Prediction of Micrometastasis (< 1 cm) to Pelvic Lymph Nodes in Prostate Cancer: Role of Preoperative MRI

Transcription

1 Genitourinary Imaging Original Research Park et al. MRI for Lymph Node Micrometastasis in Prostate Cancer Genitourinary Imaging Original Research Sung Yoon Park 1 Young Taik Oh 1 Dae Chul Jung 1 Nam Hoon Cho 2 Young Deuk Choi 3 Koon Ho Rha 3 Park SY, Oh YT, Jung DC, Cho NH, Choi YD, Rha KH Keywords: lymph node metastasis, micrometastasis, MRI, prostate cancer, tumor staging DOI: /AJR Received November 13, 2014; accepted after revision January 31, Supported through the Basic Science Research Program of the National Research Foundation of Korea (NRF) funded by Ministry of Education grant NRF-2013R1A1A and by Yonsei University College of Medicine faculty research grant Department of Radiology and Research Institute of Radiological Science, Severance Hospital, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul, , Republic of Korea. Address correspondence to Y. T. Oh (oytaik@yuhs.ac). 2 Department of Pathology, Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea. 3 Department of Urology, Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea. WEB This is a web exclusive article. AJR 2015; 205:W328 W X/15/2053 W328 American Roentgen Ray Society Prediction of Micrometastasis (< 1 cm) to Pelvic Lymph Nodes in Prostate Cancer: Role of Preoperative MRI OBJECTIVE. The purpose of this study was to retrospectively investigate whether preoperative MRI plays a key role in clinical prediction of micrometastasis (< 1 cm) to pelvic lymph nodes in prostate cancer. MATERIALS AND METHODS. One hundred one patients with prostate cancer who underwent preoperative MRI and radical prostatectomy with pelvic lymph node dissection were included. None of the patients had a pelvic lymph node with a short-axis diameter of 1 cm or larger on MRI. Both clinical (prostate-specific antigen, biopsy Gleason score, greatest percentage of biopsy core, and percentage of positive cores) and MRI parameters (tumor apparent diffusion coefficient and tumor staging) were assessed. The univariate, multivariate, and ROC curve analyses were conducted. RESULTS. Of 101 patients, nine (8.9%) had pelvic lymph node metastases. In univariate analysis, all of the clinical and MRI parameters were related to micrometastasis to pelvic lymph nodes (p < 0.05). However, multivariate analysis revealed that only preoperative MRI stage was significant (p = 0.044). AUC of preoperative MRI stage was (odds ratio, 21.7). Respective sensitivity and specificity of preoperative tumor staging by MRI were 100% and 65.2% with cutoff of T3a or more, and 88.9% and 94.6% with cutoff of T3b for predicting micrometastasis to pelvic lymph nodes. CONCLUSION. Preoperative MRI staging may play a role in prediction of micrometastasis (< 1 cm) to pelvic lymph nodes in prostate cancer. I n prostate cancer, the presence of metastasis to pelvic lymph nodes (LNs) is considered to represent systemic disease and to be related to poor prognosis after radical prostatectomy (RP) [1]. Therefore, other treatment modalities such as hormone therapy with or without radiotherapy have, in the past, often been recommended instead of surgery. More recently, RP with pelvic LN dissection (PLND) followed by adjuvant hormone therapy has been gaining in popularity because of evidence for improved prognostic outcomes in patients with pelvic LN positive prostate cancer [2, 3]. Moreover, removal of pelvic LNs alone may improve patients survival in LN-positive prostate cancer [4]. Thus, the accurate assessment of pelvic LN status is important to predict the prognosis and plan therapeutic strategies in prostate cancer. With imaging studies, it has been reported that morphologic analysis alone has limited usefulness because approximately 25% of normal-sized LNs are eventually proved to be metastatic LNs (so-called micrometastatic LNs ) after surgery [5]. Hence, researchers have investigated various imaging techniques for evaluating normal-sized pelvic LNs [6]. Although the application of superparamagnetic nanoparticles in MRI showed the best diagnostic