Presented in part at the 42nd Annual Meeting of the American Society for Nephrology held 27 October–1 November 2009 in San Diego, CA, USA.
Coronary Plaque Morphology Using Virtual Histology–Intravascular Ultrasound Analysis in Hemodialysis Patients
Article first published online: 13 SEP 2010
© 2010 The Authors. Therapeutic Apheresis and Dialysis © 2010 International Society for Apheresis
Therapeutic Apheresis and Dialysis
Volume 15, Issue 1, pages 44–50, February 2011
How to Cite
Kono, K., Fujii, H., Miyoshi, N., Kawamori, H., Shite, J., Hirata, K.-i. and Fukagawa, M. (2011), Coronary Plaque Morphology Using Virtual Histology–Intravascular Ultrasound Analysis in Hemodialysis Patients. Therapeutic Apheresis and Dialysis, 15: 44–50. doi: 10.1111/j.1744-9987.2010.00855.x
- Issue published online: 27 JAN 2011
- Article first published online: 13 SEP 2010
- Received April 2010; revised June 2010.
- Coronary artery disease;
- Hemodialysis patient;
- Vascular calcification;
- Virtual histology–intravascular ultrasound
Most dialysis patients have coronary artery disease at the initiation of dialysis therapy and these patients also have marked vascular calcification. Virtual histology–intravascular ultrasound (VH–IVUS) provides coronary tissue maps that are color coded by four major plaque components and facilitate the characterization of coronary plaque composition in vivo. The aim of this study was to identify coronary plaque characteristics in dialysis patients using VH–IVUS. Twenty-three patients with coronary artery disease were included in this study. Of these, 12 patients had normal renal function or mild renal insufficiency (control group) and 11 patients were receiving maintenance dialysis therapy (hemodialysis group). We performed coronary angiography and VH–IVUS analysis on culprit lesions of all patients in the study. The result of VH–IVUS analysis showed that the hemodialysis group had a greater plaque volume, lower percentage of fibrous plaque, and higher percentage of dense calcium plaque compared with the control group. In addition, the serum phosphate levels were significantly associated with the percentage of necrotic core and dense calcium plaque in all study patients. Our findings suggest that the amount of necrotic core and dense calcium plaques increase significantly in hemodialysis patients, and that disordered mineral metabolism may be associated with coronary plaque morphology.
Cardiovascular disease (CVD) is a major cause of death in patients with chronic kidney disease (CKD), especially in dialysis (CKD stage 5D) patients. Patients with CKD have a higher risk of developing CVD than the general population. Many reports have demonstrated that most dialysis patients have coronary artery disease (CAD) at the initiation of dialysis therapy, and these patients also have marked vascular calcification (1–5). It has been reported that coronary artery calcification predicts coronary artery events, and vascular calcification, including coronary artery calcification, was found to be associated with mortality in CKD patients (6); however, the detailed mechanism of vascular calcification remains unclear in these patients.
It is well known that hyperphosphatemia and elevated calcium–phosphorus products (Ca × P) are associated with cardiovascular mortality in CKD patients (7,8). Moreover, some reports have shown that appropriate control of serum phosphate levels by phosphate binders suppresses the progression of vascular calcification in dialysis patients (9,10). Thus, mineral metabolism disorders such as hyperphosphatemia and/or increased Ca × P are suggested to be strongly linked with vascular calcification in maintenance dialysis patients (11). In addition, a recent study has demonstrated that serum phosphate levels are significantly associated with the risk of cardiovascular events in patients with early-stage CKD (12).
Virtual histology–intravascular ultrasound (VH–IVUS) is a useful tool for the evaluation of plaque composition and morphology in vivo. Using radiofrequency backscatter signals received by the IVUS catheter, plaque images are divided into four different types of atherosclerotic plaque components: fibrous, fibro-fatty, necrotic core, and dense calcium. VH–IVUS analysis has been shown to provide structural information about coronary plaques with 87–97% accuracy in vitro (13).
The aim of the present study is to assess coronary plaque morphology and morphometry in vivo using VH–IVUS analysis and identify coronary plaque characteristics in hemodialysis patients.
PATIENTS AND METHODS
Among the 899 patients who had undergone coronary angiography and VH–IVUS analysis at our institution between May 2005 and April 2009, 29 patients were receiving maintenance hemodialysis therapy. Of these, patients with severe heart failure, cardiogenic shock, bypass graft lesions, overt infection, malignancy and inflammatory disease were excluded, and patients without sufficient data were also excluded. The remaining 11 patients were enrolled in the HD group. In addition, 12 patients with an estimated glomerular filtration rate (eGFR) ≥ 60 mL/min/1.73 m2,who were matched for age, sex, and the presence of diabetes mellitus and hypertension, were selected in the control group. Experimental protocols were approved by the appropriate institutional review committee and informed consent was obtained from all patients.
