| Objective and SignificanceHepatocellular cacinoma (HCC) is the most common cellular type of primary hepatic carcinom with high mortality and short-term survival in our country. Those are problems to be resolved urgently that early detection and diagnosis of HCC,deducing of the malignant grade and assessment of the porgnosis. CT perfusion (CTP) is a new detection method as functional imaging that can detect changes of blood flow in tissue by measuring tissue perfusion quantitatively, and show morphological as well as physiological and pathophysiological changes of organ. Clinical application of liver CT perfusion of 64 multidetector spiral CT has succeeded in realization liver imaging of CT perfusion, the method can evaluate the liver hemodynamics of micro-circulation, rapidly, accurately and non-invasively, that shows its broad perspective of clinical application.The main objectives of this research that performed CT perfusion imaging in animal model of VX2 liver cancer and clinical cases followed as: (1) To discuss technique advantages of CT perfusion imaging; (2) To explore the characteristics of blood perfusion of disease by evaluation and analysis of perfusion imaging and data in HCC; (3) To exploring futher the correlation between HCC perfusion data and neoplastic angiogenesis with combination of pathological control, to achieve the aims of non-invasive and effective evaluation of the status of neoplastic angiogensis in vivo, and to provide meaningful direction of clinical application.Materials and Methods1. aparatusSixty-four multidetector spiral CT (PHILIPS Brilliance, Holland), high pressure twein injector (MEDRAD, USA), post-image processing station is Sixty-four multidetector spiral CT equipped Mxview station and special analystic software for liver perfusion.2. Animal experment and clinical cases2.1 Animal experimental materials:2.1.1 Experimental animal groups:Eighteen New Zealand Rabbits ( provided by medical animal experimental center of Guangdong Province), weight 23kg, either male or female, were allocated randomly in trial and control groups, trail group divided in 3 groups with 5 rabbits respectively, control group divided in 3 groups with 1 rabbit.2.1.2 VX2 cell strain inocubating and preparing of modified VX2 HCC rabbit model:One New Zealand Rabbit tumor bearing was prepared by inoculation carcinoma cells in the subcutaneous of rabbit leg to form neoplasm and passage, the formed carcinoma was dissected from seeded rabbit about 2 weeks and prepared in 1～2mm3 cubes on culture dishes, and reserved as carcinoma seed sucked by 18G syring, and the method of tumor embedded directly was used. Rabbits were anesthetized intravenously by 20% Urethane (5ml/kg), and fixed supinely in operation bed, skin of operation area was prepared. After being disinfected conventionally, middle incision under xiphoid was used, skin,subcutaneous, muscles and peritoneum were cut to enter abdominal cavity, then the liver was exposed, the left and the middle lobes were found, the deepth of about 0.5cm was punctured in the liver parenchyma distant to its lower edge abour 1.0cm～1.5cm, to form small diagonal injure, and a small tumor tissue was inoculated there. After the tumour being confirmed in the liver parenchyma, gelatin sponge was pressed there to stop blooding and to prevent abdominal seeding of tumor following bood flow. The liver was put back in abdominal cavity after coagulation, antibacterial agents were spryed in wound after the incision was closed. That rabbit was sent back to animal room to be raised, after resuscitation.2.1.3 Processing of experimental animals:On 14th day, 18th day and 22th day after tumor were inoculated in that rabbits of trial groups, a trial group of rabbits and a control group of rabbites were anesthetized and fixed in CT examination bed to perform perfusion imaging; after scan was finished, trial animals were sacrificed by using air embolus immediately, the results of gross pathological appearance and quantification of immunohistochemistry were compared with the results of CTI perfusion imaging.2.2 Clinical research materialsMaterials of fifteen cases, that HCC was removed by operation and confirmed by pathological examination, were collected, patients aged from 34～76 years old, average age