Author + information
- Received January 5, 2010
- Revision received June 22, 2010
- Accepted June 28, 2010
- Published online January 11, 2011.
- Stephen J. Nicholls, MBBS, PhD⁎,†,‡,⁎ (, )
- E. Murat Tuzcu, MD⁎,
- Kathy Wolski, MPH⁎,
- Ozgur Bayturan, MD⁎,
- Andrea Lavoie, MD⁎,
- Kiyoko Uno, MD⁎,
- Stuart Kupfer, MD§,
- Alfonso Perez, MD§,
- Richard Nesto, MD∥ and
- Steven E. Nissen, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Stephen J. Nicholls, Department of Cardiovascular Medicine, Cleveland Clinic, Mail Code JJ-65, 9500 Euclid Avenue, Cleveland, Ohio 44195
Objectives The purpose of this study was to determine the factors associated with the favorable effect of pioglitazone on atheroma progression.
Background Diabetes mellitus is associated with accelerated coronary atheroma progression. Pioglitazone slowed progression compared with glimepiride in this population.
Methods In all, 360 diabetic patients with coronary artery disease were treated with pioglitazone or glimepiride for 18 months in the PERISCOPE (Pioglitazone Effect on Regression of Intravascular Sonographic Coronary Obstruction Prospective Evaluation) study. Coronary atheroma progression was evaluated by serial intravascular ultrasound. The relationship between changes in biochemical parameters, percent atheroma volume, and total atheroma volume was investigated.
Results Pioglitazone-treated patients demonstrated greater increases in high-density lipoprotein cholesterol (HDL-C) and reductions in glycated hemoglobin, triglycerides, and C-reactive protein. Significant correlations were observed between changes in percent atheroma volume and triglycerides (r = 0.15, p = 0.04), triglyceride/HDL-C ratio (r = 0.16, p = 0.03), and glycated hemoglobin (r = 0.16, p = 0.03) with pioglitazone, and changes in low-density lipoprotein cholesterol (r = −0.15, p = 0.05), apolipoprotein B (r = −0.16, p = 0.04), and apolipoprotein A-I (r = −0.20, p = 0.01) with glimepiride. Substantial atheroma regression, compared to progression, was associated with greater relative increases in HDL-C (14.2% vs. 7.8%, p = 0.04), relative decreases in triglycerides (−13.3% vs. −1.9%, p = 0.045), triglyceride/HDL-C ratio (−22.5 vs. −9.9%, p = 0.05), and decrease in glycated hemoglobin (−0.6% vs. −0.3%, p = 0.01). Multivariable analysis revealed that pioglitazone-induced effects on triglyceride/HDL-C were associated with changes in percent atheroma volume (p = 0.03) and total atheroma volume (p = 0.02).
Conclusions Favorable effects of pioglitazone on the triglyceride/HDL-C ratio correlated with delayed atheroma progression in diabetic patients. This finding highlights the potential importance of targeting atherogenic dyslipidemia in diabetic patients with coronary artery disease.
The presence of diabetes mellitus portends an adverse cardiovascular prognosis in the presence and absence of established coronary heart disease (1–3). Despite the relationship between hyperglycemia and prospective cardiovascular risk in population studies, glucose-lowering strategies generally have not reduced event rates in randomized controlled trials (4,5). This observation is supported by recent trials that report a lack of efficacy, and potential harm, in trials of intensive glycemic control (6–9). As a result, considerable debate has focused on the need to evaluate the cardiovascular safety of emerging antidiabetic therapies (10).
In contrast, therapeutic interventions that lower low-density lipoprotein cholesterol (LDL-C) and blood pressure are associated with unequivocal reductions in macrovascular events (11–15). As a result of these observations, guidelines for cardiovascular prevention in patients with diabetes mellitus emphasize the need for intensive lowering of these risk factors (16). Increasingly, attention has focused on the importance of hypertriglyceridemia, low levels of high-density lipoprotein cholesterol (HDL-C), and inflammation as factors driving diabetic cardiovascular disease.