VH–IVUS imaging and analysis
Before percutaneous coronary intervention (PCI) and after intracoronary administration of 300 µg nitroglycerin, VH–IVUS was performed for the patients with clinically suspected ischemic heart disease, such as acute coronary syndrome, acute myocardial infarction, and stable/unstable angina pectoris. For all study patients we analyzed only one of the stenotic lesions—that which was most responsible for the clinical symptoms, echocardiographic abnormalities, and positive findings of a stress electrocardiogram and myocardial scintigraphy—using VH–IVUS. These analyzed stenotic lesions were defined as “culprit lesions” in this study. Lesions which had previously undergone PCI were excluded. The VH–IVUS catheter was automatically pulled back at 0.5 mm/s throughout the culprit lesion. The site (length of 15 slices) selected from the images of each culprit lesion was analyzed using custom-built software (Volcano Therapeutics, Rancho Cordova, CA, USA). As shown in Figure 1, plaque images were reconstructed as coronary tissue maps that were color-coded by four major plaque components (fibrous, fibro-fatty, necrotic core, and dense calcium). The volume of each plaque composition was automatically calculated.
Measurement of biochemical parameters by clinical laboratory tests
We measured biochemical parameters that were reported to be related to vascular calcification in hemodialysis patients, including calcium, phosphate, and high-sensitive C-reactive protein (hsCRP). After fasting overnight, venous blood was collected from each patient following a 20-min period of supine rest in the morning. In the HD group, blood was collected before dialysis. Laboratory tests were conducted according to standardized clinical laboratory methods. The blood chemistry data were calculated as the average of three measurements. Renal function in the control group was evaluated by eGFR using the modified Modification of Diet in Renal Disease equation.
Computer software application StatView 5.0 (SAS Institute, Cary, NC, USA) was used for all statistical analyses. Values are presented as mean ± SD. The significance of differences between the two groups was analyzed using the Student's t-test for continuous variables and χ2-test for categorical variables. Relationships between variables and plaque morphology were analyzed using Pearson's correlation coefficient and Spearman's rank correlation analysis. P < 0.05 was considered statistically significant.
Table 1 compares patient characteristics between control and HD groups. Sex, age, body mass index, presence of diabetes and hypertension, as well as the smoking status of former and current smokers were comparable between the control and HD groups. The two groups had similar lipid profiles. Hemoglobin, hematocrit and serum calcium levels were significantly lower in the HD group compared with the control group, while serum phosphate levels and Ca × P were significantly higher in the HD group.
|Control group (N = 12)||Hemodialysis group (N = 11)||P|
|eGFR (mL/min/1.73 m2)||87.0 ± 16.1||–||–|
|Age (years)||61 ± 11||64 ± 8||0.381|
|Sex (male; %)||5 (42)||3 (27)||0.805|
|BMI (kg/m2)||22.5 ± 3.4||23.8 ± 3.9||0.401|
|Smoking (%)||6 (50)||3 (27)||0.285|
|Diabetes mellitus (%)||7 (58)||8 (27)||0.491|
|Hypertension (%)||9 (82)||10 (100)||0.081|
|ACEi/ARB (%)||3 (25)||5 (50)||0.066|
|Statin (%)||7 (58)||3 (27)||0.145|
|Hemoglobin (g/dL)||12.2 ± 1.9||10.2 ± 1.9||0.012|
|Albumin (mg/dL)||3.6 ± 0.7||3.4 ± 0.5||0.370|
|LDL (mg/dL)||112.5 ± 47.9||98.6 ± 40.0||0.574|
|HDL (mg/dL)||43.0 ± 12.6||42.2 ± 13.1||0.874|
|TG (mg/dL)||206.8 ± 191.1||175.5 ± 123.6||0.649|
|hsCRP (mg/L)||3.8 ± 6.7||2.6 ± 3.7||0.605|
|Ca (mg/dL)||9.5 ± 0.3||9.2 ± 0.3||0.040|
|P (mg/dL)||3.8 ± 0.6||4.9 ± 0.9||0.002|
|Ca × P||36.2 ± 6.0||45.3 ± 8.7||0.010|
|iPTH (pg/mL)||–||281.9 ± 144.6||–|
Coronary angiographic findings in the two groups
Table 2 shows the coronary angiographic findings in the control and HD groups. Culprit lesions were more commonly located in the left descending artery in all study patients. Patients in the HD group had more severe coronary artery disease. Eight of 11 patients (73%) in the HD group had triple vessel disease compared with 4 of 12 patients (33%) in the control group.
