This study identified the incidence and potential risk factors of tibial component subsidence in cementless unicompartmental knee arthroplasty (UKA) and evaluated its clinical impact. This retrospective cohort study analyzed 123 knees that underwent cementless Oxford mobile-bearing UKA. Anteroposterior and lateral radiographs were obtained preoperatively, immediately postoperatively, and at the first outpatient follow-up (3–6 months). Valgus subsidence was defined as a valgus change of > 2° in the tibial component angle between the immediate postoperative and first follow-up radiographs. The patients were categorized according to the presence or absence of subsidence. The alignment, component angles, and femoral component position relative to the tibial component in the coronal plane were compared between groups. Clinical outcomes were assessed using the Oxford Knee Score (OKS) at 3 months, 1 year, and 2 years. Tibial component subsidence incidence was 4.8%. The immediate postoperative tibial component alignment was significantly less varus in the subsidence group than in the non-subsidence group (1.0 ± 1.9° vs. 3.3 ± 2.2°, p = 0.019). The femoral component was positioned significantly more laterally than the tibial component in the subsidence group (p = 0.009). The non-subsidence group was older and had a greater postoperative posterior tibial slope than the subsidence group; however, these differences were not statistically significant (p = 0.069 and 0.052, respectively). Clinical outcomes assessed by OKS were not significantly different between groups at any time point. No revision surgery owing to subsidence was required within 2 years. Subsidence occurred in 4.8% of cases and was associated with a relative lateral placement of the femoral component and less varus of tibial component alignment. A tibial cut with an appropriate varus angle may be protective. Given the small number of events, these findings should be interpreted as exploratory and inform future prospective studies.
Kinoshita et al. (Mon,) studied this question.