performance [7 9], this technique has not yet been routinely used in prostate cancer. Similarly, choline PET/CT also showed good diagnostic potential, but its applicability in routine practice remains uncertain [10, 11]. Recently, Thoeny et al. [12] reported that DWI helps to predict the metastasis of normal-sized pelvic LNs in prostate cancer. Nevertheless, underestimation (i.e., false-negative rate) continues to be a limitation in DWI. The reported incidence of LN metastasis increases as pathologic T category of prostate cancer advances [13, 14]. However, to the best of our knowledge, it is still unclear whether determination of T category by 3-T MRI with multiparametric imaging could help to predict the presence of micrometas- W328 AJR:205, September 2015

2 MRI for Lymph Node Micrometastasis in Prostate Cancer Patients who underwent MRI and radical prostatectomy including PLND for prostate cancer (n = 162) Pathologic N0 (n = 92) Study group (n = 101) tasis to pelvic LNs in patients with only normal-sized LNs. Therefore, the purpose of our study was to retrospectively investigate the clinical role of preoperative MRI for prediction of micrometastasis (< 1 cm) to pelvic LNs in prostate cancer. Materials and Methods Patients The institutional review board at our institution approved this retrospective study, and requirement for informed consent was waived. Between January 2009 and December 2013, 162 patients with prostate cancer underwent both preoperative MRI and RP with PLND in our institution. Of those patients, 61 patients were excluded for the following reasons (Fig. 1): tumor volume was less than 0.5 cm 3 (n = 30); history of preoperative hormone therapy (n = 26); short-axis diameter of pelvic LN was 1 cm or larger (n = 4); and poor image quality of the apparent diffusion coefficient (ADC) map (n = 1). We included only patients whose prostate cancer was 0.5 cm 3 or larger in volume, because tumor volume less than 0.5 cm 3 is considered to be clinically insignificant [15, 16]. A pelvic LN with a short-axis diameter of 1 cm or larger is generally accepted as the size criterion for suggesting a metastatic node on cross-sectional imaging in spite of some controversies [14, 17]. Because the aim of our study was to predict the presence of metastatic pelvic LNs in patients with only normal-sized LNs, we excluded patients with LNs 1 cm or larger in short-axis diameter. Thus, a total of 101 patients were included in our study. Of those, nine had metastatic pelvic LNs (pathologic N1 group) and 92 had no metastatic LNs (pathologic N0 group) proven on biopsy after PLND. Exclusion Tumor volume < 0.5 cm 3 (n = 30) Neoadjuvant chemotherapy before surgery (n = 26) Pelvic lymph node 1 cm on MRI (n = 4) Poor image quality of ADC map (n = 1) Pathologic N1 (n = 9) Fig. 1 Flowchart outlines study group selection. PLND = pelvic lymph node dissection, ADC = apparent diffusion coefficient. MRI Examinations For all patients, prostate MRI was performed 3 5 weeks after the transrectal ultrasound-guided biopsy and before surgery. One of three types of 3-T MR scanners with a phased-array body coil (Intera Achieva, Philips Healthcare; Discovery MR750, GE Healthcare; or TrioTim, Siemens) was used for prostate imaging. The MRI examination included T1-weighted imaging, T2-weighted imaging, DWI, and dynamic contrast-enhanced imaging. The MRI protocols are summarized in Table 1. Before MRI, 20 mg of butylscopolamine bromide (Buscopan, Boehringer Ingelheim) was injected intramuscularly to suppress bowel peristalsis. Clinical and MRI Parameters In all patients, the following clinical parameters were recorded on the basis of the pathologic report of our institution: preoperative serum prostate-specific antigen (PSA), biopsy Gleason score, greatest percentage of biopsy core, and percentage of positive cores among all biopsy cores. Two radiologists (with 15 and 3 years of experience in TABLE 1: MRI Protocols prostate MRI, respectively), who were unaware of clinical and pathologic information, analyzed preoperative prostate MR images in consensus. With T1-weighted imaging, T2-weighted imaging, and DWI, the