|Control group (N = 12)||Hemodialysis group (N = 11)||P|
|LAD (%)||5 (42)||5 (46)||0.862|
|LCX (%)||3 (25)||2 (18)||0.708|
|RCA (%)||4 (33)||4 (36)||0.885|
|Number of diseased vessels|
|Single (%)||6 (50)||1 (9)||0.033|
|Double (%)||2 (17)||2 (18)||0.927|
|Triple (%)||4 (33)||8 (73)||0.062|
As shown in Table 3, plaque and vessel volumes were significantly greater in the HD group than those in the control group. As for the identification of coronary plaque characteristics using VH–IVUS analysis, the volumes of necrotic core and dense calcium plaques significantly increased in the HD group compared with the control group (Fig. 2a). The percentage of dense calcium plaque also significantly increased in the HD group compared with the control group (Fig. 2b).
|Control group (N = 12)||Hemodialysis group (N = 11)||P|
|Vessel volume (mm3)||87.6 ± 15.2||110.3 ± 29.4||0.039|
|Lumen volume (mm3)||26.4 ± 8.1||28.9 ± 6.7||0.458|
|Plaque volume (mm3)||61.2 ± 8.1||81.3 ± 25.1||0.036|
|Plaque volume (%)||69.7 ± 8.5||73.1 ± 4.7||0.301|
Relationship between coronary plaque morphology and clinical characteristics
The results of correlation analysis showed that the plaque volume was significantly and negatively correlated with serum albumin levels in all study patients (r = −0.515, P < 0.05). The volume of necrotic core plaques tended to be correlated with serum phosphate levels (r = 0.412, P = 0.08; Fig. 3a) and Ca × P (r = 0.410, P = 0.08; Fig. 4a). The volume of dense calcium plaques was significantly and positively correlated with serum phosphate levels (r = 0.614, P < 0.01; Fig. 3b) and Ca × P (r = 0.632, P < 0.01; Fig. 4b). In addition, the percentage of necrotic core plaques significantly correlated with serum albumin (r = −0.466, P < 0.05), hsCRP (r = 0.474, P < 0.05), serum phosphate levels (r = 0.436, P < 0.05), and Ca × P (r = 0.486, P < 0.05); and the percentage of dense calcium plaques significantly correlated with serum phosphate levels (r = 0.511, P < 0.05) and Ca × P (r = 0.432, P < 0.05). On the other hand, the correlation between traditional risk factors and plaque morphology was not statistically significant.
Our study demonstrated that: (i) total plaque volume as well as the volume and percentage of dense calcium plaque significantly increased in the HD group compared with the control group; (ii) coronary plaque burden and the number of diseased vessels were greater in the HD group compared with the control group; and (iii) serum phosphate levels were significantly correlated with the amount of necrotic core and dense calcium plaques in all the study patients.
Traditional risk factors for CVD include hypertension, hyperlipidemia, diabetes mellitus, and smoking; however, the causes of CVD in CKD patients are multifactorial. Most dialysis patients already have the traditional risk factors at the initiation of renal replacement therapy. Since most of the present study patients also had such risk factors, it might be difficult to detect a statistically significant relationship between traditional risk factors and plaque morphology; therefore, non-traditional risk factors seem to be more important in CKD patients. Among such non-traditional risk factors, mineral disorders such as hyperphosphatemia, hypercalcemia, and increased Ca × P are especially crucial for CKD patients. These abnormalities are frequently observed in CKD patients and many studies have demonstrated that they are associated not only with bone lesions, but also with soft tissue and vascular calcification (14,15). In addition, it has been reported that previous studies in non-CKD patients demonstrated that increasing serum phosphate levels significantly correlated with coronary artery disease (12,16); however, there are only a few studies that investigated coronary plaque characteristics in CKD patients. Gross et al. examined autopsy samples of coronary arteries using electron microscopy, immunohistochemistry, backscatter imaging, and X-ray analysis (17). They reported that coronary plaque volume was not increased in CKD patients compared with control patients; however, the proportion of calcified coronary plaque was significantly higher in CKD patients than in control patients. Schwarz et al. also demonstrated that coronary plaques were more calcified in the CKD group than in the control group, but the coronary plaque area was comparable between the two groups (3). This study, as well as previous ones (3,17), have demonstrated that coronary plaque calcification was more severe in hemodialysis patients than in control patients; however, in the present study plaque volumes were also greater in hemodialysis patients. We speculated that there were several reasons for this controversy. First, the two previously mentioned studies included non-renal patients as the control group and their detailed renal functions were not defined. In contrast, only patients with normal or mildly impaired renal function (CKD stage 1 or 2; eGFR ≥ 60 mL/min) were included in our study. Second, all patients in the present study had overt coronary artery disease, of which we only analyzed culprit lesions in the coronary artery. Third, our study examined living patients instead of autopsy subjects. Therefore, there is a possibility that the two previous studies might include only subjects with severe coronary artery lesions in both groups and there might be no significant differences of plaque volume between CKD patients and control patients in the previous studies.
