ARTERIAL REMODELING ON KNEE MRI AND CARTILAGE T2 RELAXATION TIME: DATA FROM THE OSTEOARTHRITIS INITIATIVE.

INTRODUCTION: Popliteal artery remodeling may contribute to knee OA through altered periarticular and subchondral perfusion. Cartilage T2 relaxation time is a quantitative MRI biomarker of cartilage composition, reflecting collagen organization and water content. Whether automated MRI-derived popliteal artery parameters are associated with cartilage T2 relaxation time and short-term T2 change remains unclear. OBJECTIVE: To evaluate cross-sectional and longitudinal associations between baseline MRI-derived popliteal artery parameters and cartilage T2 relaxation time in OAI participants. METHODS: This retrospective longitudinal analysis included 4,447 OAI participants with baseline right-knee cartilage T2 measurements; 3,854 and 3,560 participants had 12- and 24-month follow-up T2 data, respectively. Popliteal artery lumen and wall contours were derived from 3T sagittal DESS MRI using a fully automated segmentation pipeline. Arterial parameters included mean wall thickness, mean wall area, mean lumen area, mean total artery area, eccentricity ratio (maximum-to-minimum wall thickness ratio), and normalized wall index (NWI; wall area divided by total vessel area). Cartilage T2 values were obtained from automated segmentation of multi-echo spin-echo MRI in the lateral tibia, lateral femur, medial tibia, medial femur, and patella. Adjusted linear regression models tested associations between standardized arterial parameters and baseline T2 and 12-/24-month T2 change. Models adjusted for age, sex, BMI, smoking, diabetes, PASE, and comorbidity score; longitudinal models additionally adjusted for baseline T2. Analyses were stratified by KL grade <2 versus ≥2. Benjamini-Hochberg FDR-adjusted p<0.05 was considered significant. RESULTS: Mean age was 61.1 ± 9.1 years, 2,612/4,447 (58.7%) participants were female, and mean BMI was 28.5 ± 4.8 kg/m2. Mean wall area, mean lumen area, and mean total artery area were consistently associated with higher baseline cartilage T2 values across all subregions (β=0.25-0.45 per SD; all p<0.001). Mean NWI was inversely associated with baseline T2 across all subregions (β=-0.15 to -0.33 per SD; all p<0.001), with the strongest association in the patella (β=-0.33, 95% CI=-0.45 to -0.22; p<0.001). Mean wall thickness was associated only with lateral tibial (β=0.13, 95% CI=0.04 to 0.23; p=0.006) and lateral femoral T2 (β=0.11, 95% CI=0.01 to 0.21; p=0.039). Baseline arterial parameters were not associated with 12- or 24-month T2 change after baseline T2 adjustment. Effect sizes were generally larger in KL≥2. CONCLUSION: Automated knee MRI-derived popliteal artery size and remodeling parameters were associated with baseline cartilage T2 relaxation time, particularly in knees with radiographic OA, but did not predict short-term T2 change.