2.1

A transfibular approach was used for the arthroplasty system. All the patients were placed in the supine position on a radiolucent table. Osteophytes of the medial tibia were removed first in cases where osteophyte removal was deemed necessary. A 4-cm longitudinal incision was made above the medial gutter of the tibial joint. Osteophytes of the medial tibia were removed using a chisel and rongeur. A 15-cm curved longitudinal incision was made along the lateral malleolus. Subperiosteal dissection of the fibula and anterior side of the tibia was then performed to obtain a full view of the joint line and anterior osteophytes. The posterior capsule, anterior talofibular ligament (ATFL), and posterior talofibular ligament were released. Oblique fibular osteotomy was then performed. The distal fibular segment was reflected distally to visualize the ankle joint. The legs were then placed on an alignment stand. After achieving the desired alignment, the ankle was fixed to the stand with calcaneal, talar, and tibial pins. Bone resection was performed by cutting guides using the peck-drilling technique. After drilling the rail holes and trial reduction, the final components were inserted.

The fibula was reduced and repaired with a posterolateral fibular plate using either of the A.L.P.S. Distal Fibula (Zimmer-Biomet, Warsaw, IN, USA), Salus Loc posterolateral fibula (MEIRA, Aichi, Japan), or LCP Distal Fibula Plate (DePuy Synthes, West Chester, PA, USA). Syndesmosis fixation was performed with trans-syndesmostic suture tape using a zip-tight suture tape (Zimmer-Biomet, Warsaw IN, USA) in cases in which syndesmostic stabilization was deemed necessary. Finally, the ATFL was repaired using soft anchors as necessary. Percutaneous Achilles tendon lengthening using Hoke triple hemisection was performed if passive ankle dorsiflexion with the knee extended was < 5°. Ankle dorsiflexion of < 10°occurs from posterior leg compartment contracture is a common indication for surgical intervention in general . The previous studies showed average preoperative ankle dorsiflexion with knee straight was 5.1 degrees in clinically diagnosed equinus contracture . Based on the above rationale, we perform percutaneous Achilles tendon lengthening if passive ankle dorsiflexion with the knee extended is < 5°. The triple hemisection consists of 3 stab incisions centered midline over the Achilles tendon. The first hemisection is performed 2–2.5 cm proximal to the superior calcaneus whereas the second and third hemisections are sequentially performed 2.5–3 cm proximal to each preceding hemisection. The length and degree of correction can be controlled in graduated manner through a stepwise manual manipulation of the dorsiflexed foot .

Postoperatively, the patients were immobilized with a short-leg splint until sutures removal on the 14th post operative day. Squatting exercise with the splint was initiated at one week. The splint is somewhat flexible enough to allow mild dorsiflexion. This squatting motion is expected to apply vertical pressure to the ankle joint prosthesis and encourage bone ingrowth of the trabecular metal of the prosthesis. Active ankle dorsiflexion movement and stomping without a splint were started at two weeks for 30 min twice a day with a physiotherapist. Full weight-bearing gait training for 20–30 min twice a day and passive ankle dorsiflexion movement 10 min twice a day were initiated at three weeks .

2.2

All the patients were evaluated using standard weight-bearing anteroposterior and lateral radiographs of the ankle, and a lateral radiograph was obtained with the ankle in maximum plantarflexion and dorsiflexion with a established method reported by Coetzee . A radiograph was obtained in the position of maximum dorsiflexion in a weight-bearing position with the foot on the same outline. The patient flexed the knee until the foot reached maximum dorsiflexion. The radiographic view in maximum plantarflexion was taken with the patient without any weight and with the ankle in maximum active plantarflexion. Fixed anatomical landmarks were selected to create six radiographic lines to measure the sagittal motion arc ( Fig. 1 ). Ankle dorsiflexion angle, ankle plantarflexion angle, tibiotalar joint angle, subtalar joint angle, talonavicular joint angle were measured. Tibiotalar range of motion was defined as the difference between the tibiotalar joint angle with maximum dorsiflexion and the maximum plantarflexion. Subtalar range of motion was defined as the subtraction angle of the subtalar joint angle with maximum plantarflexion from the subtalar joint angle with maximum dorsiflexion. Talonavicular range of motion was defined as the difference between the talonavicular joint angle with the maximum dorsiflexion and the maximum plantarflexion ( Fig. 1 A, B).

