Integrated metabolomic and cytokine profiling reveals biomarkers across the clinical spectrum of lupus nephritis

Introduction: Lupus nephritis (LN) represents one of the most frequent and severe manifestations of systemic lupus erythematosus (SLE), highlighting an urgent need for novel non-invasive approaches to improve its diagnosis and clinical monitoring. Objectives: To characterize serum and urinary metabolomic profiles, together with urinary cytokine expression, in patients with lupus nephritis (LN) across the clinical spectrum from active disease, post-induction response, and sustained remission. Methods: We enrolled 62 patients fulfilling the 2019 ACR/EULAR SLE criteria, divided into three groups: active LN (n=21), post-induction responders at 6 months (n=21), and sustained responders with ≥2 years of remission (n=20). Untargeted metabolomic profiling of serum and urine was performed using GC-MS, and urinary cytokines were measured by luminometry. Statistical analyses included Kruskal-Wallis tests and multinomial logistic regression. Results: Median age was 36 years; 87% were female. Twelve serum metabolites and ten urinary metabolites differed significantly across groups. Active LN patients showed increased urinary threonate and elevated cytokines (15 out of 17 measured, including IL-10). Serum glycerol was significantly higher in post-induction responders compared to sustained responders, while urinary sucrose was lower in this group. Logistic regression identified urinary threonate as discriminative for active LN, whereas serum glycerol (OR 1.05, 95% CI 1.048-1.050) and urinary sucrose (OR 0.98, 95% CI 0.9802-0.981) differentiated post-induction from sustained response. Increased urinary IL-10 was strongly associated with active LN, while G-CSF discriminated post-induction from sustained responders. Discussion and conclusions: The combined analysis of serum/urinary metabolomic profiles and urinary cytokines provides discriminatory power to differentiate active LN from remission states, supporting their potential role as non-invasive biomarkers for diagnosis and treatment monitoring.