Cancer heterogeneity

Programme stream(s): Cancer discovery / underpinning research
Programme session type(s): Recent Advances session

Chair: Mariam Jamal-Hanjani, University College London Cancer Institute, UK
Speaker: Peter Van Loo, The Francis Crick Institute, UK
Speaker: Kevin Litchfield, The Francis Crick Institute, UK
Speaker: Trevor Graham, Barts Cancer Institute, UK


Room: Hall 1

This session should cover the background and add to the existing evidence for intratumour heterogeneity across different cancer types, in terms of tumour genomics and features of the tumour microenvironment, as well as the potential clinical implications of such heterogeneity in terms of therapeutic response and patient outcome. Deciphering the tumour genomic landscape relies on complex bioinformatics approaches, and so this session will also discuss the different approaches that can characterise and track tumour evolution dynamics using computational and integrative genomics.
Clinicians and scientists with an interest in tumour heterogeneity, evolutionary biology and bioinformatics analyses involving deep sequencing data should be drawn to this session.

Molecular archaeology of cancer
Speaker: Peter Van Loo
Affiliation: The Francis Crick Institute


The cancer genome contains within it an archaeological record of its past. We developed approaches to disentangle the evolutionary history of cancers through whole genome sequencing, collectively termed “molecular archaeology of cancer”. In a pan-cancer analysis of 2,658 whole-genome sequenced tumours, we are gaining insights into the extent of intra-tumour heterogeneity in different cancer types and the order and timing of acquisition of genetic aberrations. We find that characteristic copy number gains, such as trisomy 7 in glioblastoma or isochromosome 17q in medulloblastoma, are found amongst the earliest events in tumour evolution. The early phases of oncogenesis are driven by point mutations in a restricted set of cancer genes, often including biallelic inactivation of tumour suppressors. By contrast, increased genomic instability, a more than three-fold diversification of driver genes, and an acceleration of mutational processes are features of later stages. Using clock-like mutations, we obtain estimates for whole genome duplications and subclonal diversification in chronological time. Our results suggest that driver mutations often precede diagnosis by many years, and in some cases decades.Taken together, these data indicate that most cancers commonly have several predefined and ordered event trajectories, pivotal for understanding tumour biology and guiding early cancer detection.

Cancer Evolution: Insights from TRACERx Renal study
Speaker: Samra Turajlic
Affiliation: The Francis Crick Institute


The evolutionary features of clear-cell renal cell carcinoma (ccRCC) have not been systematically studied to date. We analysed 1206 primary tumour regions and 336 metastatic biopsies from 100 patients recruited into the multi-centre prospective study, TRACERx Renal. By resolving the patterns of driver event ordering, co-occurrence and mutual exclusivity at clone level, we show the deterministic nature of clonal evolution. ccRCC can be grouped into seven evolutionary subtypes, ranging from tumours characterised by early fixation of multiple mutational and copy number drivers, and rapid metastases; to highly branched tumours with >10 subclonal drivers and extensive parallel evolution, associated with attenuated metastases. We identify genetic diversity and chromosomal complexity as determinants of patient outcome. Metastatic competence was afforded by chromosome complexity and we identify 9p loss as a highly selected event driving metastasis and ccRCC related mortality (HR=7.7, p=0.0014). We observe early divergence of primitive ancestral clones and protracted latency of up to two decades as a feature of pancreatic metastases. Our insights reconcile the variable clinical behaviour of ccRCC, and suggest evolutionary potential as a biomarker for both intervention and surveillance.

Evolutionary dynamics of cancer subclones
Speaker: Trevor Graham
Affiliation: Barts Cancer Institute


The evolutionary dynamics of subclones during cancer growth remain largely undetermined as the longitudinal measurement of tumour subclones is impractical. I will describe how the underlying patterns of subclone evolution can be extracted by fitting multiscale mathematical models of cancer evolution to genomic data (n=238 tumours).  Our analysis reveals widespread effectively-neutral evolution across cancer types, due to the suppression of selective effects by tumour growth. When subclonal selection does occur, large subclones experience enormous selective advantages (>20%).  Overall, we demonstrate how population genetics theory, expressed using mathematical modelling, can be used to extract hidden temporal information about cancer evolutionary dynamics from the cancer genome.