Webinar Series. Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels. Participating experts

Transcription

1 Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels April 13, 2016 Webinar Series Brought to you by the Science/ AAAS Custom Publishing Office Participating experts P. Mickey Williams, Ph.D. Frederick National Lab for Cancer Research Frederick, MD Sponsored by Jeremy Segal, M.D., Ph.D. University of Chicago School of Medicine Chicago, IL

2 Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels April 13, 2016 Webinar Series Brought to you by the Science/ AAAS Custom Publishing Office Instructions for viewers To share webinar via social media: To share webinar via e mail: To see speaker biographies, click: View Bio under speaker name Sponsored by To ask a question, click the Ask A Question button under the slide window

3 Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels April 13, 2016 Webinar Series Brought to you by the Science/ AAAS Custom Publishing Office Participating experts P. Mickey Williams, Ph.D. Frederick National Lab for Cancer Research Frederick, MD Sponsored by Jeremy Segal, M.D., Ph.D. University of Chicago School of Medicine Chicago, IL

4 Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels P. Mickey Williams, Ph.D. Director of the Molecular Characterization Laboratory Frederick National Laboratory for Cancer Research

5 Topics Covered Basic assay development Defining the assay system and it s intended use Developing targeted NGS assays Special considerations for formalin fixed tumor tissues Analytical performance testing Frederick National Laboratory for Cancer Research

6 Defining the Assay Intended Use Assays applications: 1. Pure research 2. Clinical a) Integrated (all patients will be tested but assay results are not used to enroll or select treatment arms for a clinical study) b) Integral (all patients tested and assay results used to enroll or select treatment for a clinical study) The assay application will determine the rigor which the assay must be undergo during development Will the assay test blood, tissue blocks or biopsies Best to use the intended sample type during assay development Frederick National Laboratory for Cancer Research

7 Defining the Assay System Consider all aspects of the assay that are critical for obtaining consistent results Specimens: Formalin fixed, therefore nucleic acids are fragmented (this generally does not interfere with probe capture or PCR based library preparations, but may add noise to sequencing results due to base damage which occurs as a result of fixation) Are specimens recently collected and use neutral buffered formalin, or are the older possibly with acidic formalin? Will tumor content be enriched or will specimens be used without enrichment (this will impact detection sensitivity)? What potential clinical details of specimens and how may this impact the results (surgical resections versus biopsies, etc.) Frederick National Laboratory for Cancer Research

8 Enriching for Tumor Content Macro or Micro Dissection Acinic salivary gland tumor, 100% Hepatocellular cancer, 50% Leiomyosarcoma, 75% Renal cell carcinoma, 100% Mesothelioma, 50% Colorectal cancer, 70% Frederick National Laboratory for Cancer Research

9 Defining the Assay System (cont.) Determine the method of nucleic extraction RNA and DNA versus DNA only Define and select the NGS method Library preparation methods Sequencing read depth Impacts detection sensitivity Determine the data analysis methods Select specific parameters (minimal read depth for variant reporting) What variants will be reported? Is this discovery effort where all variants are reported or clinical application (focus on variants of clinical significance) Will germline data be used for reporting variants Helps when identifying all somatic mutations are the goal Debatable if needed for clinical applications where only clinically relevant variants are reported Will variant fraction reporting cut-points be used? If assay will be used for clinical application, will a tumor board or rules based variant annotation occur? Frederick National Laboratory for Cancer Research

10 NCI-MATCH Assay System & Work Flow Biopsy Shipped to MDACC Tissue Accession Tissue Processing/Enrichment PTEN IHC NA Extraction NA Shipped Library Prep and Sequencing Archive Tissue Blocks Slides Nucleic Acid BAM File Storage MDACC MGH Yale MoCha MOI Annotation Ion Reporter Review and Sign off MATCHbox applies rules based variant annotation/treatment selection Final Report Clinical DB

