Mapping of Pelvic Lymph Node Metastases in Prostate Cancer

Transcription

1 EUROPEAN UROLOGY 63 (2013) available at journal homepage: Platinum Priority Prostate Cancer Editorial by Alberto Briganti, Nazareno Suardi, Andrea Gallina, Firas Abdollah and Francesco Montorsi on pp of this issue Mapping of Pelvic Lymph Node Metastases in Prostate Cancer Steven Joniau a,y, *, Laura Van den Bergh b,y, Evelyne Lerut c, Christophe M. Deroose d, Karin Haustermans b, Raymond Oyen e, Tom Budiharto b, Filip Ameye a, Kris Bogaerts f, Hein Van Poppel a a Department of Urology, University Hospitals Leuven, Leuven, Belgium; b Department of Radiation Oncology, University Hospitals Leuven, Leuven, Belgium; c Department of Histopathology, University Hospitals Leuven, Leuven, Belgium; d Department of Nuclear Medicine, University Hospitals Leuven, Leuven, Belgium; e Department of Radiology, University Hospitals Leuven, Leuven, Belgium; f L-Biostat, Catholic University of Leuven, Leuven, Belgium Article info Article history: Accepted June 28, 2012 Published online ahead of print on July 6, 2012 Keywords: Lymphadenectomy Lymph node dissection template Lymph node staging Prostate cancer Abstract Background: Opinions about the optimal lymph node dissection (LND) template in prostate cancer differ. Drainage and dissemination patterns are not necessarily identical. Objective: To present a precise overview of the lymphatic drainage pattern and to correlate those findings with dissemination patterns. We also investigated the relationship between the number of positive lymph nodes (LN+) and resected lymph nodes (LNs) per region. Design, setting, and participants: Seventy-four patients with localized prostate adenocarcinoma were prospectively enrolled. Patients did not show suspect LNs on computed tomography scan and had an LN involvement risk of 10% but 35% (Partin tables) or a ct3 tumor. Intervention: After intraprostatic technetium-99m nanocolloid injection, patients underwent planar scintigraphy and single-photon emission computed tomography imaging. Then surgery was performed, starting with a sentinel node (SN) procedure and a superextended lymphadenectomy followed by radical prostatectomy. Outcome measurements and statistical analysis: Distribution of scintigraphically detected SNs and removed SNs per region were registered. The number of LN+, as well as the percentage LN+ of the total number of removed LNs per region, was demonstrated in combining data of all patients. The impact of the extent of LND on N-staging and on the number of LN+ removed was calculated. Results and limitations: A total of 470 SNs were scintigraphically detected (median: 6; interquartile range [IQR]: 3 9), of which 371 SNs were removed (median: 4; IQR: ). In total, 91 LN+ (median: 2; IQR: 1 3) were found in 34 of 74 patients. The predominant site for LN+ was the internal iliac region. An extended LND (elnd) would have correctly staged 32 of 34 patients but would have adequately removed all LN+ in only 26 of 34 patients. When adding the presacral region, these numbers increased to 33 of 34 and 30 of 34 patients, respectively. Conclusions: Standard elnd would have correctly staged the majority of LN+ patients, but 13% of the LN+ would have been missed. Adding the presacral LNs to the template should be considered to obtain a minimal template with maximal gain. Note: This manuscript was invited based on the 2011 European Association of Urology meeting in Vienna. # 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved. y Steven Joniau and Laura Van den Bergh are joint first authors of this paper. * Corresponding author. Department of Urology, UH Leuven Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium. Tel ; Fax: address: steven.joniau@uzleuven.be (S. Joniau) /$ see back matter # 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved.

