OUP user menu

★ Editor's choice ★

Increasing high-density lipoprotein cholesterol by cholesteryl ester transfer protein-inhibition: a rocky road and lessons learned? The early demise of the dal-HEART programme

Ulf Landmesser, Arnold von Eckardstein, John Kastelein, John Deanfield, Thomas F. Lüscher
DOI: http://dx.doi.org/10.1093/eurheartj/ehs182 1712-1715 First published online: 13 June 2012

Raising the level of high-density lipoprotein (HDL) cholesterol has been proposed as a potential therapeutic strategy to reduce cardiovascular risk, largely based on the epidemiological studies showing that low HDL cholesterol levels are associated with an increased risk of coronary disease and cardiovascular events.1,2 This has been supported by experimental as well as translational studies, which have demonstrated the anti-atherogenic properties of HDL.3 This led to the perception of HDL cholesterol as the ‘good cholesterol’. Notably, in the Treating to New Targets study, low HDL cholesterol levels remained a marker of increased cardiovascular risk even in coronary patients with low LDL cholesterol levels (<70 mg/dL) on statin therapy.2

Although initial studies proposed reverse macrophage cholesterol transport by HDL as the main anti-atherogenic mechanism,4 more recent studies highlighted endothelial-protective properties of HDL, including the stimulation of endothelial nitric oxide (NO) production, as well as anti-inflammatory, anti-apoptotic, and anti-thrombotic effects.5,6 Importantly, however, these studies have used HDL isolated from healthy subjects or reconstituted HDL which differ in several ways from the HDL obtained from patients with coronary disease, as discussed below.

Inhibition of cholesteryl ester transfer protein (CETP) offers an unprecedented opportunity to profoundly increase HDL cholesterol plasma levels.7 This has resulted in several large-scale programmes to explore the potential of CETP as a novel therapeutic target for cardiovascular prevention. The large randomized ILLUMINATE trial examined the effects of the CETP inhibitor torcetrapib on clinical outcomes in 15 067 patients at high cardiovascular risk. In spite of a marked increase in plasma HDL cholesterol levels of 72.1%, the trial had to be stopped due to increased mortality and morbidity in the treatment group.8 In subsequent mechanistic studies, the adverse effects of torcetrapib were attributed to off-target toxicity, in particular increased production of aldosterone as well as an increased production of endothelin and reduced expression of endothelial nitric oxide synthase in the vessel wall.9 These off-target effects of torcetrapib were held responsible for the observed increase in the arterial blood pressure with this agent.8 Two post hoc analyses, the first in ILLUMINATE and the second in ILLUSTRATE, the coronary IVUS study with torcetrapib, indicated that those participants with the highest increase in HDL cholesterol or achieved on-treatment HDL cholesterol levels did show cardiovascular benefit. This raised hopes that other CETP inhibitors without the off-target toxicity of torcetrapib would lead to the prevention of recurrent events.8,10,11

The dal-HEART programme by Roche was a carefully designed clinical development programme of another CETP inhibitor, dalcetrapib, that, to the surprise of many, was stopped on Saturday, 5 May 2012. The interim analysis of the large phase-3 dal-OUTCOMES trial, involving >15 600 patients with recent acute coronary syndrome, had revealed a lack of efficacy of the compound in reducing cardiovascular events.12 This clinical trial programme was the first development programme of a CETP inhibitor after the torcetrapib failure. With this experience in mind, two safety studies, the dal-VESSEL and the dal-PLAQUE studies, designed to exclude adverse vascular effects of the compound, were part of the early development programme.13,14 Whereas both studies did not reveal adverse vascular effects of dalcetrapib therapy neither on endothelial function as assessed by flow-mediated dilation nor on vascular structure as examined by carotid MRI, there was also not a convincing signal that the compound would exert protective effects on these parameters.13,14

Indeed, inspite of a marked increase in HDL cholesterol plasma levels by 30% and a lack of effect on blood pressure, endothelial function was not improved (as had been the case with reconstituted HDL).13,15 Similarly, despite a suggestion of potential benefit, plaque size and inflammation (as assessed by positron emission tomography using fluoro-deoxyglucose as tracer) were not largely affected by the drug.14 Unfortunately, these findings had received little attention, although they now appear in strong agreement with the neutral outcome of the phase-3 clinical trial.

Another more potent CETP inhibitor, anacetrapib, is currently developed by Merck and is being tested in a large phase 3 clinical programme (see ClinicalTrials.gov Identifier: NCT01252953) after the long-term safety study DEFINE revealed encouraging results on blood pressure and lipids.16 Furthermore, Eli Lilly's evacetrapib has also entered a clinical trial programme.17 While dalcetrapib only increased HDL cholesterol relatively modestly, both anacetrapib and evacetrapib also markedly lower LDL cholesterol, small dense LDL, lipoprotein (a), non-HDL cholesterol, and apolipoprotein B, even on top of potent statin therapy (Table 1). Moreover, a very recent and large Mendelian randomization study has confirmed that lower CETP activity, when associated with both lower LDL and higher HDL, is associated with cardiovascular benefit.18 Similarly, in the Copenhagen City Heart Study CETP gene variants associated with lower CETP activity are associated with a reduced cardiovascular event rate (Anne Tybjaerg-Hansen, personal communication).

