Author + information
- Received March 4, 2011
- Revision received April 22, 2011
- Accepted April 22, 2011
- Published online October 1, 2011.
- Esad Vucic, MD⁎,†,
- Stephen D. Dickson, MS⁎,
- Claudia Calcagno, MD, PhD⁎,‡,
- James H.F. Rudd, MD, PhD§,
- Erin Moshier, MS∥,
- Katsumi Hayashi, MD⁎,
- Jessica S. Mounessa, BS⁎,
- Michelle Roytman, BS⁎,
- Matthew J. Moon, BS⁎,
- James Lin, MS⁎,
- Sotirios Tsimikas, MD¶,
- Edward A. Fisher, MD, PhD#,
- Klaas Nicolay, PhD⁎⁎,
- Valentin Fuster, MD, PhD†,†† and
- Zahi A. Fayad, PhD⁎,†,‡,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Zahi A. Fayad, Translational and Molecular Imaging Institute, Mount Sinai School of Medicine, One Gustave L. Levy Place, Box 1234, New York, New York 10029
Objectives We sought to determine the antiatherosclerotic properties of pioglitazone using multimethod noninvasive imaging techniques.
Background Inflammation is an essential component of vulnerable or high-risk atheromas. Pioglitazone, a peroxisome proliferator-activated receptor-gamma agonist, possesses potent anti-inflammatory properties. We aimed to quantify noninvasively the anti-inflammatory effects of pioglitazone on atheroma using 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET)/computed tomography (CT) and dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI).
Methods Atherosclerotic plaques were induced in the aorta of 15 New Zealand white rabbits by a combination of a hyperlipidemic diet and 2 balloon endothelial denudations. Nine rabbits continued the same diet, whereas 6 rabbits received pioglitazone (10 mg/kg orally) in addition to the diet. Twelve animals underwent 18F-FDG-PET/CT, and 15 animals underwent DCE-MRI at baseline, 1 month, and 3 months after treatment initiation. Concomitantly, serum metabolic parameters were monitored. After imaging was completed, aortic histologic analysis and correlation analysis were performed.
Results The 18F-FDG-PET/CT imaging detected an increase in average standardized uptake value in the control group (p < 0.01), indicating progressive inflammation, whereas stable standardized uptake values were observed in the treatment group, indicating no progression. The DCE-MRI analysis detected a significant decrease in the area under the curve for the pioglitazone group (p < 0.01). Immunohistologic examination of the aortas demonstrated a significant decrease in macrophage and oxidized phospholipid immunoreactivity in the pioglitazone group (p = 0.04 and p = 0.01, respectively) with respect to control animals, underlining the imaging results. Serum metabolic parameters showed no difference between groups. Strong positive correlations between standardized uptake value and macrophage density and between area under the curve and neovessels were detected (r2 = 0.86 and p < 0.0001, and r2 = 0.66 and p = 0.004, respectively).
Conclusions Both 18F-FDG-PET/CT and DCE-MRI demonstrate noninvasively the anti-inflammatory effects of pioglitazone on atheroma. Both imaging methods seem suited to monitor inflammation in atherosclerosis.
- dynamic contrast-enhanced magnetic resonance imaging
- 18F-fluorodeoxyglucose positron emission tomography/computed tomography
Cardiovascular disease remains the major cause of morbidity and mortality worldwide (1), despite the widespread use of low-density lipoprotein cholesterol-lowering therapies. Inflammation, mainly promoted by oxidized lipids in the vessel wall, plays a critical role in all stages of atherosclerosis, but is crucial in plaque rupture and thrombosis, which causes most acute coronary events (2,3). Vulnerable plaques are characterized by a relatively large lipid core, a thin fibrous cap, and an abundance of inflammatory cells (especially macrophages) accompanied by neovascularization (4). Currently, therapies targeting reversal cholesterol transport and inflammation, in particular recombinant high-density lipoprotein therapeutics and various nuclear receptor agonists, are being developed (5). However, their translation into clinical practice has been difficult because of the lack of clinical outcome studies and validated surrogate endpoints (6).
