Extranodal Extension Is a Powerful Prognostic Factor in Bladder Cancer Patients with Lymph Node Metastasis

Transcription

1 EUROPEAN UROLOGY 64 (2013) available at journal homepage: Bladder Cancer Extranodal Extension Is a Powerful Prognostic Factor in Bladder Cancer Patients with Lymph Node Metastasis Harun Fajkovic a,b,y, Eugene K. Cha a,y, Claudio Jeldres c, Brian D. Robinson d, Michael Rink a,e, Evanguelos Xylinas a,f, Thomas F. Chromecki a,g, Eckart Breinl b, Robert S. Svatek h, Gerhard Donner b, Scott T. Tagawa a,i, Derya Tilki j, Patrick J. Bastian j, Pierre I. Karakiewicz c, Bjoern G. Volkmer k, Giacomo Novara l, Abdennabi Joual m, Talia Faison a, Guru Sonpavde n, Siamak Daneshmand o, Yair Lotan p, Douglas S. Scherr a, Shahrokh F. Shariat a,i, * a Department of Urology, Weill Cornell Medical College/New York-Presbyterian Hospital, New York, NY, USA; b Department of Urology, General Hospital Sankt Poelten, Sankt Poelten, Austria; c University of Montreal Health Center, Montreal, QC, Canada; d Departments of Urology and Pathology, Weill Cornell Medical College, New York-Presbyterian Hospital, New York, NY, USA; e University Medical Center Hamburg-Eppendorf, Hamburg, Germany; f Department of Urology, Cochin Hospital, APHP, Paris Descartes University, Paris, France; g Medical University of Graz, Graz, Austria; h University of Texas Health Science Center San Antonio, San Antonio, TX, USA; i Division of Medical Oncology, Weill Cornell Medical College/New York-Presbyterian Hospital, New York, NY, USA; j Ludwig- Maximilians-Universität München, Klinikum Grosshadern, Munich, Germany; k Klinik für Urologie, Klinikum Kassel GmbH, Kassel, Germany; l University of Padua, Padua, Italy; m University Hospital Ibn Rochd, Casablanca, Morocco; n Texas Oncology, Veterans Affairs Medical Center, Baylor College of Medicine, Houston, TX, USA; o University of Southern California, Los Angeles, CA, USA; p University of Texas Southwestern Medical Center, Dallas, TX, USA Article info Article history: Accepted July 12, 2012 Published online ahead of print on July 20, 2012 Keywords: Extranodal extension Lymph node metastasis Prognosis Bladder cancer Urothelial carcinoma Survival Recurrence Abstract Background: Lymph node metastasis (LNM) is the most powerful pathologic predictor of disease recurrence after radical cystectomy (RC). However, the outcomes of patients with LNM are highly variable. Objective: To assess the prognostic value of extranodal extension (ENE) and other lymph node (LN) parameters. Design, setting, and participants: A retrospective analysis of 748 patients with urothelial carcinoma of the bladder and LNM treated with RC and lymphadenectomy without neoadjuvant therapy at 10 European and North American centers (median follow-up: 27 mo). Intervention: All subjects underwent RC and bilateral pelvic lymphadenectomy. Outcome measurements and statistical analysis: Each LNM was microscopically evaluated for the presence of ENE. The number of LNs removed, number of positive LNs, and LN density were recorded and calculated. Univariable and multivariable analyses addressed time to disease recurrence and cancer-specific mortality after RC. Results and limitations: A total of 375 patients (50.1%) had ENE. The median number of LNs removed, number of positive LNs, and LN density were 15, 2, and 15, respectively. The rate of ENE increased with advancing pt stage ( p < 0.001). In multivariable Cox regression analyses that adjusted for the effects of established clinicopathologic features and LN parameters, ENE was associated with disease recurrence (hazard ratio [HR]: 1.89; 95% confidence interval [CI], ; p < 0.001) and cancer-specific mortality (HR: 1.90; 95% CI, ; p < 0.001). The addition of ENE to a multivariable model that included pt stage, tumor grade, age, gender, lymphovascular invasion, surgical margin status, LN density, number of LNs removed, number of positive LNs, and adjuvant y These authors contributed equally. * Corresponding author. Weill Cornell Medical College/New York-Presbyterian Hospital, Department of Urology, Starr 900, 525 East 68th St., New York, NY 10065, USA. Tel ; Fax: address: sfshariat@gmail.com (S.F. Shariat) /$ see back matter # 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved.

