Indications, Extent, and Benefits of Pelvic Lymph Node Dissection for Patients with Bladder and Prostate Cancer

Learning Objectives

1. Evaluate all patients undergoing radical pelvic surgery to determine whether they should have pelvic lymph node dissection.

2. Determine to what extent a pelvic lymph node dissection should be performed and analyze the risks involved.

3. Based on best practices, assess whether pelvic lymph node dissection is beneficial for staging and prognosis.

This article is available for continuing medical education credit at CME.TheOncologist.com.

Abstract

The role of pelvic lymph node dissection (PLND) in both bladder and prostate cancer has recently been generating renewed interest. In an attempt to avoid PLND, both nomograms and imaging studies have been evaluated; however, so far they have shown limited success because of inadequate accuracy in staging patients. The three primary objectives of this review are: to define patients in whom PLND should be performed, to define the extent and consequences of the template for PLND, and to identify the staging and prognostic benefits seen with PLND in bladder and prostate cancer. Based on the findings from this review, we conclude that PLND for bladder cancer patients is undoubtedly beneficial, whereas it is less so for prostate cancer patients, in whom a selection strategy should be employed. PLND, in particular with an extended template, seems to provide superior accuracy for postoperative staging than the presently available imaging studies and may be pivotal when considering adjuvant therapies. Furthermore, it has an impact on survival in high-risk patients, and potentially more so in low-risk cancer patients with occult metastases.

Extended pelvic lymph node dissection, Bladder cancer, Prostate cancer

Introduction

The role of lymphadenectomy in genitourinary malignancies of the bladder and prostate has recently been generating renewed interest. In the past, pelvic lymph node dissection (PLND) was considered a purely diagnostic step in which confirmation of lymph node metastases following frozen-section analysis at the time of surgery usually precluded patients from having radical prostatectomy or cystectomy. However, recently the concept of PLND has been thrust into the limelight because of its potential for optimizing outcomes. In this review, the selection of patients for PLND, the extent to which it should be performed, as well as its role in tumor staging and associated potential therapeutic benefits are discussed. To address these objectives, we reviewed the literature with a focus on studies from centers of excellence as well as from clinicians with a special interest and substantial experience dealing with PLND in bladder and prostate cancer.

Preoperative Assessment of Pelvic Lymph Nodes

Nomograms

Nomograms are tools designed to aid physicians in decision making for therapies both pre- and postoperatively based on available risk factors. However, although well constructed, nomograms do have inherent problems. The nomogram can only be as good as the data used in its development. Most nomograms have been generated by pooling patients from different institutions retrospectively. As a result, the populations are not truly comparable; for example, they have differing staging procedures, patient selection criteria, surgical techniques for PLND, and histopathological tumor characteristics as well as a lack of lymph node analysis at the molecular level (molecular markers to identify occult disease). In addition, they are often based on a limited PLND template and therefore not representative of node sampling. Finally, nomograms may predict the likelihood of nodal metastases but are unable to provide equally vital information about the number or location of these positive nodes within the pelvis.

To date, the solitary nomogram to predict lymph node metastases at cystectomy was developed by Karakiewicz and colleagues based on clinical and pathological factors in a multicenter study [1]. Although this multivariate nomogram had a predictive accuracy of 63% for lymph node metastases, questions remain with regard to its accuracy as a result of the uncertainties surrounding the staging and PLND surgical techniques employed in this study cohort. Depending on the extent of PLND undertaken, nomograms can only predict the probability of finding a positive node within the resected template and are unable to provide the true incidence of lymph node metastases in an individual patient. It is also, for example, well known that surgical technique has a major impact on outcome in bladder cancer and therefore a nomogram based on a study population from centers of excellence may not be applicable outside these centers [2].

Similarly, Briganti et al. [3] proceeded to develop and validate a nomogram for prostate cancer that predicts the probability of lymph node metastases based on clinical parameters. With the majority of patients included in that study qualifying as low-risk patients and with only 9% (a total of 71 patients) having lymph node metastases, this rather small number limits its significance and accuracy. In addition, an extended PLND was performed in only 23% of the patients. Indeed, the knowledge that primary lymphatic drainage sites of the prostate can go up to the inferior mesenteric artery, which is generally not included in an extended PLND, may again limit the reliability of this and other nomograms [4].

