Efficacy and Safety of Torcetrapib, a Novel Cholesteryl Ester Transfer

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Journal of the American College of Cardiology Vol. 48, No. 9, 2006 © 2006 by the American College of Cardiology Foundation ISSN 0735-1097/06/$32.00 Published by Elsevier Inc. doi:10.1016/j.jacc.2006.06.067 Pharmacotherapy of HDL and LDL Efﬁcacy and Safety of Torcetrapib, a Novel Cholesteryl Ester Transfer Protein Inhibitor, in Individuals With Below-Average High-Density Lipoprotein Cholesterol Levels Michael H. Davidson, MD, FACC,* James M. McKenney, PHARMD,† Charles L. Shear, DRPH,‡ James H. Revkin, MD, FACC‡ Chicago, Illinois; Richmond, Virginia; and New London, Connecticut

OBJECTIVES This study was designed to evaluate the efﬁcacy and safety of torcetrapib, a cholesteryl ester transfer protein (CETP) inhibitor, in subjects with low high-density lipoprotein cholesterol (HDL-C) levels. BACKGROUND Evidence suggests HDL-C is atheroprotective. A proven mechanism for increasing the level of HDL-C is the inhibition of CETP. METHODS A total of 162 subjects with below-average HDL-C (men Ͻ44 mg/dl; women Ͻ54 mg/dl) who were not taking lipid-modifying therapy were randomized to double-blind treatment with torcetrapib 10, 30, 60, or 90 mg/day or placebo (ϳ30 subjects per group). RESULTS The percent change from baseline to Week 8 with torcetrapib (least-squares mean difference from placebo) was dose-dependent and ranged from 9.0% to 54.5% for HDL-C (p Յ 0.0001 for 30 mg and higher doses) and from 3.0% to Ϫ16.5% for low-density lipoprotein cholesterol (LDL-C) (p Ͻ 0.01 for 90-mg dose). Low-density lipoprotein cholesterol lowering was less in subjects with higher (Ͼ150 mg/dl) versus lower levels of baseline triglycerides; at 60 mg, the change in LDL-C was 0.1% versus Ϫ22.2% (p Ͻ 0.0001), respectively. Particle size for both HDL and LDL increased with torcetrapib. There were no dose-related increases in the frequency of adverse events. Signiﬁcant blood pressure increases were noted in 2 of 140 subjects. CONCLUSIONS Torcetrapib resulted in substantial dose-dependent elevations in HDL-C, accompanied by moderate decreases in LDL-C at the higher doses. Torcetrapib was generally well tolerated. (J Am Coll Cardiol 2006;48:1774–81) © 2006 by the American College of Cardiology Foundation

High-density lipoprotein cholesterol (HDL-C) is an inde- The cholesteryl ester transfer protein (CETP) plays a pivotal pendent inverse risk factor for coronary heart disease role in cholesterol metabolism, exchanging cholesteryl esters (CHD), and elevating HDL-C is a promising strategy for (CEs) and triglycerides between lipoproteins (4). Typically, CETP transfers CEs from HDL to very low-density lipoprotein See page 1791 (VLDL) and low-density lipoprotein (LDL) in exchange for triglycerides and transfers triglycerides from VLDL to LDL and preventing cardiovascular events. However, the range of HDL in exchange for CEs (4). Thus, CETP activity is potentially drugs for elevating levels of HDL-C is limited. Statins (1) pro-atherogenic, decreasing the CE content of atheroprotective and fibrates (2) provide only modest increases in HDL-C, HDL and increasing the CE content of atherogenic VLDL and and niacin (3), although more effective, is poorly tolerated. LDL (4). Furthermore, by exchanging triglycerides in VLDL for CEs in HDL and LDL, CETP promotes the formation of small dense LDL and HDL by increasing the remodeling of From the *Rush-Presbyterian-St. Luke’s Medical Center, Chicago, Illinois; †Vir- triglyceride-enriched LDL and HDL particles by triglyceride ginia Commonwealth University, Richmond, Virginia; and ‡Pfizer Global Research and Development, New London, Connecticut. This study was sponsored by Pfizer lipases (4). Small dense LDL is particularly atherogenic (5). In Inc. Conflicts of interest: Dr. Michael H. Davidson—Research Grants: Abbott individuals with elevated triglyceride levels, such as those with Laboratories, AstraZeneca, KOS, Merck, Merck/Schering Plough, Pfizer, Reliant Pharmaceuticals, Roche, Sankyo Pharma, and Takeda Pharmaceuticals. Speakers’ diabetes, the VLDL pool is typically enlarged, and CETP-driven Bureau: Abbott Laboratories, AstraZeneca, KOS, Merck, Merck/Schering-Plough, triglyceride enrichment of HDL and LDL particles may be Pfizer, Reliant Pharmaceuticals, Sankyo Pharma, and Takeda Pharmaceuticals. especially relevant in creating an atherogenic lipid profile (6,7). Consultant/Advisory Board: Abbott Laboratories, AstraZeneca, KOS, Merck, Merck/Schering-Plough, Pfizer, Reliant Pharmaceuticals, Roche, Sankyo Pharma, Cholesteryl ester transfer protein inhibition is a potential Sumimoto Pharmaceuticals, and Takeda Pharmaceuticals. Dr. James M. strategy for elevating HDL-C and for treating cardiovascular McKenney—Speaking Honorarium: AstraZeneca, KOS, Merck/Schering Plough, disease (CVD) (4). In rabbit models, several techniques have been Pfizer, Reliant Pharmaceuticals, and Takeda Pharmaceuticals. Consultant/Advisory Board: AstraZeneca, KOS, Microbia, Pfizer, and Sankyo Pharma. Research Grants employed to suppress CETP activity, resulting in increases in (awarded to company, not individual): AstraZeneca, GSK, KOS, Merck, Pfizer, HDL-C and atherosclerotic regression (8–11). Two CETP Roche, Schering Plough, and Takeda Pharmaceuticals. Dr. James H. Revkin— Ownership Interest: Pfizer Stock/Options. Institution/Employer: Pfizer. Adjunct inhibitors, JTT-705 and torcetrapib, are in clinical development. Faculty: Yale University. Dr. Charles L. Shear—Institution/Employer: Pfizer. In preliminary trials, torcetrapib has been shown to produce JACC Vol. 48, No. 9, 2006 Davidson et al. 1775 November 7, 2006:1774–81 Torcetrapib in Subjects With Low HDL

