Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

MR Imaging of Prostate Cancer: Diffusion Weighted Imaging and (3D) Hydrogen 1 (1H) MR Spectroscopy in Comparison with Histology

MR Imaging of Prostate Cancer: Diffusion Weighted Imaging and (3D) Hydrogen 1 (1H) MR... Hindawi Publishing Corporation Radiology Research and Practice Volume 2011, Article ID 616852, 9 pages doi:10.1155/2011/616852 Research Article MR Imaging of Prostate Cancer: Diffusion Weighted Imaging and (3D) Hydrogen 1 ( H) MR Spectroscopy in Comparison with Histology 1 2 1 1 3 2 J. Yamamura, G. Salomon, R. Buchert, A. Hohenstein, J. Graessner, H. Huland, 2 1 1 M. Graefen, G. Adam, and U. Wedegaetner Department of Diagnostic and Interventional Radiology, University Medical Center Hamburg-Eppendorf, 20246 Hamburg, Germany Department of Urology, University Medical Center Hamburg-Eppendorf, 20246 Hamburg, Germany Siemens AG, 20099 Hamburg, Germany Correspondence should be addressed to J. Yamamura, j.yamamura@uke.uni-hamburg.de Received 9 November 2009; Accepted 10 May 2010 Academic Editor: David Maintz Copyright © 2011 J. Yamamura et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Purpose. To evaluate retrospectively the impact of diffusion weighted imaging (DWI) and (3D) hydrogen 1 ( H) MR-spectroscopy (MRS) on the detection of prostatic cancer in comparison to histological examinations. Materials and Methods. 50 patients with suspicion of prostate cancer underwent a MRI examination at a 1.5T scanner. The prostate was divided into sextants. Regions of interest were placed in each sextant to evaluate the apparent diffusion coefficient (ADC)-values. The results of the DWI as well as MRS were compared retrospectively with the findings of the histological examination. Sensitivity and specificity of ADC and metabolic ratio (MET)—both separately and in combination—for identification of tumor tissue was computed for variable discrimination thresholds to evaluate its receiver operator characteristic (ROC). An association between ADC, MET and Gleason 2 −3 score was tested by the non-parametric Spearman ρ-test. Results. The average ADC-value was 1.65 ± 0.32mm /s × 10 in normal 2 −3 tissue and 0.96±0.24 mm /s × 10 in tumor tissue (mean ± 1SD).MET was0.418 ± 0.431 in normal tissue and 2.010 ± 1.649 in tumor tissue. The area under the ROC curve was 0.966 (95%-confidence interval 0.941–0.991) and 0.943 (0.918–0.968) for DWI and MRS, respectively. There was a highly significant negative correlation between ADC-value and the Gleason score in the tumor-positive tissue probes (n = 62, ρ =−0.405, P = .001). MRS did not show a significant correlation with the Gleason score (ρ = 0.117, P = .366). By using both the DWI and MRS, the regression model provided sensitivity and specificity for detection of tumor of 91.9% and 98.3%, respectively. Conclusion. The results of our study showed that both DWI and MRS should be considered as an additional and complementary tool to the T2-weighted MRI for detecting prostate cancer. 1. Introduction The transrectal ultrasound (TRUS) of the prostate is the primarily used method worldwide, besides the laboratory In Europe as well as in the United States prostate cancer is modus operandi. It is used not only to gain the first a very common and frequent cancer in males. According to impression of the organ, but also to guide prostate biopsies, the American Cancer Society, it is the third leading cause of if necessary [3]. Another commonly used method is MR cancer-related death in men. In 2007, 218,890 new cases are imaging of the prostate using an endorectal coil and a pelvic assumed to be diagnosed and about 27,050 persons estimated phased-array coil, in which the malignant sites usually show to die due to this disease [1]. In Europe the incidence of the a hypointense signal compared to the normal hyperintense prostate cancer is approximately 30 per 100 000 men and also peripheral zone [4]. In contrast to the TRUS, the patients the third frequent cause of death after lung and colorectal have to undergo longer examinations, but the local staging cancer [2]. as well as the assessment of the surrounding tissues and organs has shown a better sensitivity using the MRI than Several diagnostic methods have been applied in recent years to detect the malignant changes within the prostate. the aforementioned [5]. Recent studies also showed that the 2 Radiology Research and Practice MR-guided biopsies of the prostate are possible and are of approved by the local Ethics Committee, and informed the same histopathologic quality as specimens obtained with consent was obtained from all patients. From all patients, a TRUS guided biopsies [6]. blood samples were taken to ascertain the prostate specific Over the past few years, however, other MRI tech- antigen (PSA) levels. The prerequisite for the examination niques have been developed to improve the diagnostic was that the patients would undergo a transrectal ultrasound accuracy. A number of studies have shown that the three- (TRUS) and biopsy or prostatectomy thereafter. Patients with dimensional H-spectroscopy of the prostate can ameliorate prior hormonal, surgical, or irradiation therapies as well as the anatomical and morphological situation as well as the previous biopsies within 12 weeks prior to the examination characterization of the prostate cancer [7]. date were excluded. Diffusion weighted MR Imaging (DWI) is a technique to evaluate the molecular diffusion based on the Brownian 2.2. MRI Imaging Protocol. All examinations were performed motion of the spins in biological tissues. DWI provides on a 1.5 T scanner (Symphony; Siemens Medical Solutions, information on both the perfusion and the diffusioninany Erlangen, Germany) with a combination of an endorectal organ to characterize abnormal tissue changes within the coil (MRInnervu; Medrad, Indianola, USA) and a body and sites. This method can be regarded as an additive method spine panoramic array. No contrast medium was used. For to T2-weighted MRI by developing image contrast through the morphological evaluation of the prostate including the “apparent diffusivity.” Diffusion weighted MRI is showing lymph node status of the pelvis, a T1-weighted spin echo potential for improving prostate cancer detection [8–10]. (SE) sequence was used. To evaluate prostatic changes a By adding the diffusion-weighted imaging to conventional T2-weighted fast spin echo (FSE) sequence in transversal, T2-weighted MR imaging, an improvement of detection coronal, and sagittal orientation was performed (Table 1). of prostate cancer was found [8], and diffusion-weighted imaging at 3.0 T has also showed reduced ADC values and 2.3. Diffusion-Weighted Imaging. Based on the T2w images increased fractional anisotropy in prostate cancer [11]as adiffusion weighted (DW) spin echo-planar sequence was well. generated in transversal orientation to include the whole These different methods were combined and compared prostate using the following parameters also using the above- in several studies as well, especially the MR spectroscopy mentioned coil combination: TR 3100 ms; TE 88 ms; FOV and the DW imaging. It has been shown that there is 180 × 180 mm; matrix 128 × 128 mm; Slice thickness 4 mm; a positive correlation between ADC values and the ratio intersection gap 0 mm; voxel size 1.8 × 1.5 × 4 mm; b- for choline and creatine to citrate in men with elevated factors 50, 400, 800 s/mm ; 20 slices. The duration of the prostate-specific antigen (PSA) levels [12]. Another study examination was about 4 to 6 minutes. For DW imaging the showed that if an examined voxel contained ≥70% tumor, above-mentioned coil combination was used. the combined usage of MR spectroscopy and DW imaging The ADC is given by the following equation: increased the specificity in detecting prostate cancer, while the sensitivity compared to MR spectroscopy or DW imaging S(I) = S(0)e − (bi · ADC),(1) alone retained [13]. Recently, Mazaheri et al. has reported where S(I) was the signal intensity measured on the ith a more precise study about the same issue, using the three- b-factor image, and b was the corresponding b-factor. S 1 0 dimensional (3D) hydrogen 1 ( H) MR spectroscopy and the estimates the signal intensity for a b-factor of 0 s/mm , that DW imaging. Also in this study, it could be shown that the is, without the noise induced by the MR measurement [15]. combination of these two had a significant improvement in A starting b-value of 50 s/mm was used to suppress vascular differentiation from malignant and benign tissue [14]. signal in the initial T2 weighted EPI image. The diffusion The aim of this study was to apply both (3D) hydrogen weighting was performed with a trace weighted sequence 1( H) MRS and DWI to the prostate and to determine type (3 orthogonal directions). the Choline-Citrate ratios and the ADC values of healthy According to this equation, ADC-maps were generated tissue and prostate cancer and to compare retrospectively using the software attached to the scanner on the basis of the results with histology, by means of the Gleason-Score, a voxelwise calculation and were interpolated to a 256 × in patients with questionable prostate cancer. Herewith, we 256 mm matrix. hope to assess the potentials with regard to the differentiation of cancer, and to determine the ADC values and the Choline-Citrate ratio of healthy tissue and prostate cancer in 2.4. 