11 Frederick National Laboratory for Cancer Research

12 Age Bins of Archived Ovarian Tumor FFPET Frederick National Laboratory for Cancer Research

13 QC Metrics for WES of Archived Ovarian FFPET Frederick National Laboratory for Cancer Research

14 Conclusions of Paper Sufficient quality data was retrieved from all FFPE specimens regardless of repository site or age There was a trend poorer quality sequence and increasing age of specimen In general archived FFPET provide a resource for biological research Frederick National Laboratory for Cancer Research

15 NGS Assays and Oncology NGS provides a powerful tool for massively parallel sequencing of patient tumors NGS assays are being developed, applied and acted upon for patient management in oncology Everybody is doing it There is a need for: Assay Standards (e.g., Genome in a Bottle) Guidance on clinical relevance of detected variants, levels of evidence Minimal data reporting standards such that others can understand assay bias and repeat data Public data sharing with assay details and clinical outcome Continued efforts to demonstrate clinical utility of sequencing panels Frederick National Laboratory for Cancer Research

16 Tumor Specimen: The Power of NGS Comes Along with Complexities Is the specimen archival pre-treatment or recent post treatment (resistance mutations) How does the specimen reflects whole tumor mass; i.e., tumor heterogeneity How much viable tumor is represented in sample used for sequencing (20% versus 80%) Sequencing choices: Read depth (10X versus 200X) Lower limit of detection i.e., allele frequency reported All of the above impact result interpretation Frederick National Laboratory for Cancer Research

17 NCI-MATCH Trial MATCH Trial: A national study using NGS and IHC assays to select a matched treatment Variants used for treatment selection are based on levels of evidence We attempted to follow a quality system approach for development and analytical validation of our NGS assays Frederick National Laboratory for Cancer Research

18 AKT (1 10%) mtor (5%) PIK3CA (17 18%) TSC1 or TSC2 (2.6 and 3.5%) BRAF (2.79) BRAF V600E or V600K (1 12%) GNA11 (1.6%) GNAQ (2%) NF1 (7.7%) DDR2 (2%) ALK (<2%) ckit (2%) EGFR (1 4%) EGFR T790M (1 2%) FGFR (5%) HER2 amplification (5%) HER2 mutation (2 5%) MET (4%) NF2 (2%) PTEN loss (11%) PTEN mutation or deletion (11%) ROS1 (<2%) SMO or PTCH1 (<2%) amois in NCI-MATCH and Estimated Prevalence (TCGA, c-bioportal, My Cancer Genome, MDACC, ETC) 0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0% 20.0% Frederick National Laboratory for Cancer Research

19 Each Targeted Rx Becomes a Basket Trial and Many Basket Trials are Combined Under MATCH Umbrella AKT (1 10%) mtor (5%) PIK3CA (17 18%) TSC1 or TSC2 (2.6 and 3.5%) BRAF (2.79) BRAF V600E or V600K (1 12%) GNA11 (1.6%) GNAQ (2%) NF1 (7.7%) DDR2 (2%) ALK (<2%) ckit (2%) EGFR (1 4%) EGFR T790M (1 2%) FGFR (5%) HER2 amplification (5%) HER2 mutation (2 5%) MET (4%) NF2 (2%) PTEN loss (11%) PTEN mutation or deletion (11%) ROS1 (<2%) SMO or PTCH1 (<2%) 0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0% 20.0% Frederick National Laboratory for Cancer Research

20 Selection of NGS Platform and Laboratory Network NGS platform chosen after evaluation of RFI: Ion Torrent PGM and the Oncomine Cancer Research Panel 143 genes & 4066 annotated variants SNV, indel, CNV, targeted translocations Competitively chosen lab sites: MDACC (Hamilton) MGH (Iafrate) Yale (Sklar) FNLCR - MoCha (Williams) Frederick National Laboratory for Cancer Research

