Phosphate Binder Impact on Bone Remodeling and

Original Paper Nephron Clin Pract 2008;110:c273–c283 DOI: 10.1159/000170783

Received: April 1, 2008 Accepted: July 16, 2008 Published online: November 12, 2008

Phosphate Binder Impact on Bone Remodeling and Coronary Calcification – Results from the BRiC Study Daniela Veit Barreto a Fellype de Carvalho Barreto a Aluízio Barbosa de Carvalho a Lilian Cuppari a Sérgio Antonio Draibe a Maria Aparecida Dalboni a Rosa Maria Affonso Moyses b Kátia Rodrigues Neves b Vanda Jorgetti b Marcio Miname c Raul D. Santos c Maria Eugênia Fernandes Canziani a Division of Nephrology, Department of Internal Medicine, a Federal University of São Paulo and b University of São Paulo, and c The Lipid Clinic of the Instituto do Coração (InCor, Heart Institute), University of São Paulo, São Paulo, Brazil

Abstract Background and Aims: Calcium-containing phosphate binders have been shown to increase the progression of vascular calcification in hemodialysis patients. This is a prospective study that compares the effects of calcium acetate and sevelamer on coronary calcification (CAC) and bone histology. Methods: 101 hemodialysis patients were randomized for each phosphate binder and submitted to multislice coronary tomographies and bone biopsies at entry and 12 months. Results: The 71 patients who concluded the study had similar baseline characteristics. On follow-up, the sevelamer group had higher levels of intact parathyroid hormone (498 8 352 vs. 326 8 236 pg/ml, p = 0.017), bone alkaline phosphatase (38 8 24 vs. 28 8 15 U/l, p = 0.03) and deoxypyridinoline (135 8 107 vs. 89 8 71 nmol/l, p = 0.03) and lower LDL cholesterol (74 8 21 vs. 91 8 28 mg/dl, p = 0.015). Phosphorus (5.8 8 1.0 vs. 6 8 1.0 mg/dl, p = 0.47) and calcium (1.27 8 0.07 vs. 1.23 8 0.08 mmol/l, p = 0.68) levels did not differ between groups. CAC progression (35 vs. 24%,

© 2008 S. Karger AG, Basel 1660–2110/08/1104–0273$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/nec

p = 0.94) and bone histological diagnosis at baseline and 12 months were similar in both groups. Patients of the sevelamer group with a high turnover at baseline had an increase in bone resorption (eroded surface, ES/BS = 9.0 8 5.9 vs. 13.1 8 9.5%, p = 0.05), whereas patients of both groups with low turnover at baseline had an improvement in bone formation rate (BFR/BS = 0.015 8 0.016 vs. 0.062 8 0.078, p = 0.003 for calcium and 0.017 8 0.016 vs. 0.071 8 0.084 ␮m3/␮m2/day, p = 0.010 for sevelamer). Conclusions: There was no difference in CAC progression or changes in bone remodeling between the calcium and the sevelamer groups. Copyright © 2008 S. Karger AG, Basel

Introduction

Arterial calcification is a very common complication of chronic kidney disease, and predicts cardiovascular morbidity and mortality in these patients [1, 2]. Observational and experimental studies have suggested that abnormalities in mineral metabolism, including disturbances in calcium-phosphate regulation [3–5] and changes in bone remodeling [6] may play a role in the pathogenesis of arterial calcification. Additionally, exMaria Eugênia F. Canziani Rua Pedro de Toledo, 282 São Paulo, SP 04039-000 (Brazil) Tel. +55 11 5571 3261, Fax +55 11 5572 1862 E-Mail [email protected]

Downloaded by: Sociedade Brasilieira de Nefrologia 191.177.140.167 - 9/2/2016 10:09:45 PM

Key Words Cardiovascular disease ⴢ Hemodialysis ⴢ Phosphate binders ⴢ Renal osteodystrophy ⴢ Vascular calcification

Patients assessed for eligibility (n = 595)

*

Enrollment

Not meeting inclusion criteria or refusal to participate (n = 494)

Randomization (n = 101)

Calcium acetate (n = 49)

