Author + information
- Received June 2, 2008
- Revision received August 19, 2008
- Accepted August 28, 2008
- Published online March 1, 2009.
- Tetsuro Kataoka, MD⁎,
- Verghese Mathew, MD, FACC⁎,
- Ronen Rubinshtein, MD⁎,
- Charanjit S. Rihal, MD, FACC⁎,
- Ryan Lennon, MS†,
- Lilach O. Lerman, MD, PhD‡ and
- Amir Lerman, MD, FACC⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Amir Lerman, Division of Cardiovascular Disease, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905
Objectives The current study was designed to investigate the relationship between renal arterial structure and vessel remodeling in patients with atherosclerotic renal artery stenosis (RAS), compared with that seen in coronary artery disease (CAD).
Background The nature and the tissue characterization of atherosclerotic RAS lesions have not been fully explored.
Methods Gray scale and virtual histology (VH) intravascular ultrasound imaging was used to assess 23 lesions in 14 consecutive RAS patients and 20 left main trunk lesions in age-matched CAD patients. Analysis included assessment of vessel area and atherosclerotic plaque area of the main renal artery or left main trunk. Plaque was characterized as fibrous tissue, fibro-fatty tissue, necrotic core, and dense calcium. Remodeling was assessed by means of the remodeling index (RI).
Results Positive remodeling (defined as RI ≥1.05) was present in 15 RAS and 9 CAD lesions, whereas intermediate/negative remodeling (RI <1.05) was present in 8 RAS and 11 CAD lesions. VH showed that the fibrous tissue was the most prominent plaque composition, followed by fibro-fatty, necrotic core, and dense calcium in both vascular beds. Greater vascular adaptive enlargement was observed in slices with plaque burden ≤40% compared with plaque burden >40% (p < 0.001 for all). Vessel area had a positive association with the area of all VH components (p < 0.001, for all). VH analysis shows that the most powerful determinant of adaptive vessel enlargement is dense calcium in RAS (p < 0.001), while that is necrotic core in CAD (p < 0.001). Necrotic core and dense calcium areas were greater in lesions with positive remodeling compared with intermediate/negative remodeling (p = 0.03, p = 0.03, respectively, in RAS; p = 0.005, p = 0.03, respectively, in CAD).
Conclusions The current study demonstrates in humans that plaque composition as assessed by VH intravascular ultrasound has an important role of adaptive vessel enlargement, and it is related to renal artery remodeling in RAS in a pattern similar to CAD.
- virtual histology
- atherosclerotic renal artery stenosis
- coronary artery disease
- adaptive vessel enlargement
- remodeling index
Renal artery stenosis (RAS) is one of the leading causes of secondary hypertension, and can result in refractory hypertension or ischemic renal failure. RAS is present in 0.5% to 5% of all hypertensive patients (1,2). The prevalence of atherosclerotic RAS in patients with coronary artery disease (CAD) and/or peripheral artery disease is 19% to 30% (3,4). Atherosclerosis accounts for 70% to 90% of cases of RAS (5,6). Previous human studies have established that in CAD compensatory vessel enlargement delayed luminal compromise until the plaque burden was >40% (7). Such adaptive enlargement of vessels can compensate for the accumulation of atherosclerotic plaque in vivo to preserve the vessel lumen (8). The remodeling process has not been elucidated in atherosclerotic RAS, and intravascular ultrasound imaging (IVUS) may be feasible and useful for the assessment of atherosclerotic RAS (9–11).
Recent studies ex vivo and in vivo compared with histopathological findings have reported that virtual histology (VH) allows reliable characterization of atherosclerotic plaque and distinction of 4 components: fibrous tissue, fibro-fatty tissue, dense calcium, and necrotic core (12,13). The distribution of VH plaque components of the coronary artery has been demonstrated previously (12–14). Furthermore, a relationship between coronary artery remodeling and VH plaque components has been described (15–17).
