Torcetrapib’s failure: Find more stories in the ILLUMINATE trial. (Dr. Eriksson MD, PhD)

April 2008

Libby P. J Intern Med. 2000;247:349-358.

PHASE I: Initiation PHASE II: Progression PHASE III: Complication

Atherosclerosis Is a Chronic Inflammatory Disease With LDL-C at the Core

p<0.05

p=0.05

p<0.05Statins

+

-

-

-

- - +

Only 2 Surrogate Endpoints For Cardiovascular Disease DrugsOnly 2 Surrogate Endpoints For Cardiovascular Disease Drugs

“The only surrogate endpoints currently used as a basis for approval of cardiovascular drugs are

blood pressure and serum cholesterol level”

“The only surrogate endpoints currently used as a basis for approval of cardiovascular drugs are

blood pressure and serum cholesterol level”

* Temple R: Are surrogate markers adequate to assesscardiovascular disease drugs? JAMA 1999;282:790-95.

LDL-C (apo B) Lowering As a Surrogate Endpoint for Reduced CVD risk LDL-C (apo B) Lowering As a Surrogate Endpoint for Reduced CVD risk

• Epidemiology/pathology/clinical observation

• Experiment of nature: heFH, PCsk9

• The LRC-CPPT and NHLBI trials demonstrated that resins reduced CV events and slowed progression, associated with lowering LDL-C

• Lovastatin was then approved on the basis of LDL-C reduction in 700 patients

• Ezetimibe was also approved on the basis of LDL-C reduction

• Epidemiology/pathology/clinical observation

• Experiment of nature: heFH, PCsk9

• The LRC-CPPT and NHLBI trials demonstrated that resins reduced CV events and slowed progression, associated with lowering LDL-C

• Lovastatin was then approved on the basis of LDL-C reduction in 700 patients

• Ezetimibe was also approved on the basis of LDL-C reduction

00

11

22

33

CH

D R

ISK

C

HD

RIS

K

100 100 160 160 220 220

LDLLDL-- cholesterolcholesterol(mg/(mg/dLdL))

* Men aged 50* Men aged 50--7070

85 85 65 65

45 45 25 25

HDLHDL--C C (mg/(mg/dLdL))

FRAMINGHAM: CHD RISKFRAMINGHAM: CHD RISK

0.64 0.86

Apolipoprotein B/Apolipoprotein A-I

Ris

k ra

tio

00

2

3

1

4

1.39

0

2

3

1

1.210.52 0.70 0.87

Men Women

4

1.04

p<0.0001

p<0.0001

p<0.0001

p<0.05p<0.001

AMORIS, fatal myocardial infarction

© G.Walldius, I. Jungner

Walldius G, Jungner I, Holme I, Aastveit A, Kolar W, Steiner E. Lancet 2001; 358: 2026-33

Torcetrapib – Final Results of the ILLUMINATE Trial

Philip J Barter1 for the ILLUMINATE Steering Committee and Investigators

1Heart Research Institute, Sydney, Australia

Philip J BarterPhilip J Barter11 for the ILLUMINATE Steering for the ILLUMINATE Steering Committee and InvestigatorsCommittee and Investigators

11Heart Research Institute, Sydney, AustraliaHeart Research Institute, Sydney, Australia

Steering Committee: P. Barter, Chair (Australia), M. Caulfield (UK),M. Eriksson (Sweden),S. Grundy (US), J. Kastelein (The

Netherlands),M. Komajda (France), J.-L. Lopez-Sendon (Spain), L. Mosca (US),J.-C. Tardif (Canada), D. Waters (US)

Steering CommitteeSteering Committee: P. Barter, Chair (Australia), M. Caulfield : P. Barter, Chair (Australia), M. Caulfield (UK),M. Eriksson (Sweden),S. Grundy (US), J. (UK),M. Eriksson (Sweden),S. Grundy (US), J. KasteleinKastelein (The (The

Netherlands),M. Netherlands),M. KomajdaKomajda (France), J.(France), J.--L. LopezL. Lopez--SendonSendon (Spain), L. (Spain), L. MoscaMosca (US),J.(US),J.--C. Tardif (Canada), D. Waters (US)C. Tardif (Canada), D. Waters (US)

Sponsor: Pfizer, IncSponsor: Pfizer, Inc

Statistical analysis:Statistical analysis:The Department of Biostatistics and Medical The Department of Biostatistics and Medical Informatics at the University of WisconsinInformatics at the University of Wisconsin--MadisonMadison

Published online today in the New England Journal of MedicinePublished online today in the New England Journal of Medicine

HDL protects against cardiovascular disease

A novel mechanism for raising HDL levels is to inhibit CETP. This protein transfers cholesterol from HDL to LDL fraction and thus retains cholesterol in the protective HDL.

