Genes, lifestyle, and risk for heart attack


>>GOOD AFTERNOON. WELCOME TO THE WEDNESDAY AFTERNOON LECTURE. A SPECIAL TREAT IN STORE FOR YOU TODAY. THOSE OF YOU WHO HAVE COME TO MASUR AND THOSE LISTENING IN COUNTLESS NUMBERS ON THE VIDEOCAST BECAUSE WE’RE GOING TO HEAR ABOUT GENES, LIFESTYLE AND RISK FOR HEART ATTACK FROM THE PERSON WHO I THINK I WOULD MOST WANT TO GIVE THAT LECTURE BECAUSE OF THE WAY IN WHICH HIS OWN RESEARCH HAS ILLUMINATED THOSE ISSUES IN WAYS I’M NOT SURE I WOULD HAVE IMAGINED POSSIBLE ADECADE AGO. SEKAR KATHIRESAN, BROAD INSTITUTE, AND PROFESSOR OF MEDICINE AT HARVARD MEDICAL SCHOOL, BUT THAT’S ONLY TRUE UNTIL THE MIDDLE OF JULY, WHEN IN A SOMEWHAT UNEXPECTED AND ONE MIGHT EVEN SAY BOLD MOVE, SEK HAS ANNOUNCED HE WILL BECOME THE CEO OF A NEW COMPANY WHICH HAS AS ITS GOAL THE APPLICATION OF IN VIVO GENE EDITING TO IRREVERSIBLY AND PERMANENTLY CURE CARDIOVASCULAR DISEASE BY A SOMATIC APPROACH. THIS IS PRETTY BEING SUITING STUFF, ALL OF US WHO ARE FROM TI JAZZED ABOUT GENE EDITING FOR THINGS THAT ARE CAUSED BY SINGLE GENE MEU MUTATIONS LUKE SI LIKE
SICKLE CE
LL DISEASE HAVE SEEN AN OPPORTUNITY TO HAVE THIS APPLIED TO A POLYGENIC COMMON DISEASE AND THAT’S VERY MUCH THE VISION THAT SEK IS GOING TO PURSUE IN HIS NEW ROLE IN THIS COMPANY COMING UP VERY SHORTLY. IT SOUNDS EXCITING INDEED, I HAD A CHANCE TO HEAR MORE ABOUT IT SPEAKING TO HIM EARLIER TODAY. HE IS SOMEBODY WHO HAS A DISTINGUISHED ACADEMIC RECORD, UNDERGRADUATE AT UNIVERSITY OF PENNSYLVANIA, BACHELOR’S IN HISTORY. THERE MUST BE A STORY ABOUT THAT. THEN WENT ON TO BECOME — TO RECEIVE HIS M.D. DEGREE AT HARVARD, FOLLOWING WHICH HE WAS A POSTDOC WITH CHRIS O’DONNELL WORKING ON THE FRAMINGHAM STUDY AND THEN AFTER THAT, AT THE BROAD WITH DAVID ALCHUR BUT VERY MUCH CARVING OUT THE SPACE OF THE GENETICS OF HUMAN CARDIOVASCULAR DISEASE, WHERE HE HAS BECOME AN INTERNATIONAL LEADER. HE WENT RAPIDLY THROUGH THE ACADEMIC RANKS TO HIS CURRENT ROLE AS PROFESSOR, AND ALONG THE WAY, RECEIVED A NUMB BR NUMBER
OF IMPORTANT PRIZES. I WILL PARTICULAR MENTION THE CURT STEIN AWARD FROM THE AMERICAN SOCIETY OF HUMAN GENOMICS, THE JOSEPH VITA AWARD FROM THE AMERICAN HEART ASSOCIATION AS WELL AS THE MARTIN PRIZE FOR BASIC RESEARCH FROM MASS GENERAL. HAPPY TO SAY THAT MUCH OF HIS WORK HAS BEEN FUNDED BY NIH SO WE TAKE TOTAL CREDIT FOR ALL OF IT. AND THE WORK THAT HE HAS DONE TOUCHED ON MANY DIFFERENT AREAS BUT PERHAPS YOU WILL KNOW HIM BEST FOR THE LEADING ROLE THAT HE HAS PLAYED IN INTRODUCING THE CONCEPT OF THE POLYGENIC RISK SCORE, WHICH HAS MADE IT POSSIBLE TO TAKE THE VERY COMPLEX GENETICS OF COMMON DISEASES LIKE CARDIOVASCULAR DISEASE AND DIABETES AND MOVE THAT INTO AN AREA OF REAL OPPORTUNITY FOR PREVENTIVE MD SIN IN A PROO SITION MD SIN KIND OF WAY. SO IF YOU WANT TO HEAR ABOUT CARDIOVASCULAR PRECISION MEDICINE, YOU’VE COME TO THE RIGHT PLACE. PLEASE JOIN ME IN WELCOMING TO THE PODIUM DR. SEKAR KATHIRESAN. [APPLAUSE]>>THANK YOU SO MUCH, FRANCIS, FOR THE VERY, VERY KIND INTRODUCTION. REALLY A PLEASURE AND HONOR TO BE HERE TO SPEAK TO YOU ALL TODAY. I ACTUALLY DO COUNT MYSELF AMONG THE LEGIONS OF INDIVIDUALS WHO HAVE ACTUALLY BEEN INSPIRED TO ENTER SCIENCE THROUGH AN SPEERPS AT THE NIH. I WAS HERE ACTUALLY IN THE SUMMER OF 1993, AFTER MY FIRST YEAR OF MEDICAL SCHOOL AT HARVARD WORKING WITH ALLAN RAMOLIN MOLECULAR CARDIOLOGY BRANCH AT THAT TIME I THINK LED BY JEFF HAGUE AND I HAD A TERRIFIC EXPERIENCE, AND REALLY DID INSPIRE ME TO KIND OF STAY IN SCIENCE PARTICULARLY AS A PHYSICIAN SCIENTIST SO FOR THAT I’M PARTICULARLY GRATEFUL, AND OF COURSE ALL THE FUNDING OVER THE YEARS FROM THIS PLACE IN TERMS OF ENABLING OUR WORK AS WELL, SO IT’S GREAT TO BE HERE. LET ME JUMP RIGHT IN. HERE IS MY DISCLOSURE, HERE ARE MY DISCLOSURES. IN TERMS OF ACKNOWLEDGMENTS, I WANT TO THANK MEMBERS OF MY RESEARCH LABORATORY SHOWN ON THE LEFT, AS WELL AS STACEY GABRIEL AT THE BROAD INSTITUTE, AND COLLABORATORS FROM ALL OVER THE WORLD WHO HAVE ENABLED THE STUDIES THAT I’M GOING TO DESCRIBE. OUR SCIENCE HAS BEEN INSPIRED BY PATIENTS, PATIENTS LUKE THIS ONE. THIS IS AN ACTUAL 911 CALL OF A 42 YEAR OLD Q.MAN WHO PRESENTED WITH DID YOU SEENESS AND PROFUSE SWEATING AND WHAT’S HIGHLIGHTED IS THE ACTUAL AMBULANCE SHEET, WHAT IT SAID WHAT YOU CAN READ THAT THE STRETCHER WAS BROUGHT INTO THE RESIDENCE AND THE PATIENT WAS GETTING READY FOR TRANSFER FROM THE CHAIR TO THE STRETCHER WHEN HE STARTED POSTURING AND HAVING A SEIZURE. THE PATIENT WAS TRANSPORTED INTO THE AMBULANCE AND ONCE INSIDE THE AMBULANCE, THE FIRST RHYTHM THAT WAS OBSERVED IS SHOWN HERE, AND THIS IS OF COURSE THE VENTRICULAR FIBRILLATION. THE PATIENT WAS RESUSCITATED, SHOCKED, OUT OF THIS RHYTHM INTO THE RHYTHM SHOWN BELOW, AND HERE IN THIS ECG, THERE ARE SIGNS OF AN ACUTE HEART ATTACK INVOLVING THE RIGHT CORONARY ARTERY. THE PATIENT WAS SUCCESSFULLY TAKEN TO THE HOSPITAL, HAD AN INTERVENTIONAL CARDIOLOGY PROCEDURE, HAD A STENT PLACED IN THE RIGHT CORONARY ARTERY, BUT UNFORTUNATELY HAD SUFFERED A FAIR AMOUNT OF ANOXIC BRAIN INJURY DURING THE RESUSCITATION PERIOD AT HOME AND DIED 10 DAYS INTO THE HOSPITAL COURSE. SO 42-YEAR-OLD GENTLEMAN WITH A FATAL M.I., AND HERE WERE THE M.I. RISK FACTORS DURING A VISIT TO THE DOCTOR SIX MONTHS EARLIER. WHAT CAN YOU SEE IS THAT THE LDL CHOLESTEROL, THE LIPID RISK FACTORS, LDL IS A LITTLE HIGH. THE HDL IS A LITTLE LOW, TRIGLYCERIDES ARE A LITTLE HIGH. THE NON-LIPID RISK FACTORS DO NOT STAND OUT BUT HE HAS A VERY STRONG FAMILY HISTORY, HERE HE IS, FATHER WITH EARLY DISEASE, AS WELL AS UNCLE COULD WHO DIED