Lipid assessments. The primary end point was the percent Abbreviations and Acronyms change from baseline in the levels of HDL-C after 8 weeks. AE ϭ adverse event Absolute changes from baseline in HDL-C and percent changes apo ϭ apolipoprotein and absolute changes in LDL-C, triglycerides, and total choles- ϭ CE cholesteryl ester terol were secondary end points. Additional lipid analyses included CETP ϭ cholesteryl ester transfer protein CHD ϭ coronary heart disease apolipoprotein concentrations; HDL particle type; HDL, VLDL, CVD ϭ cardiovascular disease and LDL subclass composition; phospholipid concentrations; and DBP ϭ diastolic blood pressure nuclear magnetic resonance (NMR) lipoproﬁle. HDL-C ϭ high-density lipoprotein cholesterol Analytical methods. Biochemical analyses were performed by ϭ LDL-C low-density lipoprotein cholesterol Medical Research Laboratories (Highland Heights, Kentucky). NMR ϭ nuclear magnetic resonance SBP ϭ systolic blood pressure Total cholesterol and net triglycerides were quantiﬁed by a VLDL-C ϭ very low-density lipoprotein cholesterol CDC-standardized enzymatic assay in an automated chemistry analyzer. High-density lipoprotein cholesterol was measured by

separating HDL from LDL/VLDL by heparin/MnCl2 chemical substantial elevations in HDL-C, modest decreases in LDL precipitation. Low-density lipoprotein cholesterol and VLDL cholesterol (VLDL-C) were estimated by the Friedewald formula cholesterol (LDL-C), and increases in lipid particle size (12,13). Ͼ This phase 2 study provides further data on the efficacy and (15). If total triglycerides were 400 mg/dl, LDL-C and safety of torcetrapib in a large group of individuals with below- VLDL-C were measured directly by beta-quantification using average levels of HDL-C. A study of torcetrapib administered on ultracentrifugation. Phospholipid was measured by an automated a background of atorvastatin to similar subjects is reported in this enzymatic colorimetric method. High-density lipoprotein sub- issue of the Journal (see page 1782). classes (HDL2 and HDL3) were separated by zonal ultracentrifugation. Apo A-I, A-II, and B-100 were analyzed by an auto- METHODS mated immunoturbidimetric procedure. Lipoprotein subclasses where determined using proton NMR by Liposciences Inc. Study design. This was a multicenter study (23 centers). Fol- (Raleigh, North Carolina) (16). lowing screening, eligible participants were randomized to 8 weeks Safety assessments. Safety assessments included a physical of double-blind treatment with either placebo or torcetrapib 10, examination and measurement of vital signs, electrocardiograms, 30, 60, or 90 mg once daily (Fig. 1). and standard laboratory safety tests. All adverse events (AEs) were Participants. Eligible participants were ages 18 to 65 years with recorded. low HDL-C levels (Ͻ44 mg/dl for men and Ͻ54 mg/dl for Statistical analyses. The primary statistical analysis for women) at screening (14). Exclusion criteria included major/ efficacy included all randomized participants who received at unstable concurrent illnesses, lipid-altering therapy within 30 days least 1 dose of study treatment with at least 1 before- and of screening, and an LDL-C level of Ն190 mg/dl or triglycerides after-treatment end point measurement using the last- Ն400 mg/dl during screening. observation-carried-forward approach. The analysis of the The protocol was approved by the Institutional Review Board primary end point (HDL-C percent change from baseline at or Independent Ethics Committee at each site and was conducted week 8) employed analysis of covariance using a linear in compliance with the Declaration of Helsinki. model that included a term for treatment group and baseline

