Temporomandibular joint (TMJ) ankylosis is characterized by difficulty or inability to open the mouth due to fusion of the temporal and the mandible, resulting in facial symmetry/deformity, malocclusion and dental problems. The only treatment option for TMJ ankylosis is surgical with or without condylar reconstruction. Various autogenous grafts are available for condylar reconstruction after freeing the ankylotic mass such as costochondral, sternoclavicular, fibular, coronoid, and metatarsophalangeal. Costochondral graft is preferred by surgeons, but distraction osteogenesis is slowly gaining popularity and may ultimately become the standard procedure, providing a cost-effective approach with low morbidity and excellent functional outcomes. Tissue engineering is another budding field which has shown promising results in animal studies but has not been applied to humans. To date, there is no ideal autogenous graft for condylar reconstruction that satisfies the complex anatomy and the myriad of functions of a missing condyle.
Temporomandibular joint (TMJ) ankylosis is a bony or fibrous adhesion of the anatomical joint components by an ankylotic mass with ensuing loss of function. It is primarily caused by trauma, infection, or failed surgery. S alins described the ankylotic mass as being abnormal bone that replaces the TMJ and results in restriction of mandibular movement . TMJ ankylosis is classified by location (intra-articular or extra-articular), type of tissue involved (bone, fibrous, or fibro-osseous), and the extent of fusion (complete or incomplete) .
The treatment of TMJ ankylosis is surgical, either gap arthroplasty, interpositional arthroplasty, and/or joint reconstruction using autogenous grafts or alloplastic material . Treatment of TMJ ankylosis to restore full mouth opening and normal oral function remains a challenge in children and adults. The type of operation and treatment policy vary from one country to another. Surgical treatment depends on: the extent and type of ankylosis; the age of the patient; onset and time of surgery; and whether the ankylosis is unilateral or bilateral. No single method has produced uniformly successful results. The complications most often reported are limited range of movement and recurrence of ankylosis (usually within 6 months after surgery) .
A successfully reconstructed TMJ should reproduce normal joint structure, provide functional articulation, and permit adaptive growth or remodelling . Difficulty in achieving these treatment goals is illustrated by the multiplicity of autogenous and alloplastic materials proposed or currently used to reconstruct the TMJ. Alloplastic materials have been developed claiming closer reproduction of normal TMJ anatomy, but there are potential disadvantages such as: wear and/or failure of the material with wear; particles generating a giant cell foreign body reaction with potential loosening of the implant resulting in occlusal change; displacement or fracture; high cost; dystrophic bone formation; and lack of growth, which precludes the use of such joints in children . Autogenous grafts for condylar reconstruction have drawbacks but are the most widely used grafts for condylar reconstruction. This review discusses the various modes of autogenous condylar reconstruction used in the past, common grafts currently being used and indications of what lies ahead in the field of autogenous reconstruction of the condyle.
Autogenous grafts derived from distant sites
Most widely accepted autogenous condylar reconstruction involves the use of the costochondral graft (CCG) first described by Sir Harold Gillies in the 1920s . The current technique for TMJ replacement with the CCG was popularized by P oswillo . The sixth or seventh rib is resected along with 1 cm of intact costal cartilage for CCG. Recently authors have recommended using thinner sections of cartilage (2–3 mm) due to the potential for overgrowth . CCG is biologically compatible like any autogenous graft, easily workable especially when contouring the cartilaginous part to fit into the glenoid fossa, and takes less time to heal and becomes incorporated into the new environment allowing restoration of the bone and cartilage components . CCG also has the capacity for remoulding into an adaptive mandibular condyle, and there is always a potential at the donor site to grow and regenerate . An additional advantage in children is its growth potential . One of the main disputes is the lack of predictable growth . In assessing the growth of the graft, G uyuron & L asa observed growth of the graft in four patients, hypo-growth in one patient and normal growth in three patients. In addition, three cases of postoperative ankylosis were observed, showing the variability in results following CCG . Reankylosis in the range 5–39% has been reported .
Although R oss pointed out the difficulty of accurately monitoring the long-term growth of CCG, identifying that some growth has occurred is not difficult; monitoring the precise amount of growth is virtually impossible . M ulliken et al. showed that the highest growth rate occurs about 2 years after the placement of the graft, based on radiograph measurements, and follows a slow and irregular pattern, not a linear pattern, although there are individual variations . Other disadvantages of the CCG are the poor quality of medullar and cortical bone, the possibility of resorption or infection, bone flexibility, elasticity that may cause the graft to be deformed and to produce occlusal changes with time, and the possible separation of the cartilage from the bone. Donor-site complications such as pleural tear, pneumothorax, pleural effusion and atelectasis; empyema; pneumonia and occasional fractures have also been reported . More severe complications were noted especially when a vascularized graft was harvested, as pedicled CCG is very difficult to make harvest .