2 838 EUROPEAN UROLOGY 64 (2013) chemotherapy improved predictive accuracy for disease recurrence and cancer-specific mortality from 70.3% to 77.8% ( p < 0.001) and from 71.8% to 77.8% ( p = 0.007), respectively. The main limitation of the study is its retrospective nature. Conclusions: ENE is an independent predictor of both cancer recurrence and cancerspecific mortality in RC patients with LNM. Knowledge of ENE status could help with patient counseling, clinical decision making regarding inclusion in clinical trials of adjuvant therapy, and tailored follow-up scheduling after RC. # 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved. 1. Introduction The prognosis of patients with muscle-invasive urothelial carcinoma of the bladder (UCB) treated with radical cystectomy (RC) is closely related to the pathologic stage of the primary tumor and the presence of lymph node metastasis (LNM) [1,2]. While regional LNM is associated with a significantly increased risk of cancer recurrence and mortality, in select individuals with LNM, RC with pelvic lymphadenectomy can be curative [2 5]. To better risk-stratify this heterogeneous group of patients, investigators have proposed the use of number of positive lymph nodes (LNs) and ratio of positive LNs to LNs removed (LN density) [6 10]. While some investigators have shown that extension of tumor tissue through the capsule into perinodal tissue (extranodal extension [ENE]) has prognostic value in UCB patients with LNM [11,12], other investigators could not verify these findings [13 15]. Before integration of ENE in the American Joint Committee on Cancer (AJCC) staging system and/or in an individualized clinical decision-making process, the prognostic value of ENE needs to be externally validated in large, multicenter studies. The question of whether ENE can improve the prognostic accuracy of established predictors of cancer outcome, such as number of positive LNs and LN density, necessitates more than the conventional univariable and multivariable analyses. It must be established that the use of ENE adds unique information that improves the performance of a predictive model constructed without ENE by a statistically and prognostically significant margin [16,17]. Therefore, we sought to verify the independent prognostic value of ENE in a large, multi-institutional cohort of patients treated with RC for UCB and to test whether ENE improved the accuracy of predictive models, including established clinical and pathologic predictors of cancer recurrence and mortality. computed tomography (CT) or magnetic resonance imaging of the abdomen and pelvis. A total of 423 patients (56.6%) received adjuvant chemotherapy at the investigator s discretion. A computerized databank was generated for data transfer. After combining data sets, reports were generated to identify data inconsistencies and data integrity problems. Through regular communication with all sites, resolution of all identified anomalies was achieved before analysis. Prior to analysis, the database was frozen, and the final data set was produced Pathologic evaluation All surgical specimens were processed according to standard pathologic procedures. Tumor grade and stage were assigned according to the 1973 World Health Organization grading and2002 AJCC TNM systems. Pelvic LN dissections were examined grossly, and all lymphoid tissue was submitted for histologic examination. Lymphovascular invasion (LVI) was defined as the unequivocal presence of tumor cells within an endothelium-lined space without underlying muscular walls [18]. Microscopically, each metastasis was evaluated for the presence or absence of ENE, defined as a clear-cut perforation of a microscopically visible LN capsule by tumor tissue infiltrating into perinodal tissue (Fig. 1). A genitourinary pathologist at each site re-reviewed all slides and verified the presence or absence of ENE. LN density wasdefinedas the ratioof the number ofpositive LNsto the total number of LNs removed Follow-up Follow-up was performed according to institutional protocols. Patients were generally evaluated postoperatively every 3 4 mo for the first year, semiannually forthe secondyear, and annuallythereafter. Follow-upvisits consisted of a physical examination and serum chemistry evaluation. [(Fig._1)TD$FIG] Diagnostic imaging of the upper tracts (eg, ultrasonography and/or 2. Patients and methods 2.1. Patient selection and data collection This study was approved by the institutional review board, with all participating sites providing the necessary institutional data sharing agreements prior to initiation of the study. A total of 10 centers worldwide provided data. This study comprised 3134 UCB patients who underwent RC with bilateral lymphadenectomy between 1979 and 2008, of which 748 had LNM. No patient received neoadjuvant radiotherapy or chemotherapy. The indications for RC were (1) tumor invasion into the muscularis propria or prostatic stroma or (2) high-risk non muscle-invasive UCB. No patient had distant metastatic disease at the time of RC as evidenced by chest radiography, bone scan, and Fig. 1 Representative image of extranodal extension of urothelial carcinoma (asterisk), with tumor seen in adipose tissue beyond lymph node capsule (arrows).

