This histologic investigation (light microscopy, immunohistochemistry, and enzymohistochemistry) of the insertion of the lateral and medial extraocular muscles in humans demonstrated large amounts of muscle tissue that contained only minimal islands of tendon and connected directly with the sclera. A distinct tendon was not demonstrated. These findings contrast with former studies
6 7 11 29 that reported a proper muscle tendon connecting the muscle to the sclera. However, in only one of these former studies were histologic methods used for tissue identification. Histologic evidence of the presence of a tendon was only presented in one study of the superior oblique muscle.
30 Whether the term “tendon” is appropriate for extraocular muscles has been questioned by Apt,
7 who at the same time has raised doubt about the accuracy of the measurements of the length of the tendons reported in previous articles. He has suggested that the term “aponeurosis” would be more appropriate than the term tendon. Most of the literature dealing with eye movements has not fully explained eye muscle functions based on current anatomic knowledge.
29 31 32 33 The recent discovery of the pulley system has delivered powerful functional and anatomic insight into the way eye movements take place.
31 32 34 35 36 37
With classic histologic staining such as hematoxylin-eosin or Goldner, tendon and sclera cannot be distinguished. Therefore, splitting off of scleral fibers due to inadvertent dissection could be misinterpreted to represent tendon
(Fig. 2) . We used anti-tenascin C antibodies as a reliable tissue marker for distinguishing among muscle, tendon, and sclera. Tenascin C has been demonstrated in the limbus region of the eye, conjunctiva, Descemet’s membrane, and the tunica interna of blood vessels. The high specificity of tenascin C for collagen-containing tendon tissue is reflected by the lack of reaction to the connective tissue in the endomysium and the positive reaction to the connective tissue in the walls of blood vessels and in the perineurium.
22 23 24 25 26 27 28 The reaction of these tissues was used as a positive control.
In contrast to other publications examining the “tendons” of extraocular muscles, we were unable to demonstrate the presence of a true tendon connecting muscle directly with the sclera in neither the medial nor the lateral rectus muscles. In the region of the muscle insertion (scleromuscular junction), we found muscle tissue containing only small islands of tendon tissue in the vicinity of the sclera. Such a situation was proposed by De Gottrau and Gajisin,
8 albeit without proper histologic or immunohistochemical evidence. The absence of Golgi-spindle organs in the distal portion of the extraocular muscles, as described by Blumer et al.,
38 Bruenech and Ruskell,
39 and Richmond et al.,
40 correlates with our findings of the paucity of tendon tissue in that region. The presence of motor endplates is further evidence for muscle tissue in the scleromuscular junction.
Electromyographic (EMG) recordings demonstrating EMG potentials (Killer HE, personal experience) in the vicinity of the insertion of the lateral rectus and medial rectus muscle supplies clinical evidence for the presence of muscle endplates near the insertion of the muscle into the sclera. Innervated myotendinous cylinders (IMCs) were found abundantly in the proximal segments (insertional zone) of extraocular muscles.
2 10 As functional structures, IMCs are highly important for the proprioceptive innervation of human extraocular muscles.
41 Ophthalmologists performing eye muscle surgery should be aware that when they perform muscle resection they reduce the number of IMCs as well as the number of sarcomeres and motor endplates. In comparison with recession for surgical correction of strabismus, marginal myotomy was found to be more proprioceptively “deafferenting” than recession, probably due to a greater disruption of palisade endings.
42 Our findings showing that the medial and lateral recti insert without a tendon directly into the sclera indicate that resection of muscle tissue containing proprioceptive elements may also be more “deafferenting” than recession. Eye muscles are neded to perform extremely rapid (saccades) and highly coordinated movements (versions). To perform these movements, it is desirable that the generator of force (muscle) and the object being rotated (eyeball) be in direct contact. The interposition of a tendon between muscle and eyeball would probably impair the efficiency of ocular muscle function.
The authors thank Michael C. Brodsky, MD (Arkansas Children’s Hospital, Little Rock, AR), and Jody Stähelin, MD (Department of Pediatrics, Kantonsspital, Aarau, Switzerland), for helpful comments on the manuscript.