobe FW, Jobe CM. Shoulder instability in the overhand or throwing athlete. Clin Sports Med 1995;14:917–35. [18] Neer CS. Anterior acromioplasty for the chronic impingement syndrome in the shoulder. J Bone Joint Surg A 1972;54(1): 41–51. [19] Neer CS. Impingement lesions. Clin Orthop 1983;173:70–7. [20] Riley GP, Harral RL, Constant CR, Chard MD, Cawston TE, Hazleman BL. Glycosaminoglycans of human rotator cuff
926
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
tendons: changes with age and in chronic rotator cuff tendinitis. Ann Rheum Dis 1994;53:367–76. [21] Smith RG. Montage: a system for three-dimensional reconstruction by personal computer. J Neurosci Meth 1987;21: 55–69. [22] Smith RG. Three dimensional reconstruction from serial sections. User’s Manual Version 4.4. February 25, 1988. University of PA Medical School, Philadelphia, PA. [23] Snyder SJ. Evaluation and treatment of the rotator cuff. Orthop Clin North Am 1993;1:173–92.
[24] Steiner D, Hermann B. Collagen fiber arrangement of the human shoulder joint capsule—an anatomical study. Acta Anat 1989;136: 300–2. [25] Troyer H. Principles and techniques of histochemistry. Boston: Little Brown and Company; 1980. [26] Vogel KG, Heinegard D. Characterization of proteoglycans from adult bovine tendon. J Biol Chem 1985;260:9298–306. [27] Vogel KG, Koob TJ. Structural specialization in tendons under compression. Int Rev Cytol 1989;115:267–93.
Functional morphology of the supraspinatus tendon Jonathan Fallon a, Field T. Blevins a
b,*
, Kathryn Vogel c, John Trotter
d
Department of Orthopaedics and Rehabilitation, University of New Mexico School of Medicine, Albuquerque, NM, USA b Animas Orthopaedic Associates, 1199 Main Ave. Suite #242, Durango, CO 81301, USA c Department of Anatomy and Cell Biology, University of New Mexico School of Medicine, Albuquerque, NM, USA d Department of Biology, University of New Mexico, Albuquerque, NM, USA Received 8 August 2000; accepted 23 January 2002
Abstract Grossly normal supraspinatus tendons were analyzed by stereomicroscope dissection and three-dimensional serial-section reconstruction. Four structurally independent subunits were identified: the tendon proper extended from the musculotendinous junction to approximately 2.0 cm medial to the greater tuberosity. It was composed of parallel collagen fascicles oriented along the tensional axis and separated by a prominent endotenon region. There was no interdigitation of fascicles, and an 18% incidence of fascicle convergence as the fascicles coursed from muscle toward greater tuberosity. The attachment fibrocartilage extended from the tendon proper to the greater tuberosity, consisted of a complex basket-weave of collagen fibers, and stained diffusely with alcian blue. The densely packed unidirectional collagen fibers of the rotator cable extended from the coracohumeral (CH) ligament posteriorly to the infraspinatus, coursing both superficial and deep to the tendon proper. The capsule was composed of thin collagen sheets each with uniform fiber alignment that differed slightly between sheets. These data describe a specialized tendon capable of internally compensating for changing joint angles through fascicles which are structurally independent and can slide past one another. The tendon attachment exhibits a structure adapted to tensional load dispersion and resistance to compression. Ó 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. Keywords: Rotator cuff; Supraspinatus; Tendon anotomy; Tendon function
Introduction Rotator cuff pathology is a significant cause of morbidity and disability particularly in the middle-aged and elderly population [1,14–17,23,26]. The etiology of cuff pathology remains poorly understood due at least in part to an incomplete understanding of the structure and functional anatomy of the normal supraspinatus tendon. A recent biochemical and histologic investigation done in our laboratory noted increased proteoglycan content and localized alcian blue stained ‘‘seams’’ within the supraspinatus tendon as compared with the long head of the biceps tendon [4]. Although the biochemical profile of the supraspinatus tendon was consistent with those observed in regions of tendons
*
Corresponding author. Tel.: +1-970-259-3020; fax: +1-970-2599766. E-mail address: [email protected] (F.T. Blevins).