The peroxisome proliferator-activated receptor gamma (PPAR-γ) belongs to a nuclear receptor superfamily of ligand-dependent transcription factors. Different cells involved in the pathogenesis of atherosclerosis, such as endothelial cells, smooth muscle cells, T-lymphocytes, dendritic cells, and mainly macrophages, express PPAR-γ receptor (7,8). Synthetic PPAR-γ receptor ligands, thiazolidinediones, are used in clinical practice as oral antidiabetic agents. Besides their antidiabetic properties, thiazolidinediones have demonstrated in animal studies potent antiatherosclerotic effects by inhibiting the formation of atherosclerotic plaques (9,10). However, some thiazolidinediones may increase the risk of myocardial infarction (11). Pioglitazone, in contrast, in the large clinical secondary prevention PROspective pioglitAzone Clinical Trial In macroVascular Events trial (more than 5,000 subjects), demonstrated a significant reduction of all-cause mortality, myocardial infarction, and stroke in a diabetic patient population (12). Two subsequent large trials (CHICAGO [Carotid Intima-Media Thickness in Atherosclerosis Using Pioglitazone] and PERISCOPE [Pioglitazone Effect on Regression of Intravascular Sonographic Coronary Obstruction Prospective Evaluation]) showed decreased progression of carotid intima-media thickness (as assessed by surface and intravascular ultrasounds) after 18 months of treatment with pioglitazone. These 2 studies confirmed the significant vascular benefits of pioglitazone and highlighted the potential role of imaging in tracking the progression of disease or its regression on therapeutic intervention. However, the imaging techniques used in these studies do not give any information regarding changes in vascular inflammation induced by pioglitazone, which may help to identify responsive patients much before anatomical changes become apparent. Noninvasive imaging techniques targeted toward quantifying inflammation, macrophage content, and neovascularization of atherosclerotic vessels represent an attractive clinical tool. In an ideal scenario, such techniques would allow an accurate readout of vessel wall inflammation, delineating cardiovascular risk, guiding therapeutic decisions, and defining the role of novel treatments, and therefore may aid in risk stratification and may provide surrogate therapeutic endpoints (13). Recently, a study in mice showed that noninvasive optical molecular imaging techniques can detect changes in inflammation on pioglitazone treatment (14). However, the techniques used in these studies cannot be applied in a clinical setting because of the limited depth penetration possible with optical imaging. 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET)/computed tomography (CT) and dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) are noninvasive, clinically available techniques and recently have been validated as indicators of vascular inflammation: 18F-FDG-PET/CT detected vascular macrophage content and DCE-MRI detected neovascularization (15–18). In this study, we applied 18F-FDG-PET/CT and DCE-MRI to characterize vascular inflammation and neovascularization and its changes during pioglitazone treatment in a rabbit model of atherosclerosis.
Fifteen New Zealand white rabbits (Charles River Laboratories, Inc., Wilmington, Massachusetts) at the age of 3 months (weight: 3.0 ± 0.1 kg) were fed a diet containing 0.3% cholesterol and 4.7% coconut oil for 3 months and 0.15% cholesterol and 4.7% coconut oil thereafter. The development of atherosclerosis was enhanced further with a double balloon injury to the abdominal aorta at weeks 2 and 6 (19). At 4 months, the animals were divided into 2 groups (control: n = 9, treatment: n = 6). The treatment group received pioglitazone (10 mg/kg body weight) (20) mixed with the high-fat diet for 3 months (Fig. 1). Three age-matched animals on a chow diet served as nonatherosclerotic controls. Diets were prepared by Research Diets Inc. (New Brunswick, New Jersey). All procedures were performed under anesthesia by intramuscular injection of ketamine (20 mg/kg) (Fort Dodge Animal Health, Overland Park, Kansas) and xylazine (5 mg/kg) (Bayer AG, Leverkusen, Germany). The protocol was approved by the Mount Sinai School of Medicine Institute Animal Care and Use Committee.
Serum analysis and fast protein liquid chromatography
Blood was collected from 12-h fasting animals at baseline, 4, 5, and 7 months after high-cholesterol diet and was separated into serum per standard protocols (21). Serum samples were analyzed using commercial kits for cholesterol, triglycerides, glucose, and insulin. In addition, pooled serum at baseline and after 3 months of pioglitazone treatment was subjected to fast protein liquid chromatography.
