The Impact of the Extent of Lymphadenectomy on Oncologic Outcomes in Patients Undergoing Radical Cystectomy for Bladder Cancer: a Systematic Review

EUROPEAN UROLOGY 66 (2014) 1065–1077 available at www.sciencedirect.com journal homepage: www.europeanurology.com

Review – Bladder Cancer The Impact of the Extent of Lymphadenectomy on Oncologic Outcomes in Patients Undergoing Radical Cystectomy for Bladder Cancer: A Systematic Review

Harman M. Bruins a,*, Erik Veskimae b, Virginia Hernandez c, Mari Imamura d, Molly M. Neuberger e, Philip Dahm e,f, Fiona Stewart d, Thomas B. Lam d, James N’Dow d, Antoine G. van der Heijden a, Eva Compe´rat g, Nigel C. Cowan h, Maria De Santis i, Georgios Gakis j, Thierry Lebret k, Maria J. Ribal l, Amir Sherif m, J. Alfred Witjes a a Department of Urology, Radboud University Medical Center, Nijmegen, The Netherlands; b Department of Urology, Tampere University Hospital, Tampere, Finland; c Department of Urology, Hospital Universitario Fundacio´n Alcorco´n, Madrid, Spain; d Academic Urology Unit, University of Aberdeen, Scotland, UK; e Department of Urology, University of Florida, Gainesville, FL, USA; f Malcom Randall Veterans Affairs Medical Center, Gainesville, FL, USA; g Department of Pathology, Groupe Hospitalier Pitie´—Salpeˆtrie`re, Paris, France; h Department of Radiology, Queen Alexandra Hospital, Portsmouth, UK; i 3rd Medical Department/LBI-ACR VIEnna—LBCTO and ACR-ITR VIEnna, Kaiser Franz Josef Spital, Vienna, Austria; j Department of Urology, Eberhard-Karls University, Tu¨bingen, Germany; k Department of Urology, Foch Hospital, Suresnes, France; l Department of Urology, Hospital Clinic, University of Barcelona, Barcelona, Spain; m Department of Surgical and Perioperative Science, Umea˚ University, Umea˚, Sweden

Article info Abstract

Article history: Context: Controversy exists regarding the therapeutic value of lymphadenectomy (LND) in patients Accepted May 21, 2014 undergoing radical cystectomy (RC) for muscle-invasive bladder cancer (MIBC). Objective: To systematically review the relevant literature assessing the impact of LND on oncologic and perioperative outcomes in patients undergoing RC for MIBC. Keywords: Evidence acquisition: Medline, Medline In-Process, Embase, the Cochrane Central Register of Con- trolled Trials, and the Latin American and Caribbean Center on Health Sciences Information (LILACS) Bladder neoplasms were searched up to December 2013. Comparative studies reporting on no LND, limited LND (L-LND), Radical cystectomy standard LND (S-LND), extended LND (E-LND), superextended LND (SE-LND), and oncologic and Lymphadenectomy perioperative outcomes were included. Risk-of-bias and confounding assessments were performed. Evidence synthesis: Twenty-three studies reporting on 19 793 patients were included. All but one Lymph node dissection study were retrospective. Planned meta-analyses were not possible because of study heterogeneity; Standard therefore, data were synthesized narratively. There were high risks of bias and confounding across most studies as well as extreme heterogeneity in the definition of the anatomic boundaries of LND extended templates. All seven studies comparing LND with no LND favored LND in terms of better oncologic or superextended dissection outcomes. Seven of 14 studies comparing (super)extended LND with L-LND or S-LND reported a Oncologic outcomes beneficial outcome for (super)extended LND in at least a subset of patients. No difference in outcome was reported in two studies comparing E-LND and S-LND. The comparative harms of different extents of LND remain unclear. Conclusions: Although the quality of the data was poor, the available evidence indicates that any kind of LND is advantageous over no LND. Similarly, E-LND appears to be superior to lesser degrees of dissection, while SE-LND offered no additional benefits. It is hoped that data from ongoing randomized clinical trials will clarify remaining uncertainties. Patient summary: The current literature suggests that removal of lymph nodes in bladder cancer surgery is beneficial and might result in better outcomes in terms of prolonging survival; however, the quality of the available studies is poor, and high-quality studies are needed. # 2014 European Association of Urology. Published by Elsevier B.V. All rights reserved.

* Corresponding author. Radboud University Medical Centre, Department of Urology, Geert Groo- teplein Zuid 10 (659), P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. Tel. +31 24 361 37 35; Fax: +31 24 354 10 31. E-mail address: [email protected] (H.M. Bruins). http://dx.doi.org/10.1016/j.eururo.2014.05.031 0302-2838/# 2014 European Association of Urology. Published by Elsevier B.V. All rights reserved. 1066 EUROPEAN UROLOGY 66 (2014) 1065–1077

