Genitourinar y Imaging • Original Research De Cobelli et al. Use of Apparent Diffusion Coefficient to Predict Prostate Cancer Grade
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
Genitourinary Imaging Original Research
Francesco De Cobelli1 Silvia Ravelli1 Antonio Esposito1 Francesco Giganti1 Andrea Gallina2 Francesco Montorsi2 Alessandro Del Maschio1 De Cobelli F, Ravelli S, Esposito A, et al.
Apparent Diffusion Coefficient Value and Ratio as Noninvasive Potential Biomarkers to Predict Prostate Cancer Grading: Comparison With Prostate Biopsy and Radical Prostatectomy Specimen OBJECTIVE. The purpose of this study is to test the association between diffusionweighted MRI and prostate cancer Gleason score at both biopsy and final pathologic analysis after radical prostatectomy. SUBJECTS AND METHODS. Patients with prostate cancer (n = 72) underwent diffusion-weighted MRI (b values, 0, 800, and 1600 s/mm2) with an endorectal coil. Apparent diffusion coefficient (ADC) and ADC ratio were obtained in normal and pathologic tissue and were correlated with transrectal ultrasound–guided biopsy (n = 72) and histopathologic (n = 39) Gleason scores using the ANOVA test. ADC accuracy was estimated using ROC curves. RESULTS. Lesions suspicious for prostate cancer were detected in 65 patients. The mean ADC was 1.47 and 0.87 × 10 −3 mm2 /s for normal and pathologic tissue, respectively (p < 0.001). When we divided the population into four groups (normal tissue and biopsy Gleason scores of 6, 7, and 8–10), then the mean ADC value was 1.47, 0.96, 0.80, and 0.78 × 10 −3 mm2 /s, respectively (p < 0.001). The ADC ratio decreased along with an increase in biopsy Gleason score (66.9%, 56.7%, and 51.5% for Gleason scores of 6, 7 and 8–10, respectively) (ANOVA, p = 0.003) and pathologic Gleason score (ANOVA, p < 0.001). ROC curves had an AUC of 0.94 and 0.86 for ADC and ADC ratio, respectively (p = 0.012 and 0.042, respectively). CONCLUSION. Decreasing ADC values may represent a strong risk factor of harboring a poorly differentiated prostate cancer, independently of biopsy characteristics.
P
Keywords: apparent diffusion coefficient, apparent diffusion coefficient ratio, diffusion-weighted imaging, Gleason score, prostate cancer DOI:10.2214/AJR.14.13146 Received May 6, 2014; accepted after revision July 7, 2014. 1 Department of Radiology and Center for Experimental Imaging, San Raffaele Scientific Institute, Vita-Salute University, Olgettina 60, Milan 20132, Italy. Address correspondence to F. De Cobelli (
[email protected]). 2 Department of Urology, San Raffaele Scientific Institute, Vita-Salute University, Milan, Italy.
AJR 2015; 204:550–557 0361–803X/15/2043–550 © American Roentgen Ray Society
550
rostate cancer is the most frequently diagnosed solid malignant tumor among men in the United States and Western Europe [1]. The morbidity and mortality directly attributable to this common malignancy are considerable. However, in a nonnegligible proportion of patients, the disease may be considered relatively indolent. The advent of prostate-specific antigen testing and modified prostatic biopsy schemes have led to overdiagnosis, which is the diagnosis of cancers that will not be clinically diagnosed during a patient’s life. Indeed, there are concerns about a large number of men who are currently being overtreated for their prostate malignancy, resulting in treatment-related morbidity [2, 3]. Recently, active surveillance has become a reasonable treatment option in patients with low-risk prostate cancer who tend to have excellent oncologic outcomes [4, 5]. However, even using the most stringent active surveillance inclusion criteria, some patients are misclassified as active surveillance candidates because of unfavorable findings on final path-
ologic analysis [6, 7]. The introduction of active surveillance requires an expansion of diagnostic imaging beyond staging, so that it becomes necessary to provide more accurate information regarding tumor localization and aggressiveness. The Gleason grading system is the most commonly accepted and widely used system for evaluating the biologic activity and aggressiveness of prostate cancer [8, 9]. However, biopsy determination of the Gleason score often does not provide an accurate reflection of whole-organ pathologic characteristics because of sampling errors associated with heterogeneity of multiple cancer foci [10–12]. Moreover, approximately 30% of men with localized prostate cancer diagnosed by transrectal sonography–guided biopsy have clinically insignificant prostate neoplasms when verified by a more detailed histologic analysis [13–16]. Additional consequences of the current diagnostic pathway include poor risk stratification, repeated negative biopsies, and increased incidence of multiresistant sepsis resulting from a transrectal approach performed through the contaminated rectum [17].
