SIU 2017: Molecular Predictors of RCC

He began by reviewing a clinical stratification tool his group developed using clinical variables alone to help predict progression after radical nephrectomy for patients with ccRCC – it was demonstrated to have good ability to discriminate metastases free survival. Besides his groups work, there are other clinical risk predictors already well validated.

However, with the growing number of molecular markers, the question was if there was any added benefit? To that effect, his group developed the BioScore – a biomarker panel using B7-H1 (PDL-1), surviving and Ki-67. While independently predictive of cancer-specific survival, when added to the UISS or SSIGN clinical risk models, it improved their c-index modestly. However, Dr. Leibovich was up front in saying it was not necessarily a practical addition for the amount of benefit it provided. Similarly, other molecular markers have modest c-indices that have not been shown to have added benefit to already existent clinical risk predictors.

The same story exists in mRCC setting. He reiterates what Dr. Uzzo highlighted that the clinical risk prediction tools in mRCC have lower c-indices that their localized RCC counterparts, likely due to the greater tumor heterogeneity. In a recent paper (Eur Urol 2016, Serie et al), Dr. Leibovich highlighted the identification of two clear cell RCC molecular subtypes – ccA and ccB. Interestingly, patients with ccA molecular subtype in the primary tumor could have either subtype in the metastases; patients with ccB subtype in the primary tumor often had just ccB subtype in the metastases.

They also went on to do subsequent work in mRCC and were able to identify a 5-gene panel prognostic for cancer-specific survival in a mRCC cohort. It was prognostic in both the primary tumor and in the metastases. More importantly though it was found to be additive to the clinical model. While further work is needed in this area, this is the type of work that will hopefully help move the field forward.

Next steps:

1) Genetic markers need to be validated in larger cohorts of cases

• Primary tumor
• Metastatic lesions