The Genetics of Type 2 Diabetes Learning the Right Lessons NCHPEG 12th Annual Meeting Jim Evans MD, Ph.D University of NC at Chapel Hill The Promise of Genetic Medicine “But remember throughout that no external cause is efficient without a predisposition of the body itself. Otherwise, external causes which affect one would affect all…” We now have the ability to analyze the individual’s genome deeply and define medically important variation Diabetes is an Environmental Disorder Worldwide Prevalence of Diabetes for 2000 and 2010 Zimmet et al. Nature 414:782.2001 Environmental Causation of Diabetes Obesity = BMI > 30 Mokdad et al., 2001, JAMA, 286: 1195-1200 Diabetes is an Evolutionary Disease • Neel’s “Thrifty Gene Hypothesis” – Certain human alleles were selected for which maximize metabolic efficiency & food searching behavior • In times of abundance these alleles predispose their carriers to diseases caused by excess nutritional intake, such as obesity & DM • Populations with a history of food scarcity harbor more such thrifty alleles than other populations – Accounting for the more rapid rise in T2DM incidence in societies undergoing “Westernization” Calories are Dangerously Cheap – The average citizen of NC earns $30,553/yr – The price of a Hardee’s Monster Thickburger© is $5.59 – It takes 21 minutes to earn one 1417 calories • It supplies almost a day’s worth of calories • Never in the history of our species have calories been so cheap • It is engineered to appeal to our evolutionarily derived compulsion to seek concentrated fat, salt, carbohydrates and protein • And we don’t even have to leave our car to get it T2DM results from rapid environmental changes superimposed upon an underlying evolutionarily sculpted genome Diabetes is a Genetic Disorder • A number of single-gene mutations have been shown to cause T2DM • • • • • Insulin receptor gene PPAR, HNFs and other transcription factors Mutant insulin Mitochondrial mutations in a tRNA gene Mutations in glucokinase – While of interest, these forms of diabetes are rare and relatively unimportant from a public health standpoint • Twin studies have highlighted the large genetic component in T2DM – MZ concordance for T2DM approaches 100% in some studies – Concordance has increased as mean weight and inactivity have increased Estimates of Heritability From Twin Studies • • • • • • • • • Obesity Type 2 Diabetes Schizophrenia Hypertension Alcholism Cirrhosis Atherosclerosis Breast Cancer Type 1 Diabetes 0.80 0.75 0.65 0.55 0.55 0.50 0.40 0.35 0.25 We now have the ability to analyze the genome and define individual variation Robust Genomic Analysis • At the heart of the concept of “personalized medicine” • Massively parallel genotyping now is widely available – Delineation of the individual’s genome at ~1 million sites – for << $500 • Affordable* complete sequencing of all 3 billion bases of the individual’s genome will likely be a reality within the next 1-3 years – For ~$1,000 What Can We Do With Such Information? • An avalanche of Genome Wide Association Studies (GWAS) in common diseases • Shedding light on the genetic underpinnings of every disease imaginable GWAS-Implicated T2DM Loci McCarthy et al. Current Diabetes Reports 2009, 9:164–171 There is Great Potential in Harnessing Genomic Information • Delineation of novel genes and pathways in the genesis of disease – Allowing identification of genes without underlying biochemical/physiol ogical knowledge – Identification of novel drug targets The Promise of Individualized Medicine Screening & Treatment • The current status of screening in medicine – Relatively little benefit – Actual harm to some – Tremendous waste of resources • Current drug therapy in medicine – Efficacy varies widely – Adverse effects are common and unpredictable – Wasted resources and time • Can we directly apply genetic advances in diabetes to care of the individual? Clinical Medicine is Messy The application of good ideas to the care of the individual is difficult and fraught with hazard How might we apply new genetic knowledge directly to T2DM patients? • Pharmacogenomic applications – PGx is often touted as the “low hanging fruit” of the genetics revolution – Predicated upon the idea that knowledge of one’s genome will be a useful tool to guide drug therapy • Assessment of individual risk – To enable prevention Genotype & Sulfonylurea Response • Genotype at TCF7L2 may be associated with response to sulfonylurea agents • Carriers of the risk allele (TT) were less likely to respond to sulfonylureas – As assessed by reaching target A1C <7% after initiation • Odds ratio for failure = 1.95 Pearson et al. 2007, Diabetes;56:2178 OCT1 & Metformin • The organic cation transporter 1 (OCT1) is involved in controlling intracellular metformin concentration • Could polymorphisms in OCT1 influence metformin efficacy? OCT1 & Metformin • In one small study of 20 healthy volunteers given a glucose challenge following metformin dosing, genotype at OCT1 influenced metformin response PGx and T2DM Not Ready for Prime Time • Such results are evident at an aggregate level, not the individual level • The discriminatory capabilities of genotype to assess response are very weak – 35% of the sulfonylurea “nonresponder” genotype never-theless respond to the agent – There is dramatic overlap in metformin glucose concentrations between the two genotype groups • A lack of clinical outcome data Applying Genomic Information to the Individual The Problem of Relative Risk • Numerous risk alleles for T2DM have been identified – The vast majority of RR defined by GWAS are between 1-2 • What do I do with such information? – From a clinical standpoint the information is so