RENAL CELL CARCINOMA WITH RETROPERITONEAL LYMPH NODES: ROLE OF LYMPH NODE DISSECTION

Transcription

1 /03/ /0 Vol. 169, , June 2003 THE JOURNAL OF UROLOGY Printed in U.S.A. Copyright 2003 by AMERICAN UROLOGICAL ASSOCIATION DOI: /01.ju c RENAL CELL CARCINOMA WITH RETROPERITONEAL LYMPH NODES: ROLE OF LYMPH NODE DISSECTION ALLAN J. PANTUCK, AMNON ZISMAN, FREDRICK DOREY, DEBBY H. CHAO, KEN-RYU HAN, JONATHAN SAID, BARBARA J. GITLITZ, ROBERT A. FIGLIN AND ARIE S. BELLDEGRUN From the Departments of Urology, Medicine and Pathology and Laboratory Medicine, University of California School of Medicine, Los Angeles, California ABSTRACT Purpose: We better defined the benefits and morbidity of lymph node dissection in patients with localized renal cell carcinoma using the experience of patients treated at our institution. Materials and Methods: A retrospective cohort study was performed with outcome assessment based on the chart review of demographic, clinical and pathological data in 1,087 patients with renal cell carcinoma treated at our institution. Patients with renal cell carcinoma who did not undergo nephrectomy as part of cancer treatment, those with bilateral disease and those for whom nodal status was unknown were not included in this study. A total of 900 patients meeting these criteria who underwent nephrectomy for unilateral renal cell carcinoma at our medical center form the principal study population. Results: Positive lymph nodes were associated with larger, higher grade, locally advanced primary tumors that were more commonly associated with sarcomatoid features. Positive nodes were 3 to 4 times more common in patients with metastatic disease and the majority of these patients could be identified preoperatively. The survival of patients with regional lymph node involvement only was identical to that of patients with distant metastatic disease only. Patients with regional nodes and distant metastases had significantly inferior survival to those with either condition alone. In node negative cases lymph node dissection can be performed with no additional morbidity but it confers no survival advantage. In node positive cases lymph node dissection can also be performed safely but it is associated with improved survival and a trend toward an improved response to immunotherapy. Conclusions: Regional lymph node dissection is unnecessary in patients with clinically negative lymph nodes since it offers extremely limited staging information and no benefit in terms of decreasing disease recurrence or improving survival. In patients with positive lymph nodes lymph node dissection is associated with improved survival when it is performed in carefully selected patients undergoing cytoreductive nephrectomy and postoperative immunotherapy. When lymph nodes are present, they should be resected when technically feasible. KEY WORDS: carcinoma, renal cell; lymph node dissection; kidney; survival; immunotherapy In 1969 Robson et al advocated dissection of the para-aortic and paracaval lymph nodes from the bifurcation of the aorta to the crus of the diaphragm as a necessary component of radical nephrectomy, and suggested that the improved survival of patients undergoing radical nephrectomy was due in part to this retroperitoneal lymph node dissection. 1 In this classic report written before adequate preoperative imaging 22.7% of patients undergoing radical nephrectomy had positive lymph nodes, and yet survival was equivalent to that in patients with renal vein tumoral involvement only. The Toronto group continued to advocate the benefits of lymph node dissection 2 decades later after examining long-term survival in 162 cases of renal cell carcinoma treated with radical nephrectomy 2 despite conflicting results that have appeared in the literature concerning the possible benefits of lymph node dissection. 1, 3 14 However, other aspects of radical nephrectomy, such as the necessity of removing the whole kidney 15, 16 or routine adrenalectomy, have recently been shown not to be necessary or beneficial in most adequately staged renal cell carcinoma cases. 17, 18 In the era prior to immunotherapy for advanced or metastatic disease pathological evidence of lymph node positive disease portended a striking decrease in life expectancy for Accepted for publication December 13, patients with renal cell carcinoma. 19, 20 However, the significance of lymphadenopathy and the relative benefit of retroperitoneal lymph node dissection in the context of cytoreductive nephrectomy and modern adjunctive immunotherapy has not been adequately examined. Intuitively the inclusion of retroperitoneal lymph node dissection during radical nephrectomy should at least add information relevant to staging and the determination of prognosis but more intriguing is the question of whether it exerts a survival advantage and what trade-off must be made for this possible benefit in respect to potential additional morbidity and mortality. Do patients without evidence of adenopathy experience decreased recurrence rates or improved survival due to lymph node dissection? Cytoreductive nephrectomy has been shown in 2 phase III randomized studies to improve the survival of patients with metastatic renal cell carcinoma 21, 22 but is it necessary to perform lymph node dissection in the context of distant metastatic disease? Is it necessary to perform full retroperitoneal node dissection or is more limited dissection as beneficial? We better defined the benefits and morbidity of lymph node dissection for patients with localized and metastatic renal cell carcinoma using the experience of those treated at our institution in a multidisciplinary program for kidney cancer.