Macrophages were detected on adjacent slices by standard immunohistochemistry techniques. The total macrophage-rich (RAM-11–positive) area was quantified digitally as fraction of the intimal-media area (i.e., macrophage density) per slice and was expressed as the average macrophage density per animal. Similarly, the areas for smooth muscle actin, apolipoprotein B-100, and oxidized phospholipids were evaluated. Neovessels were detected on a single DCE image matched section with an anti-CD31 antibody, as previously described (17).
Six rabbits from each group were imaged using a combined PET/CT scanner (Discovery LS, GE Healthcare). The PET images were acquired for 10 min, 3 h after intravenous injection of 18F-FDG (37 MBq [1 mCi/kg]) (19). The images were acquired in 3-dimensional mode, and the acquisition covered the area from the diaphragm to the iliac bifurcation. A Fourier iterative reconstruction of the PET data was used. In addition to a PET scan, a non–contrast-enhanced CT scan was obtained for each rabbit on the same scanner for attenuation correction and anatomical localization of the PET signal. The final reconstructed slice thickness was 4.25 mm.
The PET/CT data were displayed in 3 orthogonal planes on a GE Xeleris workstation. Maximal standard uptake values (SUVmax) and mean standard uptake values (SUVmean) were recorded on contiguous axial slices of the aorta between the left renal artery and the iliac bifurcation (average number of slices: 13.6) from regions of interest encompassing the vessel wall. Results were expressed as the average SUVmean of the aorta per animal.
Anatomic and DCE-MRI
Nine control and 6 pioglitazone-treated rabbits were imaged under anesthesia using a 1.5-T MRI system (Siemens Sonata, Siemens Medical Solutions) and a conventional knee coil. Sequential axial, 3-mm thick images of the aorta were obtained from the left renal artery to the iliac bifurcation using T1-weighted, T2-weighted, and PD (proton density)-weighted imaging sequences. The DCE-MRI was performed on 1 selected axial slice using a black blood turbo spin echo sequence previously validated against histologic analysis (17).
Anatomic multicontrast-MRI and DCE-MRI analysis
Average vessel wall area of the aorta between the left renal artery and iliac bifurcation was quantified on contiguous slices by manual tracing on T2-weighted images (VesselMass Software, Leiden University Medical Center). Then, DCE-MRI quantified the change of signal intensity in a region of interest including the entire atherosclerotic plaque during the injection of the contrast agent with a custom-made Matlab (The MathWorks Inc., Natick, Massachusetts) program. The area under the curve (AUC) was quantified at 1 min (AUC1min) and at 2 min (AUC2min) after injection of contrast agent by numerical integration (17).
Numeric values are expressed as mean ± SD. A p value <0.05 was considered statistically significant. Continuous DCE-MRI and PET outcomes were modeled as linear functions of time for both treatment and control groups. A linear mixed model with random intercepts and compound symmetric covariance structure was used to estimate the regression equations for treatment and control groups while accounting for the correlation between repeated measurements made on each rabbit (22). The models for PET outcomes were weighted by the number of slices analyzed in each animal. To test for differences among the treatment and control groups estimated by linear regression equations, an interaction term between treatment group and time was included in each mixed model along with both main effects. If the interaction term was found to be statistically significant, then pairwise comparisons of treatment group slopes were estimated. All statistical analyses were performed using Proc Mixed in SAS software version 9.2 (SAS Institute, Cary, North Carolina). Average macrophage density per animal was correlated with average animal SUVmax or AUC1min of the corresponding animal. Neovessels were counted on a single section, were matched to the DCE-MRI imaged slice, and were correlated with AUC1min or average SUVmax of the corresponding animal. The AUC1min and average SUVmax from corresponding animals were correlated. All correlations were measured using Pearson testing after normality of the data was established. All histologic data, animal characteristics data, and nonlinear regression imaging data were analyzed with a 2-tailed Student t test.
