Detection of Radiorecurrent Prostate Cancer Using ...

Genitourinar y Imaging • Original Research Rud et al. DWI and Targeted Biopsies for Radiorecurrent Prostate Cancer

Downloaded from www.ajronline.org by UiO-Universitetsbiblioteket on 11/24/14 from IP address 193.157.230.207. Copyright ARRS. For personal use only; all rights reserved

Genitourinary Imaging Original Research

Detection of Radiorecurrent Prostate Cancer Using DiffusionWeighted Imaging and Targeted Biopsies Erik Rud1 Eduard Baco 2 Diep Lien 3 Dagmar Klotz 4 Heidi B. Eggesbø5 Rud E, Baco E, Lien D, Klotz D, Eggesbø HB

Keywords: diffusion-weighted imaging (DWI), MRI, radiorecurrent prostate cancer, soft image fusion, targeted biopsy DOI:10.2214/AJR.12.10483 E. Rud and E. Baco contributed equally to this study. Received December 20, 2012; accepted after revision June 26, 2013. E. Rud received a fee from Kungshusen Medicinska AB for giving a scientific presentation about targeted biopsies for the Swedish Urological Society in Malmö, Sweden. The AKTIV Against Cancer Foundation funded the purchase of the ultrasound equipment and navigation software used in this study. 1 Department of Radiology and Nuclear Medicine, Division of Diagnostics and Intervention, Oslo University Hospital, Aker, Postboks 4959, Nydalen, 0424 Oslo, Norway. Address correspondence to E. Rud ([email protected]). 2 Department of Urology, Oslo University Hospital, Aker, Oslo, Norway. 3 Unit of Biostatistics and Epidemiology, Oslo University Hospital, Ullevål, Oslo, Norway. 4 Department of Pathology, Oslo University Hospital, Ullevål, Oslo, Norway. 5 Department of Radiology and Nuclear Medicine, Oslo University Hospital, Rikshospitalet, Oslo, Norway.

WEB This is a web exclusive article. AJR 2014; 202:W241–W246 0361–803X/14/2023–W241 © American Roentgen Ray Society

OBJECTIVE. The primary purpose of this study was to evaluate the detection rate of local radiorecurrent prostate cancer by using diffusion-weighted MR imaging (DWI) and targeted biopsies. The secondary purpose was to assess the value of performing random biopsies. MATERIALS AND METHODS. This study included 42 consecutive patients with biochemical recurrence after external beam radiation therapy (EBRT). At the time of biopsy, the mean age ± SD was 67 ± 6 years, median serum prostate-specific antigen level was 4.0 ± 3.0 ng/mL, and mean elapsed time between EBRT and biopsy was 5.6 ± 2.8 years. MRI examination included high-resolution axial T2-weighted and DWI sequences and was classified as either negative or positive. Transrectal ultrasound–guided targeted biopsies were obtained from all patients with positive findings on MRI using a soft image fusion system. Random sextant biopsies were obtained from both lobes in patients with negative findings on MRI and from the lobe contralateral to the MRI target in patients with positive findings on MRI. The biopsy results were classified as negative or positive and defined as the criterion standard. RESULTS. MRI findings were positive in 40 of 42 (95%) patients, and the overall positive biopsy rate was 79% (33 of 42 patients). Targeted biopsies were positive in 33 of 40 (83%) patients. Random biopsies were positive in 6 of 30 (20%) patients, all of whom had positive targeted biopsies. CONCLUSION. DWI is highly sensitive for detecting radiorecurrent prostate cancer, and a few targeted biopsies may confirm a positive diagnosis. However, random biopsies may assess the tumor burden more exactly.

