Clinical Study Of Diffusion Weighted Imaging And Dynamic Contrast-enhanced MRI In Histopathologic Grade Of Hepatocellular Carcinoma

1. Purpose:(1) To evaluate whether administration of gadolinium-based contrast material significantly affects diffusion-weighted imaging(DWI) and apparent diffusion coefficient(ADC) values at the focal hepatic lesions.(2) To evaluate the relationship between the signal intensity of hepatocellular carcinoma (HCC) assessed with diffusion-weighted imaging(DWI) and the apparent diffusion coefficient (ADC) with the histopathologic grade of HCC.(3) To evaluate the diagnostic value of dynamic contrast-enhanced MRI(DCE-MRI) in predicting histopathologic grade of HCC by analysis of time of intensity curve(TIC) types and semi-quantitative parameters.2. Materials and methods:(1) Study population①From November 2008 to June 2009,26 patient, including 19 males and 7 females, who were diagnosed as focal hepatic lesions. The age range was 37-67 years, with the average age of 54.50±9.12 years. The diameter of the lesions ranged from 2.1-18cm in the maximum diameter(mean,7.17±4.03 cm).②From November 2008 to March 2010,27 patients, including 24 males and 3 females, who were diagnosed as HCC. The age range was 36-38 years, with the average age of 55.90±9.52 years. The diameter of the HCCs ranged from 2.0-18cm in the maximum diameter(mean,6.19±3.10 cm). (2) MRIMRI was performed with a 3.0T system (GE Signa Excite, GE Healthcare, Milwaukee, WI) with a phased-array multicoil for the body. The MRI protocols included the T1-weighted, T2-weighted, DWI and 3D-FAME dynamic contrast-enhanced imaging. DWI was obtained using single-shot echo planar imaging with b value of 600 s/mm2.(3) Image analysis①By copying and pasting ROIs, the SI value, the SDnoise value and the ADC value were measured. SNR and CNR of each lesion was calculated using the formula SNR=SInormal/SDnoise, CNR=| Sllesion-SInormal|/SDnoise.②Visual assessmentThe same two radiologists visually assessed the signal intensity of all tumors and classified them into isointense, slightly hyperintense or obviously hyperintense by mutual agreement.ADC value and SI value of each tumor were measured.③TICs were classified into three types.MSI and SER were derived from TIC.The MSI value and SER value of HCCs were measured.(4) HistopathologyThe differentiation of an HCC was classified into well, moderate and poor according to Edmondson-Steiner's grading system. When different tumor grades coexisted within a tumor, the more predominant differentiation of the tumor was selected.(5) Statistical analysisAll the statistical calculations were performed using the statistical software SPSS16.0. The results of measured values were recorded in the format of mean±standard deviation'. p< 0.05 was considered statistically significant in all the statistical analyses.SNR, CNR and ADC values obtained on pre-contrast and post-contrast were analyzed using Student's paired t test. The difference of the ADC value, SI value, MSI value and SER value was compared with the histopathologic grade using the one-way ANOVA and LSD test. ROC analysis was used to analyse the diagnostic value of ADC value in differentiating poorly-differentiated HCC from well- and moderately-differentiated HCC. The optimal cutoff point of ADC value and the correspongding sensitivity and specificity were determined.Fisher's exact test was performed to assess the statistical difference between the histopathologic grades of HCC and visually evaluated signal intensity on DWI.3. Results:(1) The SNR and CNR of the DWI were significantly increased after contrast medium injection (t=2.356, p=0.036; t=3.170, p=0.008). ADC values of the focal hepatic lesions significantly decreased after contrast medium injection(t=2.717,P=0.019).(2) Results of histological gradesFor the histological differentiation,25 HCCs had a single histological grade and 2 HCCs had two different histological grades. Finally, well-differentiated HCCs were in 6 patients, moderately-differentiated HCCs were in 10 patients and poorly-differentiated HCCs were in 11 patients.(3) Results of visual EvaluationsOn DWI, lof 27 HCCs appeared isointense,9 tumors appeared slightly hyperintense, and the remaining 17 tumors appeared obviously hyperintense to the surrounding liver. Overall,96.3% (26/27) of HCCs showed hyperintensity to the surrounding hepatic parenchyma. Fisher's exact test revealed no significant difference between visual signal intensity and histopathologic grade (p=0.567).Results of ADC value and SI value Measurements There was a significant difference in the ADC values among the well-, moderately-and poorly-differentiated HCCs (F= 5.921, p= 0.008, one-way ANOVA test). The ADC value of the poorly-differentiated HCCs was significantly lower than that of the well-and moderately-differentiated HCCs (p= 0.004, p= 0.023, LSD test). There was no significant difference between the ADC value of the well- and moderately-differentiated HCCs(p= 0.281, LSD test). ROC analysis showed that the optimal cutoff point of ADC value in diagnosing poorly-differentiated HCC was 1.1450×10-3mm2/s. A cutoff ADC value equal to or less than 1.1450×10-3mm2/s was used to differentiate poorly-differentiated HCC from well-and moderately-differentiated with a sensitivity of 100% and specificity of 54.5%.There was a significant difference in the SI values among the well-, moderately-and poorly-differentiated HCCs (F= 4.161, p= 0.028,one-way ANOVA test). The SI value of the well- and poorly-differentiated HCCs was significantly higher than that of the moderately- differentiated HCCs (p= 0.025, p= 0.019, LSD test). There was no significant difference between the SI value of the well- and poorly-differentiated HCCs(/p= 0.791, LSD test).(4) Patterns of TICTypeⅠandⅡcurves were most seen in well- and moderately-differentiated HCCs(14/16). TypeⅢcurve was most likely seen in poorly differentiated HCCs (6/11)Results of MSI and SER MeasurementsThere was a significant difference in the MSI values among the well-,moderately-and poorly-differentiated HCCs (F= 4.071, p= 0.030,one-way ANOVA test). The MSI value of the well-differentiated HCCs was significantly higher than that of the moderately- and poorly-differentiated HCCs (p= 0.019, p= 0.014, LSD test).There was no significant difference between the MSI value of the moderately- and poorly-differentiated HCCs(p= 0.934, LSD test).No statistical difference was found in the SER values among the well-moderately-and poorly-differentiated HCCs (F= 2.758, p= 0.084,one-way ANOVA test).4. Conclusions:(1) The SNR and CNR of the DWI were significantly increased after contrast medium injection. Therefore, the lesion was easily to appeared. ADC values of the focal hepatic lesions significantly decreased after contrast medium injection. We suggest the DWI should be performed before contrast medium injection.(2) There was a significant difference in the ADC values among the well-, moderately- and poorly-differentiated HCCs. The ADC value of the poorly-differentiated HCCs was significantly lower than that of the well- and moderately-differentiated HCCs. Therefore, DWI with ADC measurement may be a valuable tool for noninvasively predicting the differentiation of HCC. Although correct prediction of the histopathologic grade of HCC is not possible because of the large overlap among the ADC values.(3)DCE-MRI could reflect the patterns of hemodynamic changes among the different histopathologic grade of HCCs. TypeⅠandⅡcurves were best seen in well-and moderately-differentiated HCCs. TypeⅢcurve was most likely seen in poorly differentiated HCCs. Patterns of TIC and semi-quantitative parameters obtained from DCE-MRI were useful in differenciating histopathologic grade of HCCs.