 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1996 The Journal of Bone and Joint Surgery, Inc.
Current Concept Review - Osteochondritis Dissecans* , SAN ANTONIO, TEXAS*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
Osteochondritis dissecans is a localized injury or condition affecting an articular surface that involves separation of a segment of cartilage and subchondral bone. It is found primarily in the ankle, knee, and elbow joints. Although relatively infrequent, osteochondritis dissecans has been reported extensively in the orthopaedic literature. Various theories about its etiology have been proposed, but none has been accepted universally or has adequately explained its occurrence; however, the lesion is probably caused by factors that vary depending on the specific joint involved and on the specific site within that joint. The treatment of osteochondritis dissecans has evolved from the removal of loose bodies, first reported more than 150 years ago108, to the resurfacing of defects with an osteochondral autogenous graft158. The results of current treatment are still not uniformly successful. Osteochondritis dissecans has been confused with other abnormalities of the joint surface, and therefore the orthopaedic literature must be read with care. Other, severe conditions described in the literature have been grouped under the heading of osteochondritis dissecans; these have included osteochondral fractures104,110,116,123,133, osteonecrosis, accessory centers of ossification, osteochondrosis, and hereditary epiphyseal dysplasia22,118,119,147. In several clinical investigations, different terms, such as osteochondrosis dissecans107 and osteochondral fragments51,70,81,83,122, have been used to describe this entity. The recognition that many disorders may result in an osteochondritis dissecans-like lesion is fundamental to an understanding of this condition. The orthopaedic literature includes descriptions of lesions of osteochondritis dissecans in the talus5,12,28, the capitellum19,96,156, the patella43,127,138, the femoral head107, the distal part of the tibia10, and the wrist52,148. The male: female ratio has been reported as two to one107 and as three to one31,101, with eight (33 per cent) of twenty-four patients having bilateral involvement in one series59. The current review will concentrate on osteochondritis dissecans at its most common sites: the knee, ankle, and elbow. Osteochondritis dissecans affects two distinct populations of patients as differentiated by the status of the physes. Children and adolescents between the ages of five and approximately fifteen years old who have open physes have the juvenile form of the disease22,59. Older adolescents who have closed physes and adults are classified as having the adult form. Osteochondritis dissecans is rare in patients who are less than ten or more than fifty years old79. It is most frequently seen in young male patients, especially athletes1. The knee is the most commonly affected joint107, being involved in about 75 per cent of patients31.
Paré108 is credited with being the first to describe the removal of loose bodies, presumably osteochondral fragments, from joints, in a report published in 1840. In 1870, Paget106 described a pathological process that he termed quiet necrosis, which resulted in the creation of loose, necrotic osteochondral fragments from articular surfaces. König74, in 1888, described a pathological condition that led to the formation of loose bodies (referred to as arthrophytes or corpora mobile in that era) of femoral origin that were found in hip and knee joints with no evidence of pre-existing trauma. He believed that an inflammatory reaction of both the bone and the cartilage, followed by spontaneous necrosis, was a major component of this condition, and he chose the term accordingly: osteochondritis, referring to an inflammation of the osteochondral joint surface, and dissecans, derived from the Latin dissec, which means to separate. Later investigators101 noted the absence of inflammatory cells in histological sections of excised osteochondral loose bodies and believed that trauma was a more important etiological factor. Nevertheless, use of the original name has persisted. Osteochondritis dissecans, as described in the earlier literature, usually was treated either non-operatively or with an arthrotomy for removal of the loose fragment13,149. Closed treatment is still preferred for patients who are first seen early in the disease process and for those who have open physes. As diagnostic capabilities and knowledge of osteochondral healing have progressed, additional options have become available for operative treatment of the lesions in later stages. Magnetic resonance imaging has become useful in the staging of lesions in the knee, elbow, and ankle. The advent of diagnostic and operative arthroscopy has resulted in more detailed pathological descriptions and, theoretically, in a lower rate of morbidity. Drilling of the site of the defect, fixation of the lesion to the site of separation, and placement of osteochondral grafts have been described4,18,57,70,82,83,87,89,129,130,141,158. Current methods of operative treatment involve fixation of the lesion or drilling of the site of the defect with removal of loose bodies.
Clinical Presentation The age of the patient at the time of presentation is the primary determinant of treatment. In children and adolescents with open physes, osteochondritis dissecans of the knee is classified as the juvenile form. The average age of the patients at the time of the diagnosis has ranged from 11.3 to 13.4 years23,24,98. Mubarak and Carroll98 reported on seventy-three patients who had juvenile osteochondritis dissecans, fifty-one (70 per cent) of whom were male. The average age at the time of the diagnosis was 13.4 years. After the physes had closed, the patients were classified as having the adult form of the disease. Although adult osteochondritis dissecans may arise de novo, it is usually the result of an incompletely healed and previously asymptomatic lesion from juvenile osteochondritis dissecans. The age at the time of presentation of adult osteochondritis dissecans has usually ranged from seventeen to thirty-six years23, but it can be seen at any adult age. The symptoms are similar to those of juvenile osteochondritis dissecans and include swelling, catching, and locking of the knee.
