getpersonalized! estonian approach - from biobanking to precision medicine, andres metspalu
TRANSCRIPT
Estonian approach - from biobanking to precision medicine
Andres MetspaluProfessor & DirectorEstonian Genome Center (Estonia)
Estonian approach: from biobanking to precision
medicine
Andres Metspalu, M.D., Ph.DEstonian Genome Center
University of Tartu
HelsinkiMay 25th, 2015
GetPersonalized! Summit in Helsinki
Per Med Definition in EU documents
• Personalised medicine refers to a medical model using molecular profiling for tailoring the right therapeutic strategy for the right person at the right time, and/or to determine the predisposition to disease and/or to deliver timely and targeted prevention
This is the EU H2020 definition of the Per Med.
For personalized medicine we’ll need:
E-infrastructure (e-health, digital prescriptions, digital signature, electronic health records etc.)Public supportPrimary care providers supportPopulation based biobankDNA analysis technologyDatabases on genotype phenotype associations
Estonian Biobank• Estonian Genome Center,
University of Tartu• A prospective, longitudinal,
population-based database with health records and biological materials
• 52,000 participants - 5% of the adult population of Estonia
• Individuals are recruited by GPs, physicians in the hospitals and medical personnel in the EGCUT recruitment offices
• Estonian Human Genes Research Act (HGRA)
§ 3. Chief processor of Gene Bank
• (1) The chief processor of the Gene Bank is the University of Tartu, whose objectives as the chief processor are to:
• 1) promote the development of genetic research;• 2) collect information on the health of the Estonian population and genetic
information concerning the Estonian population; • 3) use the results of genetic research to improve public health.
-0.05
-0.03
-0.01
0.01
0.03
0.05
0.07
-0.05 -0.03 -0.01 0.01 0.03 0.05P2
1 (4
.68%
)
PC1 (8.65%)
Austria Bulgaria Czech RepublicEstonia Finland (Helsinki) Finland (Kuusamo)France Northern Germany Southern GermanyHungary Northern Italy Southern ItalyLatvia Lithuania PolandRussia Spain SwedenSwitzerland
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-0.12 -0.07 -0.02 0.03
PC2
23.8
%
PC1 36.6%
Europe
CHB
JPT
YRI
CEU
Public opinion and awareness of the EGCUT 2001-2014
June, 2001
Sept, 2002
Feb, 2002
March, 2003
March, 2004
Sept, 2004
Dec, 2005
May, 2007
Apr, 2008
July, 2009
June, 2010
April, 2011
July, 2013
Sept, 2014
18%16%
13%
19% 18%16%
22%
33%
28%
36%
59%
55%
67%70%
38% 39% 40%
36%
43%
39%
44%
33% 35%36%
30%33%
27% 27%
4% 4% 3% 4% 4%2% 2% 2% 3% 2% 2% 2% 2% 1%
In favor of the idea of EGCUT
Never heard of EGCUT
Against it
Estonian Government
• 16.12. 2014. the Cabinet approved the Pilot Program of Personalized Medicine”
• Current government has “Personalized Medicine” as one of its activities for the next period
• From 1.03.15. the Ministry of Social Affairs has a new position “Deputy Secretary General for E-services and Innovation”
The Estonian ID card
• The ID card is a mandatory ID document for all Estonian residents from the age of 15
• Enables secure digital authentication and signing• A digital signature has the same legal consequences as a hand-written signature• Does not have any additional information
– No bank account, no health information etc.• Active cards: 1 214 428 (31.12.2013)
– Estonian Population 1 286 540 (01.01.2013)– Estonia has been issuing electronic ID cards from January 1st 2002– Also Mobile-ID
PHARMACIS AND FAMILY DOCTORS2009
X-Roads, ID-card, State IS Service Register
HEA
LTH
CAR
E BO
ARD
- Hea
lth c
are
prov
ider
s- H
ealth
pro
fess
iona
ls- D
ispe
nsin
g ch
emis
ts
STAT
E AG
ENCY
OF
MED
ICIN
ES- C
odin
g Ce
ntre
- Han
dler
s of
med
icin
es
POPU
LATI
ON
REG
ISTE
R
PHAR
MAC
IS20
10 ja
nuar
y
BUSI
NES
S RE
