Pediatric Nephrology

, Volume 24, Issue 3, pp 555–563

Progression of coronary calcification in pediatric chronic kidney disease stage 5

Authors

    • Department of Pediatric NephrologyIstanbul University Cerrahpasa Medical Faculty
    • Atakoy 7–8.kısımMimar Sinan Sitesi
  • Salim Caliskan
    • Department of Pediatric NephrologyIstanbul University Cerrahpasa Medical Faculty
  • Sebuh Kurugoglu
    • Department of RadiologyIstanbul University Cerrahpasa Medical Faculty
  • Cengiz Candan
    • Department of Pediatric NephrologyIstanbul University Cerrahpasa Medical Faculty
  • Nur Canpolat
    • Department of Pediatric NephrologyIstanbul University Cerrahpasa Medical Faculty
  • Lale Sever
    • Department of Pediatric NephrologyIstanbul University Cerrahpasa Medical Faculty
  • Ozgur Kasapcopur
    • Department of Pediatric NephrologyIstanbul University Cerrahpasa Medical Faculty
  • Nil Arisoy
    • Department of Pediatric NephrologyIstanbul University Cerrahpasa Medical Faculty
Original Article

DOI: 10.1007/s00467-008-1038-0

Cite this article as:
Civilibal, M., Caliskan, S., Kurugoglu, S. et al. Pediatr Nephrol (2009) 24: 555. doi:10.1007/s00467-008-1038-0

Abstract

Coronary artery calcification (CAC) is common in adults with chronic kidney disease (CKD) and progresses with time. However, data are limited for younger patients. We have previously reported CAC in eight of 53 children with CKD. After 2 years, CAC evaluation was repeated in 48 patients. The median CAC score (CACS) increased from 101.3 (1473.6 ± 1978.6, range 8.5–4332) to 1759.2 (2236.4 ± 2463.3, range 0–5858) Agatston units (AU). When the individual changes in CACS were evaluated one by one, we showed a mild decrease in two patients on hemodialysis (HD) and in one transplant (Tx) recipient, a moderate increase in one patient on HD, one on peritoneal dialysis (PD) and one Tx recipient, and a large increase in one HD patient. Also, CAC disappeared in one HD patient. All patients with no calcification at baseline remained calcification-free at follow-up. To obtain the individual cumulative exposure, we calculated time-averaged mean values, using the laboratory values from the beginning of dialysis to the first and second multidetector spiral computed tomography (MDCT) scans (baseline and final values, respectively). Final CACS was positively related to final calcium–phosphorus (Ca×P) product, while CAC progression was inversely associated with final serum albumin level. This report is the first study with the largest number and the youngest cohort to document the natural history of coronary calcification.

Keywords

ChildYoung adultCoronary artery calcificationChronic kidney diseaseDialysisTransplantation

Introduction

Cardiovascular disease is the leading cause of morbidity and death in patients with chronic kidney disease (CKD). With improvements in renal replacement therapy (RRT), cardiovascular disease is also increasingly recognized as a life-limiting problem in young patients with CKD, giving a 1,000-times higher risk of cardiovascular death than for the healthy age-adjusted population [1].

Coronary artery calcification (CAC) is a common finding contributing to high cardiovascular mortality in adults with CKD. Recently, a few younger patients with CKD have been examined for CAC [26]. Goodman et al. [2] did not detect CAC in patients with end-stage renal disease (ESRD) that were younger than 20 years. In contrast, a small and uncontrolled study by Eifinger et al. [3] showed the presence of CAC in two of 16 patients aged 14–39 years treated with RRT. Thereafter, two studies on young adults with childhood-onset ESRD showed different prevalences of CAC in 92% and 10% of patients [4, 5]. Another recent study, which included the largest cohort of pediatric dialysis patients, showed that evidence of CAC was present as early as the first decade of life [6]. Also, we recently screened 53 children with CKD for the presence and predisposing factors of CAC [7]. Calcification was present in 15% of patients. We showed that levels of serum phosphorus and oral intake of calcium-based phosphate binders (CBPBs) were independent predictors in the development of CAC. After this study, sevelamer and dialysates containing low calcium content were further given to our patients, when appropriate.

