Skip to main content
NCBI home page
RSNA Journals logoLink to RSNA Journals
. 2010 Dec;257(3):715–723. doi: 10.1148/radiol.10100021

Prostate Cancer: Differentiation of Central Gland Cancer from Benign Prostatic Hyperplasia by Using Diffusion-weighted and Dynamic Contrast-enhanced MR Imaging1

Aytekin Oto ✉, Arda Kayhan, Yulei Jiang, Maria Tretiakova, Cheng Yang, Tatjana Antic, Farid Dahi, Arieh L Shalhav, Gregory Karczmar, Walter M Stadler
PMCID: PMC6939960  PMID: 20843992

Abstract

Purpose

To analyze the diffusion and perfusion parameters of central gland (CG) prostate cancer, stromal hyperplasia (SH), and glandular hyperplasia (GH) and to determine the role of these parameters in the differentiation of CG cancer from benign CG hyperplasia.

Materials and Methods

In this institutional review board–approved (with waiver of informed consent), HIPAA-compliant study, 38 foci of carcinoma, 38 SH nodules, and 38 GH nodules in the CG were analyzed in 49 patients (26 with CG carcinoma) who underwent preoperative endorectal magnetic resonance (MR) imaging and radical prostatectomy. All carcinomas and hyperplastic foci on MR images were localized on the basis of histopathologic correlation. The apparent diffusion coefficient (ADC), the contrast agent transfer rate between blood and tissue (Ktrans), and extravascular extracellular fractional volume values for all carcinoma, SH, and GH foci were calculated. The mean, standard deviation, 95% confidence interval (CI), and range of each parameter were calculated. Receiver operating characteristic (ROC) and multivariate logistic regression analyses were performed for differentiation of CG cancer from SH and GH foci.

Results

The average ADCs (× 10−3 mm2/sec) were 1.05 (95% CI: 0.97, 1.11), 1.27 (95% CI: 1.20, 1.33), and 1.73 (95% CI: 1.64, 1.83), respectively, in CG carcinoma, SH foci, and GH foci and differed significantly, yielding areas under the ROC curve (AUCs) of 0.99 and 0.78, respectively, for differentiation of carcinoma from GH and SH. Perfusion parameters were similar in CG carcinomas and SH foci, with Ktrans yielding the greatest AUCs (0.75 and 0.58, respectively). Adding Ktrans to ADC in ROC analysis to differentiate CG carcinoma from SH increased sensitivity from 38% to 57% at 90% specificity without noticeably increasing the AUC (0.79).

Conclusion

ADCs differ significantly between CG carcinoma, SH, and GH, and the use of them can improve the differentiation of CG cancer from SH and GH. Combining Ktrans with ADC can potentially improve the detection of CG cancer.

© RSNA, 2010

Supplemental material: http://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.10100021/-/DC1

Introduction

Central gland (CG) carcinoma composes about 30% of all prostate cancers (1). Although CG carcinoma tends to have low Gleason scores and low pathologic stages, up to 16% of such cancers demonstrate progression if untreated (2–6). CG cancer is a cause of false-negative findings at transrectal ultrasonographically guided biopsy, and, if detected and localized accurately, CG cancer can potentially be sampled for biopsy or treated by using emerging image-guided biopsy procedures and focal therapy. In addition, preoperative awareness of CG cancer by the surgeon may decrease the risk of positive anterior surgical margins at radical prostatectomy (6).

Evaluation of the CG with conventional T2-weighted magnetic resonance (MR) imaging is difficult because of the histologic heterogeneity of the CG (7,8). Broad and overlapping ranges of metabolite ratios observed at MR spectroscopy in CG cancer preclude cancer detection on the basis of analysis of a single metabolite (9,10). Although some initial results of diffusion-weighted (DW) MR imaging and dynamic contrast material–enhanced MR imaging were promising, the CG was usually regarded as histologically uniform, and a single diffusion or perfusion parameter was calculated to represent the entire benign CG (11–15). However, the benign CG is composed of two histologically distinct types of tissue: Glandular tissue, which is hyperintense on T2-weighted images, and stromal tissue, which is hypointense and can mimic CG cancer (16,17). To our knowledge, a dedicated study comparing DW and dynamic contrast-enhanced MR imaging features of different CG benign prostatic hyperplasia (BPH) subtypes and those of CG carcinoma has not been reported. Therefore, the purpose of this study was to analyze and compare the diffusion and perfusion parameters of CG prostate cancer, stromal hyperplasia (SH), and glandular hyperplasia (GH) and to determine the role of these parameters (individually or in combination) in the detection of CG cancer.

Materials and Methods

Study Patients

This retrospective study was conducted with an Institutional Review Board–approved waiver of informed consent and was in compliance with the Health Insurance Portability and Accountability Act. We searched patient records at our institution for patients who underwent endorectal MR imaging followed by radical prostatectomy between July 2007 and February 2009. We identified a total of 79 patients, of whom 26 (average age, 63.0 years; range, 48–74 years; median serum prostate-specific antigen level, 6.2 ng/mL; range, 3.4–38.8 ng/mL) had histopathologically identified CG cancer. An additional 23 consecutive patients from our search (average age, 62.4 years; range, 46–72 years) who had peripheral zone (PZ) cancer but no CG cancer were also included in this study to equalize the numbers of SH and GH foci versus CG cancers. Therefore, our study comprised a total of 49 patients (26 with and 23 without CG cancer). On average, there were 43.7 days (range, 7–118 days) between the MR imaging examination and the surgery.

MR Imaging Protocols

All MR imaging examinations were performed by using an endorectal coil (Medrad, Warrendale, Pa) combined with a phased array surface coil in 1.5-T MR imaging units (Excite HD, GE Healthcare, Waukesha, Wis [n = 41]; Achieva, Philips Healthcare, Eindhoven, the Netherlands [n = 8]), except that with the GE Healthcare imaging unit, dynamic contrast-enhanced MR images were obtained with a phased array coil only. Immediately before the MR imaging examination, 1 mg glucagon (Lilly, Indianapolis, Ind) was injected intramuscularly. We imaged the entire prostate and oriented axial images to be perpendicular to the rectal wall, guided by sagittal images. A parallel imaging factor of two was utilized in all sequences. The following images were obtained: axial, coronal, and sagittal T2-weighted fast spin-echo (SE) images (section thickness, 3 mm); axial T1-weighted fast SE images; axial free-breathing DW images (b = 0, 1000 or 1500 sec/mm2); and axial free-breathing dynamic contrast-enhanced MR images. Acquisition of dynamic contrast-enhanced MR images (of the entire prostate) started 30 seconds before intravenous administration of 0.1 mmol gadodiamide (Omniscan; GE Healthcare, Princeton, NJ) per kilogram of body weight followed by a 20-mL saline flush at a rate of 2.0 mL/sec. Detailed acquisition protocols are given in Appendix E1 (online).

