- Email: [email protected]
Location of Extrarenal Tumor Extension Does Not Impact Survival of Patients With pT3a Renal Cell Carcinoma
Oncology: Adrenal/Renal/Upper Tract/Bladder Location of Extrarenal Tumor Extension Does Not Impact Survival of Patients With pT3a Renal Cell Carcinoma Vitaly Margulis, Pheroze Tamboli, Surena F. Matin, Matthew Meisner, David A. Swanson and Christopher G. Wood* From the Departments of Urology and Pathology (PT), University of Texas M. D. Anderson Cancer Center, Houston, Texas
Purpose: The current TNM classification for pathological pT3a renal cell carcinoma includes patients with perinephric or sinus fat invasion, suggesting that the prognoses are similar for these pathological findings. However, sinus fat invasion was proposed by some investigators to be an independent predictor of inferior cancer specific outcome following surgical treatment. To assess and improve the predictive ability of the current pT3a primary tumor classification we evaluated the prognostic significance of location of extrarenal tumor extension on cancer specific survival following surgery. Materials and Methods: The database of 3,470 patients at our institution who were treated for renal cell carcinoma from 1990 to 2006 was searched for those with pT3a tumors managed by partial or radical nephrectomy. Patients with nonrenal cell carcinoma histology, direct adrenal invasion or a followup of less then 6 months were excluded from analyses. The prognostic importance of all clinical and pathological variables was investigated using Cox proportional hazards regression. Results: A total of 365 patients with pT3a renal cell carcinoma and a mean followup of 33.5 months (range 6.1 to 158.6) met study inclusion criteria and they comprise the data set for this analysis. There was no difference in 5-year cancer specific survival between 166 patients (45.5%) with SF invasion and 199 (54.5%) with PF invasion only (50.8% and 54.1%, p ⫽ 0.782 respectively). On univariate analyses neither sinus fat invasion nor the location of extrarenal extension, assessed as perinephric fat vs sinus fat vs perinephric plus sinus fat, correlated with cancer specific survival following surgical treatment (HR 1.052, p ⫽ 0.783 and HR 1.072, p ⫽ 0.543, respectively). After adjusting for the effects of nodal and systemic metastases tumor grade and sarcomatoid differentiation remained independent predictors of renal cell carcinoma specific survival in our pT3a cohort of patients (HR 1.508, p ⫽ 0.003 and HR 1.810, p ⫽ 0.018, respectively). Conclusions: In contrast to previously reported observations, in our cohort of surgically treated patients with pT3a renal cell carcinoma the location of extrarenal extension was not an important prognosticator of cancer specific mortality. Based on our findings we confirm that perinephric and/or sinus fat should be similarly subclassified in the primary tumor staging system. Key Words: kidney; carcinoma, renal cell; nephrectomy; neoplasm staging; prognosis
ithin the emerging framework of integrative staging algorithms for predicting overall survival and CSS pathological RCC stage remains the most important determinant of oncological outcome following surgical treatment.1,2 Reflecting changing trends in RCC presentation, treatment and outcomes, the AJCC TNM tumor classification, which is the most commonly used anatomical staging system for classifying RCC, has been continuously modified to improve its prognostic accuracy.3 Based on the observations of Bonsib et al4 and experience with pediatric renal tumors5 the renal sinus compartment has been recognized as an important pathway for tumor dissemination from otherwise organ confined RCC. Subsequently in 2002 the AJCC pT3a RCC primary tumor classification category was modified to include tumors with renal SF invasion, in addition to tumors with PF invasion and
W
Submitted for publication February 26, 2007. Study received approval from the Institutional Review Board for the Protection of Human Subjects at University of Texas M. D. Anderson Cancer Center. * Correspondence: Department of Urology, Unit 1373, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, Texas 77030 (telephone: 713-792-3250; FAX: 713-7923474; e-mail: [email protected]).
0022-5347/07/1785-1878/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
direct invasion of the ipsilateral adrenal gland.6 Nonetheless, there is growing belief that the current pT3a classification incorporates tumors with a wide spectrum of oncological outcomes.7,8 For example, groups at several centers have clearly noted that tumors with direct ipsilateral adrenal invasion should be staged in the pT4 category since they behave more aggressively than tumors involving PF.9,10 Recently Thompson et al found that patients with SF invasion were 63% more likely to die of RCC than those with PF invasion only.11 These investigators urged external validation before implementing changes to future revisions of the TNM staging system. To suggest improvements to the current primary tumor classification we performed survival analysis in a large group of patients who underwent partial or radical nephrectomy for pT3a RCC. We evaluated the impact of the location of extrarenal tumor extension on CSS.
