|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 12 (April 20), 2005: pp. 2763-2771 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.07.055 Prognostic Value of Histologic Subtypes in Renal Cell Carcinoma: A Multicenter ExperienceFrom the Department of Urology; Department of Public Health; Department of Pathology, Rennes University Hospital, Rennes; Department of Urology, Centre Hospitalier Universitaire (CHU) Henri Mondor, Creteil; Department of Urology, North Hospital, CHU of Saint-Etienne, France; University of California, Los Angeles (UCLA) Department of Urology; Department of Biostatistics, UCLA School of Medicine, Los Angeles, CA; Department of Urology, Medical School of University Federico II, Naples; Department of Urology, University of Verona, Verona, Italy; University Medical Center Nijmegen, Nijmegen, the Netherlands Address reprint requests to Allan J. Pantuck, MD, UCLA Department of Urology, CHS 66-118, 10833 LeConte Avenue, Los Angeles, CA 90095-1738; e-mail: APantuck{at}mednet.ucla.edu
PURPOSE: To analyze to what extent histologic subtype is of prognostic importance in renal cell carcinoma based on a large, international, multicenter experience. PATIENTS AND METHODS: Four thousand sixty-three patients from eight international centers were included in this retrospective study. Histologic subtype (1997 International Union Against Cancer [UICC] criteria of tumor response), age, sex, TNM stage, Fuhrman grade, tumor size, Eastern Cooperative Oncology Goup performance status (ECOG PS), and overall survival were determined in all cases. The prognostic values of clear cell, papillary, and chromophobe histologic features were assessed by uni- and multivariate analysis using the Kaplan-Meier method and Cox model, respectively. RESULTS: Clear cell, papillary, and chromophobe carcinomas accounted for 3,564 (87.7%), 396 (9.7%) and 103 (2.5%) cases, respectively. In univariate analysis, a trend toward a better survival was observed when clear cell, papillary, and chromophobe histologies were considered prognostic categories (log-rank P = .0007). However, in multivariate analysis, TNM stage, Fuhrman grade and ECOG PS, but not histology, were retained as independent prognostic variables (P < .001). CONCLUSION: The stratification in three main renal cell carcinoma histologic subtypes as defined by the 1997 UICC–American Joint Committee on Cancer consensus should not be considered a major prognostic variable comparable to TNM stage, Fuhrman grade and ECOG PS.
Renal cell carcinoma (RCC) accounts for 3% of cancers in adults.1 Although TNM stage, Fuhrman grade, and Eastern Cooperative Oncology Group performance status (ECOG PS) are the most widely recognized prognostic factors in RCC, research continues to determine strong and easily available prognostic parameters that could help to classify patients in groups with different risks of death as a result of renal cancer. Currently, RCC histologic subtypes are classified according to the International Union Against Cancer (UICC) and the American Joint Committee on Cancer (AJCC) recommendations.2 This classification is based on the Heidelberg classification system, 3 which categorizes RCCs as clear cell, papillary, chromophobe, collecting duct, and unclassified RCC subtypes. Recent studies4,5 have suggested that stratification by histologic subtype could lend prognostic value. However, there is a predominance of clear cell RCCs in most series, and conclusions about a better survival in papillary and chromophobe tumors are questionable due to the small numbers of tumors in the two former groups. Series with small number of papillary and chromophobe tumors are limited in their ability to provide information regarding the potential independent prognostic value of histologic subtype compared with stage and grade. Recently, a large single-center series including 2,385 patients6 confirmed the prognostic value of a histology-based classification even when adjusted for stage and grade. Our objective was to analyze to what extent these previous results were verified when histologic subtype was stratified by tumor stage, grade and ECOG PS in a large multicenter study including 4,063 patients from eight international centers.