1. Introduction observational studies with a comparator arm, and retrospective comparative studies. Registry or database studies Lymphadenectomy (LND) combined with radical cystec- were also eligible if the analysis was clearly structured as a tomy (RC) is considered the standard of care for patients comparison between control and intervention groups. with muscle-invasive bladder cancer (MIBC). Up to 25% of Studies with no comparator group (eg, single-arm case patients harbor lymph node (LN) metastases at the time of series), noneffectiveness studies (eg, nomogram studies), RC, and the staging role of LND is unequivocal. In 1982, reviews, and studies with <10 patients in each arm were Skinner [1] was the first to report long-term survival in excluded. LN-positive patients undergoing RC and LND without The study population was limited to patients with systemic treatment. The therapeutic value of LND, however, localized muscle-invasive urothelial or squamous cell remains a topic of continuous debate. While the results of carcinoma of the bladder (cT2–4 N0M0). Studies including two ongoing randomized clinical trials (RCTs) evaluating predominantly patients with variant histology other than the impact of different LND templates on survival are squamous cell carcinoma were excluded because of low awaited, the current evidence base remains uncertain with incidence and the potentially different biologic behavior of regard to the true benefits and harms of LND. In this study these cancers. Clinical staging was preferred, but if it was we systematically reviewed the available literature to not reported, staging based on RC specimen was accepted. evaluate the impact of the extent of LND on survival and Studies with mixed populations (eg, cTa, cTis, cT1) were perioperative outcomes in patients undergoing RC for MIBC. retained for consideration for inclusion if there were no studies that included patients with MIBC exclusively. 2. Evidence acquisition Studies including patients who underwent neoadjuvant or adjuvant treatment were also retained. 2.1. Search strategy The types of interventions included LND undertaken during RC for bladder cancer (BCa). Because of the expected The review was performed in accordance with the Preferred heterogeneity in defining the extent of LND across studies, Reporting Items for Systematic Reviews and Meta-analyses the extent of LND was determined a priori based on statement and principles outlined in the Cochrane Handbook discussion in an expert panel (EAU Working Group on MIBC) for Systematic Reviews of Interventions [2,3]. Highly sensitive and was categorized as follows: (1) limited LND (L-LND): electronic searches were conducted to identify all reports of LND confined to the obturator and/or perivesical fossa only; RCTs or nonrandomized comparative studies (NRCSs) (2) standard LND (S-LND): LND performed up to the assessing LND in patients undergoing RC for MIBC. The common iliac arteries; (3) extended LND (E-LND): LND searches were not limited by language or publication date. performed up to the proximal boundary of the crossing of The databases searched were Medline (1946 to December the common iliac vessels with the ureters or the aortic 2013), Medline In-Process (December 20, 2013), Embase bifurcation, with or without the presacral LNs; and (4) (1974 to December 2013), the Cochrane Central Register of superextended LND (SE-LND): LND performed up to the Controlled Trials (The Cochrane Library, Issue 8, 2013), and proximal boundary of the inferior mesenteric artery. The the Latin American and Caribbean Center on Health primary outcome was overall survival (OS); secondary Sciences Information (LILACS; December 2013). The data- outcomes included recurrence-free survival (RFS), disease- base search was complemented by additional sources, free survival (DFS), progression-free survival, cancer- including the reference lists of included studies, which were specific survival, and perioperative outcomes (eg, operative hand searched, and additional reports identified by an time, blood loss, lymphocele). expert panel (European Association of Urology [EAU] Working Group on MIBC). Ongoing trials were identified 2.3. Assessment of risks of bias at ClinicalTrials.gov. The full search strategy is presented in Supplement 1. Two reviewers independently assessed the risk of bias (RoB) Two reviewers independently screened titles and of individual studies. Any disagreement was resolved by abstracts of all citations identified by the search strategies. discussion or reference to a third reviewer. The standard Full-text copies of all potentially relevant reports were Cochrane Collaboration RoB tool [4] was used to assess the obtained and independently assessed by the reviewers to RoB in RCTs, while for NRCSs, the RoB tool recommended by determine whether they met the predefined inclusion the Cochrane Non-Randomised Studies Methods Group was criteria. Any disagreements were resolved by consensus or used [5,6]. In addition, for NRCSs, the main confounders arbitration by a third person. A data extraction form was were identified a priori based on a study by Palmer et al. [7]. developed specifically for this assessment to collect In this study, a survey among BCa experts was performed to information on study design, characteristics of participants, identify and rank potential confounding variables and characteristics of interventions, and outcome measures. defining thresholds for imbalance for these variables. The main confounders identified are summarized in Figure 1. 2.2. Inclusion and exclusion criteria Each confounder was assessed according to whether it had been considered by the authors, whether the confounder The inclusion criterion was comparative studies only, and was balanced across the groups, and the degree to which these studies included RCTs, prospective NRCSs, prospective adjustment had been made for the confounder [7]. The risk EUROPEAN UROLOGY 66 (2014) 1065–1077 1067 [(Fig._1)TD$IG]

Fig. 1 – Risk-of-bias assessment using the Cochrane risk-of-bias tool. ASA = American Society of Anesthesiologists. [TD$INLE] = High risk of bias. [TD$INLE] = Low risk of bias; the variable was either balanced between the groups (assessed as no or small imbalance) or was adjusted for in the analysis. [TD$INLE] = Unclear risk of bias. 1068 EUROPEAN UROLOGY 66 (2014) 1065–1077 [(Fig._2)TD$IG]

Records identified through database searching n = 1897 Idenﬁcaon

Abstracts screened n = 1897

Abstracts excluded n = 1859

Screening Full-text articles assessed for eligibility

Articles through other sources:

- Hand search/reference lists n = 1

Full-text articles assessed for eligibility Eligibility Excluded full-text articles n = 16

Reasons for exclusion: - No comparative study (n = 4) -Salvage cystectomy series (n = 1) - Editorial (n = 1) Included studies - Duplicate studies (n = 10)

Included n = 23

Fig. 2 – Flow diagram of studies identified, excluded, and included.

of confounding bias was considered to be high if the 3. Evidence synthesis confounder was not described/considered, was imbalanced among the groups, or was not adjusted for in the statistical 3.1. Quantity of evidence identified and characteristics of analysis. Review Manager 5.2 was used to present these included studies results (Fig. 1) [8]. A total of 1897 abstracts were identified by the search 2.4. Data analysis (Fig. 2). Of these abstracts, 38 were selected for full-text screening. One additional study was identified through A narrative synthesis was performed [9]. Descriptive reference searching. After full-text screening, 23 studies statistics were used to summarize baseline characteristics met the inclusion criteria [10–32]. Seven studies were data. For continuous outcomes, data were summarized reported only in the form of conference meeting abstracts, using mean (plus or minus standard deviation, if available) while 16 studies were reported in full-text papers. With one and median (plus or minus interquartile range, if available); exception, all studies were retrospective comparative for categorical outcomes, data were summarized using studies. Sixteen studies were single-center studies, of proportions. For summarizing outcome data, categorical which eight studies used a historical cohort as the control outcomes were presented as proportions at 5- and 10-yr group, and seven studies were multicenter studies. time points following surgery based on crude point estimates as reported by authors, with level of significance 3.2. Risk-of-bias and confounding assessment of included set at 5%. Outcomes at other time points were narratively studies described. For time-to-event data reported by authors using univariable or multivariable Cox regression analysis, data RoB and confounding assessment for each of the individual were summarized as hazard ratios and 95% confidence studies was performed, and the results are presented intervals. in Figure 1. Because of the retrospective design in 22 of Table 1 – (a) Descriptive outcomes and (b) oncologic outcomes in patients undergoing no lymphadenectomy (LND) versus any LND

LE Type of LND Number of patients Follow-up duration pN+, no. (%) pT stage, no. (%) Neoadjuvant or (IQR/range) adjuvant therapy, no. (%)

CI C ICICIC ICI

Liu et al., 3 No LND Any 334 1606 NR N/A NR NR NR 2011 [10] Isaka et al., 4 No LND Standard 15 80 40 mo 40 mo N/A 9 (11) pT2: 10 (67) pT2: 55 (69) NR Neoadjuvant 1985 [11] (range: (range: >pT2: 5 (33) >pT2: 25 (31) rad: 58 (72) 5-143) 2-110)

