1. Is Low HDL-C a Modifiable Risk Factor or Only a Risk Marker for CHD
Is Low HDL-C a Modifiable Risk Factor or Only a Risk Marker for CHD? William E. Boden, MD, FACC, FAHA Chief of Medicine, Stratton VA Medical Center Vice-Chairman, Department of Medicine Albany Medical Center Professor of Medicine Albany Medical College, Albany, NY Cardiovascular Research Technology: 2013 Washington, DC February 25, 2013

2. Typical Patient Profile in 2013
48 y.o. overweight man with + FH and HTN comes to you with known CAD, s/p RCA stent placement for angina, and now seeks your advice. Asymptomatic at present.
Pre-Rx Lipids: TC= 185, LDL=125, HDL=25, TG=220, and non-HDL=160. You start atorvastatin 20 mg.
After 8 week Rx: TC=135, LDL=88, HDL-C=26, TG=160, non-HDL-C=109
Have you done enough?

3. Should High-Density Lipoprotein Be a Target of Therapy ?
ATP III and HDL-C
“A specific HDL cholesterol goal level to reach with HDL raising therapy is not identified”…
BUT, the epidemiology supporting the prognostic importance of low levels of HDL-C is robust

4. CHD Risk: HDL-C vs. LDL-C as Predictor*
RR of CHD After 4 Y
The Framingham Heart Study showed that the lower the level of high-density lipoprotein cholesterol (HDL-C), the greater the risk of CHD, regardless of LDL-C level. In fact, a person with a “desirable” LDL-C of 100 mg/dL but a low HDL-C of 25 mg/dL has the same risk for CHD as a person with an LDL-C of 220 mg/dL and an HDL-C of 45 mg/dL.
This information was originally published in The Canadian Journal of Cardiology: Castelli WP. Cholesterol and lipids in the risk of coronary artery disease—the Framingham Heart Study. Can J Cardiol. 1988;4(suppl A):5A-10A.
HDL-C, mg/dL
LDL-C, mg/dL
* Data represent men aged 50–70 y from the Framingham Heart Study.
Adapted with permission from Castelli WP. Can J Cardiol. 1988;4(suppl A):5A-10A.

5. Lipid Parameters and Risk of CAD Over 10 Years (Framingham)
Men
Women 20 20 16 16 12 12
Incidence (%)
Incidence (%) 8 8 4 4
The relationship between serum concentrations of HDL-C and LDL-C and the incidence of CAD is evident in both men and women, as documented by Wilson and colleagues in a 12-year study of 2489 men and 2856 women from the Framingham population.
For a given serum concentration of LDL-C or HDL-C, the absolute risk of CAD is generally lower in women than in men.
 35
 60
 130
 35
 60
 130
35-59
 160
35-59
 160
HDL-C
LDL-C
HDL-C
LDL-C
(mg/dL)
(mg/dL)
Wilson et al. Circulation. 1998;97:1837.
Wilson PWF, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation. 1998;97:

6. Impact of Low HDL-C on Clinical Events in NSTE ACS and DES Implantation
1,032 consecutive ACS patients who had DES
Low HDL-C (n = 550), < 40 mg/dl in men; < 45 mg/dl in women (mean = 32 ± 7 mg/dl)
“High” HDL-C (n = 482), > 40 mg/dl in men; > 45 mg/dl in women (mean = 55 ± 19 mg/dl)
Endpoints: Death; Q-wave MI, TLR and MACE at 30 days and 1 year
Statin use was high (98% in each group)
LDL-C (mean) was 102 mg/dl in low HDL-C and 108 mg/dl in high HDL-C groups (P = 0.26)
-Wolfram RM et al: Am J Cardiol 2006; 98:

7. Early and Late Mortality Post-DES: Low vs. High HDL at Baseline

8. 1-Year Event-Free Survival Post-DES: Low vs. High HDL at Baseline

9. Independent Predictors of Late Death and TLR MACE Post-DES for ACS

10. Anti-Atherogenic Functions of HDL

11. ARBITER 6 Trial Results: Overall Baseline Characteristics
Baseline measured variables
TC ± 26 mg/dL
LDL-C 82.1 ± 23.1 mg/dL
HDL-C ± 8.5 mg/dL
TG 134 ± 68 mg/dL
CIMT
Mean ± mm
Max ± mm
N = 208
80% male
Age: 65 ± 11 years
All on statins
42 ± 25 mg/day
6 ± 5 years duration
95% simvastatin or atorvastatin
Baseline characteristics balanced in the 2 treatment groups.
Baseline statin dose: Little room for additional statin titration.