R

ecurrence of prostate cancer after external beam radiation therapy (EBRT) is known as radiorecurrent prostate cancer. This is strictly a biochemical diagnosis, most commonly defined as an increase in serum prostate-specific antigen (PSA) level of more than 2.0 ng/ mL from the PSA nadir or increased PSA levels on two consecutive measurements after the completion of treatment [1, 2]. The diagnosis does not distinguish recurrent disease confined to the gland from metastatic disease. In case of local recurrence without metastasis, focal salvage treatment usually requires a positive biopsy from the prostate gland [3, 4]. After EBRT, the prostate gland has lost the typical zonal anatomy on MRI and instead appears with homogeneously low T2-weighted signal intensity in both peripheral and transitional zones. It is well established that T2-weighted images alone are insufficient for the detection of prostate cancer [5–8]. The European Society of Urogenital Radiology (ESUR) 2012 recom-

mends both dynamic contrast-enhanced (DCEMRI) and diffusion-weighted (DWI) MR images in addition to T2-weighted images for detection of radioreccurent prostate cancer [9]. Recent studies have shown that DWI is equally sensitive and more specific compared with DCE-MRI for detecting local recurrence after EBRT [7, 8, 10]. The restricted diffusion of water molecules in cancerous tissue compared with normal tissue is a valuable tool for identifying possible cancer and may be quantified by calculating the apparent diffusion coefficient (ADC) from DWI. In untreated highgrade prostate cancer, the inverse correlation between the ADC value and the Gleason score has been well established [11–13]. Therefore, DWI may be useful for directing the biopsy toward the areas with the most aggressive tumor [14]. Different navigation systems exist to guide the biopsy needle, and fusion of MRI and transrectal ultrasound is a promising method to obtain accurate and image-documentable targeted biopsies [15–18].

AJR:202, March 2014 W241

Rud et al. TABLE 1: MRI Sequences and Acquisition Parameters Sequence Downloaded from www.ajronline.org by UiO-Universitetsbiblioteket on 11/24/14 from IP address 193.157.230.207. Copyright ARRS. For personal use only; all rights reserved

Sagittal T2 turbo spin echo

FOV (mm)

Matrix

Voxel Size (mm)

TR/TE

Bandwidth (Hz/voxel)

380 × 272

162 × 256

1.7 × 1.2 × 5.0

2900/97

201

b (s/mm2)

Scan Time 45 s

Axial T2spc

292 × 292

384 × 387

0.9 × 0.9 × 0.9

2000/123

650

Axial DWI b50/b1000

300 × 300

128 × 128

2.3 × 2.3 × 4.0

2600/81

1648

50, 1000

3 min 15 s

7 min 2 s

Axial DWI b2000

250 × 250

68 × 114

3.7 × 2.2 × 5.0

2400/120

1512

2000

1 min 34 s

Note—T2spc = high-resolution 3D acquisition with 0.9-mm isotropic voxels, DWI = diffusion-weighted imaging.

The primary purpose of this study was to evaluate the detection rate of local radiorecurrent prostate cancer by using DWI and targeted biopsies. The secondary purpose was to assess the value of performing random biopsies. Materials and Methods The local institutional review board granted a waiver for signed informed consent and approved this study. From December 2010 to October 2012, 42 consecutive patients fulfilled the criteria for radiorecurrent prostate cancer and were included. These criteria were two increasing consecutive PSA measurements or PSA nadir of 2.0 ng/mL (or both). The following additional criterion based on PSA level was used to select candidates for focal treatment: PSA doubling time of more than 1 year without increase in PSA level during hormonal treatment. All patients underwent MRI at our institution before biopsy. Mean ± SD patient age at the time of biopsy was 67 ± 6 years (range, 49–78 years). Mean elapsed time since treatment was 5.6 ± 2.8 years (range, 1–12 years) (missing data in two patients). Median ± in-

