Atherosclerosis imaging in statin intervention trials

Cardiovascular diseases are the principal causes of mortality in middle-aged and older people worldwide. Coronary heart disease (CHD) is the most common cardiovascular disease, and atherosclerosis is considered its most important cause. Epidemiological, clinical, and experimental evidence suggests that high serum cholesterol is associated with atherosclerosis. 1 Several large randomized controlled studies of statins (drugs that inhibit 3-hydroxy-3-methylglutaryl-coenzyme A, or HMG-CoA) have demonstrated a clear reduction in the incidence of coronary events, in patients either with or without previous CHD. 2–5 Because atherosclerosis progresses over decades, intervention trials require long-term follow-up and a large number of participants. To assess modifiers of atherosclerotic disease progression such as statins or lifestyle, surrogate markers are often needed to investigate determinants of atherosclerosis. Validated surrogate markers enable the assessment of promising new drugs in a relatively short period of time, thus avoiding the need to await the outcome of trials driven by clinical events. If more sensitive and more specific results are acquired, the surrogate markers should be easy to evaluate, preferably by non-invasive means. There is therefore much current interest in the non-invasive evaluation of atherosclerosis. 6–22 In this review, we discuss current perspectives on the non-invasive evaluation of human atherosclerosis, particularly focusing on intervention trials using statins and the use of atherosclerosis imaging including carotid intima-media thickness (IMT), as measured by B-mode ultrasound and magnetic resonance imaging (MRI). IMT has been used in other intervention trials, using anti-hypertensive drugs, anti-diabetic therapies, and other hypolipidaemic therapies, but to our knowledge, MRI has not yet been used in such trials. 23–32 A significant correlation between carotid IMT regression and low-density lipoprotein-cholesterol (LDL-C) reduction due to statin use was observed in many trials with large numbers of subjects. We therefore focus particularly on intervention trials using statins. We initially detected candidate studies using a PubMed search for ‘noninvasive evaluation’ and ‘atherosclerosis’; we then identified studies using the ankle-brachial blood pressure index (ABI), pulse-wave velocity (PWV), IMT, MRI, electron beam computed tomography (EBCT), and multi-slice CT (MSCT). Lastly, we obtained data using a PubMed search for each non-invasive atherosclerosis evaluation method and ‘statins or HMG-CoA reductase inhibitor’, and by reviewing reference lists. Simple and inexpensive methods include assessment of the lower extremity arteries above the ankles. A reduced ratio of the ankle-brachial blood pressure index (ABI) indicates atherosclerosis of the lower extremity arteries, and ABI values of 0.9 or less have been reported to be associated with an increased risk of cardiovascular disease. 33 Peripheral arterial disease (PAD) is considered to be a manifestation of systemic atherosclerosis. 34 In the PAD Awareness, Risk, and Treatment: New Resources for Survival (PARTNERS) program, a total of 6979 patients aged ⩾70 years or aged 50–69 years with history of smoking or diabetes, were evaluated by history and measurement of ABI. 35 Unexpectedly, the prevalence of PAD was high, but physician awareness of the PAD diagnosis was relatively low. Atherosclerosis risk factors were very prevalent in PAD patients, suggesting that ABI might be effective for the detection of pre-clinical atherosclerosis. However, in a single-blind randomized prospective study with 85 hyperlipidaemic, hypertensive patients treated with statins for one year, statins did not ameliorate ABI data, irrespective of whether serum cholesterol levels were significantly reduced. 