Physical Examination A patient who has a lesion of the medial femoral condyle may walk with the leg externally rotated in an attempt to avoid impingement of the tibial spine on the lateral aspect of the condyle. Wilson152 described a physical sign for osteochondritis dissecans of the knee. The sign is elicited by flexing the knee to 90 degrees, internally rotating the tibia, and then slowly extending the knee. A positive Wilson sign is pain at approximately 30 degrees of flexion that is relieved by external rotation of the tibia.
Imaging Studies On radiographs, the lesion typically appears as a well circumscribed area of sclerotic subchondral bone separated from the remainder of the epiphysis by a radiolucent line (Figs. 1-A, 1-B, 1-C, 1-D through 1-E). If only anteroposterior and lateral radiographs are used, the lateral radiograph is the most helpful, as noted by Harding65. Harding described two lines drawn on the lateral radiograph: one along the posterior femoral cortex and the other along the Blumensaat line (intercondylar notch). The arc of articular surface between these two lines is the common location of lesions of osteochondritis dissecans on the medial femoral condyle. Cahill and Berg23,24 described a useful classification system in which the anteroposterior radiograph is divided into five segments: the notch is designated as segment 3; the medial tibiofemoral compartment is divided into segments 1 and 2; and the lateral tibiofemoral compartment is divided into two equal segments, 4 and 5. The lateral radiograph is divided into three segments on the basis of Harding's description: A, anterior to the Blumensaat line; B, between the Blumensaat line and a median drawn from the posterior femoral cortex; and C, the condylar area posterior to a line drawn along the posterior femoral cortex. A posteromedial femoral condylar lesion involving the entire weight-bearing surface would be classified as C1–2 (Figs. 2-A and 2-B).
Cahill and Berg24 described the appearance of osteochondritis dissecans with use of technetium bone scintigraphy, performed every six weeks. They noted the sensitivity of such bone scans as a measure of osteoblastic activity and regional blood flow. They also noted that the degree of osseous uptake seemed to be related to the remaining potential for healing of the fragment. On the basis of the degree of uptake, they described four stages of lesions: stage I—the lesion is visible on plain radiographs, but bone scans reveal normal findings; stage II—the scan reveals increased uptake in the area of the lesion, which can also be seen on plain radiographs; stage III—in addition to the stage-II findings, there is increased isotopic uptake in the entire femoral condyle; and stage IV—all of the characteristics of stage III are present and there is also uptake in the tibial plateau opposite the lesion. These investigators used technetium imaging to plan a course of treatment. They believed that stage I indicates either a healed fragment or an unhealed lesion without osteoblastic activity. They assumed that a lesion that was stage II or higher had potential for repair as a result of the apparent osteoblastic activity, and they considered bone-grafting unnecessary for such lesions. They also thought that a symptomatic patient with a stage-I or II lesion should be allowed to return to sports activities23,24. Magnetic resonance imaging has recently been used in the evaluation of patients with osteochondritis dissecans41,75,93,102,112. Mesgarzadeh et al.93 reported that magnetic resonance imaging permits visualization of loose bodies and of the degree of displacement of the major fragment. These authors recommended use of this technique to follow the healing response and the degree of revascularization of the lesion. Dipaola et al.41 studied the relationship between the appearance on arthroscopy and that on magnetic resonance imaging of osteochondritis dissecans lesions. The findings on the magnetic resonance images correlated closely with the arthroscopic classification, but they correlated poorly with plain radiographic findings. In other words, lesions that appeared similar on plain radiographs were seen to be widely disparate on magnetic resonance imaging, suggesting that magnetic resonance imaging is more sensitive in the evaluation of these lesions. Magnetic resonance imaging also allows determination of displacement of the fragment, the fluid interfaces, and the integrity of the articular surface. Magnetic resonance imaging and computed tomography are used in preoperative planning to determine if the fragment is detached. Kramer et al.75 used magnetic resonance arthrography with contrast medium to determine the stage of the lesion and whether it was detached, and the imaging findings were verified in twenty-four of twenty-five knees at the time of arthroscopy. The integrity of the surface of an osteochondritis dissecans lesion can be detected with magnetic resonance imaging, but detection is improved in the presence of an effusion because of the orthographic effect of free joint fluid on T2-weighted images. These authors noted improvement in the detection of all types of lesions after intra-articular injection of gadolinium (Magnevest; Schering, Berlin, Germany). Other authors have noted the sensitivity of a fluid interface secondary to an effusion in determining the presence of a detached osteochondritis dissecans lesion40,41,66,100,153,160. Nelson et al.102, in 1990, used magnetic resonance imaging to predict the grade of the lesion in eleven of twelve patients at the time of arthroscopy. In one patient, a grade-2 lesion was thought by both radiologists to be grade 3, according to the classification system described by Clanton and DeLee31 (grade 1, a depressed osteochondral fracture; grade 2, an osteochondral fragment attached by an osseous bridge; grade 3, a detached, non-displaced fragment; and grade 4, a displaced fragment [loose body]).