In summary, because of the inherent drawbacks, nomogram prediction for lymph node metastases might be of value in a cohort of patients but not for the individual, and for a particular template but not for the whole patient.

Imaging Studies

Cross-sectional radiological imaging in the form of magnetic resonance imaging (MRI) and computed tomography (CT) scans to evaluate lymph node metastases in the pelvis is not routinely recommended, because of the low sensitivity (0%–30%) for both bladder and prostate cancer [5, 6]. Indeed, our series at the University of Bern of 507 cystectomy patients with a negative preoperative CT scan and with a median of 22 lymph nodes removed demonstrated a 24% rate of nodal disease [7].

In recent years, functional radiological staging with the sentinel lymph node concept has been applied to prostate cancer. This concept, first introduced by Cabanas in penile cancer, outlines the theory that lymphogenic cancer spread is a gradual process and usually begins with involvement of the first lymph node encountered by lymphatic flow [8]. A negative sentinel node is thought to exclude metastatic disease and it has therefore been suggested that these patients could avoid a PLND altogether. Sentinel lymph node (primary landing site) detection for prostate cancer was first introduced by Wawroschek et al. [9], who showed that the apparent sensitivity for the presence of positive nodes in a patient seemed to be 96%, with an average of four sentinel lymph nodes identified per lymph node–patient. However, this technique has its setbacks. Although with this method at least one positive sentinel lymph node can be detected in lymph node–positive patients, not all positive nodes can be found and removed. Furthermore, only nodes in close contact with the collimator are detected and therefore there is a high probability that nodal tissue not directly accessible is missed. Because it has been shown that a significant number of nodes are found outside the area explored, for example, along the internal iliac vessels, along the common iliac vessels, and in the presacral and aortic tissue, some positive nodes may have been missed in that study, resulting in incomplete sampling [4, 10, 11]. Indeed, after applying preoperative fusion imaging of single-photon emission computed tomography (SPECT) and CT scans 1 hour after intraprostatic injection of technetium-99m-nanocolloid, a median of 10 (range, 3–19) primary landing sites was detected and 30%–40% of these sites were found outside the common area of PLND [4]. Although SPECT imaging appears to be a reliable imaging technique, it is time-consuming, expensive, and dependent on the skills of the nuclear medicine specialist [12]. In addition, technetium uptake can be compromised in bulky nodal disease, in which over one third of positive nodes remain unidentified [4].

Monoclonal antibody radioimmunoscintigraphy (i.e., ProstaScint scan; Cytogen Corporation, Princeton, NJ) has had limited accuracy in lymph node metastasis detection because of the antibody specifically targeting an intracellular epitope that is only exposed in dying or dead cells [13]. Ponsky et al. [14] found that when they evaluated 22 patients preoperatively to predict nodal disease, the ProstaScint scan had an 89% false-positive rate and yielded a sensitivity of 17%, specificity of 90%, negative predictive value of 94%, and positive predictive value of 11%. They concluded, based on these results, that ProstaScint overestimates nodal metastases and is not useful for their prediction preoperatively.

Other new and potentially promising techniques such as high-resolution MRI with lymphotrophic superparamagnetic nanoparticles are being developed and evaluated; however, detection of micrometastases remains limited and these techniques are, as yet, not available [15]. The inability to accurately diagnose smaller lymph node metastases preoperatively with imaging modalities and the important prognostic information gained by identifying nodal disease emphasize the role of PLND.

Patient Selection for PLND

It is well known that the more progressed the disease is, the higher the risk for lymph node metastases. For this reason, attempts are made to identify patients at low risk, in which PLND can be avoided.

In bladder cancer, Palapattu et al. [16] and Shariat et al. [17] showed that even patients with pT0 and Cis disease following cystectomy had a 3% rate of lymph node metastases, with 5-year bladder cancer–specific survival estimates of 95% and 90.7%, respectively. Therefore, although the risk for lymph node metastases is assumed to be low in patients with nonmuscle invasive bladder cancer requiring cystectomy, these results have demonstrated a potential benefit of PLND even in these low-risk patients. In fact, these may be the patients who benefit most from PLND.