Figure 1. Schematic representation of study design and numbers of subjects. Sample size was calculated based on earlier torcetrapib studies, with 25 subjects per group yielding 80% power to detect a 25% treatment difference in high-density lipoprotein cholesterol (HDL-C), assuming a common standard deviation of 30.5% and 2-sided t test with 5% type I error. LOCF ϭ last observation carried forward (i.e., subjects with baseline and at least one post-baseline HDL-C measurement); R ϭ randomization. 1776 Davidson et al. JACC Vol. 48, No. 9, 2006 Torcetrapib in Subjects With Low HDL November 7, 2006:1774–81

Table 1. Baseline Patient Demographics and Lipid Parameters Torcetrapib (mg/day) Parameter Placebo 10 30 60 90 Demographics n 3232313433 Men, n (%) 25 (78) 22 (69) 26 (84) 27 (79) 27 (82) Mean age, years (ϮSD) Men 47 (Ϯ10) 48 (Ϯ12) 45 (Ϯ10) 46 (Ϯ10) 49 (Ϯ11) Women 45 (Ϯ11) 48 (Ϯ9) 41 (Ϯ8) 48 (Ϯ10) 49 (Ϯ10) Race or ethnicity, n (%) White 27 (84) 25 (78) 26 (84) 28 (82) 30 (91) Black 0 (0) 0 (0) 1 (3) 3 (9) 1 (3) Hispanic 4 (13) 7 (22) 4 (13) 3 (9) 2 (6) Other 1 (3) 0 (0) 0 (0) 0 (0) 0 (0) Mean BMI, kg/m2 Men 29.4 29.4 29.3 29.6 31.1 Women 32.8 33.6 35.1 34.6 33.3 Mean SBP/DBP, mm Hg 123.0/78.4 118.7/76.6 119.0/78.0 120.0/77.6 120.4/76.5 Lipid parameters n 3232303433 Mean HDL-C, mg/dl (ϮSD)* 39 (Ϯ7) 40 (Ϯ5) 37 (Ϯ5) 37 (Ϯ6) 37 (Ϯ6) % Ͻ40 mg/dl 59 50 77 62 67 Mean LDL-C, mg/dl (ϮSD)* 125 (Ϯ31) 128 (Ϯ28) 117 (Ϯ27) 120 (Ϯ22) 127 (Ϯ29) % Ͻ130 mg/dl 59 53 63 68 52 Mean TGs, mg/dl (ϮSD)* 186 (Ϯ84) 176 (Ϯ63) 205 (Ϯ88) 192 (Ϯ88) 197 (Ϯ78) % Ͻ150 mg/dl 38 41 33 35 24 Mean TC, mg/dl (ϮSD)* 201 (Ϯ36) 202 (Ϯ35) 194 (Ϯ32) 194 (Ϯ29) 204 (Ϯ30) % Ͻ200 mg/dl 44 41 57 59 45 Ratio of LDL-C to HDL-C (ϮSD)* 3.2 (Ϯ0.8) 3.3 (Ϯ0.9) 3.2 (Ϯ0.8) 3.3 (Ϯ0.7) 3.5 (Ϯ0.9)

*Baseline values are the average of the last 2 measurements made within the 2 weeks before first treatment dose. BMI ϭ body mass index; DBP ϭ diastolic blood pressure; HDL-C ϭ high-density lipoprotein cholesterol; LDL-C ϭ low-density lipoprotein cholesterol; SBP ϭ systolic blood pressure; TC ϭ total cholesterol; TGs ϭ triglycerides. value as a continuous covariate (SAS Proc Mixed). Study well balanced across treatment groups (Table 1). Mean HDL-C center was not included as an independent variable. Least- levels for treatment groups ranged from 37 to 40 mg/dl. squares means (LS means) were computed and pairwise Efficacy: lipid parameters. HDL AND HDL-RELATED APO- treatment comparisons of torcetrapib dose group versus LIPOPROTEINS. Torcetrapib dose-dependently increased placebo were assessed for statistical significance at the p ϭ HDL-C levels (Table 2, Fig. 2). Percent changes in HDL-C 0.05 level (2-sided) using a step-down procedure to preserve from baseline to Week 8 ranged from 9.0% to 54.5% with the type 1 error across the multiple comparisons (17). A torcetrapib 10 to 90 mg/day (LS mean difference from placebo). 95% confidence interval (CI), unadjusted for multiplicity, Differences were significant at doses of 30, 60, and 90 mg/day (p was calculated for each pairwise comparison. Similar anal- Յ 0.0001). Increases in HDL-C were accompanied by dose- yses were performed for secondary end points. dependent increases in the levels of apo A-I and apo A-II, with For lipid assessments, results are presented in figures as apo A-I being the dominant change (Table 3). raw means for each time point. The percent changes in Ultracentrifugation/precipitation analysis showed that torce- lipids at 8 weeks used for hypothesis testing are presented in trapib increased HDL particles of larger size (Table 3). High- tabular form. density lipoprotein subclass changes determined by NMR spec- For vital signs, each patient’s complement of after- troscopy were consistent with the findings from baseline observations was averaged and a change from ultracentrifugation/precipitation analysis showing that torcetrapib baseline was calculated. This measure was then analyzed in produced dose-dependent increases in levels of the large HDL-C a manner analogous to the efficacy parameters discussed subclasses. At the 60-mg and 90-mg doses of torcetrapib, large earlier; that is, ANCOVA was employed using SAS Proc HDL-C (8.3 to 13 nm) increased from 12.5 (SD Ϯ 5.6) to 26.9 Mixed with a linear model, including a term for treatment mg/dl (SD Ϯ 11.3) (p Ͻ 0.05) and 13.2 (SD Ϯ 6.3) to 35.0 mg/dl group and baseline value as a continuous covariate. Least- (SD Ϯ 21.3) (p Ͻ 0.01), respectively. At these same doses of squares means were calculated and 95% CIs were computed torcetrapib, mean HDL particle size also increased from 8.3 for the within-treatment-group change from baseline. (Ϯ0.3) to 8.8 nm (Ϯ 0.4) and 8.3 (Ϯ0.3) to 9.0 nm (Ϯ0.5), Յ RESULTS respectively (p 0.0001 for both).

