PREVENTION OF CHD WITH LIPID MANAGEMENT AND ASPIRIN: MATCHING TREATMENT TO RISK Robert B. Baron MD MS Professor and Associate Dean UCSF School of Medicine Declaration of full disclosure: No conflict of interest EXPLAINING THE DECREASE IN DEATHS FROM CHD 1980 to 2000: Death rate fell from: 542.9 to 266.8 per 100K men 263.3 to 134.4 per 100K women 341,745 fewer deaths from CHD in 2000 Ford ES, NEJM, 2007 EXPLAINING THE DECREASE IN DEATHS FROM CHD Placebo-Controlled Statin Trials Reductions in Major Coronary Events Relative to Placebo 47% from CHD treatments, 44% from risk factor modification Reductions in cholesterol: 24% Ford ES, NEJM, 2007 simva 20-40 mg prava 40 mg prava 40 mg simva 40 mg prava 40 mg lova 80 mg 1
Placebo-Controlled Statin Trials Celebrating Successes but Forgetting the Majority? Remaining Major Coronary Events Relative to Placebo A RISK-BASED APPROACH Risk reduction $$ Harm Is there more we can do to identify the non- responders? simva 20-40 mg prava 40 mg prava 40 mg simva 40 mg prava 40 mg lova 80 mg The benefit from any given intervention is a function of: 1) The relative risk reduction conferred by the intervention, and 2) The native risk of the patient 63 yo woman; s/p MI LDL 115 HDL 45 TG 160 The best next step in lipid management is: Continue current therapy Begin a statin to goal LDL 100 Begin a statin to goal LDL 70 Begin a statin plus ezetimibe to LDL goal 70 27% Begin sustained release niacin 1% 70% 0% 2% Continue curre... Begin a statin... Begin a statin... Begin a statin... Begin sustaine... 2
LDL Goal and Cutpoints in Patients with CHD and CHD Risk Equivalents (10-Year Risk >20%) LDL Goal <100 mg/dl Optional : <70 LDL Level at Which to Initiate Diet 100 mg/dl LDL Level at Which to Consider Drug Therapy 100 mg/dl (<100mg/dL: drug optional) Baseline Feature LDL (mg/dl) <100 100 <130 130 ALL PATIENTS Heart Protection Study: Vascular Events by Baseline LDL-C No. Events Statin Placebo (10,269) (10,267) 285 360 670 881 1087 1365 2042 2606 (19.9%) (25.4%) Risk Ratio and 95% Cl Statin better Statin worse 0.4 0.6 0.8 1.0 1.2 1.4 24% reduction (p<0.00001) TREATING TO NEW TARGETS (TNT) TREATING TO NEW TARGETS (TNT) RCT of 10,001 patients with stable CHD; 35-75 yr LDL <130 mg/dl Atorvastatin 10 vs atorvastain 80 Followed for 4.9 years LDL Event % Death % LFTs % Atorv 10 101 10.9 2.5 0.2 Atorv 80 77 8.7 2.0 1.2 p value <0.001 0.09 Research question: safety and efficacy of lowering LDL below 100 mg/dl Larosa NEJM, 2005 3
The best next step in lipid management is: 63 yo man; s/p MI 1. Continue current therapy 2. Begin a statin to goal LDL 100 3. Begin a statin to goal LDL 70 4. Begin a statin plus ezetimibe to LDL goal 70 5. Begin sustained release niacin LDL 70 HDL 25 TG 400 The best next step in lipid management is: In my practice I routinely calculate (and treat) the non-hdl cholesterol: 1. Continue current therapy 2. Begin a statin 3. Begin fenofibrate 4. Begin fish oil 5. Begin niacin (sustained release) 18% 24% 10% 46% 1. Yes 2. No 39% 61% 2% Continue curre... Begin a statin... Begin fenofibr... Begin fish oil... Begin niacin (... Yes No 4