undergoing compression, the histologic characteristics were not. It was hypothesized that proteoglycan in the supraspinatus tendon functioned to separate and lubricate collagen bundles minimizing shear stress as fascicles moved relative to each other [4]. A natural extension of this hypothesis was to perform a detailed examination of the microanatomy of the supraspinatus tendon with the goal of identifying whether or not it was composed of structurally independent subunits. Thus, the present investigation employed stereomicroscopic dissection and three-dimensional computer reconstruction to study the anatomy of the supraspinatus tendon and test the hypothesis that the collagen fascicles are structurally independent parallel units, separated by a lubricating proteoglycan rich matrix and thus capable of movement relative to one another. Methods This investigation consisted of two parts: (1) characterization of supraspinatus tendon anatomy utilizing gross and stereomicroscopic
0736-0266/02/$ - see front matter Ó 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 2 ) 0 0 0 2 3 - 2
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926 dissection; and (2) three-dimensional computer montage reconstruction of supraspinatus tendon cross-sections to examine the course and relationship of individual collagen fascicles from musculotendinous junction to insertion on the greater tuberosity. Detailed dissections were carried out on six fresh frozen and four embalmed grossly normal human shoulders (mean age 68 yr, range 44– 89 yr) with particular attention to the rotator cuff, rotator interval, coracohumeral (CH) ligament, tendon of the long head of the biceps, and cuff insertion sites. The fresh supraspinatus tendon and muscle preparations were fixed in embalming fluid for seven days, sectioned, and reincubated in the embalming solution for 24 h. Tissue was then stored in 7% acetic acid. The embalmed preparations were bathed in 7% acetic acid for at least two weeks. Fiducials were required for accurate three-dimensional reconstruction of tissue architecture. 2-0 Ethibond polyester sutures coated with polybutilate (meniscal repair sutures) (Ethicon, Inc.) were used to create continual fiducials throughout the length of the supraspinatus tendon. Sutures were immersed in India ink prior to introduction, to aid in visualization. The 25.4 cm, straight suture needle was run through the paratenon and capsule, oriented parallel to the fiber bundle axis. The suture was cut, leaving an approximately 6 cm section within the tendon. Four to five fiducials were utilized per tendon. The tissue samples were then embedded in a dental alginate, Kromopan 100 (LASCOND SpA—Laboratori Scientifici Odontoiatria Firenze, Italy), to provide stability for slicing. The embedded tissue was then sectioned transversely with a Chefmate Deli Slicer (Globe Food Equipment Co., Dayton, OH). Slices were approximately 2 mm in thickness and were immediately stored in 7% acetic acid in preparation for histological evaluation. Alginate was discarded post-sectioning and prior to storage in acetic acid. Alcian blue staining of the endotenon and peripheral tendon tissue was used to visualize collagen fiber/fascicle organization and to qualitatively analyze glycosaminoglycan (GAG) content in tendon tissue. GAGs stain dark blue with this stain. Sectioned tissue from embalmed cadavers was stained with 0.5% alcian blue (in 7% acetic acid, pH 2.5) for 20 min, removed from the alcian blue and washed in 7% acetic acid for 45 min, then bathed in deionized water overnight. Stained tissue was removed from the water and placed in 7% acetic acid for long term storage [25]. Fixed, fresh frozen tissue was stained in the 0.5% alcian blue solution for 10 min, removed and washed in 7% acetic acid for 45 min. Tissue was then placed in 7% acetic acid for long term storage. Sections were then examined under a dissection stereomicroscope (magnification 7.5) while in a liquid media. Microdissection instruments were utilized to manually manipulate fascicles while noting areas of motion between adjacent fascicles. Specifically, microforceps were used to hold a fascicle or small group of fascicles while a microprobe was used to push longitudinally on adjacent fascicles in a direction parrallel to their orientation. Thus a shear force between adjacent fascicles was produced. For the purposes of systematic analysis of the fascicular architecture, we digitally reconstructed fascicles from or near the myotendinous junction to the attachment fibrocartilage. This was done through a series of computer programs developed for three-dimensional reconstruction from serial sections at the University of Pennsylvania [21,22]. Reconstruction of the supraspinatus tendon required photographing sections in series, digitizing a number of fascicles, and digitally reconstructing these series. A photographic montage was created at a magnification of 7.5 on a stereomicroscope with a camera mount. The transilluminated tissue was photographed using AGFAPAN APX25 film (Agfa-Gevaert AG, Germany) with a shutter speed of 1/4 of a second and an aperture of 25ð1Þ. A montage was then created of each section within the series and acetate sheets were overlaid upon each montage. Representative fascicles (6–9 per tissue) were followed through the section via stereomicroscope dissection and documented as outlines of the respective fascicles on the acetate. Acetate series were then digitized with a Summagraphics MM series digitizer bit pad (Summagraphics Corp., Fairfield, CT), using the Montage computer program. The data acquired in Montage were then analyzed by a computer program which accumulated and reconstructed the serial sections of each supraspinatus sample. All computer work was done with Red Hat Linux 4.2 operating system (Red Hat Software, Inc., Durham, NC) on a Pentium Pro SCSI system built by SW Technology (Richardson, TX).