View this table:
Table 1

Properties and effects of different cholesteryl ester transfer protein inhibitors observed in clinical trials on plasma lipid levels and blood pressure

Lipids (per cent change, %)
 LDL-C−30–40−0–7 −10–40−20–30
Increase in blood pressure
 SystolicNo effectNo effectNo effect3–6 mmHg
Molecular weight
 Chemical structurea
  • aSource: Wikipedia.

The interpretation and prediction of the impact of CETP inhibitions is complicated by the growing awareness that the effects of HDL may vary in different clinical settings. For example, the effects of HDL on macrophage cholesterol efflux and in particular the endothelial effects of HDL are altered in patients with coronary disease or diabetes, a phenomenon referred to as ‘HDL dysfunction’.3 Indeed, Khera et al. found that the cholesterol efflux capacity of apoB-depleted serum (as a measure of the capacity of HDL to accept cholesterol from macrophages) was inversely related to carotid intima-media thickness and the likelihood of angiographic coronary artery disease independent of the HDL cholesterol plasma levels.19 We have observed that the capacity of HDL to stimulate endothelial NO production and endothelial repair is substantially reduced in patients with coronary artery disease or diabetes mellitus.20,21 Of note, HDL, isolated from healthy subjects, substantially stimulated endothelial cell NO production and accelerated endothelial repair in vivo, whereas no such or even opposite effects were observed when HDL was isolated from patients with coronary disease or diabetes.20,21 The underlying mechanisms need to be further defined, but likely include increased lipid oxidation of HDL due to a reduced HDL-associated paraoxonase-1 activity, an enzyme that protects HDL from lipid oxidation, as well as modifications of the protein moiety. High-density lipoprotein cholesterol is a highly complex lipoprotein that can scavenge >70 proteins, as identified by the proteomics analysis,22 and may contain >1000 different lipid species with each of them being modifiable. The vascular effects of HDL are not necessarily predictable by simply measuring its cholesterol concentration, since cholesterol is only a non-functional cargo of this lipoprotein and hence a surrogate marker reflecting the size and number of HDL particles. These observations raise the possibility that the vascular effects of on-treatment HDL may be an important determinant of the overall cardiovascular benefits of an HDL-raising intervention. Indeed, in certain clinical settings, an increase of dysfunctional HDL particles could also be detrimental. Hence, both the on-treatment vascular effects of HDL as well as the underlying molecular mechanism leading to increased HDL cholesterol levels are likely important determinants of the overall vascular effects of an HDL-cholesterol-raising therapeutic intervention. Thus, targeting HDL by CETP inhibition remains a ‘rocky road’. It appears that we need to learn more about the biology of HDL and its relation to atherosclerotic vascular disease. Notably easy-to-measure biomarkers reflecting HDL functionality better than HDL cholesterol or apoA-I levels are urgently needed. In this regard, post hoc analyses of Dal-OUTCOME and its biobank may turn out to be a highly valuable resource towards better understanding of the dalcetrapib failure. The extensive dal-HEART programme included early safety studies with surrogate measures as an endpoint to detect potential adverse effects of dalcetrapib early on. It is noteworthy that the dal-VESSEL study neither showed evidence of adverse vascular effects of dalcetrapib, nor a signal for a beneficial effect on vascular function.13 The dal-PALQUE study showed some rather weak effects on the carotid wall.

Atherosclerosis progresses for many years before clinical events and involves dysfunctional vascular biology characterized by inflammation and endothelial dysfunction. In hindsight, the lack of convincing positive signals from the two studies of function and structure might have been given more weight. Perhaps the lack of a clear positive signal in such studies should also give reason for concern about the efficacy of a particular target in cardiovascular prevention in the future as well. Further development of atherosclerotic plaque imaging may also be valuable in this respect.


This work has been supported by a grant from the Leducq Foundation.

Conflict of interest: T.F.L. has received speaker fees and research grants from Merck, Roche and Pfizer. U.L. has received speaker fees and research grants from Merck, Roche and Pfizer. A.v.E. has received speaker and consulting fees from Roche and MSD. J.D. has received speaker fees from Roche, MSD, Takeda, Pfizer, Novartis, and research grants from Colgate and Roche. J.K. is a consultant to Roche, Merck, Eli Lilly, Boehringer Ingelheim, and Novartis for the clinical development of CETP inhibitors.


  • The opinions expressed in this article are not necessarily those of the editors of the European Heart Journal or the European Society of Cardiology.