Original Article

Can We Better Select Patients With Metastatic Renal Cell Carcinoma for Cytoreductive Nephrectomy?
Stephen H. Culp, MD, PhD1; Nizar M. Tannir, MD2; E. Jason Abel, MD1; Vitaly Margulis, MD1; Pheroze Tamboli, MD3; Surena F. Matin, MD1; and Christopher G. Wood, MD1 BACKGROUND: The benefits of cytoreductive nephrectomy (CN) in metastatic renal cell carcinoma (mRCC) should outweigh surgical morbidity. Even when the generally agreed upon selection criteria for CN are met, some patients do poorly after surgery. The objective of this study was to identify preoperative factors that were prognostic of outcome in patients who were being considered for CN. METHODS: The authors conducted a retrospective study to investigate the overall survival (OS) of patients who underwent CN using the OS of patients with mRCC who did not undergo CN as a referent group. Patients who underwent CN were divided into 2 groups based on when their OS diverged from that of nonsurgical patients. Chi-square analysis was used to identify variables that differed between the 2 surgical groups. Multivariate Cox proportional hazards regression was used to analyze those variables for the entire CN cohort. Risk factors were defined as preoperative variables that remained significant on multivariate analysis. The median OS and the overall risk of death were calculated based on the number of risk factors. RESULTS: From 1991 to 2007, 566 patients who were eligible for or received systemic therapy underwent CN, and 110 patients received medical therapy alone. On multivariate analysis, independent preoperative predictors of inferior OS in surgical patients included a lactate dehydrogenase level greater than the upper limit of normal, an albumin level less than the lower limit of normal, symptoms at presentation caused by a metastatic site, liver metastasis, retroperitoneal adenopathy, supradiaphragmatic adenopathy, and clinical tumor classification !T3. Inferior OS and an increased risk of death were correlated positively with the number of risk factors. Surgical patients who had !4 risk factors did not appear to benefit from CN. CONCLUSIONS: The authors of this report identified 7 preoperative variables that permitC 2010 American ted them to identify patients who were unlikely to benefit from CN. Cancer 2010;116:337888. V Cancer Society. KEYWORDS: renal cell carcinoma, cytoreductive nephrectomy, metastasis, survival.

Renal cell carcinoma (RCC) accounts for approximately 3% of all adult tumors. In the United States in 2008, 54,390 individuals were diagnosed with kidney cancer, and 13,010 died from the disease.1 The incidence of each stage of RCC and mortality from the disease have increased over the past few decades.2,3 In addition, up to 25% of patients continue to have metastatic disease at initial diagnosis.4 Despite recent advances in the understanding of RCC and the development of new systemic molecular agents that target the vascular endothelial growth factor (VEGF) receptor and mammalian target of rapamycin (mTOR) pathways, the prognosis for patients with metastatic RCC (mRCC) remains poor. It has been demonstrated that cytoreductive nephrectomy (CN) prolongs the survival of patients with mRCC when it is performed before interferon alpha (IFNa) treatment compared with IFNa treatment alone.5-7 To our knowledge, no prospective randomized trial has examined the role of CN in combination with systemic targeted therapy,8 however, CN continues to be central in the integrated management of patients with mRCC. Unfortunately, there are certain subgroups of patients with mRCC who do not benefit from CN. Although the group from Tufts University has published clinical criteria that can be used to select patients for surgery,9 the predictive value of those criteria remain imperfect and unvalidated, and there are small but relevant subsets of patients with rapid disease progression after surgery and, thus, derive no benefit
Corresponding author: Christopher G. Wood, MD, FACS, Department of Urology-Unit 1373, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030; Fax: (713) 792-3474; cgwood@mdanderson.org
1 Department of Urology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; 2Department of Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; 3Department of Pathology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas

DOI: 10.1002/cncr.25046, Received: August 4, 2009; Revised: September 9, 2009; Accepted: September 14, 2009, Published online May 17, 2010 in Wiley InterScience (www.interscience.wiley.com)

3378

Cancer

July 15, 2010

Patients Who Will Not Benefit From CN/Culp et al

from CN. In this retrospective study, we examined a large cohort of patients with mRCC who underwent CN at our institution to determine preoperative clinical variables that might be used to better identify patients who probably will not benefit from CN.

MATERIALS AND METHODS After we obtained approval from the Institutional Review Board for the Protection of Human Subjects at The University of Texas M. D. Anderson Cancer Center (MDACC), the institutional RCC database was queried for patients with mRCC who underwent CN between 1991 and 2007. Patients who had a previous history of treated RCC were excluded. In addition, to identify the typical overall survival (OS) of patients with mRCC who did not undergo CN, we collected data from patients with mRCC who presented with their primary tumor in place and underwent medical therapy (eg, immunotherapy, systemic targeted therapy, and/or chemotherapy) alone at MDACC. OS times of the nonsurgery and surgery groups were calculated from diagnosis to either death or the last known follow-up. For the surgery group, comprehensive clinical, perioperative, preoperative laboratory, and final pathological data variables were collected and reviewed to ensure accuracy and completeness of all data. Clinical TNM staging for patients in the surgery and nonsurgery groups was based on the American Joint Committee on Cancer 2002 TNM classification10 and was determined at the time of CN or diagnosis, respectively. Tumor grade corresponded to the Fuhrman grading system, and histology was classified according to 2004 World Health Organization criteria.11 The number and site(s) of metastasis and lymph node involvement were determined based on radiologic imaging at the time of CN or diagnosis for the surgery and nonsurgery groups, respectively. Patients with metastasis to the brain (n 20) underwent treatment of their brain metastasis (surgical resection or stereotactic radiation) before CN. The most recent laboratory values before CN were used for the surgery group, and the time between the acquisition of these laboratory values and CN was determined to control for timing variances. For all analyses, laboratory values were categorized into 3 groups (normal, above normal, and below normal) based on reference values at MDACC. Patients who had laboratory values within the normal range served as the referent group in each analysis. By using Kaplan-Meier (K-M) analysis, we divided surgical patients into 2 groups based on when OS
Cancer