Progression of CAC has been commonly documented in several longitudinal studies of adults [814]. To our knowledge, there are no published studies describing the progression of CAC in children with CKD. However, two studies have investigated the long-term consequences of CKD on carotid morphology and function [15, 16]. They showed that CKD-associated arteriopathy rapidly progresses in dialysis patients. In contrast, abolition of the uremic state by renal transplantation leads to stabilization or partial regression of vascular lesions. The aim of our study was to evaluate prospectively the change in CAC over a 2-year period and to identify the factors that might be associated with CAC progression in children and young adults with CKD.

Patients and methods

This study was carried out in a single pediatric nephrology unit. We had previously screened 53 children and adolescents [15 on hemodialysis (HD), 24 on peritoneal dialysis (PD) and 14 renal transplant (Tx) patients] with CKD stage 5 for the presence and predisposing factors of CAC. The baseline characteristics and CAC scores of these patients have been described in detail previously [7]. After 2 years, CAC evaluation was repeated in 48 patients. During the period mentioned, seven PD patients had been transferred to hemodialysis, four patients had died and one had refused a second scan. The causes of death were sepsis (2), pneumonia (1) and peritonitis (1). Five patients who had not undergone a second scan had shown no evidence of calcification at baseline. In our study the patients’ mean age was 17.7 ± 3.6 years (median 18.5 years, range 8.9–23.5 years) and 40% were girls. The primary diagnoses were as follows: hereditary nephropathies (11), reflux nephropathy (10), obstructive uropathies (9), neurogenic bladder (6), glomerulopathies (5), renal hypoplasia or dysplasia (4). The cause of renal failure was unknown in three patients. No patients had diabetes mellitus, inflammatory diseases or vasculitis and symptomatic cardiovascular disease. Five patients (with no CAC at baseline) were smokers.

The blood of all HD patients was dialyzed with standard bicarbonate-containing bath, three times a week for 4 h per session. Among the PD patients, nine were on continuous cycling peritoneal dialysis (CCPD) and six were on continuous ambulatory peritoneal dialysis (CAPD). Transplant recipients had undergone pre-emptive renal transplantation and all had experienced dialysis for a significant period (median 26.8 months). None of the transplant recipients had returned to dialysis. At the time of the investigation, the Tx patients had a mean estimated glomerular filtration rate, calculated from the Schwartz formula, of 86.3 ± 27.2 ml/min per 1.73 m2 body surface area.

All transplant patients were on the triple immunosuppressive protocol of our unit, consisting of cyclosporine or tacrolimus, mycophenolate mofetil or azathiopurine, and corticosteroids. The first dose of cyclosporine was given orally in a dose of 10 mg/kg body weight per day, with whole-blood trough concentrations aimed at 250–350 ng/ml during the first 3 months after transplantation and adjusted thereafter to obtain levels between 100 ng/ml and 200 ng/ml. Tacrolimus treatment was started with a dose of 0.1–0.15 mg/kg per day, divided into two doses and adjusted to maintain therapeutic whole-blood levels between 5 ng/ml and 10 ng/ml. Mycophenolate mofetil was given at the dose of 1,200 mg/m2 per day, divided into two doses, to recipients receiving cyclosporine, and 600 mg/m2 per day, also divided into two doses, to recipients receiving tacrolimus. All patients received methylprednisolone intravenously at a dosage of 4 mg/kg daily on the first 3 postoperative days. Methylprednisolone was tapered from 0.5 mg/kg per day to 0.15 mg/kg per day by postoperative month 3.

Height, weight and body mass index (BMI) (weight/height2) were calculated as standard deviation scores (SDSs) according to the standard growth tables for Turkish children [17].

Serum biochemical parameters were measured monthly for HD patients and every 2 or 3 months for PD and Tx patients during the study period. Blood samples were obtained in the morning after an overnight fast (before the start of a dialysis session for HD patients and at the time of routine visits for PD and Tx patients). Routine serum biochemical variables, including glucose, calcium, phosphorus, hemoglobin, albumin and lipids, were evaluated by standard laboratory methods. Serum levels of intact parathyroid hormone (iPTH) were measured by immunoradiometric assay. Serum levels of high-sensitivity C-reactive protein (hs-CRP) were measured by the nephelometric method (Dade Behring, Germany), and homocysteine levels were determined by fluorimetric high-performance liquid chromatography (HPLC).

