Update on staging controversies for locally advanced renal cell carcinoma

Tumor stage, as determined by the American Joint Committee for Cancer (AJCC) primary tumor classification system (Table 1), remains the most important and widely utilized prognosticator of progression-free and overall survival following surgical management of renal cell carcinoma (RCC) [1]. Contemporary literature reveals that 5-year cancer-specific survival (CSS) after radical nephrectomy ranges from 75–95% for patients with organ-confined disease to 0–5% for patients with metastatic disease at presentation [2,3]. Located between these clinical extremes are patients with locally advanced RCC (LARCC), considered to be at significant risk for progression and death from their renal tumors due to adverse pathologic features. Since definitions of LARCC vary throughout the literature, the authors of this text utilize the 2002 AJCC primary tumor classification to identify LARCC as nonmetastatic (M0) tumors in pT3–4 categories, which include features of perirenal fat, renal sinus and/or ipsilateral adrenal gland invasion (pT3a), associated malignant venous thrombus below (pT3b) and above (pT3c) the diaphragm and extension beyond the Gerota’s fascia (pT4) [1].

An accurate and clinically useful staging system is an essential tool, used to provide patients with counseling regarding prognosis, select treatment modalities and determine eligibility for clinical trials. Based on peer-reviewed observations in the literature and reflecting changing trends in RCC presentation, treatment and outcomes, the AJCC TNM staging system has been continuously modified to improve its accuracy [4]. Recently, several well-designed, retrospective studies have demonstrated that the current TNM staging system of LARCC lacks accuracy in stratifying patients according to their disease-specific survival [5–8]. Specifically, controversy exists regarding the prognostic significance of tumor size, ipislateral adrenal invasion, location of extrarenal tumor extension and the level of venous tumor thrombus in patients with LARCC.

This article highlights such controversies and provides an update on recently suggested revisions and improvements to the current staging system of LARCC.

Prognostic significance of tumor size in patients with LARCC

Tumor size has been established to be one of the most important predictors of cancer-specific outcome in patients with RCC. However, the current 2002 TNM primary classification defines stage T3 RCC as a tumor that invades the perinephric fat, renal sinus fat, extends into the venous system or directly invades the ipsilateral adrenal gland, regardless of tumor size. Several authors have recently demonstrated the importance of tumor size in patients with pathologic stage T3 RCC. Murphy and colleagues retrospectively analyzed oncologic outcome in 118 patients surgically treated for T2–3aN0M0 RCC and found that patients with T3aN0M0 tumors had a superior CSS, when compared with those with T2N0M0 disease, suggesting that absolute tumor size is a more powerful outcome prognosticator than extrarenal tumor extension [9]. These findings were recently confirmed by Siemer and colleagues who, in their analysis of 237 nonmetastatic pT3a renal tumors, found that tumor size cutoff of 7 cm effectively stratified pT3a RCC patients into two distinctly different prognostic groups [10]. Furthermore, these investigators demonstrated no significant differences in CSS between patients with pT3a tumors 7 cm or smaller and pT1 tumors or between pT3a tumors larger than 7 cm and pT2 tumors [11]. Finally, in a retrospective, multi-institutional analysis of 2113 patients treated by surgical resection for T2–4 RCC, Lam and colleagues evaluated the prognostic relevance of tumor size in pT3a RCC [12]. In agreement with previous reports, these authors found that the T3a staging category can be improved by substratification of these patients according to a tumor size cutoff of 7 cm. In their study, the CSS of patients with T3a larger than 7 cm RCC was similar to patients with T3b tumors, while the CSS of patients with T3a smaller than 7 cm RCC was equal to those with T2 cancers.

Based on available evidence, the future revisions of the AJCC TNM classification of RCC should incorporate tumor size into the staging of LARCC.

