(Circulation. 2006;113:2826-2834.)
© 2006 American Heart Association, Inc.
Interventional Cardiology |
From the Departments of Cardiovascular Medicine (P.S., E.M.T., K.W., I.S., S.J.N., W.A.M., A.L., T.C., T.C., S.E.N.), Radiology (P.S.), and Quantitative Health Sciences (C.A.-H., C.W., S.L.), The Cleveland Clinic Foundation, Cleveland, Ohio.
Correspondence to Paul Schoenhagen, MD, FAHA, Departments of Cardiovascular Medicine and Radiology, Cardiovascular Imaging, The Cleveland Clinic Foundation, HB-6, 9500 Euclid Ave, Cleveland OH 44195. E-mail schoenp1{at}ccf.org
Received November 22, 2005; revision received April 8, 2006; accepted April 14, 2006.
| Abstract |
|---|
|
|
|---|
Methods and Results Serial intravascular ultrasound (IVUS) data from the Reversal of Atherosclerosis with Aggressive Lipid Lowering Therapy (REVERSAL) trial were obtained during 18 months of intensive versus moderate lipid-lowering therapy. In a subgroup of 210 patients, focal coronary lesions with mild luminal narrowing were identified. Lumen area, external elastic membrane (EEM) area, and plaque area were determined at the lesion and proximal reference sites at baseline and during follow-up. The remodeling ratio (RR) was calculated by dividing the lesion EEM area by the reference EEM area. The relationship between the change in remodeling, change in plaque area, lipid profile, and inflammatory markers was examined. At the lesion site, a progression in plaque area (8.9±25.7%) and a decrease in the RR (3.0±11.2%) occurred during follow-up. In multivariable analyses, the percentage change in plaque area (P<0.0001), baseline RR (P<0.0001), baseline lesion lumen area (0.019), logarithmic value of the change in high-sensitivity C-reactive protein (P=0.027), and hypertension at baseline (P=0.014) showed a significant, direct relation with the RR at follow-up. Lesion location in the right coronary artery (P=0.006), percentage change in triglyceride levels (P=0.049), and age (P=0.037) demonstrated a significant, inverse relation with the RR at follow-up. Changes in LDL cholesterol, HDL cholesterol, and treatment group demonstrated no significant associations.
Conclusions Constrictive remodeling of the arterial wall was observed during plaque-stabilizing therapy with statin medications and appears related to their antiinflammatory effects.
Key Words: arteriosclerosis imaging inflammation intravascular ultrasonography plaque remodeling statins
| Introduction |
|---|
|
|
|---|
Editorial p 2787
Clinical Perspective p 2834
Serial IVUS data from recent multicenter trials provide the opportunity to examine the changes in vessel wall remodeling during plaque-stabilizing, lipid-lowering treatment. We hypothesized that plaque regression would be associated with constrictive wall remodeling and intended to examine the relationship to serum markers of lipid metabolism and inflammation. Consequently, in a prospectively designed substudy of the Reversal of Atherosclerosis with Aggressive Lipid Lowering Therapy (REVERSAL) population, we identified focal, mildly stenotic atherosclerotic lesions and observed the serial remodeling response during lipid-lowering therapy. In multivariable analyses, we examined the influence of changes in plaque burden, lipid profile, CRP, and other cardiovascular risk factors on the remodeling response.
| Methods |
|---|
|
|
|---|
After a 2-week placebo run-in period, patients were randomized to either 80 mg atorvastatin (intensive treatment group) or 40 mg pravastatin (moderate treatment group) daily in a blinded fashion. During routine clinic visits the patients were examined, and laboratory studies were performed every 3 months for 18 months. After the 18-month treatment period, patients underwent a repeat cardiac catheterization and IVUS examination of the matched segments under identical conditions.
In the REVERSAL trial, 654 patients were randomized at 34 centers between June 1999 and September 2001. From these, a total of 502 patients completed the protocol, 249 in the pravastatin arm and 253 in the atorvastatin arm. We identified focal atherosclerotic lesions with mild luminal narrowing in 53% (266) of the patients. Of these, a total of 210 (42% of the REVERSAL population) had complete data for a focal lesion and a reference site matched at baseline and at follow-up. This subanalysis was approved by the institutional review board at our institution.
