Objective: Despite the link between positive coronary remodelling and acute ischaemic events, no data exist about the impact of arterial remodelling on subsequent progression of coronary atherosclerosis. The objective of this study was to examine whether extent and direction of arterial remodelling are predictors of progression of coronary atherosclerosis.
Design, setting and patients: From the Reversal of Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) trial, 210 focal coronary lesions (single lesion per patient) were identified with ⩽50% angiographic diameter stenosis at baseline intravascular ultrasound (IVUS). Remodelling was categorised using the remodelling index.
Main outcome measures: Lesion sites were matched to the 18-month follow-up IVUS examination and change in atheroma area was calculated. Additionally, change in atheroma volume of the whole imaged artery was calculated.
Results: There were no relationships between baseline remodelling index and change in atheroma area at the lesion site (r = 0.004, p = 0.96) or change in atheroma volume in the whole artery (r = 0.06, p = 0.37). Change in atheroma area was not significantly different in lesions with positive, negative or no remodelling at baseline (0.4 (SD 2.1) vs 0.7 (SD 1.7) vs 0.6 (SD 1.8) mm2, p = 0.76). Similarly, change in atheroma volume in the whole artery was not significantly different among the three remodelling categories (2.2 (SD 25.0) vs 1.4 (SD 31.2) vs 2.4 (SD 27.1) mm3, p = 0.98).
Conclusions: Extent and direction of arterial remodelling do not predict subsequent progression of coronary atherosclerosis. Although positively remodelled lesions are associated with unstable clinical presentation, they are not associated with accelerated progression of atherosclerosis during lipid lowering therapy.
Statistics from Altmetric.com
In response to accumulation of atherosclerotic plaque, coronary arteries are able to increase their cross-sectional area in order to maintain the luminal size. After the first description by Glagov et al,1 this process of compensatory arterial enlargement has been referred to as the Glagov phenomenon, or arterial remodelling. The widespread use of intravascular ultrasound (IVUS) to study coronary atherosclerosis contributed substantially to the concept of arterial remodelling.2 3 IVUS studies showed that some arteries could also develop shrinkage in response to plaque accumulation (ie, negative or constrictive remodelling) instead of compensatory enlargement (ie, positive or expansive remodelling).4–6 IVUS also showed that positive remodelling is associated with an unstable clinical presentation, whereas negative remodelling is more common in patients with a stable clinical presentation.7 8 Based on these data, expansive remodelling is considered a sign of plaque vulnerability.
Despite the association of remodelling with lesion instability and numerous studies examining the risk factor-related determinants of progression of coronary atherosclerosis including plasma lipids, blood pressure and inflammatory markers,9–11 the impact of arterial remodelling on subsequent progression of coronary atherosclerosis has never been evaluated. Thus, our objective was to examine whether the extent and direction of arterial remodelling are predictors of progression of coronary atherosclerosis. The Reversal of Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) trial,12 which examined progression of atherosclerosis during moderate vs intensive lipid-lowering, provided an ideal setting for this purpose because coronary arteries were serially evaluated by IVUS, enabling calculation of the progression rate of atherosclerosis.
The study design and inclusion criteria of the REVERSAL trial have been described in detail previously.12 Briefly, patients between the ages of 30 and 75 years who underwent coronary angiography for a clinical indication and demonstrated at least one obstruction with an angiographic diameter stenosis of 20% or more were enrolled. Lipid criteria required a low-density lipoprotein (LDL) cholesterol value between 125 and 210 mg/dl at baseline. The target vessel for IVUS interrogation was required to have at least a 30 mm long segment without an angiographic luminal narrowing ⩾50% in diameter and no previous angioplasty. Between June 1999 and September 2001, 654 patients were randomised at 34 centres, and 502 patients completed the protocol. For the current study, the baseline IVUS tapes of the patients who completed the study were screened to identify focal atherosclerotic lesions. Two hundred and nine patients had diffuse atherosclerosis and 56 had no disease (defined as maximum intima-media thickness <0.5 mm) and were excluded. Focal lesions with extensive calcification (an arc of >90°) that precluded accurate measurement of EEM area (n = 15) and those with evidence of plaque rupture (eight cases at baseline and four new ruptures at follow-up) were also excluded. The remaining 210 lesions from 210 patients (single lesion per coronary artery, the most proximal lesion in case of multiple eligible lesions) met the inclusion criteria and were included in the analyses.
