BLINDBILD
Olaf Riess, MD
Institute of Medical Genetics and Applied Genomics, Tübingen, Germany
State of Play
Diagnosing Rare Diseases
Diagnostic barriere:
Routine work up in hospitals
2
• Clinical investigation by the doctor
• Routine blood and urine tests
• Imaging
• (Biopsies)
How and where does genetics fits in ?
Diagnostic Used to confirm a diagnosis based on physical signs
Predictive Used to detect gene mutations associated with disorders that
appear later in life
Includes population screening for treatable conditions
Carrier
Identification
Used by people with a family history of recessive genetic disorders
Prenatal Used to test a foetus when there is risk of bearing a child with mental
or physical disabilities
Newborn
Screening
Used as a preventative health measure once the baby is born
Forensic testing Used to identify an individual for legal purposes
Research
testing
Used for finding unknown genes and identifying the function of a
gene
Genome sequencing is expected to have the most
impact in:
 stratifying (better categorizing) patients for
appropriate cancer treatment
 diagnosing persons at risk of (treatable) genetic
disease to apply prevention concepts
 providing information about an individual’s likely
response to treatment to reduce adverse drug
reactions
Therapeutic relevance ?
Health insurance companies:
We do not need genetic analysis at all, as they have
limited diagnostic value for treatment, and
family testing is not a declared goal of the insurance.
Rare Neurometabolic Disorders
Orotic aciduria
MIM #258900
Therapy!!
Tobias Haack
Rare Neurometabolic Disorders
• A four-year-old girl, previously in a minimally conscious
state, began to communicate and walk with assistance
after nine weeks of treatment
• A three-year old girl showed developmental progress
• Immediate cessation of seizures in both
1
10
100
30 36 42 48 54 60 66
Seizure-free since
5 months
Age [month]
Tobias Haack
Rare Neurometabolic Disorders
Therapeutic Impact87 Cases (out of 1900 Exomes)
Lucia Laugwitz und Tobias Haack
Therapeutic relevance of
mutation type?
Olaf Riess, Medizinische
Genetik Uni Tübingen
Cystic Fibrosis
Mutation = Directing treatment
• Class I defects (eg G542X) disrupt synthesis of CFTR and include
nonsense and frameshift mutations that lead to premature termination
codons (PTCs) and a lack of protein production.
• Class II defects (eg F508del) result in misfolded CFTR that is then
degraded in the endoplasmic reticulum.
• Class III defects (eg G551D) result in CFTR that reaches the apical
membrane but is not activated and is therefore nonfunctional.
• Class IV defects (eg R117H) result in reduced conductance of CFTR at the
cell surface.
• Class V defects (eg A455E) result in overall reduced synthesis of normal
CFTR.
> Agents that promote ribosomal read-through of
nonsense mutations (eg ataluren, nonsense
mediated mRNA decay)
> ‘CFTR correctors’ (eg lumacaftor improves trafficking of
CFTR through cell, F508del)
> ‘CFTR potentiators’ (eg ivacaftor increases the time
the channel opens; R177H, G551D)
Small molecule CFTR modulators
Therapeutic relevance
Predict - Prevent
0
20
40
60
80
100
0 to 1 2 to 5 6 to 10 11 to 17 18 to 24 25 to 34 35 to 44 45+
Age (years)
Percent
Cystic Fibrosis Foundation Patient Registry Data. 2005
P. aeruginosa
S. aureus
MRSA
H. influenza
S. maltophilia
B. cepacia
Olaf Riess, Medizinische
Genetik Uni Tübingen
Exome sequence data for Dynactin 4 (DCTN4) revealed that 12/43 individuals
in the P. aeruginosa extreme had a missense variant in DCTN4, 9 were
heterozygous at position 150097883 (rs11954652; Phe349Leu) and 3 at
position 150110239 (rs35772018; Tyr270Cys). None of the 48 individuals in
the late P. aeruginosa extreme had missense variant‘s in DCTN4.