As a result of prostate-specific antigen (PSA)-based screening, prostate cancer has experienced a downward shift in the pathologic stage and the incidence of nodal metastases at the time of radical prostatectomy. However, it has been demonstrated that the risk for lymph node metastases even in patients with low-risk parameters varies in the range of approximately 3%–10% after extended PLND [18–22]. Schumacher et al. [21] found that, in patients with a PSA level <10 ng/ml, a 25% incidence of lymph node metastases was seen in those with a specimen Gleason score ≥7, but only 3% of patients with a specimen Gleason score ≤6 were node positive. Using the extended template, Weckermann et al. [22] similarly found the nodal metastasis risk to be in the range of 3%–5% in their series of patients with a PSA level ≤10 ng/ml and biopsy Gleason score ≤6, whereas Heidenreich and coworkers [23] reported a 2.4% incidence of lymph node metastases in their patients with a PSA level <10.5 ng/ml and biopsy Gleason score ≤6. In this context, it is of significance that Grossfeld and colleagues [24] found a 30% rate of undergrading and understaging in patients with a preoperative biopsy Gleason score ≤6 whereas Bhatta-Dhar et al. [25] found, in their study, that 40% had to be upgraded to a Gleason score >6. Therefore, although it may be permissible to refrain from performing PLND in low-risk patients with a PSA level <10 ng/ml and a biopsy Gleason score ≤6, there is still the inherent risk of biopsy-based understaging.

In summary, PLND should be performed in all bladder cancer patients requiring a cystectomy, and for prostate cancer patients it should be performed in patients requiring a prostatectomy and men with a PSA level >10 ng/ml as well as those with a PSA level ≤10 ng/ml but a Gleason score ≥7.

Extent of PLND

From the current literature, standardized universal definitions for PLND in bladder and prostate cancer are still lacking. For patients undergoing cystectomy for bladder cancer, our recommendation is for a PLND encompassing all of the lymphatic and connective tissue within the following boundaries: caudally, the deep circumflex iliac vein and femoral canal; medially, the side wall of the bladder; laterally, up to the genitofemoral nerve; posteriorly, the obturator muscle and floor of the obturator fossa down to the internal iliac vessels, where all tissue adjacent to these vessels (also termed presacral nodes) is removed; and cranially, the midcommon iliac region where the ureter crosses the common iliac artery. Our “extended” template is indicated by the triangular shaded area in Figure 1[26, 27].

Figure 1.

Diagrammatic representation of the authors' recommended triangular shaded field of extended pelvic lymph node dissection during radical cystectomy from the point where the ureter crosses the iliac vessels down. The oval shaded area represents an extension to this template whereby dissection clears tissue up to the aortic bifurcation. Primary landing sites in this oval area are likely to be rare; therefore, the clinical benefit of removing this additional nodal tissue is uncertain. Adapted from Fleischmann A, Thalmann GN, Markwalder R et al. Extracapsular extension of pelvic lymph node metastases from urothelial carcinoma of the bladder is an independent prognostic factor. J Clin Oncol 2005;23:2358–2365, with permission.

With this template, based on a series of 507 consecutive patients from our institution, a median of 22 lymph nodes (range, 10–43) was obtained [27]. In contrast, whereas Stein et al.'s [28] extended PLND recommendation includes tissue cranially up to the aortic bifurcation (oval shaded area in Fig. 1), others have proposed an even more extensive template to encompass all tissue up to the inferior mesenteric artery [27–31]. By extending the cranial border above the aortic bifurcation, not surprisingly, Vazina et al. [30] removed a median of 25 lymph nodes (range, 2–80), whereas removing lymphatics as far up as the inferior mesenteric artery enabled Leissner et al. [31] to obtain a mean of 43 lymph nodes. Of importance, close to two thirds of patients with positive lymph nodes have been noted to have these either exclusively along the internal iliac vessels or in combination with the obturator fossae and along the external iliac vessels [30]. These findings therefore stress the need for a PLND template to at least include tissue along the internal and external iliac vessels up to and around the common iliac vessels.

