The current study retrospectively reviewed consecutive 165 knees (152 patients) who underwent MOWHTO with the TomoFix® implant (DePuy Synthes, Solothurn, Switzerland) between March 2008 and May 2013. Inclusion criteria were (1) symptomatic osteoarthritis at medial compartment with Kellgren-Lawrence grade of III or IV which did not respond to non-operative treatments for at least 3 months; (2) varus deformity of > 5°; (3) preserved articular cartilage of the lateral joint compartment with International Cartilage Repair Society (ICRS) grade of I or II, without tear of lateral meniscus;[20] (4) available for minimum two-year clinical and radiological follow-up after surgery. Exclusion criteria were (1) patients who were not able to undergo second-look arthroscopy; (2) complications including nonunion, malunion, and hardware failure; (3) a history of a previous surgery to the affected knee; (4) patient who had operation on both knees; (5) patients who had undergone a concurrent cartilage restoration procedure during MOWHTO; and (6) knee joint instability of the affected joint caused by any coexisting ligament problem. Finally, 124 knees who met the above criteria were included in this study. Our institutional review board approved this study.
Included patients were divided into three groups depending on the correction angle. The medial proximal tibial angle (MPTA) was measured on knee standing anteroposterior (AP) view before initial MOWHTO and just prior to the second surgery of implant removal. The MTPA was defined as the angle obtained from an intersection between the articular surface of the proximal tibia and the tibial mechanical axis (Fig. 1). The correction angle was defined by subtracting MPTA measured before the second surgery from the initial MPTA.[21] The calculated correction angles were divided into quartiles and the patients were subsequently divided into three groups on the basis of the interquartile range of the correction angles. The first group consisted of patients who were included in the lowest quartile (< 7.8°); these individuals were defined as the small correction angle group (Group S = 31 patients). The second group of patients, who had correction angles in the middle two quartiles (7.8° to 11.6°), were named as the moderate correction angle group (Group M = 62 patients). Patients with a large correction angle, who were included in upper quartile (> 11.6°), were classified as the large correction angle group (Group L = 31 patients).
Surgical Procedure
A single surgeon performed all operations. AP full-length weight-bearing radiograph was taken for preoperative planning, and the target angle of valgus correction was measured with use of Miniaci’s method.[22] The realigned weight-bearing line was planned to pass through the Fujisawa point (62.5% from the medial edge of the tibial plateau).[23] Arthroscopic examination was initially done to assess the medial, lateral, and patellofemoral joint. And then, a skin incision was made obliquely on the medial proximal tibia and distal insertion of the superficial medial collateral ligament was stripped off the bone. Osteotomy site was usually determined near the upper border of the pes anserius. Under the guide of an image intensifier, two Kirschner wires were inserted toward just proximal to the tibiofibular joint. Along the Kirschner wires osteotomy was performed. The lateral-most 10 mm of the proximal tibia was preserved without damage in order to be used as a hinge during correction. A separate ascending osteotomy for the biplanar cut was performed about 1 cm to 1.5 cm posterior to the tibial tuberosity, and it was cut at an angle of 100° to 110° from the direction of initial osteotomy. Using four chisels, the osteotomy site was gradually opened up. To open and measure the gap of the osteotomy, the TomoFix bone spreader was inserted into the osteotomy gap. The opening process was performed such that the anterior gap of the opening was about two-thirds the size of the posterior opening to maintain the tibial slope.[24] Fixation of the osteotomy was done with the TomoFix® plate (DePuy Synthes, Solothurn, Switzerland) and locking screws.
Postoperative Rehabilitation and Implant Removal
A splint was applied for initial two weeks, and then knee brace was worn for next four weeks, during which time range of motion (ROM) of the knee joint was allowed within the tolerable range. Partial weight-bearing was initiated beginning at two weeks after the operation. The brace was discarded, and full-weight bearing was allowed at six weeks. Implant removal was performed at two year postoperatively with confirmation of bone union under simple radiography. Second-look arthroscopy to evaluate the cartilage state was also performed during the operation of removal.