Nat Genet 2013
15
Accomplishment across five domains of genomic research
Green et al. 2011, Nature
 Lack of evidence of effectiveness of genomic interpretation
 Lack of expertise and training programs in genomics and genome
analytics, but most of all in other medical disciplines
 Lack of man power (in genomics and in other clinical disciplines)
 Lack of SOPs, structure and guiding in daily routine
 High costs and lack of reimbursement
 Need for quality controled data bases with genomic variants
linked to clinical phenotypes
 Lack of infrastructure to order, receive and act on follow up
research to define impact on clinical interventions
 Ethical and legal aspects of ownership of genomic information
and transfer of data
EUROPEAN REFERENCE
NETWORKS
Complementary
Concept:
Organ focused
18
DIAGNOSTICS
New Interdisciplinary Concept
New Infrastructure
Hurdles in diagnostics
Olaf Riess, Medizinische
Genetik Uni Tübingen
Case report: 25 years old male patient
complex hereditary paraplegia with axonal polyneuropathy
MRI normal
Most frequent isoforms SPG4, SPG5, SPG3 excluded…
Gene panel of 62 genes causing paraplegic phenotype
KIF5A = SPG10 heterozygote p.R204W (VUS5) pathogenic mutation
Example 1: Decision making in diagnostics
This mutation has been described in autosomal dominant and
autosomal recessive SPG10.
What is the risk of his future children?
Should we sequence KIF5A in his wife?
Olaf Riess, Medizinische
Genetik Uni Tübingen
Case report: 25 years old male patient
complex hereditary paraplegia with axonal polyneuropathy
MRI normal
Genetically SPG4, SPG5, SPG3 excluded…
Gene panel of 62 genes causing paraplegic penotype
Additionally, compound heterozygote for GCH1 p.P23L/p.P69L
Indicating dopa-responsive Dystonia
Both mutations have been described in cis and trans.
If in trans, the patient has also dopa-responsive dystonia
and may require treatment - or he will develop symptoms later
Should we test his wife?
Example 1
KIF5A = SPG10 heterozygote p.R204W (VUS5) mutation
Olaf Riess, Medizinische
Genetik Uni Tübingen
Example 2: Hurdles of data bases / know ledge
Conclusions
„Curated“ data bases are necessary
• to avoid wrong diagnosis
• to advance diagnostics for instance by
implementing machine learning
24Heather Mason-Suares, Harvard Medical School
Olaf Riess, Medizinische
Genetik Uni Tübingen
Problems: - 27% of all data base entries are wrong for their clinical
implication (Kingsmore et al.)
- ~8% of the population are homozygotes or compound
heterozygotes for loss of function mutations
with minor allele frequencies below 2% in1171
genes (complete knockouts)
(Sulem et al. 2015 DeCode Nat Genet 47)
Disadvantage: - How to deal with „unwanted“ results
Advantages: - Identification of novel disease genes even in single
patients possible without major ressources
- Diminish categories of clearly defined phenotypes
- Discovery of at least 5% wrong diagnosis of current
genetic reports
26
Shyr and Liu 2013
Example 3: Technical hurdles in diagnostics
Park et al. AJHG, 2018
29
Neuromics: Statistics on WES/WGS
5%
7%
25%
19%
44%
WES
30%
4%
17%16%
33%
WES & WGS
Samples not yet analysed
by clinical centre
Cases solved (new genes)
Cases solved (known
genes)
Cases with VUS or novel
candidates
Cases unsolved
Statistics generated from data of 763 samples (80%)
59%
1%
7%
12%
21%
WGS
Example 4: Hurdle of sensitivity in diagnostics
50% of all patients
with a rare disease
will not have
access to health
care without
having a clear
diagnosis
150 Mio patients world wide
15 Mio patients in Europe
1.5 Mio patients in Germany
0.2 Mio patients in Jordan
Genome
Transcriptome
Epigenome
MetabolomeProteome
Cellular/
molecular
phenotypes
Exome
Solving the
unsolvable
diseases
Challange in Diagnostic Transition:
From genome analysis towards
system diagnostics
Example 5: Technological hurdles in diagnostics
Coordinators:
Olaf Riess & Holm Graessner
„Pilot Project“
Numbers Complexity Diagnostic yield Treatment Knowledge
6000-8000
different
diseases
Affect
different
organs
Complex data
sets
High
development
al demand
How many
patients/
Where?
30 Mio
patients in
Europe
Chronic
diseases
Still focussed on
Genetics
Strong
growth only
in cancer
Fragmented,
dependent
on single
experts
15 Mio RD
„unsolved“
by WES
Multi-
morbidity
MultiOmics
diagnostics
High costs No SOPs
From single
physicians
to Centers
and
networks
Inter-
disciplinary
New
structure in
health care
New type of
diagnostic
centers,
machine
learning and AI
concepts
P-P-
Partnership,
Basic and
Applied
Research
ERNs =>
integration
of diagnostic
expert
groups
Olaf Riess
olaf.riess@med.uni-tuebingen.de

The State of Play in Diagnosis

  • 1.
