Aggressive Cholesterol Management to Prevent CHD in Diabetes Wm. James Howard, M.D. Washington Hospital Center Washington, D.C. May 11. 2005

Prevalence of Obesity in the United States The prevalence of obesity among US adults increased by 61% from 1991 to 2000, according to data obtained through the Behavioral Risk Factor Surveillance System (BRFSS), which conducted telephone interviews with 184,450 people aged 18 years and older in 50 states in 2000. According to BRFSS data, 38.8 million US adults (19.6 million men, 19.2 million women) were obese (body mass index [BMI] ³30 kg/m2) in the year 2000. The prevalence of obesity in the United States was estimated to be 19.8% in 2000 (12% in 1991), the prevalence of diagnosed diabetes was 7.3% (4.9% in 1990), and the prevalence of combined obesity and diabetes was 2.9% (1.4% in 1991). In addition, 56.4% of US adults were overweight (BMI ³25 kg/m2) in 2000 (45% in 1991), and 2.1% were extremely obese (BMI ³40 kg/m2), compared with 0.9% a decade ago. Of the states participating in the BRFSS survey, only four had obesity rates of 15% or greater in 1991, but in 2000, all 50 participating states had rates of 15% or greater. None of the states had obesity rates of 20% or greater in 1991, whereas in 2000, 22 did. Mississippi had the highest rate of obesity (24.3%) in 2000, and Colorado had the lowest (13.8%). The 2000 obesity rates by race in descending order were: blacks (29.3%), Hispanics (23.4%), whites (18.5%), and other races (12%). Mokdad AH et al. JAMA. 2001;286:1195-1200.

Prevalence of Diabetes in the United States The prevalence of diagnosed diabetes among US adults increased by 49% in a single decade, from 4.9% in 1990 to 7.3% in 2000. If undiagnosed diabetes is taken into account, it is likely that as many as 10% of US adults have the disease. Based on data obtained through the Behavioral Risk Factor Surveillance System (BRFSS), which conducted telephone interviews with 184,450 people aged 18 years and older in 50 states in 2000, the investigators calculated that approximately 15 million US adults (6.3 million men, 8.7 million women) had diagnosed diabetes in 2000. The prevalence of combined diabetes and obesity was 2.9% in 2000, compared with 1.4% in 1991. Mississippi had the highest rate (8.8%) of diagnosed diabetes in 2000, and Alaska had the lowest (4.4%). In 1990, only four states had diabetes rates of 6% or greater; in 2000, 43 of 50 states had this prevalence rate. Blacks had the highest rate (11.1%) of diagnosed diabetes in 2000, followed by Hispanics (8.9%), other races (6.7%), and whites (6.6%). Analysis of the 2000 data according to education showed that people with less than a high school education had the highest rate (12.9%) of diagnosed diabetes. Mokdad AH et al. JAMA. 2001;286:1195-1200.

Mortality in People With Diabetes: Causes of Death Whereas age-adjusted mortality for many of the leading causes of death, including cardiovascular disease (CVD), stroke, and cancer, have fallen since 1980, corresponding mortality for diabetes has increased by 30%. Although mortality has decreased for CVD, heart disease remains the leading cause of death in the United States. Among people with type 1 and type 2 diabetes, CVD accounts for >50% of deaths.   ADA. Summary of the Report and Recommendations of the Congressionally Established Diabetes Research Working Group, 1999. Minino AM, Smith BL, eds. National Vital Statistics Reports. Vol 49. No. 12. Washington, DC:CDC; 2001. Geiss LS, Herman WH, Smith PJ. In: Diabetes in America. 2nd ed. 1995; chap 11.

