ZusammenfassungIdentification of subjects with or without accompanying risk factors for metabolic diseases and, accordingly, tracking of patients with metabolic-linked diseases in order to identify early markers of underlying pathomechanisms has gained increasing interest during the past decade. Large scale epidemiological studies e.g. in the United States, United Kingdom and Germany may be used for the development of novel risk scores. Magnetic resonance imaging (MRI) offers excellent reliable tools for visualization of visceral adipose tissue (VAT) and intrahepatic lipids (IHL) by state-of-the-art imaging techniques enabling volumetric quantification of VAT and detailed assessment of (even low amounts) IHL. As MRI does not go along with a radiation exposure to the subjects, this is the “method of choice” for non-invasive phenotyping in the general population. Applying 2-point and 6-point DIXON techniques in MRI as available in the NAKO, it is possible to quantify local fat such as visceral adipose tissue (VAT) depots volumetrically as well as determine the degree of organ steatosis, particular of the liver (IHL). Different segmentation methods have been developed and tested in several cohort, predominantly using 3D Statistical Shape Models and Neural Networks (machine learning). We aim to adapt existing technologies to build reliable MRI segmentation pipeline to derive more traditional MR parameters such as VAT, subcutaneous adipose tissue (SCAT) or IHL but also more novel aspects such as Radiomic features of liver steatosis or liver shape from the participants of the NAKO study. Detailed metabolic phenotypes by MRI will be a crucial cornerstone, not only in the better understanding for the pathogenesis of cardiometabolic diseases. Furthermore obesity – and mainly abdominal obesity implementing a high volume of VAT – is an independent risk factor for a number of cancers, including prostate, breast and colorectal cancers. On the other hand, non-alcoholic fatty liver disease (NAFLD) is another emerging liver disease worldwide and can be quantified using the MR-derived IHL content. Unless simple steatosis might remain harmless for the individual subject, it might progress to the advanced form called nonalcoholic steatohepatitis (NASH) which is accompanied with inflammation and fibrosis further on, resulting in liver cirrhosis, reduced liver function and potentially hepatocellular carcinoma in the end. Thus, this proposal addresses the importance of metabolic-linked diseases by first derived established and novel parameters based on MRI within the German National Cohort and second analyze these MR-based parameters in the context of traditional risk factors, laboratory parameters, anthropometric, behavioral and metabolic data as well as of MR-based disease manifestations of other organs. In a first step MRI data will be analyzed as described below. Data use for the outlined epidemiologic research questions will be applied for as soon as the required data are fully available.

Schlüsselwörter-

EinrichtungenUniversity of Freiburg, Helmholtz Zentrum München, Universität Stuttgart, Ludwig-Maximilians-Universität München, LMU, Universitätsklinikum Freiburg, Universitätsklinikum Tübingen, University Hospital Heidelberg, Lehrstuhl für Epidemiologie, Universtitätsklinikum Augsburg, Uniklinik Freiburg, Section on Experimental Radiology, DKFZ, Diagnostische und Interventionelle Radiologie, Deutsches Krebsforschungszentrum, Klinikum der Universität München, Uniklinikum Tübingen, Forschungszentrum Jülich, Universität Heidelberg, DKFZ Heidelberg