Radical cystectomy (RC) with pelvic lymph node dissection (LND) is the standard of care for refractory non-muscle-invasive and muscle-invasive bladder cancer. Although consensus exists on the need for LND, its extent is still debated.
To develop a model that allows preoperative determination of the minimum number of lymph nodes (LNs) needed to be removed at RC to ensure true nodal status.
Design, setting, and participants
We analyzed data from 4335 patients treated with RC and pelvic LND without neoadjuvant chemotherapy at 12 academic centers located in the United States, Canada, and Europe.
We estimated the sensitivity of pathologic nodal staging using a beta-binomial model and developed clinical (preoperative) nodal staging scores (cNSS), which represent the probability that a patient has LN metastasis as a function of the number of examined nodes.
Results and limitations
The probability of missing a positive LN decreased with an increasing number of nodes examined (52% if 3 nodes were examined, 40% if 5 were examined, and 26% if 10 were examined). A cNSS of 90% was achieved by examining 6 nodes for clinical Ta-Tis tumors, 9 nodes for cT1 tumors, and 25 nodes for cT2 tumors. In contrast, examination of 25 nodes provided only 77% cNSS for cT3-T4 tumors. The study is limited due to its retrospective design, its multicenter nature, and a lack of preoperative staging parameters.
Every patient treated with RC for bladder cancer needs an LND to ensure accurate nodal staging. The minimum number of examined LNs for adequate staging depends preoperatively on the clinical T stage. Predictive tools can give a preoperative estimation of the likelihood of nodal metastasis and thereby allow tailored decision-making regarding the extent of LND at RC.
Keywords: Lymph node, Radical cystectomy, Prognosis, Bladder cancer, Urothelial carcinoma, Survival.
Radical cystectomy (RC) with bilateral pelvic lymph node dissection (LND) is the standard of care for refractory non-muscle-invasive and muscle-invasive bladder cancer. Despite advancements in surgical technique, imaging, perioperative management, and chemotherapy, approximately 50% of patients develop metastases and die from their disease , , and . Lymph node (LN) positivity is a critical factor for disease-specific survival and a primary determinant of therapeutic course following surgery. Multi-institutional series of patients treated with RC have shown that approximately 80% of patients with pathologic node-positive disease experience disease recurrence, compared with 30% of patients with extravesical disease and pathologically negative LN , , and . Several studies have found that in addition to nodal status, the extent of LND (defined by number of nodes removed, the number of positive nodes detected, and LN density) can have prognostic and therapeutic implications , , and .
In an effort to reduce understaging and maximize survival, many studies have tried to establish a minimum number of LNs needed to be taken at time of RC , , and . While a minimum number of LNs needed to be removed has been proposed, the survival probability continues to improve with increasing number of LNs removed , , , , , , and . The discrepancy between studies in the number of LNs needed to be removed may be due to extent of dissection, pathologic factors (ie, the manner in which specimens are submitted or interpreted by pathologists), and variability in patients’ pelvic anatomy . The difference in disease severity between study populations likely also contributed to the discrepancies. Indeed, the rate of LN metastasis increases from a low of 5–10% in non-muscle-invasive bladder tumors (pT0, pTa, pTis, pT1) to 15–20% in superficial muscle-invasive tumors (pT2a), to 25–30% in deep muscle-invasive tumors, and to >40% in extravesical tumors (pT3-4) , , , , , and . We hypothesized that the number of LNs needed to be removed to ensure accurate nodal staging could be predicted based on clinical (preoperative) tumor stage. Recently, Gonen et al. used a beta-binomial model to estimate the probability that a colorectal cancer patient is correctly staged as node negative . Using a similar approach, we developed a model that uses the clinical (preoperative tumor stage) to determine the number of nodes needed to be removed at RC to determine the true nodal status.
2.1. Patient selection and data collection
This was an institutional review board–approved study, with all participating sites providing the necessary institutional data-sharing agreements prior to initiation. A computerized databank was generated for data transfer. After combining the data sets, reports were generated for each variable to identify data inconsistencies and other 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.