In 2004, the most comprehensive bladder cancer lymph node metastasis mapping study to date was reported by Ghoneim's group [32]. In that study, patients with one or two lymph node metastases had them within the endopelvic region, whereas involvement of extrapelvic nodes (including nodes around the aortic bifurcation) was invariably associated with not just multiple but also endopelvic nodal disease, thereby demonstrating the unlikely occurrence of skip lesions. Extending the dissection up to the aortic bifurcation increased the clearance of positive nodes by 13% in that study, whereas Leissner et al. [31] similarly demonstrated a 16% incidence of nodal involvement above the aortic bifurcation. Although more nodes would certainly be removed based on larger templates, in the absence of a definite survival benefit in this group of patients with a poor prognosis, a longer operating time and a higher risk for lymphocele formation and lymphedema, as well as disruption to the sympathetic nerves and subsequent incontinence, extending the template as yet does not seem justifiable.

Previous studies have, however, unanimously supported the need for bilateral PLND—as seen with the studies by Abol-Enein et al. [32], in which close to half of the patients with lymph node metastases had bilateral involvement and >40% of these cases with lateralized bladder cancers demonstrated contralateral nodal disease, and Leissner et al. [31], as well as our series from Bern, which identified that crossed pelvic involvement of lymph nodes on the contralateral side of the pelvis from a lateralized bladder cancer was not uncommon [31–33]. A summary of previous series from various high-volume centers detailing the extent of lymphadenectomy and their pathological findings is outlined in Table 1 [27, 28, 30, 31, 34–40].

Table 1.

Survival of bladder cancer patients with nodal disease treated by radical cystectomy and pelvic lymph node dissection (PLND)

To summarize, although it is commonly accepted that a larger PLND template extending up to the aortic bifurcation and beyond would almost certainly increase the likelihood of removing more involved lymph nodes, because of the risks and lack of definite long-term benefit, our recommendation remains a standardized bilateral extended PLND involving the removal of tissue along either side of the internal iliac vessels, laterally to the genitofemoral nerve, as well as cranially up to the ureteric crossing with the common iliac artery, being especially cautious not to damage the hypogastric nerves to preserve optimum functional outcomes.

For prostate cancer, although the minimal template only includes lymphatic tissue from the obturator fossa, a limited/standard PLND template extends proximally from the bifurcation of the common iliac vessels to the circumflex iliac vein distally and extends laterally to include lymphatic tissue along the external iliac vein. If it is to be performed, our recommendation is for an extended PLND in which the boundaries are as follows: cranially from the midcommon iliac region where the ureter crosses the iliac vessels, laterally to the external iliac vein, caudally to the deep circumflex iliac vein and femoral canal, medially to the bladder wall, and posteriorly to the pelvic wall where, similar to cystectomy patients, skeletonization of tissue should occur medial and lateral to the internal iliac vessels (presacral nodes), as schematically shown in Figure 2 [41]. Using this template in a consecutive series of 602 prostate cancer patients, our institution documented the removal of a median of 22 lymph nodes (range, 10–75), in which positive internal iliac lymph nodes alone or in combination with positive nodes in other locations were found in 70% of these men [42].

Figure 2.

Anatomical extent for our recommended extended pelvic lymph node dissection during radical prostatectomy whereby, in addition to lymphatic tissue removed from the obturator fossa and the external iliac vein, the template also encompasses tissue on the medial and lateral aspects of the internal iliac vessels as well as along the common iliac vessels up to the ureteric crossing. Adapted from Burkhard F, Schumacher M, Studer U. Lymph node dissection. In: Kirby RS, Montorsi F, Smith JS et al., eds. Radical Prostatectomy: From Open to Robotic. London; Informa Healthcare, 2007:81–91, with permission.

The rationale for extended PLND in prostate cancer stems from results comparing different templates for PLND. Heidenreich et al. [23] compared 103 patients who had an extended PLND (external, internal, and common iliac, obturator nodes, presacral nodes) with 100 men from an historical series who had received standard PLND. Lymph node metastasis was diagnosed in 27% of patients in the extended PLND group compared with only 12% of patients in the standard group, and in addition, despite negative obturator nodes, nodal metastases were identified in the internal iliac and presacral regions. With 42% of all lymph node metastases detected outside the standard template, this suggests that more important than the absolute number of nodes is the field from which these nodes are derived, as has been demonstrated by others [10]. Confirming this finding, Wawroschek and colleagues, by extending the standard PLND template to include tissue along the internal iliac vessels as well as the paravesical, presacral, and pararectal tissue, detected an additional 35% of lymph node–positive patients [43]. The importance of completely excising all tissue around the internal iliac vessels is further reinforced by findings from Bader et al [44], in which over half (58%) of the positive nodes identified were found along the internal iliac artery and 19% were exclusively at this location. A summary of lymph node metastasis detection rates based on variable templates is shown in Table 2 [20, 23, 44, 45].