Miyakawa 4 No LND Standard 45 65* NR N/A NR cT2: 6 (13) cT2: 26 (37) NR 1065–1077 (2014) 66 UROLOGY EUROPEAN et al., 1985 [12] >cT2: 24 (53) >cT2: 38 (54) cTx: 15 (33) cTx: 6 (9) Yuasa 4 No LND Limited 8 22 Minimum 1 yr N/A NR NR NR et al., 1988 [13] Abdollah 3 No LND Any 2789 8394 NR N/A 2182 (26) pT2: pT2: 4157 (49) NR et al., 2012 [14] 1787 (64) >pT2: 4237 (51) >pT2: 1002 (36) Zhang 4 No LND Any 24 63 NR N/A NR NR NR et al., 2013 [15] Brunocilla 3Noor Standard 116 94 Mean: 59.2 mo 26 (22) 28 (30) NR Excluded et al., 2013 [16] limited -> 19 No LND (range 1-171) LND -> 97 L-LND No or Extended/ 116 62 Mean: 59.2 mo 26 (22) 19 (31) NR Excluded limited superextended -> 19 No LND -> 39 E-LND (range 1-171) LND -> 97 L-LND -> 23 SE-LND * 70 patients were included of whom 65 patients underwent no LND. C = control group; E-LND = extended LND; I = intervention group; IQR = interquartile range; LE = level of evidence; L-LND = limited LND; LND = lymphadenectomy; N/A = not applicable; NR = not reported; rad = radiotherapy; S-LND = standard LND; SE-LND = superextended LND.

LND comparator OS (%) RFS/DFS (%) CSS/DSS (%) Multivariable Cox regression analysis p value 5 yr 10 yr 5 yr 10 yr 5 yr 10 yr HR 95% CI Liu et al. [10] No LND NR NR NR NR NR NR NR NR NR Any LND NR NR NR NR NR NR NR NR NR Isaka et al. [11] No LND 25.0 NR NR NR NR NR NR NR NR Standard 64.0, p < 0.05 NR NR NR NR NR NR NR NR Miyakawa et al. [12] No LND 35.0 NR NR NR NR NR NR NR NR Standard 66.0, p < 0.05 NR NR NR NR NR NR NR NR Yuasa et al. [13] No LND 50.0 NR NR NR NR NR NR NR NR Limited 67.8, p < 0.01 NR NR NR NR NR NR NR NR 1069 1070 EUROPEAN UROLOGY 66 (2014) 1065–1077

23 studies, there was high or unclear RoB across all domains. value

0.01 0.05 0.05 0.01 The issue of confounding was also poorly addressed by the p < < < < NS NS majority of studies, as it was unclear in most studies if any of the confounding factors had been considered, either pro- spectively or retrospectively through statistical adjustment.

1.24–1.44 1.16–3.79 1.18–2.17 1.47–1.91 1.01–1.33 0.96–1.28 3.3. Results of comparisons of interventions

** 3.3.1. No lymph node dissection versus lymph node dissection 3.3.1.1. Study characteristics. A total of seven studies comparing LND with no LND were identified, including a total of 13 833 patients (Table 1a) [10–16]. The intervention differed among the studies and included any LND [10,14,15], L-LND [13],S- LND [11,12,16], E-LND [16],orSE-LND[16]. All patients, 1.33 pTa–pTis, 2.09 pT1, 1.60 pT2, 1.68 pT3, 1.15 pT4, 1.11 1.00 (reference)CSS Total cohort, 0.99CSS Total cohort, 0.46 – 0.55–1.35 0.37–0.89 0.49 – 0.036 3.3.1.2. Oncologic outcomes. Table 1b summarizes the oncologic outcomes comparing no LND versus any LND. All studies reported a benefit for LND in at least one oncologic 0.01 1.00 (reference) – – < = 0.010 = 0.010 outcome. Liu et al. [10] did not report any numerical data p p p enectomy; NR = not reported; NS = not signiﬁcant; OS = overall survival; but stated that LND was associated with improved OS and 35 75, 75, Æ Æ Æ 52.5 Cancer-speciﬁc mortality 57.5, DFS in pT1 patients only compared with no LND.

3.3.1.3. Perioperative outcomes. No studies reported on periop- 0.01 erative outcomes. < = 0.010 = 0.010 p p p 58 61, 50 60, 78,

Æ Æ Æ Æ Æ 3.3.2. Limited lymph node dissection versus standard lymph node dissection No studies were identified for comparison of limited LN dissection and standard LN dissection.

3.3.3. Limited lymph node dissection versus (super)extended lymph

NR NR NR NR node dissection 3.3.3.1. Study characteristics. Five studies addressed this ques- * tion involving a total of 1 394 patients (Table 2a) [17–21].

0.01 Brossner et al. [21] focused on perioperative outcomes. < p

* Bostrom et al. [19] compared L-LND with E-LND; however, an unknown number of patients in the E-LND group 27.2 34.1, underwent SE-LND, and >50% of patients in the

* L-LND group did not undergo LND at all. 0.01 < p

5 yr 10 yr 5 yr 10 yr 5 yr 10 yr3.3.3.2. Oncologic HR outcomes. Table 95% CI 2b summarizes the oncol- *

41 47, ogic outcomes comparing L-LND with E/SE-LND. Of the five Æ Æ studies included, three studies reported improvement of at least one oncologic outcome for E/SE-LND [18–20]. Brossner et al. [21] did not report oncologic outcomes, while Hori et al. [17] found no statistically significant difference in oncologic outcomes for L-LND and E-LND performing univariable analysis.

LND comparator OS (%) RFS/DFS (%) CSS/DSS (%)3.3.3.3. Perioperative Multivariable outcomes. Cox regression analysis Jensen et al. [20] reported no prolonged operative time for E-LND compared with L-LND 1. Standard2. Extended/superextended NR NR NR NR NR NR NR NR No LNDAny LNDNo/limited NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR No LND Any LND (mean: 306 vs 302 min; p = 0.92). Brossner et al. [21],

) however, reported prolonged operative time for SE-LND compared with L-LND (median: 330 vs 277 min; p < 0.01). [16] [14] This study reported no differences in number of blood units [15]

Continued transfused (1.15 vs 0.38, respectively; p = 0.37), lympho-

symbol indicates that the Kaplan-Meier curve was used to estimate the percentage. celes (none in both groups), 30-d complication rate (11% vs Æ

Adjusting for age, gender, race tumor grade, and year of cystectomy. 9%, respectively; p = 0.28), or 30-d mortality (three vs one Overall mortality. Table 1 ( More details are availableCI in = the confidence full table interval; online. CSS = cancer-specific survival; DFS = disease-free survival; DSS = disease-specific survival; HR = hazard ratio; LND = lymphad RFS = recurrence-free survival. The Zhang et al. Brunocilla et al. Abdollah et al. * ** B. event, respectively; p = 0.57) [21]. Table 2 – (a) Descriptive outcomes and (b) oncologic outcomes in patients undergoing limited lymphadenectomy (LND) versus (super)extended LND

LE Type of LND Number of patients Follow-up duration (IQR/range) pN+, no. (%) pT stage, no. (%) Neoadjuvant or adjuvant therapy

