Featured Commentary

Issue No. 6; 2011

Cholesteryl ester transfer protein (CETP) inhibition has been a ‘hot topic’ in recent news. At the European Society of Cardiology (ESC) Congress in Paris 27-31 August, two trials with dalcetrapib were presented –dal-PLAQUE and dal-VESSEL. Results and implications from these trials are discussed here.

Dal-PLAQUE

The findings from this small trial, presented as a poster at the ESC Paris, were subsequently published in The Lancet.¹ Dal PLAQUE was designed to investigate the effects of dalcetrapib treatment on structural and inflammatory markers of plaque burden, using innovative multimodality imaging techniques.

In this phase IIb, double-blind trial 130 patients (mean age 63 years, 82% male and mean HDL cholesterol 1.2 mmol/L or 46 mg/dL at baseline) with coronary heart disease (CHD) or CHD risk equivalents and treated to a LDL cholesterol level of <2.6 mmol/L (mean 1.9 mmol/L or 74 mg/dL) were randomly allocated to treatment with dalcetrapib 600 mg/day or placebo for 24 months. The primary endpoints were indices of plaque burden from the carotid and abdominal aorta (total vessel area, wall area, wall thickness and normalised wall index) at 24 months measured using magnetic resonance imaging (MRI). Plaque inflammation was also assessed at 6 months using ¹⁸F fluoro-deoxyglucose uptake measured by positron emission tomography/computed tomography (FDG-PET/CT).

Treatment with dalcetrapib increased plasma levels of HDL cholesterol levels by 31%, consistent with previous phase II data.²

Dalcetrapib treatment was associated with reduction in structural markers of plaque burden at 24 months, as indicated by:

• Significant reduction in total vessel area (absolute change from baseline corrected for placebo, -4.01 mm², 90%CI -7.23 to -0.80, p=0.041)
• A trend for reduction in average wall area (-2.20 mm², 90%CI -4.54 to 0.13, p=0.12).
• The dalcetrapib placebo-corrected changes in total vessel area, average wall area and normalised wall index were either below the pre-specified ‘no harm’ boundary, or the change was numerically lower in the dalcetrapib group than the placebo group.

FDG-PET/CT results showed no evidence of increased vascular inflammation with dalcetrapib at 6 months. The target to background ratio (TBR) for the most diseased segment decreased over 6 months in the dalcetrapib group but did not change in the placebo group (average absolute change 0.19, 90% CI -0.29 to -0.09, p=0.001 versus -0.043, 90% CI -0.14 to 0.06, p=0.51). Exploratory analyses showed that the change in HDL cholesterol level appeared to be inversely correlated with the change in TBR for the most diseased segment at 6 months. A 4.3% reduction in arterial inflammation was observed with each increase in HDL cholesterol tertile (p=0.04).

Dalcetrapib treatment was not associated with any increase in blood pressure. There were 15 adjudicated cardiovascular events (2 on dalcetrapib vs. 13 on placebo).

It is acknowledged that the study was exploratory and therefore no correction was made for multiple statistical analyses. Additionally, patients were only randomised if they had a sufficient PET signal at baseline. Despite these methodological limitations, the data support the hypothesis that HDL cholesterol raising associated with dalcetrapib treatment might reduce inflammation in turn leading to favourable changes in structural vascular changes.

Dal-VESSEL

Dal-VESSEL³ was an exploratory phase IIB randomised, double-blind, placebo-controlled study which investigated the effects of dalcetrapib on endothelial function in patients with CHD or CHD risk equivalents and with HDL cholesterol levels <1.29 mmol/L (50 mg/dL). The primary efficacy endpoint was the change from baseline in brachial flow mediated dilation, a marker of endothelial dysfunction and atherosclerosis, after 12 weeks. This endpoint was assessed by experts who were masked to the study treatment allocation. The primary safety endpoint was 24-hour ambulatory blood pressure assessed after 4 weeks. The total duration of treatment was 36 weeks.

Of 860 patients assessed, 476 were randomised to treatment with dalcetrapib 600 mg/day or placebo in addition to their existing therapy (including statin therapy). Baseline characteristics in each treatment group were similar; mean age ≈62 years, 90% male, almost two-thirds had a history of CHD and almost one-half had type 2 diabetes mellitus. Baseline lipids (mmol/L) are summarised in Table 1.

