Genetic testing: where are we now?

Dr Melody Caramins

National Head, Genetics, Specialist Diagnostic Services

Chair, RCPA Genetics Advisory Committee

Outline

• Definitions: Genetic testing, genomic testing,

mutations and variants

• Massively parallel sequencing (MPS)

• Current implementations of MPS

– Targeted approach

– Whole Genome approach

• Challenges

• Key points for policy

Genetics vs. genomics

Genetic tests – examining on a gene by gene level –e.g. CF

Genomic tests – examining the entirety of the genetic material

Very low resolution genomics

Karyotype

Low resolution genomics

Molecular karyotype

High resolution genomics

Whole genome sequencing

Whole exome = all

known/disease

implicated genes

Gene panels =

All known genes for a

disease

Your genome: what’s included?

• 2.9 billion bases

• 3.3 million variants

• 1.5 million variants in

genes

• ~100 genuine LoF

variants

• 5-20 genuine

pathogenic findings

What is a “mutation”?

• The classical definition of a “mutation” was

once:

– A permanent change in DNA sequence or structure

that is associated with a phenotype (generally

deleterious)

– Change is recognised by comparison of sample and

reference genomes, and is described in relation to

the reference

• Usage has changed

– Can include epigenetic modifications of DNA

– “mutation” is now mostly considered a colloquialism,

with “variant” being preferred.

Why look for variants?

• Can enable specific diagnosis confirmation,

classification, or exclusion

• Can determine appropriate management based

on molecular biological classification of a disease

• Can determine risk of disease

• Can reduce risk of developing disease

• Can lead to recognition of new diseases

Why now? A history of sequencing

• Cost of HGP= 600 000 genomes today at $5000 each

MPS is disruptive innovation

• Disruptive Innovation = Allows popular or

more general access to a product or service

that was historically only accessible to

consumers with a lot of money or a lot of skill.

• Often, the value of a such technologies is

initially dismissed (e.g. the PC)

• Institutions are often blindsided as the

technology matures and suddenly gains a

larger audience, e.g. email, mobile computing/devices,

Wikipedia.

Current implementation models

• Targeted

– Panels (handful hundreds of genes)

– Exome

• Whole genome sequencing

Targeted sequencing – gene panels

• Pros

– Tractable - testing using targeted panels is

currently available both in Australia and overseas

– Clinical utility is well established for many

disorders

– Minimises risk of “incidental findings”

– Genes previously considered untestable can now

be included

– Variants identified in other related genes which

modify phenotype and prognosis

– Avoid “diagnostic odyssey”

Targeted sequencing – gene panels

Cons:

• As gene panel sizes

increase, clinical

sensitivity and diagnostic

yield increases

• However, this

improvement is tempered

by an emerging

interpretive challenge,

with inconclusive results

increasing up to 10 fold

• Unknown causative genes

not included

• ~40Mb (coding) or 60Mb (coding + UTRs)

• Greatest application in gene identification in research,

or diagnosis of unknown rare genetic disorders

Targeted sequencing - exomes

Inherited Mendelian Diseases Identified

by Exome Sequencing

Rabbani, B., et al. (2012) J. Hum. Genet. 57:621-632.

By mid-2012, ~100 genes identified.

By mid-2013, >150 genes identified.

Evaluated 500 patients

8% not genetic

92% Traditional testing model

42% diagnosed, majority at 1st

clinical consultation

50% undiagnosed. Conservative cost $25 000/ diagnosis

25-50% of this cohort could be diagnosed by

WES/WGS

(fraction of the cost)

clinically and economically viable

alternative

Whole genome sequencing

• Challenge of appropriate interpretation

• Enormous amount of genetic variation – both

inter-population and inter-individual, with

many international collaborative efforts to

catalogue variation – HapMap, DGV, HVP,

1000 genomes, ClinVar, etc...

• Spectrum of variation effect from:

no change in

discernible

phenotype

alternative

phenotypes

with no medical

consequence

medical consequences of varying

severity and penetrance

disease susceptibility→ pathogenic

Variants of unknown significance (VUS)

Clinical interpretation & pathogenicity

Recent publication evaluating coverage and concordance of clinically relevant genetic variation provided by WGS in 12 patients:

• 10-19% of inherited disease genes were not covered to acceptable standards

• Good platform concordance for SNV but poor for indel variants

• 90-127 variants/individual requiring ~55 min curation each, 2-6 disease causing variants/individual

• Variant classification agreement = kappa 0.52

• 1-3 follow-up diagnostic tests or referrals

Image by D.Shrigley

Summary of approaches

Targeted Panel

(few-hundreds

genes)

