Circulating sTLR4 and sTREM-1 in Knee Osteoarthritis: Associations with Study-Specific MRI-Defined Categories and Disease Burden

Background/Objectives: Osteoarthritis (OA) is biologically heterogeneous, and the relationship between MRI-detected inflammation and circulating innate immune activity remains uncertain. This study aimed to determine whether serum soluble Toll-like receptor 4 (sTLR4) and soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) distinguish between the study-specific MRI-defined INF-OA and non-INF-OA categories, and whether these biomarkers are associated with OA disease burden. Methods: This prospective, single-center, cross-sectional study included 30 patients with knee OA and 30 asymptomatic healthy controls. Patients with active OA flare-up, defined by a Knee Osteoarthritis Flare-Ups Score (KOFUS) ≥ 7, were excluded before MRI assessment and category assignment. Patients with OA were assigned to study-specific MRI-defined categories using MRI Osteoarthritis Knee Score (MOAKS) features: INF-OA (n = 16) required effusion-synovitis and/or Hoffa-synovitis-related signal alteration graded ≥ 1, whereas non-INF-OA (n = 14) required grade 0 for both features. Biomarkers were measured by enzyme-linked immunosorbent assay. Primary within-OA biomarker comparisons were Holm-corrected, and adjusted models included age, sex, and body mass index (BMI). Results: Serum sTLR4 and sTREM-1 did not differ between INF-OA and non-INF-OA after Holm correction (adjusted p = 0.710 for both). In age-, sex-, and BMI-adjusted models, the INF-OA versus non-INF-OA geometric mean ratios were 1.15 (95% CI 0.92–1.44; p = 0.219) for sTLR4 and 1.09 (95% CI 0.89–1.35; p = 0.398) for sTREM-1. Given the modest within-OA subgroup sizes, smaller or moderate between-category differences cannot be excluded. After covariate adjustment, the overall OA cohort had higher geometric mean concentrations than controls for sTLR4 (ratio 1.85, 95% confidence interval 1.56–2.20; p < 0.001) and sTREM-1 (ratio 1.67, 95% confidence interval 1.43–1.97; p < 0.001). None of the systemic inflammatory measurements significantly distinguished INF-OA from non-INF-OA after correction for multiple comparisons. Within the OA cohort, each 1 ng/mL increase in sTLR4 was associated with a 4.99-point higher Western Ontario and McMaster Universities Osteoarthritis Index total score after adjustment for age and body mass index (95% confidence interval 2.28–7.70; p < 0.001). Conclusions: Circulating sTLR4 and sTREM-1 concentrations were higher in patients with OA than in asymptomatic healthy controls but did not distinguish the study-specific INF-OA and non-INF-OA categories in this sample. These findings should not be interpreted as evidence of biological equivalence between the MRI-defined categories. sTLR4 warrants further evaluation as a circulating correlate of current OA burden in larger longitudinal and multimodal studies.