From March 2007 to August 2007, 33 patients (37 cases) were treated with an open wedge HTO using an Aescula open wedge plate and an allogenic bone graft by the senior author (KAJ). The study group included twenty nine women and four men with a mean age of 53 years (range, 42 to 61 years). The exclusion criteria for this procedure were symptomatic osteoarthritis of the patellofemoral joint and lateral compartment, rheumatoid arthritis, a history of fractures or previous open surgery of the lower limb, and flexion contracture. The mean follow-up duration was 15.6 months (range, 12 to 18 months).

The clinical assessments included an evaluation of the knee joint function using the knee and function scores based on the Knee Society Score, and the complications such as infection, nonunion and peroneal nerve palsy. The standardized radiography of the knee included the lower extremity scannogram (the standing anteroposterior view including the hip, knee and ankle joint), both knee standing posteroanterior 45° flexion views, both knee standing anteroposterior views and a true lateral radiograph with the knee in 30° of flexion. The pre-operative lower extremity scannogram was used to calculate the weight bearing line (WBL) percentage at the tibial intersection. The denominator was the width of the tibia, and the numerator was the tibial intersection of the WBL (with the medial tibial edge at 0% and the lateral tibial edge at 100%). The methods for determining the correction wedge preoperatively were the same as those previously described by Dugdale et al.4) The WBL was calculated to pass through the lateral tibial plateau at 62% of the width of the plateau if a two or three degrees of valgus overcorrection was required. A radiographic assessment was performed using the radiography of the lower extremity scannogram, where the femoro-tibia angle was measured using the method reported by Bauer et al.,5) and the mechanical axis of the lower extremity was the angle between the femoral and tibial mechanical axis (physiological range, 0 to 2.2° of valgus).6) The joint space distance was measured as follows. A line was drawn parallel to the both femoral condyles and a second line was parallel to both the tibial condyles. A perpendicular line from the lowest point of the femoral condyle to the tibial condyle line was drawn. The length of this perpendicular line was the joint space distance. Therefore, the degree of joint space loss and the postoperative recovery of the joint space were indirectly evaluated.7) The posterior tibial slope was assessed on the lateral X-ray. The tibial shaft axis was calculated by using a previously described technique,8) in which 2 points were identified halfway between the anterior and posterior cortices and a line drawn through those points. The tibial plateau was measured from the line of best fit on the lateral radiograph with using the medial tibial plateau. The tibial slope was the angle formed between the proximal plateau and a line drawn perpendicular to the tibial shaft axis. The Insall-Salvita (IS) ratio was calculated by dividing the length of the patellar ligament by the length of the patella.9)

The clinical indicators we used to assess union were the complaints of pain as well as the signs of tenderness, warmth and swelling at the osteotomy site. The radiographic examinations involved the anteroposterior (AP), lateral and oblique radiographs of the osteotomy site. The indicators used to assess bone union were the increasing density of the graft on the serial examination and bone bridging across the wedge opening. The signs of nonunion were the absence of the above indicators at 6 months after surgery and sclerotic margins at the periphery of the wedge opening.

Statistical comparisons between the pre- and post-operative data (at the last follow-up) were performed using a paired sample t-test (SPSS ver. 13.0; SPSS Inc., Chicago, IL, USA). A p-value < 0.05 was considered significant.

An arthroscopic examination was carried out for all the cases, which confirmed the status of the cartilage and the concurrent presence of damage. Chondroplasty and a partial meniscectomy were performed for the cases where there was damage to the articular surface or the meniscus was present.

