Variant interpretation exercise. ACGS Somatic Variant Interpretation Workshop Joanne Mason 21/09/18

Similar documents
6/12/2018. Disclosures. Clinical Genomics The CLIA Lab Perspective. Outline. COH HopeSeq Heme Panels

Reporting TP53 gene analysis results in CLL

Concurrent Practical Session ACMG Classification

A guide to understanding variant classification

The Role of Next Generation Sequencing in Solid Tumor Mutation Testing

TP53 mutational profile in CLL : A retrospective study of the FILO group.

Illumina Trusight Myeloid Panel validation A R FHAN R A FIQ

Introduction of an NGS gene panel into the Haemato-Oncology MPN service

Acute Myeloid Leukemia with RUNX1 and Several Co-mutations

Variant interpretation using the ACMG-AMP guidelines

Collaboration between CEQAS and UK NEQAS for Molecular Genetics Solid Tumour EQAs and variant interpretation

Supplemental material Mutant DNMT3A: A Marker of Poor Prognosis in Acute Myeloid Leukemia Ana Flávia Tibúrcio Ribeiro, Marta Pratcorona, Claudia

NGS in tissue and liquid biopsy

JULY 21, Genetics 101: SCN1A. Katie Angione, MS CGC Certified Genetic Counselor CHCO Neurology

Clonal Evolution of saml. Johnnie J. Orozco Hematology Fellows Conference May 11, 2012

Merging single gene-level CNV with sequence variant interpretation following the ACMGG/AMP sequence variant guidelines

Introduction to Cancer Bioinformatics and cancer biology. Anthony Gitter Cancer Bioinformatics (BMI 826/CS 838) January 20, 2015

Variant Classification: ACMG recommendations. Andreas Laner MGZ München

Characterisation of structural variation in breast. cancer genomes using paired-end sequencing on. the Illumina Genome Analyser

What we know about Li-Fraumeni syndrome

Supplementary Figure 1

Standardisation of variant interpretation across scorers. Matthew Smith Molecular Pathology Diagnostic Service University Hospital Birmingham

Molecular Pathology of Ovarian Carcinoma with Morphological Correlation

Targeted NGS in oncology and hemato-oncology using in-house designed gene panels. Joni Van der Meulen Molecular Diagnostics UZ Ghent (MDG) 24/03/2017

Variant Classification: ACMG recommendations. Andreas Laner MGZ München

Shashikant Kulkarni, M.S (Medicine)., Ph.D., FACMG Associate Professor of Pathology & Immunology Associate Professor of Pediatrics and Genetics

Importance of minor TP53 mutated clones in the clinic

Molecular Markers in Acute Leukemia. Dr Muhd Zanapiah Zakaria Hospital Ampang

Next generation sequencing analysis - A UK perspective. Nicholas Lea

AD (Leave blank) TITLE: Genomic Characterization of Brain Metastasis in Non-Small Cell Lung Cancer Patients

Please Silence Your Cell Phones. Thank You

Laboratory Service Report

Protein Domain-Centric Approach to Study Cancer Somatic Mutations from High-throughput Sequencing Studies

Next Generation Sequencing in Haematological Malignancy: A European Perspective. Wolfgang Kern, Munich Leukemia Laboratory

Clonal hematopoiesis of indeterminate potential and MDS. Siddhartha Jaiswal AAMDS Meeting 3/17/16

WHEN DO MUTATIONS OCCUR?

Corporate Medical Policy. Policy Effective February 23, 2018

PALB2 c g>c is. VARIANT OF UNCERTAIN SIGNIFICANCE (VUS) CGI s summary of the available evidence is in Appendices A-C.

Accel-Amplicon Panels

Supplementary Table 1. PIK3CA mutation in colorectal cancer

ADRL Advanced Diagnostics Research Laboratory

NGS panels in clinical diagnostics: Utrecht experience. Van Gijn ME PhD Genome Diagnostics UMCUtrecht

Using large-scale human genetic variation to inform variant prioritization in neuropsychiatric disorders

Variant Classification. Author: Mike Thiesen, Golden Helix, Inc.

Nature Genetics: doi: /ng Supplementary Figure 1. Mutational signatures in BCC compared to melanoma.

