The current study was performed to determine the impact of the presence of retroperitoneal lymphadenopathy on the survival and response to immunotherapy of patients with metastatic renal cell carcinoma (RCC).
The current study was performed to determine the impact of the presence of retroperitoneal lymphadenopathy on the survival and response to immunotherapy of patients with metastatic renal cell carcinoma (RCC).
A retrospective cohort study was performed with outcome assessment based on the chart review of demographic, clinical, and pathologic data from 1087 patients. Patients with RCC who did not present with metastatic disease, who did not undergo nephrectomy as part of their cancer treatment, and those in whom either the lymph node (N) or metastatic (M) status was unknown, were excluded. A total of 322 M1 patients who met these criteria and who underwent nephrectomy for unilateral RCC formed the principal study population.
Two hundred thirty-six patients presented with N0M1 disease and 86 patients presented with N+M1 disease. In M1 patients, the presence of positive regional lymph nodes was associated with larger sized, higher grade, locally advanced primary tumors that were more commonly associated with sarcomatoid features. N0M1 patients were more likely to achieve an objective response to systemic immunotherapy compared with N+M1 patients (P = 0.01). N+M1 patients overall had worse short-term and long-term survival compared with N0M1 patients, with a median survival of 10.5 months compared with 20.4 months, respectively. The median survival of N0M1 patients was improved to 28 months in those who received adjunctive immunotherapy (P = 0.0008), whereas the median survival of patients with N+M1 disease was the same in those treated with and those treated without adjunctive immunotherapy (P = 0.18).
Even in the modern era of systemic immunotherapy, the presence of regional lymphadenopathy exerts a detrimental effect on the survival of patients with metastatic RCC. Lymph node status is a strong predictor of the failure to achieve either an objective immunotherapy response or an improvement in survival when immunotherapy is given as an adjunctive treatment after cytoreductive nephrectomy. However, in multivariate analysis, including both clinical and pathologic variables, lymph node status was found to have less of an impact on survival than primary tumor stage and grade and patient performance status. Cancer 2003;97:2995–3002. © 2003 American Cancer Society.
Malignant tumors of the kidney and renal pelvis comprise slightly more than 2% of new cancer incidence and mortality in the U.S. with an estimated 30,800 new cases expected to occur in the year 2000, causing approximately 12,100 deaths.1 Approximately one-third of new patients with renal cell carcinoma (RCC) have metastatic disease at the time of presentation and another one-third presenting with localized disease eventually will experience disease recurrence and progression.2 Once metastatic, RCC poses a therapeutic challenge because it is highly resistant to conventional modes of therapy, including chemotherapy and radiation.3 However, for a small minority of patients with RCC, their disease is responsive to immunotherapy (IMT), with 5–10% of patients experiencing durable and complete disease remissions.2, 4 In the era prior to the use of IMT for advanced or metastatic disease, pathologic evidence of lymph node-positive disease portended a striking decrease in life expectancy for patients withRCC.5, 6 Even without concomitant distant metastases, the presence of lymph node disease was associated with poor long-term survival, with 5-year survival rates of 5–35% reported in the literature,7, 8 with the majority of studies reporting rates of approximately 15%.9–11 However, to our knowledge, the significance of lymphadenopathy within the context of cytoreductive nephrectomy and modern adjuvant IMT has not been adequately examined. The identification of prognostic factors that would permit more evidence-based selection of patients who were most likely to respond to IMT would be beneficial to maximize therapeutic efficacy and to minimize toxicity to patients who were unlikely to respond. The objective of the current study was to define better the effects of retroperitoneal, regional lymphadenopathy on IMT treatment response and survival in patients with metastatic RCC, using the experience of patients treated at the University of California at Los Angeles (UCLA) within a multidisciplinary program for renal carcinoma.