2 RENAL CELL CANCER WITH RETROPERITONEAL LYMPH NODES 2077 MATERIALS AND METHODS Patients. With approval by the University of California-Los Angeles institutional review board (99-233) a retrospective cohort study was performed with outcome assessment based on a chart review of demographic, clinical, and pathological data on patients with renal cell carcinoma treated at our institution. The records of 1,087 patients were reviewed. Patients with renal cell carcinoma who did not undergo nephrectomy as part of cancer treatment, patients with bilateral synchronous tumors whose analysis required special modeling, those with familial renal cell carcinoma syndromes, including the von Hippel-Lindau syndrome, and patients for whom nodal status was unknown were not included in this study. Between 1989 and 2000, 900 patients meeting these criteria underwent nephrectomy for unilateral renal cell carcinoma at our medical center and they form the principal study population of this report. All patients underwent radical or partial nephrectomy and the majority of those with metastases were treated with recombinant interleukin-2 (IL-2) based immunotherapy regimens within the frame of 11 clinical trials. Clinical or pathological details resulting from lymph node dissection during nephrectomy were available for 798 of the 900 patients. Stage was determined according to the 1997 UICC TNM classification of malignant tumors, 5th edition. 23 Clinical and pathological findings were gathered for staging at operation. T stage was defined by pathological examination, and the N and M components were defined according to pathological findings or by clinical data when applicable. Renal cell carcinomas were classified and graded according to current consensus standards 24, 25 by a small group of experienced pathologists. Grading was determined according to the 4 tiered Fuhrman system. 26 Eastern Cooperative Oncology Group (ECOG) performance status was determined at initial presentation and at each followup. Patients were divided into 4 pathological groups; including those with no regional lymph nodes or metastatic disease at diagnosis, those with regional lymph nodes only (fig. 1); those with distant metastatic disease only and those with regional lymph nodes as well as distant metastatic disease. These groups were subdivided into cohorts that did and did not undergo retroperitoneal lymph node dissection at nephrectomy. Clinically lymphadenopathy was defined as enlarged hilar or retroperitoneal lymph nodes 1 cm. 3 or greater on preoperative computerized tomography or magnetic resonance imaging. The extent of lymph node dissection was determined by a review of surgical reports. Clinical end points for patients without nodes or metastasis at nephrectomy were overall survival, disease specific survival and freedom from local or systemic recurrence. Clinical end points for patients with nodes and/or metastasis were overall survival and disease specific survival, and local or systemic recurrence. The impact of lymph node dissection on operative time, estimated operative blood loss, transfusion requirements, hospital stay, perioperative death and complication rates, impact on eligibility to receive post-cytoreductive nephrectomy immunotherapy and best overall response following immunotherapy was considered. Intraoperative complications that are recorded in the kidney cancer data base and are relevant to lymph node dissection include injuries to the aorta and inferior vena cava, duodenal lacerations and other enterotomies as well as hepatic and splenic tears. Acute and delayed postoperative complications recorded include lower extremity edema, deep vein thrombosis, postoperative bleeding, renal failure, adrenal insufficiency, chylous ascites, prolonged ileus and ischemic colitis. In an effort to decrease the possibility of selection bias the cohorts treated with and without lymph node dissection were compared using the ANOVA method to determine the relative distribution in each group of relevant prognostic factors, including stage, grade and ECOG performance status. Survival analysis and statistical methods. The end point of interest was survival time, defined as time from treatment initiation (nephrectomy) to the date of death or last followup. Survival distributions were estimated using the Kaplan- Meier method. Comparisons between groups were performed by the log rank method to assess the significance of the Kaplan-Meier curves. To consider potential interactions among numerous variables simultaneously clinical and pathological parameters were evaluated using univariate and multivariate Cox proportional hazard models 27 to determine the variables that independently correlated with progressive disease and cancer death. The Pearson chi-square, Wilcoxon rank sum (Mann-Whitney) and 2-sample Student t tests were used when indicated. Data was analyzed using commercially available software. For all tests a difference was considered significant at p RESULTS At diagnosis 535 patients (59%) had no evidence of regional lymph nodes or distant metastasis (N0M0), 129 (14%) had evidence of regional lymph nodes (N ) and 236 (26%) had evidence of distant metastasis only (N0M1). Of the 129 N cases 47 were pathological N1, while 82 were pathological N2. A total of 43 patients had nodal disease only (N1M0), while 86 presented with coexisting distant metastatic disease (N M1). In M0 cases there was a 7.4% overall incidence of regional lymph nodes compared with 26.7% of M1 cases that presented with lymphadenopathy. Median followup for patients alive at the FIG. 1. A, 32-year-old male who presented with hematuria and nonspecific abdominal pain had small, central right renal cell carcinoma, bulky retroperitoneal adenopathy and no evidence of distant metastatic disease. B, microscopic foci (arrows) of clear cell carcinoma within lymph node surrounding renal pedicle. Reduced from 40.

