Genomic Testing for Type 2 Diabetes Risk: A Prototype for Personalized Preventive Medicine?

1
Genomic Testing for Type 2 Diabetes Risk A
Prototype for Personalized Preventive Medicine?

• Alex Cho MD, MBA Scott Joy MD Julianne ODaniel
  MS, CGC Ley Killeya-Jones, PhD Susanne Haga
  PhD Isaac Lipkus PhD Geoff Ginsburg MD, PHD
• Center for Genomic Medicine (IGSP)
• Duke University

2
Clinical Vignette
40 yo M, BMI lt30, no FH DM, not from higher-risk
group wants screening for diabetes
3
Genomics, Direct-to-Consumer
4
Why Prevention?

• Genomic discovery finding new associations with
  disease risk
• Genomics can deliver what prevention requires
• Conditions w/ significant burden of suffering
• Conditions w/ suitable natural history
• Acceptable screening procedures
• Treatments that work better early than late
• Benefits outweigh harms
• Benefits come at reasonable cost
• Genomics thus a potentially powerful tool to help
  rationalize imprecise practice

5
At Reasonable Cost?

• Multiplex technologies have built-in economies of
  scale
• 1,000 for 1,000,000 SNPs 0.1 per SNP no
  expiration date
• even on a per-condition basis, price is
  reasonable
• vs. CT scan of the head for HA 425, good for
  12h?
• variant on 9p21 assoc w/ incr risk of
  intracranial aneurysm (OR 1.3) price 195,
  includes VTE, AF too
• Cost issue is as much about the orientation of
  our healthcare system as it is about the actual
  cost of these technologies

6
Why Diabetes?

• Huge burden
• Onset can be delayed, even prevented
• Knowledge isnt enough, but can be motivating
• Current screening practice imprecise
• FPG most common will be replaced by Hgb A1c
• USPSTF recs screening only those w/ HTN, ?chol
  ADA recs screening more broadly
• NHANES III found FPG alone misclassifies 20 of
  patients w/ diabetes, prediabetes as being
  nondiabetic
• Oral glucose tolerance testing (OGTT) is gold
  standard, but more expensive and less convenient
• Clinical inertia around borderline results
• Well-studied markers (20 SNPs), GWA studies
• Even an RCT showing benefit (DPP)

7
TCF7L2

• TCF7L2 encodes for entero-endocrine transcription
  factor
• Role in Wnt signaling pathway

• Regulates peptide hormone made by enteroendocrine
  cells (glucagon-like peptide 1)

8
TCF7L2 Diabetes Prevention Program

• Diabetes Prevention Program (DPP) not only
  showed association with risk of progression, but
  also that intervention reduces risk for
  higher-risk genotypes.

Source Florez et al. N Engl J Med
2006355241-50.
9
TCF7L2 Diabetes Prevention Program
Source Florez et al. N Engl J Med
2006355241-50.
10
Potential impact on patient behavior?

• We know that knowing is not enough
• Adherence to lifestyle changes proven to reduce
  risk for T2DM poor
• Family history motivating for some
• e.g., study of 1100 African Americans found those
  aware of FH T2DM were more likely to make
  healthier food choices
• REVEAL study
• finding of ApoE4 led to increased AD-specific
  behavior change
• Survey of patients and physicians re their
  enthusiasm for the use of genetic information for
  T2DM risk
• 71 of patients said this information would be
  motivating
• 23 of providers said it would

Source Baptiste-Roberts et al. 2007 Chao et al.
2008 Florez et al. 2009
11
Issue 1 What does this add to what we already
know?

• Recent studies suggest the addition of genomic
  testing to standard risk assessment adds little
  to risk prediction
• Meigs et al. and Lyssenko et al. found that the
  contribution of specific genetic information only
  slightly improved upon the ability of a
  constellation of other clinical factors to
  predict who would progress to T2D.
• These factors included systolic blood pressure,
  high-density lipoprotein levels, and
  triglycerides in the former and diastolic blood
  pressure, triglycerides, liver enzymes in the
  latter.
• However, Lyssenko et al. also pointed out that
  the risk prediction model from Meigs et al.
  performed worse than one based on genomic risk
  alone.

Source Meigs et al. N Engl J Med 2008 Lyssenko
et al. N Engl J Med 2008.
12
Issue 2 What if it contradicts what we
already know?
13
Issue 3 What happens when risk estimates change?
Source http//exploringmygenes.blogspot.com
14
A Three-Part Approach

• Identify markers (i.e., SNPs)
• Systematic review (TCF7L2, PPARg2, KCNJ11)
• deCODE T2D (TCF7L2, PPARg2, CDKAL1, CDKNA2A/B)
• Build a clinical prototype
• Duke Executive Health module
• Other Duke clinics (e.g., Pickett)
• Build a research program
• Clinical utility RCT pilot
• CHSRPC (Durham VAMC) study
• Multiplex pilot??

15
deCODE T2TM test for T2DM risk in 1? care

• Panel of 4 SNPs associated w/ increased risk of
  developing T2DM
• Retails for 300
• Analyzed in a CLIA-certified lab
• Can only be ordered by physician

Source deCODE genetics.
16
T2DM Genomic Risk Pilot Study
17
Risk Profile
18
Role for DNA testing in DM risk assessment?
annual OGTT (or Hgb A1c) screening stepped-up
lifestyle change
initiate early treatment w/ metformin
elevated risk
prediabetic?
screening w/ FPG every 3y if physically inactive,
?45, ? chol, HTN usual lifestyle recs
urge lifestyle change, consider early treatment
w/ metformin
baseline risk
40 yo M, BMI lt30, no FH DM, not from higher-risk
group wants screening for diabetes
OGTT oral glucose tolerance testing, FPG
fasting plasma glucose
19
Acknowledgements

• Study Team
• Ley Killeya-Jones
• Marylou Bembe
• Dana Baker
• Michael Scott
• Sarah McBane
• Gloria Trujillo
• The Duke Endowment
• deCODE genetics

20
Resources

• National Human Genome Research Institute (NIH)
• www.genome.gov
• National Office of Public Health Genomics (CDC)
• HuGE Net
• Natl Coalition for Health Professional Education
  in Genetics (NCHPEG)
• Gene Tests (www.genetests.org)
• Online Mendelian Inheritance in Man (OMIM)
• HapMap (www.hapmap.org/whatishapmap.html)
• Wellcome Trust (genome.wellcome.ac.uk)
• Guilford County Genomedical Connection
  (http//www.aheconnect.com/genomic_medicine)
• Duke IGSP (www.genome.duke.edu)

21
Thank You
22
(No Transcript)
23

• Additional Slides

24
Genomics, Direct-to-Consumer