2 EUROPEAN UROLOGY 63 (2013) Introduction The presence of lymph node (LN) metastases (positive lymph nodes [LN+]) is an important prognostic factor in prostate cancer (PCa) [1]. Until now, pelvic LN dissection (LND) has proved to be the most accurate and reliable nodal staging procedure [2,3], as currently available imaging techniques report low sensitivity [4]. LND also has therapeutic intent, since several studies have indicated the possibility of long-term survival in the presence of limited LN involvement (LNI) [5]. Good evidence exists that LND can be omitted in selected low-risk PCa patients [6]. However, when performed for intermediate- and high-risk patients, LND should be extended [7,8]. Series of extended LNDs (elnds) have shown that the actual rate of LN+ is higher than observed with limited LND (1LND) [7,9]. It has also been demonstrated that in many cases, even this template does not cover all primary landing sites. Several studies have been performed with intraprostatic injection of technetium-99m nanocolloid to obtain precise information on lymphatic drainage, and revisions of the standard template have been suggested [10,11]. Unfortunately, studies establishing the link between drainage and dissemination patterns with explicit data are rare. Although it has been shown that nodal count at LND is closely associated with metastatic rate [12], one should ensure that the yield of dissecting an additional region outweighs efforts and potential morbidity before further extending anatomic LND limits. Therefore, the aim of this anatomic mapping study was to present an overview of prostatic drainage patterns and to correlate findings with dissemination patterns in patients at high risk for LNI. We also investigated the relationship between the number of affected and resected LNs (ie, LN density) per region and hypothesized that an obvious difference could indicate the presence of a certain hierarchy in the drainage chain. 2. Patients and methods 2.1. Patients This observational study is a subanalysis of a larger prospective imaging study designed to assess sensitivity, specificity, and positive and negative predictive value of carbon 11 labeled choline positron emission tomography computed tomography and diffusion-weighted magnetic resonance imaging (MRI) for preoperative N-staging in PCa patients at high risk for LNI [13]. Between February 2008 and February 2011, Table 1 Patient and disease characteristics Characteristics Patients with negative LNs, n = 40 Patients with positive LNs, n = 34 Total, n = 74 Age, yr, median (range) 65.1 ( ) 63.9 ( ) 64.5 ( ) Preoperative PSA, ng/ml, median (range) 10.1 ( ) 10.8 ( ) 10.4 ( ) Clinical T stage (2002 TNM) 1c 0 (0) 1 (2.9) 1 (1.6) 2a 1 (2.5) 1 (2.9) 2 (2.7) 2b 2 (5.0) 1 (2.9) 3 (4.1) 2c 12 (30.0) 2 (5.9) 14 (18.9) 3a 23 (57.5) 19 (55.9) 42 (56.8) 3b 2 (5.0) 9 (26.5) 11 (14.9) 4 0 (0) 1 (2.9) 1 (1.6) Biopsy Gleason score 6 1 (2.5) 0 (0) 1 (1.6) 7 (3 + 4) 17 (42.5) 6 (17.6) 23 (31.1) 7 (4 + 3) 6 (15.0) 11 (32.4) 17 (23.0) 8 10 (25.0) 11 (32.4) 21 (28.4) 9 (4 + 5) 4 (10.0) 5 (14.7) 7 (9.5) 9 (5 + 4) 2 (5.0) 0 (0) 2 (2.7) 10 0 (0) 1 (2.9) 1 (1.6) Pathologic T stage (2002 TNM) 2b 2 (5.0) 0 (0) 2 (2.7) 2c 24 (60.0) 6 (17.6) 30 (40.5) 3a 11 (27.5) 9 (26.5) 20 (27.0) 3b 1 (2.5) 18 (52.9) 19 (25.7) 4 2 (5.0) 1 (2.9) 3 (4.1) Pathologic Gleason score 7 (3 + 4) 14 (35.0) 3 (8.8) 17 (23.0) 7 (4 + 3) 17 (42.5) 8 (23.5) 25 (33.8) 8 3 (7.5) 11 (32.4) 14 (18.9) 9 (4 + 5) 5 (12.5) 10 (29.4) 15 (20.3) 9 (5 + 4) 1 (2.5) 1 (2.9) 2 (2.7) 10 0 (0) 1 (2.9) 1 (1.6) Predicted risk for LN involvement Partin tables, %, median (range) 11.5 (10 29) 15 (10 29) 12 (10 29) Briganti nomogram, %, median (range) 54 (6 75) 59 (17 75) 56 (6 75) Nodes examined per patient, no., median (range) 21.5 (7 42) 20.5 (10 49) 21 (7 49) LN = lymph node; PSA = prostate-specific antigen. Numbers between parentheses are percentages unless indicated otherwise.