Allocation

Sevelamer (n = 52)

Unavailable at 12 months n = 19 Parathyroidectomy (n = 1) Transplanted (n = 6) Death (n = 8) Other (n = 4)

Follow-up

Unavailable at 12 months n = 11 Parathyroidectomy (n = 1) Transplanted (n = 6) Death (n = 1) Other (n = 3)

Completed 12 month follow-up n = 30 No MsCT at 12 months (n = 18) No biopsy at 12 months (n = 4) MsCT analyzed (n = 31) Bone biopsy analyzed (n = 27)

Analysis

Completed 12-month follow-up n = 41 No MsCT at 12 months (n = 11) No biopsy at 12 months (n = 4) MsCT analyzed (n = 41) Bone biopsy analyzed (n = 37)

Fig. 1. Patient distribution.

c274

Nephron Clin Pract 2008;110:c273–c283

Materials and Methods Subjects and Study Design Patients that had been on maintenance hemodialysis for at least 3 months at 4 dialysis centers in São Paulo, Brazil, were assessed for eligibility (fig. 1). All patients had been receiving 4hour hemodialysis 3 times weekly, using hollow-fiber polysulfone or acetate membranes, and at entry were on a standard dialysate calcium concentration of 3.5 mEq/l. Exclusion criteria included serious gastrointestinal disease, ethanol or drug abuse, active malignancy, human immunodeficiency virus infection, chronic inflammatory disease, current use of steroids, severe hyperparathyroidism (defined as iPTH 11,000 pg/ml), body weight 1100 kg, continuous use of antiarrhythmic or seizure drugs, pregnancy or breast-feeding, previous myocardial revascularization, uncontrolled diabetes mellitus or hypertension. After screening, 153 subjects underwent a 2-week washout period, in which all phosphate binders were withheld. All patients with hyperphosphatemia (serum phosphorus 15.5 mg/dl) at the end of the washout period were eligible for the study. One hundred and one patients were selected and randomized in a 1:1 ratio to receive either openlabel sevelamer (Renagel쏐 800-mg tablets; Genzyme Co, Cambridge, Mass., USA) or calcium acetate (PhosLo쏐, 667-mg tablets; Fresenius Medical Care, Waltham, Mass., USA). Randomization was computer generated, in blocks of 20, and the treatment was assigned by the coordinating center using concealed envelopes. The initial dose of phosphate binder was calculated for each patient according to usual label recommendations. Thereafter, medication was adjusted monthly (up to 12,000 mg daily for sevelamer, and up to 2,028 mg of elemental calcium daily for calcium

Barreto et al.


perimental data have shown that uremic rats treated with sevelamer, a non-calcium-containing phosphate binder, had reduced aortic vascular calcification compared to those treated with calcium-containing phosphate binders [7]. In clinical settings, sevelamer attenuated the progression of coronary artery calcification (CAC) both in prevalent and incident hemodialysis patients [8, 9]. Based on these results, the authors suggested that calcium-containing phosphate binders might have an adverse effect on bone remodeling, leading to extraskeletal calcification [10]. Furthermore, the management of renal osteodystrophy with calcium-containing phosphate binders, in comparison to sevelamer, has been associated with significantly greater progression of CAC and changes in intact parathyroid hormone (iPTH), which could account for the development of low-turnover bone disease, particularly in diabetic patients undergoing chronic hemodialysis [11]. The present study was conducted to test the hypothesis that calcium-containing phosphate binders have a negative impact on bone remodeling, thus contributing to the progression of CAC in hemodialysis patients. Therefore, we compared the effects of calcium acetate and sevelamer on bone remodeling and on the development and progression of CAC in hemodialysis patients.