CI C I C ICIC ICI

Hori et al., 4 Limited Extended 53 47 NR 10 (19) 14 (30) NR NR 2013 [17] Holmer et al., 4 Limited Extended 69 101 Median: Median: 38 mo 12 (17) 38 (38) pT2: 46 (67) pT2: 53 (52) Neoadj: 2009 [18] 94 mo (range: (range: 13–70) >pT2: 23 (33) >pT2: 48 (48) Excluded 61–122) Adj chemo: 13% vs 16% Bostrom et al., 4 Limited Extended or 563 NR NR (7) NR (26) NR Neoadj: 2011 [19] superextended Excluded Adj chemo: 1% vs 21% Jensen et al., 4 Limited Standard, 204 265 Median: 113 mo Median: 45 mo 43 (21) 61 (23) pT2: 112 (55) pT2: 160 (60) Excluded 2012 [20] extended or -> 202 L-LND -> 170 SE-LND (range: 86–143) (range: 24–84) >pT2: 92 (45) >pT2: 105 (40)

superextended -> 2 no LND -> 46 E-LND 1065–1077 (2014) 66 UROLOGY EUROPEAN -> 49 S-LND Brossner et al., 3 Limited Superextended 46 46 30 d 10 (22) 18 (39) pT3a: 24 (52) pT3a 28 (61) NR 2004 [21] pT3b–pT4: 22 (48) pT3b–pT4b 18 (39)

C = control group; E-LND = extended LND; I = intervention group; IQR = interquartile range; LE = Level of evidence; LN = lymph node; LND = lymphadenectomy; L-LND = limited LND; NR = not reported; S-LND = standard LND; SE-LND = superextended LND.

LND comparator OS (%) RFS/DFS (%) CSS/DSS (%) Multivariable Cox regression analysis

5 yr 10 yr 5 yr 10 yr 5 yr 10 yr HR 95% CI p value

Hori et al. [17] Limited 61 NR NR NR 73 NR NR NR NR Extended 47, p = 0.65 NR NRÆ66 NRÆ65 50, p = 0.27Æ66 NRNo cases NR 1.00 (reference) NR – NR – Æ75, p = 0.27 No cases, p = 0.27 Æ73, p = 0.45 No cases Disease-speciﬁc mortality * Holmer et al. [18] LimitedExtended NR NR NR NR All patients, 0.81 0.44–1.47 0.48 pT2, 0.80 0.28–2.32 0.68 >pT2, 0.54 0.26–1.14 0.10 * RFS All patients, 0.73 0.41–1.30 0.28 pT2, 0.88 0.34–2.30 0.79 >pT2, 0.42 0.20–0.88 0.022 Bostrom et al. [19] Limited NR NR NR NR NR NR 1.00 (reference) - - Extended or superextended NR NR NR NR NR NR DSS All patients, 0.53 0.31–0.93 0.026 Jensen et al. [20] Limited All patients: 55 NR All patients: 62 NR All patients: 67 NR 1.00 (reference) - - pT2: 77 pT2: 83 pT2: 88 >pT2: 43 >pT2: 54 >pT2: 62 LN negative: 66 LN negative: 75 LN negative: 80 LN positive: 12 LN positive: 8 LN positive: 14 Standard, extended, or superextended All patients: 67, p < 0.05 NR All patients: 64, NS NR All patients: 72, p < 0.05 NR RFS pT2: 77, NS pT2: 80, NS pT2: 88, NS All patients, 0.90 0.65–1.24 0.51 >pT2: 71, p < 0.05 >pT2: 61, NS >pT2: 62, p < 0.05 DSS LN negative: 75, NS LN negative: 74, NS LN negative: 80, NS All patients, 0.71 0.50–1.01 0.06 LN positive: 38, p < 0.05 LN positive: 29, p < 0.05 LN positive: 14, p < 0.05 OS All patients, 0.60 0.45–0.81 < 0.01 Limited NR NR NR NR NR NR NR NR NR Brossner et al. [21] Superextended NR NR NR NR NR NR NR NR NR

More details are available in the full table online. CI = confidence interval; CSS = cancer-specific survival; DFS = disease-free survival; DSS = disease-specific survival; HR = hazard ratio; LN = lymph node; LND = lymphadenectomy; NR = not reported; NS = not significant; OS = overall survival; RFS = recurrence-free survival. TheÆ symbol indicates that the Kaplan-Meier curve was used to estimate the percentage. * Adjusting for sex, age, and adjuvant chemotherapy. 1071 1072

Table 3 – (a) Descriptive outcomes and (b) oncologic outcomes in patients undergoing standard lymphadenectomy (LND) versus (super)extended LND

LE Type of LND Number of Follow-up duration (IQR/range) pN+, no. (%) pT stage, no. (%) Neoadjuvant or patients adjuvant therapy

CICIC ICICI CI

Abd-El-Latif 3 Standard Extended or 122 199 Median: 24 mo (IQR 12–36) Excluded NR Neoadj: 17 (5%) et al., 2011 [22] superextended Adj: NR Abd-El-Latif et al., 3 Standard Extended or 183 240 Median: 24 mo NR NR NR 2012 [23] superextended Dharaskar et al., 4 Standard Extended 51 27 a. Median: 24 mo Median: 6 mo 13 (26) 7 (26) pT2: 32 (63) pT2: 20 (74) NR 2011 [24] (a. en bloc; (range: 0–49) (range: 0–37) >pT2: 19 (37) >pT2: 7 (26) b. LN packets) b. Median: 14 mo

(range: 0–43) 1065–1077 (2014) 66 UROLOGY EUROPEAN Finelli et al., 3 Standard Extended 11 11 Mean: 11 mo 3 (27) 3 (27) pT2: 5 (45) pT2: 9 (82) Neoadj: NR 2004 [25] (range: 2–43) >pT2: 6 (55) >pT2: 2 (18) Adj chemo: 2 (9%) Poulsen et al., 4 Standard Extended 68 126 Median: 5.14 yr Median: 1.96 yr 15 (22) 35 (28) pT2: 34 (50) pT2: 55 (43) Excluded 1998 [26] (range: 0.17–7.94) (range: 0.29–4.73) >pT2: 33 (49) >pT2: 69 (55) pTx: 1 (1) pTx: 2 (2) Simone et al., 3 Standard Extended 584 349 NR 187 (32) 107 (31) pT2: 271 (46) pT2: 192 (55) Neoadj: Excluded 2012 [27] >pT2: 313 (54) >pT2: 157 (45) Adj chemo: 101 (11%) pT2: 200 (60) pT2: 150 (47.0) Adj rad:Neoadj: 8 (0.8%) Excluded Dhar et al., 3 Standard Extended 336 322 For RFS: 25 mo For RFS: 40 mo 44 (13) 83 (26) >pT2: 136 (40) >pT2: 172 (53.0) Adj: NR 2008 [28] (range: 1.1–166) (range: 1–229) For OS: 36 mo For OS: 51 mo Abol-Enein et al., 3 Standard Superextended 200 200(range: 50.2 1.1–166) mo (IQR: 69.0)(range: for 1–229)patients alive at 48 (24) 48 (24) pT2: 61 (31) pT2: 70 (35) 0 2011 [29] last follow–up >pT2: 139 (70) >pT2: 129 (65) Wang, 2013 [30] 3 Standard Extended 42 33 NR NR NR Excluded Adj = adjuvant; C = control group; chemo = chemotherapy; I = intervention group; IQR = interquartile range; LE = level of evidence; LN = lymph node; LND = lymphadenectomy; neoadj = neoadjuvant; NR = not reported; OS = overall survival; rad = radiotherapy; RFS = recurrence-free survival.