3 452 EUROPEAN UROLOGY 63 (2013) consecutive patients (median age: 65 yr; range: yr) with localized, biopsy-proven prostate adenocarcinoma were scheduled for radical retropubic prostatectomy (RRP) and superextended LND (selnd). The primary tumor was staged by digital rectal examination and transrectal ultrasonography (TRUS). Inclusion criteria were (1) risk of LNI 10% but 35% (Partin tables) (as nopredictivemodel existed for ct3 at the time of the study conception, ct3 tumors were considered stage ct2c); (2) no pelvic LNI on contrast-enhanced CT (ie, 8mm in the transverse dimension); (3) no bone metastasis on bone scan; (4) World Health Organization performance status <2; (5) no previous hormonal therapy, radiotherapy, or prostatectomy; and (6) no previous/other malignancy. The study was approved by the ethics committee, and written informed consents were obtained. Table 1 depicts patient and disease characteristics Scintigraphy: technique and interpretation On the day of the surgery, patients were transrectally injected with Tc-99m nanocolloid (Nanocoll; GE Healthcare, Amersham Health, USA) under TRUS guidance. Three radionuclide applications of 20 MBq (0.5 ml) were performed per lobe with a Chiba needle ( mm). Injections were performed bilaterally in the basal, middle, and apical portions of the prostate (sextant) in a standard fashion. Contrary to other tumor types, PCa is known to be notoriously multifocal within the gland. Therefore, the sentinel node (SN) procedure for PCa differs from other cancers, in which the tracer is injected peritumorally/intratumorally. A second reason for this differing SN procedure is the poor tumor visibility on TRUS. Approximately 2 h after injection, patients underwent planar scintigraphy and single-photon emission computed tomography (SPECT) imaging of the pelvis and abdomen on a Trionix BIAD dualheaded gamma camera (Biad Trionix Research Laboratories, Twinsburg, OH, USA) equipped with low-energy, ultrahigh-resolution (LEUR) collimators. Scintigraphy was performed for 10 min with a 20% energy window and a matrix with a cobalt Co 57 filled flood source. SPECT images were acquired with similar settings with continuous motion over 3608 (data binned per 38) for 20 min and reconstructed using maximum-likelihood expectation maximization with an 8.4-mm [(Fig._1)TD$FIG] Gaussian smooth without attenuation or scatter correction. To facilitate anatomic localization, SPECT images were fused with the staging computed tomography (CT) scan (110 kv[p], 85 mas) with MIMVista software (MIM Software Inc., Cleveland, OH, USA) using manual rigid registration based on visible uptake in bone marrow, liver, and kidneys. The last 17 patients were imaged on a Symbia 16 SPECT/CT dualheaded gamma camera (Siemens, Erlangen, Germany). Planar images were obtained with LEUR collimators for 10 min with a 15% energy window and a matrix with the flood source. SPECT images were acquired with similar energy settings and a matrix with continuous motion over 1808 (data binned per 38) for 17 min. Subsequently, a low-dose CT (110 kv[p], 30 mas) was performed. SPECT images were reconstructed using ordered subset expectation maximization Flash 3D with scatter correction, with and without attenuation correction based on CT. Foci of activity that were significantly higher than background activity on visual inspection unrelated to injection site, bladder, rectum, bone marrow, kidneys, or liver were considered draining LNs (Fig. 1). All scans were analyzed by an experienced nuclear medicine physician (C.D.) Surgery Surgery was performed by one experienced urologist (S.J.). First, guided by the SPECT/CT images, all pelvic drainage regions were systematically screened for the presence of SNs with a hand-held gamma-detection probe (Neo2000 Gamma Detection System; Neoprobe Corp., Dublin, OH, USA). The tip of the probe was always pointed away from the prostate to avoid signalinterference, andgood contactwith the tissues wasensured. When a focal high-intensity signal was located, this region was isolated and mapped on a template (Fig. 2a). SNs located outside the selnd template according to SPECT/CT images were resected only if technically justified. Second, a backup selnd was performed including all nodal/fibro fatty tissue at the following regions: Common iliac region. Borders: aortic bifurcation, bifurcation internal/ external iliac arteries, psoas muscle and genitofemoral nerve and medial border common iliac artery Presacral region. Triangle between medial borders of common iliac arteries and line connecting internal/external iliac arteries bifurcations; dorsal border: promontory and proximal sacrum (S1 S2) Fig. 1 (a) Sentinel node (SN) procedure in a prostate cancer patient: planar and single-photon emission computed tomography (SPECT) imaging with lead shielding of the prostate, 2 h after transrectally injecting 120 MBq of technetium-99m nanocolloid. (b) To facilitate anatomic localization of the SNs, SPECT images were fused with the pelvic staging computed tomography scan.