AJR:204, March 2015
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
Use of Apparent Diffusion Coefficient to Predict Prostate Cancer Grade
A
B
C There is thus a clear need for a noninvasive and panoramic method for accurate prediction of the true Gleason score from radical prostatectomy specimens before treatment. Recent reports [18–23] suggest that multiparametric MRI, which combines a num-
A
Fig. 1—68-year-old man with prostate-specific antigen level of 3.9 ng/ mL (biopsy Gleason score, 3 + 4; pathologic Gleason score, 4 + 3; pathologic tumor stage, 2c). A–C, Axial T2-weighted image (A), diffusionweighted image (b values, 0, 800, and 1600 mm/s2) (B), and apparent diffusion coefficient map (C) show lesion (arrow) in left midanterior peripheral zone of prostate with extracapsular extension. D, Presence of lesion (arrow) was confirmed at histopathologic analysis.
D
ber of different MRI sequences beyond the standard T1- and T2-weighted images, provides promising results by depicting and helping to rule out clinically significant cancer when whole-mount radical prostatectomy is assumed as the reference standard. In particular, diffusion-weighted MRI (DWI) has attracted attention as a possible tool to correctly classify low- and high-risk prostate cancers, even if there are still controversial data on the role of DWI in discriminating the aggressiveness of prostate cancer [20, 24–29].
We hypothesized that DWI patterns—in particular, a restricted apparent diffusion coefficient (ADC) value—may represent a clinically useful definition of tumor differentiation, providing reliable information on the Gleason score of the suspicious prostate cancer lesions. The aim of our study was to test the association between DWI and prostate cancer Gleason score at both prostate biopsy and final pathologic analysis after radical prostatectomy and to investigate the usefulness of the ADC values in predicting the aggressiveness of prostate cancer.
Fig. 2—70-year-old man with prostate-specific antigen level of 8.3 ng/mL. A and B, Axial T2-weighted image (A) and diffusion-weighted image (b values, 0, 800, and 1600 mm/s2) (B) show lesion (arrow) in right posterolateral zone of prostate. C, ROI drawn along lesion border (arrow) on apparent diffusion coefficient (ADC) map shows ADC value of 0.79 × 10 −3 mm2 /s.
B
C
AJR:204, March 2015 551
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
De Cobelli et al.
A
B
C
Fig. 3—52-year-old man with prostate cancer with prostate-specific antigen level of 4.5 ng/mL (biopsy Gleason score, 3 + 3; pathologic Gleason score, 4 + 3; pathologic tumor stage, 2a without extracapsular extension or seminal vesicle invasion). A, Axial T2-weighted sequence shows lesion (arrow) in intermediate right posterolateral zone of prostate. B, Raw dynamic contrast-enhanced image and forward volume transfer constant image derived from dynamic contrast-enhanced MRI (inset) confirm presence of lesion. C, ROIs (circles) drawn in apparent diffusion coefficient (ADC) map show lower ADC value (0.56 × 10 −3 mm2 /s) for pathologic tissue than for benign parenchyma (1.43 × 10 −3 mm2 /s).
Subjects and Methods Patient Population A total of 77 consecutive patients with newly diagnosed histologically proven (via transrectal ultrasound–guided prostate biopsies [30, 31]) adenocarcinoma of the prostate were referred for preoperative multiparametric MRI of the prostate. Patients treated with antiandrogen therapy (n = 2) and with poor image quality (n = 3) were excluded, for a total of 72 patients enrolled in the study. Complete clinical and pathologic data were recorded. To minimize postbiopsy artifacts, all MRI examinations were performed 4–12 weeks after prostate biopsy. Subsequently, 39 patients (54.16%) underwent radical prostatectomy. The interval between MRI and radical prostatectomy in this subgroup of patients was 6–12 weeks. This was a single-center prospective study that followed the Standards for Reporting of Di-
agnostic Accuracy guidelines. No industry support was given. The protocol was part of a larger study approved by our institutional review board. At our hospital, all patients who underwent multiparametric MRI, prostate biopsy, and prostatectomy were asked to have their multiparametric MRI, biopsy, and surgical histologic data entered into a prospective database. All patients included in the database gave written consent for the use of their data for research purposes and signed the institutional review board–approved consent form. Prostate specimens were processed according to the Stanford protocol [32]. Pathologic tumor stage, Gleason grade, and score were determined according to the 2010 TNM classification [33]. Furthermore, relative (percentage) measures of tumor volume were recorded. All specimens were analyzed by dedicated uropathologists.