3D- H MR Spectroscopic Imaging. The spectroscopic comparison to histology. imaging was performed with the spectroscopic software provided by the MR scanner (Symphony; Siemens Medical Solutions, Erlangen, Germany), using only the endorectal coil. This software acquires data with the point-resolved 2. Materials and Methods spatially localized spectroscopy. By using spectral-spatial 2.1. Study Population. In this study, 50 patients with clin- pulses, choline, creatine, and citrate were excited within ical suspicion of prostate cancer underwent a combined the box. Water and lipids were suppressed with a shim endorectal-body-phased-array MRI at a 1.5 T MRI scanner. around the spectral box. The box was placed on the The mean age of the examined patients was 61.8 years, transverse T2 weighted images, corresponding to the images with the range of 44 to 78 years. The study protocol was made beforehand. The magnetic field homogeneities were Radiology Research and Practice 3 Table 1 Sequence TR [ms] TE [ms] Slice Thickness [mm] FoV [mm] Matrix T1 SE 765 14 5 350 215 × 215 T2 FSE transverse 3400 98 3 180 205 × 256 T2 FSE sagittal 3000 98 3 200 205 × 256 T2 FSE coronal 3000 98 3 200 205 × 256 automatically optimized by shimming algorithms provided the apical, the mid-partial and the basis of the prostate). by the manufacturer. All biopsies were performed by urologists, and the biopsy The following parameters were acquired for the MR cores were labelled to specify the location of the biopsy. spectroscopy: TR 700 ms; TE 120 ms; Flip angle 90 ;number Histopathologic analyses were made by the Institute of of signal acquired = 1; spectral width = 1300 Hz; number of Pathology for all biopsies of the prostate and the Gleason points = 512; FOV 80 × 80 × 80 mm , and phase-encoding scores were evaluated. steps = 16× 8× 8. The voxel volume was 6.7× 6.7× 6.7mm ; SNR 100 csi-ce. The duration of the whole MR spectroscopy 2.7. Statistical Analyses. The prostate was divided into 6 was about 11.46 minutes. regions, that is, sextants: right/left apex, right/left midsec- The evaluation of the spectral data was made by utilizing tion, and right/left base. Tissue probes from each site were the manufacturer’s postprocessing software package. The classified as “normal tissue” or “tumor tissue” according postprocessing included zero filling of the raw data in the to histopathology. Tumor tissue was further categorized superior-inferior direction with a four-dimensional Fourier according to the Gleason score. transformation to yield a voxel volume of 300 mm ,spectral Univariate analysis of variance with ADC-values as apodization with a 2 Hz Lorentzian function, base line independent variable and tissue type (normal, tumor) and correction, peak registration, and an alignment of 3D HMR as intersubject factors was used to compare the ADC value spectroscopic images to the transverse T2 weighted images. between normal tissue and tumor and between different The diameter of integration was 0.3 ppm and was adjusted sextants. for each voxel, just to reach the optimal broadening of each Sensitivity and specificity of ADC-values for identifica- spectral peak. Metabolic ratio maps of choline, creatine, and tion of tumor tissue were computed for variable discrimina- citrate were generated: (choline + creatine)/citrate = MET tion thresholds to evaluate its receiver operator characteristic (metabolic ratio). (ROC). The histologically determined tissue type served The whole time duration of the MRI examination as gold standard. The area under the ROC curve was (incl. MRI, DWI and MRS) including time for patient computed as an overall performance measure. An association placement, coil placement, and localization of the prostate between the ADC value and Gleason score was tested by the was approximately 40 minutes. The image acquisition time nonparametric Spearman ρ-test. The analysis was restricted for the T1 SE, and the T2 FSE was 24 minutes and that for to the tumor-positive probes. the DWI were 5 minutes. The MRS lasted approximately 12 The statistical analyses were repeated with MET instead minutes. of ADC-value as independent variable. An effect was considered statistically significant if the significance level α = 0.05 was reached. All statistic 2.5. MR Image Analyses. The morphological and possibly computations were performed using SPSS 15.0.1 for MS pathological sites of the peripheral zone of the prostate Windows. were retrospectively assessed with the T2 weighted images In order to test whether the combination of DWI and by dividing the prostate into sextants, that is, the apex, MRS might improve the accuracy for detection of tumor the mid-portion, and the base, each right and left side. tissue compared to both DWI and MRS alone, stepwise Then, the grey value of the pixel corresponded to the ADC 2 −3 binary logistic regression was used with histopathology value [mm /s × 10 ] since a pixel-to-pixel ADC map was (normal tissue, tumor) as dependent variable and ADC automatically calculated for each slice. The value itself was values and MET as possible regressors (P for inclusion .05, calculated with the equation mentioned above. The regions P for exclusion .10). of interest (ROI) were then manually drawn in each sextant of the prostate guided by the T2 weighted images. The mean ROI size was 0.8 mm (SD ± 0.56). Additionally, 3. Results metabolic ratio maps of choline, creatine, and citrate were generated with the manufacturer’s software package for each The mean value of the prostate specific antigen (PSA) voxel, especially of suspect areas with the MR spectroscopy. taken from blood samples was 7.19 μg/L (SD ± 5.2), where Voxels were classified as suspicious if the MET was >0.86 the mean PSA level in patients with prostate cancer was [16]. 10.41 μg/L (SD ± 5.1) and those in healthy ones 4.1 μg/L (SD ± 4.2). 2.6. Biopsy/Histopathologic Analyses. All patients underwent All MR examinations were performed successfully, and TRUS guided biopsies (in sextants: right and left sites of although some images had susceptibility artifacts due to 4 Radiology Research and Practice Univariate analysis of variance revealed a highly signif- icant difference of ADC-value between normal tissue and tumor tissue (F = 224.5, df = 1, P = .000), but no difference between the sextants (F = 0.138, df = 5, P = .983) (Figure 2). There was also no significant interaction effect of tissue type and sextants on ADC values (F = 0.356, df = 5, P = .878). Averaged overall sextants, ADC-value was 1.65 ± 0.32 in normal tissue and 0.96 ± 0.23 in tumor tissue (mean ± 1 standard deviation). MET also showed a highly significant difference between normal tissue and tumor tissue (F = 198.4, df = 1, P = .000) (Figure 2). The analysis of variance suggested a significant effect of the sextants (F = 4.5, df = 5, P = .001) as well (a) as a significant interaction effect of tissue type and sextants on MRS (F = 3.4, df = 5, P = .006). However, post hoc comparison of MET between any pair of sextant did not reveal any significant effect, neither in normal tissue nor in tumor. Therefore, the sextants were not taken into account in the further analyses. MET was 0.418 ± 0.431 in normal tissue and 2.010 ± 1.649 in tumor tissue. 4. Sensitivity and Specificity of DWI and MRS Separately and Combined Sensitivity and specificity of DWI and MRS were evaluated separately using ROC analysis; a combined evaluation of both methods was performed using a stepwise binary logistic (b) regression. The ROC curves of DWI and MRS for identification Figure 1: Patient (45 y.o.) with prostate cancer (2.6 cm): PSA-level of tumor tissue irrespective of the sextants are given in 9.7 μg/L (free PSA-level 14.56 μg/L); Gleason-Score 4 + 5 = 9. In the T2w image (a) the prostate cancer is demonstrated in the right Figure 3. The area under the ROC curve was 0.966 (95%- peripheral zone. In the correspondent ADC-map (b) the prostate confidence interval 0.941–0.991) and 0.943 (0.918–0.968) for cancer is clearly shown as a hypointense area. The left peripheral DWI and MRS, respectively. For DWI a sensitivity of 0.92 zone looks hyperintense on the T2w image, but the ADC-map and a specificity of 0.93 were provided using discrimination reveals the remaining healthy prostate tissue. 2 −3 threshold of 1.208 mm /s × 10 . With a threshold providing the same sensitivity, that is, 0.92, MRS provided a specificity of 0.85. With a threshold providing the same specificity, that is, 0.93, MRS provided a sensitivity of 0.68. the endorectal coil, all images could be used for analyses. Stepwise binary logistic regression included both DWI The image acquisition time for the T1 SE, and the T2 FSE (P< .001) and MRS (P< .001) for the differentiation was 24 minutes and that for the DWI were 5 minutes. The betweennormaltissueand tumorisdemonstratedin MRS lasted approximately 11 minutes. As demonstrated in Figure 4. The regression model classified 291 of the 300 Figure 1 the ADC maps show hyperintense in benign and probes correctly (97.0%). Only 4 of 238 normal tissue probes hypointense in malignant tissue in the peripheral zone. were misclassified as tumor, and 5 of 62 tumor probes were Histopathology identified tumor tissue in 21 of the 50 misclassified as normal. Thus, the regression model provided patients (42%). All 6 sextants were infiltrated by the tumor a sensitivity and specificity for detection of tumor of 91.9 and in 3 of these patients, 4 sextants in 3 patients, 3 sextants 98.3, respectively. in 5 patients, 2 sextants in 7 patients, and in 3 patients There was a highly significant negative correlation tumor tissue was detected in only 1 sextant. Thus, in total between DWI and the Gleason score in the tumor-positive 62 of the 300 tissue probes were tumor positive according to tissue probes (n = 62, ρ =−0.405, P = .001) (Figure 5). In histopathology (20.7%). contrast, MRS did not show a significant correlation with the Therateoftumor-positivetissueprobesrangedbetween Gleason score (ρ = 0.117, P = .366). 14% (left apex) and 28% (right midsection). However, this variation was not significant statistically (Pearson’s 2 2 “portmanteau” χ test: χ = 4.066, df = 5, P = .540). 