The patients’ blood pressures and laboratory values were averaged from the beginning of dialysis to the first multidetector spiral computed tomography (MDCT) scan (baseline time-averaged mean values) and to the second scan (final time-averaged mean values). In addition, the cumulative intakes of CBPB, sevelamer and calcitriol were calculated for each patient until the first scan (baseline cumulative exposure) and between two scans (interval cumulative exposure).

Multidetector spiral computed tomography

MDCT was performed on a retrospective ECG-gating 16-detector model, the Somatom Cardiac Sensation (Siemens, Erlangen, Germany). The data acquisition parameters were: 120 kVp, 400 mAs, nominal slice width 3 mm, gantry rotation time 0.5 s, table feed 5.5 mm/s, (pitch:1). Data were reconstructed with a 180° linear interpolation algorithm providing a temporal resolution of 270 ms, retrospective ECG gating during diastole, 1.5 m longitudinal increment, 512 pixels × 512 pixels, field of view 20 cm2. Data were transferred to a workstation (Wizard). The coronary artery calcification score was calculated by the method described by Agatston et al. [18]: The principal investigator derived this score by placing a 30 mm2 region of interest over each calcified coronary arterial focus with minimal attenuation of 130 HU. A density score of 1 was applied for 130 HU to 200 HU, 2 for 201 HU to 300 HU, 3 for 301 HU to 400 HU, and 4 for >401 HU. Scores were determined for each main epicardial coronary artery, including the left main (LM), left anterior descending (LAD), left circumflex (LCX) and right coronary artery (RCA). The total calcium score was described as the sum of the values of all lesions identified. In order to minimize inter-observer variability, we mandated that the baseline and repeat scans were reviewed by the same radiologist. This investigator was unaware of the patient’s status. The absolute and relative (percent) changes in calcification score (CACS) were calculated as follows:
$$\begin{array}{*{20}c} {{\text{Absolute change = }}{\left( {{\text{CACS}}_{{\text{2}}} - {\text{CACS}}_{{\text{1}}} } \right)}} \\ {{\text{Relative change = 100}} \times {{\left( {{\text{CACS}}_{{\text{2}}} - {\text{CACS}}_{{\text{1}}} } \right)}} \mathord{\left/ {\vphantom {{{\left( {{\text{CACS}}_{{\text{2}}} - {\text{CACS}}_{{\text{1}}} } \right)}} {{\text{CACS}}_{{\text{1}}} }}} \right. \kern-\nulldelimiterspace} {{\text{CACS}}_{{\text{1}}} }} \\ \end{array} $$
where CACS1 and CACS2 are the baseline and second CAC scores, respectively.

Prospective ECG-gating MDCT is superior to retrospective-gating MDCT due to lower patient radiation dose. No doubt, we are aware of this condition. However, tachycardia and/or arrhythmia may develop in pediatric patients during scanning, despite appropriate medication. The image quality is associated with cardiac motion, cardiac arrhythmias and heart rate. It is more effortless to achieve optimal (artifact-free) image quality with the retrospective technique than with the prospective technique. Thus, in our study, we preferred to use retrospective ECG-gating to avoid repeated scans and to achieve optimal imaging.

Our examinations of the patients conformed to good medical and laboratory practices and the recommendations of the Declaration of Helsinki on Biomedical Research involving Human Subjects. This study protocol was approved by the local Research Ethics Committee. Written informed consent was obtained from subjects older than 18 years and from the parents of subjects younger than 18 years.

