HDL plays no role in the pathogenesis of

atherosclerosis

Børge G Nordestgaard

Professor, Chief Physician, MD, DMSc

University of Copenhagen & Copenhagen University Hospital

Conflict of Interest Disclosure

Consultancies or talks sponsored by AstraZeneca, Merck, Omthera, Sanofi-Aventis,

Regeneron, ISIS Pharmaceuticals, Aegerion, Dezima, Fresenius, B Braun, Kaneka, Pfizer,

Amgen, Lilly, Kowa, Denka Seiken

Rebellious

Hans Christian Andersen’s fairy tale: The Emperor's New Clothes

HDL

Borge

HDL supportersHDL supporters

”HDL mafia”

”HDL mafia”

Two weavers promise

an Emperor a new suit of

clothes that is invisible to

those who are unfit for their

positions, stupid, or

incompetent.

The good cholesterol

Low HDL

protects against

heart disease

Medical textbooksLay press

Brain washed

Main focus on LDL cholesterol

• LDL receptor – Goldstein & Brown

• Oxidized LDL – Steinberg

• Statins – Endo, 4S trial and others

• European guidelines – LDL only

• American guidelines – LDL mainly

Next focus on HDL cholesterol• Strong epidemiology – all studies

• Animal studies

• The ”HDL mafia”

• Big Pharma

• But then……

HDL-C

CVD

2007

Drug targets based on genetic evidence

LDL-C Lp(a)

HDL-C ? Remnant-C / TG

Nordestgaard Eur Society Cardiology Congress 2011

Clinical focus on lipoproteins for CVD prevention

1980 1990 20102000

100%

0%

50%

Nordestgaard 2014

2020

LDL

HDLRemnants

TGs

4S trial

Epidemiology

Clinicians

Trials

”Postprandial”

Epidemiology

Genetics & ”failed” HDL trials

Nobel prize 85 & statins

Evidence for risk factors causing

atherosclerotic disease?

LDL Many other

risk factors

Epidemiology √ Many studies √

Genetics √ FH + SNPs No

Animal models √ Many models Sometimes

Mechanism √ Understood Speculations

Intervention √ Statin trials No

Nordestgaard 2015

Randomized trial vs. Mendelian randomization

Randomization methods

Placebo Drug: (lipo)protein

levels or

Cardiovascular disease or

Confounders

evenly distributed

Random distribution of alleles

Confounders

evenly distributed

Cardiovascular disease or

Normal

allele

Allele: (lipo)protein

levels or

Reverse causation

HDL

Remnants

LDL

HDL cholesterol

Remnant or

VLDL cholesterol

LDL cholesterol

Triglycerides

total cholesterol minus LDL-C minus HDL-C

no direct assay available yet

< 1 1 – 2 2 – 3 3 – 4 4 – 5 5

2309 6040 3023 1294 527 477

Women

Triglycerides, mmol/L < 1 1 – 1.99 2 – 2.99 3 – 3.99 4 – 4.99 5

Triglycerides, mg/dL < 89 89 – 176 177 – 265 266 – 353 354 – 442 443

Observations, No. 4564 7355 2298 652 212 141

Cholesterol, mmol/L (mg/dL)

Remnant 0.35 (14) 0.64 (25) 1.08 (42) 1.53 (59) 1.62 (63) 2.33 (90)

LDL 2.95 (114) 3.41 (132) 3.64 (141) 3.70 (143) 3.92 (151) 3.54 (137)

HDL 2.11 (81) 1.86 (72) 1.57 (61) 1.41 (55) 1.32 (51) 1.13 (44)

Lipoprotein cholesterol, mmol/ L

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

Lip

opro

tein

chole

stero

l, m

mol/

L

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

Lip

opro

tein

chole

stero

l, mg/d

L

0

50

100

150

200

250

*** ***

*** *** *** ***

*** ***

*** ***

*** ***

***

***

***

*** *** ***

***

***

*** ***

*** ***

***

*** *** *** ***

***

Remnant

LDL

HDL

Remnantp < 0.001

LDLp < 0.001

HDLp < 0.001

Trend

N =

Triglycerides(mmol/L)

Men

Copenhagen General Population Study

Freiberg, Nordestgaard 2011

***

***

Remnants

Evidence for lipoproteins causing

atherosclerotic disease?