Laboratory Measurements The morning before the first weekly hemodialysis session, whole blood was collected from the subjects in a fasting state, at the respective study sites. Ionized calcium, phosphorus and hemoglobin were measured monthly. Serum iPTH (Immulite, DPC, Los Angeles, Calif., USA; reference range 10–65 pg/ml) was determined every second month. Albumin, urea, C-reactive protein (High Sensitive C-Reactive Protein Immulite쏐, immunometric assay; functional sensitivity 1 0.02 mg/dl), total cholesterol, lowdensity cholesterol, high-density cholesterol, very-low density cholesterol and triglycerides were determined at baseline, and at 6 and 12 months. Osteoprotegerin [enzyme-linked immunosorbent assay (ELISA), Immundiagnostik Laboratory, reference range 30.45 8 12.1 pg/ml], soluble RANKL [soluble receptor activator of NF-␬B ligand (ELISA), Immundiagnostik Laboratory, Bensheim, Germany, detection limit 1.5 pg/ml], 25-(OH) vitamin D (radioimmunoassay, DiaSorin쏐, Stillwater, Minn., USA; reference range 18–62 ng/ml) were determined at baseline and 12 months. Being specific serum markers of bone turnover, bonespecific alkaline phosphatase [enzyme immunoassay (EIA), Metra Biosystem Inc., Mountain View, Calif., USA; reference range 11.6–42.7 U/l] and deoxypyridinoline (EIA, Quidel Corporation, USA; reference range 3.25 8 0.66 nmol/l) were determined at baseline and at 6 and 12 months (the latter) and at baseline and 12 months (the former). Hypercalcemia was defined as ionized calcium levels higher than 1.40 mmol/l. Vitamin D deficiency was defined as 25-(OH) vitamin D levels lower than 15 ng/ml [12]. Imaging Procedure The calcification score was determined using a 16-slice MsCT (Somatron Volum Zoom Siemens AG쏐, Erlangen, Germany). A chest radiologic image without contrast was acquired while the subject was in apnea in order to determine the initial and final levels. Then, the images of each section were acquired during a 150-ms exposure, with a distance of 3 mm between each slice. The timing of image acquisition was coordinated with the diastolic phase of the cardiac cycle, at 60% of the RR interval, using electrocardiographic monitoring. All scans were analyzed with the

Phosphate Binders, Bone Remodeling and Coronary Calcification in HD

‘Workstation’ software (Indigo 02쏐 SGI쏐, Mountain View, Calif., USA) in order to determine the calcium score. This software can detect calcified lesions with a density of at least 130 Housfield units and a minimal area of 0.5 mm2. The total calcium score was calculated by measuring the total volume and the area of the calcified lesions as well as by measuring mean and peak calcium density. The total score was the sum of the scores of each coronary and was expressed in modified Agatston units (AU) [13, 14]. A single experienced investigator, who was blinded to all other patient data, read all scans. Bone Biopsy Baseline and 12-month bone specimens were obtained alternatively from the right or left iliac crests to avoid interference of repair processes from the previous biopsy. The procedure was conducted using a trephine with an inner diameter of 7 mm, which was adapted to an electrical drill (Gauthier Medical, Rochester, Minn., USA). All patients were prelabeled with oral tetracycline (20 mg/kg/day for 3 days) that was administered over two separate periods, 10 days apart. The bone fragments were submitted to the usual histological processing and analysis [15]. Bone histomorphometry was conducted using the semiautomatic method of the Osteomeasure쏐 software (Osteometrics Inc., Atlanta, Ga., USA). We used the histomorphometric parameters suggested by the American Society of Bone and Mineral Research histomorphometry nomenclature committee [16]. The reference ranges for static parameters were obtained from local controls [17], whereas the dynamic parameters followed those described elsewhere [18]. Renal osteodystrophy was classified into one of the classical types using the following criteria: (1) predominant hyperparathyroid bone disease – defined as bone formation rate (BFR/BS; reference range: 0.13 8 0.07 ␮m3/␮m2/day for men and 0.07 8 0.03 ␮m3/␮m2/day for women) more than 1 SD above the normal range, as well as either osteoblast surface (Ob.S/BS; reference range: 1.2 8 1.4% for men and 1.2 8 3.2% for women) or osteoclast surface (Oc.S/BS; reference range: 0.03 8 0.11% for men and 0.03 8 0.06% for women) more than 1 SD above the normal range, osteoid volume (OV/BV; reference range: 2.9 8 2.7% for men and 1.55 8 1.9% for women) within or above the normal range and marrow fibrosis 1 0.5%; (2) adynamic bone disease – when BFR/BS and OV/BV were more than 1 SD below the normal range and marrow fibrosis !0.5%; (3) osteomalacia – when BFR/ BS was more than 1 SD below the normal range and OV/BV was more than 1 SD above the normal range, and (4) mixed uremic osteodystrophy – when BFR/BS, OV/BV and mineralization lag time (reference range: 21.3 8 2.3 days for men and 23.7 8 2.7 for women) were more than 1 SD above the normal range and marrow fibrosis 1 0.5%. For analytical purposes, these types were grouped into one of two major patterns: high-turnover bone disease (predominant hyperparathyroid bone disease and mixed uremic osteodystrophy) or low-turnover bone disease (osteomalacia and adynamic bone disease). Aluminum intoxication was defined when aluminum bone surface (Al.S/BS) was greater than 25%. Bone histomorphometric data from 64 patients (27 from the calcium group and 37 from the sevelamer group) who had baseline and 12-month bone biopsies were included in this study. Patients who had only one biopsy were excluded from the analysis.