A

B W242

terquartile range PSA levels at the time of EBRT, PSA nadir, and biopsy, respectively, were as follows: 15.0 ± 18.4, 0.4 ± 1.1, and 4.0 ± 3.0 ng/mL (missing data in 11, 13, and zero patients, respectively). Gleason score at the time of EBRT was 6, 7a, 7b, 8, and 9 in 11, 10, 11, three, and two patients, respectively (missing data in five patients). Clinical T classification at the time of EBRT was 1c, 2, 3a, and 3b, in six, seven, 16, and five patients, respectively (missing data in eight patients). Standard random biopsy sessions had been performed in 29 of 42 (71%) patients before this study (mean, 1.1; range, 1–2); in these 29 patients, 17 (59%) had positive findings and 12 (41%) had negative findings. All patients were treated with EBRT using mean absorbed dose of 72 Gy (range, 70–78 Gy). Previous hormonal treatments had been administered as follows: neoadjuvant, five of 42 patients; adjuvant, six of 42 patients; combined neoadjuvant and adjuvant, 10 of 42 patients; and no hormonal treatment, 15 of 42 patients (missing data in six of 42 patients). Four patients received hormonal treatment at the time of the prebiopsy MRI owing to biochemical recurrence.

All patients were also examined for lymph node and skeletal metastases, and 11 patients underwent DCE-MRI in addition to DWI. These results are not considered in this article.

MRI Examination All MRI examinations were performed before the biopsy session (median time ± SD until biopsy, 0 ± 1.0 months; range, 0–13 months) on a 1.5-T Avanto MRI scanner (Siemens) using a 6-channel body matrix coil (Siemens). The following sequences were acquired: high-resolution axial T2-weighted MRI and DWI. The DWI examination consisted of two parts: b50/b1000 and ADC map; and b2000. No cleansing enema was administered. All examinations were evaluated by a single radiologist with 5 years’ cumulative experience in prostate MRI. The MRI sequences and acquisition parameters are summarized in Table 1. Postprocessing procedures—We used nordicICE (version 2.3.6, NordicNeuroLab) to colorhighlight the DWI parameters among T2-corrected b1000, b2000, and ADC.

Fig. 1—Patient with biochemical recurrence 8 years after external beam radiation therapy. MRI with positive findings at local hospital was followed by 17-core transperineal biopsy procedure with negative result. A, T2-corrected b1000 overlay shows highly suspicious tumor (arrows). B–D, Axial (B), sagittal (C), and coronal (D) images of targeted biopsy show how it appears to surgeon during procedure. Red cylinder within target = positive, green cylinder outside target = negative. Patient was later scheduled for salvage high-intensity focal ultrasound of left base.

C

D AJR:202, March 2014


DWI and Targeted Biopsies for Radiorecurrent Prostate Cancer Definition of tumor-suspicious areas—The MRI findings were considered as positive if a tumorsuspicious region was present. Because of the effect of EBRT, the prostate gland often appeared without typical zonal anatomy, showing only homogeneously low T2-weighted signal intensity. In these situations, the identification of tumor was mainly based on DWI alone. T2-weighted images were used if possible but served mainly as a morphologic map when applying the color overlays. A high signal intensity relative to that in muscle on T2-corrected b1000 and b2000, together with a corresponding low ADC signal intensity, was defined as a cancer-suspicious area. The ADC cutoff values for defining these areas were less than 0.1 × 10−3 mm2/s in the transitional zone and less than 1.2 × 10−3 mm2/s in the peripheral zone [16]. When zonal anatomy was difficult to determine, the anterior part of the gland was defined as the transitional zone and the posterior part as the peripheral zone. All tumor-suspicious regions were defined before the biopsy session. Volume estimation—The estimated volumes (in mL) of the prostate glands and the MRI targets were based on the largest measured diameters (in mm) and the volume formula: length × depth × width / 2000. Only foci with a diameter of more than 5 mm were included. The measurements were based on the ADC map and T2-weighted images when possible. Volume measurements were mainly performed retrospectively. ADC values—The lowest ADC value in a target was measured in the color overlay image using a 0.2-cm2 circular region of interest (ROI). One control value was obtained from an assumed noncancerous region in the ventral part of the transition zone, and one control value was obtained from an assumed noncancerous region in the peripheral zone. Registration of absolute values was mainly performed retrospectively.

