Long-term outcomes and factors associated with revision after medial open-wedge high tibial osteotomy in 274 knees: a Kellgren–Lawrence grade-stratified analysis

Purpose We aimed to investigate long-term outcomes and factors associated with revision after medial open-wedge high tibial osteotomy (OWHTO) according to Kellgren–Lawrence (KL) grade. Methods This retrospective study included 274 knees (158 KL grade 4 [KL4] knees and 116 KL grades 2–3 [KL2–3] knees; mean age 64.7 ± 7.2 years) with a minimum follow-up of 10 years (mean follow-up, 12.5 ± 2.6 years) after OWHTO. Revision surgery (conversion to total knee arthroplasty [TKA] or unicompartmental knee arthroplasty [UKA]) was defined as the endpoint. Clinical and radiographic outcomes were evaluated. Multivariate logistic regression analyses were performed separately for KL4 and KL2–3 knees to identify factors associated with revision. Results The mean Japanese Orthopaedic Association (JOA) score improved from 60.6 preoperatively to 91.0 at plate removal, followed by gradual deterioration during long-term follow-up. Final JOA scores were significantly lower in KL4 knees than in KL2–3 knees (71.6 versus 83.1; p < 0.001). Revision surgery was performed in 43 knees (15.7%) and was significantly more frequent in KL4 than in KL2–3 knees (22.2% versus 6.9%; p < 0.01). Ten-year joint preservation rates were 90.5% in KL4 knees and 96.5% in KL2–3 knees (log-rank p = 0.006). In KL4 knees, postoperative percentage mechanical axis (%MA) at the time of plate removal was independently associated with revision risk (odds ratio [OR] 0.94, 95% confidence interval [CI] 0.91–0.98; p = 0.004). In KL2–3 knees, an exploratory multivariate analysis suggested an association between progression of joint line convergence angle during follow-up (ΔJLCA) and revision (OR 3.31, 95% CI 1.52–7.21; p = 0.003), whereas postoperative %MA showed no significant association. Conclusions Long-term outcomes after OWHTO differed according to KL grade. KL4 knees demonstrated higher revision rates and inferior joint preservation survival, and postoperative weight-bearing line position was associated with revision risk. In contrast, revision was uncommon in KL2–3 knees, and exploratory analysis suggested a possible association with longitudinal changes in joint line convergence rather than postoperative alignment. These findings suggest that factors associated with revision may differ according to disease severity. Level of evidence This was a level IV, retrospective cohort study.