4 [(Fig._2)TD$FIG] EUROPEAN UROLOGY 63 (2013) Fig. 2 (a) Sentinel nodes were isolated and mapped separately on a standardized lymph node (LN) template. (b) Lymphadenectomy specimens were laid out on a second LN map. External iliac region. Borders: bifurcation internal/external iliac arteries, circumflex iliac vein and endopelvic fascia, psoas muscle and genitofemoral nerve and medial border external iliac artery Obturator fossa region. Borders: bifurcation internal/external iliac arteries, pelvic floor, obturator nerve, and medial border external iliac artery Internal iliac region. Borders: bifurcation internal/external iliac arteries, pelvic floor, bladder wall, obturator nerve. To limit morbidity, LND was not extended above the level of the aortic bifurcation. In most cases, all presacral nodes were removed through a right-sided approach, as the left common iliac vein prohibited a safe leftsided approach. Therefore, this area was considered as one single midline region. The pararectal nodes located behind the pararectal fascia in close contact with the rectal muscular layer were resected only in case an SN was located there that could be reached by opening the fascia. Third, all lymphatic tissue was screened ex vivo for radioactivity to rule out the presence of previously missed SNs and laid out on a second map (Fig. 2b). Figure 3 shows an intraoperative picture after LND. Finally, an open RRP was performed. fatty tissue. All stations were examined by palpation, visual inspection, and sectioning. The lamellated (1-mm) LNs were then embedded in paraffin. From these blocks, two 3 5-m sections were cut per 300 m until the whole LN was cut. An experienced uropathologist (E.L.) microscopically evaluated the presence of macrometastases and micrometastases (>0.2 mm and <2 mm, respectively) and isolated tumor cells for one section of each depth after hematoxylin and eosin staining. Prostatectomy specimens were classified according to the 2002 TNM classification, and Gleason scores were determined Statistical methodology Data are presented as percentage or median (and range) for continuous variables and rates (and percentages) for discrete variables. Ninety-five percent bootstrap confidence intervals were determined with the percentile method using 2000 bootstrap samples (SAS v.9.2). Regional LN density was calculated by dividing the number of LN+ by the total number of removed LNs for each region. 3. Results 2.4. Histologic examination All labeled specimens were delivered to the pathology department on standardized maps. LNs were fixed overnight in 6% formalin to dissolve 3.1. Sentinel nodes detected scintigraphically In 1 of 74 patients (1%), no SNs were recorded on SPECT images; in 18 patients (27%), SNs were located unilaterally.