Retrospective power was computed using SAS Proc Mixed as previously described (23).
Study design, animal characteristics, and serum analysis
The study design is depicted in Figure 1. After 4 months of a high-cholesterol diet, significantly elevated total cholesterol levels were observed (Table 1). There were no significant baseline differences in total cholesterol between the control and pioglitazone-treated groups before treatment initiation (Table 1). There were no statistically significant differences in fasting serum glucose or insulin levels and in nonfasting triglycerides in either group throughout the study. Both groups had nearly identical weights throughout the study (Table 1). Fast protein liquid chromatography at baseline and after 3 months revealed no significant differences in lipoprotein profiles after pioglitazone treatment (Online Appendix Figure 1).
SUV correlates with tissue macrophage content and AUC correlates with neovascularization
We correlated the SUVmax and AUC1min against macrophage density and neovessel counts, respectively, and found a strong positive correlation (r2 = 0.86 and p < 0.0001, and r2 = 0.66 and p = 0.004, respectively) (Figs. 2A and 2B), consistent with the results of previous studies (17,24). No strong correlation was found between SUVmax and neovessel count and between AUC1min and macrophage density (r2 = 0.09 and p = 0.37, and r2 = 0.19 and p = 0.18, respectively) (Figs. 2C and 2D). The SUVmax and AUC1min values did not correlate strongly (r2 = 0.17, p = 0.18) (Fig. 2E).
Pioglitazone arrests inflammation progression as assessed by 18F-FDG-PET/CT
On completion of baseline imaging, animals were divided into matched groups, 1 control group and 1 treatment group, with both groups displaying similar baseline SUVmax values (0.64 ± 0.05 and 0.62 ± 0.12, p = 0.70) (Figs. 3A and 3C and bar graph). The SUV versus time slope for the control group is positive, therefore indicating an increase in SUVmax (p = 0.10) and SUVmean (p = 0.02) over time. However, the SUV versus time slope for the treatment group was not statistically different from 0 (SUVmax: p = 0.70, SUVmean: p = 0.92), indicating no significant progression in plaque inflammation. Comparison of the 2 slopes indicated a trend toward significance (SUVmax: p = 0.15, SUVmean: p = 0.1). Comparison of SUV values between the 2 groups at 1 and 3 months showed significant differences (p < 0.05) for both SUVmax and SUVmean. Nonatherosclerotic aged-matched animals on chow diet show minimal SUV values.
Pioglitazone decreases neovascularization as assessed by DCE-MRI
In analogy to the 18F-FDG-PET/CT experiments, we applied DCE-MRI to assess inflammation and neovascularization in atherosclerotic plaques (Fig. 4). The AUC versus time slope for the control group was not statistically different from 0 for AUC measures (AUC1min: p = 0.33, AUC2min: p = 0.18). The AUC versus time slope for the treatment group was negative and significantly different from 0 for all AUC measures (p < 0.01). A trend toward significance was found when comparing the control and treatment groups slopes (AUC1min: p = 0.06, AUC2min: p = 0.09). There was a significant difference between baseline and 3-month AUC values in the treatment group (AUC1min: p = 0.01).
Pioglitazone has no effect on vessel wall area as assessed by multicontrast-MRI
Multicontrast-MRI showed no significant changes in aortic wall area in either group during the study (Fig. 5).
Pioglitazone decreases plaque macrophages and oxidized phospholipids
Macrophage staining at the end of the imaging study revealed a significant decrease in macrophage density within the pioglitazone group compared with the control group (0.25 ± 0.05 arbitrary unit [AU] and 0.32 ± 0.02 AU, respectively, p = 0.04) (Figs. 6 and 7)⇓⇓. A trend toward lower apolipoprotein B density was detected in the pioglitazone group compared with the control group (0.28 ± 0.03 AU and 0.34 ± 0.07 AU, respectively, p = 0.12) (Figs. 6 and 7). Pioglitazone treatment resulted in a 41% reduction in the oxidized phospholipid density compared with the control group (0.086 ± 0.026 AU and 0.146 ± 0.045 AU, respectively, p = 0.03) (Figs. 5 and 6). Average smooth muscle actin density was identical between groups (0.24 ± 0.075 AU and 0.24 ± 0.078 AU, respectively, p = 0.88) (Figs. 5 and 6). A trend toward lower neovessel content was observed in the pioglitazone-treated group (p = 0.12) (Fig. 6).
