Repair of temporomandibular joint disc perforation using a synovial membrane flap in Macaca fascicularis monkeys: Light and electron microscopy studies

J Oral Maxillofac Surg 52:259-270.1994

Repair of Temporomandibular Joint Disc Perforation Using a Synovial Membrane Flap in Macaca fascicularis Monkeys: Light and Electron Microscopy Studies MOHAMED M. SHARAWY, ROBERT A. BAYS,

BDS,

PHD,*

EMAD

S. HELMY,

DDS,$- AND VERA B. LARKE,

BDS, MS,t BS$

Previous studies have demonstrated that experimentally produced perforations in the discs of Macaca fascicularis monkeys lead to osteoarthrosis. Synovial membrane hyperplasia also was demonstrated in monkey and human joints with disc perforations. The hypothesis was advanced that a synovial flap obtained from within the affected joint would be the most appropriate tissue to repair chronic disc perforation. To test this hypothesis, four adult M fascicularis monkeys were anesthetized and 4- to 6-mm perforations were made in the posterolateral aspects of the avascular discs bilaterally. The wounds were sutured leaving the perforations open, and the animals were fed their normal diet. After 4 weeks, one joint in each monkey was reopened and a repair was performed using a double-layered flap from the synovial lining of the superior and inferior recesses. Four weeks after repair, the animals were killed and the temporomandibular joints (TMJs) were removed en bloc and decalcified. The joints were sectioned into lateral, middle, and medial sections and were photographed using a stereomicroscope and then processed for light and electron microscopy. The same processing was done to four intact joints that were used as controls. Eight weeks following perforation, the joint components showed degenerative changes consistent with osteoarthritis. Close to the perforations the disc showed loss of collagen, vacuolation of extracellular matrix, accumulation of dense proteoglycanlike material, and the appearance within the disc of type A or macrophage-like cells of the synovium. The discal tissue away from the perforation showed high cellularity and vascularity. The temporal and condylar surfaces showed denudation, fibrillation, osteophytes, and chondrocytic clustering, all characteristics of osteoarthrosis. The surgically repaired discs were intact and the articular surfaces showed no degenerative changes. Discal collagen was restored and appearance of myofibroblasts and elastogenesis were a consistent feature of the repaired disc. The vascularity of the condylar cartilage of the repaired joints appeared similar to that of embryonic cartilage. The reversibility of the degenerative alterations following discal repair in this experimental model should provide the basis for a rational and useful method for surgical repair of TMJ disc perforation using intraarticular synovial tissue.

* Professor of Anatomy and Director of Craniofacial Research, Dental Research Center, Medical College of Georgia, School of Dentistry, Augusta, GA. t Lecturer of Oral and Maxillofacial Surgery, Department of Oral Surgery, Faculty of Oral and Dental Medicine, Cairo University, Cairo, Egypt. $ Associate Professor and Chief. Department of Oral and Maxillofacial Surgery. Emory University, Atlanta, GA.

0 Electron Microscopy Technician, School of Dentistry, Medical College of Georgia, Augusta, GA. Address correspondence and reprint requests to Dr Sharawy: Department of Oral Biology, School of Dentistry, Medical College of Georgia, Augusta, GA 309 12. 0 1994 American Association of Oral and Maxillofacial Surgeons 0278-2391/94/5203-0009$3.00/O

259

260

REPAIROF DISCALPERFORATIONWITH SYNOVIUM

The temporomandibular joint (TMJ) disc functions as a shock absorber and is thought to be essential for protection of both the condylar and temporal components of the TMJ against mechanical stress.‘~2Several experimental and clinical studies support this view. Degenerative osteoarthrosis was reported to occur following the surgical removal of TMJ discs3-5 and knee menisci.6x7 In our laboratory, experimental creation of perforated TMJ discs in rhesus monkeys also caused osteoarthrosis of the affected joints8 In that experiment, among the most conspicuous regenerative responses to disc perforation were the severe hyperplastic changes of the synovial membrane and the migration of synovioblasts on the surfaces of the injured disc. Synovioblasts also line the margins of TMJ disc perforations in humans.g It was speculated that the synovial membrane was attempting unsuccessfully to repair the large perforations. Since disc perforation is considered detrimental to the health of the TMJ,8 surgical repair should be done. The present study was designed to test the hypothesis that a vascular synovial flap incorporated into the approximated edges of a perforated disc will result in a successful repair of the perforation. The results of this experiment tend to support this hypothesis. Materials and Methods ANIMALS Four male, 9- to lZyear-old, Macaca fascicdaris monkeys were used in this experiment. Two frozen heads of monkeys of the same species and of similar age were obtained from the Yerkes Primate Institute in Atlanta (GA) and their TMJs were excised and processed similar to the experimental joints for use as intact controls.

perforation was created at the posterolateral aspect of the avascular portion of the disc using an electrosurgery loop. Extreme caution was taken to avoid touching the underlying condyle or to permit bleeding into the joint spaces. The capsule was closed using 5-O Vicryl (Ethicon, Somerville, NJ) suture, and the skin incision was closed in layers using absorbable sutures. Ice was applied to the site of surgery for 10 to 15 minutes, and the animal was given antibiotics and analgesics. The same procedure was repeated on the opposite side of each animal. Following surgery, the animals were housed one per cage and provided with laboratory monkey chow supplemented with fruits and water ad libitum. There were few or no postoperative complications following surgery. Four weeks postperforation, one experimental side was surgically exposed and the disc and condylar and temporal surfaces were observed and photographed. A synovial flap procedure was done to repair the disc as described below. The opposite joint served as a perforated control. SYNOVIALFLAP PROCEDURE The edges of the perforations were trimmed, approximated, as best as possible, and sutured together using interrupted 4-O absorbable sutures. It was not always possible to close the central portion of the perforation edge to edge. The synovial membrane from the posterosuperior and posteroinferior recesses was undermined, pulled forward, and sutured to the discal wound with 4-O absorbable sutures. The disc was then sutured to the lateral collateral ligament and to the capsule. The subcutaneous tissues were closed and then the skin incision was sutured with 3.0 silk sutures. Ice was applied postoperatively and the animals were administered long-acting analgesics (Inovar [Janssen Pharmaceutics, Inc, Piscataway, NJ] for 4 days).