The REVERSAL trial complied with the Declaration of Helsinki. The Institutional Review Board of the Cleveland Clinic approved this subanalysis.
Coronary IVUS images were acquired and measured using the methodology described in detail elsewhere.9 12 Briefly, a 30 MHz, 2.6F IVUS catheter (Ultracross, Boston Scientific Scimed Inc., Maple Grove, MN, USA) was advanced into the target vessel and the transducer was positioned distal to a side branch. An automatic pull-back device was used to withdraw the catheter at a speed of 0.5 mm/s. During pull-back, IVUS images were obtained at 30 frames/second and recorded on videotape. After 18 months, a follow-up IVUS interrogation was performed in the same arterial segment and under similar conditions.
The IVUS videotapes were analysed in accordance with the standards of the American College of Cardiology13 and European Society of Cardiology14 in a core laboratory at the Cleveland Clinic. After digitisation of the videotapes, selection of a distal fiduciary site was made at a frame just proximal to a distal side branch. A proximal fiduciary site was chosen just distal to the left main coronary artery bifurcation, aorta or the right coronary ostium. Measurements were performed at every 60th frame between distal and proximal fiduciary sites representing a series of cross-sections spaced 1 mm apart. For each frame, manual planimetry was used to trace the leading edges of the luminal and EEM borders. As previously reported, the mean (SD) intraobserver differences in the REVERSAL trial were minimal for both EEM (−0.16 (0.68) mm2) and lumen areas (−0.02 (0.75) mm2) and there were close correlations between initial analysis and reanalysis (r = 0.99 for EEM area; r = 0.98 for lumen area). Atheroma area was calculated as EEM area minus lumen area. For the purpose of this study, the operator identified the frame with the largest amount of atheroma relative to the arterial size as the lesion site. The remodelling index (RI) at baseline was calculated as the ratio of the EEM area at the lesion site to that of a proximal reference site (a normal-appearing frame within the 10 mm segment proximal to the lesion site). The lesions were classified into three groups according to baseline RI: positive remodelling group, defined as an RI>1.05; no remodelling group, defined as an RI between 0.95 and 1.05; and negative remodelling group, defined as an RI<0.95.7
At the lesion site, the atheroma burden was calculated as:
The total atheroma volume (TAV) was calculated as the sum of atheroma areas in all the analysed frames spaced 1 mm apart:
To compensate for differences in the length of the arterial segments analysed, a normalised TAV was derived as:
The per cent atheroma volume (PAV) was computed as the ratio of total atheroma volume divided by total EEM volume:
Analyses were performed using SPSS 11.5 for Windows (SPSS Inc., Chicago, Illinois). Categorical variables were described as n (%) and continuous variables were reported as mean (standard deviation). Because CRP values were not normally distributed, they are expressed as median and interquartile range. To compare the mean values among three groups, one-way analysis of variance (ANOVA) test was used. Paired t test and repeated measures analysis of variance were used to compare follow-up measurements with the baseline. The relationship between baseline remodelling index and change of atheroma area or total atheroma volume was analysed by linear regression. For the non-normally distributed CRP values, natural log-transformed values were used in the analyses. Two-sided p values <0.05 were considered significant.
Assuming a balanced distribution of the three remodelling categories, and a standard deviation of 1.5 mm2 with an alpha value of 0.05, the study had 81% power to detect a 0.7 mm2 difference in the progression rate between any of the remodelling categories.
Among the 210 lesions from 210 patients, the baseline remodelling index was greater than 1.05 in 113 (positive remodelling group), between 0.95 and 1.05 in 52 (no remodelling group) and less than 0.95 in 45 (negative remodelling group). The baseline characteristics and the study treatment allocation were similar among the three groups (table 1).
