Original Study
The Prognostic Impact of Tumor Size on Cancer-Specific and Overall Survival Among Patients With Pathologic T3a Renal Cell Carcinoma

https://doi.org/10.1016/j.clgc.2014.06.011Get rights and content

Abstract

Background

We aimed to determine the prognostic role of tumor size in patients with stage pT3a renal cell carcinoma (RCC).

Patients and Methods

We analyzed our database of patients who underwent radical nephrectomy for RCC between July 2000 and December 2013. Clinical and pathologic data were obtained for each patient. Patients with stage pT3a disease were divided into 2 subgroups according to the most informative threshold for pathologic tumor dimension that was able to predict survival outcomes (group 1, ≤ 8 cm; group 2, > 8 cm).

Results

Globally, 185 consecutive patients were evaluated. The median (interquartile range [IQR]) follow-up was 32 months (18-62 months). The median (IQR) pathologic tumor size was 7.5 cm (5.7-10 cm). Seventy (34.3%) patients died of RCC during the follow-up period. Patients in group 2 experienced worse cancer-specific survival (CSS) rates compared with those in group 1, (5- and 10- year CSS, 52% and 40% vs. 67% and 63%, respectively; P = .001). Overall survival (OS) rates were significantly lower for patients included in group 2 compared with patients in group 1 (5- and 10- year OS rates, 46% and 38% vs. 60% and 57%, respectively; P = .01). Subgroup stratification (hazard ratio [HR], 3.65; P < .001), presence of positive surgical margins (HR, 3.86; P = .22), high Fuhrman grade (HR, 4.33; P < .001), and the presence of sarcomatoid cells (HR, 2.61; P = .02) were found to be independent predictors of CSS.

Conclusion

Worse oncologic outcomes are observed in patients with stage pT3a RCC tumors > 8 cm. The current TNM classification still does not precisely correlate with CSS. Tumor size should be taken into account in a future revision of the TNM staging system.

Introduction

Renal cell carcinoma (RCC) accounts for approximately 3% to 4% of all adult malignant neoplasms.1 The incidence of RCC has been increasing in recent years,2 and 20% to 30% of patients are still diagnosed with locally advanced disease.3 Surgical excision remains the mainstay of treatment in all cases of nonmetastatic RCC, even in non–organ-confined tumors, and survival rates are strongly influenced by several clinicopathologic features, such as tumor size, pathologic stage, nuclear grade, histologic tumor necrosis, and the presence of sarcomatoid features.4, 5 Pathologic stage (pT) according to the 2009 tumor, nodes, metastases (TNM) classification system,6 still represents one of the most important prognostic factors for cancer-specific survival (CSS) rates.7 In the most recent years, several revisions of the TNM staging system for RCC have been proposed to improve its prognostic accuracy,7, 8 even for locally advanced disease.9, 10 Unlike organ-confined RCC, stage pT3a RCC is classified only according to its anatomic extension (perirenal/sinus fat invasion or renal vein involvement), regardless of tumor size; consequently, small and large neoplasms are still classified together.6 The prognostic role of tumor size in stage pT3 RCC has been evaluated in several studies,11, 12 and the primary tumor dimension has been shown to be an important prognosticator in patients with RCC.4, 13 The aim of the present study is to determine whether tumor size could be an important prognostic parameter in patients with stage pT3a RCC and to identify the optimal threshold for cancer-specific and overall mortality rates in a defined cohort of patients treated with radical nephrectomy for locally advanced RCC.

Section snippets

Patients and Methods

We analyzed our prospectively collected institutional database of 786 patients who underwent radical or partial nephrectomy for renal cancer between July 2000 and December 2013. Among the 511 consecutive patients who underwent radical nephrectomy for nonmetastatic RCC, we evaluated 193 patients with unilateral stage pT3a RCC classified according to the seventh edition of the TNM.6 Of these patients, 8 were lost to follow-up, and the remaining 185 patients were included in this study. Surgical

Statistical Analysis

Continuous variables were reported as mean value ± standard deviation (SD) or median value and range/interquartile range (IQR), as appropriate. The tumor dimensions were evaluated as continuous variables and were divided into 2 groups according to the most informative threshold for survival by applying the analysis of variance test for every possible threshold value and by choosing the lowest P value. Kaplan-Meier and Cox proportional univariable and multivariable regression models were used to

Results

The mean (± SD) and the median (IQR) follow-up were 44.7 ± 37.4 and 32 months (18-62 months), respectively. The clinical and pathologic characteristics of the entire cohort of patients are summarized in Table 1. In detail, the mean (± SD) age was 63.3 ± 11.8 years; 149 patients (80.5%) were men, and in 93 cases (50.3%) the diagnosis was performed in asymptomatic individuals. The mean (± SD) and the median (IQR) pathologic tumor sizes were 8.05 ± 2.8 and 7.5 (5.7-10) cm, respectively. In 4

Discussion

Stage T3a renal cell carcinoma still encompasses an inhomogeneous group of patients with different survival outcomes. The current 2009 TNM staging system stratifies patients with stage pT3a disease according only to the tumor's anatomic extension into the perinephric/hilar fat or its invasion into the renal vein.6 Despite evidence reporting the prognostic role of tumor size among patients with stage pT3a RCC,11, 12, 14, 15 the primary tumor dimension is still not used as a determinant of T

Conclusion

Tumor size proved to be an independent prognostic factor in stage pT3a RCC, with worse oncologic outcomes observed in patients with a tumor dimension > 8 cm. The current TNM classification of stage pT3a still does not precisely correlate with CSS outcomes, and tumor size should be taken into account in a future revision of the TNM staging system. Further studies are needed to confirm our results.

Disclosure

The authors have stated that they have no conflicts of interest.

References (25)

Cited by (0)

These authors contributed equally to this work.

View full text