5. Discussion The Gleason score of the tumor-positive tissue probes was 5 in 3 sextants (4.8%), 6 in 16 sextants (25.8%), 7 in However, the fact that the MRI of the prostate might be more 22 sextants (35.5%), 8 in 6 sextants (9.7%), 9 in 11 sextants advantageous than the transrectal ultrasound for staging the (17.7%), and 10 in 4 sextants (6.5%). cancer has also been discussed in the past (especially for T2 Radiology Research and Practice 5 DWI 0.8 0.6 Area = 0.966 0.4 0.2 Right apex Right base Left midsection 0 Right midsection Left apex Left base 0 0.2 0.4 0.6 0.8 1 1-specificity (a) Normal tissue Tumor MRS (a) 0.8 0.6 Area = 0.943 0.4 2 ∗ 0.2 0 0.2 0.4 0.6 0.8 1 1-specificity Right apex Right base Left midsection (b) Right midsection Left apex Left base Figure 3: Receiver operator characteristic (ROC) curve of DWI (a) Normal tissue and MRS (b) for differentiating tumor tissue from normal tissue. Tumor The analysis included all 300 tissue probes irrespective of the ROI. (b) Figure 2: Box-and-whisker plot of DWI (a) and MRS (b) as is one of the few organs in humans which can be examined a function of tissue type (normal tissue, tumor) and region of by MRI without any contrast media. Benign tissues in the interest (ROI). Outliers (1.5–3 box lengths) are indicated by a circle; peripheral zone show hyperintense signals in T2 weighted extreme values (>3 box lengths) are indicated by an asterix. imaging, whereas malignant changes show hypointense signals, of which the reason could be the cellular density as well as the malfunction of the gland when the malignant and T3 tumours) [17, 18]. The T2 weighted MRI images change had occurred. The cause of the decrease in diffusion of the prostate has been applied more often to improve the in malignant tissue has a histopathologic origin. Some validity of the staging in prostate cancer [19]. The prostate attributes are: hypercellularity, enlargement of the nuclei, DWI MRS Sensitivity Sensitivity 6 Radiology Research and Practice 56789 10 MRS Gleason score Normal tissue (a) Tumor Figure 4: Scatter plot of DWI versus MRS in normal tissue probes and tumor-positive tissue probes. hyperchromatism, and angulation of the nuclear contour, which lead to a reduction of diffusional displacement of water molecules (Anderson JR. Muir’s textbook of pathology. London, England: Edward Arnold 1985). Commonly, the prostate produces 20 to 30% of the ejaculate secretions. In patients with known prostatic cancer, the amount of the ejaculate can be less than in healthy patients. However, this is difficult to determine since the secretions vary from 0.5 to 13 mL. Diffusion weighted MR imaging has been clinically applied in several organs. Not only is it used to show the affected tissue after a stroke, it is also to differentiate brain tumours [20, 21] or also vertebral metastases in, for example, 56789 10 prostate cancer [22]. In diffusion-weighted MRI (DWI) the Gleason score image contrast is determined by the random microscopic (b) motion of water protons, that is, the Brownian motion. The diffusion can be measured in vivo by using the MRI because Figure 5: Scatter plot of DWI (a) and MRS (b) versus Gleason score of its sensitivity to motion. This sensitivity to motion can be in tumor-positive tissue probes. increased by the addition of strong magnetic field gradient pulses to the pulse sequence [23]. Shimofusa et al. applied the DWI of the prostate with parallel imaging and with a high b-value (b = 1000) for the first time [8]. In this study since 1980s. Since then, MR spectroscopy has been effective they have not used an endorectal coil, but the sensitivity as in improving the accuracy of MR imaging in prostate cancer well as specificity was higher than in other former studies localization and staging [32–34]. In the healthy prostate, with endorectal coil [24–27]oradynamicstudy [28]. Since malign and benign tissues can be differentiated by the MR then, several prostatic MR imaging modules were performed spectroscopy on the basis of the metabolic ratio of choline, to increase the detectability of cancerous tissue. The results of creatine, and citrate. The ratio is calculated by the equation these studies demonstrate that the ADC value may provide MET = (choline + creatine)/citrate. The MET is increased in information about the malignant changes in the prostate malign tissue whereas a lower MET can be found in benign [29–31]. tissue [35, 36]. MR spectroscopy is a relatively new method in diagnos- In this present study, 50 patients with suspected cancer ing prostate cancer and has been a part of clinical routine of the prostate were examined with the DW MRI and the DWI DWI MRS Radiology Research and Practice 7 MR spectroscopy, and then compared retrospectively with tissues (0.418±0.431). This result can also be compared with pathohistological results, especially with the Gleason score. previous studies. Two recent studies compared and analyzed the combined In this present study, a relatively larger number of usage of diffusion-weighted MRI and HMRspectroscopy. patients were examined than other studies. The combination The one study examined 42 patients with prostate cancer of DWI and MRS performed significantly better in detecting using a 2D chemical shift imaging and isotropic apparent cancer in the periphery zone of the prostate than MRS diffusion coefficient (ADC) maps [13]. In this study the alone [14] in one study. In our study with 50 patients, the regions of interest were drawn around the whole gland, combination of DWI and MRS seems also to have a better central gland, and the peripheral zone tumor. The mean accuracy in detecting cancerous tissue. The regression model ADC value of the normal tissue in this study was 1.51 mm /s classified 291 of the 300 probes correctly (97.0%). Only 4 of −3 × 10 [SD: ±0.27]. 238 normal tissue probes were misclassified as tumor, and If the tumor was greater than 30% of the whole voxel 5 of 62 tumor probes were misclassified as normal. Thus, 2 −3 the mean ADC value was 1.19 mm /s × 10 [SD: ±0.24], the regression model provided sensitivity and specificity for and if the tumor was greater than 70% of the whole voxel, detection of tumor of 91.9 and 98.3, respectively (Figure 5). 2 −3 the value was 1.03 mm /s × 10 [SD: ±0.18]. The mean If the results were correlated with the Gleason-score, MET in normal gland was 0.065 ± 0.052, whereas the value there was a highly significant negative correlation between was much higher in malignant tissues; 0.814 ± 2.202 in DWI and the Gleason score in the tumor-positive tissue tumor ≥30% of the voxel and 0.917± 1.276 in tumor ≥70%, probes (n = 62, ρ =−0.405, P = .001) (Figure 4). However, respectively. The MET was significantly higher (P< .001) MRS did not show a significant correlation with the Gleason and the ADC values were significantly (P< .006) lower score (ρ = 0.117, P = .366). One explanation could be that in tumor-containing voxel. The area under the ROC curves in the present study the MET values were extracted using a using both the ADC and MET was 0.81, similar to only very short TR sequence (700 ms) which keeps the scan time MET (0.79), whereas ADC alone showed an area of 0.66 and tolerable but results in spectra that are fairly heavily T1- was inferior. An interesting point, however, is the significant weighted. Perhaps that is why the other groups [38]found improvement in specificity for the combination of ADC and a weak correlation between Gleason and MET. MET, when voxels containing 70% or tumor were considered There are some problems and limitations in our study positive and cutoffs to achieve a 90% or greater sensitivity like in other MR studies of the prostate. First limitations were chosen [13]. are based on MR technique itself. Common artefacts for the The other study performed a retrospective measurement DWI are white pixel noise, low SNR (signal-to-noise ratio), of ADC and MET in 38 patients with prostatic cancer. and susceptibility artefacts. In MR spectroscopy common ThemeanADC valueand METfor malignanttissuewere artefacts are lipid contamination and susceptibility artefacts. 2 −3 1.39 mm /s × 10 [SD: ±0.23] and 0.92± 0.32, respectively. These artefacts could be reduced by new methods in the 2 −3 For benign tissue, the values were 1.69 mm /s × 10 [SD: future. Second limitation is the discrepancy in voxel sizes ±0.24] and 0.73 ± 0.18 (P< .001 for both). In this between the ADC map and the 3D HMRspectroscopy. In study, areas under the receiver operating characteristic curves the future new spectroscopic techniques with increased spa- (AUCs) were performed to evaluate the accuracy. Obviously, tial resolution without increasing the examination duration the combination of ADC and MET performed significantly [39] could be used to overcome this problem. The present better (AUC = 0.85; P = .005) than ADC or MET alone (AUC study was a retrospective analysis of the prostate cancer. The = 0.81 and AUC = 0.09, resp.) [14]. malignant sites were known when analyzing the ADC map Analogue to these prior studies, our results of the ADC and the MET. Chronic changes, such as chronic prostatitis values were significantly lower (P< .001) and the results or atrophy of the gland itself, usually show similar changes of the mean MET were significantly higher (P< .001) in MRI [4, 27] to prostatic cancer. Prospective studies for malignant prostatic tissues than for benign tissues. In are another research task for succeeding studies with the our study, the average ADC value in benign tissue was DWI and MRS, especially in patients with several negative 2 −3 1.65 mm /s × 10 [standard deviation (SD): ±0.32] and biopsies. The last limitation of this study was the usage of 2 −3 thatinmaligntissues 0.96mm /s × 10 [SD: ±0.24]. The sextant biopsy, since biopsies can be easily false negative so significantly (P< .001) lower ADC values in malign tissues that the tumor localization might not be very exact. These compared with the benign signify some promising results methods can then be combined with the dynamic contrast- in detecting the cancer. These results were similar to other enhanced magnetic resonance imaging, since first studies prior studies, although different b-values (0 and 1000 s/mm have shown that it may be an accurate technique for detecting and 0, 300, 600 s/mm )[30, 37] were used in our study and quantifying intracapsular transition or peripheral zone (i.e., 50, 400, 800 s/mm ). The cutoff value of the mean tumor foci greater than 0.2 cc [40]. ADC value between cancerous and noncancerous tissue in 2 −3 the present was at approximately 1.2 mm /s × 10 ,which 6. Conclusion canbeseenonthe ROC-Analyses (Figure 3). Compared to the aforementioned studies our results of the ADC value The results of our study showed that both diffusion-weighted for malignant tissue were remarkably lower. The MET of Imaging and MR spectroscopy should be considered as malignant tissues showed an average value of 2.010 ± 1.649 an additional and complementary tool to the T2 weighted in our study, and is significantly higher than in healthy MRI not only for detecting prostate cancer, but also for 8 Radiology Research and Practice guiding more specific biopsies without taking several samples to increase accuracy in prostate cancer