Introduction

Several prognostic tools have been developed for patients with renal cell carcinoma (RCC). However, those calculators usually consider T3 tumors a single entity and little is known of their performance in patients with tumor thrombus. We assess the value of the Mayo Clinic Stage, Size, Grade, and Necrosis (SSIGN) score for prognosis of cancer-specific survival (CSS) in a large multicenter patient cohort with RCC and venous tumor thrombus.

Methods

The records of 846 patients collected by 17 international centers within the International Renal Cell Carcinoma-Venous Thrombus Consortium were retrospectively analyzed. Patients were treated with radical nephrectomy and tumor thrombectomy from 1971 to 2014. SSIGN scores were assigned (Table 1) and grouped in the analysis as in prior validation studies. Kaplan-Meier and Cox regression analyses examined CSS. Harrell&[prime]s concordance index was calculated.

Results

Median follow-up was 35 months (interquartile range [IQR], 11-68 months) in the 390 patients alive at last follow-up; 456 (53.9%) died of any cause and 351 (41.5%) died of RCC. The median primary tumor diameter was 9 cm (IQR, 7.0-11.4). Tumor stage, grade, and necrosis distributions are shown in Table 1. The median SSIGN score was 7 (IQR, 5-9). All SSIGN features were significantly associated with CSS in univariate analysis. All but tumor size remained significant after controlling for other factors (Table 1). Harrell's concordance index was 0.72 for prediction of CSS compared to 0.81-0.88 in prior studies, this study notably including pT3-T4 patients only. Five- and ten-year survival was significantly worse in patients with scores 7-9 and 10+. Figure 1 shows survival curves.

Conclusions

This analysis of the largest reported patient cohort with RCC and tumor thrombus demonstrates prognostic utility of the SSIGN score though pT3 subclassification was not considered in the original tool development. This may serve as a useful clinical tool in patients with tumor thrombus for follow-up counseling and clinical trial design.

Funding

none