IVUS Analysis
Lesion Identification and Matching
Tapes were reanalyzed by an experienced investigator, and focal atherosclerotic lesions were identified. Lesion sites were defined as a focal site with <50% angiographic narrowing and visible plaque accumulation (plaque thickness >0.5 mm).1012 A normal or near-normal proximal reference segment (plaque thickness <0.3 mm) was identified within a distance of 10 mm from the lesion site, without intervening side branches. One lesion was included per patient. In the follow-up IVUS analysis, the same lesion was identified. Matching of sites between baseline and follow-up analyses was guided by side branches and characteristic lesion morphology. In the final data analysis, only lesions with an identifiable lesion and a proximal reference, matched at baseline and follow-up, were included. In all patients, baseline and follow-up IVUS was performed with the same IVUS catheter type and IVUS system. Therefore, correction with previously described equations was not performed.13,14
IVUS Analysis and Standard Measurements
For each lesion and reference site, planimetry of the lumen and external elastic membrane (EEM) areas was performed by an investigator blinded to the treatment group. With the use of National Institutes of Health (NIH) Image 1.62 software (NIH public domain software; NIH, Bethesda, Md), calibration was performed by measuring 1-mm grid marks encoded in the image. Manual planimetry was then used to trace the leading edges of the luminal and EEM borders. At both the lesion and proximal reference site, EEM area, lumen area, plaque area, and percentage plaque burden were determined in accordance with international standards.10,11
Remodeling Measurements
Arterial wall remodeling was assessed by comparing the outer vessel border at the lesion and reference sites (Figures 1 and 2
) The outer vessel border is defined by the EEM area.10,11 The remodeling ratio (RR) at baseline and follow-up was defined as the ratio of the EEM area at the lesion to that at the proximal reference site. Remodeling was analyzed as a continuous variable. In addition, 3 remodeling categories were defined9: (1) Expansive remodeling, describing expansion of the EEM at the lesion site, was defined as an RR >1.05; (2) the absence of remodeling was defined as an RR between 0.95 and 1.05; and (3) constrictive remodeling, describing the constriction of the EEM at the lesion site, was defined as an RR <0.95.
|
|
Serial IVUS Measurements
Percentage Change in RR
The percentage change in RR was calculated as the difference in RR between baseline and follow-up divided by the baseline RR.
Percentage Change in Plaque Area
The percentage change in plaque area was computed as the difference in plaque area between baseline and follow-up divided by the baseline plaque area.
Serial Measurements of Markers of Lipid Metabolism and Inflammation
The biochemical analysis was performed at a central laboratory (Medical Research Laboratory, Highland Heights, Ky). Total cholesterol, LDL-C, HDL-C, and triglycerides were analyzed with an enzymatic colorimetric assay (Hitachi 747, Tokyo, Japan). High-sensitivity (hs)-CRP was analyzed by an immunonephelometric assay (Behring nephelometer 100, Westwood, Marburg, Germany).
For normally distributed variables, including total cholesterol, LDL-C, HDL-C, and triglycerides, the percentage change during follow-up was calculated as the difference between baseline and follow-up values divided by the baseline value. Because hs-CRP values were not normally distributed, logarithmic transformation was performed. The change in hs-CRP between baseline and follow-up was described by the logarithm of the ratio of follow-up hs-CRP divided by baseline hs-CRP [change in hs-CRP=log(hs-CRPfollow-up/hs-CRPbaseline)].
Statistical Methods
Simple descriptive statistics were used to summarize the data. For categorical variables, this included frequencies and percentages. For continuous variables, this included the mean and standard deviation and, in the case of nonnormally distributed variables, median and interquartile range. In addition, cross-tabs and plots were generated.
Because the patients in this subgroup analysis were not randomized, baseline variables were tested for differences in treatment groups. Univariate tests of treatment group differences at follow-up were conducted after adjusting for baseline values.
