Both Systemic and Local Lipopolysaccharide (LPS) Burden is Associated with Knee Osteoarthritis (OA)
Z. Huang,T. Stabler,F. Pei,V. Kraus
DOI: https://doi.org/10.1016/j.joca.2016.01.590
IF: 7.507
2016-01-01
Osteoarthritis and Cartilage
Abstract:Purpose: The microbiome is recognized as a new frontier in medicine with connections to a variety of diseases. We hypothesized that lipopolysaccharide (LPS), a key proinflammatory product of the microbiome, plays a role in the pathogenesis of osteoarthritis (OA). We therefore aimed to evaluate the association of LPS and its active binding partner, LPS binding protein (LBP) with knee inflammation and severity of OA symptoms and structural abnormalities. Methods: LPS was measured using a recombinant factor C (rFc) assay, carefully optimized for systemic and synovial fluid analyses by spike-in and recovery using samples from patients with knee OA (n=5). LPS and LBP were then tested in both serum and synovial fluid (SF) of 25 knee OA patients with Etarfolatide imaging (for activated macrophages) data. LPS and LBP were evaluated by Spearman correlation for associations with the macrophage produced soluble CD14 (sCD14). Multivariable linear regression models with generalized estimating equations (GEEs) were used to assess the association of LPS and LBP in serum and SF with OA phenotypic (radiographic, Etarfolatide, WOMAC and self-reported pain) outcomes. Models were adjusted for age, sex and body mass index. P values less than 0.05 were considered significant. Statistical analysis was performed with SPSS version 22.0 software (SPSS, Chicago, IL, USA) and Prism GraphPad v6.0 (La Giolla, CA, USA). Results: Based on LPS spike-in and recovery, serum and SF dilutions of 0.1% were required to achieve recovery rates of at least 75% in all test specimens (Figure 1&2). Low coefficients of variation (range 4.8% to 9.6%) were achieved with both serum and SF dilutions <0.2%. LPS, LBP and sCD14 were significantly positively correlated (Spearman r=0.39 to 0.53, p=0.026 to <0.001). Multivariable linear regression models using GEE showed that serum LPS and LBP were significantly associated with the abundance of activated macrophages in the knee joint capsule and synovium and radiographic knee osteophyte severity (Table 1). SF LPS and LBP were significantly associated with the abundance of activated macrophages in the synovium. In addition, SF LPS was positively associated with osteophyte severity (p<0.001), joint space narrowing severity (p<0.001) and total WOMAC score (p=0.008), while SF LBP was positively associated with self-reported knee pain score (p=0.034) (Table 2). Conclusions: Serum and SF dilutions of 0.1%−0.2% yield stable and reliable LPS results using the rFc assay. The quantity of LPS and LBP in serum and SF reflected the abundance of activated macrophages, structural damage and pain in OA knees. These data strongly support a role for LPS in the pathogenesis and severity of OA.Tabled 1Table 1. Associations of serum LPS and LBP concentrations with OA phenotypesOA phenotypeSerum LPS (EU/mL)Serum LBP (μg/mL)β (95% CI)†p(β (95% CI)†pActivated macrophagesJoint Capsule0.019 [0.012, 0.025]<0.0010.160 [0.068, 0.252]0.001Synovium0.014 [0.006, 0.022]0.0010.118 [0.008, 0.228]0.036Osteophyte severity0.017 [0.002, 0.033]0.0300.072 [0.013, 0.131]0.017JSN severity0.006 [−0.001, 0.013]0.1030.142 [−0.027, 0.311]0.099WOMAC pain−0.002 [−0.022, 0.017]0.8160.210 [−0.022, 0.441]0.076WOMAC total−0.058 [−0.145, 0.030]0.1960.331 [−0.945, 1.607]0.611Knee pain0.000 [−0.002, 0.002]0.7850.006 [−0.012, 0.024]0.518ˆ Multivariate linear regression models using GEEs (Generalized estimating equations) were used.† Adjusted for age, gender, and body mass index; values in bold are significantly different at p<0.05.JSN=joint space narrowing Open table in a new tab Tabled 1Table 2. Associations of synovial fluid LPS and LBP concentrations with OA phenotypesOA phenotypeSynovial fluid LPS (EU/mL)Synovial fluid LBS (μg/mL)β (95% CI)†pβ (95% CI)†pActivated macrophagesJoint capsule0.009 [−0.012, 0.030]0.6830.261 [−0.145, 0.667]0.207Synovium0.028 [0.011, 0.045]0.0010.464 [0.069, 0.859]0.021Osteophyte severity0.063 [0.041, 0.084]<0.001−0.367 [−1.042, 0.308]0.286JSN severity0.014 [0.007, 0.021]<0.001−0.088 [−0.249, 0.073]0.284WOMAC pain0.019 [−0.001, 0.040]0.0630.304 [−0.210, 0.819]0.246WOMAC total0.177 [0.046, 0.309]0.0080.275 [−2.638, 3.188]0.853Knee pain0.002 [−0.001, 0.005]0.1260.108 [0.005, 0.211]0.039ˆ Multivariable linear regression models using GEEs (Generalized estimating equations) were used.† Adjusted for age, gender, and body mass index; values in bold are significantly different at p<0.05.JSN=joint space narrowing Open table in a new tab View Large Image Figure ViewerDownload Hi-res image Download (PPT) ˆ Multivariate linear regression models using GEEs (Generalized estimating equations) were used. † Adjusted for age, gender, and body mass index; values in bold are significantly different at p<0.05. JSN=joint space narrowing ˆ Multivariable linear regression models using GEEs (Generalized estimating equations) were used. † Adjusted for age, gender, and body mass index; values in bold are significantly different at p<0.05. JSN=joint space narrowing