Tenon's Capsule: Ultrastructure of Collagen Fibrils in Normals and Infantile Esotropia

Yaacov Shouly,* Benjamin Miller,* Choim Lichrig,t Michaela Modan4 and Ewy Meyer*

No detailed information about the ultrastructure of Tenon's capsule has been published. The purpose of the present study was to compare the ultrastructural features of collagen fibrils from Tenon's capsule in a nonstrabismic control group (seven children) to those in an infantile esotropic group (10 children). Small biopsy specimens from Tenon's capsule were taken during various operations to be examined by electron microscopy. On electron microscopy, the capsule was found to be composed of groups of collagen fibrils arranged irregularly in different orientations, forming a three-dimensional network that provides tissue resistance to stress. The cross-sectioned collagen fibrils were studied by an image analyzer. In both study groups, all fibrils had a round, regular contour. In the esotropic group, the Tenon's collagen fibrils were thicker, as reflected by their significantly greater mean diameter: 101 ± 5 nm (mean ± standard deviation) compared to 86 ± 5 nm in the control group. Also, significantly greater heterogeneity was found in the collagen fibril thickness of each individual in the esotropic group compared to the control group. Moreover, the mean number of collagen fibrils per unit area was significantly higher in the esotropic group: 98 ± 13 fibrils per 106 nm2 compared to 73 ± 5 fibrils per 106 nm2 in the control group. These ultrastructural changes may be stress-induced secondary alterations of the Tenon's collagen fibrils resulting from prolonged deviation of the eye in infantile esotropia. The significantly denser collagen fibrils may cause a decrease in the elasticity of Tenon's capsule in infantile esotropia. Invest Ophthalmol Vis Sci 33:651-656,1992

Tenon's capsule is a fascia made of dense connec- portant role in the physiology of ocular motility by tive tissue that covers the eyeball and the extraocular serving as a cavity within which the eyeball may muscles. This fascia was first described in 1806 by the move. Parisian ophthalmologist and anatomist Jacques Parks4 and other authors56 stress the importance of Rene Tenon.1 There are several synonyms for this fas- Tenon's capsule in squint surgery. An inadequate sur- cia, such as the fascia bulbi and the tunica vaginalis gical technique may cause restriction in ocular motil- oculi, but the most common name is Tenon's capsule. ity due to scarring and adhesions of the capsule to the The four extraocular rectus muscles penetrate extraocular muscles. Tenon's capsule posterior to the equator, whereas the We decided to study the ultrastructural features of two oblique muscles penetrate anterior to the equa- Tenon's capsule because, in reviewing the literature of tor.2 The anterior Tenon's capsule extends from the the last 20 years, we could not find one study on this four muscle penetrations to the limbus. The posterior topic. Long-standing deviation of the eye in one posi- Tenon's capsule lies between the four rectus muscle tion, such as in infantile esotropia, may cause shorten- penetrations and the optic nerve. Not only does ing and loss of elasticity of the conjunctiva and Tenon's capsule support and protect the globe, but it Tenon's capsule.7-8 Because, to the best of our knowl- also connects the eyeball to the orbit.3 It plays an im- edge, no data about the electron microscopic basis of these changes had been previously published, elucida- tion of the ultrastructural features of the capsule seemed appropriate. From the Departments of ""Ophthalmology and fPathology, The main purposes of this research were (1) to Rambam Medical Center and Bruce Rappaport Faculty of Medi- study the ultrastructural features of the collagen fibrils cine, Technion-Israel Institute of Technology, Haifa, and tthe De- of Tenon's capsule in children without strabismus partment of Clinical Epidemiology, Sheba Medical Center, Tel (control group); (2) to study the ultrastructural fea- Hashomer, and Sackler Faculty of Medicine, Tel Aviv University, tures of these collagen fibrils in children with infantile Tel Aviv, Israel. Submitted for publication: February 22, 1991; October 9, 1991. esotropia (esotropic group); and (3) to compare the Reprint requests: Ewy Meyer, MD, Department of Ophthalmol- ultrastructural features of these collagen fibrilsi n both ogy, Rambam Medical Center, POB 9602, 31096 Haifa, Israel. groups.

