This retrospective review was approved by the ethics committee of the Peking University People’s Hospital.
Participants
We retrospectively included 307 knees from 266 patients who underwent TKA consecutively between January 2019 and August 2021 at Peking University People’s Hospital. The diagnoses were osteoarthritis for 297 knees and rheumatoid arthritis for 10 knees. Among them, 254 (82.7%) were women. The average age at the time of TKA was 67±7 years (range, 33–90 years). All knees were preoperatively planned to use a CR knee prosthesis and were implanted into the Biomet Vanguard Complete Knee System by one surgeon (DZ).
Surgical technique
All TKAs were performed under tourniquet control using a subvastus approach through a midline skin incision. Bone cuts were made using a measured resection technique. The distal femoral cut was made at 6° valgus angulation using an intramedullary guide. Approximately 7 mm of bone was removed from the distal femur. An increased amount of bone resection was applied when there was a high flexion contracture (>30°). Rotation of the femoral component was determined with reference to the transepicondylar axis. The size of the femoral implant was determined using the anterior-referencing guide. The proximal tibial cut was made using an extramedullary guide perpendicular to the long axis of the tibia. The tibial posterior slope was usually set to 5°. Medial or lateral soft tissue contracture was manually evaluated carefully and released as needed. The PCL was carefully recessed or the posterior slope of the tibia was increased when flexion tightness was indicated by lift-off or paradoxical rolling forward. In cases of mismatched flexion-extension or mediolateral gaps after these efforts, the PCL was sacrificed, and an anterior-stabilized (AS) tibial bearing was used. PCL was recessed in 89 (29.0%) knees. The Selective patella resurfacing was performed. Components were cemented for all knees.
Measurement methods
We retrospectively obtained age, sex, body mass index (BMI), preoperative diagnosis, size of the tibial and femoral components, and thickness of the polyethylene insert from patients’ medical records. The sizes of the femoral components used in our study were 55, 57.5, 60, 62.5, 65, 67.5, 70, and 72.5. The sizes of the tibial components used in our study were 63, 67, 71, 75, 79, and 83. The medial-lateral lengths of the femoral components for sizes 55, 57.5, 60, 62.5, 65, 67.5, 70 and 72.5 were 59, 61, 64, 66, 68, 71, 73, and 75 mm, respectively. The thicknesses of the polyethylene insert used in our study were 10, 12, and 14 mm.
We defined the medial-lateral width of epicondyle (MLW) as the length of femoral epicondylar axis on the antero-posterior standing view of knee. We measured the medial posterior condyle height (MPCH), the lateral posterior condyle height (LPCH), and the insall-salvati index using lateral view of knees, and mechanical axis using full limb radiograph (Figure 1). These measurements were performed by an orthopedic surgeon (YW) masked from the patients’ information using a picture archiving and communication system. The normality of all radiographs was checked.
Statistical analysis
Age was classified into 5 categories (<60, 60-65, 65-70, 70-75, and >75). BMI was classified into <=18.5, 18.5-<24.0, 24.0-<30.0, and >=30.0 kg/m2. The lower extremity mechanical axis was classified into severe varus deformity (>15° varus), mild varus deformity (varus between 5° and 15°), neutral position (within 5° varus or valgus), mild valgus deformity (valgus between 5° and 15°), and severe valgus deformity (>15° valgus). The insall-salvati index was classified into 4 quartiles. The thicknesses of the polyethylene insert used in our study were divided into two groups (10 mm and >10 mm). We calculated the difference between the MLW and the medial-lateral lengths of the femoral components to reflect the femoral coverage.
The outcome in the current study was whether the PCL was retained successfully. Patients were thus divided into CR and AS group. We compared continuous variables, such as MLW, LPCH, the ratio of MLW and MPCH, the ratio of MLW and LPCH, and the difference between the MLW and the medial-lateral width of the femoral components between the CR and AS groups using Student’s t test. We fitted the chi-squared test to compare categorical variables, such as age, sex, preoperative diagnosis, lower-extremity mechanical axis, the insall-salvati index, the size of the tibial and femoral components, the thickness of the polyethylene insert, and the resurfacing patella, between the CR and AS groups. Univariate analysis was performed to examine the relationships of risk factors to the successful retention of PCL. Then, a Poisson regression with a robust variance estimate was performed to further examine the association between preoperative clinical and radiograph factors and the successful preservation of the PCL.[9] We also performed tests for linear trends by entering the median value of each category of variables of interest as a continuous variable in the models[10]. A two-sided P value of 0.05 was considered statistically significant. All data were analyzed by IBM® SPSS® Statistics (Version 22.0) software.