MR Imaging–Histopathologic Correlation

After radical prostatectomy, the surgical specimen of the entire prostate was fixed in 5% buffered formalin for 24 hours. After dehydration, the specimen was cut serially into 4-mm-thick blocks from apex to base in transverse planes. Each block was then either halved or quartered (depending on its size), and microtome slices (thickness, 7–8 µm) were cut and stained with hematoxylin-eosin.

A genitourinary pathologist (M.T., with 7 years of experience in genitourinary pathology) reviewed hematoxylin-eosin–stained slices in the 26 patients with CG cancer, and, by using a four-quadrant (right anterior, right posterior, left anterior, and left posterior) approach, recorded on a schematic prostate diagram the size, location, Gleason score, and presence or absence of extraprostatic and seminal vesicle invasion of the CG carcinoma. A radiologist (A.O., with 7 years of experience in prostate MR imaging) determined the locations of the carcinoma on T2-weighted images on the basis of these diagrams and consultation with the pathologist. Each hematoxylin-eosin–stained slice was then visually matched to a corresponding T2-weighted MR image on the basis of the location of the ejaculatory ducts, the dimension of the prostate, any identifiable BPH nodule, and the approximate distance from the base or apex. To be considered a match, a focus of carcinoma must have been in the same anterior or posterior half of the CG and must have been at the same superior-to-inferior level of the prostate. All CG carcinoma foci larger than 5 mm in the greatest dimension histologically that were matched confidently to a T2-weighted MR image by consensus of the radiologist and the pathologist (M.T.) were included in this analysis (n = 38; average greatest dimension, 9.6 mm; range, 5–19 mm; Gleason score range, 7–9). If the same CG cancer was detected in more than one histopathologic slice and MR imaging section, it was still accepted as one tumor focus, and the MR image showing the lesion with the largest dimension was selected for the analysis. Regions of interest (ROIs) of CG carcinoma were then drawn manually on T2-weighted images and were subsequently determined automatically on other MR images by means of image-registration software in our picture archiving and communication system (i-site Radiology; Philips Healthcare).

MR images in the 26 patients with CG cancer were further reviewed to identify SH and GH foci. Histologically, SH and GH in the CG are well-defined nodular lesions with stromal or glandular tissue comprising more than 50% of the nodule, respectively. Well-defined, uniform, low- and high-signal-intensity areas within the CG, of 5 mm or larger in the greatest dimension, were identified with correlation to histopathologic findings and outlined manually on T2-weighted images as SH and GH foci, respectively. We correlated these hyperplastic foci with histopathologic features in a way that was similar to the way in which we ascertained the CG carcinoma, as detailed above. A total of 21 GH foci and 14 SH foci were identified in 17 and 14 patients with CG cancer, respectively. To equalize the numbers of SH, GH, and carcinoma foci (n = 38), an additional 17 GH and 24 SH foci in 23 patients without CG cancer were included, yielding a total of 38 CG cancers, 38 SH foci, and 38 GH foci in a total of 49 patients (26 with and 23 without CG cancer).

MR Image Analysis

T2-weighted image analysis.—The same radiologist (A.O.) interpreted T2-weighted images and recorded the following MR imaging features (7,8) for each CG carcinoma, SH focus, and CG focus: signal intensity on T2-weighted images (homogeneous or heterogeneous), margin (ill defined or well defined), capsule (present or absent), and shape (round or oval, amorphous, or lenticular). In addition, extracapsular extension (present or absent) and seminal vesicle invasion (present or absent) were recorded for each cancer focus.

Dynamic contrast-enhanced MR image analysis.—We used a previously published approximation (18) to convert the measured signal S(t) from T1-weighted dynamic contrast-enhanced MR imaging data acquired with short repetition and echo times, a median flip angle, and a standard contrast agent dose into contrast agent concentration:

graphic file with name 100021uneq1.jpg

where r1 is the relaxivity coefficient (approximately 4.5 mM−1 sec−1 at normal body temperature and 1.5 T for Omniscan). The conversion coefficient in the bracket was usually calculated by using a T1 value reported in the literature and a baseline signal intensity, both of a reference tissue. In this study, we used normal prostate tissue as the reference and used the T1 value of 1317 msec ± 85 (19).

The Tofts model (20) of time dependence of contrast agent concentration Ct(t) was applied to calculate the contrast agent transfer rate between blood and tissue (Ktrans) and the extravascular extracellular fractional volume (ve) on a voxel-by-voxel basis. Average parameters were then calculated within each ROI. When applying the Tofts model, for contrast agent arterial input function (AIF) we used the average AIFs estimated in a previous clinical dynamic contrast-enhanced MR imaging investigation (21) in which a multiple reference tissue method was used. It has been shown previously that a realistic population AIF can produce reproducible estimates of the kinetic parameters from dynamic contrast-enhanced MR imaging data (22) that correlate excellently with estimates from dynamic CT data (21). To enable visualization of the Ktrans and ve estimates, color maps of Ktrans and ve values were fused with T2-weighted images by means of an automated image-registration software developed at our institution.

DW image analysis.—Apparent diffusion coefficient (ADC) maps were generated from DW images with commercial diffusion-analysis software (Advantage Windows, version 4.2.3, GE Healthcare, Milwaukee, Wis; and Philips ViewForum, Philips Healthcare). Two radiologists (A.O. and A.K. [with 1 year of prostate MR imaging experience]) drew ROIs in consensus, using T2-weighted images as a reference and with the guidance of automatic image-registration software. In drawing the ROIs, the images were magnified, and the largest possible oval ROI (mean: 11.7 mm; range: 5.8–16.8 mm in the largest dimension) was placed on the area of interest. One ROI per each focus was selected. The average ADC within each ROI was then calculated.