MATERIALS AND METHODS All studies were performed with the approval and institutional oversight of the Institutional Review Board for the Protection of Human Subjects at University of Texas M. D. Anderson Cancer Center. The institutional nephrectomy da-
1878
Vol. 178, 1878-1882, November 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.07.011
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL tabase of 3,470 entries, including data from 1990 to 2006, was searched for patients with pT3a RCC treated with partial or radical nephrectomy. Patients with less then 6 months of followup, incomplete resection, direct adrenal invasion or nonRCC pathology were excluded from analysis. The presence or absence of synchronous metastases was routinely evaluated with computerized tomography, magnetic resonance imaging and chest x-rays. Additional studies, such as nuclear bone scans and brain imaging, were done when indicated. Table 1 lists clinical and pathological variables assessed in each patient. RCC staging was assigned according to the AJCC 2002 TNM classification.6 Tumor grade was determined using the Fuhrman grading system, while tumor histology was classified according to 2004 WHO criteria.12 PF and SF invasion were defined as RCC cells directly infiltrating PF and renal SF cells, respectively. Tumors confined by a pseudocapsule but bulging into PF and/or renal SF without clear evidence of infiltration were not classified as showing PF and/or SF invasion. Lymphovascular invasion was defined as tumor cells in an endothelium lined space. All pathological findings were reviewed by experienced genitourinary pathologists specializing in RCC who were blinded to patient outcomes. Followup consisted of physical examination, serum chemistry evaluation, liver function tests, chest radiography and abdominal computer-
1879
ized tomography, which were performed semiannually for the first 2 years and annually thereafter. The Fisher exact and chi-square tests were used to evaluate the association between clinical and pathological variables. Differences in variables with a continuous distribution across dichotomous and ranked categories were assessed using the Mann-Whitney U test and KruskalWallis nonparametric ANOVA, respectively. The KaplanMeier method was used to calculate survival functions and differences were assessed with the log rank statistic. Univariate and multivariate survival analyses were performed using the Cox proportional hazard regression model. Statistical significance in this study was considered at p ⱕ0.050. All reported p values are 2-sided. All analyses were performed with SPSS®, version 13.0. RESULTS A total of 365 patients treated with partial (10) or radical nephrectomy for pT3a RCC met protocol inclusion criteria. At a mean followup of 33.5 months (median 22.5, range 6.1 to 158.4) 131 patients (35.9%) had died of RCC, 24 (6.6%) had died of other causes, 59 (16.2%) had disease and 151 (41.4%) had no evidence of disease recurrence. Mean followup in surviving patients was 37.7 months (median 25.6, range 6.1 to 158.4).
TABLE 1. Clinical and pathological characteristics, and extrarenal extension location in patients with 365 pT3a RCC p Value
No. pts (%) Mean ⫾ SD age No. sex (%): F M No. side (%): Lt Rt No. Eastern Cooperative Oncology Group performance status (%): 0 1 No. presenting symptoms (%): None Local Systemic No. Fuhrman grade (%): 2 3 4 Mean ⫾ SD tumor size (cm) No. lymphovascular invasion (%): Absent Present No. pelvicaliceal system invasion (%): Absent Present No. histology (%): Conventional Nonconventional No. sarcomatoid differentiation (%): Absent Present No. lymph node status (%): N0, Nx N1, N2 No. systemic metastases (%): M0 M1
PF or SF vs PF ⫹ SF
All
PF
SF
PF ⫹ SF
365 58.2 ⫾ 10.9
199 (54.5) 57.5 ⫾ 10.6
96 (26.3) 60.7 ⫾ 11.1
70 (19.2) 56.4 ⫾ 10.9
0.190
0.434
101 (27.7) 264 (72.3)
49 (24.6) 150 (75.4)
33 (34.4) 63 (65.6)
19 (27.1) 51 (72.9)
0.096
0.749
176 (48.2) 189 (51.8)
93 (46.7) 106 (53.3)
42 (43.8) 54 (56.2)
41 (58.6) 29 (41.1)
0.708
0.088
310 (84.9) 55 (15.1)
170 (82.9) 29 (17.1)
80 (80.0) 16 (20.0)
60 (83.4) 10 (16.6)
0.373
0.455
242 (66.3) 102 (27.9) 74 (20.3)
130 (65.3) 52 (26.1) 41 (20.6)
65 (67.7) 31 (32.2) 22 (22.9)
47 (67.0) 19 (27.1) 11 (15.7)
0.563
0.677
71 (19.9) 174 (47.7) 120 (32.9) 8.8 ⫾ 4.1
46 (23.1) 91 (45.7) 62 (31.2) 8.7 ⫾ 4.5
18 (18.8) 53 (55.2) 25 (26.0) 8.3 ⫾ 3.8
7 (10.0) 30 (42.9) 33 (47.1) 9.5 ⫾ 3.5
0.311 0.422
0.015 0.228
318 (87.1) 47 (12.9)
183 (92.0) 16 (8.0)
76 (79.2) 20 (20.8)
59 (84.3) 11 (15.7)
0.063
0.103
331 (90.7) 34 (9.3)
194 (97.5) 5 (2.5)
82 (85.4) 14 (14.6)
55 (78.6) 15 (21.4)
⬍0.001
⬍0.001
291 (79.7) 74 (20.3)
155 (77.9) 44 (22.1)
80 (83.3) 16 (16.7)
56 (80.0) 14 (20.0)
0.354
0.866
328 (89.9) 37 (10.1)
180 (90.5) 19 (9.5)
90 (93.8) 6 (6.3)
58 (82.9) 12 (17.1)
0.236
0.125
316 (86.6) 49 (13.4)
179 (89.9) 20 (10.1)
87 (90.6) 9 (9.4)
50 (71.4) 20 (28.6)
1.000
⬍0.001
227 (62.2) 138 (37.8)
126 (63.3) 73 (36.7)
70 (72.9) 26 (27.1)
31 (44.3) 39 (55.7)
0.115
0.007
PF vs SF
1880
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL
Table 1 lists the clinical and pathological patient characteristics analyzed in the outcome analyses. PF invasion, SF invasion and the concomitant presence of the 2 pathological features were identified in 199 (54.5%), 96 (26.3%) and 70 study patients (19.2%), respectively. Except for the increased likelihood of collecting system involvement (which was observed in patients with SF invasion only) compared to patients with PF invasion only none of the other clinical and pathological variables differed between these patient groups (14.6% vs 2.5%, p ⬍0.001, table 1). Patients with concomitant tumor infiltration of PF plus SF were more likely to have high grade RCC and harbor lymph node and/or systemic metastases compared to patients with PF or SF involvement (p ⫽ 0.015, ⬍0.001 and 0.007, respectively, table 1). Kaplan-Meier analyses performed in the whole cohort of patients with pT3a revealed similar 5-year CSS in patients without SF invasion and in patients with SF invasion by RCC (53% vs 50%, log rank p ⫽ 0.782, fig. 1). Similarly no survival differences were observed when patients were stratified according to the location of extrarenal tumor extension (fig. 2). The 5-year CSS rate in patients with PF involvement only, SF involvement only and those with the 2 features was 53%, 50% and 49%, respectively. Interestingly despite a higher frequency of features consistent with aggressive RCC patients with PF plus SF invasion had CSS equivalent to that in patients with PF or SF invasion only. Subgroup analyses performed in 208 patients with nonmetastatic, node negative RCC (pT3aNx/N0M0) revealed that those with SF invasion had 5-year CSS comparable to that in patients without SF invasion by RCC (78% vs 79%, log rank p ⫽ 0.878). In addition, no 5-year CSS differences were observed when patients with pT3aNx/N0M0 were stratified according to the location of extrarenal tumor extension, including 79% for PF, 72% for SF and 83% for PF plus SF invasion (PF vs SF, PF vs PF plus SF and SF vs PF plus SF log rank p ⫽ 0.647, 0.278 and 0.173, respectively). Univariate Cox proportional hazard regression analysis of all patients with pT3a demonstrated that tumor size, grade, pathological evidence of sarcomatoid differentiation,
FIG. 1. Probability of CSS after nephrectomy for pT3a RCC stratified according to presence or absence of renal SF invasion.
FIG. 2. Probability of CSS after nephrectomy for pT3a RCC stratified according to extrarenal extension location
and nodal and systemic metastases correlated with the risk of death from RCC (table 2). SF invasion and extrarenal extension location did not correlate with kidney CSS (HR 1.052, p ⫽ 0.783 and HR 1.072, p ⫽ 0.543, respectively). After adjusting for nodal and systemic metastases tumor grade and sarcomatoid differentiation in the primary lesion remained independent predictors of CSS (HR 1.508, p ⫽ 0.003 and HR 1.81, p ⫽ 0.018, respectively, table 2). DISCUSSION The renal sinus is a fatty compartment located in the confines of the kidney that envelopes the renal collecting system and contains numerous veins and lymphatics.13 The abundant lymphovascular network of the renal SF, similar to that of the PF, provides an important avenue for metastatic tumor dissemination. Logically the most recent version of the AJCC TNM classification system for RCC incorporated otherwise organ confined pT1 or pT2 tumors with renal SF invasion into the pT3a staging category.6 The question raised by Thompson et al11 and addressed in this study is whether invasion of the renal SF by RCC portends a worse prognosis in the pT3a RCC category and whether the location of extrarenal tumor extension should be reflected in the upcoming revision of the primary tumor classification system. We found that, compared to patients with PF invasion only, patients in whom SF invasion was identified had similar 5-year CSS and were at no increased risk of death from RCC. Furthermore, similar oncological outcome was observed after stratifying all 365 patients with pT3a RCC by extrarenal extension location (PF vs SF vs PF plus SF). Likewise in subgroup analysis of 208 patients with nonmetastatic, node negative RCC SF invasion or extrarenal extension location did not correlate with the risk of death from RCC. Patients in whom tumors were found to have PF invasion and patients with tumors invading the renal SF had similar clinical and pathological features. Due to their proximity to the renal pelvis tumors with SF invasion were more likely to involve the urothelium of the renal collecting
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL
1881
TABLE 2. Univariate and multivariate Cox regression analysis of clinical and pathological features for predicting disease specific mortality in 365 patients with pT3a treated with partial or radical nephrectomy Univariate
Age Sex Histology* Eastern Cooperative Oncology Group performance status Presenting symptoms Sarcomatoid differentiation Fuhrman grade Tumor size Lymphovascular invasion Renal SF invasion Extrarenal extension location† Pelvicaliceal system invasion Lymph node metastases Systemic metastases
Multivariate
RR
95% CI
p Value
0.994 1.116 0.823 0.744 1.223 3.338 2.223 1.063 1.225 1.052 1.072 0.856 3.480 5.336
0.979–1.010 0.762–1.632 0.516–1.313 0.523–1.416 1.116–1.313 2.119–5.259 1.708–2.893 1.025–1.102 0.744–2.017 0.735–1.504 0.857–1.340 0.449–1.633 2.310–5.243 3.670–7.757
0.486 0.573 0.414 0.526 0.618 ⬍0.001 ⬍0.001 0.001 0.425 0.783 0.543 0.637 ⬍0.001 ⬍0.001
RR
95% CI
p Value
1.810 1.508 1.013
1.107–2.958 1.145–1.985 0.969–1.059
0.018 0.003 0.560
2.705 4.298
1.786–4.097 2.920–6.325
⬍0.001 ⬍0.001
* Assessed as conventional vs nonconventional histology. † Assessed as PF vs renal SF vs PF plus SF.