Statistical analysis

Statistical analysis was performed with the Statistical Package for Social Sciences for Windows, version 15.0 (SPSS Inc; Chicago, IL, USA). Data were expressed as means ± standard deviations (SDs) or medians, depending on the distribution. Comparisons of baseline and final laboratory values, and baseline and interval medications in patients with calcification vs those without, were analyzed by t-tests for continuous variables, Fisher’s exact test for discrete variables, and the Mann–Whitney U test for continuous variables that were not normally distributed. Changes from baseline to end-of-study (final) values within groups were compared with the Wilcoxon signed rank test. To investigate factors that might be associated with final CACS and change in calcification, we used Spearman’s rho correlation analysis. All variables that showed significant association in the univariate analysis were included in a stepwise multiple linear regression analysis. We used two different models to examine the independent predictors of final CACS and of change in calcification (progression of CAC). All probability values were two-tailed. P values of less than 0.05 were accepted as statistically significant.

Results

Forty-eight patients who had undergone two scans (29 male and 19 female) were selected for our report. At the time of the second MDCT scan, the mean age of the patients was 17.7 ± 3.6 years (median 18.5 years, range 8.9–23.5 years). The duration of dialysis was 77.2 ± 32.9 months (median 69.8 months, range 35.6–159.0 months) for patients on dialysis. In transplant recipients, dialysis period before transplantation was 33.8 ± 23.6 months (median 26.8 months, range 7.3–78.2 months) and post-transplantation time was 61.6 ± 17.5 months (median 64.0 months, range 36.1–96.1 months). Table 1 shows baseline and final demographic characteristics, durations of renal replacement therapy and laboratory values for the patients with CAC and those who remained calcification free. The weight-, height- and BMI-SDSs of HD and PD patients did not significantly differ (−4.0 ± 3.0 vs −3.5 ± 2.4; −3.4 ± 2.0 vs −3.4 ± 1.9; −1.2 ± 1.3 vs −1.0 ± 1.2, respectively).
Table 1

Comparison of patients with calcification at baseline and those without (CAC coronary artery calcification, SDS standard deviation score, BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, HDL high-density lipoprotein, LDL low-density lipoprotein, Ca calcium, P phosphorus, iPTH intact parathyroid hormone, hs-CRP high-sensitivity CRP, NS not significant)

Variables

Patients without CAC (n = 40)

Patients with CAC (n = 8)

Pb

Baseline

Final

Pa

Baseline

Final

Pa

Age (years)

15.7 ± 3.6

17.6 ± 3.6

 

16.4 ± 3.7

18.2 ± 3.2

 

NS

Duration of dialysis (months)

43.0 ± 29.3

58.1 ± 34.5

 

80.0 ± 23.2

96.8 ± 27.4

 

0.004

Duration of transplantation (months)

12.0 ± 20.9

19.2 ± 30.4

 

8.3 ± 16.8

13.9 ± 26.6

 

NS

Gender (male/female)

24/16

 

5/3

 

NS

Weight SDS

−3.0 ± 2.5

−3.1 ± 2.2

NS

−5.2 ± 2.4

−6.3 ± 3.7

0.036

0.007

Height SDS

−3.2 ± 1.5

−3.1 ± 1.7

NS

−4.9 ± 1.5

−5.1 ± 1.9

NS

0.011

BMI SDS

−0.7 ± 1.2

−0.9 ± 1.2

0.003

−1.2 ± 1.5

−1.5 ± 1.8

NS

NS

SBP (mmHg)

119 ± 14

120 ± 12

NS

122 ± 12

121 ± 13

NS

NS

DBP (mmHg)

73 ± 9

74 ± 9

NS

77 ± 8

76 ± 9

NS

NS

Glucose (mg/dl)

92.5 ± 8.0

94.6 ± 5.3

0.005

94.0 ± 7.6

94.0 ± 5.5

NS

NS

Albumin (g/dl)

3.9 ± 0.4

3.8 ± 0.4

NS

3.8 ± 0.5

3.7 ± 0.6

NS

NS

Hemoglobin (g/dl)

10.2 ± 1.0

10.7 ± 1.1

<0.001

9.1 ± 1.1

9.9 ± 1.7

0.036

0.049

Total cholesterol (mg/dl)

187 ± 33

180 ± 31

0.007

202 ± 39

188 ± 48

0.036

NS

HDL-cholesterol (mg/dl)

49.7 ± 12.1

49.2 ± 10.2

NS

42.6 ± 5.8

44.1 ± 4.3

NS

NS

LDL-cholesterol (mg/dl)