36 In the other pilot study, 10 normocholesterolaemic and 10 hypercholesterolaemic patients with PAD were treated with simvastatin 40 mg/day for 3 months. Significant reductions in levels of serum total cholesterol and triglyceride levels, and elongations of pain-free and total walking distance were observed in both groups. However, significant changes in ABI were not seen. 37 These reports suggest that ABI is suitable for the detection of atherosclerosis, but not enough sensitive to follow the regression or progression of atherosclerosis in intervention trials. Aortic stiffness can be assessed by measuring pulse wave velocity (PWV). The measurement of brachial-ankle PWV (baPWV) is simple and applicable to general population studies. It was measured in a study with 472 consecutive subjects who subsequently underwent coronary angiography for diagnosis of CHD. 38 The prevalence of CHD in the lowest baPWV quartile was lower than in the other three quartiles, and on multivariate logistic regression analysis, being in the lowest baPWV quartile was a significant independent variable for the absence of CHD, suggesting that low baPWV may be an independent marker of low CHD risk among subjects whose risk of CHD otherwise appears high. In a study of 22 hyperlipidaemic patients with type 2 diabetes given atorvastatin (10 mg/day) for 6 months, PWV decreased (by ∼5%) without significant changes in blood pressure, heart rate or ABI. 39 In 59 hyperlipidaemic patients who received pravastatin for 6 months, significant decreases in PWV were found in the group with a >15% reduction in total cholesterol. 40 In a double-blind randomized study (atorvastatin vs. placebo) with 23 hypertensive and hyperlipidaemic patients, after 12 weeks of treatment, atorvastatin administration significantly lowered plasma low-density lipoprotein cholesterol (LDL-C) levels, and increased PWV by 8%; there was a significant negative correlation between changes in LDL-C and PWV. 41 The results of this study differ markedly from those of the other two. It may be that at the early stage phase of hypolipidaemic therapy (but not later) PWV increases transiently as a consequence of the significant reduction of lipid vascular content. PWV has a potential as a new marker of cardiovascular risk, and may be an indicator of either atherosclerotic cardiovascular risk or severity of atherosclerotic vascular damage. It may thus be useful for screening the general population in terms of evaluating atherosclerotic vascular lesions; however, it has insufficient discriminatory power as a marker of atherosclerosis. 42 It also has some problems in the context of intervention trials, such as insufficient change for the degree of atherosclerosis progression, or changes in different manner in early or later stage during the lipid-lowering therapy, which make results difficult to interpret. 40,41 There is a wide range of surrogate markers for atherosclerosis severity and progression. Currently, the best-established method is based on carotid intima-media thickness (IMT), as measured by B-mode ultrasound. In large observational studies, carotid IMT correlated with levels of cardiovascular risk, and was an independent risk factor for cardiovascular events. 43–46 It has thus been considered a surrogate marker for clinical end-points of cardiovascular diseases. Serial measurements of carotid IMT have been used in over 20 longitudinal studies to explore the effects of lipid-lowering therapies using statins ( Table 1 ). Most showed a regression or slowing progression of carotid IMT with active treatment. Further, a significant correlation between carotid IMT regression and LDL-C reduction was observed in most studies. In a systematic review and meta-analysis of trials testing statins, carotid IMT progression was strongly correlated with reduction in serum LDL-C levels. 69 Clinical trials showing the effects of lipid-lowering therapy on carotid intima-media thickness ACAPS, the Asymptomatic Carotid Artery Progression Study; REGRESS, The Regression Growth Evaluation Statin Study; PLAC-II, Pravastatin, Lipids, and Atherosclerosis in the Carotid Arteries-II; KAPS, Kupio Atherosclerosis Prevention Study; MARS, The Monitored Atherosclerosis Regression Study; CAIUS, The Carotid Atherosclerosis Italian Ultrasound Study; LIPID, The Long-term Intervention with Pravastatin in Ischemic Disease; ASAP, The Atorvastatin vs. Simvastatin on Atherosclerosis Progression Study; BCAPS, β-Blocker Cholesterol-Lowering Asymptomatic Plaque Study; ARBITER, Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol; FAST, Fukuoka Atherosclerosis Trial; CHD, coronary heart disease; FH, familial hypercholesterolemia; NCEP II, National Cholesterol Education Program Panel II; UAE, urinary albumin