Ipsilateral adrenal involvement

Isolated ipsilateral direct adrenal invasion by RCC, without evidence of systemic metastatic disease, is a rare event, with a reported incidence of 1–5% [13–17]. Patients with direct adrenal involvement by RCC are currently classified as pT3a, despite convincing evidence from several large, single-center experiences demonstrating significantly lower CSS for patients with direct adrenal invasion than for other patients with pT3a RCC. Investigators from the University of California Los Angeles (UCLA) reviewed the CSS of 27 pT3a patients with direct adrenal invasion and 187 pT3a patients with no evidence of adrenal involvement, they found that patients with adrenal gland invasion had significantly inferior median (12.5 vs 36.0 months; p < 0.001) and 5-year CSS rates (0 vs 36%) [18]. As the median survival of patients with direct adrenal invasion was similar to those with tumors extending beyond the Gerota’s fascia, the authors concluded that RCCs invading the adrenal gland should be assigned pT4 staging category. These findings were confirmed by Thompson and colleagues, who reported a 5-year CSS probability of 20.2% in 22 pT3a–b patients with cancer invading the ipsilateral adrenal gland compared with 53.9% in pT3a and 42.7% in pT3b patients without concomitant adrenal gland involvement [19]. Similarly, these investigators found similar 5-year CSS in pT3a–b patients with direct adrenal invasion and patients with pT4 tumors (20 and 14%, respectively; p = 0.490). Inferior CSS of patients with direct adrenal invasion compared with pT3 patients without adrenal involvement by RCC has been affirmed recently in several other independent patient populations [8,12].

In summary, it is clear that direct adrenal invasion by RCC is associated with aggressive tumor biology and dismal CSS. Future revisions of the AJCC TNM classification are likely to group these patients in a separate staging category.

Location of extrarenal tumor extension

The renal sinus is a fatty compartment located within the confines of the kidney, it envelopes the renal collecting system and contains numerous veins and lymphatics [20]. The abundant lympho–vascular network of the renal sinus fat, similar to that of the perinephric fat, provides an important avenue for metastatic tumor dissemination. Based on the observations of Bonsib and colleagues and experience with pediatric renal tumors, the renal sinus compartment has been recognized as an important pathway for tumor dissemination from an otherwise organ-confined RCC [21,22]. Subsequently, in 2002 the AJCC pT3a RCC primary tumor classification category was modified to include tumors with renal sinus invasion in addition to tumors with perinephric fat invasion and direct invasion of the ipsilateral adrenal gland [1].

Recently, Thompson and colleagues examined the impact of location of extrarenal extension on CSS in pT3a RCC patients [23]. The authors from the Mayo Clinic (MN, USA) were able to demonstrate that the presence of renal sinus fat invasion was associated with a higher risk of regional lymph node metastases, higher Fuhrman grade and a greater incidence of sarcomatoid differentiation. Moreover, on multivariate analyses, after adjusting for the impact of other important prognostic features, the presence of renal sinus fat invasion was an independent predictor of oncologic outcome in pT3a patients [23]. The aforementioned findings were not corroborated by our studies at the MD Anderson Cancer Center. We found no difference in 5-year CSS between 166 (45.5%) patients with sinus fat invasion and 199 (54.5%) patients with perinephric fat invasion only (50.8 and 54.1%; p = 0.782, respectively) [24]. On univariate analyses, neither the presence of sinus fat invasion, nor the location of extrarenal extension (assessed as perinephric fat vs sinus fat vs both) correlated with CSS following surgical treatment (hazard ratio [HR]: 1.052; p = 0.783 and HR: 1.072; p = 0.543) [24]. These divergent findings may be explained by a larger cohort of patients and a higher frequency of renal sinus fat invasion in the latter series. In agreement with the findings of Bonsib and colleagues, who prospectively evaluated their specimens using thin sections of the renal sinus and found sinus fat invasion in 55% of tumors, the presence of renal sinus fat invasion was identified in 56% of the MD Anderson pT3a cohort, while Thompson and colleagues identified renal sinus fat invasion in only 21% of their pT3a patients.

In summary, there are conflicting data regarding the prognostic significance of renal sinus fat invasion in pT3 RCC patients. Current evidence suggests that the location of extrarenal tumor extension does not reflect the biology of RCC, but rather mirrors the origin of the tumor within the renal parenchyma. While pathological assessment of renal sinus fat invasion should be routine in all patients with RCC, there is no conclusive evidence at present that the location of extrarenal tumor extension should be reflected as a substratification within the TNM staging system of RCC.

