- Email: [email protected]
Reevaluation of Renal Cell Carcinoma and Perirenal Fat Invasion Only
Reevaluation of Renal Cell Carcinoma and Perirenal Fat Invasion Only Hwang Gyun Jeon,* In Gab Jeong,* Cheol Kwak, Hyeon Hoe Kim, Sang Eun Lee and Eunsik Lee† From the Departments of Urology, Seoul National University Hospital (HGJ, CK, HHK, EL) and Asan Medical Center (IGJ), Seoul and Seoul National University Bundang Hospital (SEL), Seongnam, Korea
Purpose: Controversy continues over whether perirenal fat invasion in pT3a renal cell carcinoma is a prognostic factor. We investigated the prognosis of renal cell carcinoma with perirenal fat invasion compared to the prognosis of other pathological stages by tumor size. Materials and Methods: We reviewed the medical records of 946 patients who underwent curative surgery for pT1–pT3bN0M0 renal cell carcinoma between 1988 and 2006. Patients with pT3a stage disease and perirenal fat invasion only were divided into 2 subgroups by a 7 cm tumor size cutoff. The prognostic impact of perirenal fat invasion on disease-free and cancer specific survival was investigated. Results: Patients with perirenal fat invasion and lesions greater than 7 cm had lower 5-year disease-free (49.5% vs 77.2%, p ⫽ 0.004) and cancer specific (58.5% vs 95.6%, p ⫽ 0.003) survival than those with lesions 7 cm or less. Patients with perirenal fat invasion and lesions 7 cm or less had similar 5-year disease-free and cancer specific survival compared to those with pT1 tumors (p ⫽ 0.109 and 0.602, respectively). For tumors 7 cm or less multivariate analysis showed that perirenal fat invasion was not a significant predictor of disease-free (p ⫽ 0.119) or cancer specific (p ⫽ 0.208) survival. In contrast, perirenal fat invasion was an independent prognostic factor for disease-free (p ⫽ 0.002) and cancer specific (p ⫽ 0.027) survival in patients with tumors greater than 7 cm. Conclusions: Findings suggest that the prognostic significance of perirenal fat invasion depends on primary tumor size. Perirenal fat invasion included in the pT3a stage regardless of tumor size should be reevaluated by tumor size for a more accurate patient prognosis.
Abbreviations and Acronyms CSS ⫽ cancer specific survival DFS ⫽ disease-free survival ECOG ⫽ Eastern Cooperative Oncology Group PS ⫽ performance status RCC ⫽ renal cell carcinoma Submitted for publication March 17, 2009. Study received Seoul National University Hospital institutional review board approval. * Equal study contribution. † Correspondence: Department of Urology, Seoul National University Hospital, 28, Yongondong, Chongno-gu, Seoul 110-744, Republic of Korea (telephone: 82-2-2072-2915; FAX: 82-2742-4665; e-mail: [email protected]).
Key Words: kidney; carcinoma, renal cell; prognosis; adipose tissue; neoplasm invasiveness ACCURATE cancer staging provides important prognostic information on the risk of cancer recurrence and death, and is important for physicians to make decisions on appropriate cancer treatment. The advent of modern imaging and treatment modalities to assess CSS has led to the update of the TNM classification system for RCC.
In the latest version of the American Joint Committee on Cancer TNM classification system for RCC pT3a is defined as tumors that invade adrenal gland or perirenal fat tissue but do not go beyond Gerota’s fascia.1 Recently the pT3a classification has been the subject of considerable debate. Several groups reported that pT3a tumors directly invading the ipsilateral adrenal gland
0022-5347/09/1825-2137/0 THE JOURNAL OF UROLOGY® Copyright © 2009 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 182, 2137-2143, November 2009 Printed in U.S.A. DOI:10.1016/j.juro.2009.07.065
www.jurology.com
2137
2138
RENAL CELL CANCER AND PERIRENAL FAT INVASION
have an unfavorable survival prognosis, comparable to that of pT4 tumors.2– 4 There is also debate on whether perirenal fat invasion is associated with prognosis in patients with RCC. Siemer et al suggested that perirenal fat invasion is not a prognostic predictor of CSS and the current pT3a classification should be revised.5 Another group noted that the pT3a stage is too variable and suggested removing perirenal fat invasion from the current TNM staging system.6 In contrast, recent studies show that tumor size is an important factor for predicting the outcome of pT3a RCC with perirenal fat invasion only, and tumor size and perirenal fat invasion should be included in T3a RCC staging.7,8 However, these groups analyzed the prognostic factor of perirenal fat invasion by tumor size only using the Kaplan-Meier method or did not analyze whether perirenal fat invasion is a significant factor in tumors 7 cm or less. Thus, we evaluated the prognostic significance of perirenal fat invasion for RCC and assessed whether tumor size has a role in stratifying prognosis in the pT3a stage of the current TNM system.
MATERIALS AND METHODS We retrospectively reviewed the medical records of 1,085 consecutive patients with RCC who underwent curative surgery between 1988 and 2006. This study was approved by the Seoul National University Hospital institutional review board. After excluding 139 patients with bilateral tumors, multifocal disease, pT3c–pT4 tumors, nodal involvement or distant metastasis 946 were included in our study. General health status was measured by ECOG PS criteria. Surgical procedures included open radical nephrectomy in 664 cases (70.1%), laparoscopic radical nephrectomy in 140 (14.8%), open partial nephrectomy in 139 (14.7%) and laparoscopic partial nephrectomy in 3 (0.4%). Pathological staging was done with the 2002 TNM classification.1 Histological subtyping was performed by the 2004 WHO classification9 and grading was done by the Fuhrman nuclear grading system.10 Perirenal fat invasion was defined as direct extension of tumor into perinephric and/or renal sinus fat. Of the patients 675 (71.6%), 134 (14.6%), 74 (7.8%) and 63 (6.6%) presented with pathological stage pT1, pT2, pT3a and pT3b, respectively. Four of 74 patients with pT3a tumors and adrenal gland invasion were excluded from study due to evidence suggesting an adverse prognosis based on pathological findings.2,3 Thus, 70 patients with perirenal fat invasion were identified and divided into 2 subgroups using 7 cm as the tumor size cutoff. The primary study end points were estimated DFS and CSS. Time to recurrence was defined as the interval from nephrectomy to the first evidence of disease recurrence. Survival data were obtained from the medical records by contacting the patient family or reviewing the death certificates. DFS and CSS were estimated by Kaplan-Meier methods. A Cox proportional hazards regression model was used to identify significant factors associated with DFS and CSS with the HR and 95% CI. All p values are 2-sided with p ⬍0.05 considered significant.
