400 – Table 1. Longitudinal association between baseline synovitis and effusion and cartilage status at 30-months follow-up Synovitis and effusion at baseline (514 knees) Cartilage status at follow-up Crude OR (95% CI) Adjusted OR* (95% CI) No cartilage loss (n=377) Cartilage loss (n=317) Synovitis absence (grades 0 and 1) (N=467) 348 (74.5%) 119 (25.5%) 1.0 (reference) 1.0 (reference) Synovitis presence (maximum grade ≥2) (N=47) 29 (61.7%) 18 (38.3%) 1.4 (0.7–2.7) 1.0 (0.5–2.1) p=0.89 Effusion absence ((grades 0 and 1) (N=461) 352 (76.4%) 109 (23.6%) 1.0 (reference) 1.0 (reference) Effusion presence (grade ≥2) (N=53) 25 (47.2%) 28 (52.8%) 3.4 (1.9–6.2) 2.7 (1.4–5.1) p=0.002 *Results adjusted for baseline effusion, synovitis, patellofemoral cartilage damage, meniscus damage, meniscus extrusion, body mass index, age, gender, malalignment, bone marrow lesions. the reference. Logistic regression was used to estimate the risk of cartilage loss at follow-up. Cartilage loss was defined as an increase of at least a 0.5 grade (subtle within-grade progression, that did not fulfill the criteria of a full-grade change) in any subregion. Adjustment was performed for possible confounders of future tibiofemoral cartilage damage, i.e. baseline effusion in the synovitis model, baseline synovitis in effusion model, patellofemoral cartilage damage, meniscus damage, meniscal extrusion, body mass index (BMI), age, gender, malalignment, bone marrow lesions. Results: 514 knees (1 knee per patient) were included (55.6% women, mean age 60.1±7.2, mean BMI 29.1±4.5). 47 (9.1%) knees showed synovitis, and 53 (10.3%) presented with joint effusion at baseline, 137 (26.7%) knees exhibited cartilage loss during follow-up. After adjustment, baseline synovitis was not associated with an increased risk of cartilage loss at followup (adjusted odds ratio 1.0 [95% confidence intervals 0.5-2.1, p=0.89]). Knees with baseline effusion had an increased risk for cartilage loss (adjusted odds ratio 2.7 [95% confidence intervals 1.4-5.1, p=0.002]). Conclusions: Baseline synovitis in knees without radiographic OA, as assessed on non-enhanced MRI, does not predict cartilage loss, but joint effusion. However, assessment of baseline synovitis on contrast-enhanced MRI might yield different results. Baseline effusion, which reflects synovial activation, predicts structural progression in subjects without radiographic OA.