Clinical And Experimental Study Of Renal Cell Carcinoma Perfusion Imaging With 64-Slice Spiral CT

64-slice spiral CT was applied to perform CT perfusion imaging (CTPI) in rabbit VX2 renal carcinoma model and renal cell carcinoma (RCC), measuring the CT perfusion (CTP) value respectively, comparing with that in the normal group and making the statistical analyses, and studying the hemodynamic changes of RCC. The quantitative indexes of tumor angiogenesis including microvessel density (MVD) and vascular endothelial growth factor (VEGF) of RCC and peritumoral renal cortex by immunohistochemical techniques were compared to explore the correlations between CTP parameters of RCC and the quantitative indexes of tumor angiogenesis, and consequently, to evaluate the clinical value of CTPI in RCC angiogenesis evaluation.PartⅠThe study of CTPI with 64-slice spiral CT on rabbit VX2 renal carcinoma modelObjectiveTo investigate the rule in hemodynamic changes of rabbit VX2 renal carcinoma model and provide theoretical bases for research in human renal tumor, through the establishment of rabbit VX2 renal carcinoma model and CTPI with 64-slice spiral CT on rabbit VX2 renal carcinoma model.Materials and Methods1 Major equipment and reagentsBrilliance 64-slice spiral CT scanner (Holland PHILIPS), with Mxview (with renal perfusion software) workstation and double-syringe power injector (US MEDRAD); non-ionic contrast agent, iopamidol (370mgl/ml).2 Experimental data2.1 Experimental groups A total of 20 New Zealand white rabbits, weight ranging from 2-2.5kg, male or female, were randomly divided into the experimental group and the normal control group.2.2 The establishment of rabbit VX2 renal carcinoma modelA tumor-transplanted VX2 New Zealand stud rabbit was prepared, tumor cells were inoculated in its hind leg subcutaneously to produce tumors and then the rabbit was sub-cultured to give birth to stud rabbits. Stud rabbits about two weeks were chosen and their tumors were removed; the vigorously-growing tissues surrounding the tumor masses were made into pieces about 1-2mm~3. Tumor masses transplantation by direct embedding was employed through the incisions in the posterior abdominal walls where kidneys were exposed, and the rabbit VX2 renal carcinomar model were established. Then these rabbits were fed for 16-20 days for subsequent experiments.2.3 CTPI techniques and methodsFirst the conventional full renal axial plain scan was performed, CT slice thickness and spacing both being 2.5mm, and then dynamic enhanced CT scan in a wide range and at a single level was performed, the maximal range of scan being 40mm and the injection rate of double-syringe power injector being 1.5ml/s; 4ml of iopamidol (370mgl/ml), the non-ionic contrast agent was injected, then 2ml saline was injected and CT scan was performed is later after the injection, for 50 times, 50s in total. Parameters of perfusion scan were primarily made up of 64x0.625mm of the detector array, and in non-interval continuous scan, 2.5mm of CT slice thickness, 120KV of voltage, 150mAs of electric current and 0.5s of 360-degree rotation time.Selection of ROI (region of interest), including the aorta, renal vein, tumor parenchyma and the renal cortex. The aorta and renal vein could be circular or elliptic, and the tumor parenchyma and renal cortex could be of the irregular shape. ROI should be as wide as possible to reduce the quantum noise, no fewer than 50 pixels; however, ROI could not reach the edge of the organ so as to avoid the impact of the partial volume effect (PVE); moreover, ROI of the parenchymatous area should not contain renal sinus fat and the lesion should avoid the necrotic area. 3 Image post-processing and data collection and observationIn Mxview workstation, renal perfusion software was employed for image post-processing. The abdominal aorta, taking the place of renal artery, was regarded as the afferent artery and the renal vein was regarded as the efferent artery. Regions of interest for the abdominal aorta, renal vein and renal cortex identified, the software would measure the CT value for each ROI automatically and the time-density curve (TDC) and the perfusion map were plotted; according to TDC, the software would generate and display perfusion indexes, including the renal blood flow (BF), renal blood volume (BV) and mean transit time (MTT) of the contrast agent and so on. All measurements should be executed twice during the period of study and their mean value was thus obtained.The perfusion parameters of BF, BV and MTT etc in the experimental and control groups were collected and recorded, and TDC shapes were observed.4 Pathological observationsRoutine hematoxylin-eosin (HE) staining having been performed, the normal kidney and tumor tissues were observed under the microscope.5 Statistical analysesSPSS13.0 software package was applied for statistical analyses in this study. Each parameter index of peffusion was indicated by mean±standard deviation ((?)