Impact of recurrent copy number alterations and cancer gene mutations on the predictive accuracy of prognostic models in clear cell renal cell carcinoma

Purpose Several recently reported recurrent genomic alterations in clear cell renal cell carcinoma are linked to pathological and clinical outcomes. We determined whether any recurrent cancer gene mutations or copy number alterations identified in the TCGA (The Cancer Genome Atlas) clear cell renal cell carcinoma data set could add to the predictive accuracy of current prognostic models. Materials and Methods In 413 patients who underwent nephrectomy/partial nephrectomy we investigated whole exome, copy number array analyses and clinical variables. We identified 65 recurrent genomic alterations based on prevalence and combined them into 35 alterations, including 12 cancer gene mutations. Genomic markers were modeled using the elastic net algorithm with preoperative variables (tumor size plus patient age) and in the postoperative setting using the externally validated Mayo Clinic SSIGN (stage, size, grade and necrosis) prognostic scoring system. These models were subjected to internal validation using bootstrap. Results Median followup in survivors was 45 months. Several markers correlated with adverse cancer specific survival and time to recurrence on univariate analysis. However, most of them lost significance when controlling for tumor size with or without age in the preoperative models or for SSIGN score in the postoperative setting. Adding multiple genomic markers selected by the elastic net algorithm failed to substantially add to the predictive accuracy of any preoperative or postoperative model for cancer specific survival or time to recurrence. Conclusions While recurrent copy number alterations and cancer gene mutations are biologically significant, they do not appear to improve the predictive accuracy of existing models of clinical cancer specific survival or time to recurrence for clear cell renal cell carcinoma.