Conflict of interest Cell context, mutation and common I am a - PowerPoint PPT Presentation

5/25/2018 Conflict of interest Cell context, mutation and common • I am a founder and chief medical officer of carcinomas of the uterus and ovary Contextual Genomics a company offering cancer genomics testing solutions • I will not be describing the work of Contextual Genomics and will indicate if tools used from Contextual Genomics were used to obtain any of the data David G. Huntsman presented BC Cancer Agency Vancouver General Hospital University of British Columbia Outline AGCT MUC ca LGSCa Met Ca • Genomics and histology across common ovarian and uterine cancers • Similarities and differences between histologically similar cancers of uterus and SCCOHT OCCC HGSCa ENOCa ovary • Precursors and prevention Each of these cancers reflects the interactions between specific cells of origin, mutations and • Cell context + mutation + permissive microenvironments microenvironment = cancer 1

5/25/2018 Hiking 1350 miles SE from But if Frodo started in Topeka Kansas he Hobbiton gets you to Mordor would have ended up in Jacksonville Assessing the problem “ with few exceptions, the clinical presentation, SEER treatment, and results of treatment are similar for all (cell) types of tumor ” Victoria • CDM Fletcher ed. In Diagnostic Histopathology of Tumors (2007) BC Vaughan et al Nature Reviews Cancer 2012 2

5/25/2018 Changing Paradigms for Cancer Ovarian carcinoma sub-typing 2000 Control Undifferentiated Endometrioid Generic Stratified Individualized Transitional Mucinous Clear Cell Cancer Cancer Cancer Serous Control Control Control 1990 2010 2020 LMP Breast cancer GR 1 Lymphoma GR 2 Ovarian cancer Pancreatic cancer GR 3 New classification: freq Impact of histotype changes • 300 cases centrally reviewed in 2002 • Reviewed again by same pathologist using 2014 WHO criteria : 54% concordance • New histotypes showed 98% concordance with second reviewer and stronger associations with outcome and biomarkers <<< = These cancers can be classified with 95% accuracy by microscope using WHO 2014 diagnostic criteria F Kommoss 2002 << F Kommoss 2014 = B Gilks 2014 Genomics offers a substrata of information to aid treatment decisions 3

5/25/2018 Ovarian cancer outcomes from How to define a distinct disease? 2000 Canadian patients • Different etiology • Different precursor lesions • Different molecular characteristics • Different presentation • Different clinical course • Different biomarker profiles • Different response to therapy All subtypes treated with platinum/taxane based therapy On all counts HGS, mucinous, LGS and Primary resistance in non HGSCa is primarily due to cell endometrioid/CCC are distinct diseases context and how cell context determines oncogenic opportunity Le Page et al BMC Cancer 2018 Changing Paradigms for Cancer Endometrial Cancer Control • 4 th most common Generic Stratified Individualized cancer in women Cancer Cancer Cancer Control Control Control 1990 2010 2020 Breast cancer Lymphoma Ovarian cancer Pancreatic cancer 4

5/25/2018 Endometrial cancer TCGA-2013 This is very different than in ovarian cancer TCGA Classifier for Endometrial Cancers reduced to practice Prognostic Ability of ProMisE Overall Survival Disease Specific Survival Progression Free Survival ProMisE remains a significant prognostic marker for all events even after we account for other known risk factors available at diagnosis: age, bmi, histology, Simple technologies, reproducible grade, and treatment. Talhouk, McAlpine Cancer , 2017 Talhouk, McAlpine Cancer , 2017 Kommoss, McAlpine, Talhouk in press Kommoss, McAlpine, Talhouk in press 5

