We performed a valgus opening-wedge osteotomy in nine patients (nine tibiae) who had a post-traumatic varus deformity of the ankle (Table I). We excluded any patient who had primary osteoarthrosis of the ankle. There were four male and five female patients who had an average age of thirty-five years (range, twelve to sixty-one years). As the nine patients had been managed previously in other hospitals, detailed information on the initial findings and management was not available. The initial injury consisted of an epiphyseal fracture of the distal end of the tibia in four patients, a malleolar fracture in three, and osteomyelitis of the tibia that had developed after an injury in two. Two (Cases 4 and 7) of the four ankles with an epiphyseal injury were treated with immobilization in a cast, and two (Cases 3 and 5) had open reduction and internal fixation. One (Case 3) of those had had an open fracture. A comparison of the initial radiographs, which were available for three of the ankles that had an epiphyseal injury, with those made when the patients were first seen by us showed an increase in the varus deformity (Figs. 1-A, 1-B, 1-C and 1-D).
Two of the three patients who had a malleolar fracture had an associated fracture of the tibial plafond. One patient (Case 1), who had a bimalleolar fracture, was managed with immobilization in a cast for two months, and the other (Case 6), who had a trimalleolar fracture, was managed with open reduction and internal fixation with a plate. The third patient (Case 2) had a supination-adduction type of bimalleolar fracture (involving the medial and lateral malleoli), according to the classification system of Lauge-Hansen5. We believed that the varus deformity in these patients was due to the persistent malalignment that had begun at the time of the initial treatment.
Of the two patients who had osteomyelitis, one (Case 8) had chronic osteomyelitis; she was eight years old at the time of the initial injury and was fifty-six years old when we first saw her. Although the details of the initial treatment were not available, the scar on the medial side of the leg and the radiographic findings of an irregular cortex with sclerosis of the tibial metaphysis suggested that she may have had an open fracture. The patient stated that she had had purulent drainage from the distal part of the leg for six months. The second patient (Case 9) had acute osteomyelitis after an open pilon fracture; it had been treated with antibiotics and irrigation by local perfusion five years earlier in another hospital.
A corrective osteotomy was indicated when the angular deformity between the tibial shaft and the tibial joint surface on the anteroposterior radiograph (the TAS angle6,10,11) was less than 80 degrees, the patient had mild or moderate pain in the ankle as well as the subtalar joint, and non-operative treatment had failed. Osteotomy was indicated for children only when they had pain after walking a long distance (one patient had pain after walking two kilometers and two, after walking four kilometers), had difficulty participating in sports activities, and had a progressive deformity as well as uneven wear of the sole of the shoe, with more rapid wear of the lateral part.
Preoperative problems, in addition to the varus deformity of the ankle, included a medial scar (two patients), a callosity (two patients), claw toes (two patients), and hallux valgus (one patient). One patient (Case 8) had had transplantation of a vascularized peroneal flap to cover the scar on the medial side of the leg before the corrective osteotomy. The callosity and claw toes were secondary to a malunited fracture (Case 1) or to osteomyelitis (Cases 8 and 9), whereas the hallux valgus was an incidental finding. The average interval between the initial treatment of the injury and the index operation was twelve years and four months (range, two years and seven months to forty-eight years and five months).
We used weight-bearing radiographs of the ankle to classify the osteoarthrosis as stage 1 (no narrowing of the joint space but evidence of sclerosis and the formation of osteophytes), stage 2 (narrowing of the joint space medially), stage 3 (obliteration of the medial aspect of the joint space and contact between adjacent subchondral bone), or stage 4 (complete obliteration of the joint space with bone-on-bone contact). In reviewing the results, we designated stage 1 as early osteoarthrosis, stages 2 and 3 as intermediate osteoarthrosis, and stage 4 as late osteoarthrosis.
We used a 100-point system12 to evaluate the function of the ankle. Pain was given a maximum of 40 points, and the ability to walk, the ability to perform the activities of daily living, and the range of motion, a maximum of 20 points each (Table II). A score of 90 points or more was considered to be an excellent result; 80 to 89 points, a good result; 70 to 79 points, a fair result; and 69 points or less, a poor result.
