SEARCH

SEARCH BY CITATION

Keywords:

  • renal cell carcinoma;
  • lymph node dissection;
  • lymph node invasion;
  • extended lymphadenectomy

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

What's known on the subject? and What does the study add?

  • In renal cell carcinoma the role of lymphadenectomy (LND) is still controversial. Moreover, no firm consensus exists regarding the minimum number of lymph nodes that should be removed to obtain a satisfactory staging LND at the time of surgery.
  • Our findings demonstrate that, when clinically indicated, staging LND in renal cell carcinoma should be extended. The removal of 15 lymph nodes might represent the lowest threshold to define a staging LND as adequate. More extended LND should be pursued if unfavourable clinical and pathological characteristics are evident at diagnosis and/or during surgery.

Objective

  • To investigate the staging of lymphadenectomy in renal cell carcinoma. No convincing data exist regarding the minimum number of lymph nodes that should be removed at the time of nephrectomy to ensure an accurate staging.

Methods

  • Between 1987 and 2011, 850 patients with renal cell carcinoma underwent either partial or radical nephrectomy plus lymph node dissection (LND) at a single tertiary care institution (TanyN0–1Many). Receiver operating characteristic curve coordinates were used to graph the probability of finding lymph node invasion according to the number of removed lymph nodes. Assuming that the likelihood of finding lymph node invasion according to the number of lymph nodes removed may be affected by patient characteristics, analyses were further stratified for clinical and pathological characteristics.

Results

  • The rate of lymph node metastases strongly correlated with the clinical and pathological characteristics of the patients. Fifteen lymph nodes need to be removed to achieve a 90% probability of detecting at least one metastatic lymph node. Only slight differences were recorded after stratification for clinical nodal status, the presence of metastases at diagnosis and pathological T stage. Finally, 13, 16 and 21 lymph nodes need to be removed to achieve a 90% probability of detecting lymph node invasion, if present, in the low risk (score 0–1), intermediate risk (score 2–3) and high risk (score 4–5) Mayo Clinic classification, respectively.

Conclusion

  • The removal of 15 lymph nodes represents the lowest threshold for considering a staging LND as adequate. More extended LND should be pursued if unfavourable clinical and pathological characteristics are evident at diagnosis and/or during surgery.

Abbreviations
RCC

renal cell carcinoma

LND

lymphadenectomy

LN

lymph node

LNI

lymph node invasion

ROC

receiver operating characteristic

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

In renal cell carcinoma (RCC), the role of lymphadenectomy (LND) at the time of nephrectomy is still controversial [1]. Although further efforts are needed to clarify whether LND may affect cancer control, progression and survival, an adequate LND maintains its fundamental aim in correctly staging the disease. LND represents the most accurate staging procedure in the RCC setting, due to the limited capacity of preoperative imaging and statistical predictive models to identify patients with nodal metastases [1].

In this context, LND provides knowledge of the real status of regional lymph nodes (LNs) and consequently an accurate prognosis; it also permits the programming of an adequate follow-up and eventually, if there are metastases, the consideration of possible further treatment. However, no firm consensus exists regarding the minimum number of LNs that should be removed to obtain a satisfactory LND at the time of surgery. We decided to fill this gap by identifying the likelihood of finding one or more positive LNs according to the number of LNs removed at radical surgery for RCC.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

Study Population

The current study was undertaken with the approval and institutional oversight of the institutional ethics committee review board. For each patient, comprehensive clinical and pathological data were collected and entered into a computerized prospective database. Patient records were retrieved and yielded 1849 patients with RCC treated with partial or radical nephrectomy between 1987 and 2011 at a single tertiary care institution. For the aim of the study, only patients submitted to nephrectomy plus LND were considered (n = 850; TanyN0–1Many RCC included).

Clinical and Pathological Evaluation

TNM stages were assigned according to the 2009 American Joint Committee on Cancer/Union Internationale Contre le Cancer classification [1]. Cases before the introduction of the most updated classification were reclassified. Tumour size definition was based on the largest tumour diameter in centimetres derived from computed tomography. Dedicated genitourinary pathologists examined all surgical specimens. Sarcomatoid and tumour necrosis features were systematically gathered in the last decade. Older cases were retrospectively reviewed. All removed LNs were examined for the presence of nodal metastases. Lymph node invasion (LNI) was invariably defined as one or more metastatic LNs.

The Surgical Template

A regional LND (hilar region plus, on the right side, paracaval nodes from the adrenal vein to the level of the inferior mesenteric artery or, on the left side, para-aortic nodes) was systematically performed. In 203 of 850 patients (23.9%), an extended LND was performed according to the preference of the operating surgeon. In those cases, extended LND included, on the left, para-aortic and pre-aortic nodes from the crus of the diaphragm to the inferior mesenteric artery and, on the right, paracaval, retrocaval and precaval nodes from the adrenal vein to the level of the inferior mesenteric artery. Interaortocaval nodes were always removed when an extended LND was sought.