Table 2.

Lymph node metastasis detection rates based on pelvic lymph node dissection (PLND) templates

Recent data using intraprostatic technetium injections and fused SPECT combined with either CT or MRI to map lymphatic drainage from the prostate have also shown that the prostate has multiple landing sites and that if PLND is limited to the iliac vessels distal to the bifurcation of the common iliac artery, 30%–40% of the lymphatic landing sites are potentially missed (Fig. 3) [4, 12]. By including tissue along the common iliac vessels at least up to the ureteric crossing, approximately 75% of all nodes potentially harboring metastases can be removed.

Figure 3.

Lymphatic drainage map based on intraprostatic technetium injections and fused single-photon emission tomography combined with either computed tomography or magnetic resonance imaging demonstrating multiple landing sites from the prostate, with 30%–40% of these sites located proximal to the bifurcation of the common iliac artery. Reprinted from Mattei A, Fuechsel FG, Bhatta Dhar N et al. The template of the primary lymphatic landing sites of the prostate should be revisited: Results of a multimodality mapping study. Eur Urol 2008;53:118–125, copyright 2008, with permission from Elsevier.

Therefore, based on the current evidence, we recommend, for prostate cancer patients, a standardized bilateral extended PLND with the cranial extent up to the ureteric crossing, the lateral extent to the external iliac vein, and the posterior extent to include tissue on both sides of the internal iliac vessels.

PLND Influence on Staging and Survival

PLND provides staging information that may be used to give patients a better idea of prognosis as well as to help select patients who may benefit from additional treatment. The accuracy of staging based on the number of lymph nodes removed appears to relate to not only the completeness of lymph node dissection but also interindividual variation [46]. PLND studies on bladder and prostate cancers have clearly demonstrated a significant correlation between the number of lymph nodes removed and the presence of lymph node metastases [18, 20, 23, 47]. Although the minimum number of lymph nodes required for optimal staging still remains unknown, because of the interindividual variation in lymph node count, it is probably more important to clear a certain template (extended PLND) and submit the lymph nodes in separate packets for histological assessment than to focus only on the number of lymph nodes retrieved. In addition, various prognostic factors characterized only through PLND may also improve risk stratification during staging and better direct the need for adjuvant treatment. These factors, which are highlighted later, represent the volume of metastatic disease and biological aggressiveness and include the number and volume of positive nodes, extranodal growth, the ratio of positive nodes to total number of nodes removed, and molecular occult metastases.

Presently, although there are as yet no randomized trials that have demonstrated differences in survival with or without lymphadenectomy, there is strong evidence supporting some form of lymph node removal. Previously, high rates of local recurrence were seen even in the presence of organ-confined primary tumors and this was probably related to unresected nodal disease.

Bladder Cancer

Kerr and Colby first instigated the potential benefit of a PLND in 1950 [48]. They reported a 2-year survival duration for two patients following cystectomy and PLND despite nodal disease. Although the practice of PLND gradually gained acceptance, it was not until Skinner's data in 1982 that PLND became firmly established as a standard treatment in bladder cancer [49]. Skinner indicated that a “meticulous” PLND could provide cure and control of pelvic disease in some patients with regional lymph node metastases without increasing morbidity. Prior to the practice of PLND, earlier series reported dismal 5-year survival rates of 4%–7% in lymph node–positive patients [34, 35]. However, since Skinner's findings, more contemporary data originating from centers including the University of Southern California in Los Angeles, the Memorial Sloan Kettering Cancer Center in New York, and the University of Bern have confirmed that, with extended PLND, approximately one third (31%–35%) of bladder cancer patients with lymph node metastases are alive at 5 years [28, 36, 38, 40, 50]. In the largest cystectomy series to date by Stein and colleagues (1,054 patients with extended PLND), of the 246 (23%) patients who were node positive, 31% were found to be alive at 5 years [28]. Excellent long-term data from these series therefore confirm the feasibility and reproducibility of extended PLND and its survival benefit even in patients with regional lymph node metastases. A summary table of 5-year outcomes is shown in Table 1.

