| PurposeThe aim of this study was to investigate the value of the Multi-slice CT (MSCT) indiagnosing gastrointestinal stromal tumors (GIST) before operating. To improve thecognitive ability about diagnose and differential diagnose of GIST. To analyze thecorrelation between site, size, shape, border and so on and the clinical risk of GIST, inorder to provide significant reference information about clinical tumors grading andtreatment plan.Materials and Methods1. Subjects The data of all the95cases of GIST in Shandong university Qiluhospital were collected from January2013to January2014, including50males and45females, aged20-79years old, average aged of54.3. The clinical symptoms includedabdominal pain or discomfort, anorexia or vomit, gastrointestinal bleeding and black stools,height losing, abdominal mass, progressive dysphagia, intestinal obstruction and so on. Allof the patients have had surgery treatment and have been confirmed by clinical pathologyand immunohistochemical makers as GIST.2. Pathological results Gastroenterology specialists determined the degree ofrisk of tumors based on tumor size, cell mitosis and cell dysplasia and so on. Grade oftumors was established by using United States National Institutes of Health (NIH) standardin2002. GISTs were classified as high risk group, medium risk group, low risk group andthe extremely low risk group. The extremely low risk groups which had similar clinicalbehavior with low risk groups, to facilitate the study, were merged into the low risk groups.3. CT Image Analysis All CT images of the95patients were independentlyanalyzed by two experienced radiologists, who were blinded to the clinical details. Therelationship between imaging findings and clinical pathology of gastrointestinal stromal tumors was evaluated in details.(a) Accurately positioning gastrointestinal stromal tumors in the axial, sagittal andcoronal images by Multi-planar imaging technique (MPR): the sites of the gastrointestinalstromal tumors were divided into in gastrointestinal and out of gastrointestinal, followedby organs accurately locating. Analyzed blood supplies of gastrointestinal stromal tumorsby using maximum intensity projective technology.(b) Summarized the general CT imaging characteristics of gastrointestinal stromaltumors. Further more, analyzed the relationship between CT images and pathologicalinformation of tumors. CT evaluation factors include: tumor site; shape or border; size ordiameters; tumor morphology and growth pattern; density; tumor enhancement patterns; fatspace around the tumors and surrounding organs.4. Statistical analysis Statistical analysis was performed by using the commercialSPSS16.0. The statistical differences between the risk of tumors and CT imagingcharacteristics, such as the site; size or diameters; shape or border; tumor growth patternand morphology; density (particular density include bleeding, calcification, cystic necrosis,etc.); tumor enhancement patterns; fat space around the tumor and surrounding organswere tested by using chi-square or Fisher’s exact test. Then with low risk, medium risk andhigh risk of tumors as the dependent variability (Y), there being differences incharacteristics as independent variability on the risk of tumor were more variable stepwiseLogistic regression analysis. P<0.05, there was significant difference between them.ResultPathological information confirmed there were2cases tumors of extremely low risk,27cases of low risk,19cases of moderate risk and47cases of high-risk. The vast majorityof tumors were in the gastrointestinal tract. There were48cases tumors originated instomach, including15cases in gastric fundus,18cases in gastric body,5cases in antrum;5cases originated in the esophagus;11cases originated in the duodenum cases;16caseslocated in the jejunum and ileum;2cases sited in colon;6cases sited in the rectal. Few oftumors were located outside the gastrointestinal tract. There were two cases located in theprostate gland, one case located in the left corner of the uterine, two cases originated in thevaginal, one case sited in omental sac and one case located in mesenteric. The tumorswhich were outside the gastrointestinal tract were all of high risk. The number of tumorsvaried from a single focal to multi-focal growth.The sizes of tumors were varied the smallest diameter of about1cm from the largestof about32×33×10cm. There were63cases larger than5cm and32cases smaller than 5cm. There were39cases growing intraluminal,30cases of extraluminal growth, one caseof wall-type growth,25cases of mixed-type growth.43cases of tumors had regular shapeand neat edges,52cases of tumors had irregular shape, part of which had lobulated edges.There were33cases with uniform density, and62cases with uneven density, part of whichhad visible calcification or hemorrhagic or necrosis liquefied zone.34cases of tumorsenhanced homogeneously and61cases of tumors enhanced heterogeneously. Thosenecrosis liquefied zones were not enhanced. There were62cases having clear fat spacearound the tumors and33cases having surrounding infringement, involved the gapbetween fat and tumors or surrounding organs. The reactive hyperplasia of lymph nodewas discovered in10cases of tumors but lymph node metastases were not confirmed.Analyzed by using the chi-square test, we can see the size of tumor, morphology,density of tumor, fat space, growth pattern, strengthen pattern of tumors had statisticallydifference from the risk of tumors. Multivariate stepwise linear regression analysis, therewas statistically significant difference between growth pattern, size and shape, cysticnecrosis and the degree of risk of tumors.ConclusionMSCT can be considered an essential tool for the diagnosing and staging of GIST. Itcan reveal some special characteristics of the gastrointestinal stromal tumors, which canhelp locate and determine the nature of the tumors before operation. MSCT has a veryimportant value for differential diagnosis of gastrointestinal stromal tumors, and candetermine the risk of GISTs, which can provide significant value for the clinicalclassification and treatment. |
PDF Full Text Request