Meaningful and reliable predictors of disease progression in PSC are currently lacking despite efforts over the past three decades. Patients with PSC often demonstrate increased plasma bile acids (BA) concentrations due to ongoing cholestasis, a hallmark of the disease. However, the relationship between composition of the BA in blood and future development of advanced liver disease among patients with PSC remains unclear. In fact, an assessment of plasma BA as predictors of clinical progression in PSC has not been demonstrated. In our present study presented by Omar Mouse MD, et al., at the 2019 Annual conference of the American Associations for the Study of Liver Diseases (AASLD) we use machine learning, an artificial intelligence application, to establish a non-invasive, reliable BA-based model, that is predictive of events of liver disease advancement. More than 400 PSC patients participated from across the US and 108 patients with PSC from Norway were included to make this study possible. Our BA-based model could be useful in PSC prognostication and stratification into clinical trials.
Read the full abstracts from the PSC team