SPIO Enhanced And MT Contrast MR Imaging For The Rat With Cirrhosis And HCC: Relationship With Pathologic Changes

ObjectiveTo study relationship of pathologic changes of hepatocellular carcinoma (HCC) and cirrhosis with magnetic resonance imaging (MRI) in the rat model induced by DENA. To correlate the signal intensity on superparamagnetic iron oxide (SPIO) enhanced MR images with Kupffer cells (KCs) of liver. To discuss the application value of magnetization transfer (MT) contrast MR imaging for hepatic lesions of rat. Materials and Methods1. Object and the animal model establishment: In this study, forty cleanly male Wistar rats were feed freely Diethylnitrosamine (DENA, American sigma company, 0.95g/ml) solution being diluted as 100ppm. After twelve weeks, the water without DENA was feed to these rats. Ten normal wistar male rats that bought from the animal center of Nan fang hospital of the first military medical university served as the control group.2. Instrument and contrast agents: MRI was performed with a 1.5-Tesla whole-body imager (Magnetom vision plus 4, Siemens, Germany) with a head coil. All images were analyzed and data were measured on the monitors of a picture archiving and communication system (PACS). SPIO (11.2mg Fe/ml) was triturated and offered by professor Xu Yi-kai in imaging center of Nan fang hospital. The gadolinium-Diethylene triamine pentaaceticacid (Gd-DTPA, 0.496g/ml) wasproduced by Kangcheng Pharmacy Company.3. MR sequences and parameters: the images obtained before and after SPIO enhancement consisted of Tl-weighted spin echo (SE) images (repetition time [TR] msec/echo time [TE] msec, 450/15; field of view (FOV), 100mm X 140mm; three signals acquired; thickness of slice, 4mm; matrix, 134X256) and T2-weighted fast SE images (TR/TE, 5000/128; 138X256 matrix) and fat suppression (FS) T1WI (539/12; 194X256 matrix) and FS T2WI (5000/128; 138X256 matrix) images, PD-weighted spoiled gradient-echo (fast low-angle shot [FLASH], 550/18; 120X256 matrix; flip angle, 20° ) and double FSE (T2-weighted: 4020/90; PDWI: 4020/20) and EPI SE T2WI (0/81.3). After fifteen minutes SPIO administration (0.05mmol/kg) was injected, MR imaging was performed. Double enhanced MRI was that Dynamic Gd-DTPA (0.2ml/kg) enhanced FLASH (TR/TE: 160/4.7) T1WI was carried out immediately after SPIO enhancement, and delayed period was Tl-weighted SE imaging. MRI with MT presaturation pulse was performed in each sequence.4. Pathological examination: after MRI, fresh liver tissues were immediately chipped into a piece of 1.0mm3 and were fixed, embedded, sliced up and the slices were observed and taken pictures under an transmission electron microscope (TEM). The liver were transected into slices with 5.0mm and taken count of the lesions. All liver tissues were fixed in neutral formalin solution and embedded with paraffin and sliced up. These slices were preformed with HE and peri's staining. Two pathologists made the pathological diagnosis. HCCs with regions of varying histological grades were classified as belonging to the predominating histological characteristic. Two ultrastructure pathologists observed ultrastructures of cells and KCs. HE staining were used to qualitative diagnosis. On peri's staining slices, the amount of KCs was counted under five high power lens of optical microscope (X40). The ultrastructure of hepatic cells and KCs were observed in the different liver tissues under TEM.5. Image analysis: three MR doctors read MR images. Maximal diameters of lesions were measured and lesions were determinant as hypointensity, isointensity or hyperintensity compared with the adjacent hepatic parenchyma on MR images. Average signal intensity (SI) over region-of-interest (ROI) drawn on lesions(Maximal diameter 255.0mm) and adjacent hepatic parenchyma were measured onMR images. Background noise (N) was measured in each image and ROIs wereplaced adjacent background outside abdomen and coding direction was same to thelesions. Signal to noise ratio (SNR), contrast to noise ratio (CNR) of lesions and liverwere calculated on all images and percentage of signal intensity loss (PSIL) afterSPIO enhanced images and MT ratio and MT effect on MTC images.6. Statistical analysis: statistical package for