DISC PERFORATIONPROCEDURE

FIXATIONAND PREPARATIONOF TISSUESFOR LIGHT AND ELECTRON MICROSCOPY

Each monkey was anesthetized with an intramuscular injection of ketamine (20 mg/kg body weight [SW]) and acepromezine (1 mg/kg BW). The preauricular skin was shaved at both sides of the head, scrubbed, and sterilized with antiseptic solution, and the animal was draped except to expose the surgical field. To control bleeding and ensure anesthesia, a subcutaneous injection of 2% xylocaine containing l/lOO,OOO epinephrine was given. A tranzygomatic incision was made and the fascia dissected to expose the lateral joint capsule. The jaw was then manipulated to locate the condyle. An incision parallel to the zygomatic arch was made through the capsule to expose the upper joint space. The angle of the mandible was grasped with a towel clamp and pulled inferiorly to allow better access to the joint space. A 4- to 6-mm through-and-through

Four weeks following repair the four experimental monkeys were anesthetized and perfused through the carotids with Karnovsky’s fixative. Skull blocks containing the TMJs were excised and immersed in the same fixative for at least 48 hours. The joints with intact discs, perforated discs, and repaired discs were decalcified in ethylenediamine tetraacetic acid for 5 to 6 weeks. The decalcified blocks were sectioned in a sagittal direction into lateral, middle, and medial sections with a sharp razor blade. The sections were studied and photographed with a Zeiss stereomicroscope. Each of the thick sections were mounted flat on a metal chuck and one or two approximately 200~pm sections were cut using a vibratome sectioning system (Lancer Vibratome Series 1000, Redding, CA). These thin sections were used for obtaining l- to 2-mm3 samples for

SHARAWY

261

ET AL

the electron microscopic portion of the study, as explained below. The remainder of the thick sections were dehydrated in ascending grades of alcohol, cleared in xylene, and infiltrated and embedded in paraffin. The vibratome sections were viewed under a Zeiss stereomicroscope and 2- to 3-mm3 pieces of tissue from different portions of the disc (eg, anterior band, intermediate zone, posterior band, bilaminar zone, etc), temporal articular surface (articular eminence and glenoid fossa), and condylar surface were obtained and kept in separate bottles containing cold .2-mol/L sodium cacodylate buffer. The tissues were then dehydrated in ascending grades of alcohol, cleared in propylene oxide, infiltrated in a mixture of propylene oxide and araldite, and embedded in araldite. LIGHT MICROSCOPY The paraffin blocks were cut into 5-pm-thick sections and stained with hematoxylin-eosin (II&E), al&n blue (pH 2.5), resorcin fuchsin, and periodic acid-schiff reagent (PAS). The stained sections were observed and photographed using a Zeiss photomicroscope II. ELECTRONMICROSCOPY One-micrometer-thick araldite sections were obtained, stained with a mixture of crystal violet and alcian blue, and observed for orientation purposes and for further trimming of the blocks without cutting away valuable tissues. The trimmed blocks were then cut using a diamond knife and Reichert OMU 3 Ultramicrotome. The ultrathin sections (500 to 700 nm) were mounted on copper grids and stained with uranyl acetate and lead citrate. Some grids were stained to demonstrate elastin using uranyl acetate in methanol alcohol.” The stained grids were observed and photographed using a Joel X 100 electron microscope. Results CONTROLTMJ Stereomicroscopy

The covering of the condyle, articular eminence, and glenoid fossa was smooth and glistening. The superior and inferior compartments were free of debris and their recesses were well defined. The disc had a smooth, glistening surface and could easily be discerned into anterior band, intermediate thin zone, posterior band, and bilaminar zone (Fig 1). A few fibers of the superior head of the lateral pterygoid muscle seemed to be inserted into the inferior aspect of the anterior band. The surface of the fibrous covering of the condyle and temporal bony component was smooth and uninterrupted (Figs 1, 2).

Histology

Both the condylar and temporal articular surfaces were covered with compact, avascular, fibroelastic connective tissue, with collagen constituting the most predominant component (Fig 3). The condylar covering contained a few fusiform fibroblasts and occasional rounded cells with lacunae and tutoria1 alcian blue-positive matrix best described as chondrocyte-like cells (Fig 4). A two-to-three-cellthick layer of prechondrocytic resting or undifferentiated mesenchymal cells was observed at the interface between the condylar fibrous covering and the underlying hyaline cartilage. The latter was characteristically alcian blue-positive (Fig 4). Sections of the condyle stained with resorcin fuchsin (not presented) demonstrated fine elastic fibers interwoven with the collagenous bundles of the fibrous covering. The fibrous covering of the temporal articular surface was thicker at the descending slope of the articular eminence and thinner at the depth of the glenoid fossa. The fibrous tissue consisted of woven collagen bundles and fine elastic fibers. The cells were mostly fibroblasts, but a few chondrocyte-like cells (alcian blue-positive) were consistently seen. Unlike the condylar covering, there was no subfibrous hyaline cartilage or reserve cell layer observed in the temporal fibrous covering (Fig 5). The TMJ disc seemed to be composed primarily of dense collagen bundles that were interwoven, partio ularly at the anterior and posterior band areas. Blood vessels were only seen in the bilaminar zone and the anterior extension of the anterior band. The majority of cells found in the control discs were fibroblasts. Some chondrocyte-like cells were also observed, and were more predominant in the intermediate zone and posterior band of the disc. The resorcin fuchsin-stained sections showed a remarkable network of elastic fibers in both the superior stratum and inferior stratum of the bilaminar zone and the posterior capsule. In contrast, far fewer elastic fibers, which ran parallel to the collagen bundles of the avascular disc, were seen (Fig 3). The density of the elastic fibers, however, increased at the most anterior portion of the anterior band and at its anterior extension. Synovial membrane was found in all the recesses of the TMJ spaces. It consisted of one-to-two-cell layers of intima facing the joint spaces and a fibroelastic vascular subintima (Fig 6). The cells of the synovium extended for short distances onto the adjacent discal, condylar, and temporal surfaces. Electron Microscopy