Extent of venous tumor thrombus

The incidence of venous tumor extension in patients with RCC is reported to be between 4 and 10% [25–27]. In up to 25% of patients, the tumor extends above the confluence of the hepatic veins and, in approximately 5%, the thrombus is supradiaphragmatic or intracardiac. While the knowledge of tumor thrombus extent is critical for appropriate surgical planning, the prognostic implication of tumor thrombus level has been debated extensively. Based on earlier reports, which suggested decreased CSS in patients with supradiaphragmatic venous tumor thrombus, the current 2002 TNM classification of renal tumors groups these patients into a separate staging category (pT3c) from patients with renal vein or infradiaphragmatic caval involvement (pT3b) [28–30]. Subsequently, most studies have found no difference in survival based on the level of inferior vena cava involvement [31]. Much of the current debate centers on prognostic significance of the renal vein-only thrombi versus thrombi with intracaval extension. Investigators from UCLA reviewed their experience with a nephrectomy and tumor thrombectomy in 117 patients with renal vein-only and 109 patients with inferior vena cava involvement, and found no difference in 3-year CSS between these groups of patients (36 vs 35%, respectively; p = 0.883) [31]. These findings were not substantiated by Moinzadeh and colleagues who, with longer median follow-up, found significantly decreased 10-year CSS in patients with vena caval thrombi compared with patients with renal vein thrombi only (66 vs 29%; p < 0.0001) [32]. Similarly, the Mayo group recently reported that patients with pT3b RCC and intracaval tumor thrombus extension were significantly more likely to die of RCC compared with patients with pT3b RCC and renal vein tumor thrombus (risk ratio: 1.62; p < 0.001) [5]. Unfortunately neither of the previous studies adjusted for clinico–pathologic features important in RCC, which may have accounted for differences in CSS seen between patients with renal vein and caval thrombi.

There is ample evidence to suggest that oncologic outcome of RCC patients with venous tumor thrombus correlates with biological tumor aggressiveness, as determined by tumor grade, perinephric fat invasion, nodal and metastatic status, rather than with the thrombus level. While several recent reports suggest that patients with caval tumor thrombi have a significantly decreased survival compared with those with thrombi in the renal vein only, these data need to be further validated in a multicenter collaborative effort.

Proposals for revision of the current TNM staging system for LARCC

Keeping up with the modern demand for evidence-based medical practice, and based on compelling clinical evidence, several investigators have proposed changes to the current TNM staging system of LARCC (Table 2).

Based on previously reported observations that the invasion of renal fat portends a worse prognosis among patients with venous tumor thrombi, Thompson and colleagues proposed a system where patients with locally advanced RCC were reclassified into five subcategories: pT3a (renal vein thrombus without extrarenal extension), pT3b (perinephric fat invasion only), pT3c (renal vein thrombus with extrarenal extension or vena caval thrombus below diaphragm without extrarenal extension) and pT3d (vena cava thrombus below diaphragm with extrarenal extension or presence of supradiaphragmatic thrombus) [6]. Patients with direct adrenal invasion were grouped into the pT4 category. The predictive ability of their proposed classification was superior to that of the 2002 classification (concordance indexes of 0.635 and 0.561, respectively) [6].

After a survival analysis of a large group of LARCC treated in the modern era, investigators from the MD Anderson Cancer Center found that patients with extrarenal tumor extension only and patients with any level of venous tumor thrombus without concomitant extrarenal tumor extension, were at a similar risk of death from RCC and had overlapping 5- and 10-year CSS [8]. Conversely, patients with both venous thrombus and extrarenal tumor extension were at a significantly increased risk of death from RCC. Based on these findings, we proposed a clinically simple reclassification of LARCC patients, where patients with extrarenal extension only or venous tumor thrombus only were grouped into a separate prognostic category from patients with both extrarenal tumor extension and venous tumor thrombus. Again, patients with ipsilateral adrenal invasion were assigned pT4 classification. Despite its simplicity, the LARCC staging system proposed in this study demonstrated significantly improved prognostic accuracy, compared with the 2002 AJCC TNM staging system (concordance indexes of 0.625 and 0.580, respectively) [8].