Coronary artery imaging with intravascular ultrasonography has demonstrated accelerated progression of atherosclerosis in patients with diabetes, despite the use of intensive lipid-lowering regimens (17). These findings suggest that additional therapies will be required to achieve more effective prevention of cardiovascular events in diabetic patients. The PERISCOPE (Pioglitazone Effect on Regression of Intravascular Sonographic Coronary Obstruction Prospective Evaluation) study compared the impact of 2 glucose-lowering strategies, pioglitazone and glimepiride, on disease progression and demonstrated that the peroxisome proliferator activated receptor-γ agonist pioglitazone halted progression of coronary atherosclerosis (18). The current analysis was designed to determine the traditional biochemical factors that were associated with this antiatherosclerotic activity.
The details of the PERISCOPE study have been described in detail previously (18). In brief, 543 patients with type 2 diabetes mellitus with a diagnosis of coronary artery disease, defined as the presence of at least 1 20% lumen stenosis in a major epicardial coronary artery on an angiogram performed for a clinical indication were enrolled in the study. Patients were treated for 18 months with either glimepiride 1 to 4 mg or pioglitazone 15 to 45 mg with dose titration to achieve a fasting plasma glucose <140 mg/dl. Coronary intravascular ultrasonography imaging was performed at baseline and at the end of the study in 360 patients. The study was approved by each participating center's institutional review board.
Image acquisition and analysis
The details of ultrasonic image acquisition and analysis have been described in detail previously. Briefly, after anticoagulation therapy and administration of intracoronary nitroglycerin, a high frequency (40 MHz) ultrasound transducer (Atlantis SR Pro, Boston Scientific, Maple Grove, Minnesota) was placed as distally as possible within the target coronary artery. Imaging was acquired while continuously withdrawing the catheter through the artery back to the aorta at a constant rate of 0.5 mm/s by a motorized pullback. Images were digitized, and analysis of each segment was selected by using proximal and distal side branches as reference points to enable subsequent analysis of the same segment at follow-up. Images spaced 1 mm apart were selected for analysis. The leading edges of the lumen and external elastic membrane (EEM) were traced by manual planimetry. Total atheroma volume (TAV) was calculated as the summation of all measured plaque areas and normalized by adjusting for the median number of images analyzed in the entire cohort to account for heterogeneity of segment length: Percent atheroma volume (PAV) was calculated as the percentage of vessel wall occupied by atherosclerotic plaque:
Clinical characteristics and baseline medication use were expressed as mean ± SD for continuous variables and percentage for categorical variables. The chi-square test or Fisher exact test were used, where appropriate, for categorical variables, and t tests were used for continuous variables. Biochemical parameters with a non-normal distribution (triglycerides, triglyceride/HDL ratio, apolipoprotein [apo]B, apoA-I, apoB/apoA-I ratio, and C-reactive protein [CRP]) were summarized using median and interquartile range. Absolute and percentage change for biochemical parameters and atheroma burden were summarized by treatment group, and the percentage change from baseline was tested within each treatment group and between treatment groups using an analysis of covariance model to adjust for the baseline value. For analysis purposes, biochemical parameters with a non-normal distribution were log-transformed. Percentage change from baseline was calculated as: log (follow-up value/baseline value). The relationship between either baseline levels or changes in biochemical parameters and changes in PAV in the entire cohort were characterized by correlation coefficients. Substantial regression or progression was defined as >5% relative decrease or increase in PAV, respectively. Factors independently associated with disease progression in the entire study cohort, and patients treated with either pioglitazone or glimepiride, were determined by multivariable linear regression analysis. In addition to baseline characteristics (age, sex, race, body mass index, smoking status, and self-reported history of hypertension, hypercholesterolemia, myocardial infarction, coronary intervention, and concomitant medication use), biochemical parameters considered for inclusion in the model included change in glycated hemoglobin (HbA1c), LDL-C, HDL-C, triglycerides, triglycerides/HDL-C ratio, apoB, apoA-I, apoB/A-I ratio, and CRP. Baseline percent atheroma volume was forced into the model, and variables with a p value <0.15 were retained. All statistical analyses were performed using SAS version 9.1 (SAS Institute, Cary, North Carolina).