AT THE AGE OF 42 AS WELL AND PATERNAL GRANDMOTHER WITH EARLY M.I. AS WELL. SO STRONG FAMILY HISTORY. NOW, WHEN A PATIENT LIKE THIS IS SEEN IN THE CLINIC NOWADAYS, WE’RE ASKED TO CALCULATE 10-YEAR RISK FOR HAVING A HEART PROBLEM USING A GLOBAL RISK CALCULATOR. THE CALCULATOR USED IN THE UNITED STATES IS SPONSORED BY THE AMERICAN COLLEGE OF CARDIOLOGY, AMERICAN HEART ASSOCIATION. THIS IS A 10-YEAR CALCULATOR [email protected] SCLEROTIC CARDIOVASCULAR DISEASE. IF YOU PUT THIS PATIENT’S NUMBERS INTO THE RISK CALCULATOR, YOU END UP GETTING A 1.7% 10-YEAR RISK OF HAVING A PROBLEM. THAT’S CATEGORIZED AS LOW RISK BUT CLEARLY THIS PATIENT WAS NOT LOW RISK, HAD DIED SIX MONTHS LATER. SO THIS KIND OF SPEAKS TO SOME OF THE LIMITATIONS OF THE CURRENT APPROACH TO GLOBAL RISK SCORING. SO THE PATIENT HIGHLIGHTS A GENERAL ISSUE THAT WE’VE ADDRESSED WHICH IS WHAT IS THE INHERITED BASIS FOR MYOCARDIAL INFARCTION, AND WE’VE LOOKED AT THIS PROBLEM FROM TWO DIFFERENT LENSES, TWO DISTINCT LENSES. SO HERE’S THE POPULATION AT LARGE, THE GENOME, AND AVERAGE RISK OF MYOCARDIAL INFARCTION. A GIVEN INDIVIDUAL MAY HARBOR A GENETIC FACTOR THAT MAY INCREASE THE RISK FOR DISEASE SO THE QUESTION HERE IS WHAT’S GENETIC BASIS FOR RISK. ALTERNATIVELY, THAT SAME FACTOR MIGHT ACTUALLY PROTECT AGAINST DISEASE AND SO WHAT’S THE GENETIC BASIS FOR RESISTANCE? SO THESE ARE THE TWO ISSUES THAT I’M GOING TO HIGHLIGHT IN THE NEXT 40 MINUTES OR SO. SO TO GET EVERYBODY UP TO SPEED ON THE DISEASE PROCESS, WE STUDIED MYOCARDIAL INFARCTION OR MORE COMMONLY KNOWN AS HEART ATTACK. THIS IS A DISEASE OF THE WALL OF THE ARTERIES THAT RUN ON THE SURFACE OF THE HEART, THERE ARE THREE MAJOR BRANCHES. THIS IS A CROSS-SECTION OF THAT ARTERY, AND THE DISEASE HAS TWO PHASES. THE FIRST PHASE IS A CHRONIC PHASE, SEVERAL DECADES. THERE’S BUILDUP OF ATHEROSCLEROTIC PLAQUE IN THE WALL OF THE ARTERY, AND THEN THERE’S AN ACUTE PHASE WHERE THERE’S BASICALLY A TEAR, A RUPTURE OF THE ENDOTHELIAL LINING ON THE INNER WALL OF THE ARTERY. THAT RUPTURE EXPOSES THE FLOWING BLOOD TO SUBENDOTHELIAL COLLAGEN WHICH IS A POTENT TRIGGER FOR BLOOD CLOTTING, THROMBOSIS. SO IN A GIVEN TIME POINT, THERE COULD BE A BLOOD CLOT THAT SUPERIMPOSES ON TOP OF THE ATHEROSCLEROTIC PLAQUE ESSENTIALLY BLOCKING THE FLOWING BLOOD FROM GOING DOWNSTREAM. IF THAT BLOOD CLOT STOPS BLOOD FLOW FOR MORE THAN 20 MINUTES, THEN THE PART OF THE HEART SERVED BY THE ARTERY COULD DIE. THAT DEATH OF HEART TISSUE OR MYOCARDIAL NECROSIS IS A HEART ATTACK. THAT CAN BE DETECTED BY SYMPTOMS, EKG CHANGE AND ELEVATION IN BLOOD PROTEIN LEVELS. THIS DISEASE LIKE MANY OF THE DISEASES WE ALL STUDY IS A CLASSIC COMPLEX TRAIT THAT HAS BOTH THE HERITABLE COMPONENTAS WELL AS LIFESTYLE COMPONENTS, AND FOR ABOUT HALF OF ALL MYOCARDIAL INFARCTION, THE FIRST PRESENTATION IS FATAL LIKE THE PATIENT I PRESENTED. NOW THE BEST UNDERSTOOD RISK FACTORS FOR HEART ATTACK ARE LIPOPROTEINS. SHOWN HERE ARE DIFFERENT LIPOPROTEINS, TYPICALLY SPHERICAL PARTICLES WITH A PHOSPHOLIPID MONO LAYER AND THEN A LIPID IN THE CENTER, A NEUTRAL LIP ED IN THE SN TER, AND THE LIPID CAN EITHER BE TRIGLYCERIDE OR CHOLESTEROL. AND THE LOU PO LIPOPROTEINS ARE
NAMED ON THE DENSITY, HIGH-DENSITY LIPOPROTEIN, LOW DENSITY OR THE ONE IN THE CENTER. THESE COLLECTIVELY ARE CALLED TRIGLYCERIDE-RICH LIPOPROTEINS BECAUSE THEY MOSTLY CARRY TRIGLYCERIDE IN THE CENTER. WITH THAT BACKGROUND, THE INHERITED BASIS OF RISK, TO UNDERSTAND THIS, IN 1997, WE STARTED ENROLLING PATIENTS WHO HAD A HEART ATTACK AT A YOUNG AGE WHEN I WAS AN INTERN AS MASS GENERAL HOSPITAL DURING INTERNAL RESIDENCE — WE FOCUSED ON EARLY M.I. PATIENTS BECAUSE THE ROLE OF INHERITANCE IS LARGER, SO WHEN DISEASE OCCURS AT A YOUNGER AGE AND WE USED A CUTOFF OF MEN LESS THAN 50, WOMEN LESS THAN 60 IN TERMS OF FOR THIS RECRUITMENT. SO OVER A FEW YEARS, WE RECRUITED SEVERAL HUNDRED PEOPLE AND THEN COLLABORATED WITH PEOPLE ALL OVER THE WORLD OF SIMILAR COLLECTIONS TO GET TO THE RESULTS I’M GOING TO DESCRIBE. SO IT TURNS OUT THEY’RE IN THREE MAIN GENETIC PATHS TO M.I. RISK. ONE IS THE MONOGENIC MODEL WHERE MUTATION OF A SINGLE GENE IS SUFFICIENT TO LEAD TO EARLY DISEASE. THE SECOND IS THE POLYGENIC MODEL, THE ADDITIVE EFFECT OF MANY VARIANTS IN AGGREGATE CAN LEAD TO EARLY DISEASE, AND THEN LASTLY, SOMATIC MUTATIONS. SO THESE ARE MUTATIONS THAT ARISE IN YOUR OWN BODY DURING YOUR ADULT LIFE THAT CAN PREDISPOSE DISEASE. I’M GOING TO TAKE OOCH OF THOOS
EACH OF TH
OOS MO DLS IN TURN. SO THE FIRST OF THE MO NO JNG MODEL IS A GENE WHERE A SINGLE MUTATION CAN DRIVE THE PATIENT TO EARLY M.I. THE EXPERIMENT HERE SPONSORED BY THE NHLBI SEQUENCING PROJECT A FEW YEARS BACK REALLY LOOKED AT ABOUT 5,000 CASES WITH EARLY M.I., SO 5,000 INDIVIDUALS LIKE THE PATIENT I DESCRIBED, AND 5,000 CONTROLS, AND WE WERE ABLE TO SEQUENCE EVERY GENE IN THE GENOME, THE PROTEIN CODING REGION FOR ALL 10,000 INDIVIDUALS. AS YOU KNOW, THAT’S ROUGHLY 30 MILLION BASES, 180,000 EXONS, AND OF THE 3 BILLION IN THE HUMAN GENOME SEQUENCE, ROUGHLY — GOES FOR PROTEIN. THE ANALYSIS IS RELATIVELY STRAIGHTFORWARD, YOU’RE BASICALLY COMPARING THE FREQUENCY OF MUTATIONS IN CASES IN THE GIVEN GENE AND COMPARING THAT FREQUENCY OF CONTROLS. WITH THE EXPECTATION THAT MOST GENES IN THE GENOME, THERE WILL BE NO DIFFERENCE IN FREQUENCY OF MUTATIONS IN CASES OF CONTROLS BUT FOR A FEW, THERE WILL BE IN EXCESS IN CASES — ACROSS EACH OF 18,000 GENES, AND AT THE END OF THIS KIND OF ANALYSIS, WE FOUND THERE ARE FOUR MAIN GENES THAT STOOD OUT IN TERMS OF WHERE MUTATIONS AND CODING SEQUENCE HAD A LARGE EFFECT ON M.I. RISK. THE GENES ARE SHOWN HERE. THE CARRIER FREQUENCY IN THE POPULATION, THEN THE BLOOD BIOMARKER THAT’S CHANGED AS A RESULT OF THE MUTATION SO IT’S EITHER LDL, TRIGLYCERIDE-RICH LIPOPROTEINS FOR THESE TWO GENES AND LPA KIND OF A COUSIN OF LDL FOR THIS GENE AND THEN CLINICAL EFFECT ON M.I. RISK, ROUGHLY 2 TO 5 FOLD INCREASED RISK-BASED ON WHETHER YOU CARRY MUTATION OR NOT. OKAY, THAT’S THE MONOGENIC MODEL. SO OF THE 5,000 INDIVIDUALS IN THIS STUDY, WITH HEART ATTACK AT A YOUNG AGE, WHAT FRACTION CAN YOU EXPLAIN, CAN YOU FIND ONE OF THESE MONOGENIC MUTATIONS? THIS WAS ACTUALLY NOT KNOWN PRIOR TO OUR STUDY. AND WE FOUND THAT ABOUT 4% OF THE CASES CARRIED A MONOGENIC MUTATION. SO THAT’S ROUGHLY 200 OUT OF THE 5,000 PEOPLE. SO WHAT ABOUT EVERYBODY ELSE? WHAT ABOUT THE REMAINING 4800? WHY DID THEY HAVE A HEART ATTACK IN THEIR 30s, 40s OR 50s? SO THIS TO EXPLAIN THIS, WE TURN TO A POLYGENIC MODEL, AND HERE THE QUESTION IS, DO COMMON DNA POLYMORPHISMS CONTRIBUTE TO M.I. RISK? HERE A MUCH LARGER NUMBER, 60,000 CASES, 120,000 CONTROLS OVER ABOUT 10 YEARS OF WORK IN TERMS OF A META-ANALYSIS. AND THE YES NE TICK JEA KNOW
TYPING MILLI