Preoperatively, the average angle (and standard deviation) between the tibial shaft and the tibial joint surface on the anteroposterior radiograph (the TAS angle6,10,11) was 70.0 ± 7.8 degrees (range, 54 to 78 degrees), confirming the presence of a severe varus deformity. The average TAS angle in Japanese individuals who have normal ankles is 88.1 degrees6,10,11. The average preoperative angle of the tibial joint surface on the lateral radiograph (the TLS angle6,10,11) was 75.1 ± 5.8 degrees (range, 66 to 85 degrees). The average angle in Japanese individuals who have normal ankles is 81 degrees6,10,11. The average limb-length discrepancy, as measured from the center of the hip joint to the center of the distal joint surface of the tibia on a scanogram, was 2.3 ± 1.7 centimeters (range, 0.5 to 4.2 centimeters), and the average shortening of the tibia was 2.2 ± 1.1 centimeters (range, 0.6 to 3.5 centimeters) (Table I). In one patient (Case 2), the length of the femur on the affected side was 49.2 centimeters and the length of the ipsilateral tibia was 35.4 centimeters, whereas the length of the femur on the unaffected side was 48.3 centimeters and the length of the ipsilateral tibia was 37.4 centimeters.
An opening-wedge valgus osteotomy was performed in all nine patients with use of a template made from the preoperative measurements. The tibial osteotomy in the adult patients was done at the level of the initial site of the fracture of the tibia if possible. If this was not possible, the osteotomy was performed five centimeters proximal to the distal end of the medial malleolus. In three patients who had open growth plates, the osteotomy was performed two to three centimeters proximal to the epiphyseal plate. The oblique osteotomy of the fibula was performed first, through a lateral incision, and the tibial osteotomy was done through a longitudinal medial incision. The horizontal osteotomy through the tibia was performed along a line parallel to the distal joint surface of the tibia, preserving a small part of the posterolateral tibial cortex to act as a fulcrum for the opening wedge and to provide stability. The planned anteromedial opening was carefully enlarged, and the space was filled with an appropriately shaped bone graft obtained from the ilium. The size and thickness of the tricortical bone graft were determined on the basis of the preoperative measurements. The height of the bone graft averaged 12.4 ± 4.0 millimeters; the width, 24.6 ± 2.9 millimeters; and the depth, 11.8 ± 3.3 millimeters.
Kirschner wires were used to stabilize the site of the tibial osteotomy in eight patients: two wires were used in six patients, and three wires and four wires were used in one patient each. One Kirschner wire was used to stabilize the site of the fibular osteotomy in four patients. A plate and screws was used for fixation of the tibia in one patient (Case 1), who had had a malleolar fracture, because of initial concern about the fixation.
All patients reported postoperative pain over the iliac crest, which usually decreased after five to seven days. At the time of the latest evaluation, none of the patients had pain over the iliac crest.
The average duration of follow-up was seven years and four months (range, two years and four months to thirteen years and two months) (Table I). The average time to osseous union in eight patients was six and one-half weeks (range, four to nine weeks). One patient (Case 9), who was fifty-nine years old, had osseous union at six months. The Kirschner wires were removed an average of four months (range, two to eight months) after the osteotomy.
Postoperatively, the range of motion of the ankle was decreased in six patients (Cases 1, 2, 3, 5, 6, and 8) and remained unchanged in three (Cases 4, 7, and 9) (Table I). No patient had any limitation in the activities of daily living that was attributable to the decreased range of motion of the ankle. Two patients (Cases 1 and 9) reported pain over the medial aspect of the ankle, and two others (Cases 5 and 8) had pain in the region of the subtalar joint. The pain in the latter two patients was alleviated by the injection of a local anesthetic and a steroid into the region of the sinus tarsi and the posterior aspect of the talocalcaneal joint.