Table 1. Patient characteristics No. Pts (%) Gender: M F ECOG PS: 0 1 or Greater Presentation: Incidental Symptoms Pathological T stage: T1 T2 T3a T3b Fuhrman grade: G1 G2 G3 G4 Histological type: Clear cell Chromophobe Papillary Collecting duct Unclassified
658 (69.9) 284 (30.1) 767 (85.0) 135 (15.0) 458 (49.1) 475 (50.9) 675 (71.7) 134 (14.2) 70 (7.4) 63 (6.7) 33 (4.2) 420 (53.1) 287 (36.3) 51 (6.4) 788 (83.7) 72 (7.6) 47 (5.0) 3 (0.3) 32 (3.4)
Mean ⫾ SD age 54.5 ⫾ 11.8
RESULTS Table 1 shows characteristics in 942 patients, of whom 675 (71.7%), 134 (14.2%) and 70 (7.4%) were classified with pT1, pT2 and pT3a disease, respectively. Of patients with pT3a disease 38 (54.2%) had tumors 7 cm or less and 32 (45.8%) had tumors greater than 7 cm. There was no statistically significant difference (p ⬎0.05) between pT1 and pT3a in the 7 cm or less group based on age, gender, histology, Fuhrman grade, ECOG scale and presenting symptoms except for tumor size (3.8 vs 4.7 cm p ⫽ 0.003, table 2). There was no statistically significant difference between pT2N0M0 and pT3aN0M0 in the greater than 7 cm group based on age, gender, histology, Fuhrman grade, ECOG scale and presenting symptoms, including tumor size (table 2). Mean postoperative followup was 61.4 months (median 53, range 1 to 248). At analysis 150 patients had disease recurrence at a mean of 37.9 months (median 26, range 1 to 194), including 58 of 675 (8.6%) with pT1, 36 of 134 (26.9%) with pT2, 23 of 70 (32.9%) with pT3a and 33 of 63 (52.4%) with pT3b. At last followup 87 patients (9.2%) had died of RCC at a mean of 50.3 months after surgery (median 41.7, range 3.1 to 190.2). Figure 1 shows the Kaplan-Meier survival curve for DFS and CSS by pathological stage. Patients with pT1, pT2, pT3a and pT3b tumors had 92.4%, 74.9%, 63.3% and 49.1% 5-year DFS, and 96.6%, 85.4%, 75.9% and 69.4% 5-year CSS, respectively. The difference in the DFS among the different T stages was significant (pT1 vs pT2 p ⬍0.001 and pT3a vs pT3b
RENAL CELL CANCER AND PERIRENAL FAT INVASION
2139
Table 2. Clinical and pathological characteristics in 942 study patients
Mean ⫾ SD age No. Gender (%): M F No. ECOG PS (%): 0 1 or Greater No. Presentation (%): Incidental Symptoms No. Fuhrman grade (%): G1 G2 G3 G4 No. Histological type (%): Clear cell Chromophobe Papillary Collecting duct Unclassified
pT1
pT3a 7 cm or Less
p Value
pT2
pT3a Greater Than 7 cm
p Value
54.2 ⫾ 12.0
57.9 ⫾ 12.6
0.066 0.285
53.0 ⫾ 10.3
56.3 ⫾ 12.1
0.094 0.608
90 (67.2) 44 (32.8)
23 (71.9) 9 (28.1)
97 (72.4) 37 (27.6)
24 (75.0) 8 (25.0)
51 (38.9) 80 (61.1)
8 (25.8) 23 (74.2)
4 (3.9) 47 (46.1) 42 (41.2) 9 (8.8)
0 11 (40.7) 11 (40.7) 5 (18.5)
106 (79.1) 17 (12.7) 2 (1.5) 0 10 (7.4)
26 (81.3) 3 (9.4) 1 (3.1) 0 1 (3.1)
481 (71.3) 194 (28.7)
24 (63.2) 14 (36.8)
569 (89.2) 69 (10.8)
31 (88.6) 4 (11.4)
379 (56.3) 294 (43.7)
21 (58.3) 15 (41.7)
29 (5.0) 329 (56.7) 201 (34.7) 21 (3.6)
0 18 (64.3) 7 (25.0) 3 (10.7)
565 (83.7) 48 (7.1) 42 (6.2) 2 (0.3) 18 (2.7)
34 (89.5) 0 (4.2) 1 (2.6) 0 3 (7.9)
0.910
0.764
0.800
0.172
0.441
0.144
0.955
p ⫽ 0.029) except for pT2 and pT3a (p ⫽ 0.137). CSS significantly differed for pT1 vs pT2 (p ⬍0.001) but the difference between pT2 and pT3a (p ⫽ 0.458), and pT3a and pT3b (p ⫽ 0.066) was not statistically significant. There was no difference in DFS and CSS in patients with pT3a RCC invading renal sinus fat and those with pT3a RCC invading perinephric fat (data not shown). Of patients with pT3a those with a tumor greater than 7 cm had lower 5-year DFS (49.5% vs 77.2%, p ⫽ 0.004) and CSS (58.5% vs 95.6%, p ⫽ 0.003) than those with tumors 7 cm or less. Of patients with tumors 7 cm or less there was
0.679
no significant difference in DFS and CSS between stages pT1 and pT3a (p ⫽ 0.109 and 0.602, respectively, fig. 2). However, of patients with tumors greater than 7 cm those with pT3a had poorer DFS and CSS than those with pT2 (p ⫽ 0.004 and 0.003, respectively, fig. 2). There was no difference in DFS and CSS between patients with pT3a tumors greater than 7 cm and those with pT3b tumors. Univariate analysis showed that perirenal fat invasion was significantly associated with DFS and CSS (each p ⬍0.001, fig. 3, A). Perirenal fat invasion in tumors 7 cm or less was significantly associated with
Figure 1. Kaplan-Meier survival curves by current T stage classification. A, DFS (pT2 vs pT3a p ⫽ 0.137). B, CSS (pT2 vs pT3a p ⫽ 0.458).
2140
RENAL CELL CANCER AND PERIRENAL FAT INVASION
Figure 2. Kaplan-Meier RCC survival estimates. A, DFS (pT1 vs pT3a 7 cm or less p ⫽ 0.109 and pT2 vs pT3a greater than 7 cm p ⫽ 0.004). B, CSS (pT1 vs pT3a 7 cm or less p ⫽ 0.602 and pT2 vs pT3a greater than 7 cm p ⫽ 0.003).
DFS but not CSS (p ⫽ 0.005 and 0.161), but in tumors greater than 7 cm it was significantly associated with DFS and CSS (p ⬍0.001 and 0.002, respectively, fig. 3, B and C). In pT3a cases with perirenal fat invasion only tumor size was significantly associated with DFS and CSS (HR 1.17, 95% CI 1.08 –1.27 and HR 1.28, 95% CI 1.16 –1.42, respectively, each p ⬍0.001). Multivariate analysis of the entire cohort showed that perirenal fat invasion independently predicted DFS and CSS (HR 2.52, 95% CI 1.55–4.08, p ⬍0.001 and HR 2.38, 95% CI 1.24–4.57, p ⫽ 0.009, respectively). In tumors 7 cm or less multivariate analysis showed that age (p ⫽ 0.033 and 0.003), tumor size (p ⫽ 0.001 and 0.016), microvascular invasion (p ⫽ 0.009 and 0.002) and ECOG PS (p ⫽ 0.033 and 0.006, respectively)
independently predicted DFS and CSS. In this subgroup perirenal fat invasion did not independently predict DFS (p ⫽ 0.119) and CSS (p ⫽ 0.208). However, in tumors greater than 7 cm tumor size (p ⫽ 0.012 and 0.022) and perirenal fat invasion (p ⫽ 0.002 and 0.027, respectively) independently predicted DFS and CSS (table 3). Also, Fuhrman grade independently predicted CSS (p ⫽ 0.028, table 3).
