The Value Of Diffusion Weighted Imaging And Dynamic Contrast-Enhanced Imaging In Diagnosing Prostate Cancer And The Correlation Between The Two Methods At3T MR

Part I The Value of Diffusion Weighted MRI in the Diagnosis and Assessing the Aggressiveness of Prostate CancerPURPOSE:1. To evaluate the value of diffusion weighted imaging (DWI) in prostate cancer at3T MR, and to investigate the proper ADC value threshold to differentiate cancerous and benign prostatic tissues.2.To retrospectively determine the relationship between apparent diffusion coefficients (ADCs) obtained with3T diffusion-weighted (DW) magnetic resonance (MR) imaging and Gleason grades in prostate cancer.MATERIALS AND METHODS:99cases with prostatic carcinoma and23cases with BPH nodules were included in the study. All cases were proved by sextant biopsy and/or prostatectomy after MR examinations. The DWI were obtained with b-values of0,800,1000s/mm2. All cancerous lesions from99cases with prostatic carcinoma, together with all BPH nodules and normal PZ tissues from all cases were chosen. Then ADC values of the three kinds of tissues were compared using one-way ANOVA. The difference of ADC values between PZ cancer and CG cancer were compared using independent sampled t test. ROC analysis was used to investigate the proper ADC value threshold to differentiate cancerous and benign prostatic tissues. According to pathological Gleason scores (GS), tumors were classified into3qualitative grade groups as low-grade (GS3+3), intermediate-grade (GS3+4), or high-grade (GS4+3,3+5,4+4,4+5,5+4) tumors. The ADCs of the three groups were compared with Kruskal Wallis method. And the ADC in the tumors was related to Gleason score or qualitative grade groups with Spearman correlation analysis. Then ROC analysis was used to investigate the ability for ADC values to differentiate high-grade group of prostate cancer from low-and intermediate-grade group.RESULTS:In99cases with prostate cancer,130cancerous foci were found,77in peripheral zone, and53in transition zone. Significant ADC differences were noted among the three tissues (F=921.332, P<0.001), despite an inter-patient overlap. ROC analysis revealed a discriminatory performance in discerning tumor and benign prostatic tissues of AUC=0.989for ADCs (with a threshold of1.14×10-3mm2/s, and sensitivity of92.5%and specificity of97.1%). According to pathological Gleason scores (GS),130cancerous foci were classified into3groups, with13low-grade tumors,33intermediate-grade, and84high-grade. The mean and standard deviation of the ADCs were(0.92±0.16)×10-3mm2/s (low-grade group),(0.91±0.16) X10-3mm2/s (intermediate-grade group) and (0.84±0.14) X10-3mm2/s (high-grade group). There was a weak correlation between ADCs and Gleason score (r=-0.229, P=0.009), and Gleason grade group (r=-0.200, P=0.023). And significant differences of the ADCs were seen among the three groups (F=3.793, P=0.025). ROC analysis showed an intermediate ability for ADC value to differentiate high-grade group of prostate cancer from low-and intermediate-grade group, with AUC of0.620.CONCLUSIONS:DWI plays an important performance in discriminating cancerous and benign prostatic tissues both in peripheral zone and central gland. The ADC value less than1.14X10-3mm2/s was a preferable threshold to differentiate prostate cancer from non-cancerous tissue in our study. ADC values at3T MR had advantages to achieve a discriminatory performance in the differentiation of low-, intermediate-, and high-grade cancer, while the same results could not be seen in the differentiation of Gleason Grade, thus ADC values would be helpful for prediction of the tumor differentiations. However, an ideal threshold was hard to set off to discerning the different Gleason Grade Group of prostate cancer. Part ? The Value of dynamic contrast enhanced MRI (DCE-MRI) in the Diagnosis of Prostate Cancer and the correlation between DCE-MRI and DWIPURPOSE:To evaluate the diagnosis value of dynamic contrast enhanced MRI (DCE-MRI) and the correlation between DCE-MRI and DWI.MATERIALS AND METHODS:Fifty-two patients with prostatic carcinoma in Part ? were examined both by DCE-MRI after diffusion-weighted imaging (DWI) at3T MR. Regional analysis of cancerous and benign prostatic tissue were delineated based on the sextant biopsy histopathology. And then, several semi-quantitative parameters on DCE-MRI, including maximum signal intensity(SImax), time to peak(TTP), relative maximum signal intensity(rSImax), the maximum enhancement ratio (SImax%) and washout were calculated and compared with one-way ANOVA method among the3types of prostatic tissues and3groups of cancerous tissues (except TTP for Mann-Whitney U test). ROC analysis was used to investigate the best SI-T parameters and proper threshold to differentiate cancerous and benign prostatic tissues. Additionally, correlations between Gleason score or Gleason grade group and each semi-quantitative parameter with Spearman correlation analysis, and correlations between ADC and each semi-quantitative parameter in the three different tissues were evaluated with Pearson correlation analysis.RESULTS:173cancerous ROIs were drawn in55patients with prostate cancer,91in peripheral zone, and82in central gland. Despite a considerable inter-patient overlap, all SI-T parameters among3types of prostatic tissues had significant differences, except SImax%between cancerous tissues and BPH nodules, P=0.699. However, no significant differences were seen in all SI-T parameter between PZ cancer and CG cancer(P>0.05). ROC analysis revealed a discriminatory performance of AUC=0.802for rSImax in discerning tumor and benign prostatic tissue (with a threshold of98.7, the diagnostic sensitivity was77.2%, and specificity was71.7%). In cancerous tissues, there were significant positive correlations between ADC and SImax (r=0.375, P<0.001) or SImax%(r=0.322, P<0.001), while significant negative correlations were seen in BPH nodules and normal peripheral zone. No significant correlation was found between other DCE-MRI parameters and ADC in all prostatic tissues. In cancerous tissue, no significant correlation was found between all of the SI-T parameters and Gleason Grade. And no significant correlation was found either between all SI-T parameters and Gleason grade groups except rSImax, with a significant positive correlation to Gleason Grade Group (r=0.284, P=0.020). And there were no significant differences of all of the SI-T parameters between the three Gleason Grade Groups.CONCLUSIONS:rSImax plays discriminatory performance between cancerous and benign prostatic tissues in PZ, when rSImax is not available, SImax could be a good choice. While in CG, the best parameter is TTP, and the second is washout. To differentiate the grade of prostate cancer, there are no significant advantages for all SI-T parameters. However, significant different types of correlations exist between DWI and DCE-MRI of cancerous and benign prostatic tissues, which is helpful for discovering the physical and pathological characteristics of prostatic carcinoma.