921
Results Supraspinatus tendon gross and microscopic anatomy Based on gross and microscopic dissections, four structural subunits within the supraspinatus tendon and the associated structures were identified: tendon proper, attachment fibrocartilage, rotator cable, and capsule. The tendon proper extended from the supraspinatus musclo-tendinous junction (approximately 5 cm medial to the greater tuberosity) to the attachment fibrocartilage (2 cm medial to the greater tuberosity. The tendon proper enlarged and broadened towards the distal attachment manifesting as a thick, ‘rope-like’ structure in the anterior tendon and a thin ‘strap-like’ region that spread posteriorly, creating a broad point of attachment. The internal structure of the tendon proper consisted of collagen bundles, grouped as fibers and fascicles, running parallel to the axis of tension. The fascicles were separated by a thick endotenon region that stained prominently with alcian blue (Fig. 1), indicating the presence of negatively charged GAG [22]. The loose connective tissue of the endotenon was a soft meshwork that allowed the fascicles to easily be separated and maneuvered. When fascicles in cross-sections of the tendon proper were manipulated with microdissection instruments under the stereomicroscope, the stout, solid fascicles slid past one another, and the motion occurred in the region of the alcian blue staining endotenon. As the collagen network converted from the fascicular structure of the tendon proper to the basket-weave of the attachment fibrocartilage, it appeared to compress and reorganize. This process originated in the thick anterior portion of the tendon proper, occurring closer to the bony attachment in the thinner, posterior regions. The attachment fibrocartilage of the supraspinatus extended from the tendon proper to the greater tuberosity (Fig. 2A). The average fibrocartilage length was 1:8 0:5 cm. This region encompassed the ‘‘critical zone’’ described by Codman where tears are often seen [10,11]. The collagenous ultrastructure of the fibrocartilage was a basket-weave of undetectable pattern with diffuse alcian blue staining (Fig. 2B), histologically resembling fibrocartilage subject to compression. Rotator Cable. Our observations correlated well with those of Burkhart et al. [6,7] who originally described this structure. We observed the rotator cable to be the thicker, deeper one of two extensions of the CH ligament into the supraspinatus structure. The first projection of the CH ligament was a thin layer of tissue enveloping the superior border the tendon proper. The second contribution of the CH ligament (the rotator cable) to the supraspinatus extended perpendicular to the axis of the tendon proper, deep to the tendon and superficial to the joint capsule (Fig. 2). This thick collagenous strap traversed the tendon proper and
922
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
Fig. 1. Cross-sectional photomicrograph of alcian stained tendon proper (magnification 7.5). (A) Note the parallel fascicles of the tendon proper (TP), the dense rotator cable (RC) inferior to the tendon proper, and the thin capsule on the articular (inferior) surface. (B) Cross-sectional photomicrograph of tendon proper fascicle and GAG rich endotenon, (magnification 60). Localization of the GAG containing matrix is primarily around the funcional unit of the fascicle.
Fig. 2. Longitudinal photomicrograph of sectioned supraspinatus tendon and its attachment to the greater tuberosity. (A) Attachment fibrocartilage (AF), rotator cable (arrow), greater tuberosity (GT), tendon proper (TP). (B) Cross-section photomicrograph of fibrocartilage in the plane of the dashed line (magnification 10), alcian blue stain. Note the diffuse GAG stain and basketweave collagen orientation (GAG darkly stained).
