| Objective: To investigate the value of the 64-slice Spiral CT imagine with virtual colonoscopy in diagnosis on preoperative colorectal cancer staging and in surgical forecast evaluation.Subjects and Methods :1. 74 cases who had diagnosed as colorectal cance were collected from October 2008 to October 2008. 2. All the patients were confirmed by colonoscopy and biopsy. The scan was done by GE lightSpeed 64-slice spiral CT machine(64-SCT) and all the 74 patients were given the enhanced scan. Routine bowel preparation was done before the examine, intramuscular injection has been done with 20mg ofanisodamine(654-2) 10 minutes before the scanning to reduce the tension of colon and spasms of colon .It also can reduce the peristalsis artifacts. The anal canal or Foley insertion was completed when the patients lying on the right side, then about 1000～1500ml gas were injection into the anal canal, the specific value is based on the degree that the patient's tolerance. The volume of the whole belly was scaned(ranging from top of the diaphragm to the lower edge of pubic symphysis). Scan parameters: 140 kV ,300～600 mA, The collimated slice thickness was 64×0.625 mm , the speed of going into to bed was 39 mm , the pitch was 0.98, the pitch speed of X-ray tube was 0.6 s/ r . 70～90 ml non-ionic contrast agent(iohexol)3(50 mg/ ml ) were injected into the cubital vein by high-pressure syringe when do the enhanced scanning. The injection rate was 3～5 ml/ s, the delay time was in the 25～30 s of arterial phase or 55～70 s of intravenous phase. San data then passed to the workstation 4.2 to the image reconstruction and post processing. 3.The workstation 4.2 and dedicated CTVE processing software were used to do the post processing for the data .The post processing includes CT virtual endoscopy(CT virtual endoscopy, CTVE), shaded surface imaging(SSD), transparent technology(RaySum), multi-planar reconstruction(MPR), volume rendering(VR) technology and other post-processing method to obtain a variety of different images, We do the post processing to analysis comprehensively, to record the preoperative CT-TNM stage and to plan the surgical programs so we can do the comparision with the pathological staging after the operation and with the actual operation program. 4. Colorectal cancer staging were according to the latest UICC/AJCC staging manual (seventh edition) for pathologic stage. The pathological type of tumor, the intrusive degree around the intestine, the length of cancer were assessed respectively by two experienced radiologists and pathologists blindly before and after operation. The preoperative clinical stage and pathologic TNM stage after the operation were compared by 64-SCT. SPSS 13.0 software was used for statistical analysis. The date analyzed byχ2 test, Kappa test was applied when the observations were consistent. We calculated the accuracy rate, sensitivity, specificity, positive predictive value and negative predictive value for the correctness of the test indicators when analysing. Test level: p=0.05.Results:1. General information: There were 48 male cases, 26 female cases aged from 30 to 85, the median age was 67.0, the average age was 65.34±12.4 of all the cases including 43 cases of colon cancer, 4 cases of rectosigmoid junction cancer and 27 cases of rectal cancer.2. The comparison of 64-SCT staging and pathological staging after operation.(1). T staging: According to the latest international stage the T stage is divided into six levels. Compared with the pathological staging the sensitivity of Tis, T1, T4b for T staging were all 100%, the sensitivity of T2 stage was lower(66.7%). The overall accuracy rate of T stage was 87.8%. It showed by the diagnostic consistency test that preoperative T staging were appreciated coincidently with pathological T staging(κ=0.818,P=0.000). There's no significant difference in accuracy between preoperative 64-SCT T staging and pathologic T staging(Z=-0.333,P=0.739). (2) N staging: N stage is divided into six levels. Compared with the postoperative pathology it indicated that the sensitivity of N0,N1c,N2b for N staging were higher which were 88,1%,100%,100% respectively and the N1a,N1b,N2a for N staging were lower (58.3%,60%,66.7% ). The overall accuracy rate of N stage was 82.4%. It showed by the by the diagnostic consistency test thatκ=0.716, P=0.000. It indicated that there's no significant difference between the accuracy of preoperative 64-SCT staging and postoperative pathological staging of N stage by the rank sum test (Z=-0.178,P=0.858). It showed that the accuracy rate of lymph node which diameter≥5mm was higher(89.5%) by comparing the preoperative lymph node metastasis with different diameters with postoperative pathologic results. The sensitivity was 89.5% and the specificity was 80.0%. It showed by the univariate and multivariate analysis for the clinicopathological factors of lymph node metastasis that the lymph node metastasis was highly correlated with the tumor location, histological type, histological grade and lymphatic invasion. Logistic regression model was applied to analysis. There were 3 risk factors in the logistic regression model which were tumor location, histological grade and lymphatic invasion. We determine the effect according to the demarcation point when prediction probability was 0.5. The overall rate which can judge correctly was 71.6%.(3). M staging: The sensitivity and accuracy rate for M stage were higher(100%,98.6%). It showed by the by the diagnostic consistency test thatκ=0.951,P=0.000. It indicated that there's significant difference between the accuracy of preoperative 64-SCT and postoperative pathological staging of M staging (Z=-1.000,P=0.317).(4). The overall staging of TNM: according to the latest clinical staging of colorectal cancer there're four phases ,and we stratified according to the sub-period. The overall accuracy of cTMN was 78.4% which were appreciated coincidently with the pathological TNM staging(Kappa=0.740,P=0.000).3. It indicated by the post-proccessing technique results of three- dimensional reconstruction of 64-SCT that the accuracy was high for T staging when using CTVE(70/74). RaySum can display the length of lesion, range(74/74) accurately , but have a low prediction accuracy for the T-staging(50/74). MPR can show the degree of violation around the intestine, length of tumor lesion, relations or tumor and the organs surrounding it accurately, it also has a high accuracy in T and N staging(68/74, 70/74). MRP and MRP-CTVE technique were all have high accuracy in violation around the intestine (74/74) and lesion type(72/74).4. It indicated by the surgical way that by the preoperative CT comprehensive CT staging and three dimensional imaging prediction that in all the cases 2 the rectal cancer need the surgical way of Mile's , 1rectal cancer need the surgical way of Dixon and 1case of sigmoid colon cancer need the surgical way of sigmoid resection. Finally rectal and colonic local excision +125I radioactive seed implantation were done when we found the cancer invasion and metastasis which is in curable by the laparotomy. For other cased, there're good consistency by prediction, the diagnostic consistency test showed thatκ=0.936,P=0.000。Conclusion1. It can achieve high accuracy and can predict the transfer factors of lymph node metastasis accurately by compare the TNM stage of 64-SCT imaging with postoperative pathological staging. Also it's a good guidance for the surgical approach and treatment options and prognosis.2. The post-processing technique of 64-SCT three-dimensional reconstruction can compensate the shortconings of colonoscopy effectively and can do a more accurate staging for colorectal cancer. Certain imaging methods had high accuracy for the TNM staging. So it is an important basis for clinical treatment strategies.3. TNM staging imaging depends on the skilled post-processing technology of 64-MSCT reconstruction which is a integrated application results. |
PDF Full Text Request