3 [(Fig._2)TD$FIG] EUROPEAN UROLOGY 64 (2013) Fig. 2 Kaplan-Meier curves of recurrence-free survival stratified according to extranodal extension (ENE) in 748 patients with lymph node metastases treated with radical cystectomy for urothelial carcinoma of the bladder. intravenous pyelography, CT urography) and chest radiography were performed at least annually, or earlier when clinically indicated. Additional radiographic evaluation, such as bone scan and/or CT, was performed atthe discretion ofthe treatingphysician. Detectionofcancerin the ureter and/or urethra was coded as a second (metachronous) primary and not as a local or distant recurrence. When patients died, the cause of [(Fig._3)TD$FIG] death was determined by the treating physicians, by chart review corroborated by death certificates, or by death certificates alone [19]. All patients who were identified as havingdied ofucb hadprogressive, widely disseminated, and often highly symptomatic metastases at the time of death. Perioperative mortality (death within 30 d of surgery) was censored at time of death for UCB-specific survival analyses. Fig. 3 Kaplan-Meier curves of cancer-specific survival stratified according to extranodal extension (ENE) in 748 patients with lymph node metastases treated with radical cystectomy for urothelial carcinoma of the bladder.

4 840 EUROPEAN UROLOGY 64 (2013) Statistical analysis 3. Results The Fisher exact test and the x 2 test were used to evaluate the association between categorical variables. Differences in variables with a continuous distribution across dichotomous categories were assessed using the Mann-Whitney U test. Time-to-event calculations started at the date of RC. The Kaplan-Meier method was used to calculate survival functions, and differences were assessed with the log-rank statistic. Univariable and multivariable Cox regression models addressed time to disease recurrence and cancer-specific mortality after RC. In all models, proportional hazards assumptions were systematically verified using the Grambsch-Therneau residual-based test. Predictive accuracy was quantified with the Harrell concordance index [20,21]. The DeLong test was used to compare the predictive accuracy of various models [22]. Internal validation was performed using 200 bootstrap samples. We performed subgroup analyses in patients treated with or without adjuvant chemotherapy to minimize any potential selection bias and in patients with 25 LNs or <25 LNs removed to mitigate the effects of the varying extent of lymphadenectomy. The stratification cut-off of 25 LNs removed was consistent with a recent study demonstrating that removal of 25 LNs results in a 75% probability of finding 1 positive LNs in patients treated with RC [23]. All reported p values are two-sided, and statistical significance was set at p < All statistical tests were performed with S-Plus Professional (MathSoft, Inc.) or SPSS v.17.0 (IBM Corp., Armonk, NY, USA) Association of extranodal extension and other lymph node parameters with clinical and pathologic characteristics A total of 375 patients (50.1%) had ENE. Table 1 shows the clinicopathologic characteristics of the patients and their association with ENE. The median number of LNs removed, number of positive LNs, and LN density were 15, 2, and 15, respectively. The rate of ENE increased with advancing pt stage ( p < 0.001). ENE was not associated with the number of LNs removed, number of positive LNs, or LN density Association of extranodal extension and other lymph node parameters with clinical outcomes in all patients The median follow-up was 27 mo (mean: ; interquartile range: 44). occurred in 420 patients (56.1%); 438 patients (58.6%) were dead at last follow-up, and 353 patients (47.2%) died of UCB. Actuarial recurrence-free survival estimates at 2, 5, and 10 yr after RC were 37% 2 (standard error), 29% 2, and 26% 2, Table 1 Association of extranodal extension with clinical and pathologic characteristics in 748 patients with positive lymph nodes at radical cystectomy for urothelial carcinoma of the bladder All patients, no. (%) Extranodal extension p value Negative, no. (%) Positive, no. (%) All (49.9) 375 (50.1) Gender 0.93 Male 609 (81.4) 303 (81.2) 306 (81.6) Female 139 (18.6) 70 (18.8) 69 (18.4) Age 0.35 Median (IQR) (14) (14) (14) Soft tissue surgical margin status 0.47 Positive 110 (14.7) 51 (13.7) 59 (15.7) Negative 638 (85.3) 322 (86.3) 316 (84.3) Adjuvant chemotherapy 0.89 Present 423 (56.6) 210 (56.3) 213 (56.8) Absent 325 (43.4) 163 (43.7) 162 (43.2) pt stage <0.001 <pt2 44 (5.9) 38 (10.2) 6 (1.6) pt2 121 (16.2) 73 (19.6) 48 (12.8) pt3 377 (50.4) 183 (49.1) 194 (51.7) pt4 206 (27.5) 79 (21.2) 127 (33.9) Tumor grade (2.4) 8 (2.1) 10 (2.7) (15.2) 60 (16.1) 54 (14.4) (82.4) 305 (81.8) 311 (82.9) Lymphovascular invasion 0.18 Present 457 (61.1) 219 (58.7) 238 (63.5) Absent 291 (38.9) 154 (41.3) 137 (36.5) Concomitant carcinoma in situ 0.56 Present 287 (38.4) 147 (39.4) 140 (37.3) Absent 461 (61.6) 226 (60.6) 235 (62.7) Lymph node density, %, median 0.23 (IQR) 15 (26) 14 (26) 17 (25) No. of lymph nodes removed, median 0.17 (IQR) 15 (13) 15 (14) 14 (12) No. of positive lymph nodes, median 0.68 (IQR) 2 (3) 2 (3) 2 (3) IQR = interquartile range.

5 EUROPEAN UROLOGY 64 (2013) Table 2 Univariable Cox regression analyses assessing the association between predictor variables and disease recurrence and cancerspecific mortality in 748 patients with positive lymph nodes at radical cystectomy for urothelial carcinoma of the bladder Age (continuous) Female gender pt stage <0.001 <0.001 pt pt pt Tumor grade (grade 3 vs grade 1/2) Lymphovascular invasion Adjuvant chemotherapy Soft tissue surgical margin positivity < <0.001 Lymph node density < <0.001 No. of lymph nodes removed No. of positive lymph nodes < <0.001 Extranodal extension < <0.001 CI = confidence interval; HR = hazard ratio; ref = referent. respectively. Actuarial cancer-specific survival estimates at 2, 5, and 10 yr after RC were 48% 2, 36% 2, and 29% 3, respectively. In univariable analyses (Figs. 2 and 3, Table 2), ENE was associated with a higher risk of disease recurrence (hazard ratio [HR]: 1.92; 95% confidence interval [CI], ; p < 0.001) and cancer-specific mortality (HR: 2.01; 95% CI, ; p < 0.001). ENE had the highest C-index for predicting both disease recurrence and cancer-specific mortality (71.0% and 68.1%, respectively), followed by pathologic stage (61.4% and 63.7%, respectively), LN density (61.1% and 58.5%, respectively), soft tissue surgical margin status (56.9% and 59.2%, respectively), and number of positive LNs (57.9% and 56.8%, respectively). In multivariable Cox regression analyses that adjusted for the effects of established clinicopathologic features, ENE was associated with disease recurrence (HR: 1.89; 95% CI, ; p < 0.001) and cancer-specific mortality (HR: 1.90; 95% CI, ; p < 0.001) (Table 3). The C-indexes of base multivariable models including pt stage, tumor grade, gender, age, LVI, adjuvant chemotherapy, soft tissue surgical margin status, LN density, number of LNs removed, and number of positive LNs for prediction of disease recurrence and cancer-specific mortality were 70.3% and 71.8%, respectively. The addition of ENE to the base model improved its accuracies for predicting both disease recurrence and cancer-specific mortality to 77.8% ( p < 0.001) and 77.8% ( p = 0.007), respectively Association of extranodal extension with clinical outcomes in patients stratified by adjuvant chemotherapy Adjuvant chemotherapy was administered to 423 patients (56.6%). We performed subgroup analyses on patients Table 3 Multivariable Cox regression analyses assessing the association between predictor variables and disease recurrence and cancerspecific mortality in 748 patients with positive lymph nodes at radical cystectomy for urothelial carcinoma of the bladder pt stage pt pt pt Tumor grade (grade 3 vs grade 1/2) Female gender Age (continuous) Lymphovascular invasion Adjuvant chemotherapy Soft tissue surgical margin positivity < <0.001 Lymph node density No. of lymph nodes removed No. of positive lymph nodes Extranodal extension < <0.001 CI = confidence interval; HR = hazard ratio; ref = referent.