Baseline demographics. Baseline demographic characteristics APO B-RELATED LIPOPROTEINS. At Week 8, moderate de- and lipid proﬁles of the randomized participants (n ϭ 162) were creases in LDL-C levels from baseline relative to placebo were JACC Vol. 48, No. 9, 2006 Davidson et al. 1777 November 7, 2006:1774–81 Torcetrapib in Subjects With Low HDL

Table 2. Change in Standard Lipid Parameters Mean Values at Baseline and Week 8 (Baseline, Final mg/dl) Torcetrapib (mg/day) Lipid Parameter Placebo 10 30 60 90 HDL-C 39, 39 40, 43 37, 47 37, 53 37, 58 LDL-C 125, 127 128, 128 117, 122 120, 115 127, 108 Triglycerides 186, 204 176, 173 205, 211 192, 172 197, 184 Total cholesterol 201, 207 202, 206 194, 209 194, 202 204, 202 Ratio LDL-C/HDL-C 3.2, 3.3 3.3, 3.0 3.2, 2.7 3.3, 2.3 3.5, 2.1 Ratio apo B-100/apo A-I 1.0, 1.0 1.0, 1.0 1.0, 0.9 1.0, 0.8 1.1, 0.8 Percent Change From Baseline at Week 8 (LS Mean Differences Relative to Placebo Using LOCF Approach)

Torcetrapib (mg/day) Lipid Parameter 10 30 60 90 HDL-C (95% CI) 9.0 (Ϫ2.8, 20.8) 27.5‡ (15.5, 39.6) 45.1‡ (33.4, 56.9) 54.5‡ (42.8, 66.3) LDL-C (95% CI) Ϫ1.1 (Ϫ10.2, 8.0) 3.0 (Ϫ6.3, 12.3) Ϫ8.1 (Ϫ17.1, 0.9) Ϫ16.5† (Ϫ25.5, Ϫ7.4) Triglycerides (95% CI) Ϫ11.9 (Ϫ29.0, 5.2) Ϫ4.1 (Ϫ21.5, 13.3) Ϫ16.1 (Ϫ32.9, 0.8) Ϫ17.8 (Ϫ34.8, Ϫ0.8) Total cholesterol (95% CI) Ϫ0.9 (Ϫ6.5, 4.7) 4.9 (Ϫ0.8, 10.6) 0.3 (Ϫ5.2, 5.8) Ϫ3.8 (Ϫ9.3, 1.8) Ratio LDL-C/HDL-C (95% CI) Ϫ9.5* (Ϫ18.2, Ϫ0.9) Ϫ16.9† (Ϫ25.7, Ϫ8.1) Ϫ34.9‡ (Ϫ43.4, Ϫ26.3) Ϫ43.3‡ (Ϫ51.9, Ϫ34.7) Ratio apo B-100/apo A-I (95% CI) Ϫ5.2 (Ϫ13.2, 2.7) Ϫ10.5* (Ϫ18.5, Ϫ2.44) Ϫ21.5‡ (Ϫ29.3, Ϫ13.7) Ϫ31.2‡ (Ϫ39.2, Ϫ23.3)

