Methods

We retrospectively reviewed 107 knees in 65 patients with a minimum of six months of follow-up. The 107 knees were operated by two surgeons using BS4+ (Bisurface 4 plus) implant (Japan Medical Materials, Japan); the femoral component was cemented, and the tibial component was either cemented or not cemented by using four screws. All the replacements were performed under same operative procedure with medial para-patellar approach and measured bone technique. The knees were classified into two groups (: I and U) on the basis of postoperative radiological findings that indicate the loosening of tibial components. First, there were not any apparent loosening symptom like radio-lucent lines nor sinking; group-I (intact, n=75). Second, there were some radio-lucent lines around tibial component; subgroup-R (radio-lucent lines, n=25), or some subsidence of component over 2 mm; subgroup-S (subsidence, n=7), and the latter two subgroups were put into group-U (unstable, n=32) all together. We measured preoperative and postoperative alignment (overall FTA, correction of malalignment, and alignment of the tibial and the femoral component in the coronal plane). Furthermore, each thickness of tibial osteotomy was measured with use of preoperative and postoperative radiographs of the knee. These parameters including patient's BMI were compared between two major groups statistically to evaluate the factor influencing the stability of tibial components. Moreover, the thickness of tibial osteotomy were compared between two subgroups.