6 842 EUROPEAN UROLOGY 64 (2013) Table 4 Multivariable Cox regression analyses assessing the association between predictor variables and disease recurrence and cancerspecific mortality in 423 lymph node positive patients who received adjuvant chemotherapy and 325 lymph node positive patients who did not receive adjuvant chemotherapy after radical cystectomy for urothelial carcinoma of the bladder Patients who received adjuvant chemotherapy pt stage pt pt pt Tumor grade (grade 3 vs grade 1/2) Female gender Age (continuous) Lymphovascular invasion Soft tissue surgical margin positivity Lymph node density No. of lymph nodes removed <0.001 No. of positive lymph nodes Extranodal extension < <0.001 Patients who did not receive adjuvant chemotherapy pt stage pt pt pt Tumor grade (grade 3 vs grade 1/2) Female gender Age (continuous) Lymphovascular invasion Soft tissue surgical margin positivity Lymph node density No. of lymph nodes removed No. of positive lymph nodes Extranodal extension CI = confidence interval; HR = hazard ratio; ref = referent. treated with or without adjuvant chemotherapy as an adjunct to the multivariable analyses to minimize any potential effects of bias in patients selected for adjuvant chemotherapy (Table 4). In patients treated with adjuvant chemotherapy, ENE was independently associated with disease recurrence (HR: 2.03; 95% CI, ; p < 0.001) and cancer-specific mortality (HR: 1.97; 95% CI, ; p < 0.001). Similarly, in patients not treated with adjuvant chemotherapy, ENE was independently associated with disease recurrence (HR: 1.82; 95% CI, ; p = 0.001) and cancer-specific mortality (HR: 1.87; 95% CI, ; p = 0.002) Association of extranodal extension with clinical outcomes in patients stratified according to the number of lymph nodes removed We stratified patients into individuals with >25 and <25 LNs removed (Table 5), consistent with a recent study demonstrating that removal of 25 LNs results in a 75% probability of finding 1 LNs in patients treated with RC [23]. A total of 162 patients (21.7%) had 25 LNs removed, and 586 patients (78.3%) had <25 LNs removed. In patients with <25 LNs removed, ENE was independently associated with disease recurrence (HR: 1.86; 95% CI, ; p < 0.001) and cancer-specific mortality (HR: 2.00; 95% CI, ; p < 0.001). Similarly, in patients with 25 LNs removed, ENE was independently associated with disease recurrence (HR: 2.25; 95% CI, ; p = 0.001) and cancer-specific mortality (HR: 1.95; 95% CI, ; p = 0.007). 4. Discussion There is considerable variability in the clinical outcomes of patients with LN-positive UCB treated with RC. Several risk factors, such as tumor and LN stages, number and density of positive LNs, and number of LNs removed, have been evaluated as a means of risk-stratifying patients with LNM. Approximately 20 30% of patients with muscle-invasive UCB present with LNM, which is the strongest adverse pathologic feature. Extension of metastatic deposits in LNs