5 454 [(Fig._3)TD$FIG] EUROPEAN UROLOGY 63 (2013) Fig. 3 (a) Intraoperative picture illustrating superextended lymph node dissection that was performed in all patients; (b) external iliac region (blue), obturator fossa region (green), internal iliac region (yellow), common iliac region (purple), and presacral region (red). In total, 470 SNs were detected (median: 6; interquartile range [IQR]: 3 9) and were distributed as follows: internal iliac (n = 107, 23%), common iliac (n = 88, 19%), obturator fossa (n = 78, 17%), external iliac (n =77,16%),paraaortic (n = 49, 10%), presacral (n = 35, 7%), aortic bifurcation (n = 17, 4%), pararectal (n = 14, 3%), paravesical (n =2,0%), mesenteric fat (n = 2, 0%), and inguinal (n = 1, 0%) (Fig. 4a). Based on this drainage pattern, 56% of SNs are located within the standard elnd template (external iliac, obturator fossa, and internal iliac regions) Sentinel nodes detected intraoperatively In three patients (4%), no SN was detected intraoperatively. Including these patients, the median number of SNs removed was 4 (IQR: ), with a total of 371 SNs. Figure 4b illustrates the distribution: obturator fossa [(Fig._4)TD$FIG] (n = 94, 25%), internal iliac (n = 91, 25%), external iliac (n = 71, 19%), common iliac (n = 53, 14%), presacral (n = 49, 13%), pararectal (n = 6, 2%), paraaortic (n = 3, 1%), mesenteric fat (n = 2, 1%), aortic bifurcation (n = 1, 0%), and paravesical (n = 1, 0%). SNs located outside the selnd template were resected only if technically justified (eg, SNs above the level of the aortic bifurcation), which explains the 79% resection rate. It is possible that not all SNs were preoperatively mapped in the exact region in which they were intraoperatively located because of limited mismatches between SPECT images and CT scan Lymph node dissection and positive lymph nodes The bilateral LND combined with an SN procedure resulted in 1656 LNs (median: 21; IQR: 16 27). A total of 91 LN+ (median: 2; IQR: 1 3) were found in 34 of 74 patients (46%). Fig. 4 (a) Number of sentinel nodes (SNs) scintigraphically detected per region and (b) number of SNs removed per region. Dimensions of the circles correlate with the numbers.

6 [(Fig._5)TD$FIG] EUROPEAN UROLOGY 63 (2013) Fig. 5 (a) Number of positive lymph nodes (LN+) per region and (b) percentage of LN+ of the total number of removed lymph nodes per region in 74 patients. Dimensions of the circles correlate with the numbers. The predominant site for LN+ was the internal iliac region (n = 32, 35%), followed by the external iliac region (n = 24, 26%) and the obturator fossa region (n = 23, 25%). Remaining metastases were located in the presacral region (n = 8, 9%), common iliac region (n = 3, 3%), and aortic bifurcation region (n = 1, 1%) (Fig. 5a). An llnd (external iliac and obturator fossa regions) would have correctly staged 26 of 34 patients (76%) and would have removed all LN+ in only 10 of 34 patients (29%). An elnd would have correctly staged 32 of 34 patients (94%) but would have removed all LN+ in only 26 of 34 patients (76%). Table 2 demonstrates the impact of the extent of LND on N-staging, including a new standard elnd template that we propose based on these results Lymph node density and hierarchy The hypothesis of a certain hierarchic order with a predilection for the internal iliac region is confirmed when regional LN density is determined. Figure 5b shows that in this region, LN density is twice the rate observed in the second predominant site: internal iliac region (11%), external iliac region (6%), obturator fossa region (5%), presacral region (5%), and common iliac region (1%). In five patients (15%), four patients (12%), three patients (9%), one patient (3%), and one patient (3%), LN+ were exclusively located in the external iliac, obturator fossa, internal iliac, presacral, and paraaortic regions, respectively. All three patients with common iliac region metastases also had LN+ within the elnd template. Table 2 The impact of the extent of lymph node dissection on nodal staging, including a proposed new standard extended lymph node dissection template based on these results N+ patients correctly staged, no. (%, 95% CI) N+ patients in whom all N+ removed, no. (%, 95% CI) N+ removed, no. (%) Nodes removed, no., median (IQR) Obturator LND 16/34 (47, 29 62) 5/34 (15, 3 26) 23/91 (25) 6 (4 9) Limited LND (external iliac plus 26/34 (76, 62 88) 10/34 (29, 15 44) 47/91 (52) 12 (8 17) obturator regions) Extended LND (external and internal 32/34 (94, ) 26/34 (76, 62 88) 79/91 (87) 16 (10 21) iliac plus obturator regions) New suggested LND template (elnd plus 33/34 (97, ) 30/34 (88, 76 97) 87/91 (96) 18 (12 23) presacral regions) Superextended LND (elnd plus presacral 33/34 (97, ) 33/34 (97, ) 90/91 (99) 21 (16 27) plus common iliac regions) Superextended LND plus SN 34/34 (100) 34/34 (100) 91/91 (100) 21 (16 27) N+ = node-positive; CI = confidence interval; IQR = interquartile range; LND = lymph node dissection; elnd = extended lymph node dissection; SN = sentinel node.