Integrated staging algorithms

Despite their improved predictive ability, the purely anatomic staging systems discussed previously do not incorporate alternative clinico–pathologic factors important in RCC prognosis, such as tumor grade, histologic subtype, presence of coagulative or gross tumor necrosis, sarcomatoid differentiation or patient performance status. In an attempt to improve risk assessment in RCC, combinations of the TNM staging system with various other prognostic factors have allowed for the emergence of new comprehensive staging or risk assessment paradigms. Investigators at UCLA have developed the UCLA Integrated Staging System (UISS), a novel staging system based upon TNM stage, grade and Eastern Cooperative Oncology Group performance status [33]. Patients are stratified into three risk groups according to the probability of tumor recurrence and survival. This staging system has been internally and externally validated and has been demonstrated to be superior to TNM staging [34]. The Mayo Clinic group developed the stage, size, grade and necrosis (SSIGN) score, in which patients with clear cell RCC were assigned a score based on tumor stage, tumor size, nuclear grade and the presence of necrosis [35]. Using a SSIGN score, CSS at 1–10 years post treatment can be estimated for an individual patient. The predictive ability of the SSIGN score was recently validated externally and also has been shown to be superior to TNM staging alone [36]. Researchers from Memorial Sloan–Kettering Cancer Center combined tumor stage, tumor size, histologic subtype, and symptoms at presentation into a nomogram, predicting the probability of freedom from recurrence at 5 years after treatment [37]. Finally, Yaycioglu and colleagues and Cindolo and colleagues independently devised preoperative clinical predictive models based on patient symptoms at presentation and radiographic tumor size, stratifying patients into high and low risk for recurrence [38,39]. Neither Yaycioglu nor Cindolo’s models have been validated in an independent patient cohort.

It is certain that the future of RCC staging will rely on a combined and structured analysis of clinical information, anatomic extent of the disease, histopathologic criteria and molecular markers. A glimpse into the future of RCC staging has recently been offered by the UCLA group, who demonstrated that the addition of expression data from five markers, including carbonic anhydrase IX, Ki-67, gelsolin, vimentin and p53, significantly improved the discriminating ability of the UISS [40].

Conclusion

Patients with LARCC harbor adverse pathologic features and are at significant risk of disease relapse. Consequently, there is a clear-cut need for improved ability to predict an individual tumor’s behavior. Within the emerging framework of nomograms and algorithms for the prediction of overall survival and CSS, the pathologic stage of RCC remains the most important determinant of disease-specific outcome. The shortcomings and controversies of the current TNM RCC staging system, as discussed in this article, will undoubtedly lead to its continuous improvement. However, as we move into the future, combined and structured analysis of clinical information, histopathologic criteria, as well as molecular and genetic markers of disease, will enable the stratification of LARCC patients into more sophisticated risk categories and, ultimately, permit delivery of individualized therapies for targeted patient populations.

Table 1. The American Joint Committee on Cancer 2002 staging system for renal cell carcinoma.

Primary tumor (T)
TX	Primary tumor cannot be assessed (information not available)
T0	No evidence of a primary tumor
T1a	Tumor is 4 cm in diameter or smaller, limited to the kidney
T1b	Tumor is larger than 4 cm but smaller than 7 cm, limited to the kidney
T2	Tumor is larger than 7 cm, limited to the kidney
T3a	Direct tumor invasion into the adrenal gland, perinephric fat or renal sinus fat, tumor confined within Gerota’s fascia
T3b	Tumor invasion of the renal vein or the inferior vena cava below the diaphragm
T3c	Tumor invasion of the vena cava above the diaphragm
T4	Tumor invasion beyond Gerota’s fascia
Regional lymph nodes (N)
NX	Regional lymph nodes cannot be assessed
N0	No regional lymph node metastasis
N1	Metastasis to one regional lymph node
N2	Metastasis to more than one regional lymph node
Distant metastasis (M)
MX	Presence of distant metastasis cannot be assessed
M0	No distant metastasis
M1	Distant metastasis present; includes metastasis to nonregional lymph nodes and/or to other organs
Stage grouping
I	T1a–1b, N0, M0
II	T2, N0, M0
III	T1a–3b, N1, M0 or T3a–3c, N0, M0
IV	T4, N0–1, M0 or any T, N2, M0 or any T, any N, M1

Table 2. Proposals for the revision of the current TNM staging system of locally advanced renal cell carcinoma.

Primary tumor stage	Pathologic tumor characteristics
	Mayo Clinic	MD Anderson Cancer Center
pT3a	Renal vein thrombus without extrarenal tumor extension	Venous thrombus (any level) without extrarenal tumor extension; or extrarenal tumor extension without venous thrombus
pT3b	Extrarenal tumor extension without venous thrombus	Venous thrombus (any level) and extrarenal tumor extension
pT3c	Renal vein thrombus and extrarenal tumor extension; or infradiaphragmatic caval thrombus without extrarenal tumor extension
pT3d	Infradiaphragmatic thrombus with extrarenal tumor extension; or supradiaphragmatic thrombus
pT4	Direct ipsilateral adrenal invasion; or tumor extension beyond Gerota’s fascia	Direct ipsilateral adrenal invasion; or tumor extension beyond Gerota’s fascia