The clinical characteristics and concomitant medication use of patients are summarized in Table 1. Patients had a mean age of 60 years, were predominantly male, and had a high prevalence of obesity. Pioglitazone-treated patients were less likely to be current smokers and were slightly less likely to have a history of hypertension. No differences were noted between groups with regard to concomitant medication use. In particular, there was no difference in the background use of statin therapy, regardless of dose.
Biochemical parameters and measures of atheroma burden
Levels of biochemical parameters and measures of atheroma burden at baseline and the serial change in these parameters are summarized in Table 2. Pioglitazone-treated patients demonstrated slightly greater reductions in HbA1c (−0.6 ± 0.01% vs. −0.4 ± 0.1%, p = 0.01) and greater relative reductions in triglycerides (−15.3 ± 2.8% vs. 0.6 ± 2.7%, p < 0.001), triglyceride/HDL-C ratio (−26.4 ± 3.5% vs. −2.6 ± 3.3%, p < 0.001), apoB (−5.0 ± 1.7% vs. +1.7 ± 1.6%, p = 0.003), apoB/A-I ratio (−11.0 ± 2.0% vs. −4.2 ± 1.8 %, p = 0.005), and CRP (−44.9 ± 7.6% vs. −18.0 ± 7.3%, p < 0.001). A greater relative increase in HDL-C (16.1 ± 1.4% vs. 4.2 ± 1.3%, p < 0.001) was also observed with pioglitazone, despite no differential effect on apoA-I. Pioglitazone had a favorable effect on progression of PAV (−0.16 ± 0.21% vs. +0.73 ± 0.20%, p = 0.002) and TAV (−5.5 ± 1.6 mm3 vs. −1.5 ± 1.5 mm3, p = 0.06) compared with glimepiride.
Relationship between biochemical parameters and change in atheroma burden
The relationship between changes in biochemical parameters and changes in measures of atheroma burden are summarized in Table 3. Significant univariate correlations were observed between changes in PAV and changes in triglycerides, the triglyceride/HDL-C ratio, and HbA1c in pioglitazone-treated patients and changes in LDL-C, apoB, and apoA-I in glimepiride-treated patients. Significant correlations were also observed between changes in TAV and changes in HDL-C, triglyceride/HDL-C, and HbA1c with pioglitazone and changes in apoA-I with glimepiride. For the entire cohort, changes in biochemical parameters in patients who underwent a substantial degree of plaque regression (n = 59, >5% relative decrease in PAV) or progression (n = 76, >5% relative increase in PAV) were also investigated. Substantial regression, compared to progression, was associated with a greater relative increase in HDL-C (14.2 ± 2.4% vs. 7.8 ± 2.1%, p = 0.04), relative reductions in triglyceride (−13.3 ± 4.7% vs. −1.9 ± 4.1%, p = 0.045), and triglyceride/HDL-C ratio (−22.5 ± 5.9% vs. −9.9 ± 5.2%, p = 0.05), and decrease in HbA1c (−0.6 ± 0.1% vs. −0.3 ± 0.1%, p = 0.01). The relationship between changes in biochemical parameters and measures of atheroma burden in the treatment groups was explored further by examination of tertiles of changes in these parameters. Significant univariate relationships were noted between decreasing levels of the triglyceride/HDL-C ratio and HbA1c with slowing of PAV in pioglitazone-treated patients. In contrast, there was no significant association between changes in biochemical parameters and PAV in patients treated with glimepiride (Table 4). Similarly, greater decreases in the triglyceride/HDL-C ratio and HbA1c were associated with slower rates of progression of TAV, although the latter just failed to meet statistical significance (Table 5).