diverged from that of nonsurgical patients. Chi-square analysis was used to identify the variables that differed between the 2 surgical groups. Variables that differed significantly were then entered into a multivariate Cox proportional hazards regression model that was controlled for year of surgery, time from diagnosis to surgery, and time from acquisition of laboratory values to surgery. The variables that remained significant on multivariate analysis and could be established preoperatively were labeled as risk factors. Complete preoperative laboratory data were missing for 98 of 566 patients (17.3%). Therefore, to compensate for missing values, we performed multivariate Cox proportional hazards regression analysis using only those patients who had complete laboratory data and including all significant variables from chi-square analysis. From this, we identified the preoperative laboratory variables that remained significant on multivariate analysis. Then, we performed multivariate Cox proportional hazards regression analysis using the entire surgical cohort and included all significant variables from chi-square analysis except for preoperative laboratory values. However, we did include in the multivariate analysis a binary variable corresponding to preoperative laboratory variables that were significant in the initial analysis. Additional Cox proportional hazards regression and K-M analyses were performed to categorize surgical patients based on the number of preoperative risk factors and using the OS of the nonsurgical cohort as a referent. For all analyses, Stata software (version 10.1; Stata Corp., College Station, Tex) was used. All P values .05 (2sided) were considered significant.

RESULTS We identified 566 patients with mRCC who underwent CN and 110 patients who received medical therapy alone between 1991 and 2007 with a median follow-up of 20 months (range, 1.5-166.8 months) and 9.4 months (range, 1-102.6 months), respectively. Significant differences between the surgical and nonsurgical groups are listed in Table 1. Of the 110 patients who received medical therapy alone, most (86 patients; 78.2%) were diagnosed between 2002 and 2007, and a total of 54 patients (49.1%) received 1 or more targeted agents at some time during treatment. On the basis of K-M analysis, patients who underwent CN and died within 8.5 months of diagnosis did not appear to have benefited from surgery compared with patients who received medical therapy alone

July 15, 2010

3379

Original Article
Table 1. Characteristics That Differed Significantly Between Patients With Metastatic Renal Cell Carcinoma Who Underwent Cytoreductive Nephrectomy or Received Medical Therapy Alone (1991-2007) Table 1. (Continued)

Variable

No. of Patients (%)

CN Medical Therapy Only
7 (6.1) 1 (0.9) 15 (13.2)

Variable

No. of Patients (%)

CN Medical Therapy Only
88 (74.6)

Metastatic and systemic All

.031 <.001 History of embolization to primary tumor

Caucasian race

470 (83)

15 (2.7) 1 (0.2) 33 (5.9)

.006

ECOG PS
0 1 2 3 4
375 176 15 0 0 (66.3) (31.1) (2.7) (0) (0) 22 50 27 14 1 (19.3) (43.9) (24.7) (12.3) (0.9)

Creatinine, mg/dL
0.8-1.5a <0.8 >1.5
449 (80.8) 75 (13.5) 32 (5.8) 92 (80) 10 (8.7) 13 (11.3)

.047

CN indicates cytoreductive nephrectomy; ECOG PS, Eastern Cooperative Oncology Group performance. a The normal laboratory value range was based on M. D. Anderson Cancer Center reference standards. b Patients who had no organ sites but had clinical evidence of lymph node involvement (retroperitoneal and/or supradiaphragmatic) at the time of diagnosis or CN.

Lactate dehydrogenase, IU/L

313-618a <313 >618
366 (72.2) 34 (6.7) 107 (21.1) 61 (54.5) 10 (8.9) 41 (36.6)

.001

Platelet count, K/UL

140-441a <140 >441
434 (77.6) 9 (1.6) 116 (20.8) 76 (66.1) 3 (2.6) 36 (31.3)

.031

No. of metastatic organ sitesb

0-1 2-3 >3
402 (71) 158 (27.9) 6 (1.1) 56 (47.5) 62 (52.5) 0 (0)

<.001

Site of metastasis
Liver Brain Adrenal
Retroperitoneal lymphadenopathy Supradiaphragmatic lymphadenopathy Clear cell histology 44 20 77 151 149 (7.8) (3.5) (13.7) (26.7) (26.3) 33 13 25 58 14 (28) (11) (21.2) (52.3) (70) <.001 .001 .038 <.001 <.001 <.001 <.001 290 (51.2) 275 (48.6) 1 (0.2) 60 (50.9) 51 (43.2) 7 (5.9) <.001 2 49 194 321 (0.4) (8.7) (34.3) (56.7) 0 21 39 24 (0) (25) (46.4) (28.6) <.001 73 182 193 64 16 17 (13) (32.4) (34.4) (11.4) (2.9) (4.3) 19 31 24 19 1 12 (16.7) (27.2) (21.1) (16.7) (0.9) (10.5)

487 (86)

63 (53.4)

