A multivariate analysis of prognostic factors from assure (E2805): Adjuvant sorafenib or sunitinib for unfavorable renal carcinoma.

Abstract

618 Background: Previous post-surgical prognostic nomograms for clear cell RCC are based upon TNM stage, PS, Fuhrman grade, and sometimes presence of necrosis. 1943 patients (pts) with locally advanced renal cell carcinoma (RCC) participated in the ASSURE trial of one year of sunitinib daily for 4 wks of a 6 wk cycle, sorafenib daily, or placebo, then treated for up to 9 cycles. Of these, 1541 pts had clear cell predominant features. There were no significant differences in DFS (median 5.7 years) between either of the experimental arms and placebo, in the total or clear cell population. We investigated site-reported clinical and surgical factors associated with prolonged DFS. Methods: Patients were staged according to the AJCC 2007 version. Factors considered in addition to those in the final model shown below were age, race/ethnicity, histology, risk category and PS as stratified, and 10 other specific baseline symptoms. A Cox proportional hazards model was built including factors that significantly changed the -2 log likelihood ratio (Collett). Results: Shorter DFS was associated with higher age, stage and Fuhrman grade, male sex, PS > 0, open vs. laparoscopic surgical approach, low hemoglobin and diagnosis based on presence of symptoms. Conclusions: The association of stage, grade, age, and PS with outcome confirms previous analyses in resectable disease. The same factors which predict outcome in metastatic disease now clearly confer relapse risk in the postoperative setting. Clinical trial information: NCT00326898. [Table: see text]