±s), data were indicated by two independent samples t-test and P＜0.05, was regarded as the significant difference.Results1 General shapes and pathological observation of tumor-transplanted modelsThe ten New Zealand white rabbits all gave birth to tumors, sizes ranging from 0.6-1.4cm; the diameter of the smallest tumor was 0.6cm, borders smooth and density uniform and the diameter of the largest was 1.4cm, most borders smooth, but a small part not smooth, due to the small flaky necrotic region observed in the center and some part of the tumor body protruding from the renal capsule.HE staining for the VX2 renal carcinomar model group found that cell nucleus of tumors was large, abnormal or deep dyed, ranking in cluttered, clustered or mamillary state; by the HE staining of renal tissues, the comparatively normal glomerulus and renal tubule observed were regular in structure and relatively well in cell shape, and inflammatory cell infiltration could be observed in some part.2 TDC featuresAorta TDC was divided into four segments, baseline segment, rising segment, declining segment and horizontal segment, based on the sequence. The baseline segment was relatively straight; then the curve suddenly rose to become wave crest, and the rising and declining segments showed sudden rise and sudden decline; in the end, the curve smoothly became the horizontal segment, declining gradually.Renal vein TDC: the baseline segment was relatively straight, lasting for a long time; then the wave rose to form wave crest shortly after the formation of aorta wave crest, and declined fast, giving the impression of fast rise and fast decline; the crest value of renal vein was obviously lower than that of the aorta, but its time span was longer than that of the aorta; later, the curve declined gradually, when the small recycle crest could be seen.Renal TDC in the normal control group: the baseline segment was longer than that in aortic TDC; the rising segment was comparatively steep, but gradient lower than that of the aorta; when the curve reached its crest value, it gradually rose for some period, later gradually declined, impressing us with fast rise and fast decline, and then it became the gradually declining horizontal segment, which was relatively smooth and straight.TDC of the tumor-transplanted group: similar with that in the normal group, but the start point of rising was postponed; the rising speed slowed, so the rising segment was evidently smooth and the extent lowered; in addition, the crest value appeared later than that in the normal group.3 Comparisons between the perfusion parameters of BF, BV and MTT in the normal control group and tumor-transplanted groupBF, BV and MTT in the normal control group was respectively (114.2±58.6)ml/ (min.100mg), (107.9±99.1)ml/100mg and (22.4±19.1)s, and in the tumor-transplanted group, (24.6±14.3)ml/(min.100mg), (106.7±64.5)ml/100mg and (40.9±13.1)s, respectively. P＜0.05, the reference value derived from the BF and MTT comparisons between the two groups suggested the significant difference; P＞0.05, the reference value obtained by comparing BV of the two groups suggested no evident significance.Conclusions1 Through transplantation by embedding method by incision in the posterior abdominal wall, steady rabbit VX2 renal carcinoma model can be obtained.2 When the CT perfusion parameter of BF in rabbit VX2 renal carcinoma model, is lower in that of the normal control group, M'IT shall be higher than that in the normal group; CT peffusion parameters may reflect the hemodynamic changes of VX2 renal carcinoma model to some extent, providing theoretical bases for research in human RCC.PartⅡThe study of correlations between CTPI with 64-slice spiral CT and MVD, VEGF in renal