Background/Purpose: 3TR (taxonomy, treatment, targets and remission) aims to provide insights into the mechanisms of response and non-response to treatment in systemic autoimmune diseases, including SLE. The lupus arm of 3TR focuses on identifying biomarkers of disease activity. Volatile organic compounds (VOCs) can be generated by metabolic processes in the body. VOCs can diffuse from their point of origin into the blood and be emitted through breath, providing a potential non-invasive method to assess whole-body metabolism and pathophysiological associations.

Methods: Thirty SLE patients and 29 age- and sex-matched healthy controls were recruited to a single-site, case-control observational study. Breath VOC sampling was performed using the ReCIVA® breath sampler, linked to a clean air supply (CASPER®). Samples were collected on sorbent tubes and analyzed by thermal desorption gas chromatography-mass spectrometry (TD-GC-MS) by Owlstone Medical. VOCs were chemically identified in alignment with the Metabolomics Standards Initiative (MSI) criteria, with blank air samples analyzed to discern VOCs genuinely present in patients’ breath. Associations with disease activity (SLEDAI-2K, BILAG-2004, PGA, PhGA), LLDAS, DORIS remission, and fatigue measures were assessed using robust covariate-adjusted models.

Results: Demographic and clinical characteristics of patients and controls are summarized in the Table. Among 1,433 detected VOCs, 539 exceeded background levels. Focusing on high-confidence annotations (MSI tier 1 or 2), several significant correlations emerged with clinical parameters, as depicted in Figure 2. Based on compound correlations, five distinct VOC clusters were identified: 1) The anti-inflammatory compound methyl acetate stood apart as the only compound negatively associated with disease activity indices and failure to achieve LLDAS or DORIS remission; consistently, its levels in healthy controls were higher and comparable to those in non-active SLE patients (figure 3); 2) A cluster of four VOCs containing by-products of lipid peroxidation showed a coherent pattern of association with disease activity, lack of remission, and fatigue; 3) A cluster, composed of nitrogen-containing heterocycles of exogenous origin and suggestive of gut barrier disruption, showed moderate associations with disease activity; 4) A cluster, whose main representative a member of the Cyclopentenediones family with pro-inflammatory prostaglandin properties, related with disease activity and fatigue measures; 5) A small heterogeneous group with weak associations with SLEDAI-2K and LLDAS.

Conclusion: These data demonstrate the potential of breath-based VOC analysis in detecting non-invasively pathophysiological changes in SLE. Our findings highlight the intricate role of unresolved inflammation, oxidative stress, and gut barrier dysfunction in driving disease activity, as well as in influencing the patient’s perception of fatigue and overall disease severity. Future work should focus on validating these markers and exploring their associations with additional inflammatory indicators as well as assessing their potential for monitoring changes, therapy response, and prognostication.

Demographic and clinical characteristics

Figure 1 – Strength of associations with selected compounds (robust Cohen’s d)

Figure 2 – Concentrations of methyl acetate in HC and in relation to LLDAS status

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/breathomics-in-systemic-lupus-erythematosus-uncovering-non-invasive-markers-of-disease-activity/