c275


acetate) in order to achieve the study endpoints, which comprised: serum phosphorus levels of 3.5– 5.5 mg/dl, ionized calcium levels of 1.11–1.40 mmol/l and iPTH levels between 150 and 300 pg/ml. Investigators were also encouraged to alter the calcium dialysate concentration and the vitamin D treatment during the study based on baseline bone biopsy diagnosis. Therefore, patients with low-turnover bone disease were withdrawn from the vitamin D treatment and the calcium dialysate concentration was set at a level of 2.5 mEq/l, regardless of iPTH levels. No tablet count was performed and no patient received calcimimetics during the course of the clinical trial. None of the selected patients had undergone prior parathyroidectomy. Blood samples were drawn periodically to assess serum markers of mineral metabolism. Patients underwent multislice coronary tomographies (MsCT) at baseline and at 12 months to assess the CAC status. Bone biopsies were performed at baseline and at 12 months to assess long-term alterations of bone remodeling. All patients signed an informed consent form. The study protocol was reviewed and approved by the local institutional ethics board, in accordance with the ethical principles of the Declaration of Helsinki.

Table 1. Demographic and laboratory characteristics of the study participants

Baseline

Age, years Gender, male, % Ethnic, white, % Length on hemodialysis, months Body mass index, kg/m2 Comorbid conditions Hypertension, % Diabetes mellitus, % Smoking, % Calcitriola, % Dialysate Ca2+ 2.5 mEq/lb, % Phosphorus, mmol/l Ionized calcium, mmol/l Bone-specific alkaline phosphatase, U/l iPTH, pg/ml Osteoprotegerin, pg/ml sRANKL, pg/ml Deoxypyridoline, nmol/l 25-(OH)vitamin D, ng/ml Hemoglobin, g/dl Kt/V High-density lipoprotein cholesterol, mg/dl Low-density lipoprotein cholesterol, mg/dl Very low-density lipoprotein cholesterol, mg/dl Triglycerides, mg/dl Albumin, g/dl C-reactive protein, mg/dl

Follow-up

calcium (n = 30)

sevelamer (n = 41)

47814 70 63 38823 2484

47813 66 58 36827 2484

73 13 13 17 0 2.380.45 1.2380.08 27819 3438347 160848 6.4810.8 90890 31815 1181.5 1.3780.2 40810 89827 29814 153885 3.780.2 0.9080.90

66 15 32 15 0 2.380.7 1.2380.08 34838 4048343 172874 4.286.2 1368177 33816 1181.6 1.3380.23 43810 88825 26814 139891 3.880.3 0.9681.99

calcium (n = 30)

34 53 1.8780.3 1.2780.07 28815 3268236 266894 10.0814.1 89871 29812 1181.2 1.3380.15 4089 91828 31816 161888 3.980.2 1.1381.06

sevelamer (n = 41)

65* 37 1.7180.3 1.2880.07 38824* 4988352* 2798109 7.188.8 1358107* 28811 1181.3 1.3080.16 43811 74821* 28814 147885 4.080.2 0.8280.89

Baseline demographic characteristics. Baseline and postbaseline laboratory values. Follow-up values represent the mean of all measures of each parameter during the study. * p < 0.005 between group comparisons. a Patients on calcitriol for at least 1 month. b Patients on Ca 2+ 2.5 mEq/l dialysate for at least 3 months.

c276


Two-tailed p values ^0.05 were considered statistically significant. Analyses were conducted using SPSS 13.0 for Windows쏐 (SPSS Inc., Chicago, Ill., (USA).