LND comparator OS (%) RFS/DFS (%) CSS/DSS (%) Multivariable Cox regression analysis

5 yr 10 yr 5 yr 10 yr 5 yr 10 yr HR 95% CI p value

Abd El-Latif et al. [22] Standard NR NR NR NR NR NR 1.00 (reference) - - Extended or superextended NR NR NR NR NR NR OS All patients, 2.50 0.98–6.5 NR CSS All patients, 2.00 0.4–11.1 NR RFS Æ58 NR NR NR NR NR NRAll patients, 2.60 NR0.7–9.9 NRNR Abd El-Latif et al. [23] ExtendedStandard or superextended Æ40, p = 0.14 NR NR NR NR NR NR NR NR Dharaskar et al. [24] Standard No cases No cases NR NR NR NR NR NR NR Extended No cases No cases NR NR NR NR NR NR NR Finelli et al. [25] Standard NR NR NR NR NR NR NR NR NR Extended NR NR NR NR NR NR NR NR NR Poulsen et al. [26] Standard NR NR All patients: 56 NR NR NR NR NR NR pT3: ** 85 >pT3: 27 pN0: 71 pN+: 7 Extended NR NR All patients: 62, p = 0.33 NR NR NR NR NR NR ** pT3: 64, p < 0.02 >pT3: 39, p = 0.87 pN0: 90, p < 0.02 pN+: 24, p = 0.15 Simone et al. [27] Standard NR NR 42.6 NR 50.9 NR 1.00 (reference) – – Extended NR NR 63.1, p < 0.01 NR 68.8, p < 0.01 NR DFS * 0.40–0.65 <0.01 All patients, 0.51 pT2, 0.41 0.22-0.78 0.004 pT3, 0.55 0.40–0.77 0.001 PT4, 0.33 0.20–0.55 <0.001 pN0, 0.63 0.4– 0.91 0.011 pN1, 0.19 0.08–0.49 <0.001 0.24–0.67 <0.001 pN2, 0.40 1065–1077 (2014) 66 UROLOGY EUROPEAN CSS * All patients, 0.56 0.42–0.73 <0.01 pT2, 0.40 0.19–0.83 0.011 pT3, 0.65 0.44–0.94 0.019 PT4, 0.33 0.19–0.58 <0.001 pN0, 0.54 0.30–0.95 0.034 pN1, 0.15 0.05–0.41 <0.001 pN2, 0.45 0.25–0.80 <0.001 Dhar et al. [28] Standard pT2N0: 68 NR pT2N0: 67 NR NR NR NR NR NR pT2N0–2: 64 pT2N0–2: 63 pT3N0: 26 pT3N0: 23 pT3N0–2: 22 pT3N0–2: 19 Extended pT2N0: 66, p = 0.12 NR pT2N0: 77, p = 0.12 NR NR NR NR NR NR pT2N0–2: 61, p = 0.10 pT2N0–2: 71, p = 0.22 pT3N0: 46, p = 0.002 pT3N0: 57, p < 0.0001 pT3N0–2: 42, p = 0.002 pT3N0–2: 49, p < 0.0001 *** 1.06–1.99 0.02 Abol-Enein et al. [29] Standard NR NR All patients: 54.7 NR NR NR DFS, 1.45 LNÀ: 66.2 LN+: 28.2 Superextended NR NR All patients: 66.6, p = 0.04 NR NR NR 1.00 (ref) - - LNÀ: 72.3, p = 0.24 LN+: 48.0, p = 0.029 Wang [30] Standard NR NR NR NR 59.0 NR NR NR NR Extended NR NR NR NR 77.0, p = 0.08 NR NR NR NR

More details are available in the full table online. CI = confidence interval; CSS = cancer-specific survival; DFS = disease-free survival; DSS = disease-specific survival; HR = hazard ratio; LN = lymph node; LND = lymphadenectomy; NR = not reported; OS = overall survival; RFS = recurrence-free survival. The Æ symbol indicates that the Kaplan-Meier curve was used to estimate the percentage. * HR was recalculated using the same reference group (standard LND) as in the subgroup analysis. ** pT3a indicated organ-confined disease in this study. *** Adjusting for pT stage, pN stage, age, gender, tumor grade, and histology. 1073 1074 EUROPEAN UROLOGY 66 (2014) 1065–1077

3.3.4. Standard lymph node dissection versus (super)extended lymph node dissection value p 3.3.4.1. Study characteristics. Nine studies were identified involving 3104 patients (Table 3a) [22–30]. Four studies used data from the Cleveland Clinic [22,23,25,28]. Abd El-Latif et al. [23] differed from their previous study [22] by extending the study period by 2 yr (2004–2010 vs 2006– 2010). One study specifically looked at the outcomes of laparoscopic LND [25].

3.3.4.2. Oncologic outcomes. Table 3b summarizes the oncologic outcomes of S-LND compared with E/SE-LND, and contradicting results were reported. Four studies noted no difference in oncologic outcomes between S-LND and E-LND [22–24,30], although only one study reported on

data from multivariable analysis [22]. Three studies Excluded Excluded Chemo: 46 (11) Chemo: 195 (35) reported a benefit for E-LND, and one study reported a benefit for SE-LND for at least one oncologic outcome. NRNR NR NR NR NR NR NR NR NR

Subgroup analysis in these studies revealed no consistent y; NR = not reported. subgroup that benefited most from E-LND. For example, pT2: 253 (46) pT3: 301 (54) Poulsen et al. [26] reported an RFS benefit for E-LND in enectomy; NR = not reported; OS = overall survival; RFS = recurrence-free survival. patients with organ-confined disease, while Dhar et al. [28] found an RFS benefit only for patients with >pT2 disease. = 0.75 p pT3: 236 (58) 59 3.3.4.3. Perioperative outcomes. Poulsen et al. [26] reported a 59, NRNR NR NR NR NR NR NR NR NR NR NR Æ Æ lymphocele rate of 1.6% for E-LND and 1.5% for S-LND. One patient in the E-LND group (0.8%) died perioperatively from complications unrelated to LND. Finelli et al. [25], performing laparoscopic LND, reported an estimated increase in = 0.41 operative time, from 30 to 45 min for S-LND to 90 min for = 0.75 p p 114 (28) 195 (35) pT2: 169 (42) 59 64, E-LND (no p value reported). 61 61, Æ Æ Æ Æ

