Because of its critical role in HDL metabolism, CETP represents an attractive target for HDL-raising therapies. Two inhibitors of CETP (JTT-70533
have been shown to significantly increase plasma HDL-C concentrations in humans. The main goal of the present study was to assess the effect of CETP inhibition with torcetrapib on HDL apoA-I metabolism in patients with low levels of HDL-C.
Analysis of the HDL apoA-I kinetic data revealed that the increases in apoA-I PS observed in subjects treated with torcetrapib were primarily attributable to reductions in HDL apoA-I clearance rate. Torcetrapib reduced HDL apoA-I FCR by 7% (P
=0.10) in the atorvastatin cohort and by 8% (P
<0.001) and 21% (P
<0.01) in the 120 mg nonatorvastatin QD and BID cohorts, respectively. Torcetrapib did not significantly alter apoA-I PR in any of the cohorts. Thus, torcetrapib modulates plasma HDL apoA-I concentrations primarily via its effects on apoA-I catabolism. This finding is consistent with that of Ikewaki et al,13
who reported that the elevated concentrations of apoA-I observed in patients with CETP deficiency were due solely to delayed apoA-I catabolism.
The major role of CETP is to mediate the transfer of CE from HDL to apoB-containing lipoproteins and, in turn, of TG from apoB-containing lipoproteins to HDL.8,9
Therefore, it is not surprising that torcetrapib significantly influenced the lipid composition of HDL particles. Specifically, torcetrapib increased the CE and PL content of HDL while decreasing HDL TG content. These changes in HDL composition led to significant increases in mean HDL particle diameter in these subjects on torcetrapib.16
It has been demonstrated consistently that HDL particle size can influence the metabolism of apoA-I.34–37
In fact, one study has reported that as much as 70% of the variability in apoA-I FCR is attributable to variability in estimates of HDL size or density.37
Thus, the differences in apoA-I FCR observed in our study were likely attributable to torcetrapib-induced changes in HDL composition and, ultimately, HDL particle size.
We hypothesize that the relationship between increased particle size and decreased apoA-I FCR is not one of cause and effect, but rather, that the alterations in catabolism may be secondary to changes in apoA-I conformation. In vivo and in vitro work support this concept. Horowitz et al38
clearly demonstrated that HDL composition and size can affect the metabolic fate of apoA-I. Their data suggest that increased renal clearance of apoA-I occurs when HDL are cholesterol depleted and relatively TG enriched because of the dissociation of apoA-I from HDL. The in vitro experiments of Liang et al39
have further shown that the dissociation of apoA-I from HDL is a concentration-dependent phenomenon, such that increasing the level of HDL decreases the reduction in HDL particle size and the dissociation of apoA-I.
Torcetrapib also significantly influenced the distribution of apoA-I among HDL subpopulations. The most dramatic changes were observed in concentrations of the largest apoA-I–containing subpopulation of HDL: α
-1. Relative to placebo, 120 mg torcetrapib QD increased levels of apoA-I in the α
-1 subpopulation of HDL from 9.0 to 20.1 mg/dL in the atorvastatin cohort and from 9.5 to 23.1 mg/dL in the nonatorvastatin cohort, whereas the twice per day dose increased α
-1 from 10.0 to 39.8 mg/dL. In comparison, we reported previously a mean apoA-I concentration in α
-1 of 19.6 mg/dL for 79 healthy, normolipidemic subjects having a mean age of 53±13 years.40
Thus, in our subjects with low HDL-C, the once-daily dose of torcetrapib normalized the amount of apoA-I in α
-1. We believe that this is a key observation in light of the fact that we have shown previously that levels of α
-1 are significantly reduced in patients with CHD.40
Moreover, we reported that increased concentrations of α
-1 observed during simvastatin–niacin combination therapy were inversely associated with progression of coronary stenosis in patients from the HDL-Atherosclerosis Treatment Study.41
Under normal metabolic conditions, the α
-1 sub-population of HDL does not contain apoA-II.42
With this in mind, it is important to note that torcetrapib did not alter the distribution of apoA-II among α
-migrating HDL subspecies (data not shown). In contrast, we reported recently that apoA-II was localized to the α
-1 HDL subpopulation in patients with heterozygous and homozygous CETP deficiency.43
A crucial question is whether CETP inhibition influences the rate of reverse cholesterol transport (RCT). Schwartz et al44
reported recently that in normolipidemic humans, most HDL CE excreted into the bile gets there via apoB-containing lipoproteins and, therefore, presumably via CETP-mediated transfer. Hence, there is theoretical concern that CETP inhibition may impair RCT and, ultimately, be proatherogenic. Preliminary studies with CP-456,643, a close analog of torcetrapib, have shown that the removal of HDL CE from the plasma of rabbits is not reduced,45
suggesting that the RCT pathway is not compromised by CETP inhibition. Additional support for the latter concept comes from the work of Morehouse et al,46
which showed that the percentage of aortic surface covered with lesions was 60% lower in rabbits treated with torcetrapib relative to controls. Statistical analysis revealed that the reduction of aortic atherosclerosis in the torcetrapib-treated rabbits was significantly associated with the increase in HDL-C concentrations. Although the results of these studies are promising, whether or not the marked increases in HDL-C seen in humans treated with torcetrapib will translate into reduced cardiovascular disease risk remains to be determined.
In the present study, RCT was assessed indirectly, using fecal concentrations of neutral sterols and bile acids as surrogate measures. Torcetrapib did not substantially alter concentrations of either fecal sterols or bile acids, indicating that it did not influence fecal sterol excretion in these subjects. In the atorvastatin cohort, it should be noted that the dominant effect observed on these parameters was that associated with statin treatment because patients on atorvastatin had significantly lower serum concentrations of lathosterol (reduced cholesterol synthesis) and, in turn, decreased fecal sterol content, on placebo than did subjects in the nonatorvastatin cohorts. Although torcetrapib did not increase fecal sterol excretion, as did high doses of recombinant apoA-I in the study of Eriksson et al,47
increased flux of cholesterol from the periphery to the liver may not increase fecal sterol secretion or plasma lathosterol levels if it leads to a decrease in hepatic cholesterol synthesis and a proportional increase in cholesterol synthesis in the peripheral tissues. Moreover, important methodological differences prevent a direct comparison of these 2 studies. Included among these differences are: (1) modes of therapeutic delivery (oral versus intravenous infusion); (2) dosages of therapeutic agents (small daily doses of torcetrapib versus large doses of apoA-I, in effect equivalent to a dramatic increase in apoA-I synthesis); and (3) the duration of stool collection (3 days after 4 weeks of therapy versus 9 days before and 9 days after therapy).
To summarize, our data indicate that the CETP inhibitor torcetrapib increases plasma concentrations of HDL apoA-I by delaying apoA-I catabolism. The delayed apoA-I catabolism is likely attributable to CE enrichment of HDL which, in turn, leads to increased HDL particle size. Torcetrapib markedly increased concentrations of α
1-migrating HDL, which we have consistently shown to be inversely associated with atherosclerotic risk.40,41
When given either as monotherapy or in combination with atorvastatin, torcetrapib did not significantly alter fecal sterol excretion or bile acid synthesis. In conclusion, CETP inhibition has multiple effects on the dynamic aspects of HDL metabolism that may be relevant to its ultimate effects on atherosclerotic cardiovascular disease in humans.