Table 2 summarises plasma lipid and C-reactive protein levels in the three groups. The baseline total cholesterol was somewhat lower in patients with no remodelling. Other baseline and follow-up laboratory data and the changes from baseline to follow-up were similar among the three groups. Baseline and follow-up blood pressure levels were not significantly different among the three remodelling categories either (p>0.05 by ANOVA).
IVUS measurements at the lesion site
Fifteen (7%) lesions were located in the left main coronary artery, 68 (32%) in the left anterior descending, 69 (33%) in the left circumflex, and 58 (28%) in the right coronary artery. Table 3 summarises the IVUS measurements at the lesion site. Baseline and follow-up atheroma area, EEM area, lumen area, and atheroma burden were similar among the three groups. The progression rate (change in atheroma area) was also similar among the three groups (p = 0.76). During the 18-month follow-up period, the atheroma area increased significantly in all three groups (positive remodelling group, from 7.7 (SD 3.0) mm2 to 8.2 (SD 3.5) mm2, p = 0.023; no remodelling group, from 7.2 (SD 2.4) mm2 to 7.8 (SD 2.6) mm2, p = 0.015; negative remodelling group, from 7.5 (SD 2.4) mm2 to 8.1 (SD 2.8) mm2, p = 0.013). However, the degree of change in atheroma volume was similar in all three groups (p = 0.76). The EEM area and lumen area also increased significantly from baseline and again there was no difference among the three groups. The atheroma burden did not change from baseline to follow-up.
As significant determinants of disease progression at the lesion site, follow-up total cholesterol (p = 0.02) and LDL cholesterol (p = 0.01) were higher in those with progression of disease (ie, an increase in atheroma area). There were also trends for more frequent allocation to pravastatin 40 mg (p = 0.11) and lower follow-up HDL cholesterol (p = 0.09) in patients with progression of disease.
IVUS measurements in the whole imaged arterial segment
Table 4 summarises the ultrasonographic measurements in the whole imaged arterial segment.
There were no differences in the baseline and follow-up normalised total atheroma volume or percent atheroma volume among the three groups. Similar to change in atheroma area at the lesion site, the changes in normalised total atheroma volume and per cent atheroma volume were not different among the three groups (p = 0.98 and p = 0.38, respectively). Neither normalised total atheroma volume nor per cent atheroma volume changed significantly from baseline to follow-up in any of the groups.
The baseline remodelling index did not correlate with the change in atheroma area at the lesion site (r = 0.004, p = 0.96) (fig 1A), or with the change in the normalised total atheroma volume (r = 0.06, p = 0.37) (fig 1B). There was no interaction between the statin regimen used (ie, moderate vs intensive) and the relationship of remodelling with progression of disease (interaction p value = 0.72 for change in atheroma area and p = 0.18 for change in nTAV).
Our findings demonstrate that, during 18 months of follow-up, progression of coronary atherosclerosis at the lesion site and in the entire coronary segment is unrelated to the extent and direction of arterial remodelling at baseline. These data provide insight into the relationship of lesion vulnerability and plaque progression during lipid-lowering therapy.