Side of primary tumor

Left Right Bilateral

Fuhrman nuclear grade

1 2 3 4

(P < .05). Therefore, we divided the surgical patients into 2 groups based on a survival of 8.5 months. Clinical, preoperative laboratory, perioperative, and pathologic variables of the 2 surgical groups are listed in Table 2. In the multivariate analysis, we identified 7 variables that remained significant (P < .05) and could be assessed preoperatively that were negative predictors of survival (Table 3). These included a serum albumin below the lower limit of normal, a serum lactate dehydrogenase (LDH) level above the upper limit of normal, a clinical tumor classification of T3 or T4, symptoms at presentation caused by a metastatic site (eg, bone pain, neurologic symptoms, etc), the presence of liver metastasis, and radiographic evidence (!1 cm) of retroperitoneal or supradiaphragmatic adenopathy at the time of CN. The number of preoperative risk factors was correlated with the overall risk of death and was inversely proportional to the median survival of patients who underwent CN (Table 4, Fig. 1). Patients who underwent CN and had !4 risk factors did not benefit from CN when using the survival of patients who received medical therapy alone as a referent group (Table 4, Fig. 2).

Symptoms at presentation
None Local only Metastasis only Systemic only Local and metastatic Local and systemic

(Continued)

DISCUSSION Despite advancements in understanding the biology that underlies RCC and the development of systemic targeted therapy, patients with mRCC continue to have a poor prognosis. Historically, the grim prognosis for patients with mRCC was because of the lack of effective systemic therapy. The use of CN, although not curative, does improve OS for selected patients with mRCC. Before the
Cancer

3380

July 15, 2010

Patients Who Will Not Benefit From CN/Culp et al

Table 2. Clinical Characteristics of Patients With Metastatic Renal Cell Carcinoma Who Underwent Cytoreductive Nephrectomy Based on an Overall Survival of 8.5 Months (1991-2007)

Variable

No. of Patients (%)

OS <8.5 Mo OS >8.5 Mo
316 (69.3) 306 (67.1) 141 (30.9) 9 (2) 27.5 [15.8-49.3] 377 (82.7) 121 103 109 104 62 150 163 46 10 8 11 1 333 120 3 106 (27.7) (23.6) (29.9) (23.8)

Clinical variables
Men ECOG PS
0 1 2 76 (69.1) 69 35 6 28 93 23 28 22 31 11 32 30 18 7 8 4 0 69 38 3 45 (62.7) (31.8) (5.5) [19.5-48.9] (84.6) (22.1) (26.9) (21.2) (29.8) (10) (29.1) (27.3) (16.4) (6.4) (7.3) (3.6) (0) (62.7) (34.6) (2.7) (40.9) .966 .115

Median BMI [range], kg/m2 Caucasian race Age at diagnosis, y

<50 50-57 57-65 >65

.243 .639 .379

Symptoms at time of diagnosis

None Local only Metastatic only Systemic only Local and systemic (13.8) (33.3) (36.1) (10.2) (2.2) (1.8) (2.4) (0.2)

.003

Local and metastatic

Metastatic and systemic All

No. of organs involved with metastasis at time of CNa

0-1 2-3 <3 (73) (26.3) (0.7) (23.3)

.029

Retroperitoneal adenopathy at time of CN Supradiaphragmatic adenopathy at time of CN Clinical tumor classification at time of CN
T1 T2 T3 T4

<.001 .041 <.001

34 (30.9)

99 (21.7)

6 38 59 7

(5.5) (34.6) (53.6) (6.4)

86 177 182 11

(18.9) (38.8) (39.9) (2.4) .676

Side of primary tumor

Right Left Bilateral 57 (51.8) 53 (48.2) 0 (0) 16 27 83 3 20 (14.6) (24.6) (75.5) (2.7) (18.2) 275 (48.6) 237 (52) 1 (0.2) 28 143 308 17 57 178 185 77 22 230 58 159 24 (6.1) (31.4) (67.5) (3.7) (12.5) (40.5) (42.1) (17.5) (4.9) (50.4) (12.7) (34.9) (5.3)

Organ site of metastasis at time of CN

Liver Bone Lung Brain Adrenal .003 .162 .107 .610 .119 .415

Smoking status at diagnosis

Never Former Current 49(47.1) 37 (35.6) 18 (17.3) 9 (8.4) 40 (36.4) 4 (3.6) 7 (6.4) 5 (4.6)

Family history of RCC Metastasectomy at any timeb Neoadjuvant targeted therapy Adjuvant targeted therapy History of secondary malignancy

.155 .008 .006 <.001 .873

(Continued)

Cancer

July 15, 2010

3381

Original Article
Table 2. (Continued)

Variable
Perioperative variables
Laparoscopy Partial nephrectomy Lymph node dissection Median surgery duration [range], min Surgical blood loss: Mean 6 SD, cc Blood transfusion during CN

No. of Patients (%)

OS <8.5 Mo
5 (4.6) 0 (0) 38 (34.6) 240.5 [76-1060] 2508.6 3302.8 80 (72.7)

OS >8.5 Mo
57 (12.5) 9 (2) 168 (36.8) 196 [61-865] 1126.5 1847.9 190 (41.7) .016 .137 .653 .004 <.001 <.001

Pathologic variables
Clear cell histology Sarcomatoid elements Fuhrman nuclear grade
1 2 3 4 85 (77.3) 36 (32.7) 0 7 22 81 33 18 (0) (6.4) (20) (73.6) (30) (16.4) 402 (88.2) 72 (15.8) 2 42 172 240 110 49 (0.4) (9.2) (37.7) (52.6) (24.4) (10.9) .003 <.001 .001