It is well known that a diagnosis of CAD is one of the most important prognostic factors in CKD patients. Many studies have reported that most CKD patients have vascular calcification, and CAD events in CKD patients are closely associated with calcification (18–20). Vascular calcification in CKD patients is divided into two major types based on the mechanism of atherosclerotic formation. One mechanism involves atherosclerotic calcification occurring in the intimal layer of the artery. This type of calcification, which tends to progress around a lipid core lesion, involves cellular apoptosis and inflammation. Previous studies have demonstrated that hyperphosphatemia causes endothelial cell apoptosis and vascular calcification (21). The other mechanism, known as Monckeberg sclerosis, occurs in the medial layer of the artery. This type of calcification is often observed in diabetes mellitus, aged, and advanced stage CKD patients. The supposed mechanism of Monckeberg sclerosis involves vascular smooth muscle cells that induce osteogenesis and apoptosis by the uptake of phosphate into cells through sodium-dependent phosphate co-transporters (22). In the present study, serum phosphate and hsCRP levels were significantly and positively associated with necrotic core plaque, whereas the serum albumin level was significantly and negatively associated with necrotic core plaques. Thus, elevated serum phosphate levels, malnutrition, and/or inflammation are suggested to result in the initiation of atherosclerotic calcification in CKD patients.
In the present study we demonstrated that hemodialysis patients had significantly larger volumes of dense calcium and necrotic core plaques than patients with normal or mildly impaired renal function. Furthermore, hemodialysis patients had more severe coronary plaque burden and a greater number of diseased vessels. These results show that HD patients had more severe coronary artery lesions than the control patients.
Taken together, considering that both the absolute and relative volume of necrotic core and dense calcium were correlated with mineral disorders, serum phosphate may affect not only the transformation of coronary plaque, but also the progression of plaque burden.
There are only a few studies that have evaluated coronary plaque morphology in CKD patients. Some previous studies included only autopsy subjects and examined the characteristics of coronary artery plaque using an in vitro histopathological method (3,17). Recently, sophisticated radiological techniques, such as electron-beam computed tomography and multidetector computed tomography have been used in many studies to quantify vascular calcification (23–26). Though there are many studies using such techniques, it is impossible to identify the detailed characteristics of plaque with their use. On the other hand, VH–IVUS provides detailed information regarding plaque components with a high predictive accuracy. Furthermore, VH–IVUS can be performed in living patients and is a useful modality for real-time characterization of clinically relevant plaque components (27). This newly developed imaging tool enabled us to clarify that living hemodialysis patients had not only calcium-rich plaques, but also necrotic plaques.
There are several limitations in this study that must be mentioned. First, although VH–IVUS provides detailed information regarding coronary plaque morphology, it is difficult to differentiate medial from intimal calcification by this technique; we hope to develop new software in the near future. Second, this study included only those patients who had normal or mildly impaired renal function and were on hemodialysis; therefore, in the present study it was impossible to determine whether coronary calcification increases linearly as renal function decreases. A previous study using VH–IVUS has demonstrated that coronary calcification increases with decreasing renal function (28). In addition, we could not perform multivariate analysis because the number of study patients was relatively small; however, this study clearly demonstrated the differences in coronary plaque morphology between living control patients and hemodialysis patients. We believe that this information is valuable. Third, we could not evaluate other key factors, such as fetuin-A, several cytokines, pyrophosphate, or matrix Gla protein in terms of their relationship with coronary plaque morphology. We also could not measure intact parathyroid hormone levels in the control group. Further studies are necessary to resolve these issues.
Our results indicate that the amount of necrotic core and dense calcium plaques increased significantly in hemodialysis patients. In addition, we speculate that disordered mineral metabolism, mainly serum phosphate, may play an important role in the pathophysiology of CAD. Further studies are required to explore a specific strategy for the prevention of coronary artery calcification in CKD patients.