excretion; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; IMT, intima-media thickness; NC, no significant change. Clinical trials showing the effects of lipid-lowering therapy on carotid intima-media thickness ACAPS, the Asymptomatic Carotid Artery Progression Study; REGRESS, The Regression Growth Evaluation Statin Study; PLAC-II, Pravastatin, Lipids, and Atherosclerosis in the Carotid Arteries-II; KAPS, Kupio Atherosclerosis Prevention Study; MARS, The Monitored Atherosclerosis Regression Study; CAIUS, The Carotid Atherosclerosis Italian Ultrasound Study; LIPID, The Long-term Intervention with Pravastatin in Ischemic Disease; ASAP, The Atorvastatin vs. Simvastatin on Atherosclerosis Progression Study; BCAPS, β-Blocker Cholesterol-Lowering Asymptomatic Plaque Study; ARBITER, Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol; FAST, Fukuoka Atherosclerosis Trial; CHD, coronary heart disease; FH, familial hypercholesterolemia; NCEP II, National Cholesterol Education Program Panel II; UAE, urinary albumin excretion; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; IMT, intima-media thickness; NC, no significant change. Measurements of carotid IMT can accurately describe the process of arterial wall changes due to atherosclerosis, can identify low-risk patients in high-risk populations for CHD, and can provide information on the efficacy of lipid-lowering. Carotid IMT measurement thus appears to be a validated surrogate endpoint for atherosclerosis or the risk of CHD. However, there was clinically important inter-observer variability in the assessment of IMT, 70 and conventional ultrasonic assessment of IMT does not differentiate between intima and media (and thus not between hypertension-related intima hypertrophy and atherosclerosis-induced intima changes). 71,72 Therefore, this method cannot follow changes in plaque composition in intervention trials. Further, few studies have conclusively linked improvement in carotid IMT to reduction in cardiovascular events, and this needs to be elucidated in future studies. Recently, magnetic resonance imaging (MRI) has become a useful and promising tool for the non-invasive evaluation of atherosclerosis. To determine the effects of sustained normocholesterolaemia on advanced atherosclerosis and whether changes in the size and composition of plaque can be detected by MRI, Trogan and colleagues transplanted aortic arch segments containing advanced atherosclerotic lesions from apolipoprotein E (apo E)-deficient mice into wild-type and apoE-deficient mice. 73 Mice underwent serial MRI at 3, 5, 7 and 9 weeks after transplantation. Compared with baseline, normalization of dyslipidaemia resulted in a monotonic reduction of lesion burden, whereas continued dyslipidaemia in apoE-deficient mice increased the lesion burden. MRI and histological measurements were closely correlated (correlation coefficient 0.937). In humans, MRI findings have been also extensively validated against pathology in ex vivo studies of carotid, aortic and coronary artery specimens obtained at autopsy, and MRI finding of carotid arteries of patients referred for endarterectomy has shown a significant correlation with pathology. 74 Recently, Takaya et al . suggested that contrast-enhanced MRI might provide an important tool for clinical trials assessing lipid-rich necrotic core of atherosclerotic plaques progression and regression. 75 MRI differentiates plaque components on the basis of biophysical and biochemical parameters such as chemical composition and concentration, water content, physical state, molecular motion and diffusion. 22 It can thus be used to discriminate lipid core, fibrosis, calcification, and intra-plaque haemorrhage deposits. 74 The ability to detect changes in plaque composition by MRI could allow insights in the biological effects of interventions. Corti et al . demonstrated atherosclerosis regression with simvastatin using MRI ( Table 2 ). 76 Significant reductions in vessel wall thickness (VWT) and vessel wall area (VWA) (10% and 11% for thoracic aortic, and 8% and 11% for carotid arterial plaques, respectively) were observed at 12 months, whereas lumen area (LA) did not change significantly at this time point. Further decreases in VWT and VWA were seen after 24 months, and finally, a significant increase in LA was also found ( Table 2 ). In the early stages of atherogenesis, lipid deposition was associated with a positive outer remodelling of the arterial wall (compensatory enlargement), whereas at later stages, continuous accumulations of lipid and cell accumulation begin to compromise the lumen. 