The primary outcomes studied were the RR at follow-up and the categorization of that RR as <0.95, 0.95 to 1.05, and >1.05. Potential predictors included baseline IVUS characteristics, demographic information, traditional cardiovascular risk factors, and serum markers of lipid metabolism and inflammation. For serum markers of lipid metabolism, the percentage change from baseline to follow-up was used. Because CRP values were not normally distributed, the change in hs-CRP was analyzed as the natural logarithm of the ratio between follow-up and baseline values. Missing values were imputed from the mean value (no more than 3 values for each of 12 possible predictor variables).
For identification of the factors predicting remodeling, bootstrap analysis was performed with all baseline clinical, biochemical, and IVUS variables as well as the change in biochemical variables as possible predictors (see online-only Data Supplement). Based on this bootstrap analysis, variables were chosen for the initial multivariable analysis.15 The best final multivariable model included only those variables that significantly added to the model in the presence of the other factors. An ordinal logistic regression was used to determine whether the categorized follow-up RR was affected by the variables. Unless otherwise stated, statistical testing was conducted with 2-sided alternatives, with a type I error level of 0.05. SAS, version 8.2, was used to conduct all analyses (SAS Institute, Cary, NC).
The authors had full access to the data and take full responsibility for its integrity. All authors have read and agree to the manuscript as written.
| Results |
|---|
|
|
|---|
|
Laboratory Results
Table 2 summarizes laboratory values at baseline and follow-up and the changes at follow-up for the entire population and the treatment groups. As expected, significant changes in lipid levels and hs-CRP were found, with a more pronounced effect in the intensive treatment group.
|
IVUS Measurements at Baseline and Follow-Up
Table 3 summarizes IVUS measurements at baseline and follow-up for the overall group of patients and the treatment groups. In the overall group, there was a small but significant increase in both EEM and lumen area between baseline and follow-up. The baseline mean percentage plaque burden at the lesion and the reference site was 49.9±9.9% and 29.5±7.8%, respectively. It was unchanged at follow-up, measuring 49.8±10.1% and 29.8%±7.8% (P=0.46 and 0.47), respectively. Plaque area at the lesion site increased from 7.6±2.8 mm2 to 8.1±3.2 mm2, which represents a significant increase in percentage change in plaque area (8.9%±25.7, P<0.0001 for comparison of baseline to follow-up).
|
Remodeling at Baseline and Follow-Up
Table 3 also summarizes the RR at baseline and follow-up for the overall group and the treatment groups. In the overall group, the RR decreased from 1.07±0.15 to 1.03±0.14, which represents a significant decrease in the percentage change of the RR (3.0±11.2%, P<0.0001 for comparison of baseline to follow-up). Table 4 shows the frequency of remodeling categories at baseline and follow-up. There was a shift to constrictive remodeling categories during follow-up.
|
Influence of Treatment Group
There were no significant differences in IVUS measurements between the intensive and moderate treatment groups. In both groups, there was a small increase in plaque area between baseline and follow-up (7.9±25.8% and 9.9±25.7%, respectively, P=0.57). The change in RR was not significantly different in the intensive and moderate lipid-lowering arms (3.3±10.4% and 2.7±12.0%, respectively, P=0.68; Table 3).
Relation Between Changes in Remodeling, Plaque Area, and Other Variables
In multivariable analyses, the percentage change in plaque area (P<0.0001), baseline RR (P<0.0001), baseline lesion lumen area (0.019), logarithmic change in hs-CRP (P=0.027), and hypertension at baseline (P=0.014) showed a significant, direct relation with the RR at follow-up. Lesion location in the right coronary artery (P=0.006), percentage change in triglyceride levels (P=0.049), and age (P=0.037) demonstrated a significant, inverse relation with the RR at follow-up (Table 5). In the presence of the aforementioned variables, the change in HDL-C, the change in LDL-C, and treatment assignment did not add to the final model.