MRI Technique All patients were examined using a 1.5-T MRI scanner (Achieva, Philips Healthcare) combined with a commercially available balloon-covered expandable endorectal coil filled with air for signal reception. A semianesthetic gel (lidocaine) was used. Gastrointestinal peristalsis was suppressed by intramuscular administration of 20 mg of scopolamine-butylbromide (Buscopan, Boehringer Ingelheim), in the absence of contraindications. The imaging protocol consisted of multiplanar turbo spin-echo T2-weighted images, turbo spinecho T1-weighted images, 3D fast field-echo dynamic contrast-enhanced (DCE) MRI, and echo-planar DWI. Imaging protocol details are listed in Table 1. First, turbo spin-echo T1-weighted images were performed in the axial plane, to exclude the presence of postbiopsy hemorrhagic foci in the prostate gland that potentially might mimic or conceal cancer as a
TABLE 1: MRI Scan Protocols Parameters TR (ms) TE (ms) FOV (mm) Thickness (mm) In-plane resolution (mm)
TSE T2-Weighted Axial
TSE T2-Weighted Sagittal
TSE T2-Weighted Coronal
TSE T1-Weighted Axial
DWI
DCE
Shortest
Shortest
Shortest
Shortest
2500
Shortest
125
125
120
13
80
Shortest
180 × 119
180 × 109
180 × 119
180 × 149
180 × 165
180 × 125
3
3
3
3
3
3
0.50
0.50
0.50
0.50
2
1
TSE factor
17
17
17
4
—
60
Flip angle (°)
90
90
90
90
90
10
No. of acquisitions
6
6
6
2
6
1
Acquisition time
5 min 41 s
5 min 19 s
4 min 42 s
3 min 16 s
5 min 30 s
5 min 0 s
b values (s/mm2)
—
—
—
—
0, 800, and 1600
—
No. of dynamic scans
—
—
—
—
—
50
Dynamic scan time (s)
—
—
—
—
—
6
Note—Dash (—) indicates data not available. TSE = turbo spin-echo, DWI = diffusion-weighted imaging, DCE = dynamic contrast enhanced.
552
AJR:204, March 2015
Use of Apparent Diffusion Coefficient to Predict Prostate Cancer Grade
p < 0.001
p < 0.001
2.00
1.50
1.47 ADC
ADC
1.50
1.00
1.47
1.00
0.96 0.80
0.87
*
0.50
0
0.50
BPH
0
Tumor
Fig. 4—Box-plot of normal (benign prostatic hyperplasia [BPH]) and pathologic prostatic tissue as function of apparent diffusion coefficient (ADC). Mean ADC value in supposedly neoplastic regions at multiparametric MRI examination was significantly lower than mean ADC value in nonneoplastic area. Center line denotes median, top of box denotes 75th percentile, bottom of box denotes 25th percentile, whiskers denote 10th and 90th percentiles, asterisk denotes extreme values (more than 3 interquartile ranges), and circles denote outliers (between 1.5 and 3 interquartile ranges).
result of interfering artifacts. Turbo spin-echo T2weighted images were acquired in the axial plane, covering the entire prostate gland and seminal vesicles, and in the sagittal and coronal planes. DWI was acquired with three different b values (0, 800, and 1600 s/mm2), and an ADC map was automatically elaborated on a pixel-by-pixel basis with the use of all b values. ADC values were calculat-
100
*
≤6 7 Biopsy Gleason Score
BPH
ed by monoexponential regression with the following formula: S = S0(-bADC), where S and S0 represent the pixel values with and without the application of diffusion-weighting gradients, respectively, and b is the diffusion-weighting factor of the applied gradient. For DCE-MRI, an IV bolus of 0.1 mmol/kg gadobutrol (Gadovist, Bayer Schering Pharma) followed by a 200-mL flush of normal saline solution (NaCl,
p = 0.003
Image Analysis Consensus reads of T2-weighted, DCE, and DWI datasets were performed by two radiologists (with 10 and 5 years of experience, respectively, in
p < 0.001
1.50
1.47
*
51.5%
ADC
66.9% 56.7%
1.17
1.00
40
0.83 0.50
20
0
8–10
0.9%) was performed with an automatic injector (Spectris MR, Medrad Europe) at a rate of 2 mL/s. Total imaging time was approximately 30 minutes.
2.00
60
0.78
Fig. 5—Box-plot of normal (benign prostatic hyperplasia [BPH]) and pathologic biopsy-proved prostatic tissue as function of apparent diffusion coefficient (ADC). Mean ADC value of tumor decreased significantly in parallel to increase in biopsy Gleason score, with progressive reduction of ADC. Center line denotes median, top of box denotes 75th percentile, bottom of box denotes 25th percentile, whiskers denote 10th and 90th percentiles, asterisk denotes extreme values (more than 3 interquartile ranges), and circles denote outliers (between 1.5 and 3 interquartile ranges).
