Epidemiologi, biomarkører og nye drug targets ved bruk av norske register, kohorter og biobanker - Sett i konteksten av persontilpasset medisin og legemiddelutvikling - Big data og biomedisinsk næringsutvikling fra genomikk til real world data Oslo, 7.mai 2015 Christian Jonasson, CSO Lifandis AS cj@lifandis.com, +4790936941
Agenda Lifandis hvem er vi og hva gjør vi? Hva menes med persontilpasset medisin og biomarkører? Biobanker, kohorter og register. Hva er det som gjør Norge unikt? Anvendelse av biobanker i legemiddelutvikling Case studies
ifandis Profesjonell industri interface for forskning på norske biobanker, kohorter og register Offentlig eiet kommersielt drevet AS (NTNU, HMN, NTFK) Overskudd pløyes tilbake til eierne Våre kunder er legemiddel- og diagnostiske firmaer Tre forretningsområder; biomarkører, epidemiology (RWD) og drug discovery
Epidemiology Clinical Epidemiology Pharmacoepidemiology Molecular Epidemiology Genetic Epidemiology Understand standard of care Disease natural course Disease Risk factors Burden of illness Cost of illness Drug utilization studies Outcome in real-life Comparative Effectiveness Safety (PASS) Disease mechanism Pathophysiology/disease understanding Molecular level data based on omictechnologies Finding new drug targets based on genomics (genotyping and sequencing) RWD RWD Biomarkers Drug Targeting
HUNT Biobank HUNT Databank Clinical medicine Epidemiology Core Genomics Facility Core Proteomics Facility Lifandis collaborator network Bioinformatics Nordic RWD EMR extraction Biostatistics Bioinformatics Biostatistics Genomics
Persontilpasset medisin og biomarkører
23andMe Teams With Big Pharma to Find Treatments Hidden in Our DNA Jan 12, 2015 Wired Magazine In the news on precision medicine Obama calls for major new precision medicine initiative Jan 20, 2015 Reuters News The Genomics Network for Enterprises (GENE) Consortium Drug companies unite to mine genetic data Mar 26, 2015 FT In Iceland s DNA, New Clues to Disease-Causing Genes Mar 25, 2015 NY times
Kjært barn har mange navn.
Effekt Toksisitet Diagnostisk test Biomarkør Effekt Ingen toksisitet Ingen effekt Ingen toksisitet Pasienter med samme diagnose Hva er persontilpasset medisin? Ingen effekt Toksisitet Riktig behandling, riktig dosering, til riktig pasient, til riktig tidspunkt An approach to treating illnesses that takes into account a patient's individual genetic make-up as well as molecular subtypes of diseases to improve the chances of successful treatment."
Hva er en biomarkør? Blodtrykk PET/CT LDL kolesterol HbA1c PSA Kreftmutasjoner (BRCA1&2, EGFR, ALK, KRAS) CYP mutasjoner (PK/PD) CYP2C9, CYP2C19, CYP2D6 Definition of a biomarker: A characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacological response to a therapeutic intervention ctdna mirna proteom metabolom
Hvordan skal vi bruke biomarkører? Prognostiske biomarkører for stratifisering og design av RCT trial enrichment
Need of new blood-based biomarkers for prodromal Alzheimer s disease Cognitive function Pre-clinical AD MCI Dementia Normal aging Many failures in large phase III trials in the last decade Neuropathophysiological process of AD begins years before clinical onset Are we targeting AD at the wrong stage? Need of non-invasive biomarkers (metabolome, exosomes, proteome, mirna) Years
Longitudinal 20-years cohort with dementia validated diagnoses and predementia biobanked samples
Avansert og offentlig helsevesen Mange og gode helseregistre (landsdekkende) Digitalisert helsevesen, inkl EPJ Lover og etisk regelverk som tillater og støtter helseforskning Personnummer Tradisjon for store helseundersøkelser og biobanker Ekspertsentre for omics og biostatistikk Sterke akademiske miljøer
Phenotype matters! Phenotype data associated with omics data is of critical importance Longitudinal data vital for chronic diseases Higher resolution phenotype than codes (ICD10) often required Self-reported data must be used with caution
1973 1975 1984 1995 1999 2002 2011 2015 2017 Norwegian population based cohorts - A unique foundation for tailored services > 1 million Norwegians have so far been recruited into consent based research biobanks based on population based studies only +25-30 million samples in clinical biobanks, including archival tumor tissue
The HUNT Studies (adults more than 20 years old) 1984 86 95 97 06 08 2017 HUNT1 HUNT2 HUNT3 HUNT4 n=77,212 n=65,237 n=50,807 Planning ongoing Questionnaires Questionnaires Questionnaires Clinical Exam Clinical Exam Clinical Exam Open to Blood Samples Blood Samples industry collaboration 10 year follow-up 47,316 10 year follow-up 37,071 20 year follow-up, HUNT 1-3: 27,992 Krokstad S et al, Int. J Epidemiol, 2013
Selected Endpoints 4,000 8,000 800 1,500 8,000 COPD Asthma RA Atrial Fibrillation Fractures 6,000 3,000 1,600 14,000 Myocardial Infarction Type 2 diabetes with dementia with obesity (class I-III) 13,100 2,400 1,800 1,600 2,400 Cancer Prostate cancer Breast cancer Lung Cancer Colorectal cancer