DISCUSSION Results show that CSS in patients with stage pT3a RCC and tumors 7 cm or less was similar to that of pT1 while CSS in patients with stage pT3a RCC and tumors greater than 7 cm was equal to that of pT3b.
Figure 3. Kaplan-Meier estimated CSS. A, all tumors. B, tumors 7 cm or less. C, tumors greater than 7 cm. (⫺), negative. (⫹), positive.
RENAL CELL CANCER AND PERIRENAL FAT INVASION
2141
Table 3. Multivariate Cox regression analysis of factors influencing DFS and CSS DFS
Overall: Age* ECOG PS (1 or greater vs 0) Presenting symptoms Size* Fuhrman nuclear grade Histological type (nonconventional/conventional) Perirenal fat invasion Microvascular invasion Tumor necrosis 7 cm or Less: Age* ECOG PS (1 or greater vs 0) Presenting symptoms Size* Fuhrman nuclear grade Histological type (nonconventional/conventional) Perirenal fat invasion Microvascular invasion Tumor necrosis Greater than 7 cm: Age* ECOG PS (1 or greater vs 0) Presenting symptoms Size* Fuhrman nuclear grade Histological type (nonconventional/conventional) Perirenal fat invasion Microvascular invasion Tumor necrosis
CSS
HR (95% CI)
p Value
1.015 (0.996–1.034) 2.270 (1.191–4.329) 1.088 (0.702–1.685) 1.215 (1.149–1.284) 1.311 (0.965–1.782) 0.689 (0.375–1.265) 2.520 (1.556–4.083) 1.647 (0.995–2.726) 1.378 (0.885–2.145)
0.129 0.013 0.707 ⬍0.001 0.083 0.229 ⬍0.001 0.052 0.156
1.029 (1.002–1.056) 2.704 (1.084–6.744) 1.180 (0.636–2.189) 1.430 (1.167–1.751) 1.156 (0.743–1.799) 0.701 (0.271–1.816) 2.093 (0.827–5.300) 2.693 (1.281–5.662) 1.272 (0.663–2.440) 1.006 (0.976–1.036) 1.811 (0.699–4.688) 1.045 (0.562–1.944) 1.149 (1.031–1.281) 1.422 (0.917–2.203) 0.672 (0.300–1.508) 2.530 (1.408–4.546) 1.240 (0.656–2.343) 1.394 (0.762–2.552)
HR (95% CI) 1.048 2.482 1.354 1.238 1.487 0.807 2.389 2.105 0.674
p Value
(1.019–1.078) (1.126–5.471) (0.717–2.558) (1.148–1.335) (0.980–2.256) (0.355–1.838) (1.248–4.573) (1.061–4.176) (0.349–1.303)
0.001 0.024 0.350 ⬍0.001 0.062 0.610 0.009 0.033 0.240
0.033 0.033 0.599 0.001 0.520 0.465 0.119 0.009 0.468
1.070 (1.203–1.119) 5.453 (1.636–18.175) 1.416 (0.524–3.823) 1.497 (1.079–2.077) 1.086 (0.548–2.150) 0.848 (0.190–3.787) 2.541 (0.596–10.844) 5.074 (1.805–14.261) 0.686 (0.227–2.071)
0.003 0.006 0.493 0.016 0.814 0.829 0.208 0.002 0.503
0.701 0.221 0.890 0.012 0.116 0.335 0.002 0.508 0.281
1.037 1.267 1.412 1.174 1.862 0.934 2.363 1.115 0.827
0.079 0.711 0.463 0.022 0.028 0.894 0.027 0.803 0.637
(0.996–1.079) (0.362–4.440) (0.562–3.545) (1.023–1.347) (1.069–3.242) (0.344–2.540) (1.104–5.056) (0.475–2.615) (0.376–1.820)
* Continuous variable.
Also, perirenal fat invasion was an independent prognostic factor for DFS and CSS in patients with lesions greater than 7 cm but not in those with lesions 7 cm or less. By adding tumor size to the evaluation of prognosis in patients with RCC and perirenal fat invasion the predictive value of the next TNM staging system could be considerably improved in regard to survival. Under the current 2002 TNM classification system stages pT1 and pT2 are defined as tumor size 7 cm or less and greater than 7 cm, respectively, but confined to the kidney while pT3a is defined as extension into perirenal fat independent of tumor size as well as adrenal invasion. Based on the current system tumors 7 cm or less and tumors greater than 7 cm are classified together as stage pT3a as long as perirenal fat invasion is found on histology. There are several reports of DFS and CSS for pT3a tumors not limited by size and other pathological groups. Several studies show that the pT3a disease survival rate is similar to or better than that of pT1 or pT2.11,12 In contrast, Lam et al recently noted that pT3a disease with perirenal fat invasion only had poorer CSS than pT2 disease.7 However, we observed no significant difference in DFS and CSS
between stages pT2 (mean tumor size 9.7 cm) and pT3a (mean tumor size 7.7 cm). These survival rate discrepancies may have resulted from differences in pT3a tumor size. Tumor size is one of the most significant independent prognostic factors for DFS and CSS in patients with RCC,13–17 consistent with our multivariate analysis findings. However, the role of tumor size in patients with pT3a RCC has attracted little attention in the literature. More recently pT3a DFS and CSS have been assessed by comparing tumor size and other pathological stages. Siemer et al noted no significant differences in CSS between pT3a tumors 7 cm or less and pT1 tumors, and suggested that pT3a 7 cm or less should be included in stage T1.5 Furthermore, Yoo et al found that patients with pT3a tumors 7 cm or less had poorer DFS but similar CSS compared to patients with pT1 tumors.8 These results are in accordance with the results of our study that DFS and CSS for pT3a tumors 7 cm or less are similar to those for pT1 tumors. A recent multicenter study in 623 patients with perirenal fat invasion identified the ideal tumor size cutoff as 7 cm.7 When comparing CSS in patients with pT3a tumors 7 cm or less to CSS in those with pT2 dis-
2142
RENAL CELL CANCER AND PERIRENAL FAT INVASION
ease, the pT3a 7 cm or less group tended to have a poor CSS prognosis, although this difference was not statistically significant. A possible explanation for this discrepancy is the variable composition of Fuhrman grade for pT3a tumors 7 cm or less. Lam et al noted higher Fuhrman grade (56.1%)7 than in other reported studies (29% to 41%).12,18 For pT3a greater than 7 cm tumors Siemer et al found no significant difference in CSS between pT3a tumors greater than 7 cm and pT2 tumors, and suggested that pT2 should include all pT3a tumors greater than 7 cm.5 In contrast, Lam et al reported that CSS for T3a tumors greater than 7 cm was similar to CSS for pT3b, as in our finding that patients with pT3a greater than 7 cm had DFS and CSS similar to those in patients with pT3b. In our study considering that pT3a stage tumors 7 cm or less had a prognosis similar to that of pT1 while pT3a stage tumors greater than 7 cm had a prognosis similar to that of pT3b, perirenal fat invasion may have important prognostic value for larger tumors. There are few studies of whether perirenal fat invasion is an independent prognostic factor for DFS and CSS on multivariate analysis. Patard et al reported that perirenal fat invasion is an independent prognostic factor.19 Our results show that perirenal fat invasion was an independent predictor of DFS (HR 2.52, p ⬍0.001) and CSS (HR 2.38, p ⫽ 0.009) in the entire cohort, consistent with the current TNM classification system. On the other hand, Gofrit et al reported that the current definition of pT3a includes a highly variable group of patients with marked differences in DFS and perirenal fat invasion does not necessarily predict aggressive biological behavior.6 Yoo et al reported that perirenal fat invasion is an independent prognostic factor for DFS but not for CSS.8 Recent studies showed the significance of pT3a tumor size on multivariate analysis. Siddiqui et al