ONS OF POLYMORE CHISMS ACROSS THE GENOME. NOW WE’RE SEARCHING FOR ALLELE FREQUENCY DIFFERENCES BETWEEN CASES AND CONTROLS. THE ANALYSIS I MENTIONED EARLIER IS THE GENE, HERE THE UNIT OF ANALYSIS NOT A GENE BUT ACTUALLY SITE BY SITE, POLYMORPHISM BY POLYMORPHISM. THIS IS SO CALLED MAN MANHATTAN PLOT. EACH DOT HERE REPRESENTS A SPECIFIC SPOT IN THE YEE NOME SEQUENCE, A POLYMORPHIC SITE. ALL THE DOTS ABOVE THIS GENOME-WIDE STATISTICAL THRESHOLD REPRESENT — THERE ARE ABOUT 95 GENE REGIONS THAT HAVE THAT PROPERTY. NOW EACH OF THESE GENE REGIONS, EACH OF THESE VARIANTS HAS A VERY SMALL EFFECT ON DISEASE RISK. ROUGHLY ANYWHERE FROM 5% INCREASE IN RISK TO MAYBE 30% INCREASE IN RISK IF YOU CARRY THE RISK FLAVOR. BUT THERE ARE MANY OF THEM, THERE ARE 95 OF THEM, SO PEOPLE HAVE WONDERED WHETHER MAYBE YOU COULD ADD THESE UP INTO A POLYGENIC SCORE TO UNDERSTAND RISK. SO IT DOES A POLYGENIC MODEL CONTRIBUTE TO EARLY M.I. SO THE EARLY EFFORTS FOCUSED ON THE RESULTS AT THE TOP OF THE LIST, LET’S SAY THE TOP 95 POLYMORPHISMS THAT I MENTIONED. THE CONCEPT IS QUITE SIMPLE, A SCORE RANGING FROM ZERO TO 2N FOR EACH PERSON. THAT SCORE CAN BE WEIGHTED BASED ON EACH VARIANT’S AFFECT SIZE ON DISEASE RISK AND WE SHOWED OVER THE LAST EIGHT TO 10 YEARS OR SO THAT THIS APPROACH CAN WORK IN AID GUYING AT RISK INDIVIDUALS. BUT THERE WAS AN IMPORTANT CONCEPTUAL ADVANCE PROPOSED ACTUALLY BASED ON WORK IN LIVESTOCK, ANIMAL BREEDING, AND THIS IS LED BY PETER VISHER, NAOMI RAY AND DAVID EVANS BACK IN 2009. THEY REASONED THAT ACTUALLY TAKING THE ANALYSIS FROM THE TOP OF THE LIST TO ACTUALLY LOOKING AT THE FULL GENOME MAY ACTUALLY BE BETTER IN TERMS OF RISK PREDICTION. AND SOON AFTER IT WAS SHOWN THAT ACTUALLY HERITABILITY OF ANY GIVEN TRAIT IS EXPLAINED MUCH BETTER BY LOOKING AT THE FULL LIST RATHER THAN JUST THE TOP OF THE LIST. AND SO THERE HAVE BEEN ALGORITHMS, APPROACHES DEVELOPED TO TAKE ADVANTAGE OF THE FULL SET OF COMMON POLYMORPHISMS TO DO RISK PREDICTION, AND THAT’S WHAT WE LEVERAGED LAST YEAR, ACTUALLY, TO DEVELOP WHAT WE CALL GENOME-WIDE POLYGENIC SCORES, WHERE WE MOVE FROM THE TOP SNPS, JUST LOOKING AT THE TOP SNPS HERE TO REALLY THE FULL BREADTH OF GENETIC DATA, 6.6 MILLION COMMON VARIANTS FOR PREDICTION. SO I WANT TO WALK YOU THROUGH THAT DATA. SO THE HYPOTHESIS HERE THAT WE STARTED WITH WAS THAT A POLYGENIC SCORE INCLUDING A GENOME-WIDE SET OF SNPs COULD EXPLAIN SOME OF THOSE 98% OF PEOPLE WHO HAD A HEART ATTACK AT A YOUNG AGE WHERE WE DIDN’T FIND A MONOGENIC MUTATION. IN FACT, WE THOUGHT THAT MAYBE THERE ARE A SET OF PEOPLE WHO ARE GETTING TO MONOGENIC LEVELS OF RISK USING THIS GENETIC MODEL RATHER THAN A SINGLE MUTATION. SO THAT’S WHAT WE SET OUT TO TEST, AND THIS WORK IN REALLY FOUR STEPS. THE FIRST STEP IS A LARGE GENOME-WIDE ASSOCIATION STUDY WHERE WE HAVE MILLIONS OF POLYMORPHISMS, FOR EACH WE HAVE A RISK FLAVOR AND AN EFFECT SIZE. AND THAT’S THE TRAINING DATASET SO WE HAVE THE EFFECT SIZES FOR 6.6 MILLION SITES, COMMON VARIANTS IN THE GENOME. STEP TWO IS TO USE THAT INFORMATION TO CONSTRUCT SCORES AND WE CONSTRUCTED ROUGHLY 23 DIFFERENT SCORES USING FOUR DIFFERENT ALGORITHMS AND EVALUATED EACH SCORE IN A VALIDATION DATA IS THE. SO THIS IS RAO LU WHAT’S CALLED OUT OF SAMPLE PREDICTION, AND THIS WAS DONE IN A LARGE PERSPECTIVE COHORT CALLED THE UK BIOBANK, HALF A MILLION PEOPLE THAT HAVE BEEN RECRUITED IN THE U.K., CLINICAL DATA, THEY ANSWERED QUESTIONS, BLOOD MEASUREMENTS HAD BEEN MADE, WE KNOW WHO HAS DEVELOPED A HEART ATTACK AND WHO HASN’T OVER TIME, THEY’VE BEEN FOLLOW ON AVERAGE ABOUT EIGHT YEARS AND THEY’VE ALL BEEN GENOTYPES AND WE HAVE ALL THE COMMON VARIANTS REALLY OPENLY AVAILABLE, REALLY A MODEL FOR DATA ACCESS. AND WE WENT TO THAT STUDY, LOOKED AT THE HE FIRST 125,000 PEOPLE, THERE ARE ABOUT 4,000 CASES, 120,000 CONTROLS, AND WE APPLIED ALL 23 SCORES TO PICK THE BEST SCORE. ONE STOOD OUT, AND THE BEST, AND THAT SCORE INVOLVED THE INCLUSION OF 6.6 MILLION COMMON VARIANTS FOR PREDICTION. THEN WE TOOK THAT SCORE TO AN INDEPENDENT SET OF 300,000 PEOPLE FROM U.K. BIOBANK, ABOUT 9,000 CASES, 288,000 CONTROLS, AND LOOKED AT THE PERFORMANCE OF THAT BEST SCORE IN THIS NEW SET OF INDIVIDUALS AND REMARKABLY THE PREDICTION PERFORMANCE WAS IDENTICAL IN THIS GROUP COMPARED TO THIS GROUP. SO I’M GOING TO SHOW YOU DATA FOR THAT SCORE GOING FORWARD. FIRST IS WHAT IS THE SCORE DOES
THE SCORE LOOK LIKE IN THE POPULATION? SO EVERYBODY HAS A NUMBER AND I’M GOING TO SHOW YOU A PLOT OF THAT NUMBER IN 300,000 PEOPLE. AND HERE IT IS. A BEAUTIFUL BELL CURVE. LIKE MANY OTHER TRAITS IN THE POPULATION LIKE HEIGHT OR CHOLESTEROL OR BLOOD PRESSURE, IF YOU PLOTTED THOSE TRAITS, THIS IS WHAT IT WOULD LOOK LIKE. SO WE NOW HAVE A QUANTITATIVE RISK FACTOR FOR HEART ATTACK, REALLY A MEASURE OF ONE’S GENETIC LIABILITY TO DISEASE. NOW, THIS SCORE, A COUPLE OF OTHER QUESTIONS COME TO MIND. HOW DOES THIS SCORE RELATE, THEN, TO THINGS THAT WE USE EVERY DAY IN CLINICAL PRACTICE CURRENTLY TO MEASURE RIS BE FOR HEART ATTACK, FOR EXAMPLE, THE CHOLESTEROL OR PLOOP. AND BLOOD PRESSURE. HERE IS THE CORRELATION OF THE — THE Y AXIS IS THE AMERICAN HEART ASSOCIATION POOLED EQUATION SCORE, AND YOU CAN SEE THE CORRELATION COEFFICIENT IS VERY LOW, LIKE ALMOST NO CORRELATION. SO THIS NEW METRIC OF GENETIC LIABILITY IS CAPTURING SOMETHING QUITE ORTHOGONAL TO THE THINGS WE USE EVERY DAY IN CLINICAL PRACTICE. ANOTHER QUESTION THAT COMES UP IS HOW MUCH HIGHER RISK ARE INDIVIDUALS BASED ON WHERE THEY STAND IN THE SCORE DISTRIBUTION? THERE ARE A COUPLE OF WAYS TO LOOK AT THIS. ONE IS A VERY SIMPLE ANALYSIS WHICH IS ON THE X AXIS, I’M GIVING YOU THE PERCENTILE OF SCORE WHERE EACH PERSON STANDS IN THE DISTRIBUTION COMPARED TO EVERYBODY ELSE. SO THESE ARE #% 1% BINS OF POLYGENIC SCORE. ON THE Y AXIS IS FOR ANY GIVEN BIN, WHAT’S THE PREVALENCE OF CORONARY ARTERY DISEASE, WHAT FRACTION HAVE CORONARY ARTERY DISEASE IN THAT BIN. NOW BECAUSE THIS IS 300,000 PEOPLE, EACH BIN IS ROUGHLY 3,000 INDIVIDUALS, AND WHAT YOU CAN SEE IS IF YOU’RE IN THE TOP PERCENTILE OF SCORE, ROUGHLY 11% OF INDIVIDUALS HAD CORONARY ARTERY DISEASE. IF YOU’RE IN THE BOTTOM PERCENTILE, NO ONE, ALMOST NO ONE HAD CORONARY DISEASE. SO THIS IS A VERY REMARKABLE GRADIENT, I THINK, THAT YOU CAN HAVE IN DISEASE PREVALENCE BASED ON A NUMBER MEASURABLE AT BIRTH. SO WE WERE IMPRESSED PI BY
THIS. THEN WE ASKED, WHAT ABOUT THIS QUESTION OF COULD YOU GET A SET OF INDIVIDUAL WHO WOULD BE AT SIMILAR LEVEL OF RISK TO MONOGENIC MODEL? SO CAN WE IDENTIFY A GROUP WITH RISK FOR M.I. EQUIVALENT TO MONOGENIC? SO REMEMBER THE MONOGENIC COMPONENT WAS ROUGHLY 1 IN 200, 1 IN 250 ABOUT 3 FOLD INCREASED FOR M.I., CARRIER VERSUS NON-CARRIER. SO WE WANTED TO DO AN APPLES TO APPLES COMPARISON BETWEEN THE TWO MODELS, SO WE ARBITRARILY LABELED THE TOP 5% OF THE DISTRIBUTION AS CARRIER OF A GENETIC FACTOR, AND THE REMAINDER AS THE COMPARISON GROUP BASICALLY NON-CARRIERS. AND SAID WHAT’S THE RISK IN THIS GROUP COMPARED TO THIS GROUP? THIS ANALYSIS CAN BE DONE IN MANY DIFFERENT WAYS. YOU CAN MAKE THE REFERENCE GROUP JUST THE MIDDLE, TOWARD THE END. IT’S ALL THE SAME MATH. IT DOESN’T CHANGE THE END RESULT THAT MUCH. IF YOU LOOK AT THIS TOP GROUP IS HIGH, EVERYBODY ELSE AS NON-CARRIER, THE ODDS RATIO IS 3.3. IF YOU LOOK AT THE TOP 1%