63 yo man; s/p MI {LDL 70, HDL 25, TG 400, Total 175} Total - HDL = Non-HDL 175-25 = 150 NCEP non-hdl goal: LDL goal + 30 = 100 HDL AS A THRAPEUTIC TARGET Systematic review of 31 RCTs Currently available therapies (drug and non-drug) can increase by 20-30% proof that increasing HDL confers reduction in CV outcomes independent of changes in LDL or TG changes has been elusive. Conclusion: only modest evidence to support aggressively increasing HDL beyond lifestyle interventions Singh, JAMA 2007 Management of Low HDL-C Therapeutic lifestyle changes Smoking cessation Regular aerobic exercise Weight loss Alcohol use? CETP INHIBITORS TO INCREASE HDL Inhibition of cholesteryl ester transfer protein markedly raises HDL (and raise BP) Two trials show no benefit on coronary atherosclerosis or carotid intimal medial thickness (compared to atorvastatin alone) Clinical trial of 15,000 stopped prematurely due to excess mortality and events in those on torcetrapib vs. placebo Pfizer abandons development of torcetrapib Nissen, NEJM 2007; Tall, NEJM 2007 5
VA-HIT: Major Coronary Events in Gemfibrozil vs. Placebo Groups Cumulative Incidence (%) 0 22% reduction P = 0.006 Placebo Gemfibrozil 1 2 3 4 5 6 Year Extended-release Niacin vs. Ezetimibe RCT of 208 patients with CHD or CHD riskequivalent on statin at goal LDL Niacin 2000 vs Ezetimibe 10 Outcome: carotid intima-media thickness (IMT) Results: Niacin better Reduction in IMT Major cardiovascular events, 1% vs 5% Taylor NEJM, 2009 Rubins HB et al. N Engl J Med 1999;341:410-418. Fenofibrate plus statin vs. statin alone: ACCORD RCT of 5518 patients with type 2 DM The best next step in lipid management is: Fenofibrate Placebo p CV events: 2.24 2.41.32 Death 1.47 1.61.33 No difference in any secondary outcome Results do not support routine use of combination therapy in DM 1. Continue current therapy 2. Begin a statin 3. Begin fenofibrate 4. Begin fish oil 5. Begin sustained release niacin ACCORD, NEJM 2010 6
If, however, primary prevention? 63 yo woman, no risk factors 1. Continue current therapy 2. Begin a statin 3. Begin fenofibrate 4. Begin fish oil 5. Begin niacin (sustained release) LDL 175 HDL 45 TG 160 The best next step in lipid management is: 1. Continue current therapy 2. Begin a statin 3. Begin fenofibrate 4. Begin a stain plus ezetimibe 5. Begin sustained release niacin Continue curre... 30% Begin a statin... 68% Begin fenofibr... 0% 0% Begin a stain... 2% Begin sustaine... Truth About CVD Risk Prevention Health professionals are not good at judging CV risk Counting risk factors is a blunt instrument and often leads to misclassification Calculate 10-year risk of hard CHD events (CHD death or non-fatal MI) using the Framingham Risk Score 7
DOES GIVING PATIENTS GLOBAL RISK DATA IMPROVE OUTCOMES? Framingham Risk -Limitations 20 studies (14 RCTs) Risk information plus counseling: Improves accuracy of risk perception Increases intent to initiate CHD prevention But, single interventions ineffective Not accurate in patients under 30 or over 75 Provide risk over 4-12 years only Relatively few patients with diabetes; no family history BUT Good discrimination for future CHD events Validated in several populations and found to be relatively transportable for risk ordering but calibration varies Sheridan SL, Archives Intern Med 2010 How Well Does the Framingham Risk Score Perform? COHORT AUC- Men AUC - Women NHANES I 0.73 0.84 NHANES II 0.78 0.80 Framingham 0.80 0.86 Tecumseh 0.81 0.88 Honolulu 0.80 -- LRC Follow-Up 0.83 -- HDFP 0.73 0.79 Examples of Proposed Novel Risk Markers C-reactive protein Fibrinogen vwf Factor VII Homocysteine Lipoprotein a LDL sub-fractions ST segment depression Heart rate variability Carotid Doppler Ankle-brachial index EBCT for coronary calcium Platelet activity Ref: Heart 2002; 88: 222-228. Above models all include only age, SBP, cholesterol, smoking, and diabetes 8
