Whole genome analysis to support cancer treatment decision making: The BC Cancer Agency Personalized Oncogenomics (POG) Project. Marco Marra, PhD, FRSC. Professor and Head, Medical GeneJcs, University of BriJsh Columbia. Director, Genome Sciences Centre, BC Cancer Agency.
Impact of cancer on Canada and Canadians • 2/5 Canadians will get cancer Canadian Cancer Society StaJsJcs 2015 2
Age is the major risk factor for cancers Canadian Cancer Society Sta;s;cs 2015 Current Biology 22, #17, p R741-R752 (2012) 3
Treatments are increasingly expensive
BC Cancer Agency • Provincial Mandate for Cancer Care and Research • Standard opera<ng procedures. • Radia<on therapy. • Chemotherapy. • 6 Regional Cancer Centres • Employees: 2,877 • Medical oncologists: 103 • Chemo. drug costs: $221M • Radia<on therapy: $57M • Es Est. ne t. new c w cases (2014): 25,170 ases (2014): 25,170 • Me Metas asta7 a7c >10,000 c >10,000 • Es Est. ne t. new c w cases (2028): 35,450 ases (2028): 35,450 5
Cancer is a gene;c disease • Mutations (“mistakes” in the genetic code) can cause cancers. • Inherited “predisposing” mutations • Sporadic “acquired” mutations • In the tumor DNA, not in the normal DNA • Environmental mutagens • DNA replication errors • …AATCGCGCTACCG… à …AATCGCGCT C CCG…
DNA can be thought of as the “hard drive” of the cell Nucleus Cytoplasm RNA PROTEIN DNA 7
Cancer muta;ons can inform treatment and prognosis bcr bcr Nature Medicine 15 , 1149 - 1152 (2009)
The first “draGs” of the human genome sequence NATURE |VOL 409 | 15 FEBRUARY 2001 16 FEBRUARY 2001 VOL 291 9
The evolution of DNA Sequencing 10
BCCA Genome Sciences Centre: Sequencing Capacity &Throughput • 335 staff including 13 senior scien<sts. • Total data generated to date: >1.4 petabases (14,000 human genomes) • Annual capacity 1.2 petabases • > 12,000 30X human genomes/year 2 secured data centres • Compute clusters 1: 800 nodes, 24,000 • hyper-threaded cores 16 – 48 GB RAM per node • High memory (1.5TB RAM) computers • >11 Petabytes on-line disk storage • 14,305 Human Genome Equivalents (30X)
Cancer genomes are complex • Base pairs (A-T, G-C) in one genome: ~3 billion • Genes: > 20,000 • Genes mutated / dysregulated in cancers: 1,000s 12
Tumors are communi;es of cells 100,000,000 cells / cm 3
Tumors are communi;es of cells 100,000,000 cells / cm 3
Observa;ons from cancer genome studies • The constellaJon of somaJc & gene expression alteraJons in individual cancer paJents cannot be predicted. These must be measured and their impact on pathways assessed. • Cancers can “evolve” to become treatment resistant. • • How can we align the right paJent to the right drug at the right Jme? • What are the geneJc properJes of treatment resistant disease?
“Medical oncology is an educated guessing game.” 16
DNA sequencing can be used to align pa;ents to treatments Sequence paJent DNA Drug A Drug D Drug B Drug C
“Panels” survey small numbers of genes / muta;ons 18
Hypothesis Comprehensive whole genome analysis of treatment resistant cancers can explain treatment resistance and reveal cryp7c therapeu7c vulnerabili7es. 19
The first case 20
Acknowledgements Medical Oncology Janessa Laskin Deepa Wadhwa Lawrence Lee Simon Chan Andy Mungall Abdul Al-Tourah Lyly Le Tamana Walia Hector Li Chang Carolyn Ch'ng Brad Nelson Helen Anderson Christopher Lee HuiLi Wong Pedro Farinha Eric Chuah Cydney Nielsen Vanessa Bernstein Ursula Lee Muhammad Zulfigar Malcolm Hayes Richard CorbeO Julie Nielsen Sylvie Bourque Howard Lim Ann Tan Tadaaki Hiruki An He Jacquie Schein Barbara Campling Jenny Ko Sara Taylor Hugo Horlings MarSn Jones Colin Schlosser Angela Chan ChrisSan Kollmannsberger Brian Thiessen David Huntsman Steven Jones Sohrab Shah Theresa Chan Caroline Lohrisch Anna Tinker Diana Ionescu Katayoon Kasaian Liz Starks Sylvia Cheng Nicol Macpherson Dorothy Uhlman Hoang Lien Ji-Young Kim Yongjun Zhao Winson Cheung Barb Melosky Medical GeneScs Nikita Makretsov Sreeja Leelakumari Social Science Kim Chi John Paul McGhie Linlea Armstrong Nissreen Mohammad Jake Lever Anita Charters Stephen Chia Corey Metcalf Ian Bosdet Greg Naus Yvonne Li Peter Chow-White Joseph Connors Deepu Mirchandani Gillian Mitchell Tony Ng William Long Dung Ha Janine Davies Nevin Murray Sean Young Torsten Nielsen Yussanne Ma Dean Regier Rebecca Deyell Sujaatha Narayanan Intan Schrader Tomo Osako Karen