Introduction to genetic variation. He Zhang Bioinformatics Core Facility 6/22/2016

Frequency(%) KRAS G12 KRAS G13 KRAS A146 KRAS Q61 KRAS K117N PIK3CA H1047 PIK3CA E545 PIK3CA E542K PIK3CA Q546. EGFR exon19 NFS-indel EGFR L858R

Concomitant WT1 mutations predicted poor prognosis in CEBPA double-mutated acute myeloid leukemia

Tumor suppressor genes D R. S H O S S E I N I - A S L

Molecular Testing Updates. Karen Rasmussen, PhD, FACMG Clinical Molecular Genetics Spectrum Medical Group, Pathology Division Portland, Maine

Acute leukemia and myelodysplastic syndromes

Out-Patient Billing CPT Codes

NGS for Cancer Predisposition

Proposal form for the evaluation of a genetic test for NHS Service Gene Dossier

TP53 ABERRATIONS Methodical considerations

GENETIC TESTING FOR FLT3, NPM1 AND CEBPA VARIANTS IN CYTOGENETICALLY NORMAL ACUTE MYELOID LEUKEMIA

Time series analysis of TP53 gene mutations in recurrent HPV-negative vulvar squamous cell carcinoma

BWA alignment to reference transcriptome and genome. Convert transcriptome mappings back to genome space

EXAMPLE. - Potentially responsive to PI3K/mTOR and MEK combination therapy or mtor/mek and PKC combination therapy. ratio (%)

Functional analysis of DNA variants

Chapter 12-4 DNA Mutations Notes

New drugs in Acute Leukemia. Cristina Papayannidis, MD, PhD University of Bologna

Proposal form for the evaluation of a genetic test for NHS Service Gene Dossier/Additional Provider

Proposal form for the evaluation of a genetic test for NHS Service Gene Dossier

Laboratory Service Report

variant led to a premature stop codon p.k316* which resulted in nonsense-mediated mrna decay. Although the exact function of the C19L1 is still

Breast and ovarian cancer in Serbia: the importance of mutation detection in hereditary predisposition genes using NGS

Molecular Markers. Marcie Riches, MD, MS Associate Professor University of North Carolina Scientific Director, Infection and Immune Reconstitution WC

Mutational Impact on Diagnostic and Prognostic Evaluation of MDS

UNIVERSITY OF TORINO DEPARTMENT OF ONCOLOGY. Giorgio V. Scagliotti University of Torino Dipartment of Oncology

The mutations that drive cancer. Paul Edwards. Department of Pathology and Cancer Research UK Cambridge Institute, University of Cambridge

SUPPLEMENTAL APPENDIX METZELER ET AL.: SPECTRUM AND PROGNOSTIC RELEVANCE OF DRIVER GENE MUTATIONS IN ACUTE MYELOID LEUKEMIA

Transform genomic data into real-life results

LESSON 3.2 WORKBOOK. How do normal cells become cancer cells? Workbook Lesson 3.2

Clinical Grade Genomic Profiling: The Time Has Come

6.3 DNA Mutations. SBI4U Ms. Ho-Lau

CANCER GENETICS PROVIDER SURVEY

Analysis of Massively Parallel Sequencing Data Application of Illumina Sequencing to the Genetics of Human Cancers

Issues arising from UKNEQAS schemes. Ottie O Brien, Northern Genetics Service, Newcastle, UK 15 th May 2014

CITATION FILE CONTENT/FORMAT

THE UMD TP53 MUTATION DATABASE UPDATES AND BENEFITS. Pr. Thierry Soussi

CHAPTER IV RESULTS Microcephaly General description

Key determinants of pathogenicity

West Midlands Regional Genetics Laboratory

Supplementary Tables. Supplementary Figures

Secuenciación masiva: papel en la toma de decisiones

COSMIC - Catalogue of Somatic Mutations in Cancer

Juan Ma 1, Jennifer Dunlap 2, Lisong Shen 1, Guang Fan 2 1

School of Pathology and Laboratory Medicine: Current and New Research Interests

CHR POS REF OBS ALLELE BUILD CLINICAL_SIGNIFICANCE

Correspondence to Nature Genetics: Exploring pediatric cancer mutation information using ProteinPaint

Blastic Plasmacytoid Dendritic Cell Neoplasm with DNMT3A and TET2 mutations (SH )

Objectives. Genetics and Rett syndrome: As easy as apple pie! Chromosome to gene to protein

REGULATED SPLICING AND THE UNSOLVED MYSTERY OF SPLICEOSOME MUTATIONS IN CANCER

Should Mutational Status in Primary Myelofibrosis (PMF) Guide Therapy..YES!!!