*p Ͻ 0.05; †p Ͻ 0.01; ‡p Յ 0.0001. apo ϭ apolipoprotein; CI ϭ conﬁdence interval; LOCF ϭ last observation carried forward; LS ϭ least squares; other abbreviations as in Table 1. observed with torcetrapib 60 mg (Ϫ8.1%) and 90 mg (Ϫ16.5%; p was increased in a dose-dependent manner. Torcetrapib 60 Ͻ 0.01) (Table 2, Fig. 3). At these doses, LDL-C decreases were mg and 90 mg increased mean LDL particle size from 20.4 accompanied by signiﬁcant (p Ͻ 0.01) decreases in apo B-100 (Ϯ0.7) to 21.2 nm (Ϯ0.6) and 20.4 (Ϯ0.7) to 21.3 nm levels, suggesting a reduction in the concentration of circulating (Ϯ0.6), respectively (p Յ 0.0001 for both). LDL particles (Table 3). Interestingly, LDL-C lowering with Very low-density lipoprotein cholesterol levels showed a dose- torcetrapib was less apparent in subjects with high baseline responsive decrease from baseline, with a maximal change of Ͼ triglycerides ( 150 mg/dl) compared with those with low baseline Ϫ25% with torcetrapib 90 mg at week 8. Very low-density Ͻ triglycerides 150 mg/dl (Table 4). lipoprotein cholesterol phospholipid levels, triglyceride levels, and Nuclear magnetic resonance analysis demonstrated a subclass sizing patterns did not demonstrate any consistent dose- trend to reduction in the concentration of the small LDL-C related trends. subclass. Torcetrapib 60 mg and 90 mg decreased small Changes in non–HDL-C levels at week 8 were consistent LDL-C (18.3 to 19.7 nm) from 35.6 (SD Ϯ 39.5) to 11.9 with changes in apo B-100 levels (Table 3). mg/dl (SD Ϯ 17.3) and from 42.1 (SD Ϯ 44.2) to 10.3 mg/dl (SD Ϯ 12.0), respectively (p ϭ NS for both). Nuclear TOTAL CHOLESTEROL, TRIGLYCERIDES, AND LIPID magnetic resonance spectroscopy showed LDL particle size RATIOS. There were no appreciable changes in total cholesterol

Figure 2. Mean change in high-density lipoprotein cholesterol (HDL-C) over the course of the study—all subjects. 1778 Davidson et al. JACC Vol. 48, No. 9, 2006 Torcetrapib in Subjects With Low HDL November 7, 2006:1774–81

Table 3. Change in Other Lipid Parameters Mean Values at Baseline and Week 8 (Baseline, Final mg/dl) Torcetrapib (mg/day) Parameter Placebo 10 30 60 90 Lipoproteins Apo A-I 131.0, 129.5 133.5, 139.9 130.4, 144.8 128.0, 146.9 130.0, 160.2 Apo A-II 31.4, 30.7 31.2, 31.3 31.2, 34.1 30.9, 35.0 31.8, 35.6 Apo B-100 129.6, 132.2 131.3, 134.5 127.7, 131.9 129.2, 121.7 135.6, 117.9 Non–HDL-C 162.2, 167.3 162.9, 162.3 157.2, 162.3 157.6, 149.1 166.3, 143.4 Ultracentrifugation/precipitation analysis HDL-2 cholesterol 11.7, 11.1 12.0, 13.1 11.2, 14.6 10.9, 17.5 11.5, 22.6 HDL-3 cholesterol 27.7, 27.9 28.3, 30.2 26.5, 32.1 26.2, 35.3 26.2, 36.5 Percent Change From Baseline at Week 8 (LS Mean Differences Relative to Placebo Using LOCF Approach)

Torcetrapib (mg/day) Parameter 10 30 60 90 Lipoproteins Apo A-I (95% CI) 6.0 (Ϫ1.4, 13.5) 12.3† (4.7, 19.8) 15.8‡ (8.4, 23.1) 23.5‡ (16.1, 31.0) Apo A-II (95% CI) 2.0 (Ϫ3.4, 7.3) 11.3‡ (5.9, 16.8) 14.4‡ (9.1, 19.6) 14.3‡ (8.9, 19.6) Apo B-100 (95% CI) 0.5 (Ϫ6.0, 7.1) 1.0 (Ϫ5.7, 7.6) Ϫ8.5† (Ϫ15.0, Ϫ2.1) Ϫ15.5‡ (Ϫ22.0, Ϫ8.9) Non–HDL-C (95% CI) Ϫ3.4 (Ϫ10.6, 3.8) Ϫ0.2 (Ϫ7.5, 7.1) Ϫ9.8† (Ϫ16.9, Ϫ2.7) Ϫ18.0‡ (Ϫ25.1, Ϫ10.8) Ultracentrifugation/precipitation analysis HDL-2 cholesterol (95% CI) 14.1 (Ϫ17.7, 45.9) 34.2* (1.9, 66.6) 66.9‡ (35.5, 98.3) 108.0‡ (76.2, 140.8) HDL-3 cholesterol (95% CI) 7.5 (Ϫ1.8, 16.8) 19.1† (9.6, 28.5) 30.9‡ (21.7, 40.1) 35.8‡ (26.4, 45.1)