7 456 EUROPEAN UROLOGY 63 (2013) In total, 46 of 91 LN+ (51%) presented as SNs and, thus, primary landing sites. Of 371 SNs, 46 (12%) were LN+. When analyzing LN density exclusively in these primary landing sites to rule out potential confounding factors, the order of sequence remained identical Morbidity Seventeen of 74 patients (23%) had postoperative LNDrelated morbidity. Eight patients (11%) developed a lymphocele, requiring percutaneous drainage in five cases. Eight patients (11%) had limited lower limb edema, which was transient in half of them. Two patients (3%) were diagnosed with osteitis pubis, one patient (1%) presented with deep venous thrombosis, and one patient (1%) experienced transient obturator nerve malfunctioning. 4. Discussion At present, nodal staging in PCa is still an unresolved issue. There is consensus that until more accurate and patientspecific alternatives come along, patients at intermediate and high risk for LNI should undergo staging elnd [3]. Although interesting studies have been published about this topic, the ideal anatomic template is still under debate. Several authors have shown that up to two-thirds of patients with LNI have LN+ along the internal iliac vessels, explaining the higher rates of LN+ detected with elnd than with llnd [14]. The predominance for this region was confirmed, as it was affected in 59% of our N+ patients and was the unique metastatic site in 12% of these men. We obtained similar results compared with a large study by Bader and colleagues, who reported 58% and 19% of N+ patients, respectively, when addressing the same question [7]. One of the major strengths of this study is that we not only identified drainage patterns for each patient but also have detailed information about the location of the LN+. This specific study design led to the finding that although drainage can be extremely variable among patients, certain regions appear to be more likely to harbor LN+ than others. For example, it is striking that although 19% of SNs are preoperatively detected and 14% intraoperatively resected in the common iliac region, only 3% of LN+ are found there. Compared with the presacral region, for example, these numbers are 7% (scintigraphically detected), 13% (resected), and 9% (LN+), respectively. The power of this study lies not only within the combination of lymphatic drainage information and histopathologic data but also in the assessment of regional LN density. With our standardized LND approach, we aimed to test the hypothesis that different areas have a different hierarchy in the drainage chain. It seems plausible that PCa metastases preferentially disseminate to certain regions, and such information could guide us in dissecting regions with the highest priority. Combining data of all patients, the highest yield is obtained in areas that are covered by an elnd. Next is the presacral region, with 5% LN+, thereby equaling the obturator fossa. Conversely, we demonstrated that removing the common iliac nodes resulted in 1% LN+, so it is questionable whether this procedure was worth the effort and morbidity when aiming at adequate staging. All three patients with common iliac LN+ presented with (lower) pelvic LN+ as well. A mapping study published by Briganti and colleagues also demonstrated an ascending pathway for metastatic PCa cells, with a predilection for certain regions before disseminating to others [15]. These researchers showed that all patients with retroperitoneal LN+ also presented with metastases in the common iliac region and that lymphatic spread can be divided into two main levels: pelvic and common iliac plus retroperitoneal LNs. Based on these results, we want to introduce a new standard elnd template that includes the presacral region in addition to the current elnd template. Although this template is technically challenging, we are convinced that its use could further improve metastases detection rates and result in more accurate staging. We are aware that this template differs from the revisited template proposed by Mattei and colleagues [11]. In their study, SPECT-CT and SPECT-MRI completed with intraoperative gamma-probe detection was used in a group of 34 organ-confined PCa patients. The authors