Table 1. Baseline lipids for patients in dal-VESSEL

		Dalcetrapib	Placebo
LDL-C		2.11 ± 0.55	2.05 ± 0.46
HDL-C		1.01 ± 0.19	0.99 ± 0.19
Triglycerides		1.82 ± 0.92	1.65 ± 0.73

 
Treatment with dalcetrapib reduced CETP activity by almost 50% and increased plasma levels of HDL cholesterol by 31% after 36 weeks. This magnitude of HDL cholesterol raising was consistent with dal-PLAQUE and phase II study findings.^1,2 There was no significant change in plasma levels of LDL cholesterol or apolipoprotein (apo) B.

There was no significant change in nitric-oxide-dependent endothelial function, or in any biomarkers of inflammation and oxidative stress.

There was no significant increase in 24-hour ambulatory blood pressure at 4 weeks, or any other safety concerns during dalcetrapib treatment. Treatment was well tolerated. The adverse event profile of dalcetrapib was consistent with phase II trials.² There were 23 pre-specified positively adjudicated events (11 with dalcetrapib and 12 with placebo).

Placing these trials in context

Whether CETP inhibition is a viable strategy for preventing cardiovascular events has been the subject of much debate since the demise of torcetrapib. One key question relates to safety.

• Do other CETP inhibitors share the adverse effects of torcetrapib on blood pressure and aldosteronism? The results of both dal-PLAQUE and dal-VESSEL,added to those from recent phase II studies,^2,4 provide reassurance that this is not the case. These data, together with findings from experimental and genetic studies,^5-7 strengthen the argument that the adverse findings with torcetrapib in ILLUMINATE were due to ‘off-target’ pharmacological effects specific to this drug.

The other key question relates to the nature of the HDL produced by CETP inhibition.

• Are these HDL functionally defective? Recent in vitro studies do not support this.^8,9 Indeed, dal-PLAQUE provides further in vivo reassurance. While taking into account the usual limitations associated with interpretation of findings from a single, small imaging trial, it appears that treatment with dalcetrapib was not associated with an increase in plaque burden, as assessed by total vessel area. Furthermore, the decrease in FDG uptake in the carotid wall in the dalcetrapib-treated group implies that HDL associated with dalcetrapib treatment might have anti-inflammatory effects and therefore are unlikely to be functionally defective.

It would be of much interest to repeat dal-PLAQUE using anacetrapib, the other CETP inhibitor in advanced development. However, anacetrapib also lowers LDL cholesterol and Lp(a), as well as raising HDL cholesterol,⁴ which may complicate interpretation. In contrast, the lipid-modifying effects of dalcetrapib are defined by raising HDL cholesterol in the absence of effects on plasma levels of LDL cholesterol or Lp(a). Indeed, recent data show that dalcetrapib specifically increases markers of cholesterol absorption, indicative of nascent HDL lipidation by intestinal ABCA1, without affecting markers of synthesis.¹⁰

Taken together, the results of dal-PLAQUE and dal-VESSEL provide further reassurance regarding the safety profile of dalcetrapib, the more advanced of the newer CETP inhibitors. However, neither study provides any insights into possible cardiovascular benefits of dalcetrapib. Additionally, whether improvement in glycaemic control in statin-treated diabetic patients shown for torcetapib in a post hoc analysis of ILLUMINATE,¹¹ is also evident with other CETP inhibitors, is unknown.
 
For all of these outstanding questions, we need to wait for the results of the large cardiovascular outcome study, dal-OUTCOMES, in 2014.

Lipid measures in cardiometabolic risk

Data from the Swedish National Diabetes Register¹² provide further support for recommendations for the value of non-HDL cholesterol, a measure of the total number of atherogenic particles in plasma, in patients with diabetes or metabolic syndrome. In the study, the ratio of non-HDL :HDL cholesterol was the best predictor of coronary heart disease (CHD) risk in patients with type 2 diabetes.

This observational study included data from 18,673 patients with type 2 diabetes (mean age 60 years, 60% male, mean diabetes duration 7 years). Overall, 42% were receiving lipid-modifying therapy.

The adjusted hazard ratio (HR) for CHD per 1 standard deviation (SD) increment was consistently higher for non-HDL: HDL cholesterol than LDL cholesterol (Table 2). This is because this ratio takes account of apo B-containing lipoproteins, more so than LDL cholesterol, as well as HDL cholesterol, both of which are features of diabetic dyslipidemia.