Etiology
Trauma Fairbank47 proposed that repetitive impingement of the tibial spine on the lateral aspect of the medial femoral condyle during internal rotation of the tibia is a causative factor in osteochondritis dissecans. This theory was later supported by Smillie131, who thought that concomitant ligamentous laxity is an associated factor. Another indirect traumatic etiology was postulated by Aichroth1, who noted that the odd facet of the patella articulates with the classic site of osteochondritis dissecans on the medial femoral condyle when the knee is in full flexion. He thought that the lesion is caused by an ununited subchondral fracture in association with pathological patellofemoral contact. The classic osteochondral fracture of the lateral femoral condyle secondary to a patellofemoral dislocation has been described separately from these lesions and possible etiological factors. Makin90 reported on osteochondral fractures of the lateral femoral condyle secondary to reduction of the patella. He correctly considered these to be acute transchondral fractures rather than osteochondritis dissecans lesions per se72,77,92. Cahill23 recently described osteochondritis dissecans as a stress fracture in 204 patients who recalled no obvious traumatic event. Mubarak and Carroll98 noted that trauma is an unlikely cause of osteochondritis dissecans since affected sites are located at specific locations in the knee and not randomly, as would be expected with trauma. The role of indirect microtrauma as an etiological factor remains theoretical.
Ischemia Osteonecrosis of the knee may produce an osteochondral lesion similar to that of osteochondritis dissecans49. An osteochondral fracture that is secondary to osteonecrosis must be classified as a clinical entity separate from that of osteochondritis dissecans, although some early reports of osteochondritis dissecans may have included such fractures31,149. Chiroff and Cooke30 studied the histological characteristics of six excised lesions of osteochondritis dissecans. Two specimens were loose bodies, and four were fragments that had been partially detached. They found no signs of avascular necrosis in the sections of bone, additional evidence against a vascular etiology for osteochondritis dissecans.
Defects of Ossification Caffey et al.21 reviewed the radiographs of children who had marginal irregularities in the epiphyseal outline of the distal part of the femur and divided the patients into three groups. Patients in group I had varying degrees of roughening of the epiphyseal surface and small areas of calcification beyond the irregular margin, those in group II had larger indentations in the epiphysis, and those in group III had the same characteristics as those in group II as well as independent islands of bone in the marginal indentation or crater. These investigators thought that the radiographs of the patients in groups II and III resembled those of patients who have osteochondritis dissecans. Sontag and Pyle134, however, reported that irregularity or roughening of the epiphyseal outline of the distal femoral epiphysis may be a normal variant. They were able to associate these irregularities with periods of rapid skeletal growth. Their data may constitute an argument against the theory that abnormal ossification is associated with osteochondritis dissecans, as the epiphyseal defects seen in many patients with early lesions may have been a result of normal development. Mubarak and Carroll, in two studies97,98, carefully examined patients who had juvenile osteochondritis dissecans of the knee. In 1979, they reported on a family in which twenty of thirty-one members had the lesion97. Interestingly, radiographs of all of the children who were less than twelve years old showed irregularity of ossification bilaterally in at least one location. These authors thought that juvenile osteochondritis dissecans is usually the result of irregularities of ossification. They later reported on seventy-three patients who had osteochondritis dissecans of the knee, thirty-four (47 per cent) of whom had multiple lesions98. They concluded that multiple lesions may represent a forme fruste of epiphyseal dysplasia. They found little support for the concept that either internal trauma (to the tibial spine or the cruciate ligaments) or external trauma causes the disorder.
Genetic Causes In a review in 1977, Petrie111 found no definite genetic etiology for osteochondritis dissecans. First-degree relatives of thirty-four patients who had osteochondritis dissecans were studied for the presence of the disease, and only one was found to have it. A hereditary influence in some patients who have osteochondritis dissecans has been reported by Stougaard136,137, Gardiner54, and others53,63,103,113,114,132. However, in most patients, hereditary influence is probably slight.