Radiographic measurement of ankle motion(A) Radiographic assessment of maximum dorsiflexion and plantarflexion. (B) Radiographic lines for the calculation of sagittal arc of movement. Line AB was the tibial axis, and line CD was drawn from the inferior aspect of the sesamoid bone to the inferior aspect of the calcaneal tuberosity. Ankle dorsiflexion angle was defined as 90° minus the angle created by lines AB and CD at maximum dorsiflexion. Ankle plantarflexion angle was defined as 90° minus the angle created by lines AB and CD with maximum plantarflexion. Line EF was drawn from the anterior to the posterior aspect of tibia plafond. Line GH was drawn from the lateral to the posterior process of the talus. Tibiotalar joint angle was defined as the angle created by the EF and GH lines. Tibiotalar range of motion was defined as the difference between the tibiotalar joint angle with maximum dorsiflexion and the tibiotalar joint angle with maximum plantarflexion. Line ID was drawn from the inferior anterior process of the calcaneus to the inferior aspect of the calcaneal tuberosity. Subtalar joint angle was defined as the angle created by the GH and ID lines. Subtalar range of motion was defined as the subtraction angle of the subtalar joint angle with maximum plantarflexion from the subtalar joint angle with maximum dorsiflexion. Line JK was drawn from the superior and inferior proximal corners of the navicular region. Talonavicular joint angle was defined as the angle created by lines GH and JK. Talonavicular range of motion was defined as the difference between the talonavicular joint angle with maximum dorsiflexion and maximum plantarflexion. (C) Radiographs showing measurement techniques for preoperative tibiotalar surface angle. (D) Radiographs showing measurement techniques for postoperative α angle, β angle, γ angle, and tibio-talar ratio (BC/AC). The α angle is the angle between the anatomic axis of the tibia and articular surface of the tibial component on the anteroposterior view. The β angle is the angle between the anatomic axis of the tibia and articular surface of the tibial component on the lateral view. The γ angle is the angle between a line connecting the ends of the talar component and a line along the center of the talar neck on the lateral view. The tibiotalar ratio is the ratio of the length of the posterior part of the talus (BC) to its full length (AC) and is expressed as a percentage. (E) Radiographic image of the anterior talar implant depth. The depth measured from the anterior inferior aspect of the talar implant to the dorsal talar cortex. (F) Computed tomography image showing postoperative residual medial tibial osteophytes (black arrows).

Tibiotalar surface angle, tibiotalar ratio, α angle, β angle, γ angle, and anterior talar implant depth were also recorded ( Fig. 1 C-D). Anterior talar implant depth was defined as the depth measured from the anterior inferior aspect of the talar implant to the dorsal talar cortex ( Fig. 1 E).

The presence of the remaining postoperative osteophytes of the medial tibia were determined using postoperative CT images ( Fig. 1 F).

To remove the influence of interobserver variability, all the measurements were performed by an independent board-certified orthopedic foot and ankle surgeon.

2.3

Descriptive analysis was used to summarize baseline characteristics. The Shapiro–Wilk test was used to verify the normality of distribution for all variables. For subgroup comparisons, the Student’s t test or Mann–Whitney U test was performed for continuous variables, depending on the normality of data distribution. The Pearson’s chi-square test or Fisher’s exact test was used to analyze categorical variables. Statistically significant was considered at P < 0.05. All statistical analyses were performed using EZR (ver.4.0.2) (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria) .