TABLE 2: Number of Biopsy Cores According to the Biopsy Procedure in 42 Patients No. of Biopsy Cores Total

Positive

Targeted (n = 12)

Biopsy Procedure(s)

3.5 ± 2.4

1.5 ± 2.2

Random (n = 2)

14.0 ± 2.8

0 ± 0

Targeted and random (n = 28)

13.0 ± 4.8

3.0 ± 4.3

Note—Data are median ± interquartile range.

gets were 4–6 mm (corresponding volume, 0.03–0.1 mL). Second, a transrectal ultrasound examination obtained a 3D volume of the prostate gland using an ultrasound probe with an internally rotating head. The third and final step involved real-time software fusion of the MRI and ultrasound volumes. After each biopsy, the biopsy track was displayed in a 3D image, giving immediate feedback as to whether the biopsy was inside the target. This is in accordance with the process previously described by Rud et al. [16]. The key features of the navigation system are shown in Figures 1 and 2. All biopsies were performed by a single operator using a real-time transrectal ultrasound endfire probe (3D Accuvix V10, Medison) with an 18-gauge × 25-cm biopsy needle (Tru-Core II, Angiotech). All needle tracks were saved in 3D for retrospective evaluation.

Histopathologic Procedure The biopsies were examined by staff uropathologists and used as the criterion standard. Because performance of Gleason scoring secondary to EBRT is debated and no consensus exists, biopsies were classified only as positive or negative [20]. Atypical small acinar cell proliferation,

Classification of Biopsy Groups Targeted biopsies—A minimum of a single biopsy (range, 1–3 biopsies) was obtained from each MRI target. No random sextant biopsy procedure was performed. The procedure findings were considered as positive if a minimum of one biopsy was classified as positive. Random biopsies—A random sextant biopsy procedure was performed from both lobes in patients with negative findings on MRI. The procedure findings were considered as positive if a minimum of one biopsy in either lobe was classified as positive. Targeted and random biopsies—A minimum of a single biopsy (range, 1–3 biopsies) was obtained from each MRI target and classified as described earlier. In addition, a random sextant biopsy procedure was performed in the contralateral lobe, and findings were considered as positive if a minimum of one positive biopsy was obtained. All patients in the groups that underwent targeted biopsy and that underwent targeted and ran-

TABLE 3: Clinical Data and Biopsy Results in 40 Patients With Positive Findings on MRI Biopsy Results

MRI and Transrectal Ultrasound Soft Image Fusion and Targeted Biopsies Soft image fusion–guided targeted biopsies were performed using Urostation (Koelis). The targeting error ± SD of the biopsy needle using this navigation system is 0.76 ± 0.52 mm [19]. Three essential steps were involved in the fusion process. First, before the biopsies, the MRI T2weighted sequence was loaded into the Urostation navigation system. In a semiautomatic process, the system registered the prostate volume in 3D, and the cancer-suspicious area was manually highlighted as a red spherical target on the axial T2-weighted images. The targets were made small enough to ascertain that a successful biopsy truly was inside the suspected tumor and they were directed toward the areas with the lowest ADC value. The diameters of the tar-

inflammation, and atrophy were considered as negative results. All patients with negative biopsies were reexamined by a single pathologist not primarily involved in the initial examination.

Parameter and Study

Positive (n = 33)

Negative (n = 7)

p

Ultrasound

20 ± 7

20 ± 7

0.93a

MRI

31 ± 12

28 ± 11

0.62a

0.7 ± 0.6

0.4 ± 0.3

0.03b

Target

0.8 ± 0.2

0.9 ± 0.2

0.08a

Peripheral zone

1.3 ± 0.1

1.2 ± 0.3

0.56a

Transitional zone

1.3 ± 0.2

1.2 ± 0.2

0.37a

Prostate volume (mL)

Target volume (mL) MRI Apparent diffusion coefficient (x 10 –5 mm2 /s)

Note—Except as otherwise indicated, data are mean ± SD for prostate volume and apparent diffusion coefficient and median ± interquartile range for target volume. aBy independent-samples Student t test. bBy Mann-Whitney U test.