detection,” American Journal of Roentgenology, vol. 188, no. 1, pp. 91–98, 2007. from the organ. The combination of these methods can [14] Y. Mazaheri, A. Shukla-Dave, H. Hricak et al., “Prostate improve the specificity and may prevent uncomfortable as cancer: identification with combined diffusion-weighted MR well as painful biopsies for patients. The DWI is furthermore imaging and 3D H MR spectroscopic imaging—correlation advantageous over the present examination methods, such with pathologic findings,” Radiology, vol. 246, no. 2, pp. 480– as the MR spectroscopy, regarding the considerably shorter 488, 2008. examination time (approximatly 5 minutes in case of this [15] D. Le Bihan, C. Poupon, A. Amadon, and F. Lethimonnier, study) and the correlation with the Gleason score. “Artifacts and pitfalls in diffusion MRI,” Journal of Magnetic Resonance Imaging, vol. 24, no. 3, pp. 478–488, 2006. [16] J. Kurhanewicz, D. B. Vigneron, H. Hricak et al., “Prostate References cancer: metabolic response to cryosurgery as detected with 3D H-1 MR spectroscopic imaging,” Radiology, vol. 200, no. 2, pp. [1] M.J.Hayat,N.Howlader, M. E. Reichman,and B. K. 489–496, 1996. Edwards, “Cancer statistics, trends, and multiple primary cancer analyses from the surveillance, epidemiology, and end [17] R. A. H. Boni, ¨ B. E. Hutter, F. Trinkler, W. Jochum, D. results (SEER) program,” Oncologist, vol. 12, no. 1, pp. 20–37, Pestalozzi, and G. P. Krestin, “Preoperative T-staging of prostate carcinoma: endorectal MRI compared with other imaging and clinical methods,” RoF ¨ o, vol. 165, no. 2, pp. 152– [2] F. Levi, F. Lucchini, E. Negri, P. Boyle, and C. La Vecchia, 158, 1996. “Leveling of prostate cancer mortality in Western Europe,” The Prostate, vol. 60, no. 1, pp. 46–52, 2004. [18] A. Wetter, A. N. Ajdukovic, K. Fliessbach et al., “Staging of prostate cancer: value of the combined information of [3] R. S. Purohit, K. Shinohara, M. V. Meng, and P. R. Carroll, “Imaging clinically localized prostate cancer,” Urologic Clinics endorectal MRI, biopsy Gleason score, and preoperative PSA level,” RoF ¨ o, vol. 178, no. 4, pp. 385–390, 2006. of North America, vol. 30, no. 2, pp. 279–293, 2003. [19] M. R. Engelbrecht, G. J. Jager, R. J. Laheij, A. L. M. Verbeek, H. [4] S. Ikonen, L. Kivisaari, P. Tervahartiala, T. Vehmas, K. Taari, and S. Rannikko, “Prostatic MR imaging: accuracy in J. van Lier, and J. O. Barentsz, “Local staging of prostate cancer using magnetic resonance imaging: a meta-analysis,” European differentiating cancer from other prostatic disorders,” Acta Radiology, vol. 12, no. 9, pp. 2294–2302, 2002. Radiologica, vol. 42, no. 4, pp. 348–354, 2001. [5] M. Sanchez-Chapado, J. C. Angulo, C. Ibarburen et al., [20] T. L. Chenevert, P. E. McKeever, and B. D. Ross, “Monitoring early response of experimental brain tumors to therapy “Comparison of digital rectal examination, transrectal ultra- sonography, and multicoil magnetic resonance imaging for using diffusion magnetic resonance imaging,” Clinical Cancer Research, vol. 3, no. 9, pp. 1457–1466, 1997. preoperative evaluation of prostate cancer,” European Urology, vol. 32, no. 2, pp. 140–149, 1997. [21] F. Yamasaki, K. Kurisu, K. Satoh et al., “Apparent diffusion [6] T. Franiel, F. Fritzsche, A. Staack, J. Rost, B. Hamm, and D. coefficient of human brain tumors at MR imaging,” Radiology, vol. 235, no. 3, pp. 985–991, 2005. Beyersdorff, “Histopathologic quality of prostate core biopsy specimens: comparison of an MR-compatible biopsy needle [22] T. Hacklander ¨ , C. Scharwa¨chter,R.Golz,and H. Mertens, and a ferromagnetic biopsy needle used for ultrasound-guided “Value of diffusion-weighted imaging for diagnosing vertebral prostate biopsy,” RoF ¨ o, vol. 178, no. 12, pp. 1212–1218, 2006. metastases due to prostate cancer in comparison to other [7] J. Kurhanewicz, D. B. Vigneron, H. Hricak, P. Narayan, primary tumors,” RoF ¨ o, vol. 178, no. 4, pp. 416–424, 2006. P. Carroll, and S. J. Nelson, “Three-dimensional H-1 MR [23] M. Cercignani, M. A. Horsfield, F. Agosta, and M. Filippi, spectroscopic imaging of the in situ human prostate with high “Sensitivity-encoded diffusion tensor MR imaging of the (0.24-0.7-cm3) spatial resolution,” Radiology, vol. 198, no. 3, cervical cord,” American Journal of Neuroradiology, vol. 24, no. pp. 795–805, 1996. 6, pp. 1254–1256, 2003. [8] R. Shimofusa, H. Fujimoto, H. Akamata et al., “Diffusion- [24] K. K. Yu and H. Hricak, “Imaging prostate cancer,” Radiologic weighted imaging of prostate cancer,” Journal of Computer Clinics of North America, vol. 38, no. 1, pp. 59–85, 2000. Assisted Tomography, vol. 29, no. 2, pp. 149–153, 2005. [25] H. Hricak, S. White, D. Vigneron et al., “Carcinoma of [9] K. Hosseinzadeh and S. D. Schwarz, “Endorectal diffusion- the prostate gland: MR imaging with pelvic phased-array weighted imaging in prostate cancer to differentiate malignant coils versus integrated endorectal-pelvic phased-array coils,” and benign peripheral zone tissue,” Journal of Magnetic Radiology, vol. 193, no. 3, pp. 703–709, 1994. Resonance Imaging, vol. 20, no. 4, pp. 654–661, 2004. [26] L. E. Quint, J. S. Van Erp, P. H. Bland et al., “Prostate [10] B. Issa, “In vivo measurement of the apparent diffusion cancer: correlation of MR images with tissue optical density at coefficient in normal and malignant prostatic tissues using pathologic examination,” Radiology, vol. 179, no. 3, pp. 837– echo-planar imaging,” Journal of Magnetic Resonance Imaging, 842, 1991. vol. 16, no. 2, pp. 196–200, 2002. [27] S. Ikonen, P. Kar ¨ kkainen, ¨ L. Kivisaari et al., “Endorectal [11] P. Gibbs, M. D. Pickles, and L. W. Turnbull, “Diffusion magnetic resonance imaging of prostatic cancer: comparison imaging of the prostate at 3.0 tesla,” Investigative Radiology, between fat-suppressed T2-weighted fast spin echo and three- vol. 41, no. 2, pp. 185–188, 2006. dimensional dual-echo, steady-state sequences,” European Radiology, vol. 11, no. 2, pp. 236–241, 2001. [12] V. Kumar, N. R. Jagannathan, R. Kumar et al., “Apparent diffusion coefficient of the prostate in men prior to biopsy: [28] G. J. Jager, E. T. G. Ruijter, C. A. van de Kaa et al., “Dynamic determination of a cut-off value to predict malignancy of the TurboFLASH subtraction technique for contrast-enhanced peripheral zone,” NMR in Biomedicine, vol. 20, no. 5, pp. 505– MR imaging of the prostate: correlation with histopathologic 511, 2007. results,” Radiology, vol. 203, no. 3, pp. 645–652, 1997. [29] V. A. Morgan,S.Kyriazi,S.E.Ashley, andN.M.DeSouza, [13] S. A. Reinsberg, G. S. Payne, S. F. Riches et al., “Combined use of diffusion-weighted MRI and HMRspectroscopy “Evaluation of the potential of diffusion-weighted imaging in Radiology Research and Practice 9 prostate cancer detection,” Acta Radiologica,vol. 48, no.6,pp. 695–703, 2007. [30] C. Sato, S. Naganawa, T. Nakamura et al., “Differentiation of noncancerous tissue and cancer lesion by apparent diffusion coefficient values in transition and peripheral zones of the prostate,” Journal of Magnetic Resonance Imaging, vol. 21, no. 3, pp. 258–262, 2005. [31] H. Miao, H. Fukatsu, and T. Ishigaki, “Prostate cancer detection with 3-T MRI: comparison of diffusion-weighted and T2-weighted imaging,” European Journal of Radiology, vol. 61, no. 2, pp. 297–302, 2007. [32] J. Scheidler, H. Hricak, D. B. Vigneron et al., “Prostate cancer: localization with three-dimensional proton MR spectroscopic imaging—clinicopathologic study,” Radiology, vol. 213, no. 2, pp. 473–480, 1999. [33] A. E. Wefer, H. Hricak, D. B. Vigneron et al., “Sextant localization of prostate cancer: comparison of sextant biopsy, magnetic resonance imaging and magnetic resonance spectro- scopic imaging with step section histology,” JournalofUrology, vol. 164, no. 2, pp. 400–404, 2000. [34] K. K. Yu, J. Scheidler, H. Hricak et al., “Prostate cancer: prediction of extracapsular extension with endorectal MR imaging and three-dimensional proton MR spectroscopic imaging,” Radiology, vol. 213, no. 2, pp. 481–488, 1999. [35] J. Kurhanewicz, D. B. Vigneron,R.G.Males,M.G.Swanson, K. K. Yu, and H. Hricak, “The prostate: MR imaging and spectroscopy: present and future,” Radiologic Clinics of North America, vol. 38, no. 1, pp. 115–138, 2000. [36] R. G. Males, D. B. Vigneron, J. Star-Lack et al., “Clinical application of BASING and spectral/spatial water and lipid suppression pulses for prostate cancer staging and localization by in vivo 3D H magnetic resonance spectroscopic imaging,” Magnetic Resonance in Medicine, vol. 43, no. 1, pp. 17–22, [37] A. Tanimoto, J. Nakashima, H. Kohno, H. Shinmoto, and S. Kuribayashi, “Prostate cancer screening: the clinical value of diffusion-weighted imaging and dynamic MR imaging in combination with T2-weighted imaging,” Journal of Magnetic Resonance Imaging, vol. 25, no. 1, pp. 146–152, 2007. [38] K. L. Zakian, K. Sircar, H. Hricak et al., “Correlation of proton MR spectroscopic imaging with gleason score based on step-section pathologic analysis after radical prostatectomy,” Radiology, vol. 234, no. 3, pp. 804–814, 2005. [39] T. W. J. Scheenen,D.W.J.Klomp,S.A.Rol ¨ l, J. J. Futt ¨ erer, J. O. Barentsz, and A. Heerschap, “Fast acquisition-weighted three- dimensional proton MR spectroscopic imaging of the human prostate,” Magnetic Resonance in Medicine,vol. 52, no.1,pp. 80–88, 2004. [40] A. Villers, P. Puech, D. Mouton, X. Leroy, C. Ballereau, and L. Lemaitre, “Dynamic contrast enhanced, pelvic phased array magnetic resonance imaging of localized prostate cancer for predicting tumor volume: correlation with radical prostatec- tomy findings,” JournalofUrology, vol. 176, no. 6, pp. 2432– 2437, 2006. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 International Journal of Journal of Immunology Research Endocrinology Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Submit your manuscripts at http://www.hindawi.com BioMed PPAR Research Research International Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Research and Treatment Cellular Longevity Neurology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Radiology Research and Practice Hindawi Publishing Corporation

MR Imaging of Prostate Cancer: Diffusion Weighted Imaging and (3D) Hydrogen 1 (1H) MR Spectroscopy in Comparison with Histology

Loading next page...
