1841: Differential metabolomic signatures along lupus nephritis spectrum

Background/Purpose: Lupus nephritis (LN) is a common feature in systemic lupus erythematosus (SLE) affecting up to 50% of the individuals with this disease. The gold standard for diagnosis is renal biopsy, nonetheless this procedure can entail complications due to its invasive nature. Metabolomic profile seems to be a promising alternative as a noninvasive diagnostic tool, as it has demonstrated to be useful distinguishing between active LN, SLE without renal involvement and healthy controls. However, scarce information exists about metabolomic profile differences in LN active patients, patients achieving renal response and patients with sustained response. The aim of this study is to characterize the metabolomic profile of three distinct groups of patients representing a broad spectrum of LN, from activity to sustained response.

Methods: Adult patients that met the 2019 ACR/EULAR criteria for SLE were eligible for inclusion and assigned into groups according to LN activity and treatment response. Group 1 included patients with active LN diagnosed by renal biopsy and/or 24-hour urinary protein-creatinine ratio greater than 0.5, active urinary sediment and creatinine increase of 0.3 mg/dl from baseline. Group 2 encompassed patients with either complete or partial renal response after 6 months of starting induction therapy. Group 3 encompassed patients with sustained response defined as a period greater than 2 years since achieving complete response to induction therapy without relapse. Urine metabolomic analysis was performed using a non-targeted method based on liquid chromatography tandem mass spectrometry.

Results: Urine metabolomic profile was obtained from 21, 21 and 20 patients in groups 1, 2 and 3 respectively. The population included adults with a median age of 36 years out of which 90% were female. Among 37 identified metabolites, 10 differential metabolites were detected, of which 7 were increased in group 1. Most of these correspond to sugar acids such as threonic acid (median concentration of 125 in group 1 vs. 89 and 82 in group 2 and 3 respectively, p < 0.001), as reported in previous literature reports, has been shown to differentiate between SLE with renal involvement from non-renal involvement. Amino acids were also increased in group 1, such as vanillylmandelic acid (median concentration 5.4 vs 4.6 and 4.1 group 2 and 3 respectively, p= 0.010) and norvaline (5.5 vs 4.8 and 4.4, group 2 and 3 respectively, p= 0.025), which differs from previous data. Citrate was diminished in active LN, but increased in group 2 and 3 patients (9 vs. 27 and 53 respectively, p=0.008). This observation has been related to a dampened oxidative phosphorylation in activated immune cells which normalize after treatment.

Conclusion: Urine metabolomic profile differentiates active LN patients from individuals that have achieved and maintained response to treatment and may serve as a noninvasive diagnostic tool for LN. Active LN is characterized by increased sugar acids and lower citrate levels. These findings give insight into the metabolomic shifts underlying renal activity in SLE and may resolve once response to therapy has been achieved.