|
| Discussion |
|---|
|
|
|---|
Our results provide important new insights into the pathophysiology of focal vessel wall changes during plaque stabilization. A relationship of arterial wall remodeling with plaque stability/vulnerability has been observed in previous studies. In histological studies, the expansion in vessel size or expansive remodeling has been associated with an increase in inflammatory cells and proteolytic enzymes at the lesion site.6,7 In clinical IVUS studies, expansive wall remodeling has consistently been associated with plaque instability and unstable clinical presentation.8,9 In the present study, the shift to constrictive wall remodeling during lipid-lowering treatment is intriguing but consistent with previous IVUS reports. In studies describing changes in advanced lesions after percutaneous coronary intervention, constrictive wall remodeling has been associated with restenosis.16,17 Experimental studies with histological analysis after percutaneous intervention have further demonstrated that constrictive remodeling is associated with plaque compositional changes characterized by increased fibrosis.18 In native coronary lesions of nonhuman primates, histological studies have demonstrated lipid depletion and fibrosis during lipid-lowering therapy.1921
Based on these results, lipid depletion and fibrosis are considered to be related to the concept of plaque regression and stabilization.8,22,23 In clinical IVUS studies, the stabilizing effect of lipid-lowering therapy is primarily reflected in changes of plaque burden.2 However, standard gray-scale analysis in serial clinical IVUS studies has demonstrated a shift to a more fibrotic lesion morphology during lipid-lowering therapy,24 and quantitative analysis of vessel wall composition is emerging.25 Other studies have suggested that disease regression might in fact be associated with constrictive remodeling.2628 In analogy to the process of vessel wall expansion during plaque progression as described by Glagov et al,5 vessel constriction during regression has been termed "reverse remodeling."
However, the determinants of the remodeling response during lipid-lowering therapy are incompletely understood. Recent IVUS studies analyzing the determinants of plaque progression/regression during statin therapy demonstrated an independent effect of LDL and CRP reduction on plaque burden.3,4 Our results suggest that the determinants of remodeling are different. Particularly interesting is the lack of an independent relation to LDL-C and the positive relation to CRP. The lack of a relation with LDL-C may indicate that the beneficial effect of LDL lowering is primarily mediated through the reduction of plaque burden. In contrast, the independent, direct relation with CRP may demonstrate that the antiinflammatory, plaque-stabilizing effect of statin therapy is associated with structural changes in the vessel wall that are not reflected in the reduction of plaque burden. These results suggest that arterial remodeling is an independent factor in disease progression and stabilization with a close relationship to inflammation. It is an attractive hypothesis that the systemic marker CRP reflects focal inflammatory processes, including the effect of matrix metalloproteinases, affecting both remodeling and plaque stability.2931
The relation between baseline hypertension and expansive remodeling is intriguing in the context of the recently published Comparison of AMlodipine vs Enalapril to Limit Occurrences of Thrombosis (CAMELOT) study that described a relationship between blood pressure lowering and attenuated changes in plaque burden in normotensive patients.32 The relations between an increase in triglycerides, older age, and lesion location in the right coronary artery with constrictive remodeling are similar to those found in previous studies.27,33 The relation between larger baseline lumen area and expansive remodeling is interesting in the context of the original observations of Glagov et al that expansive remodeling is limited by increasing plaque area at the lesions site, leading to increasing luminal stenosis beyond a 40% cross-sectional plaque area.5 Table 4 shows complex changes in the remodeling categories. These subgroups are too small to determine the association with clinical variables, and further studies are necessary to understand these relationships.34 Lastly, we found no difference between the intensive and moderate statin treatment groups. However, even though the result is not significant, the shift toward constrictive remodeling was greater in the intensive group than in the moderate group. Larger studies may be necessary to show a significant difference. On the other hand, our results could suggest that the achieved extent of the statin effect (on-treatment LDL-C and HDL-C, lowering of CRP, reduction of plaque growth) is more important than the treatment assignment itself.
Limitations
To attain the low variability in plaque burden measurements required for serial pharmacological studies, the REVERSAL trial and other similar studies have used a volumetric analysis approach of entire coronary segments, thus avoiding matching of individual lesions. In contrast, the conventional methodology used for assessment of remodeling, which was also applied in this analysis, relies on matching of focal lesion sites. Matching of the lesion and the reference site and subsequent calculation of the RR are expected to be associated with increased measurement variability. We calculated the reproducibility for the remodeling calculation in a separate variability analysis. Both the interobserver and intraobserver variability was relatively poor, with the probability values for a paired t test of the difference in the change in RR approaching statistical significance (P=0.162 and 0.073, respectively).