LDL TG &

Remnants

Epidemiology √ √ Many studies

Genetics √ √ Remnant

hyperlipidemia

Animal models √ √ Many models

Mechanism √ √ Similar to LDL

Intervention √ (√) More needed

Nordestgaard 2015

Copenhagen General Population Study (CGPS)

Copenhagen City Heart Study (CCHS)

N=15,000

N=110,000+

37 yrs follow-up

10 yrs follow-up

No losses to

follow-up

1977-2014

Copenhagen

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Nonfasting triglycerides, mmol/L27% 46%

27% 0.1%

Frac

tio

n o

f p

op

ula

tio

n

Reference In extreme groupsNordestgaard & Varbo, Lancet 2014; 384: 626-635

Nonfasting triglycerides, mmol/L

Copenhagen City Heart Study and Copenhagen General Population Study

Haz

ard

rat

io (

95

%C

I)

Nordestgaard & Varbo, Lancet 2014; 384: 626-635

Myocardial infarctionN = 96,394 (Events = 3,287)

0

1

2

3

4

5

6

1 2 3 4 5 6 7

Nonfasting triglycerides, mmol/L

Copenhagen City Heart Study and Copenhagen General Population Study

Haz

ard

rat

io (

95

%C

I)

Ischemic Heart DiseaseN = 93,410 (Events = 7,183)

Haz

ard

rat

io (

95

%C

I)

Mainly fasting triglycerides, mmol/L

Emerging Risk Factors Collaboration

JAMA 2009

Coronary Heart DiseaseN = 302,430 (Events = 12,785)

0

1

2

3

4

5

0

1

2

3

4

5

1 2 3 4 5 6 7 1 2 3 4 5 6 7

Nordestgaard & Varbo, Lancet 2014; 384: 626-635

In decilesIn extreme groups

Nonfasting triglycerides, mmol/L

Copenhagen City Heart Study and Copenhagen General Population Study

Haz

ard

rat

io (

95

%C

I)

Ischemic StrokeN = 97,442 (Events = 2,994)

Haz

ard

rat

io (

95

%C

I)

Mainly fasting triglycerides, mmol/L

Ischemic StrokeN = 173,312 (Events = 2,534)

0

1

2

3

4

5

0

1

2

3

4

5

1 2 3 4 5 6 7 1 2 3 4 5 6 7

Nordestgaard & Varbo, Lancet 2014; 384: 626-635

Emerging Risk Factors Collaboration

JAMA 2009

In quintilesIn extreme groups

Nonfasting triglycerides, mmol/L

Copenhagen City Heart Study and Copenhagen General Population Study

Haz

ard

rat

io (

95

%C

I)

All-cause mortalityN = 98,515 (Events = 14,547)

0

1

2

3

4

5

1 2 3 4 5 6 7

Nordestgaard & Varbo, Lancet 2014; 384: 626-635

In extreme groups

Lipoprotein Cardiovascular

Disease Risk

Genotype

Mendelian randomization hypotheses

established but causal?

effect size?

pleiotropic effects?statistical power?

