Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P741 The gut microbiota in primary sclerosing cholangitis and inflammatory bowel disease: correlations with disease and diet

J. Torres*1, C. Palmela2, X. Bao3, A.P. Krieger4, V. Teixeira5, S. Velho5, C. Vieira6, A. Oliveira7, J. Pereira da Silva8, P. Moura Santos9, S. Itzkowitz1, I. Peter3, M. Cravo2, J. Hu3

1Icahn School of Medicine at Mount Sinai, Division of Gastroenterology, Department of Medicine, New York, United States, 2Hospital Beatriz Ângelo, Division of Gastroenterology, Loures, Portugal, 3Icahn School of Medicine at Mount Sinai, Department of Genetics and Genomic Sciences, New York, United States, 4University of Lisbon, Lisbon, Portugal, 5Hospital Beatriz Ângelo, Nutrition, Loures, Portugal, 6Centro Hospitalar Barreiro Montijo, Division of Gastroenterology, Montijo, Portugal, 7Hospital Fernando Fonseca, Division of Gastroenterology, Amadora, Portugal, 8Instituto Português de Oncologia, Division of Gastroenterology, Lisbon, Portugal, 9University of Lisbon, Hospital Santa Maria, Division of Gastroenterology, Lisbon, Portugal


PSC-IBD patients present a special phenotype, the reason for which remains obscure; potentially the microbiota could have a role. Our goal was to compare gut microbiota between PSC-IBD and IBD patients, integrating clinical and diet information.


In the study, 30 patients with extensive colitis (15 w/ PSC) were recruited. The patients provided stool samples, clinical information, and a food-frequency questionnaire. All patients were off antibiotics and ursodeoxycholic acid. Stool microbiome profiling using bacterial 16S rRNA sequencing was performed. Rpackage [Vegan] was used to compare the overall microbiota diversity, and the LEfSe method to find differential taxa features. Pearson correlation was used to assess the relationship between nutrients and microbiota. P-values were adjusted for false discovery rate.


Median age was 44; 16 were male. Further, 70% of patients had their IBD in clinical remission; 42% of PSC patients had an intermediate/high risk PSC Mayo score. No PSC patient had undergone liver transplant. PSC-IBD patients had a significantly lower BMI than IBD (23.8 vs 30.4 Kg/m2, p = 0.002). No overall differences in the overall microbiota diversity were found (Figure 1A and 1B). At the genus level, we found 5 genera differentially expressed in PSC-IBD vs IBD (logarithmic LDA score > 2): Ruminococcus, Fusobacterium, and uncultured genus from Christensenellaceace were more abundant in PSC-IBD, whereas Veilonella and Roseburia were less abundant (Figure 1C).

Figure 1. Beta diversity (1A); alpha diversity (1B) in PSC-IBD and IBD patients; (1C) taxa features differentially abundant in PSC-IBD vs IBD (green colour, taxa enriched; red colour, taxa reduced in PSC-IBD).

No specific genera were associated with disease severity in PSC. No significant differences were found in the daily intake of macro and micronutrients amongst groups. However, correlations between certain taxa and dietary components were disease specific. In IBD, Dialister (p = 0.00028, r = 0.93) and unknown genus from Corioabacteriaceae correlated with complex carbohydrate intake (p = 0.00062, r = 0.92), and Veillonella correlated with saturated fatty acids (eicosanoic acid, p = 0.044, r = 0.84); (docosanoic acid, p = 0.043, r = 0.84). In PSC-IBD, SMB53 genus was correlated with alcohol intake (p = 0.012, r = 0.87); Bifidobacterium with omega-6 (p = 0.044, r = 0.84); f_Enterobactericae with boron (p = 0.0025, r = 0.90); and Parabacteroides with manganese (p = 0.0071, r = 0.88); and selenium intake (p = 0.0067, r = 0.88).


Specific microbiota features were found in PSC-IBD compared with IBD alone. No genera could be associated with PSC disease severity. Correlation analyses established unique links to dietary components. Those connections suggest the possibility of using dietary interventions to modulate disease-associated microbiota.