3.3.5. Extended lymph node dissection versus superextended lymph node dissection 3.3.5.1. Study characteristics. Two multi-institutional studies, Median: 10.9 yr (range: 0–24.1) involving 1462 patients, were included (Table 4a) [31,32]. = 0.45 p 42 40, 3.3.5.2. Oncologic outcomes. Table 4b summarizes the oncol- Æ Æ ogic outcomes comparing E-LND with SE-LND. Both studies reported no statistically significant difference in survival Follow-up duration (IQR/range) pN+, no. (%) pT stage, no. (%) Neoadjuvant therapy Adjuvant therapy, no. (%) outcomes between E-LND and SE-LND, irrespective of (range: 0–22.3)

tumor stage or nodal status. = 0.45 p 5 yr 10 yr 5 yr 10 yr 5 yr 10 yr HR 95% CI 54 50, Æ Æ

3.3.5.3. Perioperative outcomes. No studies reporting on peri- patients operative outcomes were identified.

3.4. Discussion

3.4.1. Principal findings To the best of our knowledge, this study represents the most Extended NR NR LND comparator OS (%) RFS/DFS (%) CSS/DSS (%) Multivariable Cox regression analysis SuperextendedExtended NR NR robust literature review focusing on the impact of the Superextended II CCICIC I ICIC CICIC anatomic extent of LND on post-RC oncologic and perioperative outcomes. The findings of this study suggest that in LE Type of LND Number of terms of oncologic outcomes, any extent of LND is better 3 Extended3 Superextended Extended Superextended 503 405 554 Median: 9.9 yr NR NR NR Excluded Excluded [32] than no LND for patients undergoing RC for MIBC. [31] indicates that the Kaplan-Meier curve was used to estimate the percentage. Æ

Additionally, E-LND might improve oncologic outcomes [32] compared with lesser degrees of dissection, although [31] 2011 extending the dissection beyond E-LND is unlikely to 2012 Table 4 – (a) Descriptive outcomes and (b) oncologic outcomes in patients undergoing extended lymphadenectomy (LND) versus superextended LND More details are availableCI in = the confidence full interval; table CSSThe online. = prefix cancer-specific survival; DFS = disease-free survival; DSS = disease-specific survival; HR = hazard ratio; LND = lymphad Zehnder et al., Simone et al., Simone et al. C = control group; chemo = chemotherapy; I = intervention group; IQR = interquartile range; LE = level of evidence; LN = lymph node; LND = lymphadenectom Zehnder et al. A. yield any further benefits. With respect to perioperative B. EUROPEAN UROLOGY 66 (2014) 1065–1077 1075 outcomes, a secondary outcome of this study, SE-LND consequently affect reproducibility [37,40,41]. The LN resulted in increased operative time compared with less count cannot adequately be determined intraoperatively, extended LND templates but does not appear to substan- whereas surgeons can adhere to anatomic templates, tially increase postoperative morbidity. making studies comparing LND templates more clinically relevant. For these reasons, only studies describing ana- 3.4.2. Clinical implications of our study findings tomic templates for the extent of LND were included in our The data in this study support the routine performance of review. In addition, although Tilki et al. [35] described some LND in patients undergoing RC. Whether the reported studies comparing LND templates (references 26, 28, 29, beneficial oncologic outcomes are a result of stage and 32 in Tilki et al.), an additional 19 studies were included migration (the so-called Will Rogers phenomenon), repre- in this study, providing a more comprehensive overview of sent a true therapeutic benefit of LND, or a combination of studies comparing different LND templates. both remains uncertain. There is, however, a clear staging The attempt by Fan et al. [36] to perform a meta-analysis role of LND, as supported by LN mapping studies [33,34]. is noteworthy, but the results of this study should be Thus, in spite of the lack of RCTs, the current evidence base interpreted with caution. Aside from the low-quality is sufficiently convincing to recommend LND for patients studies included in the analysis with its associated bias, undergoing RC for MIBC. differences in the definition of the extent of LND were not While L-LND may contribute to disease staging, per- adjusted for in this study. Reflecting the lack of consensus forming LND outside the true pelvis (ie, S-LND or more on what constitutes an L-LND, S-LND, E-LND, or SE-LND, extended LND) should be considered a potential therapeutic there was significant heterogeneity in the definition intervention, as skip nodal lesions are rare and therefore regarding the extent of LND across studies. To illustrate, unlikely to contribute to disease staging [33,34]. To date, Abol-Enein et al. [29] and Dhar et al. [28] were both however, questions remain about the potential therapeutic classified as E-LND studies, although the proximal bound- value of LND and what extent of LND is the most effective. aries were the inferior mesenteric artery and crossing of the Based on the current data, consisting of retrospective ureter with the common iliac vessels, respectively. For this studies with a significant RoB and confounding, the reason, we chose to define the LND templates a priori and, if evidence base is not strong enough to provide firm necessary, reclassify accordingly if sufficiently large num- recommendations regarding the most optimal extent of bers of studies did not match our chosen definitions. LND. Conversely, these studies are currently the best Although the definitions chosen for each of the LND available evidence and fairly consistently report an templates may not be universally accepted by all clinicians, oncologic benefit for E-LND compared with less extended their use at least allows for a certain degree of standardiza- LND templates. In addition, E-LND appears not to increase tion, which enables a comparison of outcomes among perioperative morbidity. Collectively, there is accumulating different LND templates. evidence that E-LND may be beneficial for patients undergoing RC for MIBC, and can be therefore recom- 3.4.4. Strengths and limitations of the review mended for patients undergoing RC for MIBC. The strength of the current study is the comprehensive literature review that evaluates the impact of the extent of 3.4.3. How does this systematic review compare with other recent LND on post-RC outcomes using a robust and transparent reviews? methodological approach based on Cochrane Review To our knowledge, two systematic reviews on the impor- principles and that incorporates the assessment of RoB tance of LND in BCa have been published [35,36]. Fan et al. and confounding, which are essential in any review [36] performed a systematic review and meta-analysis of involving nonrandomized studies. The search strategy studies comparing E-LND and nonextended LND and its was complemented by additional sources for potentially impact on RFS. The authors concluded that E-LND was important articles, which included an expert panel (EAU associated with improved RFS compared with nonextended Working Group on MIBC). The review was limited to LND. Subgroup analysis revealed that patients with pT3 comparative studies to maintain at least moderate levels of BCa, independent of LN status, benefit from E-LND. Tilki evidence. Throughout the entire review process, peer et al. [35] performed a systematic review only and review was obtained from the expert panel, which concluded that the extent of LND may influence DFS after represents a reference group of international experts. This RC, independent of LN status and pT stage. approach ensured a comprehensive review of the literature The outcomes of our present study are in line with these while maintaining methodological rigor and enabled reviews. However, there are important methodological the authors to put into clinical context the relevance and differences that deserve discussion. Tilki et al. [35] included implication of the review findings. studies using the LN count as a surrogate for the extent of The major limitation of the review is the quality of LND. Although an association between LN count, the extent included studies; except for one prospective study, all of LND, or even post-RC outcomes has been suggested studies were retrospective, nonstandardized, comparative [37–39], using the LN count as a surrogate for the extent of studies with high risks of bias and confounding. In LND has limitations, as acknowledged by the authors. particular, selection bias may have affected clinical out- Differences in surgical technique, sample processing, and comes (eg, cases with apparent nodal disease intraopera- pathologic assessment greatly influence the LN count and tively in which no LND was performed or less extended LND 1076 EUROPEAN UROLOGY 66 (2014) 1065–1077