*p Ͻ 0.05; †p Ͻ 0.01; ‡p Յ 0.0001. Abbreviations as in Table 2. levels. Total triglyceride levels were generally decreased in discontinuation from treatment (torcetrapib 30 mg/day) and 1 torcetrapib-treated subjects relative to placebo, although there was because of mild asymptomatic abnormal liver function tests that variability in response over course of the study (Table 2). Dose- resolved without intervention following discontinuation from related decreases in the LDL-C/HDL-C ratio (Table 2, Fig. 4) treatment (placebo). Two subjects receiving torcetrapib 90 mg/day and the apo B-100/apo A-I ratio (Table 2) were consistent with withdrew from the study temporarily because of treatment-related the observed decreases in LDL-C and increases in HDL-C. AEs (gastroesophageal reﬂux disease and rash) but resumed SAFETY AND TOLERABILITY. Torcetrapib was generally well treatment and completed the study without recurrence of the AE. tolerated (Table 5). Treatment-related discontinuations were rare. The incidence of all-causality AEs was similar across Two subjects withdrew from the study permanently: 1 because of placebo and torcetrapib treatment groups, with no evidence severe diarrhea and vomiting that resolved following permanent of a dose-related response (Table 5). Most treatment-

Figure 3. Mean change in low-density lipoprotein cholesterol (LDL-C) over the course of the study—all subjects. JACC Vol. 48, No. 9, 2006 Davidson et al. 1779 November 7, 2006:1774–81 Torcetrapib in Subjects With Low HDL

Table 4. Mean Percent Change (95% CI) in LDL-C Analyzed by Baseline Total Triglyceride Levels (LS Mean Difference Relative to Placebo at Week 8, LOCF) Torcetrapib (mg/day) 10 30 60 90 TG Յ150 mg/dl 0.8 (Ϫ9.5, 11.1) Ϫ2.9 (Ϫ14.0, 8.3) Ϫ22.2 (Ϫ32.7, Ϫ11.6)* Ϫ32.9 (Ϫ44.3, Ϫ21.4)* n1310129 TG Ͼ150 mg/dl Ϫ1.5 (Ϫ14.1, 11.0) 6.8 (Ϫ5.6, 19.2) 0.1 (Ϫ12.0, 12.2) Ϫ10.3 (Ϫ22.2, 1.5) n19202224

*p Յ 0.0001. Abbreviations as in Table 2. related AEs were mild or moderate, with headache, diar- Hg change from baseline or DBP Ն105 mm Hg with a Ն15 rhea, and flatulence being the most common. There were no mm Hg change from baseline at a single visit. No subject treatment-related serious AEs or deaths. permanently discontinued treatment because of elevated Laboratory test abnormalities showed no dose-related blood pressure. trends. One subject receiving placebo demonstrated elevated liver transaminase levels (alanine aminotransferase/aspartate DISCUSSION aminotransferase Ͼ3.0 ϫ upper limits of normal [ULN]) and was withdrawn from treatment. No subject had creatine This study provides further information about the lipid- kinase elevations Ͼ10.0 ϫ ULN (Table 5). modifying benefits and safety of torcetrapib. In individuals Although in some treatment groups at Week 8 there were with low HDL-C levels, torcetrapib 30 to 90 mg/day elevations from baseline in systolic and diastolic blood resulted in substantial and significant dose-dependent ele- pressure (SBP and DBP), over the course of the study, vations in HDL-C (54.5% at 90-mg dose) and, at higher changes from baseline in SBP and DBP were highly variable doses, moderate decreases in LDL-C (Ϫ16.5% at 90-mg in all treatment groups, with no evidence of a dose-response dose). These changes in HDL and LDL are consistent with relationship with torcetrapib (Fig. 5). When all follow-up prior reports of torcetrapib (12,13). measures were averaged, mean SBP changes ranged from Of importance, albeit post-hoc and in non-randomized Ϫ0.2 mm Hg (placebo group) to 1.3 mm Hg (torcetrapib subgroups, was the observation that LDL-C lowering with 60-mg group), with none of the changes in any group torcetrapib was almost completely lost in subjects with high achieving statistical significance (all 95% CIs overlapped baseline triglyceride levels. This suggests that CETP inhi- zero) (Table 6). Mean DBP changes ranged from Ϫ0.7 mm bition may be of limited utility as a monotherapy in those Hg (torcetrapib 10-mg group) to 0.9 mm Hg (torcetrapib with high triglycerides (a highly prevalent concurrent pre- 60-mg group); again, no change in any group was significant sentation in those with low HDL and/or metabolic syn- (Table 6). drome) requiring LDL lowering. One explanation may be Of the patients receiving torcetrapib, 1.6% (2 of 129) that compositional changes in VLDL-1, in the presence of experienced elevations in blood pressure defined as 1) CETP inhibition, may lead to enhanced conversion of SBP Ն15 mm Hg or DBP Ն10 mm Hg from baseline at 3 VLDL to LDL via lipoprotein lipase. Without correspond- consecutive visits or 2) SBP Ն180 mm Hg with a Ն20 mm ing up-regulation in LDL receptor activity, there may be an

Figure 4. Change in low-density/high-density lipoprotein cholesterol (LDL-C/HDL-C) ratio over the course of the study—all subjects. 1780 Davidson et al. JACC Vol. 48, No. 9, 2006 Torcetrapib in Subjects With Low HDL November 7, 2006:1774–81