Whole Exome

(~90% coverage

of 20 000 genes)

Whole genome

Size 500kb – 3 Mb 45-62 Mb 3.1 Gb

Advantages Greater sensitivity,

Greater specificity

High clinical utility

Easier analysis

Quick and cost

effective

Higher chance of

finding a mutation

in all the protein-

coding regions

Covers everything

Disadvantages Pathogenic mutations

could be in genes not

covered

High drop out rate

of some exons and

incomplete

coverage (~80-

90%)

Very challenging

and expensive

analytics

Success rate High Good but complex

and challenging to

implement

In nascent phase

of utility evaluation

for diagnostics

Wright C BMJ 2013

Which approach?

The new genomics will save money?

• Paradigm: a less expensive test/procedure, utilisedin a larger swath of indications.

• A previous example: laparascopic cholecystectomy.– Hypothesised savings due to shorter hospital stays,

decreased complications

– Increased costs due to 20% increase in surgery uptake.

• Genomics will lower costs?– Pharma companies may increase prices due to reduced

target market

– Physicians may continue to treat inappropriately (although not borne by MBS if clear indications for rebate)

– Cost of follow up confirmatory testing in inherited disease, and testing of relatives?

Things change….

Things change…

• The use of genomic technologies (genome-

wide arrays or WGS) requires a justification

in terms of necessity (the need to solve a

clinical problem) and proportionality (the

balance of benefits and drawbacks for the

patient).

Challenges (not a comprehensive list)

Technical

How much to sequence

How complete is coverage

Scalability of interpretive practices

Ethical

Who to sequence

Incidental findings –report, or

not?

Biological

Which sequence

variant findings are relevant and contribute to

current diagnosis?

Which sequence

variant findings are relevant to clinical care

but not necessarily

to presenting diagnosis?

(clinically actionable

incidental data)

Informatics

What sequence

information should be

stored

Where can it be stored,

who should have access

to it?

• How do we most responsibly communicate

results to patients and their doctors?

• How do we best balance individuals’ right to

privacy/security of health information vs.

community benefit of sharing information?

• Many of the issues are not entirely new, but the

scale of the challenges is

• Our frameworks and guidelines for good clinical

services are based on experiences to date;

these are surpassed by the amount of

information available

– Should current frameworks dictate developments?

– Should existing frameworks be reconsidered?

Differing international opinionsACMG ESHG

Diagnostic laboratories should

routinely screen all clinical

exomes/genomes for a list of

known variants in genes

associated with medically

important conditions

Preferable to use a targeted

approach to avoid unsolicited or

uninterpretable findings.

Genomic screening is not

specifically advocated. Variants

with limited/no clinical utility

should be filtered out (neither

analyzed nor reported

Patients cannot opt out of

genetic screening, it is the

responsibility of the clinical team

to provide appropriate pre and

post test counseling

Guidelines for informed consent

need to be developed, but the

individual’s right not to know

does not automatically over-ride

professional responsibilities

The genomes of minors should

be screened for variants offering

clinical utility to their parents

Guidelines for minors need to be

developed regarding what

unsolicited information should

be disclosed

Key points of awareness for policy

• Individuals vary in their tolerance for uncertainty

• Individuals vary in their understanding of predictive

value (or lack thereof) of genomics

• Autonomy of patients should be balanced with

autonomy of relatives

• Resolve whether healthcare providers have a duty

to search for and act on clinically actionable

variants that are not directly pertinent to the clinical

question

• Resolve whether healthcare providers have a duty

to interpret data with uncertain predictive value

THANK YOU!

Overview of the MPS process

• Aim to provide improved constitutional and

somatic mutation detection in diagnosis and

research

1 2 3 4 5 6

Assay

development

DNA

extraction

Library

preparation

Sequence

generation

Primary

data

analysis

data

analysis

Secondary

Primer design

Primer validationTarget enrichment

Sample indexing

-SOLiD 5500 XL

-Ion PGM

-OtherImage or signal

processing

Mapping

SNP calling

SV

MDT meetings

Altman A, Weber P et al

Human Genetics 2012

• Next Generation Sequencing Data

• Primary Sequence Alignment• BWA

• Refined Sequence Alignment• GATK/Picard

• Variant Calling• SAMTools/GATK

Variant Annotation

• Annovar

Candidate genes/potential causative variants

Ashley et al. Lancet 2010, 375:1525

An approach to comprehensive analysis of

a human genome in a defined clinical

context

Kaiser J. Science (2012) 338:1016-1017.