Patients The study included patients from eight international academic centers: Rennes University Hospital, Rennes; North Hospital, Centre Hospitalier Universitaire (CHU), Saint Etienne; Créteil, France; Medical School of University Frederico II, Naples; University of Verona, Verona, Italy; University of California, Los Angeles, Los Angeles, CA; The University of Texas M.D. Anderson Cancer Center, Houston, TX; and University Medical Center, Nijmegen, the Netherlands. Patient records were extracted from each institutional database. Data collected from each patient included TNM stage, Fuhrman grade, ECOG PS, survival, and histologic subtype. Other variables analyzed included age, sex, and tumor size. All data were labeled with their respective institution and pooled. Patients included in the study were staged preoperatively with computed tomography (CT) of the abdomen and pelvis, chest CT or chest x-ray, serum electrolytes, and liver function tests. Postoperative follow-up was according to protocols established at each institution. Tumor stage was determined according to the 1997 UICC TNM classification of renal tumors.7 Tumor histology was classified according to the Heidelberg classification,3 and tumors were graded according to the Fuhrman grading scheme8 by the pathologists at each of the eight institutions. ECOG PS was assigned for each patient according to the original criteria set forth by Oken et al.9 Of the 4,204 patient records resulting from the fusion of the eight databases, 16 (0.4%), 31 (0.7%) and 94 (2.2%) histologies were labeled as "collecting duct carcinomas," "unclassified tumors," and "missing values," respectively. Because our objective was to correlate histologic groups to stage, grade, ECOG PS, and survival, we decided to focus on the three main histologic subtypes and to delete other records (141 exclusions). Patients were diagnosed between 1984 and 2001.
Histologic Features In conventional clear cell RCCs, the neoplastic cells were characterized by a clear cytoplasm surrounded by a distinct cell membrane, but eosinophilic cells could be predominant. Various architectures were described. Usually tumors had an alveolar or acinar pattern. There was a network of delicate interconnecting capillaries or sinusoidal stroma supported by a network of thin reticulin fibers. Papillary RCC consisted predominantly of papillary or tubulopapillary architecture. The neoplastic cells were cuboidal or columnar with eosinophilic granular cytoplasm, but foci of clear cells could be observed. Usually fine fibrovascular stalks were present. The stroma contained plasma cells, lymphocytes, neutrophils, and macrophages with focal or diffuse accumulation of foamy macrophages in the stromal spaces. Psammoma bodies were identifiable. Chromophobe RCCs usually had a solid growth pattern, but could show some foci of tubular, tubuloalveolar, and cribriform areas. Neoplastic cells showed an abundant and pale or eosinophilic cytoplasm. A narrow zone of cytoplasmic condensation was found frequently along the cell periphery. The nuclei were central, and some were binucleated. The cytoplasm was slighty positive for Periodic Acid Schiff and strongly positive for Hale's acid iron colloid stain. In all of these tumors, areas of necrosis, hemorrhage, and fibrosis could be identified.
Statistical Analysis
Patient and tumor characteristics are given in Table 1. Three thousand five hundred sixty-four tumors (87.7%), 396 (9.7%) and 103 (2.6%) were classified as conventional clear cell, papillary, and chromophobe carcinoma, respectively. Among centers, clear cell features were observed in 635 (93.4%), 198 (96.1%), 263 (86.5%), 454 (86.1%), 583 (86.4%), 580 (79.6%), 596 (91.7%), and 255 (87.6%) cases, respectively. Papillary tumors were diagnosed in 31 (4.6%), three (1.5%), 24 (7.9%), 57 (10.8%), 75 (11.1%), 124 (17%), 50 (7.7%), and 32 (11%) cases, respectively. Chromophobe carcinoma accounted for 14 (2.1%), five (2.4%), 17 (5.6%), 17 (3.2%), 17 (2.5%), 25 (3.4%), four (0.6%) and four (1.4%) cases, respectively.
Tumors were classified as localized and distant disease in 3,056 (75.2%) and 1,007 (24.8%) cases, respectively. The median survival time was 19 months for 1,398 patients who died during follow-up. The median follow-up for 2,665 surviving patients was 43 months. An association was found among histologic subtype, tumor size, sex, TNM stage and grade (Table 2). Clear cell, papillary, and chromophobe carcinomas were classified as stage I to II in 50.3%, 62.9%, and 74.8% of the cases, respectively (P < .001). The association between stage and histologic subtype was particularly evident when considering distant metastasis that occurred in 765 (21.5%), 59 (14.9%), and 3 (2.9%) patients with clear cell, papillary and chromophobe tumors, respectively (P < .001). Similarly, Fuhrman grades 1 and 2 were found in 55.8%, 66.9%, and 68% of the three histologic subtypes, respectively (P < .001). Mean tumor size tended to be lower in papillary tumors compared to clear cell and chromophobe tumors. (6.30 ± 3.6 cm v 7.1 ± 3.8 and 6.8 ± 4.2 cm; P < .001). A male predominance was observed mainly in papillary (75%) and in clear cell tumors (66.5%), whereas it was less pronounced in chromobophe tumors (55.3%). Conversely, no association was found between histologic subtype and ECOG PS.