651

Downloaded from iovs.arvojournals.org on 09/29/2021 652 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / March 1992 Vol. 33

Materials and Methods In each case, a 2 X 2 mm biopsy specimen was taken from the anterior Tenon's capsule overlying the Subjects medial rectus muscle 8-10 mm from the limbus. On The study included 17 children in two study removal, the specimens were immediately cut into groups. small pieces and fixed with ice-cold 3% glutaralde- The control group included seven children without hyde in cacodylate buffer, pH 7.3, for 1 hr, postfixed strabismus who were operated on under general anes- with 2% osmium tetroxide, dehydrated through a se- thesia for congenital cataract (n = 3), traumatic cata- ries of graded alcohols (50-100%), and embedded in ract (n = 2), retinal detachment (n = 1), or intraocular epon. At least four blocks were cut from each biopsy foreign body (n = 1). Their mean age at operation was specimen. Sections 1 ^m thick were stained with tolu- 5.9 ± 5.6 yr (mean ± standard deviation; range, 1 mo idine blue and examined under the light microscope. to 12 yr). The mean refraction of four children in the Thin sections were stained with uranyl acetate and control group was +1.9 ± 1.75 diopters (mean ± SD; lead citrate and were examined with a Zeiss (Oberko- range, -2.0 to +3.25 D). chen, Germany) 9S electron microscope. Small biopsy specimens were taken from Tenon's capsule during the above-mentioned operations, Electron Microscopy and Individual Image Analysis which were associated with tenotomy as an integral part of the surgical procedure. Electron micrographs of Tenon's capsule, enlarged The esotropic group included 10 children operated 3800 times, were taken from areas away from blood on because of infantile esotropia. Their mean age at vessels. Micrographs of cross-sectioned and longitudi- operation was 2.9 ± 2.2 yr (range 1.5-8 yr). Their nally sectioned collagen fibrils, enlarged 56,000 times, mean refraction was +2.25 ± 2.0 D (range -1.5 to also were taken from these regions. For image analysis +5.5 diopters), and their mean angle of deviation was and morphometry, micrographs of the cross-sec- 54 ± 15 prism D (range 30-80 prism D). Five had tioned fibrilsthe n were enlarged an additional 1.9881 dissociated vertical deviation, five had overaction of times (so the final enlargement was 111,333.6 times). inferior oblique muscles, four had amblyopia, two Image analysis was performed using the Stereology had nystagmus, and one had anomalous head pos- program of the Morphomat 30 image analyzer (Zeiss, ture. Oberkochen, Germany). A histologic window with a In all cases, the diagnosis of infantile esotropia was constant area of 2,420,297 nm2 was defined in each made by an ophthalmologist in the pediatric ophthal- electron micrograph. Each histologic window in- mology clinic of Rambam Medical Center. In all cluded 163-304 collagen fibrils. For each patient, cases, manifest esotropia was diagnosed prior to the image analysis of three histologic windows was per- age of 6 mo. The children were treated and followed formed. Final image analysis was based on the pooled up in our pediatric ophthalmology clinic. Only those data of the three histologic windows because prelimi- patients with clinical characteristics of essential infan- 9 nary analysis showed very little variability between tile esotropia were included in the study. the characteristics of the three windows of each indi- Seven patients underwent bilateral recession of the vidual. medial rectus muscles combined with bilateral reces- The pooled area of three windows is denoted as Ar. sion of the conjunctiva and Tenon's capsule over the For each pooled image (j), the number of fibrils (Nj) medial rectus muscles. Three patients underwent re- as well as the area (Aij) and the diameter (Dij) of each section of the lateral rectus muscle and recession of fibril (ij) was measured. Five individual summary fi- the medial rectus muscle combined with recession of bril characteristics were calculated for each pooled the conjunctiva and Tenon's capsule. Small biopsies image, as follows: from Tenon's capsule were taken during these operations. All operations were performed under general 1. Point density (PDYj): The number of fibrils per 6 2 anesthesia. 10 nm of the tissue section, ie, The study was approved by the Helsinki Commit- PDYj = (Nj X 106)/Ar tee of Rambam Medical Center. In all subjects, preop- erative parental informed consent was obtained after 2. Mean fibril diameter (nm) of the individual (Dj): the nature of the biopsy procedure had been fully explained. Qj = 2 DiJ/Nj u=i Tissue Processing 3. Mean fibril area (nm2) of the individual (Aj): Biopsies from Tenon's capsule always were taken during the same procedure and by the same surgeon Aj = 2 Aij/Nj (E.M.) during the various operations. ij

