| Background and ObjectiveGastric carcinoma is the most common malignant tumor of digestive tract in China. Gastrointestinal stromal tumor is the most common gastrointestinal non-epithelial tumor, and the incidence increased year by year. In recent years, with the rapid development of imaging techniques, spiral CT scanning has become the preferred preoperative routine examination of gastric cancer and gastric stromal tumors, and also widely used in the preoperative TNM staging of gastric cancer and the benign and malignant diagnosis of gastric stromal tumors. The concept of perfusion CT is first introduced in early 90's, which developed the imaging technology from conventional morphologic diagnosis to functional evaluation. Multi-slice spiral CT perfusion imaging as a new functional imaging technology, through displaying changes of the contrast medium concentration in the tissues and organs, indirectly reflects the perfusion characteristics of tissue and organ. Nowadays, CT perfusion imaging has been widely used in the research of head, neck, lung, liver, pancreas, kidneys and other organs. And the clinical applications mainly focus on the diagnosis, treatment, curative effect evaluation and prognosis monitoring of acute stroke and tumor diseases.Because of the gastric anatomical structure and physiological characteristics, there have few specific reports of CT perfusion in stomach, and most of the studies are still in the preliminary exploration stage. In this study, through the study of 64-slice spiral CT perfusion imaging in normal stomach, gastric cancer and gastric stromal tumors, and to approach the clinical value of MSCT perfusion imaging in the stomach by the following two aspects. First, analyze the perfusion characteristics and perfusion parameters values of normal gastric tissue, gastric cancer and gastric stromal tumor, and compare the differences of each other, to approach the clinical value of CT perfusion imaging in diagnosis and reflecting neoplastic perfusion features. Second, analyze the differences of the perfusion parameters between differentiated gastric cancer and undifferentiated gastric cancer, the different general types of gastric carcinoma, and the different differentiated degree of gastric stromal tumor, to investigate the application value of CT perfusion imaging in reflecting the biological behavior of tumor.Materials and methods1. Clinical datas and cases for groupSeventy-three cases gastric carcinoma and 43 cases gastric stromal tumor from June 2009 to December 2010 which were confirmed by pathology were collected. Among 73 cases of gastric cancer, there were 55 cases for men and 18 cases for women, and their age range was between 23 and 77, average at 56.90. There were 57 cases of cardial carcinoma,6 cases of gastric body carcinoma,8 cases of gastric antrum carcinoma and 2 cases of linitis plastica. The cases were divided into two groups according to the general classification:28 cases for localized gastric cancer group (including 7 cases of Borrman I type and 21 cases of Borrman II type), and 45 cases for infiltrative gastric cancer group(including 43 cases of Borrman III type and 2 cases of Borrman IV type). According to the pathological differentiation, all gastric cancer cases were divided into two groups:30 cases for differentiated group (6 cases of well differentiated adenocarcinoma and 24 cases of moderately differentiated adenocarcinoma), and 43 cases for undifferentiated group (34 cases of poorly differentiated adenocarcinoma, signet-ring cell carcinoma in 5 cases and 4 cases of mucinous adenocarcinoma). Among 43 cases of gastric stromal tumor, there were 20 cases for men and 23 cases for women, and their age range was between 40 and 69, average at 51.67. According to the pathological extent of tumor cell differentiation of gastric stromal tumors, they will be divided into two groups:17 cases forⅠgroup (low potential malignancy in 17 cases),26 cases forⅡgroup (moderate malignancy and high malignancy in 26 cases). All the cases were not receiving any preoperative chemotherapy, and all cases underwent 64-slice spiral CT gastric perfusion scanning within one week