Table 2. Adjusted HR per 1 SD increment for different lipid measures. Data from the Swedish National Diabetes Register.¹²

Lipid measure		Adjusted hazard ratio (95% CI)*
Non-HDL: HDL cholesterol		1.23 (1.17-1.30)
Non-HDL cholesterol		1.20 (1.14-1.27)
LDL cholesterol		1.17 (1.10-1.24)
TG:HDL		1.15 (1.08-1.21)

* Adjusted for age, sex, diabetes duration, antidiabetic therapy, HbA_1c, systolic blood pressure, smoking, albuminuria and history of cardiovascular disease

Compared with LDL cholesterol, both the non-HDL:HDL cholesterol and triglyceride (TG): HDL cholesterol ratios showed higher and almost linear increases in CHD event rates across their range (Fig 1.).

Fig. 1 5-year CHD event rates by non-HDL: HDL cholesterol and TG:HDL cholesterol ratios. Data from the Swedish National Diabetes Register.¹²



Additionally, tertile analysis showed that the TG:HDL ratio was almost two-fold lower in patients in tertile 1 for non-HDL:HDL cholesterol than in tertile 1 for LDL cholesterol (0.82 ±0.5 vs. 1.49 ±1.0).

These data show that measurement of non-HDL: HDL cholesterol provides a better estimation of coronary risk compared with LDL cholesterol in patients with type 2 diabetes. From a practical perspective, both measures can be assessed in nonfasting samples.¹³ The study underlines the relevance of atherogenic dyslipidemia, elevated triglyceride-rich lipoproteins (TRL) and low HDL cholesterol to coronary risk in type 2 diabetes patients, as highlighted by the recent EAS Consensus Panel paper.¹⁴

HDL: treatment target or risk factor?

Epidemiologic data clearly support HDL cholesterol as a strong risk factor for cardiovascular disease. Probably the most robust evidence is provided by the Emerging Risk Factors Collaboration, which showed that HDL cholesterol was strongly associated with coronary risk even after adjustment for non-HDL-C, logeTG and non-lipid risk factors.¹³ In high-risk individuals with acute coronary syndromes, those with low HDL cholesterol are almost 3 times more likely to die from a recurrent event during follow-up than individuals with higher values.¹⁵ Even intensive statin therapy fails to abrogate the cardiovascular risk associated with low HDL cholesterol.¹⁶

In cardiometabolic disease, low HDL cholesterol is closely associated with elevated TRL. This association is driven by CETP which mediates the hetero-exchange of TG from apoB-containing lipoproteins with cholesteryl ester from apo A-I lipoproteins, together with dissociation of apo A-I from TG-enriched HDL. The EAS Consensus Panel¹⁴ has previously highlighted the high cardiovascular risk associated with this dyslipidemic profile, which is further supported by the above study based on data from the Swedish National Diabetes Register.

In the recent ESC/EAS guidelines on dyslipidemia management,¹⁷ HDL cholesterol is recognised as a strong cardiovascular risk factor and recommended for use in risk estimation. However, consistent with the EAS Consensus Panel, the guidelines do not recommend HDL cholesterol as a treatment target, given the lack of supportive evidence from clinical intervention trials.^14,17 Furthermore, there is also controversy whether HDL cholesterol is the most appropriate marker in risk estimation.

Assessment of HDL functionality may be more relevant, given emerging experimental evidence of the pleiotropic potentially atheroprotective functions of HDL. Apart from its role in cholesterol efflux and lipid homeostasis, the HDL particle has been shown to exhibit a wide range of activities which include anti-thrombogenic, anti-inflammatory, anti-oxidative, anti-platelet and vasodilatory functions.¹⁸ A recent study (reviewed in EAS newsletter January 2011) showed that cholesterol efflux capacity had a strong inverse association with carotid intima-media thickness and the risk of angiographic coronary artery disease, irrespective of plasma lipid levels.¹⁹

Clearly, data from three key ongoing outcomes studies - HPS2-THRIVE with niacin/laropiprant, dal-OUTCOMES with dalcetrapib and REVEAL HPS-3 TIMI-55 with anacetrapib – are crucial for resolving this controversy. Lipidomic analysis may also provide useful insights, as suggested by recent data in patients with unstable coronary artery disease.²⁰ There are also emerging data for a number of innovative HDL raising therapies targeted to the acute management of high-risk patients (see Fig. 2). The findings from these studies will help in resolving whether HDL is indeed a target for therapy.

Fig 2. Possibilities for emerging HDL therapies