Assessment of focal remodeling may also be confounded by disease and remodeling at the reference site.35 To minimize that possibility, we chose the reference site as one with no significant disease, defined as a maximal intimal thickness <0.3 mm.12 Despite its limitations, remodeling analysis of individual lesions may allow the identification of focal, pathophysiological processes that determine remodeling. In contrast, definitions of remodeling in entire vessel segments, describing the volumetric response of the EEM during lesion progression/regression, may allow us to understand the systemic response of the vessel wall during disease progression.10,2628 The differences between these approaches are incompletely understood and will need to be addressed in future studies.34
Other limitations are related to the patient population. The data were obtained for a subgroup of the REVERSAL population (210 of 654 patients) in whom focal lesions could be identified. The definition of focal lesion as described in the literature is to some degree arbitrary.1012 Although baseline characteristics were not different from those of the overall REVERSAL population, the smaller population raises questions about the generalizability of our findings. In particular, the results may differ in lesions with more diffuse plaque distribution, which would require a volumetric analysis.2628 The influence of CRP is expected to be more pronounced in patients with acute coronary syndromes, and such a population would therefore be best suited to evaluate the relationship between positive remodeling and inflammation. However, the CRP level in the REVERSAL subjects, who were initially referred for percutaneous intervention, identified a high-risk population according to current guidelines.36 Because the lipid inclusion criteria in the REVERSAL trial specified an LDL range that included only patients with hypercholesterolemia (125 to 210 mL/dL), our findings may not apply to vessel segments of normocholesterolemic subjects. In addition, narrowing the range of LDL may have affected the correlation with remodeling. Therefore, the lack of a relationship between the change in LDL and remodeling will need confirmation in unselected patient populations.
Conclusion
Analyzing serial IVUS data from the REVERSAL trial, we have demonstrated that constrictive arterial wall remodeling was the predominant response during plaque-stabilizing treatment with statin medications. In multivariable analyses, the percentage change in plaque area, baseline RR, baseline lesion lumen area, logarithmic change in hs-CRP, and hypertension at baseline showed a significant, direct relation with the RR at follow-up. Lesion location in the right coronary artery, percentage change in triglyceride levels, and age demonstrated a significant, inverse relation with the RR at follow-up. In the presence of these variables, changes in LDL-C and HDL-C were not found to have an independent relation.
These findings have important implications for the understanding of plaque stability and remodeling. Together with previous findings describing the association of expansive wall remodeling with lesion inflammation and unstable clinical presentation,8,9 our results are consistent with the hypothesis that constrictive remodeling is associated with plaque stabilization. Arterial wall remodeling is an independent factor in disease progression and stabilization and appears to be related to the inflammatory activity of atherosclerotic lesions.
| Acknowledgments |
|---|
This work was supported by an award from the Ohio Affiliate of the American Heart Association (to P.S.); NIH National Center for Research Resources, General Clinical Research Center grant MO1 RR-018390 (to P.S.); a Ralph Reader Research Fellowship, National Heart Foundation of Australia (to S.J.N.); and an educational grant from Pfizer (I.S.).
Disclosures
Dr Tuzcu has received support from Pfizer, Merck, and Takeda. Dr Nicholls has received support from Pfizer. Dr Nissen has received research support from Pfizer. The remaining authors report no conflicts.
| References |
|---|
|
|
|---|
12 months) follow-up intravascular ultrasound studies in left main coronary arteries. Am J Cardiol. 2004; 93: 11071113.[CrossRef][Medline]
[Order article via Infotrieve]
| Footnotes |
|---|
Related Article:
Circulation 2006 113: 2783.