social sciences (SPSS 10.0) and theIndependent-Samples T Test, repeated measurements One-way analysis of variance(ANOVA) and Curve estimation were used. AP value of 0.05 or less was consideredto indicate a significant difference.Results1. Pathologic findings of the rat model with cirrhosis and HCC1.1 Macropathologic findings: in forty rats, twenty-two (55%) were induced successfully in which the cirrhosis was six and the HCC was sixteen, and the mortality was 45%. The death reasons in early process were hemorrhage of liver, spleen and lung and dysfunction of multiple organs. In late period, the causes of death were large HCCs were broken and hemorrhage and dysfunction of multiple organs. The size of cirrhotic liver was small and its surface was rough. Gray-white nodules (lmm~2mm) distributed equably and diffusely in cirrhosis liver. Yellow-white HCCs were various sizes in liver. The maximal diameter of HCC ranged from 5mm to 25mm and the diameter was beyond 30mm in the maximal tumor. Of 115 lesions (maximal diameter < 3.0mm), Ninety-eight were determined as HCC and seventeen were cirrhotic nodule (diameter < 5.0mm).1.2 HE stainning findings: the hepatic structure was destroyed and the size of hepatic cells was slightly larger and some were swelling, and connective tissues increased and many inflammatory cells were in portal area on cirrhotic HE staining slices. Of HCC, the structure was abnormal and loose and tumor cells were obviously larger and were heterotype, and basophil of karyon was increased and had multiple or heterotypic nuclear. In HCC tissues, blood sinusoids were ecstatic and increased.1.3 On perl's staining slices, irregular KCs distributed equably in which had manyblue particles in normal liver. The amount of KCs was slightly reduced in cirrhosis in which the blue particles were heterogeneous distribution. KCs in well-differentiated HCC were lower and of the poorly differentiated HCC were obviously reduced or hardly disappeared than the normal and the cirrhosis. The number of KCs of HCCs was significantly lower than the normal and the cirrhosis (P=0.000), while there was no significant difference between of the normal and of the cirrhosis (P=0.088). 1.4 On TEM slices, hepatic cells had abundant mitochondria and KCs had ample lysosomes in which black SPIO particles were taken up and large SPIO clusters were formatted in cytoplasm in normal livers. In cirrhotic tissues, intercellular space was broadened and collagenous fibers remarkably increased. The organelles were heterogeneous distribution and the mitochondrion thread granules were argument and appeared vacuolar degeneration, and the structure of rough endoplasmic reticulum (RER) was destroyed in the cirrhotic hepatic cells. In cirrhosis liver, the amount of KCs slightly decreased in which the amount of lissome was lower than the normal liver, and in lysosomes small SPIO particles were taken up and large SPIO clusters were formatted in cytoplasm. In well-differentiated HCC, the appearance of tumor cells was normal and in which organelles and KCs decreased. In poorly differentiated HCC, the cytoplasm obviously decreased and the organelles disappeared and. karyon was excursionl and KC was almost absent. 2. MR imaging findings2.1 Ninety-eight HCC showed commendably on each sequence. On before and after SPIO enhanced T2WI and double enhanced T1WI, Twenty-four small lesions (maximal diameter < 3.0mm) were showed clearly and the detection rate was 100%. On FS T2W images, all lesions were showed but some were covered up by the rough hepatic parenchyma. On T1WI and FS Tl WI, the detection rate of small lesions (42%, 58%, respectively) was obviously lower than the T2WI. On FLASH PD-weighted images, the contrast of lesions was no better because the signal intensity of parenchyma increased so that the detection rate of small lesions (75%) was low. On SPIO enhanced T1WI, the contrast of lesion was decreased obviously so that the small lesion was hardly seen. On all MR images, cirrhosis nodules was unclear andshowed that hepatic parenchyma was coarse or unclear.2.2 On TIWI, HCC was hypointensity or iso-hypointensity. The signal intensity of the hepatic parenchyma was comparatively increased so that the edge and inside of tumor was