GIS
TER
HO
SPIT
ALS
2009
FAM
ILY D
OCT
ORS
2009
SCH
OO
L N
URS
ES20
10 s
epte
mbe
r
EMER
GEN
CY M
EDIC
AL S
ERVI
CE20
13
NATIONAL HEALTH INFORMATION SYSTEM2008 december
PRESCRIPTION CENTRE2010 january
PATIENT PORTAL2009
XROADS GATEWAY SERVICE2009
Architectural “Big picture”
From Artur Novek (eHealth)
Documents total – 10.8 mio
1.1 mio persons medical data (growth 20% during 2012)
National Patient Portal
15
Acceptance• ePrescription covers 94% of issued
prescriptions.• Over 90% of Hospital discharge letters –
digital• Over 97% of stationary case summaries
have sent to the central e-Health DB1 mio person have documents (82% of population)
Health Insurance Fund
Estonian Myocardial Infarction Registry
Tuberculosis Registry
Estonian Cancer Registry Estonian Causes of Death Registry
Population Registry
Tartu University Hospital
North Estonia Medical Center National Digital Health Record Database
The broad informed consent.Phenotype data
Phenotype database
Coding center High security area.
Data collector:baseline phenotype data (demographics, health, behavior,...)
From questionnaires to the national registries
17
Example of an individual disease trajectory
18
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
1970
1977
1980
1981
1981
1983
1984
1985
1988
1998
1999
1999
1999
2001
2003
2003
2006
2014
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
1970
2008
2008
2008
2008
2008
2009
2010
2010
2010
2010
2010
20102011
2012
2013
2014
Recruitment
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
1970
2010
2010
2014
Male, born 1970, joined 2007
Smoker, BMI=27.5, low physical activity
Health Insurance Fund
NE Med Center
19
2007 2008 2009 2010 2011 2012 20132007
2008
2008
2008
2008
2008
2009
2010
2010
2010
2010
2010
2010
2010
2010
2011
2012
2013
2013
Polygenic risk scores: questions to address
• Which markers to include?– We use 7500 independent markers for prediction of
genetic risk of Type 2 diabetes and 10200 markers for Coronary Artery Disease
• Optimal weights?- We show that optimal risk scores are obtained with double weights: regression coefficients from a large-scale meta-analysis are additionally weighted, to give more weight to more significant markers (correction for „winners curse“)
Polygenic risk scoresCalculated as S = β1X1 + β2X2 + … + βkXk,
X2,…, Xk - allele dosages for k independent markers (SNP-s),
β1, β2, … , βk – weights
Individuals at high genetic risk
Polygenic risk score for type II diabetes: histogram of the score in 7462 individuals (Estonian Biobank)
Score
Fre
qu
en
cy
3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
05
00
10
00
15
00
Practical usefulness of the polygenic risk score?
BMI <25 BMI 25..30 BMI 30..35 BMI >35
T2D prevalence in individuals aged 45-80
0%
5%
10%
15%
20%
25%Below-average (<50%) genetic riskAbove-average (50-90%) genetic riskHighest 10% of the genetic risk
24 27 580
134
52119
198
84
116
154 47
(n=1810) (n=2020) (n=1337) (n=685)
Incident T2D: analysis of 264 incident cases in overweight individuals free of T2D at recruitment
0 1 2 3 4 5 6
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Follow-up time (years)
Cumulative risk of T2D in 3421 individuals BMI at recruitment >24 kg/m2, age 35-74
Highest genetic risk score quintile (>80%)Average risk score quintiles (20-80%)Lowest risk score quintile (<20%)
p 8.310 8
81 cases
p 1.910 3
153 cases
30 cases
No significant sex difference, genetic risk score is the strongest predictor after BMI
(fasting glycose and insulin measurements are not available for this cohort)
Coronary Artery Disease
Males Females
CAD prevalence in individuals aged 40-75
0%
5%
10%
15%
20%
25%Below-average genetic riskAbove-average (50-90%) gen. riskHighest decile of the gen. risk
146
144
52
81
97
34
What is Clinical Decision Support Software (CDSS)?