Five-year survival rates for localized clear cell, papillary, and chromophobe carcinoma were 73.2%, 79.4%, and 87.9% respectively. Among the 3,056 patients with localized disease, patients with chromophobe carcinoma (n = 97; 3.2%) were found to have a better outcome compared with patients with papillary or clear cell tumors (log-rank P = .03; Fig 1). Only six patients with chromophobe carcinomas had distant disease (three N1to N2, three M1), and they were not considered for survival analysis in metastatic patients. One thousand one patients with nonlocalized clear cell (n = 919; 91.8%) or papillary carcinomas (n = 82; 8.2%) were available for survival analysis. No survival difference was observed between these two groups (log-rank P = .6; Fig 2). Five-year survival rates were 10.5% and 10.3% for the two groups, respectively.
When histologic subtypes were adjusted to TNM stage, no significant survival difference remained among the three histologic groups (log-rank P = .81 and .16 for stages I to II and III to IV, respectively; Fig 3). Similarly, when stratifying histology according to tumor grade the three histologic groups had the same outcome for grade 1 and 2 tumors (log-rank P = .3; Fig 4), whereas chromophobe carcinoma had a better outcome in grade 3 and 4 tumors (log-rank P = .02; Fig 4). When stratifying histology by ECOG PS, a trend toward a better outcome was found for chromophobe carcinoma compared to papillary and clear cell histologies in both PS 0 and PS  1 groups (log-rank P = .02 and .06, respectively; Fig 5). Finally when adjusting histologic subtype to University of California, Los Angeles, Integrated Staging System (UISS) level,10 which is a prognostic system that simultaneously combines TNM stage, Fuhrman grade, and ECOG PS, no significant survival difference was found among the three main histologic types (Fig 6).
Finally we considered histologic subtype as a single categoric variable for univariate and multivariate analysis. Overall, a trend toward a better outcome was observed when histologies were identified as clear cell, papillary, and chromophobe, respectively. Median survival times were 119 months, 153 months and not reached for the three categories, respectively (log-rank P = .0007; Table 3). In both overall and localized disease, only TNM stage, Fuhrman grade, and ECOG PS remained significant for predicting survival in multivariate analysis (Table 4).
TNM stage, nuclear grade, and ECOG PS are the three most recognized prognostic factors in RCC. Currently, efforts are being made to identify new prognostic variables that could help stratify patients into risk groups in addition to the usual prognostic parameters. In this large multicenter study including 4,063 patients from eight international centers, we tried to assess whether histologic subtype could be retained as a powerful independent prognostic parameter as well as TNM stage, Fuhrman grade, or ECOG PS in both localized and metastatic RCC tumors. From our study, it is clear that although histologic subtype is correlated to TNM stage and grade and is significant as a single variable, the histologic subtype was removed from the multivariate model by TNM stage, tumor grade, and ECOG PS. Papillary and clear cell tumors were not found to have a significantly different outcome in distant disease. When adjusted to TNM stage and grade, chromophobe carcinoma was found to have a better outcome when compared with other subtypes only in the case of high-grade tumors. Since the first report11 suggesting a favorable outcome for patients with papillary tumors compared with clear cell histology, significant advances have been made in morphologic and genetic characterization of papillary tumors, which are now recognized as a well-defined histologic entity.12 The frequency found in our study (9.7%) is consistent with what is generally reported (10% to 15%).13 We also confirmed the tendency for papillary tumors to present at an earlier stage and grade when compared with clear cell tumors.5,12,13 Similar to Cheville et al,6 we found a close correlation between papillary and clear cell tumors with TNM stage, Fuhrman grade, and tumor size. We did not perform a specific search for prognostic factors in papillary tumors; however, most studies agree on the prognostic value of TNM stage, nuclear grade, and sarcomatoid component but not tumor necrosis4,6,12 in this subset of patients. Finally, our reported 79.4% overall 5-year survival rate is consistent with the 82% to 90% survival rate that is generally reported.12 However, some authors have reported less favorable outcomes in papillary tumors.14 Recently, two distinct papillary histologic profiles have been described (type 1 and 2) which possess distinct clinical behaviors. Mejean et al15 reported 5-year survival rates of 95% and 66% in type 1 and 2 papillary tumors respectively. Chromophobe RCC, first described by Thoenes et al,16 also has distinguishable clinicopathologic and molecular