Atherosclerosis in Diabetes About 80% of all diabetic mortality (75% from coronary atherosclerosis; 25% from cerebral or peripheral vascular disease) >75% of all hospitalizations for diabetic complications >50% of patients with newly diagnosed NIDDM have CHD

Evolution of the Treatment Approach 1970s NCEP ATP I Guidelines 1988 ATP II 1993 ATP III 2001 Framingham MRFIT LFC-CPPT Coronary Drug Project Helsinki Heart CLAS (anglo) Angiographic Trials (FATS, POSCH, SCOR, STARS, Omish, MARS) Meta-Analyses (Holme, Rossouw) 4S, WOSCOPS CARE, LIPID, AFCAPS/TexCAPS, VAHIT, Others

New Features of ATP III 1. Type 2 Diabetes Mellitus CHD Risk Equivalents: 1. Type 2 Diabetes Mellitus 2. Non-Cardiac Forms of Athero. 3. Framingham Projection of 10 yr. Risk >20% (identifies individuals with multiple risk factors in need of more aggressive lipid lowering) The Metabolic Syndrome

Diabetes Mellitus As CHD Risk Equivalent Increased CHD Risk: Women -- 4-6 fold, Men-- >2 fold Risk for a person with DM having 1st MI is equal to a non-DM having 2nd MI DM more likely to die before reaching hospital with 1st MI DM confers worse prognosis in hospital and during first year after discharge >50% of DM have CHD at diagnosis

The Metabolic Syndrome AKA: Pleuri-Metabolic Syndrome Insulin Resistance Syndrome Syndrome X (Metabolic) Deadly Quartet Multiple Metabolic Syndrome

The Metabolic Syndrome General Features of the Metabolic Syndrome: Abdominal obesity Atherogenic dyslipidemia Elevated triglycerides Small Dense LDL particles Low HDL cholesterol Raised blood pressure Insulin resistance ( glucose intolerance) Proatherosclerotic state Prothrombotic State Proinflammatory State

HAFFNER’S TICKING CLOCK HYPOTHESIS: The “Atherosclerosis Clock” starts ticking when Insulin Resistance develops. The “Clock” advances faster when hyperglycemia develops. The “Clock” begins to run-away when overt diabetes develops. Hence, by time of diagnosis, > 50% of DM have clinical CHD. Haffner

NHANES III Conclusions The Metabolic Syndrome: Prevalence per ATP III definition Overall: 23.7% Mexican-Americans: highest age-adjusted Prevalence: 31.9% 2000 census data Approximately 47 million Americans Ford ES, et al. JAMA. 2002;287:356-359.

Prevalence of Diabetes Strong Heart Study, by Gender and Center Women Men %

Non-HDL Cholesterol (Non-HDL Chol. = TC - HDL) Known predictor of CHD in epidemiology Equivalent to total apo B-100, and TC/HDL Represents the sum of LDL, Lp(a), IDL, and VLDL: All atherogenic apo B containing lipoproteins Lipid Equivalent of “HbA1C”

Diabetes Prevention Program (DPP 2) Lifestyle changes consisting of diet and exercise reduced the conversion of IGT to Type 2 Diabetes by 58% NEJM, 346; 393; 2002

Effect of Pioglitazone on Lipid Levels In a 26-week double-blind, placebo-controlled trial, 408 patients with type 2 diabetes were randomized to treatment with either placebo or one of the following doses of pioglitazone daily—15 mg, 30 mg, or 45 mg. Administration of any previous antidiabetic agent was discontinued 8 weeks prior to treatment. The effect of pioglitazone on blood lipid levels was evaluated. At doses of 15 mg, 30 mg, and 45 mg daily, pioglitazone was associated with mean reductions in TG levels ranging from 9% to 9.6% of baseline values. This was compared with 4.8% increase in TG level with placebo. In this monotherapy study, the group taking 45 mg once daily showed a statistically significant (P£0.05) increase in HDL-C vs placebo. The change from baseline was 19.1%. Increases in LDL-C of 5.2% to 7.2% were observed with pioglitazone. Aronoff S et al. Diabetes Care. 2000;23:1605-1611.

Non-Diabetic Patients HOT Trial Effect of Diastolic Target on Cardiovascular Events - 4 Years 30 48% Risk Reduction 24.4 20 Events /1000 Pt-Yrs 18.6 10 11.9 9.9 10.0 9.3 10 10 10 10 10 10 Diabetic Patients n=1, 501, P=0.016 Non-Diabetic Patients n=18, 790, P=NS

Common Lipoprotein Abnormalities Diabetic Dyslipidemia

CVD Hazard Ratios by Quartile of LDL Cholesterol in Diabetes The Strong Heart Study Howard et al. ATVB 2000;20(3):830

The Pyramid of Recent Trials Relative Size of the Various Segments of the Population