Table 5. Summary of Safety—Number of Subjects (%) Torcetrapib (mg/day) Placebo 10 30 60 90 (33 ؍ n) (34 ؍ n) (31 ؍ n) (32 ؍ n) (32 ؍ n) Treatment-related withdrawals 1 (3) 0 (0) 1 (3) 0 (0) 2 (6)* Subjects with AEs All-causality 24 (75) 19 (59) 19 (61) 22 (65) 18 (55) Treatment related 6 (19) 8 (25) 6 (19) 4 (12) 7 (21) Serious AEs All-causality 0 (0) 0 (0) 1 (3) 0 (0) 0 (0) Treatment related 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) (33 ؍ n) (31 ؍ n) (37 ؍ n) (34 ؍ n) (32 ؍ n) Clinical laboratory tests ALT/AST Ͼ3 ϫ ULN 1 (3) 0 (0) 0 (0) 0 (0) 0 (0) CK Ͼ 10 ϫ ULN 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

*Both subjects were only temporarily discontinued from treatment and completed the study. AEs ϭ adverse events; ALT ϭ alanine aminotransferase; AST ϭ aspartate aminotransferase; CK ϭ creatine kinase; ULN ϭ upper limit of normal. inability to clear the rapidly generated LDL. Statin therapy decreases in apo B are indicative of a reduction in the may resolve this issue, which would support strategies to number of circulating LDL particles. Nuclear magnetic administer torcetrapib with a statin. resonance analysis showed that torcetrapib also increased Ultracentrifugation/precipitation analysis conducted dur- the size of LDL particles. ing this study suggests that torcetrapib affects both the The effect of torcetrapib on LDL particle size may be number and size of circulating HDL and LDL particles. particularly important in reducing atherosclerosis, as the Increases in the levels of apo A-I and A-II and increases in characteristics of small dense LDL make it more athero- HDL2 and HDL3 cholesterol are indicative of increased genic than larger, less dense LDL (5). Even if there was no numbers of the larger subspecies of HDL particles. This was significant effect of torcetrapib on HDL-C levels, the effects confirmed by NMR HDL subclass analysis. Conversely, on LDL-C, including modest decreases in overall levels and a shift in particle size from small to large, might be expected to provide benefit. In addition to further elucidating the beneficial effects of torcetrapib on lipid metabolism, this trial provides important safety data. Generally, torcetrapib was well tolerated, discontinuations from treatment were rare, there were no apparent dose-related trends in the incidences of AEs, and most AEs were mild or moderate in nature. Although previously published studies of torcetrapib have not reported effects on blood pressure (12,13), increases in blood pressure were observed in some individuals in this study. However, the lack of a consistent treatment-related pattern over time and dose suggests the effect to be of limited magnitude within this dose range. Further studies are underway to define the magnitude and clinical relevance of these blood pressure changes. The link between lower LDL-C levels and decreased cardiovascular risk has been clearly demonstrated in CVD prevention trials with statins (18). Furthermore, recent statins trials provide evidence that aggressive versus moderate LDL lowering is associated with additional benefits (19,20). Yet there is less clinical trial data showing the benefits of increasing HDL-C levels and a distinct paucity of clinical trial data to show the impact of aggressively elevating HDL-C on clinical end points. This may partly be Figure 5. Least-squares mean change in systolic (A) and diastolic (B) due to the current lack of well-tolerated drugs that can blood pressure over the course of the study. substantially increase HDL-C (21). Thus, although current JACC Vol. 48, No. 9, 2006 Davidson et al. 1781 November 7, 2006:1774–81 Torcetrapib in Subjects With Low HDL

