We previously reported that arterial micro-calcification (AMC) of vascular access was common and closely associated with early failure of vascular access in hemodialysis (HD) patients (1). We also reported that early vascular access failure was associated with cardiovascular mortality in HD patients (2). Although vascular calcification seen on radiologic studies such as aortic and coronary artery calcification has been known as an independent risk factor for cardiovascular morbidity and mortality, the effect of AMC of vascular access has not been reported yet. Vascular calcification is a complication of advanced arteriosclerosis and the presence or extent of vascular calcification is an important predictor of cardiovascular outcomes (3). There was a positive correlation between vascular calcification and arterial stiffness evaluated by aortic pulse wave velocity (PWV), and an inverse correlation between coronary artery calcification and endothelial function by flow-mediated dilatation (FMD) (4,5). Moreover, decreased FMD was related to increased PWV (6). Vascular calcification, PWV and FMD were significantly correlated with each other and measurements of these parameters were good surrogate markers of clinical arteriosclerosis. This study was performed to evaluate the impact of AMC of vascular access by histological evaluation on aortic arch calcification, arterial stiffness and endothelial dysfunction, which are risk factors for cardiovascular disease in HD patients.
Vascular calcification of the coronary arteries or aorta is an independent risk factor for cardiovascular outcome, but clinical significance of arterial micro-calcification (AMC) of vascular access is unclear in hemodialysis (HD) patients. Sixty-five patients awaiting vascular access operation were enrolled. We compared surrogate markers of cardiovascular morbidity such as aortic arch calcification (AoAC) by chest radiography, arterial stiffness by brachial-ankle pulse wave velocity (baPWV) and endothelial dysfunction by flow-mediated dilatation (FMD) between patients with and without AMC of vascular access on von Kossa staining. AMC of vascular access was detected in 36 (55.4%). The AMC-positive group had significantly higher incidence of AoAC (63.9% vs. 20.7%, p < 0.001) and higher baPWV (26.5 ± 9.4 m/s vs. 19.8 ± 6.6 m/s, p = 0.006) than the AMC-negative group. There was no significant difference in FMD between the two groups (5.4 ± 2.6% vs. 5.7 ± 3.5%, p = 0.764). The AMC-positive group had higher incidence of diabetes mellitus, higher systolic blood pressure and wider pulse pressure than the AMC-negative group. This study suggests that AMC of vascular access may be associated with cardiovascular morbidity via AoAC and arterial stiffness in HD patients.
Materials and Methods
Informed consent was obtained from all patients and this study was approved by the Institutional Review Board of Uijeongbu St. Mary’s Hospital, Uijeongbu-city, Korea. This retrospective study enrolled 65 patients awaiting vascular access surgery between April 2008 and April 2010 in Uijeongbu St. Mary’s Hospital. Of the 65 patients, 55 were undergoing HD via a central venous catheter and 10 were in a predialysis state. Primary renal diseases included 46 (70.7%) patients with diabetic nephropathy, 12 (18.5%) with hypertensive nephropathy, 4 (6.2%) with chronic glomerulonephritis, and 3 (4.6%) with other diseases.
Laboratory and Radiologic Study
Before the vascular access operation, we measured serum blood urea nitrogen (BUN), creatinine, lipid profile (total cholesterol, triglyceride, and high density lipoprotein (HDL) cholesterol), calcium, phosphate, intact PTH (iPTH), and high sensitivity C-reactive protein (hsCRP). Glycosylated hemoglobin (HbA1c) was measured in patients with diabetes mellitus (DM). Dyslipidemia was defined as total cholesterol >200 mg/dl, triglyceride >150 mg/dl or HDL cholesterol <40 mg/dl (for males, <50 mg/dl for females) after fasting for 12 hours, or use of lipid-lowering medication according to the NCEP ATP III guidelines (7).
We investigated gross calcification of the aortic arch and radial artery with chest and wrist plain X-rays. The aortic arch calcification (AoAC) score was determined by a specific scale according to the method of Ogawa et al. (8). to measure the severity of AoAC. The scale, which is divided into 16 circumferences, was attached to the aortic arch on chest X-ray and then the number of sectors with calcification was divided by 16. The AoAC score was multiplied by 100 to express it as a percentage. The presence of AoAC per se is known as an independent risk factor for cardiovascular events, therefore the AoAC score higher than 0% is considered abnormal.