HMG CoA Reductase Inhibitors (Statins) Statin Dose Range Lovastatin 20–80 mg Pravastatin 20–40 mg Simvastatin 20–80 mg Fluvastatin 20–80 mg Atorvastatin 10–80 mg Rosavustatin 5--40mg Cerivastatin 0.4–0.8 mg

Relation Between CHD Events and LDL-C in Recent Statin Trials 30 4S-PI 2° Prevention 25 4S-Rx 20 % with LIPID-Rx 15 CHD event LIPID-PI CARE-Rx 1° Prevention CARE-PI 10 WOSCOPS-PI AFCAPS/TexCAPS-PI 5 WOSCOPS-Rx AFCAPS/TexCAPS-Rx 90 110 130 150 170 190 210 Mean LDL-C level at follow-up (mg/dL) PI=placebo; Rx=treatment Shepherd J et al. N Engl J Med. 1995;333:1301-1307. 4S Study Group. Lancet. 1995;345:1274-1275. Sacks FM et al. N Engl J Med. 1996;335:1001-1009. Downs JR et al. JAMA. 1998;279:1615-1622. Tonkin A. Presented at AHA Scientific Sessions, 1997.

Non-Statin Lipid Lowering Drugs Niacin—extended release, OTC immediate Bile Acid Sequestrants—colesevelam Fibric Acids—gemfibrozil, fenofibrate Intestinal acting—ezetimibe Omega 3 fatty acids—fish oil (EPA, DHA) Dietary adjuncts—plant sterol/stanol ester margerines, viscous fiber supplements

Key Point: Lipid-lowering therapy using complementary mechanisms modulates cholesterol metabolism on 2 fronts, synthesis and absorption, using agents with complementary actions. 23

Mechanism of Intestinal-Acting Agents Intestinal-acting agents lower cholesterol by distinct mechanisms. Bile acid sequestrants inhibit bile acid reabsorption in the ileum, causing hepatic bile acid deficiency. This deficiency promotes compensatory increase in bile acid synthesis from hepatic cholesterol, which is replenished through increased hepatic uptake of LDL/chylomicron from plasma and increased hepatic cholesterol synthesis. Therefore, bile acid sequestrants reduce LDL-C through increased clearance of LDL particles by the liver. Plant stanols and sterols are thought to displace cholesterol from micelles, preventing its uptake at the brush border membrane and reducing the amount of cholesterol transported to the liver. Reduced delivery of dietary/biliary cholesterol to the liver effectively increases clearance of LDL (and LDL-C) particles from plasma. Ezetimibe selectively inhibits the uptake of micellar cholesterol into intestinal epithelial cells and substantially reduces the amount of cholesterol from diet and bile that is transported to the liver, thereby resulting in compensatory increase in LDL clearance by the liver and reduction in plasma LDL-C levels. The mechanism of action of ezetimibe is thought to involve selective inhibition of the putative sterol transporter on the brush border surface of intestinal epithelial cells.

Ezetimibe: Inhibition of Cholesterol Absorption LDL Apo B100 Liver Carey 1994, p 722-A Duodenum Jejunum VLDL Apo B100 Ezetimibe Inhibits Absorption (—) Ileum Key Point: Reduced absorption of intestinal cholesterol results in decreased delivery of cholesterol to the liver. Additional Background Information: Inhibition of cholesterol absorption leads to a decrease in the delivery of intestinal cholesterol to the liver and a reduction of hepatic cholesterol stores, resulting in an increase in clearance of cholesterol from the blood. Reducing hepatic cholesterol results in fewer VLDL particles being secreted into the circulation and converted to LDL particles.1 Inhibition of cholesterol absorption has been associated with increases in cholesterol synthesis.2,3 References: 1. McKenney J. Combination therapy for elevated low-density lipoprotein cholesterol: the key to coronary artery disease risk reduction. Am J Cardiol. 2002;90(suppl):8K–20K. 2. Sudhop T, Lutjohann D, Kodal A, et al. Inhibition of intestinal cholesterol absorption by ezetimibe in humans. Circulation. 2002;106:1943–1948. 3. Altmann SW, Davis HR Jr, Zhu LJ, et al. Niemann-Pick C1 Like 1 protein is critical for intestinal cholesterol absorption. Science. 2004;303:1201–1204. CM Apo B48 CM Remnant Apo B48 Colon Adapted with permission from Carey MC, Duane WC. In: Arias IM et al, eds. The Liver: Biology and Pathobiology. Raven Press; 1994. 35