108 ± 25

103 ± 23

0.016

113 ± 28

103 ± 33

NS

NS

Triglycerides (mg/dl)

131 ± 45

135 ± 43

NS

206 ± 81

184 ± 81

0.017

NS

Serum Ca (mg/dl)

9.5 ± 0.6

9.6 ± 0.5

0.004

9.6 ± 0.6

9.4 ± 0.4

NS

NS

Serum P (mg/dl)

5.4 ± 0.8

5.2 ± 0.8

0.047

7.2 ± 0.9

6.4 ± 0.9

0.012

0.004

Serum Ca×P (mg2/dl2)

51.3 ± 8.1

50.0 ± 8.2

NS

68.8 ± 9.8

59.8 ± 9.0

0.012

0.008

iPTH (pg/ml)

272 ± 133

307 ± 194

NS

542 ± 324

633 ± 368

NS

0.013

hs-CRP (mg/l)

3.7 ± 4.6

2.9 ± 3.6

<0.001

2.7 ± 3.0

3.2 ± 2.7

NS

NS

Homocysteine (µmol/l)

20.3 ± 15.6

18.3 ± 13.0

<0.001

13.2 ± 4.5

13.3 ± 4.2

NS

NS

a Baseline vs final values

b Without CAC vs with CAC, final values

All patients that had no calcification at baseline remained calcification-free at follow-up. For only those patients with calcification at baseline, CACS had increased from 101.3 AU (range 8.5–4,332 AU, mean 1,473.6 ± 1,978.6 AU) to 1,759.2 AU (range 0–5,858 AU, mean 2,236.4 ± 2,463.3 AU) after 2 years. The absolute change in mean values was 862.8 AU, and the relative increase was 59% for CAC.

We had determined CAC in eight patients (3 on HD, 3 on PD and 2 Tx recipients) at baseline. The individual changes in CACS are shown in Fig. 1. Calcification remained in seven patients and had disappeared in one patient (patient 7) after 2 years. Two dialysis patients (patients 5 and 6) and one transplant recipient (patient 8) had a mild decrease, while two dialysis patients (patients 2 and 3) and one transplant recipient (patient 4) had a moderate increase in their CACS. Of interest, one dialysis patient (patient 1) showed a large increase in CACS over time (from 36.7 AU to 5,099 AU). This was the youngest patient with CAC (12.8 years old). He had been treated by PD for 115 months and had been transferred to HD 3 months before having the second scan.
https://static-content.springer.com/image/art%3A10.1007%2Fs00467-008-1038-0/MediaObjects/467_2008_1038_Fig1_HTML.gif
Fig. 1

The individual changes in coronary artery calcification with time

The cumulative drug intakes at baseline (until the first scan) and in the interval (between the two scans) are shown in Table 2. Among the dialysis patients, patient 3 (CACS 8.5 AU → 138 AU) was treated only with calcium, and patient 1 (CACS 36.7 AU → 5099 AU) was treated with either calcium (18 months) or sevelamer (6 months) after the first scan. Another four dialysis patients had been moved from calcium to sevelamer during the interval period.
Table 2

Comparison of medications at baseline and during the interval period (CBPB calcium-based phosphate binder, NS not significant)

Medication

Patients without CAC (n = 40)

Patients with CAC (n = 8)

Pb

Baseline

Interval

Pa

Baseline

Interval

Pa

CBPB intake (elemental calcium, g/kg)

19.5 ± 18.1

11.9 ± 12.8

<0.001

73.4 ± 38.5

9.1 ± 9.1

0.012

NS

Sevelamer intake (g/kg)

3.3 ± 13.9

13.7 ± 26.5

0.013

11.2 ± 15.8

63.4 ± 71.8

0.043

0.018

Calcitriol intake (ng/kg)

3.6 ± 4.2

6.9 ± 8.5

0.040

16.3 ± 8.1

9.4 ± 8.4

NS

NS

a Baseline vs interval values

b Without CAC vs with CAC, interval values

The final CACS and the change in CACS were associated with several variables in univariate analyses (Table 3). Stepwise multiple linear regression analysis (Table 4) revealed that the final time-averaged mean Ca×P product was independently associated with final CACS (β coefficient = 0.880, P = 0.004) and final time-averaged mean serum albumin level was the only independent predictor of change in CACS (β coefficient = −0.811, P = 0.009).
Table 3