subdivided T3a tumors into 3 groups by tumor size, including T1a— 4 cm or less, T1b— 4 to 7 and T2— greater than 7.20 Regardless of tumor size perirenal fat invasion was a significant independent predictor of CSS on multivariate analysis, consistent with the current TNM staging system. However, our findings contradict findings in previous studies and show that perirenal fat invasion is not an independent prognostic factor for DFS and CSS in tumors 7 cm or less. Rather, it was an independent factor for DFS and CSS only in tumors greater than 7 cm. We previously noted that capsular invasion is a poor prognostic factor only for pT2 localized tumors.21 Since capsular invasion precedes perirenal fat invasion, our finding that perirenal fat invasion was only seen in tumors larger than 7 cm as an independent prognostic factor is in accordance with our previous report.21 Our findings support the 2002 current TNM classification but suggest that pT3a substratification by tumor size may improve the predictive value of the TNM staging system. Our study has limitations. 1) Data were collected retrospectively and reflect a single institution experience. 2) We included a relatively small number of T3a cases. Nevertheless, results show that perirenal fat invasion is a marker for aggressive RCC behavior for lesions greater than 7 cm. Multicenter studies are needed to determine the prognostic implications of perirenal fat invasion by tumor size in nephrectomy specimens. A larger sample size and longer followup are needed to confirm our findings.
CONCLUSIONS Results show that perirenal fat invasion has prognostic significance for tumors greater than 7 cm but not 7 cm or less. These findings suggest that perirenal fat invasion should be considered an important parameter in the next TNM staging system. Also, patient with pT3a tumors greater than 7 cm require more stringent followup and intensive treatment.
REFERENCES 1. Greene FL: AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag 2002.
cinoma according to the cancer-related outcome. Eur Urol 2007; 51: 722.
2. Han KR, Bui MH, Pantuck AJ et al: TNM T3a renal cell carcinoma: adrenal gland involvement is not the same as renal fat invasion. J Urol 2003; 169: 899.
5. Siemer S, Lehmann J, Loch A et al: Current TNM classification of renal cell carcinoma evaluated: revising stage T3a. J Urol 2005; 173: 33.
3. Thompson RH, Leibovich BC, Cheville JC et al: Should direct ipsilateral adrenal invasion from renal cell carcinoma be classified as pT3a? J Urol 2005; 173: 918.
6. Gofrit ON, Shapiro A, Pizov G et al: Does stage T3a renal cell carcinoma embrace a homogeneous group of patients? J Urol 2007; 177: 1682.
4. Ficarra V, Novara G, Iafrate M et al: Proposal for reclassification of the TNM staging system in patients with locally advanced (pT3– 4) renal cell car-
7. Lam JS, Klatte T, Patard JJ et al: Prognostic relevance of tumour size in T3a renal cell carcinoma: a multicentre experience. Eur Urol 2007; 52: 155.
8. Yoo C, Song C, Hong JH et al: Prognostic significance of perinephric fat infiltration and tumor size in renal cell carcinoma. J Urol 2008; 180: 486. 9. Lopez-Beltran A, Scarpelli M, Montironi R et al: 2004 WHO classification of the renal tumors of the adults. Eur Urol 2006; 49: 798. 10. Fuhrman SA, Lasky LC and Limas C: Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982; 6: 655. 11. Gilbert SM, Murphy AM, Katz AE et al: Reevaluation of TNM staging of renal cortical tumors: recurrence
RENAL CELL CANCER AND PERIRENAL FAT INVASION
and survival for T1N0M0 and T3aN0M0 tumors are equivalent. Urology 2006; 68: 287.
analysis of 643 patients using the revised 1997 TNM staging criteria. J Urol 2000; 163: 1090.
12. Murphy AM, Gilbert SM, Katz AE et al: Reevaluation of the Tumour-Node-Metastasis staging of locally advanced renal cortical tumours: absolute size (T2) is more significant than renal capsular invasion (T3a). BJU Int 2005; 95: 27.
15. Howard GE and Wood CG: Staging refinements in renal cell carcinoma. Curr Opin Urol 2006; 16: 317.
13. Frank I, Blute ML, Cheville JC et al: An outcome prediction model for patients with clear cell renal cell carcinoma treated with radical nephrectomy based on tumor stage, size, grade and necrosis: the SSIGN score. J Urol 2002; 168: 2395. 14. Tsui KH, Shvarts O, Smith RB et al: Prognostic indicators for renal cell carcinoma: a multivariate
16. Steiner T, Knels R and Schubert J: Prognostic significance of tumour size in patients after tumour nephrectomy for localised renal cell carcinoma. Eur Urol 2004; 46: 327. 17. Karakiewicz PI, Lewinshtein DJ, Chun FK et al: Tumor size improves the accuracy of TNM predictions in patients with renal cancer. Eur Urol 2006; 50: 521. 18. Roberts WW, Bhayani SB, Allaf ME et al: Pathological stage does not alter the prognosis
2143
for renal lesions determined to be stage T1 by computerized tomography. J Urol 2005; 173: 713. 19. Patard JJ, Leray E, Rodriguez A et al: Correlation between symptom graduation, tumor characteristics and survival in renal cell carcinoma. Eur Urol 2003; 44: 226. 20. Siddiqui SA, Frank I, Leibovich BC et al: Impact of tumor size on the predictive ability of the pT3a primary tumor classification for renal cell carcinoma. J Urol 2007; 177: 59. 21. Jeong IG, Jeong CW, Hong SK et al: Prognostic implication of capsular invasion without perinephric fat infiltration in localized renal cell carcinoma. Urology 2006; 67: 709.