Dosage and Administration Patients should be on a standard cholesterol-lowering diet* Dosage should be individualized according to baseline LDL-C, recommended goal of therapy, and patient response* Dosage range: 10/10 mg/day–10/80 mg/day Usual recommended starting dose: 10/20 mg/day Patients requiring a larger reduction in LDL-C (>55%) may be started at a dose of 10/40 mg/day; 10/10 mg/day may be considered for patients requiring less aggressive LDL-C reductions Lipid levels may be analyzed and dosage adjusted 2 or more weeks following initiation or titration of therapy VYTORIN should be taken once daily in the evening, with or without food *See NCEP ATP III Guidelines.

VYTORIN Lowered LDL-C by 52% at the Starting Dose 10/20 mg (n = 156) 10/40 mg (n = 147) 10/80 mg (n = 154) VYTORIN 10/20 mg Mean Baseline LDL-C 176 mg/dL Mean End Point LDL-C 84 mg/dL VYTORIN 10/80 mg Mean Baseline LDL-C 178 mg/dL Mean End Point LDL-C 70 mg/dL Mean Decrease in LDL-C From Untreated Baseline, % –25 –50 –52%*,† –55%*,† –60%* VYTORIN lowered LDL-C more than simvastatin across the dosage range Simvastatin lowered LDL-C by 34% at the 20-mg dose, 41% at the 40-mg dose, and 49% at the 80-mg dose Comparison With Simvastatin *P<0.001 for VYTORIN vs each corresponding dose of simvastatin. †P<0.001 for VYTORIN vs next highest dose of simvastatin monotherapy.

VYTORIN Provided Significantly Greater LDL-C Reductions vs Atorvastatin Starting Doses (mg) 10/20 (n = 250) 10 (n = 262) 20 (n = 246) 10/40 mg (n = 482) 40 mg (n = 237) 10/80 mg (n = 459) 80 mg (n = 228) –10 VYTORIN 10/20 mg Mean Baseline LDL-C 179 mg/dL Mean End Point LDL-C 89 mg/dL Atorvastatin 10 mg Mean Baseline LDL-C 181 mg/dL Mean End Point LDL-C 112 mg/dL –20 Mean Decrease in LDL-C From Untreated Baseline, % –30 –40 –37% –44% –50 –50%* –49% –53% –60 –56%† –59%‡ –70 *P<0.001 for VYTORIN 10/20 vs atorvastatin 10 mg; P≤0.05 for VYTORIN 10/20 vs atorvastatin 20 mg. †P≤0.05 for VYTORIN 10/40 vs atorvastatin 40 mg. ‡P<0.001 for VYTORIN 10/80 vs atorvastatin 80 mg. The clinical significance of comparative lipid effects has not been established. Ballantyne CM et al. Am J Cardiol. 2004;93:1487–1494.

VYTORIN Provided Excellent HDL-C Efficacy Starting Doses (mg) 10/20 (n = 250) 10 (n = 262) 20 (n = 246) 10/40 mg (n = 482) 40 mg (n = 237) 10/80 mg (n = 459) 80 mg (n = 228) 12%‡ 11%* 9%*,† Mean Increase in HDL-C From Untreated Baseline, % 8% 7% 6% 5% *P≤0.05 for VYTORIN 10/20 mg vs atorvastatin 10 mg and for VYTORIN 10/40 mg vs atorvastatin 40 mg. †P=NS for VYTORIN 10/20 vs atorvastatin 20 mg. ‡P<0.001 for VYTORIN 10/80 vs atorvastatin 80 mg. The clinical significance of raising HDL-C has not been established.