Torcetrapib inhibits CETP and had been shown in humans to raise the level of HDL-C and lower that of LDL-C.

Inhibiting CETP with torcetrapib protects against atherosclerosis in rabbits.

The ILLUMINATE trial was designed to test the hypothesis that inhibiting CETP with torcetrapib would also protect against cardiovascular disease in humans.

HDL protects against cardiovascular disease

A novel mechanism for raising HDL levels is to inhibit CETP. This protein transfers cholesterol from HDL to LDL fraction and thus retains cholesterol in the protective HDL.

Torcetrapib inhibits CETP and had been shown in humans to raise the level of HDL-C and lower that of LDL-C.

Inhibiting CETP with torcetrapib protects against atherosclerosis in rabbits.

The ILLUMINATE trial was designed to test the hypothesis that inhibiting CETP with torcetrapib would also protect against cardiovascular disease in humans.

Rationale

Torcetrapib + titrated atorvastatin dose

Titrated atorvastatin dose

Planned 4.5 years of treatment Planned 4.5 years of treatment

Investigation of Lipid Level Management to Understand Investigation of Lipid Level Management to Understand its Impact in Atherosclerotic Eventsits Impact in Atherosclerotic Events

♦15,067 ♦7 countries

Subjects♦Major cardiovascular

events ♦Power=0.9 for 21%

reduction

♦Men or postmenopausal women

♦Statin eligible♦Any HDL-C level♦CHD or risk equivalent

(type 2 DM)

Primary End PointPatient Population

Atorvastatin run-in to LDL <100

mg/dL(2.6 mmol/L)4-10 weeks

ILLUMINATE: Long-term Outcomes in Patients With CHD or CHD Risk Equivalence

On Dec 2, 2006, after a median follow-up on-treatment of 550 days, the sponsor was informed by the Steering Committee, acting on advice from the independent Data Safety Monitoring Board, to terminate the trial based on a statistically significant excess of deaths and cardiovascular events in the group treated with torcetrapibThe sponsor immediately terminated the trialAll data were censored for the primary analyses on December 2, 2006

On Dec 2, 2006, after a median follow-up on-treatment of 550 days, the sponsor was informed by the Steering Committee, acting on advice from the independent Data Safety Monitoring Board, to terminate the trial based on a statistically significant excess of deaths and cardiovascular events in the group treated with torcetrapibThe sponsor immediately terminated the trialAll data were censored for the primary analyses on December 2, 2006

Premature Trial Termination:December 2nd, 2006

Primary Endpoint: Time to First MCVE*: Kaplan-Meier Plot

Primary Endpoint: Time to First MCVE*: Kaplan-Meier Plot

**Major cardiovascular event: CHD death, nonMajor cardiovascular event: CHD death, non--fatal MI, stroke or hospitalization for unstable fatal MI, stroke or hospitalization for unstable anginaangina

00 9090 180180 270270 360360 450450 540540 630630 720720 810810

Days from RandomizationDays from Randomization

Eve

nt F

ree

(%)

Eve

nt F

ree

(%)

9090

9292

9494

9696

9898

100100

Atorvastatin (A) events = 373

Torcetrapib/Atorvastatin (T/A) events = 464

P=0.001

Hazard Ratio 1.25

0.1430.130.2BMI, kg/m2, mean0.8261.361.3Age, yrs, mean0.8293.293.3White, n (%)0.9077.777.8Male (%)

P-value

Torcetrapib/ Atorvastatin

(n=7,533)

Atorvastatin(n=7,534)

Baseline Demographics

48.548.5 48.248.2 48.948.9 49.249.2 48.948.9

79.379.3 80.580.5 80.580.5 80.780.7

128128 127127 128128 130130 129129

79.979.9

00

2020

4040

6060

8080

100100

120120

140140

BaselineBaseline 11 33 66 1212

Study MonthStudy Month

Lipi

ds (m

g/Li

pids

(mg/

dLdL))

On Trial Lipid Levels By Study MonthAtorvastatinAtorvastatin Group (Post RunGroup (Post Run--In)In)

HDLHDL--CCLDLLDL--CCTGTG

48.648.6

71.871.8 77.577.5 80.980.9 82.982.979.779.7

127127115115 112112 112112 112112

58.358.358.258.259.359.359.759.7

00

2020

4040

6060

8080

100100

120120

140140

BaselineBaseline 11 33 66 1212

HDLHDL--CC

LDLLDL--CC

TGTG

On Trial Lipid Levels By Study MonthTorcetrapibTorcetrapib/ / AtorvastatinAtorvastatin Group (Post RunGroup (Post Run--In)In)

Study MonthStudy Month

Lipi

ds (m

g/Li

pids

(mg/

dLdL))

