Pre-operative predictive factors of residual varus on the mechanical axis after Oxford unicompartmental knee arthroplasty

BACKGROUND: Residual varus after Oxford unicompartmental knee arthroplasty (UKA) happens frequently. This study aims to evaluate the pre-operative contributing factors of residual varus. METHODS: A total of 1,002 knees (880 patients, 201 patients were male, and 679 were female) underwent Oxford UKA in the Orthopedic Surgery Department of the Beijing Jishuitan Hospital from March 2018 to April 2021. The mean age of the patient was 64.7 ± 7.7 years. To assess residual varus, the full-length lower extremity is placed upright for EOS imaging, with the knee fully extended. The angle of post-operative residual varus was measured as described by Noyes et al. Of the knees studied, they were either categorized into an under-corrected group (post-operative Noyes angle >5°) or a corrected group (post-operative Noyes angle ≤5°). Age, gender, body mass index (BMI), range of motion (ROM), Clinical American Knee Society Score (Clinical AKSS), and Function American Knee Society Score (Function AKSS) were compared. The following additional parameters were measured: pre-operative Noyes angle, lateral distal femoral angle (LDFA), medial proximal tibial angle (MPTA), the posterior slope of the proximal tibia angle (PPTA), joint line converge angle (JLCA), and fixed flexion deformity (FFD). RESULTS: There was no statistically significant difference between the two groups in regards to gender (p = 0.428), surgical leg (p = 0.937), age (p = 0.851), BMI (p = 0.064), pre-operative Clinical AKSS (p = 0.206) and Function AKSS (p = 0.100). However, pre-operative ROM statistically differed between the two groups (p < 0.001). The contributing factors of post-operative residual varus were determined to be the following parameters: pre-operative MPTA (p < 0.001, OR = 4.522, 95% CI: 2.927–6.984), pre-operative Noyes (p < 0.001, OR = 3.262, 95% CI: 1.802–5.907) and pre-operative FFD (p = 0.007, OR = 1.862, 95% CI: 1.182–2.934). The effects of pre-operative LDFA (p = 0.146), JLCA (p = 0.942), and pre-operative PPTA (p = 0.899) on the post-operative mechanical axis did not show statistical significance. CONCLUSIONS: Patients with severe pre-operative varus, particularly varus deformity mainly from the tibial side or pre-operative FFD, are more prone to get extremity mechanical axis residual varus after UKA with Oxford.