Purpose: Controversy continues over whether perirenal fat invasion in pT3a renal cell carcinoma is a prognostic factor. We investigated the prognosis of renal cell carcinoma with perirenal fat invasion compared to the prognosis of other pathological stages by tumor size. Materials and Methods: We reviewed the medical records of 946 patients who underwent curative surgery for pT1–pT3bN0M0 renal cell carcinoma between 1988 and 2006. Patients with pT3a stage disease and perirenal fat invasion only were divided into 2 subgroups by a 7 cm tumor size cutoff. The prognostic impact of perirenal fat invasion on disease-free and cancer specific survival was investigated. Results: Patients with perirenal fat invasion and lesions greater than 7 cm had lower 5-year disease-free (49.5% vs 77.2%, p ⫽ 0.004) and cancer specific (58.5% vs 95.6%, p ⫽ 0.003) survival than those with lesions 7 cm or less. Patients with perirenal fat invasion and lesions 7 cm or less had similar 5-year disease-free and cancer specific survival compared to those with pT1 tumors (p ⫽ 0.109 and 0.602, respectively). For tumors 7 cm or less multivariate analysis showed that perirenal fat invasion was not a significant predictor of disease-free (p ⫽ 0.119) or cancer specific (p ⫽ 0.208) survival. In contrast, perirenal fat invasion was an independent prognostic factor for disease-free (p ⫽ 0.002) and cancer specific (p ⫽ 0.027) survival in patients with tumors greater than 7 cm. Conclusions: Findings suggest that the prognostic significance of perirenal fat invasion depends on primary tumor size. Perirenal fat invasion included in the pT3a stage regardless of tumor size should be reevaluated by tumor size for a more accurate patient prognosis.
Abbreviations and Acronyms CSS ⫽ cancer specific survival DFS ⫽ disease-free survival ECOG ⫽ Eastern Cooperative Oncology Group PS ⫽ performance status RCC ⫽ renal cell carcinoma Submitted for publication March 17, 2009. Study received Seoul National University Hospital institutional review board approval. * Equal study contribution. † Correspondence: Department of Urology, Seoul National University Hospital, 28, Yongondong, Chongno-gu, Seoul 110-744, Republic of Korea (telephone: 82-2-2072-2915; FAX: 82-2742-4665; e-mail: [email protected]).
Key Words: kidney; carcinoma, renal cell; prognosis; adipose tissue; neoplasm invasiveness ACCURATE cancer staging provides important prognostic information on the risk of cancer recurrence and death, and is important for physicians to make decisions on appropriate cancer treatment. The advent of modern imaging and treatment modalities to assess CSS has led to the update of the TNM classification system for RCC.
In the latest version of the American Joint Committee on Cancer TNM classification system for RCC pT3a is defined as tumors that invade adrenal gland or perirenal fat tissue but do not go beyond Gerota’s fascia.1 Recently the pT3a classification has been the subject of considerable debate. Several groups reported that pT3a tumors directly invading the ipsilateral adrenal gland
0022-5347/09/1825-2137/0 THE JOURNAL OF UROLOGY® Copyright © 2009 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 182, 2137-2143, November 2009 Printed in U.S.A. DOI:10.1016/j.juro.2009.07.065
www.jurology.com
2137
2138
RENAL CELL CANCER AND PERIRENAL FAT INVASION
have an unfavorable survival prognosis, comparable to that of pT4 tumors.2– 4 There is also debate on whether perirenal fat invasion is associated with prognosis in patients with RCC. Siemer et al suggested that perirenal fat invasion is not a prognostic predictor of CSS and the current pT3a classification should be revised.5 Another group noted that the pT3a stage is too variable and suggested removing perirenal fat invasion from the current TNM staging system.6 In contrast, recent studies show that tumor size is an important factor for predicting the outcome of pT3a RCC with perirenal fat invasion only, and tumor size and perirenal fat invasion should be included in T3a RCC staging.7,8 However, these groups analyzed the prognostic factor of perirenal fat invasion by tumor size only using the Kaplan-Meier method or did not analyze whether perirenal fat invasion is a significant factor in tumors 7 cm or less. Thus, we evaluated the prognostic significance of perirenal fat invasion for RCC and assessed whether tumor size has a role in stratifying prognosis in the pT3a stage of the current TNM system.
MATERIALS AND METHODS We retrospectively reviewed the medical records of 1,085 consecutive patients with RCC who underwent curative surgery between 1988 and 2006. This study was approved by the Seoul National University Hospital institutional review board. After excluding 139 patients with bilateral tumors, multifocal disease, pT3c–pT4 tumors, nodal involvement or distant metastasis 946 were included in our study. General health status was measured by ECOG PS criteria. Surgical procedures included open radical nephrectomy in 664 cases (70.1%), laparoscopic radical nephrectomy in 140 (14.8%), open partial nephrectomy in 139 (14.7%) and laparoscopic partial nephrectomy in 3 (0.4%). Pathological staging was done with the 2002 TNM classification.1 Histological subtyping was performed by the 2004 WHO classification9 and grading was done by the Fuhrman nuclear grading system.10 Perirenal fat invasion was defined as direct extension of tumor into perinephric and/or renal sinus fat. Of the patients 675 (71.6%), 134 (14.6%), 74 (7.8%) and 63 (6.6%) presented with pathological stage pT1, pT2, pT3a and pT3b, respectively. Four of 74 patients with pT3a tumors and adrenal gland invasion were excluded from study due to evidence suggesting an adverse prognosis based on pathological findings.2,3 Thus, 70 patients with perirenal fat invasion were identified and divided into 2 subgroups using 7 cm as the tumor size cutoff. The primary study end points were estimated DFS and CSS. Time to recurrence was defined as the interval from nephrectomy to the first evidence of disease recurrence. Survival data were obtained from the medical records by contacting the patient family or reviewing the death certificates. DFS and CSS were estimated by Kaplan-Meier methods. A Cox proportional hazards regression model was used to identify significant factors associated with DFS and CSS with the HR and 95% CI. All p values are 2-sided with p ⬍0.05 considered significant.