Lymphovascular invasion Extra lymph node extension

.231 .115

Preoperative serum laboratory measurements

Creatinine, mg/dL
0.8-1.5c <0.8 >1.5 83 (76.9) 18 (16.7) 7 (6.5) 55 (57.3) 41 (42.7) 0 (0) 52 (52.5) 6 (6.1) 41 (41.4) 85 (86.7) 4 (4.1) 9 (9.2) 92 (86) 0 (0) 15 (14) 16 (15) 90 (84.1) 1 (0.9) 80 (74.8) 1 (0.9) 26 (24.3) 56 (50.9) 54 (49.1) 0 (0) 366 (81.7) 57 (12.7) 25 (5.6) <.001 321 (82.1) 67 (17.1) 3 (0.8) 314 (77) 28 (6.9) 66 (16.2) 328 (83.3) 14 (3.6) 52 (13.2) .346 393 (87) 7 (1.6) 52 (11.5) .004 139 (30.8) 311 (68.8) 2 (0.4) .517 354 (78.3) 8 (1.8) 90 (19.9) .007 296 (64.9) 160 (35.1) 0 (0) <.001 .502

Albumin, g/dL
3.5-4.7c <3.5 >4.7

Lactate dehydrogenase, IU/L

313-618c <313 >618

Total calcium, mg/dL

8.4-10.2c <8.4 >10.2

.551

White blood cell count, K/UL

4-11c <4 >11

Hemoglobin, g/dL
14-18 (Men) or 12-16 (women)c <14 (Men) or <12 (women) >18 (Men) or >16 (women)

Platelet count, K/UL

140-441c <140 >441

Alkaline phosphatase (IU/L)

38-126c >126 <38

OS indicates overall survival; ECOG PS, Eastern Cooperative Oncology Group performance status; BMI, body mass index; CN, cytoreductive nephrectomy; RCC, renal cell carcinoma; SD, standard deviation. a Patients with 0 organ sites had clinical evidence of lymph node involvement (retroperitoneal and/or supradiaphragmatic) at the time of CN. b Includes stereotactic radiation treatment of brain metastasis. c Normal laboratory value range based on M. D. Anderson Cancer Center reference standards.

3382

Cancer

July 15, 2010

Patients Who Will Not Benefit From CN/Culp et al

Table 3. Multivariate Cox Proportional Hazards Regression Analysis of the Entire Cytoreductive Nephrectomy Cohort (1991-2007)a

Variable
Serum albumin, g/dL
3.5-4.7b <3.5 >4.7

HR
Referent 1.59 1.18

95% CI
1.21-2.10 0.28-4.86

.001 .823

Serum alkaline phosphatase, IU/L

38-126b >126 313-618b <313 >618
Referent 0.89 0.70-1.14 .359

Serum lactate dehydrogenase, IU/L

Referent 0.73 1.66 0.46-1.15 1.26-2.18 .169 <.001

Hemoglobin, g/dL
14-18 (Men) or 12-16 (women)b <12 (Men) or <14 (women) >16 (Men) or >18 (women)
Metastasectomy at any timec Liver metastasis Referent 0.99 1.00 0.92 1.47 0.76-1.29 0.24-4.25 0.75-1.14 1.02-2.13 .920 .999 .458 .039

Clinical tumor classification

T1 T2 T3 T4
Fuhrman nuclear grade Referent 1.28 1.37 2.05 1.19 0.95-1.72 1.01-1.87 1.13-3.72 1.01-1.40 .108 .045 .019 .041

No. of metastatic sites at CNd

0-1 2-3 >3
Sarcomatoid dedifferentiation Clear cell histology Referent 1.13 1.17 1.40 0.75 0.89-1.43 0.47-2.96 1.08-1.80 0.55-1.02 .305 .734 .011 .063

Symptoms at presentation
Metastatic symptoms Systemic symptoms Local symptoms
Laparoscopic nephrectomy Retroperitoneal lymphadenopathy Supradiaphragmatic lymphadenopathy Neoadjuvant targeted therapy Adjuvant targeted therapy Surgery duration Surgical blood loss Blood transfusion at surgery 1.35 1.12 1.14 1.04 1.29 1.48 0.90 0.42 1.00 1.00 1.28 1.03-1.75 0.85-1.49 0.89-1.46 0.73-1.50 1.01-.63 1.18-1.86 0.60-1.34 0.32-0.55 0.99-1.00 1.00-1.00 1.00-1.64 .028 .413 .306 .816 .040 .001 .597 <.001 .694 .007 .048

HR indicates hazard ratio; CI, confidence interval. a Analyses were adjusted for year of surgery, time from diagnosis to cytoreductive nephrectomy (CN) and the time between laboratory examination and CN. b Normal laboratory value range was based on M. D. Anderson Cancer Center reference standards. c Includes stereotactic radiation treatment of brain metastasis. d Patients with 0 organ sites had clinical evidence of lymph node involvement (retroperitoneal and/or supradiaphragmatic) at the time of CN.

