The Study Of MRI New Techniques In Cervical Cancer Staging

Part Ⅰ Comparison of Plain and Gadolinium-enhanced MRI Imaging at3.0T for assessing cervical cancer stagingpurpose:To compare two MRI techniques at3.0T-T2-weighted and contrast-enhanced T1-weighted images-by using macroscopic sections to determine which image type enables the most accurate assessment of cervical carcinoma.Material and method:seventy-five consecutive patients who came to cancer hospital for treatment from January2010to July2013(mean age,46.8years) with biopsy-proven cervical carcinoma were included.All patients finished Plain and Gadolinium-enhanced MRI examination,then underwent radical hysterectomy and lymphadenectomy.The surgical operation process and associated pathology were recorded in minute detail. Each MRI sequence was assessed for tumour localisations,tumour margins, cancer extent and MRI staging with the consensus of two readers with more than10years’experience in gynecological oncology imaging. The readers were blinded to the clinical data and histopathological results but they were aware that each patient had at least one cervical cancer. Two reviewing sessions were performed in one-month interval to diminish recall bias. MRI findings were correlated with histopathological findings. Four-point scales assessing tumour margin definition on T2-weighted and contrast-enhanced T1-weighted images was compared using the Wilcoxon signed-ranks test. Contrast-to-noise ratios (CNRs) obtained with each image were compared using nonparametric tests. The sensitivity, specificity and accuracy of each sequence for evaluating tumour at stage1B or higher was alculated, and the difference between the two and preoperative clinical staging were assessed with paired X2test. All statistical analyses were performed using SPSS17.0. P values less than0.05were considered statistically significant.Result:Tumour margins appeared significantly more distinct on contrast-enhanced T1-weighted images than on T2-weighted images (P<0.001). The CNRs obtained using contrast-enhanced T1-weighted images were significantly higher (P<0.001) than those obtained using T2-weighted images. The accuracy of staging by contrast-enhanced T1-weighted images,T2-weighted images and the clinical examinations was85.3%,77.3%and66.7%, respectively.On Ⅰa,Ⅱ a and Ⅱb, the accuracy of staging by contrast-enhanced T1-weighted imaging is89.3%、85.3%and96%,higher than T2-weighted imaging which is84.0%、77.3%and93.3%,but not statistically significant within Paired X2test(P>0.05).Conclusion:Although Contrast-enhanced T1-weighted imaging is more accurate for cervical carcinoma staging than T2-weighted imaging,the difference is not signifancant stastically. Part II Study of Diffusion-weighted imaging in lymph node staging in patients with cervical cancerPurpose:This study aims to determine the diagnostic value of diffusion-weighted imaging (DWI) in the differentiation of metastatic lymph nodes from non-metastatic lymph nodes in uterine cervical cancer.Material and method:eighty-two consecutive patients who came to cancer hospital for treatment from January2010to July2013with biopsy-proven cervical carcinoma were included.In81patients who underwent radical hysterectomy and lymph node dissection for uterine cervical cancer, conventional MRI and DWI examinations were performed before surgery. Of the2585total dissected pelvic lymph nodes,144enlarged nodes with a short-axis diameter of5mm or greater were included for further analysis. The resected nodes were anatomically labeled in left or right internal,external, and common iliac regions by the surgeons. Each of the size-based criteria (i.e, short-axis diameter,long-axis diameter and S/L ratio) and ADC-based criteria (i.e., mean ADC, minimum ADC, mean rADC (relative ADC) and minimum rADC) were compared between metastatic lymph nodes and non-metastatic lymph nodes. Statistical analyses were performed by using statistical software (SPSS, version17.0) with the exception of the ROC analysis for the comparison of different criteria in the differentiation of metastatic from non-metastatic lymph nodes, which were performed using Medcalc for windows(version13.0; MedCalc software). Mann-Whitney test and a ROC analysis were used to compare the diagnostic performance of each ADC-based criteria on DWI and each of the conventional size-based criteria on T2WI for the differentiation of metastatic lymph nodes from non-metastatic lymph nodes. The resulting threshold values were then used to calculate the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy.Result:There were statistically significant differences between metastatic and non-metastatic lymph nodes in the short-axis diameter, long-axis diameter,S/L ratio, mean ADC, minimum ADC, mean rADC and minimum rADC (P<0.05). The Az of the minimum ADC (0.906) was greater than that of the other ADC-based criteria,(0.865,0.832,0.830for minimum rADC,mean ADC,mean rADC, respectively)and all size-based criteria(0.856,0.780,0.652for short-axis diameter,long-axis, S/L Ratio,respectively)(P<0.05). Using the minimum ADC criteria (≤0.923×10-3mm2/s), the sensitivity and specificity for differentiating metastatic from non-metastatic lymph nodes were95.1%and82.5%, respectively. On a single node basis, the sensitivities for the detection of malignancy from ADCmin and S/L ratio were95.1%and92.7%, both higher than obtained fromADCmean(82.9%),rADCmean(70.7%),rADCmin(87.8%),short axis (85.4%), long axis (61%), but the specificity of S/L ratio was too low (38.8%). The accuracy of ADCmin(86.1%).Conclusion:DWI is feasible for differentiating metastatic from non-metastatic pelvic lymph nodes in patients with uterine cervical cancer and minimum ADC could be served as a representative marker.