Statistical Analyses

The relationship between the number of removed and examined LNs and the probability of finding one or more metastatic LNs was examined in receiver operating characteristic (ROC) analyses. Specifically, the ROC curve coordinates were used to define the probability of LNI, according to the number of removed and examined LNs. The ROC curve coordinates were then used to graph the probability of finding LNI according to the number of removed and examined LNs.

Assuming that the likelihood of finding LNI according to the number of LNs may be affected by patients' characteristics, we tested our hypothesis in different clinical and pathological scenarios. Sensitivity analyses were further stratified for clinical nodal status, presence of metastases at diagnosis, pathological T stage and LNI risk according to the Mayo Clinic LNI risk score as previously described by Blute et al. [2] and Crispen et al. [3]. Statistical tests were performed using SPSS v. 19 (IBM Corp., Somers, NY, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

Table 1 shows the descriptive characteristics of the entire cohort (n = 850). Among the patients who were candidates for nephrectomy plus LND, the LNI prevalence was 14.6% (124 of 850). The rate of LN metastases strongly correlated with the clinical and pathological characteristics of the patients (Table 1). LNI prevalence was 50.8% vs 4.5% in patients with or without suspicious nodal metastases respectively at preoperative imaging, and 33.3% vs 9.7% in patients with or without metastases at presentation (all P < 0.001). The rate of LN metastases significantly correlated with pathological T stage, differentiation of cancer cells and tumour size. LNI prevalence in particular gradually increased with increase of pathological T stage, higher Fuhrman grade and larger tumours (Table 1; all P < 0.001). Age at surgery was not statistically significantly different between patients with or without pathologically confirmed nodal metastases (P = 0.4).

Table 1. Clinical and pathological characteristics of patients treated with partial/radical nephrectomy and lymphadenectomy; data were further stratified according to lymph node status
 Overall (n = 850)pN0 (n = 726)pN1 (n = 124)P
Clinical characteristics    
Mean age, years (median)60.6 (61)60.5 (61)61.5 (62)0.4
Range24–8924–8929–86
Clinical N status (at preoperative imaging)   <0.001
cN0666 (78.4%)636/666 (95.5%)30/666 (4.5%)
cN1183 (21.5%)90/183 (49.2%)93/183 (50.8%)
Clinical M status (at preoperative imaging)   <0.001
cM0673 (79.2%)608/673 (90.3%)65/673 (9.7%)
cM1177 (20.8%)118/177 (66.7%)59/177 (33.3%)
     
Pathological characteristics    
pT   <0.001
1a122 (14.4%)119/122 (97.5%)3/122 (2.5%)
1b192 (22.6%)188/192 (97.9%)4/192 (2.1%)
2a96 (11.3%)89/96 (92.7%)7/96 (7.3%)
2b46 (5.4%)39/46 (84.8%)7/46 (15.2%)
3a276 (32.5%)216/276 (78.3%)60/276 (21.7%)
3b49 (5.8%)41/49 (83.7%)8/49 (16.3%)
3c32 (3.8%)17/32 (53.1%)15/32 (46.9%)
436 (4.2%)17/36 (47.2%)19/36 (52.8%)
Fuhrman   <0.001
188 (10.4%)87/88 (98.9%)1/88 (1.1%)
2405 (47.6%)382/405 (94.3%)23/405 (5.7%)
3269 (31.6%)201/269 (74.7%)68/269 (25.3%)
486 (10.1%)55/86 (64.0%)31/86 (36.0%)
Mean tumour size, cm3 (median)7.6 (7.0)7.2 (6.5)10.0 (9.6)<0.001
Range0.7–23.00.7–21.00.9–23.0
Mayo Clinic LNI risk score   <0.001
0249 (29.3%)246/249 (98.8%)3/249 (1.2%)
1185 (21.8%)172/185 (93.0%)13/185 (7.0%)
2142 (16.7%)128/142 (90.1%)14/142 (9.9%)
3146 (17.2%)103/146 (70.5%)43/146 (29.5%)
4108 (12.7%)67/108 (62.0%)41/108 (38.0%)
520 (2.4%)10/20 (50.0%)10/20 (50.0%)
Mean number of nodes removed (median)7.8 (6)7.1 (6)12.1 (9)<0.001
Range1–581–391–58

The mean number of nodes removed was 7.8 (median 6) in the overall population. The number of nodes removed in patients with pathologically confirmed nodal metastases was higher relative to their pN0 counterparts (12.1 vs 7.1; P < 0.001). Mean LN density was 7.4% and 50.6% in the overall population and pN1 cases, respectively. Regarding those patients who received a systematic extended LND (n = 203), 62 (30.5%) showed LNI. Of those, 40 (64.5%) harboured lymph node metastases in landing sites outside the ipsilateral hilar boundaries (e.g. interaortocaval or contralateral retroperitoneal nodes). Moreover, eight out of 62 patients (12.9%) showed metastases solely outside the ipsilateral hilar boundaries and they would have been understaged as pN0 if a limited LND had been performed.