clearer. On FSE T2WI, HCC showed hyperintensity and was similar to the human's HCC. The large lesions easily appeared putrescence that showed as higher signal intensity on T2WI. The contrast of HCC was improved and tumor was high hyperintensity but inside structure showed unclear, and small lesions detected difficult due to the hepatic parenchyma was rough on FS T2WI. FSE-echo planar T2WI, the contrast of HCC was highest but the image quality was bad and the inside of HCC was showed worst.2.3 The signal intensity of surrounding hepatic parenchyma was decreased so that some lesions showed unclear due to the contrast of the lesions was depressive on SPIO enhanced TIWI. On SPIO enhanced T2WI, the signal intensity of the hepatic parenchyma was obviously decreased and the signal intensity of well-differentiated HCCs was decreased so that the contrast of some HCCs were decreased, while poorly differentiated HCCs retained hyperintensity so that HCCs was clearer. On GRE FLASH TIWI, HCCs was heterogeneous enhancement in arterial and portal period of dynamic Gd-DTPA enhanced imaging, and in delay period the enhancement was still obvious and lesions were clearer. On SPIO enhanced T2WI, the signal intensity of the normal hepatic parenchyma and cirrhosis were obviously lower than before SPIO enhanced images and PSIL were 42% and 38% respectively. PSIL of SPIO enhanced images was no significant difference between the two groups (P=0.409). The signal intensity of HCC on SPIO enhanced T2WI was similar to those images before SPIO enhancement. PSIL of HCC (12%) was significant lower than the normal liver (i>=0.000) and cirrhosis (P=0.000). CNR of HCC on SPIO enhanced T2WI was significant higher than before SPIO enhancement (P=0.002). On SPIO enhanced TIWI, PSIL of normal liver and cirrhosis was 15% and 6% respectively and the HCC was increased by 9%. Some small HCCs were slightly inhomogeneous hyperintensity on SPIO enhanced Tl WI. CNR of HCC was obviously lower than the images before enhanced (P=0.000). SNR of the normal and cirrhosis on SPIO enhanced TIWI andT2WI were significant lower than before enhancement (P=0.000). SNR of HCC was no significant difference between after and before SPIO enhanced T2WI (P=0.187).2.4 PSIL was increased on SPIO enhanced T2WI along with increaseing of the amount of KCs and was curve trend in hepatic tissue. R2 (Rsq) of Cubic models of curve estimation was 0.920 and had significant (P=0.000). The amount of KCs of norml liver was abundant and phagocyte power was normal so that the signal intensity decreased obviously. The number of KCs in cirrhosis was similar to the normal but phagocyte function was abnormal so that PSIL was slightly lower than the normal liver. In HCC, the amount of KCs obviously decreased so that signal intensity was no significantly decreased and maintained hyperintensity correspondingly.2.5 MT findings: on T1WI, the signal intensity decreased on MTC images and SNR of MTC images were lower than non-MTC in normal liver (P=0.002). The signal intensity of cirrhosis was obviously decreased on MTC images and SNR was significant difference than non-MTC images (P=0.000). In HCC, SNR of T1WI with MTC was slightly lower than non-MTC images but was no significant difference between the two groups (P=0.158). On T2WI, the signal intensity of normal liver with MTC was lower than non-MTC and SNR was significant lower than non-MTC (P=0.021). In cirrhosis liver, SNR of T2WI with MTC images was remarkably low than non-MTC (P=0.000). The signal intensity of HCC on T2WI with MTC images was similar to non-MTC and SNR was no significant difference between MTC and non-MTC images (F=0.549), while CNR was higher than non-MTC but was no statistical significance between the two groups (P=0.205).MT ratio (MTR) of cirrhosis (0.49) was significant higher than the normal liver (0.17) and HCC (0.19)(P=0.000), while MTR of normal liver was similar to HCC (P=0.788) on T1WI. On T2WI, MTR of cirrhosis was remarkably higher than the normal (F=0.002) and HCC (P=0.000), while MTR was no significant difference between the normal and the HCC (P=0.352). SPIO enhanced T1WI; MTR of cirrhosis was higher than the normal and HCC (P=0.000). MTR of cirrhosis was similar to the HCC, while MTR of normal liver