tumor ADC and MRI stage were investigated. The prostate cancer was localized once all of the following MRI features were seen within the same area: low-signal intensity on T2-weighted imaging (without high signal intensity on T1- weighted imaging, which finding would be suggestive of hemorrhage at the corresponding area); low ADC value; and earlier enhancement compared with adjacent prostatic tissues on dynamic contrast-enhanced imaging [18 20]. For measurement of tumor ADC, an ellipsoid ROI was drawn on the ADC map within the area suspected to be cancerous on the basis of the aforementioned MRI features. When multiple cancer foci 0.5 cm 3 or larger were suspected in a patient, the ADC measurement was performed only for the lesion with the lowest ADC value (because, among multiple cancerous lesions, one with a lower ADC would likely represent a more aggressive cancer focus) [21]. Preoperative tumor staging by MRI was performed according to the American Joint Committee on Cancer method of pathologic tumor categorization [22]. Extracapsular extension was considered to be present when at least one of the following MRI findings was seen: asymmetric neurovascular bundle; obliterated rectoprostatic angle; irregular bulging contour of the prostatic capsule; low signal intensity in the rectoprostatic fat, suggestive of cancer; or overt extracapsular cancer extension [23, 24]. Seminal vesicle invasion was considered to be present when at least one of the following MRI features was seen on T2-weighted imaging: loss of normal architecture of the seminal vesicle; abnormal homogeneous low signal intensity in the seminal vesicle (without high signal intensity on T1-weighted imaging, suggestive of hemorrhage at the same area); or evident tumor at the prostate base extending to the seminal vesicle [25 27]. Parameter T1-Weighted MRI T2-Weighted MRI DWI a DCE-MRI Orientation Axial Three planes b Axial Axial TR/TE / / / /1.7 Flip angle ( ) Matrix No. of excitations FOV (cm) Slice thickness (mm) Note DCE-MRI = dynamic contrast-enhanced MRI. a Two b values (0 and 1000 s/mm 2 ) were used. b Consisting of axial, sagittal, and coronal images. AJR:205, September 2015 W329

3 Park et al. The axial T1-weighted images covered the following pelvic LN chains: the obturator, internal iliac, external iliac, and common iliac chains. On the basis of per-patient analysis, micrometastasis to LNs was considered to be definitively present when metastatic cancer foci were pathologically confirmed in normal-sized LNs. Reference Standard The surgical specimens of the prostate gland and pelvic LNs were examined by a urogenital pathologist (with > 10 years of experience in prostate cancer histopathology). The findings of whole-mount step-section analysis of the resected prostate gland were used as the standard of reference for pathologic tumor categorization and assessment of tumor volume. Among 101 patients, 64 underwent limited PLND (obturator, internal iliac, and external iliac LN dissection), and 37 underwent extended PLND that was extended to the level of the common iliac bifurcation from the aorta. Statistical Analysis For comparison of clinical, radiologic, and pathologic parameters between pathologic N1 and N0 groups, the t test and Fisher exact test were used. Both univariate and multivariate analyses using logistic regression analysis were conducted to assess whether preoperative clinical and MRI parameters could predict the presence of micrometastasis to pelvic LNs. By use of ROC curve analysis, AUCs of preoperative clinical and MRI parameters were assessed for prediction of the presence of micrometastasis to pelvic LNs. Sensitivity and specificity of those preoperative parameters at the youden-selected cutoff were also analyzed. Statistical analyses were performed using SPSS (version 20.0, SPSS) and MedCalc (version 13.0, MedCalc Software). A p value of less than 0.05 was considered to be statistically significant. Results Of 101 patients, nine patients (8.9%) had a total of 33 pelvic LN metastases. The locations of those 33 metastatic pelvic LNs were as follows: obturator chain (n = 12); internal