Radiologic Evaluation
For radiological assessments, standing true anteroposterior full-length lower extremity radiograph, anteroposterior and lateral view of the both knee weight bearing radiographs, and 45° Merchant view were taken. In addition to MPTA, the weight bearing line ratio, hip–knee–ankle angle, modified Blackburn-Peel ratio, and the International Knee Documentation Committee (IKDC) radiographic scale value were evaluated by two orthopedic surgeons. Their evaluations were performed independently of each other, and were completed preoperatively and at the last follow-up for all patients. For both the preoperative and last evaluations, each surgeon measured these parameters twice in all knees, with an interval of two weeks between the measurements. The average of these two measurements was used for final analysis.
The weight bearing line ratio was obtained by first drawing a weight bearing line from the center of the femoral head to the midpoint of the talar superior articular surface; subsequently, the horizontal distance from the medial edge of the tibia plateau to the tibial intersection of the weight bearing line was recorded. With the medial edge at 0% and the lateral edge at 100%, the weight bearing line ratio was set as a ratio of the horizontal distance from the medial edge of the tibia divided by the whole width of the tibia (Fig. 2-A, B). The hip-knee-ankle angle was measured by drawing a line from the femoral head center to the tibial spine midpoint and another line extended from this midpoint to the talar surface center of the ankle joint. The angle made by the intersection of these two lines was determined as the hip-knee-ankle angle (Fig. 2-C). The modified Blackburn-Peel ratio was assessed with standing lateral view of 30° knee flexion. The ratio was set as the perpendicular distance between the tibial articular surface and the inferior end point of the patellar articular surface divided by the entire length of the articular surface of patella (Fig. 2-D). The radiologic evaluation of the patellofemoral joint on osteoarthritis was performed with the IKDC radiographic assessment scale on 45° Merchant view, and the scale was categorized as follows: A = normal; B = nearly normal, showing > 4 mm of a joint space with evidence of minimal changes in the joint including early osteophytes, slight sclerosis, or flattening of the femoral condyle; C = abnormal, showing a joint space of between 2 mm and 4 mm in height or up to 50% of joint-space narrowing; D = severely abnormal, showing a joint space with less than 2 mm or less than 50% of joint space remaining.[25] All radiologic evaluations were performed with the picture archiving and communication system (PACS) (Centricity® PACS viewer; GE Healthcare Co., Piscataway, NJ, USA).
Functional Evaluation
Functional evaluation of the patients was performed with Lysholm score[26] and Knee Injury and Osteoarthritis Outcome score (KOOS), including its subscales of KOOS Pain, KOOS Symptoms, KOOS Activities of Daily Living, KOOS Sports/Rec, and KOOS Quality of Life,[27] while clinical outcome of the patellofemoral joint was measured with the Shelbourne and Trumper questionnaire.[28] An independent examiner who was blinded to group assignment assessed the all of the above-mentioned clinical evaluations that were performed both preoperatively and at the time of the last follow-up.
Arthroscopic Examination of the Articular Cartilage Status
Changes in patellofemoral articular cartilage between initial and second-look arthroscopy were compared. Grades of the cartilage status were evaluated for both the patellar and femoral trochlea surfaces with ICRS grade, (Fig. 3-A, B). Grading results were first determined by a single operator, with the finding subsequently needing to be concurred with by two trained fellows. When the cartilage showed improvement with partial or full recovery at the defect area or if there was no interval change between the findings of initial and second-look arthroscopy, it was categorized into the non-progressed group. Cases with any evidence of progressive change in either the patellar or femoral trochlear cartilage were categorized in the progressed group.
Statistical Analysis
One-way analysis of variance (ANOVA) was used for normally distributed continuous variables. If the result of the ANOVA test had statistical significance, subsequent post-hoc analysis (Bonferroni's test) was performed. The Kruskal–Wallis test was first applied for non-normal distribution and the Mann–Whitney U test was then carried out as the post-hoc analysis for any result that had statistical significance. The chi-square test or Fisher's exact test was employed to compare categorical variables. Logistic regression analysis was used to determine any effect of independent variables such as age, sex, body mass index, preoperative hip-knee-ankle angle, postoperative modified Blackburn-Peel ratio, and correction angle on the cartilage status progression of patellofemoral joint. The intraclass correlation coefficient (ICC) was calculated to evaluate both intraobserver and interobserver reliabilities of the measured values. Statistical analysis was done using SPSS software (version 23.0; IBM Co., Armonk, NY, USA). A p-value of less than 0.05 was set to be statistically significant. The statistical power was calculated with the use of PASS software (version 11, NCSS, LLC, Kaysville, Utah, USA).