100,000 Genomes Project Update: The Cancer Programme Cancer Testing Strategy & NHS Genomic Medicine Service

Transcription

1 100,000 Genomes Project Update: The Cancer Programme Cancer Testing Strategy & NHS Genomic Medicine Service Nirupa Murugaesu Consultant Medical Oncologist at St George s Hospital Clinical Lead for Molecular Oncology at Genomics England 17 March

2 Genomics The science of all the DNA in the genome How it is sequenced analysed interpreted

3 The first human genome sequence Completed in cost $3 billion and took more than 10 years

4 How many genes would they find in the human genome? 50,000? 100,000? 150,000

5 We have about 20,000 genes.. The same number as a starfish

6 Human genome ~3 billion nucleotides For comparison, there are around 5,500,000 (5.5 million) letters in the whole of the Harry Potter book series.

7 Cost of sequencing Today it takes 24 hours costs less than $1000

8 The 100,000 Genomes Project Announced by the Prime Minister in December 2012 An Olympic Legacy Genomics England announced by Secretary of State for Health in speech during NHS 65 th Anniversary Celebrations, July July

9 The 100,000 Genomes Project Focus on rare disease and cancer NHS patients

10 100,000 Genomes Project Cancer Programme February: 50,000 genomes sequenced ,000 patients recruited to Cancer Main Programme The Cancer Programme was initiated in 2012 to develop infrastructure for routine high throughput whole genome tumour sequencing for NHS cancer patients. 1. To establish a national research platform of molecular data with linkage to longitudinal clinical data. 2. Transform delivery of molecular testing in NHS clinical cancer care. Fast track cancer results returned within 18 working days First main programme cancer results returned Genomics England Clinical Interpretation Partnership (GeCIP) established: >2600 researchers registered in 41 domains First 11 GMCs (Genomic Medicine Centres) announced Illumina wins competitive tender for sequencing Prime Minister announces 100,000 Genomes Project First research users access data (WGS and clinical) within research environments Sequencing moves to new NHS sequencing centre at Hinxton, Cambridgeshire First recruitment to main cancer programme First collection of samples for cancer programme pilot (FFPE and FF) Genomics England set up as a Department of Health-owned company 10

11 How it works

12 NHS Genomic Medicine Centres 13 Genomic Medicine Centres covering England, and Northern Ireland, Scotland and Wales now on board Responsible for identifying and recruiting participants and for clinical care following results 13 July

13 Cancer Analysis & Interpretation for Main Programme (>95% FF samples) GMC selfreported recruitment to 9 Mar ,799 cancer samples (inc. tumour sample and germline) 16,000 14,000 12,000 10,000 8,000 6,000 13,042 14,799 10,289 12,386 8,380 10,768 5,370 10,158 6,400 6,400 4,332 4,332 4 weekly average = 360 samples 4,000 2,000 2,6322, Samples at GMCs 2.1 Samples at Biobanks 2.2 DNA Dispatched 3.1 DNA passed QC 3.2 WGS Completed 4.1 WGS received to illumina by GEL Dispatched to GMCs 13

14 Registered participants by tumour type as of 8 th March 2018 Testicular Germ Cell Tumours, 45 Upper Gastrointestinal, 157 Adult Glioma, 399 Sinonasal, 1 Bladder, 325 Sarcoma, 681 Renal, 910 Breast, 2207 Prostate, 722 Ovarian, 582 Oral Oropharyngeal, 108 Childhood, 41 Nasopharyngeal, 7 Malignant Melanoma, 219 Colorectal, 1887 Lung, 1397 Hepatopancreatobiliary, Endometrial Carcinoma, Haemonc, 198 Tumour Type Carcinoma of unknown primary, 116 Adult Glioma 399 Bladder 325 Breast 2207 CUP 116 Childhood 41 Colorectal 1887 Endometrial Carcinoma 541 Haemonc 198 Hepatopancreatobiliary 294 Lung 1397 Malignant Melanoma 219 Nasopharyngeal 7 Oral Oropharyngeal 108 Ovarian 582 Prostate 722 Renal 910 Sarcoma 681 Sinonasal 1 Testicular Germ Cell 45 Upper Gastrointestinal 157 n 13 July