cell carcinomaObjectiveTo investigate with CTP techniques the rule of hemodynamic changes for RCC, especially the correlations between CT perfusion parameters BF, BV and MTT etc, and MVD and VEGF.Materials and Methods1 Major equipment and reagents Ditto!2 SubjectsTen cases in the normal control group: divided into two subgroups by the left kidney and right kidney.Fifteen cases in RCC group: all identified as RCC by surgical-pathological diagnoses, pathological types including eight cases of clear cell carcinoma, four cases of granulosa cell carcinoma and three cases of undifferentiated carcinoma; RCC was compared with ipsilateral peritumoral renal cortex (over 2.0cm away from the tumor border).3 The scanning program and parameters Routine scan: first the conventional middle abdominal axial plain scan was performed, then the full renal perfusion pattern scan was performed (by the set procedures, bed-board shifting in circulatory and reciprocating modes); the double-syringe power injector was employed for injection, rate being 5ml/s, first the injection of 40ml iopamidol (370mgI/ml), the non-ionic contrast agent and then the injection of 20ml saline, and 8s later after injection, scanning was started, 15 times in total. Parameters of scanning consisted of 64×0.625mm of the detector array, 5.0mm of CT slice thickness, 5.0mm of spacing, Pitchl.156, 120KV, 100mAs, 0.4s of 360-degree rotation time, 4.7-5.6s of scan interval (minimum being the default value) and 85-90s of total time.4 Image post-processing and data collectionDitto! Image drift would occur to individual subjects due to respiratory effects, so the images should be adjusted layer by layer manually through delicate adjustment, to ensure the correctness and integrity of TDC.5 Pathological examination and immunohistochemical observation5.1 Pathological preparationsThe positions of tumor tissues derived were in accordance with that of CTPI ROI. Each tissue mass was sliced into 3 pieces and CT slice thickness was 4μm. One piece of HE staining was used for conventional pathological diagnosis, one for MVD rabbit immunohistochemical staining and the other for VEGF immunohistochemical staining.5.2 MVD calculation methodsBy the evaluation criteria of Weidner etc, first the whole tomogram was observed under 100xmicroscope, to detect the high-density regions of tumor blood vessels, then four highest-density regions of blood vessels were chosen for observation under the high power microscope (200x), enumerating within 0.25μm~2 of ocular mesh micrometer scale, averaging the number of blood vessels in the four fields of vision and obtaining the average value.5.3 Standards for the preparation of specimens for VEGF examination and result analyzing In detection of VEGFs, streptavidin-biotin-peroxidase (SP) method was applied for immunohistochemical staining, and the color development kit of diamino benzidine tertrahydrochloride (DAB) was applied for VEGF detection and quantitative analysis. All color development kits were purchased from Roche Company. DAB color development, hematoxylin re-staining, and routine dehydration, transparency and mounting were compared with existed positive specimens for positive contrast. In the control group, phosphate buffer solution (PBS) was applied to replace the first antibody.The evaluation for results of VEGF expression applied the criterion of balancing the intensity of VEGF positive staining and the percentage of positive cells. First the staining intensity was graded, 0 point for colorless, 1 point for light yellow, 2 points for brown yellow and 3 points for brown and then the percentage of positive cells was graded, 0 point for negative, 1 point for (positive cells =10%), 2 points for 11%～50%, 3 points for 51%～75% and 4 points for (＞75%). The arithmetic product of staining intensity and positive percentage＞3 points, the immune reaction was positive. Results of VEGF expression were then classified into four degrees by arithmetic product scores: - (0, 1, 2 points), + (3, 4points), ++ (6, 8points) and +++ (9, 12points).6 Statistical analysesAll enumeration data were indicated by ((?)