The physiology of the renal artery is different from the coronary artery, as it shows functional responses, some of which are unique for the renal circulation, and renal blood flow far exceeds coronary blood flow. Hence, the remodeling process in the renal artery during atherogenesis might conceivably exhibit a differential pattern than that observed in coronary circulation. In the present study, we evaluated the plaque composition and adaptive vessel enlargement and remodeling index (RI) in atherosclerotic renal arterial and coronary arterial lesions using VH-IVUS.
Patient population and data collection
The Mayo Clinic Institutional Review Board approved the present study. Fourteen consecutive patients (age 72 ± 7 years [mean ± SD], range 61 to 86 years; 9 women) who were referred to renal angiography for evaluation of atherosclerotic RAS (>30% diameter stenosis), and 20 additional age-matched patients (age 69 ± 6 years [mean ± SD], range 50 to 80 years; 6 women) with known left main trunk (LMT) narrowing (>30% diameter stenosis) diagnosed by coronary angiography were enrolled in the study. Overall, 23 lesions (10 right renal arteries, 13 left renal arteries), and 20 LMT lesions were examined by VH-IVUS and included for analysis. Patients with in-stent restenosis, total vessel occlusion, or acute coronary syndromes were excluded from this study. Angiographic stenosis was measured by a caliper method.
IVUS image protocol and data analysis
IVUS studies were performed using a 30-MHz 2.9-F IVUS catheter (Volcano Therapeutics, Rancho Cordova, California). The IVUS catheter was positioned distally to the stenosis in the main renal artery or LMT. VH-IVUS data were acquired using slow manual pullback, by a dedicated VH-IVUS console (Volcano Therapeutics). The VH-IVUS data were stored on DVD-R and sent to the imaging core lab for offline analysis.
External elastic membrane cross-sectional area (vessel area), lumen cross-sectional area (lumen area), plaque plus media cross-sectional area (plaque area: vessel area − lumen area), and plaque burden (plaque area ÷ by vessel area × 100) were measured with the use of custom-built software (IVUSLab, Volcano Corp.). Manual contour detection of both the lumen and the media-adventitia interface was performed by independent and experienced investigators in the imaging core lab. The image slice with the minimum lumen as well as the largest plaque burden was selected for analysis. The lesion vessel area and lesion plaque area were reported from the slice with minimum lumen area. The reference site was identified from the slice with the largest lumen area and the smallest plaque burden in the distal main renal artery or LMT. Lesion plaque burden was defined as the plaque burden at the slice with minimum lumen area. Percent area stenosis was calculated as the difference between the minimum lumen area and the lumen area at the reference site divided by the lumen area at the reference site.
Details regarding the validation of VH-IVUS acquisition and analysis have previously been reported (12,18,19). Briefly, VH-IVUS uses spectral analysis of IVUS radiofrequency data to construct tissue maps that classify plaque into 4 major components (fibrous [labeled dark green], fibro-fatty [labeled light green], necrotic core [labeled red], and dense calcium [labeled white]), correlating a specific spectrum of the radiofrequency signal with assigned color codes. VH-IVUS analysis was reported in absolute area and as a percentage (fractional) of plaque area.
The definition of adaptive vessel enlargement associated with plaque accumulation
With regard to adaptive vessel enlargement associated with plaque accumulation, the association between vessel area and plaque area was assessed according to a previous report (20). The degree of adaptive vessel enlargement was defined as the expected increase in vessel area per 1 mm2 increase in plaque area or each plaque component.
The definition of RI and remodeling pattern
The RI was calculated as the vessel area at the minimum lumen area ÷ the reference vessel area. Positive remodeling was defined as an RI ≥1.05, and intermediate/negative remodeling as an RI <1.05.