Statistical Analysis

We calculated the mean, standard deviation, 95% confidence interval, and range of each MR imaging parameter. The linear mixed model was used to determine whether the distributions of the three types of lesions differed significantly and to identify which differences between the distributions of the lesion types were statistically significant (two-tailed P values). We also conducted receiver operating characteristic (ROC) analysis to measure the effectiveness of each MR imaging parameter in differentiating CG carcinoma from SH or GH foci. Note that GH (high-signal-intensity) and SH (low-signal-intensity) lesions are clearly distinguishable on T2-weighted MR images. Area under the ROC curve (AUC) was used as a summary index of diagnostic effectiveness, while sensitivity and specificity values were also calculated. The maximum-likelihood ROC curve estimation algorithms developed by Charles E. Metz were used in the ROC analysis (23–25).

We tested with multivariate logistic regression whether combining Ktrans values with ADCs improved the differentiation of CG cancer from SH foci. We also conducted an ROC analysis of combining the ADC and Ktrans parameters in the differentiation of CG carcinoma from SH foci by taking their average values (26). Because the ADC and Ktrans values differed by more than an order of magnitude, in the ROC analysis we first transformed the ADC and Ktrans values monotonically (but not necessarily linearly) to the domain of the latent decision variable of the ROC analysis (24) (as described previously [27]), such that numeric averaging of the ADC and Ktrans values became more meaningful, before obtaining their average values. P < .05 was considered to indicate a significant difference.

Results

Lesion Characteristics on T2-Weighted Images

T2-weighted MR imaging features of CG carcinoma, SH, and GH are summarized in Table 1. All 38 cancer foci demonstrated homogeneously low signal intensity (Fig 1). The majority of the cancer foci had an ill-defined margin (33 of 38 foci, or 87%) and an amorphous shape (30 of 38 foci, or 79%). Extracapsular extension was suspected in three cancer foci on the basis of T2-weighted MR imaging appearance but was confirmed histopathologically in only two (5%) of 38 foci. No cancer focus with seminal vesicle invasion was evident on T2-weighted MR images; this was confirmed histopathologically.

Table 1.

T2-weighted MR Imaging Characteristics of CG Carcinoma, SH, and GH

graphic file with name 100021t01.jpg

Open in a new tab

Note.—Data are numbers of lesions, with percentages in parentheses. NA = not applicable.

*

Confirmed histologically in two of three carcinoma foci.

Figure 1a:

Figure 1a:

Open in a new tab

Images in 76-year-old man with CG carcinoma with a Gleason score of 9 (5 + 4). (a) T2-weighted fast SE axial MR image shows ill-defined, hypointense carcinoma focus (arrows) occupying most of the left side of the CG and extending to the midline. The carcinoma (arrows) is dark on (b) ADC map, demonstrating restricted diffusion (measured ADC of the carcinoma = 0.73 × 10−3 mm2/sec). (c) Ktrans map color overlain on T2-weighted MR image reveals increased Ktrans values over the areas corresponding to the carcinoma (arrows) (measured Ktrans of the carcinoma = 0.13/min). (d) Histopathologic slide from the prostatectomy specimen shows the cancer (arrows) on the left side of the CG. (Hematoxylin-eosin stain.)

Figure 1b:

Figure 1b:

Open in a new tab

Images in 76-year-old man with CG carcinoma with a Gleason score of 9 (5 + 4). (a) T2-weighted fast SE axial MR image shows ill-defined, hypointense carcinoma focus (arrows) occupying most of the left side of the CG and extending to the midline. The carcinoma (arrows) is dark on (b) ADC map, demonstrating restricted diffusion (measured ADC of the carcinoma = 0.73 × 10−3 mm2/sec). (c) Ktrans map color overlain on T2-weighted MR image reveals increased Ktrans values over the areas corresponding to the carcinoma (arrows) (measured Ktrans of the carcinoma = 0.13/min). (d) Histopathologic slide from the prostatectomy specimen shows the cancer (arrows) on the left side of the CG. (Hematoxylin-eosin stain.)

Figure 1c:

Figure 1c:

Open in a new tab

Images in 76-year-old man with CG carcinoma with a Gleason score of 9 (5 + 4). (a) T2-weighted fast SE axial MR image shows ill-defined, hypointense carcinoma focus (arrows) occupying most of the left side of the CG and extending to the midline. The carcinoma (arrows) is dark on (b) ADC map, demonstrating restricted diffusion (measured ADC of the carcinoma = 0.73 × 10−3 mm2/sec). (c) Ktrans map color overlain on T2-weighted MR image reveals increased Ktrans values over the areas corresponding to the carcinoma (arrows) (measured Ktrans of the carcinoma = 0.13/min). (d) Histopathologic slide from the prostatectomy specimen shows the cancer (arrows) on the left side of the CG. (Hematoxylin-eosin stain.)

Figure 1d:

Figure 1d:

Open in a new tab

Images in 76-year-old man with CG carcinoma with a Gleason score of 9 (5 + 4). (a) T2-weighted fast SE axial MR image shows ill-defined, hypointense carcinoma focus (arrows) occupying most of the left side of the CG and extending to the midline. The carcinoma (arrows) is dark on (b) ADC map, demonstrating restricted diffusion (measured ADC of the carcinoma = 0.73 × 10−3 mm2/sec). (c) Ktrans map color overlain on T2-weighted MR image reveals increased Ktrans values over the areas corresponding to the carcinoma (arrows) (measured Ktrans of the carcinoma = 0.13/min). (d) Histopathologic slide from the prostatectomy specimen shows the cancer (arrows) on the left side of the CG. (Hematoxylin-eosin stain.)

Like CG cancer, the majority of SH foci also demonstrated homogeneously low signal intensity (32 of 38 foci, or 84%) and had ill-defined margins (28 of 38 foci, or 74%) (Fig 2). When these foci were heterogeneous, they were still predominantly hypointense, with focal, small areas of hyperintensity. Most SH foci had an amorphous shape (22 of 38 foci, or 58%), and the rest had a round or oval shape. There was no SH focus with a lenticular shape, and only three (8%) of 38 foci had a capsule.