80 ADC Ratio (%)
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
2.00
≤6
7 Biopsy Gleason Score
8–10
Fig. 6—Box-plot of pathologic biopsy-proved prostatic tissue as function of apparent diffusion coefficient (ADC) ratio. ADC ratio decreased significantly along with increase of biopsy Gleason score. Center line denotes median, top of box denotes 75th percentile, bottom of box denotes 25th percentile, whiskers denote 10th and 90th percentiles, and circles denote outliers (between 1.5 and 3 interquartile ranges).
0
0.74
*
BPH
≤6 7 8–10 Radical Prostatectomy Gleason Score
Fig. 7—Box-plot of pathologic Gleason score after surgery as function of apparent diffusion coefficient (ADC). When considering only 36 (92.31%) patients with visible multiparametric MRI lesions, ADC value decreased significantly in parallel to increase in biopsy Gleason score. Center line denotes median, top of box denotes 75th percentile, bottom of box denotes 25th percentile, whiskers denote 10th and 90th percentiles, asterisks denote extreme values (more than 3 interquartile ranges), and circles denote outliers (between 1.5 and 3 interquartile ranges).
AJR:204, March 2015 553
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
De Cobelli et al. reading prostate multiparametric MRI) who were blinded to clinical characteristics (prostate-specific antigen level, clinical stage, number of positive specimens, percentage of tumoral tissue per specimen, and localization of the lesion) and to the biopsy histopathologic results. Multiparametric MRI was assessed according to the European Society of Urogenital Radiology guidelines according to the Prostate Imaging and Reporting and Data System score [34]. The criterion for prostate cancer detection on the T2-weighted images was the evidence of a nodular hypointense lesion in the peripheral zone or a very pronounced hypointense sickle-shaped lesion in the transitional zone of the prostate gland. DCE images were analyzed by means of visual interpretation of the raw dynamic enhanced T1-weighted images, evaluating the curve morphologic features and type (I to III). Prostate cancer was defined, in both the peripheral zone and the transitional zone, as a nodular focus that displayed early strong enhancement with or without rapid washout (Fig. 1). On DWI, nodular foci in the peripheral zone and the transitional zone were considered positive for prostate cancer when showing high signal intensity on the highest b value images in combination with low signal intensity on the ADC map (Fig. 2). For the quantitative analysis of DWI parameters, ROIs were drawn on the ADC map by referring to the T2-weighted images in visible prostate cancer. The ROIs were placed on the ADC map by using the freehand drawing tool to encompass as much of the inner aspect of the lesion as possible without contacting the edges. The ADC ratio (expressed as a percentage) was obtained by dividing the ADC of a cancer ROI by the ADC of an area of noncancerous tissue similar in size to the cancer ROI, in the same prostate zone (i.e., peripheral or transition) in mirror position to the tumor (Fig. 3).
Statistical Analysis Clinical and demographic data were summarized using descriptive statistics. The independent sample Student t test was used to assess the differences between tumor ADC and noncancerous ADC. The relationship between the tumor ADC and ADC ratio with the biopsy Gleason score was tested using ANOVA. The diagnostic accuracy of ADC to differentiate low-grade from intermediate- and highgrade tumors at final pathologic analysis was quantified using the AUC. Statistical significance was set at p < 0.05. All analyses were performed with the SPSS software package (version 17.0, IBM).
Results Table 2 shows clinical characteristics and biopsy Gleason score distribution of the population enrolled in the study.