before surgery. Besides of these, we also collected 10 cases of normal stomach, in which 7 males and 3 females, aged 26 to 70 years, mean age 57.00 years, all cases underwent gastric CT perfusion scanning. All cases were to sign CT perfusion scanning informed consent before scanning.2. Scanning methodSubjects to be fasting 8-12h or more, and were injected 654-2 10mg at 10～15min before the scanning, and were asked to drink water 800～1200ml. Breath-hold training exercises in patients before the examination, all subjects taken supine position, using GE Lightspeed 64-VCT for perfusion scanning. Specific operation is as follows:Firstly,all subjects were underwent the routine abdominal plain scan, the thickness and spacing:5mm, pitch:0.984:1, the scanner speed:0.8sec. According to the plain scan image to determine 8 target layers. Perfusion scan mode: axis-scan mode, tube rotation speed:1s, thickness:5.0mm, spacing 0,120kV/80mA, the total exposure time is 25s. All cases were injected non-ionic contrast agent (Ioversol 320mgI/ml), and the injection rate is 4～5ml/s, the amount of contrast agent is 48～50ml. All subjects conventionally took abdominal belt and held breath for 50s during scanning. The aged people who couldn't hold breath for a long time were asking to take quiet breathing. The day of perfusion end, the patients were injected 80～100ml non-ionic contrast agent at the speed of 2.5～3.0ml/s to complete the enhanced CT scan. The scanning time were started after injection of contrast agent 30s,65s and 120s, to complete arterial phase, venous phase and delayed phase.3. Data processing and analysisThe perfusion data were sent to the GE ADW4.3 workstations, using the GE CT perfusion-3 perfusion package of pancreas perfusion protocol to calculate, map, analyze and generate the corresponding perfusion parameter values. Select the abdominal aorta as the arterial input, the size is 2～6 pixels to avoid partial volume effects. Then defined the threshold, usually is 0-200HU, removing the influence of bone, fat, air and other tissues. Finally, selected the regions of interest (ROI), the selection of ROI should be the right size, it usually selected more than 50 pixels, and the region of interest shouldn't extend out the edge of the lesion, and also should avoid the large blood vessels and cystic necrosis. Through the selected regions of interest to obtain the corresponding time-density curves (TDC), perfusion imaging and to further calculated various perfusion parameters. Main perfusion parameters include:blood flow (Blood Flow, BF), blood volume (Blood Volume, BV), mean transit time (Mean Transit Time, MTT), permeability surface (Permeability Surface, PS). Measurement of all perfusion parameters were repeated for 3 times and chosen the average value as a result.All data were handled by SPSS 15.0 statistical package. Measurement data demonstrated inχ±s and the reference range with 95% confidence interval. Comparison between two groups was done using independent sample t-test. Significance level a=0.05, when p<0.05 represented statistical differences, when p <0.01 represented significant statistical differences.Results1. TDC of 10 cases of normal gastric wall:it can be divided into three parts, flat baseline segment, rising slowly ascending and down slowly horizontal section. TDC of 73 cases of gastric cancer can be divided into three types:Ⅰtype (41 cases), the initial section showed a slow rise, to reach the peak, keeping the peak levels trend until the end of perfusion.Ⅱtype (27 cases), the initial section of perfusion scan showed a rapid rise, to reach the peak, and keeping the peak levels of horizontal movements until the end of perfusion.Ⅲtype (5 cases), the initial section rise rapidly to get the peak value, and after a short phase of rapid decline forming a crest, then remained in platform stage until the scanning was ended. TDC of 43 cases of gastric stromal tumor can be divided into two types:Ⅰtype (36 cases), the initial segment rise slowly, and reach the peak showing a slow decline, forming a crest, then showed a horizon trend to the end of perfusion.