AJR:202, March 2014 W243

dom biopsies had positive findings on MRI. In each patient, we registered the total number of biopsies, including both targeted and random biopsies; and the overall number of positive biopsy cores, including all positive biopsies.

ADC Values (×10−5mm2/s)


Rud et al.

Statistical Analysis The MRI results were compared with the biopsies, using the biopsy results as the criterion standard, and the measure of agreement was calculated by kappa value. The patients were stratified according to their biopsy results (negative or positive), and differences between the two groups were tested with an independent-samples Student t test or the Mann-Whitney U test, depending on the distribution of the continuous data in each group (target volumes, prostate volumes, and ADC values). The differences in the total number of biopsies and the number of positive biopsies between the biopsy groups were assessed using the MannWhitney U test. The differences in ADC values obtained from MRI targets and control values from peripheral and transitional zones were tested using a paired-samples t test. The significance level was set to p = 0.05, and the two-sided p values were given. The analyses were performed using SPSS (version 20, IBM), and Prism (version 6.0a, GraphPad) was used to create Figure 3.

Results MRI findings were positive in 40 of 42 (95%) patients. The overall positive biopsy rate was 79% (33 of 42 patients; κ = 0.31). Targeted biopsies were obtained from all MRI targets (n = 40). Of the 40 targets, 33 (83%) were classified as positive. Random biopsies were performed in 30 of the 42 (71%) patients. The two patients with negative findings on MRI also had negative random biopsies. Of the 28 of 30 (93%) patients with positive findings on MRI, six of these 28 (21%) patients had positive random biopsies. All of these were found in patients with positive targeted biopsies. Both a higher number of overall biopsy cores (13.0 vs 3.5; p < 0.001) and a higher number of positive biopsy cores (3.0 vs 1.5; p = 0.04) were found when random biopsies were performed in addition to targeted biopsies. A slightly higher number of biopsy cores were sampled in patients with positive biopsies compared with those with negative biopsies (11.4 vs 9.9; p = 0.70). The results according to the biopsy groups are summarized in Table 2. The mean ADC value for the MRI targets was significantly lower than the mean con-

W244

200

MRI Targets n = 40

Control Values n = 42

150 100 50 0

e iv

y

ps

o Bi

ve

iti

at

g Ne

y

ps

o Bi

s Po

r

he

ip

r Pe

e

on

Z al

ne

Zo

tio

si

an Tr

l na

Fig. 2—Patient with radiorecurrent prostate cancer 10 years after EBRT. Coronal fusion image shows cancer-suspicious target (yellow sphere). Two targeted biopsies from left lobe were positive (red cylinders), whereas six random biopsies from right lobe were negative (green cylinders). Superimposed grid allows random biopsies to be distributed properly, to sample all regions of gland.

Fig. 3—Bar chart shows mean ± SD apparent diffusion coeffecient (ADC) values in 40 MRI targets according to biopsy results. Mean ADC value for MRI targets was significantly lower than mean control values obtained from presumed noncancerous regions in both peripheral and transition zones (p < 0.001). No significant differences were found between biopsy-positive and -negative MRI targets (p = 0.08).

trol ADC values obtained from presumed noncancerous regions (p < 0.001). The mean ADC value for the biopsy-positive targets was also lower than the biopsy-negative targets, although this difference was not significant (p = 0.08). The mean ADC values in 40 MRI targets according to biopsy results and control values in assumed noncancerous areas are illustrated in Figure 3. The median volumes of the biopsy-positive targets were significantly higher than those of the biopsy-negative targets (0.7 mL vs 0.4 mL; p = 0.03). The distributions of prostate volumes, target volumes, and ADC values are summarized according to the biopsy results in Table 3.