    BLINDBILD Olaf Riess, MD Instituteof Medical Genetics and Applied Genomics, Tübingen, Germany State of Play Diagnosing Rare Diseases
  • 2.
    Diagnostic barriere: Routine workup in hospitals 2 • Clinical investigation by the doctor • Routine blood and urine tests • Imaging • (Biopsies) How and where does genetics fits in ?
  • 3.
    Diagnostic Used toconfirm a diagnosis based on physical signs Predictive Used to detect gene mutations associated with disorders that appear later in life Includes population screening for treatable conditions Carrier Identification Used by people with a family history of recessive genetic disorders Prenatal Used to test a foetus when there is risk of bearing a child with mental or physical disabilities Newborn Screening Used as a preventative health measure once the baby is born Forensic testing Used to identify an individual for legal purposes Research testing Used for finding unknown genes and identifying the function of a gene
  • 4.
    Genome sequencing isexpected to have the most impact in:  stratifying (better categorizing) patients for appropriate cancer treatment  diagnosing persons at risk of (treatable) genetic disease to apply prevention concepts  providing information about an individual’s likely response to treatment to reduce adverse drug reactions
  • 5.
    Therapeutic relevance ? Healthinsurance companies: We do not need genetic analysis at all, as they have limited diagnostic value for treatment, and family testing is not a declared goal of the insurance.
  • 6.
    Rare Neurometabolic Disorders Oroticaciduria MIM #258900 Therapy!! Tobias Haack
  • 7.
    Rare Neurometabolic Disorders •A four-year-old girl, previously in a minimally conscious state, began to communicate and walk with assistance after nine weeks of treatment • A three-year old girl showed developmental progress • Immediate cessation of seizures in both 1 10 100 30 36 42 48 54 60 66 Seizure-free since 5 months Age [month] Tobias Haack
  • 8.
    Rare Neurometabolic Disorders TherapeuticImpact87 Cases (out of 1900 Exomes) Lucia Laugwitz und Tobias Haack
  • 9.
  • 10.
    Olaf Riess, Medizinische GenetikUni Tübingen Cystic Fibrosis Mutation = Directing treatment
  • 11.
    • Class Idefects (eg G542X) disrupt synthesis of CFTR and include nonsense and frameshift mutations that lead to premature termination codons (PTCs) and a lack of protein production. • Class II defects (eg F508del) result in misfolded CFTR that is then degraded in the endoplasmic reticulum. • Class III defects (eg G551D) result in CFTR that reaches the apical membrane but is not activated and is therefore nonfunctional. • Class IV defects (eg R117H) result in reduced conductance of CFTR at the cell surface. • Class V defects (eg A455E) result in overall reduced synthesis of normal CFTR. > Agents that promote ribosomal read-through of nonsense mutations (eg ataluren, nonsense mediated mRNA decay) > ‘CFTR correctors’ (eg lumacaftor improves trafficking of CFTR through cell, F508del) > ‘CFTR potentiators’ (eg ivacaftor increases the time the channel opens; R177H, G551D) Small molecule CFTR modulators
  • 12.
  • 13.
    0 20 40 60 80 100 0 to 12 to 5 6 to 10 11 to 17 18 to 24 25 to 34 35 to 44 45+ Age (years) Percent Cystic Fibrosis Foundation Patient Registry Data. 2005 P. aeruginosa S. aureus MRSA H. influenza S. maltophilia B. cepacia
  • 14.
    Olaf Riess, Medizinische GenetikUni Tübingen Exome sequence data for Dynactin 4 (DCTN4) revealed that 12/43 individuals in the P. aeruginosa extreme had a missense variant in DCTN4, 9 were heterozygous at position 150097883 (rs11954652; Phe349Leu) and 3 at position 150110239 (rs35772018; Tyr270Cys). None of the 48 individuals in the late P. aeruginosa extreme had missense variant‘s in DCTN4. Nat Genet 2013
  • 15.
    15 Accomplishment across fivedomains of genomic research Green et al. 2011, Nature
  • 16.