After the arthroscopy, the anteromedial aspect of the tibia was exposed through a 6 cm vertical skin incision, and a sharp dissection was performed to expose the deep fascia over the pes anserineus, which was retracted distally using a periosteal elevator to expose the superficial fibers of the medial ligament (Fig. 1). A blunt Hohmann retractor was then passed deep to the medial collateral ligament around the posteromedial corner of the proximal tibia, and also along the posterior cortex of the tibia to protect the posterior neurovascular structures. For a biplanar medial open wedge osteotomy,10) two Kirschner wires were placed under fluoroscopic guidance to mark the saw cut. The start point on the medial tibia is normally approximately 3-4 cm distal to the medial joint line. The anteroposterior plane of the osteotomy should be in the same plane as the tibial slope. The lateral aiming point was the upper third of the proximal tibiofibular joint. Behind the tibial tuberosity, the osteotomy was performed with using a thin osteotome to protect the patellar tendon in a plane parallel to the anterior cortex of the tuberosity. A tibial osteotomy was performed just distal to the guide pin, initially with an oscillating saw and then with a thin osteotome. The extent of the osteotomy was checked by fluoroscopy to ensure an appropriate depth and direction of the cut (up to 1 cm medial to the lateral cortex). Opening of the osteotomy and adequate correction in the coronal plane were confirmed by placing a rod from the center of the hip joint to the center of the ankle joint (the weight bearing axis) under fluoroscopic guidance. The rod should be located at 62% of the tibial width and just lateral to the lateral tibial spine.4) The tibial slope was controlled using two plates with different sized space blocks (B. Braun, Aesculap) (Fig. 2). First, a properly sized open wedge plate with a space block was inserted posteromedially with gentle valgus force. The tibial slope was identified on the fluoroscopic image after placing the first plate within the posteromedial osteotomy gap. The anterior plate was placed behind the oblique tuberosity osteotomy and the anterior plate was 2 to 4 mm shorter depending on the size of the posterior plate. The fluoroscopic image was then compared with the preoperative radiographic image. Osteosynthesis was performed as follows. Two screws were applied to the first plate that was placed posteromedially. A smaller plate was then inserted within the anteromedial gap and locked with another two screws. A final fluoroscopic assessment was performed to ensure proper alignment of the lower extremities and adequate positioning of the open wedge plate. An allogenic bone graft was then inserted within the osteotomy gap. In each case, a proximal tibial corticocancellous wedge allograft (hemi-femoral head) and cancellous pieces (Fig. 3) were provided by the Musculoskeletal Transplant Foundation (the Korean Bone Bank). The wedge (ordered to match the size of the predetermined osteotomy opening and it was made up of the medial calcar portion of allogenic bone) was then measured and cut to precisely fill the osteotomy defect. Cancellous pieces were then inserted into the posteromedial (the same size of space block as the Aescula plate) and middle areas of the ostotomy tines, and a corticocancellous piece was inserted into the anterior area between the plates (Fig. 4). In addition, a careful radiographic evaluation of the lateral hinge bone was used to detect the presence of any fracture. If there was a fracture, then supplementary lateral fixation with a 6.5 mm cannulated screw or staple was carried out. The wound was closed with leaving a hemovac drain; the wound was dressed with a compression wrap and a long-leg splint was applied.

The postoperative protocol included quadriceps and range-of motion exercises on the first day after surgery to obtain knee motion. A long leg cylinder cast was applied three to four weeks after surgery and partial weight-bearing was then started. This was followed by full weight-bearing eight weeks after surgery. The cast was removed after no changes in the opening angle of the osteotomy site were noted with full weight-bearing (Fig. 5).

The mean knee score significantly improved from 52.19 ± 11.82 preoperatively to 92.49 ± 5.10 at the last follow-up (p < 0.001). The mean functional score significantly improved from 52.84 ± 6.23 preoperatively to 89.05 ± 5.53 at the last follow-up (p < 0.001).

The femoro-tibia angle, which was measured using Bauer's method,5) had a mean varus of -1.86 ± 2.76° preoperatively. This was corrected to a mean valgus of 10.93 ± 2.50° at the final follow-up (p < 0.001). The mechanical axis of the lower extremity showed a mean varus of -7.34 ± 2.86° preoperatively. This was corrected to a mean valgus of 4.38 ± 2.30° at the final follow-up (p < 0.001). The mean posterior tibial slope was 8.20 ± 1.80° preoperatively and 8.04 ± 1.30° at the final follow-up (p = 0.437) (Fig. 4). The mean joint space distance increased from 4.05 ± 1.30 mm pre-operatively to 4.83 ± 1.33 mm at the final follow-up (p < 0.001). The pre- and postoperative Insall-Salvati ratios were 1.03 ± 0.11 and 1.09 ± 0.11, respectively (p = 0.018) (Table 1). The time needed to achieve complete clinical and radiographic bone union was 12.69 ± 1.5 weeks.

No complications were encountered, such as delayed union, nonunion, superficial infection, deep infection, skin necrosis, peroneal nerve and vascular injury. There was one case of skin irritation due to screw back-out (Fig. 6). However, this was resolved after removing the plate and screws after bone union.