extended into the infraspinatus tendon. It surrounded the fibrocartilage and the area correlating with the
critical zone. Manual tension applied to the medial extreme of the CH ligament demonstrated a structural connection to the rotator cable in all except one shoulder. There was no distinguishable CH ligament in one of ten shoulders. The joint capsule was a thin collagenous structure that lined the articular surface of the rotator cuff. The capsule was a composite of thin collagen sheets; each individual sheet having a uniform fiber alignment, which differed slightly between sheets (Fig. 3), combining to form a tough structure of varying fiber orientation. The capsule became inseparable from the attachment fibrocartilage just medial to the point of attachment to the greater tuberosity. The capsule and the infraspinatus tendon remained distinct structural entities until close to the point of attachment, where the entire rotator cuff became a solid structure, in the vicinity of the supraspinatus tendon. An additional area of interest was the rotator interval (the region between the supraspinatus and subscapularis). A thick, soft, collagenous tissue that was well vascularized, connected the supraspinatus and subscapularis tendons and stained heavily with alcian blue. A solid, porous structure was observed in two of the five (5) fresh shoulders in which this region was examined and sectioned. The deposit was 4–6 mm in thickness with no consistent shape, and was confined to the area abutting the tendon proper of the supraspinatus.
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
923
Fig. 3. Polarized photomicrograph of joint capsule (magnification 40). Collagen fibril direction alternating with different layers (arrowheads).
cant difference was observed between embalmed cadaver tissue and fixed fresh frozen tissue.
Three-dimensional reconstruction A systematic three-dimensional investigation of the fascicular interrelationships demonstrated that fascicles within the supraspinatus tendon proper do not interdigitate (Table 1). Fascicles were functionally defined as fibrillar structures surrounded by endotenon that retain their integrity during manual manipulation under stereomicroscope dissection. The ability to manipulate fascicles allowed structural integrity to be observed through the section of tissue. This technique, in addition to gross and stereomicroscopic observations, was used to investigate possible collagenous interaction between adjacent fascicles. Six to nine fascicles were digitized in each tendon proper along a length of approximately 1:6 0:4 cm. In all tendons examined, fascicles remained distinct parallel structures, demonstrating no interdigitation. Convergence of smaller fascicles into large fascicles (proximal to distal) was seen in approximately 18% of all fascicles reconstructed (Table 1). No signifi-
Table 1 Fascicle convergence and interdigitation Sample
Fascicles
Convergence
Interdigitation
% of convergence (%)
S5 S8 S13 S14 S15 S16 S18 S19
6 7 8 8 6 9 7 6
2 0 2 3 1 1 0 1
0 0 0 0 0 0 0 0
33 0 25 38 17 11 0 17
Average
18
Discussion This study was designed as a descriptive investigation of supraspinatus anatomy and fascicular architecture. When correlated with data from recent histologic and biochemical investigations; structure–function relationships may be proposed [4,5,7,9,20]. A potential limitation of the stereomicroscopic portion of this study is that fresh tissue could not be used, and embalming may have affected the ability of the fascicles to move in relation to one another. The sectioning methods in this study required the use of formalin treated tissues. Formalin increases protein cross-links tending to stiffen collagenous tissue. If there was an effect of formalin on the movement between fascicles of the supraspinatus tendon, it would likely be to decrease the ability of the fascicles to move in relation to each other (due to increased cross-links), leading to an underestimation of the true in vivo motion which occurs. The preservation of tissue samples was done in a manner to prevent dehydration, which could have caused distortion, affecting the endotenon regions. The tendon proper acts in series with the attachment fibrocartilage to transmit tensional force from the muscle to the bone. It has a distinct fascicular structure with independent, parallel tendons oriented along the line of tension. The parallel fascicles observed in this region correspond well with Clark and Harryman’s layer #2 [9]. Fascicles are separated by a prominent endotenon tissue comprised of loose connective tissue that stains heavily for negatively charged GAGs. The loose endotenon
924
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
Fig. 4. Differential movement of different regions of the supraspinatus tendon with various movements. (A) During horizontal adduction, the posterior region (X) of the supraspinatus tendon has a greater excursion than the anterior region (Y). (B) During abduction, the superior region (X) has a greater excursion then the inferior (articular) surface (Y).