than anticipated was performed). This review highlights the Cancer Guideline panel and EAU Guidelines Ofﬁce can be found at http:// lack of high-quality and reliable evidence concerning the www.uroweb.org/guidelines/eau-guidelines-board-and-working-panels/ benefits and harms of LND during RC in terms of oncologic ?id=58&gid=39. All coauthors who are not members of the panels (Bruins, and perioperative outcomes. The results, however, are Veskimae, Hernandez, Imamura, Neuberger, Dahm, Stewart) do not have any conﬂicts of interest. supported by the fact that these studies are fairly consistent in reporting an oncologic benefit. Currently, two phase 3 Funding/Support and role of the sponsor: None. RCTs evaluating the impact of different LND templates on survival—one in Germany and one initiated by SWOG (S1011)—are ongoing. The final results of these studies may Appendix A. Supplementary data provide a more definitive answer to some aspects of this important clinical question. Standardization of the LND Supplementary data associated with this article can be templates and surgeon expertise, however, are of critical found, in the online version, at doi:10.1016/j.eur- importance for the success of these trials. uro.2014.05.031.

4. Conclusions References

This systematic review set out to determine the evidence [1] Skinner DG. Management of invasive bladder cancer: a meticulous base in regard to the comparative effectiveness of LND in pelvic node dissection can make a difference. J Urol 1982;128: patients undergoing RC for MIBC in terms of oncologic 34–6. benefits and perioperative outcomes. The findings reveal a [2] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. lack of randomized studies and an evidence base derived BMJ 2009;339:b2535. mainly from retrospective studies with significant risks of [3] Higgins JPT, Green S, editors. Cochrane handbook for systematic bias and confounding. Nevertheless, the data indicate that reviews of interventions, version 5.1.0 (updated March 2011). any form of LND produces more favorable oncologic Cochrane Collaboration Web site. http://www.cochrane.org/hand- outcomes compared with no LND. There was no evidence book. that LND results in increased perioperative adverse events [4] Higgins JP, Altman DG, Gøtzsche PC, et al. The Cochrane Collabora- compared with no LND. In terms of how different extents tion’s tool for assessing risk of bias in randomised trials. BMJ of LND influence outcomes, the findings indicate that E- 2011;343:d5928. LND might be superior to lesser degrees of dissection from [5] Reeves BC, Shea B, Wells GA. Classifying non-randomised studies an oncologic perspective; however, further extending the (NRS) and assessing the risk of bias for a systematic review. Presented at: 18th Cochrane Colloquium; October 18–22, 2010; dissection (eg, SE-LND) is not beneficial. The results of Keystone CO, USA. ongoing RCTs, it is hoped, will clarify the remaining [6] Reeves BC, Deeks JJ, Higgins JPT, Wells GA. Chapter 13. Including uncertainties regarding the role of LND during RC for non-randomized studies. In: Higgins JPT, Green S, editors. Cochrane MIBC. handbook for systematic reviews of interventions, version 5.1.0 (updated March 2011). Cochrane Collaboration Web site. http:// Author contributions: Harman M. Bruins had full access to all the data in www.cochrane.org/handbook. the study and takes responsibility for the integrity of the data and the [7] Palmer DA, MacLennan S, Imamura M, et al. Initial experience with accuracy of the data analysis. a pilot Cochrane tool for applying a web-based survey of content experts to derive criteria for imbalance. Presented at: 19th Study concept and design: Bruins, Veskimae, Hernandez, Imamura, Dahm, Cochrane Colloquium; October 19–22, 2011; Madrid, Spain. Lam, N’Dow, Witjes. [8] Review Manager (RevMan) v.5.2 [computer program]. Copenhagen, Acquisition of data: Bruins, Veskimae, Hernandez, Imamura. Denmark: Nordic Cochrane Centre, Cochrane Collaboration; 2012. Analysis and interpretation of data: Bruins, Veskimae, Hernandez, [9] Schu¨ nemann HJ, Oxman AD, Higgins JPT, Vist GE, Glasziou P, Guyatt Imamura, Lam, Witjes. GH. Chapter 11. Presenting results and ‘‘summary of findings’’ Drafting of the manuscript: Bruins, Veskimae, Hernandez, Imamura, tables. In: Higgins JPT, Green S, editors. Cochrane handbook for Stewart, Lam. systematic reviews of interventions, version 5.1.0 (updated March Critical revision of the manuscript for important intellectual content: Dahm, 2011). Cochrane Collaboration Web site. http://www.cochrane.org/ Neuberger, Lam, N’Dow, van der Heijden, Compe´rat, Cowan, De Santis, handbook. Gakis, Lebret, Ribal, Sherif, Witjes. [10] Liu J, Leppert J, Shinghal R, Gill H, Presti J, Gonzalgo M. Practice Statistical analysis: Bruins, Lam, Dahm. patterns of pelvic lymph node dissection for radical cystectomy Obtaining funding: None. from Veterans Affairs Central Cancer Registry (VACCR). J Urol Administrative, technical, or material support: Stewart, Lam. 2011;185:e562. Supervision: Lam, Dahm, Witjes. [11] Isaka S, Okano T, Sato N, Shimazaki J, Matsuzaki O. Pelvic lymph Other (specify): None. node dissection for invasive bladder cancer. Nihon Hinyokika Financial disclosures: Harman M. Bruins certifies that all conflicts of GakkaiZasshi 1989;80:402–6. interest, including specific financial interests and relationships and [12] Miyakawa M, Oishi K, Okada Y, Takeuchi H, Okada K, Yoshida O. affiliations relevant to the subject matter or materials discussed in the Results of the multidisciplinary treatment of invasive bladder manuscript (eg, employment/affiliation, grants or funding, consultan- cancer. Hinyokika Kiyo 1986;32:1931–9. cies, honoraria, stock ownership or options, expert testimony, royalties, [13] Yuasa M, Yamamoto A, Kawanishi Y, Higa I, Numata A, Imagawa A. or patents filed, received, or pending), are the following: The conflict of Clinical evaluation of total cystectomy for bladder carcinoma: a interest statements of all members of the EAU Muscle Invasive Bladder ten-year experience. Hinyokika Kiyo 1988;34:975–81. EUROPEAN UROLOGY 66 (2014) 1065–1077 1077