Downloaded from iovs.arvojournals.org on 09/29/2021 No. 3 TENON'5 CAPSULE: ULTRASTRUCTURE OF COLLAGEN FIBRILS / Shouly er ol 653

4. Within-individual standard deviation (nm2) of fibril area (SD Aij): Zjjl, (Aij ~ Aj)2 SD Aij = Nj- 1 5. Volume density (VDYj): The fraction of the tissue volume occupied by the fibrils, ie,

VDYj = 2 Aij/Ar

Statistical Comparison Between the Study Groups Mean values ± SD (followed by ranges in parenthe- fmrn ses) pertaining to each of the fiveindividua l summary fibril characteristics were computed for each of the Fig. 2. Cross-sectioned collagen fibrils of Tenon's capsule from a study groups. Statistical comparison of the distribu- child without strabismus. Note the round, regular contour of the fibrils and the variation in size between the fibrils. TEM, original tions of the summary fibril characteristics between magnification X56,000; scale bar = 100 nm. the control group and the esotropic group was performed by Wilcoxon's rank-sum test because of in- equality of the variances between the two groups. posed of groups of collagen fibrilsarrange d irregularly in different orientations (Fig. 1). The longitudinal, Results horizontal, and oblique groups of collagen fibrils Light Microscopy formed a three-dimensional network. The longitudi- nally sectioned collagen fibrils showed a delicate, Examination of Tenon's capsule by light micros- wavy appearance. The collagen fibrilsforme d the ma- copy showed dense connective tissue with abundant jor component of the tissue, with only a few fibro- collagen fibers, irregularly arranged around a few fi- blasts interspersed between. broblasts. Collagen fibrils:O n high magnification (X 56,000), most cross-sectioned collagen fibrils in both groups Electron Microscopy had a round, regular contour (Figs. 2 and 3). Variabil- Tenon's capsule (general view): On electron micros- ity of diameter and of area of the collagen fibrils was copy with 3800-power magnification, we did not no- evident. tice a remarkable difference between the basic structure of Tenon's capsule in the control group and in the esotropic group. In both groups the tissue was com-

Fig. 1. Tenon's capsule from a child without strabismus. Note Fig. 3. Cross-sectioned collagen fibrilso f Tenon's capsule from a that the groups of collagen fibrils are arranged irregularly in various child with infantile esotropia. Note the larger diameter and area of orientations. F, nucleus of fibroblast.Transmissio n electron micros- the fibrils, as compared to Figure 2. TEM original magnification copy (TEM), original magnification X3800; scale bar = 2 jim. X 56,000; scale bar = 100 nm.

Downloaded from iovs.arvojournals.org on 09/29/2021 654 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / March 1992 Vol. 33

