INTRODUCTION: The prevalence of lumbar spinal stenosis (LSS) and lower extremity osteoarthritis (OA) is increasing globally due to population aging.1 Comorbid hip and knee OA have been reported in 2-35% and 5-41% of patients undergoing LSS surgery, respectively, highlighting their significance in surgical planning.2 However, evidence regarding the impact of concomitant OA on postoperative outcomes in LSS patients remains limited. This study aimed to evaluate the impact of severe hip or knee OA on patient-reported outcome measures (PROMs) and global alignment in LSS patients undergoing surgery.
METHODS: This retrospective study included 121 patients who underwent decompression with or without interbody fusion for LSS (≤3 levels) with a minimum 1-year follow-up. Upright whole-body EOS imaging was performed preoperatively and 1-year postoperatively. Two independent reviewers graded hip and knee OA severity using the Kellgren-Lawrence (KL) classification. Patients with KL grades 3-4 were classified as the severe OA group and those with KL grades 0–2 as the non-severe OA group. PROMs included (1) Visual Analog Scale (VAS) for low back pain, leg pain, and leg numbness; (2) Japanese Orthopedic Association Back Pain Evaluation Questionnaire (JOABPEQ) scores for health-related quality of life (HRQOL); and (3) Oswestry Disability Index (ODI) for disability. Demographics, radiographic parameters, and PROMs were compared between groups using the chi-square or Mann-Whitney U tests. Multivariable regression analysis was performed to assess the effect of severe OA on 1-year PROMs, adjusting for age, sex, ASA classification, fusion, surgical levels, sagittal vertical axis (SVA), and severe OA.
RESULTS: Of 121 patients (mean age 69.4 years, 52 females), 67 patients had severe hip or knee OA. Patients with severe OA were older, had higher ASA classifications, and underwent more multilevel surgeries. The incidence of implant-related complications was higher in the severe OA group (10.5% vs. 1.9%, P = 0.043). Radiographically, the severe OA group had a greater preoperative SVA (69.5 ± 54.7 mm vs. 45.4 ± 37.5 mm, P = 0.008), which persisted at 1-year (60.7 ± 54.4 mm vs. 37.1 ± 34.1 mm, P = 0.036). At 1 year, the severe OA group had greater C7-central sacral vertical line offset (12.8 ± 12.6 mm vs. 7.0 ± 5.1 mm, P = 0.039). Multivariable regression showed that severe OA was independently associated with higher VAS scores for low back pain (β = 0.234, P = 0.030) and leg pain (β = 0.272, P = 0.012), lower JOABPEQ scores in pain disorder (β = -0.227, P = 0.035), lumbar function (β = -0.266, P = 0.012), walking ability (β = -0.216, P = 0.019), and social life function (β = -0.249, P = 0.017), and higher ODI scores (β = 0.269, P = 0.008).
DISCUSSION: Severe hip or knee OA was present in 55% of patients undergoing LSS surgery and was independently associated with residual pain, poorer HRQOL, and greater disability at 1-year postoperatively. Despite comparable spinopelvic mismatch, patients with severe OA exhibited greater trunk anteversion and coronal imbalance postoperatively. These findings underscore the importance of addressing severe OA to optimize outcomes in LSS management.