Ⅱtype (7 cases), the initial segment showed a slight upward trend to a peak level, and then remained the level of the peak until the end of perfusion scan.2. The BV and PS maps of gastric carcinoma showed red and yellow in the part of the lesion, which reflected changes of high blood perfusion and high permeability. The BV and PS maps of gastric stromal tumors showed red and yellow on blue background, which demonstrated that the blood supply and permeability of the lesion are increased, but compared with gastric cancer, the blood supply and vascular permeability of gastric cancer are higher than gastric stromal tumors'.3. The mean BF, BV, MTT and PS of the group of normal stomach were 43.66±10.44ml/100g/min,7.82±2.52ml/100g,14.82±6.89s,8.06±2.15ml/100g/min.The mean BF, BV, MTT and PS of the group of gastric carcinoma were respectively 121.6±61.06ml/(100g/min),10.45±7.09ml/100g,6.89±3.59S,26.31±14.99ml/(100g.mi) .The BF, BV, MTT and PS of the group of gastric stromal tumors were respectively 74.83±18.98ml/100g/min,8.05±2.71ml/100g,10.23±2.67s,15.03±5.32ml/100g/min. he perfusion parameters BF, MTT and PS had statistical significance between normal stomach and gastric carcinoma, also normal stomach and gastric stromal tumors (p<0.05). And the perfusion parameters BV had no statistical significance(p>0.05).4. The mean BF, BV, MTT and PS of the group of localized gastric carcinoma were respectively 105.30±40.28ml/100g/min,10.50±4.33ml/100g,8.02±4.86s, 19.08±4.55 ml/100g/min. The mean BF, BV, MTT and PS of the group of infiltrative gastric carcinoma were respectively 123.56±56.88ml/100g/min,13.15±6.40ml/100g, 9.98±5.65s,27.84±5.77ml/100g/min. The perfusion parameter PS had significant statistical differences between localized gastric carcinoma and infiltrative gastric carcinoma (p=0.002), and BF, BV, MTT had no statistical differences (p>0.05)5. The mean BF, BV, MTT and PS of the group of differentiated gastric carcinoma were respectively 118.49±36.35ml/100g/min,11.35±3.42ml/100g,6.89±3.10s,20.68±5.26ml/100g/min. The mean BF, BV, MTT and PS of the group of undifferentiated gastric carcinoma were respectively 128.15±62.23ml/100g/min, 13.30±7.48ml/100g,8.75±6.31s,28.79±6.12ml/100g/min. The perfusion parameter PS had significant statistical differences between differentiated gastric carcinoma and undifferentiated gastric carcinoma (p=0.000), and BF, BV, MTT had no significant statistical differences (p>0.05)6. The BF,BV,MTT and PS of I group of gastric stromal tumors were 77.28±25.57ml/(100g/min),7.91±2.84ml/100g,7.37±2.38s,13.99±4.06ml/(100g.min), and the BF, BV, MTT and PS of II group of gastric stromal tumors were 95.13±29.75ml/(100g/min),8.21±2.65 ml/100g,7.33±1.83s,14.15±4.04ml/(100g.min). There was no correlation between the different pathological differentiations of gastric stromal tumors in BV, BF, MTT and PS. The perfusion parameters BF, BV, PS had significant statistical differences between gastric carcinoma and midrange and high malignant gastric stromal tumors (p<0.01). No significant statistical differences were found in mean MTT between gastric carcinoma and moderate malignant gastric stromal tumors and high malignant gastric stromal tumors.Conclusions1.64-slice spiral CT perfusion imaging reflects the hemodynamic characteristic of normal stomach and stomach neoplasm on microcirculatory levels, and provides more quantitative information for the diagnosis of gastric carcinoma and gastric stromal tumors on functional aspects.2. The perfusion parameter PS value of gastric carcinoma is useful to detect the general classification and differentiation degree, and the BV, BF and PS values between midrange and high malignant gastric stromal tumors and gastric cancer had statistical differences, and reflected the differences of blood flow perfusion between malignant gastric stromal tumors and gastric cancer, provided significant clinical value on the diagnosis and differential diagnosis.3. CT perfusion imaging, as a new functional imaging technique, to some extent make up the deficient of traditional morphological imaging. And it will be feasible in the study of clinical treatment, the assessment of curative effect and prognostic monitoring. |
PDF Full Text Request