 
/lp/hindawi-publishing-corporation/mr-imaging-of-prostate-cancer-diffusion-weighted-imaging-and-3d-xCPlkLHilK
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2011 J. Yamamura et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
2090-1941
eISSN
2090-195X
DOI
10.1155/2011/616852
Publisher site
See Article on Publisher Site

Abstract

Hindawi Publishing Corporation Radiology Research and Practice Volume 2011, Article ID 616852, 9 pages doi:10.1155/2011/616852 Research Article MR Imaging of Prostate Cancer: Diffusion Weighted Imaging and (3D) Hydrogen 1 ( H) MR Spectroscopy in Comparison with Histology 1 2 1 1 3 2 J. Yamamura, G. Salomon, R. Buchert, A. Hohenstein, J. Graessner, H. Huland, 2 1 1 M. Graefen, G. Adam, and U. Wedegaetner Department of Diagnostic and Interventional Radiology, University Medical Center Hamburg-Eppendorf, 20246 Hamburg, Germany Department of Urology, University Medical Center Hamburg-Eppendorf, 20246 Hamburg, Germany Siemens AG, 20099 Hamburg, Germany Correspondence should be addressed to J. Yamamura, j.yamamura@uke.uni-hamburg.de Received 9 November 2009; Accepted 10 May 2010 Academic Editor: David Maintz Copyright © 2011 J. Yamamura et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Purpose. To evaluate retrospectively the impact of diffusion weighted imaging (DWI) and (3D) hydrogen 1 ( H) MR-spectroscopy (MRS) on the detection of prostatic cancer in comparison to histological examinations. Materials and Methods. 50 patients with suspicion of prostate cancer underwent a MRI examination at a 1.5T scanner. The prostate was divided into sextants. Regions of interest were placed in each sextant to evaluate the apparent diffusion coefficient (ADC)-values. The results of the DWI as well as MRS were compared retrospectively with the findings of the histological examination. Sensitivity and specificity of ADC and metabolic ratio (MET)—both separately and in combination—for identification of tumor tissue was computed for variable discrimination thresholds to evaluate its receiver operator characteristic (ROC). An association between ADC, MET and Gleason 2 −3 score was tested by the non-parametric Spearman ρ-test. Results. The average ADC-value was 1.65 ± 0.32mm /s × 10 in normal 2 −3 tissue and 0.96±0.24 mm /s × 10 in tumor tissue (mean ± 1SD).MET was0.418 ± 0.431 in normal tissue and 2.010 ± 1.649 in tumor tissue. The area under the ROC curve was 0.966 (95%-confidence interval 0.941–0.991) and 0.943 (0.918–0.968) for DWI and MRS, respectively. There was a highly significant negative correlation between ADC-value and the Gleason score in the tumor-positive tissue probes (n = 62, ρ =−0.405, P = .001). MRS did not show a significant correlation with the Gleason score (ρ = 0.117, P = .366). By using both the DWI and MRS, the regression model provided sensitivity and specificity for detection of tumor of 91.9% and 98.3%, respectively. Conclusion. The results of our study showed that both DWI and MRS should be considered as an additional and complementary tool to the T2-weighted MRI for detecting prostate cancer. 1. Introduction The transrectal ultrasound (TRUS) of the prostate is the primarily used method worldwide, besides the laboratory In Europe as well as in the United States prostate cancer is modus operandi. It is used not only to gain the first a very common and frequent cancer in males. According to impression of the organ, but also to guide prostate biopsies, the American Cancer Society, it is the third leading cause of if necessary [3]. Another commonly used method is MR cancer-related death in men. In 2007, 218,890 new cases are imaging of the prostate using an endorectal coil and a pelvic assumed to be diagnosed and about 27,050 persons estimated phased-array coil, in which the malignant sites usually show to die due to this disease [1]. In Europe the incidence of the a hypointense signal compared to the normal hyperintense prostate cancer is approximately 30 per 100 000 men and also peripheral zone [4]. In contrast to the TRUS, the patients the third frequent cause of death after lung and colorectal have to undergo longer examinations, but the local staging cancer [2]. as well as the assessment of the surrounding tissues and organs has shown a better sensitivity using the MRI than Several diagnostic methods have been applied in recent years to detect the malignant changes within the prostate. the aforementioned [5]. Recent studies also showed that the 2 Radiology Research and Practice MR-guided biopsies of the prostate are possible and are of approved by the local Ethics Committee, and informed the same histopathologic quality as specimens obtained with consent was obtained from all patients. From all patients, a TRUS guided biopsies [6]. blood samples were taken to ascertain the prostate specific Over the past few years, however, other MRI tech- antigen (PSA) levels. The prerequisite for the examination niques have been developed to improve the diagnostic was that the patients would undergo a transrectal ultrasound accuracy. A number of studies have shown that the three- (TRUS) and biopsy or prostatectomy thereafter. Patients with dimensional H-spectroscopy of the prostate can ameliorate prior hormonal, surgical, or irradiation therapies as well as the anatomical and morphological situation as well as the previous biopsies within 12 weeks prior to the examination characterization of the prostate cancer [7]. date were excluded. Diffusion weighted MR Imaging (DWI) is a technique to evaluate the molecular diffusion based on the Brownian 2.2. MRI Imaging Protocol. All examinations were performed motion of the spins in biological tissues. DWI provides on a 1.5 T scanner (Symphony; Siemens Medical Solutions, information on both the perfusion and the diffusioninany Erlangen, Germany) with a combination of an endorectal organ to characterize abnormal tissue changes within the coil (MRInnervu; Medrad, Indianola, USA) and a body and sites. This method can be regarded as an additive method spine panoramic array. No contrast medium was used. For to T2-weighted MRI by developing image contrast through the morphological evaluation of the prostate including the “apparent diffusivity.” Diffusion weighted MRI is showing lymph node status of the pelvis, a T1-weighted spin echo potential for improving prostate cancer detection [8–10]. (SE) sequence was used. To evaluate prostatic changes a By adding the diffusion-weighted imaging to conventional T2-weighted fast spin echo (FSE) sequence in transversal, T2-weighted MR imaging, an improvement of detection coronal, and sagittal orientation was performed (Table 1). of prostate cancer was found [8], and diffusion-weighted imaging at 3.0 T has also showed reduced ADC values and 2.3. Diffusion-Weighted Imaging. Based on the T2w images increased fractional anisotropy in prostate cancer [11]as adiffusion weighted (DW) spin echo-planar sequence was well. generated in transversal orientation to include the whole These different methods were combined and compared prostate using the following parameters also using the above- in several studies as well, especially the MR spectroscopy mentioned coil combination: TR 3100 ms; TE 88 ms; FOV and the DW imaging. It has been shown that there is 180 × 180 mm; matrix 128 × 128 mm; Slice thickness 4 mm; a positive correlation between ADC values and the ratio intersection gap 0 mm; voxel size 1.8 × 1.5 × 4 mm; b- for choline and creatine to citrate in men with elevated factors 50, 400, 800 s/mm ; 20 slices. The duration of the prostate-specific antigen (PSA) levels [12]. Another study examination was about 4 to 6 minutes. For DW imaging the showed that if an examined voxel contained ≥70% tumor, above-mentioned coil combination was used. the combined usage of MR spectroscopy and DW imaging The ADC is given by the following equation: increased the specificity in detecting prostate cancer, while the sensitivity compared to MR spectroscopy or DW imaging S(I) = S(0)e − (bi · ADC),(1) alone retained [13]. Recently, Mazaheri et al. has reported where S(I) was the signal intensity measured on the ith a more precise study about the same issue, using the three- b-factor image, and b was the corresponding b-factor. S 1 0 dimensional (3D) hydrogen 1 ( H) MR spectroscopy and the estimates the signal intensity for a b-factor of 0 s/mm , that DW imaging. Also in this study, it could be shown that the is, without the noise induced by the MR measurement [15]. combination of these two had a significant improvement in A starting b-value of 50 s/mm was used to suppress vascular differentiation from malignant and benign tissue [14]. signal in the initial T2 weighted EPI image. The diffusion The aim of this study was to apply both (3D) hydrogen weighting was performed with a trace weighted sequence 1( H) MRS and DWI to the prostate and to determine type (3 orthogonal directions). the Choline-Citrate ratios and the ADC values of healthy According to this equation, ADC-maps were generated tissue and prostate cancer and to compare retrospectively using the software attached to the scanner on the basis of the results with histology, by means of the Gleason-Score, a voxelwise calculation and were interpolated to a 256 × in patients with questionable prostate cancer. Herewith, we 256 mm matrix. hope to assess the potentials with regard to the differentiation of cancer, and to determine the ADC values and the Choline-Citrate ratio of healthy tissue and prostate cancer in 2.4. 