Hepatobiliary and pancreatic: Liver histopathology: Fourier transform infrared microspectroscopic imaging for objective and quantifiable assessment of liver biopsies

Abstract

Histopathological changes in the liver are conventionally assessed using histochemical stains and examination by an experienced histopathologist. Sampling error and interobserver variation were addressed by Bedossa et al. who called for objective measurements by imaging technologies. We are investigating the applicability of FourierTransform Infrared Microspectroscopic Imaging (FTIRMI) to the analysis of liver biopsies. One previous study has been reported. Moreover, microimaging could find additional applications, including investigating lesions not amenable to histochemical study such as drug accumulations in liver cells. The applications of infrared spectroscopy range from analysis of interstellar dust particles to detection of cancer cells in cervical biopsies. The infrared spectrum of any compound is in effect a unique molecular fingerprint. The positions of observed peaks depend on the masses of the atoms, the strengths of the bonds between them, and the interaction with the surrounding molecular environment. Furthermore, the absorbance is directly proportional to analyte concentration. Lewis et al. revolutionized infrared imaging analysis by combining a focal plane array detector with a FTIR microscope, allowing insight into the molecular structure of tissues. The diagnostic and monitoring capacity of FTIRMI has the premise that a chemical change must accompany morphological expression. Thus, morphological changes are readily detected by FTIRMI as changes in the concentration and conformational orientation of functional groups associated with proteins, lipids, nucleic acids and carbohydrates. Techniques such as Unsupervised Hierarchical Cluster Analysis (UHCA) can be used to analyze the ‘spectral data hypercubes’ that constitute the individual pixels in the focal plane array. Our images exemplify the technique applied to a cirrhotic liver. Figure 1 is a Masson stain (original magnification ¥ 100) while Figure 2 is an image derived by performing UHCA on spectral data obtained with the DigiLab ‘Stingray’ FPA system. The images show regions both with measurably high levels of collagen (fibrosis) and glycogen (liver parenchyma).