21 MATCH Assay Oncomine Cancer Panel Gene List Gene SNV/Indel CNV Fusion Hotspot genes, n=73 (hotspot coverage) CDS, n=26 (full gene) Copy gain, n=49 Fusion drivers, n=22 (183 assays) ABL1 AKT1 ALK AR ARAF BRAF BTK CBL CDK4 CHEK2 CSF1R CTNNB1 DDR2 DNMT3A EGFR ERBB2 ERBB3 ERBB4 ESR1 EZH2 FGFR1 FGFR2 FGFR3 FLT3 FOXL2 GATA2 GNA11 GNAQ GNAS HNF1A HRAS IDH1 IDH2 IFITM1 IFITM3 JAK1 JAK2 JAK3 KDR KIT KNSTRN KRAS MAGOH MAP2K1 MAP2K2 MAPK1 MAX MED12 MET MLH1 MPL MTOR MYD88 NFE2L2 NPM1 NRAS PAX5 PDGFRA PIK3CA PPP2R1A PTPN11 RAC1 RAF1 RET RHEB RHOA SF3B1 SMO SPOP SRC STAT3 U2AF1 XPO1 APC ATM BAP1 BRCA1 BRCA2 CDH1 CDKN2A FBXW7 GATA3 MSH2 NF1 NF2 NOTCH1 PIK3R1 PTCH1 PTEN RB1 SMAD4 SMARCB1 STK11 TET2 TP53 TSC1 TSC2 VHL WT1 ACVRL1 AKT1 APEX1 AR ATP11B BCL2L1 BCL9 BIRC2 BIRC3 CCND1 CCNE1 CD274 CD44 CDK4 CDK6 CSNK2A1 DCUN1D1 EGFR ERBB2 FGFR1 FGFR2 FGFR3 FGFR4 FLT3 GAS6 IGF1R IL6 KIT KRAS MCL1 MDM2 MDM4 MET MYC MYCL MYCN MYO18A NKX2-1 NKX2-8 PDCD1LG2 PDGFRA PIK3CA PNP PPARG RPS6KB1 SOX2 TERT TIAF1 ZNF217 ALK RET ROS1 NTRK1 ABL1 AKT3 AXL BRAF CDK4 EGFR ERBB2 ERG ETV1 ETV4 ETV5 FGFR1 FGFR2 FGFR3 NTRK3 PDGFRA PPARG RAF1 Courtesy of Thermo Fisher 143 Genes Frederick National Laboratory for Cancer Research

22 Rules of Evidence for Actionable Variants Used for Treatment Selection Frederick National Laboratory for Cancer Research

23 Feasibility Testing Non harmonized SOPs used by each lab IR v 4.2 used for data analysis 44 FFPE clinical samples tested within 4 laboratories 10 Cancer cell line genomes x4 labs 3 Hapmap genomes x3 replicates x 4 labs Frederick National Laboratory for Cancer Research

24 Feasibility Data Reproducibility Variant allele frequencies are very close across four lab replicates Frederick National Laboratory for Cancer Research Not detected

25 70 60 Variant Distribution in Sensitivity Study Total Variant Number in each variant type MDACC MGH NCI Yale 0 SNV Indel Large Indel CNV Fusion Total 149 Unique Variants Fusion, 13 CNV, 15 Indel, 26 Large indel, 19 SNV, total variants 149 unique variants Frederick National Laboratory for Cancer Research

26 Multiple Tissues Tested for Performance Stomach, 1 Thyroid, 2 Bladder, 8 Soft Tissue, 3 Blood, 1 Pancreas, 2 Source unknown, 2 Skin, 10 Bone, 2 Ovary, 4 Brain, 38 Breast, 8 Lung, 69 Colon, 13 Liver, 2 Head and Neck, 2 Esophagus, 3 GI Tract, Unique clinical specimen samples from 15 tissue types in in sensitivity study Frederick National Laboratory for Cancer Research

27 Resources for NCI-MATCH Main Webpages: cancer.gov/nci-match ecog-acrin.org/nci-matcheay131 Protocol Documents: ctsu.org (password required) Spanish: cancer.gov/espanol/nci-match Inquiries: Patient Brochure: EA website (above) Site Process Brochure: EA website (above) NCI s Cancer Information Service: CANCER and cancer.gov/contact This slide presentation is updated regularly. For the latest version, visit ecogacrin.org. Frederick National Laboratory for Cancer Research

