Aims: The non-alcoholic fatty liver disease (NAFLD) activity score (NAS) is the histological tool used to assess disease severity based on steatosis, inflammation and hepatocyte ballooning. As steatosis contributes up to three of a potential eight points to NAS, it is important to quantify steatosis accurately. We sought to determine the optimum histological technique for identifying fat in tissue.
Methods and results: Using tissue from a mouse model of NAFLD, with validation in human liver biopsies, the percentage steatosis and fat droplet size were assessed in haematoxylin and eosin (H&E)- and Oil Red-O (ORO)-stained sections by light microscopy and digital image analysis (DIA). Results were compared to biochemical tissue triglyceride content and MRI assessment of hepatic lipid content. H&E steatosis assessment correlated poorly with tissue triglyceride concentration. However, ORO DIA exhibited much higher sensitivity and specificity for steatosis and correlated very well with triglyceride concentration in mouse and human liver (R = 0.706, P = 0.001 and R = 0.894, P =0.041, respectively). MRI-based assessment of steatosis was inaccurate.
Conclusions: ORO DIA is the most accurate method for detecting and quantifying steatosis. Although H&E-based NAS remains clinically valid in both clinical research and experimental situations, ORO DIA is a more robust technique to assess liver steatosis accurately for NAS scoring.
© 2012 Blackwell Publishing Ltd.