3 2078 RENAL CELL CANCER WITH RETROPERITONEAL LYMPH NODES last followup was 30 months. Median time to death was 16 months. Figure 2 shows the survival of the 4 groups. The best survival was seen in patients without nodal or metastatic disease. The survival of patients with nodal disease only (N ) was equivalent to that in patients presenting with distant metastases (p 0.59), while the survival of patients presenting with nodal and metastatic disease was significantly worse (p 0.01). Table 1 lists the demographics of patients presenting with and without regional lymph nodes at diagnosis. Patients with lymph nodes tended to present at a younger age, had larger primary tumors that more commonly occurred on the right side, were more likely to have metastatic disease at presentation, presented with higher pathological T stage, and had worse ECOG performance status and higher grade tumors. Patients with regional nodes were 3 times more likely to have a tumor with predominantly sarcomatoid histology. Compared with 54% of N0 cases only 18% of N cases presented with organ confined disease and 67% versus 32% presented with low grade (1 to 2) tumors, respectively. The median survival of patients with nodal and distant metastatic disease was 10 months compared with twice that in patients with nodal or distant metastases only (p 0.002). Long-term 5-year survival of patients with node positive disease was 23%. Table 2 shows clinical characteristics and outcomes of patients who were treated with and without regional lymph node dissection at nephrectomy. Overall 54% of the whole study cohort underwent some degree of lymph node dissection, including limited dissection in 64%, full template dissection in 20% and lymph node dissection of unspecified extent in 16%. For patients without metastasis there were no statistically significant differences between those who did and did not undergo lymph node dissection in terms of surgery time, estimated blood loss, transfusion requirements, perioperative complication rates or hospitalization. Patients with metastatic disease had greater blood loss, transfusion requirements, surgery time, hospital time, and surgical complication rates than patients with nonmetastatic disease. However, an ANOVA test including metastatic disease and lymph node dissection as variables did not demonstrate a greater morbidity of lymph node dissection in M1 compared with M0 cases. The differences in outcomes for patients with M1 disease who did and who did not undergo lymph node dissection did not achieve statistical significance. TABLE 1. Clinical characteristics of patients with and without regional retroperitoneal lymphadenopathy No. Node Neg. (%) No. Node Pos. (%) p Value Total pts Males 509 (66) 76 (59) Rt. tumor 415 (54) 64 (50) 0.03 Smokers 407 (53) 73 (57) Mean pack yrs Metastatic disease 236 (31) 86 (66) N1 Not applicable 43 (33) N2 Not applicable 86 (67) Clear cell 595 (77) 84 (65) Papillary 101 (13) 19 (14.7) Sarcomatoid 42 (5.4) 20 (15.5) Stage (%): T1 (less than (38) 7 (6) cm.) T2 (greater than 116 (16) 15 (12) 7 cm.) T3 310 (42) 88 (69) T4 33 (4) 17 (13) Organ confined (54) (18) ECOG performance status (%): (45) 30 (25) (52) 81 (66) 2 19 (3) 11 (9) Grade (%): /2 466 (67) 38 (32) 3/4 236 (33) 83 (68) The overall, local, and distant recurrence-free survival of 257 patients with node negative disease who underwent no lymph dissection was the same as the survival of 238 with node negative disease who underwent lymph node dissection (p 0.42, fig. 3). Survival was equivalent in those who underwent limited, full and an unspecified degree of lymph node dissection (data not shown). However, there was a statistically significant survival advantage in 112 patients with node positive disease who underwent lymph node dissection compared with 17 who did not undergo lymph node dissection on univariate analysis (p , fig. 4). Median survival was approximately 5 months better in those undergoing lymphadenectomy. A similar proportion in each group received immunotherapy following nephrectomy. Of those who did not undergo lymph node dissection FIG. 2. Kaplan-Meier survival curves for 4 patient cohorts. Similar survival was noted in patients with lymphadenopathy or distant metastases, while those with positive lymph nodes and distant metastatic disease had significantly worse survival.