Spearman’s correlation analysis (only significant correlations shows) (CACS coronary artery calcification score, Ca calcium, P phosphorus, LDL low-density lipoprotein)

Parameter 

r

P

Final CACS

 Baseline glucose

0.707

0.025

 Baseline phosphate

0.810

0.007

 Final phosphate

0.667

0.035

 Baseline Ca×P product

0.810

0.007

 Final Ca×P product

0.643

0.043

CAC progression (ΔCACS)

 Baseline total cholesterol

0.738

0.018

 Baseline LDL-cholesterol

0.643

0.043

 Final LDL-cholesterol

0.738

0.018

 Final albumin

−0.707

0.025

Table 4

Independent predictors of final coronary artery calcification score (CACS) and CAC progression (stepwise linear regression analysis) (Ca calcium, P phosphorus)

Dependent variable

Independent variable

β coefficient

Adjusted R2

P

Final CACS

Final Ca×P product

0.880

0.736

0.004

CAC progressiona

Final albumin

−0.811

0.601

0.009

a Change in coronary artery calcification (ΔCACS)

Discussion

Patients with chronic kidney disease are at increased risk for both morbidity and death from cardiovascular diseases. In these patients accelerated atherosclerosis, left ventricular hypertrophy, arrhythmias and CAC contribute to high cardiovascular morbidity [1, 1921]. CAC is significantly more common in adults and young adults with CKD than in healthy subjects of the same age and gender [24, 2224]. We screened for CAC 53 children and adolescents with CKD stage 5 2 years ago and demonstrated the presence of CAC in eight of them (3 on HD, 3 on PD and 2 Tx recipients). Our patients with CAC had undergone dialysis for longer, and had higher values of time-integrated serum phosphorus, calcium- phosphate product, iPTH and cumulative CBPB intake [7]. Our results confirmed the previous observations and extended to childhood. In our study described now, coronary calcification was re-evaluated in 48 patients 24 months after their first scan.

All our patients with no calcification at baseline remained calcification-free at follow-up, as in some publications on adults [8, 9]. In our study final time-averaged mean levels of serum phosphorus, calcium–phosphate product, total cholesterol and low-density lipoprotein (LDL)-cholesterol were lower, and serum hemoglobin level was higher, than baseline time-averaged mean levels. These findings could have been the result of better treatment of the patients during the interval period. Unfortunately, the duration of dialysis is generally long for patients with CKD stage 5 in Turkey, since the number of renal transplants is not sufficient. Moreover, sevelamer and dialysates that contain low calcium content have only recently become available in our country. Thus, our patients had been exposed mostly to dialysates containing high calcium levels and CBPBs until their baseline MDCT scans, while sevelamer use was significantly higher during the interval period (Table 2). Consequently, we suggest that the higher use of sevelamer as a phosphate binder could have prevented the development of new calcification.

The pathogenesis of CAC in CKD stage 5 is most probably multifactorial [1929]. The mechanisms include either stimulation of calcium phosphate precipitation into the vessel or a decrease in the inhibitory process that prevents precipitation [21]. However, increased Ca×P product or hyperphosphatemia may be the key promoter of calcification. Inhibitors of calcification, such as fetuin-A, osteoprotegerin and matrix Gla protein, play a crucial role in the prevention of a serum calcium phosphate complex. Race and/or genetic factors may also influence the development of CAC. Another promoter of vascular calcification is 1,25-dihydroxycholecalciferol [1,25(OH)2 D3], which may have a direct effect on calcium deposition in vascular smooth muscle cells. Some papers have noted a causal relationship between calcitriol and vascular calcification in children with CKD [6, 28, 30]. Several observational studies have reported improved clinical outcomes for patients receiving vitamin D therapy compared with those who did not [3133]. Some evidence also exists that vitamin D2 analogues may be associated with better outcomes than calcitriol [3436]. In recent papers on vitamin D levels and vascular outcomes, London et al. [37] indicated that the majority of patients with ESRD are vitamin D deficient and that this deficiency is independently associated with abnormal function of large arteries. In contrast, Shroff et al. [38] reported that there is a bimodal association of vitamin D levels and vascular disease in children on dialysis. In their study, both low and high 1,25(OH)2 D levels were associated with adverse morphological changes in large arteries.

