Abstract 434: 100,000 Genomes Project: Cancer program

Abstract

Abstract The 100,000 Genomes Project aims to improve cancer care for patients in the UK's National Health Service by applying personalized medicine paradigms based on whole-genome sequencing (WGS). The program has been designed to enable return of WGS-based results to clinicians in a clinically meaningful timeframe in order to enable diagnosis and treatment decisions for patient. Validation of WGS analysis results against standard-of-care tests: WGS analysis of tumors has a potential for replacing multiple standard-of-care tests, such as gene panels, genotyping arrays, FISH and karyotyping. For structural variants, WGS also provides much higher genomic resolution in comparison to FISH, the current standard of care. An important challenge when validating the analytical performance of WGS against existing technologies is to be able to systematically compare across a wide range of structural variant classes, low resolution events with WGS ascertained variants, which have often base pair resolution. We present here the results of successful validation for WGS analysis against targeted sequencing, genotyping array and FISH for a cohort of patients with chronic lymphocytic leukaemia (CLL). Quality control of sequencing data: To ensure high sensitivity and specificity of reported variants we developed a protocol for DNA quality control based on PCA analysis of six metrics derived from sequencing data: AT and CG dropout, unevenness of local genome coverage, percentage of mapped reads, percentage of chimeric DNA fragments and average fragment size. For each new sample, these six metrics are projected into the transformed space and outliers are determined by comparing with a cohort of 672 fresh frozen samples. Samples having a p below 10-4 (p: probability density after multivariate normal fitting) are manually reviewed. This procedure allowed us to quickly spot the sets of samples that were extracted with the violation of DNA-extraction protocol and to prevent polluting results with high rates of false positive variant calls. Reporting outcomes: Currently our bioinformatics analysis of WGS includes somatic small variants and somatic structural variants, germline pertinent findings, mutational signatures and mutational burden. Somatic small variants are linked to approved therapies, UK clinical trials and prognosis guidelines. We are working on implementing recently derived genomic signatures for predicting therapeutic response for specific tumor types. We are presenting here an overview of clinical utility for our reported outcomes. Bioinformatics reports for WGS with links to potentially relevant therapies and clinical trials have been already delivered for hundreds of cancer patients in the UK. We are collecting feedback from clinicians and working on improving our analyses to provide better benefits for cancer patients. Citation Format: Alona Sosinsky, Pavlos Antoniou, John Ambrose, Martina Mijuskovic, Antonio Rueda-Martin, Daniel Perez-Gil1, Enric Serra, Shirley Henderson, Angela Hamblin, Anna Schuh, Nirupa Murugaesu, Clare Turnbull, Augusto Rendon, Mark Caulfield. 100,000 Genomes Project: Cancer program [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 434.