allowed for easy manual separation of individual fascicles, and the presence of such a distinct endotenon suggests an important function within tendon physiology. GAGs have been implicated in such mechanisms as assisting nutrient flow, simple occupation of space, protection and resistance from compressional forces, and lubrication against shear and frictional forces [4,26,27]. Although any of these hypotheses are possible functions for this GAG containing medium such as the endotenon, the latter appears especially plausible. The kinematics of the shoulder joint and shape of the supraspinatus tendon dictate that different regions of the supraspinatus tendon move independently in relation to each other. In Fig. 4, the tendon structure in area X would have a different excursion than that of in area Y during horizontal adduction (Fig. 4A), and abduction (Fig. 4B). Because of differing degrees of excursion the muscle–tendon structure would have to compensate at some level in order to maintain tension throughout the tendon during active motion. The architectural data provide indirect evidence of a tendon that allows fascicles to slide relative to each other providing a mechanism of compensation. Data from digital reconstruction of the supraspinatus tendon further strengthen this theory. Both fresh and cadaver tissue demonstrated no interdigitation in the six to nine fascicles followed throughout the length of each tendon proper, indicating a structural independence of fascicles. Convergence of smaller fascicles into larger fascicles did occur in a significant number of tendons (Table 1, Fig. 5). The convergence observed in the tendon proper demonstrates the formation of fascicles on a larger scale, with subfibrils grouping to form a large fascicle. Although these converging fascicles may be
Fig. 5. Representation of tendon fascicles converging into one functional unit (18% incidence) and interdigitation between fascicles (not seen).
thought to correspond to Clark and Harryman’s layer #3, their location with respect to the parallel fascicles is different. Clark and Harryman’s layer #3 is deep to layer #2; the converging fascicles in the present investigation were observed throughout the tendon proper, intermixed with the parallel fascicles [9]. Convergence of small fascicles into a larger fascicle may suggest multiple muscle fibers acting in concert to affect a particular joint movement. It may also be a necessary mechanism by which the volume of the tendon is reduced as it courses from the musculotendinous junction through the supraspinatus outlet to the greater tuberosity. Biochemical data from previous studies also are consistent with this theory. Histological assays in the tendon
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
proper region of the supraspinatus have localized GAG to the endotenon, between fascicular structures. The isolation of this high quantity of GAG to the tissue separating fascicles indicates its functional significance. While the precise function has not been determined, hyaluronic acid and chondroitin sulfate have been implicated in protection against compression and shear forces throughout the body [4,26,27]. However, the histological profile of the tendon proper does not fit that of tendon under compression. The placement, composition, and quantity of the GAG does make it ideal for the lubrication of moving structures. The synthesis of these results yields a tendon proper region with structurally independent units, separated by a thick loose connective tissue rich in lubricant. This schematic fits into a tissue model that allows fascicles to move independently in order to compensate for differential joint angles. The tendon proper attaches to the greater tuberosity through an extended attachment fibrocartilage. The morphology of the attachment fibrocartilage seen in the supraspinatus tendon is unlike that found in most epiphysial tendon attachments. This fibrocartilagenous region extends well beyond the normal 500–700 lm [3,12] to approximately 2 cm. The histological characteristics are similar to those described in compressional fibrocartilage [4]; a basket-weave collagenous structure with diffuse GAG staining. Fibrocartilage within a tendon is thought to be an adaptation to compression. The tensional strength of such a disorganized collagen structure is inferior to that of tensional tendon, but is better able to resist compression [2]. For this reason, it is to the advantage of the tendon to adapt this morphology only when compressional strain is present. Studies have implicated this region as being compressed between the humeral head and acromion [8,18,19]. Thus, the morphology and composition of the attachment fibrocartilage region may function to resist compression. A possible second function for the attachment fibrocartilage pertains to disbursement of stress in the region of the tendon insertion into the greater tuberosity. An extended attachment fibrocartilage may allow for increased load disbursement at the point of attachment. Anatomical dissections have determined the existence of the rotator cable as a collagenous extension of the CH ligament [6,7]. Our findings regarding this structure are in agreement with those Clarke and Harryman [9], Burkhart et al. [7], and Gohlke et al. [13]. Our findings of the joint capsule agree with those of Gohlke et al. [13], and Steiner and Hermann [24].
Conclusion Four functional subunits were observed within the supraspinatus tendon. Tendon fascicles surrounded by
925
proteoglycan rich material appear to have the ability to move independent of one another.