--9% (9% (--27,+13)27,+13)**

+72.1% (34.7) +72.1% (34.7) ††

--24.9% (28.5) 24.9% (28.5) ††

**median % change (IQR) for TG at month 12; p<0.001 median % change (IQR) for TG at month 12; p<0.001 vsvs atorvastatinatorvastatin†† mean % change (SD) for LDLmean % change (SD) for LDL--C, HDLC, HDL--C at month 12; p<0.001 C at month 12; p<0.001 vsvs atorvastatinatorvastatin

Secondary EndpointsSecondary Endpoints

0.0061.58 (1.14, 2.19)93 (1.2)59 (0.8)All-cause mortality

0.0011.35 (1.13, 1.62)270 (3.6)201 (2.7)Hospitalization for unstable angina

0.741.08 (0.70, 1.66)43 (0.6)40 (0.5)Stroke

0.131.21 (0.95, 1.54)142 (1.9)118 (1.6)Non-fatal MI (excluding procedure related)

0.411.21 (0.77, 1.92)40 (0.5)33 (0.4)CHD death

P- valueHazard Ratio (95% CI)

Torcetrapib/ Atorvastatin

n (%)

Atorvastatinn (%)

Estimated hazard ratio for pre-specified cardiovascular outcomes

Causes Of DeathCauses Of Death

Fatal stroke

44Reason unknown34Trauma/suicide/homicide

42Other non-cardiovascular90Infection2414Cancer

4020Any non-cardiovascular32Other vascular death/procedure related MI

21Fatal heart failure

41Other cardiac death60

86Fatal MI - not procedure related

2625Sudden cardiac death

4935Any cardiovascular death

Torcetrapib/ Atorvastatin (n=93)

Atorvastatin(n=59 )

Investigator-reported SAEs of neoplasmsInvestigator-reported SAEs of neoplasms

AtorvastatinAtorvastatin GroupGroup 136136

Torcetrapib/atorvastatinTorcetrapib/atorvastatin GroupGroup 128128

Investigator-reported SAEs of infections/infestations

AtorvastatinAtorvastatin GroupGroup 177177

Torcetrapib/atorvastatinTorcetrapib/atorvastatin GroupGroup 182182

On-Trial Blood Pressure By Study MonthOn-Trial Blood Pressure By Study Month

AtorvastatinAtorvastatin GroupGroup

122.9122.9 128.2*128.2*

73.773.7 75.7*75.7*

00 11 33 66 99 1212

Study MonthStudy Month

4040

6060

8080

100100

120120

140140

Blo

od P

ress

ure

(mm

Hg)

Blo

od P

ress

ure

(mm

Hg)

123.0123.0 123.9123.9

73.973.9 73.773.7

00 11 33 66 99 1212

TorcetrapibTorcetrapibAtorvastatinAtorvastatin GroupGroup

** p<0.001 p<0.001 vsvs atorvastatinatorvastatin at month 12at month 12

Patients receiving torcetrapib/atorvastatin had changes in serum electrolytes that differed significantly from those in patients taking atorvastatin alone.

♦ Serum potassium was reduced (p<0.001)♦ Serum bicarbonate was increased (p<0.001)♦ Serum sodium was increased (p<0.001)

Patients receiving torcetrapib/atorvastatin had changes in serum electrolytes that differed significantly from those in patients taking atorvastatin alone.

♦ Serum potassium was reduced (p<0.001)♦ Serum bicarbonate was increased (p<0.001)♦ Serum sodium was increased (p<0.001)

Changes in Serum Electrolytes

Serum Aldosterone

Percentage of patients with aldosterone > 8 ng/dl

21.6%17.8%Month 316.3%17.1%Baseline

A GroupA Group T/A GroupT/A Group P value*P value*

0.2< 0.001

Performed after study termination on the 87% of patients with stPerformed after study termination on the 87% of patients with stored ored baseline and month 3 samplesbaseline and month 3 samples55% of the analyzed samples had 55% of the analyzed samples had aldosteronealdosterone levels below the lower levels below the lower limit of quantification by the techniquelimit of quantification by the techniqueIt was, however, possible to determine unambiguously whether orIt was, however, possible to determine unambiguously whether or not not an analyzed sample had an an analyzed sample had an aldosteronealdosterone level of 8 level of 8 ng/dLng/dL or more. or more.