Arterial remodelling, first described by Glagov et al,1 is a heterogeneous response ranging from compensatory enlargement (positive or expansive remodelling) to paradoxical shrinkage (negative or constrictive remodelling). Despite its beneficial effect on maintaining luminal size, positive remodelling has been associated with several markers of plaque vulnerability, including higher lipid content and macrophage count, and medial thinning.15 16 In concert with these histopathological observations, positive remodelling has been clinically associated with acute coronary syndromes.7 8 On the other hand, arteries with negative remodelling frequently have pronounced fibrosis and calcification and less cellularity.17 Clinically, negative remodelling is associated with a more benign presentation of coronary artery disease, that is, chronic stable angina.6–8 15
Arterial remodelling is known to be a dynamic phenomenon, which takes place alongside atherosclerotic plaque development. While remodelling is thought to occur in response to growth of an atherosclerotic plaque, no study has evaluated whether pre-existing arterial remodelling can predict subsequent progression of atherosclerosis– that is, the answer to the question “Do lesions with positive remodelling progress differently from lesions with negative remodelling?” was not known. Based on the above-mentioned histopathological and clinical IVUS studies, it would appear logical to find more progression in positively remodelled lesions. In this study, we examined whether arterial remodelling can predict subsequent progression of atherosclerosis using a validated methodology for assessment of remodelling in the setting of a large IVUS trial.18 19 During 18 months’ follow-up of 210 lesions, changes in atheroma area at the lesion site and the change in atheroma volume in the whole imaged segment were essentially the same in lesions with positive remodelling, no remodelling and negative remodelling. Thus, our data demonstrates that arterial remodelling, per se, does not determine subsequent progression of atherosclerosis. While the standard deviation for change in atheroma area was higher than the values used in the initial power calculations and the remodelling categories were unbalanced, using the observed standard deviation and the observed sample size for each group, the study still had a 70% power to detect a 0.7 mm2 difference between the remodelling categories.
Negative remodelling has been frequently viewed as a burnt-out stage of coronary atherosclerosis, given its association with fibrotic changes and calcification in the vessel wall. Given these pathological findings and the association of negative remodelling with more benign presentations of coronary disease, it would be intuitive to find less progression in negatively remodelled and more progression in positively remodelled lesions. Our findings disagree with this. It is probable that disease progression during lipid-lowering therapy is more complex, because the most vulnerable, active lesions may be the ones most responsive to therapy, while the most stable may be the more resistant to therapy. This differential response to lipid-lowering therapy may nullify the effect of remodelling on subsequent progression in this setting. This concept is supported by our recent finding that there is less regression of disease in more overtly calcified, and therefore the more stable, lesions.20
Due to the protocol of the original study, all patients were undergoing either a moderate or an aggressive lipid-lowering regimen throughout the study period. Therefore, our finding of a lack of a relationship between remodelling and subsequent progression of disease may not reflect the natural history of atherosclerosis. However, interaction analyses revealed that the intensity of lipid lowering (ie, moderate vs intensive) did not affect the lack of relationship between baseline remodelling and progression of disease, suggesting that our findings are independent of statin therapy. On the other hand, our findings may not be applicable to severely stenotic, diffuse or overtly calcified lesions, none of which were included in this study.
Extent and direction of arterial remodelling do not predict subsequent progression of coronary atherosclerosis in patients undergoing statin therapy. Although positively remodelled lesions are more commonly associated with an unstable clinical presentation, they are not associated with a more accelerated progression of atherosclerosis. Negative remodelling is not a burnt-out stage of atherosclerosis and lesions with negative remodelling can progress to the same extent as those with positive remodelling.
We thank Eva Balazs, Craig Balog, Sorin Brener, Marlene Goormastic, Thomas Ivanc, Roman Poliszczuk and Kathy Wolski at the Cleveland Clinic Cardiovascular Coordinating Center and Michele Norton and Fady Ntanios at Pfizer Inc. for their contributions.
Funding: The REVERSAL (Reversal of Atherosclerosis with Aggressive Lipid Lowering) trial was funded by Pfizer. This substudy is partially funded by a National Institutes of Health, National Center for Research Resources, General Clinical Research Center Grant MO1 RR-018390. PS was supported by a postdoctoral fellowship grant of the Ohio Affiliate of the American Heart Association. A Ralph Reader Overseas Research Fellowship from the National Heart Foundation of Australia supports SJN.
Competing interests: IS received lecture honoraria and an educational grant from Pfizer. EMT received grant support from Pfizer and Takeda and lecture honoraria from Pfizer. SJN has received lecture honoraria from Pfizer and AstraZeneca. PS has consulted for Takeda without financial compensation. All honoraria are paid directly to higher education, so that neither income nor any tax deduction is received. SEN has received research support from AstraZeneca, Eli Lilly, Pfizer, Takeda, Sankyo and Sanofi-Aventis. He has also 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.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.