Power analysis of 18F-FDG-PET/CT and DCE-MRI
18F-FDG-PET/CT determined SUVmax values had 30% power to detect a difference between individual group slopes, 43% power to detect a group effect, and 15% power to detect a time effect. The SUVmean values had 40% power to detect a difference between individual group slopes, 37% power to detect a group effect, and 45% power to detect a time effect.
DCE-MRI determined AUC1min values had 46% power to detect a difference between individual group slopes, 14% power to detect a group effect, and 87% power to detect a time effect. The AUC2min values had 39% power to detect a difference between individual group slopes, 15% power to detect a group effect, and 92% power to detect a time effect.
In this study, we showed significant modulation in inflammation in aortic plaques of atherosclerotic rabbits as measured noninvasively by 18F-FDG-PET/CT and DCE-MRI on treatment with pioglitazone. The imaging results are confirmed by the histologic findings, which show a strong correlation of macrophage density and neovessel content with imaging parameters. In addition, this study provides insight to the diagnostic strengths of both FDG-PET and DCE-MRI while tracking antiatherosclerotic therapeutic intervention.
Previous studies have tested the potential of FDG-PET to track therapeutic intervention. Tahara et al. (25) investigated the effect of simvastatin on plaque inflammation by FDG-PET in the thoracic aorta of human subjects, whereas Ogawa et al. (26) investigated the effect of probucol on the aortic plaques of Watanabe heritable hyperlipidemic rabbits. In contrast with the work presented here, both of these studies showed no significant increase in FDG uptake in the control groups, whereas regression of inflammation was demonstrated by a significant decrease of FDG uptake over time in the treatment groups. This behavior is different from what we reported in our study, where we showed progression of inflammation in the control group and no change in FDG uptake over time in the treatment group. However, some significant differences between this study and the ones described above must be noted. For example, the dietary differences (dietary restrictions for patients and chow diet for rabbits as opposed to hyperlipidemic diet in this study with or without drug, depending on the group) and the different rabbit model used (Watanabe heritable hyperlipidemic rabbits versus New Zealand white rabbits). The only study in the literature investigating the use of DCE-MRI to track therapeutic intervention in atherosclerosis shows results comparable with those presented here, although the time frame of the 2 studies is considerably different (48 h vs. 3 months of treatment) (27).
In addition to showing the ability of FDG-PET and DCE-MRI to track therapeutic intervention with pioglitazone, this study sheds light on the possible complementary role in future clinical applications of the 2 methods. For example, FDG-PET showed an increase in plaque inflammation in the control group over time, whereas DCE-MRI measures did not change. Given the established correlation between FDG-PET and DCE-MRI with macrophages and neovessels, respectively, this may reflect a difference in the change of these histologic variables over time during plaque progression and may provide insight to the natural progression of aortic atherosclerosis in this specific rabbit model. More natural progression studies are needed to clarify this behavior and to plan future pre-clinical drug trials. The difference in FDG-PET and DCE-MRI measures in the treatment groups (impaired progression vs. regression) may reflect either the different sensitivity of the techniques or a different impact of pioglitazone on the related histologic variables (macrophages and neovessels, respectively). In this last case, the relationship between the 2 imaging techniques may be different for every new drug study planned with other therapeutic interventions.
In summary, we demonstrated the effectiveness of the PPARγ-agonist pioglitazone in promoting modulation of vascular inflammation in a rabbit model of atherosclerosis as evaluated by noninvasive imaging techniques and validated by histologic analysis. The 18F-FDG-PET/CT and DCE-MRI analyses were used in this case as surrogate imaging markers of vascular inflammation. Both techniques already are available clinically, which stresses the translational aspect of this study. Dedicated prospective outcome studies are underway (HRP Initiative scheduled report in 2011 ) that will clarify the role of plaque inflammation imaging in patient risk stratification. Furthermore, novel antiatherosclerotic therapies are being developed and their role in clinical practice will need to be established. In this sense, the imaging tools described herein may become potent armaments for improved drug development and their translation into clinical practice. Dedicated future clinical studies in nondiabetic patient populations will be needed to evaluate pioglitazone's effect in atherosclerotic disease.