Natural History and Treatment The treatment of osteochondritis dissecans remains somewhat controversial for several reasons4. Because this disorder is relatively uncommon, most physicians have not had substantial clinical experience with a large number of patients. Healing of created osteochondral defects with cartilaginous tissue has been noted in experimental animal models124, but there has been limited success in humans. The ability to induce mechanically sound hyaline cartilage to cover the surface of subchondral bone is truly a therapeutic challenge17. Furthermore, osteochondritis dissecans affects many different joints, and the lesions vary with regard to size, location, stage of progression, and mechanism of injury. These many variables of presentation make it difficult to draw meaningful conclusions from the literature. To our knowledge, no prospective, randomized clinical trials have been performed to compare various regimens of treatment for damage to the articular surface. Furthermore, there is confusion because of the presentation in the literature of juvenile and adult osteochondritis dissecans as the same condition23. DePalma et al.39 and Mitchell and Shepard95 described the reparative process that articular cartilage undergoes during the healing response. Defects of articular cartilage do not heal with the regeneration of normal cartilage; instead, they are filled with mesenchymal tissue, which is converted into fibrocartilage. This newly formed fibrocartilage possesses a decreased amount of proteoglycans compared with that found in hyaline cartilage. Fibrocartilage is less resilient than normal articular hyaline cartilage, and continued trauma to the joint eventually leads to degeneration of the articular surface76. Green and Banks59 used a cast or brace in an attempt to prevent fragmentation of the joint surface so that healing could progress. In their series of twenty-four patients who ranged in age from four to fifteen years old (seventeen patients were less than twelve years old), they reported an excellent result (normal physical and radiographic findings and no symptoms) in seventeen of eighteen children who had been managed non-operatively with a brace or cast for an average of four months. Eight patients had a bilateral lesion about the knee. Chiroff and Cooke30 studied the histological and microradiographic characteristics of osteochondritis dissecans lesions in older adolescents and young adults. They found no completely detached lesions, and the underlying bone was viable, being covered by a bed of fibrocartilage that was undergoing endochondral ossification. Because signs of healing were present, these authors recommended non-operative management, unless the lesion was completely separated, for this subset of patients. Crawfurd et al.34 noted the importance of the location of the lesion within the joint. They studied twenty-one undetached lesions of the knee that had been left in situ and that had been determined to be stable on arthroscopy. Follow-up arthroscopy at an average of 7.5 years showed healing of thirteen lesions. Ten of the twenty-one lesions were in the common location on the lateral aspect of the medial femoral condyle and only three of these had healed, compared with ten of eleven stable lesions located elsewhere in the joint. These authors concluded that stable osteochondritis dissecans lesions in the common location in the knee often do not heal spontaneously, and they recommended that these lesions be treated with internal fixation. Linden80 described the natural history of osteochondritis dissecans lesions in sixty-seven joints of fifty-eight patients, who had been followed clinically and radiographically. At an average of thirty-three years, no loose fragments had been replaced and no internal fixation had been used. Children who had open physes did not have secondary degenerative changes; however, thirty-eight of forty-eight patients who had had the first manifestation of the disease after closure of the physes had gonarthrosis. Linden stated that "symptoms and roentgenographic gonarthrosis become more frequent and approach 100 per cent with time." Löfgren85 also noted spontaneous healing of osteochondritis dissecans lesions in children. The basic tenets of operative treatment include restoration of the congruity of the joint surfaces, enhancement of the local blood supply to the fragment or crater, rigid fixation of unstable fragments, and protected weight-bearing with motion of the joint as soon as possible postoperatively. Accepted operative indications include symptomatic loose bodies; detachment that occurs during observation or non-operative treatment; predicted physeal closure within six to twelve months; juvenile lesions that remain symptomatic despite adequate non-operative treatment; and an established non-union of a fragment, determined on the basis of bone scans, magnetic resonance imaging, and clinical judgment22. Most children who have juvenile osteochondritis dissecans and open physes can be successfully managed non-operatively14,22,23,29,42,59,80,130,139,146,151. The chance that such treatment will be successful decreases with closure of the physes. Cahill22 recommended non-operative treatment only if progression of the lesion and the clinical response of the patient could be closely monitored and if at least six to twelve months remained before closure of the physes. Cahill proposed limitation of activities until the patient was free of symptoms as well as protected weight-bearing with use of splints or crutches22. He did not believe that a cast should be used because of the deleterious effects of immobilization, an observation supported by Hughston et al.67. Cahill22 recommended that non-operative treatment be abandoned if symptoms persist for three