Abstract
BACKGROUND AND HYPOTHESIS: Sarcoidosis (SAR) and lymph-node tuberculosis (LNTB) are granulomatous diseases that present diagnostic challenges, especially in TB-endemic regions. We hypothesized that serum-metabolic profiles would help in differentiating SARs from LNTBs. OBJECTIVE: This study aimed to identify serum metabolic biomarkers to distinguish SAR from LNTB using NMR-based metabolomics analysis. METHODS: Serum samples were collected from 26 SAR and 22 LNTB patients. The serum metabolic profiles were measured using 800 MHz NMR spectroscopy and quantified using the commercial software CHENOMX. The serum metabolic profiles were compared using multivariate partial least squares discriminant analysis (PLS-DA), and potential discriminatory metabolites were identified using variable importance in projection (VIP) scores and subsequently evaluated for statistical significance using a volcano plot. The diagnostic potential of the discriminatory metabolites was evaluated using receiver operating characteristic (ROC) curve analysis. RESULTS: PLS-DA demonstrated significant metabolic disparity between the SAR and LNTB groups. The key metabolic features identified included elevated levels of glutamate, pyroglutamate, acetate, and leucine and a decreased glutamate-to-glutamine ratio (EQR) and decreased levels of glutamine, pyruvate, and myo-inositol in TB patients. These metabolic changes suggest that TB-infection involves activated glutaminolysis and elevated host lipid metabolism. ROC curve analysis revealed several metabolites with high diagnostic potential (AUC > 0.8), including glutamate, pyroglutamate, and glutamine (AUC > 0.98). CONCLUSION: In conclusion, this study underscores the potential of serum metabolic profiling as a noninvasive tool for distinguishing SARs from LNTBs. However, further studies are imperative to validate these findings on independent patient cohorts and to facilitate their integration into routine clinical practice.