Introduction to Genetics

Winship Cancer Institute of Emory University New Determinants and Approaches for MPN

The lymphoma-associated NPM-ALK oncogene elicits a p16ink4a/prb-dependent tumor-suppressive pathway. Blood Jun 16;117(24):

MEDICAL GENOMICS LABORATORY. Next-Gen Sequencing and Deletion/Duplication Analysis of NF1 Only (NF1-NG)

MUTATIONS, MUTAGENESIS, AND CARCINOGENESIS. (Start your clickers)

Transcription:

Variant interpretation exercise ACGS Somatic Variant Interpretation Workshop Joanne Mason 21/09/18

Format of exercise Compile a list of tricky variants across solid cancer and haematological malignancy. Also include a couple of more straightforward variants. Include referral reason/tumour site, gene, HGVS for variant and VAF (all >20%). Limit cells for ease of compilation. - Classification (clinically significant, potentially clinically significant, VUS, likely benign, benign) - Would the variant would be reported externally? (yes/no/maybe) - Ask for a brief summary for the reasoning behind the responses Distribute to everyone on mailing list/slack 5-14 responses per variant

Solid tumours

Colorectal Dukes C1 adenocarcinoma PIK3CA frameshift Potentially clinically significant: - Nonsense variant, COSMIC similar variants at penultimate aa in solids - We think this is a frameshift/ premature stop in the last codon of the exon, not a recognised driver mutation therefore would not report out - There are clinvar entries and we report fs routinely, regardless of where they are located in the gene. - Variant has been previously reported in adenocarcinoma and functional studies suggest that it is an activating mutation (Ng et al. Cancer Cell. 2018; 33:450)

Colorectal Dukes C1 adenocarcinoma PIK3CA frameshift VUS: - Not included in COSMIC or My Cancer Genome and clinical significance is unclear. Besides this mutation is located in the codon before the STOP codon, resulting in an elongated protein with only extra 4 amino acids. - Some evidence activating- frameshift causing protein extention (PMID: 29533785 ). No evidence clinically actionable Likely benign - Truncating mutations in PIK3CA show little recurrence outside COSMIC and are not known to be pathogenic. Mutation is very close to C-terminus.

Colorectal Dukes C1 adenocarcinoma PIK3CA frameshift Potentially CS: 2/5 labs Yes, 1/5 Maybe, 2/5 No ( we don t currently report variants in PIK3CA outside of specific codons ). VUS: 3/4 No, 1/4 Maybe ( No evidence clinically actionable ) Likely benign: No

Colorectal Dukes C1 adenocarcinoma PIK3CA c.3201_3202insa p.asn1068lysfs*5 Summary points Lack of consensus on likely effect of insertion. Some labs think this frameshift creates a premature stop in the last codon of the exon, one lab called it a nonsense variant and other labs consider this results in an elongated protein with 4 extra amino acids. Many labs commented on it s position in the last codon and next to the C-terminus Some labs would not report on basis that this is not a recognised driver mutation, despite assessing it as pathogenic (i.e. not clinically actionable)

Lung tumour BRAF missense Clinically significant/potentially CS: - Cannonical driver mutation - Clinvar and multiple COSMIC entries - Located in highly conserved motif in KD domain, other variants reported- ClinVar - Described as oncogenic in OncoKB, associated with gain of function - Reported in multiple papers in NSCLC but insufficient evidence regarding actionability

Lung tumour BRAF missense VUS: - This is not an activating mutation in codon 600/exon 15, the standard of care test for NICE-approved BRAF/MEK inhibitor therapy but may be one of the less common BRAF inactivating mutations. Not clinically significant based on current knowledge. - Well documented activating mutation in multiple tumour types, but prognostic significance/actionability of this variant in lung cancer is unclear.

Lung tumour BRAF missense CS: 2/2 Yes Potentially CS: 5/6 Yes, 1/6 Maybe When assessing this type of somatic variant in solid tumours an additional level of assessment is required. The variant may be established as likely / pathogenic but the effect on the kinase domain may be different depending on the amino acid substitution and thus the response to targeted treatments may differ. It is important to establish the effect on kinase activity and then source information on response to targeted drugs. VUS: 1/3 Yes, 2/3 No

Lung tumour BRAF c.1406g>t p.(gly469val) Summary - All agreed this was a pathogenic activating mutation, but some labs annotated as VUS/would not report because of questionable actionability in lung tumours.