     Lack ofevidence of effectiveness of genomic interpretation  Lack of expertise and training programs in genomics and genome analytics, but most of all in other medical disciplines  Lack of man power (in genomics and in other clinical disciplines)  Lack of SOPs, structure and guiding in daily routine  High costs and lack of reimbursement  Need for quality controled data bases with genomic variants linked to clinical phenotypes  Lack of infrastructure to order, receive and act on follow up research to define impact on clinical interventions  Ethical and legal aspects of ownership of genomic information and transfer of data
  • 17.
  • 18.
  • 19.
  • 20.
    Olaf Riess, Medizinische GenetikUni Tübingen Case report: 25 years old male patient complex hereditary paraplegia with axonal polyneuropathy MRI normal Most frequent isoforms SPG4, SPG5, SPG3 excluded… Gene panel of 62 genes causing paraplegic phenotype KIF5A = SPG10 heterozygote p.R204W (VUS5) pathogenic mutation Example 1: Decision making in diagnostics This mutation has been described in autosomal dominant and autosomal recessive SPG10. What is the risk of his future children? Should we sequence KIF5A in his wife?
  • 21.
    Olaf Riess, Medizinische GenetikUni Tübingen Case report: 25 years old male patient complex hereditary paraplegia with axonal polyneuropathy MRI normal Genetically SPG4, SPG5, SPG3 excluded… Gene panel of 62 genes causing paraplegic penotype Additionally, compound heterozygote for GCH1 p.P23L/p.P69L Indicating dopa-responsive Dystonia Both mutations have been described in cis and trans. If in trans, the patient has also dopa-responsive dystonia and may require treatment - or he will develop symptoms later Should we test his wife? Example 1 KIF5A = SPG10 heterozygote p.R204W (VUS5) mutation
  • 22.
    Olaf Riess, Medizinische GenetikUni Tübingen Example 2: Hurdles of data bases / know ledge
  • 23.
    Conclusions „Curated“ data basesare necessary • to avoid wrong diagnosis • to advance diagnostics for instance by implementing machine learning
  • 24.
  • 25.
    Olaf Riess, Medizinische GenetikUni Tübingen Problems: - 27% of all data base entries are wrong for their clinical implication (Kingsmore et al.) - ~8% of the population are homozygotes or compound heterozygotes for loss of function mutations with minor allele frequencies below 2% in1171 genes (complete knockouts) (Sulem et al. 2015 DeCode Nat Genet 47) Disadvantage: - How to deal with „unwanted“ results Advantages: - Identification of novel disease genes even in single patients possible without major ressources - Diminish categories of clearly defined phenotypes - Discovery of at least 5% wrong diagnosis of current genetic reports
  • 26.
    26 Shyr and Liu2013 Example 3: Technical hurdles in diagnostics
  • 28.
    Park et al.AJHG, 2018
  • 29.
    29 Neuromics: Statistics onWES/WGS 5% 7% 25% 19% 44% WES 30% 4% 17%16% 33% WES & WGS Samples not yet analysed by clinical centre Cases solved (new genes) Cases solved (known genes) Cases with VUS or novel candidates Cases unsolved Statistics generated from data of 763 samples (80%) 59% 1% 7% 12% 21% WGS Example 4: Hurdle of sensitivity in diagnostics
  • 30.
    50% of allpatients with a rare disease will not have access to health care without having a clear diagnosis 150 Mio patients world wide 15 Mio patients in Europe 1.5 Mio patients in Germany 0.2 Mio patients in Jordan
  • 31.
    Genome Transcriptome Epigenome MetabolomeProteome Cellular/ molecular phenotypes Exome Solving the unsolvable diseases Challange inDiagnostic Transition: From genome analysis towards system diagnostics Example 5: Technological hurdles in diagnostics
  • 32.
    Coordinators: Olaf Riess &Holm Graessner „Pilot Project“
  • 33.
    Numbers Complexity Diagnosticyield Treatment Knowledge 6000-8000 different diseases Affect different organs Complex data sets High development al demand How many patients/ Where? 30 Mio patients in Europe Chronic diseases Still focussed on Genetics Strong growth only in cancer Fragmented, dependent on single experts 15 Mio RD „unsolved“ by WES Multi- morbidity MultiOmics diagnostics High costs No SOPs From single physicians to Centers and networks Inter- disciplinary New structure in health care New type of diagnostic centers, machine learning and AI concepts P-P- Partnership, Basic and Applied Research ERNs => integration of diagnostic expert groups
  • 34.