Comparison of the Two Study Groups The number of collagen fibrils per unit area, the size of the individual fibril, and the within-individual variability of fibril size all were larger in the esotropic group (Fig. 3) than in the control group (Fig. 2). Point density: There was complete separation of the two study groups regarding the point density of each individual PDYj (Fig. 4). The mean PDYj of the esotropic group was higher than that of the control group. The group mean PDYj was 98 ± 13 fibrils per 106 nm2 (79-117 fibrils per 106 nm2) in the esotropic group, compared to 73 ± 5 fibrils per 106 nm2 (68-78 fibrils per 106 nm2) in the control group. The differ- 4000 ence in the distributions between the two groups was Control Group Esotropic Group highly significant (P < 0.001). Collagen fibril diameter: The group mean of mean Fig. 5. Mean collagen fibril area of each individual (Aj) in the control group and in the esotropic group. fibril diameter of each individual (Dj) in the esotropic group was larger than that in the control group—101 ± 5 nm (94-108 nm) versus 86 ± 5 nm (77-91 nm). (670-848 nm2) in the esotropic group, compared to The difference in the distributions between the two 587 ± 79 nm2 (428-673 nm2) in the control group. In groups was highly significant (P < 0.001). addition, there was almost complete separation of the Collagen fibril area: As expected from the differ- two study groups in regard to the individual SD of the ence in the fibril diameter, fibril area of the esotropic collagen fibril area (Fig. 6). The finding indicates that group was larger than that of the control group. The more heterogeneity existed in the collagen fibril area group mean of the mean fibril area of each individual of each individual in the esotropic group than in the (Aj) was 7352 ± 749 nm2 (6432-8536 nm2) in the control group. esotropic group, compared to 5299 ± 561 nm2 (4313- Volume density: The VDYj of the esotropic group 5934 nm2) in the control group. Moreover, there was was higher than that of the control group. The group complete separation of the two study groups with re- mean VDYj was 0.72 ± 0.08 (0.62-0.81) in the eso- spect to the mean fibril area of each individual (Fig. tropic group, compared to 0.39 ± 0.03 (0.34-0.43) in 5). The difference in the distributions between the two the control group. The difference in the distributions study groups was highly significant (P < 0.001). between the two study groups was highly significant The within-individual standard deviation of fibril (P< 0.001). area (SD Aij) in the esotropic group was significantly The greater volume density in the esotropic group higher than that in the control group (P < 0.001). The resulted from their greater number of fibrils per unit group mean of this characteristic was 767 ± 54 nm2 area and from their larger mean fibrilarea . Thus, simi-

900

115 " 800 -

1_ 105 0> Q. E 700 S •^ S- 600 • o 85 Q> E 75 500 (nu >- Q Q. 65 400 Control Group Esotropic Group Control Group Esotropic Group

Fig. 4. Point density of each individual (PDYj) in the control Fig. 6. Standard deviation of collagen fibril area of each individ- group and in the esotropic group. ual (SD Aij) in the control group and in the esotropic group.

Downloaded from iovs.arvojournals.org on 09/29/2021 No. 3 TENON'S CAPSULE: ULTRASTRUCTURE OF COLLAGEN FIBRILS / Shouly er ol 655