The disc consisted primarily of collagen organized into interlacing bundles, particularly at the anterior and posterior bands of the disc (Figs 7, 8). The inter-

262

REPAIR OF DISCAL PERFORATION

WITH SYNOVIUM

FIGURE I. Stereoscopic photomicrograph of a sagittal section of control TMJ showing anterior band (A), intermediate zone (I), posterior band (P), bilaminar zone (BL), articular eminence (AE), and condyle (c) (original magnification X2.4).

FIGURE 2. Stereoscopic photomicrograph of control joint at higher magnification showing the even, smooth covering of the condyle (c) and the lack of debris in the joint spaces (original magnification X4.0).

FIGURE 3. Sagittal section of control joint. Note the compact nature and the even thickness of the fibrous coverings of the condyle (c) and articular eminence (AE) (hematoxylin-eosin stain, original magnification X51).

FIGURE 4. Light micrograph of a portion of the superior aspect of articulating surface of control condyle showing the compact, smooth nature of the fibrous covering, the presence of reserve cell layers (arrows), secondary cartilage (cr), and bone and bone marrow (arrowhzds) (hematoxylin-eosin stain, original magnification X94).

FIGURE 5. Light micrograph of a portion of a control articular eminence stained with toluidine blue showing the compact, smooth nature of the surface collagen next to joint space (*). A moderate number of fibroblasts (arrows) are scattered among the collagen bundles. Chondroid bone (ChB) is seen under the fibrous covering (periodic acid-Schiff stain, original magnification X374).

FIGURE 6. Histologic section of the posterosuperior recess of control normal joint showing the cellular intima and the vascular subintimal connective tissue components of synovial membrane (periodic acid-Schiff stain, original magnification X94).

FIGURE 7. Electron micrograph of a portion of the anterior band of a control disc. Note the abundance of collagen, which runs in interwoven bundles, and the presence of a tibroblast (F) (original magnification X5,000).

FIGURE 8. Electron micrograph of a portion of anterior band of a control disc. Fibroblasts with long cell processes (arrows) were the most common cells seen in the disc (original magnification X4.000).

FIGURE 9. Electron micrograph of a portion of the intermediate zone of a disc showing the compact nature of the collagen fibrils and the presence of scant amounts of ground substance between the fibrils (original magnification x20,313).

SHARAWY

263

ET AL

mediate zone of the disc contained collagen fibers that ran parallel to one another (Fig 9). In the anterior and posterior bands, the collagen fibers varied in their crosssectional diameters (Figs 7,8, 10). The spaces between the collagen fibers contained scant amounts of electrondense material that had the fine structural characteristics of proteoglycans seen in cartilage (Fig 9). Elastic fibers with both amorphous and microfibrillar components were also seen. The majority of cells were fibroblasts characterized by abundant rough endoplasmic reticulum and numerous cell processes. The latter extended for long distances among the collagen fibers (Figs 7, 8). Occasionally, chondrocyte-like cells with a typical territorial matrix were observed (Fig 10). The fibrous covering of the temporal component consisted primarily of avascular collagen fibers running in different directions and a few fibrocytes. The cells had large nuclei and scant cytoplasm (Fig 11). The surface of the temporal fibrous covering was smooth and made of compact parallel collagen bundles. The fibrous covering of the condyle was avascular, consisting of compact collagen layers arranged perpendicular to one another, and a few fibroblasts (Fig 12). Between the fibrous covering and the underlying bone, layers of reserve cells and cartilage were seen (Fig 13). The reserve cells were spindle shaped, with large nuclei and thin cytoplasm that contained few profiles of rough endoplasmic reticulum (Fig 13).

EXPERIMENTALJOINTS 8 WEEKS FOLLOWINGPERFORATION

The stereoscopic and histopathologic features of the monkey TMJ following disc perforation have been described in detail in a previous publication (Helmy et a18). The results of the present study were similar to the previous findings; therefore, the stereomicroscopy and histopathology will be only briefly described and the emphasis will be on the electron microscopic results.