On multivariable analysis of pioglitazone-treated patients, factors that were independently associated with progression of PAV included the change in the triglyceride/HDL-C in the pioglitazone-treated patients (p = 0.03). In contrast, changes in these metabolic parameters did not predict changes in PAV in glimepiride-treated patients (Table 6). Similarly, baseline disease burden (p = 0.001) and changing the triglyceride/HDL-C ratio (p = 0.02), but not HbA1c (p = 0.16) was independently associated with changes in TAV in pioglitazone-treated patients, whereas changes in these parameters were not associated with changes in TAV in the glimepiride group. In the adjusted model, the p value for the interaction term between treatment group and change in triglyceride/HDL-C ratio and change in PAV did not meet statistical significance (p = 0.14).
The optimal therapeutic approach towards reduction of cardiovascular risk in patients with diabetes and established coronary heart disease continues to be a source of considerable debate. In the current post hoc analysis of diabetic patients with coronary artery disease, changing levels of both the triglyceride/HDL-C ratio and HbA1c were each associated with reduction in disease progression. However, multivariable analysis revealed that lowering the triglyceride/HDL-C ratio was the only parameter independently associated with slowing of disease progression with pioglitazone. This observation highlights the potential importance of atherogenic dyslipidemia as a therapeutic target to reduce cardiovascular risk in patients with diabetes.
The association between changes in the triglyceride/HDL-C ratio and atheroma progression provides further evidence of the role of abnormal lipid homeostasis in the propagation of diabetic cardiovascular disease. Atherogenic dyslipidemia, characterized by hypertriglyceridemia, low levels of HDL-C, and normal levels of LDL-C, is commonly encountered in patients with diabetes (19). The influence of these lipid abnormalities on the arterial wall may explain the observed benefit with pioglitazone treatment. This drug has favorable effects on triglycerides and HDL-C, but minimal effects on levels of LDL-C. These findings are consistent with a prior report that raising HDL-C predicted slowing progression of carotid intimal-medial thickness with pioglitazone in diabetic patients (20). While the lack of a similar association in patients treated with glimepiride is likely to reflect their lack of effect on triglycerides and HDL-C, it remains to be determined whether there are other factors underlying the differences between the agents.
The finding that LDL-C is not a predictor of disease progression in the current analysis of pioglitazone-treated patients, most of whom were treated with a statin at baseline, is likely to reflect the relative importance of atherogenic dyslipidemia in diabetes and does not detract from reports that LDL-C is independently associated with progression in broader populations with coronary artery disease (21–23). Furthermore, the findings do not diminish the importance of lowering LDL-C in diabetic patients with coronary artery disease.
The relationship between the degree of glycemic control and disease progression is controversial. The current findings support previous reports that HbA1c, a measure of chronic glucose homeostasis, weakly correlates with the degree of disease progression (17,24) and cardiovascular risk (25,26) in population-based studies. Although a statistically significant difference was observed between glimepiride and pioglitazone in effects on HbA1c, the absolute difference was relatively small and unlikely accounted for the marked benefit of pioglitazone in arresting disease progression.
The finding that changes in HbA1c were not independently associated with disease progression, after controlling for additional cardiovascular risk factors, is consistent with recent reports demonstrating no benefit of intensive glucose lowering on macrovascular outcomes (6–9). Although a recent meta-analysis of glucose lowering trials suggested cardiovascular benefit (27), this has yet to be demonstrated in an appropriately powered randomized clinical trial. We did observe a weak relationship between changes in measures of glycemic control and atheroma progression, but that was predominantly driven by the finding of accelerated disease progression in those subjects whose HbA1c increased. These findings suggest that marked and sustained lowering of glucose might have a beneficial impact on coronary atherosclerosis. The current findings do not exclude the possibility that the benefit of pioglitazone was not due, in part, to a reduction in insulin resistance. This relationship requires additional investigation in further studies. Accordingly, there remains considerable interest in determining whether antidiabetic therapies can have a cardiovascular benefit.