Figure 2a:

Figure 2a:

Open in a new tab

Images in 64-year-old man with an SH nodule on the right side of the CG. (a) Axial T2-weighted fast SE MR image shows a round, well-defined, homogeneously hypointense nodule (arrows) representing SH. The nodule (arrows) can be detected on (b) a DW image (b = 1000 sec/mm2) but cannot be delineated on (c) an ADC map (ADC of nodule = 1.33 × 10−3 mm2/sec). (d) Fused Ktrans map and T2-weighted image reveals increased Ktrans throughout the SH nodule (arrow). (e) On the histopathologic slide (hematoxylin-eosin stain), the predominantly stromal component of the nodule (white arrow) is noted. Incidentally noted is a mixed glandular and stromal nodule (arrowhead in d and black arrow in e) on the left side of the CG that demonstrates similar Ktrans features.

Figure 2b:

Figure 2b:

Open in a new tab

Images in 64-year-old man with an SH nodule on the right side of the CG. (a) Axial T2-weighted fast SE MR image shows a round, well-defined, homogeneously hypointense nodule (arrows) representing SH. The nodule (arrows) can be detected on (b) a DW image (b = 1000 sec/mm2) but cannot be delineated on (c) an ADC map (ADC of nodule = 1.33 × 10−3 mm2/sec). (d) Fused Ktrans map and T2-weighted image reveals increased Ktrans throughout the SH nodule (arrow). (e) On the histopathologic slide (hematoxylin-eosin stain), the predominantly stromal component of the nodule (white arrow) is noted. Incidentally noted is a mixed glandular and stromal nodule (arrowhead in d and black arrow in e) on the left side of the CG that demonstrates similar Ktrans features.

Figure 2c:

Figure 2c:

Open in a new tab

Images in 64-year-old man with an SH nodule on the right side of the CG. (a) Axial T2-weighted fast SE MR image shows a round, well-defined, homogeneously hypointense nodule (arrows) representing SH. The nodule (arrows) can be detected on (b) a DW image (b = 1000 sec/mm2) but cannot be delineated on (c) an ADC map (ADC of nodule = 1.33 × 10−3 mm2/sec). (d) Fused Ktrans map and T2-weighted image reveals increased Ktrans throughout the SH nodule (arrow). (e) On the histopathologic slide (hematoxylin-eosin stain), the predominantly stromal component of the nodule (white arrow) is noted. Incidentally noted is a mixed glandular and stromal nodule (arrowhead in d and black arrow in e) on the left side of the CG that demonstrates similar Ktrans features.

Figure 2d:

Figure 2d:

Open in a new tab

Images in 64-year-old man with an SH nodule on the right side of the CG. (a) Axial T2-weighted fast SE MR image shows a round, well-defined, homogeneously hypointense nodule (arrows) representing SH. The nodule (arrows) can be detected on (b) a DW image (b = 1000 sec/mm2) but cannot be delineated on (c) an ADC map (ADC of nodule = 1.33 × 10−3 mm2/sec). (d) Fused Ktrans map and T2-weighted image reveals increased Ktrans throughout the SH nodule (arrow). (e) On the histopathologic slide (hematoxylin-eosin stain), the predominantly stromal component of the nodule (white arrow) is noted. Incidentally noted is a mixed glandular and stromal nodule (arrowhead in d and black arrow in e) on the left side of the CG that demonstrates similar Ktrans features.

Figure 2e:

Figure 2e:

Open in a new tab

Images in 64-year-old man with an SH nodule on the right side of the CG. (a) Axial T2-weighted fast SE MR image shows a round, well-defined, homogeneously hypointense nodule (arrows) representing SH. The nodule (arrows) can be detected on (b) a DW image (b = 1000 sec/mm2) but cannot be delineated on (c) an ADC map (ADC of nodule = 1.33 × 10−3 mm2/sec). (d) Fused Ktrans map and T2-weighted image reveals increased Ktrans throughout the SH nodule (arrow). (e) On the histopathologic slide (hematoxylin-eosin stain), the predominantly stromal component of the nodule (white arrow) is noted. Incidentally noted is a mixed glandular and stromal nodule (arrowhead in d and black arrow in e) on the left side of the CG that demonstrates similar Ktrans features.

The majority of GH foci were homogeneously hyperintense (26 of 38 foci, or 68%) (Fig 3). Heterogeneous foci were still predominantly hyperintense, containing small hypointense areas. Unlike SH and CG cancer, the majority of GH foci had capsules (33 of 38 foci, or 86.8%) and a round or oval shape (33 of 38 foci, or 87%).

Figure 3a:

Figure 3a:

Open in a new tab

Images in 67-year-old man with a GH nodule on the left side of the CG. (a) Axial T2-weighted fast SE MR image shows the well-defined, hyperintense nodule (white arrow) in the CG. (b) The GH nodule (white arrow) is bright on ADC map (ADC = 1.89 × 10−3 mm2/sec) and does not show any restricted diffusion. (c) Fused Ktrans map and T2-weighted image reveals an area of decreased Ktrans corresponding to the location of the GH nodule (white arrow). Incidentally noted is cancer (black arrow) on the left posterior side of the PZ on all images. (d) Histopathologic slide from the prostatectomy specimen reveals the GH on the left side of the CG (white arrow) and the cancer in the left PZ (black arrow). (Hematoxylin-eosin stain.)

Figure 3b:

Figure 3b:

Open in a new tab

Images in 67-year-old man with a GH nodule on the left side of the CG. (a) Axial T2-weighted fast SE MR image shows the well-defined, hyperintense nodule (white arrow) in the CG. (b) The GH nodule (white arrow) is bright on ADC map (ADC = 1.89 × 10−3 mm2/sec) and does not show any restricted diffusion. (c) Fused Ktrans map and T2-weighted image reveals an area of decreased Ktrans corresponding to the location of the GH nodule (white arrow). Incidentally noted is cancer (black arrow) on the left posterior side of the PZ on all images. (d) Histopathologic slide from the prostatectomy specimen reveals the GH on the left side of the CG (white arrow) and the cancer in the left PZ (black arrow). (Hematoxylin-eosin stain.)