Histopathologic and Electron Microscopic Observations

The four experimental joints showed changes in the condyle, temporal surfaces, and the disc consistent with a diagnosis of secondary osteoarthrosis. Changes in the Condyle

The condyles were deformed and adherent to the perforated discs (Fig 16). Significant condylar bone resorption and degeneration of the secondary condylar cartilage were seen under the perforation. Vascular granulation tissue was consistently seen adjacent to areas of bone resorption (Fig 17). In one joint, focal areas of eburnated, denuded bone facing the joint space were seen (Fig 18). Bone microcysts open into the joint cavity also were seen (Fig 18). At the electron microscopic level, necrotic cells were observed close to the surface of the deformed condyle (Fig 19). Many enlarged bone marrow spaces within the condyle, with no intervening cartilage, were seen (Fig 20). In decalcified sections, the denuded bone matrix directly faced the joint space (Fig 2 1). Changes in the Temporal Surface

In contrast to the control, the temporal fibrous covering was remarkably thickened, highly cellular, and contained areas of hyalinization. Close to the overlying bone, vascular granulation tissue with a large number of osteoclasts adjacent to areas of bone resorption was consistently seen (Fig 22). In one joint, focal areas of bone denudation and eburnation were seen adjacent to an area covered by thick fibrocartilage with chondrocytic clustering (Fig 18). At the electron microscopic level, it was not uncommon to see numerous fibroblasts and blood vessels in the temporal fibrous covering, in contrast to the control joint that never showed vascularity at this site (Fig 23). Fibrillation or spacing of collagen bundles, particularly of the surface collagen, was a common finding (Fig 23).

Stereoscopic Observations

Discal Changes

The discal perforation in all four joints was larger both mediolaterally and anteroposteriorly than what was created surgically. Adhesion of the lips of the perforation to the underlying condyle and/or to the overlying temporal component was a consistent finding. Thickening and irregularities of the condylar and temporal at-titular coverings were seen in all four joints with perforated discs (Fig 14). Focal areas of polished, denuded bone of the condyle and the temporal articular surface were seen (Fig 15). Large, globular masses of what appeared to be hyperplastic synovial tissue filled the joint recesses (Fig 15).

Stereoscopically, the perforated discs were noted to be adherent to the articulating surfaces of the joint. The margin of the perforation appeared thin and translucent (Figs 24,25). Histologically, the perforated discs were abnormally vascular and cellular (Fig 26) particularly in the areas close to the margin of the perforation (Fig 27). The surface of the disc was covered by one or two layers of cells that could be traced to the hyperplastic synovium (Figs 26, 27). There were little or no inflammatory cells. Under the light microscope, the margins were abnormally vascular and cellular, and contained areas with minimal collagen (Fig 27). Sy-

264

REPAIR OF DISCAL PERFORATION

WITH SYNOVIUM

FIGURE 10. Electron micrograph of a portion of posterior band of a control disc. Chondroblast-like cells surrounded by amorphous territorial matrices are occasionally seen. Note the nonuniform cross-sectional diameters of the collagen fibrils, more noticeable in anterior and posterior bands of the disc (original magnification X 10,050).

FIGURE 11. Electron micrograph of a portion of the fibrous covering of the articular eminence. Note the interwoven collagen fibers, the presence of a fibrocyte (F),and the uniform thickness of the collagen librils (original magnification X6,000).

FIGURE 12. Electron micrograph of a portion of the fibrous covering of a control condyle. Note that the collagen bundles run perpendicular to each other, and the presence of tibroblast (F)(original magnification X2,700).

FIGURE 13. Electron micrograph of a portion of control condyle showing the undifferentiated nature of the reserve cells (R),interlacing collagen, and portion of a chondrocyte (Ch)(original magnification X2,869).

FIGURE 14. Stereoscopic photomicrograph of a joint 8 weeks postperforation. Note the adhesion of the lips of the perforation to the condyle, thinning of the fibrous covering of condyle, irregularities of condyle surface, and thickening of most of temporal fibrous coverings (original magnification X2.3). Compare with control (Fig 1).

FIGURE 15. Stereoscopic photomicrograph of a joint 8 weeks postperforation. In this joint the osteoarthritic changes are severe. Note the absence of the disc, which resulted in bone-tobone contact between the disfigured condyle and the temporal bony component. The bulbous masses in the joint space (arrows) are hyperplastic synovial membrane (original magnification X4.0).

FIGURE 16. Histologic sagittal section of a joint 8 weeks postperforation. Notice the alteration in shape of the condyle, fibrillation of the disc, and the presence of hyperplastic synovial membrane in all joint recesses (arrows) (hematoxylin-eosin stain, original magnification X21).

FIGURE 17. Light micrograph of superior aspect of condyle under discal perforation. Note the marked resorption of bone and adjacent granulation tissue (arrows), the presence of pyknotic nuclei of cells near the condylar surface (arrowheads), and the absence of secondary cartilage. Compare with control (Fig 4) (hematoxylin-eosin stain, original magnification X94).

FIGURE 18. Histologic section of the joint seen in Fig 15. The bone of condyle (c) and articular eminence (AE)is denuded for the most part. A small area of fibrillated fibrous covering, with chondrocytic clusters (arrows), can be seen. The opposing denuded condylar bone contains bone microcysts (arrowheads) that open into the joint cavity (*) (hematoxylin-eosin stain, original magnification X2 1).

FIGURE 2 1. Electron micrograph of decalcified section of denuded condyle showing the bone collagen directly exposed to the joint cavity (*) (original magnification X4.250).

FIGURE 19. Electron micrograph of portion of the condyle 8 weeks postperforation. Note the presence of degenerated chondrocytes of secondary cartilage in the osteoartheritic joint. Compare with control (Figs 11, 12) (original magnification X2,700).

FIGURE 20. Electron micrograph of condyle 8 weeks postperforation. Bone marrow spaces are enlarged and came in close contact with articulating bone, with no intervening secondary cartilage (original magnification X2,784).

FIGURE 22. Histologic section of the articular eminence from joint 8 weeks after discal perforation. The fibrous covering is thickened, abnormally cellular, vascular, and fibrillated. Note the large number of osteoclasts (arrows) next to areas of bone resorption. Asterisk (*) denotes joint space (periodic acid-Schiff stain, original magnification X 120).