77 Lipid-lowering thus appears to influence remodelling of the vessel wall before changing the luminal area. Such observations have been facilitated by MRI. Effects of simvastatin on atherosclerotic plaques detected by magnetic resonance imaging LDL-C, low density-lipoprotein cholesterol; MRI, magnetic resonance imaging; VWA, vessel wall area; VWT, vessel wall thickness; LA, lumen area; APA, atherosclerotic plaque area; APV, atherosclerotic plaque volume; LV, lumen volume; NC, no significant change. Effects of simvastatin on atherosclerotic plaques detected by magnetic resonance imaging LDL-C, low density-lipoprotein cholesterol; MRI, magnetic resonance imaging; VWA, vessel wall area; VWT, vessel wall thickness; LA, lumen area; APA, atherosclerotic plaque area; APV, atherosclerotic plaque volume; LV, lumen volume; NC, no significant change. Lima et al. found similar reductions in atherosclerotic plaque volume and area in response to lipid-lowering by simvastatin ( Table 2 ). 78 They also demonstrated a significant correlation between atherosclerotic plaque regression and LDL-C reduction. Pathological studies have documented statin-induced changes in plaque composition 3 months after therapy initiation. 79 Conversely, results from clinical trials using carotid ultrasound 80 or intravascular ultrasound (IVUS) 81 suggest that plaque regression may be detectable at least 1 year after starting statin therapy. However, Lima et al . demonstrated that atherosclerotic plaque regression, as well as reversal of remodelling, could be detected accurately by MRI just 6 months after statin therapy initiation. In a recent double-blind study, hypercholesterolaemic patients were randomized to conventional (20 mg) vs. aggressive (80 mg) simvastatin groups, respectively, and their atherosclerotic plaques were detected and sequentially followed by MRI at 0 and 12 months ( Table 3 ). LDL-C levels in the conventional and aggressive simvastatin groups decreased by 36% vs. 46%, respectively. 82 In aortic lesions, VWA decreased by 10% and 15% at 12 and 24 months, respectively, VWT decreased by 9% and 15% at 12 and 24 months, respectively, and LA increased by 5% and 6% at 18 and 24 months, respectively. Similar changes were observed for carotid arterial lesions. However, there were no significant differences in these parameters for plaque between the conventional and aggressive therapeutic group. Patients whose LDL-C reached ⩽100 mg/dl showed a greater decrease in plaque size compared with those who did not achieve this goal. Patients with on-treatment LDL-C levels ⩽100 mg/dl also had a 17% reduction in aortic VWA, vs. a 12% reduction observed in the group with on-treatment LDL-C levels >100 mg/dl. This suggests that the degree of LDL-C reduction may be a good predictor of plaque regression, as do the results of the IVUS-based Reversal of Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) trial 83 and the Pravastatin or Atorvastatin Evaluation and Infection Therapy (PROVE-IT) trial. 84 Effects of aggressive versus conventional lipid-lowering therapy using statins on atherosclerosis, as evaluated by magnetic resonance imaging LDL-C, low density-lipoprotein cholesterol; MRI, magnetic resonance imaging; VWA, vessel wall area; VWT, vessel wall thickness; LA, lumen area. Effects of aggressive versus conventional lipid-lowering therapy using statins on atherosclerosis, as evaluated by magnetic resonance imaging LDL-C, low density-lipoprotein cholesterol; MRI, magnetic resonance imaging; VWA, vessel wall area; VWT, vessel wall thickness; LA, lumen area. Yonemura et al . investigated the effects of 20 mg vs. 5 mg atorvastatin on thoracic and abdominal aortic plaques in 40 hypercholesterolaemic patients, using MRI ( Table 3 ). 