Figure 3c:

Figure 3c:

Open in a new tab

Images in 67-year-old man with a GH nodule on the left side of the CG. (a) Axial T2-weighted fast SE MR image shows the well-defined, hyperintense nodule (white arrow) in the CG. (b) The GH nodule (white arrow) is bright on ADC map (ADC = 1.89 × 10−3 mm2/sec) and does not show any restricted diffusion. (c) Fused Ktrans map and T2-weighted image reveals an area of decreased Ktrans corresponding to the location of the GH nodule (white arrow). Incidentally noted is cancer (black arrow) on the left posterior side of the PZ on all images. (d) Histopathologic slide from the prostatectomy specimen reveals the GH on the left side of the CG (white arrow) and the cancer in the left PZ (black arrow). (Hematoxylin-eosin stain.)

Figure 3d:

Figure 3d:

Open in a new tab

Images in 67-year-old man with a GH nodule on the left side of the CG. (a) Axial T2-weighted fast SE MR image shows the well-defined, hyperintense nodule (white arrow) in the CG. (b) The GH nodule (white arrow) is bright on ADC map (ADC = 1.89 × 10−3 mm2/sec) and does not show any restricted diffusion. (c) Fused Ktrans map and T2-weighted image reveals an area of decreased Ktrans corresponding to the location of the GH nodule (white arrow). Incidentally noted is cancer (black arrow) on the left posterior side of the PZ on all images. (d) Histopathologic slide from the prostatectomy specimen reveals the GH on the left side of the CG (white arrow) and the cancer in the left PZ (black arrow). (Hematoxylin-eosin stain.)

Analysis of Dynamic Contrast-enhanced and DW MR Imaging Data

Examples of CG cancer, SH, and GH are shown in Figures 1–3, respectively. Table 2 shows results of the comparison of the diffusion and perfusion parameters ADC, Ktrans, and ve between CG carcinoma, SH foci, and GH foci. The mixed-model analysis of variance revealed that the distribution of ADCs differed significantly among CG carcinoma, SH foci, and GH foci (P < .001). In this analysis, multiple lesions in a given patient were subjected to random selection so that the analysis contained exactly one lesion of a given type from each prostate. The estimates of fixed effects showed that the distributions of ADCs differed significantly between CG carcinoma and SH foci (P = .003), between CG carcinoma and GH foci (P < .001), and between SH foci and GH foci (P < .001). Figure 4 depicts the individual ADCs in CG carcinoma, SH foci, and GH foci. Note the considerable overlap between CG carcinoma and SH foci, despite the significantly different average ADCs. The mixed-model analysis of variance showed that the distribution of Ktrans values differed significantly among CG carcinoma, SH foci, and GH foci (P = .005). In this analysis, multiple lesions in a given patient were subjected to random selection so that the analysis contained exactly one lesion of a given type from each prostate. The estimates of fixed effects showed that the distributions of Ktrans values differed significantly between SH foci and GH foci (P = .018) and between CG carcinoma and GH foci (P = .002), but not between CG carcinoma and SH foci (P = .288).

Table 2.

Results of Comparison of Diffusion and Perfusion Parameters between CG Carcinoma, SH Foci, and GH Foci

graphic file with name 100021t02.jpg

Open in a new tab

Figure 4:

Figure 4:

Open in a new tab

Scatterplot of ADCs for CG carcinoma, SH foci, and GH foci. Data points = individual ADCs, horizontal lines = average ADCs and associated 95% confidence intervals.

The ROC curves of the ADCs yielded AUCs of 0.99 and 0.78, respectively, for differentiation of carcinoma from GH and SH (Fig 5). Note that GH (high-signal-intensity) and SH (low-signal-intensity) lesions are clearly distinguishable on T2-weighted MR images. The perfusion parameters of CG carcinoma and SH foci were similar, with Ktrans yielding the greatest AUC among them (AUC = 0.58).

Figure 5:

Figure 5:

Open in a new tab

ROC curves of diffusion and perfusion parameters in the differentiation of CG carcinomas. Left: CG carcinomas versus SH foci. Right: CG carcinomas versus GH foci.

ADC and Ktrans Combined Analysis

Multivariate logistic regression showed that combining Ktrans values with ADCs did not significantly improve (P > .3) differentiation between CG carcinoma and SH foci compared with ADC alone. The ROC curves for combining the ADC and Ktrans parameters are shown in Figure 5. The AUC of the two parameters in combination (0.79 ± 0.05) was similar to the AUC for the ADC alone (0.78 ± 0.05). However, the shape of the ROC curve of the parameters in combination differed from that of the ADC alone: At a specificity 90%, the ROC curve of the ADC alone had a sensitivity value of 38%, whereas the ROC curve of the ADC and Ktrans combined achieved a sensitivity value of 57%. (This increase in sensitivity at a 90% specificity value did not reach statistical significance; P = .29 [25,28].) These results indicate that combining ADC and Ktrans possibly offers higher sensitivity at a low false-positive rate compared with using the ADC alone in the detection of CG carcinoma.

Discussion

Our results show that the performance of diffusion and perfusion parameters in differentiating cancer from BPH in the CG varies depending on histologic BPH subtypes. The ADC differs significantly between CG carcinoma, SH foci, and GH foci. Although quantitative perfusion parameters alone are not effective in the differentiation of CG cancer from SH, combining the Ktrans parameter with the ADC shows the possibility of improving this differentiation compared with using the ADC alone.

Conventional MR imaging is generally considered inadequate for evaluating CG cancers because of the heterogeneous T2 signal intensity in the normal transition zone. It has been suggested that homogeneous low T2 signal intensity, ill-defined margins, and lack of a capsule can be useful in the identification of CG cancer (7,8). Akin et al (7) proposed lenticular shape and invasion of the anterior fibromuscular stroma as additional features that can help differentiate cancer from benign CG. Even after inclusion of these new criteria, the detection rate of CG cancer on the basis of T2-weighted images alone remained between 56% and 63% (7). Similar to the results in the literature, all of the CG foci in our study showed homogeneously low signal intensity on T2-weighted images, and the majority (33 of 38) had ill-defined borders. GH foci could be easily differentiated from CG cancer owing to their predominantly hyperintense T2 signal and the presence of a capsule. However, SH foci had T2-weighted imaging characteristics (predominantly hypointense, ill-defined margins, and mostly amorphous shape without a capsule) that were similar to those of CG cancer. Differentiation of cancer foci from other phenomena that lead to hypointense foci on T2-weighted images, such as SH, atypical adenomatous hyperplasia, scarring, bleeding, and infection, still remains a problem that contributes to poor sensitivity and specificity.