FIGURE 23. Electron micrograph of a tem- FIGURE 24. Stereoscopic photomicrograph poral surface 8 weeks postperforation. The of a joint 8 weeks postperforation presented for normally avascular tissues contain a large ven- orientation of Fig 26. The scored area is shown ular capillary (vc), large numbers of cells (ar- histologically in Fig 26 (original magnification rows), and show fibrillation of collagen (large X3.0). arrows). Contrast to Fig 11 (original magnification X4,137).

FIGURE 25. Stereoscopic photomicrograph of a joint 8 weeks postperforation showing the previously described osteoarthritic changes. The edges of the perforation (arrows) are thin and transparent. The same area is presented histologically in Fig 27 (original magnification X5.0).

FIGURE 26. Histologic section of a portion of the disc away from the edge of the perforation (area scored in Fig 24). Note the abnormally high cellularity and vascularity. Also note the presence of synovial cells on the surface of disc (arrows) (periodic acid-Schiff stain, original magnification X 120).

FIGURE 27. Histologic section of the edge of a perforation showing the high celhrlarity, vascularity, and washed-out appearance of the connective tissue of the disc, Note the presence of synovioblasts on both surfaces of the disc (arrows) (periodic acid-S&IT stain, original magnification X 120).

266

REPAIROF DISCALPERFORATIONWITH SYNOVIUM

novial cells consistently covered both surfaces of the disc and the margins of the perforation (Figs 26, 27). These cells could be traced to the hyperplastic synovial membrane that filled the recesses and grossly appeared as globular masses (Figs 15, 28). At the electron microscopic level, the normal, interlacing, compact, dense collagen bundles were replaced by thin, loosely arranged, collagen fibers that were surrounded by large amounts of electron-dense material that morphologically resembled the proteoglycans of other connective tissues. Vacuoles surrounded by electron-dense granular material were observed (Fig 29). Type A or macrophage-like cells of the synovium were present within the injured disc (Fig 29). Away from the margin of the perforation, the remaining portion of the disc was highly vascular, highly cellular, and rich in collagen (Fig 26). At the electron microscopic level, the discal cells had the fine structural characteristics of fibroblasts, and they were surrounded by dense collagen fibers running in different directions (Fig 30). EXPERIMENTALJOINTS 4 WEEKSAFTER REPAIR OF DISCAL PERFORATION: STEREOSCOPICOBSERVATIONS In contrast to the control joints, the superior and inferior compartments appeared smaller in the anteroposterior dimension. However, all the other joint features appeared normal. The area of the suture could be easily discerned (Fig 31). In contrast to the joints with discal perforation, the articular surfaces of the condyle and temporal component appeared smooth, with no gross evidence of pathologic alterations (Fig 3 1). Histopathologicand Electron Microscopic Observations The middle portion of the disc appeared highly cellular and vascular. The vessels consisted mostly of large capillaries and venules. The superior and inferior surfaces of the disc were covered by synovial cells (Fig 32). At the electron microscopic level, although the density of collagen fibers was less than in the control, and the fibers appeared thinner, there were bundles of varying diameter collagen fibers and healthy fibroblasts similar to normal discal tissue (Fig 33). Medial to where the perforation was, areas of the disc had a fibrocartilagenous appearance with chondroblast-like cells (Fig 34). At the electron microscopic level, the cells had the fine structural characteristics of chondroblasts and were surrounded by fine, fibrillar, territorial matrix (Fig 35). At the site where sutures were placed, a line of demarcation between the old collagen and the young, highly cellular, regenerating portion of the disc could be observed (Figs 36, 37). Large amounts of electron-dense

material, probably connective tissue ground substance, were observed in between the collagen fibers (Fig 38). Areas of microfibrillar accumulation that had the fine structural characteristic of oxytalan or preelastin were observed (Figs 38, 39). An amorphous, electron-dense component of elastin, combined with microfibrillar components, appeared thin and curled next to cells that had the fine structural characteristics of myofibroblasts (Fig 39). The cells were typically surrounded by a basal lamina and contained the characteristic electron-dense myosin and thin microfibrillar cytoplasmic components (Fig 39). Using a 1X lens, an entire repaired joint could be seen under the light microscope (Fig 36). At this level, the fibrous covering of the surface of the glenoid fossa and posterior tubercle appeared abnormally thick; the same was true for the fibrous covering of the condyle, particularly under the repaired area or the area where the perforation was created (Fig 36). In the repaired joints, although at the dissecting microscope level the condylar and temporal bony components appeared normal, at the light microscopic level the fibrous covering of the condyle appeared abnormally cellular (Fig 39). Also, the prechondroblastic layer was not well defined and the area of secondary cartilage appeared abnormally vascular (Fig 40). At the depth of the glenoid fossa, underneath the fibrous covering, focal areas of bone resorption and granulation tissue, similar to what was described in the joints with disc perforation, were observed (Fig 41). However, next to the area of bone resorption chondroid bone and alcian blue-positive cartilage were observed (Figs 42, 43). Discussion The anatomy of the rhesus monkey TMJ is remarkably similar to the human joint, which makes it an excellent animal model for experimental joint studies. The avascular portion of the disc consists mainly of collagen fibers that exhibit a variety of cross-sectional diameters, particularly in the anterior and posterior bands. A similar observation was made by Silva’ in the pig’s disc and by Ghadially’ in human knee men&i. Such variations in the thickness of the collagen fibers are thought to make these structures withstand a broad variation of stress and make them function as shock absorbers. ‘* The cullagen bundles and the individual collagen fibers are surrounded by a ground substance, or proteoglycans, which together with the elastic fibers convey the viscoelastic property to the disc that is necessary for its function as shock absorber under compressive forces.13 Recently we demonstrated the presence of these glycoconjugates between collagen bundles and collagen fibers in rabbit discs using lectin histochemistry and lectin-colloidal gold techniques.14 The ma-

SHARAWY ET AL

FIGURE 28. Histologic section ofjoint recess of a joint with perforated disc. Note the marked hyperplasia of the synovium (periodic acidSchiff stain, original magnification X 187).