4 RENAL CELL CANCER WITH RETROPERITONEAL LYMPH NODES 2079 TABLE 2. Clinical characteristics of patients with and without retroperitoneal lymph node dissection during nephrectomy 71% remained eligible to receive systemic immunotherapy compared with 63% who underwent lymph node dissection. ANOVA analysis showed no difference in the distribution of tumor stage or ECOG performance status between these 2 groups. However, there was an imbalance between the 2 groups in the distribution of tumor grade with patients who underwent lymph node dissection more likely to have higher grade tumors. In N cases 50% of grade 1, 72% of grade 2, 92% of grade 3 and 100% of grade 4 tumors were associated with concomitant lymph node dissection at nephrectomy. Thus, the survival of patients with N disease undergoing lymph node dissection was better than those not undergoing lymphadenectomy despite a worse predicted prognosis since patients with higher grade tumors have been shown to have worse survival than those with lower grade disease. Despite the positive impact on survival there was no difference in local or systemic recurrence rates following lymph node dissection and recurrence rates were similar regardless of the extent of lymph node dissection (p 0.57). A number of Cox multivariate regression models were performed to analyze variables that were independently predictive of survival in N cases, including lymph node dissection, tumor grade, ECOG performance status, immunotherapy, metastatic disease and tumor stage (table 3). In each model the absence of lymph node dissection remained a significant predictor of decreased survival since patients not undergoing lymph node dissection were 3 times more likely to die than those who underwent the procedure. Patients with N disease who underwent lymph node dissection had only a 44% incidence of progressive disease following 1 course of immunotherapy compared with 56% of those in whom lymph nodes were left in place. In the multivariate model immunotherapy had a protective effect on the survival of those in whom the nodes were removed that approached statistical significance (p 0.057). DISCUSSION No Node Dissection Node Dissection p Value No. pts. (%) 365 (46) 433 (54) No. disease (%): M1 108 (36) 195 (64) M0 257 (52) 238 (48) % Dissection: Full Not applicable 20 Limited Not applicable 64 Unspecified Not applicable 16 Mean estimated blood loss SE (ml.): M ,526 1,132 1,753 M ,301 % Transfusion: M M Mean operative time SE (mins.): 0.32 M M Mean hospital stay SE (days): 0.65 M M No. periop. deaths 4 4 % Complications: M M % Recurrence: Local Any Surgery remains the mainstay of treatment for locally confined renal cell carcinoma and it has recently been shown to be a fundamental part of an integrated treatment plan along with immunotherapy for patients with metastatic renal cell carcinoma. 21, 22 The 5-year survival rate for all disease stages in patients with renal cell carcinoma has continued to improve. 28 However, it remains unclear what portion, if any, of this improved survival can be attributed to retroperitoneal lymph node dissection as advocated by Robson et al in their classic study describing the surgical management of renal cell carcinoma. 1 Recognizing the unpredictable lymphatic drainage from renal tumors along with the incidence of concomitant blood borne metastases, 29 the practice of systematic, extended lymphadenectomy has decreased at many institutions and many patients are left with unknown pathological nodal status. 30 The published literature has been scant and contradictory in its conclusions regarding the benefits of lymphadenectomy 1, 3 13 with many earlier studies promoting the value of lymphadenectomy, while more recent studies failed to show any benefit. Positive lymph nodes are commonly associated with metastatic disease and yet the literature on lymph node dissection in patients with known metastatic disease is even more scant. 14 Johnsen and Hellsten studied 554 cases involving renal cell carcinoma diagnosed at autopsy and found only 80 (14%) with positive lymph nodes, including 75 with additional distant metastases, of which only 5 showed nodal disease confined to the tissues surrounding the great vessels. 31 Thus, fewer than 1% of cases were theoretically curable by nephrectomy and lymph node dissection alone. The incidence of positive lymph nodes at nephrectomy has decreased dramatically with the introduction of computerized tomography 1, 12 and improved preoperative staging. Lymph node dissection has been proposed to offer several possible benefits, including staging and prognosis, decreasing local or systemic recurrence, and improved survival. The original study of Robson et al showed a 23% incidence of positive nodes in patients who underwent extensive lymphadenectomy, which was associated with 10-year survival that approached 35%. 1 Several contemporaneous series showed 5-year survival rates of only 15% but the extent of lymph node dissection was not specified. 3 5 Phillips and Messing reported that in a series of 37 patients who underwent radical nephrectomy and later had progression local control was improved by lymph node dissection without extending hospital stay or adding to surgical morbidity. 