5/25/2018 High Grade Serous Carcinomas Genomic Stratification of cancers B • In ovarian cancer genomic data offers a A 1 2 3 4 5 1 2 3 4 5 substra of information to refine our 6 7 8 9 10 11 6 7 8 9 10 11 understanding of subtypes 12 13 14 15 16 17 18 12 13 14 15 16 17 18 19 20 21 22 X 19 20 21 22 X • In endometrial cancer genomic data will C D 1 2 3 4 5 1 2 3 4 5 replace histotyping 6 7 8 9 10 11 12 6 7 8 9 10 11 13 14 15 16 17 18 • In brain and lymphoid cancer integration 12 13 14 15 16 17 18 • Chromosomal instability 19 20 21 22 X 19 20 21 22 X has occurred • P53 mutations (ubiquitous) E F 1 2 3 4 5 1 2 3 4 5 • BRCA or BRCA associated 6 7 8 9 10 11 • In breast cancer molecular and histologic 6 7 8 9 10 11 12 13 14 15 16 17 18 12 13 14 15 16 17 18 analysis are parallel tracks with no 19 20 21 22 X 19 20 21 22 X Ahmed et al J path 2010; Press et al BMC Cancer 2008 integration Reproducible expression-based subtypes of HGSC but clinical Not prognostic with current therapy; may predict response to future therapies (antiangiogenics) relevance is unknown Tumour/gene groups TCGA Tothill et al. 489 tumours 245 tumours Differentiated Immunoreactive Mesenchymal Proliferative 1000 genes Gene expression Low High TCGA = The Cancer Genome Atlas. Ribeiro et al. 2014 Frontiers in Oncology The Cancer Genome Atlas Research Network. Nature 2011;474:609–15; Tothill RW, et al. Clin Cancer Res 2008;14:5198–208. 6

5/25/2018 Targeting Cyclin E1 in high grade serous cancer <10% of cases have highly level amplification This is mutually exclusive with BRCA mutation David Bowtell Peter MacCallum Cancer Centre Melbourne Imperial College London Nature 2011 d.bowtell@petermac.org Targeting the 500lb gorilla BRCA2 BRCA1 Germline Germline 6% 8% BRCA1 Somatic BRCA2 3% Somatic Other 3% 34% BRCA1 Methylation 11% EMSY Amplification MMR 6% Germline PTEN Loss 2% CCNE1 5% Other HRD Amplification 7% 15% Genetic Deficits in Homologous Recombination are found in 50% of Serous Ovarian Cancer: multiple PARP inhibitor trials ongoing – this is truly exciting Patch et al, Nature 2015 7

5/25/2018 PARP inhibition with Olaparib as maintenance therapy for HGSOC e Ledermann et al NEJM April 2014 Iglehart NEJM 2009 How best to identify potential PARP responders and non responders Final efficacy analysis Ariel2: PFS in BRCA mut and LOH-high versus LOH-low patients Platinum response 60–65% Somatic and germline BRCA1 and TCGA Analysis 51% 2 mutations c.20%  10000 1000 CA125 100 10 0 50 100 150 200 250 Time (Days) TCGA, Nature (2011) 474:609 RAD51 assay 50% Mukhopadhyay, Cancer Res (2012) 72:5675 Swisher et al Lancet Oncology 2017 Courtesy of Iain McNeish 8

5/25/2018 Given the paucity of specific features Final efficacy analysis NOVA: PFS in gBRCA mut and HRD positive versus HRD negative patients beyond tp53 mutation and BRCA: Can the genome be the biomarker? gBRCA WT HRD gBRCA WT HRD gBRCA mut 100 negative 100 100 positive Niraparib Niraparib Niraparib Progression–free Survival (%) Placebo Progression–free Survival (%) Placebo Progression–free Survival (%) Placebo • Higher level view of cancer genome 75 75 75 50 enables identification of signatures that 50 50 25 25 25 point to mutational process 0 0 2 4 6 8 10 12 14 16 18 20 22 24 0 0 Time Since Randomization (months) 0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24 Time Since Randomization (months) Time Since Randomization (months) HR= 0.27 HR= 0.38 HR= 0.58 Mirza et al NEJM (2016) ePub 8 th Oct 2016 vs We still can not adequately predict lack of response to PARP inhibition Mutational signatures of ovarian Signatures as well as specific mutations track with cancer types cancer histotypes Signatures of mutational processes in human cancers: Alexandrov et al Nature 2014 Yikan Wang et al Nature Genetics 2017 9