Lung tumour CCND1 del ins Benign / likely benign - Apparent indels can be the result of sequencing artefacts - Not included in COSMIC, gnomad/exac or My cancer genome or found in literature search. No evidence found to support clinical reporting or significance. - Not recurrent

Lung tumour CCND1 del ins VUS - Not much evidence - The only indicator to report is the VAF, so it might be reportable as a simple clonal marker. - Variant affects a well conserved residue within the CCND1 protein - 2/8 might report externally

Lung tumour CCND1 c.364_366delinsaaa p.glu122lys Summary points - Suggestion this could be an artefact what are labs doing to reassure themselves they are excluding NGS artefacts? - Is non-recurrence / absence from databases sufficient to annotate as benign/likely benign? - One lab may report as evidence of clonality on VAF alone

Ovarian cancer TP53 (dup) nonsense Clinically significant/potentially - Nonsense - Frameshift in tumour suppressor, predict likely pathogenic. - Classic disrupting mutation - Presumed LOF. Always report fs/stops. - R175 is a hotspot for mutations - may be eligible for NICE-approved PARP inhibitor therapy

Ovarian cancer TP53 c.518_522duptgagg p.(arg175*) Summary points - Good consensus on clinical significance LOF mutation in TS gene - Lack of consensus on nomenclature regarding this duplication: frameshift or nonsense or either? - One lab stated in IARC, one lab could only find point mutation of this codon in IARC, not exact variant, a third lab stated not in mutation database/s

Haem Malignancy

Monocytosis MPL missense Clinically significant/potentially - Known activating mutation with supporting functional evidence (Ding J, et al. (2004) - Known clonal marker for MPN (entries on COSMIC, literature). Recurrent mutation found in a range of myeloid malignancies including myelofibrosis - Clinvar shows pathogenic/likely pathogenic (can be germline) - absent in GnomAD

Monocytosis MPL c.1514g>a p.(ser505asn) 12/12 would report Summary points - Good consensus lots of evidence available - This is one of a rare class of mutations than can be seen as either a germline or a somatic event (see Beer et al 2008 PMID: 18451306). It is a pathogenic mutation in both circumstances, associated with inherited thrombocytosis or ET respectively. Given the VAF, germline testing should be recommended

?transformation to AML TET2 missense Potentially clinically significant - Confirmed somatic variant in haematopoietic tissue (COSMIC, gnomad/exac) - Not common variant, highly conserved amino acid, in silico not tolerated. TET2 mutations common in AML. Likely benign - Likely passenger event: no support outside of COSMIC.

?transformation to AML TET2 missense VUS - Not seen in Clinvar; COSMIC: 5 calls at this position; large physicochemical difference; moderately conserved nucleotide; highly conserved amino acid. - No good evidence of pathogenicity, although predictive scores are high. - not a confirmed SNP (GnomAD/dbSNP) however note VAF - VAF would suggest a heterozygous variant, not clonal. - Insufficient evidence that the variant is actionable. - Does not fit MoP for gene (LOF).

?transformation to AML TET2 missense Reflecting the fact that most labs think VUS

?transformation to AML TET2 c.5609c>t p.(ser1870leu) Summary points - Labs are using same evidence to come to different conclusions - Some labs are anti-vaf ( reluctant to place any reliance on the VAF), others are using VAFs as evidence of potential germline origin (either for SNP or for cancer associated syndrome) - Caution with VAFs if using amplicon NGS - Very few defined actionable variants (in terms of treatment) in haem malignancy (examples include FLT3-ITDs and TKD mutations and midostaurin, IDH1(R132) and IDH2(R140) and IDH inhibitors). Clinical significance often at the gene level once pathogenicity established >diagnosis and prognosis, only occasionally treatment.

?MDS TET2 missense Potentially clinically significant /clinically significant - COSMIC: 44 calls at this position, 21 calls corresponding to this variant. - Recurrent across multiple databases with multiple different alternative amino acids, both of which are strong evidence for pathogenicity - strong predictive scores - Hu et al describe this as a 'critical residue' in Loop L1 of the DNA binding domain - VAF is not heterozygous/homozygous

?MDS TET2 missense VUS - COSM87130; Not found in dbsnp. ExAC ~0.005% for p.arg1261his - Metzeler et al. 2011 - Somatically acquired mutation in 2 patients (3rd patient had no matched normal tissue) - however all three patients also had truncating mutations in TET2 - therefore its significance is uncertain

?MDS TET2 missense 2 labs would not report externally 1 VUS and 1 Potentially CS

?MDS TET2 c.3781c>t p.(arg1261cys) Summary points - Majority of labs agreed this is (potentially) clinically significant. Consistent with a diagnosis of MDS. Not enough to diagnose MDS in absence of other findings. - No-one mentioned possibility of CHIP. How many labs mention this in their reports, and when? - Some labs again mentioned the relevance of the VAF (21%) in helping them decide this was probably a clonal marker, probably/possibly a driver