iliac chain (n = 5); external iliac chain (n = 7); and common iliac chain (n = 9). The use of preoperative MRI allowed correct tumor staging in 72 of 101 patients (overall accuracy, 71.3%). The ratio of upgraded and downgraded MRI staging was 15.8% (16/101) and 12.9% (13/101), respectively, based on the surgical results as the reference standard. Table 2 summarizes the clinical, radiologic, and pathologic characteristics between TABLE 2: Patient Characteristics Parameter Pathologic N0 (n = 92) Pathologic N1 (n = 9) p Preoperative characteristics Age at MRI (y) 69.0 ± ± a Prostate-specific antigen (ng/dl) 13.7 ± ± 48.4 < a Biopsy Gleason score (n) b Greatest percentage of biopsy core 48.3 ± ± a Percentage of positive cores c 28.9 ± ± 32.4 < a Tumor apparent diffusion coefficient (s/mm 2 ) 0.84 ± ± a MRI stage (n) < b 2a c a b 5 8 Postoperative characteristics Tumor volume (cm 3 ) 2.7 ± ± 4.4 < a Gleason score (n) b Pathologic T category (n) < b 2a c a b 5 5 Note Except where otherwise indicated, data are given as mean ± SD. N0 = absence of metastatic pelvic lymph node, N1 = presence of metastatic pelvic lymph node. a Derived using t test. b Derived using Fisher exact test. c Among all biopsy cores. pathologic N1 and N0 groups. For clinical parameters, the serum PSA of the pathologic N1 group (40.5 ± 48.4 ng/dl) was significantly higher than that of the pathologic N0 group (13.7 ± 15.1 ng/dl) (p < 0.001). The ratio of Gleason score was significantly different between the two groups (p = 0.001). The greatest percentage of the biopsy core and the percentage of positive cores in the pathologic N1 group were significantly higher than those in the pathologic N0 group (respectively 80.0% vs 48.3%, for greatest percentage of biopsy core [p = 0.002], and 62.1% vs 28.9%, for percentage of positive cores [p < 0.001]). For preoperative MRI parameters, mean (± SD) tumor ADC of the pathologic N1 group was significantly lower than that of the pathologic N0 group (respectively 0.72 ± 0.11 vs 0.84 ± 0.15 s/mm 2 ; p = 0.021). The ratio of MRI stage was significantly different between the two groups (p < 0.001). On MRI, the tumor stage of the pathologic N1 group (n = 9) consisted of T3a in one patient and T3b in eight patients. No patient with MRI-diagnosed T2 prostate cancer had metastasis to LNs. For postoperative parameters, the tumor volume, Gleason score, and ratio of pathologic tumor categorization were all different between pathologic N1 and N0 groups (p < 0.05). W330 AJR:205, September 2015

4 MRI for Lymph Node Micrometastasis in Prostate Cancer In univariate analysis, all clinical (PSA, biopsy Gleason score, greatest percentage of biopsy core, and percentage of positive core) and MRI (tumor ADC and MRI stage) parameters were related to the presence of micrometastasis to pelvic LNs (p < 0.05) (Table 3). However, the preoperative MRI stage was significant only for prediction of the presence of micrometastasis to pelvic LNs in multivariate analysis (odd ratio, ; p = 0.044) (Fig. 2 and Table 4). Among preoperative clinical and MRI parameters, MRI stage yielded the highest AUC (0.954) for predicting the presence of micrometastatic pelvic LNs, followed by biopsy Gleason score (AUC = 0.846) (Fig. 3 and Table 5). Respective sensitivity and specificity of the MRI stage for prediction of micrometastasis to pelvic LNs were 100% and 65.2% with a cutoff of T3a or higher and 88.9% and 94.6% with a cutoff of T3b. Discussion The incidence of LN metastasis increases as the T category of prostate cancer increases [13, 14]. Concordantly, in our study, surgery (RP with PLND) revealed that there were no metastatic LNs in T2 prostate cancer, whereas 20% of patients with T3 prostate cancer had metastatic LNs. In patients with T3 prostate cancer, the incidence of LN metastasis was 13% (4/31) in T3a and 50% (5/10) in T3b in our study. With the use of MRI, similar data were also found in our study; patients with T2 prostate cancer diagnosed by MRI had no metastatic LN, whereas 22% (9/41) of patients with T3 prostate cancer diagnosed by MRI had metastatic LNs. In prostate cancer, there are several predictive