Table 1. Patient characteristics No. Pts (%) Gender: M F ECOG PS: 0 1 or Greater Presentation: Incidental Symptoms Pathological T stage: T1 T2 T3a T3b Fuhrman grade: G1 G2 G3 G4 Histological type: Clear cell Chromophobe Papillary Collecting duct Unclassified
658 (69.9) 284 (30.1) 767 (85.0) 135 (15.0) 458 (49.1) 475 (50.9) 675 (71.7) 134 (14.2) 70 (7.4) 63 (6.7) 33 (4.2) 420 (53.1) 287 (36.3) 51 (6.4) 788 (83.7) 72 (7.6) 47 (5.0) 3 (0.3) 32 (3.4)
Mean ⫾ SD age 54.5 ⫾ 11.8
RESULTS Table 1 shows characteristics in 942 patients, of whom 675 (71.7%), 134 (14.2%) and 70 (7.4%) were classified with pT1, pT2 and pT3a disease, respectively. Of patients with pT3a disease 38 (54.2%) had tumors 7 cm or less and 32 (45.8%) had tumors greater than 7 cm. There was no statistically significant difference (p ⬎0.05) between pT1 and pT3a in the 7 cm or less group based on age, gender, histology, Fuhrman grade, ECOG scale and presenting symptoms except for tumor size (3.8 vs 4.7 cm p ⫽ 0.003, table 2). There was no statistically significant difference between pT2N0M0 and pT3aN0M0 in the greater than 7 cm group based on age, gender, histology, Fuhrman grade, ECOG scale and presenting symptoms, including tumor size (table 2). Mean postoperative followup was 61.4 months (median 53, range 1 to 248). At analysis 150 patients had disease recurrence at a mean of 37.9 months (median 26, range 1 to 194), including 58 of 675 (8.6%) with pT1, 36 of 134 (26.9%) with pT2, 23 of 70 (32.9%) with pT3a and 33 of 63 (52.4%) with pT3b. At last followup 87 patients (9.2%) had died of RCC at a mean of 50.3 months after surgery (median 41.7, range 3.1 to 190.2). Figure 1 shows the Kaplan-Meier survival curve for DFS and CSS by pathological stage. Patients with pT1, pT2, pT3a and pT3b tumors had 92.4%, 74.9%, 63.3% and 49.1% 5-year DFS, and 96.6%, 85.4%, 75.9% and 69.4% 5-year CSS, respectively. The difference in the DFS among the different T stages was significant (pT1 vs pT2 p ⬍0.001 and pT3a vs pT3b
RENAL CELL CANCER AND PERIRENAL FAT INVASION
2139
Table 2. Clinical and pathological characteristics in 942 study patients
Mean ⫾ SD age No. Gender (%): M F No. ECOG PS (%): 0 1 or Greater No. Presentation (%): Incidental Symptoms No. Fuhrman grade (%): G1 G2 G3 G4 No. Histological type (%): Clear cell Chromophobe Papillary Collecting duct Unclassified
pT1
pT3a 7 cm or Less
p Value
pT2
pT3a Greater Than 7 cm
p Value
54.2 ⫾ 12.0
57.9 ⫾ 12.6
0.066 0.285
53.0 ⫾ 10.3
56.3 ⫾ 12.1
0.094 0.608
90 (67.2) 44 (32.8)
23 (71.9) 9 (28.1)
97 (72.4) 37 (27.6)
24 (75.0) 8 (25.0)
51 (38.9) 80 (61.1)
8 (25.8) 23 (74.2)
4 (3.9) 47 (46.1) 42 (41.2) 9 (8.8)
0 11 (40.7) 11 (40.7) 5 (18.5)
106 (79.1) 17 (12.7) 2 (1.5) 0 10 (7.4)
26 (81.3) 3 (9.4) 1 (3.1) 0 1 (3.1)
481 (71.3) 194 (28.7)
24 (63.2) 14 (36.8)
569 (89.2) 69 (10.8)
31 (88.6) 4 (11.4)
379 (56.3) 294 (43.7)
21 (58.3) 15 (41.7)
29 (5.0) 329 (56.7) 201 (34.7) 21 (3.6)
0 18 (64.3) 7 (25.0) 3 (10.7)
565 (83.7) 48 (7.1) 42 (6.2) 2 (0.3) 18 (2.7)
34 (89.5) 0 (4.2) 1 (2.6) 0 3 (7.9)
0.910
0.764
0.800
0.172
0.441
0.144
0.955
p ⫽ 0.029) except for pT2 and pT3a (p ⫽ 0.137). CSS significantly differed for pT1 vs pT2 (p ⬍0.001) but the difference between pT2 and pT3a (p ⫽ 0.458), and pT3a and pT3b (p ⫽ 0.066) was not statistically significant. There was no difference in DFS and CSS in patients with pT3a RCC invading renal sinus fat and those with pT3a RCC invading perinephric fat (data not shown). Of patients with pT3a those with a tumor greater than 7 cm had lower 5-year DFS (49.5% vs 77.2%, p ⫽ 0.004) and CSS (58.5% vs 95.6%, p ⫽ 0.003) than those with tumors 7 cm or less. Of patients with tumors 7 cm or less there was
0.679
no significant difference in DFS and CSS between stages pT1 and pT3a (p ⫽ 0.109 and 0.602, respectively, fig. 2). However, of patients with tumors greater than 7 cm those with pT3a had poorer DFS and CSS than those with pT2 (p ⫽ 0.004 and 0.003, respectively, fig. 2). There was no difference in DFS and CSS between patients with pT3a tumors greater than 7 cm and those with pT3b tumors. Univariate analysis showed that perirenal fat invasion was significantly associated with DFS and CSS (each p ⬍0.001, fig. 3, A). Perirenal fat invasion in tumors 7 cm or less was significantly associated with
Figure 1. Kaplan-Meier survival curves by current T stage classification. A, DFS (pT2 vs pT3a p ⫽ 0.137). B, CSS (pT2 vs pT3a p ⫽ 0.458).
2140
RENAL CELL CANCER AND PERIRENAL FAT INVASION
Figure 2. Kaplan-Meier RCC survival estimates. A, DFS (pT1 vs pT3a 7 cm or less p ⫽ 0.109 and pT2 vs pT3a greater than 7 cm p ⫽ 0.004). B, CSS (pT1 vs pT3a 7 cm or less p ⫽ 0.602 and pT2 vs pT3a greater than 7 cm p ⫽ 0.003).
DFS but not CSS (p ⫽ 0.005 and 0.161), but in tumors greater than 7 cm it was significantly associated with DFS and CSS (p ⬍0.001 and 0.002, respectively, fig. 3, B and C). In pT3a cases with perirenal fat invasion only tumor size was significantly associated with DFS and CSS (HR 1.17, 95% CI 1.08 –1.27 and HR 1.28, 95% CI 1.16 –1.42, respectively, each p ⬍0.001). Multivariate analysis of the entire cohort showed that perirenal fat invasion independently predicted DFS and CSS (HR 2.52, 95% CI 1.55–4.08, p ⬍0.001 and HR 2.38, 95% CI 1.24–4.57, p ⫽ 0.009, respectively). In tumors 7 cm or less multivariate analysis showed that age (p ⫽ 0.033 and 0.003), tumor size (p ⫽ 0.001 and 0.016), microvascular invasion (p ⫽ 0.009 and 0.002) and ECOG PS (p ⫽ 0.033 and 0.006, respectively)
independently predicted DFS and CSS. In this subgroup perirenal fat invasion did not independently predict DFS (p ⫽ 0.119) and CSS (p ⫽ 0.208). However, in tumors greater than 7 cm tumor size (p ⫽ 0.012 and 0.022) and perirenal fat invasion (p ⫽ 0.002 and 0.027, respectively) independently predicted DFS and CSS (table 3). Also, Fuhrman grade independently predicted CSS (p ⫽ 0.028, table 3).