10 Herrlinger et al compared the survival of 511 patients undergoing radical nephrectomy, of whom 320 underwent extended dissection and 191 underwent only facultative or staging lymphadenectomy. 9 They found 56% 10-year survival in those who underwent extended dissection compared with only 41% in those treated with facultative lymphadenectomy only (p 0.01). More recently, Minervini et al evaluated a series of 167 patients, including 108 who underwent nephrectomy alone and 59 who underwent nephrectomy with regional lymph node dissection limited to the anterior, posterior and lateral sides of the ipsilateral great vessel from the level of the renal vessels down to the inferior mesenteric artery. 13 In this series the 5-year survival rate for the 108 patients who underwent nephrectomy alone was 79%, which was the same as the 78% rate for those who also underwent lymph node dissection, suggesting no clinical benefit. The only randomized phase III trial to address the benefits of lymph node dissection during radical nephrectomy for patients with resectable nonmetastatic renal cell carcinoma was performed by the European Organization for the Research and Treatment of Cancer Genitourinary Group in study In this study dissection extended from the crus of the diaphragm inferior to the bifurcation of the aorta. The incidence of unsuspected lymph nodes after proper preoperative staging was only 3.3% and complication rates were similar between the 2 groups. Only the preliminary results of this study have been published since in 1999 only 17% of patients enrolled had achieved a survival end point. In the current University of California-Los Angeles data

5 2080 RENAL CELL CANCER WITH RETROPERITONEAL LYMPH NODES FIG. 3. Survival of 495 patients undergoing nephrectomy for clinically localized disease was same regardless of whether they underwent lymph node dissection (LND) at nephrectomy. FIG. 4. Survival of patients with metastatic disease undergoing cytoreductive nephrectomy prior to planned systemic immunotherapy was significantly better in those undergoing lymph node dissection (LND) as part of nephrectomy. presented despite the addition of essentially no significant additional morbidity or mortality lymph node dissection during nephrectomy in patients with clinically negative lymph nodes did not add a measurable benefit. We could not detect a decrease incidence of local or systemic recurrence, or an added benefit in terms of survival in clinically node negative cases. Furthermore, in the current study population of 900 patients the incidence of patients presenting with regional nodal metastases in the absence of distant disease was less than 5% (43 of 900). It is likely that many of these patients are identified preoperatively by computerized tomography or intraoperatively by surgical exploration. In the remainder in whom disease is unrecognized it is just as likely that lymph nodes may be found in the perirenal fat and surrounding the pedicle as along the great vessels. Fewer than 8% of node positive cases in this study were discovered incidentally in the sense that nodes were found in the pathological specimen without a clinical impression of nodal involvement. Clearly the majority of node positive cases are identified prior to lymph node dissection. Furthermore, only 5 patients (less than 0.5%) in the whole University of California-Los Angeles data base presented with a tumor of less than 7 cm. and without coexisting distant metastases, and only 10 presented with organ confined disease without metastatic disease. This knowledge has great implications regarding the management of small (less than 4 cm.) incidental renal mass, for which current trends have shifted toward minimally invasive as well as percutaneous ablative techniques that may pre-

6 RENAL CELL CANCER WITH RETROPERITONEAL LYMPH NODES 2081 TABLE 3. Cox multivariate survival analysis of patients with retroperitoneal lymph nodes treated with and without retroperitoneal lymph node dissection during cytoreductive nephrectomy Variable HR SE (95% CI) p Value No dissection ( ) Grade ( ) ECOG performance status ( ) Immunotherapy ( ) Metastases ( ) clude adequate lymphadenectomy. The current data suggest that lymph node dissection can safely be omitted in these patients in the absence of evidence of metastatic disease. In the classic study of nephrectomy for metastatic renal cell carcinoma in the pre-immunotherapy era dekernion et al found that survival was significantly decreased in patients with the addition of positive nodes, positive regional nodes were strongly associated with local recurrence in the renal fossa and patients in whom regional disease was not removed during nephrectomy were dead within 1 year with only 20% alive at 6 months. 19 A more recent experience presented by investigators at the National Cancer Institute described 154 patients with metastatic renal cell carcinoma who underwent cytoreductive nephrectomy at the National Institutes of Health as preparation for IL-2 based therapy. 