Professor Stephen

RobertsonCure Kids Chair

Child Health Research

University of Otago

Dunedin

14:15 - 14:30 Direct to the Consumer Genetics in Primary Care - An

Approach

Direct to the Consumer Genetics in Primary Care - An Approach

Stephen RobertsonCurekids Professor of Paediatrics

University of OtagoNEW ZEALAND

Individual genomic profiling for multiple health-related traits

Many tests offered online: “direct-to-consumer”

Pharmacogenetics

Nutrigenetics

“Genomic wellness”

Direct to Consumer TestingThe ground rules

• Patient initiates the test

• Company regulated as consumer product

• Patient controls and manages the data

• Patient often chooses a third party to interpret the data

• Information interpreters may not be accredited

• Test quality may be unregulated

• Multiple questions being asked by the customer

• Data may be sold for secondary use.

Direct to Consumer TestingCategories of Genomic Test

Health related traitsPenetrant Mendelian Diseases (includes carrier status)Common morbiditiesPharmacogenetics

Lifestyle related factorsDiet and food preferencesExercise regimens(Traits exhibit complicated interactions between genes and environmental

factors; Evidence is lagging behind provision of such testing)

Ancestry(Accuracy contingent on size and breadth of population databases)

Direct to Consumer TestingCategories of Genomic Test

Health related traitsPenetrant Mendelian Diseases (includes carrier status)Common morbiditiesPharmacogenetics

Direct to Consumer TestingCategories of Genomic Test

Health related traits

1. Penetrant Mendelian Diseases (includes carrier status)BRCA1/2, haemochromatosis, 𝛂1-antitrypsin deficiency

carriers for cystic fibrosis….dozens of recessive diseases,

2. Common morbiditiesDementia, Parkinsons disease, type 2 diabetes, coronary artery disease

3. Pharmacogeneticsantithrombotics, anti-depressants, statins……

Managing personal genomic testing in general practice?Advising patients who want Personalised Genomic Testing

Step 1 The questionDo they have symptoms, signs or a family history of the condition?

Consider definitive testingIs the trait mendelian or have multiple genetic or environmental determinants?

For the former, risks usually only rise or fall incrementally after testingIs their aim to rule out a given diagnosis or predisposition?

Very rare; most risks are just modified; common alleles only testedAre they at population risk for a suite of diseases being considered?

Absolute risks will be altered only minimally in the vast majority

Step 2 Consequences and potential adverse outcomesTesting may impact ability to obtain life and health insuranceIs the laboratory accredited? Will they interpret the results?

Confusion a frequent experience for most. Third party providers variable qualityWill the data be retained (and sold on) to third parties?

A prime motivator for larger companies

Outcomes and Results - common scenarios - 1Mendelian Disorders

Off the Shelf ScreensFor focused assessment of single conditions

- family history, segregation key: recommend a local providerA screen that excludes the presence of any pathogenic mutation is very, very rare

- residual risk remains. Risk of false reassurance highMost tests look for common causal variants - they are not comprehensive.Specific disease-causing variants sought for some tests (e.g. BRCA1/2 mutations, haemochromatosis, macular degeneration).

-Referral to genetics services is appropriate if the provider is validated

Products that screen for disease-causing mendelian variants• Searching for pre-specified known causative variants for Mendelian Disorders• Very long list of rare conditions

Outcomes and Results - common scenarios - 1Common Disorders

Traits:• Diabetes• Heart disease• Cancer susceptibility• ………..

Products that screen for variants statistically associated with disease statesTens, if not hundreds of variants for each disease; each individually weakTests only measure a fewEstimates of risk that they confer vary wildly between providers

A few tests are useful:• Coeliac disease (for exclusionary purposes)• Macular degeneration (especially if family history)

Common MorbiditiesPatients seeking a “genetic warrant of fitness” for the big killers

(typical examples: Parkinsons, Alzheimer)

• Summed genetic contributions to these diseases are large

• Named explanatory genetic factors are few

• Tests measure a very small number of these factors

• Each genetic variant typically has a weak effect

• These results report very small increases in absolute risk

• Some tests have little evidence for clinical validity (e.g. MTHFR variants for miscarriage, vascular disease, thrombophilia).

A very small number of variants confer large explanatory effects but caveats remain

Late-onset Alzheimer disease and ApoE4 genotyping– We all have two ApoE alleles: ApoE2, 3 or 4.

– 1% of the population is ApoE4/ApoE4 homozygous.

– ApoE4/ApoE4 homozygotes have increased risk of developing AD (~ 12-fold)

BUT

– Positive predictive value of E4/E4 genotype is low ( penetrance).

– Negative predictive value is limited: 42% with AD don’t have an ApoE4 allele

– Example of an FDA accredited DCT that leaves a high degree of uncertainty.

– Not generally offered in the public sector.

Direct to Consumer Genomics and Prescribing

Relling and Evans (2015)

The evidence behind actionability

• Warfarin – conflicting studies of clinical utility – CYP2C9, VKORC1

Ancestry specific effects clear

• Carbamazepine - hypersensitivity (risk ~10%)– HLA-A*31:01 (NNT caucasian 47)

– HLA-B*15:02 Toxic epidermal necrolysis in SE Asians

• Clopridogel– variants that confer loss and gain of function

– LoF alleles common ++ in Maori and Pacific Islanders clinical in-efficacy (as judged by in vitro platelet assays)

– a priori case for pharmacogenomics in clinical practice

– Still requires systematic clinical evaluation

Direct to the Consumer Genetic TestingThe Report Card

Analytic validity – does the test measure the genetic change of

interest well? ✓ (some 3rd party providers, ✘)

Clinical validity - how well the variant being tested correlates with

risk of disease? ✓✘

Clinical utility - is it medically useful ✓✘

Personal utility - is it useful even if it doesn’t change medical

management? ✓✘

Will the market improve?

– Consumer driven testing with health professional involvement

– Success: low-yield pick-up of rare penetrant traits with high positive predictive value

– Essentially a validated screening programme to define risk

A litmus test question to ask:Is the genetic test I am interested in available through any mainstream

health system globally for someone in my clinical situation?

If so, is cost the barrier? - good wrap-around options are availableIf not: Does the offering fail on validity and utility grounds?

Thank you