554
Considering all 72 patients, multiparametric MRI was able to diagnose tumors in 65 patients (90.27%). In all seven (9.72%) patients with a negative multiparametric MRI examination, the histologic analysis showed a biopsy Gleason score of 3 + 3, indicating a well-differentiated tumor, with two or fewer pathologic cores, each one containing 10% or less malignant cells. When considering only visible lesions, the mean ADC value in the supposedly neoplastic regions at the multiparametric MRI examination was significantly lower compared with the mean ADC value in the nonneoplastic area (0.87 vs 1.47 mm2 /s, respectively; p < 0.001; Fig. 4). Furthermore, the mean ADC value of the tumor decreased significantly in parallel to an increase in the biopsy Gleason score, with a progressive reduction of the ADC from a value of 1.47 mm2 /s in the nonneoplastic tissue to 0.96 mm2 /s in tumors with a biopsy Gleason score of 6, to 0.80 mm2 /s and 0.78 mm2 /s in tumors with a biopsy Gleason score of 7 and 8–10, respectively (ANOVA p < 0.001; Fig. 5). In contrast, the ADC value of the nonneoplastic prostate gland was not statistically significantly different between patients with a biopsy Gleason score of 6 (1.45 mm2 /s), 7 (1.46 mm2 /s), and 8–10 (1.52 mm2 /s; ANOVA p = 0.3). Furthermore, we calculated the ratio between the ADC values of the tumor and of nonneoplastic tissue, expressed as a percentage (ADC ratio). Although no significant differences were found in the absolute ADC values in the nonneoplastic tissue, the ADC ratio decreased significantly along with an increase in the biopsy Gleason score from 66.9% for tumors with a Gleason score of 6, to 56.7% for tumors with a Gleason score of 7, to 51.5% for tumors with a Gleason score of 8–10 (ANOVA p = 0.003) (Fig. 6). In 39 of 72 patients (54.17%), radical prostatectomy was performed, and it was possible to obtain the final histologic response (pathologic Gleason score). The pathologic characteristics of the prostate tumors of these patients are summarized in Table 3. In 36 of the 39 patients (92.31%) who underwent radical prostatectomy, the multiparametric MRI was able to clearly detect the tumor. In the remaining three (7.69%) patients, the multiparametric MRI was negative—namely, it was not able to detect the lesion in none of the performed sequences. However, in the three (7.69%) patients with a negative multiparametric MRI examination, the pathologic Gleason score was 3 + 3 with
TABLE 2: Patient Demographic, Clinical, and H istopathologic Data Considered in the Study Population (n = 72) Characteristic
Value
Age (y) Mean
64.6
Median
64.0
Range
41.0–83.0
Total prostate-specific antigen level (ng/mL) Mean
9.2
Median
6.2
Range
0.6–57.0
Biopsy Gleason score, no. (%) of patients 6
37 (51.4)
3+4
16 (22.2)
4+3
8 (11.1)
8–10
11 (15.3)
No. of biopsy cores collected Mean
15.6
Median
16.0
Range
8–21
TABLE 3: Histopathologic Data of All Patients (n = 39) Referred for Radical Prostatectomy Characteristic
Value
Pathologic stage, no. (%) of patients T2
24 (61.5)
T3a (extracapsular extension)
7 (17.9)
T3b (seminal vesicle invasion)
8 (20.5)
Pathologic Gleason score, no. (%) of patients ≤6
8 (20.5)
3+4
13 (33.3)
4+3
9 (23.1)
8–10
9 (23.1)
Tumor volume (cm3) Mean
5.2
Median
2.7
Range
0.4–36
a limited tumor volume (≤ 1% of the prostate gland volume). When considering only the 36 (92.31%) patients with visible multiparametric 1.5-T MRI lesions, the ADC value decreased significantly from 1.17 mm2/s in
AJR:204, March 2015
Discussion Our results confirmed that the ADC value of lesions suspicious for prostate cancer is significantly lower than the ADC values of the normal prostatic tissue [35–37]. This finding is strikingly important in the diagnostic setting of prostate cancer because the evaluation of DWI seems to be a valuable tool in the detection and characterization of prostate cancer, increasing the specificity of standard MRI evaluation. According to our results, lower ADC values are associated with a more-aggressive histologic pattern with reference to the biopsy Gleason score. As previously found by
Fig. 8—Box-plot of pathologic Gleason score after surgery as function of apparent diffusion coefficient (ADC) ratio. ADC ratio decreased significantly along with increase of pathologic Gleason score. Center line denotes median, top of box denotes 75th percentile, bottom of box denotes 25th percentile, whiskers denote 10th and 90th percentiles, asterisks denotes extreme values (more than 3 interquartile ranges).
100
p = 0.01
80 ADC Ratio (%)
tumors with a Gleason score of 3 + 3, to 0.83 mm2/s in tumors with a Gleason score of 7, and to 0.74 mm2/s in tumors with a Gleason score of 8–10 (ANOVA p < 0.001; Fig. 7). Similarly, also the ADC ratio decreased significantly along with an increase of the pathologic Gleason score (ANOVA p < 0.001) (Fig. 8). Finally, we derived the ROC curves both for the tumor ADC value and the ADC ratio, to assess their accuracy in the prediction of the tumor aggressiveness, distinguishing nonaggressive (pathologic Gleason score ≤ 6) from aggressive tumors (pathologic Gleason score > 6). The ability of the tumor ADC and ADC ratio to predict tumor aggressiveness at final histologic analysis, considering all the patients who underwent surgery, was very high (AUC for ADC and ADC ratio, 0.94 and 0.86; p = 0.012 and 0.042, respectively) (Fig. 9).