Findings from previous bladder cancer studies have demonstrated unanimously that greater nodal burden results in a more adverse survival outcome [33, 51]. Skinner's group was one of the first to report a higher risk for cancer-related progression and death associated with six or more lymph node metastases identified at cystectomy [51]. Similarly, our group in Bern also demonstrated that patients with fewer than five involved nodes had a statistically significant survival advantage over those with five or more involved nodes [33]. In addition, we also demonstrated that the size of the nodal tumor deposit from bladder cancer has a significant effect on survival. Patients with a maximal nodal metastasis ≤0.5 cm had a median survival duration of 84 months, compared with only 16 months in those with a metastasis >0.5 cm.

More recently, we also demonstrated that lymph node capsular perforation by metastatic tumor is associated with a worse survival prognosis [27]. In that study, although the tumor stage, number of positive nodes (fewer than five versus five or more), and percentage of nodes containing metastatic disease (<20% versus ≥20%) all had a significant influence on recurrence-free survival, evidence of extracapsular lymph node extension was the strongest and the only independent prognostic factor for recurrence-free survival on multivariate analysis. Patients with extracapsular disease had a median recurrence-free survival duration of 12 months, compared with 60 months for their counterparts without extracapsular invasion.

As seen with our results, Herr was the first to demonstrate that the ratio of the number of positive to total number of lymph nodes removed (lymph node density), better defined surgical outcome than conventional lymph node staging and was a significant prognosticator for disease-specific survival and local recurrence [52]. Soon after, Stein and colleagues from the University of Southern California were able to confirm this finding, whereby patients with a ≤20% lymph node metastasis density had a 43% 10-year recurrence-free survival rate, compared with a 17% 10-year survival rate for those with a >20% density [53]. The prognostic validity of lymph node density is superior to that of the absolute number because it probably represents, more importantly, the extent/quality of surgical dissection, and therefore the likelihood of positive nodes left behind. The risk for residual lymph node metastases is greater if one or two positive nodes are identified in five rather than 50 removed nodes. This, however, only stands if nodal dissection is undertaken using an appropriate template.

Significantly, Stein et al. [53] also reported that patients who had removal of >15 lymph nodes had a 10-year recurrence-free survival rate of 36%, compared with 25% in patients who had ≤15 lymph nodes removed. In other words, not only are lymph node metastases influential to survival outcome but the total number of lymph nodes dissected at the time of PLND is also relevant. Herr reported that survival following cystectomy for patients with both lymph node–negative and lymph node–positive disease was better (lower local recurrence rate) when a higher number of nodes was removed [54]. Similarly, in a study by Leissner et al. [55], it was demonstrated that, regardless of nodal metastases, 65% of patients who had removal of >16 lymph nodes were disease free and alive at 5 years, compared with 51% of those who had <16 lymph nodes resected.

Therefore, even patients with histologically negative lymph nodes seem to benefit from PLND in the long term. This may be because many of these patients may, in fact, harbor occult metastatic disease at a molecular level. Copp et al. [56], in their study on molecular staging of bladder cancer, analyzed uroplakin II mRNA as a molecular marker to detect occult bladder cancer cells following PLND. That group demonstrated that 46% of lymph nodes that were positive for this molecular marker were pathologically node negative. Noteworthy, 5% of the pathological and molecular lymph node–negative patients suffered disease recurrence, in comparison with 91% of those who were pathologically node negative but node positive based on the molecular assessment. Finally, although results from the European Organization for Research and Treatment of Cancer trial are pending, the consequence of accurately identifying either pathologically proven or even occult lymph node metastases is highlighted by some evidence suggesting that adjuvant/salvage chemotherapy may provide slight survival benefit in patients with nodal disease following cystectomy [57, 58].

In summary, aside from more accurate staging and stratification for consideration of adjuvant therapy, radical surgery with PLND has been shown to cure up to approximately 30% of patients with node-positive disease, where most survivors only have one or two microscopically involved positive lymph nodes rather than grossly positive or multiple nodal involvement. It is for these reasons and the limitations of intraoperative lymph node frozen-section analysis that all cystectomy patients should directly undergo an extended PLND, particularly those without gross nodal disease who may have occult metastases and for whom there may potentially be a greater benefit in their long-term prognosis.