Continuous variables are expressed as the mean ± 1 SD. Discrete variables are expressed as percentage frequencies. Comparison of continuous variables was performed by the 2 sample Student t tests for normally distributed data, and by the Wilcoxon rank sum test for non-normally distributed data. The chi-square, or Fisher exact test for sparse data, was used for comparing the frequency of occurrence. Mixed regression models were used to make inferences on data from VH-IVUS slices (vessel area, plaque burden, plaque composition). For all such models, the correlation within patients, and within arteries within patients (for those with both right and left renal arteries analyzed) were accounted for by including random intercept terms for each unique patient and each artery (nested within patients). A p value of <0.05 was considered to indicate statistical significance. All statistical tests were 2-sided. Most statistical analyses were performed using JMP version 6.0.0 (SAS Institute, Cary, North Carolina), except the mixed models, which were performed with SAS-STAT version 9.1.3 (SAS Institute).
Baseline patient characteristics in RAS compared with those in CAD patients
Twenty-three main renal artery lesions and 20 LMT lesions were studied in 14 RAS patients and 20 CAD patients, whose baseline characteristics are presented in Table 1. A total of 488 slices (mean 21 slices per vessel, range 7 to 32 slices per vessel) in RAS lesions and a total of 338 slices (mean 17 slices per vessel, range 12 to 20 slices per vessel) were analyzed. The proportion of female subjects in RAS patients was greater than that of CAD patients (Table 1). The other factors were similar between the 2 groups, except that hypertension tended to be more prevalent in RAS patients (p = 0.05).
Inter- and intraobserver variability
Inter- and intraobserver variability of IVUS data were obtained (vessel area 4 ± 3% and 2 ± 2%, respectively, and plaque area 5 ± 6% and 3 ± 3%, respectively).
Angiographic data and IVUS data in RAS compared with that in CAD patients
Angiographic stenosis was similar between RAS patients and CAD patients. The reference plaque burden and percent area stenosis of RAS patients were smaller than those of CAD patients (Table 2). For RAS patients, the stenotic lesion including the slice with minimum lumen area existed in the proximal part of the main renal artery near ostium of aorta in most cases, in keeping with the known proximal distribution of atherosclerotic RAS.
Distribution of VH plaque component
VH of RAS and CAD lesions showed that atherosclerotic plaques contained 61 ± 11% and 57 ± 11% for fibrous plaque, 21 ± 13% and 18 ± 10% for fibro-fatty plaque, 6 ± 4% and 11 ± 6% for dense calcium, and 11 ± 9% and 14 ± 7% for necrotic core, respectively (Fig. 1).
The mean percent area of dense calcium in RAS lesions was significantly smaller than that in CAD lesions (p = 0.003). The fibrous tissue, fibro-fatty tissue, and necrotic core were similar between RAS lesions and CAD lesions (Table 2).
Adaptive vessel enlargement associated with plaque accumulation
The degree of adaptive vessel enlargement (the expected increase in vessel area per 1 mm2 increase in plaque area) was significantly greater in the slices with plaque burden ≤40% than in those with plaque burden >40% in both RAS and CAD lesions (p < 0.001).
There were significant positive associations between vessel area and plaque area. In RAS lesions, for every 1-mm2 increase in plaque area, vessel area increased by 1.11 mm2 at the slices with plaque burden ≤40% and by 0.63 mm2 in those with plaque burden >40% (Fig. 2). For CAD lesions, vessel area increased by 1.32 mm2 for every 1-mm2 increase in plaque area at the slices with plaque burden ≤40% and by 0.83 mm2 in those with plaque burden >40% (Fig. 2).
Associations between vessel area and VH plaque components
There were significant positive associations between vessel area and each area of fibrous tissue, fibro-fatty tissue, necrotic core, and dense calcium in both RAS and CAD lesions (p < 0.001 for all).
In RAS lesions, the unadjusted associations between plaque components and vessel area are as follows: A 1 mm2 increase of fibrous tissue was associated with a 0.52-mm2 increase in vessel area. A 1-mm2 increase of fibro-fatty tissue was associated with a 0.42-mm2 increase in vessel area. A 1 mm2 increase of necrotic core was associated with a 1.65-mm2 increase in vessel area. A 1-mm2 increase of dense calcium was associated with a 4.33-mm2 increase in vessel area (Fig. 3). Therefore, the greatest determinant of vessel enlargement was dense calcium, followed by necrotic core, fibro-fatty, and fibrous tissue in RAS lesions.