A retrospective cohort study was performed with approval from the UCLA Institutional Review Board (IRB # 99-233), with outcome assessment based on chart review of demographic, clinical, and pathologic data in patients with RCC who were treated at UCLA. A database containing the records of 1087 patients, including 466 who had metastatic RCC at the time of diagnosis, was reviewed. Patients with RCC who did not have metastatic disease at the time of presentation, patients who did not undergo nephrectomy as part of their cancer treatment, patients with bilateral synchronous tumors whose analysis required special modeling, patients with familial RCC syndromes including the von Hippel–Lindau disease (VHL), and patients for whom either the lymph node or metastatic status was unknown were not included in the current study. Between 1989 and 2000, 322 patients meeting these criteria underwent nephrectomy for unilateral RCC at the UCLA Medical Center and form the principal study population of this report. All patients underwent radical or partial nephrectomy and the majority of the patients were treated after cytoreductive nephrectomy with recombinant interleukin-2 (IL-2)-based IMT regimens within the frame of 11 clinical trials.
Stage was determined according to the 5th edition (1997) of the International Union Against Cancer TNM classification of malignant tumors.12 The clinical and pathologic findings were gathered for staging at the time of surgery. The T classification was defined by pathologic examination and the N and M classifications were defined according to pathologic findings or by clinical data when applicable. RCCs were classified and graded according to the current consensus standards13, 14 by a small group of experienced pathologists. Although sarcomatoid tumors are no longer classified as a separate histologic entity, the percentage of tumors having secondary sarcomatoid features was recorded. Grading was determined according to the four-tiered Fuhrman system. The Eastern Cooperative Oncology Group performance status (ECOG PS) was determined at the time of the initial presentation and at each follow-up visit.
Patients were divided into two pathologic groups: those with metastatic RCC but no positive regional lymph nodes at the time of diagnosis, and those with positive regional lymph nodes as well as distant metastatic disease at the time of presentation. These groups were subdivided into cohorts of patients who did and patients who did not undergo systemic IMT after cytoreductive nephrectomy. The extent of lymph node dissection was determined by review of surgical reports. The determination of the extent of lymph node dissection retrospectively based on surgical reports is dependent on the completeness of the surgeons' dictation and is subject to subjective interpretation, although these data were collected for database entry by reviewers blinded to the purpose of the current study. Clinically, lymphadenopathy was defined as enlarged hilar or retroperitoneal lymph nodes measuring ≥ 1 cm3 on preoperative computerized tomography or magnetic resonance imaging studies. The clinical endpoints examined included best response after systemic IMT, and overall and disease-specific survival.
Response to IMT was determined by standard practice of clinical measurements of indicator lesions according to the following definitions. All measurable and evaluable lesions and sites were assessed using the same technique as baseline. A complete response (CR) was the complete disappearance of all measurable and evaluable disease, without the appearance of any new lesions and without evidence of nonevaluable disease, including normalization of markers and other abnormal laboratory values. A partial response (PR) was defined as a ≥ 50% decrease from baseline in the sum of the products of the perpendicular dimensions of all measurable lesions and the appearance of no new lesions. Progressive disease (PD) was defined as a 25% increase or an increase of 10 cm2 (whichever was smaller) in the sum of the products of measurable lesions compared with the smallest sum observed (over baseline if no decrease), a clear worsening of any evaluable disease, or the appearance of any new lesion/site. Patients were defined as having stable disease when they did not qualify for CR, PR, or PD.
The endpoint of interest was survival time, defined as the time from treatment initiation (nephrectomy) to the date of death or last follow-up. 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 the potential interactions between numerous variables simultaneously, clinical and pathologic parameters were evaluated using univariate and multivariate Cox proportional hazards models15 to determine the variables that were correlated independently with PD and cancer death. The Pearson chi-square test, Wilcoxon rank-sum test (Mann–Whitney U test), and two-sample Student t test were used when indicated. Data were analyzed using Stata 7.0 software (Stata Corporation, College Station, TX). For all tests, a difference was considered statistically significant when P < 0.05.