Prostate Cancer

Initially, in 1987, Golimbu and coworkers, when analyzing 42 patients with occult nodal disease following PLND, discovered that those with low tumor bulk and a single lymph node metastasis had a higher survival rate than their matched control counterparts after a mean follow-up of 5 years [59]. In 1988, Catalona et al. [60] showed that treatment could be curative even in lymph node–positive disease. In a small series of 12 men with lymph node metastases without adjuvant therapy, 75% remained recurrence free at 5 years and 58% were recurrence free at 7 years. Similarly, Pound et al. [61], from their series, confirmed a 68% 10-year metastasis-free survival rate in patients with lymph node micrometastasis who were managed without adjuvant therapy.

Joslyn et al. [18], as seen in bladder cancer, showed that the presence of two or more prostate cancer lymph node metastases conferred a greater risk for cancer-related death than a solitary metastasis. Recently, Boorjian and colleagues retrospectively reviewed a large series of patients from the Mayo Clinic with positive lymph nodes with or without adjuvant hormonal therapy, in which 56% and 86% of the lymph node–positive patients had a 10-year biochemical progression-free survival and cancer specific survival duration, respectively [62]. They also demonstrated that radical prostatectomy patients with a single nodal metastasis were at a threefold higher risk for systemic progression and fourfold higher risk for prostate cancer death, but in comparison, those with two or more lymph node metastases were twice as likely to experience systemic progression as well as prostate cancer death. Similarly, Skinner's group confirmed a 10-year recurrence-free survival rate following a radical prostatectomy of 70%–73% in patients with one or two lymph node metastases, but only a 49% rate in those with five or more involved nodes [63]. In addition, when pathological specimens were reviewed, Griebling et al. [64] found that the 5-year overall survival rate for patients with extracapsular nodal extension was 55%, compared with 71% for those with histologically confined prostatic nodal metastases.

In terms of the lymph node density status, an advantage in PSA progression-free survival was reported in men with <15% compared with those with ≥15% involved nodes [20]. In addition, a <15% density in patients conferred a 5-year PSA progression-free survival rate of 43% following an extended PLND, compared with only 10% after a limited PLND. Daneshmand et al. [63] also identified that patients with a ≤20% lymph node density had a 72% 10-year recurrence-free survival rate, compared with a 47% 10-year survival rate for those with >20% density.

Similar to what was seen in bladder cancer studies, the group from the Memorial Sloan-Kettering Cancer Center found that, in men without nodal involvement, a higher number of nodes removed correlated significantly with freedom from biochemical recurrence and survival rates [65, 66]. It was shown that, even by removing ≥11 negative lymph nodes, the 5-year survival rate was approximately 50%, compared with the 20% rate seen in those who had <11 negative nodes dissected. Based on the analysis of the Surveillance, Epidemiology, and End Results program database in 1,923 patients, Josyln and colleagues also concluded that an extended PLND reduced the long-term risk for prostate cancer–related death, even in patients with negative nodes, compared with those who had no PLND [18]. Here again, the benefit of removing negative nodes is possibly related to the presence of occult lymph node metastases. When Miyake et al. [67], Ferrari et al. [68], and Pagliarulo et al. [69] studied molecular markers of micrometastasis for prostate cancer using real-time reverse transcriptase polymerase chain reaction (RT-PCR) and immunostaining, all three studies found that the presence as well as the load of these markers was significantly associated with a higher rate of cancer recurrence and shorter survival. These studies have re-emphasized the potential benefit of PLND for occult disease during radical prostatectomy. It is in such patients that leaving residual tumor would deprive them of the opportunity for potentially curative treatment.