system. We were surprised that, despite harboring features of aggressive disease, patients with PF plus SF invasion had CSS similar to that in patients with PF or SF invasion. Unequal distribution of patients in the study groups and the relatively small number of patients with PF plus SF invasion could have affected statistical analyses. After adjusting for the effects of lymph node and systemic metastasis only nuclear grade and sarcomatoid differentiation independently predicted the risk of death from RCC. Taken as a whole our data suggest that extrarenal extension location does not reflect the biology of RCC, but rather mirrors the origin of the tumor in the renal parenchyma. In what is to our knowledge the only other study of the implication of location extrarenal tumor extension Thompson et al reported significantly decreased cancer specific survival in 43 patients (21%) with SF invasion compared to that in 162 (79%) with PF invasion only.11 The diametrically opposite conclusion supported by our data may be explained by a larger cohort of patients and higher frequency of renal SF invasion in our series. In agreement with the findings of Bonsib et al, who prospectively evaluated specimens using thin sections of renal sinus and found SF invasion in 55% of tumors,4 we documented renal SF invasion in 56% of patients with extrarenal tumor extension, while Thompson et al identified renal SF invasion in only 21% of their T3a cohort.11 A potential limitation of the current study should be noted. Pathology specimens obtained from patients in the current series were not submitted for pathological reevaluation. Consequently a standardized definition of SF invasion may not have been uniformly applied and interobserver variability likely exists. Nonetheless, we report one of the largest single center series of pT3a RCC treated in the modern era, in which the presence and location of extrarenal tumor extension was prospectively recorded by genitourinary pathologists. CONCLUSIONS Survival analysis of a large group of patients who underwent surgical therapy for pT3a RCC revealed that, contrary to another recently published analysis, SF invasion in the RCC specimen and the location of extrarenal tumor exten-
sion were not associated with an increased risk of death from renal cancer. Based on our findings we suggest that tumors found to have PF and/or SF invasion should be similarly subclassified in the RCC primary tumor staging system.
Abbreviations and Acronyms AJCC CSS PF RCC SF
⫽ ⫽ ⫽ ⫽ ⫽
American Joint Committee on Cancer cancer specific survival perinephric fat renal cell carcinoma sinus fat
REFERENCES 1.
2.
3.
4. 5.
6.
7.
8.
9.
Frank I, Blute ML, Leibovich BC, Cheville JC, Lohse CM and Zincke H: Independent validation of the 2002 American Joint Committee on cancer primary tumor classification for renal cell carcinoma using a large, single institution cohort. J Urol 2005; 173: 1889. Tsui KH, Shvarts O, Smith RB, Figlin RA, deKernion JB and Belldegrun A: Prognostic indicators for renal cell carcinoma: a multivariate analysis of 643 patients using the revised 1997 TNM staging criteria. J Urol 2000; 163: 1090. Gospodarowicz MK, Miller D, Groome PA, Greene FL, Logan PA and Sobin LH: The process for continuous improvement of the TNM classification. Cancer 2004; 100: 1. Beckwith JB: National Wilms Tumor Study: an update for pathologists. Pediatr Dev Pathol 1998; 1: 79. Bonsib SM: The renal sinus is the principal invasive pathway: a prospective study of 100 renal cell carcinomas. Am J Surg Pathol 2004; 28: 1594. Greene FL: American Joint Committee on Cancer, American Cancer Society. AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag 2002; pp xiv and 421. Siemer S, Lehmann J, Loch A, Becker F, Stein U, Schneider G et al: Current TNM classification of renal cell carcinoma evaluated: revising stage T3a. J Urol 2005; 173: 33. Leibovich BC, Cheville JC, Lohse CM, Zincke H, Kwon ED, Frank I et al: Cancer specific survival for patients with pT3 renal cell carcinoma-can the 2002 primary tumor classification be improved? J Urol 2005; 173: 716. Thompson RH, Leibovich BC, Cheville JC, Lohse CM, Frank I, Kwon ED et al: Should direct ipsilateral adrenal invasion
1882
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL
from renal cell carcinoma be classified as pT3a? J Urol 2005; 173: 918. 10. Alamdari FI and Ljungberg B: Adrenal metastasis in renal cell carcinoma: a recommendation for adjustment of the TNM staging system. Scand J Urol Nephrol 2005; 39: 277. 11. Thompson RH, Leibovich BC, Cheville JC, Webster WS, Lohse CM, Kwon ED et al: Is renal sinus fat invasion the same as
12.