nomograms, including the Partin nomogram, for preoperative prediction of final pathologic status (i.e., organ-confined prostate cancer, extracapsular extension, seminal vesicle invasion, or LN metastasis) by use of clinical parameters [28 30]. In these nomograms, serum PSA, clinical stage, and biopsy findings are usually investigated because they are known to be associated with the final pathologic stage of prostate cancer. However, those clinical parameters do not show the anatomic status directly but rather suggest the probability of pathologic state when assessed overall. By contrast, MRI offers the advantage of visualizing the anatomic status of pelvic organs. For tumor staging of prostate cancer, it was reported that MRI provides good diagnostic performance regardless whether TABLE 3: Univariate Analysis of Preoperative Clinical and MRI Parameters for Predicting Pelvic Lymph Node Metastasis < 1 cm TABLE 4: Multivariate Analysis of Preoperative Clinical and MRI Parameters for Predicting Pelvic Lymph Node Metastasis < 1 cm TABLE 5: Diagnostic Performance of Preoperative Clinical and MRI Parameters for Predicting Pelvic Lymph Node Metastasis < 1 cm Cutoff Sensitivity (%) Specificity (%) AUC Prostate serum antigen > 13.5 mg/dl Biopsy Gleason score > Greatest percentage of biopsy core > 50% Percentage of positive cores a > 33% Tumor apparent diffusion coefficient 0.74 s/mm MRI stage higher than 3a a Among all biopsy cores. Parameter Odds Ratio (95% CI) p Prostate serum antigen ( ) Biopsy Gleason score ( ) Greatest percentage of biopsy core ( ) Percentage of positive cores a ( ) Tumor apparent diffusion coefficient ( ) MRI stage ( ) a Among all biopsy cores. Parameter Odds Ratio (95% CI) p Prostate serum antigen ( ) Biopsy Gleason score ( ) < Greatest percentage of biopsy core ( ) Percentage of positive cores a ( ) < Tumor apparent diffusion coefficient < ( ) MRI stage ( ) < a Among all biopsy cores. a primary tumor is confined to the prostate gland [24 26]. In our study, overall accuracy of MRI for tumor staging was 71.3%, and these data are in line with data from a previous study (which reported 72% accuracy) using a 3-T MRI unit with a phased-array coil [31]. Furthermore, a previous study showed the feasibility of MRI as a preoperative diagnostic tool for one-step determination of TNM category in high-risk prostate cancer [32]. For these reasons, MRI staging might be the only independent parameter (odds ratio, ; p = 0.044) and would show the highest AUC for predicting micrometastatic pelvic LNs in our study because the incidence of LN metastasis is closely related to the prostate cancer stage [13, 14]. Previously, Wang et al. [33] reported that MRI findings suggestive of T3 (namely extracapsular extension or seminal vesicle invasion) and abnormally increased size of a LN ( cm in short-axis diameter) had an incremental predictive value for LN metastasis in prostate cancer when either was combined with the Partin nomogram. Our results are in line with those of their study; however, there are several differences in our study design. First, we excluded pelvic LNs of 1.0 cm or larger. Independent from the status of the primary tumor, LN size alone could be an important clue for suggesting the likelihood of metastasis [14, 34]. Thus, radiologists often have difficulty interpreting normal-sized LNs, for which reason we focused on small LNs. Second, multiparametric imaging, such as DWI and dynamic contrast-enhanced MRI, was used for evaluating prostate cancer in our study. For assessment AJR:205, September 2015 W331