months and if serial bone scans show little or no increase in uptake. Smillie130,131 was apparently the first to recommend open drilling, replacement, and internal fixation for a loose, separated fragment in the knee joint. Internal fixation of the loose fragment has also been advocated by Anderson et al.4, Cahill22, Green58, Hughston et al.67, Lipscomb et al.84, and Outerbridge105. Many authors have discussed fixation of loose osteochondritis dissecans lesions to allow healing26,32,82. Treatment of osteochondritis dissecans with simple excision of the loose or partially attached fragment in the knee has been reported to be unsuccessful by several authors3,22,23,58,60,67. Green58 noted poor results after excision but recognized the technical difficulty associated with replacement of a loose fragment. He thought that the results of replacement, if not precise, may be worse than those of excision because of residual irregularity of the articular surface; therefore, he recommended replacement of a large fragment but excision of a smaller one if it cannot be replaced anatomically. Hughston et al.67 thought that open arthrotomy and internal fixation with Kirschner wires to fix a large fragment yielded better results than did excision and curettage of the base of the crater. Almgård and Wikstad3 also recognized the poor results of excision and recommended the routine use of arthrograms to guide treatment. They advocated limitation of activity and protected weight-bearing for patients who had intact articular cartilage. For those who had disruption of the cartilage, they recommended internal fixation of an attached fragment and extraction of loose bodies with reconstruction of the intra-articular defect. These authors did not elaborate on their technique of reconstruction of the defect3. Cahill22,24 also noted categorically poor results with simple excision of a fragment about the knee. Smillie130 was one of the first to describe fixation of a separated osteochondritis dissecans lesion with use of pins. Lipscomb et al.84 used Kirschner wires to reattach a large fragment and emphasized the importance of removal of the fibrous tissue at the base of the crater, drilling of the bed of the defect to increase vascularity, and use of bone graft to restore the congruity of the joint surfaces if necessary. Cahill22 suggested that, since the results of excision are so poor, this procedure should be considered ineffective and every effort should be made to replace a loose fragment unless the degree of maceration prevents reattachment. With his method, the lesion is curetted down to bleeding subchondral bone, with care being taken to avoid overexcavation. The base of the crater is then drilled in an attempt to enhance vascularity, and the fragment is trimmed, replaced, and securely fixed with Kirschner wires. The Kirschner wires are later removed in a retrograde fashion to avoid disturbance of the healing osteochondral segment. Bone-grafting is used to elevate a fragment that would remain depressed when replaced. Fixation with bone pegs57,70,83,128,129 involves the use of autogenous corticocancellous bone graft for internal fixation of the osteochondritis dissecans lesion to the site of the defect. Matchstick-sized strips of corticocancellous bone are obtained from the proximal part of the tibia and placed into the fragment and the site of the defect. Cancellous bone-graft material is packed as needed into the lesion to restore articular congruity. Lindholm et al.83 reported good results with this method for sixteen of eighteen lesions, with eventual healing of the defect in all seventeen patients who were studied. Johnson and McLeod70 used this technique in two patients; they noted that it provided easy access to bone-graft material and obviated the need for removal of the hardware while providing biological fixation, which may stimulate revascularization of the subchondral bone. They also noted that Kirschner wires can provide poor fixation and may be bent or broken in the postoperative period and protrude into the joint. Scott and Stevenson128 used this technique with supplemental Kirschner-wire fixation in fifty-four patients; forty-seven were able to return to active military duty. Slough et al.129 recently reported good results with use of bone-peg fixation in eight of ten knees (nine patients), at an average of 2.9 years. Attention has recently been focused on the reconstruction of osteochondritis dissecans defects with use of osteochondral allogeneic and autogenous grafts. McDermott et al.87 inserted allogeneic grafts, within twenty-four hours after they had been obtained, to reconstruct 100 osteochondral defects resulting from various conditions, including osteonecrosis, osteoarthrosis, trauma, and osteochondritis dissecans. Four of the defects were the result of osteochondritis dissecans, and the treatment was successful for only one of them. McDermott et al. concluded that this technique may be better suited for indications other than osteochondritis dissecans. Garrett55 reported the results in twenty-four patients who had an osteochondral defect of the knee that had been treated with an osteochondral allogeneic graft; all twenty-four had improvement postoperatively. Radiographs showed osseous bridging at six to twelve weeks, and arthroscopy revealed apparent viable articular cartilage without chondromalacia55. Yamashita et al.158 used internal fixation with AO screws in the transplantation of osteochondral autogenous grafts in two patients. The grafts were taken from a non-weight-bearing portion of the medial femoral condyle. Transplantation was followed by a period of non-weight-bearing and by subsequent removal of the screws. A biopsy performed at the time of the removal of the screws showed viability of the graft.