VERSUS& EVERYBODY ELSE, THE ODDS RATIO IS CLOSER TO 5. SO THERE ARE A SET OF INDIVIDUALS WHO ARE GETTING TO LEVELS OF RISK SIMILAR TO MONOGENIC FH, FAMILIAL HYPER CHOLESTEROL I’M YA. ANOTHER QUESTION THAT COMES UP IS HOW DOES IT ADD TO TRADITIONAL RISK FACTORS. SO I’M SHOWING YOU HERE ON THE X AXIS, THIS IS WORK OF MIKE INOUYE FROM YOU A AUSTRALIA. ON THE X AXIS ARE INDIVIDUAL RIS BE FACTORS WE CURRENTLY USE TO PREDICT HEART ATTACK RISK. ON THE Y AXIS IS A METRIC OF PREDICTION PERFORMANCE CALLED THE C INDEX. SO THIS IS SMOKING HERE, THIS IS DIABETES, FAMILY HISTORY, AND THIS LABELED META GRS, THAT’S A MOLLYGENIC POLYGENIC SCORE, JUST A DIFFERENT NAME USED HERE. WHAT YOU CAN SEE IF YOU TAKE EACH INDIVIDUAL RISK FACTOR IN TURN, THE POLYGENIC SCORE HAS A PREDICTION CHARACTERISTIC SIMILAR TO CHOLESTEROL OR BLOOD PRESSURE INDIVIDUALLY.& IN FACT, IT’S A BIT BETTER. THEN IF YOU THINK ABOUT WHAT ABOUT ALL THE CONVENTIONAL RISK FACTORS TOGETHER WITHOUT THE GENETICS, THAT IS ABOUT .67 IN TERMS OF C INDEX, CONVENTIONAL RISK FACTORS PLUS THE POLYGENIC SCORE IS ABOUT .7. SO AGAIN MAKING THIS POINT THAT THIS IS ADDING INCREMENTAL INFORMATION ON TOP OF WHAT WE ALREADY USE IN CLINICAL PRACTICE. BUT AGAIN KEY DIFFERENCE I THINK BETWEEN THIS FACTOR AND OTHER THINGS IS THEY CAN BE MEASURED EARLY IN LIFE AND HAS PREDICTED PROPERTY FROM THEN ON. NOW, THAT WAS ALL DATA IN U.K. WHAT ABOUT EXTERNAL VALIDATION? WHAT ABOUT PEOPLE IN THE UNITED STATES? WHAT ABOUT BROADER ETHNICITIES BEYOND THOSE IN THE U.K., EUROPEAN ANCESTRY? SO TO LOOK AT THE EXTERNAL VALIDATION, WE TURNED TO A STUDY IN THE UNITED STATES OF ABOUT 2,000 POOM WITH 2,000 PEOPLE WITH EARLY HEART ATTACK, 3,000 CONTROL, AND IN THIS CASE THROUGH THE SUPPORT FROM THE NATIONAL HUMAN GENOME RESEARCH INSTITUTE, CENTERS FOR COMMON DISEASE GENOMICS GRANTS, WE WERE ABLE TO OBTAIN 30X WHOLE GENOME SEQUENCES IN ALL INDIVIDUALS AND WE WERE ABLE TO SIMULTANEOUSLY LOOK AT THE CONTRIBUTION OF POLYGENIC AND MONOGENIC IN THE SAME SET OF INDIVIDUALS. SO I’M GOING TO SHOW YOU THOOS RESULTS JUST IN PICTORIAL FORM. SO FOR EVERY 100 PATIENTS WITH RECALL M. HUH IN THIS STUDY SIMILAR TO THER TO THE EARLIER
STUDY I MENTIONED, ROUGHLY TWO HAD A MONOGENIC MUTATION THAT YOU COULD READILY SEE BECAUSE THEY HAD THAT MUTATION IN THE FH GENE THEY HAD ABOUT A FOUR FOLD RISK FOR HEART ATTACK. IN CONTRAST, ALMOST 17% WERE BASICALLY HIGH POLYGENIC RISK. THEY WERE IN THAT TOP 5% OF THE POLYGENIC SCORE DISTRIBUTION SO REALLY A TENFOLD HIGHER NUMBER OF INDIVIDUALS AFFECTED BY THIS MODEL COMPARED TO THE OTHER MODEL. AND A SIMILAR LEVEL OF RISK COMPARED TO THE MONOGENIC MODEL. NOW HERE ARE THE TWO MODELS IN CONTRAST. WE TALKED ABOUT THE PREVALENCE AND THE ODDS RATIOS BEING SIMILAR. NOW THE MODE OF DETECTION, HOW ARE WE FINDING THESE PEOPLE RIGHT NOW? FOR THE MONOGENIC MODEL, WE TYPICALLY ARE FINDING THEM IN CLINIC BECAUSE OF THE ELEVATED LIPIDS, THEY GENERALLY INCREASE LDL AND/OR TRIGLYCERIDES. THE POLYGENIC MODEL, THIS IS THE MOST IMPORTANT POINT I THINK OF THIS SECTION OF THE TALK IS, THEY’RE CURRENTLY UNAWARE OF THEIR RISK. THEY’RE NOT BEING PICKED UP BY OTHER BIOMARKERS BECAUSE I SHOWED YOU THAT IT’S VERY POORLY CORRELATED WITH THE EXISTING RISK SCORES. AND ANOTHER POINT IS ABOUT THE MECHANISM OF RISK. FOR MONOGENIC, WE KNOW WHAT IT IS, IT’S THE APOE LIPOPROTEINS, TO POLLLYGENIC, IT REALLY IS MANY, MANY THINGS IN AGGREGATE LEADING TO DISEASE. THEN A VERY IMPORTANT POINT IN TERMS OF RISK ASSESSMENT, IT’S IMPORTANT TO FIND PEOPLE AT RISK BUT EVEN MORE IMPORTANT TO BE ABLE TO OFFER THEM INTERVENTIONS THAT CAN REDUCE RUSK. SO FOR THE MONOGENIC LIFESTYLE, WE MONOGENIC — FOR POLYGENIC WE DO NOT KNOW IF THAT RISK WAS MODIFIABLE. WE WENT ON TO LOOK AT THIS QUESTION AND THE SHORT ANSWER IS YES. THE RISK CONFERRED BY THE POLYGENIC MODEL IS MODIFIABLE FOR M.I., AND IN TERMS OF EITHER LIFESTYLE AND/OR MEDICINES. SPECIFICALLY STATIN MEDICATIONS. IN THE INTEREST OF TIME, I’M GOING TO SKIP THROUGH — ACTUALLY LET ME JUST DO THIS, THE LIFESTYLE PORTION QUICKLY. SO THIS IS THE LIFESTYLE DATA. IN INDIVIDUALS AT HIGH POLYGENIC RISK, CAN A FAVORABLE LIFESTYLE COUNTERBALANCE THE INHERITED RISK? SO WE DID A LIFESTYLE SCORE VERY SIMILAR TO THE POLYGENIC SCORE EXCEPT THIS IS JUST FOUR FACTORS. NO SMOKING, BODY MASS INDEX LESS THAN 30, REGULAR PHYSICAL ACTIVITY OR A HEALTHY DIET PATTERN, AND GAVE PEOPLE A SCORE, ONE POINT FOR EACH OF THESE GOOD THINGS, 3 OR 4 IS A FAVORABLE SCORE, FAVORABLE LEUF STEUL, SOOR OR 1 LIFESTYLE, 2 IS INTERMOOD YACHT. THEN WE LOOKED AT THE RELATIONSHIP OF LIFESTYLE SCORE TO THE INHERITED POLYGENIC RISK. AND HERE ON THE X AXE US IS THE JE IN TICK RISK. THIS IS HIGH POLYGENIC RISK. ON THE Y AXIS IS THE EVENT RATE IN AGAIN THE ERIC STUDY. WHAT YOU CAN SEE IS A THANK YOU SET — UNFAVORABLE LIFESTYLE, FAVORABLE LIFESTYLE AND INTERMEDIAT. IF YOU’RE A HIGH POLYGENIC RISK AND UNFAVORABLE LIFESTYLE, THE 10-YEAR EVENT RATE WAS 11% IN THIS STUDY. IF YOU’RE HIGH POLYGENIC RISK AND A FAVORABLE LIFESTYLE, THE 10-YEAR EVENT RATE WAS ONLY 5% SO ALMOST A 50% OFFSET IN THE INHERITED RISK-BASED ON BEING OPTIMAL IN TERMS OF LIFESTYLE. SO DNA WAS NOT DESTINY HERE, BUT DO YOU HAVE CONTROL OVER THIS GENETIC RISK. OKAY. ANOTHER OPTION TO HELP PATIENTS WHO MAY HAVE HIGH POLYGENIC RISK WOULD BE MEDICATIONS. SPECIFICALLY STATIN MEDICATIONS. SO WE LOOKED AT THE RELATIONSHIP OF POLYGENIC RISK STRATA TO BENEFIT FROM STATINS TO PREVENT A FIRST HEART ATTACK FROM THREE COMPLETED RANDOMIZED CONTROLLED TRIALS WITH STATIN THERAPY WHERE WE HAD ACCESS TO THE DNA OF THE PARTICIPANTS IN BOTH ARMS AT BASELINE. AND WE EVALUATED THE CLINICAL BENEFIT OF STATIN THERAPY IN HIGH RISK GROUP VERSUS ALL OTHERS AND HERE ARE THE RESULTS. AMONG THOSE WHO ARE HIGH GENETIC RISK, THE RELATIVE RISK REDUCTION WAS 44%, THE ABSOLUTE RISK REDUCTION WAS 8%. IN CONTRAST AMONG EVERYBODY ELSE, RELATIVE RISK REDUCTION DUX WAS LOWER AT 24% AND — AT 2.7%. SO AMONG THOSE AT HIGH POLYGENIC RISK, STATINS CONFER A GREATER BENEFIT TO PREVENT A FIRST M.I. THAN ALL OTHERS. SO WE HAVE BASICALLY TWO OPTIONS TO MODIFY THE RISK COMING FROM POLYGENICITY. SO WHAT I’D LIKE TO SUMMARIZE IN TERMS OF THE POLYGENIC SCORE WORK IS, THIS IS IN SOME SENSE THE FIRST RISK FACTOR, IT’S THE SOIL ON WHICH ALL THE OTHER RISK FACTORS ARE KIND OF PLANTED, SO WE CAN NOW DISTILL INHERITED SUSCEPTIBILITY TO M.I. INTO A SINGLE NUMBER, THAT HAS CONTRIBUTION IN THE POPULATION AND ADDS INCREMENTAL VALUE TO WHAT WE ALREADY MEASURE. I THINK IN A FEW YEARS, MOST KNOW THIS NUMBER SIMILAR TO THE WAY WE KNOW LDL RIGHT NOW. THIS APPROACH OF POLYGENIC RISK SCORES ISN’T GENERALIZABLE TO OTHER DISEASES. SO WE LOOKED AT THIS AS WELL AND LOOKED FOR ATRIAL FIBRILLATION, INFLAMMATORY BOWEL DISEASE, TYPE 2 DIABETES, BREAST CANCER, IN EACH CASE THERE’S A SET OF INDIVIDUALS IN THE TOP, LET’S SAY, 4 OR 5% OF SCORES THAT ARE EXTRAORDINARILY HIGH RISK FOR DISEASE COMPARED TO EVERYBODY ELSE. AND IT’S IMPORTANT NOW TO THINK THROUGH WHAT ARE THE INTERVENTIONS, EITHER LIFESTYLES, MEDICINES OR DIAGNOSTIC SCREENING THAT MIGHT BE YOU A PLEUD APPLIED TO THESE