3D- H MR Spectroscopic Imaging. The spectroscopic comparison to histology. imaging was performed with the spectroscopic software provided by the MR scanner (Symphony; Siemens Medical Solutions, Erlangen, Germany), using only the endorectal coil. This software acquires data with the point-resolved 2. Materials and Methods spatially localized spectroscopy. By using spectral-spatial 2.1. Study Population. In this study, 50 patients with clin- pulses, choline, creatine, and citrate were excited within ical suspicion of prostate cancer underwent a combined the box. Water and lipids were suppressed with a shim endorectal-body-phased-array MRI at a 1.5 T MRI scanner. around the spectral box. The box was placed on the The mean age of the examined patients was 61.8 years, transverse T2 weighted images, corresponding to the images with the range of 44 to 78 years. The study protocol was made beforehand. The magnetic field homogeneities were Radiology Research and Practice 3 Table 1 Sequence TR [ms] TE [ms] Slice Thickness [mm] FoV [mm] Matrix T1 SE 765 14 5 350 215 × 215 T2 FSE transverse 3400 98 3 180 205 × 256 T2 FSE sagittal 3000 98 3 200 205 × 256 T2 FSE coronal 3000 98 3 200 205 × 256 automatically optimized by shimming algorithms provided the apical, the mid-partial and the basis of the prostate). by the manufacturer. All biopsies were performed by urologists, and the biopsy The following parameters were acquired for the MR cores were labelled to specify the location of the biopsy. spectroscopy: TR 700 ms; TE 120 ms; Flip angle 90 ;number Histopathologic analyses were made by the Institute of of signal acquired = 1; spectral width = 1300 Hz; number of Pathology for all biopsies of the prostate and the Gleason points = 512; FOV 80 × 80 × 80 mm , and phase-encoding scores were evaluated. steps = 16× 8× 8. The voxel volume was 6.7× 6.7× 6.7mm ; SNR 100 csi-ce. The duration of the whole MR spectroscopy 2.7. Statistical Analyses. The prostate was divided into 6 was about 11.46 minutes. regions, that is, sextants: right/left apex, right/left midsec- The evaluation of the spectral data was made by utilizing tion, and right/left base. Tissue probes from each site were the manufacturer’s postprocessing software package. The classified as “normal tissue” or “tumor tissue” according postprocessing included zero filling of the raw data in the to histopathology. Tumor tissue was further categorized superior-inferior direction with a four-dimensional Fourier according to the Gleason score. transformation to yield a voxel volume of 300 mm ,spectral Univariate analysis of variance with ADC-values as apodization with a 2 Hz Lorentzian function, base line independent variable and tissue type (normal, tumor) and correction, peak registration, and an alignment of 3D HMR as intersubject factors was used to compare the ADC value spectroscopic images to the transverse T2 weighted images. between normal tissue and tumor and between different The diameter of integration was 0.3 ppm and was adjusted sextants. for each voxel, just to reach the optimal broadening of each Sensitivity and specificity of ADC-values for identifica- spectral peak. Metabolic ratio maps of choline, creatine, and tion of tumor tissue were computed for variable discrimina- citrate were generated: (choline + creatine)/citrate = MET tion thresholds to evaluate its receiver operator characteristic (metabolic ratio). (ROC). The histologically determined tissue type served The whole time duration of the MRI examination as gold standard. The area under the ROC curve was (incl. MRI, DWI and MRS) including time for patient computed as an overall performance measure. An association placement, coil placement, and localization of the prostate between the ADC value and Gleason score was tested by the was approximately 40 minutes. The image acquisition time nonparametric Spearman ρ-test. The analysis was restricted for the T1 SE, and the T2 FSE was 24 minutes and that for to the tumor-positive probes. the DWI were 5 minutes. The MRS lasted approximately 12 The statistical analyses were repeated with MET instead minutes. of ADC-value as independent variable. An effect was considered statistically significant if the significance level α = 0.05 was reached. All statistic 2.5. MR Image Analyses. The morphological and possibly computations were performed using SPSS 15.0.1 for MS pathological sites of the peripheral zone of the prostate Windows. were retrospectively assessed with the T2 weighted images In order to test whether the combination of DWI and by dividing the prostate into sextants, that is, the apex, MRS might improve the accuracy for detection of tumor the mid-portion, and the base, each right and left side. tissue compared to both DWI and MRS alone, stepwise Then, the grey value of the pixel corresponded to the ADC 2 −3 binary logistic regression was used with histopathology value [mm /s × 10 ] since a pixel-to-pixel ADC map was (normal tissue, tumor) as dependent variable and ADC automatically calculated for each slice. The value itself was values and MET as possible regressors (P for inclusion .05, calculated with the equation mentioned above. The regions P for exclusion .10). of interest (ROI) were then manually drawn in each sextant of the prostate guided by the T2 weighted images. The mean ROI size was 0.8 mm (SD ± 0.56). Additionally, 3. Results metabolic ratio maps of choline, creatine, and citrate were generated with the manufacturer’s software package for each The mean value of the prostate specific antigen (PSA) voxel, especially of suspect areas with the MR spectroscopy. taken from blood samples was 7.19 μg/L (SD ± 5.2), where Voxels were classified as suspicious if the MET was >0.86 the mean PSA level in patients with prostate cancer was [16]. 10.41 μg/L (SD ± 5.1) and those in healthy ones 4.1 μg/L (SD ± 4.2). 2.6. Biopsy/Histopathologic Analyses. All patients underwent All MR examinations were performed successfully, and TRUS guided biopsies (in sextants: right and left sites of although some images had susceptibility artifacts due to 4 Radiology Research and Practice Univariate analysis of variance revealed a highly signif- icant difference of ADC-value between normal tissue and tumor tissue (F = 224.5, df = 1, P = .000), but no difference between the sextants (F = 0.138, df = 5, P = .983) (Figure 2). There was also no significant interaction effect of tissue type and sextants on ADC values (F = 0.356, df = 5, P = .878). Averaged overall sextants, ADC-value was 1.65 ± 0.32 in normal tissue and 0.96 ± 0.23 in tumor tissue (mean ± 1 standard deviation). MET also showed a highly significant difference between normal tissue and tumor tissue (F = 198.4, df = 1, P = .000) (Figure 2). The analysis of variance suggested a significant effect of the sextants (F = 4.5, df = 5, P = .001) as well (a) as a significant interaction effect of tissue type and sextants on MRS (F = 3.4, df = 5, P = .006). However, post hoc comparison of MET between any pair of sextant did not reveal any significant effect, neither in normal tissue nor in tumor. Therefore, the sextants were not taken into account in the further analyses. MET was 0.418 ± 0.431 in normal tissue and 2.010 ± 1.649 in tumor tissue. 4. Sensitivity and Specificity of DWI and MRS Separately and Combined Sensitivity and specificity of DWI and MRS were evaluated separately using ROC analysis; a combined evaluation of both methods was performed using a stepwise binary logistic (b) regression. The ROC curves of DWI and MRS for identification Figure 1: Patient (45 y.o.) with prostate cancer (2.6 cm): PSA-level of tumor tissue irrespective of the sextants are given in 9.7 μg/L (free PSA-level 14.56 μg/L); Gleason-Score 4 + 5 = 9. In the T2w image (a) the prostate cancer is demonstrated in the right Figure 3. The area under the ROC curve was 0.966 (95%- peripheral zone. In the correspondent ADC-map (b) the prostate confidence interval 0.941–0.991) and 0.943 (0.918–0.968) for cancer is clearly shown as a hypointense area. The left peripheral DWI and MRS, respectively. For DWI a sensitivity of 0.92 zone looks hyperintense on the T2w image, but the ADC-map and a specificity of 0.93 were provided using discrimination reveals the remaining healthy prostate tissue. 2 −3 threshold of 1.208 mm /s × 10 . With a threshold providing the same sensitivity, that is, 0.92, MRS provided a specificity of 0.85. With a threshold providing the same specificity, that is, 0.93, MRS provided a sensitivity of 0.68. the endorectal coil, all images could be used for analyses. Stepwise binary logistic regression included both DWI The image acquisition time for the T1 SE, and the T2 FSE (P< .001) and MRS (P< .001) for the differentiation was 24 minutes and that for the DWI were 5 minutes. The betweennormaltissueand tumorisdemonstratedin MRS lasted approximately 11 minutes. As demonstrated in Figure 4. The regression model classified 291 of the 300 Figure 1 the ADC maps show hyperintense in benign and probes correctly (97.0%). Only 4 of 238 normal tissue probes hypointense in malignant tissue in the peripheral zone. were misclassified as tumor, and 5 of 62 tumor probes were Histopathology identified tumor tissue in 21 of the 50 misclassified as normal. Thus, the regression model provided patients (42%). All 6 sextants were infiltrated by the tumor a sensitivity and specificity for detection of tumor of 91.9 and in 3 of these patients, 4 sextants in 3 patients, 3 sextants 98.3, respectively. in 5 patients, 2 sextants in 7 patients, and in 3 patients There was a highly significant negative correlation tumor tissue was detected in only 1 sextant. Thus, in total between DWI and the Gleason score in the tumor-positive 62 of the 300 tissue probes were tumor positive according to tissue probes (n = 62, ρ =−0.405, P = .001) (Figure 5). In histopathology (20.7%). contrast, MRS did not show a significant correlation with the Therateoftumor-positivetissueprobesrangedbetween Gleason score (ρ = 0.117, P = .366). 14% (left apex) and 28% (right midsection). However, this variation was not significant statistically (Pearson’s 2 2 “portmanteau” χ test: χ = 4.066, df = 5, P = .540). 