12. Results: Lipid Concentrations
Niacin
Ezetimibe
Niacin: HDL increased by 18.4% to 50 mg per deciliter
 LDL and TG
Ezetimibe: LDL decreased by 19.2%, to 66 mg per deciliter
Δ LDL-C
Δ HDL-C
Δ TG (median)
Ezetimibe
−17.6±20.1 mg/dL
−2.8±5.7 mg/dL
-9 mg/dL
Niacin
−10.0±24.5 mg/dL
+7.5±9.2 mg/dL
-36 mg/dL

13. Results: Primary Endpoint Between-Group Change in CIMT
Niacin was superior to ezetimibe for the primary endpoint of the between group difference in carotid intima-media thickness.
P = 0.003
GLM for repeated measures

14. Results: Major Cardiovascular Events
Major adverse cardiovascular events occurred at a significantly lower incidence in the niacin (2/160 patients [1.2%] vs. the ezetimibe group (9/165 patients [5.5%])
Chi-square p=0.04; Log-rank p = 0.047

15. FIELD: Primary End Point
Nonfatal MI or CHD death at 5 years
Full study cohort:
Fenofibrate (n=4895)
Placebo (n=4900)
The primary end point of nonfatal MI or CHD death was not significantly lower in the fenofibrate group compared with the placebo group
P = .16
5.9% 6
5.2% 4
Nonfatal MI or CHD death at 5 years, % 2
FIELD (Fenofibrate Intervention and Event Lowering in Diabetes) was a double-blind, randomized, placebo-controlled trial designed to assess the effect of long-term treatment with fenofibrate (200 mg/d; n=4895) or placebo (n=4900) on the incidence of cardiovascular disease events in patients with diabetes and total blood cholesterol concentrations of less than 6.5 mmol/L. The primary end point was the first occurrence of either nonfatal myocardial infarction (MI) or coronary heart disease (CHD) death at 5-year follow-up.
As shown on the slide, no difference in the primary end point of CHD death or nonfatal MI was found between treatment groups (5.2% vs 5.9%, HR 0.89, 95% CI , P=.16). This corresponded to a significant 24% relative reduction in nonfatal MI, with a nonsignificant increase in fatal CHD.
Patients in the placebo group were treated more frequently with other lipid-lowering therapy during follow-up, which may have attenuated any treatment effect differences between the fenofibrate and placebo groups.
Fenofibrate (n=256)
Placebo (n=288)
Keech A, et al. Lancet. 2005;366(9500):
Keech A, Simes RJ, Barter P, et al; FIELD study investigators. Effects of long-term fenofibrate therapy on cardiovascular events in 9795 people with type 2 diabetes mellitus (the FIELD study): a randomised controlled trial. Lancet. 2005;366(9500):

16. ACCORD Lipid Trial (n=5,518): Primary Outcome
Rate
(%/yr)
HR (95% CI)
P Value
Primary Outcome:
Major Fatal or Nonfatal
Cardiovascular Event
291
2.24
310
2.41
0.92
( )
0.32
Fenofibrate
Placebo
(N=2765)
(N=2753)
N of
Events