INDIVIDUA
LS WHO ARE AT HIGH RISK. OKAY. IN TERMS OF THE THIRD PATH, THE SOMATIC MUTATION PATH, WHAT WE’VE BEEN FOCUSING ON SO FAR ARE HEREDITARY MUTATIONS, THAT YOU INHERIT FROM YOUR PARENTS BUT THERE’S ANOTHER SET OF MUTATIONS YOU MIGHT AQEUR DURING YOUR LIFETIME, THAT ARE TYPICALLY THOUGHT OF IN THE CON TEBS OF CANCER. AND ONE OF THE DOOZ — ONE OF THE SO MA TIB MUTATION PROCESSES IS CALLED TONAL HEMATOPOIESIS. SO WE ALL HAVE HEMATOPOIETIC STEM CELLS IN OUR BONE MARROW. THOSE CELLS, ONE OF THOSE CELLS MIGHT REQUIRE A MUTATION — ACQUIRE A MUTATION DURING THE COURSE OF AN INDIVIDUAL’S LIFE, AND THAT MUTATION MIGHT GIVE THAT CLONE, THAT CELL THE GROWTH ADVANTAGE, THAT CLONE EXPANDS AND MAY ACQUIRE ADDITIONAL MUTATIONS ALONG THE WAY. SO THIS CLONAL HEMATOPOIESIS, TURNS OUT IT’S ACTUALLY DETECTABLE BY EXOME SEQUENCING OF BLOOD, AND SYD JAZZWALL AND BENNY BERT IN 2014 SHOW THIS IS AN AGE-DEPENDENT PROCESS. ON THE X AXIS IS DECILES OF AGE, 10-YEAR WINDOWS, ON THE Y AXIS IS THE FREQUENCY OF CLONAL HEMATOPOIESIS SO BY THE TIME PEOPLE ARE IN THEIR 70s AND EIGHTIES, ALMOST 10% OF INDIVIDUALS WILL HAVE ONE DOMINANT CLONE. NOW, IT TURNS OUT THESE MUTATIONS ARE NOT JUST SCATTERED AT GENES ALL YOU A CROSS THE GENOME BUT THEY ACTUALLY FALL IN JUST A FEW GOONS. THESE ARE THE GENES, THESE ARE ALL EPIGENETIC DRIVER GENES AND THEY’RE FREQUENTLY MEU TAUGHTED AND AGAIN, THIS PROCESS CAN BE DETECTABLE WITH SEQUENCING OF BLOOD TURN EU. JUST WHOLE BLOOD. NOW, THE PRESENCE HE OF SOMATIC MUTATION UPPED THE RISK FOR HEMATOLOGIC CANCER, THAT’S NO SURPRISE BECAUSE IT’S A MUTATION PROCESS THAT’S GIVING THE CLONE A GROWTH ADVANTAGE, BUT WHAT WAS A SURPRISE IS THEY FOUND INCREASED RISK FOR HEART ATTACK, CORONARY HEART DISEASE. WE WENT ON TO REPLICATE THIS SITUATION AND ASK IF ANY OF THEM MIGHT HAVE CLONES. AND WE FOUND THAT CLONAL HEMATOPOIESIS IS IDENTIFIED IN EARLY HE — NO YOUNG PERSON BASICALLY SHOULD HAVE A CLONE BUT WE FOUND ABOUT 2% OF EARLY M.I. PATIENTS DID AND THE RISK CONFERRED BUT HAVING A COLON EXAWRED TO NOT HAVING IT WAS ABOUT FOUR FOLD, SIMILAR TO THE MOLLYGENIC AND POLYGENIC MODEL. SO WHAT’S THE MECHANISM, HOW DOES HAVING A CLONE HERE INCREASE RISK FOR HEART ATTACK? AND THIS IS A WORK IN PROGRESS, BUT EARLY EVIDENCE SUGGESTS THAT THIS IS BASICALLY HEIGHT EBBED
BASICALLY
HEIGHTENED INFLAMMATION. S THAT THE ROONE FOR THE AK SL ACCELERATED ATHEROSCLEROSIS IN PATIENTS THAT CARRY CLONES. MICE ARE GIVEN WILD TYPE BONE MARROW OR FROM TATTOO KNOCKOUT ANIMALS, TA TATTOO BEING ONE OF THOSE GENES AND IF YOU GIVE THEM — TYPE 2 LEVEL, WILD TYPE BONE MARROW. THE INFLAMMASOME PART OF IT, THERE’S HEIGHTENED INFLAMMASOME ACTIVATION. SO MONOGENIC, POLYGENIC AND SOMATIC. SO WHAT DOES ALL THIS MEAN FOR THE ARCHITECTURE OF DISEASE, COMMON COMPLEX DISEASE. I WANT TO TAKE A MINUTE TO GIVE YOU OUR THOUGHTS ON IMPLICATIONS OF THIS WORK FOR HOW WE THINK ABOUT COMMON JEEZ DISEASE
GENERALLY. THERE ARE LARGELY TWO MO DLS OF COMMON COMPLEX DISEASE. I’M GOING TO CALL THEM THE FRUIT SALAD MO DL AND THE SMOOTHIE MO DL. HAVE SIGN TUSK. THE FRUIT SALAD MO DL MODEL IS
HERE. THE COLORS REPRESENT DIFFERENT DRIVING PATHWAYS, AND IN ANY GIVEN PERSON, THERE’S ACTUALLY ONE DRIVING PATHWAY. AND OUR JOB IS TO MILWAUKEE MEASUREMENTS TO RESOLVE A COMMON COMPLEX DISEASE LIKE HEART ATTACK INTO THESE MULTIPLE DISEASES. SO THAT’S ONE MODEL. THE OTHER MODEL IS THIS, WHERE IN ANY GIVEN PERSON, IT’S ACTUALLY A QUANTITATIVE BLEND OF DIFFERENT DRIVE PATHWAYS WITH RELATIVE EXCESS OF ONE FLAVOR VERSUS ANOTHER. NOW IT’S PRETTY CLEAR THAT THERE’S BEEN A LOT OF WORK AND HYPOTHESES THAT ACTUALLY THIS IS REALLY WHAT’S GOING ON FOR COMMON COMPLEX DISEASE. AND I WOULD SAY THAT THE GENETIC DATA IS REALLY SUGGESTING THAT THE SMOOTHIE MODEL IS WHAT’S MOST APPROPRIATE TO REFLECT OR DESCRIBE COMMON COMPLEX DISEASE. AND OUR JOB IS TO REALLY FIND THESE DRIVING PATHWAYS, THESE COLORS, AND THEN UNDERSTAND IN ANY GIVEN PERSON, DO THEY HAVE A RELATIVE EXCESS VERSUS ONE OR ANOTHER, AND CAN WE DEVELOP THERAPIES IMENS EACH AGAINST
EACH OF THES
E DRIVING PATHWAYS AND YOU MIGHT UNDERSTAND THAT THE THERAPY WOULD HAVE DIFFICULT REPRESENTATION BENEFIT BASED ON HOW MUCH THAT PATHWAY IS OPERATIVE, BUT IT’S NOT AN ALL OR NOTHING PHENOMENON. OKAY. WE NEED TO MOVE ON TO THE NEXT TOPIC WHICH IS RESISTANCE. WHAT’S THE GENETIC BASIS FOR RAO SUS TANS. SO EUM GOING TO WORK THROUGH THIS IN THE CONTEXT OF LIE PO PROTEINS. THE RISK FACTORS THAT I’VE DID SCRIBED RECALLIER. RESISTANCE MUTATIONS CAN BE VERY HELPFUL TO GUIDE THE DEVELOPMENT OF NEW MEDICINES. A GREAT EXAMPLE HERE IS LDL. THIS IS A PLOT OF LDL, HDL, TRIGLYCERIDES ON THE X AK
ACCESS M.I. RISK. HIGHER LEVEL OF LDL IS CORRELATED WITH INCREASED DISEASE RISK. FROM GENETICS AND THERAPY, WE KNOW THAT LDL IS A CAUSAL PATHWAY, SO THERE ARE MUTATIONS IN SFERL SEVERAL GENES THAT
LOWER THE LEVEL OF LDL LIFELONG THAT — WOULD WORK IN THE CLINIC AND TWO CASES, MEDICINES THAT HAVE BEEN DEVELOPED BASEDDOPATHIES
DOPATHIESON THESE GENETIC — DO WORK. CAUSAL PATHWAY FOR RESIST AT THAT PARTICULAR TIME — LOW LDL RESISTANCE TO HART ATTACK. NOW WHAT ABOUT HDL? THANK YOU LEVEL HIGH LFL LEVEL
OF HDL IS S
OSHTED WITH LORIS ASSOCIATED WITH LOWER RISK OF HEART ATTACK. SHOULD BE PROTECTIVE. SO IS HIGH HDL A PROTECTIVE PATHWAY? WE LOOKED AT THIS FROM A GENETICS PERSPECTIVE OF THE HDL HYPOTHESIS AND AGAIN WE FOUND A MUTATION CARRIED BY 1 IN 40 INDIVIDUALS IN THIS GEEP CALLED NFL LIPASE. AS A RESULT OF CARE OOOING THAT MEU TAWTION, THIS HAVE ABOUT 15% HIGH CHOLESTEROL LIFELONG. SHOULD BE A GOOD THING IF HDL WAS CAUSAL. THESE INDIVIDUALS CARRYING THE MUTATION SHOULD BE PROTECTED FROM HEART ATTACK. SO WE LOOKED AT THAT, WE TESTED WHETHER THE MUTATION HAD AN ASSOCIATION WITH HEART ATTACK, AND LOOKING A LITTLE OVER 100,000 PEOPLE, THE INDIVIDUALS WHO KEAR HE’D CARRIED THE