28 MATCH Acknowledgements FNLCR Jason Lih David Sims Robin Harrington Kneshay Harper Patty Runge Vivekananda Datta JoyAnn Phillips Rohan Courtney Bouk Thermo Fisher Leslie Evans Jingwei Ni Peter Wyngaard Seth Sadis Jeff Smith Oncomine team MDACC Rajesh Singh Johnny Yao Raja Luthra Mark Routbort Geeta Mantha Stanley Hamilton Yale Kayn Ronski Sandra Canosa Jeff Sklar MGH Hayley Robinson Amelia Raymond John Iafrate NCI CBIIT Brent Coffey Mary Anderson Frank Spina David Patton NCI DCTD Barbara Conley Eric Polley Lisa McShane Larry Rubenstein Erin Souhan Sanita Bharti Rita Misra Alice Chen Jeff Abrams Jim Doroshow ECOG-ACRIN Keith Flaherty Peter O Dwyer Bob Comis Shuli Li Bob Gray Kamalia Sazali Jeff Zhang Donna Marinucci AND MANY OTHERS Frederick National Laboratory for Cancer Research 28

29 Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels April 13, 2016 Webinar Series Brought to you by the Science/ AAAS Custom Publishing Office Participating experts P. Mickey Williams, Ph.D. Frederick National Lab for Cancer Research Frederick, MD Sponsored by Jeremy Segal, M.D., Ph.D. University of Chicago School of Medicine Chicago, IL

30 Jeremy P. Segal, MD, PhD Co-Director, Clinical Genomics Laboratory Division of Genomic and Molecular Pathology University of Chicago NGS Cancer Profiling: Options, Best Practices and Pitfalls

31 Indications for Clinical Cancer Genomics Inborn genetics: Genetic disease Risk factors Disease Genetics: Early screening Disease Genetics: Diagnosis Prognosis Therapy Disease Genetics: Residual disease testing Resistance mutation surveillance

32 Cancer Genomics Targets Mutations Point mutations Insertions and deletions (indels) Structural Variations Large scale deletions/duplications Fusions/Rearrangements Aneuploidy Chromothripsis RNA Analysis Altered expression Pathway activation MicroRNAs LncRNAs Alternative Splicing Allele-specific expression RNA binding protein interactions Epigenetics Altered DNA methylation Altered histone methylation Altered DNA-protein interactions Altered chromatin structure

33 Cancer Genomics Targets Mutations Point mutations Insertions and deletions (indels) Structural Variations Large scale deletions/duplications Fusions/Rearrangements Aneuploidy Chromothripsis RNA Analysis Altered expression Pathway activation MicroRNAs LncRNAs Alternative Splicing Allele-specific expression RNA binding protein interactions Epigenetics Altered DNA methylation Altered histone methylation Altered DNA-protein interactions Altered chromatin structure NGS

34 Clinical Example: Lung Cancer Demanding Clinicians! This is now too many tests to do in the traditional manner. NGS advantage: cover all these analytes with a single test process. Point Mutations EGFR L858R, T790M, KRAS, PIK3CA BRAF, etc. Small Deletions EGFR exon 19 Large Deletions MET exon 14 deletion Copy Number Alterations MET amplification EGFR amplification Lung Adenocarcinoma Gene Fusions ALK (e.g. EML4-ALK) RET ROS1 NTRK1

35 What s the right assay? Clinical Genomics Laboratory Smaller Targeted Assays Some clinicians Cancer specimens Validation effort Cost Reimbursement Larger Comprehensive Assays Most clinicians Clinical requirements Translational research Lab Competition Technology

36 Depth: Key Consideration #1 COVERAGE DEPTH COVERAGE x DEPTH = Sequencing $$

37 Cancer Low % Mutations Tumor cell percentage Tumor heterogeneity Low mutation allelic percentage Yates and Campbell, Nat. Rev. Genetics

38 Increased Depth Improves Mutation Detection Estimated Sensitivity Sequencing Depth

39 Sampling: Key Consideration #2 GOOD Sampling Library Prep & Sequencing If QC is adequate, this is a believable/reproducible result.