85 The 20 mg dose reduced VWT and VWA of thoracic aortic plaques by 12 and 18%, respectively, and increased LA by 5%, whereas 5 mg did not. LDL-C reduction and degree of plaque regression in thoracic aorta were significantly correlated, supporting the idea that degree of LDL-C reduction is the most important factor for plaque regression in thoracic aorta. In abdominal aortic plaques, even 20 mg could not reduce either the VWA or VWT, and surprisingly, 5 mg atorvastatin actually increased both VWA (+5%) and VWT (+12%). This difference between thoracic and abdominal aortic plaques may be explained by the high prevalence of calcified and fibrous plaques in abdominal aorta, 85,86 for which statins may be less effective. In addition, the abdominal aorta may be highly susceptible to plaque formation associated with aging and hypertension, irrespective of LDL-C levels. However, the underlying mechanisms, for this difference merit further investigation. Taken together, these studies using MRI present exciting and promising results, and MRI may make great advances in the non-invasive evaluation of atherosclerosis in the intervention trials. However, further prospective studies are still needed to ascertain the predictive value of asymptomatic atherosclerosis detected by MRI for future cardiovascular events. MRI also has several safety issues, including the effects of high magnetic fields and radiofrequency pulses on the human body, and on implanted devices such as pacemakers, the side-effects of contrast agents, toxicity during pregnancy, claustrophobia, and hearing loss. 87 Calcification of coronary arteries is a specific marker of atherosclerosis, and is a predictive factor of subsequent cardiac events. 88,89 There has been a progressive increase in the use of EBCT scanners to identify and quantify the amount of calcified coronary arteries. The calcium volume score (CVS) seems to improve CHD risk prediction. 90,91 Callister et al . studied changes in coronary plaque volume with lipid-lowering therapy using EBCT in asymptomatic, hyperlipidaemic patients. 92 In this study, the CVS of statin-treated patients was compared with that of patients with no statins. CVS was significantly reduced by 7% in patients who had achieved a LDL-C level < 120 mg/dl, which indicates regression of atherosclerosis in this group. In patients who failed to reach this target point, an increase in CVS was observed. Another EBCT study, the Beyond-Endorsed Lipid-Lowering with EBCT Scanning (BELLES) Trial is currently ongoing, and will compare the effects of atorvastatin and pravastatin on atherosclerosis progression in asymptomatic postmenopausal women. 93 MSCT has recently attracted much attention. This technique provides non-invasive visual evaluation of the coronary arteries, and comparisons with invasive coronary angiography have shown good correlations. 94–96 However, to improve this technology, optimization of patient preparation, data acquisition and reconstruction is required, and standardization of data analysis is also needed. EBCT and MSCT are promising, non-invasive evaluation techniques for assessing the effect of therapy on coronary atherosclerosis. However, at present, results from intervention trials are very limited and their applicability in trials needs to be evaluated in more detail. Increased availability of EBCT and MSCT may further boost the use of these novel imaging modalities in atherosclerosis research. ABI and PWV are simple and well-suited to general population studies, and able to detect of atherosclerosis to some extent, but are not suitable for following the regression or progression of atherosclerosis in intervention trials. IMT measurement is the best-established, and longitudinal studies have presented the regression of atherosclerosis by lipid-lowering with statins. MRI has recently been used to evaluate atherosclerosis, and several intervention trials with statins suggest that it can detect atherosclerotic plaque regression accurately. A significant correlation between LDL-C reduction and atherosclerosis regression, supported by studies defined invasively by angiography or IVUS, has been demonstrated by both IMT and MRI. Therefore, both these imaging methods can be seen as validated surrogate endpoints for atherosclerosis or the risk of CHD. However, because few studies have conclusively linked improvements in atherosclerosis as detected by these methods to reductions in cardiovascular events, predictive values of cardiovascular events estimated by these methods need future study. The increased availability of EBCT and MSCT may further boost the use of these novel imaging modalities in atherosclerosis intervention trials.