[14] Abdollah F, Sun M, Schmitges J, et al. Stage-specific impact of pelvic [28] Dhar NB, Klein EA, Reuther AM, Thalmann GN, Madersbacher S, lymph node dissection on survival in patients with non-metastatic Studer UE. Outcome after radical cystectomy with limited or ex- bladder cancer treated with radical cystectomy. BJU Int 2012;109: tended pelvic lymph node dissection. J Urol 2008;179:873–8. 1147–54. [29] Abol-Enein H, Tilki D, Mosbah A, et al. Does the extent of [15] Zhang F, Wang F, Weng Z. Clinical significance of standard lymphadenectomy in radical cystectomy for bladder cancer influ- lymphadenectomy in radical cystectomy for bladder cancer. ence disease-free survival? A prospective single-center study. Eur J Practical Oncol 2013;28:284–6. Urol 2011;60:572–7. [16] Brunocilla E, Pernetti R, Schiavina R, et al. The number of nodes [30] Wang G. Clinical significance of radical cystectomy with extended removed as well as the template of the dissection is independently lymphadenectomy and influencing factors associated with recurrence correlated to cancer-specific survival after radical cystectomy for of bladder cancer [abstract MP08-19]. Presented at: World Congress of muscle-invasive bladder cancer. Int Urol Nephrol 2013;45:711–9. Endourology; October 22–26, 2013; New Orleans, LA, USA. [17] Hori J, Tamaki G, Kita M, Iwata S, Matsumoto S, Kakizaki H. Survival [31] Simone G, Abol-Enein H, Ferriero M, et al. Extended versus super- impact of the extent of pelvic lymph node dissection in patients extended PLND during radical cystectomy: comparison of two undergoing radical cystectomy for bladder cancer. Eur J Cancer prospective series. J Urol 2012;187(Suppl 4):e708. 2013;49:S1–22. [32] Zehnder P, Studer UE, Skinner EC, et al. Super extended versus [18] Holmer M, Bendahl PO, Davidsson T, Gudjonsson S, Ma˚ nsson W, extended pelvic lymph node dissection in patients undergoing Liedberg F. Extended lymph node dissection in patients with radical cystectomy for bladder cancer: a comparative study. J Urol urothelial cell carcinoma of the bladder: can it make a difference? 2011;186:1261–8. World J Urol 2009;27:521–6. [33] Vazina A, Dugi D, Shariat SF, et al. Stage specific lymph node [19] Bostrom PJ, Mirtti T, Nurmi M, et al. Extended lymphadenectomy metastasis mapping in radical cystectomy specimens. J Urol and chemotherapy offer survival advantage in muscle-invasive 2004;171:1830–4. bladder cancer. J Urol 2011;185:e640. [34] Abol-Enein H, El-Baz M, Abd El-Hameed MA, Abdel-Latif M, [20] Jensen JB, Ulhøi BP, Jensen KM. Extended versus limited lymph Ghoneim MA. Lymph node involvement in patients with bladder node dissection in radical cystectomy: impact on recurrence pat- cancer treated with radical cystectomy: a patho-anatomical tern and survival. World J Urol 2009;27:521–6. study—a single center experience. J Urol 2004;172:1818–21. [21] Brossner C, Pycha A, Toth A, Mian C, Kuber W. Does extended [35] Tilki D, Brausi M, Colombo R, et al. Lymphadenectomy for bladder lymphadenectomy increase the morbidity of radical cystectomy? cancer at the time of radical cystectomy. Eur Urol 2013;64:266–76. BJU Int 2004;93:64–6. [36] Bi L, Huang H, Fan X, et al. Extended versus non-extended pelvic [22] Abd El-Latif A, Miocinovic R, Stephenson AJ, et al. Impact of ex- lymph node dissection and their influence on recurrence-free tended (e) versus standard lymph node dissection (sLND) on post- survival in patients undergoing radical cystectomy for bladder cystectomy survival (PCS) among patients with LN-negative cancer: a systematic review and meta-analysis of comparative urothelial bladder cancer (UBC). J Urol 2011;185(Suppl 4):e759. studies. BJU Int 2014;113:E39–48. [23] Abd El-Latif A, Miocinovic R, Stephenson AJ, et al. Impact of ex- [37] Stein JP, Cai J, Groshen S, et al. Risk factors for patients with pelvic tended versus standard lymph node dissection on overall survival lymph node metastases following radical cystectomy with en bloc among patients with urothelial cancer of the bladder. J Urol pelvic lymphadenectomy: concept of lymph node density. J Urol 2012;187(Suppl 4):e707. 2003;170:35–41. [24] Dharaskar A, Kumar V, Kapoor R, Jain M, Mandhani A. Does ex- [38] Herr HW, Bochner BH, Dalbagni G, et al. Impact of the number of tended lymph node dissection affect the lymph node density and lymph nodes retrieved on outcome in patients with muscle inva- survival after radical cystectomy? Ind J Cancer 2011;48:230–3. sive bladder cancer. J Urol 2002;167:1295–8. [25] Finelli A, Gill IS, Desai MM, Moinzadeh A, Magi-Galluzzi C, Kaouk JH. [39] Leissner J, Hohenfellner R, Thuroff JW, Wolf HK. Lymphadenectomy Laparoscopic extended pelvic lymphadenectomy for bladder can- in patients with transitional cell carcinoma of the urinary bladder; cer: technique and initial outcomes. J Urol 2004;172:1809–12. significance for staging and prognosis. BJU Int 2000;85:817–23. [26] Poulsen AL, Horn T, Steven K. Radical cystectomy: extending the [40] Dorin RP, Daneshmand S, Eisenberg MS, et al. Lymph node dissection limits of pelvic lymph node dissection improves survival for technique is more important than lymph node count in identifying patients with bladder cancer confined to the bladder wall. J Urol nodal metastases in radical cystectomy patients: a comparative 1998;160:2015–9. mapping study. Eur Urol 2011;60:946–52. [27] Simone G, Papalia R, Ferriero M, et al. Stage-specific impact of [41] Meijer RP, Nunnink CJ, Wassenaar AE, et al. Standard lymph node extended versus standard pelvic lymph node dissection in radical dissection for bladder cancer: significant variability in the number cystectomy. Int J Urol 2013;20:390–7. of reported lymph nodes. J Urol 2012;187:446–50.