biologic features.17 Our reported chromophobe carcinoma frequency (2.5%) is slightly less than previously published reports in surgical series (4% to 10%).4-6,13,18 However we found an 88% 5-year survival rate, which is consistent with the excellent outcome that is generally reported in most series with 5-year survival probability ranging from 80% to 100%.5,6,13 Five previous reports including a total of 4,621 patients have focused on the prognostic value of the histologic subtype classification derived from the 1997 UICC/AJCC consensus. Ljungberg et al14 first reported significantly different outcomes among histologic subtypes for 186 patients including 145 (78%) with conventional clear cell, 25 (13%) with papillary, and 12 (7%) with chromophobe carcinoma. The 5-year cancer-specific survival rates for patients with conventional, papillary, and chromophobe RCC were 43%, 61%, and 91%, respectively. Moch et al4 reported a better outcome for chromophobe carcinoma compared to clear cell tumors (P = .05) in 588 patients including 487 (83%) with clear cell, 64 (11%) with papillary, and 31 (5%) with chromophobe carcinoma, but showed no difference between papillary and chromophobe tumors, as well as no statistical significant stratification among the three groups (P = .1). Amin et al5 reported 76%, 86% and 100% 5-year cancer-specific survival rates, respectively, in 377 patients including 255 (63%) with clear cell, 75 (18.5%) with papillary, and 24 (5.9%) with chromophobe tumors. Although significant in univariate analysis, the histologic subtype was not retained in multivariate analysis. Independent variables that were significantly associated with survival in this study were TNM stage, nuclear grade, and tumor necrosis. Cheville et al6 found 5-year survival rates of 68.9%, 87.4%, and 86.7%, respectively, in 2,385 patients including 1,985 (83.2%) with clear cell, 270 (11.3%) with papillary, and 102 (4.3%) with chromophobe carcinoma. Patients with clear cell tumors had a significantly poorer outcome compared with patients with tumors of the two other types (P < .001) even when adjusted to TNM stage and grade. However, no survival difference was observed between papillary and chromophobe tumors (P = .9). Beck et al13 found 5-year survival rates of 73.3%, 81.7%, and 80.1%, respectively, in a 1,057-patient series with localized disease including 784 (75%) with clear cell, 157 (15%) with papillary, and 106 (10%) with chromophobe tumors. Patients with clear cell tumors had a significantly worse survival compared with patients with papillary (P = .015) and chromophobe carcinoma (P = .013). However, patients with papillary and chromophobe carcinoma did not have a significantly different outcome (P = .6). In summary, only two previous studies have conducted a multivariate analysis to determine whether histologic subtype could be retained as an independent prognostic variable. Amin et al failed to demonstrate such an independent prognostic value for the histologic type. Beck showed that the only combination that could be retained in the multivariate model was chromophobe versus clear cell histology, whereas papillary feature was not retained. In the former study, it should be noted that nuclear grade was not included in the Cox model. Thus, it is not certain that the chromophobe feature would have been retained as an independent prognostic variable if grade had been included. Although associated in most studies with stage and grade and providing a significant survival stratification in univariate analysis, there is no strong evidence from all these studies that the histologic subtype should be considered an independent prognostic parameter when compared with TNM stage and grade. In our study including 4,063 patients with 3,564 (87.7%) clear cell, 396 (9.7%) with papillary, and 103 (2.5%) with chromophobe tumors, no survival difference was observed between papillary and clear cell histologies in metastatic tumors, and only TNM stage, Fuhrman grade and ECOG PS were retained as independent prognostic variables in the overall population. There is a need in patients with metastatic RCC to define prognostic factors that enable stratification of patients into risk groups and predict response to therapy. It is noteworthy that none of the prognostic models that have been developed recently in metastatic RCC has integrated histologic subtype as a prognostic variable.19,20 The only histologic feature that has been retained as a prognostic variable in predicting response to immunotherapy is the presence of a sarcomatoid component.21 Motzer et al22 found that patients with chromophobe tumors had a longer survival compared with papillary and collecting duct histology in 64 metastatic patients with non–clear cell histology. However, as shown in our study, the chromophobe feature is extremely rare in metastatic series (0.6%). Thus, this variable has little utility in decision making in metastatic RCC. Additionally, in our