17. HATS: Angiographic and Clinical Endpoints After 3 Years
89% reduction 25 20 15 10 5
Mean Change in Stenosis, %
Composite Event Rate, % † ‡
The HDL-Atherosclerosis Treatment Study (HATS) was a 3-year double-blind, placebo-controlled, NHLBI-sponsored study that, in part, evaluated niacin therapy of HDL-C management in patients also on an LDL-C–lowering statin regimen.
The study included 160 patients with coronary artery disease (CAD) who had low HDL-C and normal LDL-C levels. Patients were randomly assigned to 1 of 4 regimens – statin (S; simvastatin) and niacin (N), S and N plus antioxidant vitamins (AV), antioxidants, or placebo. For the purpose of this graph, we will be discussing the effects of N on HDL-C parameters.
Niacin was started at 250 mg twice daily and increased linearly to 1000 mg twice daily at 4 weeks; the mean dose was 2.4 g/d. Antioxidants (total daily doses: 800 IU vitamin E, 1000 mg vitamin C, 25 mg -carotene, and 10 µg selenium) were given twice daily.
The primary outcomes were angiographic evidence of a change in coronary stenosis and the occurrence of a first cardiovascular event (death, MI, stroke, or revascularization).
Patients receiving N had a significant increase in HDL-C of 26%. The addition of AV blunted the magnitude of this increase in HDL-C.
For the angiographic primary endpoint – the change in severity of the most severe stenosis in 9 proximal coronary segments – slight regression was observed with the addition of N (-0.4%) and progression was slowed with N and AV (+0.7%) treatment when compared to placebo (+3.9% mean change in stenosis).
The composite clinical endpoint of death from coronary causes, confirmed MI or stroke, or revascularization was reduced by 89% in patients treated with N compared with placebo (p=0.04). There were no statistically significant differences in clinical endpoints in the group receiving both N and AV compared with placebo.
The authors concluded that the addition of N in CAD patients with low HDL-C and “normal” LDL-C resulted in slight regression of coronary atherosclerosis and a significant reduction (89%) in clinical coronary events over 3 years compared with placebo.
Brown BG, Zhao X-Q, Chait A, et al. Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. N Engl J Med. 2001;345: *
9 Proximal Lesions
Coronary Death, MI, Stroke, or Revascularization
HATS = HDL-C-Atherosclerosis Treatment Study; S = simvastatin; N = niacin; AV = antioxidant vitamins.
*p<0.001 vs. placebo; †p<0.005 vs. placebo; ‡p=0.04 vs. placebo.
Adapted from Brown BG et al. N Engl J Med. 2001;345:

18. Effects of LDL-C and HDL-C Changes in 28 Lipid Trials on Primary Clinical Trial Outcomes
Reduction During Trial in 1ry CV Event Rate, vs Placebo
(%)
%  HDL-C plus %  LDL-C in Rx (%), placebo-adjusted
Compiled by Greg Brown, Am J. Cardiol, 2006

19. Evidence from Prior RCTs Supporting Niacin or Fibrate Clinical Benefit of ↑HDL-C or ↓TG
Coronary Drug Project (1975) 5-year follow-up
Immediate-release niacin (3,000 mg/day)
Reduced CHD Death/MI by 14%
Reduced non-fatal MI by 26%
Reduced stroke/TIA by 21%
VA-HIT (1999) 5-year follow-up
Gemfibrozil vs. placebo (no statin therapy)
Reduced CHD Death/MI by 22%
HATS (2001) 3-year follow-up (single-site)
niacin + simvastatin
regression of QCA angiographic coronary stenoses and 70% reduction in clinical events
Coronary Drug Project (1975) showed that immediate-release niacin (3,000 mg/day) significantly reduced CHD Death/MI by 14% and stroke/TIA by 26% vs. placebo at 5 years of follow-up
VA-HIT (1999) showed that gemfibrozil reduced CHD Death/MI by 22% vs. placebo in men with low HDL-C but with no LDL-C change (111 mg/dL at BL and 5.1 years follow-up) in the absence of statin therapy
HATS (2001) showed that niacin + simvastatin resulted in significant regression of angiographic coronary stenoses and reductions in clinical events at 2.5 years of follow-up
Seems too wordy.
Perhaps just 3 bullets with less detail. The detail can be spoken to in the presentation.

20. AIM-HIGH Trial
ATHEROTHROMBOSIS DRUG INTERVENTION for METABOLIC SYNDROME With Low HDL/HIGH TRIGLYCERIDES and Its IMPACT on GLOBAL HEALTH OUTCOMES

21. Entry Criteria Patients Age ≥ 45 Years with And Dyslipidemia
Coronary Heart Disease (CHD), or
Cerebrovascular Disease (CVD), or
Peripheral Arterial Disease (PAD)
And Dyslipidemia
Low Levels of Baseline HDL-C
<40 mg/dL for men; < 50 mg/dL for women;
Triglycerides mg/dL;
LDL-C < 180 mg/dL