Compared with the literature on PZ carcinoma, the literature on DW imaging characteristics of CG carcinoma and BPH is limited. In small studies (11–15,29), CG cancers demonstrated restricted diffusion, with smaller ADCs (0.93–1.37 × 10−3 mm2/sec) compared with BPH in the CG (1.34–1.79 × 10−3 mm2/sec). The average ADC of CG carcinoma in our series was 1.05 × 10−3 mm2/sec, which is concordant with values in the literature. Despite a consensus in multiple series of restricted diffusion in CG cancers, there is substantial variation in reported ADCs of benign CG tissue. Although there can be many technical and physiologic sources for this variation, we believe that the failure to recognize the histologic variability of the tissue and the use of a single ADC to represent the entire benign CG also contribute to this problem. As our study results demonstrate, the average ADCs differ significantly in SH and GH (1.27 vs 1.73 × 10−3 mm2/sec). Our results are similar to the differences in average ADC reported by Noworolski et al (30) between stromal and glandular prostatic tissue in the CG (1.18 vs 1.57 × 10−3 mm2/sec). SH is significantly more cellular and more dense and has less extracellular fluid than GH; these differences could explain the smaller ADCs in SH (31). A major challenge with T2-weighted images is to differentiate both hypointense SH and hyperintense GH from CG cancer. Our results indicates that, despite some overlap between the ADCs of CG cancer and those of SH, the ADC can help differentiate CG cancer from GH and, to a lesser degree, from SH. Recognition of the different histologic subtypes of BPH in the CG is important for radiologists in the quest to improve the diagnosis of CG cancer.

The role of dynamic contrast-enhanced MR imaging in the diagnosis of CG cancer is controversial. Padhani et al (32) and Deering et al (33) described a complete overlap in enhancement characteristics between carcinoma and BPH and explained this similarity by noting the increased microvessel density in BPH, which is similar to that in cancer. Contradictory to their results, Turnbull et al (34) reported that quantitative analysis of dynamic contrast-enhanced MR imaging data (specifically the amplitude of the initial contrast agent upslope and contrast agent exchange rate) can help distinguish carcinoma from fibroglandular and fibromuscular BPH. Later, Engelbrecht et al (35) suggested relative peak enhancement and washout as optimal parameters for discriminating carcinoma from benign CG, with carcinoma enhancing more and washing out faster than benign tissue. Our results show that SH and GH differ in enhancement characteristics and that SH enhances in way that is similar to CG cancer. For this reason, we believe that quantitative dynamic contrast-enhanced MR imaging parameters alone are not helpful for differentiating CG cancer from SH.

To date, there is not a single MR technique that has been shown to help detect and characterize prostate cancer adequately. However, in studies where diagnostic information from multiple techniques was combined (36–42), better performance was achieved than with use of a single parameter. This was true even when the individual parameters alone did not perform well enough (36). Langer et al (36) developed a multiparametric model using T2-weighted, DW, and dynamic contrast-enhanced MR imaging parameters to identify prostate cancer in the PZ. In their prospective series of 25 patients with PZ cancer, even though the P value of the Ktrans parameter did not reach statistical significance in their logistic regression model, and even though the AUC of Ktrans was only 0.592, the addition of Ktrans contributed significantly to the goodness of fit of their model. Similarly, in our study, Ktrans was ineffective in the differentiation of CG carcinoma from SH foci. But when Ktrans was combined with the ADC, performance improved, with an increase in sensitivity at a low false-positive rate, although this was not statistically significant. This improvement in performance was possible in part because of weak correlation between the ADC and Ktrans. This improvement in diagnostic performance achieved by combining diagnostic information from different sources can be explained on theoretic grounds (26). Prospective tests of our combination of ADC and Ktrans are necessary to confirm our findings and to evaluate the clinical usefulness of this approach to detecting CG carcinoma.

Our study had several limitations. First, it was a retrospective study with a relatively small sample size, which could be influenced by selection and verification biases. Second, we did not measure the T1 relaxation rate of CG tissue in each case. The lack of case-specific T1 relaxation rates could have affected the calculation of the quantitative perfusion parameters. Third, the spatial correlation of a lesion between MR images and histologic slices carries inherent limitations. Last, our case series included only carcinoma foci with Gleason scores between 7 and 9, in part because the urologists at our institution mainly refer patients with high risk of locally advanced tumors for preoperative MR imaging examinations.

In summary, our study results suggest that the combination of DW and dynamic contrast-enhanced MR imaging has the potential to improve the differentiation of CG cancer from BPH in the CG. Prospective studies are needed for further investigation of this potential advantage. Our results underscore the importance of recognizing different histologic subtypes of BPH because GH and SH have very different characteristics on T2-weighted, DW, and dynamic contrast-enhanced MR images. Their distinction in clinical and research studies will lead to more accurate analysis. Our results also show that although dynamic contrast-enhanced MR imaging alone is not effective in the differentiation of CG cancer from SH in the CG, the addition of Ktrans to the ADC could possibly improve the performance of this differentiation task.

Advances in Knowledge.

  • •. 

    Diffusion and perfusion MR imaging parameters of central gland stromal (apparent diffusion coefficient [ADC]: 1.27 × 10−3 sec/mm2; Ktrans: 0.094 per minute) and glandular hyperplasia (ADC: 1.73 × 10−3 sec/mm2; Ktrans: 0.068 per minute) differ significantly (P < .001); therefore, it can be misleading to calculate a single diffusion or perfusion parameter to represent the entire benign central gland.

  • •. 

    Diffusion-weighted imaging and calculation of ADC can potentially improve the detection of central gland carcinoma.

  • •. 

    Although dynamic contrast-enhanced MR imaging alone is not effective in the differentiation of central gland carcinoma from stromal hyperplasia (area under the receiver operating characteristic curve [AUC]: 0.58), combining the Ktrans and the ADC could possibly improve this differentiation (AUC: 0.79).

Implication for Patient Care.

  • •. 