introduction of immunotherapy, it was believed that there was no benefit from CN,12 and surgery was performed only in the context of palliation of local symptoms (eg, pain, intractable hematuria) or paraneoplastic symptoms (eg, fever, hypertension, hypercalcemia), and the latter often would persist beyond surgery because of the continued presence of metastatic lesions. Despite anecdotal reports of spontaneous regression of metastasis after nephrectomy,13,14 this was a rare event (<2%), and the morbidity of surgery far outweighed any foreseeable benefit. However, with the introduction of immunotherapy (eg, IFNa, interleukin-2), multiple retrospective series suggested that survival could be improved with the use of CN in highly selected patients before immunotherapy.9,15-23 One study by Fallick et al. demonstrated that, when strict criteria were applied (eg, >75% debulking of tumor, clear cell histology, an Eastern Cooperative Oncology Group [ECOG] performance status of 0 or 1, and the absence of central nervous system, bone, or liver metastasis), the median survival was 20.5 months, and >90% of patients were able to receive adjuvant interleukin-2 therapy.9 More important, a combined analysis of 2 randomized trials (European Organization for the Research and Treatment of Cancer trial 30947 and Southwest Oncology Group trial 8949) that examined patients with mRCC who either received IFNa alone or underwent nephrectomy followed by IFNa demonstrated that the median OS was longer in patients who underwent nephrectomy compared with patients who received IFNa alone (13.6 months; 95% confidence interval [CI], 9.7-17.4 months; vs 7.8 months; 95% CI, 5.9-9.7 months; hazard ratio [HR], 0.69; 95% CI, 0.55-0.87; P .002).5-7 Perioperative mortality was low (1.4%), and >90% of patients who underwent nephrectomy were able to receive adjuvant immunotherapy, although there was no difference in response to immunotherapy between the nonsurgical and surgical groups.5 Although prospective randomized trials have yet to evaluate the role of CN in combination with systemic targeted therapy,8 CN remains a central component in the treatment paradigm for patients with mRCC. The mechanisms underlying the survival benefit of CN remain unknown but may involve the removal of tumorpromoting factors, such as VEGF and/or immunosuppressive cytokines, from the primary tumor as well as decreasing the total tumor burden, thereby increasing the time interval before a lethal burden exists.24-29 Although CN does improve survival for patients with mRCC, subsets of patients do exist who fail to benefit from surgery. Multiple retrospective studies have tried

Cancer

July 15, 2010

3383

Original Article
Table 4. Median Overall Survival and Overall Risk of Death in Patients With Metastatic Renal Cell Carcinoma Who Underwent Cytoreductive Nephrectomy Based on the Number of Preoperative Risk Factorsa

Patient Group
Medical therapy only

No. (%)
110

HR
Referent

95% CI

Median OS, mo
9.6

CN group
No. of preoperative risk factors
0 1 2 3 4 5 6 3 4 70 194 153 88 45 13 3 505 61 (12.4) (34.3) (27) (15.5) (8) (2.3) (0.1) (89.2) (10.8) 0.22 0.33 0.45 0.66 0.78 1.57 0.98 0.39 0.89 0.15-0.31 0.26-0.43 0.34-0.58 0.49-0.88 0.55-1.13 0.88-2.81 0.24-3.99 0.31-0.48 0.64-1.24 <.001 <.001 <.001 .005 .191 .125 .982 <.001 .499 40.6 27.9 20.2 12.6 13.8 7.5 4.3 22.7 12.2

HR indicates hazard ratio; CI, confidence interval; OS, overall survival. a Factors that were identified as significant were serum albumin below the lower limit of normal, serum lactate dehydrogenase above the upper limit of normal, clinical T3 or T4 tumor classification, symptoms at presentation caused by a metastatic site, the presence of liver metastasis, retroperitoneal lymphadenopathy, and supradiaphragmatic adenopathy (!1 cm).

to identify factors that may be associated with a poorer response after surgery.30,31 This is important, because the response to both CN and systemic therapy can be variable, with some patients undergoing rapid progression of their disease despite multimodal therapy. In general, these studies indicated that positive prognostic factors included the ability to remove a significant proportion (>75%) of tumor burden, a favorable ECOG performance status (0 or 1), adequate surgical resectability, and the absence of central nervous system, bone, or liver metastases. In addition, negative prognostic factors included systemic symptoms (eg, weight loss, fever) at the time of CN, multiple sites of metastatic disease, a Fuhrman nuclear grade of 4, sarcomatoid dedifferentiation, coagulative necrosis in the tumor, abnormally high thyroid-stimulating hormone (TSH) levels, retroperitoneal adenopathy, or tumor thrombus.30,31 The objective of the current study was to identify clinical factors that could be known before CN and could be used to help identify those patients who most likely would not benefit from undergoing CN. Most patients in our surgical cohort met the generally accepted criteria for surgery based on previous studies9 that examined clinical parameters. Each patient had complete removal of their primary tumor, which accounted for the majority of their tumor burden. In addition, the majority of patients in our surgical population (97.3%) had an ECOG performance status of 0 or 1 at the time of surgery. Although a selection bias invariably existed in the nonsurgical cohort that was used in our study, the OS of these patients allowed us to establish a referent by which we could identify those surgical patients who clearly did not benefit from CN. From

Figure 1. This chart illustrates a Kaplan-Meier analysis of overall survival for patients with metastatic renal cell carcinoma (mRCC) who underwent cytoreductive nephrectomy (CN) based on the number of preoperative risk factors. The solid line represents patients with mRCC who underwent medical therapy alone (reference line).

our results, we identified 7 variables that would be known before CN and that had a negative association with OS: a serum albumin level below the lower limit of normal, a serum LDH level above the upper limit of normal, radiographic evidence (!1 cm) of retroperitoneal or supradiaphragmatic adenopathy, symptoms at presentation caused by a site of metastasis, clinical T3 or T4 disease, and the presence of liver metastasis. Patients who had !4 factors did not benefit from surgery, because their OS was no different from the cohort of patients with mRCC who received medical therapy alone.

3384

Cancer

July 15, 2010

Patients Who Will Not Benefit From CN/Culp et al

Figure 2. This chart illustrates a Kaplan-Meier analysis of overall survival for patients with metastatic renal cell carcinoma (mRCC) who underwent cytoreductive nephrectomy (CN) based on the number of preoperative risk factors ( 3 vs !4; P < .001). The solid line represents patients with mRCC who underwent medical therapy alone (reference line).