13.
perinephric fat invasion for pT3a renal cell carcinoma? J Urol 2005; 174: 1218. Eble JN: World Health Organization: Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. Lyon: IARC Press 2004; p 359. Satyapal KS: Classification of the drainage patterns of the renal veins. J Anat 186: 329, 1995.
Purpose: The current TNM classification for pathological pT3a renal cell carcinoma includes patients with perinephric or sinus fat invasion, suggesting that the prognoses are similar for these pathological findings. However, sinus fat invasion was proposed by some investigators to be an independent predictor of inferior cancer specific outcome following surgical treatment. To assess and improve the predictive ability of the current pT3a primary tumor classification we evaluated the prognostic significance of location of extrarenal tumor extension on cancer specific survival following surgery. Materials and Methods: The database of 3,470 patients at our institution who were treated for renal cell carcinoma from 1990 to 2006 was searched for those with pT3a tumors managed by partial or radical nephrectomy. Patients with nonrenal cell carcinoma histology, direct adrenal invasion or a followup of less then 6 months were excluded from analyses. The prognostic importance of all clinical and pathological variables was investigated using Cox proportional hazards regression. Results: A total of 365 patients with pT3a renal cell carcinoma and a mean followup of 33.5 months (range 6.1 to 158.6) met study inclusion criteria and they comprise the data set for this analysis. There was no difference in 5-year cancer specific survival between 166 patients (45.5%) with SF invasion and 199 (54.5%) with PF invasion only (50.8% and 54.1%, p ⫽ 0.782 respectively). On univariate analyses neither sinus fat invasion nor the location of extrarenal extension, assessed as perinephric fat vs sinus fat vs perinephric plus sinus fat, correlated with cancer specific survival following surgical treatment (HR 1.052, p ⫽ 0.783 and HR 1.072, p ⫽ 0.543, respectively). After adjusting for the effects of nodal and systemic metastases tumor grade and sarcomatoid differentiation remained independent predictors of renal cell carcinoma specific survival in our pT3a cohort of patients (HR 1.508, p ⫽ 0.003 and HR 1.810, p ⫽ 0.018, respectively). Conclusions: In contrast to previously reported observations, in our cohort of surgically treated patients with pT3a renal cell carcinoma the location of extrarenal extension was not an important prognosticator of cancer specific mortality. Based on our findings we confirm that perinephric and/or sinus fat should be similarly subclassified in the primary tumor staging system. Key Words: kidney; carcinoma, renal cell; nephrectomy; neoplasm staging; prognosis
ithin the emerging framework of integrative staging algorithms for predicting overall survival and CSS pathological RCC stage remains the most important determinant of oncological outcome following surgical treatment.1,2 Reflecting changing trends in RCC presentation, treatment and outcomes, the AJCC TNM tumor classification, which is the most commonly used anatomical staging system for classifying RCC, has been continuously modified to improve its prognostic accuracy.3 Based on the observations of Bonsib et al4 and experience with pediatric renal tumors5 the renal sinus compartment has been recognized as an important pathway for tumor dissemination from otherwise organ confined RCC. Subsequently in 2002 the AJCC pT3a RCC primary tumor classification category was modified to include tumors with renal SF invasion, in addition to tumors with PF invasion and
W
Submitted for publication February 26, 2007. Study received approval from the Institutional Review Board for the Protection of Human Subjects at University of Texas M. D. Anderson Cancer Center. * Correspondence: Department of Urology, Unit 1373, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, Texas 77030 (telephone: 713-792-3250; FAX: 713-7923474; e-mail: [email protected]).
0022-5347/07/1785-1878/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
direct invasion of the ipsilateral adrenal gland.6 Nonetheless, there is growing belief that the current pT3a classification incorporates tumors with a wide spectrum of oncological outcomes.7,8 For example, groups at several centers have clearly noted that tumors with direct ipsilateral adrenal invasion should be staged in the pT4 category since they behave more aggressively than tumors involving PF.9,10 Recently Thompson et al found that patients with SF invasion were 63% more likely to die of RCC than those with PF invasion only.11 These investigators urged external validation before implementing changes to future revisions of the TNM staging system. To suggest improvements to the current primary tumor classification we performed survival analysis in a large group of patients who underwent partial or radical nephrectomy for pT3a RCC. We evaluated the impact of the location of extrarenal tumor extension on CSS.