During the past few years, several observations have heightened interest in the progression of CAC. The natural history of coronary calcification has been commonly documented in adults with CKD [814]. However, to date, there are no published studies describing the progression of CAC in childhood. Ours is the first investigation to determine the progression of CAC in children and young adults with CKD. In good agreement with studies of adults, there was an increase in calcification of the coronary arteries in our population during the 2 years. Unfortunately, the subjects, methods, imaging procedures and statistical analyses are significantly different in the previous studies, and the reported absolute changes ranged between −46 AU and 637 AU, while relative rates of progression ranged between −45% and 104% [814]. In our patients, the absolute change was 862.8 AU and the relative increase was 59% for CAC. One patient (patient 1) showed a large increase (from 36.7 AU to 5,099 AU) in CACS.

When the individual CACS changes were evaluated one by one, the final CACS decreased in three patients on dialysis (patients 5, 6 and 7). Furthermore, CAC disappeared in one of them. We speculated that greater use of sevelamer might have contributed to a potential benefit on the CAC process in these patients. A few studies have demonstrated that progressive CAC can be largely abolished by the use of sevelamer in place of CBPBs [11, 12, 25, 39, 40]. Also, the CACS increased in our three patients on dialysis (patients 1, 2 and 3). The first patient was the youngest one with CAC (12.8 years old) and he showed a large increase in CAC score over time. He had been treated by PD for 115 months and had been transferred to HD 3 months before having the second scan. Two possible explanations may be suggested for the dramatic increase in CACS in this patient: long-term calcium use during the interval period and/or a genetic predisposition to CAC progression. The second patient was a 15-year-old girl who had undergone dialysis for 124 months (97 months on PD and 27 months on HD). She had uncontrolled hypertension, very high levels of time-integrated Ca×P product, severe hyperparathyroidism and dyslipidemia due to noncompliance with her medication. The third patient showed a moderate increase in CACS and had been on PD for 79 months. Unfortunately, the National Insurance Foundation of Turkey does not allow the prescription of sevelamer to PD patients. Thus, we had to give CBPBs to patients 1 and 3. This could have been a probable factor for progression of CAC in both patients.

Our patients with CAC could not be randomly chosen to receive calcium or sevelamer, due to the small sample size. When the use of calcium or sevelamer was evaluated on an individual patient basis, patient 3 (CACS 8.5 AU→138 AU) had been treated only with calcium and patient 1 (CACS 36.7 AU→5,099 AU) had been treated with calcium for a long time (18 months) after the first scan. Another four dialysis patients had been moved from calcium to sevelamer during the interval period. Among patients who had ingested sevelamer, the CACS decreased in three patients and increased in one patient. On the basis of these findings, we hypothesized that sevelamer would be less likely to lead to progressive calcification than would calcium-based, orally administered, phosphate binders.

Renal transplantation appears to slow down or arrested the calcification process in most patients [8, 9]. In our study, CAC had been detected in two transplant recipients at baseline. At follow-up, CACS had decreased mildly in one patient (patient 8) and increased greatly in another patient (patient 4). We could not suggest a clear explanation for the latter case. Some studies on adults have shown that CAC progresses after transplantation [4142]. Immunosuppressive drugs contribute importantly to cardiovascular risks such as CAC, left ventricular hypertrophy, endothelial dysfunction and increased thickness of the carotid intima media in transplant recipients, because of their hypertensive and hyperlipidemic effects [4044]. Furthermore, some papers have reported that cardiovascular morbidity is more prominent in patients on cyclosporine treatment and that tacrolimus may preferable to cyclosporine for optimization of the cardiovascular risk profile [44, 45]. However, our two transplant recipients with CAC were on cyclosporine treatment.