Our results demonstrated that both liver metastasis and clinical T3 or T4 disease were negative predictors of survival in patients who underwent CN. These results were consistent with previous studies.30 Multiple preoperative laboratory variables have been identified that influence the survival of patients with mRCC. These include thrombocytosis,32 hemoglobin levels less than the lower limit of normal,33 elevations in C-reactive protein (CRP),34,35 TSH,31 the erythrocyte sedimentation rate (ESR),36 LDH (1.5 times the upper limit of normal),15 total serum calcium,15 and corrected serum calcium. In the current study, we observed that both a preoperative serum albumin level below the lower limit of normal and a serum LDH level above the upper limit of normal were associated negatively with OS in patients who underwent CN. Although few reports have examined serum albumin levels in the context of survival among patients with mRCC, our results make biologic sensepatients with low preoperative serum albumin levels likely suffer from malnourishment and depleted reserves, factors that theoretically would correlate with surgical recovery and survival. It is noteworthy that, although preoperative alkaline phosphatase and hemoglobin levels differed between the 2 surgical groups, neither was significant in multivariate analysis. In addition, unlike previous reports, we did not observe that preoperative serum total calcium or corrected calcium levels predicted the OS of patients after CN, even when those variables were included in the multivariate analysis. Unfortunately, few patients underwent TSH, CRP, or ESR testing before CN; thus, we could not

include these variables in our analyses. Therefore, in the current study, we could not assess whether or not preoperative values of 1 or more of these variables also may be associated with the OS of patients who undergo CN. Multiple series have demonstrated that lymph node status is an important predictor of survival in patients with mRCC.37,38 Vasselli et al., in a retrospective review of 154 patients with mRCC who underwent CN, observed that the median survival was shorter for patients who had preoperative evidence (>1 cm) of retroperitoneal lymph node involvement (8.5 months vs 14.7 months; P < .001). In addition, patients who had a larger retroperitoneal lymph node volume (>50 cm3) and those who had unresectable lymph nodes had a median survival (5.3 months vs 10.5 months and 3.3 months vs 8.6 months, respectively; P < .05).37 Pantuck et al., in a retrospective review of 900 patients who underwent nephrectomy for RCC, reported that patients who had regional lymph node involvement alone had the same survival as patients who had distant metastases only, and patients who had both regional lymph node involvement and distant disease had a significant decrease in survival compared with patients who had either factor alone.38 Our results are consistent with those studies, in that preoperative evidence of retroperitoneal lymph node involvement was a poor prognostic indicator. However, in addition to retroperitoneal lymph node disease, we observed that preoperative evidence of supradiaphragmatic adenopathy was a separate but equally important negative prognostic indicator, a factor that was not reported in prior studies. Although most of our findings are consistent with previous studies, there are important differences that deserve highlighting. It is noteworthy that, although previous reports have indicated that multiple sites of metastasis are a poor prognostic factor in patient survival after CN,39 we did not observe this in our study, despite 29% of the patients in our series having !2 involved sites. Of the patients who underwent CN with an OS <8.5 months, 62.7% had <2 sites of disease, indicating that factors other than the number of disease sites affected survival in these patients. In addition, although some series have reported that bone metastasis was a poor prognostic factor,9,19,40 we did not note this finding. One potential explanation for this is that we included symptoms at presentation caused by a metastatic site (ie, bone pain) as a variable, and this inclusion may have abrogated the negative effect of having a bone metastasis. However, when we removed this variable from our model, bone metastasis still was not significant in multivariate analysis. Finally,

Cancer

July 15, 2010

3385

Original Article

systemic symptoms at presentation have been associated with poor survival in previous reports,31 but they were not a significant predictor in our study. However, the presence of symptoms because of specific metastasis was significant. Few studies have evaluated symptoms at presentation caused by a metastatic site as a separate variable, and this may explain in part the differences observed in our study compared with previous reports. Yet, when we excluded this variable from our analysis, systemic symptoms at presentation still did not predict inferior survival. Most studies that have examined the outcomes of CN were based on patients who had conventional RCC, because that histology comprises the majority of patients who present with mRCC. In a study by Motzer et al., patients with mRCC and nonclear cell histology represented <10% of those enrolled in clinical trials and demonstrated a worse prognosis.41 In addition to demonstrating a decreased disease-specific survival in patients with mRCC and nonclear cell histology, Kassouf et al., in a retrospective review of 92 patients, observed that those with nonclear cell histology were more likely to be younger, have lymph node involvement, and display sarcomatoid dedifferentiation in the primary tumor.42 In addition to nonclear cell histology, the presence of sarcomatoid features in the tumor was a negative predictor of survival in patients with RCC, even with surgery.43 In our study, 14% and 19.1% of patients who underwent CN had nonclear cell histology and sarcomatoid features on final pathology, respectively; and, although both were negative predictors of survival on multivariate analysis, only the presence of sarcomatoid features was statistically significant (P < .05). In addition, consistent with previous reports,30 a Fuhrman nuclear grade of 4 also was a negative predictor of OS when it was included as a separate variable in multivariate analysis (HR, 1.38; 95% CI, 1.11-1.72; P < .01). Although each of these factors may be known before surgery, percutaneous biopsy of the primary tumor would be required; and recent data based on our own institutions experience suggest that percutaneous biopsy may not be accurate in predicting these factors, particularly in the case of sarcomatoid dedifferentiation and a Fuhrman nuclear grade of 4.44 Because of these findings, along with the finding that only a minority of patients undergo a biopsy before CN, we elected not to include these 3 variables in our risk factor scheme. The current study is limited by its retrospective nature and single-institution experience. Furthermore, our results have yet to be validated in a separate prospective