was 45-year, among them 11 cases were male, 4 cases were female. CT perfusion examination was performed before operation, routine pathological HE and immunohistochemistrical testing were done after operation. Two parts of carcinoma lesion and near carcinoma lesion (more than 2 cm from the margin of carcinoma lesion) were obtained for every pathological sample for the sake of control.3. CT scan protocol and data3.1 Animal trial groupThat trial animal was fasted for 12h on the day of test, venous indwelling needle (24GA×0.75IN) was placed in Auricular vein of rabbit before scan. After being anesthetitized, that rabbit was fixed on a wood board supinely, abdominal belt was used to reduce respiration movement. Conventional CT plain scan of general liver was performed first, when disease and the area and slice of interesting lesion were confirmed, the mode of static slice and dynamic cine perfusion scan was applied, 4 ml of Iopamiro (370mg/ml) was injected at the rate of 1.5ml/s by high pressure twein injector, and at the same rate 2 ml of normal saline was injected, scan was started when 1s was postponed after the contrast agent was injected, continuous volume scan mode was used for image acquisition. Scan data: detectors were configured as 64×0.625mm , slice-thickness was 2.5mm, and non-spiral mode scan, 120KV,80mAs,360°rotation time: 1.2s, scan time per rotation was 0.75s, 50 scans, total scan time: 60s.3.2 Clinical materialsPatients fasted, 20G venous integrated catheter was indewlled in median cubital vein, 500 ml of pure water was administered 10 min before scan, abdominal belt was used during scan, patients were told to take stable breathing throughout scan. Epigastric plain scan was performed first to confirm liver position, total liver perfusion model was applied with 40 ml of Iopamiro was injected at the rate of 5ml/s by high pressure twein injector, and 20 ml of normal saline was injected at the same rate meanwhile, total liver perfusion was started after the contrast agent was injected for 8s. The main scan data: detector configuration was 64×0.625mm,slice thickness was 5.0mm, scan interval 5.0mm, Pitch 1.165, 120KV,100mAs,360°rotation time:0.4s, scan interval time 4.7～5.6s (defaulted as the minimum), 15 scans, total scan time was controlled between 85s～95s, to get the contrast agent dynamic curve of the complete circulation of contrast agent in liver.3.3 The method of CT perfusion imaging3.3.1 Processing of the imagingThe serial images of tow groups were transferred to Mxview post-processing station, software of CT perfusion (CTP) was input, abdominal aretery was confirmed as perfusion artery substitued the liver artery, the portat vein or its main branche was confirmed as perfusion vein. Post-image processing was performed then.3.3.2 Selection fo region of interest (ROI)ROI was selected according to different aims, generally including aorta, portal vein,liver and spleen etc. The aorta and the portal vein were showed as round and circulate, solid organs (the liver and the spleen) were showed as irregular shaps. ROI was selected as large as possible, could not be less than 50 pixel to reduce quantum noise. But the edge of solid organs could be reached to avid the influence of part volum effects, it is better that larege vessels were included in ROI of parenchymal area as possible, necrosed region was avoided.3.3.3 Time desity curve (TDC)After proper ROI having been selected, TDC was output by perfusion software automatically, including TDC of the aorta, the portal vein, the liver and the spleen etc. The tendence of TDC was observed.3.3.4 Caculation method and main perfusion dataThe software was base on convolution method to caculate the main data of liver perfusion imaging as followed: the indices of the hepato-artery perfusion (HAP), the hepeto-portal perfusion (HPP) and the total liver perfusion (TLP) etc. The caculatiom formula of HPI: HPI=HAP/(HAP+HPP)×100%, and TLP=HAP+HPP3.3.5 Output of perfusion color imagesCT perfusion software was used with the enchanced pitch values of spleen as threshholds, and the liver perfusion was divided into the arterial stage and the veinal stage. Image reconstruction and pseudo-color processing were performed for collected data, the blood perfusion color images were divided