DISCUSSION Results show that CSS in patients with stage pT3a RCC and tumors 7 cm or less was similar to that of pT1 while CSS in patients with stage pT3a RCC and tumors greater than 7 cm was equal to that of pT3b.
Figure 3. Kaplan-Meier estimated CSS. A, all tumors. B, tumors 7 cm or less. C, tumors greater than 7 cm. (⫺), negative. (⫹), positive.
RENAL CELL CANCER AND PERIRENAL FAT INVASION
2141
Table 3. Multivariate Cox regression analysis of factors influencing DFS and CSS DFS
Overall: Age* ECOG PS (1 or greater vs 0) Presenting symptoms Size* Fuhrman nuclear grade Histological type (nonconventional/conventional) Perirenal fat invasion Microvascular invasion Tumor necrosis 7 cm or Less: Age* ECOG PS (1 or greater vs 0) Presenting symptoms Size* Fuhrman nuclear grade Histological type (nonconventional/conventional) Perirenal fat invasion Microvascular invasion Tumor necrosis Greater than 7 cm: Age* ECOG PS (1 or greater vs 0) Presenting symptoms Size* Fuhrman nuclear grade Histological type (nonconventional/conventional) Perirenal fat invasion Microvascular invasion Tumor necrosis
CSS
HR (95% CI)
p Value
1.015 (0.996–1.034) 2.270 (1.191–4.329) 1.088 (0.702–1.685) 1.215 (1.149–1.284) 1.311 (0.965–1.782) 0.689 (0.375–1.265) 2.520 (1.556–4.083) 1.647 (0.995–2.726) 1.378 (0.885–2.145)
0.129 0.013 0.707 ⬍0.001 0.083 0.229 ⬍0.001 0.052 0.156
1.029 (1.002–1.056) 2.704 (1.084–6.744) 1.180 (0.636–2.189) 1.430 (1.167–1.751) 1.156 (0.743–1.799) 0.701 (0.271–1.816) 2.093 (0.827–5.300) 2.693 (1.281–5.662) 1.272 (0.663–2.440) 1.006 (0.976–1.036) 1.811 (0.699–4.688) 1.045 (0.562–1.944) 1.149 (1.031–1.281) 1.422 (0.917–2.203) 0.672 (0.300–1.508) 2.530 (1.408–4.546) 1.240 (0.656–2.343) 1.394 (0.762–2.552)
HR (95% CI) 1.048 2.482 1.354 1.238 1.487 0.807 2.389 2.105 0.674
p Value
(1.019–1.078) (1.126–5.471) (0.717–2.558) (1.148–1.335) (0.980–2.256) (0.355–1.838) (1.248–4.573) (1.061–4.176) (0.349–1.303)
0.001 0.024 0.350 ⬍0.001 0.062 0.610 0.009 0.033 0.240
0.033 0.033 0.599 0.001 0.520 0.465 0.119 0.009 0.468
1.070 (1.203–1.119) 5.453 (1.636–18.175) 1.416 (0.524–3.823) 1.497 (1.079–2.077) 1.086 (0.548–2.150) 0.848 (0.190–3.787) 2.541 (0.596–10.844) 5.074 (1.805–14.261) 0.686 (0.227–2.071)
0.003 0.006 0.493 0.016 0.814 0.829 0.208 0.002 0.503
0.701 0.221 0.890 0.012 0.116 0.335 0.002 0.508 0.281
1.037 1.267 1.412 1.174 1.862 0.934 2.363 1.115 0.827
0.079 0.711 0.463 0.022 0.028 0.894 0.027 0.803 0.637
(0.996–1.079) (0.362–4.440) (0.562–3.545) (1.023–1.347) (1.069–3.242) (0.344–2.540) (1.104–5.056) (0.475–2.615) (0.376–1.820)
* Continuous variable.
Also, perirenal fat invasion was an independent prognostic factor for DFS and CSS in patients with lesions greater than 7 cm but not in those with lesions 7 cm or less. By adding tumor size to the evaluation of prognosis in patients with RCC and perirenal fat invasion the predictive value of the next TNM staging system could be considerably improved in regard to survival. Under the current 2002 TNM classification system stages pT1 and pT2 are defined as tumor size 7 cm or less and greater than 7 cm, respectively, but confined to the kidney while pT3a is defined as extension into perirenal fat independent of tumor size as well as adrenal invasion. Based on the current system tumors 7 cm or less and tumors greater than 7 cm are classified together as stage pT3a as long as perirenal fat invasion is found on histology. There are several reports of DFS and CSS for pT3a tumors not limited by size and other pathological groups. Several studies show that the pT3a disease survival rate is similar to or better than that of pT1 or pT2.11,12 In contrast, Lam et al recently noted that pT3a disease with perirenal fat invasion only had poorer CSS than pT2 disease.7 However, we observed no significant difference in DFS and CSS
between stages pT2 (mean tumor size 9.7 cm) and pT3a (mean tumor size 7.7 cm). These survival rate discrepancies may have resulted from differences in pT3a tumor size. Tumor size is one of the most significant independent prognostic factors for DFS and CSS in patients with RCC,13–17 consistent with our multivariate analysis findings. However, the role of tumor size in patients with pT3a RCC has attracted little attention in the literature. More recently pT3a DFS and CSS have been assessed by comparing tumor size and other pathological stages. Siemer et al noted no significant differences in CSS between pT3a tumors 7 cm or less and pT1 tumors, and suggested that pT3a 7 cm or less should be included in stage T1.5 Furthermore, Yoo et al found that patients with pT3a tumors 7 cm or less had poorer DFS but similar CSS compared to patients with pT1 tumors.8 These results are in accordance with the results of our study that DFS and CSS for pT3a tumors 7 cm or less are similar to those for pT1 tumors. A recent multicenter study in 623 patients with perirenal fat invasion identified the ideal tumor size cutoff as 7 cm.7 When comparing CSS in patients with pT3a tumors 7 cm or less to CSS in those with pT2 dis-
2142
RENAL CELL CANCER AND PERIRENAL FAT INVASION
ease, the pT3a 7 cm or less group tended to have a poor CSS prognosis, although this difference was not statistically significant. A possible explanation for this discrepancy is the variable composition of Fuhrman grade for pT3a tumors 7 cm or less. Lam et al noted higher Fuhrman grade (56.1%)7 than in other reported studies (29% to 41%).12,18 For pT3a greater than 7 cm tumors Siemer et al found no significant difference in CSS between pT3a tumors greater than 7 cm and pT2 tumors, and suggested that pT2 should include all pT3a tumors greater than 7 cm.5 In contrast, Lam et al reported that CSS for T3a tumors greater than 7 cm was similar to CSS for pT3b, as in our finding that patients with pT3a greater than 7 cm had DFS and CSS similar to those in patients with pT3b. In our study considering that pT3a stage tumors 7 cm or less had a prognosis similar to that of pT1 while pT3a stage tumors greater than 7 cm had a prognosis similar to that of pT3b, perirenal fat invasion may have important prognostic value for larger tumors. There are few studies of whether perirenal fat invasion is an independent prognostic factor for DFS and CSS on multivariate analysis. Patard et al reported that perirenal fat invasion is an independent prognostic factor.19 Our results show that perirenal fat invasion was an independent predictor of DFS (HR 2.52, p ⬍0.001) and CSS (HR 2.38, p ⫽ 0.009) in the entire cohort, consistent with the current TNM classification system. On the other hand, Gofrit et al reported that the current definition of pT3a includes a highly variable group of patients with marked differences in DFS and perirenal fat invasion does not necessarily predict aggressive biological behavior.6 Yoo et al reported that perirenal fat invasion is an independent prognostic factor for DFS but not for CSS.8 Recent studies showed the significance of pT3a tumor size on multivariate analysis. Siddiqui et al