Acknowledgements Funding for this project was received from a University of New Mexico School of Medicine Research Allocations Committee Grant; and from a University of New Mexico School of Medicine, Department of Orthopaedics and Rehabilitation George E. Omer, MD Grant. No authors have professional or financial affiliations, which would be perceived to have biased the presentation. References [1] Beach WR, Caspari RB. Arthroscopic management of rotator cuff disease. Orthopaedics 1993;16:1007–15. [2] Benjamin M, Evans EJ. Fibrocartilage. J Anat 1990;171:1–15. [3] Benjamin M, Evans EJ, Copp L. The histology of tendon attachments to bone in man. J Anat 1986;149:89–100. [4] Berenson MC, Blevins FT, Plaas AH, Vogel KG. Proteoglycans of human rotator cuff tendons. J Orthop Res 1996;14:518–25. [5] Blevins FT, Djurasovic M, Flatow EL, Vogel KG. Biology of the rotator cuff tendon. Orthop Clin North Am 1997;28:1–16. [6] Burkhart SS. Fluoroscopic comparison of kinematic patterns in massive rotator cuff tears. Clin Orthop Rel Res 1992;284:144–52. [7] Burkhart SS, Esch JC, Jolson RS. The rotator crescent and rotator cable: an anatomic description of the shoulder’s ‘‘Suspension Bridge’’. J Arth Rel Surg 1993;9:611–6. [8] Chansky HA, Iannotti JP. The vascularity of the rotator cuff. Clin Sports Med 1991;10:807–22. [9] Clark J, Harryman. Tendons, ligaments, and capsule of the rotator cuff. J Bone Joint Surg A 1992;74:713–25. [10] Codman EA. The shoulder; rupture of the supraspinatus tendon and other lesions in or about the subcromial bursa. Boston: T. Todd Company; 1934. [11] Codman EA, Akerson IB. The pathology associated with rupture of the supraspinatus tendon. Ann Surg 1931;94:348–59. [12] Cooper RR, Misol S. Tendon and ligament insertion. A light and electron microscopic study. J Bone Joint Surg B 1970;52:1– 20. [13] Gohlke F, Essigkrug B, Schmitz F. The pattern of the collagen fiber bundles of the capsule of the glenohumeral joint. J Shoulder Elbow Surg 1994;3:111–28. [14] Hagberg M. Neck and shoulder disorders. In: Rosenstock L, Cullen MR, editors. Textbook of occupational and environmental medicine. Philadelphia: WB Saunders; 1994 (Chapter 14.2). [15] Hagberg M. Neck and arm disorders. Br Med J (BMJ) 1996;313:419–22. [16] Hagberg M, Wegman DH. Prevalence rates and odds ratios of shoulder–neck diseases in different occupational groups. Br J Ind Med 1987;44:602–10. [17] Kvitne RS, Jobe FW, Jobe CM. Shoulder instability in the overhand or throwing athlete. Clin Sports Med 1995;14:917–35. [18] Neer CS. Anterior acromioplasty for the chronic impingement syndrome in the shoulder. J Bone Joint Surg A 1972;54(1): 41–51. [19] Neer CS. Impingement lesions. Clin Orthop 1983;173:70–7. [20] Riley GP, Harral RL, Constant CR, Chard MD, Cawston TE, Hazleman BL. Glycosaminoglycans of human rotator cuff
926
J. Fallon et al. / Journal of Orthopaedic Research 20 (2002) 920–926
tendons: changes with age and in chronic rotator cuff tendinitis. Ann Rheum Dis 1994;53:367–76. [21] Smith RG. Montage: a system for three-dimensional reconstruction by personal computer. J Neurosci Meth 1987;21: 55–69. [22] Smith RG. Three dimensional reconstruction from serial sections. User’s Manual Version 4.4. February 25, 1988. University of PA Medical School, Philadelphia, PA. [23] Snyder SJ. Evaluation and treatment of the rotator cuff. Orthop Clin North Am 1993;1:173–92.
[24] Steiner D, Hermann B. Collagen fiber arrangement of the human shoulder joint capsule—an anatomical study. Acta Anat 1989;136: 300–2. [25] Troyer H. Principles and techniques of histochemistry. Boston: Little Brown and Company; 1980. [26] Vogel KG, Heinegard D. Characterization of proteoglycans from adult bovine tendon. J Biol Chem 1985;260:9298–306. [27] Vogel KG, Koob TJ. Structural specialization in tendons under compression. Int Rev Cytol 1989;115:267–93.