LDL Goal and Cutpoints Patients with 0 1 Risk Factor 63 yo woman, no risks LDL Goal <160 mg/dl LDL Level at Which to Initiate Diet 160 mg/dl LDL Level at Which to Consider Drug Therapy 190 mg/dl (160 189 mg/dl: LDL-lowering drug optional) LDL 175 HDL 45 TG 160 SBP 120 Nonsmoker NCEP: no treat 10 yr risk: 3% no The best next step in lipid management is: Is this Threshold for Statins Too High? Benefits: reduction in CHD events and strokes 1. Continue current therapy 2. Begin a statin 3. Begin fenofibrate 4. Begin fish oil 5. Begin niacin (sustained release) Downsides: muscle pain, costs, need to take a pill daily, high NNT Recent data on efficacy, long-term safety, and sharply reduced costs have reduced treatment threshold 9
Modeling Cost Effectiveness Assumptions Pletcher and colleagues recently examined thresholds for treating adults with elevated LDL (>130 mg/dl) They found treating adults with 10 year risk over 5% warranted at statin cost of $20 per month or less Low-cost statin Non-drug costs of statin therapy modest Limited lipid, LFT, CK testing Limited lipid-related visits Adherence support not expensive No important adverse effects, apart from muscle pain No downside to daily medication usage Pletcher, Annals Int Med 2009 63 yo woman, no risk factors LDL 175 HDL 45 TG 160 If not a statin, should we add aspirin? Aspirin for Primary Prevention of CHD Events in Men Aspirin reduces relative risk of CHD events in men by 23% Aspirin causes gastrointestinal bleeding in 1-5 of 1000 users over 5 years Aspirin may also cause bleeding strokes in 1 of 1000 users over 5 years Determining benefit / harm ratio requires estimate of underlying CHD risk ATT Collaboration Lancet 2009 10
Effects of ASA in 1000 men over 5 years Implications 2.5% CHD risk 5% CHD risk 10% CHD risk MI prevented 6 12 28 When 10 yr. CHD risk < 5%, harms of aspirin are of similar magnitude to the benefits -- aspirin not routinely indicated GI bleeds caused 3-5 3-5 3-5 Bleeding stroke caused 1 1 1 When 10 yr. CHD risk >10%, benefits of aspirin likely exceed the harms -- aspirin should be considered for routine use For patients with intermediate risk, providers and patients should decide together, based on risk tolerance Aspirin in Women: USPTF 2009 ASA reduces stroke (24%), but not MI, CV mortality, all cause mortality Women 55-79 should use ASA when potential benefit of ischemic stroke outweighs harms of GI hemorrhage Not recommended under age 55; insufficient evidence over 80. CONCLUSIONS Patients with CHD or CHD equivalent: Treat aggressively with statin independent of LDL level (to LDL <70 in most cases) Treat other risk factors aggressively as well, especially easy ones (HTN, Aspirin use) Treat elevated non-hdl cholesterol and low HDL Patients at high risk are undertreated 11
CONCLUSIONS Patients without CHD: Assess overall risk with Framingham Risk Score Novel markers (hscrp, coronary calcium score) may be useful for further discrimination among those with intermediate risk High hscrp may identify additional patients who are candidates for statins Do a better job with what we have: HTN, smoking, weight, exercise, diet Patients without CHD: CONCLUSIONS Use medications at thresholds based on risk: LDL goal LDL drug treatment threshold High Risk (>20%) <100 (<70 optional) 100 (<100 optional) Mod high risk (10-20%) <130 130 (100-129 optional) Moderate risk (5-10%) <130 160 (130-159 optional) Lower risk (<5%) <160 190 (160-189 optional) Engage patient in shared decision making, especially if risk <10% CONCLUSIONS Patients without CHD: Use medications at thresholds based on risk: ASA STATIN High Risk (>20%) YES YES Mod high risk (10-20%) YES YES Moderate risk (5-10%) NO Maybe YES Lower risk (<5%) NO Usually NO 12