7 EUROPEAN UROLOGY 64 (2013) Table 5 Multivariable Cox regression analyses assessing the association between predictor variables and disease recurrence and cancerspecific mortality in 162 lymph node positive patients who had I25 lymph nodes removed and 586 lymph node positive patients who had <25 lymph nodes removed at radical cystectomy for urothelial carcinoma of the bladder Patients with 25 lymph nodes removed pt stage pt pt pt Tumor grade (grade 3 vs grade 1/2) Female gender Age (continuous) Lymphovascular invasion Soft tissue surgical margin positivity Lymph node density No. of positive lymph nodes Adjuvant chemotherapy Extranodal extension Patients with <25 lymph nodes removed pt stage <pt (ref.) 1.00 (ref.) pt pt pt Tumor grade (grade 3 vs grade 1/2) Female gender Age (continuous) Lymphovascular invasion Soft tissue surgical margin positivity Lymph node density No. of positive lymph nodes Adjuvant chemotherapy Extranodal extension < <0.001 CI = confidence interval; HR = hazard ratio; ref = referent. beyond the capsule may reflect the inherent biologic aggressiveness of the tumor and/or the time span the tumor had to grow and spread to other areas. Using a large multi-institutional database, we demonstrated that ENE is a strong independent prognostic factor in UCB patients with LNM who underwent RC. These findings are in agreement with previous studies showing an independent predictive value of ENE in UCB. In a study of 101 patients with LNM after RC, Fleischmann et al. reported that ENE was an independent predictor of disease recurrence [11]. In a larger series from the same institution, Seiler et al. recently confirmed that ENE is an independent prognostic factor for disease recurrence, cancer-specific mortality, and mortality [12]. In contrast, analyses of other similarly sized series, such as that of Kassouf et al. [13], did not reveal an independent prognostic value for ENE in UCB [14,15]. To address this discrepancy in findings, we performed an analysis of a data set comprising four times as many patients as the previous studies. We found that ENE improved the accuracy of multivariable models that included established predictors by a clinically and statistically significant margin when evaluated in all RC patients. Previous studies did not assess the additive value of ENE beyond standard clinicopathologic features in predicting outcomes [11,12]. Interestingly, the improvement in prediction of recurrence-free and cancerspecific survival was significant in all patients and in the subgroups of patients who received adjuvant chemotherapy and patients with <25 LNs removed. We examined the prognostic value of several established LN parameters, including LN density, number of LNs removed, and number of positive LNs. On univariable analyses, LN density and number of positive LNs were inferior to ENE in terms of predicting disease recurrence and cancer-specific mortality. In contrast to previously published reports, our study did not demonstrate an independent prognostic value for LN density [6,8,24,25]. Our multivariable model included additional variables such as gender, LVI, and ENE, each of which were independent predictors of disease recurrence and cancer-specific mortality, potentially explaining the discrepancy in findings. Our study demonstrated that the number of positive LNs

8 844 EUROPEAN UROLOGY 64 (2013) was associated with clinical outcomes following RC [11,26]. However, when combined into a multivariable model including pt stage, gender, LVI, soft tissue surgical margin status, and ENE, the number of positive LNs no longer remained an independent prognostic factor. This finding is in accordance with previous studies [9]. We also found a higher magnitude of effect for ENE on cancer-specific mortality in patients with 25 LNs removed compared with the whole cohort. Since the number of LNs removed is a proxy of the dissection extent, one could conclude that there is a need for more extended pelvic LN dissection. However, the overlapping CIs for ENE comparing 25 LNs and <25 LNs removed limit conclusions based on these findings. In our cohort of UCB patients with LNM treated with RC and lymphadenectomy, we also found that soft