In CAD lesions, 1-mm2 increase of fibrous tissue was associated with a 0.92 mm2 increase in vessel area. A 1-mm2 increase of fibro-fatty tissue was associated with a 1.25-mm2 increase in vessel area. A 1-mm2 increase of necrotic core was associated with a 2.01-mm2 increase in vessel area. A 1-mm2 increase of dense calcium was associated with a 1.04-mm2 increase in vessel area (Fig. 4).
A multivariate model for vessel area for both RAS and CAD lesions revealed several comparisons between the 2 groups of arteries. First, the expected increase in vessel area per 1-mm2 increase in fibrous tissue was significantly less (p = 0.005) in the RAS slices. It was also significantly less (p = 0.03) in RAS slices for increase in necrotic core. However, the expected increase in vessel area per 1-mm2 increase in dense calcium was significantly greater (p < 0.001) in the RAS lesions. Furthermore, in RAS lesions, the expected increase in vessel area per 1-mm2 increase in dense calcium was significantly greater than for the other 3 plaque composition measures (p < 0.001), although all 4 were significantly associated with vessel area (p < 0.001, p < 0.03, p < 0.03, and p < 0.001, respectively). In CAD lesions, only fibrous tissue and necrotic core were significantly associated with vessel area in the model (p < 0.001 and p < 0.001, respectively), and the coefficients for these 2 were not significantly different (p = 0.14).
Comparison of baseline patient characteristics between positive remodeling and intermediate/negative remodeling
The levels of serum triglycerides were significantly lower in RAS patients who had lesions with positive remodeling. There were no other significant differences between lesions with positive remodeling and intermediate/negative remodeling in RAS and CAD patients (Table 3).
Comparisons of angiographic data and IVUS data between positive remodeling and intermediate/negative remodeling
There were no significant differences in the degree of angiographic diameter stenosis between lesions with positive remodeling and intermediate/negative remodeling in both RAS and CAD patients. Minimum lumen area was also similar in vessels with positive remodeling or intermediate/negative remodeling. However, there were significant differences in lesion vessel area and lesion plaque area between the 2 groups in RAS patients. In addition, there were also significant differences for lesion plaque burden and percent area stenosis between the 2 groups in CAD patients (Table 4).
Comparisons of VH-IVUS data between positive remodeling and intermediate/negative remodeling
Figures 5 and 6⇓⇓ show examples of VH-IVUS images with positive and negative remodeling in RAS and CAD patients, respectively. There was greater necrotic core and dense calcium in lesions with positive remodeling compared with that seen with negative remodeling.
In RAS lesions, the area and fractional area of both the necrotic core and dense calcium were significantly larger in vessels with positive remodeling than in those with intermediate/negative remodeling. There were no significant differences between the groups in the area and fractional area of fibrous and fibro-fatty tissue (Fig. 7).
Similarly, in CAD lesions, the area and fractional area of the necrotic core and dense calcium were significantly larger in lesions with positive remodeling than in those with intermediate/negative remodeling. However, in addition, the fractional area of the fibrous tissue in lesions with positive remodeling was smaller than intermediate/negative remodeling. There were no significant differences for the area of the fibrous tissue, and the area and the fractional area of fibro-fatty tissue (Fig. 7).
The main findings of this study
The main findings of this study are:
1. Comparison of plaque composition and vessel remodeling between RAS and CAD
2. The characteristics of atherosclerotic RAS assessed by VH-IVUS
3. The impact of VH plaque components on adaptive vessel enlargement
4. The impact of VH plaque components on remodeling pattern
This study describes for the first time the characteristics of atherosclerotic RAS assessed by VH-IVUS, and demonstrates that the distribution of the plaque components and vessel response of the renal arteries is very similar to the findings in CAD patients with LMT lesions. Hence, this study supports the concept that atherosclerosis is a diffuse and systemic disease imposing similar plaque composition in multiple vascular beds.