FIGURE 29. Electron micrograph of an area of perforated disc similar to that shown histologically in Fig 27. Note the remarkable loss of collagen, presence of vacuoles (v) surrounded by electron-dense amorphous material, large amount of ground substance (arrows), and the presence of macrophage-like, or type A, cells of the synovium (M) (original magnification X2,784).

FIGURE 30. Electron micrograph of a portion of disc far from the edge of perforation showing the high cellularity and thin collagen bundles between the cells. Most of the cells have the fine structural characteristics of hibroblasts (F)(original magnification X2,700).

FIGURE 3 1. Stereoscopic photomicrograph of a joint 4 weeks after repair of discal perforation. Except for alteration in shape of the condyle and apparent shrinkage of anteroposterior dimensions of joint cavities, the joint appears normal. The site of healed surgical repair is noticeable (arrows) (original magnification X12).

FIGURE 32. Histologic section of repaired disc. Unlike the control disc, this disc is highly cellular, vascular, and is covered with cells (arrows) that could be traced back to the synovial membrane tissue in the joint recesses (hematoxylin-eosin stain, original magnification X 198).

FIGURE 33. Electron micrograph of repaired disc. Note the restoration of collagen, which exhibits variation in the cross-sectional diameter ofthe fibers similar to control tissue. A tibroblast (F)can be seen (original magnification ~4,125).

FIGURE 34. Histologic section of repaired disc medial to the repaired area showing a large number of chondrocyte-like cells (arrows) surrounded by alcian blue-positive territorial matrix giving the tissue the appearance of fibrocartilage (original magnification X234).

FIGURE 35. Electron micrograph of a partion of a repaired disc medial to the repaired area showing chondrocyte-like cells embedded in pairs and surrounded by tibrillar territorial matrix (arrows) similar to cartilage cells (original magnification X2,784).

FIGURE 36. Histologic section of the same joint seen in Fig 3 1. The condyle seems to be undergoing considerable remodeling The area of surgical repair is noticeable. Note the thickened fibrous covering in the glenoid fossa (arrows) and postglenoid tubercle (arrowhead). Notice the absence of fibrillation, adhesion, debris in joint cavities, and bone resorption (hematoxylin-eosin stain, original ma$nification x IO).

268

REPAIR OF DISCAL PERFORATION

WITH SYNOVIUM

FIGURE 37. Histologic section of repaired disc next to the area that was perforated. Note the old discal tissue to the right interfaced with highly cellular and regenerating discal tissue to the left. Note the presence of synovial cells (arrows) overlying the regenerating discal tissue (original magnification X25 1).

FIGURE 38. Electron micrograph of a portion of repaired disc. Note the presence of ground substance between the collagen bundles and the presence of microfibrilar proelastin or oxytalin components (arrows) (original magnification X 14,438).

FIGURE 39. Electron micrograph of a portion of repaired disc. A myofibroblast (MF) with its characteristic microfibrihu content and basal lamina (arrows) surrounded by elastin fibers (arrowheads) is seen. Each elastic fiber showed both microfibrillar and electron-dense components (original magnification X8,250).

FIGURE 40. Histologic section of repaired joint showing portion of the condyle. Note that the fibrous covering is highly cellular and similar to the control. An area of secondary cartilage is well developed but is abnormally vascular (VS and arrows) (original magnification X 100).

FIGURE 4 1. Histologic section of articular eminence of a joint with a repaired disc. Although the fibrous covering is compact, smooth, and moderately cellular, an area of bone resorption (arrows) and vascularity can be seen (original magnification X 100).

FIGURE 42. Histologic section of postglenoid tubercle area of a joint with a repaired disc. Note bone lines indicative of bone remodeling (arrows) and the presence of chondroid bone (ChB) (original magnification X 100).

FIGURE 43. Histologic section of postglenoid tubercle area similar to the one shown in Fig 42 but stained with alcian blue to show alcian blue-positive matrix of cartilage cells in chondroid bone (arrows) (original magnification X100).

jority of the cells in the normal disc are fibroblasts, which are known for their ability to secrete both collagen and proteoglycans of the ground substance.‘5 At the electron microscopic level, the elongated processes of the disc fibroblasts were found to extend for long distances between the collagen fibers and, therefore,

seem to be responsible for the turnover of collagen and ground substance farther away from the cell bodies. Similar to the covering of the condyle, the covering of the articular eminence consists mostly of collagen that runs in different directions and is uniform in hbrillar diameter. The collagen bundles are compact and offer a smooth surface facing the joint space. In this regard, the TMJ differs from other synovial joints in which the bone is covered with hyaline cartilage. The presence of a reserve cell layer under the fibrous covering of the monkey condyle is also found in TMJs of other animal species, including humans.‘6T’7These cells are known to be highly proliferative,‘* and thought to be able to differentiate into chondroblasts and fibroblasts as an adaptive response to overloading.‘9*20 In our study, perforation of the disc precipitated an increase in thickening of the condylar fibrous covering, perhaps due to an increase in differentiation of the reserve cells into fibroblasts. When the mechanical overload exceeds the physiologic tolerance of these tissues, fibrillation and denudation of focal areas occur, as we showed in our previous study on disc perforation and