into four kinds: the hepato-arterial perfusion image, the hepato-portal vein perfusion image, the total liver perfusion image and the index figure of hepato-arterial blood perfusion, and the status of organ perfusion was evaluated completely.4. Pathological evaluation4.1 MVD caculation methodAccording to the criteria of Weidner et al, a microvessel is regarded as that any endothelial cells or endothlial cell clusters were stained as brown-yellow isolately by CD34 antibody, and can be separated from near microvesssel, neoplastic cells and other connective tissue; If structure is not connected, its branch structure can be regarded as microvessel counter. Those vessels can not be counted when cells are blurred or stained without good visibility, muscular layer rather thick or the lumen square is more than the diameter of 8 red blood cells. That section was scanned generally under 100 magnification light microscope to find high desity area of neoplastic vessels, 4 highest desity area were selected and observed under high magnification, then vessels were counted in 0.25μm2 area under grid micro scale of eye lens, the final mean value was MVD when those counts of 4 fields of view were averaged.4.2 Evalution method of VEGF staining resultsThe criteria of VEGF expression results adopted combination of the desity of positive VEGF staining and the percent of positively stained cells. The criteria of VEGF positive results: VEGF positive expression was located in the cytoplasm of neoplastic cells, positive cells showed brown-yellow or brown-yellow granule, the whole section was scanned under high magnification fo 10 fields of view, to count 1000 neoplastic cells, grade 0, if the positive cells =10%, then it was detected as negative; the positive cells =10%, then it was detected as positive; grade 1, among the total cell count, the positive cells count between 10%～25%; grade 2, the positive cells count between 26%-50%; grade 3, the positive cells count between 51%-75%; grade 4, the positive cells count between 76%-100%. That was simplified as 0,1,2,3,4 when score was filled in tables.5. Statistic analysisCT perfusion data, MVD ect were expressed as x|-±s. Statistical analyses were carried out with one-way anova or LSD for K Independent Samples; the Mann-Whitney U-test for two samples; Nonparanetric Tests were carried out for the two independent samples of rank data VEGF ; Pearson Correlate tests and Spearman's Correlate tests were carried for CT perfusion data HAP, MVD and VEGF; using SPSS 11.0 for Windows. A probability of p<0.05 was accepted as significant.Results1. Results of animal experiment1.1 Analystic results of morphology of normal control groups and HE, immunohistochemistric observation by light microscope.Neoplasm and other abnormal pathological changes were not observated in 3 rabbit of the control group. Primary results of HE staining under light microscope: Hepatocyte were seen distributing regulately with clear margin and normal morphology, showing typical structure of hepatic lobule, hepatocytic size was uniform, circulating the central vein and distributing radially, hepatic sinusoid was seen clear; nucleolus size and morphology were regular with uniform staining and without heterotypic mitosis. Analystic results of VEGF in normal control group under light microscope: Positive cells were not seen in normal hepatic tisssue, and all showed negative appearance.1.2 Analystic results of morphological observation, HE and immunohistochemistry in VX2 HCC model of rabbit by light microscope.Neoplasm growed in 16 New Zealand Rabbits of the trial groups successfully. Total of 26 HCC lesions were dynamically measured on the 14th day, 18th day and 22th day, there were 4, 11,6, 5 HCC lesions respectively; the size of HCC lesions varied from 0.3 cm～3.9cm, among those HCC lesion, sixteen located in the left lobe, ten located in the right lobe; in the trial group, single HCC lesion happened in nine rabbits, multiple HCC lesions happened in six rabbits.Results of light microscopic observation: Under low magnification, HCC was seen as invasive carcinoma in the liver, there was no clear margin between neoplasm and the hepatic parenchyma, mesenchyma blurred, the amount of connective tissue varied, invasive hepatic cord structure