A total of 12 academic centers worldwide provided data. This study comprised 4335 patients who underwent RC with bilateral pelvic LND between 1980 and 2008. No patient received preoperative radiotherapy or chemotherapy. No patient had distant metastatic disease at the time of RC.
2.2. Pathologic evaluation
All surgical specimens were processed according to standard pathologic procedures as discussed by Shariat et al. . Genitourinary pathologists assigned pathologic stage, which was reassigned according to the 2002 American Joint Committee on Cancer (AJCC) Tumor, Node, Metastasis (TNM) staging system. All lymphoid tissue removed was submitted for histologic examination. The extent of LND was at the surgeon's discretion. Extended LND was not routinely performed. Clinical stage was assigned based on the information from the pathologic evaluation of the transurethral resection (TUR) specimen, bimanual examination, and imaging study results.
2.3. Statistical analysis
The method we followed is identical to an earlier work on colon cancer . Briefly, we were concerned with the probability of incorrect nodal staging as a function of the number of examined nodes. The true nodal status is unascertainable, but information from node-positive patients can used to determine if the number of nodes examined and the number of these that are negative are sufficient to classify a patient as node negative. Consider a patient with a large number of examined nodes and small, positive k, where k is the number of positive nodes from patients with node involvement: If fewer nodes were examined, there would be a chance that this patient would be incorrectly deemed node negative. Conversely, for a patient with a small number of examined nodes and large k, it is unlikely that nodal disease would have been missed, even though fewer nodes were examined. Hence, the data from node-positive patients were used to interpret the data for the node-negative patients.
The probability that a node-negative patient has nodal disease can be computed using the following algorithm: Compute the probability of missing a positive node, compute the prevalence, and compute the nodal staging score from sensitivity and prevalence.
2.3.1. Probability of missing a positive node
The probability of missing a positive node (1, the sensitivity) is inherent to the process of pathologic detection and, as such, depends on the number of examined nodes but not on patient characteristics. We used a beta-binomial model for this purpose, allowing for heterogeneity in the intensity of nodal spread across patients. Two key assumptions underlie this step: (1) There are no false positives (if the specimen contains a positive node, it will be correctly identified by the pathologist), and (2) sensitivity is the same for node-positive and node-negative patients. These assumptions may not be completely tenable, but we find them to be sufficient approximations to our biologic understanding of nodal spread and clinical practice of nodal staging.
2.3.2. Estimation of prevalence of nodal disease
The observed prevalence was an underestimate and needed to be adjusted for false negatives. This was done in two steps. The first step invokes Assumption 1 and estimates #FNk as a function of k:
In this step, #TPk is the number of true positives for a given k. Since prevalence is not a function of k, the second step obtains the adjusted prevalence by averaging over k:
Estimation of prevalence was stratified by T stage for clinical (preoperative) nodal staging scores (cNSS), but this is not explicitly noted in the above formula to avoid cumbersome notation.
2.3.3. Nodal staging score
Adequate staging was assessed by computing NSS, the probability that a pathologically node-negative patient is indeed free of nodal disease:
2.3.4. Confidence intervals
The precision of the reported estimates was assessed by creating 2000 bootstrap samples from the entire data set and replicating the estimation process . We formed 95% confidence intervals (CIs) using this bootstrap estimate of the corresponding sampling distributions. Analyses were performed with SPSS 17 (IBM Corp., Armonk, NY, USA).
Table 1 shows the descriptive characteristics of the patients. The median number of removed LNs was 18 (quartiles: 11–31) and 74.2% of the patients were deemed node negative. LN metastases were present in 32 of 774 (4.1%) pT0/Ta/Tis patients, 40 of 585 (6.8%) pT1 patients, 188 of 1042 (18.0%) pT2 patients, and 859 of 1934 (44.4%) pT3-4 patients.