Although it is unlikely that removing all positive nodes will cure all men with high-risk disease, it is possible that progression-free survival can be prolonged when radical surgery and adjuvant therapy are combined [70]. Messing et al. [71] demonstrated, in their study, that early adjuvant androgen-deprivation therapy following radical prostatectomy in men with node-positive disease is associated with longer survival. Should we, therefore, request molecular assessment of lymph nodes in all our radical prostatectomy patients and initiate androgen deprivation? Probably not, since the Messing et al. [71] trial was based on patients with mostly gross nodal disease, the majority of whom also had positive margins and seminal vesicle infiltration. Its results, therefore, may not necessarily apply to patients with micrometastatic lymph nodes detected by RT-PCR or immunostaining, or to patients with minimal metastatic disease. Even though the majority of the node-positive patients in the study by Boorjian et al. [62] received androgen-deprivation therapy, patients with a single positive node were not at a higher risk for local recurrence than node-negative patients. Furthermore, in our study of 122 lymph node–positive patients who underwent an extended PLND, two thirds of the patients with a solitary positive lymph node and half of those with two positive nodes experienced no evident tumor progression and therefore did not require androgen-deprivation therapy during a follow-up period >10 years [42]. When androgen deprivation should be initiated, therefore, remains unclear. Because a postoperative PSA doubling time <12 months is a negative prognostic factor for survival in prostate cancer, these patients may well be the group who are most likely to benefit from immediate androgen-deprivation therapy [72–74].

To summarize, PLND with an extended template will remove more nodes and improve the accuracy of staging, although even our extended template may be inadequate in some patients because approximately 25% of the primary lymphatic landing sites of the prostate lie outside this template [4]. However, there must be optimization between benefits and risks when defining the extent of PLND. Increasing evidence also points to the potential for the removal of occult molecular lymph node metastases with an extended PLND. Because patients with minimal nodal involvement can have a good prognosis, we recommend that no frozen sections be performed and the radical prostatectomy be continued independently of the intraoperative lymph node status. In patients with palpable and multiple nodal disease, however, it still remains to be seen if PLND, whether extended or not, represents not just tumor mass reductive, but also curative, surgery.

Complications of PLND

Opponents of PLND argue that performing a PLND, particularly using the extended template, may result in higher costs and greater morbidity [75]. Although completing an extended PLND could add, on average, 30–45 minutes to the total operating time, as of yet there are still no reports on a cost–benefit analysis for PLND in bladder and prostate cancers. Previous reports have shown complication rates for limited/standard PLND in the range of 2%–9.8%, whereas extended PLND complication rates varied in the range of 19.8%–75% [8–10]. As expected, the most frequently reported complication is lymphocele formation, with rates in the range of 4.6%–10.6% [3, 10]. As would be expected, performance at high-volume centers of excellence as well as a higher level of surgical experience usually results in lower complication rates. To reduce or to prevent PLND-associated morbidity, several authors have provided advice regarding meticulous surgical techniques. First, instead of using clips, which are often turned away during subsequent surgery, all the lymphatic vessels coming from the lower extremities are tied using ligatures. Second, all lymphatics lateral to the external artery are saved. Third, two drains are placed in each side of the pelvis and are not removed until the total amount of fluid is <50 ml in 24 hours. Fourth, low molecular weight heparin is injected into the upper arm to avoid damaging lower extremity lymphatics.

Conclusions

Presently, because of the inherent drawbacks of nomograms and the limited staging accuracy of imaging modalities, PLND remains the most reliable option for identifying lymph node metastases. If PLND is performed, then an extended template should be undertaken to include tissue cranially up to the crossing of the ureter with the iliac vessels, posteriorly where the tissue adjacent to either side of the internal iliac vessels is excised, and laterally up to the external iliac vein in prostate cancer and up to the genitofemoral nerve in bladder cancer. In doing so, staging is optimized and, with risk stratification, patients potentially may be more appropriately considered for adjuvant therapy. Indeed, removing histopathologically negative nodes may perhaps be most beneficial to patients because such nodes may, in fact, harbor metastases at a molecular level. Therefore, with a reduced metastatic load as a result of PLND, the prognosis may be optimized in bladder and prostate cancer patients.

Author Contributions

Conception/design: Fiona C. Burkhard, Ramesh Thurairaja, Urs E. Studer

Manuscript writing: Fiona C. Burkhard, Ramesh Thurairaja, Urs E. Studer

Author notes

Disclosures

Ramesh Thurairaja: None; Urs E. Studer: None; Fiona C. Burkhard:  Consultant/advisory role: Medtronic

Section editors Chris Parker and Matthew R. Smith have disclosed no financial relationships relevant to the content of this article.

The content of this article has been reviewed by independent peer reviewers to ensure that it is balanced, objective, and free from commercial bias.

Target audience: Physicians who wish to advance their current knowledge of clinical cancer medicine in genitourinary cancer.