15 Progress: Cancer Programme Delivering Clinical WGS Technical/Logistical Challenges LOCAL Pathways for tissue acquisition and processing Tissue preparation (FF/FFPE) [QUALITY] Sample type (biopsies/resections) [QUANTITY] Molecular pathology (Tumour assessment, Digital Pathology) [PURITY] Pathways for patients Consent Pathways for results and decision-making Tumour sequencing Boards vs tumour-specific MDTs CENTRAL High throughput, cost-effective sequencing DNA requirement [QUANTITY] Turnaround time Automated analysis, annotation

16 Progress: Cancer Programme Delivering Clinical WGS Technical/Logistical Challenges LOCAL Pathways for tissue acquisition and processing Tissue preparation (FF/FFPE) [QUALITY] Sample type (biopsies/resections) [QUANTITY] Molecular pathology (Tumour assessment, Digital Pathology) [PURITY] Pathways for patients Consent Pathways for results and decision-making Tumour sequencing Boards vs tumour-specific MDTs CENTRAL High throughput, cost-effective sequencing DNA requirement [QUANTITY] Turnaround time Automated analysis, annotation

17 NHS Molecular pathology transformation Removing formalin from theatres via vacuum-packing Biopsies into genomics-friendly transport media (PBS, alcohol) Snap-freezing in OPD and pathology: moving away from liquid nitrogen Digital Pathology assessment driving tissue sampling Lead by Professor Louise Jones & Dr Clare Craig Cellular pathology samples (in vivo/ ex vivo)

18 Sample Evenness Handling of coverage AT drop 4.4 Evenness of coverage 12.9 FF AT drop 15.8 Evenness of coverage 50.7 FFPE 100 kb Calculated as median for the root mean square deviation (RMSD) of coverage calculated in non-overlapping 100 kb windows 7

19 Biopsy overview Total 180 of Biopsy samples have been submitted in 2017 and 2018 so far 7 GMCs: South London (110), East of England (13), North East and Cumbria (9), North Thames (38) Greater Manchester (6) and Wessex (1), Oxford (3) 12 cancer types: Adult Glioma, Breast, Bladder, Carcinoma of unknown primary, Childhood, Colorectal, Haemonc, Lung, Ovarian, Prostate, Sarcoma, Testicular Germ Cell Breakdown of biopsies by tumour type Adult Glioma Bladder Breast Carcinoma of unknown primary Childhood Colorectal Haemonc Lung Ovarian Prostate Sarcoma Testicular Germ Cell Tumours 19

20 Whole genome sequencing of biopsies A total of 180 biopsies have been received in 2017 and 2018 so far 2 breast biopsies have failed QC at UKB 132 samples have reached Illumina for sequencing so far 102 samples have passed QC at Illumina, 11 have failed QC and 19 are awaiting results 68 WGS have been delivered back to Genomics England Breakdown of samples at UKB Pass No QC available Fail Breakdown of samples at Illumina Pass No QC available Fail 20

21 Progress: Cancer Programme Delivering Clinical WGS Technical/Logistical Challenges LOCAL Pathways for tissue acquisition and processing Tissue preparation (FF/FFPE) [QUALITY] Sample type (biopsies/resections) [QUANTITY] Molecular pathology (Tumour assessment, Digital Pathology) [PURITY] Pathways for patients Consent Pathways for results and decision-making Tumour sequencing Boards vs tumour-specific MDTs CENTRAL High throughput, cost-effective sequencing DNA requirement [QUANTITY] Turnaround time Automated analysis, annotation

22 Implementation of the Joint Consensus Statement Consent after confirmed cancer diagnosis Samples pertaining directly to the diagnostic function of the 100,000 Genomes Project i.e. tumour samples and germline samples are diagnostic samples. These samples do not fall under the Human Tissue Act 2004 and their storage is not subject to licensing. Consensus Statement Version /03/17 Lead by Dr Clare Craig

23 Progress: Cancer Programme Delivering Clinical WGS Technical/Logistical Challenges LOCAL Pathways for tissue acquisition and processing Tissue preparation (FF/FFPE) [QUALITY] Sample type (biopsies/resections) [QUANTITY] Molecular pathology (Tumour assessment, Digital Pathology) [PURITY] Pathways for patients Consent Pathways for results and decision-making Tumour sequencing Boards vs tumour-specific MDTs CENTRAL High throughput, cost-effective sequencing DNA requirement [QUANTITY] Turnaround time Automated analysis, annotation