Arthroscopy In an early report, Guhl61 reviewed the use of various arthroscopic procedures directed at preservation of the fragment and noted healing in eighteen (75 per cent) of twenty-four patients. Recently, Bradley and Dandy18 used an arthroscope in conjunction with drilling with Kirschner wires for fragments that were attached. They reported healing in nine of eleven patients at one year. They also reported prompt relief of pain and good results in children but found that adults managed with the same procedure had no relief of pain. Ewing and Voto46 used an arthroscope to aid in removal of the fragment in conjunction with drilling and abrasion of the base of the crater in twenty-nine young adults who had osteochondritis dissecans. At thirty-five months, twenty-one patients (72 per cent) had a satisfactory result as assessed subjectively and fifteen (52 per cent) had associated intra-articular abnormality. Of ten patients who had so-called second-look arthroscopy (re-examination after treatment), eight had healing with fibrocartilage. Anderson et al.4 used diagnostic arthroscopy, followed by antegrade curettage (reflection or temporary removal of the fragment with curettage of the crater), local femoral bone-grafting, and pinning of large lesions of the medial femoral condyle through a medial arthrotomy to treat osteochondritis dissecans lesions in skeletally mature patients. They immobilized the knee for six weeks postoperatively and removed the pins eight to twelve weeks after healing. Sixteen of seventeen lesions healed, with two excellent, nine good, and five fair results. Johnson et al.71 recently reported on arthroscopically assisted fixation with screws for the treatment of osteochondritis dissecans in thirty-five knees (thirty-two young adults). They used cannulated, partially threaded 4.0-millimeter AO screws and performed a second arthroscopy for removal of the screws two months postoperatively (Figs. 3-A and 3-B). They found that thirty-three lesions (94 per cent) had healed by the time of the second arthroscopy, and twenty-eight patients (88 per cent) had an excellent or good result. Three types of lesions were encountered: those in which the articular cartilage was intact, with the lesion being either stable or mobile with compression; those in which the articular cartilage had separated, with the lesion remaining either in situ or hinged; and completely loose fragments. They noted erosion of the head of the screw into the opposing tibial articular cartilage in patients who did not comply with the protocol of strict postoperative non-weight-bearing.
Thomson141 reported radiographic healing in sixteen of eighteen patients who had been managed with Herbert compression-screw fixation. Wombwell and Nunley155 reported healing in one patient with use of this technique. Mackie et al.89 inserted Herbert screws under arthroscopic guidance in three patients who had early-stage osteochondritis dissecans. The patients were allowed to bear weight immediately and had a decrease in symptoms at the one-year follow-up evaluation. Dipaola et al.41 used a number of arthroscopic techniques to treat osteochondritis dissecans. Magnetic resonance imaging was helpful for the prediction of a detached lesion at the time of the operation. On the basis of our review of the literature, we support the following treatment plan. The age of the patient is of primary importance in clinical decisions. Children who have a symptomatic lesion should be managed initially with limitation of activity until they no longer have symptoms. Protected weight-bearing is advised to allow healing of the subchondral bone and to prevent further displacement or degeneration of the fragment. Because it is difficult to restrict the activity of a child, a knee immobilizer, splint, or cast may be needed. Close clinical observation with serial radiographs is also important. Serial magnetic resonance images may also be useful for following healing. Operative intervention is warranted if loose bodies are present, if symptoms persist for six to twelve months, or if healing is unlikely as determined by radiographic indices, especially if skeletal maturity is imminent. Arthroscopy allows direct visualization of the defect, and drilling of intact lesions with soft articular cartilage promotes revascularization. Children who have a completely separated, displaced fragment should be managed in a manner similar to that used for adults. A more intensive operative approach should be used in adults if healing after non-operative treatment is unlikely. Guhl62 provided useful guidelines for the arthroscopic treatment of osteochondritis dissecans of the knee. He suggested that lesions with intact articular cartilage can be drilled in a retrograde fashion under arthroscopic guidance. Separated lesions should be treated with curettage and drilling of the base of the crater followed by replacement of the fragment61,62. Stable fixation with compression screws, bone pegs, or biodegradable pins is essential for osseous healing, and bone-grafting may be needed to restore the congruity of the articular surfaces. Excision should be reserved for small fragments or for lesions that cannot be reconstructed. Strict postoperative non-weight-bearing is necessary until healing has been documented, but early, active range-of-motion exercises are encouraged. Young patients who have a completely detached, loose fragment and advanced degenerative joint disease present a difficult and challenging therapeutic dilemma135. The treatment decisions are even more difficult when the lesion involves a major portion of the weight-bearing surface. The use of allografts to reconstruct intra-articular defects appears promising, but this technique is unproved. Osteochondral autogenous grafts have been associated with a good rate of success, but experience with this procedure is limited. Older patients who have an advanced lesion may be candidates for a total knee arthroplasty or a tibial osteotomy.
Clinical Presentation The lesions are found in either the posteromedial or the anterolateral aspect of the talus. Posteromedial lesions are frequently asymptomatic, and the patient may have no history of trauma. Bauer et al.10 noted almost no symptoms after long-term follow-up of osteochondritis dissecans lesions of the ankle in thirty patients. In that series, twenty-seven (87 per cent) of the thirty-one joints had a medial lesion. Rödén et al.120 reported that medial talar lesions were not secondary to trauma, were associated with fewer symptoms and little or no osteoarthrosis, and healed spontaneously. In contrast, lateral lesions are frequently symptomatic and result from trauma2,25,86,91,159.