The distribution of VH plaque components
According to previous reports from the coronary artery and carotid artery circulations, fibrous tissue was the most prominent plaque composition (58% to 71%), followed by fibro-fatty (7% to 31%), necrotic core (7% to 29%), and dense calcium (2% to 10%) (14–17,21). In our subjects, the distribution of VH plaque composition was similar to previous reports, although the fractional area of dense calcium of CAD lesions was somewhat greater than that of RAS lesions. The degree of angiographic stenosis was similar between RAS and CAD lesions, but the percent area stenosis and the references plaque burden of CAD lesions were greater than those of RAS lesions. The greater lesions with advanced atherosclerosis might be included in CAD lesions.
Adaptive vessel enlargement to plaque accumulation
The adaptive vessel enlargement to plaque accumulation has been observed in coronary arteries. Glagov et al. (7) showed that compensatory vessel enlargement delayed luminal compromise until plaque burden becomes >40%. The present study extends these previous observations to renal arteries and demonstrated in vivo in humans a significant positive association between vessel area and plaque area. The degree of adaptive vascular enlargement of slices with plaque burden >40% was attenuated compared with that seen with slices with plaque burden ≤40%, which suggests that in the renal artery, adaptive vessel enlargement delays luminal narrowing, similar to what has been reported in the coronary circulation.
Association of vessel area with VH plaque components
Histopathological studies have previously described the plaque composition of CAD (22), and the relationship between vessel remodeling and plaque composition (23,24). VH-IVUS allows identification of 4 different types of atherosclerotic plaques in vivo, and its accuracy has been validated against histopathological assessments (12,13).
This is the first study to demonstrate the relationship between vessel area and VH plaque components in atherosclerotic RAS. Our study demonstrated that adaptive vessel enlargement as assessed by VH was associated with the area of all components in atherosclerotic RAS and CAD. The most powerful determinant of adaptive vessel enlargement was dense calcium in RAS, while that of vessel enlargement was necrotic core in CAD. The reason for this impact of VH plaque components on adaptive vessel enlargement is unknown. Compensatory enlargement is part of an adaptive mechanism that regulates blood flow to organs. The severity of lesions as well as blood flow might affect the remodeling response. Since renal blood flow is markedly higher than coronary blood flow, it may impose different hemodynamic forces (e.g., shear stress) on the vessel wall that favor deposition of calcium rather than development of a necrotic core. In addition, variable development of lesions has been recently shown to be site-specific and correlate with atherogenic gene expression in the coronary compared with the peripheral circulation (25).
The definition and prevalence of the remodeling pattern
We investigated main renal arteries with a wide range of angiographic stenosis. The stenotic lesion including the slice with minimum lumen area was typically in the proximal part of the main renal artery near ostium of aorta in most cases. Therefore, in our study, the reference site was selected in the distal part of the main renal artery, and positive remodeling was defined as RI >1.05, as previous reports suggested (11,15,17,26).
Leertouwer et al. (11) studied symptomatic RAS in vivo, and reported that coarctation of renal arteries, which was defined as RI <0.85, was prevalent in 50% of cases. Pasterkamp et al. (26) reported that the prevalence of renal vascular enlargement was significantly higher compared with that of vascular shrinkage (8%) in mild or moderate atherosclerosis. A previous study demonstrated an increased prevalence of failure of adaptive vessel enlargement with severe luminal narrowing in coronary arteries (27). The prevalence of negative remodeling may increase if more lesions with severe luminal narrowing are particularly selected, although the lesions with positive remodeling dominated in our subjects.
