© 2007 THE AUTHORS 444 JOURNAL COMPILATION © 2 0 0 7 B J U I N T E R N A T I O N A L | 1 0 1 , 4 4 4 – 4 4 9 | doi:10.1111/j.1464-410X.2007.07257.x Original Articles ANALYSIS OF T3B RCCKLAVER et al. Analysis of renal cell carcinoma with subdiaphragmatic macroscopic venous invasion (T3b) Sjoerd Klaver, Steven Joniau, Raphael Suy*, Raymond Oyen† and Hein Van Poppel Departments of Urology, *Vascular Surgery and †Radiology, University Hospital Gasthuisberg KU Leuven, Belgium Accepted for publication 20 July 2007 nephrectomy with resection of subdiaphragmatic tumour thrombus (TT) between October 1990 and May 2006. The log-rank and Cox uni- and multivariate regression analysis were used to evaluate predictive factors for CSS. RESULTS In all, 101 cases were identified. In the N0M0 group, univariate Cox regression analysis confirmed that ipsilateral adrenal gland invasion, Mayo Clinic level of TT, histological subtype and fat invasion were significantly associated with worse CSS. In multivariate Cox regression analysis, only Mayo Clinic level of T T was an independent predictor for CSS. In the subgroup with renal vein involvement only, the median CSS was not reached. In the subgroups with level I, II and III TT involvement, the median CSS was 69, 26 and 21 months, respectively. In the N + and/or M + group, only tumour size and type were independent predictors of CSS, while the level of TT was not. Radical nephrectomy yielded poor results with a median CSS of 13 months. CONCLUSION The Mayo Clinic level of T T is an independent prognostic predictor for CSS in non- metastatic T3b RCC. We strongly support the need for re-classification of the currently applied 2002 Tumour-Node-Metastasis staging system, which in its present form does not discriminate between levels of subdiaphragmatic venous invasion. KEYWORDS renal cell carcinoma, T3b, tumour thrombus, Mayo Clinic level, cancer-specific survival Study Type – Prognosis (individual case series) Level of Evidence 4 OBJECTIVE To review our institutional experience of surgery for renal cell carcinoma (RCC) with subdiaphragmatic macroscopic venous invasion (T3b) and to assess variables associated with cancer-specific survival (CSS), as the stratification of RCC with venous involvement (T3b and T3c) is subject to debate. PATIENTS AND METHODS We retrospectively reviewed the hospital records of patients who underwent a radical INTRODUCTION Up to 10% of patients with newly diagnosed RCC have tumour thrombus (T T) involving the renal vein or the inferior vena cava (IVC) [1]. Once, this phenomenon was thought to be a poor prognostic factor and was not considered an indication for surgery because of the major surgical risk. With improvement in surgical technique, it became possible to remove a TT with significantly reduced surgical risk [2,3]. Recent studies report 5-year survival rates of 45–69% for patients with RCC and venous T T if the tumour is otherwise confined to the kidney [4–6]. While perinephric or renal sinus tumour extension, direct ipsilateral adrenal gland invasion, lymph node (LN) involvement and/or distant metastasis, incomplete tumour or thrombus excision, and tumour invasion in the caval wall have adverse prognostic impact, the impact of the extent of venous RCC TT in the renal vein and/or IVC remains subject to debate [1,5–12]. The TNM staging system is the most widely used method to classify the local extension of the primary tumour, loco-regional LN involvement and the presence of metastasis [13,14]. Recently, different proposals have been made for the re-classification of patients with pT3–T4 RCC and patients with LN metastasis [11,14–19]. Furthermore, no distinction is made in the pT3b group between patients with and with no fat invasion and/or adrenal gland invasion and the level of subdiaphragmatic TT involvement is not included in the TNM classification. We retrospectively analysed our database of 101 patients who underwent surgery for T3b RCC and assessed variables associated with cancer-specific survival (CSS). The aim of this study was to identify independent predictors of CSS in this specific group and to propose arguments for a revision of the 2002 TNM staging system to improve its prognostic accuracy. We also assessed to what extent metastatic disease determines survival in patients with TT to evaluate whether surgery is still beneficial in these patients. PATIENTS AND METHODS Our institutional database contained 987 patients who underwent surgery for RCC from October 1990 until May 2006. The record contained 101 (10.2%) patients with RCC and subdiaphragmatic TT, who were included in the analysis (56 men and 45 women, age range 21–82 years). We evaluated the presence of synchronous metastases, LN involvement, level of TT, nuclear grade according to the Fuhrman classification, tumour size (cm), histological subtype according to the Heidelberg guidelines, perirenal fat and/or renal sinus invasion, and ipsilateral adrenal gland invasion. Radiological staging was done by contrast-enhanced CT of the abdomen and