269

SHARAWY ET AL

also in this present study.’ In the control intact joints, synovial membrane was demonstrated in all the recesses of TMJ spaces. It consisted of intimal cells of type A, or macrophage-like cells, and type B, or fibroblast-type cells. In the joints with unrepaired perforated discs, the synovial membrane became hyperplastic and synovial cells migrated on the surface of the perforated discs and lined the circumference of the perforations. The synovial cells were supported by subintimal capillaries at areas of the disc that are normally nonvascularized and hypocellular. These results were remarkably similar to our observations on degenerative joints following experimental disc perforation in monkeys, and also in human joints with perforated discs. 8.9It seems reasonable to assume that the synovial membrane contributes to the high cellularity and vascularity of the perforated disc. Angiogenic factors have been found in synovial fluid obtained from patients with osteoarthritic joints.21,22 At the electron microscopic level we observed that the discal cells resembled fibroblasts or type B cells of the synovium. A similar response has been found in injured knee menisci.23,24Since the osteoartheritic changes we observed at the light microscopic level in the present study 8 weeks following disc perforation were similar to those we previously reported,’ we will highlight the discussion of the electron microscopic changes in joints with perforated discs. A large part of the discal tissue surrounding the perforation showed vacuolation, loss of collagen, and an increase in electron-dense material resembling disaggregated proteoglycans. Similar results were reported by Ghadially in the torn knee menisciz5 Mechanical injury may lead to an increase in lysosomal enzyme release from the macrophage-like cells of the synovium and cytokines from injured joint cells, which could cause damage to the extracellular matrix of the disc and the covering of the joint surfaces.26-28In our study, synovial macrophage-like cells were found to invade the perforated disc. However, the role of the initial inflammatory response resulting from the surgical intervention and the creation of disc perforation in causing degradation of collagen and extracellular matrix cannot be excluded, since we only observed the joints 8 weeks following discal perforation. At this time, little or no inflammatory change was observed. A similar observation was reported in experimental knee osteoarthrosis in rabbits as inflammation declined 4 weeks following surgery.29 The diffuse electron-dense material that was present in the perforated disc is suggestive of disaggregation of proteoglycan, and along with the noted loss of collagen, point out the loss of the shock absorber function of the perforated disc and the subsequent mechanical injury to the fibrous coverings of the articular bony compo-

nents. The latter trauma may have contributed to fibrillation and the surface destruction of the condylar and articular eminence coverings. Wounds in the avascular portion of monkey discs away from the bilaminar zone were shown not to heal while wounds close to the bilaminar zone and synovium were found to heal.” Since the approximated edges of the perforation were in the avascular portion of the disc, one would not expect it to heal on its own without the synovium. When the perforation was excised and the disc was repaired with the aid of synovium, the degenerative changes attributed to biomechanical trauma were no longer observed at the gross level. At the electron microscopic level, new collagen, preelastin (oxytalan), myofibroblasts, and numerous fibroblasts were found in the repaired disc. Myofibroblasts were previously observed in human retrodiscal tissues from patients with internal derangement,’ in torn human menisci,‘2.25and in injured knee menisci repaired by synovium.30 The origin of discal myofibroblasts is not certain, but they are thought to be derived from synoviocytes.30 These cells may undergo contraction at the time of collagen formation in regenerating mesenchymal tissue.3’ It is also possible that the synovial flap has contributed libroblasts and blood vessels to the repaired disc. However, experimental evidence is needed to confirm this speculation. Without synovium sutured into the knee meniscal wound, the tom men&i of rabbits, dogs, pigs, and sheep did not heal.30 Ghadially and colleagues considered synovial implantation as an alternative to menisectomy in the treatment of tom menisci3’ The results of this study in the TMJ support Ghadially and coworkers’ concept for the treatment of torn menisci. We found that the repaired disc protected the underlying condyle and the overlying temporal components and permitted the regeneration of cartilage, bone, and fibrous covering. It is remarkable to note that the vascularization of the young, regenerating, condylar cartilage 4 weeks following discal repair is reminiscent of fetal cartilage.3’ Also, the appearance of cartilage and chondroid bone in the posterior aspect of the glenoid fossa in the repaired joints suggests a remodeling effort of the joint in restoring its normal function. Finally, our study provides evidence for the regenerative capabilities of the TMJ components, particularly the synovium, in repairing the damages induced by osteoarthrosis. The reversibility of the degenerative alterations following discal repair in this experimental model should provide a rational and a useful method for surgical repair of TMJ disc perforations. References 1. Silva DG: Further ultrastructural studies on the temporomandibular joint of the guinea pig. J Ultrastruct Res 26: 148, 1969