showed at the margin of carcinoma; neoplastic cells distributed straggly, there was separating fibre with abundant capillary of newly developed; under high magnification, the volume of neoplasm was large with irregular morphology, cells distributed irregularly with abundant cytoplasm and light red staining. Nucleus was hypertrophic with different size and shape, staining was non-uniform, phase of nucleous mitosis showed in many cells, a great number of lymphocytes and plasma cells could be seen infiltrating in the mesenchyma.VEGF positive staining could be seen most in cytoplasm as uniform or non-uniform granule, clot in brown-yellow color; as HCC cells proliferated actively, positive cells were more abundant, and increased not only in cell number but stronger in staining.1.3 Correlation between the perfusion results of trial groups and their pathological results1.3.1 Analysis of CT perfusion data of the trial groups and the control groupsHAP,HPP and HPI data of the trial groups were 59.15±16.49,92.07±31.96 and 39.90±6.82 respectively; HAP,HPP and HPI data of the control groups were 32.38±8.48,148.54±41.23,18.46±5.66 respectively. There were significant differences between CT perfusion data of the trial groups and the control groups by Tow Independent-Samples T Test: HAP (t=-6.690), HPP(t=5.169), HPI(t=-12.074)的P=0.000.1.3.2 Control analysis of CT perfusion data at different time points of the trial groupsROI were measured three times in 5 rabbits of each trail group of VX2 rabbit cancer model on the 14th day, the 18th day and the 22th day. HAP,HPP and HPI data of the 14th day were 42.03±22.32,113.30±13.99,25.90±11.02 respectively; HAP,HPP and HPI data of the 18th day were 48.76±17.04,59.41±15.04,44.34±9.90 respectively; HAP,HPP and HPI data of the 22th day were 65.99±15.96,23.78±11.18,73.33±11.04 respectively. Results analysed by one-way ANOVA, showed different significantly HAP(F=6.543),HPP(F=167.919),HPI(F=75.029),P=0.000, at different time points of the trial groups of CT perfusion data. 1.3.3 Control analysis of CT perfusion data in different parts of the trial groupsNecrosis occurred often in disease of VX2 rabbit cancer model, to guarantee the accurancy of results, we selected the region of tumor margin as ROI to collect data where tumor cells proliferated actively and necrosis occurred little. The paper compared the perfusion data of margin region of liver and the liver tissue near carcinoma. HAP,HPP,HPI of CT perfusion data of carcinoma margin area were 53.89±20.15,66.62±43.96,48.45±18.56 respectively; HAP,HPP,HPI of CT perfusion data of the liver tisssue near carcinoma were 27.35±11.17,132.08±21.87,17.12±6.74 respectively. Data analysed by Two Independent-Samples T Test showed that HAP(t=-7.360), HPP(t=8.596), HPI(t=-10.439), P=0.000, there were significant differences.1.3.4 Control analysis of VEGF pathology.VEGF changes in HCC tissue of the trial group were 1.00±0.38,1.47±0.64,2.07±0.88 on the 14th day, the 18th day and the 22th day respectively; that was 0.13±0.35 in the control group.Statistic analysis showed different significantly (p<0.05) by Multiple comparison LSD excerpt those on the 14th day, the 18th day and the 22th day. VEGF expressed in the trial groups compared with that in the control group, there was significant difference statisitcally (F=27.39,P=0.000). Comparision between of CT perfusion of HAP, HPI and VEGF expression showed positive correlation between HAP and VEGF expression (r=0.610,P=0.000) analysed by pearson correlation analyse. And there was positive correlation between HAP and VEGF expression (r=0.636,P=0.000),there was significant difference, P<0.05.2. Results of clinical research 2.1 Analystic results of immunohistochemistry (MVD,VEGF) by light microscopeCD34 was expressed in HCC and tissue near HCC, but MVD of HCC tissue was higher significantly than that of the tissue near HCC. Various degree of yellow staining could be seen in the epithelial cells of hepatic sinusoid in the HCC section, individual vessels could be seen stained near HCC.VEGF expressed mainly in the cytoplasm of neoplastic cells, positive staining was showed as light yellow, brown-yellow and dark brown grains; a small number of VEGF could be seen in some liver samples near HCC. VEGF positive neoplastic cells