COMPARATIVE COST per Month Vytorin 10/10,20,40,80 $84.24 Zetia 77.77 Zocor 10 mg. 79.02 20 mg. 137.87 40 mg. 137.87 80 mg. $ 137.87 Medical Letter (9/13/04); 46,73,2004

Current Medical Research and Opinions, 2002; 18: 220-227 GREek Atorvastatin and Coronary Heart Disease Evaluation Study GREACE TRIAL Current Medical Research and Opinions, 2002; 18: 220-227

GREACE TRIAL RESULTS: Total Mortality -43% CHD Mortality -47% non fatal MI -59% Revascularization -51% CHF -50% Stroke -47% Women -54% Diabetics -58% 60-75 yoa -49%

Heart Protection Study (HPS) Design Large, multicenter, placebo-controlled, double-blind study Mean duration: 5 years Patients (N=20,536, 97% Caucasian) allocated* to Simvastatin 40mg/day (n=10,269) Placebo (n=10,267) Mean age 64 years (range 40 to 80 years) Patients were at high risk of a major coronary event because of Existing coronary heart disease (CHD) (65%) Diabetes (type 2, 26%; type 1, 3%) History of stroke or other cerebrovascular disease (16%) Peripheral vessel disease (33%) Hypertension in males aged 65 years and older (6%) * Patients were allocated to treatment using a covariate adaptive method, which took into account the distribution of 10 important baseline characteristics of patients already enrolled and minimized the imbalance of those characteristics across the groups.

HPS: MCE by Metabolic History Baseline Characteristics n Incidence(%) MCE Risk Ratio (95% CI) Simvastatin Placebo Diabetes mellitus 5,963 9.4 12.6 Without CHD 3,982 5.5 8.4 With CHD 1,981 17.4 21.0 Without diabetes mellitus 14,573 8.5 11.5 0.4 0.6 0.8 1.0 1.2 ▼ Favors simvastatin Favors placebo

HPS: Primary and Secondary Prevention Implications CARE 4S LIPID 25 20 15 % with CAD event HPS (estimated) 10 WOSCOPS AFCAPS Results of HPS, together with the results from the WOSCOPS and AFCAPS/ TexCAPS trials, show that the benefits of LDL-C lowering with statin therapy in primary prevention of CAD are seen across the broad range of baseline plasma LDL-C concentrations. Together with the results from the 4S, CARE, and LIPID trials, HPS has also established the benefits of LDL-C reduction with statin therapy in secondary prevention of CAD in patients with a wide range of baseline plasma LDL-C concentrations. HPS (estimated) 5 50 70 90 110 130 150 170 190 210 LDL-C (mg/dL) Adapted from Illingworth. Med Clin North Am. 2000;84:23. At: http://www.hpsinfo.org. Illingworth DR. Management of hypercholesterolemia. Med Clin North Am. 2000;84:23-42. At: http://www.hpsinfo.org.

Additional Background Information: There is a linear relationship between LDL-C and the risk for CHD events in individuals with and without CHD. This relationship extends to very low levels of LDL-C.1,2 Additional Background Information: Results from a large number of primary and secondary prevention studies with clinical events (eg, CHD, MI) as the primary end point, showed that the risk for specific CHD events decreased in the statin-treated group vs the placebo-treated group. Importantly, combined analysis of the results from these studies indicated no clear lower limit to the relationship between reduced LDL-C and decreased CHD risk. This linear relationship is strongly supported by the aggregate clinical trial evidence.1 Not all statins are indicated to lower risk of CHD events. See prescribing information for specific product uses. References: 1. Ballantyne CM. Low-density lipoproteins and risk for coronary artery disease. Am J Cardiol. 1998;82:3Q–12Q. 2. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20 536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002;360:7–22. 13

New Category: Very High Risk Patients Definite CHD plus additional risk factors, such as diabetes, significant hypertension etc. LDL goal < 100 mg/dl with optional goal of <70 mg/dl. Initiate drug therapy if LDL > 100 mg/dl with consideration for drug therapy to reach optional goal of < 70 mg/dl when baseline LDL < 100mg/dl Lower LDL by at least 30%. Circulation 2004;110:227-239

CARDS Primary Prevention Study: 2838 T2DM randomized atorva. 10 mg. or placebo. (+ Additional risk factor) Terminated at 3.9 years—2 years early. End of study LDL: atorva = 78 mg/dl and placebo = 120 mg/dl. End of Study non-HDL: atorva = 100 mg/dl and placebo = 155 mg/dl. No excess of adverse events in atorvastatin group