40 But.don t be fooled! BAD Sampling any amplification error becomes dangerous Library Prep & Sequencing There is no cure for a badly controlled wet-lab assay.

41 Variable Specimens, Total DNA Loading Multiplex PCR Product

42 QPCR Assessment of Effective/Amplifiable DNA

43 Variable Specimens, Amplifiable DNA Loading Multiplex PCR Product

44 Amplicon vs. Hybrid Capture Template Probe Design modified from Nature Reviews Genetics 2011; 12:

45 Simplified Assay Type Comparisons Amplicon Systems Hybrid Capture Low DNA Input Broad Coverage Small/Medium Indels (<100 bp) Copy Number Alterations Structural Alterations

46 Amplicon Assays for Minute Specimens 50 gene amplicon panel ~5 ng FFPE DNA, 10-15% tumor cells: EGFR mutation negative KRAS c.34g>t, p.g12c (NM_033360) 5% MAF TP53 c.818g>t, p.r273l (NM_000546) 5% MAF 1 mm

47 FNAs: 1 ng DNA from Diffquik Smear KRAS G13C

48 OncoPlus Comprehensive Cancer Panel Tier 1 = 316 genes Tier 2 = 896 genes ~1200 genes I. II. Actionable Possibly Actionable III. Cancer Research Interest Tier 1 Validated by CLIA standards Included in Clinical Report Progressive Validation Tier 2 Discovery content (masked without IRB approval and patient consent)

49 Capture Assay Flexibility Mutations/Indels Rearrangements EGFR Exon 19del TP53 11kb deletion Copy Number Events Gene Fusions EGFR/MET/other amplification KIF5B-RET Fusion

50 Scant Mutations A critical emerging target in cancer diagnostics: Mutation detection in plasma (ctdna), urine, pancreatic secretions, etc. Minimal residual disease detection (heme, etc.) Bettegowda C et al. Sci Transl Med 2014

51 NGS Data Is Not Perfect

52 Molecular Barcode Proof-Reading Kinde I et al. PNAS 2011;108:

53 Without Molecular Barcode Proof-Reading

54 With Molecular Barcode Proof-Reading NPM1 c.860_863dup AF = 0.1%

55 Summary: NGS in Cancer Diagnostics NGS offers affordable, broad coverage of many anomaly types. Optimal assay selection depends on a variety of institutional and laboratory-related factors. Many options for preparation of targeted sequencing libraries, each with different pros and cons: Amplicon: Low input, less expensive, more targeted, limited applications Capture: Higher input, more expensive, more scalable, more applications DNA sampling is ALWAYS a critical concern. Every assay produces assay-specific data with assay-specific artifacts. NGS oncology assays are ALWAYS critically dependent on high quality informatics systems. Low % mutation targets (e.g. ctdna) are addressable with NGS using molecular barcodes to increase specificity. The majority of described NGS applications have not yet reached the clinic, and are only awaiting demonstration of clinical relevance.

56 Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels April 13, 2016 Webinar Series Brought to you by the Science/ AAAS Custom Publishing Office Participating experts P. Mickey Williams, Ph.D. Frederick National Lab for Cancer Research Frederick, MD Sponsored by Jeremy Segal, M.D., Ph.D. University of Chicago School of Medicine Chicago, IL To ask a question, click the Ask A Question button under the slide window

57 Characterizing cancers from liquid biopsies and FFPE samples: The rise of targeted sequencing panels April 13, 2016 Webinar Series Brought to you by the Science/ AAAS Custom Publishing Office Look out for more webinars in the series at: webinar.sciencemag.org To provide feedback on this webinar, please your comments to Sponsored by For related information on this webinar topic, go to: sequencing.roche.com