5. Discussion The Gleason score of the tumor-positive tissue probes was 5 in 3 sextants (4.8%), 6 in 16 sextants (25.8%), 7 in However, the fact that the MRI of the prostate might be more 22 sextants (35.5%), 8 in 6 sextants (9.7%), 9 in 11 sextants advantageous than the transrectal ultrasound for staging the (17.7%), and 10 in 4 sextants (6.5%). cancer has also been discussed in the past (especially for T2 Radiology Research and Practice 5 DWI 0.8 0.6 Area = 0.966 0.4 0.2 Right apex Right base Left midsection 0 Right midsection Left apex Left base 0 0.2 0.4 0.6 0.8 1 1-specificity (a) Normal tissue Tumor MRS (a) 0.8 0.6 Area = 0.943 0.4 2 ∗ 0.2 0 0.2 0.4 0.6 0.8 1 1-specificity Right apex Right base Left midsection (b) Right midsection Left apex Left base Figure 3: Receiver operator characteristic (ROC) curve of DWI (a) Normal tissue and MRS (b) for differentiating tumor tissue from normal tissue. Tumor The analysis included all 300 tissue probes irrespective of the ROI. (b) Figure 2: Box-and-whisker plot of DWI (a) and MRS (b) as is one of the few organs in humans which can be examined a function of tissue type (normal tissue, tumor) and region of by MRI without any contrast media. Benign tissues in the interest (ROI). Outliers (1.5–3 box lengths) are indicated by a circle; peripheral zone show hyperintense signals in T2 weighted extreme values (>3 box lengths) are indicated by an asterix. imaging, whereas malignant changes show hypointense signals, of which the reason could be the cellular density as well as the malfunction of the gland when the malignant and T3 tumours) [17, 18]. The T2 weighted MRI images change had occurred. The cause of the decrease in diffusion of the prostate has been applied more often to improve the in malignant tissue has a histopathologic origin. Some validity of the staging in prostate cancer [19]. The prostate attributes are: hypercellularity, enlargement of the nuclei, DWI MRS Sensitivity Sensitivity 6 Radiology Research and Practice 56789 10 MRS Gleason score Normal tissue (a) Tumor Figure 4: Scatter plot of DWI versus MRS in normal tissue probes and tumor-positive tissue probes. hyperchromatism, and angulation of the nuclear contour, which lead to a reduction of diffusional displacement of water molecules (Anderson JR. Muir’s textbook of pathology. London, England: Edward Arnold 1985). Commonly, the prostate produces 20 to 30% of the ejaculate secretions. In patients with known prostatic cancer, the amount of the ejaculate can be less than in healthy patients. However, this is difficult to determine since the secretions vary from 0.5 to 13 mL. Diffusion weighted MR imaging has been clinically applied in several organs. Not only is it used to show the affected tissue after a stroke, it is also to differentiate brain tumours [20, 21] or also vertebral metastases in, for example, 56789 10 prostate cancer [22]. In diffusion-weighted MRI (DWI) the Gleason score image contrast is determined by the random microscopic (b) motion of water protons, that is, the Brownian motion. The diffusion can be measured in vivo by using the MRI because Figure 5: Scatter plot of DWI (a) and MRS (b) versus Gleason score of its sensitivity to motion. This sensitivity to motion can be in tumor-positive tissue probes. increased by the addition of strong magnetic field gradient pulses to the pulse sequence [23]. Shimofusa et al. applied the DWI of the prostate with parallel imaging and with a high b-value (b = 1000) for the first time [8]. In this study since 1980s. Since then, MR spectroscopy has been effective they have not used an endorectal coil, but the sensitivity as in improving the accuracy of MR imaging in prostate cancer well as specificity was higher than in other former studies localization and staging [32–34]. In the healthy prostate, with endorectal coil [24–27]oradynamicstudy [28]. Since malign and benign tissues can be differentiated by the MR then, several prostatic MR imaging modules were performed spectroscopy on the basis of the metabolic ratio of choline, to increase the detectability of cancerous tissue. The results of creatine, and citrate. The ratio is calculated by the equation these studies demonstrate that the ADC value may provide MET = (choline + creatine)/citrate. The MET is increased in information about the malignant changes in the prostate malign tissue whereas a lower MET can be found in benign [29–31]. tissue [35, 36]. MR spectroscopy is a relatively new method in diagnos- In this present study, 50 patients with suspected cancer ing prostate cancer and has been a part of clinical routine of the prostate were examined with the DW MRI and the DWI DWI MRS Radiology Research and Practice 7 MR spectroscopy, and then compared retrospectively with tissues (0.418±0.431). This result can also be compared with pathohistological results, especially with the Gleason score. previous studies. Two recent studies compared and analyzed the combined In this present study, a relatively larger number of usage of diffusion-weighted MRI and HMRspectroscopy. patients were examined than other studies. The combination The one study examined 42 patients with prostate cancer of DWI and MRS performed significantly better in detecting using a 2D chemical shift imaging and isotropic apparent cancer in the periphery zone of the prostate than MRS diffusion coefficient (ADC) maps [13]. In this study the alone [14] in one study. In our study with 50 patients, the regions of interest were drawn around the whole gland, combination of DWI and MRS seems also to have a better central gland, and the peripheral zone tumor. The mean accuracy in detecting cancerous tissue. The regression model ADC value of the normal tissue in this study was 1.51 mm /s classified 291 of the 300 probes correctly (97.0%). Only 4 of −3 × 10 [SD: ±0.27]. 238 normal tissue probes were misclassified as tumor, and If the tumor was greater than 30% of the whole voxel 5 of 62 tumor probes were misclassified as normal. Thus, 2 −3 the mean ADC value was 1.19 mm /s × 10 [SD: ±0.24], the regression model provided sensitivity and specificity for and if the tumor was greater than 70% of the whole voxel, detection of tumor of 91.9 and 98.3, respectively (Figure 5). 2 −3 the value was 1.03 mm /s × 10 [SD: ±0.18]. The mean If the results were correlated with the Gleason-score, MET in normal gland was 0.065 ± 0.052, whereas the value there was a highly significant negative correlation between was much higher in malignant tissues; 0.814 ± 2.202 in DWI and the Gleason score in the tumor-positive tissue tumor ≥30% of the voxel and 0.917± 1.276 in tumor ≥70%, probes (n = 62, ρ =−0.405, P = .001) (Figure 4). However, respectively. The MET was significantly higher (P< .001) MRS did not show a significant correlation with the Gleason and the ADC values were significantly (P< .006) lower score (ρ = 0.117, P = .366). One explanation could be that in tumor-containing voxel. The area under the ROC curves in the present study the MET values were extracted using a using both the ADC and MET was 0.81, similar to only very short TR sequence (700 ms) which keeps the scan time MET (0.79), whereas ADC alone showed an area of 0.66 and tolerable but results in spectra that are fairly heavily T1- was inferior. An interesting point, however, is the significant weighted. Perhaps that is why the other groups [38]found improvement in specificity for the combination of ADC and a weak correlation between Gleason and MET. MET, when voxels containing 70% or tumor were considered There are some problems and limitations in our study positive and cutoffs to achieve a 90% or greater sensitivity like in other MR studies of the prostate. First limitations were chosen [13]. are based on MR technique itself. Common artefacts for the The other study performed a retrospective measurement DWI are white pixel noise, low SNR (signal-to-noise ratio), of ADC and MET in 38 patients with prostatic cancer. and susceptibility artefacts. In MR spectroscopy common ThemeanADC valueand METfor malignanttissuewere artefacts are lipid contamination and susceptibility artefacts. 2 −3 1.39 mm /s × 10 [SD: ±0.23] and 0.92± 0.32, respectively. These artefacts could be reduced by new methods in the 2 −3 For benign tissue, the values were 1.69 mm /s × 10 [SD: future. Second limitation is the discrepancy in voxel sizes ±0.24] and 0.73 ± 0.18 (P< .001 for both). In this between the ADC map and the 3D HMRspectroscopy. In study, areas under the receiver operating characteristic curves the future new spectroscopic techniques with increased spa- (AUCs) were performed to evaluate the accuracy. Obviously, tial resolution without increasing the examination duration the combination of ADC and MET performed significantly [39] could be used to overcome this problem. The present better (AUC = 0.85; P = .005) than ADC or MET alone (AUC study was a retrospective analysis of the prostate cancer. The = 0.81 and AUC = 0.09, resp.) [14]. malignant sites were known when analyzing the ADC map Analogue to these prior studies, our results of the ADC and the MET. Chronic changes, such as chronic prostatitis values were significantly lower (P< .001) and the results or atrophy of the gland itself, usually show similar changes of the mean MET were significantly higher (P< .001) in MRI [4, 27] to prostatic cancer. Prospective studies for malignant prostatic tissues than for benign tissues. In are another research task for succeeding studies with the our study, the average ADC value in benign tissue was DWI and MRS, especially in patients with several negative 2 −3 1.65 mm /s × 10 [standard deviation (SD): ±0.32] and biopsies. The last limitation of this study was the usage of 2 −3 thatinmaligntissues 0.96mm /s × 10 [SD: ±0.24]. The sextant biopsy, since biopsies can be easily false negative so significantly (P< .001) lower ADC values in malign tissues that the tumor localization might not be very exact. These compared with the benign signify some promising results methods can then be combined with the dynamic contrast- in detecting the cancer. These results were similar to other enhanced magnetic resonance imaging, since first studies prior studies, although different b-values (0 and 1000 s/mm have shown that it may be an accurate technique for detecting and 0, 300, 600 s/mm )[30, 37] were used in our study and quantifying intracapsular transition or peripheral zone (i.e., 50, 400, 800 s/mm ). The cutoff value of the mean tumor foci greater than 0.2 cc [40]. ADC value between cancerous and noncancerous tissue in 2 −3 the present was at approximately 1.2 mm /s × 10 ,which 6. Conclusion canbeseenonthe ROC-Analyses (Figure 3). Compared to the aforementioned studies our results of the ADC value The results of our study showed that both diffusion-weighted for malignant tissue were remarkably lower. The MET of Imaging and MR spectroscopy should be considered as malignant tissues showed an average value of 2.010 ± 1.649 an additional and complementary tool to the T2 weighted in our study, and is significantly higher than in healthy MRI not only for detecting prostate cancer, but also for 8 Radiology Research and Practice