Pulse Wave Velocity (PWV)
Brachial-ankle PWV (baPWV) measurement were performed with the patient in the supine position. PWV was determined using an automated device, a Colin Waveform analyzer (PWV/ABI, Colin Co. Ltd., Komaki, Japan), which measures the pulse volume wave forms of the brachial and tibial arteries by connecting to a plethysmographic sensor. The distance of each segment (Lb–La) was calculated based on the patient’s height and the time delay from the ascending point of the right brachial waveform to the ascending point of each ankle waveform (ΔTba) was also automatically determined. The baPWV was calculated as the pulse wave propagation distance (Lb–La) divided by the pulse wave propagation time (ΔTba) and expressed in m/second. The normal range of the baPWV was defined as less than 12 m/second (9).
Endothelial function was determined by recording the dilator response of the brachial artery to increased blood flow during reactive hyperemia of the downstream forearm. We measured FMD of the brachial artery according to guideline by Mary et al. (10). Patients were asked to lie supine with the arm in a comfortable position for imaging the brachial artery. A pressure cuff was placed above the antecubital fossa of the nondominant arm and inflated to at least 50 mmHg above systolic pressure for 5 minutes to induce ischemia. Longitudinal images of the radial artery were recorded continuously using a 10 MHz transducer (Voluson I: GE Healthcare, Milwaukee, Wisconsin, USA) from 30 seconds before to 2 minutes after cuff deflation. The brachial artery diameter was measured at baseline (D0) and 1 minute after cuff deflation (D1, maximum diameter). FMD was calculated as (D1 − D0)/D0 × 100 and expressed as a percentage. The normal range for FMD was defined as ≥10% (11).
Arterial Micro-Calcification of Vascular Access
We acquired arterial specimens by the same method as previously reported (12). During the vascular access operation, 5 mm long partial arterial specimens were obtained from the incision sites of the artery in an elliptical form, which were fixed in formalin and embedded in paraffin blocks. They were stained with hematoxylin and eosin (H&E) and von Kossa. The slides were examined by an experienced pathologist blinded to the clinical data. AMC was defined as the presence of von Kossa-positive areas. We divided the patients into two groups according to the presence or absence of AMC of vascular access and then compared clinical and laboratory findings, radiologic findings including aorta and radial artery, baPWV, and FMD.
Continuous variables were reported using means and standard deviations (SD). Categorical variables were described as counts and percentages. Two continuous variables were compared using Student’s t-test in case of a normal distribution. The chi-squared test or Fisher’s exact test were used to compare categorical variables. Univariate analyses of associations between AMC of vascular access and clinical or biochemical profiles were performed. A multivariate regression analysis was done to identify risk factors for AMC of vascular access. A p-value less than 0.05 was considered statistically significant.
Clinical Characteristics of the Subjects
Mean age of the patients was 60.5 ± 12.8 years and the number of males was 40 (61.5%). The incidence of DM was 70.7%. Other demographic and laboratory findings are summarized in Table 1. Types of vascular access were native arteriovenous fistula (AVF, 89.2%) and arteriovenous graft (AVG, 10.8%). AVF consisted of radiocephalic AVF (n = 46, 79.3%) and brachiocephalic AVF (n = 12, 20.7%).