14 They found a 8.5-month median survival in patients with positive nodes, which was inferior to 15-month survival in those without nodes. Furthermore, 8.6-month median survival in those in whom lymph nodes were completely resected was similar to 8.5-month survival in those with incomplete resection and each result was better than 3.3-month survival in patients deemed unresectable. These values are in general agreement with the data on patients with metastatic renal cell carcinoma currently presented. In the current data patients in whom positive regional lymph nodes were left in place had statistically inferior survival to similar patients in whom regional disease was resected by lymphadenectomy. Furthermore, in the National Cancer Institute and University of California-Los Angeles series no real difference could be detected in survival that was related to the extent of node dissection performed. It is possible that this fact merely reflects the lack of prospectively defined templates dictating surgical technique and the extent of dissection performed was individualized to suit each particular patient with those who had clinically evident but minimal nodal disease not undergoing formal dissection from the diaphragm to the aortic bifurcation. It is possible that a clear statement describing exactly what surgical boundaries represent adequate node dissection can be defined only by prospective clinical trials. It is interesting that immunotherapy appeared to benefit patients with nodal disease who underwent surgical resection of these lymph nodes since the University of California-Los Angeles data suggest that patients who have regional lymphadenopathy do not respond in a meaningful way to systemic immunotherapy and experience improved survival following treatment. 28 There are several limitations to the current study. Data were collected in a retrospective manner by chart review, the extent of lymph node dissection was not standardized and the decision of whether to perform lymph node dissection was left to surgeon choice rather than being assigned through a randomization process. In addition, determining the extent of node dissection retrospectively based on operative reports depends on the completeness of surgeon dictation and is subject to subjective interpretation. For these reasons we hesitate to overstate the comparisons of the benefits of full versus limited dissection, although it should be stated that these data were collected for data base entry by reviewers blinded to the purpose of the current study. However, more significantly, because of the nature of these limitations, we were concerned that the group not undergoing lymph node dissection was highly selected and the survival advantage noted for node positive cases in which lymph node dissection was done may have been due to unrecognized selection bias rather than to a treatment effect. For example, were patients who were more ill or had more extensive disease selected not to undergo lymph node dissection? Several strategies were used in an effort to minimize the potential for selection bias. Surgical reports were reviewed to determine the reason that node dissection was deferred. From this review it became clear that few patients had the nodes left in situ because they were believed by the primary surgeon to be unresectable. Many patients were treated during the earlier half of the 1990s, when a common practice was to leave the nodes in place in patients who had N M0 disease as a marker lesion that could be serially imaged to assess the response to immunotherapy. Poorer survival of these patients with N M0 disease in whom nodes were left in place contrasts with the overall better survival of this group as a whole. In most patients with metastatic disease in addition to nodal disease, the decision to leave the nodes in place was based on the primary surgeon perception that node dissection would not provide an added benefit in light of the distant disease that was left in place. ANOVA was performed to determine the relative distribution of prognostic factors in the dissection and no dissection groups at the original presentation. In the University of California-Los Angeles data set ECOG performance status, tumor grade and tumor stage are the 3 variables that are most significant for determining patient outcome and survival and they make up the University of California-Los Angeles Integrated Staging System, which has been previously described. 32 ANOVA revealed equal distributions for stage and ECOG performance status between the groups, while it was noted that the group that underwent lymph node dissection was more likely to have higher grade disease, which would normally portend worse survival. However, the same trend for improved survival in the lymph node dissection group was noted when patients were evaluated across each grade level separately (data not shown). Furthermore, it is unlikely that the survival difference noted can be attributed merely to overall tumor burden since data have previously been presented showing no survival difference between patients with microscopic and macroscopic nodal involvement in the same cohort. 