Urinary metabolomic biomarker candidates for skeletal muscle wasting in patients with rheumatoid arthritis.

BACKGROUND: Rheumatoid arthritis (RA) is an autoimmune disease that affects the joints, leading to chronic synovial inflammation and local tissue destruction. Extra-articular manifestations may also occur, such as changes in body composition. Skeletal muscle wasting is often observed in patients with RA, but methods for assessing loss of muscle mass are expensive and not widely available. Metabolomic analysis has shown great potential for identifying changes in the metabolite profile of patients with autoimmune diseases. In this setting, urine metabolomic profiling in patients with RA may be a useful tool to identify skeletal muscle wasting. METHODS: Patients aged 40-70 years with RA have been recruited according to the 2010 ACR/EULAR classification criteria. Further, the Disease Activity Score in 28 joints using the C-reactive protein level (DAS28-CRP) determined the disease activity. The muscle mass was measured by Dual X-ray absorptiometry (DXA) to generate the appendicular lean mass index (ALMI) by summing the lean mass measurements for both arms and legs and dividing them by height squared (kg/height2 ). Finally, urine metabolomic analysis by 1 H nuclear magnetic resonance (1 H-NMR) spectroscopy was performed and the metabolomics data set analysed using the BAYESIL and MetaboAnalyst software packages. Principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) were applied to the 1 H-NMR data, followed by Spearman's correlation analysis. The combined receiver operating characteristic curve (ROC) was calculated, as well as the logistic regression analyses to establish a diagnostic model. The significance level at P < 0.05 was set for all analyses. RESULTS: The total set of subjects investigated included 90 patients with RA. Most patients were women (86.7%), with a mean age of 56.5 ± 7.3 years old and a median DAS28-CRP of 3.0 (IQR 1.0-3.0). Fifteen metabolites were identified in the urine samples with high variable importance in projection (VIP scores) by MetaboAnalyst. Of these, dimethylglycine (r = 0.205; P = 0.053), oxoisovalerate (r = -0.203; P = 0.055), and isobutyric acid (r = -0.249; P = 0.018) were significantly correlated with ALMI. Based on the low muscle mass (ALMI ≤6.0 kg/m2 for women and ≤8.1 kg/m2 for men) a diagnostic model have been established with dimethylglycine (area under the curve [AUC] = 0.65), oxoisovalerate (AUC = 0.49), and isobutyric acid (AUC = 0.83) with significant sensitivity and specificity. CONCLUSIONS: Isobutyric acid, oxoisovalerate, and dimethylglycine from urine samples were associated with low skeletal muscle mass in patients with RA. These findings suggest that this group of metabolites may be further tested as biomarkers for identification of skeletal muscle wasting.

Metabolomic Biomarker Candidates for Skeletal Muscle Loss in the Collagen-Induced Arthritis (CIA) Model.

There is no consensus for diagnosis or treatment of RA muscle loss. We aimed to investigate metabolites in arthritic mice urine as biomarkers of muscle loss. DBA1/J mice comprised collagen-induced arthritis (CIA) and control (CO) groups. Urine samples were collected at 0, 18, 35, 45, 55, and 65 days of disease and subjected to nuclear magnetic resonance spectroscopy. Metabolites were identified using Chenomx and Birmingham Metabolite libraries. The statistical model used principal component analysis, partial least-squares discriminant analysis, and partial least-squares regression analysis. Linear regression and Fisher's exact test via the MetaboAnalyst website were performed (VIP-score). Nearly 100 identified metabolites had CIA vs. CO and disease time-dependent differences (p < 0.05). Twenty-eight metabolites were muscle-associated: carnosine (VIPs 2.8 × 102) and succinyl acetone (VIPs 1.0 × 10) showed high importance in CIA vs. CO models at day 65; CIA pair analysis showed histidine (VIPs 1.2 × 102) days 55 vs. 65, histamine (VIPs 1.1 × 102) days 55 vs. 65, and L-methionine (VIPs 1.1 × 102) days 0 vs. 18. Carnosine was fatigue- (0.039) related, creatine was food intake- (-0.177) and body weight- (-0.039) related, and both metabolites were clinical score- (0.093; 0.050) and paw edema- (0.125; 0.026) related. Therefore, muscle metabolic alterations were detected in arthritic mice urine, enabling further validation in RA patient's urine, targeting prognosis, diagnosis, and monitoring of RA-mediated muscle loss.