24 Phase 2 Fast Track summary - weeks NHS GMCs have been part of the first 25 weeks of Fast Track. We have recently increased capacity to 40 slots per week, this is split across the 4 GMCs. It has been agreed that 96 more samples will be processed in a medium turnaround time (< 20 days) at Illumina from December. The current average TAT from sample submission to sequence return is 12.9 days 12.9 Any samples not meeting the QC parameters at UKB have been transitioned to the normal turnaround time pipeline. Discussions with further GMCs are underway for a further stage of the rollout Samples sent by NHS GMCs were sequenced (89.6%) NHS GMC Samples sent Samples passed QC at UKB Samples sequenced East of England North Thames Oxford South London

25 Variant-level and gene-level actionability in Domain 1 Variants in a virtual panel of 132 potentially actionable genes Actionable genes are defined as genes in which small variants (SNVs and indels <50bp) have reported therapeutic, prognostic or clinical trial association, as defined by the GenomOncology Knowledge Management System

26 Domain 1 somatic variants

27 Domain 1 somatic variants

28 Potential clinical trials

29 Supplementary Content Structural variants Mutational density Coverage and copy number Mutational signatures Mutation context Hypermutation rain plots

30 1260 cancer whole genomes analysed Jan July

31 Colorectal genomes (potential actionability) Trials and therapies available for CRC at the gene level Coding SNVs per Mb (log10) COSMIC cancer signature Signatures defective DNA mismatch repair/msi tumours Signature 10 - recurrent POLE somatic mutations/ultrahypermutator tumours Signatures correlate with high mutation burden. Population MSI deficient and proficient very well distinguished

32 KRAS BRAF NRAS Colorectal genomes (potential actionability) Variant level Coding SNVs per Mb (log10) COSMIC cancer signature This is a subset of variant-level associations with therapies and trials Only therapies are associated with specific variants KRAS/NRAS variants are mutually exclusive with BRAF

33 KRAS BRAF NRAS Colorectal genomes Colorectal genomes (potential genomes (potential actionability) actionability) Trials and therapies available Variant for level CRC at the Trials and therapies available Phase Lung cancer 2 - Fast Track summary gene (weeks level for lung cancer at the gene level (log10) 1-22) 4 NHS GMCs have been part of the first 22 weeks of Fast Track. Signature Signatures 4 We 6 - tobacco have + 15 recently - defective increased capacity to 40 slots per week, this is split carcinogens This is a subset of variant-level DNA associations mismatch across with repair/msi the therapies 4 GMCs. and It has trials tumours been agreed that 96 more samples will be Signatures processed correlate in a medium turnaround time (< 20 days) at Illumina from with Onlyhigh therapies Signature December. mutation are associated with 10 - recurrent POLE burden. specific somatic Mutation variants mutations/ultrahypermutator current average tumours TAT burden distribution is The continuous KRAS/NRAS variants are mutually from exclusive sample with submission BRAF to Signatures sequence correlate return with is high We are planning to 11.8 mutation Highlighted days burden. are Population implement cases where WGS with interpretation high mutation clonal can burden gois neoantigen analysis beyond very standard well distinguished into our pipeline gene panels. High mutation burden and signature analysis may 614 indicate Samples eligibility sent by NHS for immunotherapy GMCs were treatment. sequenced 684 (89.8%) Coding SNVs per Mb COSMIC cancer COSMIC Coding SNVs per Mb (log10) signature cancer signature

34 Strategy for utilising WGA Current: 1. Enhance recruitment to existing clinical trials - Better curation of existing national stratified oncology trials: FOCUS-4, MATRIX - Enhance recruitment to Phase 2 targeted therapy trials Next steps: 2. Pan genomic markers - Utilise WGA for stratification as a companion diagnpostic (dependent on biopsy pathways, TAT) - Better molecular characterisation of tumours/patients 3. Designing clinical trials - Utilise WGA for stratification (dependent on biopsy pathways, TAT) - To better identify potential biomarkers for stratification - Better molecular characterisation of tumours/patients 34

35 Facilitating recruitment of patients to potential clinical trials PIK3CA in lung vs colorectal, breast trials PIK3CA trials in lung are curated with specific variants, e.g. Q546E, D549N, E542K, H1047L, etc. In colorectal, breast cancer are trials associated with PIK3CA Mutation will all appear in the gene-level actionability category Need for better curation of UK clinical trials At the moment we relay on curation of clinicaltrials.gov provided by GenomOncology Trial status on clinicaltrials.gov is poorly updated Most trials in clinicaltrials.gov are annotated at gene level 13 July