biopsies. The remaining three patients never had random biopsies performed and may be subject to insufficient biopsy sampling. Two of these three patients experienced time intervals of 11 and 13 months between MRI and biopsy due to small volumes (< 0.5 mL) and low PSA level (mean, 1.3 ng/dL). ESUR 2012 recommends a combination of T2-weighted, DCE-MRI, and DWI using three b values (0, 100, and 800–1000 s/mm2) for the detection of treated and untreated prostate cancer. MR spectroscopy (MRS) is recommended only at experienced centers [9]. Rouvière et al. [6] showed that DCE-MRI was significantly more sensitive than T2weighted MRI (70–74% vs 26–44%) for detecting local recurrence after EBRT. Studies using only T2-weighted and DWI report sensitivities ranging from 62% to 93% [7, 8]. To our knowledge, only one study (n = 16) [10] has compared T2-weighted, DWI, and DCEMRI after brachytherapy, and this study reported 77% sensitivity when combining all three sequences, in contrast to 68% when using only T2-weighted MRI and DWI (p > 0.05). Our study did not include DCE-MRI, and this may have influenced the estimation of total tumor burden. ADC measurements are highly susceptible to the choice of sequence, b values, and size of ROI, which makes it difficult to compare the ADC values in the present study with those of other studies [22]. Our protocol used an echo-planar imaging sequence and

Discussion This study represents, to our knowledge, the largest evaluation of DWI and targeted biopsies for detecting local radiorecurrent prostate cancer after EBRT. Our results show that a few targeted, rather than multiple random, biopsies are sufficient for obtaining a positive diagnosis. This is in accordance with recently published studies [15, 21]. However, random biopsies detected a higher number of positive biopsy cores and revealed cancer foci in the contralateral lobe in 21% of patients. For this reason, we suggest performing random biopsies from the contralateral lobe if focal treatment is applied. Seven patients had negative targeted biopsies, in whom four also had negative random

AJR:202, March 2014


DWI and Targeted Biopsies for Radiorecurrent Prostate Cancer two b values (50 and 1000 s/mm2) to create an ADC map. The value of an isolated b2000 sequence is also believed to increase the sensitivity of tumor detection in untreated patients [23]. According to a recent study, using b50 instead of b0 provides more contrast in the ADC map, thereby improving the sensitivity for tumor detection [24]. The guidelines provide no recommendations regarding how to measure ADC, and caution is advised when using absolute values. To obtain ADC values from regions with the lowest diffusion, we used a small circular ROI (0.2 cm2). This is in accordance with Mazaheri et al. [25], who used voxel-by-voxel analyses, whereas other studies have used larger ROIs of different sizes and shapes [8, 24]. Although the ADC map of each MRI target was visually assessed prospectively by means of the ADC color overlay images, many values were measured retrospectively. This could have influenced the results. In six of 30 (20%) patients with positive random biopsies, the ADC values of the corresponding prostate region were above our cutoff values, reflecting the known overlap in ADC between cancerous and noncancerous tissues [26]. Consequently, the ADC map cannot rule out cancer but still is the most important tool for directing biopsies toward the most suspicious regions. Performing MRS in addition to DWI improved the sensitivity for detecting radiorecurrent cancer, according to a recent study by Westphalen et al. (n = 26) [27]. However, MRS prolongs the scanning time and is technically more challenging. A navigation system that guides and documents the position of the targeted biopsy relative to the MRI target plays an important role when performing targeted biopsies. However, all navigation systems based on image fusion hold a range of targeting error that may cause false-negative biopsies. This error can be eliminated by using directly MRI-guided biopsies, but this method is usually limited by availability and cost-benefit issues. By use of MRI and transrectal ultrasound soft image fusion, the biopsy procedures can be performed as usual in the outpatient clinic— and with a higher detection rate of high-grade cancer compared with a biopsy method without using a navigation system [15]. A potential targeting error of our navigation system might have caused false-negative biopsies in small tumor volumes on MRI, because the median tumor volume in positive biopsy targets was 0.7 mL compared with 0.4

