Background/Purpose:

To explore the mechanisms by which an altered gut microbiota might predispose to ANCA-associated vasculitis (AAV), a comprehensive metabolic profiling of fecal and plasma bile acids, amino acids, and short chain fatty acids was performed in patients with AAV (granulomatosis with polyangiitis; microscopic polyangiitis; and eosinophilic granulomatosis with polyangiitis) and in healthy controls.

Methods:

Using cross-sectional and longitudinal designs, the fecal and plasma metabolomes of patients with newly diagnosed AAV (active and in remission) were compared to chronic AAV (active and in remission), and to healthy controls. All samples were collected using standardized methods, and analyzed by Liquid Chromatography/Mass Spectrometry for bile acids, aminoacids, and short chain fatty acids. The gut microbiome was analyzed on the same fecal samples by sequencing the bacterial 16S rRNA gene (V1-V2 region). The association between bacterial taxa and fecal metabolites was studied using logistic regression models, correcting for multiple comparisons. Bacterial taxa were tested if their mean abundance was >1%.

Results:

78 fecal samples were studied: 37 active AAV (20 new diagnosis and 17 chronic), 25 remission, and 16 controls. 32 plasma samples were studied: 20 active AAV, 6 remission, and 6 controls.

Compared to remission states, the fecal amino acids phenylalanine and tyrosine were significantly diminished in active disease. Patients with chronic AAV in remission had higher levels of plasma phenylalanine and tyrosine compared to patients with a new diagnosis of AAV in remission (p=0.02 for both).

There was no statistical difference in fecal bile acids between the disease states of AAV. Patients with a chronic active AAV had higher levels of plasma taurolithocholic bile acid compared to patients with an active new diagnosis of AAV (Figure 1A). Faecalibacterium was found to be associated with this plasma bile acid (p=0.02). Plasma glycholic and glycodeoxycholic acids were strongly associated with active disease (p<0.01 and p=0.01 for both).

Patients with chronic AAV in remission had higher levels of plasma taurochenodeoxycholic bile acid compared to patients with a new diagnosis of AAV in remission (Figure 1B). Phascolarctobacterium, Sutterella, and Ruminococcus were found to be associated with this plasma bile acid (p<0.05).

There was no statistical difference in plasma or fecal short chain fatty acids.

Conclusion:

Active AAV is associated with an altered fecal and plasma metabolome. Plasma taurochenodeoxycholic bile acid, and plasma and fecal amino acids are higher in chronic remission states compared to new diagnosis remission states, suggesting potential anti-inflammatory effects of these metabolites. Diminished metabolic diversity may be a feature of active AAV and potential biomarker to predict disease activity in AAV.