Thrombocytosis DNMT3A nonsense CS/potentially CS - Nonsense variant presumed pathogenic. Fits MoP for DNMT3A. - Not found in COSMIC, dbsnp, HMGD, ExAC; but likely oncogenic due to lose of function. - MoP can be dominant negative (e.g. R882H common missense) or haploinsufficiency, the latter acting over a more extended period of time (Cole et al)

Thrombocytosis DNMT3A nonsense VUS - Similar evidence quoted as labs assigning clinical significance or potential clinical significance

Thrombocytosis DNMT3A c.1443c>g p.(tyr481*) 1 lab would not report (VUS) Summary points - Good consensus - some labs seem to have a higher threshold for assigning clinical significance or potential clinical significance

?MDS DNMT3A missense Potentially CS Not common variant, absent in GnomAD (PM2). Highly conserved amino acid, in silico not tolerated. Ser663 in MTase domain of protein (?PM1), catalytic activity, aa 664 binding site. Benign - Not recurrent

?MDS DNMT3A missense VUS Not seen in population databases, seen once in COSMIC Highly conserved amino acid. Missense variants S669F, A662G reported as oncogenic while Y660C reported as possibly oncogenic in AML. Variant is in the methyltransferase domain and in silico tools suggest it could be pathogenic. MoP is dominant negative with missense variants being common in methyltransferase domain

?MDS DNMT3A c.1988c>g p.(ser663trp) 11/13 labs assigned VUS > Yes, No and Maybe! Summary points Most labs agree not enough evidence to call either way One lab thought enough evidence (of lack of recurrence) to call benign One lab mentioned possibility of CHIP

?biphenotypic leukaemia TP53 missense CS/ Potentially CS Multiple Clinvar and COSMIC entries IARC: pathogenic/non-functional Canonical driver mutation 12/12 would report externally

?AML RUNX1 frameshift CS/ Potentially CS Frameshift mutation at the 3' end of RUNX1 predicted to disrupt the transactivation domain resulting in LOF: MoP in RUNX1 This particular variant not listed on COSMIC but a cluster of other frame shifting indels in the region This would be reported as 'likely pathogenic', given the diagnosis, the allele fraction and that it is a truncating mutation. This is a difficult one as the mutation is very close to the C-terminus. Needs to be considered carefully in the clinical context. WHO provisional entity AML with mutated RUNX1, as a reflection of the evidence base.

?AML RUNX1 frameshift VUS Frameshift mutation located in the Runx inhibition domain. Majority of the pathogenic mutations described are missense. This mutation is localized in a CG-rich region, which is prone to originate sequencing artefacts. Not reported in COSMIC, mycancer genome or gnomad/exac Insufficient evidence that the variant is actionable.

?AML RUNX1 c.1214_1218dup p.(tyr407valfs*194) 10/10 labs CS / Potentially CS would report externally 2/2 labs VUS would not report

?AML RUNX1 c.1214_1218dup p.(tyr407valfs*194) Summary points Most labs agree that frameshift mutations, even at the C- terminus of RUNX1, are likely to disrupt the TAD >pathogenic. One lab stated this is a truncating mutation (?unfamiliar with gene). Actually predicted to result in a longer last exon with removal of VWRPY motif in TAD (fitting LOF model). In AML, RUNX1 (pathogenic) variants are highly relevant: WHO provisional entity for classification, associated with poor prognosis (ELN Guidelines, Döhner et al 2017).

Summary There is some consensus already. There are differences in emphasis between solids and haem. Emphasis on actionability in terms of targeted therapies much more evident for solid tumours. Haem malignancy once pathogenicity established, variants in relevant genes most often used to corroborate diagnosis or for risk stratification. There is plenty of evidence that labs are approaching SVI with much caution. When a variant was assigned to have clinical significance by any lab, this was associated with a general consensus for clinical significance or potential clinical significance i.e. only when the evidence was strong. There were also high levels of consensus for the use of VUS i.e. when there was lack of sufficient evidence.

Summary Many labs are using the same evidence commonly used in germline variant annotation to support assignment of pathogenicity (or not) to somatic variants (nucleotide/aa conservation; functional domain?; in silico prediction tools etc) Evidence that some labs are using VAFs to aid interpretation (or at least report clonality) On occasion, labs are highlighting the potential for a germline variant rather than somatic (can we establish when?) Are haem labs highlighting potential for CHIP?

Summary There is inconsistency in reporting out perhaps due to pathogenicity v actionability Potentially clinically significant variants may not always be reported VUS sometimes reported, sometimes not. (Concept of hot 3s?)

Thanks Many thanks to everyone that participated