The exact mechanism of CAC progression is not yet completely understood in patients with CKD. Several longitudinal studies have documented the risk factors associated with progression of CAC in adults with CKD [814]. The triggers for CAC progression include a number of uremia-related factors such as duration of dialysis and/or time since transplantation, pressure and volume overload, hyperphosphatemia, high Ca×P product, hyperparathyroidism, dyslipidemia, malnutrition, chronic inflammation, serum levels of calcification inhibitors, and medications(vitamin D, calcium, sevelamer, immunosuppressive drugs). In our study progression of CAC was associated with baseline levels of total cholesterol, baseline and final levels of LDL-cholesterol, and final serum albumin levels in univariate analysis. In several studies of adults with CKD, dyslipidemia, such as increased triglyceride level, total cholesterol and LDL-cholesterol levels or decreased high-density lipoprotein (HDL)-cholesterol level, has been implicated in the progression of CAC [25, 39, 46, 47]. The results of the Treat to Goal Study [11] suggest that the lowering of LDL-cholesterol level in adults with ESRD may also result in amelioration of the progression of CAC. The progression of CAC can be decreased with LDL-cholesterol reduction by statins and possibly sevelamer [39].

In stepwise regression analysis, the only significant independent predictor of final CACS was the final Ca×P product. This finding confirmed the notion that increased Ca×P product is the key promoter of vascular calcification. Also, the final time-averaged mean level of serum albumin was inversely correlated with change of CACS. Serum albumin has been identified as a mediator of vascular damage, and an inverse correlation between serum albumin and progression of vascular calcification in adults with ESRD has been reported [48, 49].

Serum albumin level is commonly used as a marker of nutritional status. Besides inadequate dietary protein intake and increased protein losses, impaired protein assimilation (digestion and absorption) may contribute to protein malnutrition in CKD [50]. However, the definition and assessment of malnutrition have not been standardized for children with CKD. Traditional measures such as height, weight, BMI and serum albumin concentration may not be accurate indicators to assess the nutritional status of children. It is a pity that, in our study, the patients had not been examined for more specific anthropometric measures, such as mid-arm circumference and normalized protein catabolic rate (nPCR), as better markers of nutritional status. Also, hypoalbuminemia reflects a possible increase in inflammatory mediators. Recently, the role of malnutrition and chronic inflammation in vascular damage has been increasingly recognized. Indeed, chronic inflammation can induce an atherogenic and catabolic state, while low food intake can lead to malnutrition and inflammation, but can also result from it. The exact mechanisms responsible for the interplay between the entities of the so-called malnutrition–inflammation–atherosclerosis (MIA) syndrome are still being discovered [5052]. In our study, final serum albumin level was significantly correlated with BMI, but we could not detect a relationship between malnutrition (defined either by BMI or biochemical markers such as albumin) and inflammation (defined by hs-CRP). There may be several reasons why a state of chronic inflammation may promote vascular calcification. For example, fetuin-A is down regulated during inflammation [13]. Unfortunately, values for fetuin-A levels were not available in this study. Recent evidence suggests that each component of MIA syndrome may be an important contributor to vascular calcification [5153]. Also, the syndrome is associated with high cardiovascular mortality and accounts for most of the premature deaths in adult dialysis patients [54].

In conclusion, ours is the first study with the largest number and the youngest cohort to document the change in CKD-related coronary calcification. In this preliminary study we showed that there was a trend towards an increase in the young generation CKD patients with time, as in adults. Another important result was that the patients with no calcification at baseline remained calcification-free, suggesting the presence of protective factors such as higher use of sevelamer and/or probably genetical factors and higher levels of calcification inhibitors. Since most of the patients without CAC remained calcification-free at follow-up, we do not recommend routine screening for CAC by MDCT in all pediatric CKD patients, due to high radiation exposure. Levels of calcification inhibitors, fetuin-A, osteoprotegerin and matrix Gla protein can help to identify patients at risk of CAC. It is possible that patients with lower levels of these proteins will be screened by MDCT in future. Further clinical trials are warranted to confirm this idea.

Acknowledgments

Our research was supported by the Turkish Pediatric Association.

Copyright information

© IPNA 2008