|Patients, no.||Patients, %|
|Clinical T stage||Ta||138||3.2|
|Pathologic T stage||T0-Ta-Tis||774||17.9|
|Pathologic N stage||Negative||3216||74.2|
|Examined nodes, no.||18||1–136|
|Removed nodes in patients with positive nodes, no.||18||1–136|
|Positive nodes in patients with positive nodes, no.||2||1–93|
Using our model, the beta-binomial parameters α and β were estimated to be 0.194 (95% CI, 0.189–0.199) and 0.306 (95% CI, 0.301–0.311). The resulting probability of missing nodal disease (1 − the sensitivity) as a function of the number of LN examined is plotted in Figure 1. As expected, the probability of missing nodal disease decreased as the number of nodes examined increased (Table 2): If only a single node was examined in all patients, 76% of nodal disease would be missed. Even with 11 nodes examined (observed median), 24% of the node-positive patients would be incorrectly staged. Sensitivity of nodal staging exceeded 80% only when ≥15 nodes were examined.
|Nodes examined, no.||1||3||5||8||10||15||20||25|
|Probability of missing nodal disease, %||76||52||40||30||26||19||15||13|
The clinical nodal staging score is presented in Table 3 and Figure 2. For Ta-Tis tumors, six examined nodes provide 90% confidence that the patient was, indeed, node negative. For the same level of confidence, one would need 9 nodes for T1, and 25 for T2 tumors. Even with 30 examined nodes, the probability of incorrect nodal staging remains 20%. Bootstrap CIs for all the estimates reported in Table 2 and Table 3 are all within ±1% (in absolute terms) of the estimates (data not shown).
|Nodes examined, no.||2||5||8||10||15||20||25||30|
|Tumor stage, %|
* This score can be used to find the probability of having nodal disease despite a pathologic N0 classification. The number corresponding to patient's T stage and the number of examined nodes is the percent probability of patient having node-negative disease.
Researchers have tried to identify the minimum necessary number of LNs needed to be removed at RC. However, the analysis of a large tertiary care center's database revealed that the probability of survival continues to rise as the number of LNs removed increases and that no minimum number of LNs can be determined . One potential limitation of that and previous studies is that they did not adjust for clinical factors. To address this need, the primary aim of our study was to assess whether every patient needs the same extent of LND, and if not, whether we could identify a minimal number of LN needed based on clinical tumor stage.
We found that the number of LNs needed to be removed varies largely among patients according to their tumor stage. However, in accordance with previous studies, we found that every patient treated with RC for bladder cancer, even those with cTa-Tis, needs an LND to ensure accurate nodal staging. On the other hand, even an extended LND does not ensure 100% accuracy with regard to nodal status. Indeed, different studies have shown ≤8% extrapelvic skip lesions above the aortic bifurcation, although these skip lesions have not been reported exclusively above the extended LND template  and . However, not all patients benefit from an extended LND. Furthermore, extended LND may incur morbidity. Therefore, an estimation of the minimum number of LNs needed to be removed to ensure detection of possible cancer-burdened LNs could help establish an individualized risk-based determination of the extent of LND each patient should undergo.
We found that clinical tumor stage is a powerful predictor of the number of LNs needed to be removed to ascertain LN status. In patients with clinical Ta-Tis disease, at least six LNs need to be removed to achieve 90% confidence that the patient is node negative. Up to 6% of patients with refractory cTis treated with RC harbor LN metastasis . Moreover, ≤71% and 32–48% of patients with cTa and cTis, respectively, are upstaged at RC to pathologic T1 or higher  and . One potential reason for this relatively high range of discrepancy between clinical and pathologic stage could be the lack of routine restaging TUR and currently inadequate imaging technology . While one cannot extrapolate from the number of LNs removed to the surgical template, an LND limited to the true pelvis or even to the obturator fossa in patients with cTa-Tis seems adequate. Conversely, in patients with cT1, at least 10 LNs need to be removed to ensure 90% probability of ascertaining true nodal status. In cT2 bladder cancer patients, removal of 25 LNs results in a >90% probability of ascertaining true nodal status. In cT3-4 bladder cancer, the 90% probability is not reached, with an asymptotic approach to a maximum 80% probability of detecting a positive LN. One possible reason for this is the higher likelihood of skip lesions outside of the regular template in patients with locally advanced tumor stage . Taking these data together, it seems that a safe approach would be to recommend an extended LND in patients with stage cT1 and higher bladder cancer. In patients with cT3-4 bladder cancer, one has to realize that even with an extended LND, there is the probability of missing 20% of LN metastasis. This further reinforces the need of multimodal therapy in patients with pT2 or greater disease.