70.8%
60
58.6%
48.8%
20
0
≤6
Tamada et al. [25], there is an inverse correlation between the cancer ADC value and the biopsy Gleason score. The interesting aspect of our study is that, in addition to the ADC value, we evaluated the ratio between the ADC value of the prostate cancer and the ADC value of the normal prostatic tissue—that is, the ADC ratio. We found that the ADC ratio decreases significantly with the increase of the cancer aggressiveness, regardless of the ADC values. To our knowledge, there are no evident data regarding the ADC ratio in the literature and its value in the evaluation of the aggressiveness of the prostate cancer. In fact, it is easily understandable that this new parameter could have some clinical advantages
80
60
40
20
0
0
20
40 60 1 – Specificity (%)
80
100
* * *
40
100
Sensitivity (%)
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
Use of Apparent Diffusion Coefficient to Predict Prostate Cancer Grade
Fig. 9—ROC curves for pathologic apparent diffusion coefficient (ADC) (black line) and ADC ratio (gray line). Ability of tumor ADC and ADC ratio to predict tumor aggressiveness at final histology, considering all patients who underwent surgery, was very high (AUC for ADC and ADC ratio, 0.94 and 0.86, respectively; p = 0.012 and 0.042, respectively).
7 8–10 Final Pathologic Gleason Score
compared with the absolute ADC value, being less prone to absolute quantification bias. In fact, because the ADC ratio is not an absolute value but is expressed as a percentage (also including the ADC value of healthy tissue), it can be supposed that this parameter may provide a greater standardization and a more intuitive and functional cutoff. Multiparametric MRI was able to detect a prostate cancer in 36 of the 39 patients (92.31%) who later underwent radical prostatectomy. In the remaining three patients (7.69%), the multiparametric MRI was not able to detect the tumor; in particular, there were no areas of restricted diffusivity, even with DWI. However, all these multiparametric MRI–undetected cancers had a pathologic Gleason score of 3 + 3 and a small tumor volume (≤ 1% of the prostate gland volume). We also assessed the relationship between the ADC value and the histologic Gleason score in patients with positive multiparametric MRI findings who later underwent radical prostatectomy. Also in this subgroup of patients, the mean ADC value of the lesions with a Gleason score of 6 was significantly higher than the ADC value of the lesions with a Gleason score of 7 or 8–10, showing that our results are not influenced by possible errors in sampling and histologic results of biopsies. However, in our results, we observed that ADC values in biopsy Gleason score 3 + 3 were substantially lower than ADC values in prostatectomy Gleason score 3 + 3 (0.96 vs 1.17 mm2 /s). It may be postulated that this difference reflects the underestimation of the final Gleason score related to prostate biopsy limitation. If these findings are confirmed by larger studies, the role of multiparametric
AJR:204, March 2015 555
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
De Cobelli et al. MRI may significantly affect the assessment of newly diagnosed prostate cancer, providing a new tool for reducing the rate of Gleason score misclassification. Moreover, we assessed the ROC analysis of the ADC value and the ADC ratio in the 39 (54.17%) patients with positive multiparametric MRI who later underwent radical prostatectomy. These multiparametric MRI parameters can distinguish between patients with less- or more-aggressive lesions. Interestingly, we observed that not only the mean pathologic ADC value decreases steadily with the increase of the Gleason score but also that this parameter can also predict, with high accuracy, the presence of an aggressive neoplastic lesion (pathologic Gleason score ≥ 7) [38]. In this study, we observed that the AUC of conventional ADC values was higher than that of ADC ratios; this can be explained by assuming that, in patients with more-aggressive prostate cancer, the prostatic tissue might be already modified by chronic inflammation (thus, with a lower ADC value). Conversely, in patients with lower Gleason score prostate cancer, the prostatic tissue is not affected by inflammation or fibrosis (i.e., higher ADC in the normal tissue and, therefore, lower ADC ratio values). In summary, we can confirm the hypothesis that MRI, and in particular diffusion patterns of lesions suspicious for prostate cancer, can help clinicians to properly identify candidates for active surveillance, thus limiting the risk of misclassifying those with apparently favorable clinical characteristics who, however, will harbor an aggressive cancer at final pathologic analysis. By evaluating some relatively simple