The Failure of Torcetrapib -- The Search for the Reason Why

Linda Brookes, MSc


February 14, 2008

In This Article

ILLUMINATE: Study Results

Details of the main results of the ILLUMINATE trial as presented at the AHA by Prof. Barter have been widely reported elsewhere and were published simultaneously in The New England Journal of Medicine.[3] Briefly, the trial involved 15,067 patients (mean age 61 years, 78% male, 93% white) with coronary heart disease (CHD) or CHD risk equivalent (type 2 diabetes). Patients underwent a run-in period of 4-10 weeks during which they received lifestyle counseling with or without atorvastatin to achieve a low-density lipoprotein cholesterol (LDL-C) goal of <100 mg/dL. Patients who achieved that target at the end of the run-in period were randomized to atorvastatin at the dose established during run-in plus torcetrapib 60 mg or placebo. The trial was planned to run for 4.5 years to achieve a given number of events.

By termination, median follow-up in each treatment group was 550 days. At 12 months, patients who received torcetrapib showed an unprecedented panel of results, including a:

  • Mean increase of 72.1% in high-density lipoprotein cholesterol (HDL-C);

  • Mean decrease of 24.9% in LDL-C; and

  • Small mean decrease of 9% in triglycerides compared with baseline (all P <.001 vs atorvastatin only).

These lipid effects were consistent with those in previous studies.

However, the primary endpoint of the trial, a composite of first major cardiovascular event, defined as CHD death, nonfatal myocardial infarction (MI), stroke, or hospitalization for unstable angina, diverged between the 2 treatment groups from early on in the trial. At termination, the torcetrapib group showed a 25% increased risk over the groups that received atorvastatin alone (hazard ratio [HR], 1.25; 95% confidence interval [CI], 1.09-1.44; P = .001). A total of 464 events occurred in the torcetrapib group vs 373 in the atorvastatin-only group. Among the secondary endpoints, the torcetrapib group showed significantly increased risks of all-cause mortality (HR, 1.58; 95% CI, 1.14-2.19; P = .006) and hospitalization for unstable angina (HR, 1.35; 95% CI, 1.13-1.62; P = .001) and trends toward increased risk for the other components of the primary endpoint.

In the torcetrapib group vs the atorvastatin-only group, there was an increased risk of death from both cardiovascular causes (49 vs 35) and noncardiovascular causes (40 vs 20). The increase in cardiovascular risks was not attributable to a single cause of death. Among the noncardiovascular deaths, there were more cancer deaths in the torcetrapib group than in the atorvastatin-only group (24 vs 14), but this difference was not statistically significant and should not be a cause for concern, according to Prof. Barter. There were significantly more deaths from infection with torcetrapib vs atorvastatin alone (9 vs 0), and despite the small numbers this should be assumed to be a real phenomenon, Prof. Barter stated. However, from examination of the numbers of investigator-reported neoplasms and/or infections/infestations there was no evidence that torcetrapib was associated with development of new cancer or new infections.

At 12 months, systolic blood pressure (SBP) increased from baseline by a mean of 5.4 mm Hg in the torcetrapib group, a greater effect than had been observed in earlier short-term studies,[4,5,6] but consistent with that seen in 3 phase III imaging trials.[7,8,9] The SBP increase in the torcetrapib group was significantly greater than the increase (0.9 mm Hg) seen in the atorvastatin group (P <.001). Changes in diastolic blood pressure (DBP) were 2.0 mm Hg and -0.2 mm Hg, respectively. An unexpected finding in the torcetrapib group was that the relationship between changes in blood pressure and clinical outcome was counterintuitive, with an apparent increased risk of death in patients whose increase in SBP was less than the median. However, it appeared that an increase in blood pressure above the median identified a group with lower baseline blood pressure levels, making it difficult to interpret the relationship without further analysis, Prof. Barter explained.

Patients taking torcetrapib also showed small but significant changes in serum electrolytes compared with patients on atorvastatin alone, with a reduction in potassium and increases in sodium and serum bicarbonate ( Table 1 ).

These changes in serum electrolytes in the torcetrapib group suggested that the increase in blood pressure may have been due to mineralocorticoid excess. This proposition gained further support from a post hoc analysis of serum samples from patients with levels of aldosterone above the lower limit for quantitation. At Month 3, there was no significant change in the percentage of patients with serum aldosterone >8 ng/dL in the atorvastatin-only group, but there was a significant increase observed in the torcetrapib group ( Table 2 ). This was seen as clear evidence that serum aldosterone was increased with torcetrapib, although the mechanism by which this occurs remains unknown.