demonstrated that common iliac LNs at least up to the ureteric crossing are primary landing sites and should therefore be included to remove approximately 75% of all potential LN+. Unfortunately, a major drawback was that selecting a pn0 patient population prevented the authors from detecting any predilection regarding dissemination patterns. Some striking similarities and differences are seen between the two studies. The nanocolloid used, as well as the injection template, was identical. Imaging studies and the surgical approach to the SNs were almost identical. Nevertheless, LN region definitions were somehow different. In the Mattei et al. publication, the external iliac and obturator regions were jointly considered as one region; thesameistrueforthepresacralandpararectalareas.in our study, the external iliac region resection included tissue up to the genitofemoral nerve, while in the Mattei publication, the extent was limited to the medial border of the external iliac artery. Our presacral area appears to be overlapping with the Mattei et al. internal iliac, common iliac, and presacral regions. Differences may be minimal: higher resection along the common iliac and more lateral of the external iliac portion and similar extent of presacral/internal iliac or common iliac nodes. In another study assessing the incidence of LN+ in 103 clinically localized PCa patients, Heidenreich and coworkers concluded that dissecting the quaternary lymphatics was unnecessary, since only 3.1% of patients had LN+ in the presacral and common iliac group [16]. Contrarytothese reports, there are also authors who acknowledge the importance of the presacral drainage area. For example, an older study by Golimbu et al. counted the presacral and presciatic region as a first echelon for LN+, with this area involved almost as often as the external iliac obturator group [17]. An N-staging review supports our idea that the presacral region is of greater importance than the common iliac LNs through reporting an incidence of solitary LN+ at one

8 EUROPEAN UROLOGY 63 (2013) single level in 14 15% in the former region and 4% in the latter region [18]. We are aware that our study has certain limitations. First, we did not take into account tumor location within the prostate gland, which has been demonstrated to potentially affect dissemination patterns [19]. Second, some studies suggested that (larger) LN+ can obstruct lymphatic flow [20,21]. Since approximately half of the LN+ were macrometastases, this possibility could have influenced our results. However, 21 of 46 resected macrometastases were defined as SNs and did appear to have taken up a sufficient amount of radionuclide to be detected. Third, the current anatomic model was based on data of patients at clearly elevated risk for LNI and therefore may not be applicable in lower-risk patients. Finally, we have no pathologic information on LNs at the aortic bifurcation or higher. Nevertheless, given the data of Briganti et al, we assume that our number of patients with potential retroperitoneal LN+ is negligible [15]. The present study has important merits because of its prospective design, the highly standardized selnd template for both imaging and surgery, and its standardized pathologic review. The high event rate (46% N+ patients) and single operator approach ensure an extremely valuable dataset. Opinions about the optimal N-staging template differ, and although we also have therapeutic intent, the question of whether LND translates into a therapeutic benefit remains unanswered [5]. Therefore, we focused on defining the minimal template for maximal gain as a staging procedure, as well as for potential therapeutic purposes. 