mL in the negative targets (p = 0.03). These patients might be candidates for rebiopsy. The navigation system permits immediate feedback to the operator regarding needle placement. A successful targeted biopsy can be difficult to define when the biopsy edges the target. In such a case, or if a biopsy clearly misses the MRI target, as seen in Figure 1, more biopsies can be performed. Atrophy, fibrosis, and inflammation cause restricted diffusion, and might therefore appear suspicious for tumor at MRI [28, 29]. These histologic alterations were verified in all patients with negative target biopsies and might explain the false-positive MRI results. It is also known to be difficult to differentiate these findings from cancer when evaluating biopsies. Limitations Limitations of the present study include the selected patient cohort, because 17 of 40 (42.5%) patients already had a positive biopsy before our study. This may artificially improve the sensitivity of MRI owing to a known high prevalence of local recurrence. Furthermore, because no patients underwent salvage prostatectomy, the true prevalence of local disease is unknown, especially in patients with negative biopsies. Only one MRI reader was involved in this study, causing a limitation in the evaluation of interobserver variability and methodologic robustness. Conclusion This article presents a detection protocol consisting of T2-weighted MRI and DWI. It describes a sensitive method for detecting radiorecurrent prostate cancer, although the complete tumor burden is underestimated. A few targeted biopsies are sufficient to confirm a positive diagnosis, whereas random biopsies are required for more accurate assessment of the total tumor burden. References 1. Roach M, Hanks G, Thames H, et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG-ASTRO Phoenix Consensus Conference. Int J Radiat Oncol Biol Phys 2006; 65:965–974 2. Heidenreich A, Aus G, Bolla M, et al.; European Association of Urology. EAU guidelines on prostate cancer. Eur Urol 2008; 53:68–80 3. Mottet N, Bellmunt J, Bolla M, et al. EAU guidelines on prostate cancer. Part II. Treatment of advanced, relapsing, and castration-resistant pros-

tate cancer. Eur Urol 2011; 59:572–583 4. Boukaram C, Hannoun-Levi JM. Management of prostate cancer recurrence after definitive radiation therapy. Cancer Treat Rev 2010; 36:91–100 5. Rouvière O, Vitry T, Lyonnet D. Imaging of prostate cancer local recurrences: why and how? Eur Radiol 2010; 20:1254–1266 6. Rouvière O, Valette O, Grivolat S, et al. Recurrent prostate cancer after external beam radiotherapy: value of contrast-enhanced dynamic MRI in localizing intraprostatic tumor—correlation with biopsy findings. Urology 2004; 63:922–927 7. Kim CK, Park BK, Lee HM. Prediction of locally recurrent prostate cancer after radiation therapy: incremental value of 3T diffusion-weighted MRI. J Magn Reson Imaging 2009; 29:391–397 8. Morgan VA, Riches SF, Giles S, Dearnaley D, deSouza NM. Diffusion-weighted MRI for locally recurrent prostate cancer after external beam radiotherapy. AJR 2012; 198:596–602 9. Barentsz JO, Richenberg J, Clements R, et al.; European Society of Urogenital Radiology. ESUR prostate MR guidelines 2012. Eur Radiol 2012; 22:746–757 10. Tamada T, Sone T, Jo Y, et al. Locally recurrent prostate cancer after high-dose-rate brachytherapy: the value of diffusion-weighted imaging, dynamic contrast-enhanced MRI, and T2-weighted imaging in localizing tumors. AJR 2011; 197:408–414 11. Verma S, Rajesh A, Morales H, et al. Assessment of aggressiveness of prostate cancer: correlation of apparent diffusion coefficient with histologic grade after radical prostatectomy. AJR 2011; 196:374–381 12. Bittencourt LK, Barentsz JO, Miranda LCD, Gasparetto EL. Prostate MRI: diffusion-weighted imaging at 1.5T correlates better with prostatectomy Gleason grades than TRUS-guided biopsies in peripheral zone tumours. Eur Radiol 2012; 22:468–475 13. Hambrock T, Somford DM, Huisman HJ, et al. Relationship between apparent diffusion coefficients at 3.0-T MR imaging and Gleason grade in peripheral zone prostate cancer. Radiology 2011; 259:453–461 14. Kim CK, Park BK, Kim B. Diffusion-weighted MRI at 3 T for the evaluation of prostate cancer. AJR 2010; 194:1461–1469 15. Delongchamps NB, Peyromaure M, Schull A, et al. Prebiopsy magnetic resonance imaging and prostate cancer detection: comparison of random and targeted biopsies. J Urol 2013; 189:493–499 16. Rud E, Baco E, Eggesbø HB. MRI and ultrasound-guided prostate biopsy using soft image fusion. Anticancer Res 2012; 32:3383–3389 17. Ukimura O, Faber K, Gill IS. Intraprostatic targeting. Curr Opin Urol 2012; 22:97–103 18. Pinto PA, Chung PH, Rastinehad AR, et al. Mag-