33 We next questioned whether the no dissection group had greater surgical complications or became ineligible for immunotherapy due to more rapidly progressive disease following nephrectomy. However, a similar incidence of patients in each group remained eligible and received immunotherapy following cytoreductive nephrectomy with 71% of the no dissection group going on to immunotherapy compared with only 63% of the dissection group. Since generally only patients with an ECOG performance status of 0 or 1 are selected for high dose IL-2 therapy, this finding indicates that the 2 groups were similar in performance status following nephrectomy as well as diagnosis. Furthermore, several Cox multivariate models were created to examine the potential interactions of multiple prognostic variables simultaneously, including lymph node dissection, tumor stage, tumor grade, ECOG performance status, distant metastatic disease, treatment with immunotherapy and University of California-Los Angeles Integrated Staging System category. In each model failure to perform lymph node dissection remained a significant and independent predictor of decreased survival after accounting for these other variables (RR 3.0). Recently much has been learned to explain the molecular mechanisms behind the clinical patterns of observed metastatic spread. For example, it is well established that angiogenesis (the formation of new blood vessels) is necessary for

7 2082 RENAL CELL CANCER WITH RETROPERITONEAL LYMPH NODES the growth and metastatic spread of solid tumors. The vascular endothelial growth factor (VEGF) family of genes includes unique isoforms, of which each is capable of binding to unique receptors. Evidence is accumulating that different VEGF family members are important in determining the route of metastatic spread through classic angiogenesis via blood vessels or through lymphangiogenesis. Experimental systems have demonstrated the importance of VEGF and VEGF receptor-2 in tumor progression through the promotion of blood vessel formation. 34 VEGF-D is capable of activating VEGF receptor-2 and VEGF receptor-3, a receptor expressed by endothelial cells of the lymphatic channels, and it is thought to be a molecular signal for lymphangiogenesis 35 and lymphatic tumor spread. 36 Stacker et al noted that VEGF-D expression in a mouse tumor model led to spread of the tumor to lymph nodes, which could be blocked by an antibody specific for VEGF-D, whereas the expression of the angiogenic growth factor VEGF, which activates only VEGFR-2 did not. 37 The factors that promote the regulation of the expression of different VEGF genes is not entirely clear. Gunningham et al recently observed that VEGF, VEGF-B and VEGF-D gene expression is differentially regulated by the level of tumor hypoxia and by von Hippel- Lindau gene expression in renal cell carcinoma cell lines. 38 It is possible that different tumor genetics and different risk factor exposures may lead to differential tumor gene expression, resulting in unique tumor biologies with unique patterns of spread and response to treatment. Currently a renal cancer tissue array study is underway at our institution to determine expression of the VEGF genes and receptors, which will be correlated with clinical patterns of tumor spread, treatment response and survival. CONCLUSIONS The morbidity associated with retroperitoneal lymph node dissection during nephrectomy appears to be acceptable. However, regional lymph node dissection is unnecessary in patients with clinically negative lymph nodes because it offers extremely limited staging information and no benefit in terms of decreasing disease recurrence or improving survival. When lymph nodes are positive, lymph node dissection is associated with improved survival when it is performed in carefully selected patients undergoing cytoreductive nephrectomy and postoperative immunotherapy. Although defining the optimal surgical boundaries would require a prospective comparison of various possible templates, the current data suggest that when lymph nodes are present, they should be resected when technically feasible. REFERENCES 1. Robson, C. J., Churchill, B. M. and Anderson, W.: The results of radical nephrectomy for renal cell carcinoma. J Urol, 101: 297, Robson, C. J.: Results of radical thoraco-abdominal nephrectomy in the treatment of renal cell carcinoma. Prog Clin Biol Res, 100: 481, Flocks, R. H. and Kadesky, M. C.: Malignant neoplasms of the kidney: an analysis of 353 patients followed five years or more. J Urol, 79: 196, Skinner, D. G., Colvin, R. B., Vermillion, C. D., Pfister, R. C. and Leadbetter, W. F.: Diagnosis and management of renal cell carcinoma. A clinical and pathological study of 309 cases. Cancer, 28: 1165, Rafla, S.: Renal cell carcinoma. Natural history and results of treatment. Cancer, 25: 26, Siminovitch, J. P., Montie, J. E. and Straffon, R. A.: Lymphadenectomy in renal adenocarcinoma. J Urol, 127: 1090, Giuliani, L., Martorana, G., Giberti, C., Prescatore, D. and Magnani, G.: Results of radical nephrectomy with extensive lymphadenectomy for renal cell carcinoma. J Urol, 130: 664, Pizzocaro, G., Piva, L. and Salvioni, R.: Lymph node dissection in radical nephrectomy for renal cell carcinoma: is it necessary? Eur Urol, 9: 10, Herrlinger, A., Schrott, K. M., Schott, G. and Sigel, A.: What are the benefits of extended dissection of the regional renal lymph nodes in the therapy of renal cell carcinoma? J Urol, 146: 1224, Phillips, E. and Messing, E. M.: Role of lymphadenectomy in the treatment of renal cell carcinoma. Urology, 41: 9, Elsasser, D., Stadick, H., Stark, S., van de