Abstract PR008: Timing complex copy number gains in whole genome duplicated tumors

Abstract

Abstract The accumulation of genetic copy number aberrations is a process common to the development of many cancers. Tumors that have undergone a whole genome duplication (WGD) are known to have high degrees of genomic instability and often contain genomic regions that have undergone a series of copy number gains resulting in multiple copies of both alleles. Metastases are known to have an even greater rate of WGDs and general genomic instability. WGDs provide a powerful handle to time genetic events, such as single nucleotide variants or copy-number aberrations, in tumor evolution. However, the relative timing of copy number gains in whole genome duplicated tumors is complicated by the fact that there are multiple plausible route histories that could give rise to them. Typically, the most parsimonious event history is assumed to take place but this has never been validated. Here we describe a method, GRITIC, that overcomes this problem by evaluating all possible histories that could result in a given copy number state and determining the timing of the most likely series of gains. GRITIC allows the gains leading to any copy number state to be inferred. On a representative simulated cohort, it accurately measures the timing of complex copy number gains. By applying GRITIC to 6,010 primary and metastatic tumor samples from the Pan-Cancer Analysis of Whole Genomes and Hartwig Medical Foundation datasets, we find that the principle of maximum parsimony is violated in at least 25% of all copy number gains in whole genome duplicated tumors, with gains occurring both much earlier and later than thought under this assumption. We also find evidence for punctuated bursts of gains in WGD tumors, independent of the duplication itself. Finally, GRITIC allows us to look at the influence of WGD on copy-number events by comparing pre- and post-WGD copy number aberration landscapes. We find that the frequency with which different chromosome arms are gained before a WGD is highly correlated with the frequency that they are gained post-WGD. In contrast, the correlation between pre- and post-WGD losses is much smaller. We suggest that this indicates that a genome duplication changes the evolutionary trajectories of copy number losses but not gains. As an exception to this trend, chromosome 8 was disproportionately likely to be gained pre-WGD in metastatic prostate cancer as well as 1q in both breast and liver cancers. Altogether, GRITIC allows for a more accurate inference of evolutionary histories in different cancer types and better insights into the early genomic events that occur in tumors that underwent a WGD. Citation Format: Toby M. Baker, Siqi Lai, Stefan Dentro, Maxime Tarabichi, Peter Van Loo. Timing complex copy number gains in whole genome duplicated tumors [abstract]. In: Proceedings of the AACR Special Conference: Cancer Metastasis; 2022 Nov 14-17; Portland, OR. Philadelphia (PA): AACR; Cancer Res 2022;83(2 Suppl_2):Abstract nr PR008.