MUTATION HAD TH
E SAME RISK OF HEART ATTACK AS THOSE WHO DIDN’T CARRY THE MUTATION DESPITE THE FACT THEY HAD HIGHER HDL LIFELONG. THE ODDS RATIO WAS .99, SO PRETTY CLOSE TO 1. SO THIS WAS A REAL SURPRISE TO US. AND TOOK US A WHILE TO ACTUALLY PUBLISH THIS. AND SO WHAT’S GOING ON HERE? HERE’S THE DAW TA. TURNS OUT YOU COULD SAY IT’S JUST ONE WAY OF RAISING HDL, WE LOOKED AT SEVERAL OTHER RARE — SAME THING, HDL HIGHER, NO EFFECT OP HEART ATTACK, WE LOOGED AT A SET OF COMMON VARIANTS, A POLYGENIC SCORE FOR HIGH HDL, HIGHER HDL BUT NO EFFECT ON M.I. SO THESE GENETIC DATA WOULD ACTUALLY SAY THAT MEDICINES DESIGNED TO RAISE HDL CLE TROLL WON’T WORK IN THE CLINIC. THAT THEY WILL NOT LOWER RISK OF HEART ATTACK. THAT’S WHAT THE PREDICTION WOULD BE HERE. SO THAT PREDICK HAS BEEN
PREDICTION HAS BE
EN TESTED NOW. HERE I’M GOING TO SHOW YOU ONE, A MEDICINE THAT INHIBIT A PROTEIN CALLED CETP. HERE ON THE X AXIS IS 15,000 PEOPLE, HALF OF THEM HAVE BEEN GIVEN MEDICINE, HALF OF THEM NOT, AND THE Y AXIS IS THE HDL CHOLESTEROL, X AXIS IS TIME, AND YOU CAN ZOO IF YOU GOT THE SEE
— HERE WHE
RE THE OOFNT OH CURSE FOR THE TWO GROUPS IN THE RANDOM USED KROLL TRIAL. COMPLETELY SUE SUPER IMPOSABLE. IT LEADS TO THIS CONCEPT THE EMERGING VIEW THAT IT’S A NON-CAUSAL MARKER RISK. SO IT’S STILL VERY USEFUL TO IDENTIFY PATIENTS AT RISK BUT NOT USEFUL TO TARGET IN TERMS OF THERAPY BECAUSE IT SEEMS TO BE MARKING SOMETHING ELSE. SO IF HDL IS A NON-CAUSAL MARKER THN WHY IS THE EPIDEMIOLOGY SO ROBUST, WHY IS IT IN EVERY STUDY HIGH HDL IS CORRELATED WITH LOWER M.I. RISK, WHAT IS IT MARKING? RANGE OF POSSIBILITIES. THE ONE WE’VE EXPLORED MOST CLOSELY IS THIS IDEA OF IT’S MARKING THE TRIGLYCERIDE-RICH LIPOPROTEINS. AS YOU KNOW WHEN YOU HAVE HIGH HDL IT OFTEN TRACKS WITH LOW LEVEL OF LDL, OVER THE LAST 30, 40 YEARS, THE FIELD HAS DISCOUNTED LARGELY FROM OBSERVATIONAL EPIDEMIOLOGY BUT WE MAY HAVE HAD IT BACKWARDS FOR THE LAST 30 OR 40 YEARS. THE QUESTION IS ARE THERE MORE PROTECTIVE PATHWAYS BEYOND LOW LDL. WE JUST TURNED AWAY HIGH HDL, WHAT ABOUT LOW LEVEL OF LOW TRIGLYCERIDE-RICH LIPOPROTEINS. THE STORY I’M GOING TO TELL YOU IS ABOUT APOC3. THIS STORY INVOLVES THE LANCASTER AMISH. SO THESE ARE INDIVIDUALS WHO IMMIGRATED TO THE UNITED STATES IN THE 1700s, A FEW HUP PEOPLE, AND THAT POPULATION HAS GROWN TO I THINK A COUPLE HUNDRED THOUSAND PEOPLE IN LANCASTER COUNTY. WHAT TONY POLIN FROM THE UNIVERSITY OF M.D. MARYLAND
SHOWED BACK IN 2008 WAS THAT ONE IN 20 AMISH CARRY A MUTATION THAT BROKE ONE COPY OF APOC3. AS A RESULT OF HAVING THAT MUTATION, INDIVIDUALS HAD LOWER TRIGLYCERIDES. WE CAME ALONG AND SHOWED THAT ONE IN 150 IN THE GENERAL POPULATION ACTUALLY CARRIED A NULL MUTATION IN THIS GENE, THEY HAVE LOWER TRIGLYCERIDES, AND LOWER RISK OF CLINICAL EVENT, HEART ATTACK AS WELL, REALLY SUGGESTING THAT BASICALLY GETTING RID OF A WORKING COPY OF TERMS OF MEDICINES DEVELOPMENT. THAT’S GREAT, HETEROZYGOUS AND ALL, ONE COPY BROKEN, GOOD, BUT WHAT ABOUT CAN YOU FIND PEOPLE WITH BOTH COPIES GONE, KNOCKOUT. IF YOU CAN FIND INDIVIDUALS AND SHOW THAT THEY’RE HEALTHY, AGAIN, IT WOULD ADD EVEN MORE CREDENCE TO THIS TARGET IN TERMS OF MEDICINE DEVELOPMENT BECAUSE YOU WOULD PRESUMABLY — IT WOULD GIVE YOU CONFIDENCE IN TERMS OF SAFETY OF INHIBITION OF THIS PROTEIN. SO WE LOOKED FAR AND WIDE FOR HOMOZYGOUS NULLS FOR APOC3, DIDN’T FIND A SINGLE PERSON. THEN WE TURNED TO AN IMPORTANT CONCEPT, CONSANGUINITY, WHEN PARENTS ARE CLOSELY RELATED TO EACH OTHER THE CHILDREN ARE MORE LIKELY TO BE HOMOZYGOUS AT ANY GIVEN SIDE OF THE GENOME. THE DARKER COLORS ARE HIGHER RATES, AND THE HIGHEST RATE IS IN PAKISTAN, ROUGHLY 40 TO 50% OF ALL MARRIAGES INVOLVE RELATIONSHIPS AS CLOSE AS FIRST COUSINS. AND WE HAD A COLLABORATOR, WE HAVE BEEN WORKING WITH FOR SEVERAL YEARS FROM PAKISTAN WHO HAD COLLECTED — A BUN OF PEOPLE INTO A RESEARCH STUDY NOR GENETICS OF CARDIOBOLG DISEASE. SO WE ENDED UP AGAIN THROUGH GENEROUS FUNDING FROM NHLBI AND NHGRI WE WERE ABLE TO EXOME SEQUENCE ALL 7,000 INDIVIDUALS AND LOOKED AT THE RELATIONSHIP OF MUTATION STATUS IN THIS GENE TO BLOOD LEVEL OF APOC3, THIS IS A CIRCULATING PROTEIN. ON THE X AXIS ARE THE THREE GENOTYPES. WILD TYPE, HETEROZYGOUS NULL, HOMOZYGOUS NULL ON THE Y AK CYST IS APOC3 AND WOULDN’T YOU KNOW IT, WE FOUND FOUR INDIVIDUALS IN THE FIRST 7,000 THAT WERE HUMAN KNOCKOUTS FOR APOC3. THEY HAD NO CIRCULATING APOC3. OKAY. SO NOW WE HAVE THESE INDIVIDUALS, WE WANTED TO TRY TO UNDERSTAND WHY THEY MIGHT BE PROTECTED FROM HEART ATTACK. THERE’S A FAIR AMOUNT KNOWN ABOUT APOC3, SO IT’S A PROTEIN THAT PUTS A BREAK ON YOUR BODY’S ABILITY TO CLEAR DIETARY TRIGLYCERIDES SO KIND OF A BAD ACTOR IN SOME SENSE. SO WE WANTED TO BRING INDIVIDUALS BACK THAT HAVE THIS GENOTYPE AND MAYBE TO PERTURBATIONAL TESTING TO BETTER UNDERSTAND THEIR PHYSIOLOGY, SO WE ENDED UP RECRUITING ONE OF THE FOUR INDIVIDUALS HERE, HE’S MINUS-MINUS, HOMOZYGOUS NULL. THIS IS THE SPOUSE, AND NINE — THEY HAD NINE CHILDREN BE AND WE WERE SHOCKED TO LEARN THAT THE SPOUSE WAS ALSO HOMOZYGOUS NULL. MAKING ALL NINE CHILDREN OBLIGATE HOMOZYGOUS. SO TO FINDING 11 JUST IN THIS ONE FAMILY. AND WE WERE ABLE TO FURTHER EXTEND OUT THIS FAMILY, ALL THREE GINO TYPES IN THIS FAMILY, AND THEN WE DID A PROVOCATIVE TEST TO UNDERSTAND WHY THESE NULL MUTATION CARRIERS MIGHT BE PROTECTED. THE PROVOCATIVE TEST WAS IN INDIVIDUALS WITHOUT THE MUTATION AND WITH HOMOZYGOUS NULL AND THE TEST WAS A BIG MILK SHAKE. 