±s), analyzed by SPSS13.0 software, and the comparison between mean values of each two samples underwent t-test; comparisons between mean values of multiple samples underwent variance analysis. The comparison between classified data underwent non-parametric test, the relationships between BF of RCC parenchyma and MVD as well as VEGF underwent Pearson correlation analysis and Spearman's correlation analysis. P＜0.05 suggested the significant difference.Results1 TDC shapes1.1 TDC shapes of the aorta and renal vein: in the same TDC segment, higher than that of experimental rabbits. 1.2 Normal renal TDC shapesRenal cortex TDC rose fast and declined fast, its crest value occurred after that in the aorta; the shape of rising segment was similar with that of the aorta, but smoother and lower in amplitude; the declining segment first declined fast and slightly, and then it turned into plateau period, maintaining its steady declining.1.3 RCC TDC shapesAccording to the CTP TDC of RCC, the 15 cases with RCC were classified into two types with rich and poor blood supply, respectively, among which eight cases had rich blood supply RCC, criteria as below: the tumor body TDC shape was similar with the aorta TDC, and the amplitude and steep degree of the curve were lower than that of the abdominal aorta; seven cases had poor blood supply RCC, criteria as below: the tumor parenchyma TDC was evidently different from that of the abdominal aorta, and the amplitude and steep degree of the curve were lower than that of the abdominal aorta obviously, with slow rising segment and indistinct or even wavy declining segment.2 Comparisons between perfusion parameters in normal left and right renal cortexes BF, BV and MTI" of the left renal cortex were respectively, (182.99±25.89)ml/ (min.100mg), (109.74±13.07)ml/100mg and (9.50±3.13)s, and in the right renal cortex, (190.68±33.68 (ml/(min.100mg), (122.84±19.75)ml/100mg and (8.70±3.30)s, respectively. Comparisons between BF, BV and MTT of the left and right renal cortexes indicated P＞0.05, no significant statistical difference.3 Comparisons between perfusion parameters of the RCC group and ipsilateral peritumoral renal cortex groupIn the RCC group, BF (133.0±29.9)ml/ (min.100mg), BV (107.2±23.1)ml/100mg and MTT (15.6±6.5)s were obtained, and in the ipsilateral peritumoral renal cortex group, BF (166.1±26.0)ml/ (min.100mg), BV (84.8±31.6)ml/100mg and MTT (11.3±3.3)s were obtained, P＜0.05, indicating significant statistical difference.4 Comparisons between MVD as well as VEGF and perfusion parameters4.1 VEGF expressions in RCC and ipsilateral peritumoral renal cortex In the RCC group with 15 cases, 12 cases had positive VEGF expression, but in the ipsilateral peritumoral renal cortex group, no positive VEGF expression. After non-parametric test, we found that P＜0.05, suggesting significant statistical difference.4.2 CD34 expressions in RCC and ipsilateral peritumoral renal cortexThe measurements of MVD for the RCC and ipsilateral peritumoral renal groups under 200×microscope were respectively, 122.2±25.4 stripes and 28.2±14.7 strips. After independent sample t-test, we found that P<0.05, indicating significant statistical difference.5 The correlation between CTPI parameters and VEGF as well as MVD5.1 The correlation between VEGF expression and MVDMVD value of RCC patients with positive VEGF expression was evidently higher than that with negative VEGF expression; Spearman correlation analysis showed that VEGF expression was positively correlated with MVD (r=0.870, P=0.000).5.2 Correlations between RCC CTP parameters, BF and MTT, and MVDPearson correlation analysis for MVD and BF as well as MTT in the RCC group found that BF value of tumor parenchyma was positively correlated with MVD (r=0.565, P=0. 028) and MTT was negatively correlated with MVD (r=-0.878, P=0.000).5.3 The correlation between RCC CTP parameter BF and VEGFSpearman correlation analysis for RCC VEGF and BF found that (P=0.06＞0.05), indicating that the correlation between VEGF expression and the perfusion parameter BF was unclear.Conclusions1 Among RCC CTP parameters, BF were lower than that of the normal renal cortex, but MTI" and BV increased. CTP parameters may reflect the evident differences lying between hemodynamics of RCC tissues and the normal renal cortex.2 Among RCC CTP parameters, BF Was positively correlated with MVD and MTT was negatively correlated with MVD; in addition, VEGF had a high expression in RCC tissues, but had a low expression in the peritumoral renal cortex. Therefore, CTPI might reflect the angiogenesis situation of renal cell carcinoma.