22. Concomitant Medications at Entry
On a Statin
94%
Duration of Statin Therapy*
≥ 1 year
76%
≥ 5 years
40%
Prior Niacin Use
20%
ASA/Antiplatelet Therapy
98%
Βeta-Blocker
80%
ACEI / ARB
74%
Unable to ascertain years on statin therapy on 204 (6%)
Use of all secondary prevention therapies was
well-balanced between treatment groups
*Duration of statin therapy not ascertained in 6%

23. Baseline Lipids (mg/dL)
On Statin
Off Statin
LDL-C (mean)
(n=3,196) 71
(n=218)
125
HDL-C (mean) 35 33
Triglycerides (median)
161
215
Non-HDL (mean)
107
165
Apo-B (mean) 81
111

24. Primary & Secondary Endpoints
Hazard Ratio
95% CI
Primary Endpoint
1.02
0.87, 1.21
Secondary Endpoints
CHD Death, MI, Ischemic Stroke, High-Risk ACS
1.08
0.87, 1.34
CHD Death, MI, Ischemic Stroke
1.13
0.90, 1.42
Cardiovascular Death
1.17
0.76, 1.80

25. Primary Outcome HR 1.02, 95% CI 0.87, 1,21 Log-rank P value= 0.79 50
Combination Therapy 40
Monotherapy 30
HR 1.02, 95% CI 0.87, 1,21
Log-rank P value= 0.79
Cumulative % with Primary Outcome
16.4% 20
16.2% 10 1 2 3 4
N at risk
Time (years)
Monotherapy
1696
1581
1381
910
436
Combination Therapy
1718
1606
1366
903
428

26. Updated Niacin Studies Meta-Analysis – Coronary Events

27. Updated Niacin Studies Meta-Analysis: Stroke

28. HPS 2-THRIVE (2004 – 2013): Differences From AIM-HIGH
25,000 pts with history of CAD or diabetes
UK, Scandinavia, and (~50%) China
No lipid entry criteria other than LDL-C<160 mg/dL
No pre-defined HDL-C entry criteria
Randomized to simva 40 mg or simva + extended release niacin/laropiprant
Use of ezetimibe allowed
Trial terminated because of futility; 5-6 SAEs were significantly higher in N/L-treated pts; results to be presented at ACC on 3/9/13

29. Extant Questions
Why have recent clinical outcomes trials shown no benefit?
Baseline HDL-C levels not low enough?
Use of concomitant secondary prevention therapies may make it difficult to discern incremental benefit?
Unanticipated off target toxicities (CETP inhibitors; laropiprant)?
How can we reconcile the positive outcomes with surrogate measures and the absence of benefit in clinical outcomes trials (e.g. niacin)?
Are surrogate endpoints (e.g. QCA, CIMT and IVUS) informative measures of atherosclerosis stabilization/regression?
Different patient populations studied?
To what extent does the fact that all CETP inhibitor trials to date have taken ‘all comers’, without specific HDL entry criteria, contribute to their lack of observed benefit?

30. Extant Questions (cont’d)
Should HDL-P or specific measures of HDL function be used as inclusion criteria in future studies?
-If so, what is the most appropriate assay for HDL functionality?
To what extent does the occurrence of a recent acute coronary syndrome potentially alter the relationship between HDL levels and outcomes?

31. Clinical Recommendations
There is currently insufficient evidence from clinical trials to recommend HDL-targeted therapy
In patients with established CHD who are able to achieve and maintain optimal levels of LDL-C on statins, current data do not support incremental clinical benefit with additional lipid-altering agents
Patients who are unable to achieve their LDL-C goals on a statin should continue to be considered for combination therapy with fibrates, niacin, fish oil, etc.
The results of published clinical trials to date should not be extended to patient populations not represented by the study populations tested in the trials.

32. Concluding Thoughts
There is robust epidemiologic evidence from human population studies that low levels of HDL-C are associated with increased CV events
Robust evidence from animal studies support this
There is robust evidence that raising HDL has pleiotropic properties with the potential to protect against atherosclerosis experimentally and clinically.
But the hypothesis that HDL-targeted therapies in humans will reduce clinical cardiovascular events has still not been substantiated. In fact, 4 recent human trials with HDL-targeted agents have not shown a reduction in CV events. Absent such evidence, HDL-raising therapy cannot be recommended at present.