80 GRAMS OF FAT, AND TO TRY TO UNDERSTAND THEIR CLEARANCE OF DIETARY FAT, WE MEASURED BLOOD EVERY HOUR FOR SIX HOURS, AFTER THE FAT CHALLENGE. ON THE X AXIS IS TIME, Y AXIS IS BLOOD TRY GLIS RYE. THIS IS THE WILD TYPE INDIVIDUALS IN THE FAMILY. WHAT HAPPENS PRESUMABLY IN ALL OF US IN THE ROOM IS AFTER A FATTY MEAL, THE BLOOD TRIGLYCERIDES, BLOOD FAT RISES ROUGHLY TWO TO THREE FOLD FROM ABOUT 100 TO LET’S SAY 300. THIS IS WHAT HAPPENS IN THE HOMOZYGOUS NULL INDIVIDUALS. THERE’S BASICALLY NO POST POST PRANDIAL BUDGE IN DIETARY TRIGLYCERIDES SO NOW WITH A HIVE TIME OF MEALS YOU CAN UNDERSTAND WHY THESE INDIVIDUALS ARE BEING PROTECTED, THE ARTERIES ARE SEEING LESS FAT THROUGHOUT THEIR LIFETIME. ALL RIGHT. SO THESE DATA THEN SUGGEST THAT MEDICINES THAT MIGHT MIMIC THE NULL MUTATION, SO BLOCK APOC3, WOULD HAVE AN EFFECT ON TRIGLYCERIDES. A COUPLE COMPANIES THAT DID THAT WITH SEVERAL APPROACHES, AND THOSE MEDICINES NOW HAVE BEEN GIVEN TO FIRST IN HUMAN STUDIES AND THEY DO LOWER TRIGLYCERIDES AND TIME WILL TELL WHETHER THEY ALSO LOWER RISK OF M.I. SO IN ADDITION TO APOC #, THERE ARE TWO OTHER GENES THAT WE HAVE ESTABLISHED WHERE NULL MUTATIONS REDUCE TRIGLYCERIDES AND REDUCE RISK OF HEART ATTACK AND ALL THREE OF THESE GENES, BELIEVE IT OR NOT, FALL INTO ONE PATHWAY, AND THAT PATHWAY INVOLVES YOUR BODY’S ABILITY TO CLEAR DIETARY TRIGLYCERIDES, SO CALLED LIPOLYSIS SHOWN HERE IS THE KEY ENZYME THAT BASICALLY CLEARS OR HYDROLYZES FROM THE DIET AND THE KEY PROTEINS THAT ACTUALLY ARE ENDOGENOUS REGULATORS OF HYPOPROTEIN LIPASE IN EACH OF THESE THREE GENES, THERE ARE NULL MUTATIONS AND AS A RESULT OF NOT HAVING A COPY, WORKING COPY OF THESE GENES, INDIVIDUALS HAVE LOWER LEVEL BLOOD TRIGLYCRIDES, LOWER RISK OF HEART ATTACK, AND MEDICINES ARE BEING DEVELOPED AGAINST EACH OF THESE GENES TO MIMIC THE RESISTANCE MUTATIONS IN THIS PATHWAY. SO WHAT WE CAN SAY IS BEYOND LOW LDL, LIPOLYSIS OF TRIGLYCERIDE RICH LIPOPROTEINS BY LPL AND ITS REGULATORS IS A PROTECTIVE PATHWAY FOR M.I. LET ME CLOSE BY COMING BACK TO PATIENT I DESCRIBED EARLIER. SADLY THIS IS ACTUALLY THE STORY OF MY OWN BROTHER WHO PASSED AWAY ABOUT SEVEN YEARS AGO DUE TO THIS TERRIBLE DISEASE. I REALLY WANT TO ASK, WHAT CAN WE DO TO PREVENT SUCH TRAGEDIES IN TERMS OF OTHER FAMILIES, AND THERE ARE REALLY THREE THINGS THAT EMERGE FROM OUR WORK. ONE IS INTERPRET THE GINO EARLY IN HIVE LIFE TO IDENTIFY
INDIVIDUALS AT RISK FOR PREMATURE M.I. SECOND IS DELIVER PROVEN RISK REDUCING INTERVENTIONS, LIFESTYLE AND/OR MEDICATIONS TO THOSE INDIVIDUALS, AND LASTLY, DEVELOP NEW TREATMENTS THAT MIMIC NATURALLY OCCURRING RESISTANCE MUTATIONS, AND THIS THIRD POSSIBILITY IS WHAT I REALLY HOPE TO ACCOMPLISH IN MY NEW EFFORT. SO LET ME CLOSE THEN HERE, WHAT I’VE TRIED TO TALK ABOUT IS THE GENETIC BASIS OF RISK AND THE GENETIC BASIS OF RESISTANCE. RISK MUTATIONS AND REALLY IDENTIFY THOSE LIKELY TO BENEFIT FROM EARLY INTERVENTION AND PROTECTIVE MUTATIONS CAN POINT THE WAY TO CAUSAL PATHWAYS IN NEW MEDICINES. THANK YOU VERY MUCH.>>THANKS FOR A TERRIFIC PRESENTATION. WE HAVE TIME FOR SOME QUESTIONS. THERE ARE MICROPHONES IN THE AISLES. I SEE DR. GREEN APPROACHING THE MICROPHONE.>>THAT WAS TERRIFIC. I HAVE A QUESTION. YOU SAID THERE WERE SEVERAL DIFFERENT APPROACHES OR ALGORITHMS USED FOR CALCULATING THE POLYGENIC RISK SCORE BUT THEN ONE EMERGED AS BETTER THAN THE OTHERS. I GUESS MY QUESTION WAS, IN VERY SIMPLE TERMS, WHAT WERE THE DIFFERENCES BETWEEN THEM AND WHAT MAKES YOU THINK THAT THE ONE THAT YOU’RE USING NOW IS THE OPTIMAL OR BEST ONE THAT THERE MIGHT NOT BE BETTER METHODS THAT WOULD EMERGE?>>I THINK IT’S CERTAINLY POSSIBLE THAT BETTER METHODS WILL EMERGE. THIS IS A VERY ACTIVE VARIANT DEVELOPMENT. I THINK THE KEY DISTINGUISHING POINT AT THIS POINT AT THIS STAGE IS SHOULD YOU LIMIT YOUR ANALYSIS TO THE TOP SNPS OR SHOULD YOU GO GENOME-WIDE. AND IT’S VERY CLEAR THAT IF YOU GO GENOME-WIDE, ROUGHLY YOU GET TWICE AS GOOD PREDICTION AS IF YOU LIMIT TO JUST THE TOP SNPs. AND THAT’S BECAUSE WHEN YOU IF
GO GENOME-WIDE — UP ARE LUMENT PREDICTION ACTUALLY IS THE HERITABILITY OF THE DISEASE. SO THAT’S KIND OF WHERE WE ARE RIGHT NOW BUT THE ACTUAL ALGORITHMS ARE VERY ACTIVE AREA OF DEVELOPMENT.>>YOU SAID THERE’S NO REAL CORRELATION AND YOU SHOWED A PICTURE OF THAT BETWEEN THE RISK FACTORS WE TRADITIONALLY IDENTIFY FOR CARDIOVASCULAR DISEASE AND THE POLYGENIC RISK SCORE BUT YOU WERE LUMPING THEM ALL TOGETHER. I GUESS I WOULD BE SURPRISED IF THERE’S NOT SOME CORRELATION WITH FAMILY HISTORY AT THE POLYGENIC RISK SCORE, IS THERE?>>ONE WAY TO LOOK AT THIS IS TAKE THE TOP 5% SCORE, TAKE EVERYBODY ELSE AND ASK WHAT’S THE FRACTION OF PEOPLE WHO — FAMILY HISTORY IN THE TWO GROUPS. TURNS OUT IF YOU TAKE THE TOP 5% OF POLYGENIC SCORE, 37% SAY THEY HAVE A FAMILY HISTORY OF HEART ATTACK AND EVERYBODY ELSE, ABOUT 30%. SO THIS IS — THERE ARE A NUMBER OF REASONS, ONE IS SELF-REPORTED FAMILY HISTORY IS QUITE INACCURATE. EVERYBODY THINKS SOMEBODY IN THE FAMILY HAD A HEART ATTACK AND IT’S RARELY THE CASE. OBJECTIVELY SHOWN IN THE FRAMINGHAM HEART STUDY WHERE YOU HAVE BOTH THE PARENTS AND THE OFFSPRING AND YOU’VE ASKED OFFSPRING IF THEY HAVE A FAMILILY HISTORY OF HEART ATTACK AND YOU KNOW IF THE PARENT DID OR DID NOT AN THERE’S MISCLASIFICATION ROUGHLY 40% OF THE TIME. BUT THE OTHER POINT I THOUGHT YOU WERE GOING TO ASK, THERE HAS TO BE SOME CORRELATION BETWEEN THE SCORE AND KNOWN RISK FACTORS LIKE LDL AND SO FORTH, AND THERE; ACTUALLY, AND THE CORRELATION BETWEEN THE ACCH — EQUATION AND THE POLYGENIC SCORE IS VERY LOW. 0.02. THAT’S BECAUSE THE MAIN DRIVER OF THE POOLED CORDS EQUATION IS AGE. THERE’S ALMOST NO CORRELATION BETWEEN AGE AND GENOTYPE, RIGHT, SO THAT’S KIND OF — YOU KNOW, BUT IF YOU TOOK RISK FACTOR BY RISK FACTOR AND LOOKED AT THE RELATIONSHIP. YOU’LL SEE SOME.>>BACK IN 1992, I PUBLISHED A STUDY IN THE ARCHIVES OF PATHOLOGY AND LABORATORY MEDICINE ABOUT VERY HIGH LEVELS OF HDL THAT WE FOUND PEOPLE WITH HDL OVER 100 MILLIGRAMS PER KES LITER AND ABOUT 40 SOME PEOPLE IN NEW YORK AND I FOUND BACK IN THOSE DAYS WE DIDN’T HAVE MONEY TO DO THINGS SO WE JUST WENT THROUGH THE CHARTS AND WE FOUND THAT THEY HAD THINGS LIKE THEY WERE PERIMENOPAUSAL FEMALES OR ON ANTIEPILEPTIC MEDICATION OR IT WAS A DRUNKEN STUPOR OR THEY WERE EXPOSED TO INSECTICIDES AS THE CAUSE, BUT TO MAKE A LONG STORY SHORT, MOST OF THEM DIDN’T HAVE — THEY SHOWED IN THEIR HISTORIES THAT SOME OF THEM HAD MYOCARDIAL INFARCTIONS IN SPITE OF HAVING THIS VERY HIGH LEVEL AND THERE WERE SOME — I REMEMBER THERE WAS SOME PAPER THAT CITED FAMILIAL GROUPS THAT HAD ELEVATED GENETICALLY HIGH HDL WHICH THEY THOUGHT ACCOUNTED FOR THEIR INCREASED LIFE EXPECTANCY. AND APPARENTLY YOU’VE DEBUNKED ALL THAT, BUT AREN’T THERE SOME FAMILIAL STUDIES THAT STILL MIGHT SUPPORT THAT AND HOW ABOUT PEOPLE WHO HAVE GENETICALLY LOW HDL, WOULD THAT HAVE AN EFFECT ON THE RISK?>>THAT’S A GREAT QUESTION. SO WE HAVE LOOKED AT THE LOW, WE AND ACTUALLY OTHERS BEFORE US, FROM COPENHAGEN. THE NUMBER ONE CAUSE OF LOW HDL FROM A GENETICS PERSPECTIVE IS A GENE CALLED ADCA1. IF YOUR HETEROZYGOUS NULL FOR ADCA1 YOU HAVE QUITE LOW HDL. AND WE LOOKED TO SEE WHETHER THOSE INDIVIDUALS WERE AT HIGHER RISK FOR HEART ATTACK AND THEY’RE NOT. SO I THINK IT’S PRETTY SAFE TO SAY THAT THE HDL CHOLESTEROL, WHICH IS WHAT STARTED THE WHOLE FIELD, THAT IT’S NOT A CAUSAL PATH. NOW THERE’S A NEW HYPOTHESIS NOW WHERE IT’S NOT ABOUT THE CHOLESTEROL BUT IT’S ABOUT THE FUNCTION OF HDL, WHICH IS AGAIN THE NEW HYPOTHESIS AND THAT WOULD NEED TO BE TESTED.