ROC curve analyses were used to explore graphically the relationship between the number of removed and examined LNs and the probability of finding one or more metastatic LN in the overall population (n = 850; Fig. 1). The curve indicated that 15 and 19 LNs need to be removed and examined to achieve a 90% and 95% probability respectively of detecting one or more metastatic LNs,. The maximal increase in the probability of finding positive LNs occurred when 5–12 LNs were removed (from 40% to 80%, respectively). Substantially smaller gains were achieved between 12 and 19 LNs (from 80% to 95%). Virtually no gain was recorded with the removal of >20 LNs.

figure

Figure 1. ROC-derived plot depicting the relationship between the number of lymph nodes removed and the probability of finding one or more positive lymph nodes (overall population, n = 850).

Download figure to PowerPoint

When the analyses were stratified according to clinical and pathological characteristics, only slight differences were recorded among the sensitivity analyses stratified for clinical nodal status (Fig. 2a,b), the presence of metastases at diagnosis (Fig. 3a,b) and pathological T stage (Fig. 4a–c). Specifically, if unfavourable characteristics are evident (cN1 or cM1 or pT3–4), a few more LNs (from two to four) need to be removed to achieve the same probability of detecting LNI of patients without unfavourable features.

figure

Figure 2. ROC-derived plot depicting the relationship between the number of lymph nodes removed and the probability of finding one or more positive lymph nodes according to the presence of suspected lymph node metastasis at preoperative imaging: a, green curve, n = 666, cN0; b, red curve, n = 183, cN1.

Download figure to PowerPoint

figure

Figure 3. ROC-derived plot depicting the relationship between the number of lymph nodes removed and the probability of finding one or more positive lymph nodes according to the presence of metastases at presentation: a, green curve, n = 673, cM0; b, red curve, n = 177, cM1.

Download figure to PowerPoint

figure

Figure 4. ROC-derived plot depicting the relationship between the number of lymph nodes removed and the probability of finding one or more positive lymph nodes according to pathological T stage: a, green curve, n = 314, pT1; b, yellow curve, n = 142, pT2; c, red curve, n = 393, pT3-4.

Download figure to PowerPoint

Finally, when the analyses were stratified according to the Mayo Clinic LNI risk classification, additional meaningful differences were recorded (Fig. 5a–c). The curves indicated that 13, 16 and 21 LNs need to be removed and examined to achieve a 90% probability of detecting LNI in low risk (score 0–1), intermediate risk (score 2–3) and high risk (score 4–5) patients, respectively.

figure

Figure 5. ROC-derived plot depicting the relationship between the number of lymph nodes removed and the probability of finding one or more positive lymph nodes according to Mayo Clinic LNI risk classification: a, green curve, n = 185, risk score 1; b, yellow curve, n = 142, risk score 2; c, red curve, n = 108, risk score 4.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

Relative to other urological settings, such as bladder and prostate cancer [4-6], the effect of LND and its extension on progression and cancer-specific survival has, to date, not been adequately addressed [7]. The issue of whether LND might affect survival after nephrectomy for RCC led to the European Organization for Research and Treatment of Cancer prospective trial in 1988 [8], which failed to demonstrate a significant difference between patients treated with nephrectomy plus LND vs nephrectomy alone. However, although the report represents the only high level of evidence data addressing the effect of LND on survival, most patients included in the trial were clinically T1–T2 N0 RCC cases, namely at very low risk of developing LN metastases [9]. In contrast, many retrospective reports suggested that LND might be beneficial in T3–4 cases or in the presence of unfavourable conditions (i.e. the presence of coagulative tumour necrosis or sarcomatoid features, high Fuhrman grade, larger tumours, suspected LN metastases at preoperative imaging, or palpable nodes at surgery) [3, 10-12].

The most widely used clinical guidelines (Table 2) consider LND merely as a staging procedure in the RCC setting [7]. Undeniably, adequate information regarding the presence of LN metastases is of key importance because clinical understaging and overstaging may lead to significant consequences in the neoadjuvant, treatment and adjuvant setting, where an inaccurate staging may result in incorrect clinical decisions [10].