was significant lower than the cirrhosis and HCC (P=0.032, 0.014 respectively). MT effect of cirrhosis was remarkable and hadsignificant difference between the normal liver (0.14) and HCC (0.14) on TIWI with MTC images (P=0.000). On T2WI, MT effect of cirrhosis (0.28) was higher than the normal (0.17, P=0.003) and the HCC (0.13, P=0.000), while MT effect had no statistical significance between the normal and the HCC groups (P=0.170). Conclusions1. The rat model carrying cirrhosis and HCC induced by lOOppm DENA was a multiple centers and nodules of HCC accompanied with cirrhosis. The processes of commencement and development of the rat model were similar to the cirrhosis type of HCC in human. The appearance rate of HCC was high and the method was performed easily. The rat model of HCC was a perfect animal model to study the imaging for cirrhosis and HCC.2. In the normal liver of rat, hepatic cells arranged into cord-like and had abundant mitochondrion thread granules and RERs were layer arrangement. The structure of cirrhotic tissues was abnormal and fibers increased and hepatic cells were slightly large and some were swelling. Organelles were heterogeneous distribution and the mitochondrion thread granules were argument and appeared vacuolar degeneration, and the normal layer structure of RER was destroyed in the cirrhotic hepatic cells. The structure of HCC was abnormal, loose and blood sinusoids were ecstatic. Well-differentiated HCC cells were large in which organelles decreased. Cytoplasm of poorly differentiated HCC cells obviously decreased and organelles disappeared and, karyon was excursion and chromatin was abnormal.3. Signal characteristic of cirrhosis and HCC of the rat model on MR images were similar to of the humans. On Tl-weighted spin echo images, HCCs were hypo- or iso-intensity, while were hyperintensity on T2-weighted fast spin echo images. In detection the small lesion (maximum diameter^3.0mm), FSE sequence T2-weighted imaging was most sensitive. SPIO-enhanced T2-weighted imaging and double enhanced Tl-weighted imaging could obviously improve the rate of the detection and characterization in small HCC and were very important to the diagnosis and differentiated diagnosis of the hepatic focal lesions.4. Signal changes of liver on SPIO enhanced T2WI correlated with the amount andfunction of KCs and PSIL increased along with increaseing of the amount of KCso KCs of normal liver could take up much SPIO particles so that the signal intensity was obviously decreased on SPIO enhanced T2WI. In cirrhosis liver, the amount of KCs was similar to the normal liver but the function was slightly incomplete so that PSIL was slightly lower than the normal on SPIO enhance T2WI. The amount of KCs of well-differentiated HCCs decreased so that the signal intensity decreased, while the poorly differentiated HCCs had few KCs and could not take up SPIO particles so that the signal intensity decreased obviously than before SPIO enhanced images. The contrast of HCC was improved. SPIO enhanced MRI could improve the detection of focal hepatic lesions, including metastasis and primary hepatic malignancies, and could indirectly reflect the amount of KCs in liver tosses and was helpful for forecast of the histological stages in HCC.5. Cirrhotic liver had a strong MT effect due to the abundance of collagen. MT contrast imaging might indirectly reflect the changes of collagen content and could provide some indirect evidence about diagnosis and stages of cirrhosis. MTC might be regarded as effective method to evaluate cirrhosis.6. The contrast of HCC with the adjacent hepatic parenchyma was improved due to the signal intensity of surrounding cirrhotic tissue decreased. MTC was in favor of in the detection and charactezation of HCC in cirrhotic background. When MT effect of liver cirrhosis had an abnormal weaker than it should have, the possibility of diffuse hepatoma or coexistent fatty change should be considered. MTC imaging was a potential tool for tissue characterization. MT contrast and SPIO-enhanced MR imaging might have some concurrent effect and were combined them might ulteriorly improve the contrast of HCC and provided more information about the diagnosis and differentiated diagnosis for HCC.