AJR:202, March 2014 W245


Rud et al. netic resonance imaging/ultrasound fusion guided prostate biopsy improves cancer detection following transrectal ultrasound biopsy and correlates with multiparametric magnetic resonance imaging. J Urol 2011; 186:1281–1285 19. Baumann M, Mozer P, Daanen V, Troccaz J. Prostate biopsy tracking with deformation estimation. Med Image Anal 2012; 16:562–576 20. Osunkoya AO. Update on prostate pathology. Pathology 2012; 44:391–406 21. Haffner J, Lemaitre L, Puech P, et al. Role of magnetic resonance imaging before initial biopsy: comparison of magnetic resonance imaging-targeted and systematic biopsy for significant prostate cancer detection. BJU Int 2011; 108:E171–E178 22. Babourina-Brooks B, Cowin GJ, Wang D. Diffusion-weighted imaging in the prostate: an appar-

W246

ent diffusion coefficient comparison of half-Fourier acquisition single-shot turbo spin-echo and echo planar imaging. Magn Reson Imaging 2012; 30:189–194 23. Katahira K, Takahara T, Kwee TC, et al. Ultrahigh-b-value diffusion-weighted MR imaging for the detection of prostate cancer: evaluation in 201 cases with histopathological correlation. Eur Radiol 2011; 21:188–196 24. Thörmer G, Otto J, Reiss-Zimmermann M, et al. Diagnostic value of ADC in patients with prostate cancer: influence of the choice of b values. Eur Radiol 2012; 22:1820–1828 25. Mazaheri Y, Hricak H, Fine SW, et al. Prostate tumor volume measurement with combined T2weighted imaging and diffusion-weighted MR: correlation with pathologic tumor volume. Radi-

ology 2009; 252:449–457 26. Pickles MD, Gibbs P, Sreenivas M, Turnbull LW. Diffusion-weighted imaging of normal and malignant prostate tissue at 3.0T. J Magn Reson Imaging 2006; 23:130–134 27. Westphalen AC, Reed GD, Vinh PP, Sotto C, Vigneron DB, Kurhanewicz J. Multiparametric 3-tesla endorectal MR imaging after external beam radiation therapy of prostate cancer. J Magn Reson Imaging 2012; 36:430–437 28. Prando A, Billis A. Focal prostatic atrophy: mimicry of prostatic cancer on TRUS and 3D-MRSI studies. Abdom Imaging 2009; 34:271–275 29. Shukla-Dave A, Hricak H, Eberhardt SC, et al. Chronic prostatitis: MR imaging and 1H MR spectroscopic imaging findings—initial observations. Radiology 2004; 231:717–724

AJR:202, March 2014