The impact of VH plaque components on remodeling pattern
In human renal arteries, previous comparisons between the IVUS image and histopathological features have been reported (9,28); however, plaque components were not fully described. Our findings showed that necrotic core and dense calcium had significant correlations with positive remodeling in both RAS and CAD lesions. These findings are in keeping with the data of Burke et al. (23) in coronary artery atherosclerosis. The impact of plaque components on remodeling might conceivably vary within the arterial system; nevertheless, the vessel remodeling of RAS was similar to that of CAD. This study supports the concept that atherosclerosis is a diffuse and systemic disease with multifocal manifestations.
A limitation of this study is that it is the retrospective analysis of a limited number of cases, although the data were collected prospectively by independent investigators. There are no classifications for thrombus or blood on VH-IVUS. VH-IVUS data were acquired using slow manual pullback; therefore, plaque volumes were not acquired.
Classifications of lesion types by VH-IVUS lack histopathological validation in renal arteries, and we extrapolated VH-IVUS analysis that was validated in the coronary circulation. However, it may be speculated that the histopathological manifestation will be similar to the coronary and carotid circulation. Further studies should try to validate the findings of this study comparing histopathological data.
Our study suggests that plaque accumulation in the renal artery and coronary artery results in adaptive vascular enlargement, which is accelerated in segments with plaque burden ≤40% compared with plaque burden >40%. VH analysis shows that the most powerful determinant of adaptive vessel enlargement is dense calcium in RAS, and necrotic core in CAD. Vessels with positive remodeling have greater necrotic core and dense calcium as compared with vessels with intermediate/negative remodeling in patients with atherosclerotic RAS. The vessel remodeling of RAS was similar to that of CAD.
Tissue characterization using VH may provide more insights into the prognosis and natural history of patients with RAS and into the effect of conventional and emerging treatment modalities for RAS.
- Abbreviations and Acronyms
- coronary artery disease
- intravascular ultrasound imaging
- left main trunk
- renal artery stenosis
- remodeling index
- virtual histology
- Received June 2, 2008.
- Revision received August 19, 2008.
- Accepted August 28, 2008.
- American College of Cardiology Foundation
- Ramsay L.E.,
- Waller P.C.
- Hamasaki S.,
- Higano S.T.,
- Suwaidi J.A.,
- et al.
- Dangas G.,
- Laird J.R. Jr..,
- Mehran R.,
- Lansky A.J.,
- Mintz G.S.,
- Leon M.B.
- Leertouwer T.C.,
- Gussenhoven E.J.,
- van Dijk L.C.,
- et al.
- Nair A.,
- Kuban B.D.,
- Tuzcu E.M.,
- Schoenhagen P.,
- Nissen S.E.,
- Vince D.G.
- Nasu K.,
- Tsuchikane E.,
- Katoh O.,
- et al.
- Rodriguez-Granillo G.A.,
- Serruys P.W.,
- Garcia-Garcia H.M.,
- et al.
- Surmely J.F.,
- Nasu K.,
- Fujita H.,
- et al.
- Konig A.,
- Klauss V.
- Rodriguez-Granillo G.A.,
- Garcia-Garcia H.M.,
- McFadden E.P.,
- et al.
- Diethrich E.B.,
- Pauliina Margolis M.,
- Reid D.B.,
- et al.
- Virmani R.,
- Kolodgie F.D.,
- Burke A.P.,
- Farb A.,
- Schwartz S.M.
- Burke A.P.,
- Kolodgie F.D.,
- Farb A.,
- Weber D.,
- Virmani R.
- Varnava A.M.,
- Mills P.G.,
- Davies M.J.
- Mohler E.R. 3rd.,
- Sarov-Blat L.,
- Shi Y.,
- et al.
- Pasterkamp G.,
- Schoneveld A.H.,
- van Wolferen W.,
- et al.
- Nishioka T.,
- Luo H.,
- Eigler N.L.,
- Berglund H.,
- Kim C.J.,
- Siegel R.J.
- Yasuda G.,
- Takizawa T.,
- Takasaki I.,
- Shionoiri H.,
- Umemura S.,
- Shimoyama K.