Study Of Spio-enhanced MR Imaging In The Diagnosis Of Dysplastic Nodule And Hepatocellular Carcinoma

Objective:Previous investigates proved that the carcinogenesis of hepatocellular carcinoma (HCC) is significantly correlated to hepatocirrhosis. Cirrhotic liver parenchyma contains series of nodules, including benign regenerative nodule (RN), pre-malignant dysplastic nodule (DN) and malignant HCC. Viral hepatitis, mainly caused by the hepatitis B and hepatitis C viruses, is currently the most important etiologic factor leading to hepatocirrhosis in China. There is a big population of hepatitis B in China, the incidence of HCC has been increasing, and long-term survival is poor. Resection and transplantation are the most effective treatment for HCC at present, substantially increases survival. It is therefore critical to detect nodules that contain HCC at an early stage. During the past 2 decades, MR imaging has emerged as an important imaging modality for assessing cirrhosis and its complications, such as hepatic nodules. For assessment of hepatic nodules, MR imaging is more useful than any other imaging modality currently available. The multiphasic dynamic sequences show the enhancement patterns of liver nodules, which is very helpful in characterization of the nodules. But some questions remain after gadolinium-enhanced MR imaging. In addition, MR imaging with tissue-specific contrast media should be advocated to answer certain questions that might remain after MR imaging with gadolinium contrast material. Superparamagnetic iron oxide (SPIO), which is reticuloendothelial system-specific, has been used more and more widely in recent years. Former researches demonstrated that spio-enhanced MR imaging can increase detectability of hepatic lesions, differentiate between benign and malignant lesions, and help to assess the differentiation of HCC. This study was designed to evaluate SPIO-enhanced MR imaging in the diagnosis of DN and HCC by inducing hepatocirrosis, DN and HCC in rats. Materials and methods:1. Diethylnitrosamine (DEN, 1%) was used to induce hepatocirrosis, DN and HCC in rats. SPIO (ferucarbotran, SHU-555A , Rosevist, 0.5mmolFe/ml) was used as contrast media for MR imaging.2. Forty-five male Wistar rats were randomly divided into two groups: the experimental group (n=40) and the control group (n=5). Diethylnitrosamine that was diluted by pure water to 100ppm was given as drinking water to induce hepatocirrhosis, DN and HCC in the experimental group in the first 16 weeks. Pure water was given after 16 weeks. Non-enhanced and SPIO-enhanced MR imaging were performed at week 22. Pure water was given in the control group all the time of the experiment.3. MR imaging was performed with 1.5 MR imaging units (Signa, GE medical system, USA). All images were obtained in the traverse plane by using surface coil. A thickness 3mm, spacing 1mm, FOV140mm, and matrix 192×160 was held constantly for all the sequences. Plain scan sequences were: T1WI FSPGR (TE160,TR8), T2WI FSE (TR4000,TE103),NEX=3. T2WI FSE was used in SPIO-enhanced MR imaging. The same parameters were applied in the control group. Rats were sacrificed after scanning for pathological examination.4. We chose nodules with diameter larger than 3mm to avoid partial volume effect. Signal intensity (SI) of non-tumorous liver parenchyma, DNs and HCC was measured, percentage of signal intensity loss (PSIL) and SPIO intensity ratio (SIRSPIO) of DNs and HCC were then calculated.5. DNs and HCC were confirmed by HE staining and immunohistochemical staining. Kupffer-cell count of lesions and liver parenchyma was determined by Perls staining, Kupffer- cell-count ratio was then analyzed.Results:1. The signal intensity of normal and cirrhotic liver parenchyma decreased significantly, and normal liver parenchyma demonstrated greater decrease than cirrhotic liver parenchyma (65.16±10.12 vs. 47.74±7.92, P<0.05) . 2. On precontrast images, DNs were hyperintense or isointense to the surrounding liver on T1WI and hypointense or isointense on T2WI. Most moderately- and poorly-differentiated HCC was hypointense to the surrounding liver on T1WI and hyperintense on T2WI. The signal intensity of well-differentiated HCC varied greatly. Well-differentiated HCC was hyperintense, isointense or hypointense to the surrounding liver on T1WI and hypointense, isointense or hyperintense on T2WI. On spio-enhanced T2WI, DN was slightly hyperintense or isointense to the surrounding liver parenchyma, moderately- and poorly-differentiated HCC were hyperintense, and well-differentiated HCC was hyperintense or isointense.3. From DN to well-, moderately- and poorly- differentiated HCC, Kupffer-cell count (Kupffer-cell-count ratio) decreased. PSIL was directly and SIRSPIO was inversely correlated to Kupffer-cell count.Conclusion: SPIO-enhanced MR imaging may reflect Kupffer-cell count of DN and HCC, and its efficacy is damaged by hepatocirrhosis. A comprehensive application of signal-intensity characteristics, SIRSPIO and PSIL is useful for the diagnosis of DN and HCC.