Characterization of ¹H NMR Plasma Glycoproteins as a New Strategy To Identify Inflammatory Patterns in Rheumatoid Arthritis

Abstract

Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory disease associated with a high index of morbidity and mortality from cardiovascular diseases. We used ¹H NMR to characterize the plasma glycoprotein and lipoprotein profiles of a cohort of patients with RA ( n = 210) versus healthy individuals ( n = 203) to associate them with the RA disease and its severity. Using ¹H NMR, we developed a line-shape method to characterize the two peaks associated with glycoproteins (GlycA and GlycB) and its derived variables: areas of GlycB (Area GlycB) and GlycA (Area GlycA), shape factors of these two peaks (H/W = height/width), and the distance between them (Distance GlycB-GlycA). We also used the advanced lipoprotein test Liposcale (CE) to characterize the lipoprotein subclasses. The standard lipid panel and traditional inflammatory markers such as C-reactive protein, the erythrocyte sedimentation rate, fibrinogen, the rheumatoid factor, anticitrullinated peptide antibodies, and the DAS28 index have also been determined. RA patients presented a significant 10.65% increase in the GlycA associated area compared with the control group ( p = 2.21 × 10^-10). They also presented significantly higher H/W GlycA and GlycB ratios than the control population (H/W GlycB p = 7.88 × 10^-8; H/W GlycA p = 5.61 × 10^-8). The prediction model that uses the traditional inflammatory variables and the ¹H NMR-derived parameters presented an AUC that was almost 10% higher than the model that only uses the traditional inflammatory variables (from 0.7 to 0.79 AUC). We have demonstrated that GlycA and GlycB variables derived from ¹H NMR, along with classic inflammatory parameters, help to improve the classification of individuals with high RA disease activity.