Table 6. Longitudinal Analysis of Changes in Blood Pressure: Average of All Follow-Up Measures Over the Course of the Study (All Subjects, Observed Cases) Torcetrapib (mg/day) Placebo 10 30 60 90 (33 ؍ n) (34 ؍ n) (30 ؍ n) (32 ؍ n) (32 ؍ n) Systolic blood pressure Least-squares mean change Ϫ0.17 Ϫ0.15 0.98 1.29 0.85 95% confidence interval Ϫ2.01, 1.66 Ϫ1.93, 1.62 Ϫ0.85, 2.82 Ϫ0.46, 3.04 Ϫ0.89, 2.59 Diastolic blood pressure Least-squares mean change 0.37 Ϫ0.7 0.66 0.88 0.83 95% confidence interval Ϫ0.86, 1.60 Ϫ1.90, 0.49 Ϫ0.57, 1.90 Ϫ0.30, 2.06 Ϫ0.34, 2.01 guidelines for CVD prevention recognize low HDL-C 7. Guerin M, Le Goff W, Lassel TS, Van Tol A, Steiner G, Chapman levels as a risk factor, they continue to place most emphasis MJ. Atherogenic role of elevated CE transfer from HDL to VLDL(1) and dense LDL in type 2 diabetes: impact of the degree of triglycer- on decreasing LDL-C levels (22). This fact, combined with idemia. Arterioscler Thromb Vasc Biol 2001;21:282–8. the relatively modest decreases in LDL-C observed with 8. Sugano M, Makino N, Sawada S, et al. Effect of antisense oligonu- torcetrapib and JTT-705, means that CETP inhibitors are cleotides against cholesteryl ester transfer protein on the development of atherosclerosis in cholesterol-fed rabbits. J Biol Chem 1998;273: likely to be used in combination with statin therapy. A 5033–6. separate phase 2 study evaluating the efficacy and safety of 9. Rittershaus CW, Miller DP, Thomas LJ, et al. Vaccine-induced torcetrapib when administered on a background of atorva- antibodies inhibit CETP activity in vivo and reduce aortic lesions in a statin to subjects with below-average HDL-C levels is rabbit model of atherosclerosis. Arterioscler Thromb Vasc Biol 2000; 20:2106–12. reported in this issue of the Journal. 10. Okamoto H, Yonemori F, Wakitani K, Minowa T, Maeda K, Shinkai H. A cholesteryl ester transfer protein inhibitor attenuates atheroscle- Acknowledgments rosis in rabbits. Nature 2000;406:203–7. 11. Morehouse LA, Sugarman ED, Bourassa PA, Milici AJ. The CETP- We are indebted to the study investigators, coordinators, inhibitor torcetrapib raises HDL and prevents aortic atherosclerosis in and subjects whose participation made this study possible. rabbits. Presented at the XVth International Symposium on Drugs We also acknowledge the staff of Development Operations Affecting Lipid Metabolism. Venice, Italy; October 24–27, 2004. at Pfizer Global Research and Development for their 12. Clark RW, Sutfin TA, Ruggeri RB, et al. Raising high-density lipoprotein in humans through inhibition of cholesteryl ester transfer assistance in conducting the study and the Pfizer Clinical protein: an initial multidose study of torcetrapib. Arterioscler Thromb Statistics/Data Management Team (including Michael Li, Vasc Biol 2004;24:490–7. Clio Wu, and Michael Fetchel) for providing the data. 13. Brousseau ME, Schaefer EJ, Wolfe ML, et al. Effects of an inhibitor of cholesteryl ester transfer protein on HDL cholesterol. N Engl J Med 2004;350:1505–15. Reprint requests and correspondence: Dr. Michael Davidson, 14. Bell TJ. Statistical Analysis of NHANES III Data. Research Triangle Radiant Research, 515 North State Street, Suite 2700, Chicago, Park, NC: Research Triangle Institute, 2000. Illinois 60610. E-mail: [email protected]. 15. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499–502. REFERENCES 16. Otvos JD, Jeyarajah EJ, Bennett DW. Quantification of plasma lipoproteins by proton nuclear magnetic resonance spectroscopy. Clin 1. Brewer HB Jr. Benefit-risk assessment of rosuvastatin 10 to Chem 1991;37:377–86. 40 milligrams. Am J Cardiol 2003;92:23–9K. 17. Ruberg SJ. Dose response studies. II. Analysis and interpretation. 2. Chapman MJ. Fibrates in 2003: therapeutic action in atherogenic J Biopharm Stat 1995;5:15–42. dyslipidaemia and future perspectives. Atherosclerosis 2003;171:1–13. 18. Maron DJ, Fazio S, Linton MF. Current perspectives on statins. 3. Chapman MJ, Assmann G, Fruchart JC, Shepherd J, Sirtori C. Circulation 2000;101:207–13. Raising high-density lipoprotein cholesterol with reduction of cardio- 19. Cannon CP, Braunwald E, McCabe CH, et al. Intensive versus vascular risk: the role of nicotinic acid—a position paper developed by the European Consensus Panel on HDL-C. Curr Med Res Opin moderate lipid lowering with statins after acute coronary syndromes. 2004;20:1253–68. N Engl J Med 2004;350:1495–504. 4. Barter PJ, Brewer HB Jr., Chapman MJ, Hennekens CH, Rader DJ, 20. LaRosa JC, Grundy SM, Waters DD, et al. Intensive lipid lowering Tall AR. Cholesteryl ester transfer protein: a novel target for raising with atorvastatin in patients with stable coronary disease. N Engl HDL and inhibiting atherosclerosis. Arterioscler Thromb Vasc Biol J Med 2005;352:1425–35. 2003;23:160–7. 21. Rosenson RS. Low HDL-C: a secondary target of dyslipidemia 5. Krauss RM. Heterogeneity of plasma low-density lipoproteins and therapy. Am J Med 2005;118:1067–77. atherosclerosis risk. Curr Opin Lipidol 1994;5:339–49. 22. Expert Panel on Detection Evaluation and Treatment of High Blood 6. Hayek T, Azrolan N, Verdery RB, et al. Hypertriglyceridemia and Cholesterol in Adults. Third report of the National Cholesterol cholesteryl ester transfer protein interact to dramatically alter high Education Program (NCEP) Expert Panel on Detection, Evaluation, density lipoprotein levels, particle sizes, and metabolism. Studies in and Treatment of High Blood Cholesterol in Adults (Adult Treat- transgenic mice. J Clin Invest 1993;92:1143–52. ment Panel III). J Am Med Assoc 2001;285:2486–97.