MATERIALS AND METHODS All studies were performed with the approval and institutional oversight of the Institutional Review Board for the Protection of Human Subjects at University of Texas M. D. Anderson Cancer Center. The institutional nephrectomy da-
1878
Vol. 178, 1878-1882, November 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.07.011
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL tabase of 3,470 entries, including data from 1990 to 2006, was searched for patients with pT3a RCC treated with partial or radical nephrectomy. Patients with less then 6 months of followup, incomplete resection, direct adrenal invasion or nonRCC pathology were excluded from analysis. The presence or absence of synchronous metastases was routinely evaluated with computerized tomography, magnetic resonance imaging and chest x-rays. Additional studies, such as nuclear bone scans and brain imaging, were done when indicated. Table 1 lists clinical and pathological variables assessed in each patient. RCC staging was assigned according to the AJCC 2002 TNM classification.6 Tumor grade was determined using the Fuhrman grading system, while tumor histology was classified according to 2004 WHO criteria.12 PF and SF invasion were defined as RCC cells directly infiltrating PF and renal SF cells, respectively. Tumors confined by a pseudocapsule but bulging into PF and/or renal SF without clear evidence of infiltration were not classified as showing PF and/or SF invasion. Lymphovascular invasion was defined as tumor cells in an endothelium lined space. All pathological findings were reviewed by experienced genitourinary pathologists specializing in RCC who were blinded to patient outcomes. Followup consisted of physical examination, serum chemistry evaluation, liver function tests, chest radiography and abdominal computer-
1879
ized tomography, which were performed semiannually for the first 2 years and annually thereafter. The Fisher exact and chi-square tests were used to evaluate the association between clinical and pathological variables. Differences in variables with a continuous distribution across dichotomous and ranked categories were assessed using the Mann-Whitney U test and KruskalWallis nonparametric ANOVA, respectively. The KaplanMeier method was used to calculate survival functions and differences were assessed with the log rank statistic. Univariate and multivariate survival analyses were performed using the Cox proportional hazard regression model. Statistical significance in this study was considered at p ⱕ0.050. All reported p values are 2-sided. All analyses were performed with SPSS®, version 13.0. RESULTS A total of 365 patients treated with partial (10) or radical nephrectomy for pT3a RCC met protocol inclusion criteria. At a mean followup of 33.5 months (median 22.5, range 6.1 to 158.4) 131 patients (35.9%) had died of RCC, 24 (6.6%) had died of other causes, 59 (16.2%) had disease and 151 (41.4%) had no evidence of disease recurrence. Mean followup in surviving patients was 37.7 months (median 25.6, range 6.1 to 158.4).
TABLE 1. Clinical and pathological characteristics, and extrarenal extension location in patients with 365 pT3a RCC p Value
No. pts (%) Mean ⫾ SD age No. sex (%): F M No. side (%): Lt Rt No. Eastern Cooperative Oncology Group performance status (%): 0 1 No. presenting symptoms (%): None Local Systemic No. Fuhrman grade (%): 2 3 4 Mean ⫾ SD tumor size (cm) No. lymphovascular invasion (%): Absent Present No. pelvicaliceal system invasion (%): Absent Present No. histology (%): Conventional Nonconventional No. sarcomatoid differentiation (%): Absent Present No. lymph node status (%): N0, Nx N1, N2 No. systemic metastases (%): M0 M1
PF or SF vs PF ⫹ SF
All
PF
SF
PF ⫹ SF
365 58.2 ⫾ 10.9
199 (54.5) 57.5 ⫾ 10.6
96 (26.3) 60.7 ⫾ 11.1
70 (19.2) 56.4 ⫾ 10.9
0.190
0.434
101 (27.7) 264 (72.3)
49 (24.6) 150 (75.4)
33 (34.4) 63 (65.6)
19 (27.1) 51 (72.9)
0.096
0.749
176 (48.2) 189 (51.8)
93 (46.7) 106 (53.3)
42 (43.8) 54 (56.2)
41 (58.6) 29 (41.1)
0.708
0.088
310 (84.9) 55 (15.1)
170 (82.9) 29 (17.1)
80 (80.0) 16 (20.0)
60 (83.4) 10 (16.6)
0.373
0.455
242 (66.3) 102 (27.9) 74 (20.3)
130 (65.3) 52 (26.1) 41 (20.6)
65 (67.7) 31 (32.2) 22 (22.9)
47 (67.0) 19 (27.1) 11 (15.7)
0.563
0.677
71 (19.9) 174 (47.7) 120 (32.9) 8.8 ⫾ 4.1
46 (23.1) 91 (45.7) 62 (31.2) 8.7 ⫾ 4.5
18 (18.8) 53 (55.2) 25 (26.0) 8.3 ⫾ 3.8
7 (10.0) 30 (42.9) 33 (47.1) 9.5 ⫾ 3.5
0.311 0.422
0.015 0.228
318 (87.1) 47 (12.9)
183 (92.0) 16 (8.0)
76 (79.2) 20 (20.8)
59 (84.3) 11 (15.7)
0.063
0.103
331 (90.7) 34 (9.3)
194 (97.5) 5 (2.5)
82 (85.4) 14 (14.6)
55 (78.6) 15 (21.4)
⬍0.001
⬍0.001
291 (79.7) 74 (20.3)
155 (77.9) 44 (22.1)
80 (83.3) 16 (16.7)
56 (80.0) 14 (20.0)
0.354
0.866
328 (89.9) 37 (10.1)
180 (90.5) 19 (9.5)
90 (93.8) 6 (6.3)
58 (82.9) 12 (17.1)
0.236
0.125
316 (86.6) 49 (13.4)
179 (89.9) 20 (10.1)
87 (90.6) 9 (9.4)
50 (71.4) 20 (28.6)
1.000
⬍0.001
227 (62.2) 138 (37.8)
126 (63.3) 73 (36.7)
70 (72.9) 26 (27.1)
31 (44.3) 39 (55.7)
0.115
0.007
PF vs SF
1880
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL
Table 1 lists the clinical and pathological patient characteristics analyzed in the outcome analyses. PF invasion, SF invasion and the concomitant presence of the 2 pathological features were identified in 199 (54.5%), 96 (26.3%) and 70 study patients (19.2%), respectively. Except for the increased likelihood of collecting system involvement (which was observed in patients with SF invasion only) compared to patients with PF invasion only none of the other clinical and pathological variables differed between these patient groups (14.6% vs 2.5%, p ⬍0.001, table 1). Patients with concomitant tumor infiltration of PF plus SF were more likely to have high grade RCC and harbor lymph node and/or systemic metastases compared to patients with PF or SF involvement (p ⫽ 0.015, ⬍0.001 and 0.007, respectively, table 1). Kaplan-Meier analyses performed in the whole cohort of patients with pT3a revealed similar 5-year CSS in patients without SF invasion and in patients with SF invasion by RCC (53% vs 50%, log rank p ⫽ 0.782, fig. 1). Similarly no survival differences were observed when patients were stratified according to the location of extrarenal tumor extension (fig. 2). The 5-year CSS rate in patients with PF involvement only, SF involvement only and those with the 2 features was 53%, 50% and 49%, respectively. Interestingly despite a higher frequency of features consistent with aggressive RCC patients with PF plus SF invasion had CSS equivalent to that in patients with PF or SF invasion only. Subgroup analyses performed in 208 patients with nonmetastatic, node negative RCC (pT3aNx/N0M0) revealed that those with SF invasion had 5-year CSS comparable to that in patients without SF invasion by RCC (78% vs 79%, log rank p ⫽ 0.878). In addition, no 5-year CSS differences were observed when patients with pT3aNx/N0M0 were stratified according to the location of extrarenal tumor extension, including 79% for PF, 72% for SF and 83% for PF plus SF invasion (PF vs SF, PF vs PF plus SF and SF vs PF plus SF log rank p ⫽ 0.647, 0.278 and 0.173, respectively). Univariate Cox proportional hazard regression analysis of all patients with pT3a demonstrated that tumor size, grade, pathological evidence of sarcomatoid differentiation,
FIG. 1. Probability of CSS after nephrectomy for pT3a RCC stratified according to presence or absence of renal SF invasion.
FIG. 2. Probability of CSS after nephrectomy for pT3a RCC stratified according to extrarenal extension location
and nodal and systemic metastases correlated with the risk of death from RCC (table 2). SF invasion and extrarenal extension location did not correlate with kidney CSS (HR 1.052, p ⫽ 0.783 and HR 1.072, p ⫽ 0.543, respectively). After adjusting for nodal and systemic metastases tumor grade and sarcomatoid differentiation in the primary lesion remained independent predictors of CSS (HR 1.508, p ⫽ 0.003 and HR 1.81, p ⫽ 0.018, respectively, table 2). DISCUSSION The renal sinus is a fatty compartment located in the confines of the kidney that envelopes the renal collecting system and contains numerous veins and lymphatics.13 The abundant lymphovascular network of the renal SF, similar to that of the PF, provides an important avenue for metastatic tumor dissemination. Logically the most recent version of the AJCC TNM classification system for RCC incorporated otherwise organ confined pT1 or pT2 tumors with renal SF invasion into the pT3a staging category.6 The question raised by Thompson et al11 and addressed in this study is whether invasion of the renal SF by RCC portends a worse prognosis in the pT3a RCC category and whether the location of extrarenal tumor extension should be reflected in the upcoming revision of the primary tumor classification system. We found that, compared to patients with PF invasion only, patients in whom SF invasion was identified had similar 5-year CSS and were at no increased risk of death from RCC. Furthermore, similar oncological outcome was observed after stratifying all 365 patients with pT3a RCC by extrarenal extension location (PF vs SF vs PF plus SF). Likewise in subgroup analysis of 208 patients with nonmetastatic, node negative RCC SF invasion or extrarenal extension location did not correlate with the risk of death from RCC. Patients in whom tumors were found to have PF invasion and patients with tumors invading the renal SF had similar clinical and pathological features. Due to their proximity to the renal pelvis tumors with SF invasion were more likely to involve the urothelium of the renal collecting
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL
1881
TABLE 2. Univariate and multivariate Cox regression analysis of clinical and pathological features for predicting disease specific mortality in 365 patients with pT3a treated with partial or radical nephrectomy Univariate
Age Sex Histology* Eastern Cooperative Oncology Group performance status Presenting symptoms Sarcomatoid differentiation Fuhrman grade Tumor size Lymphovascular invasion Renal SF invasion Extrarenal extension location† Pelvicaliceal system invasion Lymph node metastases Systemic metastases
Multivariate
RR
95% CI
p Value
0.994 1.116 0.823 0.744 1.223 3.338 2.223 1.063 1.225 1.052 1.072 0.856 3.480 5.336
0.979–1.010 0.762–1.632 0.516–1.313 0.523–1.416 1.116–1.313 2.119–5.259 1.708–2.893 1.025–1.102 0.744–2.017 0.735–1.504 0.857–1.340 0.449–1.633 2.310–5.243 3.670–7.757
0.486 0.573 0.414 0.526 0.618 ⬍0.001 ⬍0.001 0.001 0.425 0.783 0.543 0.637 ⬍0.001 ⬍0.001
RR
95% CI
p Value
1.810 1.508 1.013
1.107–2.958 1.145–1.985 0.969–1.059
0.018 0.003 0.560
2.705 4.298
1.786–4.097 2.920–6.325
⬍0.001 ⬍0.001
* Assessed as conventional vs nonconventional histology. † Assessed as PF vs renal SF vs PF plus SF.