Prof. Barter and his colleagues concluded that these findings, together with other evidence presented at the AHA meeting (see below), show that activation of RAAS is a very clear off-target pharmacologic effect of torcetrapib. Prof. Barter noted, however, that the possibility of an adverse effect of CETP inhibition, with the generation of dysfunctional or even proatherogenic HDL-C, could not be excluded by the results of this randomized trial.

Post hoc exploratory analyses of ILLUMINATE data are underway and Prof. Barter presented the results of 2 analyses that have already provided clues as to the mechanism behind the effects of torcetrapib. One analysis, using changes in potassium and bicarbonate as indicators of the aldosterone effect, showed that in the torcetrapib group there were more deaths in patients with a serum potassium decrease greater than the median and/or with a serum bicarbonate increase greater than the median ( Table 3 ). This was consistent with these markers of an aldosterone effect being predictive of an excess in the rate of events. Prof. Barter said that although this finding requires further formal analysis corrected for other factors, it is consistent with the off-target pharmacology being 1 explanation for the harm caused by torcetrapib.

The ILLUMINATE investigators have also addressed the speculation that the HDL-C produced by inhibiting CETP may be dysfunctional or even proinflammatory or proatherogenic. If that were the case, Prof. Barter pointed out, patients with the greatest increases in HDL-C would be expected to do worst. However, HRs calculated for endpoints of CHD death or nonfatal MI analyzed by quintile of HDL-C and adjusted for age, gender, and baseline HDL-C showed that as HDL-C increased, the event rate decreased ( Table 4 ). Thus, the patients with the highest increase HDL-C (>93 mg/dL) had an event rate 57% lower than those who did not achieve this level. Although this finding should not be over-interpreted, Prof. Barter cautioned, it does not support the theory that CETP inhibition produces dysfunctional HDL-C, but it is consistent with other in vitro studies showing that the HDL-C produced by CETP is functional.

Prof. Barter emphasized that these post hoc observations were only suggestive and do not rule out HDL-C dysfunctionality, nor do they rule out the possibility that other unknown effects of CETP inhibition may have contributed to a mechanism-related adverse outcome.

In the opinion of AHA-designated discussant, Robert M. Califf, MD (Duke Translational Medicine Institute, Durham, North Carolina), the ILLUMINATE trial served as a "major warning to the biological and medical community" about the need to identify and understand the off-target toxicity early in drug development. The blood pressure and electrolyte effects associated with torcetrapib might have been identified earlier, signaling that more work needed to be done with respect to off-target evaluation, he suggested. He believes that HDL-C remains a good target and that CETP inhibition could be beneficial in raising HDL-C.

In an editorial accompanying publication of the ILLUMINATE results in The New England Journal of Medicine,[10] Daniel J. Rader, MD (University of Pennsylvania School of Medicine, Philadelphia) said that "the ILLUMINATE trial will undoubtedly stand as a watershed event in the field of HDL-targeted therapies. It may ultimately be seen as the study that brought about the rejection of the 'HDL hypothesis'." At a minimum, "it will have been responsible for shifting the focus from HDL concentration to HDL function and raising the bar for approval of new HDL-targeted therapies." The increased rate of cardiovascular events and death associated with torcetrapib could be due either to CETP inhibition itself, or to off-target effects of the drug, or to a combination of the 2, Dr. Rader believes. "It is tantalizing that in the ILLUMINATE study, a greater increase in HDL-C levels in torcetrapib-treated patients was associated with fewer major cardiovascular events; it will be important to confirm this observation in the atherosclerosis imaging trials of torcetrapib," he said.

Although the off-target effects of torcetrapib probably contributed in important ways to the increased rate of cardiovascular events and death in this trial, it is much more difficult to explain the increased rate of death from noncardiovascular causes by either CETP inhibition or off-target effects of torcetrapib, Dr. Rader noted. He suggested that the doubling in the rate of death from noncardiovascular events in patients receiving torcetrapib compared with those on atorvastatin alone might be due to changes in the composition of the HDL-C particle that might alter the immune or inflammatory function of HDL in such a way as to increase the risk of death from cancer or infection. Alternatively, off-target effects of torcetrapib might account for the excess rate of death from noncardiovascular causes through an unknown mechanism, he proposed.