Abstract

Background

Large-scale sequencing studies have revealed the molecular landscape of clear cell renal cell carcinoma (ccRCC), highlighting frequent inactivation of VHL tumour suppressor gene, alterations in the SWI/SNF complex, histone-modifying genes and the PI3K/AKT pathway, coupled with recurrent somatic copy number alterations (SCNAs). We previously observed a significant degree of mutational and SCNA intratumour heterogeneity (ITH) in ten cases of advanced ccRCC. However, an in-depth exploration of evolutionary trajectories across all stages and clinical phenotypes in ccRCC has not been performed to date.

Method

Cases were recruited under a multi-centre prospective cohort study TRAcking renal cell Cancer Evolution through Therapy (TRACERx) Renal. Target population are patients undergoing nephrectomy for ccRCC. Primary surgical specimens, as well as material from synchronous and/or metachranous metastases were subject to multi-region sampling and molecular profiling. We used a targeted panel which captures mutations in known driver genes, and SCNAs across the genome. Individual region mutational data were integrated to infer the evolutionary history of each tumour distinguishing clonal drivers from subclonal drivers.

Results

We analysed 1209 primary tumour regions from 101 patients and 169 metastatic regions from from a subset of 38 patients with metastatic disease. Analysis of driver event co-occurrence, mutual exclusivity and timing revealed deterministic patterns of ccRCC evolution, which can be grouped into seven distinct evolutionary subtypes. These subtypes ranged from mono-clonal tumours achieving high selective fitness immediately within the truncal clone, to highly branched tumours with >10 driver subclones and extensive parallel evolution. Profiling of tumour subclones achieving metastatic dissemination identified strong enrichment for SCNA events including loss of 9p21.3 and 14q31.1. 

Conclusion

Collectively these observations show deterministic patterns in the course of ccRCC evolution, with acquisition of SCNA events including loss of 9p21.3 and 14q31.1 critical in driving metastasis.