Postoperatively, the average TAS angle was 87.1 ± 4.3 degrees (range, 78 to 92 degrees), which did not exceed the normal angle. The average postoperative TLS angle was 81.0 ± 4.8 degrees (range, 72 to 88 degrees), which was similar to normal values.
The length of the tibia increased an average of 1.0 ± 0.4 centimeter (range, 0.4 to 1.5 centimeters) postoperatively, which was an improvement from an average discrepancy of 2.2 ± 1.1 centimeters (range, 0.6 to 3.5 centimeters) preoperatively to a postoperative average of 1.2 ± 0.1 centimeters (range, 0.2 to 2.4 centimeters).
All patients had a reduction in pain and an improvement in the ability to walk even though the range of motion of the joint had decreased. Four patients had an excellent result; two, a good result; and three, a fair result. None of the patients had a poor result. The three patients (Cases 1, 8, and 9) who had a fair result were more than fifty years old. The average score for pain improved from 20.0 ± 7.1 points preoperatively to 34.4 ± 5.3 points postoperatively; that for the ability to walk, from 15.6 ± 4.6 points to 18.9 ± 2.2 points; and that for the ability to perform the activities of daily living, from 14.7 ± 3.0 points to 16.9 ± 2.7 points. The average score for the range of motion decreased from 18.6 ± 2.2 points to 17.0 ± 4.0 points. The average total score improved from 68.9 ± 15.3 points to 87.2 ± 12.0 points.
The four patients (Cases 2, 3, 4, and 7) who had been operated on when they were children or adolescents were followed for an average of eight years (range, six years and nine months to nine years and eight months); they were able to participate in physical education classes at school, and two of them were able to participate in competitive basketball and athletic activities.
Osteoarthrosis of the ankle usually develops secondary to trauma. Primary osteoarthrosis—that is, osteoarthrosis without a history of trauma or infectious arthritis—is rare. Patients who have a severe varus deformity have increased pain and a decreased range of motion that becomes even more limited after the distal tibial epiphyseal plate has closed. The only treatment for advanced osteoarthrosis of the ankle is either arthrodesis or total ankle arthroplasty11,12. However, in the early stage, before the hyaline articular cartilage is damaged, realignment of the tibia should be considered in an effort to prevent the progression of osteoarthrosis.
An injury of the ankle may be associated with a residual deformity that sometimes persists or becomes manifest as the patient grows2-4,9. The anatomical configuration of the ankle joint may lead to a varus deformity after a malleolar fracture or an epiphyseal injury. We noted a progressive deformity after an injury of the distal tibial epiphysis in the children and adolescents. In contrast, a varus deformity after a malleolar, tibial plafond, or pilon fracture was usually secondary to inadequate reduction at the time of the initial treatment. We have managed only two patients who had a post-traumatic valgus deformity; they had mild pain or no pain, and the deformity was slight.
Most patients who had a varus deformity of the ankle had some evidence of shortening of the extremity, as there often are changes in the subtalar joint to compensate for a long-standing varus deformity. A corrective osteotomy that does not address this problem could overload the subtalar joint and lead to pain. Therefore, we recommend an opening-wedge osteotomy to lengthen the tibia rather than a closing-wedge osteotomy, which would shorten the tibia. A large bone graft is necessary to perform a one-stage opening-wedge osteotomy. We initially performed internal fixation with use of a plate and screws (Case 1). However, the bone graft made it possible to achieve an appropriate amount of compression with use of two, three, or four Kirschner wires. Lengthening of the extremity should be considered for the management of adolescent patients in whom the tibia has shortened more than four centimeters.
The range of motion of the ankle was decreased postoperatively in most patients. This complication is unavoidable when performing an opening-wedge osteotomy. However, no patient reported any limitation in the activities of daily living that was attributable to the decreased range of motion, and the adolescent patients could participate in most sports at school. Therefore, on the basis of our results, we suggest that a one-stage opening-wedge corrective osteotomy can be effective for the treatment of varus deformity of the ankle associated with early osteoarthrosis.