5. Conclusions This study is the first to provide a fundamental insight into the hierarchic pattern of lymphatic spread, which cannot be discerned by mapping alone. Based on this dataset, it is confirmed that a predilection for the internal iliac region exists. Although elnd would have correctly staged the majority of patients, we suggest adding the presacral LNs to the standard elnd to obtain a minimal template with maximal gain. Author contributions: Steven Joniau had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Joniau, Lerut, Oyen, Ameye, Haustermans, Van Poppel. Acquisition of data: Joniau, Van den Bergh, Budiharto. Analysis and interpretation of data: Van den Bergh, Joniau, Deroose, Oyen, Haustermans, Van Poppel. Drafting of the manuscript: Van den Bergh, Joniau. Critical revision of the manuscript for important intellectual content: Haustermans. Statistical analysis: Van den Bergh, Bogaerts. Obtaining funding: Joniau, Lerut, Oyen, Ameye, Haustermans, Van Poppel. Administrative, technical, or material support: Van den Bergh, Budiharto. Supervision: Haustermans, Van Poppel. Other (specify): None. Financial disclosures: Steven Joniau certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None. Funding/Support and role of the sponsor: This work was supported through a research grant of IWT Institute for the Promotion of Innovation by Science and Technology in Flanders IWT TBM References [1] Gervasi LA, Mata J, Easley JD, et al. Prognostic significance of lymph nodal metastases in prostate cancer. J Urol 1989;142: [2] Briganti A, Blute ML, Eastham JH, et al. Pelvic lymph node dissection in prostate cancer. Eur Urol 2009;55: [3] Heidenreich A, Bellmunt J, Bolla M, et al. EAU guidelines on prostate cancer, I: screening, diagnosis, and treatment of clinically localised disease. Eur Urol 2011;59: [4] Hovels AM, Heesakkers RA, Adang EM, et al. The diagnostic accuracy of CT and MRI in the staging of pelvic lymph nodes in patients with prostate cancer: a meta-analysis. Clin Radiol 2008;63: [5] Schumacher MC, Burkhard FC, Thalmann GN, Fleischmann A, Studer UE. Good outcome for patients with few lymph node metastases after radical retropubic prostatectomy. Eur Urol 2008;54: [6] Hyndman ME, Mullins JK, Pavlovich CP. Pelvic node dissection in prostate cancer: extended, limited, or not at all? Curr Opin Urol 2010;20: [7] Bader P, Burkhard FC, Markwalder R, Studer UE. Is a limited lymph node dissection an adequate staging procedure for prostate cancer? J Urol 2002;168: [8] Heidenreich A, Ohlmann CH, Polyakov S. Anatomical extent of pelvic lymphadenectomy in patients undergoing radical prostatectomy. Eur Urol 2007;52: [9] Touijer K, Rabbani F, Otero JR, et al. Standard versus limited pelvic lymph node dissection for prostate cancer in patients with a predicted probability of nodal metastasis greater than 1%. J Urol 2007;178: [10] Ganswindt U, Schilling D, Muller AC, Bares R, Bartenstein P, Belka C. Distribution of prostate sentinel nodes: a SPECT-derived anatomic atlas. Int J Radiat Oncol Biol Phys 2011;79: [11] Mattei A, Fuechsel FG, Bhatta DN, et al. The template of the primary lymphatic landing sites of the prostate should be revisited: results of a multimodality mapping study. Eur Urol 2008;53: [12] 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: [13] Budiharto T, Joniau S, Lerut E, et al. Prospective evaluation of 11C-choline positron emission tomography/computed tomography and diffusion-weighted magnetic resonance imaging for the nodal staging of prostate cancer with a high risk of lymph node metastases. Eur Urol 2011;60: [14] Burkhard FC, Studer UE. The role of lymphadenectomy in high risk prostate cancer. World J Urol 2008;26: [15] Briganti A, Suardi N, Capogrosso P, et al. Lymphatic spread of nodal metastases in high-risk prostate cancer: the ascending pathway from the pelvis to the retroperitoneum. Prostate 2012;72: [16] Heidenreich A, Varga Z, Von KR. Extended pelvic lymphadenectomy in patients undergoing radical prostatectomy: high incidence of lymph node metastasis. J Urol 2002;167:

9 458 EUROPEAN UROLOGY 63 (2013) [17] Golimbu M, Morales P, Al-Askari S, Brown J. Extended pelvic lymphadenectomy for prostatic cancer. J Urol 1979;121: [18] Malmstrom PU. Lymph node staging in prostatic carcinoma revisited. Acta Oncol 2005;44: [19] Tokuda Y, Carlino LJ, Gopalan A, et al. Prostate cancer topography and patterns of lymph node metastasis. Am J Surg Pathol 2010; 34: [20] Holl G, Dorn R, Wengenmair H, Weckermann D, Sciuk J. Validation of sentinel lymph node dissection in prostate cancer: experience in more than 2,000 patients. Eur J Nucl Med Mol Imaging 2009;36: [21] Weckermann D, Dorn R, Holl G, Wagner T, Harzmann R. Limitations of radioguided surgery in high-risk prostate cancer. Eur Urol 2007; 51: