Characterization and Diagnosis of Liver Cirrhosis and Hepatocellular Carcinoma by Native Fluorescence Spectroscopy of Human Urine Coupled with Multivariate Statistical and Chemometric Methods.

Early diagnosis of hepatocellular carcinoma (HCC), especially on the background of cirrhosis is crucial in the management of liver disease and providing proper therapeutic protocol. In this study, urine samples of cirrhosis and HCC patients were characterized using fluorescence excitation emission matrix to optimize excitation wavelength. Fluorescence spectroscopic measurement at 405 nm excitation was carried out to monitor molecular level changes associated with liver pathologies. Parallel factor analysis, principal component analysis and spectral deconvolution identified the presence of NAD(P)H, riboflavin, flavins and porphyrins in urine. Receiver operator characteristics of fluorescence intensity ratio - I464/619 provided better result in the binary classification of liver abnormal group (cirrhosis and HCC) from normal group with sensitivity of 95.74% and specificity of 98% in comparison with other fluorescence intensity ratios. Linear discriminant analysis of various fluorescence intensity ratios provided better result in the classification of normal vs. cirrhosis vs. HCC with 86.6% accuracy.