We developed a simple probabilistic model to predict the number of LNs needed to be removed as a function of clinical determinants of stage, including pathologic examination of the TUR specimen, bimanual examination, and imaging studies. Generally, the extent of LND is performed based on the surgeon's intuitive experience integrating his beliefs and patient factors such as health status and tumor features. Our model is a simple tool that could guide preoperative clinical decision-making regarding the extent of LND. Performing an extended LND in all patients would result in overtreatment of many patients, with resulting side effects and cost. While an extended pelvic LND is not associated with increased morbidity compared with a limited LND in experienced hands, the risk of longer operative time and associated complications needs to be determined for each individual patient  and . In contrast, removing too few LNs may result in inaccurate staging and possibly inferior survival , , , and . However, neither intuition nor nomograms  can give a personalized risk/benefit analysis integrating threshold probabilities based on individual preferences. Our tool may enable the physician and the patient to engage in shared decision making and determine risks by incorporating the risk of missing a positive node compared with the risk/benefit of an extended LND. Therefore, our cNSS can be incorporated to individualize treatment and demonstrates the incremental improvement of nodal stage accuracy with each individual node resected.
Nevertheless, our study has some limitations. First and foremost are limitations inherent to its retrospective design and multicenter nature, as well as the lack of routine repeat TUR , failure to control for the quality of the TUR , and failure to control for other preoperative prognostic factors, such as lymphovascular invasion . Other limitations include the variability in preoperative imaging and pathologic evaluation, as well as differences in treatment decisions and surgical technique. Newer imaging techniques such as positron emission tomography/computer tomography and magnetic resonance imaging with ultrasmall superparamagnetic iron oxide may allow better decision making about the extent of LND . In addition, the number of LNs removed is not only a factor of the extent of LND but also is dependent on the pathologic evaluation and inherent differences between patients. Moreover, the number of LNs removed is not an exact surrogate for the extent of LND. Central pathology review was not performed, which might have an undefined impact, given differences in the rigor used by different pathologists to identify LNs. Conversely, our data reflect a real-world multicenter experience and pathologic examination was performed by genitourinary pathologists in major academic centers. Our study did not examine the differential impact of the extent of LND on long-term outcomes; two randomized trials are comparing a standard with extended extent of LND and may shed more light on this issue. Finally, the location of LNs is important. Removing nodes from an area of high likelihood of malignancy may be more valuable than removing nodes less likely to be involved with cancer .
We found that the risk of LN metastases and the number of LNs needed to be removed to ensure true node-negative status increases with advancing clinical stage. There is no one-size-fits-all for LND in patients treated with RC for bladder cancer. While all patients need an LND, a limited LND seems sufficient in patients with cTa-Tis, while an extended LND should be recommended for those with cT1 and higher stage. In patients with cT3-4, even an extended LND still misses about 20% of LN metastasis. We developed a simple cNSS to aid preoperative clinical decision making about the extent of LND in patients for whom RC for bladder cancer is planned. After validation, such a tool could help physicians decide treatment strategies prior to RC and tailor the extent of LND at RC.
Study concept and design: Shariat, Ehdaie, Gonen.
Acquisition of data: Svatek, Novara, Lotan, Sagalowsky, Fradet, Kassouf, Fritsche, Bastian, Burger, Izawa, Tilki, Abdollah, Scherr, Shariat.
Analysis and interpretation of data: Shariat, Gonen.
Drafting of the manuscript: Shariat, Ehdaie, Rink, Cha.
Critical revision of the manuscript for important intellectual content: Shariat, Ehdaie, Rink, Cha, Svatek, Chromecki, Fajkovic, Novara, David, Daneshmand, Fradet, Lotan, Sagalowsky, Clozel, Bastian, Kassouf, Fritsche, Burger, Izawa, Tilki, Abdollah, Chun, Sonpavde, Karakiewicz, Scherr, Gonen.
Statistical analysis: Shariat, Gonen.
Obtaining funding: None.
Administrative, technical, or material support: None.
Supervision: Gonen, Shariat.
Other (specify): None.