To date, CCG is the most widely used graft for condylar reconstruction. To maximize performance and minimize complications, authors have recommended meticulous dissection of the periosteum and perichondrium off the transplanted rib as well as harvesting alternate ribs if multiple grafts are required. This leads to reduced postoperative pain at the donor site and ensures that the parietal layer of the pleura is not perforated. Similarly, retaining an intact piece of periosteum and perichondrium adherent to and bridging the costochondral junction, and sectioning the chondral part before the osseous part, has reduced the rate of fracture at the costochondral junction. Inset of CCG with a relatively shorter (2–3 mm) chondral portion helps to avoid unwanted overgrowth. Patients who have undergone previous release of ankylosis and who have developed recurrent ankylosis are not good candidates for CCG grafting . In spite of the potential problems with the CCG, most authors consider that it should be the preferred graft in the growing child and as initial reconstruction in many adult deformities .
S nyder et al. reported the first sternoclavicular (SCJ) whole joint graft consisting of a portion of the manubrium, the intact capsule, and a portion of the clavicle on a 70-year-old man with cylindroma in 1971 . In 1986, R eid et al. reported a free-flap technique that included the clavicular head of the pectoralis major muscle and overlying skin, to provide a vascularized clavicular bone graft. They suggested splitting the clavicle longitudinally and repositioning it with the attached flap as the entire head of the clavicle was too large to fit into the glenoid fossa . Later in 1994, W olford et al. reported splitting the clavicle head and applying only the superior half of the clavicle for condylar reconstruction .
The SCJ and TMJ are similar anatomically and physiologically . The head of the clavicle contains layers of cartilage that are similar to the mandibular condyle . The SCJ articulation has a growth centre and an interarticular fibrocartilage articular disc that simulates the meniscus of the TMJ . When a whole joint is used, the two adjacent synovial compartments and the strong fibrous capsule resemble those in the TMJ. Its absence is also of no great anatomical, functional, or aesthetic consequence . Complete regeneration of the clavicle at the donor site has also been reported, significantly reducing patient morbidity postoperatively . Authors observed an improved fit, especially when only the superior half of the clavicle was used, particularly if the graft was placed on the posterior border of the mandible, and there was easier vascularisation because of the direct exposure of the medullary bone of the graft to adjacent soft tissues . An animal study by E llis & C arlson showed the close histomorphological similarities between the SCJ and the TMJ in the monkey at all stages of postnatal development . D aniels et al. showed that when implanted in the TMJ area, SCJ graft undergoes remodelling and resembles the native condyle unlike CCG which does not.
SCJ grafting may seem to be a benign procedure because of its superficial location, but the surgery is comparatively complex and can damage the great vessels at the time of harvest. Postoperatively, instability of the clavicle under stress with resulting shoulder instability or even fractures may damage the underlying structures . Clavicle fracture with fracture impinging on the brachial plexus has been reported . The scar is in a prominent part of the body which might displease aesthetically concerned patients. Some authors have suggested that in female patients, the supraclavicular incision be kept 1–2 cm inferiorly for cosmetic reason . W olford et al. suggested the use of a figure-of-eight bandage for up to 3 months to support the shoulders, as for clavicular fractures, as well as the use of arm slings to immobilize the arm on the side of the removed graft; they note that this reduces postsurgical fracture of the clavicle . Patients must be also be cautioned not to lift anything heavy, use the arm for lifting themselves out of bed, or sustain large amounts of weight, such as when getting out of a chair, until 3–4 months postoperatively. SCJ can be a viable alternative for CCG, but long-term studies comparing both grafts are lacking.
The metatarsophalangeal (MTP) joint as a graft for condylar replacement was reported in 1909 by Bardenheur, as a half-joint transplant of the fourth metatarsal using the metatarsal head for replacement of the mandibular condyle . In 1971, metatarsal as a nonvascularized free bone graft for bilateral reconstruction of the TMJ was reported by D ingman , who failed to observe any longitudinal growth in the graft . In 1985, T ing et al. described the use of a free vascularized second metatarsal for reconstruction of the TMJ in four cases of ankylosis .
MTP complex can provide a good supply of articular cartilage combined with up to 7 cm of vascularized bone. In addition to replacement of condyles, it can also be used for simultaneous reconstruction after wide tumour resections . The risk of degeneration and reankylosis of the graft is low, especially when used as a vascularized graft. The MTP joint is smaller than the TMJ, so it easily fits within the confines of glenoid fossa . The intact epiphysis in the transplanted MTP joint, which contains an epiphyseal growth plate, ensures the normal growth of the transplant in young patients . There is an acceptable cosmetic result at the donor site, which can be well hidden.
