ABSTRACT
Steatosis is the most important prognostic histologic feature in the setting of liver procurement. The currently utilized diagnostic methods, including gross evaluation and frozen section examination, have important shortcomings. Novel techniques that offer advantages over the current tools could be of significant practical utility. The aim of this study is to evaluate the accuracy of surface color spectrophotometry in the quantitative assessment of steatosis in a murine model of fatty liver. C57BL/6 mice were divided into a control group receiving normal chow (n = 19), and two steatosis groups receiving high-fat diets for up to 20 weeks-mild steatosis (n = 10) and moderate-to-severe steatosis (n = 19). Mouse liver surfaces were scanned with a hand-held spectrophotometer (CM-600D; Konica-Minolta, Osaka, Japan). Spectral reflectance data and color space values (L*a*b*, XYZ, L*c*h*, RBG, and CMYK) were correlated with histopathologic steatosis evaluation by visual estimate, digital image analysis (DIA), as well as biochemical tissue triglyceride measurement. Spectral reflectance and most color space values were very strongly correlated with histologic assessment of total steatosis, with the best predictor being % reflectance at 700 nm (r = 0.91 [0.88-0.94] for visual assessment, r = 0.92 [0.88-0.95] for DIA of H&E slides, r = 0.92 [0.87-0.95] for DIA of oil-red-O stains, and r = 0.78 [0.63-0.87] for biochemical tissue triglyceride measurement, p < 0.0001 for all). Several spectrophotometric parameters were also independently predictive of large droplet steatosis. In conclusion, hepatic steatosis can accurately be assessed using a portable, commercially available hand-held spectrophotometer device. If similarly accurate in human livers, this technique could be utilized as a point-of-care tool for the quantitation of steatosis, which may be especially valuable in assessing livers during deceased donor organ procurement.