Of the 322 patients comprising the study cohort, 236 had metastatic RCC and no evidence of retroperitoneal lymphadenopathy (N0M1), whereas 86 patients had both positive regional lymph nodes as well as distant metastatic disease (N+M1). Thus, the incidence of concomitant lymph node disease in patients with metastatic RCC was 26.7%. This incidence was more than 3 times greater than the 7% incidence of lymph node disease in M0 patients. The 86 patients in the N+M1 group could be stratified further into 28 patients (33%) with pathologic N1 disease and 58 patients (67%) with pathologic N2 disease. The demographics for the N0M1 and N+M1 groups are presented in Table 1. Although both groups presented at a similar average age, women were more likely to present with N+M1 disease. The ratio of men to women for the entire UCLA Kidney Cancer Database is 2.2:1, which is similar to the 2.4:1 ratio in all patients presenting with metastatic disease. However, the ratio of men to women for N0M1 patients was 3:1, whereas the ratio for N+M1 patients was 1.5:1 (P = 0.01). Patients with positive lymph nodes (all N+ patients with or without concomitant M1 disease) were less likely than their N0M1 counterparts to be either current or former smokers. There was a 63% incidence of a history of tobacco use in all M1 patients, which was similar to the 62% incidence reported in the N+M1 group. However, tobacco use decreased to 45%% in patients presenting with lymph node-only disease (N+M0) and was reduced further to 20% in patients with pathologic N1M0 disease (P < 0.05). This trend was similar for patients who were current smokers at the time of diagnosis, with 24% of patients with M1 disease, 11% of patients with N+M0 disease, and only 5% of patients with N1M0 disease being current smokers. Patients with lymph node disease had slightly larger primary tumors than their N0M1 counterparts (9.6 cm vs. 8.9 cm), yet this difference did not reach statistical significance. However, if the N2M1 subgroup was considered alone, the average tumor size increased to 10.2 cm (P = 0.02). Overall, the N1M1 patients were more likely to have locally advanced pT disease (P = 0.03), higher grade tumors (P = 0.0001), and sarcomatoid or undifferentiated histologic features (20% vs. 15%). Comparing the N0M1 group with the N1M1 group, 78% of cases versus 67% were clear (conventional) cell carcinomas, 6% of cases versus 13% were chromophil (papillary), and 1% of cases versus 0% were chromophobe, respectively (Table 1). However, the overall distribution of histologic types was similar between the 2 groups (P = 0.11), as was the distribution of patients according to the ECOG PS at baseline (P = 0.95).
|Males||59 (25%)||34 (40%)|
|Females||177 (75%)||52 (60%)|
|Smoking history||149 (63%)||53 (62%)|
|Average tumor size (cm)||8.9 cm||9.6 cm||0.18|
|(N2M1 = 10.2 cm)||0.016|
|pT1 (< 7 cm)||29 (12%)||3 (3%)|
|pT2 (> 7 cm)||32 (14%)||10 (12%)|
|pT3||151 (64%)||61 (71%)|
|pT4||24 (10%)||12 (14%)|
|Organ confined||61 (26%)||13 (15%)|
|Locally advanced||175 (74%)||73 (85%)|
|Primary T classification||0.028|
|Low-grade, 1/2||112 (47%)||21 (25%)|
|High-grade, 3/4||124 (53%)||63 (75%)|
|Clear cell||184 (78%)||58 (67%)|
|Chromophil||15 (6%)||11 (13%)|
|Sarcomatoid/undifferentiated||35 (15%)||17 (20%)|
|Chromophobe||2 (1%)||0 (0%)|
|ECOG 0||42 (19%)||18 (21%)|
|ECOG 1||169 (75%)||58 (69%)|
|ECOG > 1||15 (6%)||8 (10%)|
|Treated with IMT||147 (63%)||56 (65%)|
|IMT objective response||44 (30%)||6 (11%)|
|IMT stable disease||52 (35%)||21 (37%)|
|IMT progressive disease||51 (35%)||29 (52%)|
Both the N0M1 and N+M1 groups were as likely to be treated after cytoreductive nephrectomy with adjunctive IMT, with 64% and 65%, respectively, of each group going on to be treated with IMT. N0M1 patients were three times more likely than N+M1 patients to achieve an objective response after IMT. With a 30% objective response rate, N0M1 patients experienced approximately twice the response rate normally quoted for M1 RCC patients after IMT.2 N0M1 patients also were significantly less likely to experience PD compared with M1 patients with positive lymph nodes (P = 0.01).