manner. Although 86% of patients in our study had clear cell RCC, we evaluated patients with all histologies. This heterogeneity may introduce some bias into the results, it nevertheless reflects the actual clinical scenario, because tumor histology is not always known before surgery. Finally, the nonsurgical cohort that we used in our study as a comparator likely represents a selection bias. At our institution, cytoreductive surgery is pursued aggressively in appropriate candidates; therefore, the finding that these patients did not undergo CN speaks to their poor prognosis. We hypothesized that, if the introduction of surgery could not improve survival over this group, then the patient probably did not benefit from surgery. The majority (78.2%) of nonsurgical patients was diagnosed within the last 5 years of the study, and almost 50% received treatment with 1 or more targeted agents. Therefore, the OS of the nonsurgical cohort represents what currently would be expected in patients we would consider not eligible for CN who received targeted therapy. Thus, patients who underwent CN and had the same or a worse survival compared with those who received medical therapy alone clearly did not benefit from surgical intervention. Despite the limitations of the current study, our results are based on a large sample of patients who underwent CN, and our analyses included not only preoperative clinical but also perioperative and pathologic variables to control for factors that may influence patient survival. The 7 preoperative risk factors identified from this study can be used at the time of initial evaluation to help identify patients with mRCC who will not benefit from CN. Therefore, the treatment of these patients can be directed toward upfront systemic therapy, helping them to avoid the unnecessary morbidity of surgical intervention.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES
1. American Cancer Society: Cancer Facts and Figures, 2009. Atlanta, Ga: American Cancer Society; 2009. 2. Chow WH, Devesa SS, Warren JL, Fraumeni JF Jr. Rising incidence of renal cell cancer in the United States. JAMA. 1999;281:1628-1631. 3. Hollingsworth JM, Miller DC, Daignault S, Hollenbeck BK. Five-year survival after surgical treatment for kidney cancer: a population-based competing risk analysis. Cancer. 2007;109:1763-1768. 4. Lam JS, Shvarts O, Leppert JT, Figlin RA, Belldegrun AS. Renal cell carcinoma 2005: new frontiers in staging, prognostication and targeted molecular therapy. J Urol. 2005;173:1853-1862.

3386

Cancer

July 15, 2010

Patients Who Will Not Benefit From CN/Culp et al

5. Flanigan RC, Mickisch G, Sylvester R, Tangen C, Van Poppel H, Crawford ED. Cytoreductive nephrectomy in patients with metastatic renal cancer: a combined analysis. J Urol. 2004;171:1071-1076. 6. Flanigan RC, Salmon SE, Blumenstein BA, et al. Nephrectomy followed by interferon alfa-2b compared with interferon alfa-2b alone for metastatic renal-cell cancer. N Engl J Med. 2001;345:1655-1659. 7. Mickisch GH, Garin A, van Poppel H, de Prijck L, Sylvester R. Radical nephrectomy plus interferon-alfa-based immunotherapy compared with interferon alfa alone in metastatic renal-cell carcinoma: a randomised trial. Lancet. 2001;358:966-970. 8. Rini BI, Campbell SC. The evolving role of surgery for advanced renal cell carcinoma in the era of molecular targeted therapy. J Urol. 2007;177:1978-1984. 9. Fallick ML, McDermott DF, LaRock D, Long JP, Atkins MB. Nephrectomy before interleukin-2 therapy for patients with metastatic renal cell carcinoma. J Urol. 1997;158:16911695. 10. Greene FL, Page DL, Fleming ID, et al., eds. AJCC Cancer Staging Manual. 6th ed. New York: Springer-Verlag; 2002. 11. Eble J, Sauter G, Epstein J, Sesterhenn I, eds. Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. Lyon, France: IARC Press; 2004. 12. Dekernion JB, Ramming KP, Smith RB. The natural history of metastatic renal cell carcinoma: a computer analysis. J Urol. 1978;120:148-152. 13. Montie JE, Stewart BH, Straffon RA, Banowsky LH, Hewitt CB, Montague DK. The role of adjunctive nephrectomy in patients with metastatic renal cell carcinoma. J Urol. 1977;117:272-275. 14. Marcus SG, Choyke PL, Reiter R, et al. Regression of metastatic renal cell carcinoma after cytoreductive nephrectomy. J Urol. 1993;150(2 pt 1):463-466. 15. Motzer RJ, Mazumdar M, Bacik J, Berg W, Amsterdam A, Ferrara J. Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol. 1999;17:2530-2540. 16. Motzer RJ, Russo P. Systemic therapy for renal cell carcinoma. J Urol. 2000;163:408-417. 17. Minasian LM, Motzer RJ, Gluck L, Mazumdar M, Vlamis V, Krown SE. Interferon alfa-2a in advanced renal cell carcinoma: treatment results and survival in 159 patients with long-term follow-up. J Clin Oncol. 1993;11:1368-1375. 18. Muss HB, Costanzi JJ, Leavitt R, et al. Recombinant alfa interferon in renal cell carcinoma: a randomized trial of 2 routes of administration. J Clin Oncol. 1987;5:286-291. 19. Mani S, Todd MB, Katz K, Poo WJ. Prognostic factors for survival in patients with metastatic renal cancer treated with biological response modifiers. J Urol. 1995;154:35-40. 20. Rackley R, Novick A, Klein E, Bukowski R, McLain D, Goldfarb D. The impact of adjuvant nephrectomy on multimodality treatment of metastatic renal cell carcinoma. J Urol. 1994;152(5 pt 1):1399-1403. 21. Wolf JS Jr, Aronson FR, Small EJ, Carroll PR. Nephrectomy for metastatic renal cell carcinoma: a component of systemic treatment regimens. J Surg Oncol. 1994;55:7-13. 22. Walther MM, Yang JC, Pass HI, Linehan WM, Rosenberg SA. Cytoreductive surgery before high dose interleukin-2 based therapy in patients with metastatic renal cell carcinoma. J Urol. 1997;158:1675-1678.