** WilcoxonWilcoxon comparisons after truncating the data below 8 comparisons after truncating the data below 8 ngng/dl/dl

Reported events occurring after Dec 2, 2006Reported events occurring after Dec 2, 2006

The following events were reported up to July 15th 2007:

DeathsAtorvastatin: 20 (n=7,475)Torcetrapib/atorvastatin: 14 (n=7,440)

MCVEs

Torcetrapib/atorvastatin: 38 (n=7,475)Atorvastatin: 38 (n=7,440)

Cox proportional hazard model adjusted for age, gender and baselCox proportional hazard model adjusted for age, gender and baseline HDLine HDL--C. Excludes 265 patients with C. Excludes 265 patients with missing month 3 HDLmissing month 3 HDL--C. Preliminary analysis initiated and authorised by P Barter anC. Preliminary analysis initiated and authorised by P Barter and conducted by Pfizerd conducted by Pfizer

Hazard ratios for CHD Death or NonHazard ratios for CHD Death or Non--Fatal MI Fatal MI by quintile of onby quintile of on--trialtrial HDLHDL--C C

(referent group is HDL(referent group is HDL--C < 60 mg/C < 60 mg/dLdL stratum)stratum)

1.001.00

0.670.67

0.470.470.570.57

0.430.43

00

0.20.2

0.40.4

0.60.6

0.80.8

1.01.0

CH

D D

eath

or N

onC

HD

Dea

th o

r Non

-- Fat

al M

IFa

tal M

I(H

azar

d R

atio

)(H

azar

d R

atio

)

<60<60 6060--7070 7171--8080 8181--9393 >93>93Quintiles of HDLQuintiles of HDL--C (mg/C (mg/dLdL) at Month 3) at Month 3

**P<0.05P<0.05**

****

Post-hoc Exploratory Analyses in the Torcetrapib/Atorvastatin Group

Effect of nicotinic acid, cholestyramine, clofibrate and combination on biliary lipids

and lithogenic index

Effect of nicotinic acid, cholestyramine, clofibrate and combination on biliary lipids

and lithogenic index

0.92 +/- 0.10 p<0.059.0 +/- 1.0 p<0.01Combination with cholstyramine

1.30 +/- 0.13 p<0.0112.1 +/- 0.9 p<0.001during

0.85 +/- 0.3 8.5 +/-0.3Clofibrate n=11before

0.71 +/- 0.05 p<0.0017.0 +/- 0.4during

0.87 +/- 0.058.7 +/- 0.5Cholestyramine n=19 before

1.00 +/- 0.09 p<0.0110.0 +/- 0.9 p<0.01during0.85 +/- 0.078.5+/-0.7Nicotinic acid n=13 before

Lithogenic indexBile cholesterol (molar %)

Angelin et al 1979

HDL and atherosclerosis

Does the way HDL is increasedinfluence clinical outcome?

ApoA-I infusion

HDL

Fecal BA

Fecal NS

Regression of atherosclerosis!

CETP inhibition

Torcetrapib

Brousseau et al. NEJM 2004

HDL 106%

Torcetrapib

HDL 2XC4LathoFecal BAFecal NS

Torcetrapib

Negative outcome in illuminate-patient end-point study

LDL-cholesterol cardiovascular disease(CVD) – secondary preventive studies

LDL-cholesterol

Pati

ient

s w

ith

CVD

eve

nts

(%)

25

20

15

10

5

0

1,3 1,8 2,3 2,9 3,4 3,9 4,4 4,9 5,5

LIPID-SCARE-S

4S-S

CARE-P

LIPID-P

4S-P

30

mmol/l

HPS-P

HPS-SLIPS-P

LIPS-S

AtoZ 80TNT 80

PROVE-IT AAtoZ 20

PROVE-IT P

TNT 10

Ballantyne CM. Am J Cardiol. 1998, O’Keefe JH et al, JACC 2004

Summary and ConclusionsSummary and ConclusionsUse of the CETP inhibitor, torcetrapib, in the ILLUMINATE trial was associated

with:♦ a predicted substantial increase in HDL-C and decrease in LDL-C ♦ an off-target pharmacology consistent with activation of the renin-

angiotensin-aldosterone system♦ an increased risk of all-cause mortality and CVD morbidity that triggered a

premature termination of the trial

The adverse clinical outcome associated with use of torcetrapib may have been the consequence of an off-target pharmacology but the possibility of an adverse effect of CETP inhibition cannot be excluded by the results of this randomized trial.

Possibilities: 1. Blood pressure2. No effect on the excretion of sterols from the human body3. Enrichment of cholesterol in HDL4. 40 % of the population had diabetes-is increasing HDL among

a good thing per se.

Use of the CETP inhibitor, torcetrapib, in the ILLUMINATE trial was associated with:♦ a predicted substantial increase in HDL-C and decrease in LDL-C ♦ an off-target pharmacology consistent with activation of the renin-

angiotensin-aldosterone system♦ an increased risk of all-cause mortality and CVD morbidity that triggered a

premature termination of the trial

The adverse clinical outcome associated with use of torcetrapib may have been the consequence of an off-target pharmacology but the possibility of an adverse effect of CETP inhibition cannot be excluded by the results of this randomized trial.

Possibilities: 1. Blood pressure2. No effect on the excretion of sterols from the human body3. Enrichment of cholesterol in HDL4. 40 % of the population had diabetes-is increasing HDL among

a good thing per se.