A slight foot splaying has been reported, but the main donor-site morbidity is the aesthetic loss of a toe . The MTP joint is a simple hinge joint that does not follow the same movements as the TMJ, and therefore lateral excursion is restricted. In a study using whole MTP joint graft, V ilkki et al. were able to observe lateral jaw movement as well . Contemplating the transfer of vascularized MTP to the temporomandibular area is usually complex and a lengthy procedure, and is not the first choice for condylar reconstruction in cases of ankylosis . It is a useful option where vascularized tissue is required for autogenous reconstruction of the TMJ along with simultaneous mandible reconstruction. Using the metatarsal head for articulation in the glenoid fossa with the tarsal plate interposed between the articular surface and the head of the graft can simulate the condylar head and the interpositional cartilage disc, so MTP grafts can be a viable option in selected cases of ankylosis .
As a new procedure for bilateral reconstruction of mandibular condyles, S mith & R obinson used iliac bone grafts in a recurrent case of bilateral TMJ dislocation, in a patient who had previously undergone condylectomy for the same problem . An iliac graft is a chondro-osseous graft, and the graft consists of a full-thickness piece of the iliac crest, including the overlying cartilage layer as a cap. M atukas et al. used a piece of iliac bone with a cartilage cap secured to the glenoid fossa as a stump to prevent recurrence of TMJ ankylosis in a child . K ummoona suggested that the cartilaginous rim of the iliac crest be grafted to the ramus of the mandible with an osseous element cut from the ilium in continuity with the cap of cartilage in the same technical manner used to position the CCG for reconstruction of ankylosed TMJ in children .
One major advantage especially in growing patients is that the vertical growth pattern of the ilium is converted in the graft to a multidirectional pattern, to be adaptive to the functional demands of the TMJ, restoring the normal growth of the condyle . K ummoona did not observe any creeping substitution of the cartilage columns by osteoid tissue nor any cases of reankylosis, although the cases were followed for only 18 months . The location of the scar can be easily concealed even with minimal clothing unlike other grafts, such as fibula and sternoclavicular . Harvest of the iliac crest can cause prolonged postoperative pain, altered gait, sensory nerve damage, poor scar/bone contour, delayed healing, herniation of abdominal contents, ilium fracture, peritonitis, and retroperitoneal hematoma . Large amounts of bone can be harvested from the iliac bone with adequate height and length and it is more suitable for mandibular reconstruction. Iliac grafts can be used to reconstruct condyle, so iliac grafts can be a viable option in selected cases for condylar reconstruction .
P osnick et al. first reported a series of cases using the free fibula flap for immediate reconstruction of paediatric mandibular tumours in 1993 . In 2000, W ax et al. reported placement of the distal portion of the flap directly into the glenoid fossa for condylar reconstruction . The fibula is tubular in shape and is densely cortical. It can be easily adapted to passively fit in the glenoid fossa, and its narrow shape allows an easy fit . G uyot et al. noticed rounding off of the fibula in the glenoid fossa in patients who were followed up for 6 years . Alternatively, the fibular head can be harvested and used as the neocondyle to avoid bleeding from the medullary space of the fibula shaft and subsequent scar formation and temporofibular ankylosis . Fibular head amputation is detrimental to knee joint stability; but it can be avoided if the collateral fibular ligament is fixed carefully to the tibial condyle .
The main disadvantage of this graft is that unlike sternoclavicular, costochondral or metatarsophalangeal grafts, it lacks articular cartilage. Donor-site morbidity resulting from fibula flap harvest appears to be more pronounced in the paediatric population than in adults leading to complications, such as great toe flexion contractures and valgus deformity with a tibiotalar tilt, requiring, in some cases, Ilizarov lengthening and bone grafting . Other notable complications are: ankle stiffness, instability and weakness; numbness of the lateral side of the leg; pedal ischaemia and foot oedema; and partial dehiscence of the fibular donor site . G uo et al. recommend that if an adequate length of residual fibula is left behind with careful dissection that does not disrupt the ankle syndesmosis, paediatric patients can tolerate this donor site quite well .
There are mixed results regarding growth of vascularized fibula flap. Some say it grows as the patient grows with minimal disturbance to the growth pattern of the midface whereas others disagree . The survival rate of the graft is usually high, especially when used as a vascularized graft . It currently represents the best treatment option in paediatric patients with large mandibular defects . The fibular osseocutaneous free-flap is nowadays considered the workhorse donor site for mandibular reconstruction, as large segments of bone can be harvested from the fibula with robust axial blood supply, permitting multiple osteotomies . TMJ reconstruction can effectively and successfully be performed by direct transposition of fibula into the fossa. P otter & D ierks recommend fibula as the vascularized graft of choice for defects of the mandibular condyle when the wound beds are compromised or there are other conditions that preclude the use of nonvascularized tissues . Fibular reconstruction in cases pertaining to TMJ ankylosis has not been reported so far, so this does not seem to be the first option for condylar reconstruction in TMJ ankylosis.