subdivided T3a tumors into 3 groups by tumor size, including T1a— 4 cm or less, T1b— 4 to 7 and T2— greater than 7.20 Regardless of tumor size perirenal fat invasion was a significant independent predictor of CSS on multivariate analysis, consistent with the current TNM staging system. However, our findings contradict findings in previous studies and show that perirenal fat invasion is not an independent prognostic factor for DFS and CSS in tumors 7 cm or less. Rather, it was an independent factor for DFS and CSS only in tumors greater than 7 cm. We previously noted that capsular invasion is a poor prognostic factor only for pT2 localized tumors.21 Since capsular invasion precedes perirenal fat invasion, our finding that perirenal fat invasion was only seen in tumors larger than 7 cm as an independent prognostic factor is in accordance with our previous report.21 Our findings support the 2002 current TNM classification but suggest that pT3a substratification by tumor size may improve the predictive value of the TNM staging system. Our study has limitations. 1) Data were collected retrospectively and reflect a single institution experience. 2) We included a relatively small number of T3a cases. Nevertheless, results show that perirenal fat invasion is a marker for aggressive RCC behavior for lesions greater than 7 cm. Multicenter studies are needed to determine the prognostic implications of perirenal fat invasion by tumor size in nephrectomy specimens. A larger sample size and longer followup are needed to confirm our findings.
CONCLUSIONS Results show that perirenal fat invasion has prognostic significance for tumors greater than 7 cm but not 7 cm or less. These findings suggest that perirenal fat invasion should be considered an important parameter in the next TNM staging system. Also, patient with pT3a tumors greater than 7 cm require more stringent followup and intensive treatment.
REFERENCES 1. Greene FL: AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag 2002.
cinoma according to the cancer-related outcome. Eur Urol 2007; 51: 722.
2. Han KR, Bui MH, Pantuck AJ et al: TNM T3a renal cell carcinoma: adrenal gland involvement is not the same as renal fat invasion. J Urol 2003; 169: 899.
5. Siemer S, Lehmann J, Loch A et al: Current TNM classification of renal cell carcinoma evaluated: revising stage T3a. J Urol 2005; 173: 33.
3. Thompson RH, Leibovich BC, Cheville JC et al: Should direct ipsilateral adrenal invasion from renal cell carcinoma be classified as pT3a? J Urol 2005; 173: 918.
6. Gofrit ON, Shapiro A, Pizov G et al: Does stage T3a renal cell carcinoma embrace a homogeneous group of patients? J Urol 2007; 177: 1682.
4. Ficarra V, Novara G, Iafrate M et al: Proposal for reclassification of the TNM staging system in patients with locally advanced (pT3– 4) renal cell car-
7. Lam JS, Klatte T, Patard JJ et al: Prognostic relevance of tumour size in T3a renal cell carcinoma: a multicentre experience. Eur Urol 2007; 52: 155.
8. Yoo C, Song C, Hong JH et al: Prognostic significance of perinephric fat infiltration and tumor size in renal cell carcinoma. J Urol 2008; 180: 486. 9. Lopez-Beltran A, Scarpelli M, Montironi R et al: 2004 WHO classification of the renal tumors of the adults. Eur Urol 2006; 49: 798. 10. Fuhrman SA, Lasky LC and Limas C: Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982; 6: 655. 11. Gilbert SM, Murphy AM, Katz AE et al: Reevaluation of TNM staging of renal cortical tumors: recurrence
RENAL CELL CANCER AND PERIRENAL FAT INVASION
and survival for T1N0M0 and T3aN0M0 tumors are equivalent. Urology 2006; 68: 287.
analysis of 643 patients using the revised 1997 TNM staging criteria. J Urol 2000; 163: 1090.
12. Murphy AM, Gilbert SM, Katz AE et al: Reevaluation of the Tumour-Node-Metastasis staging of locally advanced renal cortical tumours: absolute size (T2) is more significant than renal capsular invasion (T3a). BJU Int 2005; 95: 27.
15. Howard GE and Wood CG: Staging refinements in renal cell carcinoma. Curr Opin Urol 2006; 16: 317.
13. Frank I, Blute ML, Cheville JC et al: An outcome prediction model for patients with clear cell renal cell carcinoma treated with radical nephrectomy based on tumor stage, size, grade and necrosis: the SSIGN score. J Urol 2002; 168: 2395. 14. Tsui KH, Shvarts O, Smith RB et al: Prognostic indicators for renal cell carcinoma: a multivariate
16. Steiner T, Knels R and Schubert J: Prognostic significance of tumour size in patients after tumour nephrectomy for localised renal cell carcinoma. Eur Urol 2004; 46: 327. 17. Karakiewicz PI, Lewinshtein DJ, Chun FK et al: Tumor size improves the accuracy of TNM predictions in patients with renal cancer. Eur Urol 2006; 50: 521. 18. Roberts WW, Bhayani SB, Allaf ME et al: Pathological stage does not alter the prognosis
2143
for renal lesions determined to be stage T1 by computerized tomography. J Urol 2005; 173: 713. 19. Patard JJ, Leray E, Rodriguez A et al: Correlation between symptom graduation, tumor characteristics and survival in renal cell carcinoma. Eur Urol 2003; 44: 226. 20. Siddiqui SA, Frank I, Leibovich BC et al: Impact of tumor size on the predictive ability of the pT3a primary tumor classification for renal cell carcinoma. J Urol 2007; 177: 59. 21. Jeong IG, Jeong CW, Hong SK et al: Prognostic implication of capsular invasion without perinephric fat infiltration in localized renal cell carcinoma. Urology 2006; 67: 709.