CARDS RESULTS All Cause Mortality: - 27 % CHD Events: - 36 % Revascularizations: - 31 % Stroke: - 48 % Lancet 2004: 364; 685-696

Clinical Trials of Lipid Lowering to Prevent CHD in Diabetes Trial Results HPS ALL HAT ASCOT CARDS Prevention No Prevention

SANDS Stop Atherosclerosis in Native Diabetics Study

What we learned from SHS Most CVD in SHS communities occurs in those with diabetes LDL cholesterol is a strong predictor even though levels are generally low in Indians Blood pressure is a strong predictor, and it leads to nephropathy which also causes CVD

Inclusion Criteria Diabetic Men and Women >40 yrs LDL>100 mg/dl SBP>130 mm Able to measure carotid IMT

Four Clinical Centers Phoenix area (Charlton Wilson, MD, Marie Russell, MD, Damon Davis, RN) Oklahoma (Brice Poolaw, MD) South Dakota (Jeffrey Henderson, MD) Chinle (Jim Galloway, MD) 496 Men and Women (124/center)

HYPOTHESIS Lowering LDL cholesterol and Blood Pressure to lower targets than are currently recommended will retard CVD Control Intervention LDL chol (mg/dl) <100 <70 SBP (mm) 130/80 115/75

Measure CVD using carotid and cardiac ECHO at baseline Men and women with diabetes over 40 yrs old Usual targets Lower targets Measure CVD using carotid and cardiac ECHO at baseline and after 3 yrs FU

Algorithm for LDL Therapy LDL > target Algorithm for LDL Therapy A Statin (Dose per LDL level) LDL < target Non HDL < target LDL < target Non HDL >target LDL > target Fish Oil Increase Statin Monitor Follow Protocol B LDL <target Non HDL < target LDL <target Non HDL > target Monitor Follow Protocol B LDL < target B LDL < target Non HDL > target LDL < target Non HDL > target LDL < target Non HDL< target Fish Oil Add Fenofibrate or Niacin Monitor

SUMMARY There is a rising tide of CVD in diabetes LDL and blood pressure are strong risk factors We believe SANDS will validate a strategy to prevent/retard CVD in diabetes SHS will continue to work to identify future strategies for therapy or prevention of CVD in diabetes

Diabetes and an Excess of Fat “With an excess of fat diabetes begins and from an excess of fat diabetics die…” - EP Joselin, 1927

Insulin Resistance: Inherited and Acquired Influences Insulin resistance may result from inherited and/or acquired influences on insulin action. Rare forms of severe insulin resistance are caused by mutations coding for proteins in the insulin signaling pathway (eg, mutated insulin receptors). There is evidence that more common forms of insulin resistance have inherited influences, but the precise genetic defects are largely unidentified. Acquired influences on insulin resistance include physical inactivity, overeating and obesity (especially abdominal obesity); aging; certain medications used at pharmacologic doses (eg, glucocorticoids, nicotinic acid); hyperglycemia, which worsens insulin resistance in diabetic patients; and elevated free fatty acids (FFAs). Much of the insulin resistance in the United States and other developed countries today appears to result from a combination of genetic and acquired influences (eg, genes that were selected over millennia to favor survival under conditions of limited food supply now lead to obesity in an environment that provides access to high-calorie foods with little physical exertion).

ASA 20-30% Beta Blockers 20-35% ACE inhibitors 22-25% Statins 25-50% Comprehensive Medical Therapy For Patients with CHD or Other Vascular Disease Risk Reduction ASA 20-30% Beta Blockers 20-35% ACE inhibitors 22-25% Statins 25-50% The four medications every atherosclerosis patient should be treated with, unless contraindications exist and are documented Adapted from the UCLA CHAMP Guidelines 1994

CHAMP ~ Impact on Clinical Outcomes in the First Year Post Hospital Discharge RR 0.43 p<0.01 256 AMI pts discharged in 92/93 pre-CHAMP compared to 302 pts in 94/95 post-CHAMP ASA 78% vs 92%; Beta Blocker 12% vs 61%; ACEI 4% vs 56%; Statin 6% vs 86% Fonarow Am J Cardiol 2001;87;819-822