23. Levy DA, Swanson DA, Slaton JW, Ellerhorst J, Dinney CP. Timely delivery of biological therapy after cytoreductive nephrectomy in carefully selected patients with metastatic renal cell carcinoma. J Urol. 1998;159:1168-1173. 24. Rini BI, Campbell SC, Escudier B. Renal cell carcinoma. Lancet. 2009;373:1119-1132. 25. Freed SZ. Nephrectomy for renal cell carcinoma with metastases. Urology. 1977;9:613-616. 26. Spencer WF, Linehan WM, Walther MM, et al. Immunotherapy with interleukin-2 and alpha-interferon in patients with metastatic renal cell cancer with in situ primary cancers: a pilot study. J Urol. 1992;147:24-30. 27. Robertson CN, Linehan WM, Pass HI, et al. Preparative cytoreductive surgery in patients with metastatic renal cell carcinoma treated with adoptive immunotherapy with interleukin-2 or interleukin-2 plus lymphokine activated killer cells. J Urol. 1990;144:614-617; discussion 617-618. 28. Rayman P, Wesa AK, Richmond AL, et al. Effect of renal cell carcinomas on the development of type 1 T-cell responses. Clin Cancer Res. 2004;10(18 pt 2):6360S6366S. 29. Jacobsen J, Rasmuson T, Grankvist K, Ljungberg B. Vascular endothelial growth factor as prognostic factor in renal cell carcinoma. J Urol. 2000;163:343-347. 30. Leibovich BC, Cheville JC, Lohse CM, et al. A scoring algorithm to predict survival for patients with metastatic clear cell renal cell carcinoma: a stratification tool for prospective clinical trials. J Urol. 2005;174:1759-1763; discussion 1763. 31. Leibovich BC, Han KR, Bui MH, et al. Scoring algorithm to predict survival after nephrectomy and immunotherapy in patients with metastatic renal cell carcinoma: a stratification tool for prospective clinical trials. Cancer. 2003;98:2566-2575. 32. Suppiah R, Shaheen PE, Elson P, et al. Thrombocytosis as a prognostic factor for survival in patients with metastatic renal cell carcinoma. Cancer. 2006;107:1793-1800. 33. Choueiri TK, Garcia JA, Elson P, et al. Clinical factors associated with outcome in patients with metastatic clearcell renal cell carcinoma treated with vascular endothelial growth factor-targeted therapy. Cancer. 2007;110:543550. 34. Fujikawa K, Matsui Y, Oka H, Fukuzawa S, Takeuchi H. Serum C-reactive protein level and the impact of cytoreductive surgery in patients with metastatic renal cell carcinoma. J Urol. 1999;162:1934-1937. 35. Karakiewicz PI, Hutterer GC, Trinh QD, et al. C-reactive protein is an informative predictor of renal cell carcinomaspecific mortality: a European study of 313 patients. Cancer. 2007;110:1241-1247. 36. Lee SE, Byun SS, Han JH, Han BK, Hong SK. Prognostic significance of common preoperative laboratory variables in clear cell renal cell carcinoma. BJU Int. 2006;98:1228-1232. 37. Vasselli JR, Yang JC, Linehan WM, White DE, Rosenberg SA, Walther MM. Lack of retroperitoneal lymphadenopathy predicts survival of patients with metastatic renal cell carcinoma. J Urol. 2001;166:68-72. 38. Pantuck AJ, Zisman A, Dorey F, et al. Renal cell carcinoma with retroperitoneal lymph nodes. Impact on survival and benefits of immunotherapy. Cancer. 2003;97:2995-3002. 39. Han KR, Pantuck AJ, Bui MH, et al. Number of metastatic sites rather than location dictates overall survival of patients

Cancer

July 15, 2010

3387

Original Article
with node-negative metastatic renal cell carcinoma. Urology. 2003;61:314-319. 40. Tatokoro M, Saito K, Iimura Y, Fujii Y, Kawakami S, Kihara K. Prognostic impact of postoperative C-reactive protein level in patients with metastatic renal cell carcinoma undergoing cytoreductive nephrectomy. J Urol. 2008;180: 515-519. 41. Motzer RJ, Bacik J, Mariani T, Russo P, Mazumdar M, Reuter V. Treatment outcome and survival associated with metastatic renal cell carcinoma of nonclear-cell histology. J Clin Oncol. 2002;20:2376-2381. 42. Kassouf W, Sanchez-Ortiz R, Tamboli P, et al. Cytoreductive nephrectomy for metastatic renal cell carcinoma with nonclear cell histology. J Urol. 2007;178:1896-1900. 43. Cangiano T, Liao J, Naitoh J, Dorey F, Figlin R, Belldegrun A. Sarcomatoid renal cell carcinoma: biologic behavior, prognosis, and response to combined surgical resection and immunotherapy. J Clin Oncol. 1999;17:523-528. 44. Abel EJ, Culp SH, Stamatakis L, et al. Percutaneous primary tumor biopsy is unreliable to predict high risk pathologic features in patients with metastatic renal cell carcinoma. J Urol. 2009;181(1 suppl):501.

3388

Cancer

July 15, 2010