distributed non-uniform, area invaded by neoplastic cells stained stronger, due to disseminated or concentrated VEGF expression in carcinoma, carcinoma near the margin or the capsular of HCC showed more positive and stronger VEGF expression. VEGF staining degree in HCC tissue was obviously higher than that in the tissue nearby. Among those fifteen cases, 14 cases showed positive VEGF expression in the HCC tissue, only in 2 cases, VEGF expressed in the tissue near HCC. CD34 expressed in the HCC and the tissue near HCC, Various degree of yellow staining could be seen in the epithelial cells of hepatic sinusoid in the HCC section, individual vessels could be seen stained near HCC.2.3 Statistic analysis2.3.1 CD34 expressed in the HCC was 50.40±12.56; CD34 expressed in the tissue near HCC was 13.73±5.53. Two Independent-Samples T Test was carried out to analyse the data, result showed that there was significant difference statistically, p<0.05.2.3.2 VEGF expression in the HCC and the tissue near HCCIn HCC tissue, VEGF expressed negative in one case, positive in 14 cases, among them 4 cases showed positive strangely; In the tissue near HCC, VEGF expressed negative in 13 cases, positive in 2 cases weakly. Nonparametric Test of Mann-whitney Test was carried out to analyse that two groups' data, result showed that there was significant difference statistically, p<0.05.2.3.3 Correlate analysis between VEGF expression and MVDMVD in patients with positive VEGF expression was significantly higher than that in patients with negative VEGF expression, the amount and stronge degree of VEGF expression in HCC tissue was higher than that in the tissue near HCC, and MVD in HCC tissue was higher than that in the tissue near HCC, the data were analysed by Spearman Rank Correlation, results showed that there was positive correlation between VEGF expression and MVD (r=0.889,P=0.000) .2.3.4 Correlate analysis between MVD and CT perfusion data HAP etc in HCC and in the tissue near HCCCT perfusion data of HCC tissue HAP, HPP and HPI were 41.72±9.97,24.37±7.34,58.78±15.32 respectively; CT perfusion data of the tissue near HCC HAP, HPP and HPI were 24.3±15.7,54.10±19.06,30.03±15.77 respectively. Those data analysed by Two Independent-Samples T Test, results showed that HAP, HPP and HPI were significant difference statistically (p<0.05) ; the data were analysed by Pearson's correlation coefficient, results showed that there was positive correlation between HAP and MVD (r=0.664,P=0.000); the data of HAP and VEGF expression were analysed by Spearman Rank Correlation, results showed that there was positive correlation between HAP and VEGF expression (r=0.633,P=0.000) .Conclusion1. Conclusion of animal experiment 1.1 The rabbit model of VX2 liver carcinoma created by tumor enbedded directly, the time of carcinoma formation was short, the rate of carcinoma formation was high. The volum of tumor growed steadily in 2w3w, there was no obvious necrosis, and that was the best period for CT perfusion imaging.1.2 CT perfusion imaging had advantages in observation of the changes of hemodynamics of the rabbit model of VX2 liver carcinoma with accurancy, that could provide quantification information about the hepato-arterial perfusion and the hepato-port veinal perfusion.1.3 There was positive correlation between VEGF expression and the data of CT perfusion HAP, HPI in rabbit hepatic VX2 tumor, that explained that carcinoma angiogenesis could be reflected by the data of CT perfusion at some degrees.2. Conclusion of clinical application study2.1 Total liver CT perfusion imaging with 64 multidetector spiral CT is simple and practical with the advantages of total organ perfusion imaging, that overcomes those shortcomings of past perfusion imaging mode such as narrowness of scanning area, limitation by respiration movement etc.2.2 MVD and VEGF expression in HCC tissue were higher than that in the liver tissue near HCC, and that was significant difference statistically.2.3 There was positive correlation between VEGF expression and perfusion data of HAP and MVD, that meant there was correlation between CT perfusion imaging and neoplastic angiogenesis, therefore CT perfusion imaging could be applied to evaluate the angiogenic status of HCC in vivo.
|
PDF Full Text Request