The development of magnetic resonance (MR) microscopy has provided new approaches to histology and histopathology. Recent work has shown the promise of increased sensitivity in animal models of chemically induced hepatotoxicity. However, the field is so new that there is little experience to relate changes seen in MR micrographs to the more traditional optical images stained with hematoxylin and eosin. This work compares the sensitivity and reproducibility of MR microscopy with conventional histopathology in detecting bromobenzene-induced hepatotoxicity in the rat. A time-course study was undertaken to provide a range of histopathologies. Specimens were studied at 24, 48, 72, and 96 hours after exposure to 10% of the median lethal dose of bromobenzene. Using 4 animals per group (a total of 32 rats) added statistical significance to the study and defined a range of interanimal variability over 96 hours. This work shows that MR microscopy, besides being nondestructive and three-dimensional, is at least as sensitive as conventional hematoxylin-eosin staining in detecting bromobenzene-induced centrilobular lesions and recovery of the hepatocellular architecture in the rat. This study further suggests that, as we begin to understand the underlying mechanisms of contrast in MR histology, MR may, in fact, supply even higher specificity than more traditional studies: variations were observed in MR images of treated livers at a given time point that could be not be differentiated based on the grading of necrosis and inflammation on hematoxylin-eosin-stained sections. (HEPATOLOGY 1998;27: 526-532.)
Methods
Animals. Thirty-two female Fisher 344 rats (Charles River Laboratories, Wilmington, MA) weighing 175 to 200 g were used. The protocol for this study conformed to Duke University' s Institutional Animal Care Committee-approved procedures. Rats were gavaged under light isoflurane anesthesia using an 18-gauge curved gavage Abbreviations: MR, magnetic resonance; LD 50 , median lethal dose; RF, radiofrequency.