The survival of N0M1 and N+M1 patients is presented in Table 2 and Figure 1. The median survival for all N0M1 patients regardless of subsequent treatment was 20.4 months compared with 10.5 months for N+M1 patients (P = 0.002). Figure 2 presents the Kaplan–Meier survival curves for each group broken into subgroups according to whether each received postoperative, adjunctive IMT. N0M1 patients who were not treated with adjunctive IMT were found to fare as poorly as N+M1 patients, whereas the N0M1 patients treated with IMT improved their median survival to 8–28 months. The group with the worst median survival was the N+M1 patients who did not receive postoperative IMT; their reported median survival was 4.5 months. However, the survival of N+M1 patients who were treated with IMT was not found to be significantly better than the survival of either N0M1 patients who were not treated with IMT or the survival of N+M1 patients who were treated with IMT (P = 0.18). The distribution of the percentage in each group with solitary versus multiple organ sites of metastatic disease was examined because it is known that the survival of patients with multiple metastatic sites is inferior to that for patients with solitary metastases (median survival of 11 months vs. 27 months; unpublished data). In the N0M1 group, 52% of patients had metastases to a solitary organ, whereas 66% of the N+M1 patients had metastases to a solitary organ. Thus, the inferior survival of the N+M1 group of patients cannot be attributed to greater tumor burden or number of metastatic sites. Furthermore, the survival of 56 patients with pathologic N2M1 disease was not found to be significantly different from the survival of 25 patients with pathologic N1M1 disease (P = 0.69) (Fig. 3). In Cox multivariate modeling (Table 3), the impact of lymphadenopathy on survival was less than the impact of tumor grade, tumor stage, and patient ECOG PS, findings that are consistent with previous publications derived from the UCLA kidney data set.16
|2-year||5-year||Median (mos)||P value|
|All N0M1 (236)||44 ± 4%||23 ± 4%||20.4|
|All N+M1 (86)||26 ± 5%||17 ± 5%||10.5|
|N0M1, No IMT (85)||34 ± 6%||15.6 ± 6%||12|
|N0M1 + IMT (151)||55 ± 5%||28 ± 5%||28|
|N+M1, No IMT (26)||32 ± 9%||10.7 ± 9%||4.5|
|N+M1 + IMT (60)||31 ± 7%||18 ± 7%||10.8|
|Variable||Hazards ratio||P value|
|Primary T classification||2.6||0.000|
Since 1950, there has been a 126% increase in the incidence of RCC accompanied by a 36.5% increase in annual mortality in the U.S.17 Although in recent years the majority of these new cases involve incidental, localized tumors that are detected through the increased and routine use of ultrasound and computerized tomography, the number of cases of advanced disease, including those with regional extension and distant metastases, also increased in all race and gender categories.17 The initial reports that recombinant interferon-α produced responses of 15–20% with a duration of up to 10 months provided early enthusiasm for the use of IMT for RCC.18 To our knowledge, whether interferon-induced tumor regression is achieved by an indirect effect involving activation of the host immune response or by a direct antiproliferative effect on the tumor remains unknown. Recombinant human IL-2 monotherapy has well documented activity in patients with metastatic RCC and was approved by the U.S. Food and Drug Administration for use in this disease in 1992. A recent update on the use of high-dose, bolus IL-2 in RCC patients demonstrated a CR rate of 7% and a PR rate of 8%;19 the median duration of response for all cases of CR had still not been reached at the time of last follow-up but was at least 80 months, with a small subset of responding patients who remained alive for 11 years after treatment.
