Abstracts Division 3

73. The detection of Primary Sclerosing Cholangitis using an optimized methodology for Fecal VOC analysis.

R. van Vorstenbosch1, A. Mommers1, G. Stavropoulos1, K. Munster2, C. Ponsioen2, FJ van Schooten1, A. Smolinska1

1 Department Pharmacology and Toxicology, Maastricht University, Maastricht, the Netherlands
2 Department of Gastroenterology and Hepatology, Amsterdam University Medical Centres, Amsterdam, the Netherlands

Background
In Primary Sclerosing Cholangitis (PSC) inflammation of the bile ducts results in accumulation of waste products in the liver, causing liver damage and cirrhosis. Approximately 8% of Irritable Bowel Disease (IBD) patients will develop PSC, highlighting a need for non-invasive screening tests. Recently, exhaled breath could successfully distinguish IBD from PSC. In the current study, we show similar utility of fecal VOC profiles.

Methods
Fecal samples of 24 PSC patients and 49 IBD patients were sampled using optimized settings of the Micro-Chamber/Thermal Extractor (MC). Next, fecal VOC profiles were analyzed using Gas Chromatography coupled to Mass spectrometry (GC-MS). Due to high variation in the data, data normalization was unfeasible. Therefore, a log-ratio approach combined with Random Forest was applied to analyze the data. To improve predictions, data fusion of exhaled breath and fecal VOC profiles was performed using proximity stacking.

Results
A semi-targeted approach to distuinguish PSC from IBD cases resulted in an AUC ROC of 0.82 based on 8 fecal VOC markers. Proximity stacking further increased these predictions.

Conclusions
Fecal VOC profiles were shown to distuinguish PSC from IBD and offer complementary information compared to exhaled breath profiles, thereby offering unique insights into the microbiome-gut-liver axis related to PSC.

NUTRIM | School of Nutrition and Translational Research in Metabolism
NUTRIM aims to contribute to health maintenance and personalised medicine by unraveling lifestyle and disease-induced derangements in metabolism and by developing targeted nutritional, exercise and drug interventions. This is facilitated by a state of the art research infrastructure and close interaction between scientists, clinicians, master and PhD students.
www.maastrichtuniversity.nl/nutrim