Physical Examination
Imaging Studies Radiographs show a well circumscribed area of sclerotic subchondral bone separated from the remainder of the epiphysis by a radiolucent line. Radiolucency of the talar chondral surface is also seen. Computed tomography is helpful for ruling out conditions similar to osteochondritis dissecans, such as small lytic or cystic lesions or osteonecrosis163. Although plain radiographs are usually sufficient to make the diagnosis, computed tomography can be beneficial, in preoperative planning, for accurately determining the size and location of the lesion. Magnetic resonance imaging is also useful for identifying the location of the suspected lesion. Dipaola et al.41 correlated the findings on magnetic resonance imaging without contrast medium with the stage of the lesion as determined arthroscopically. These authors found magnetic resonance imaging to be extremely useful for distinguishing a stage-II from a stage-III lesion on the basis of a fluid interface. Yulish et al.160 found magnetic resonance imaging without contrast medium to be accurate on the basis of operative findings at arthroscopy or arthrotomy. Of eight patients who were managed operatively, five had disruption of the articular cartilage overlying the talar defect; magnetic resonance imaging had detected the disruption preoperatively in four of the five. De Smet et al.40 also documented the usefulness and value of magnetic resonance imaging of the ankle without the use of contrast agents for predicting the integrity of the cartilage surface of osteochondritis dissecans lesions. Partially detached lesions were diagnosed by the presence of an irregular signal zone at the interface between the fragment and the bone of the talus on T2-weighted images. The four unattached fragments were completely surrounded by fluid.
Etiology
Trauma
Canale and Belding28, in a clinical study of medial and lateral lesions of the talus, concurred that lateral talar lesions are traumatic in origin and most likely result from injury of the talus against the fibular articular surface. However, they noted that nine of fourteen medial lesions were not associated with trauma. Furthermore, the medial lesions were different morphologically from the anterolateral lesions; the medial lesions were dome-shaped, whereas the lateral lesions had a wafer-like shape. The lateral and medial lesions were identical on histological examination; both had a viable hyaline surface and a necrotic osseous section. These authors concluded that all lateral lesions are traumatic in origin, wafer-shaped, and associated with more persistent symptoms and degenerative changes than are medial lesions. Canale and Belding thought that lateral lesions are true transchondral fractures, which are traumatic in origin. Medial talar lesions can be related to trauma but, in their series, they may have been the result of another factor. Athanasiou et al.6 evaluated the compressive stiffness of portions of the articular surfaces of the tibiotalar joint. They found significant differences in stiffness in the posteromedial talar articulations in fourteen cadaveric specimens (p < 0.05). These investigators hypothesized that a mismatch of stiffnesses of the articulating surfaces results in a posteromedial lesion because of repetitive microtrauma.
Natural History and Treatment Flick and Gould51 studied the results of arthroscopy in the treatment of osteochondritis dissecans of the talus. They noted the ease of arthroscopic access to lateral talar lesions through the standard anterolateral portal. Posteromedial lesions presented a greater technical challenge, and these authors described a technique of grooving of the anteromedial corner of the tibia to allow better exposure. Successful arthroscopic treatment has also been reported by others7,45,109,115. Baker et al.7 reported generally good results in ten ankles, eight of which had a posteromedial lesion. Ewing45 recently described his technique of excision of the fragment followed by arthroscopic curettage of the base of the crater with use of an abrasion shaver; however, he did not report the results in that paper. Angermann and Jensen5 described their operative experience with a medial malleolar osteotomy for exposure and curettage of a medial talar lesion in sixteen patients. These authors noted deterioration of the results over time, with onset of pain and degenerative changes in the joint. Bryant and Siegel20 described the drilling of talar lesions with use of a meniscal repair cannula to allow accurate and safe placement of the pins. Arthroscopy is very useful; however, depending on the presentation of the lesion, arthrotomy may be needed to perform the procedure thoroughly.
Clinical Presentation Pappas107 divided osteochondritis dissecans of the elbow into three categories on the basis of the age of the patient. Category I includes patients less than thirteen years old; category II, those between thirteen years old and adulthood; and category III, adults. Similar to previous investigators who have treated osteochondritis dissecans of the knee, Pappas found that younger age-groups have better results with non-operative treatment. He also noted worse results in association with a more advanced stage of the lesion. Patients who have an osteochondritis dissecans lesion of the elbow describe an insidious onset of poorly localized, diffuse joint pain. Stiffness and swelling are variably reported. Approximately 20 per cent of patients (three of twenty-one in one study19 and five of fourteen in another88) have a history of a single traumatic event1,15,27,36-38,58,80,128,142,161. The mechanism of injury is usually overuse, with repeated microtrauma being the most common factor19,88. In patients who have more advanced irregularity of the articular surface, more dramatic symptoms, such as catching, clicking, giving-way, and grinding, may be reported. If loose bodies are present, true locking of the joint may occur. Symptoms are usually intermittent and occur with exertion or throwing activities.