Although it is clear that major strides have been made in the struggle against metastatic RCC, the overall response rates of IMT remain approximately 20% at best. Furthermore, these response rates often are obtained at the cost of significant toxicities to the patient. One major goal in the treatment of cancer is the identification of prognostic factors that permit the selection of patients most likely to respond to specific therapeutic interventions to maximize therapeutic efficacy and to minimize toxicity to those patients unlikely to respond. The current system of RCC classification is partially predictive of outcomes, and remains based primarily on morphologic criteria such as tumor grade, tumor stage, and patient performance status, which remain the most useful, clinically available predictors of patient outcome.16 However, the mechanisms by which immunologic therapies produce antitumor responses in patients are not well understood, and to the best of our knowledge no clinical or laboratory markers have been established to date to differentiate between responding and nonresponding patients. Although some aspects of the molecular genetics of RCC have begun to be appreciated, little is known regarding the molecular mechanisms or other prognostic factors underlying its unpredictable clinical course and whether an individual tumor will respond to IMT.
A number of studies have been undertaken to define patient- or tumor-related determinants of response, with varying success. Determinants that have been explored include the induction of autoimmune thyroid disease,20 the magnitude of rebound lymphocytosis, nadir platelet count and serum tumor necrosis factor levels after IL-2 therapy, and pretreatment serum IL-6 levels.21–24 Studies involving major histocompatibility phenotype have been contradictory, with some finding no association with response25 and others suggesting that the histocompatibility leukocyte antigen (HLA)-A32 phenotype on patients' peripheral blood mononuclear cells26 or HLA-DR expression in tumor lesions27 could be correlated with response to IL-2. Previous work at UCLA failed to demonstrate any significant differences between responders and nonresponders with regard to their immune status vis-á-vis the in vitro characteristics of tumor-infiltrating lymphocytes (TIL), including the degree of intratumoral lymphocyte infiltrate, amount or time of TIL expansion, specific lysis of autologous tumor, phenotype, quantity of TIL infused, or cytokine release profile.28 Other basic immune parameters, including pretreatment levels of signal transduction molecules,29 leukocyte and lymphocyte subset counts, cytolytic activity, or serum cytokine concentration during treatment, have failed to yield a reliable predictive parameter.29 Although a number of recent studies characterizing genetic aberrations in RCC have implicated a number of chromosomal loci as being significant in cancer development and progression,30, 31 to our knowledge no data are available relating specific genetic events to IMT response in patients with RCC. Several small cytogenetic studies of RCC patients undergoing immune-based therapy have suggested both a better response associated with an aneuploid DNA pattern32 as well as a correlation between hyperdiploidy and tumor progression.33
The data presented herein confirm that, in a large single institution experience with > 300 patients with metastatic RCC who were treated aggressively with both cytoreductive surgical extirpation as well as systemic IMT, a simple clinical and pathologic variable—regional lymphadenopathy—portends a striking decrease in survival and that in a univariate model, the presence of positive lymph nodes is a significant predictor of failure to respond or to benefit from IMT in terms of increased survival after treatment. It is possible that patients who experience disease spread early on into the lymphatics are unable to control their tumor locally because of an intrinsic immune dysfunction. A recent article published by investigators from the National Cancer Institute examined 154 patients with metastatic RCC who underwent cytoreductive nephrectomy at the National Institutes of Health as preparation for IL-2-based therapy.34 They found a median survival of 8.5 months in patients with positive lymph nodes, which was inferior to the 15-month median survival reported in patients without positive lymph nodes. These numbers are in general agreement with the data regarding patients with metastatic RCC presented in the current study. In this study, the 20-month median survival of patients with N0M1 disease was significantly better than the median survival of 10 months reported in patients with N1M1 disease, and