tissue surgical margin status was an independent predictor of disease recurrence and cancer-specific mortality. In previous studies of ENE, this variable was either not considered [11,12,15] or was not a significant predictor in multivariable analyses of these end points [13,14]. As previously stated, these prior studies may have been limited by smaller sample size. Similarly, in other studies of UCB patients with LNM, surgical margin status is not universally reported or included in statistical analyses [25]. Future studies should include this variable to further examine this finding. There are several limitations to our study, including limitations common to all retrospective analyses. Variability in determination to identify ENE could represent a limitation, as we did not perform a central pathology review but rather re-review at each individual site. Additionally, the template for lymphadenectomy varied between sites and surgeons, which may have a significant impact on our findings. Therefore, we adjusted for the number of LNs removed and LN density in our analyses. Nevertheless, these limitations would more likely limit the prognostic value of ENE. We could not address the potential impact of tumor or LN diameter, as we did not have those data [14]. However, in previous UCB studies and in other cancers, ENE was more important than diameter of LN metastases [12]. Some of our findings are based on subgroup analyses, which may limit their validity. Since patients in our study were treated by various physicians over a long time period and specimens were evaluated by several pathologists, our data are heterogeneous and representative of the diversity of current practice. 5. Conclusions We found that ENE was a powerful prognosticator of oncologic outcomes in UCB patients treated with RC with LNM, regardless of receipt of adjuvant chemotherapy and number of LNs removed. In addition, ENE improved the accuracy of currently used predictors of disease recurrence and cancer-specific mortality by a statistically and prognostically significant margin. In contrast, number of positive LNs, number of LNs removed, and LN density did not add any prognostic information. Standardization of reporting of ENE in UCB should be considered, as knowledge of this information could allow for improved patient counseling and might be used to alter follow-up scheduling and decision making regarding adjuvant therapy. Author contributions: Shahrokh F. Shariat 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: Shariat. Acquisition of data: Fajkovic, Cha, Jeldres, Robinson, Rink, Chromecki, Breinl, Svatek, Donner, Tagawa, Tilki, Bastian, Karakiewicz, Volkmer, Novara, Joual, Faison, Sonpavde, Daneshmand, Lotan, Scherr, Shariat, Xylinas. Analysis and interpretation of data: Fajkovic, Cha, Jeldres, Karakiewicz, Shariat. Drafting of the manuscript: Fajkovic, Cha, Shariat. Critical revision of the manuscript for important intellectual content: Fajkovic, Cha, Jeldres, Robinson, Rink, Chromecki, Breinl, Svatek, Donner, Tagawa, Tilki, Bastian, Karakiewicz, Volkmer, Novara, Joual, Faison, Sonpavde, Daneshmand, Lotan, Scherr, Shariat, Xylinas. Obtaining funding: None. Administrative, technical, or material support: None. Supervision: Shariat. Other (specify): None. Financial disclosures: Shahrokh F. Shariat 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: None. References [1] Margulis V, Lotan Y, Montorsi F, Shariat SF. Predicting survival after radical cystectomy for bladder cancer. BJU Int 2008;102: [2] Stenzl A, Cowan NC, De Santis M, et al. Treatment of muscleinvasive and metastatic bladder cancer: update of the EAU guidelines. Eur Urol 2011;59: [3] Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol 2001;19: [4] Shariat SF, Karakiewicz PI, Palapattu GS, et al. Outcomes of radical cystectomy for transitional