A number of caveats in the interpretation of the current study should be noted. All patients had presented for a clinically indicated coronary angiogram. It remains uncertain whether the current findings can be extrapolated to asymptomatic subjects. The data are derived from study of a relatively small number of patients who were followed up for only 18 months. It remains to be determined to what degree these associations persist on longer follow-up. Conventional gray scale ultrasonography has limited resolution and cannot characterize atheroma morphology. Accordingly, it remains uncertain whether changes in measures of atherogenic dyslipidemia and glycemic control can result in changes in plaque composition. Traditional lipid parameters were evaluated in the PERISCOPE study. Changes in lipid subspecies may have contributed to these results, and this possibility will require further investigation in future clinical trials of peroxisome proliferator activated receptor-γ agonists. Finally, although increasing data support a relationship between disease progression and clinical outcome (28), that requires ongoing validation. It remains to be determined how the current findings can be applied to assessment of the risk of adverse cardiovascular outcomes.
In summary, a favorable effect on the triglyceride/HDL-C ratio was independently associated with slowing of progression of coronary atherosclerosis in patients with diabetes and established coronary artery disease. This finding supports the hypothesis that pioglitazone halted disease progression predominantly because of its properties beyond glycemic control. These findings are also consistent with clinical outcome data indicating the importance of atherogenic dyslipidemia as a target for therapeutic manipulation in patients with diabetes mellitus to achieve more effective prevention of cardiovascular disease.
The authors thank the technical expertise of the Intravascular Ultrasound Core Laboratory at the Cleveland Clinic.
The PERISCOPE study was sponsored by Takeda Pharmaceuticals. Dr. Nicholls has received honoraria and consulting fees from AstraZeneca, Takeda, Pfizer, Merck Schering-Plough, and Roche; and research support from Novartis, AstraZeneca, Eli Lilly, and Resverlogix. Dr. Tuzcu has received speaking honoraria and research support from Pfizer and AstraZeneca. Dr. Lavoie is speaker for Boehringer and AstraZeneca. Drs. Kupfer and Perez are employees of Takeda Pharmaceuticals. Dr. Nesto has received consulting fees and research support from Takeda Pharmaceuticals, is a consultant for GlaxoSmithKline, and serves on the advisory board of Sanofi-Aventis. Dr. Nissen has received research support from AstraZeneca, Eli Lilly, Pfizer, Takeda, Sankyo, and Sanofi-Aventis; he has consulted for a number of pharmaceutical companies without financial compensation. All honoraria, consulting fees, or other payments from any for-profit entity are paid directly to charity, so that neither income nor any tax deduction is received. All other authors have reported that they have no relationships to disclose.
- Abbreviations and Acronyms
- C-reactive protein
- external elastic membrane
- glycated hemoglobin
- high-density lipoprotein cholesterol
- low-density lipoprotein cholesterol
- percent atheroma volume
- total atheroma volume
- Received January 5, 2010.
- Revision received June 22, 2010.
- Accepted June 28, 2010.
- American College of Cardiology Foundation
- Barr E.L.,
- Zimmet P.Z.,
- Welborn T.A.,
- et al.
- Malmberg K.,
- Ryden L.,
- Wedel H.,
- et al.
- Skyler J.S.,
- Bergenstal R.,
- Bonow R.O.,
- et al.
- UK Prospective Diabetes Study Group
- ↵(2009) Executive Summary: Standards of medical care in diabetes—2009. Diabetes Care 32(Suppl 1):6–12.
- Nicholls S.J.,
- Tuzcu E.M.,
- Kalidindi S.,
- et al.
- Davidson M.,
- Meyer P.M.,
- Haffner S.,
- et al.
- Stratton I.M.,
- Adler A.I.,
- Neil H.A.,
- et al.
- Nicholls S.J.,
- Hsu A.,
- Wolski K.,
- et al.