lar to these two latter fibril characteristics, there was a reflected by their higher volume density. In addition, complete separation between the two study groups re- there was greater heterogeneity in fibril thickness of garding individual values of volume density (data not the esotropic group, as demonstrated by their higher shown). within-individual SD of fibril area. Periodicity of collagen fibrils: In longitudinal sec- Generally, most collagen fibrils are characterized tions, the periodical banding of the fibrils was uni- by fairly uniform shape and diameter in most mature form, and the length of each period was 60-64 nm. normal tissues.11 However, abnormal variations in No difference of periodicity existed between the con- the size and shape of the fibrilshav e been described in trol group and the esotropic group. various diseases.1617 The exact causes of increased fibril diameter and greater heterogeneity of fibril thick- Discussion ness in the esotropic group are unknown. Our hy- pothesis is that the ultrastructural changes of Tenon's In reviewing the literature of the last 20 years, we capsule are secondary changes associated with infan- could not find a single study concerning the ultrastruc- tile esotropia. Prolonged deviation of the eye in one ture of Tenon's capsule. The main contribution of the position may exert tension or stress on the capsule, present study is to present for the first time detailed stimulating fibrosis or healing processes associated information on the ultrastructural features of the col- with increased diameter and greater heterogeneity in lagen fibrils of Tenon's capsule in nonstrabismic eyes fibril thickness. and in eyes with infantile esotropia. After corneal wounds heal, the diameter of the re- Using transmission electron microscopy, Tenon's generated collagen fibrils is greater than normal.15 capsule was found to be composed of groups of colla- Also, fibrils of large and small diameter have been gen fibrils arranged irregularly in different orienta- found in the dermis of patients undergoing wound tions (Fig. 1). The three-dimensional network formed repair.16 Also, some studies have shown there may be by groups of collagen fibrils may provide tissue resis- a relationship between mechanical stress and ultra- tance to stress. This basic formation also may enable structural features of fibrils in different connective tis- Tenon's capsule to play an important role in support- sues. Matthew et al18 studied quantitatively the ultra- ing, protecting, and connecting the eyeball to the or- structure of collagen fibrils in the healing extensor di- bit. A different arrangement of the collagen fibrils is gitorum longus tendon of the rat. Collagen fibrils with found in the corneal lamellae, which are composed of significantly larger diameters were synthesized only in very regularly spread, uniform collagen fibrils that response to increased levels of stress, in contrast to provide corneal transparency.10 smaller fibrils, which were synthesized in areas of low In the control group, which we believe represents and high stress levels. normal Tenon's capsule tissue, mean collagen fibril Similarly, Stransky et al19 reported on heterogene- diameter was 86 ± 5 nm and mean area was 5299 ity in the diameter of collagen fibrils in idiopathic car- ± 561 nm2. Quantitative analysis of electron micro- pal tunnel syndrome. They demonstrated the pres- graphs has shown that the diameter of collagen fibrils ence of fibrils with small diameters comparable to may range from 20-120 nm in most connective tis- that in normal tissue, as well as fibrils with large diam- sues.11 The mean diameter of the collagen fibrils in eter that could not be observed in normal tissue. normal human skin is considered by most authors to Thus, we suggest that the ultrastructural alterations in 12 14 be 90-100 nm. " Thus, normal collagen fibrils of the collagen fibrils of Tenon's capsule in infantile eso- the reticular dermis are slightly larger than those of tropia may be stress-induced secondary changes, simi- Tenon's capsule. In contrast, most collagen fibrils of lar to those described in other connective tissues. 15 the normal cornea measure 27-35 nm and are much Lack of size uniformity and significantly larger colla- smaller than those of normal Tenon's capsule. Mean gen fibrils forming the fibrotic lateral rectus muscle point density of the Tenon's collagen fibrils in the 6 2 also have been described in patients with infantile eso- control group was 73 ± 5 fibrils per 10 nm , and tropia and myopia.20 mean volume density was 0.39 ± 0.03. Von Noorden7 and Helveston8 point out that long- In the esotropic group, the Tenon's collagen fibrils standing esotropia may cause shortening and de- were significantly thicker than those in the control creased elasticity of the conjunctiva and Tenon's cap- group, as reflected by their larger cross-sectional sule. Our study may provide the ultrastructural basis areas. In addition, there were more fibrils per unit of for their clinical observations. It is known that the Tenon's tissue volume in the esotropic group, indi- collagen fibers, in contrast to the elastic fibers of the cated by their higher point density compared to the connective tissue, are not extensible.21 Therefore, the control group. As a result, the Tenon's collagen fibrils increase in the number and thickness of the collagen of the esotropic group were more densely arranged, as fibrils, resulting in their greater density, may cause a

Downloaded from iovs.arvojournals.org on 09/29/2021 656 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / March 1992 Vol. 33

