Objective: The aim of this study was to externally validate in an up-to-date setting the predictive ability of the model for recurrence after radical treatment of clear cell renal cell carcinoma (CCRCC) published by Leibovich in 2003.
Materials and methods: The study included a total of 386 consecutive patients with CCRCC between January 1997 and May 2013, treated with partial or radical nephrectomy. All patients were scored with points between 0 and 11, and further subdivided into low-, intermediate- and high-risk groups according to the original paper. Well-recognized statistical methods for the evaluation of Cox's proportional hazard-based prognostic models were applied. To validate the discriminative ability, Harrell's concordance (c) index and hazard ratios (HRs) between risk groups were used, and calibration was graphically explored.
Results: The 10 year recurrence-free survival rates for the low-, intermediate- and high-risk groups were 87.3%, 63.8% and 19.8%, respectively Harrell's c index was 0.864. The HRs across risk groups for the intermediate- and high-risk groups were 5.29 and 21.56, respectively, with the low-risk group as a reference category. A gross comparison of the survival estimates between the patients showed an overall similarity. However, differences within the intermediate- and high-risk groups were seen in the first year of follow-up.
Conclusions: The Leibovich model seems to discriminate well between risk groups, but for the intermediate- and high-risk groups the calibration is not optimal. This study validates the model in a present-day Nordic patient population. The model can be used as a risk stratification tool for follow-up after radical treatment of CCRCC.
Keywords: Clear cell renal cell carcinoma; external validation; follow-up; integrated staging systems; prognostic risk groups; recurrence.