Winkel, J. G., Gramatzki, M., Schrott, K. M. et al: Preclinical studies combining bispecific antibodies with cytokine-stimulated effector cells for immunotherapy of renal cell carcinoma. Anticancer Res, 19: 1525, Blom, J. H., van Poppel, H., Marechal, J. M., Jacqmin, D., Sylvester, R., Schroder, F. H. et al: Radical nephrectomy with and without lymph node dissection: preliminary results of the EORTC randomized phase III protocol EORTC Genitourinary Group. Eur Urol, 36: 570, Minervini, A., Lilas, L., Morelli, G., Traversi, C., Battaglia, R., Cristofani, R. et al: Regional lymph node dissection in the treatment of renal cell carcinoma: is it useful in patients with no suspected adenopathy before or during surgery? BJU Int, 88: 169, Vasselli, J. R., Yang, J. C., Linehan, W. M., White, D. E., Rosenberg, S. A. and Walther, M. M.: Lack of retroperitoneal lymphadenopathy predicts survival of patients with metastatic renal cell carcinoma. J Urol, 166: 68, Steinbach, F., Stöckle, M., Müller, S. C., Thüroff, J. W., Melchior, S. W., Stein, J. et al: Conservative surgery of renal cell tumors in 140 patients: 21 years of experience. J Urol, 148: 24, Licht, M. R. and Novick, A. C.: Nephron sparing surgery for renal cell carcinoma. J Urol, 149: 1, Tsui, K.-H., Shvarts, O., Barbaric, Z., Figlin, R., dekernion, J. B. and Belldegrun, A.: Is adrenalectomy a necessary component of radical nephrectomy? UCLA experience with 511 radical nephrectomies. J Urol, 163: 437, Kletscher, B. A., Qian, J., Bostwick, D. G., Blute, M. L. and Zincke, H.: Prospective analysis of the incidence of ipsilateral adrenal metastasis in localized renal cell carcinoma. J Urol, 155: 1844, dekernion, J. B., Ramming, K. P. and Smith, R. B.: The natural history of metastatic renal cell carcinoma: a computer analysis. J Urol, 120: 148, Thrasher, J. B. and Paulson, D. F.: Prognostic factors in renal cancer. Urol Clin North Am, 20: 247, Flanigan, R. C., Salmon, S. E., Blumenstein, B. A., Bearman, S. I., Roy, V., McGrath, P. C. et al: Nephrectomy followed by interferon alfa-2b compared with interferon alfa-2b alone for metastatic renal-cell cancer. N Engl J Med, 345: 1655, Mickisch, G. H., Garin, A., van Poppel, H., de Prijck, L. and Sylvester, R.: Radical nephrectomy plus interferon-alfa-based immunotherapy compared with interferon alfa alone in metastatic renal-cell carcinoma: a randomised trial. Lancet, 358: 966, Sobin, L. and Wittekind, C.: TNM Classification of Malignant Tumors. New York: International Union Against Cancer, p. 180, Storkel, S., Eble, J. N., Adlakha, K., Amin, M., Blute, M. L., Bostwick, D. G. et al: Classification of renal cell carcinoma: Workgroup No. 1. Union Internationale Contre le Cancer (UICC) and the American Joint Committee on Cancer (AJCC). Cancer, 80: 987, Kovacs, G., Akhtar, M., Beckwith, B. J., Bugert, P., Cooper, C. S., Delahunt, B. et al: The Heidelberg classification of renal cell tumours. J Pathol, 183: 131, Fuhrman, S. A., Lasky, L. C. and Limas, C.: Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol, 6: 655, Cox, D. R. and Oakes, D.: Analysis of Survival Data. New York: Chapman and Hall, Pantuck, A. J., Zisman, A. and Belldegrun, A. S.: The changing natural history of renal cell carcinoma. J Urol, 166: 1611, DeKernion, J. B.: Lymphadenctomy for renal cell carcinoma. Therapeutic implications. Urol Clin North Am, 7: 697, Ward, J. F., Blute, M. L., Cheville, J. C., Lohse, C. M., Weaver, A. L. and Zincke, H.: The influence of PNx/PN0 grouping in a multivariate setting for outcome modeling in patients with clear cell renal cell carcinoma. J Urol, 168: 56, 2002

8 RENAL CELL CANCER WITH RETROPERITONEAL LYMPH NODES Johnsen, J. A. and Hellsten, S.: Lymphatogenous spread of renal cell carcinoma: an autopsy study. J Urol, 157: 450, Zisman, A., Pantuck, A. J., Dorey, F., Said, J. W., Schvarts, O., Quintana, D. et al: Improved prognostication of renal cell carcinoma using an integrated staging system. J Clin Oncol, 19: 1649, Pantuck, A., Zisman, A., Chao, D., Dorey, F., Gitlitz, B. J., dekernion, J. B. et al: Regional lymphadenopathy during cytoreductive nephrectomy predicts IL-2 failure in patients with metastatic renal cell carcinoma. J Clin Oncol, 20: 172a, Saleh, M., Stacker, S. A. and Wilks, A. F.: Inhibition of growth of C6 glioma cells in vivo by expression of antisense vascular endothelial growth factor sequence. Cancer Res, 56: 393, Taipale, J., Makinen, T., Arighi, E., Kukk, E., Karkkainen, M. and Alitalo, K.: Vascular endothelial growth factor receptor-3. Curr Top Microbiol Immunol, 237: 85, Stacker, S. A., Caesar, C., Baldwin, M. E., Thornton, G. E., Williams, R. A., Prevo, R. et al: VEGF-D promotes the metastatic spread of tumor cells via the lymphatics. Nat Med, 7: 186, Stacker, S. A., Vitali, A., Caesar, C., Domagala, T., Groenen, L. C., Nice, E. et al: A mutant form of vascular endothelial growth factor (VEGF) that lacks VEGF receptor-2 activation retains the ability to induce vascular permeability. J Biol Chem, 274: 34884, Gunningham, S. P., Currie, M. J., Han, C., Turner, K., Scott, P. A., Robinson, B. A. et al: Vascular endothelial growth factor-b and vascular endothelial growth factor-c expression in renal cell carcinomas: regulation by the von Hippel-Lindau gene and hypoxia. Cancer Res, 61: 3206, 2001