decrease in the connective tissue elasticity of Tenon's 6. O'Donoghue HN and Smith AB: Importance of Tenon's cap- capsule in infantile esotropia. sule in squint surgery. Trans Ophthalmol Soc UK 102:492, The main surgical procedures for the treatment of 1982. 7. von Noorden GK: Principles of surgical treatment. In Binocu- infantile esotropia are bilateral recession of the medial lar Vision and Ocular Motility, 4th ed, Klein EA and Falk KH, rectus muscles or recession of the medial rectus mus- editors. St Louis, C.V. Mosby, 1990, pp. 525-526. cle combined with resection of the lateral rectus mus- 8. Helveston EM: Atlas of Strabismus Surgery. St Louis, C.V. cle. The effectiveness of medial rectus muscle reces- Mosby, 1985, p. 234. sion may decrease if the conjunctiva and Tenon's cap- 9. von Noorden GK: A reassessment of infantile esotropia (44th Edward Jackson Memorial Lecture). Am J Ophthalmol 105:1, sule are returned to their original position at the end 1988. of the operation. On the other hand, recession of the 10. Edelhauser HF, Van Horn DL, and Records RE: Cornea and conjunctiva and Tenon's capsule can significantly en- sclera. In Biomedical Foundations of Ophthalmology, Tasman hance the effectiveness of medial rectus muscle reces- W and Jaeger EA, editors. Philadelphia, J. B. Lippincott, 1990, sion by decreasing the restrictive effect caused by vol. 2, p. 13. 11. Ghadially FN: Ultrastructural Pathology of the Cell and Ma- shortened and inelastic Tenon's capsule. As men- trix. London, Butterworths, 1988, vol. 2, pp. 1215-1225. tioned above, the decreased elasticity of Tenon's cap- 12. Vogel A, Holbrook KA, Steinmann B, Gitzelmann R, and sule in infantile esotropia may be explained according Byers PH: Abnormal collagen fibril structure in the gravis form the ultrastructural findings reported here. (type I) of Ehlers-Danlos syndrome. Lab Invest 40:201, 1979. 13. Hayes RL and Rodnan GP: The ultrastructure of skin in pro- Further research with a larger number of subjects is gressive sclerosis (scleroderma). Am J Pathol 63:433, 1971. needed to determine more definitely the characteris- 14. Ludatscher RM, Friedman-Birnbaum R, Bergman R, and tics of collagen fibrils in Tenon's capsule in nonstra- Lichtig C: Structural alterations of collagen fibrils in the reac- bismic patients and in those with infantile esotropia. tive type of elastosis perforans serpiginosa. Arch Dermatol Res 280:319, 1988. Key words: Tenon's capsule, ultrastructure, image analysis, 15. Rodrigues MM, Warnig GO III, Hackett J, and Donohoo P: collagen fibrils,infantil e esotropia Cornea. In Biomedical Foundations of Ophthalmology, Tas- man W and Jaeger EA, editors. Philadelphia, J. B. Lippincott, 1990, vol. l,p. 6. Acknowledgments 16. Holbrook KA and Byers PH: Structural abnormalities in the The authors wish to thank Mrs. Zelda Weinberg for excel- dermal collagen and elastic matrix from the skin of patients lent technical assistance and Miss Ruth Singer for typing the with inherited connective tissue disorders. J Invest Dermatol manuscript. 79(suppl l):7s, 1982. 17. Meyer E, Ludatscher RM, and Zonis S: Collagen fibril abnormalities in the extraocular muscles in Ehlers-Danlos syn- References drome. J Pediatr Ophthalmol Strabismus 25:67, 1988. 18. Matthew C, Moore MJ, and Campbell L: A quantitative ultra- 1. Duke-Elder S: System of Ophthalmology. London, Henry structural study of collagen fibril formation in the healing ex- Kimpton, 1961, vol. 2, p. 450. tensor digitorum longus tendon of the rat. J Hand Surg [Br] 2. Parks MM: Extraocular muscles. In Duane's Clinical Ophthal- 12:313, 1987. mology, Tasman W and Jaeger EA, editors. Philadelphia, J. B. 19. Stransky G, Weis S, Neumuller J, Hakimzadeh A, Firneis F, Lippincott, 1990, vol. 1, p. 7. Ammer K, Partsch G, and Eberl R: Morphometric analysis of 3. von Noorden GK: Summary of the gross anatomy of the extra- collagen fibrils in idiopathic carpal tunnel syndrome. Exp Cell ocular muscles. In Binocular Vision and Ocular Motility, 4th Biol 55:57, 1987. ed, Klein EA and Falk KH, editors. St. Louis, C.V. Mosby, 20. Meyer E, Ludatscher RM, Lichtig C, Shauly Y, and Gdal-On 1990, pp. 49-50. M: End-stage fibrosis of the lateral rectus muscle in myopia 4. Parks MM: The role of fascia in muscle surgery. International with esotropia: An ultrastructural study. Ophthalmic Res Ophthalmology Clinics 16:17, 1976. 22:259, 1990. 5. Price RL: Role of Tenon's capsule in postoperative restrictions. 21. Lever WF and Lever GS: Histopathology of the Skin, 6th ed. International Ophthalmology Clinics 16:197, 1976. Philadelphia, J. B. Lippincott, 1983, pp. 30-31.

Downloaded from iovs.arvojournals.org on 09/29/2021