• CDSS provides clinicians with knowledge presented at appropriate times
• It encompasses a variety of tools such as computerized alerts, clinical guidelines, and order sets
• CDSS has the potential provide the necessary level of personalized guidance to providers at the point of care that will be necessary in the era of genomic medicine
• This is a tool to advise PCP (like radiology report)
Virtuous Cycle of Clinical Decision Support
Measure
Guideline
CDS
Practice
Registry
http://www2.eerp.usp.br/Nepien/DisponibilizarArquivos/tomada_de_decis%C3%A3o.pdf
Approved at the Estonian Government Research and Development Council on 17.12.2013.- Health care - Educating health care professionals - Educating the patients - Further development of the eHealth incl. decision support
systems- Research and Development - Sequencing individuals from the Estonian Biobank (ongoing), decision support analysis software, microarray analyze of 50 000 people from the biobank and return the important information to all who want by the PCP or in hospitalCommercialization - IPR - Business agreements
The Estonian Program for Personal Medicine I
Currently the pre-PILOT PROJECT is underway until the end of June 2015. Then the pilot will last up to 2018.
The Estonian Program for Personal Medicine II
The precision medicine method should be seenbasic as a additional “instrument” for physicians in diagnosing, treating and preventing disease, but also for research and clinical investigations, drug trials etc.
*Committee on a Framework for Developing a New Taxonomy of Disease; ‘Towards Precision Medicine’, National Research Council November 2011
Future
Clinical Care
Research Cohort
From Prof. Böttinger
Evidence Generating Medicine
• The next step beyond evidence-based medicine
• The systematic incorporation of research and quality improvement considerations into the organization and practice of healthcare
• To advance biomedical science and thereby improve the health of individuals and populations.
Challenges and issues
• Awareness executives, doctors and patients• New technologies and data empower patient with
more possibilities to manage own health• Ethical issues
– Right to know and right not to know– Treatable and non-treatable conditions– Big data
• Not enough knowledge about associations between DNA variants and diseases
• Large work-load to keep database of known risk markers updated
Conclusions• Estonia has great potential to plan and implement
personalized medicine solutions for the whole country, starting with the pilot project for 50 000 gene donors
– Genetic research with 5% of population genetic and continuously updated phenotype information
– Nation wide Health Information Exchange platform– 10 years of experience of national level e-services (PKI, X-Road,
ID-card, security framework)– High level public trust and acceptance
Estonian government:25. We will establish a course towards personalised medicine based on modern gene technology.
Timeline: linking to registries
38
2007 2008 2009 2010 2011 2012 2013 2014
Pop reg
03/2010
Pop Reg
11/2010
Pop reg
10/2011
Pop reg
4/2012
Pop reg
1/2013
NDHRD2/2014
Pop reg
3/2014
HIF6/2014
Cancer Reg12/2010
Cancer Reg
6/2012
Tub Reg
12/2012
University Hospital
8/2013
MI Reg
2/2014
North Estonia Medical Center
5/2014
01/2002 - 31/2010
Baseline
1/2011 - 6/2014
2nd visit
1/2003 - ...
Electronic follow-up
2002
Death Reg
9/2010
Death Reg
5/2011
Death Reg1/2012
Death Reg12/2012
Death Reg
8/2013
Death Reg6/2014
Prevented deaths of CHD in Finland
By treating 2144 persons who have >20% risk when the genetic information is known 135 deaths would be prevented in 10 years!
Tikkanen, E. 2013. Genetic risk profiles for coronary heart disease
GD in national registries
• Population Registry 51800 GD99.8%
• Health Insurance Fund 51607 GD99.5%
• Estonian eHealth 39880 GD76.9%
• Tartu University Hospital 22492 GD43.4%
• North Estonia Medical Center 21202 GD40.9%
• Cancer Registry 2644 GD 5.1%
• Causes of Death Registry 2349 GD 4.5%
• Myocardial Infarction Registry 945 GD1.8%
• Tuberculosis Registry 260 GD0.5%
40