series no survival difference was observed between 59 metastatic patients with papillary and 765 patients with clear cell tumors. Thus, we failed to confirm the general belief that metastatic papillary tumors have a more aggressive profile than metastatic clear cell tumors. The main limitation of our study is the lack of central pathologic review. It is likely that it resulted in misclassifications. However, it is worth noting that, although due to the relative infeasibility of centrally reviewing more than 4,000 RCC cases occurring across eight independent medical centers on two continents, and although cases were not reviewed across centers, within each center cases were often reviewed by a single pathologist, lending a uniformity to the diagnoses of pathologic variables within sites. In other sites, diagnoses were rendered by a small group of pathologists experienced in the classification of RCC using the Heidelberg and Furhman classifications. Nevertheless, it cannot be denied that differences in the relative prevalence of the three histologic subtypes that were found across the eight institutions may be due to histologic misclassification rather than to differences in patient recruitment or cancer epidemiology in each place. It could also explain why our reported percentages of unclassified renal cell carcinomas (0.7%) and chromophobe cell carcinomas (2.5%) were lower compared to that of the literature and why the overall percentage of clear cell carcinoma was higher than anticipated. It is also conceivable that misclassifying good prognostic tumors together with more aggressive tumors could somewhat "neutralize" the prognostic effect of the histologic type. This issue was recently addressed by a group at Harvard that recently presented relevant data at the American Society of Clinical Oncology. To assess the accuracy and completeness of pathology data contributed to large databases, Carloss et al23 collected hematoxylin and eosin–stained pathology slides and outside pathology reports of RCC primary tumors from patients with RCC. Central review was performed according to preset criteria by a single pathologist blinded to the outside pathology reading, and the degree of concordance between the two readings was compared. In this study, 99% of those classified as clear cell by local review were confirmed to be the same by central review, while there was an 84% concordance between the two reviews in describing papillary features. Overall, therefore, we don't believe that a centrally performed classification would have significantly modified our results since differences obviously existed in univariate analysis. Nevertheless, it is not surprising that those differences disappeared in multivariate analysis due to strong correlations that exist among stage, grade, and histologic types. Even though tools for distinguishing histologic features have evolved during recent years, rendering more uniform the evaluation of histology across centers, it is likely that in the oldest cases, pathologists tended to lump all tumors with clear cell features including some papillary renal cell carcinomas and chromophobe renal cell carcinomas into the category of carcinoma with clear cell features. Another limitation of the study due to the lack of central pathologic review is the concern about reliability of the tumor grading. It is known that inter- and intra-observer discrepancies exist concerning Fuhrman grade classification.24,25 It is likely that such inaccuracies occurred in our study since different pathologists from different centers evaluated nuclear grade during a long period of time. Conversely, the lack of centralized pathology review perhaps lends a better insight into the prognostic value of histologic type in general practice. Finally, the prognostic issue is not the only reason that it is important to recognize different histologic subtypes in RCC. As we have demonstrated, different subtypes have different predilections for stage as well as different metastatic profiles; different subtypes have different propensities for multicentricity and renal vein invasion and may respond differently to different therapies. Given the fact that each type of cancer has its own genetic and protein signature and has corresponding syndromic and familial associations, those differences are probably driven at the molecular level. Although substantial progress has been made during the last 10 years, the story of histologic subtyping is not complete. Further understanding of the molecular basis of histologic subtyping may provide us with greater insight into the various tumorigenic pathways of kidney carcinogenesis. In summary, histologic subtyping of RCC along morphologically and genetically/molecularly well-characterized subtypes has prognosis significance in a univariate setting; however, in an independent analysis including TNM stage, Fuhrman nuclear grade, and ECOG PS, the prognostic significance does not hold true.