The authors thank Ash Rafique, RT(N) CNMT, BS, and Suzanna Zata, CNMT, for their support with the PET acquisitions.
For an expanded Methods section, please see the online version of this article.
Pioglitazone Modulates Vascular Inflammation in Atherosclerotic Rabbits Monitored Noninvasively with 18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography and Black Blood Dynamic Contrast-Enhanced Magnetic Resonance Imaging
This investigation was supported in part by an investigator-initiated grant from Takeda Pharmaceutical (to Dr. Vucic); the Fondation Leducq (to Dr. Tsimikas); grants R01 HL071021 and R01 HL078667 from the National Heart, Lung and Blood Institute, National Institutes of Health (to Dr. Fayad); and grant EB009638 from the NBIB (National Institute of Biomedical Imaging and Bioengineering), National Institutes of Health (to Dr. Fayad). Dr. Tsimikas holds patents on oxidation-specific antibodies through the University of California San Diego. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Mr. Dickson and Dr. Calcagno contributed equally to this work. H. William Strauss, MD, served as Guest Editor for this article.
- Abbreviations And Acronyms
- arbitrary unit
- area under the curve
- AUC1 min
- area under the curve quantified at 1 minute
- AUC2 min
- area under the curve quantified at 2 minutes
- computed tomography
- dynamic contrast-enhanced
- magnetic resonance imaging
- positron emission tomography
- peroxisome proliferator-activated receptor gamma
- standard uptake value
- maximal standard uptake value
- mean standard uptake value
- Received March 4, 2011.
- Revision received April 22, 2011.
- Accepted April 22, 2011.
- American College of Cardiology Foundation
- Libby P.,
- Ridker P.M.,
- Hansson G.K.
- Fuster V.,
- Moreno P.R.,
- Fayad Z.A.,
- Corti R.,
- Badimon J.J.
- Natarajan P.,
- Ray K.K.,
- Cannon C.P.
- deGoma E.M.,
- deGoma R.L.,
- Rader D.J.
- Marx N.,
- Mach F.,
- Sauty A.,
- et al.
- Takata Y.,
- Kitami Y.,
- Yang Z.-H.,
- Nakamura M.,
- Okura T.,
- Hiwada K.
- Collins A.R.,
- Meehan W.P.,
- Kintscher U.,
- et al.
- ↵Nissen SE, Wolski K. Rosiglitazone revisited: an updated meta-analysis of risk for myocardial infarction and cardiovascular mortality. Arch Intern Med;170:1191–201.
- Dormandy J.A.,
- Charbonnel B.,
- Eckland D.J.A.,
- et al.
- Rudd J.H.,
- Hyafil F.,
- Fayad Z.A.
- Chang K.,
- Francis S.A.,
- Aikawa E.,
- et al.
- Tawakol A.,
- Migrino R.Q.,
- Bashian G.G.,
- et al.
- Rudd J.H.F.,
- Warburton E.A.,
- Fryer T.D.,
- et al.
- Calcagno C.,
- Cornily J.C.,
- Hyafil F.,
- et al.
- Rudd J.H.,
- Myers K.S.,
- Bansilal S.,
- et al.
- Hyafil F.,
- Cornily J.C.,
- Rudd J.H.,
- Machac J.,
- Feldman L.J.,
- Fayad Z.A.
- Joner M.,
- Farb A.,
- Cheng Q.,
- et al.
- McVicar J.P.,
- Kunitake S.T.,
- Hamilton R.L.,
- Kane J.P.
- Littell Ramon C.,
- Miliken G.A.,
- Stroup W.W.,
- Wolfinger R.D.
- Stroup W.W.
- Tahara N.,
- Kai H.,
- Ishibashi M.,
- et al.
- Ogawa M.,
- Magata Y.,
- Kato T.,
- et al.
- HRP initiative