lacking in robustness that it is of questionable utility Frayling T; Nature Reviews 8:657.2007 Applying Genetic Risk Information to the Individual • Most genotyping may soon be performed outside of the traditional medical setting • Numerous companies are now offering “boutique” genotyping – Most include estimates of T2DM risk • Heavily covered by the media • Is such information useful towards the goal of improving health? • “Buying is more American than thinking” Aggregate Risk Scores • One purveyor of such testing (deCODE Genetics; $370) offers a T2DM risk score using one’s genotype at 4 loci – ~40% of population have increased relative risk (RR >1) – 8% have RR 1.5-2.2 – ~3% have RR 1.8-2.2 • Calculation of aggregate risks assumes no interactions • But the real problem is how little useful clinical information is provided Multiple Genetic Variants How Much Added Value? Shifting Knowledge • We are rapidly discovering more genes that influence the risk of T2DM • We have yet to define the majority of T2DM’s the genetic component • Thus future discoveries, like the discovery of modifiers and novel loci may dramatically shift risk assessments – It may eventually make genetic prediction more robust – Today’s “low risk” genotype may well be tomorrow’s “high risk” genotype, and vice versa • Proceed cautiously in the midst of a data avalanche Good Ideas That Weren’t • Hormone Replacement Therapy • Sleeping Babies & prevention of SIDS • Anti-arrhythmics for PVCs • PSAs may cause more harm than help • Good ideas are not enough to guide medical care – We have the power to harm – Even through “non-invasive” testing • Such information has the potential to put our patients on a trajectory that leads to dangerous and harmful interventions We must demand outcome data before embracing good ideas The Paradox of Risk Information • “Knowledge is Power”…or is it? • It is often maintained that knowledge of one’s genetic risk will benefit patients • Does knowledge of increased risk of obesity, diabetes, CVD lead to improved lifestyle? – We’re all at “increased risk” for these maladies – We already strongly recommend healthy lifestyles • Little evidence that genetic information per se is more effective than other types of information in getting people to change their long-term behavior • And for everyone I find who is at increased risk, I’ll find those at decreased risk – Will such information give such individuals license to not pursue healthy lifestyles? The Coming Deluge • The $1,000 genome is coming • We will be awash with new highly complex genetic information • It will often be touted as ready to revolutionize healthcare • It does hold great promise – Especially for drug development and laying bare the molecular foundation of disease • But the rules of medicine haven't changed – There’s nothing magical about genetic information – Take a good family history! – Insist on actual outcome data before embracing attractive ideas You Know You’re a Genetics Nerd When… Google News Alert for: genetics From: Google Alerts [mailto:googlealerts-noreply@google.com] Sent: Thursday, September 10, 2009 6:29 AM To: Evans, Jim Subject: Google Alert - genetics Toward Personalized Diabetes Treatment Pharmaceutical Technology Magazine These are the types of breakthroughs the industry has been waiting for; the links between genetics, biology, and medicine are solidifying. ... This once a day Google Alert is brought to you by Google. See all stories on this topic Remove this alert. Create another alert. Manage your alerts. • How might we apply new genetic knowledge directly to T2DM patients? • Pharmacogenomic applications • Estimate Disease Predisposition The Need for Clinical Outcome Data • The history of medicine is riddled with the corpses of good ideas that didn’t pan out – EC/IC bypass to prevent stroke – HRT to prevent every possible bad outcome of female aging – PSA? • Good ideas are not enough to guide medical care – We have the power to harm – Even through “non-invasive” testing • Such information has the potential to put our patients on a trajectory that leads to dangerous and harmful interventions • We need to insist on data to prove that our good ideas actually result in improved outcomes • We have to be leery of shortcuts and attractive theories Genotype Adds Little to Conventional Risk Estimation C Statistic Model Parameters No Genotype With Genotype P Value Sex & FHx 0.595 0.615 0.11 Age, Sex, FHx, BMI, FBS, BP, HDL, TG 0.900 0.901 0.49 Meigs; N Engl J Med 2008;359:2208-19 What Else Can We Do With Such Information? • Make money!!! – After all, this is the USA • Several companies are now offering “boutique” genotyping • Their success threatens to prove Andy Warhol’s adage that “buying is more American than thinking” Worldwide Prevalence of Diabetes for 2000 and 2010: An Epidemic ZIMMET P, ALBERTI KGMM & SHAW J. Nature 414, 782 - 787 (2001) Published Genome-Wide Associations through 3/2009, 398 published GWA at p < 5 x 10-8 NHGRI GWA Catalog www.genome.gov/GWAStudies Controlling the Genetic Genie • Genetic Discrimination – There is no federal legislative protection…yet • Gene Patenting – Most of our genes have patent claims on them – BRCA1/2 are under restrictive patents • Privacy Issues – “Privacy is dead. Get over it.” – What happens when 23andMe is subpoenaed? – Problems compounded by the acceleration of Information Technology and the World Wide Web – Corporate, government and public genetic databases are being formed all over the world • Who will control this information? – Do you want to know ? – Do you want others to know? • • • • • Your insurance / managed care plan Your employer Your family Dick Cheney Your neighbor who surfs the web