The authors indicated no potential conflicts of interest.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Jemal A, Murray T, Samuels A, et al: Cancer Statistics, 2003. CA Cancer J Clin 53:5-26, 2003 2. Storkel S, Eble JN, Adlakha K, et al: Classification of renal cell carcinoma: Workgroup No. 1. Union Internationale Contre le Cancer (UICC) and the American Joint Committee on Cancer (AJCC). Cancer 80:987-989, 1997[CrossRef][Medline] 3. Kovacs G, Akhtar M, Beckwith BJ, et al: The Heidelberg classification of renal cell tumours. J Pathol 183:131-133, 1997[CrossRef][Medline] 4. Moch H, Gasser T, Amin MB, et al: Prognostic utility of the recently recommended histologic classification and revised TNM staging system of renal cell carcinoma: A Swiss experience with 588 tumors. Cancer 89:604-614, 2000[CrossRef][Medline] 5. Amin MB, Tamboli P, Javidan J, et al: Prognostic impact of histologic subtyping of adult renal epithelial neoplasms: An experience of 405 cases. Am J Surg Pathol 26:281-291, 2002[CrossRef][Medline] 6. Cheville JC, Lohse CM, Zincke H, et al: Comparisons of outcome and prognostic features among histologic subtypes of renal cell carcinoma. Am J Surg Pathol 27:612-624, 2003[CrossRef][Medline] 7. Guinan P, Sobin LH, Algaba F, et al: TNM staging of renal cell carcinoma: Workgroup No. 3. Union International Contre le Cancer (UICC) and the American Joint Committee on Cancer (AJCC). Cancer 80:992-993, 1997[CrossRef][Medline] 8. Fuhrman SA, Lasky LC, Limas C: Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 6:655-663, 1982[Medline] 9. Oken MM, Creech RH, Tormey DC, et al: Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol 5:649-655, 1982[Medline]
10. Zisman A, Pantuck AJ, Dorey F, et al: Improved prognostication of renal cell carcinoma using an integrated staging system. J Clin Oncol 19:1649-1657, 2001 11. Mancilla-Jimenez R, Stanley RJ, Blath RA: Papillary renal cell carcinoma: A clinical, radiologic, and pathologic study of 34 cases. Cancer 38:2469-2480, 1976[CrossRef][Medline] 12. Amin MB, Corless CL, Renshaw AA, et al: Papillary (chromophil) renal cell carcinoma: Histomorphologic characteristics and evaluation of conventional pathologic prognostic parameters in 62 cases. Am J Surg Pathol 21:621-635, 1997[CrossRef][Medline]
13. Beck SD, Patel MI, Snyder ME, et al: Effect of papillary and chromophobe cell type on disease-free survival after nephrectomy for renal cell carcinoma. Ann Surg Oncol 11:71-77, 2004 14. Ljungberg B, Alamdari FI, Stenling R, et al: Prognostic significance of the Heidelberg classification of renal cell carcinoma. Eur Urol 36:565-569, 1999[CrossRef][Medline] 15. Mejean A, Hopirtean V, Bazin JP, et al: Prognostic factors for the survival of patients with papillary renal cell carcinoma: Meaning of histological typing and multifocality. J Urol 170:764-767, 2003[CrossRef][Medline] 16. Thoenes W, Storkel S, Rumpelt HJ, et al: Chromophobe cell renal carcinoma and its variants—A report on 32 cases. J Pathol 155:277-287, 1988[CrossRef][Medline] 17. Nagashima Y: Chromophobe renal cell carcinoma: Clinical, pathological and molecular biological aspects. Pathol Int 50:872-878, 2000[CrossRef][Medline] 18. Crotty TB, Farrow GM, Lieber MM: Chromophobe cell renal carcinoma: Clinicopathological features of 50 cases. J Urol 154:964-967, 1995[CrossRef][Medline]
19. Motzer RJ, Mazumdar M, Bacik J, et al: Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol 17:2530-2540, 1999 20. Atzpodien J, Royston P, Wandert T, et al: Metastatic renal carcinoma comprehensive prognostic system. Br J Cancer 88:348-353, 2003[CrossRef][Medline] 21. 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 98:2566-2575, 2003[CrossRef][Medline]
22. Motzer RJ, Bacik J, Mariani T, et al: Treatment outcome and survival associated with metastatic renal cell carcinoma of non-clear cell histology. J Clin Oncol 20:2376-2381, 2002 23. Carloss E, Regan M, Upton M, et al: Concordance of local pathology interpretation (LR) with central pathology review (CR) in primary tumor specimens from patients with renal cancer (RCC). J Clin Oncol, 22: 4548, 2004 (suppl 14) 24. Al-Aynati M, Chen V, Salama S, et al: Interobserver and intraobserver variability using the Fuhrman grading system for renal cell carcinoma. Arch Pathol Lab Med 127:593-596, 2003[Medline]
25. Lohse CM, Blute ML, Zincke H, et al: Comparison of standardized and nonstandardized nuclear grade of renal cell carcinoma to predict outcome among 2,042 patients. Am J Clin Pathol 118:877-886, 2002 Submitted July 12, 2004; accepted January 12, 2005.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 (Fisher's exact test) and Student t tests were used to compare the qualitative and quantitative variables, respectively. For the purpose of survival analysis, histologic subtype was considered a single categoric variable. Estimations of the cumulative survival distributions were calculated according to the Kaplan-Meier method, and the log-rank test was used to compare groups. The Cox proportional hazards model was used for multivariate analysis. All P values were two sided, and a P < .05 was considered significant. All data analysis was processed through the SPSS statistical software version 10.0 (SPSS Inc, Chicago, IL). 