Mungall Deirdre Weymann Thuan Do Thao Nguyen Clinical Ethics Amir Rahemtulla Brandon Pierce Project Management/CoordinaSon Bernhard Eigl Conrad Oja Alice Virani David Schaeffer Erin Pleasance Leslie Alfaro Susan Ellard Gary Pansegrau Radiology Brandon Sheffield Cara Reisle Charlene Appleby Xiaolan Feng Maryse Power Francois Bernard Sona Sihra Yaoqing Shen Balvir Deol David Fenton Bradley Proctor Colin Mar Brian Skinnider Greg Taylor Nancy Ferguson Daygen Finch Sanjay Rao Montgomery MarSn Graham Slack Nina Thiessen Colleen Fitzgerald Paul Galbraith Rod Rassekh John Myo Peyman Tavassoli Tina Wong Cathy Fitzpatrick Karen Gelmon Daniel Renouf Pharmacy Basile Tessier-ClouSer Wei Zhang Alexandra Fok Alina Gerrie Paul Rogers Shirin Abadi Tom Thomson Eric Zhao Colleen Jantzen Sharlene Gill David Sanford Pathology Tracy Tucker Amir Zadeh Jas Kandola Karmjit Gill Delia Sauciuc Yazeed Alwalaie Emilija Todorovic Kelsey Zhu Julie LoreOe Anagha Gurjal Kerry Savage Daiana Becker-Santos Dirk van Niekerk Genome Science Katherine Mui Edward Hardy Ravinder Sawhney Ian Bosdet Suzanne Vercauteren Sam Aparicio Jessica Nelson Jason Hart Asif Shaikh Kathy Ceballos Carlos Vilamil ScoO Brown Robyn Roscoe Cheryl Ho Wen Wen Shan Andy Churg Joanne Wright Robin Coope Payal Sipahimalani Donna Hogge Tamara Shenkier Bakul Dalal Stephen Yip Peter Eirew June Song Paul Hoskins ChrisSne Simmons Christopher Dunham Chen Zhou Bruno Grande Isabel TrapagaAvancena Michael Humphreys Kevin Song John English BioinformaScs MarSn Hirst Peggy Tsang Bal Johal Caron Strahlendorf Patrice Eydoux Jianghong An Rob Holt Hagen Kennecke Sophie Sun Anthony Karnezis DusSn Bleile Christopher Hughes Kong Khoo Isabella Tai Aly Karsan Melika Bonakdar Richard Moore All pa7ents & their families Meg Knowling Joanna Vergidis Helga Klein-Parker Pinaki Bose Gregg Morin Doran Ksienski Diego Villa Anna Lee Morgan Bye Ryan Morin
POG process Informed Sample WGS / WTS Targeted consent acquisi7on sequencing alignment analysis Biopsy (Metasta7c site) Tumour (80x) Oncology consult & consent Pathology review Normal (40x) ‘In silico panel’ report Sample prepara7on RNA (200M) Genomic data genera7on Integra7ve analysis Tumour board discussion Clinical ac7on SNV, CNV, SV Genomic events of poten7al biological Review of genomic findings Expression and therapeu7c relevance (in context to Discussion of poten7al for Follow up consult & clinical decision Other analyses pa7ent disease) clinical ac7on 1-22
Comprehensive knowledgebase GOAL: Deliver comprehensive, high-quality genomic characteriza<on for clinical interpreta<on Leverages publically available BioFX GSC Knowledgebase resources where possible Sequencing Tumour & Tumour Alignment Transcriptome Normal transcriptome ReposiJoning bams fastqs fastq Splilng Events References 9400 Cancer ‘Events’ De novo Expression Alignment & SNV & Indel Gene & Exon Gene annota<ons (1108) structural Biological Merging calling expression variant calling Variants Targeted Gene muta<ons (6473) alignment Copy number variants Structural Merged (412) DiagnosJc Microbial Merged Transcriptome Tumour content & Variants normal bam indel file bam Structural Variants (1342) integraJon Curated Gene expression Expression Variants (65) Literature databasec Copy number PrognosJc 1093 References Variants Occurrence (42) Germline Expression SomaJc CNV SomaJc SNV Indel variant SV merging cohort Gene & LOH calling calling merging Biological (289) calling correlaJon TherapeuJc MutaJons Diagnos<c (28) Prognos<c (75) Germline Combined Combined Combined Therapeu<c (428) CNV & LOH SomaJc SNV Microbial Review structural indel expression summariy summary report summary summary summary Updated from research and outcome of POG case analyses Drug Target Analysis POG_AGBT_2017 23
Clinical ac;on 386 pa<ents sequenced and analyzed (23 pediatric) Erin Pleasance PhD. 79% acJonable 100% (n=307) 90% 80% Not evaluable 70% Progressive disease 36% not acJoned 26% wait & watch 38% treated 60% Mixed response (n=112) (n=76) (n=119) 50% Stable disease 40% ParJal response Complete response 30% 20% 10% 0% All Clinical Trial Standard Tx Off-label 52 % SOC 34% off-label 14% Trials (n=62) (n=40) (n=17) ReposiJoned to • >40% pa<ents treated demonstrated par<al more appropriate response (30 pts) or stable disease (22 pts) standard of care POG_AGBT_2017 24
Recommend
More recommend