European Research Biobank of the Year 2013 Biological Samples Stored in HUNT Biobank Material Individuals Storage temp. ( C) Isolated DNA (HUNT2 and/or 3) 73,085-20 Serum (HUNT2 and/or 3) 77,788-80/-196 Plasma (HUNT3) 49,072-80/-196 RNA Tempus Tubes (HUNT 3) 14,874-80 Urine Samples (Fresh Frozen) 11,879-80
HUNT biobank for genotyping and NGS studies 30,000 subjects already genotyped on different platforms Genotyping of the whole cohort (n >70,000) ongoing (Illumina s HumanCoreExome BeadChip) with industry access once completed (in 2015) Consent allows industry-sponsored genotyping and NGS
Genomic case study: Rare causal variants for dyslipidemia Exome chip genotyping for new genes involved in dyslipidemia traits performed on 10,000 subjects Identified a new causal variant, TM6SF2 Overexpression and knock-down associated with elevated and reduced lipid levels in mice model
Norske biobanker og legemiddelutvikling
Population-based biobanks Pharmacogenomics Clinical phenotype Tissue banks and diagnostic biobanks for target expression in large patient cohorts clinical trial design Tissue, serum, plasma, and urine for biomarker validation (companion diagnostics) Tissue banks and diagnostic biobanks for target expression in disease and healthy organs Tissue banks for validation of animal model relevance to human disease Tissue, serum, plasma, and urine for biomarker discovery Target ID Target Val Hit Lead Lead Optim Preclinic Phase I Phase II Phase III Research Discovery Development Biobanker er viktig for industri FoU Muligheter for persontilpasset legemiddelutvikling Identifisering og validering av nye targets (farmakogenomikk) Pasientstratifisering basert på biomarkører for å øke sjansen for positivt utfall av kliniske studier Identifisering av nye biomarkører for diagnostiske tester Nye IVD bruksområder, f.eks. tidlig deteksjon biomarkører
Biomarkørforskning Høyteknologisk Biologiske prøver Fenotypedata Ekspertise Genomics Proteomics Metabolomics Biobanker Prediagnostiske prøver Vevsprøver Longitudinelle data Databanker Elektroniske pasientjournaler Helseregistre Epidemiologi Biologi/genetikk Statistikk Bioinformatikk Data modellering
Why do drugs fails? High attrition rates: 95% i fase I 66% i fase II 30% i fase III Main reasons: Lack of efficacy Safety
Until recently, it was very difficult to establish accurate genetic maps of human disease now we can! Cost of genome sequencing continues to drop rapidly which results in many more human genomes being sequenced and a more accurate molecular understanding of human disease. Slide courtesy of R Plenge, Merck
GWAS to RVAS Common genetic variants to rare genetic variants Genome-wide association studies: - Have identified thousands of genetic loci associated with outcome(s) for common and complex traits - These are common variants ( 5% frequency) Missing heritability Sequencing technology (NGS, WGS, WES exome arrays) Rare variant association studies: NGS allows the identification of lower frequency rare genetic variants that has a potential for a larger effect contribution Drug R&D PCSK9 dyslipidemia APP Alzheimer's Etc.
Human genetics helps to identify potential drug targets to start drug discovery Drug Target? But, tens of thousands of potential targets and which one causes disease? Disease Healthy and how do you perturb the target? vs Drug The key steps are: 1. Map genetic differences in those with disease vs healthy; 2. Understand how these genetic differences lead to disease; 3. Develop drugs against these targets that reverse disease processes in the population. Disease Healthy Slide courtesy of R Plenge, Merck
Experiments of nature human knock-outs. There are encouraging examples of this principle working in drug discovery example of PCSK9 Many genes influence cholesterol levels and risk of heart disease Thanks to new technologies, we can now find these disease genes Atherosclerotic Plaque Blood Flow Disease Healthy and design studies to find drugs that fix the underlying molecular defects for example, blocking PCSK9 lowers LDL (or bad ) cholesterol in the blood. PCSK9 LDLR LDL-C mab Lysosome LDLR Recycling Slide courtesy of R Plenge, Merck
Øket satsing på biobanker og helsedata gir win-win Offentlige Bidra til persontilpasset behandling Forbedret helsetjeneste Kunnskap om helsetjenesten og klinisk praksis Bedre helse Verdensledende ressurs for drug discovery og biomarkør utvikling Internasjonal helseindustri investerer i Norge Ny norsk virksomhet etableres Industri Akademia Unik plattform for epidemiologisk og genetisk forskning Publikasjoner Internasjonalt nettverk Økt finansiering
Hindringer for øket industri utnyttelse av norske biobanker Etiske problemstillinger, bla knyttet til samtykke Biobank kvalitet Synlighet og markedsføring Ikke kultur for forskningssamarbeid mellom industri og akademia
Epidemiology Takk for oppmerksomheten! Extensive life style and environmental data Access to external registries e.g. hospital records, prescription and socio economic registries to create a comprehensive picture Longitudinal data allowing long term follow up of outcomes cj@lifandis.com Thank you: www.lifandis.com Population based cohort including e.g. data on family relations Lifandis -