1

2 3

Causality: Instrumental Variable Analysis

Remnant cholesterol

mmol/L

HDL cholesterol

mmol/L

Hazard ratio for IHD

<0.4

0.4-0.6

0.6-0.7

0.7-1.1

>1.1

>2.0

1.7-2.0

1.4-1.7

1.2-1.4

<1.2

Quintiles

1.0 1.5 2.0 2.5N=57.000

CCHS+CGPS

Varbo et al JACC

2013; 61: 427-36

N=66.000 CCHS+CGPS+CIHDS

Remnant cholesterol

Plasma: observational

Genetic: causal

Remnant / HDL-C

Plasma

Genetic

HDL cholesterol

Plasma

Genetic

Hazard ratio for IHD per 1mM or

1.0 2.0 4.0

15 selected genetic variants

LDL cholesterol

Plasma

Genetic

12.000 IHD

Varbo et al JACC

2013; 61: 427-36

N=87.000 CARDIoGRAM

Triglycerides

Genetic unadjusted

Genetic LDL+HDL adjust

HDL cholesterol

Genetic unadjusted

Genetic LDL+TG adjust

Effect size (β) for CAD per 1SD or

0 0.3 0.6

Genome wide 185 SNPs

LDL cholesterol

Genetic unadjusted

Genetic TG+HDL adjust22.000 CAD

Do et al Nat Genet

2013; 45: 1345-52

APOC3Jørgensen et alNEJM 2014

01

01

APOC3TG and HDL Working GroupNEJM 2014

PCSK9CGPS & CCHS

0

1

Ischemic vascular disease

N alleles N total N events Risk estimate

75,465 10,770

260 27 0.59

110,472 33,889

498 113 0.60

64,492 10,665

1,697 250 0.90

- 44%

- 39%

- 12%

LDL cholesterol, mmol/L

Triglycerides, mmol/L Hazard ratio (95% CI)

Hazard ratio (95% CI)

0.0 0.5 1.0

0.0 0.5 1.0

0 1 2

0 1 2 3 4

Nordestgaard & Varbo, Lancet 2014; 384: 626-635

LDL

Remnants

LDL

Remnants

Plasma Intima

Triglycerides

Cholesterol Chylomicron

Inflammation

Macrophage

Foam cells

LPL LPL

FFA +Monoacylglycerol

Nordestgaard & Varbo, Lancet 2014; 384: 626-635

HDL cholesterol

Evidence for lipoproteins causing

atherosclerotic disease?

LDL TG &

Rem-

nants

Low HDL

Epidemiology √ √ √

Genetics √ √ None

Animal models √ √ (√)

Mechanism √ √ ?

Intervention √ (√) None

Nordestgaard 2015

Evidence for lipoproteins causing

atherosclerotic disease?

LDL TG &

Rem-

nants

Low HDL

Epidemiology √ √ √

Genetics √ √ None

Animal models √ √ (√)

Mechanism √ √ ?

Intervention √ (√) None

Nordestgaard 2013

Reverse cholesterol transport?

HDL+cholesterol

Human

evidence?

Evolution?

Causal factor

with variation

Glucose

Longterm

monitoring

HgbA1c

TG

RemnantsHDL

Nordestgaard et al. Current Drug Targets, 2009, 10, 328-335

Evidence for lipoproteins causing atherosclerotic disease?

LDL TG &

Rem-

nants

Low

HDL

HDL

function

ality

Epidemiology √ √ √ None

Genetics √ √ None None

Animal models √ √ (√) None

Mechanism √ √ (√) ?

Intervention √ (√) None None

Present Future The past?

HDL hypothesis

HDL functionality

Remnant cholesterol

HDL-C for risk

prediction

SCORE-HDL

Proportion classified as high CVD mortality risk

Martin Mortensen, Afzal, Nordestgaard, Falk. Eur Heart J 2015; in press

2003-2008: 46092 individuals without CVD, diabetes or statin use

6.8 years of follow-up

Copenhagen General Population Study

Martin Mortensen, Afzal, Nordestgaard, Falk. Eur Heart J 2015; in press

-20

-10

0

10

20

%R

ecla

ssif

icati

on

Inappro

pri

ate

A

ppro

pri

ate

-20

-10

0

10

20

%R

ecla

ssif

icati

on

Inappro

pri

ate

A

ppro

pri

ate

Fatal CVD Myocardial infarctionA C

Cases Non-cases Cases Non-cases

NRI, %

p value

NRI, %

p value

-16 +4 -10 +4

0.002 < 0.0001 < 0.0001 < 0.0001

Combined Combined

-12

0.02 0.006

-5

-20

-10

0

10

20

%R

ecla

ssif

icati

on

Inappro

pri

ate

A

ppro

pri

ate

B

Cases Non-cases

NRI, %

p value

-9 +4

<0.0001 < 0.0001

Combined

-5

<0.0001

Fatal CVD + nonfatal MI or stroke

Reclassification across 5% 10-year risk of fatal CVD

Using SCORE-HDL instead of SCORE

Copenhagen General Population Study

HDL

LDL

Rem-

nant

”GOOD”

BAD

UGLY

=innocent

Hans Christian Andersen’s fairytale: The Emperor's New Clothes

HDL

Borge

HDL supportersHDL supporters

”HDL mafia”

”HDL mafia”

Rebellious