270

DISCUSSION

2. Ghadially Flv: Intraarticular discs and minsci (normal and tom), in Ghadially FN (ed): Fine Structure of Synovial Joints. London, England, Butterworths, 1983, pp 103-144 3. Yaillen DM, Shapiro PA, Ludcher RS: Temporomandibular joint menisectomykffects on joint structure and masticatory function in Macca fasicularis. Maxillofac Sum 7:255. 1979 4. Yaillen DM, Shapiro PA, Ludcher ES Temp&omandibuiar joint menisectomy effects on joint structure and masticatory function in Macca fasicularis. J Maxillofac Surg 7:264, 1979 5. Hinton RJ: Alterations in rat condylar cartilage following discectomy. J Dent Res 71:1292, 1992 6. Appel H: Late results after meniscetomy in the knee joint. Acta Orthop Stand 133: 1, 1970 (Suppl) 7. McDevitt C, Gilbertson E, Muir H: An experimental model of osteoarthritis, early morphological and biochemical changes. J Bone Joint Surg [Br] 59:24, 1977 8. Helmy E, Bays R, Sharawy M: Osteoarthrosis of the TMJ following experimental disc perforation in Macaca fascicularis. J Oral Maxillofac Surg 46:979, 1988 9. Helmy E, Bays R, Sharawy M: Histopathological study of human TMJ perforated discs with emphasis on synovial~membrane resnonse. J Oral Maxillofac Sure. 47: 1048. 1989 10. Helm; E, Ingalls G, Sharawy M: H&ling of the TMJ disc wound created at the anterior band. Case Reports and Outlines of Scientific Sessions of the American Association of Oral and Maxillofacial Surgeons, 69th Annual Meeting, Anaheim, CA, September 16-20, 1987, pp 68-69 11. Franc S, Garrone R, Bosch A: A routine method for contrasting elastin at the ultrastructural level. J Histochem Cytochem 32: 251, 1984 12. Ghadially FN, LaLoude JMA, Wedge JH: Ultrastructure of normal and tom men&i of the human knee joint. J Anat 136: 773, 1983 13. Blaustein DI, Scapino RP: Remodeling of the temporomandibular joint disk and posterior attachment in disk displacement specimens in reiation to glycosaminoglycan content. Plast Reconstr Surg 78:756, 1986 14. Sharawy MM, Linatoc AJ, O’Dell NL, et al: Morphological study of lectin binding in the rabbit temporomandibular joint disc. Histochem J 23:132. 1991 15. Leblond CP: Synthesis and secretion of collagen by cells of connective tissue. bone. and dentin. Anat Ret 224:123. 1989 16. Ben-Ami Y, Lewinson D, Silbermann M: Structural characterJ Oral Maxillofac 52:270-271,

ization of the mandibular condyle in human fetuses: Light and electron microscopy studies. Acta Anat 145:79, 1992 17. Bibb CA, Pullinger AG, Baldioceda F: The relationship of undifferentiated mesenchymal cells to TMJ articular tissue thickness. J Dent Res 7 1:I8 16, 1992 18. Hall BK: Selective proliferation and accumulation of chondroprogenitor cells as the mode of action of biomechanical factors during secondary chondrogenesis. Teratology 20:8 1, 1979 19. McNamara JA, Hinton RI, Hoffman DL: Histologic analysis of temporomandibular joint adaptation to protrusive function in young adult rhesus monkeys (Macacca mulata). Am J Orthod 82:288. 1982 20. McNamara JA; Carlson DS: Quantitative analysis of temporcmandibular joint adaptations to protrusive function. Am J Orthod 761593, 1979 21. Brown RA, Tomlinson IW, Hill CR, et al: Relationship of angiogenesis factor in synovial fluid to various joint diseases. Ann Rheum Dis 42:301, 1983 22. Brown RA, Weiss JB: Neovascularization and its role in the osteoarthritic process. Ann Rheum Dis 47:8X1. 1988 23. Bruce J, Walmsley R: Replacement of semilunar cartilage of the knee after operating excisions. Br J Surg 25: 17, 1937 24. Cox JS, Corden LD: The degenerative effects of medial meniscus tears in dogs’ knees. Clin Orthop 125:236? 1977 25. Ghadially FN_ Intraarticular discs and minisci’(normal and tom), in Ghadially FN (ed): Fine Structure of Synovial Joints. London, England, Butterworths, 1983, pp 103-144 26. Ali SY, Evans J: Enzymatic degradation of cartilage in osteoarthritis. Fed Proc 32: 149 1. 1973 27. Pugh JW, Radin EL, Rose RM: Qualitative studies of human subchondral cancellous bone: Its relationship to the state of its overlying cartilage. J Bone Joint Surg [Am] 56:313, 1974 28. Stockwell RA: Cartilage failure in osteoarthritis: Relevance of normal structure and function. A review. Clin Anat 4: 161, 1991 29. Lukoschek M, Schaffler MB, Burr DB, et al: Synovial membrane and cartilage changes in experimental osteoarthrosis. J Orthop Res 6:475, 1988 30. Ghadially FN, Wedge JH, LaLoude JMA: Experimental methods of repairing injured menisci. J Bone Joint Surg [Br] 68: 106, 1986 3 1. Thilander B, Carlsson GE, Ingervall B: Postnatal development of human temporomandibular joint. Acta Odontol Stand 34: 117, 1976

Surg

1994

,

Discussion Repair of Temporomandibular Joint Disc Perforation Using a Synovial Membrane Flap in Yacaca Fascicularis Monkeys: Light and Electron Microscopy Studies Myron R. Tucker, DDS Charlotte, NC

This study represents one of the most elaborate documentations of temporomandibular joint anatomy of the normal, damaged, and repaired temporomandibular joint. The stereomicroscopy, histology, and electron microscopy are meticulously done and provide excellent supporting documentation. The authors should be complimented for this effort. There are, however, several problems with the methodological approach used to evaluate repair of disc perforations. The authors’ hypothesis was that a synovial flap would be the most appropriate technique to repair chronic disc perforation. However, the experimental design does not adequately test this hypothesis. In this study, bilateral disc per-

forations were created in monkeys. At 4 weeks postoperatively one joint was opened and a primary repair of the perforation attempted in addition to the use of a synovial flap to repair the perforation. The left side underwent no second surgical procedure. To adequately test the advantages of the synovium, an operation should have been performed on the control joint which simulated the experimental side with the exception of use of synovium. Ideally, this would have included an attempt to repair the perforation primarily (as was done in the experimental joint). At the very least the control joint should have been opened, examined, and closed in a “sham” procedure. Given the experimental design as presented one could argue that the improvement in the outcome of the experimental joint was due simply to a second operation or closure of the disc perforation and not necessarily the result of the use of synovium. In this type of study the treatment of the control joint should exactly parallel the treatment of the experimental joint with the exception of the one variable being evaluated, in this ease the use of synovium. A second criticism involves the use of only four monkeys killed at the same time. Due to the significant cost involved