Imaging Studies Magnetic resonance imaging is extremely useful for evaluating the presence and stage of the lesion. Murphy100 described the use of magnetic resonance imaging in three patients who had osteochondritis dissecans of the elbow. Detachment of the fragment was noted on the basis of intervening fluid seen on T2-weighted images. The operative findings showed that the stage of the lesion had been correctly predicted with magnetic resonance imaging.
Etiology
Trauma Tullos et al.144,145 described the biomechanics of the throwing motion and noted the abnormal valgus stress placed on the elbow, especially during the acceleration and follow-through phases of pitching. They suggested that osteochondritic changes in the capitellum are secondary to compressive forces between the radial head and the capitellum as a result of tensile stresses exerted on the medial aspect of the elbow during throwing. Jobe and Nuber69 postulated a similar mechanism of injury. Brown et al.19 evaluated eighteen patients who had osteochondritis dissecans of the elbow and stated that all of the lesions were related to repeated episodes either of throwing or of participation in racquet sports. They concluded that the "constant impingement by the radial head against the capitellum is in part responsible for the ultimate breakdown of the surface of the capitellum and the irregularity and hypertrophy of the radial head." Mitsunaga et al.96 found that thirty-eight (67 per cent) of fifty-seven patients who had osteochondritis dissecans of the capitellum were involved in competitive sports, mostly baseball, and postulated the high stresses placed on the elbow during pitching as the etiology. Brown et al.19 noted that three patients (athletes involved in throwing activities) of twenty-one had a history of acute injury to the area; these three patients had first been seen with a non-united capitellar osteochondral fracture six months after the initial injury. Proponents of the theory of a traumatic etiology have postulated that the subchondral bone is weakened by cyclic loading as a result of repetitive microtrauma or, less commonly, by a direct traumatic event140. In a treatise on articular trauma, Tallqvist140 noted that the injury eventually results in fatigue failure of the subchondral osseous elements. If the fatigue fracture fails to heal, resorption of bone at the site of the fracture leads to early separation of a fragment from its underlying bed. The separated segment of bone becomes avascular. The overlying articular cartilage is subjected to shear stresses and breaks down because osseous structural support has been lost. This breakdown of cartilage leads to fragmentation and partial or complete separation of the osteochondral fragment and to formation of a loose body. The cartilage is nourished by diffusion from the synovial fluid and remains viable, but the attached bone fragment is avascular, having lost its blood supply. Histological sections of a fragment caused by microtrauma should therefore reveal viable articular cartilage and a variably sized segment of necrotic subchondral bone. Schenck et al.126 found a mechanical disparity of compressive stiffness between the radial head and the capitellum in cadaveric elbows. They theorized that a stiffer radial head articulating with a softer capitellum produces greater strain at the capitellar surface during compressive stresses in the radiocapitellar joint and can result in an osteochondritis dissecans lesion.
Ischemia
Natural History and Treatment McManama et al.88 reported the results of treatment of osteochondritis dissecans involving the capitellum in fourteen adolescents. An excellent or a good result was noted in thirteen patients after a loose but attached capitellar chondral fragment had been shaved down to bleeding bone. Through a lateral arthrotomy, small defects were excised and multiple drill-holes were made with Kirschner wires to promote revascularization. Bone-grafting was not performed, and no attempt was made to preserve loose, attached fragments. These authors found arthrography to be of little value, and they cautioned against use of procedures such as local bone-grafting, internal fixation, and excision of the lateral condyle in the treatment of capitellar osteochondritis dissecans. Woodward and Bianco156 reported on forty-two male patients who had been managed for osteochondritis dissecans of the elbow between 1935 and 1970. Of twenty-two elbows for which radiographs were made, all had involvement of the capitellum. These authors' indication for operative treatment was the presence of a loose body. Curettage or drilling of the capitellar defect did not provide a better result than did removal of the loose body alone. Limitation of extension (less than 20 degrees) was a consistent finding at the long-term follow-up evaluation, yet the patients described little functional limitation. Brown et al.19 reported on twenty-one patients who had osteochondritis dissecans of the elbow. Three patients had had an acute traumatic injury; in the remaining eighteen, the lesion was the result of overuse, with a throwing-type mechanism of injury. Radiographic findings of capitellar radiolucency were common. Removal of a loose body was the most frequently performed procedure. Functional limitations were common, and extension of the elbow was routinely limited. In summary, most authors have noted osteochondritis dissecans of the elbow in athletes who are involved in throwing activities, as a result of repetitive overuse (microtrauma). Treatment usually involves débridement of loose bodies. The capitellar bed or crater should be curetted to remove any remaining loose fragments and then should be drilled with a Kirschner wire to stimulate vascular ingrowth. Degenerative changes and limitations of motion are to be expected in these patients.
|
Top
Introduction
History