Financial disclosures: I certify 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: M. Burger has consulted and lectured for Astellas Pharma, GE Healthcare, and Photocure ASA.
Funding/Support and role of the sponsor: None.
-  S. Madersbacher, W. Hochreiter, F. Burkhard, et al. Radical cystectomy for bladder cancer today—a homogeneous series without neoadjuvant therapy. J Clin Oncol. 2003;21:690-696 Crossref.
-  S.F. Shariat, P.I. Karakiewicz, G.S. Palapattu, 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;176:2414-2422 discussion 2422 Crossref.
-  J.P. Stein, G. Lieskovsky, R. Cote, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol. 2001;19:666-675
-  P.I. Karakiewicz, S.F. Shariat, G.S. Palapattu, et al. Nomogram for predicting disease recurrence after radical cystectomy for transitional cell carcinoma of the bladder. J Urol. 2006;176:1354-1361 discussion 1361–2
-  S.F. Shariat, P.I. Karakiewicz, G.S. Palapattu, et al. Nomograms provide improved accuracy for predicting survival after radical cystectomy. Clin Cancer Res. 2006;12:6663-6676 Crossref.
-  P.I. Karakiewicz, S.F. Shariat, G.S. Palapattu, et al. Precystectomy nomogram for prediction of advanced bladder cancer stage. Eur Urol. 2006;50:1254-1262 discussion 1261–2 Abstract, Full-text, PDF, Crossref.
-  U. Capitanio, N. Suardi, S.F. Shariat, et al. Assessing the minimum number of lymph nodes needed at radical cystectomy in patients with bladder cancer. BJU Int. 2009;103:1359-1362 Crossref.
-  J. Leissner, M.A. Ghoneim, H. Abol-Enein, et al. Extended radical lymphadenectomy in patients with urothelial bladder cancer: results of a prospective multicenter study. J Urol. 2004;171:139-144 Crossref.
-  M. May, E. Herrmann, C. Bolenz, et al. Association between the number of dissected lymph nodes during pelvic lymphadenectomy and cancer-specific survival in patients with lymph node-negative urothelial carcinoma of the bladder undergoing radical cystectomy. Ann Surg Oncol. 2011;18:2018-2025 Crossref.
-  H.W. Herr, B.H. Bochner, G. Dalbagni, et al. Impact of the number of lymph nodes retrieved on outcome in patients with muscle invasive bladder cancer. J Urol. 2002;167:1295-1298
-  B.R. Konety, S.A. Joslyn, M.A. O’Donnell. Extent of pelvic lymphadenectomy and its impact on outcome in patients diagnosed with bladder cancer: analysis of data from the Surveillance, Epidemiology and End Results Program data base. J Urol. 2003;169:946-950 Crossref.
-  T.M. Koppie, A.J. Vickers, K. Vora, G. Dalbagni, B.H. Bochner. Standardization of pelvic lymphadenectomy performed at radical cystectomy: can we establish a minimum number of lymph nodes that should be removed?. Cancer. 2006;107:2368-2374 Crossref.
-  C.M. Hugen, A.J. Polcari, M.P. Fitzgerald, et al. Risk factors for recurrence following radical cystectomy for pathologic node negative bladder cancer. J Surg Oncol. 2010;102:334-337 Crossref.
-  B.H. Bochner, D. Cho, H.W. Herr, et al. Prospectively packaged lymph node dissections with radical cystectomy: evaluation of node count variability and node mapping. J Urol. 2004;172:1286-1290 Crossref.
-  W. Kassouf, P.K. Agarwal, H.W. Herr, 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;26:121-126 Crossref.
-  M. Gonen, D. Schrag, M.R. Weiser. Nodal staging score: a tool to assess adequate staging of node-negative colon cancer. J Clin Oncol. 2009;27:6166-6171 Crossref.
-  B. Efron, R. Tisbshirani. An Introduction to the Bootstrap. (Chapman and Hall, New York, NY, 1994)
-  B. Roth, M.P. Wissmeyer, P. Zehnder, et al. A new multimodality technique accurately maps the primary lymphatic landing sites of the bladder. Eur Urol. 2010;57:205-211 Abstract, Full-text, PDF, Crossref.