    Diffusion-weighted and dynamic contrast-enhanced MR imaging can potentially improve the detection of central gland carcinoma.

APPENDIX

Appendix E1 (PDF)
appendix.pdf (41.9KB, pdf)

Acknowledgments

We thank Mrs Chuanhong Liao at the University of Chicago (Chicago, Ill) for her valuable assistance in the statistical analysis of the manuscript.

Received January 4, 2010; revision requested February 26; revision received April 7; accepted May 6; final version accepted June 28.

From the 2009 RSNA Annual Meeting.

Authors stated no financial relationship to disclose.

Abbreviations:

ADC
apparent diffusion coefficient
AUC
area under the ROC curve
BPH
benign prostatic hyperplasia
CG
central gland
DW
diffusion weighted
GH
glandular hyperplasia
Ktrans
contrast agent transfer rate between blood and tissue
PZ
peripheral zone
ROC
receiver operating characteristic
ROI
region of interest
SE
spin echo
SH
stromal hyperplasia
ve
extravascular extracellular fractional volume

References

  • 1.McNeal JE, Redwine EA, Freiha FS, Stamey TA. Zonal distribution of prostatic adenocarcinoma: correlation with histologic pattern and direction of spread. Am J Surg Pathol 1988;12(12):897–906. [DOI] [PubMed] [Google Scholar]
  • 2.Greene DR, Wheeler TM, Egawa S, Weaver RP, Scardino PT. Relationship between clinical stage and histological zone of origin in early prostate cancer: morphometric analysis. Br J Urol 1991;68(5):499–509. [DOI] [PubMed] [Google Scholar]
  • 3.Grignon DJ, Sakr WA. Zonal origin of prostatic adenocarcinoma: are there biologic differences between transition zone and peripheral zone adenocarcinomas of the prostate gland? J Cell Biochem Suppl 1994;19:267–269. [PubMed] [Google Scholar]
  • 4.Shannon BA, McNeal JE, Cohen RJ. Transition zone carcinoma of the prostate gland: a common indolent tumour type that occasionally manifests aggressive behaviour. Pathology 2003;35(6):467–471. [DOI] [PubMed] [Google Scholar]
  • 5.Blute ML, Zincke H, Farrow GM. Long-term followup of young patients with stage A adenocarcinoma of the prostate. J Urol 1986;136(4):840–843. [DOI] [PubMed] [Google Scholar]
  • 6.Epstein JI, Paull G, Eggleston JC, Walsh PC. Prognosis of untreated stage A1 prostatic carcinoma: a study of 94 cases with extended followup. J Urol 1986;136(4):837–839. [DOI] [PubMed] [Google Scholar]
  • 7.Akin O, Sala E, Moskowitz CS, et al. Transition zone prostate cancers: features, detection, localization, and staging at endorectal MR imaging. Radiology 2006;239(3):784–792. [DOI] [PubMed] [Google Scholar]
  • 8.Li H, Sugimura K, Kaji Y, et al. Conventional MRI capabilities in the diagnosis of prostate cancer in the transition zone. AJR Am J Roentgenol 2006;186(3):729–742. [DOI] [PubMed] [Google Scholar]
  • 9.Zakian KL, Eberhardt S, Hricak H, et al. Transition zone prostate cancer: metabolic characteristics at 1H MR spectroscopic imaging—initial results. Radiology 2003;229(1):241–247. [DOI] [PubMed] [Google Scholar]
  • 10.Shukla-Dave A, Hricak H, Eberhardt SC, et al. Chronic prostatitis: MR imaging and 1H MR spectroscopic imaging findings—initial observations. Radiology 2004;231(3):717–724. [DOI] [PubMed] [Google Scholar]
  • 11.Kim CK, Park BK, Han JJ, Kang TW, Lee HM. Diffusion-weighted imaging of the prostate at 3 T for differentiation of malignant and benign tissue in transition and peripheral zones: preliminary results. J Comput Assist Tomogr 2007;31(3):449–454. [DOI] [PubMed] [Google Scholar]
  • 12.Kim JH, Kim JK, Park BW, Kim N, Cho KS. Apparent diffusion coefficient: prostate cancer versus noncancerous tissue according to anatomical region. J Magn Reson Imaging 2008;28(5):1173–1179. [DOI] [PubMed] [Google Scholar]
  • 13.Tamada T, Sone T, Jo Y, et al. Apparent diffusion coefficient values in peripheral and transition zones of the prostate: comparison between normal and malignant prostatic tissues and correlation with histologic grade. J Magn Reson Imaging 2008;28(3):720–726. [DOI] [PubMed] [Google Scholar]
  • 14.Sato C, Naganawa S, Nakamura T, et al. Differentiation of noncancerous tissue and cancer lesions by apparent diffusion coefficient values in transition and peripheral zones of the prostate. J Magn Reson Imaging 2005;21(3):258–262. [DOI] [PubMed] [Google Scholar]
  • 15.Kim CK, Park BK, Lee HM, Kwon GY. Value of diffusion-weighted imaging for the prediction of prostate cancer location at 3T using a phased-array coil: preliminary results. Invest Radiol 2007;42(12):842–847. [DOI] [PubMed] [Google Scholar]
  • 16.Schiebler ML, Tomaszewski JE, Bezzi M, et al. Prostatic carcinoma and benign prostatic hyperplasia: correlation of high-resolution MR and histopathologic findings. Radiology 1989;172(1):131–137. [DOI] [PubMed] [Google Scholar]
  • 17.Ishida J, Sugimura K, Okizuka H, et al. Benign prostatic hyperplasia: value of MR imaging for determining histologic type. Radiology 1994;190(2):329–331. [DOI] [PubMed] [Google Scholar]
  • 18.Medved M, Karczmar G, Yang C, et al. Semiquantitative analysis of dynamic contrast enhanced MRI in cancer patients: Variability and changes in tumor tissue over time. J Magn Reson Imaging 2004;20(1):122–128. [DOI] [PubMed] [Google Scholar]
  • 19.de Bazelaire CMJ, Duhamel GD, Rofsky NM, Alsop DC. MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results. Radiology 2004;230(3):652–659. [DOI] [PubMed] [Google Scholar]