36 Facilitating recruitment of potential patients National Lung MATRIX Trial SMP2 n=122 patients Matrix Trial Centre Patients North West Coast Greater Manchester Matrix trial centre Metastatic Primary Unknown if Primary or Metastatic West Midlands Oxford East of England South London Wessex matrix trial: lung cancer patients n = 108 (N.B. 14/122 patients do not ha ve LDP code available in opencga at p 13 July

37 Oxford Wessex South London West Midlands East of England North Thames Greater Manchester number of patients with mutations in focus4 eligible genes=229 Focus4 Metastatic Primary Facilitating recruitment of potential patients 13 July n=229 patients FOCUS-4 Trial number of patients with mutations in focus4 eligible genes=229 Focus4 Metastatic Primary KRAS/NRAS 40% BRAF 13% PIK3CA 10% 2+ Genes (KRAS/NRAS/PI K3CA/BRAF 37%

38 Mutation burden across different tumour types POLE mutation potential MSI 13 July

39 Facilitating recruitment of potential patients GMC High MTB POLE Potential Immunotherapy clinical trials using MSI as a biomarker High mutational burden High PolE signature Greater Manchester West Midlands Oxford North Thames East of England West of England South London Wessex 13 July COLORECTAL = 208, ENDOMETRIAL_CARCINOMA = 34

40 Pan genomic markers Mainstream TMB testing through Genomics England and existing infrastructure Obtain supporting evidence: Fast Track all lung biopsies (focus on Fast Track sites) 40

41 Molecular Tumour Boards federated in space and time: South London GMC as an exemplar Molecular Tumour Boards Patient Identification Patient Identification Consent Consent Sample collection Sample collection Sample processing Sample processing Sample tracking Sample tracking Results validation Results validation Clinical review & Clinical feedback review & feedback Clinical review and feedback Molecular Tumour Boards established at three laboratory sites for rapid lab validation and clinical interpretation of results in line with local pathways and agreed GMC standards Clinical review and feedback Molecular Tumour Boards established at three laboratory sites for rapid lab validation All results and sent clinical from tumour interpretation boards to of tumour results in specific line with MDTs local and pathways leads and agreed GMC wide GMC Molecular standards Tumour Board in place to ensure standardisation, All consistency results sent and from sharing tumour of good boards practice to tumour specific MDTs and leads GMC Development wide Molecular of standardised Tumour Board pathways, in place integrated to ensure molecular/pathology standardisation, consistency reports and and generic sharing patient of good letters practice Development Networks being of developed standardised across pathways, GMC region integrated to support molecular/pathology LDPs reports and generic patient letters Example of generic patient letter: Networks being developed across GMC region to support LDPs Example of generic patient letter: LDP LDP SGH SGH LDP LDP GMC GMC GST GST LDP LDP LDP LDP KCH KCH LDP LDP Slide courtesy of Professor Sean Whittaker Slide courtesy of Professor Sean Whittaker Page 32

42 Moving to a NHS genomics service Molecular testing: how, when, why? WHEN? HOW? WHY? Single gene/ Standalone test Small panel (eg Amplicon hot spot) Large panel (eg Hybridisationcapture) Exome Genome mrnaseq methylation proteinomics diagnostic biopsy surgery Biopsy (diagnostic/recurrence) +/-biopsy +/-biopsy LOCAL/REGIONAL DISEASE METASTATIC DISEASE neoadjuvant chemo-radiorx Adjuvant chemo-radiorx chemo-rx targeted drug chemo-rx Phase II/III Clinical trial Phase I clinical trial/ Experimental/compassi onate use drug Standard of Care actionability Clinical Trials Research Diagnostic Predictive Prognostic Targeted Drugs Single new agent vs SOC Multi-arm umbrella/basket Experimental n of 1 genomics-drug matching Discovery Longitudinal patient studies: stratifiers of response

43 Implementation of a Cancer Testing Strategy

44 Acknowledgements Genomics England Cancer Team Bioinformatics Team Informatics Team Clinical Data Team GeCIPs NHS England NHS Genomic Medicine Centres Health Education England Cancer Working Group Validation & Feedback Working Group Illumina/R&D Team 44

45 Thank you Any questions?

46 Following the #genomes100k Like the Genomics England page Follow Genomics England 46