5 Park et al. C D Fig year-old man with pathologic T3aN1 prostate cancer. Serum prostate serum antigen, biopsy Gleason score, greatest percentage of biopsy core, and percentage of positive cores were 7.4 ng/dl, 9, 70%, and 50%, respectively. A C, MR images show prostate cancer. Focal area of low signal intensity on T2-weighted image (A), low apparent diffusion coefficient (ADC) value (0.92 s/mm 2 ) on ADC map (B), and relatively earlier enhancement on dynamic contrast-enhanced MR image (C) were seen in left peripheral zone of prostate apex (arrows). Two radiologists came to consensus that left extracapsular extension is present without seminal vesicle invasion (MRI stage T3a). D, On surgical specimen, cancer focus with left extracapsular extension was identified at corresponding site of prostate gland (circled region). Pelvic lymph node dissection revealed single metastatic lymph node in left obturator chain. of tumor stage, the accurate detection and localization of cancer foci are requisites. It has been reported that multiparametric imaging allows better diagnostic performance for detection and localization when combined with conventional T2-weighted imaging [19]. Third, the primary-tumor ADC was also analyzed. On the basis of the background of the relationship between tumor cellularity and ADC (i.e., the higher the cellularity is, the lower will be the tumor ADC value), many studies have shown that tumor ADC reflects tumor aggressiveness in prostate cancer [21]. In our study, the univariate analysis revealed that tumor ADC is related to LN metastasis A (p = 0.010), although it became insignificant in the multivariate analysis. In our study, the optimal cutoff value of MRI stage for prediction of micrometastasis to pelvic LNs was T3b (sensitivity, 88.9%; specificity, 94.6%), which was the youden-selected cutoff. However, with a cutoff of T3a or higher, the sensitivity of MRI stage increased up to 100%, although specificity decreased to 65.2%. Moreover, in a previous study that analyzed 5274 patients with prostate cancer, the pathologic categorization of T3a or higher was a significant factor for the presence of LN metastasis in the multivariate analysis (odds ratio, 8.58; p < 0.001) [13]. From this point of view, we think that PLND could be omitted in MRIdiagnosed T2 prostate cancer. Meanwhile, more data with a large population are required in order to determine which MRI stage (T3a vs T3b) is the optimal threshold at which PLND is indicated. Although preoperative MRI stage showed promise in nodal staging, biopsy Gleason score may also be a powerful parameter for prediction of pelvic LN metastasis. In our study, the ROC curve analysis revealed that biopsy Gleason score (AUC = 0.846) has the second highest AUC, and its odds ratio was in multivariate analysis in spite of statistical insignificance (p = 0.076). Because B W332 AJR:205, September 2015

6 MRI for Lymph Node Micrometastasis in Prostate Cancer Sensitivity (%) of the small number of pathologic N1 cancers (n = 9) in our study population, a further investigation with a large population may be needed. There is evidence to support the clinical significance of biopsy Gleason score for prediction of pelvic LN metastases in prostate cancer [29, 33]. Our study had limitations. First, the number of patients with LN metastasis was small (n = 9). However, the incidence rate in our study (8.9%) was in line with those of previous reports (~ 5 10%) [14, 29, 33]. A multicenter study with a large population may be required. Second, extended PLND was not performed for all patients (limited PLND, n = 64; extended PLND, n = 37). Thus, the incidence of true pelvic LN metastasis may be underestimated because limited PLND has a potential risk for missing metastatic LNs [35]. Third, ADC was calculated with the monoexponential model. The biexponential model may depict a more accurate state of diffusion restriction than the monoexponential model because it reflects both fast and slow components of biologic tissues in calculation of ADC [36]. Nevertheless, for evaluation of prostate cancer, several previous studies also reported the feasibility of the monoexponential model with two b values of 0 and 1000 s/mm 2 [37, 38]. Finally, the quantitative analysis for pelvic LNs was not performed with multiparametric imaging such as the ADC of a LN. We considered that the ADC measurement with a small ROI for the small LNs (< 1 cm) may not represent a small Prostate serum antigen Biopsy Gleason score Greatest percentage of biopsy core Percentage of positive cores Tumor apparent diffusion coefficient MRI stage Specificity (%) Fig. 3 ROC curve analysis shows diagnostic performance of preoperative clinical