  • 20.Tofts PS, Brix G, Buckley DL, et al. Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: standardized quantities and symbols. J Magn Reson Imaging 1999;10(3):223–232. [DOI] [PubMed] [Google Scholar]
  • 21.Yang C, Stadler WM, Karczmar GS, Milosevic MF, Yeung I, Haider MA. Comparison of quantitative parameters in cervix cancer measured by dynamic contrast-enhanced MRI and CT. Magn Reson Med 2010;63(6):1601–1609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Parker GJ, Roberts C, Macdonald A, et al. Experimentally-derived functional form for a population-averaged high-temporal-resolution arterial input function for dynamic contrast-enhanced MRI. Magn Reson Med 2006;56(5):993–1000. [DOI] [PubMed] [Google Scholar]
  • 23.PROPROC and CLABROC http://tinyurl.com/236aoq. Accessed June 3, 2010.
  • 24.Metz CE, Pan X. “Proper” binormal ROC curves: theory and maximum-likelihood estimation. J Math Psychol 1999;43(1):1–33. [DOI] [PubMed] [Google Scholar]
  • 25.Metz CE, Wang PL, Kronman HB. A new approach for testing the significance of differences between ROC curves measured from correlated data. In: , Deconinck F, ed. Information processing in medical imaging. The Hague, the Netherlands: Nijhoff, 1984; 432–445. [Google Scholar]
  • 26.Metz CE, Shen JH. Gains in accuracy from replicated readings of diagnostic images: prediction and assessment in terms of ROC analysis. Med Decis Making 1992;12(1):60–75. [DOI] [PubMed] [Google Scholar]
  • 27.Jiang Y, Metz CE. A quadratic model for combining quantitative diagnostic assessments from radiologist and computer in computer-aided diagnosis. Acad Radiol 2006;13(2):140–151. [DOI] [PubMed] [Google Scholar]
  • 28.CLABROC http://tinyurl.com/236aoq. Accessed June 3, 2010.
  • 29.Yoshizako T, Wada A, Hayashi T, et al. Usefulness of diffusion-weighted imaging and dynamic contrast-enhanced magnetic resonance imaging in the diagnosis of prostate transition-zone cancer. Acta Radiol 2008;49(10):1207–1213. [DOI] [PubMed] [Google Scholar]
  • 30.Noworolski SM, Vigneron DB, Chen AP, Kurhanewicz J. Dynamic contrast-enhanced MRI and MR diffusion imaging to distinguish between glandular and stromal prostatic tissues. Magn Reson Imaging 2008;26(8):1071–1080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.McNeal JE, Bostwick DG. Anatomy of the prostate: implications for disease. In: Pathology of the prostate. New York, NY: Churchill-Livingstone, 1990; 1–14. [Google Scholar]
  • 32.Padhani AR, Gapinski CJ, Macvicar DA, et al. Dynamic contrast enhanced MRI of prostate cancer: correlation with morphology and tumour stage, histological grade and PSA. Clin Radiol 2000;55(2):99–109. [DOI] [PubMed] [Google Scholar]
  • 33.Deering RE, Bigler SA, Brown M, Brawer MK. Microvascularity in benign prostatic hyperplasia. Prostate 1995;26(3):111–115. [DOI] [PubMed] [Google Scholar]
  • 34.Turnbull LW, Buckley DL, Turnbull LS, Liney GP, Knowles AJ. Differentiation of prostatic carcinoma and benign prostatic hyperplasia: correlation between dynamic Gd-DTPA-enhanced MR imaging and histopathology. J Magn Reson Imaging 1999;9(2):311–316. [DOI] [PubMed] [Google Scholar]
  • 35.Engelbrecht MR, Huisman HJ, Laheij RJ, et al. Discrimination of prostate cancer from normal peripheral zone and central gland tissue by using dynamic contrast-enhanced MR imaging. Radiology 2003;229(1):248–254. [DOI] [PubMed] [Google Scholar]
  • 36.Langer DL, van der Kwast TH, Evans AJ, Trachtenberg J, Wilson BC, Haider MA. Prostate cancer detection with multi-parametric MRI: logistic regression analysis of quantitative T2, diffusion-weighted imaging, and dynamic contrast-enhanced MRI. J Magn Reson Imaging 2009;30(2):327–334. [DOI] [PubMed] [Google Scholar]
  • 37.Kozlowski P, Chang SD, Jones EC, Berean KW, Chen H, Goldenberg SL. Combined diffusion-weighted and dynamic contrast-enhanced MRI for prostate cancer diagnosis: correlation with biopsy and histopathology. J Magn Reson Imaging 2006;24(1):108–113. [DOI] [PubMed] [Google Scholar]
  • 38.Mazaheri Y, Shukla-Dave A, Hricak H, et al. Prostate cancer: identification with combined diffusion-weighted MR imaging and 3D 1H MR spectroscopic imaging—correlation with pathologic findings. Radiology 2008;246(2):480–488. [DOI] [PubMed] [Google Scholar]
  • 39.Reinsberg SA, Payne GS, Riches SF, et al. Combined use of diffusion-weighted MRI and 1H MR spectroscopy to increase accuracy in prostate cancer detection. AJR Am J Roentgenol 2007;188(1):91–98. [DOI] [PubMed] [Google Scholar]
  • 40.van Dorsten FA, van der Graaf M, Engelbrecht MR, et al. Combined quantitative dynamic contrast-enhanced MR imaging and (1)H MR spectroscopic imaging of human prostate cancer. J Magn Reson Imaging 2004;20(2):279–287. [DOI] [PubMed] [Google Scholar]
  • 41.Tanimoto A, Nakashima J, Kohno H, Shinmoto H, Kuribayashi S. Prostate cancer screening: the clinical value of diffusion-weighted imaging and dynamic MR imaging in combination with T2-weighted imaging. J Magn Reson Imaging 2007;25(1):146–152. [DOI] [PubMed] [Google Scholar]
  • 42.Chen M, Dang HD, Wang JY, et al. Prostate cancer detection: comparison of T2-weighted imaging, diffusion-weighted imaging, proton magnetic resonance spectroscopic imaging, and the three techniques combined. Acta Radiol 2008;49(5):602–610. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix E1 (PDF)
appendix.pdf (41.9KB, pdf)

Articles from Radiology are provided here courtesy of Radiological Society of North America

RESOURCES