This article has been cited by other articles:
![]() |
G. S. Mintz Diabetic Coronary Artery Disease: How Little We Know and How Little Intravascular Ultrasound Has Taught Us J. Am. Coll. Cardiol., July 22, 2008; 52(4): 263 - 265. [Full Text] [PDF] |
||||
![]() |
G. Pasterkamp and I. Hoefer The visualization of the remodelling paradox Eur. Heart J., July 2, 2008; 29(14): 1704 - 1705. [Full Text] [PDF] |
||||
![]() |
I Sipahi, E M Tuzcu, K-W Moon, S J Nicholls, P Schoenhagen, J Zhitnik, T D Crowe, S Kapadia, and S E Nissen Do the extent and direction of arterial remodelling predict subsequent progression of coronary atherosclerosis? A serial intravascular ultrasound study Heart, May 1, 2008; 94(5): 623 - 627. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Aoki, G. S. Mintz, N. J. Weissman, J. T. Mann, L. Cannon, J. Greenberg, E. Grube, A.R. Z. Masud, J. Koglin, L. Mandinov, et al. Chronic Arterial Responses to Overlapping Paclitaxel-Eluting Stents: Insights From Serial Intravascular Ultrasound Analyses in the TAXUS-V and -VI Trials J. Am. Coll. Cardiol. Intv., April 1, 2008; 1(2): 161 - 167. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. A. Rodriguez-Granillo, S. de Winter, N. Bruining, J. M.R. Ligthart, H. M. Garcia-Garcia, M. Valgimigli, P. J. de Feyter, and on behalf of the EUROPA/PERSPECTIVE Investigators Effect of perindopril on coronary remodelling: insights from a multicentre, randomized study Eur. Heart J., October 1, 2007; 28(19): 2326 - 2331. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Schoenhagen and I. Sipahi Arterial remodelling: an independent pathophysiological component of atherosclerotic disease progression and regression. Insights from serial pharmacological intervention trials Eur. Heart J., October 1, 2007; 28(19): 2299 - 2300. [Full Text] [PDF] |
||||
![]() |
L. O. Jensen, P. Thayssen, G. S. Mintz, M. Maeng, A. Junker, A. Galloe, E. H. Christiansen, S. K.S. Hoffmann, K. E. Pedersen, H. S. Hansen, et al. Intravascular ultrasound assessment of remodelling and reference segment plaque burden in type-2 diabetic patients Eur. Heart J., July 2, 2007; 28(14): 1759 - 1764. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. S. Chatzizisis, A. U. Coskun, M. Jonas, E. R. Edelman, C. L. Feldman, and P. H. Stone Role of Endothelial Shear Stress in the Natural History of Coronary Atherosclerosis and Vascular Remodeling: Molecular, Cellular, and Vascular Behavior J. Am. Coll. Cardiol., June 26, 2007; 49(25): 2379 - 2393. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Schoenhagen Plaque Temperature, Arterial Remodeling, and Inflammation: Understanding "Hot-Spots" in the Coronary Arteries J. Am. Coll. Cardiol., June 12, 2007; 49(23): 2272 - 2273. [Full Text] [PDF] |
||||
![]() |
K. Toutouzas, A. Synetos, E. Stefanadi, S. Vaina, V. Markou, M. Vavuranakis, E. Tsiamis, D. Tousoulis, and C. Stefanadis Correlation Between Morphologic Characteristics and Local Temperature Differences in Culprit Lesions of Patients With Symptomatic Coronary Artery Disease J. Am. Coll. Cardiol., June 12, 2007; 49(23): 2264 - 2271. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Sanz, P. R. Moreno, and V. Fuster The Year in Atherothrombosis J. Am. Coll. Cardiol., April 24, 2007; 49(16): 1740 - 1749. [Full Text] [PDF] |
||||
![]() |
L. W. Klein Atherosclerosis Regression, Vascular Remodeling, and Plaque Stabilization J. Am. Coll. Cardiol., November 8, 2006; (2006) j.jacc.2006.10.039v1. [Full Text] [PDF] |
||||
![]() |
A. E. Silver and J. A. Vita Shear Stress-Mediated Arterial Remodeling in Atherosclerosis: Too Much of a Good Thing? Circulation, June 20, 2006; 113(24): 2787 - 2789. [Full Text] [PDF] |
||||
| ||||||||||