Abstract 2373: Assessing intra-tumor heterogeneity and tracking longitudinal and spatial clonal evolution by next-generation sequencing

Abstract Cancer is a disease driven by genetic and epigenetic alterations that follows Darwinian evolution. Recently, there have been increasing efforts to sequence the tumor from the same patient at multiple time points and/or from multiple spatially separated resections. Different snapshots of the same tumor have proved invaluable for identifying subclonal populations and inferring the tumor's history. We propose a method, Canopy, for estimating the clonal history and for longitudinal and spatial comparison of mutation profiles from one or more samples derived from a single patient. Canopy accounts for normal cell contamination and reconstructs subclonal phylogeny utilizing both somatic copy number alterations (CNAs) and somatic single nucleotide alterations (SNAs). Canopy provides a general mathematical framework that enumerates all possible CNA-SNA phases and temporal orderings. Taking as input the mutant and reference allele frequencies for SNAs, and depth of coverage for CNAs, Canopy gives confidence assessments of all possible configurations of the cancer's clonal evolution. Canopy is applied to three cancer sequencing datasets of varying study design, as well as to an extensive simulation study. On a whole-exome study of a transplantable metastasis model derived from human breast cancer cell line MDA-MB-231, Canopy successfully deconvolutes the mixed cell sublines, using the single cell sublines as ground truth, and identifies DNA signatures that can be prognostic of distant metastasis. On a whole-genome sequencing dataset of the primary tumor and relapse genome of a leukemia patient, Canopy predicts phylogenetic histories in concordance with existing knowledge. On a whole-genome sequencing dataset of the breast cancer tumor and its subsequent metastatic xenograft, Canopy's inferred clonal phylogeny is concordant with genomic markers of major clonal genotype and is confirmed by single-cell sequencing. Finally, through simulations, we explore the effects of various parameters on deconvolution accuracy, and evaluate performance with comparison against existing methods. Collectively, Canopy provides a rigorous foundation for statistical inference on repeated sequencing experiments from evolving populations delineated temporally and spatially. Citation Format: Yuchao Jiang, Yu Qiu, Andy J. Minn, Nancy R. Zhang. Assessing intra-tumor heterogeneity and tracking longitudinal and spatial clonal evolution by next-generation sequencing. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2373.