cell carcinoma of the bladder: a contemporary series from the Bladder Cancer Research Consortium. J Urol 2006;176: , discussion [5] Madersbacher S, Hochreiter W, Burkhard F, et al. Radical cystectomy for bladder cancer today a homogeneous series without neoadjuvant therapy. J Clin Oncol 2003;21: [6] Herr HW. Superiority of ratio based lymph node staging for bladder cancer. J Urol 2003;169: [7] Stein JP, Cai J, Groshen S, Skinner DG. Risk factors for patients with pelvic lymph node metastases following radical cystectomy with en bloc pelvic lymphadenectomy: concept of lymph node density. J Urol 2003;170: [8] Kassouf W, Agarwal PK, Herr HW, et al. Lymph node density is superior to TNM nodal status in predicting disease-specific survival after radical cystectomy for bladder cancer: analysis of pooled data from MDACC and MSKCC. J Clin Oncol 2008;26: [9] Kassouf W, Svatek RS, Shariat SF, et al. Critical analysis and validation of lymph node density as prognostic variable in urothelial carcinoma of bladder. Urol Oncol. In press. [10] Karl A, Carroll PR, Gschwend JE, et al. The impact of lymphadenectomy and lymph node metastasis on the outcomes of radical cystectomy for bladder cancer. Eur Urol 2009;55:

9 EUROPEAN UROLOGY 64 (2013) [11] Fleischmann A, Thalmann GN, Markwalder R, Studer UE. Extracapsular extension of pelvic lymph node metastases from urothelial carcinoma of the bladder is an independent prognostic factor. J Clin Oncol 2005;23: [12] Seiler R, von Gunten M, Thalmann GN, Fleischmann A. Extracapsular extension but not the tumour burden of lymph node metastases is an independent adverse risk factor in lymph node-positive bladder cancer. Histopathology 2011;58: [13] Kassouf W, Leibovici D, Luongo T, et al. Relevance of extracapsular extension of pelvic lymph node metastasis in patients with bladder cancer treated in the contemporary era. Cancer 2006;107: [14] Stephenson AJ, Gong MC, Campbell SC, Fergany AF, Hansel DE. Aggregate lymph node metastasis diameter and survival after radical cystectomy for invasive bladder cancer. Urology 2010;75: [15] Jeong IG, Ro JY, Kim SC, et al. Extranodal extension in node-positive bladder cancer: the continuing controversy. BJU Int 2011;108: [16] Shariat SF, Karakiewicz PI, Godoy G, et al. Survivin as a prognostic marker for urothelial carcinoma of the bladder: a multicenter external validation study. Clin Cancer Res 2009;15: [17] Shariat SF, Lotan Y, Vickers A, et al. Statistical consideration for clinical biomarker research in bladder cancer. Urol Oncol 2010;28: [18] Shariat SF, Svatek RS, Tilki D, et al. International validation of the prognostic value of lymphovascular invasion in patients treated with radical cystectomy. BJU Int 2010;105: [19] Rink M, Fajkovic H, Cha EK, et al. Death certificates are valid for the determination of cause of death in patients with upper and lower tract urothelial carcinoma. Eur Urol 2012;61: [20] Harrell Jr FE, Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the yield of medical tests. JAMA 1982;247: [21] Harrell Jr FE, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15: [22] DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44: [23] Capitanio U, Suardi N, Shariat SF, et al. Assessing the minimum number of lymph nodes needed at radical cystectomy in patients with bladder cancer. BJU Int 2009;103: [24] Herr HW. The concept of lymph node density is it ready for clinical practice? J Urol 2007;177:1273 5, discussion [25] Wright JL, Lin DW, Porter MP. The association between extent of lymphadenectomy and survival among patients with lymph node metastases undergoing radical cystectomy. Cancer 2008;112: [26] Lerner SP, Skinner DG, Lieskovsky G, et al. The rationale for en bloc pelvic lymph node dissection for bladder cancer patients with nodal metastases: long-term results. J Urol 1993;149:758 64, discussion