parameters, multiparametric MRI proved to be a very useful tool to discriminate between aggressive and indolent tumors and, as a consequence, to confidently choose the most appropriate treatment for each patient. In this context, the routine use of multiparametric MRI could obviate surgery in those patients with indolent malignancy, for whom the potential risks of surgery may outweigh the survival benefits. van As et al. [39] previously found that in 86 low-risk patients with prostate cancer in an active surveillance program, the ADC value of the lesion could be a useful marker of disease progression, defined as an increase in the Gleason score at the second biopsy, or an indication to proceed to surgery. In that study, the multiparametric MRI was performed during, not at the beginning of, the active
556
surveillance program and, therefore, a lower ADC value in patients with disease progression could also be explained as an incorrect Gleason scoring at the first biopsy. However, Morgan et al. [40] also found that the cancer ADC value decreased along with the progression of the disease in a population of 50 patients under active surveillance. Our study has some limitations. Because of the small number of patients enrolled, it has to be regarded as a preliminary study and it needs to be supported by a larger series of examinations. Owing to the small sample size, it was not possible to perform multivariate analysis to assess the predictive ability of the multiparametric MRI after having corrected all the confounding factors. However, the high statistical significance of our results makes them interesting even without a multivariate analysis. Another limitation of our study may be the fact that we performed a per-patient analysis, focusing our attention on the most aggressive lesion (index lesion) of prostate cancer. A possible explanation might be that multiparametric MRI is not able to detect all pathologic microfoci. In conclusion, multiparametric MRI seems to be a valuable tool in the detection and characterization of prostate cancer. A multiparametric MRI examination could be useful for all patients with prostate cancer, especially those with a biopsy Gleason score of 6. In fact, a multiparametric MRI evaluation could help the urologist in deciding on the most appropriate therapeutic strategy, especially for those patients under active surveillance. References 1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J Clin 2010; 60:277–300 2. Lilja H, Ulmert D, Vickers AJ. Prostate-specific antigen and prostate cancer: prediction, detection and monitoring. Nat Rev Cancer 2008; 8:268–278 3. Etzioni R, Penson DF, Legler JM, et al. Overdiagnosis due to prostate-specific antigen screening: lessons from U.S. prostate cancer incidence trends. J Natl Cancer Inst 2002; 94:981–990 4. Schröder FH, Hugosson J, Roobol MJ, et al.; ERSPC Investigators. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med 2009; 360:1320–1328 5. Bill-Axelson A, Holmberg L, Ruutu M, et al.; Scandinavian Prostate Cancer Group Study No. 4. Radical prostatectomy versus watchful waiting in early prostate cancer. N Engl J Med 2005; 352:1977–1984 6. Cooperberg MR, Lubeck DP, Meng MV, Mehta
SS, Carroll PR. The changing face of low-risk prostate cancer: trends in clinical presentation and primary management. J Clin Oncol 2004; 22:2141–2149 7. van den Bergh RC, Roemeling S, Roobol MJ, et al. Outcomes of men with screen-detected prostate cancer eligible for active surveillance who were managed expectantly. Eur Urol 2009; 55:1–8 8. Gleason DF, Mellinger GT; Veterans Administration Cooperative Urological Research Group. Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging, 1974. J Urol 2002; 167(2 pt 2):953–958; discussion, 959 9. Epstein JI, Allsbrook WC Jr, Amin MB, Egevad LL; ISUP Grading Committee. The 2005 International Society of Urological Pathology (ISUP) consensus conference on Gleason grading of prostatic carcinoma. Am J Surg Pathol 2005; 29:1228–1242 10. Kvåle R, Møller B, Wahlqvist R, et al. Concordance between Gleason scores of needle biopsies and radical prostatectomy specimens: a population-based study. BJU Int 2009; 103:1647–1654 11. Rajinikanth A, Manoharan M, Soloway CT, Civantos FJ, Soloway MS. Trends in Gleason score: concordance between biopsy and prostatectomy over 15 years. Urology 2008; 72:177–182 12. Stav K, Merald H, Leibovici D, Lindner A, Zisman A. Does prostate biopsy Gleason score accurately express the biologic features of prostate cancer? Urol Oncol 2007; 25:383–386 13. Bastian PJ, Carter BH, Bjartell A, et al. Insignificant prostate cancer and active surveillance: from definition to clinical implications. Eur Urol 2009; 55:1321–1330 14. Suardi N, Capitanio U, Chun FK, et al. Currently used criteria for active surveillance in men with low-risk prostate cancer: an analysis of pathologic features. Cancer 2008; 113:2068–2072 15. Lee MC, Dong F, Stephenson AJ, Jones JS, MagiGalluzzi C, Klein EA. The Epstein criteria predict for organ-confined but not insignificant disease and a high likelihood of cure at radical prostatectomy. Eur Urol 2010; 58:90–95 16. Epstein JI, Walsh PC, Carmichael M, Brendler CB. Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 1994; 271:368–374 17. Gallina A, Suardi N, Montorsi F, et al. Mortality at 120 days after prostatic biopsy: a populationbased study of 22,175 men. Int J Cancer 2008; 123:647–652 18. Akin O, Hricak H. Imaging of prostate cancer. Radiol Clin North Am 2007; 45:207–222 19. Kirkham AP, Emberton M, Allen C. How good is MRI at detecting and characterising cancer within the prostate? Eur Urol 2006; 50:1163–1174; discussion, 1175
AJR:204, March 2015
Downloaded from www.ajronline.org by Ospedale San Raffaele SRL on 08/04/16 from IP address 192.167.192.100. Copyright ARRS. For personal use only; all rights reserved
Use of Apparent Diffusion Coefficient to Predict Prostate Cancer Grade 20. Vargas HA, Akin O, Franiel T, et al. Diffusionweighted endorectal MR imaging at 3 T for prostate cancer: tumor detection and assessment of aggressiveness. Radiology 2011; 259:775–784 21. Wang L, Mullerad M, Chen HN, et al. Prostate cancer: incremental value of endorectal MR imaging findings for prediction of extracapsular extension. Radiology 2004; 232:133–139 22. 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; 42:842–847 23. Miao H, Fukatsu H, Ishigaki T. Prostate cancer detection with 3-T MRI: comparison of diffusionweighted and T2-weighted imaging. Eur J Radiol 2007; 61:297–302 24. Itou Y, Nakanishi K, Narumi Y, Nishizawa Y, Tsukuma H. Clinical utility of apparent diffusion coefficient (ADC) values in patients with prostate cancer: can ADC values contribute to assess the aggressiveness of prostate cancer? J Magn Reson Imaging 2011; 33:167–172 25. 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; 28:720–726 26. Turkbey B, Shah VP, Pang Y, et al. Is apparent diffusion coefficient associated with clinical risk scores for prostate cancers that are visible on 3-T
MR images? Radiology 2011; 258:488–495 27. 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 28. Yoshimitsu K, Kiyoshima K, Irie H, et al. Usefulness of apparent diffusion coefficient map in diagnosing prostate carcinoma: correlation with stepwise histopathology. J Magn Reson Imaging 2008; 27:132–139 29. Langer DL, van der Kwast TH, Evans AJ, et al. Prostate tissue composition and MR measurements: investigating the relationships between ADC, T2, Ktrans, ve, and corresponding histologic features. Radiology 2010; 255:485–494 30. Scattoni V, Raber M, Abdollah F, et al. Biopsy schemes with the fewest cores for detecting 95% of the prostate cancers detected by a 24-core biopsy. Eur Urol 2010; 57:1–8 31. Scattoni V, Zlotta AR, Nava L, Roscigno M, Montorsi F, Rigatti P. Prostatic transrectal ultrasound (TRUS) guided biopsy schemes and TRUS prostatic lesion-guided biopsies. Eur Urol Suppl 2002; 1:28–34 32. Srigley JR, Humphrey PA, Amin MB, et al.; Members of the Cancer Committee, College of American Pathologists. Protocol for the examination of specimens from patients with carcinoma of the prostate gland. Arch Pathol Lab Med 2009; 133:1568–1576 33. Edge SB, Byrd DR, Compton CC, et al. American Joint Committee on Cancer Staging Manual, 7th ed. New York, NY: Springer, 2010
34. Barentsz JO, Richenberg J, Clements R, et al.; European Society of Urogenital Radiology. ESUR prostate MR guidelines 2012. Eur Radiol 2012; 22:746–757 35. Lim HK, Kim JK, Kim KA, Cho KS. Prostate cancer: apparent diffusion coefficient map with T2-weighted images for detection—a multireader study. Radiology 2009; 250:145–151 36. 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 37. Haider MA, van der Kwast TH, Tanguay J, et al. Combined T2-weighted and diffusion-weighted MRI for localization of prostate cancer. AJR 2007; 189:323–328 38. Kitajima K, Takahashi S, Ueno Y, et al. Do apparent diffusion coefficient (ADC) values obtained using high b-values with a 3-T MRI correlate better than a transrectal ultrasound (TRUS)-guided biopsy with true Gleason scores obtained from radical prostatectomy specimens for patients with prostate cancer? Eur J Radiol 2013; 82:1219–1226 39. van As NJ, de Souza NM, Riches SF, et al. A study of diffusion-weighted magnetic resonance imaging in men with untreated localised prostate cancer on active surveillance. Eur Urol 2009; 56:981–987 40. Morgan VA, Riches SF, Thomas K, et al. Diffusionweighted magnetic resonance imaging for monitoring prostate cancer progression in patients managed by active surveillance. Br J Radiol 2011; 84:31–37
AJR:204, March 2015 557