and MRI parameters for predicting pelvic lymph node metastasis less than 1 cm. MRI stage yielded highest AUC (0.954) for predicting presence of micrometastasis to pelvic LNs, followed by biopsy Gleason score (AUC = 0.846). burden of focally deposited cancer cells [39]. In addition, the potential for measurement error exists in analyzing such small lesions because DWI has limited spatial resolution. In conclusion, preoperative MRI stage may play a role in prediction of micrometastasis (< 1 cm) to pelvic LNs in prostate cancer. For patients with T2 prostate cancer diagnosed by MRI, PLND could be omitted during surgery because of the low incidence of positive LNs. Nevertheless, further prospective studies with a large population and extended PLND as the reference standard are required in order to validate our results. References 1. Gervasi LA, Mata J, Easley JD, et al. Prognostic significance of lymph nodal metastases in prostate cancer. J Urol 1989; 142: Cheng L, Zincke H, Blute ML, Bergstralh EJ, Scherer B, Bostwick DG. Risk of prostate carcinoma death in patients with lymph node metastasis. Cancer 2001; 91: Messing EM, Manola J, Yao J, et al. Immediate versus deferred androgen deprivation treatment in patients with node-positive prostate cancer after radical prostatectomy and pelvic lymphadenectomy. Lancet Oncol 2006; 7: Abdollah F, Gandaglia G, Suardi N, et al. More extensive pelvic lymph node dissection improves survival in patients with node-positive prostate cancer. Eur Urol 2015; 67: Giannarini G, Petralia G, Thoeny HC. Potential and limitations of diffusion-weighted magnetic resonance imaging in kidney, prostate, and bladder cancer including pelvic lymph node staging: a critical analysis of the literature. Eur Urol 2012; 61: Fortuin A, Rooij M, Zamecnik P, Haberkorn U, Barentsz J. Molecular and functional imaging for detection of lymph node metastases in prostate cancer. Int J Mol Sci 2013; 14: Harisinghani MG, Barentsz J, Hahn PF, et al. Noninvasive detection of clinically occult lymphnode metastases in prostate cancer. N Engl J Med 2003; 348: Thoeny HC, Triantafyllou M, Birkhaeuser FD, et al. Combined ultrasmall superparamagnetic particles of iron oxide-enhanced and diffusionweighted magnetic resonance imaging reliably detect pelvic lymph node metastases in normalsized nodes of bladder and prostate cancer patients. Eur Urol 2009; 55: Triantafyllou M, Studer UE, Birkhauser FD, et al. Ultrasmall superparamagnetic particles of iron oxide allow for the detection of metastases in normal sized pelvic lymph nodes of patients with bladder and/or prostate cancer. Eur J Cancer 2013; 49: Vag T, Heck MM, Beer AJ, et al. Preoperative lymph node staging in patients with primary prostate cancer: comparison and correlation of quantitative imaging parameters in diffusion-weighted imaging and 11 C-choline PET/CT. Eur Radiol 2014; 24: Heck MM, Souvatzoglou M, Retz M, et al. Prospective comparison of computed tomography, diffusion-weighted magnetic resonance imaging and [ 11 C]choline positron emission tomography/ computed tomography for preoperative lymph node staging in prostate cancer patients. Eur J Nucl Med Mol Imaging 2014; 41: Thoeny HC, Froehlich JM, Triantafyllou M, et al. Metastases in normal-sized pelvic lymph nodes: detection with diffusion-weighted MR imaging. Radiology 2014; 273: Abdollah F, Suardi N, Gallina A, et al. Extended pelvic lymph node dissection in prostate cancer: a 20-year audit in a single center. Annals of Oncology: official journal of the European Society for Medical Oncology / ESMO 2013; 24: Pano B, Sebastia C, Bunesch L, et al. Pathways of lymphatic spread in male urogenital pelvic malignancies. RadioGraphics 2011; 31: Mazzucchelli R, Barbisan F, Scarpelli M, et al. Is incidentally detected prostate cancer in patients undergoing radical cystoprostatectomy clinically significant? Am J Clin Pathol 2009; 131: Wolters T, Roobol MJ, van Leeuwen PJ, et al. A critical analysis of the tumor volume threshold for clinically insignificant prostate cancer using a data set of a randomized screening trial. J Urol 2011; 185: AJR:205, September 2015 W333