|Total, n = 65||AMC-positive, n = 36||AMC-negative, n = 29||p|
|Age (years)||60.5 ± 12.8||62.4 ± 9.5||58.2 ± 15.8||0.194|
|Old age (>65 years), n (%)||26 (40.0)||15 (41.6)||11 (37.9)||0.760|
|Men, n (%)||40 (61.5)||21 (58.3)||19 (65.5)||0.554|
|DM, n (%)||46 (70.7)||33 (91.6)||13 (44.8)||<0.001|
|Hypertension, n (%)||53 (81.5)||31 (86.1)||22 (75.8)||0.290|
|BMI (kg/m2)||24.3 ± 3.5||24.4 ± 3.7||24.1 ± 3.2||0.770|
|Obesity, n (%)||42 (64.6)||25 (69.4)||17 (58.6)||0.364|
|Dyslipidemia, n (%)||46 (76.6)||27 (81.8)||19 (70.3)||0.297|
|BUN (mg/dl)||45.9 ± 19.4||44.7 ± 13.8||47.5 ± 24.8||0.559|
|Creatinine (mg/dl)||5.8 ± 2.4||5.4 ± 1.6||6.3 ± 3.0||0.170|
|Total cholesterol (mg/dl)||166.0 ± 46.9||166.2 ± 52.4||165.9 ± 40.3||0.982|
|Triglyceride (mg/dl)||155.3 ± 127.8||145.3 ± 103.7||167.5 ± 153.5||0.507|
|hsCRP (mg/l)||2.6 ± 3.2||2.5 ± 3.3||2.6 ± 3.2||0.897|
|HbA1c (%)||6.9 ± 1.3 (n = 32)||7.2 ± 1.4 (n = 21)||6.4 ± 0.9 (n = 11)||0.125|
|Calcium (mEq/l)||8.1 ± 0.9||8.1 ± 0.7||8.0 ± 1.1||0.638|
|Phosphate (mEq/l)||4.3 ± 1.6||4.3 ± 1.4||4.2 ± 1.8||0.845|
|CaP product (mEq/l)2||34.9 ± 13.3||35.7 ± 12.7||33.8 ± 14.1||0.576|
|iPTH (pg/ml)||165.9 ± 110.4||140.6 ± 84.4||199.1 ± 132.0||0.045|
AoAC, baPWV and FMD
AoAC was detected in 29 of 65 patients (44.6%) and the mean AoAC score in these patients was 27.4 ± 13.6%. Of the 65 patients, baPWV and FMD values were measured in 52 and 38 patients, respectively. Mean baPWV was 23.5 ± 8.9 m/second (11.8–61.4 m/second) and baPWV was increased in all but one of the patients (n = 51, 98.0%). Mean FMD was 5.1 ± 2.5% (0.6 to 10.1%) and abnormal FMD was also detected in all except one of the patients (n = 37, 97.3%).
Gross and Micro-Calcification of Vascular Access
Plain wrist X-rays were performed in 27 patients and radial artery calcification was seen in five patients (18.5%) (Fig. 1). Of the five patients with gross calcification of the radial artery, AMC of vascular access was observed in four patients and there was one false-negative case without AMC despite definite gross calcification. AMC was detected in 45.4% (n = 10) of 22 patients without gross calcification of the radial artery. Arterial specimens were acquired from the radial artery (n = 46, 70.7%) and brachial artery (n = 19, 29.3%). Of the 65 patients, AMC of vascular access by von Kossa staining was observed in 36 patients (55.4%). AMC was observed in only the medial layer of the arterial wall and not seen in the intimal layer. Of the 36 patients with AMC, multiple spotted calcification and diffuse calcification were observed in 10 (15.4%) and 21 (32.3%) patients, respectively (Fig. 2).
Comparison of Clinical and Laboratory Findings According to the Presence of AMC
The AMC-positive group had a higher incidence of DM (91.6% vs. 44.8%, p = 0.001) and a lower level of iPTH (140.6 ± 84.4 pg/ml vs. 199.1 ± 132.0 pg/ml, p = 0.045), than the AMC-negative group. There was no significant difference in age, gender, BMI, obesity, and dyslipidemia between the two groups. In a subgroup analysis, we divided patients into DM and non-DM group and compared iPTH levels according to the presence of AMC in each group. There was no significant difference in iPTH between the AMC-positive and -negative groups in both DM (n = 46, 138.6 ± 86.8 pg/ml vs. 170.7 ± 83.0 pg/ml, p = 0.824) and non-DM groups (n = 19, 160.9 ± 62.7 pg/ml vs. 225.4 ± 164.2 pg/ml, p = 0.090).