patients with metastatic RCC without positive regional lymph nodes who were treated successfully with surgery and adjunctive IMT were reported to have a median survival approaching 2.5 years (28 months), with nearly 30% of patients reportedly still alive at 5 years. This survival clearly is superior to the range of 6–12 months that is classically quoted in the literature5 and which is similar to the survival of patients with lymph node-positive metastatic disease in the current study. Furthermore, the presence of lymph nodes was associated with the same survival regardless of whether patients went on to receive adjunctive IMT. Unlike N1M1 patients who did not achieve a survival benefit from IMT, patients with N0M1 disease were reported to have a significant improvement in survival for those able to receive treatment versus those who did not receive treatment after nephrectomy. Regional lymphadenopathy is associated with other poor prognostic covariates including larger tumor size, higher tumor grades, higher pathologic tumor stage, and the presence of undifferentiated histology. We previously published an integrated staging system that demonstrated that three variables (tumor stage, tumor grade, and ECOG PS) significantly stratify RCC patient survival and that these three factors account for the majority of the survival prediction in a multivariate model.16 However, unlike the presence of lymphadenopathy, tumor grade and stage are determined only through pathologic examination after nephrectomy. Although the sensitivity, specificity, and predictive values of imaging for enlarged lymph nodes are not 100%,35 modern imaging permits excellent identification of patients with regional lymphadenopathy prior to surgery, and this information can be used to counsel patients prior to the initiation of aggressive, potentially morbid treatment regarding their potential to respond to therapy. The limitations of a retrospective database-based study, including the potential for unrecognized selection bias, the lack of predetermined lymph node dissection templates, and the nonrandomized nature of treatment assignment preclude a definitive statement regarding whether patients with N1M1 RCC should be denied the opportunity to receive IMT. However, clearly, the knowledge of lymphadenopathy as a prognostic factor should help with clinical trial design, patient stratification, and interpretation.
Clinically, patients with advanced RCC may present with lymph node spread only, with visceral or osseous metastases only, or with a combination of both. Recent advances in our knowledge of the molecular mechanisms involved in an individual tumor's propensity to spread via the bloodstream versus via the lymphatic channels may reflect basic differences in the basic biology of individual tumors as a result of unique genetic abnormalities or secondary to unique carcinogenic exposures. These differences in tumor biology may account for differences in treatment response or patient survival. The vascular endothelial growth factor (VEGF) genes encompass a family of unique isoforms, each capable of binding to a different pattern of receptors that are associated with different routes of tumor spread, through either classic angiogenesis and spread through blood vessels or through lymphangiogenesis. Several experimental systems have demonstrated the importance of VEGF and VEGF receptor-2 in tumor progression via the promotion of new blood vessel formation36, 37 whereas VEGF-D is capable of binding VEGF receptors on both blood vessels as well as lymphatic endothelium, and it is believed to be a molecular signal for lymphangiogenesis38 and lymphatic tumor spread39 that can be blocked by antibodies specific for VEGF-D.40 To our knowledge, the factors that promote the regulation of the expression of different VEGF gene family members are not well understood. However, Gunningham et al. recently demonstrated that VEGF, VEGF-B, and VEGF-D gene expression were differentially regulated by levels of tumor hypoxia and by the presence or absence of wild-type VHL gene expression in RCC cell lines.41 Currently, a renal carcinoma tissue array study is underway at UCLA to determine the expression of the VEGF genes and receptors, the results of which will be correlated with clinical patterns of tumor spread, treatment response, and survival.
Even in the modern era of systemic IMT, the presence of regional lymphadenopathy exerts a detrimental effect on the survival of patients with metastatic RCC. Lymph node status is a strong predictor of the failure to achieve either an objective response to IMT or an improvement in survival when IMT is given as an adjunctive treatment after cytoreductive nephrectomy.