7 Park et al. 17. Roy C, Le Bras Y, Mangold L, et al. Small pelvic 242: endorectal and phased-array MRI in the predic- lymph node metastases: evaluation with MR im- 26. Soylu FN, Peng Y, Jiang Y, et al. Seminal vesicle tion of pelvic lymph node metastasis in prostate aging. Clin Radiol 1997; 52: invasion in prostate cancer: evaluation by using cancer. AJR 2006; 186: Barentsz JO, Richenberg J, Clements R, et al. multiparametric endorectal MR imaging. Radiol- 34. Jager GJ, Barentsz JO, Oosterhof GO, Witjes JA, ESUR prostate MR guidelines Eur Radiol ogy 2013; 267: Ruijs SJ. Pelvic adenopathy in prostatic and uri- 2012; 22: Sala E, Akin O, Moskowitz CS, et al. Endorectal nary bladder carcinoma: MR imaging with a 19. Hoeks CM, Barentsz JO, Hambrock T, et al. Prostate cancer: multiparametric MR imaging for detection, localization, and staging. Radiology 2011; 261: Verma S, Turkbey B, Muradyan N, et al. Overview of dynamic contrast-enhanced MRI in prostate cancer diagnosis and management. AJR 2012; 198: Hambrock T, Hoeks C, Hulsbergen-van de Kaa C, et al. Prospective assessment of prostate cancer aggressiveness using 3-T diffusion-weighted magnetic resonance imaging-guided biopsies versus a systematic 10-core transrectal ultrasound prostate biopsy cohort. Eur Urol 2012; 61: Edge S, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, eds. AJCC cancer staging manual, 7th ed. New York: Springer, Fütterer JJ, Heijmink SW, Scheenen TW, et al. Prostate cancer: local staging at 3-T endorectal MR imaging early experience. Radiology 2006; 238: 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: Wang L, Hricak H, Kattan MW, et al. Prediction of seminal vesicle invasion in prostate cancer: incremental value of adding endorectal MR imaging to the Kattan nomogram. Radiology 2007; MR imaging in the evaluation of seminal vesicle invasion: diagnostic accuracy and multivariate feature analysis. Radiology 2006; 238: Eifler JB, Feng Z, Lin BM, et al. An updated prostate cancer staging nomogram (Partin tables) based on cases from 2006 to BJU Int 2013; 111: Briganti A, Larcher A, Abdollah F, et al. Updated nomogram predicting lymph node invasion in patients with prostate cancer undergoing extended pelvic lymph node dissection: the essential importance of percentage of positive cores. Eur Urol 2012; 61: Kluth LA, Abdollah F, Xylinas E, et al. Clinical nodal staging scores for prostate cancer: a proposal for preoperative risk assessment. Br J Cancer 2014; 111: Park BK, Kim B, Kim CK, Lee HM, Kwon GY. Comparison of phased-array 3.0-T and endorectal 1.5-T magnetic resonance imaging in the evaluation of local staging accuracy for prostate cancer. J Comput Assist Tomogr 2007; 31: Pasoglou V, Larbi A, Collette L, et al. One-step TNM staging of high-risk prostate cancer using magnetic resonance imaging (MRI): toward an upfront simplified all-in-one imaging approach? Prostate 2014; 74: Wang L, Hricak H, Kattan MW, et al. Combined three-dimensional TI-weighted magnetizationprepared-rapid gradient-echo sequence. AJR 1996; 167: Briganti A, Chun FK, Salonia A, et al. Critical assessment of ideal nodal yield at pelvic lymphadenectomy to accurately diagnose prostate cancer nodal metastasis in patients undergoing radical retropubic prostatectomy. Urology 2007; 69: Shinmoto H, Tamura C, Soga S, et al. An intravoxel incoherent motion diffusion-weighted imaging study of prostate cancer. AJR 2012; 199:[web]W496 W Jung SI, Donati OF, Vargas HA, Goldman D, Hricak H, Akin O. Transition zone prostate cancer: incremental value of diffusion-weighted endorectal MR imaging in tumor detection and assessment of aggressiveness. Radiology 2013; 269: Woodfield CA, Tung GA, Grand DJ, Pezzullo JA, Machan JT, Renzulli JF 2nd. Diffusion-weighted MRI of peripheral zone prostate cancer: comparison of tumor apparent diffusion coefficient with Gleason score and percentage of tumor on core biopsy. AJR 2010; 194:[web]W316 W Scolyer RA, Li LX, McCarthy SW, et al. Micromorphometric features of positive sentinel lymph nodes predict involvement of nonsentinel nodes in patients with melanoma. Am J Clin Pathol 2004; 122: W334 AJR:205, September 2015