Comparison of AoAC, baPWV and FMD According to the Presence of AMC
AoAC by plain chest radiography was detected in 63.9% of the AMC-positive group and 20.7% of the AMC-negative group (p < 0.001). The AoAC score by specific scale with 16 circumferences of the AMC-positive group was significantly higher than that of the AMC-negative group (17.8 ± 17.1% vs. 5.1 ± 12.3%, p = 0.001) (Table 2). The baPWV value was significantly higher in the AMC-positive group than the AMC-negative group (26.5 ± 9.4 m/second vs. 19.8 ± 6.6 m/second, p = 0.006) (Fig. 3). But the difference in FMD values between the two groups was not significant (5.4 ± 2.6% vs. 5.7 ± 3.5%, p = 0.764) (Fig. 4).
|AMC-positive, n = 36||AMC-negative, n = 29||p|
|Systolic BP (mmHg)||156.4 ± 23.7||143.1 ± 25.0||0.036|
|Diastolic BP (mmHg)||89.0 ± 18.3||87.7 ± 14.6||0.772|
|Pulse pressure (mmHg)||67.4 ± 19.5||55.3 ± 17.6||0.014|
|Positive AoAC, n (%)||23 (63.9)||6 (20.7)||0.001|
|AoAC score (%)||17.8 ± 17.1||5.1 ± 12.3||0.001|
|baPWV (m/second)||26.5 ± 9.4 (N = 29)||19.8 ± 6.6 (N = 23)||0.006|
|FMD (%)||5.7 ± 3.5 (N = 21)||5.4 ± 2.60 (N = 17)||0.764|
Evaluation of Risk Factors for AMC of Vascular Access
Univariate analysis was performed to identify risk factors associated with AMC. The presence of DM, systolic blood pressure, pulse pressure, iPTH, the presence of AoAC, and baPWV were possible risk factors. Multivariate analysis of these risk factors was performed to determine which factors might be associated with AMC of vascular access. As shown in Table 3, only DM was an independent risk factor associated with AMC of vascular access.
|Systolic blood pressure||0.020||0.004||0.922|
|Presence of AoAC||0.229||0.138||0.105|
It is clinically difficult to obtain an arterial specimen without an operation; therefore, there are limited studies on histological evaluation of vascular calcification. However, it is easy to acquire an arterial specimen from HD patients because most of the patients undergo a vascular access operation. We previously reported that an arterial biopsy during the vascular access surgery did not influence the vascular access patency (13). We undertook this study to identify the relationship between AMC of vascular access by histological evaluation and cardiovascular outcomes by investigating the association between AMC of vascular access, aortic stiffness and endothelial dysfunction. This is the first study to demonstrate the value of assessing AMC of vascular access to predict cardiovascular morbidity in ESRD patients.
Wang et al. (14) reported that the incidence of AMC of vascular access accounted for 36.6% of HD patients. In this study, it was found in 54.5% of the patients. Although there was a difference in incidence depending on biopsy sites, a higher incidence of AMC in this study might be due to the high incidence of DM at a rate of 70.8%. Vascular calcification is known to be more common and more severe in DM patients (15). DM patients had a higher incidence of AMC than non-DM patients in this study. Although patients with AMC had lower levels of iPTH than those without, in subgroup analysis of DM patients, there was no significant difference in iPTH levels between the AMC-positive and negative groups. In multivariate analysis, the presence of DM was the only significant risk factor for AMC of vascular access. Therefore, we postulate that DM of the pro-calcification factors had a critical role in the development of AMC in this study.
Schlieper et al. (16) suggested that vascular access calcification on plain X-ray was seen in 23% and it was a predictor of mortality in HD patients. In this study, gross calcification of vascular access site was seen in 18.5% and there was no significant association between gross calcification at vascular access site and baPWV, or FMD. However, the number of patients who had a wrist X-ray was only 27 of the 65 patients, so further studies will be required. In our opinion, biopsy at a vascular access site during surgery is technically simple and not risky. Moreover, it allows early detection and more precise information of the arteriosclerosis status of patients, compared with plain radiography. In this study, AMC was detected in about a half of patients without gross calcification of vascular access site.