>>OKAY, THANK YOU.>>OVER HERE.>>I WAS REALLY INTRIGUED BY YOUR SMOOTHIE MODEL. SO FOR EXAMPLE, YOU VERY NICELY ASSURED THAT THIS ENVIRONMENT CAN ATTRACT THE GENOME PEOPLE LIKE HAVING HARMLESS MUTATIONS CAN LEAD TO THESE KIND OF PROBLEMS BUT I WAS WONDERING, YOU SHOWED MOSTLY FOR THE DIET. COULD YOU COMMENT ON THE ROLE OF SLEEP OR — FEEDING, WHICH HAS A LOT OF BENEFITS ON MANY OTHER DISEASES, SO THIS HAS BEEN SHOWN IN THE MICE AS WELL AS IN HUMANS.>>SO THERE’S SO MANY IMPORTANT LIFESTYLE FACTORS THAT CLEARLY INTERACT WITH GENETICS TO LEAD TO COMMON COMPLEX DISEASE. IN FACT THE MAJORITY OF THE RISK FOR HEART ATTACK IS NON-GENETIC. AND I THINK A FEW OF THOSE FACTORS HAVE BEEN VERY WELL INVESTIGATED, SMOKING, PHYSICAL ACTIVITY, FOR EXAMPLE, BUT THERE’S A LOT MORE LIKE SLEEP AND OTHERS WHERE IT’S JUST GETTING STARTED. IN FACT, LIKE THE ALL-OF-US RESEARCH PROGRAM OR U.K. BIOBANK, THERE’S VERY LARGE STUDIES WHERE THERE’S BEEN VERY NICE ASSESSMENTS THAT HAVE BEEN ALREADY DONE OR STARTING TO BE DONE, SHOULD HELP US TEASE OUT THESE FACTORS BUT ALSO SOCIAL DETERMINANTS OF HEALTH WHICH IS OF COURSE A KEY COMPONENT FOR MANY COMMON COMPLEX DISEASES AS WELL.>>I WAS GOING TO ASK YOU A QUESTION RELATED TO DR. COLLINS’ QUESTION BECAUSE IF I UNDERSTOOD YOU CORRECTLY, THE CORRELATION BETWEEN THE POLYGENIC RISK SCORE AND THE ACCHA PREDICTION IS VERY LOW, THE QUOTIENT IS 4, BUT STILL THE ADDED PREDICTIVE VALUE OF THE POLYGENIC RISK SCORE ON TOP OF WHAT WE CAN ALREADY PREDICT SEEMED RELATIVELY SMALL, AND DO YOU THINK THAT’S JUST BECAUSE ASIDE FROM AGE, IT DOES ACTUALLY CORRELATE WITH A LOT OF THE OTHER RISK FAG TORES THAT WE
FACTORS
WE KNOW ABOUT?>>IT’S KIND OF THE EYE OF THE BEHOLDER, WHAT IS THE PORP
IMPORTANT INCREMENTAL VALUE. IF YOU GO BACK TO THAT, IT ALL DEPENDS ON WHAT YOU START WITH, OKAY? SO IF YOU START WITH — YOU KNOW, YOU COULD ACTUALLY ARGUE WHAT WE’RE BORN WITH IS WHAT YOU KNOW INITIALLY IS THE GENOTYPES, AGE AND GENDER. IF YOU JUST LOOK AT THOSE THREE VALUES IN A PREDICTIVE MODEL AND START THERE, YOU’RE GOING TO END UP GETTING TO AN AUC PRETTY CLOSE TO BASICALLY HERE. IT’S GOING TO BE VERY CLOSE TO CONVENTIONAL RISK FACTORS. SO YOU CAN SEE IF YOU JUST LOOK AT EACH INDIVIDUAL RISK FACTOR — IT’S ALL WHAT YOU START WITH. RIGHT NOW WE’RE NOT MEASURING GENOTYPE BUT YOU CAN IMAGINE AN ERA, IN FACT, ACTUALLY WE DO HAVE THE POSSIBILITY IN 20 MILLION PEOPLE IN THE UNITED STATES, RIGHT NOW, THEY ALREADY HAVE THEIR GENOTYPES, EITHER THROUGH “NATIONAL GEOGRAPHIC,” ANCESTRY OR 20 ME, 23 ME,
THAT’S ALL YOU NEED TO CALCULATE THESE SCORES. IT ALL DEPENDS ON A COUPLE THINGS, WHAT YOU START WITH AND WHAT YOU CONSIDER TO BE THE APPROPRIATE THRESHOLD IN TERMS OF INCREMENTAL VALUE.>>THINKING ABOUT THE GENETIC ARCHITECTURE OVERALL, HOW DO YOU CONCEPTUALIZE THE POLYGENIC RISK BE SCORE IN THE COP CONTEXT OF
THE MONOGENIC SO WHEN YOU HAVE BOTH, WHAT DO YOU HAVE? DO YOU IGNORE THE PRS WHEN YOU HAVE THE MONOGENIC OR IS THE PRS A MODIFIER OF THE MONO JEPG OR HOW ARE YOU THINKING ABOUT THAT?>>WE’RE JUST STARTING TO GET DATA ON THAT. IT’S A QUESTION THAT CAN EMPIRICALLY ADDRESSED AND IT TURNS OUT IT’S ENTIRELY ADDITIVE OR SUBJECTIVE. SO WHAT SUBTRACTIVE. IF YOU ARE A HIGH POLYGENIC RISK, IT’S LIKE 4 PLUS 4 IS 8, BUT IF YOU CARRY AP FH MUTATION AND YOU’RE LOW POLYGENIC RISK, YOU ACTUALLY COME BACK DOWN TO THE BASELINE LEVEL. AND THE SAME FOR BRCA. SO THIS EXPLAINS, I THINK — CAN EXPLAIN A FAIR BIT OF THE PENETRANCE ISSUE FOR THE MONOGENIC MUTATIONS.>>AND NO EPISTASIS.>>IT’S ADDITIVE.>>ONE MORE THING, AND LIFESTYLE IS ADDED ON TOP OF THAT.>>ONE MORE QUESTION.>>MY QUESTION IS, BE WOULD YOU BE ABLE TO MAKE THOSE DATA PUBLICLY AVAILABLE? BECAUSE YOU CAN ONLY — YOU’RE LUCKY IF YOU CAN GET A THOUSAND SUBJECTS BUT YOU HAVE 10,000, 100,000, THAT WOULD BE GREAT, CAN YOU MAKE DATA AVAILABLE FOR PEOPLE LIKE CARDIOVASCULAR DISEASE, THAT WOULD BE GREAT. THANK YOU.>>YEAH, THAT’S A GREAT QUESTION. SO ALL THE WORK THAT — MUCH OF THE WORK I DESCRIBED IS BASED ON U.K. BIOBANK DATA WHICH IS PUBLICLY AVAILABLE AND THEN THE MODELS THAT WE’VE USED IN TERMS OF THE WEIGHTS AND THE UP? s THAT SNPs THAT GO INTO EACH OF THE SCORES, THOS ARE ALL PLUCKILY AVAILABLE AS WELL.>>I SAID LAST QUESTION BUT SINCE WIN ARIAS CAME TO THE MICROPHONE THIS, REALLY IS THE LAST QUESTION.>>IS C REACTIVE PROTEIN AND OTHER METHODS, OTHER MANIFESTATIONS OF INFLAMMATION UNDER PENT VARIABLES IN THE PRODUCTION OF A MYOCARDIAL INFARCTION?>>YES. SO IF YOU ADDED CRP INTO THE MODEL, IT WOULD BE AN INDEPENDENT VARIABLE. SO IT HASN’T BEEN SHOWN IN THESE — IN MY PLOTS BUT YES, IT WOULD BE. IT’S CAPTURING SOMETHING SLIGHTLY DIFFERENT FROM WHAT ELSE IS HAPPENING, AND JUST RECENTLY THERE’S BEEN SOME VERY GOOD EVIDENCE THAT DIRECTLY ANTAGONIZING INFLAMMATION WITH A MONO MONOCLONAL ANTIBODY AGAINST IL1 BETA IN A RANDOMIZED CONTROL TRIAL OF PATIENTS WHO HAD ALREADY HAD A HEART ATTACK ON OPTIMAL THERAPY FOR LIPID, HAVE OF THEM GOT THIS MONOCLONAL ANTIBODY THAT TARGETS INFLAMMATION, ONE INFLAMMATORY PROTEIN, THE OTHER HALF GOT PLACEBO. THERE WAS A REDUCTION IN CLINICAL EVENTS. REALLY THE FIRST HUMAN EVIDENCE FROM AN RCT PERSPECTIVE, RANDOMIZED CONTROL TRIAL PERSPECTIVE, SHOWING THAT TARGETING INFLAMMATION DIRECTLY CAN MAKE A DIFFERENCE.>>WELL, THIS HAS BEEN A WONDERFUL PRESENTATION WITH A GREAT DISCUSSION, AND WE CAN CONTINUE THE DISCUSSION IN THE MEDICAL LIBRARY IF YOU’D LIKE WITH COFFEE AND COOKIES, AND I’M PLEASE, LET’S THANK OUR SPEAKER, SEKAR KATHIRESAN, AGAIN. [APPLAUSE]

3 thoughts on “Genes, lifestyle, and risk for heart attack

  1. This was a great presentation. Thank you NIH for posting it. The most interesting part for me was learning that high HDL cholesterol, while correlated with reduced risk of heart attack, is NOT itself causal. It is what is called a non-causal marker. Triglyerides, specifically low triglyceride-rich lipoproteins (TRL), are really what is important for reducing heart attack risk. That part of the presentation starts about 41 minutes in.

  2. Thank you NIH for sharing this informative video~! 😃 For anyone who's interested, Dr.Sekar's researches are available at: http://www.kathiresanlab.org/ 👈

Leave a Reply

Your email address will not be published. Required fields are marked *