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

Linda Brookes, MSc


February 14, 2008

In This Article

Further Evidence From ILLUSTRATE, the Imaging Study

Prof. Barter reported that during analysis of the ILLUMINATE data, the finding of changes in serum electrolytes associated with torcetrapib were communicated to the investigators involved in the 3 torcetrapib imaging studies. These were the Investigation of Lipid Level management using coronary UltraSound To assess reduction of Atherosclerosis by CETP Inhibition and HDL Elevation (ILLUSTRATE)[7] and the Rating Atherosclerotic Disease change by Imaging with A New CETP inhibitor (RADIANCE) 1[8] and 2[9] trials, all supported by Pfizer. The main results of these imaging studies showed that torcetrapib was associated with a substantial increase in HDL-C and decrease in LDL-C. However there no significant decrease in the progression of coronary atherosclerosis as measured by intravascular ultrasonography (IVUS), and in all 3 trials torcetrapib was associated with an increase in blood pressure.

At the AHA, Stephen J. Nicholls, MB BS, PhD (The Cleveland Clinic Foundation, Cleveland, Ohio), one of the ILLUSTRATE investigators, reported that analysis of electrolytes and HDL-C changes in that trial appeared to confirm the findings of ILLUMINATE. More regression in atheroma volume was observed with greater elevations of HDL-C and elevations in serum electrolytes were identified in patients taking torcetrapib.[11]

The ILLUSTRATE trial involved 910 patients with angiographic coronary artery disease , mean age 57 years, predominantly male and white, overweight, with a high prevalence of atherosclerotic risk factors and high concomitant use of established medical therapies. Patients underwent IVUS imaging within a single coronary artery before and during 24 months of daily treatment with torcetrapib 60 mg (n = 464) or placebo (n = 446) on a background of atorvastatin therapy. The percent change in atheroma burden, the study's primary endpoint, was measured in matched arterial segments by IVUS and investigated relationship between changes in biochemical parameters and plaque burden with treatment.

Torcetrapib, added to atorvastatin therapy, was associated with a significant incremental lowering of LDL-C to 70.3 mg/dL, an increase in HDL-C to 72.3 mg/dL, associated changes in apolipoproteins, and a significant reduction in C-reactive protein , all of which translated into a 0.12% increase in percent atheroma volume (PAV) and a 9 mm3 in total atheroma volume. Torcetrapib was also associated with an increase in SBP of 4.6 mm Hg.

Analysis of the relationship between biochemical parameters and changes in percent atheroma volume, the primary endpoint of ILLUSTRATE study, revealed a clear inverse relationship between raising HDL-C and beneficial impact in slowing the rate of progression of coronary atherosclerosis in the torcetrapib group ( Table 5 ). There was no significant correlation between changes in any lipid parameter and change in percent atheroma volume for atorvastatin. In contrast there was a highly significant inverse correlation between changes in HDL-C and apolipoprotein (apo)A-1 and the change in percent atheroma volume. When patients were stratified according to whether they were regressors (with any decrease in PAV) or progressors (any increase in PAV), there was no difference between the 2 groups in patients treated with atorvastatin alone. In torcetrapib group, however, regressors had higher levels of HDL-C at follow-up and higher percentile increases in the level of HDL-C during the treatment period ( Table 5 ). Similar results were seen with apoA-1.

Analysis of change in atheroma burden by quartiles of achieved HDL-C again showed no relationship with increasing levels of HDL-C in the patients on atorvastatin alone, but a highly statistically significant slowing of progression with increasing HDL-C levels in patients given torcetrapib. The rate of progression was highest in the patients in the lowest quartile of HDL-C (<56 mg/dL) on treatment, whereas in patients who achieved the highest levels of HDL-C (>80 mg/dL), there was clear regression of atherosclerosis. Similar analysis by quartiles of percent change in HDL-C showed the same effect, ie, no significant relationship in the atorvastatin-alone patients, and a highly statistically significant trend for benefit with increasing quartiles of percent increases in HDL-C. Dr. Nicholls pointed out that these findings were consistent with the concept that the greatest levels of HDL-C in these patients were associated with regression, which would also be consistent with the generation of functional HDL particles removing cholesterol from the arterial wall.

Further analysis of the ILLUSTRATE data revealed electrolyte abnormalities associated with torcetrapib similar to those seen in the ILLUMINATE trial. Torcetrapib was associated with statistically significant reductions in serum potassium, and increases in serum sodium and bicarbonate. These changes, together with the rise in blood pressure, suggested activation of the RAAS system as a likely effect of treatment with torcetrapib. This was consistent with the preliminary ILLUMINATE findings and other data reported at the meeting that suggested increasing release of aldosterone from the adrenal gland increased in response to administration of torcetrapib.[12,13] At the time of this analysis, the ILLUSTRATE investigators did not have access to data on aldosterone levels, so using potassium as a surrogate for aldosterone activation, they examined changes in atheroma burden with torcetrapib by percent changes in potassium and HDL-C. In patients with increases in potassium levels above the median, raising levels of HDL-C was associated with a statistically significant trend for a beneficial effect on plaque progression ( Table 6 ). In contrast, a borderline significant effect was seen in the patients with decreased levels of potassium, consistent with increasing aldosterone activity.

Investigation in a similar fashion of percent change in HDL-C and change in PAV in patients stratified according to their tertiles of achieved levels of potassium on treatment with torcetrapib showed highly statistically significant relationships in the patients with potassium levels in the top 2 tertiles. However in the patients with the lower levels of potassium on treatment, the relationship between percent change in HDL-C and plaque progression rates was no longer significant. This was also consistent with the concept that in patients on torcetrapib, potential RAAS activation mitigates the benefit of HDL-C on plaque progression.

On multivariate analysis, percent change in HDL-C was found to be an independent predictor of change in PAV and total atheroma volume (TAV) (P = .001 and P = .004, respectively). Other independent predictors of PAV were atorvastatin dose, baseline PAV, and African American race. Additional predictors of the change in TAV included baseline TAV, diabetes, female gender, use of angiotensin-converting enzyme inhibitors, and previous MI or coronary artery bypass graft.Change in SBP was not an independent predictor of the lack of effect of torcetrapib on PAV or TAV.

Dr. Nicholls noted that both angiotensin II and aldosterone are known to have direct effects on the artery wall and may have an influence on plaque progression that is not related to the blood pressure effect.