Table 2. Currently available guidelines regarding the need for the extent of lymphadenectomy in renal cell carcinoma
 Limited LND*Regional/extended LND
  1. *Hilar region. From the crus of the diaphragm inferiorly to the bifurcation of the aorta or vena cava. For right-sided tumours, dissections of the lateral caval, precaval, postcaval and interaortocaval nodes. For left-sided tumours, dissections of the left para-aortic nodes, the left diaphragmatic nodes and the pre-aortic nodes.

European Association of Urology [1]For staging purposes, in patients with palpable or CT-detected enlarged lymph nodes, resection of the affected lymph nodes should be performed to obtain adequate staging informationFor staging purposes if patients show palpable or CT-detected enlarged lymph nodes, to obtain adequate staging information
American Urological AssociationNot coveredNot covered
National Comprehensive Cancer Network [16]Not coveredOptional
Asian Urological Association [17]Not coveredWhen lymph node metastases are expected, suspected or clinically enlarged
Canadian Kidney Consensus guidelinesOptional for clinical N0M0In patients with palpable or CT-detected enlarged lymph nodes, resection of the affected lymph nodes should be performed and submitted separately for stagingNot covered

In this context, no convincing data exist regarding the minimum number of LNs that should be removed at the time of nephrectomy to obtain an adequate staging LND. Terrone et al. [13] and Joslyn et al. [14] found a positive correlation between the increasing number of nodes examined and the number of positive LNs detected. Similarly, Schafhauser et al. [15] retrospectively analysed patients treated with nephrectomy and extended LND. In such a cohort, when an extended LND was sought, a mean number of 14 nodes were removed. Correspondingly, Terrone et al. [13] suggested that 13 represents the most informative cut-off of LNs removed that predicts the presence of a nodal metastasis. Specifically, pN1 prevalence between patients with organ-confined disease was 3.4% vs 10.5% if <13 or ≥13 nodes were removed. Among patients with locally advanced RCC, pN1 prevalence was 19.7% vs 32.2% if <13 or ≥13 nodes were examined [13].

Although of note, none of those studies systematically analysed the likelihood of finding one or more positive LNs according to the number of LNs removed at radical surgery for RCC. Interestingly, by relying on a systematic statistical methodology, we found a very similar number of nodes (n = 15) to be removed as suggested by Terrone et al. and Schafhauser et al. ROC curve analyses allowed us to depict the probability of finding LN metastases according to each individual number of removed and examined LNs. Specifically, we demonstrated that only slight differences were recorded among the sensitivity analyses stratified for clinical nodal status, the presence of metastases at diagnosis and pathological T stage (Figs 1-4).

Finally, we showed that in stratifying patients according to the recently published Mayo Clinic LNI risk classification [2, 3], additional meaningful differences were recorded. Interestingly, we confirmed the results of Blute et al. and Crispen et al. that demonstrated a higher LNI prevalence when tumour diameter is >10 cm (Table 2). In our study population the presence of an organ-confined tumour >10 cm virtually doubled LNI prevalence (7.3 vs 15.2% in pT2a and pT2b cases, respectively; P < 0.001). Finally, we confirmed that the presence of metastases at diagnosis, higher T stages and Fuhrman grade 3–4 represent the main risk factors for nodal invasion at final pathology. Moreover, when considering only patients who received a retroperitoneal extended LND, 64.5% harboured LN metastases in landing sites outside the ipsilateral hilar boundaries. However, only 12.9% of patients with LN metastases showed metastases exclusively outside the ipsilateral hilar boundaries. Therefore, in the majority of patients with LN metastases, a limited LND would have correctly staged the disease as pN1 but it would have severely underestimated the actual extent of LN involvement.

Our findings confirm that, when clinically indicated, LND in RCC should be extended. We demonstrated that, if unfavourable clinical or pathological characteristics are evident at diagnosis and/or during surgery, an even more extended LND should be pursued. Finally, we found that the greatest accuracy in staging the disease is achieved when about 20 LNs are removed.

Although this report represents the first attempt to assess systematically the minimum number of LNs that should be removed at the time of nephrectomy, some limitations apply. Specifically, because patient variables and pathological specimen handling could affect the yield of LND, the information provided in our probability plots cannot be directly converted into surgical guidelines about the required nodal yield at LND. However, our plot provides a valid approximation of the number of nodes that should be removed to consider LND satisfactory.

Our findings demonstrate that the extent of LND affects the ability to find nodal metastases. If clinically indicated, staging LND should be extended to achieve an adequate accuracy. Limited (hilar) LND can no longer be considered adequate for staging purposes in the RCC setting.

In conclusion, our findings confirm that, when clinically indicated, staging LND in RCC should be extended. The removal of 15 LNs might represent the lowest threshold to define a staging LND as adequate. More extended LND should be pursued if unfavourable clinical and pathological characteristics are evident at diagnosis and/or during surgery.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References