Aortic arch calcification is significantly associated with cardiovascular disease, which is the leading cause of death in ESRD patients (17,18). Assessment of aortic arch calcification on plain chest radiography is easily performed and useful, because its grade reflects the magnitude of calcified change in the whole aorta and is highly correlated with aortic arch calcification volume on computed tomography (CT) (8,19). Vascular calcification induces arterial wall stiffness and reduces vascular compliance, which is associated with an increased left ventricular afterload and hypertrophy (20). Therefore, many previous studies have demonstrated a strong association between vascular calcification and pulse wave velocity, which serves as useful tool for cardiovascular risk assessment in ESRD patients as well as in the general population (4,21–23). Our data show that patients with AMC of vascular access have a greater average systolic blood pressure, pulse pressure, AoAC score and baPWV than those without.
Recent studies showed a direct relationship between vascular calcification and endothelial dysfunction. Decreased serum fetuin-A, the major inhibitor of vascular calcification, was associated with the development of endothelial dysfunction in patients with CKD (24). Another study showed that the use of sevelamer, which is a calcium-free phosphate binder and can help to reduce vascular calcification, led to a significant increase in serum fetuin-A concentrations with improvement of endothelial function in CKD patients (25). In addition, endothelial function was also inversely correlated with coronary artery calcification by multidetector CT in patients with coronary risk factor (5).
However, in this study, AMC of vascular access was not related to endothelial dysfunction. Vascular calcification may be one of the important mechanisms involved in endothelial dysfunction. Endothelial dysfunction, characterized by decreased bioavailability of nitric oxide, is an early event in arteriosclerosis and is observed even as early as stage 1 CKD patients (24), therefore, it should be further aggravated in ESRD patients. We assumed that there was a relative lack of influence of the degree of vascular calcification on endothelial function in advanced arteriosclerosis status in ESRD patients. Almost all patients in this study showed abnormal PWV and FMD, which must be considered as a significant sign of advanced arteriosclerosis. Kopec et al. (26) demonstrated that endothelial dysfunction is associated with the aortic stiffness and that this association is limited by the progression of atherosclerosis. The correlation between baPWV and FMD was not significant in patients with significant coronary artery stenosis (26). Kobayashi et al. (6) reported similar results, there was no correlation between endothelial dysfunction and arterial stiffness in the lowest tertile of the FMD group. As a result, this reverse correlation between arterial stiffness (increased PWV) and endothelial function (decreased FMD) may be limited in advanced arteriosclerosis status such as ESRD patients.
In our results, patients with AMC of vascular access had lower levels of iPTH, compared with patients without AMC. These results are consistent with data, suggesting that low-turnover bone disease, characterized by a low iPTH level (<150 pg/ml) is more likely to be associated with vascular calcification, partly due to the decreased calcium phosphate-buffering capacity of bone (27–29). However, the incidence of low iPTH was not significantly different between patients with and without AMC (60.6% vs. 48.0%, p = 0.339). In contrast to previous studies (30), we found no association between mineral metabolism and AMC of vascular access. Chronic mineral dysregulation could contribute to the progression of vascular calcification. There was no difference in serum calcium and phosphate levels between patients with and without AMC, and the mean serum mineral concentrations were almost within the normal range, which may explain the lack of a significant relationship between mineral metabolism and AMC of vascular access in this study.
The duration of dialysis is also known to be a strong determinant of vascular calcification (31,32). In this study, 55 (84.6%) of 65 patients started unplanned HD with a central venous catheter. The mean duration of HD before the operation for vascular access was 28.8 days, and there was no difference of HD duration between the two groups (26.7 days vs. 31.1 days, p = 0.690). Having a HD duration of less than a month might be a lesser influence on vascular calcification.
In conclusion, this study demonstrates that AMC of vascular access is related to aortic arch calcification and arterial stiffness but not to endothelial dysfunction in ESRD patients. Such parameters including aortic arch calcification, PWV and FMD were known as good surrogate markers of clinical arteriosclerosis. This result indicates that AMC of vascular access site reflects, not just about vascular access patency, but also, the degree of arteriosclerosis. The sample size of this study was relatively small and this was not a prospective controlled study. Further studies involving a larger sample size and with follow up of the cardiovascular morbidity and mortality of the patients with AMC of vascular access are required. Nevertheless, our data emphasize that AMC of vascular access may be associated with cardiovascular morbidity and mortality in ESRD patients.
A part of this study was orally presented at the Annual Meeting of the American Society of Nephrology, Denver, Colorado, 2010.
This study was supported by a grant of the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (A102065).