guiding more specific biopsies without taking several samples to increase accuracy in prostate cancer detection,” American Journal of Roentgenology, vol. 188, no. 1, pp. 91–98, 2007. from the organ. The combination of these methods can [14] Y. Mazaheri, A. Shukla-Dave, H. Hricak et al., “Prostate improve the specificity and may prevent uncomfortable as cancer: identification with combined diffusion-weighted MR well as painful biopsies for patients. The DWI is furthermore imaging and 3D H MR spectroscopic imaging—correlation advantageous over the present examination methods, such with pathologic findings,” Radiology, vol. 246, no. 2, pp. 480– as the MR spectroscopy, regarding the considerably shorter 488, 2008. examination time (approximatly 5 minutes in case of this [15] D. Le Bihan, C. Poupon, A. Amadon, and F. Lethimonnier, study) and the correlation with the Gleason score. “Artifacts and pitfalls in diffusion MRI,” Journal of Magnetic Resonance Imaging, vol. 24, no. 3, pp. 478–488, 2006. [16] J. Kurhanewicz, D. B. Vigneron, H. Hricak et al., “Prostate References cancer: metabolic response to cryosurgery as detected with 3D H-1 MR spectroscopic imaging,” Radiology, vol. 200, no. 2, pp. [1] M.J.Hayat,N.Howlader, M. E. Reichman,and B. K. 489–496, 1996. Edwards, “Cancer statistics, trends, and multiple primary cancer analyses from the surveillance, epidemiology, and end [17] R. A. H. Boni, ¨ B. E. Hutter, F. Trinkler, W. Jochum, D. results (SEER) program,” Oncologist, vol. 12, no. 1, pp. 20–37, Pestalozzi, and G. P. Krestin, “Preoperative T-staging of prostate carcinoma: endorectal MRI compared with other imaging and clinical methods,” RoF ¨ o, vol. 165, no. 2, pp. 152– [2] F. Levi, F. Lucchini, E. Negri, P. Boyle, and C. La Vecchia, 158, 1996. “Leveling of prostate cancer mortality in Western Europe,” The Prostate, vol. 60, no. 1, pp. 46–52, 2004. [18] A. Wetter, A. N. Ajdukovic, K. Fliessbach et al., “Staging of prostate cancer: value of the combined information of [3] R. S. Purohit, K. Shinohara, M. V. Meng, and P. R. Carroll, “Imaging clinically localized prostate cancer,” Urologic Clinics endorectal MRI, biopsy Gleason score, and preoperative PSA level,” RoF ¨ o, vol. 178, no. 4, pp. 385–390, 2006. of North America, vol. 30, no. 2, pp. 279–293, 2003. [19] M. R. Engelbrecht, G. J. Jager, R. J. Laheij, A. L. M. Verbeek, H. [4] S. Ikonen, L. Kivisaari, P. Tervahartiala, T. Vehmas, K. Taari, and S. Rannikko, “Prostatic MR imaging: accuracy in J. van Lier, and J. O. Barentsz, “Local staging of prostate cancer using magnetic resonance imaging: a meta-analysis,” European differentiating cancer from other prostatic disorders,” Acta Radiology, vol. 12, no. 9, pp. 2294–2302, 2002. Radiologica, vol. 42, no. 4, pp. 348–354, 2001. [5] M. Sanchez-Chapado, J. C. Angulo, C. Ibarburen et al., [20] T. L. Chenevert, P. E. McKeever, and B. D. Ross, “Monitoring early response of experimental brain tumors to therapy “Comparison of digital rectal examination, transrectal ultra- sonography, and multicoil magnetic resonance imaging for using diffusion magnetic resonance imaging,” Clinical Cancer Research, vol. 3, no. 9, pp. 1457–1466, 1997. preoperative evaluation of prostate cancer,” European Urology, vol. 32, no. 2, pp. 140–149, 1997. [21] F. Yamasaki, K. Kurisu, K. Satoh et al., “Apparent diffusion [6] T. Franiel, F. Fritzsche, A. Staack, J. Rost, B. Hamm, and D. coefficient of human brain tumors at MR imaging,” Radiology, vol. 235, no. 3, pp. 985–991, 2005. Beyersdorff, “Histopathologic quality of prostate core biopsy specimens: comparison of an MR-compatible biopsy needle [22] T. Hacklander ¨ , C. Scharwa¨chter,R.Golz,and H. Mertens, and a ferromagnetic biopsy needle used for ultrasound-guided “Value of diffusion-weighted imaging for diagnosing vertebral prostate biopsy,” RoF ¨ o, vol. 178, no. 12, pp. 1212–1218, 2006. metastases due to prostate cancer in comparison to other [7] J. Kurhanewicz, D. B. Vigneron, H. Hricak, P. Narayan, primary tumors,” RoF ¨ o, vol. 178, no. 4, pp. 416–424, 2006. P. Carroll, and S. J. Nelson, “Three-dimensional H-1 MR [23] M. Cercignani, M. A. Horsfield, F. Agosta, and M. Filippi, spectroscopic imaging of the in situ human prostate with high “Sensitivity-encoded diffusion tensor MR imaging of the (0.24-0.7-cm3) spatial resolution,” Radiology, vol. 198, no. 3, cervical cord,” American Journal of Neuroradiology, vol. 24, no. pp. 795–805, 1996. 6, pp. 1254–1256, 2003. [8] R. Shimofusa, H. Fujimoto, H. Akamata et al., “Diffusion- [24] K. K. Yu and H. Hricak, “Imaging prostate cancer,” Radiologic weighted imaging of prostate cancer,” Journal of Computer Clinics of North America, vol. 38, no. 1, pp. 59–85, 2000. Assisted Tomography, vol. 29, no. 2, pp. 149–153, 2005. [25] H. Hricak, S. White, D. Vigneron et al., “Carcinoma of [9] K. Hosseinzadeh and S. D. Schwarz, “Endorectal diffusion- the prostate gland: MR imaging with pelvic phased-array weighted imaging in prostate cancer to differentiate malignant coils versus integrated endorectal-pelvic phased-array coils,” and benign peripheral zone tissue,” Journal of Magnetic Radiology, vol. 193, no. 3, pp. 703–709, 1994. Resonance Imaging, vol. 20, no. 4, pp. 654–661, 2004. [26] L. E. Quint, J. S. Van Erp, P. H. Bland et al., “Prostate [10] B. Issa, “In vivo measurement of the apparent diffusion cancer: correlation of MR images with tissue optical density at coefficient in normal and malignant prostatic tissues using pathologic examination,” Radiology, vol. 179, no. 3, pp. 837– echo-planar imaging,” Journal of Magnetic Resonance Imaging, 842, 1991. vol. 16, no. 2, pp. 196–200, 2002. [27] S. Ikonen, P. Kar ¨ kkainen, ¨ L. Kivisaari et al., “Endorectal [11] P. Gibbs, M. D. Pickles, and L. W. Turnbull, “Diffusion magnetic resonance imaging of prostatic cancer: comparison imaging of the prostate at 3.0 tesla,” Investigative Radiology, between fat-suppressed T2-weighted fast spin echo and three- vol. 41, no. 2, pp. 185–188, 2006. dimensional dual-echo, steady-state sequences,” European Radiology, vol. 11, no. 2, pp. 236–241, 2001. [12] V. Kumar, N. R. Jagannathan, R. Kumar et al., “Apparent diffusion coefficient of the prostate in men prior to biopsy: [28] G. J. Jager, E. T. G. Ruijter, C. A. van de Kaa et al., “Dynamic determination of a cut-off value to predict malignancy of the TurboFLASH subtraction technique for contrast-enhanced peripheral zone,” NMR in Biomedicine, vol. 20, no. 5, pp. 505– MR imaging of the prostate: correlation with histopathologic 511, 2007. results,” Radiology, vol. 203, no. 3, pp. 645–652, 1997. [29] V. A. Morgan,S.Kyriazi,S.E.Ashley, andN.M.DeSouza, [13] S. A. Reinsberg, G. S. Payne, S. F. Riches et al., “Combined use of diffusion-weighted MRI and HMRspectroscopy “Evaluation of the potential of diffusion-weighted imaging in Radiology Research and Practice 9 prostate cancer detection,” Acta Radiologica,vol. 48, no.6,pp. 695–703, 2007. [30] C. Sato, S. Naganawa, T. Nakamura et al., “Differentiation of noncancerous tissue and cancer lesion by apparent diffusion coefficient values in transition and peripheral zones of the prostate,” Journal of Magnetic Resonance Imaging, vol. 21, no. 3, pp. 258–262, 2005. [31] H. Miao, H. Fukatsu, and T. Ishigaki, “Prostate cancer detection with 3-T MRI: comparison of diffusion-weighted and T2-weighted imaging,” European Journal of Radiology, vol. 61, no. 2, pp. 297–302, 2007. [32] J. Scheidler, H. Hricak, D. B. Vigneron et al., “Prostate cancer: localization with three-dimensional proton MR spectroscopic imaging—clinicopathologic study,” Radiology, vol. 213, no. 2, pp. 473–480, 1999. [33] A. E. Wefer, H. Hricak, D. B. Vigneron et al., “Sextant localization of prostate cancer: comparison of sextant biopsy, magnetic resonance imaging and magnetic resonance spectro- scopic imaging with step section histology,” JournalofUrology, vol. 164, no. 2, pp. 400–404, 2000. [34] K. K. Yu, J. Scheidler, H. Hricak et al., “Prostate cancer: prediction of extracapsular extension with endorectal MR imaging and three-dimensional proton MR spectroscopic imaging,” Radiology, vol. 213, no. 2, pp. 481–488, 1999. [35] J. Kurhanewicz, D. B. Vigneron,R.G.Males,M.G.Swanson, K. K. Yu, and H. Hricak, “The prostate: MR imaging and spectroscopy: present and future,” Radiologic Clinics of North America, vol. 38, no. 1, pp. 115–138, 2000. [36] R. G. Males, D. B. Vigneron, J. Star-Lack et al., “Clinical application of BASING and spectral/spatial water and lipid suppression pulses for prostate cancer staging and localization by in vivo 3D H magnetic resonance spectroscopic imaging,” Magnetic Resonance in Medicine, vol. 43, no. 1, pp. 17–22, [37] A. Tanimoto, J. Nakashima, H. Kohno, H. Shinmoto, and S. Kuribayashi, “Prostate cancer screening: the clinical value of diffusion-weighted imaging and dynamic MR imaging in combination with T2-weighted imaging,” Journal of Magnetic Resonance Imaging, vol. 25, no. 1, pp. 146–152, 2007. [38] K. L. Zakian, K. Sircar, H. Hricak et al., “Correlation of proton MR spectroscopic imaging with gleason score based on step-section pathologic analysis after radical prostatectomy,” Radiology, vol. 234, no. 3, pp. 804–814, 2005. [39] T. W. J. Scheenen,D.W.J.Klomp,S.A.Rol ¨ l, J. J. Futt ¨ erer, J. O. Barentsz, and A. Heerschap, “Fast acquisition-weighted three- dimensional proton MR spectroscopic imaging of the human prostate,” Magnetic Resonance in Medicine,vol. 52, no.1,pp. 80–88, 2004. [40] A. Villers, P. Puech, D. Mouton, X. Leroy, C. Ballereau, and L. Lemaitre, “Dynamic contrast enhanced, pelvic phased array magnetic resonance imaging of localized prostate cancer for predicting tumor volume: correlation with radical prostatec- tomy findings,” JournalofUrology, vol. 176, no. 6, pp. 2432– 2437, 2006. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 International Journal of Journal of Immunology Research Endocrinology Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Submit your manuscripts at http://www.hindawi.com BioMed PPAR Research Research International Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Research and Treatment Cellular Longevity Neurology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal

Radiology Research and PracticeHindawi Publishing Corporation

Published: Jul 20, 2010

References