system. We were surprised that, despite harboring features of aggressive disease, patients with PF plus SF invasion had CSS similar to that in patients with PF or SF invasion. Unequal distribution of patients in the study groups and the relatively small number of patients with PF plus SF invasion could have affected statistical analyses. After adjusting for the effects of lymph node and systemic metastasis only nuclear grade and sarcomatoid differentiation independently predicted the risk of death from RCC. Taken as a whole our data suggest that extrarenal extension location does not reflect the biology of RCC, but rather mirrors the origin of the tumor in the renal parenchyma. In what is to our knowledge the only other study of the implication of location extrarenal tumor extension Thompson et al reported significantly decreased cancer specific survival in 43 patients (21%) with SF invasion compared to that in 162 (79%) with PF invasion only.11 The diametrically opposite conclusion supported by our data may be explained by a larger cohort of patients and higher frequency of renal SF invasion in our series. In agreement with the findings of Bonsib et al, who prospectively evaluated specimens using thin sections of renal sinus and found SF invasion in 55% of tumors,4 we documented renal SF invasion in 56% of patients with extrarenal tumor extension, while Thompson et al identified renal SF invasion in only 21% of their T3a cohort.11 A potential limitation of the current study should be noted. Pathology specimens obtained from patients in the current series were not submitted for pathological reevaluation. Consequently a standardized definition of SF invasion may not have been uniformly applied and interobserver variability likely exists. Nonetheless, we report one of the largest single center series of pT3a RCC treated in the modern era, in which the presence and location of extrarenal tumor extension was prospectively recorded by genitourinary pathologists. CONCLUSIONS Survival analysis of a large group of patients who underwent surgical therapy for pT3a RCC revealed that, contrary to another recently published analysis, SF invasion in the RCC specimen and the location of extrarenal tumor exten-
sion were not associated with an increased risk of death from renal cancer. Based on our findings we suggest that tumors found to have PF and/or SF invasion should be similarly subclassified in the RCC primary tumor staging system.
Abbreviations and Acronyms AJCC CSS PF RCC SF
⫽ ⫽ ⫽ ⫽ ⫽
American Joint Committee on Cancer cancer specific survival perinephric fat renal cell carcinoma sinus fat
REFERENCES 1.
2.
3.
4. 5.
6.
7.
8.
9.
Frank I, Blute ML, Leibovich BC, Cheville JC, Lohse CM and Zincke H: Independent validation of the 2002 American Joint Committee on cancer primary tumor classification for renal cell carcinoma using a large, single institution cohort. J Urol 2005; 173: 1889. Tsui KH, Shvarts O, Smith RB, Figlin RA, deKernion JB and Belldegrun A: Prognostic indicators for renal cell carcinoma: a multivariate analysis of 643 patients using the revised 1997 TNM staging criteria. J Urol 2000; 163: 1090. Gospodarowicz MK, Miller D, Groome PA, Greene FL, Logan PA and Sobin LH: The process for continuous improvement of the TNM classification. Cancer 2004; 100: 1. Beckwith JB: National Wilms Tumor Study: an update for pathologists. Pediatr Dev Pathol 1998; 1: 79. Bonsib SM: The renal sinus is the principal invasive pathway: a prospective study of 100 renal cell carcinomas. Am J Surg Pathol 2004; 28: 1594. Greene FL: American Joint Committee on Cancer, American Cancer Society. AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag 2002; pp xiv and 421. Siemer S, Lehmann J, Loch A, Becker F, Stein U, Schneider G et al: Current TNM classification of renal cell carcinoma evaluated: revising stage T3a. J Urol 2005; 173: 33. Leibovich BC, Cheville JC, Lohse CM, Zincke H, Kwon ED, Frank I et al: Cancer specific survival for patients with pT3 renal cell carcinoma-can the 2002 primary tumor classification be improved? J Urol 2005; 173: 716. Thompson RH, Leibovich BC, Cheville JC, Lohse CM, Frank I, Kwon ED et al: Should direct ipsilateral adrenal invasion
1882
EXTRARENAL TUMOR EXTENSION AND RENAL CELL CANCER SURVIVAL
from renal cell carcinoma be classified as pT3a? J Urol 2005; 173: 918. 10. Alamdari FI and Ljungberg B: Adrenal metastasis in renal cell carcinoma: a recommendation for adjustment of the TNM staging system. Scand J Urol Nephrol 2005; 39: 277. 11. Thompson RH, Leibovich BC, Cheville JC, Webster WS, Lohse CM, Kwon ED et al: Is renal sinus fat invasion the same as
12.
13.
perinephric fat invasion for pT3a renal cell carcinoma? J Urol 2005; 174: 1218. Eble JN: World Health Organization: Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. Lyon: IARC Press 2004; p 359. Satyapal KS: Classification of the drainage patterns of the renal veins. J Anat 186: 329, 1995.