-  D. Tilki, O. Reich, R.S. Svatek, et al. Characteristics and outcomes of patients with clinical carcinoma in situ only treated with radical cystectomy: an international study of 243 patients. J Urol. 2010;183:1757-1763 Crossref.
-  S.F. Shariat, G.S. Palapattu, P.I. Karakiewicz, et al. Discrepancy between clinical and pathologic stage: impact on prognosis after radical cystectomy. Eur Urol. 2007;51:137-149 discussion 149–51
-  R.S. Svatek, S.F. Shariat, G. Novara, et al. Discrepancy between clinical and pathological stage: external validation of the impact on prognosis in an international radical cystectomy cohort. BJU Int. 2011;107:898-904 Crossref.
-  H.W. Herr, S.M. Donat. A re-staging transurethral resection predicts early progression of superficial bladder cancer. BJU Int. 2006;97:1194-1198 Crossref.
-  A. Vazina, D. Dugi, S.F. Shariat, et al. Stage specific lymph node metastasis mapping in radical cystectomy specimens. J Urol. 2004;171:1830-1834 Crossref.
-  C. Brossner, A. Pycha, A. Toth, C. Mian, W. Kuber. Does extended lymphadenectomy increase the morbidity of radical cystectomy?. BJU Int. 2004;93:64-66 Crossref.
-  N.B. Dhar, E.A. Klein, A.M. Reuther, et al. Outcome after radical cystectomy with limited or extended pelvic lymph node dissection. J Urol. 2008;179:873-878 discussion 878 Crossref.
-  K. Steven, A.L. Poulsen. Radical cystectomy and extended pelvic lymphadenectomy: survival of patients with lymph node metastasis above the bifurcation of the common iliac vessels treated with surgery only. J Urol. 2007;178:1218-1223 discussion 1223–4
-  R.T. Divrik, A.F. Şahin, Ü. Yildirim, M. Altok, F. Zorlu. Impact of routine second transurethral resection on the long-term outcome of patients with newly diagnosed pT1 urothelial carcinoma with respect to recurrence, progression rate, and disease-specific survival: a prospective randomised clinical trial. Eur Urol. 2010;58:185-190 Abstract, Full-text, PDF, Crossref.
-  P. Mariappan, A. Zachou, K.M. Grigor, for the Edinburgh Uro-Oncology Group. Detrusor muscle in the first, apparently complete transurethral resection of bladder tumour specimen is a surrogate marker of resection quality, predicts risk of early recurrence, and is dependent on operator experience. Eur Urol. 2010;57:843-849 Abstract, Full-text, PDF, Crossref.
-  K.S. Cho, H.K. Seo, J.Y. Joung, et al. Lymphovascular invasion in transurethral resection specimens as predictor of progression and metastasis in patients with newly diagnosed T1 bladder urothelial cancer. J Urol. 2009;182:2625-2630
-  N.C. Cowan, J.P. Crew. Imaging bladder cancer. Curr Opin Urol. 2010;20:409-413 Crossref.
a Weill Cornell Medical College, New York, NY, USA
b University Medical Center Hamburg-Eppendorf, Hamburg, Germany
c University of Texas San Antonio, San Antonio, TX, USA
d Medical University of Graz, Graz, Austria
e St. Poelten General Hospital, St. Poelten, Austria
f University of Padua, Padua, Italy
g University of Southern California, Los Angeles, CA, USA
h Laval University, Québec City, Québec, Canada
i University of Texas Southwestern Medical Center, Dallas, TX, USA
j Ludwig-Maximilians-Universität München, Klinikum Grosshadern, Munich, Germany
k McGill University Health Centre, Montréal, Québec, Canada
l Caritas St. Josef Medical Centre, University of Regensburg, Regensburg, Germany
m University of Western Ontario, London, Ontario, Canada
n University of Montréal, Montréal, Québec, Canada
o Baylor College of Medicine, Houston, TX, USA
p Memorial Sloan-Kettering Cancer Center, New York, NY, USA
Both authors contributed equally to the manuscript.
© 2011 European Association of Urology, Published by Elsevier B.V.