Results

The patient distribution by treatment group is shown in figure 1. Seventy-one patients concluded the study protocol, with 2 available MsCTs. Reasons for dropout during follow-up were similar between treatment groups, except for the number of deaths, that was greater in the calcium group than in the sevelamer group (16 and 2%, respectively, p = 0.014). Median CAC at baseline (150 vs. Barreto et al.


Statistical Analysis Baseline and follow-up characteristics were compared between the sevelamer and the calcium groups using the ␹2 test or Fisher’s exact test for the categorical variables and either the t test or the Mann-Whitney test for the continuous variables. The absolute increase in CAC was calculated as the difference between 12-month and baseline scores. Relative increase indicates the ratio: absolute increase/baseline score ! 100. As the calcium score was not normally distributed, the results regarding this variable were presented as median and range. The changes from baseline to end-of-study CAC and histomorphometric parameters were compared within groups using the Wilcoxon signed-rank test. The Kruskal-Wallis test was performed to evaluate the differences in CAC between the calcium and sevelamer groups which were subdivided according to calcium dialysate concentration. Missing laboratory values and bone biopsy results were not imputed.

Phosphorus (mg/dl)

7.5

*

5.0

2.5

Sevelamer Calcium

0 0

1

2

3

4

5

6

7

8

9 10 11 12 13

Time (months)

Ionized calcium (mmol/l)

1.5 1.4 1.3 1.2 1.1 1.0 0

1

2

3

4

5

6

7

8

9 10 11 12 13

Time (months) 1,250

*

1,000

*

750 500 250 0 1 –250

2

3

4

5

6

7

8

9 10 11 12 13

Time (months)

Fig. 2. Control of serum concentrations of phosphorus, ionized

calcium and iPTH by treatment group. * p ^ 0.05.

bone formation and resorption parameters remained unchanged. In the sevelamer group, a significant increase in trabecular thickness (Tb.Th) and a decrease in trabecular number (Tb.N) was observed, while no change was observed on BV/TV. In this group, there was a significant increase in bone formation rate (BFR/BS) and in bone resorption, expressed by eroded surface (ES/BS). There was no difference in the absolute changes of any of the Nephron Clin Pract 2008;110:c273–c283

c277



10.0

iPTH (pg/ml)

123 AU, p = 0.803), demographic and laboratory characteristics were similar between patients who died and those who survived. Five deaths in the calcium group were attributable to cardiovascular disease and happened on days 66, 174, 280, 289 and 372 of follow-up. A single death caused by cardiovascular disease was observed in the sevelamer group on day 285 of follow-up. Table 1 depicts demographic and laboratory characteristics at baseline and follow-up (mean of all measured values). There were no differences regarding demographic or laboratory parameters between groups. The proportion of patients on calcitriol was similar in the calcium and the sevelamer groups at baseline, but on follow-up, patients of the sevelamer group were treated more frequently with calcitriol. The rate of patients on 2.5 mEq/l dialysate was similar between treatment groups at baseline and during follow-up. During the follow-up period, patients of the sevelamer group presented significantly lower serum levels of low-density lipoprotein cholesterol, and higher serum levels of iPTH, bone-specific alkaline phosphatase and deoxypyridinoline. Biochemical endpoints are shown in figure 2. A similar serum phosphorus control was observed in both groups during most of the follow-up period, except at 10 and 11 months, at which times it was poorer in the calcium group. Mean ionized calcium concentrations did not differ between groups, nor did the incidence of hypercalcemia. Mean serum iPTH levels tended to be higher during follow-up in the sevelamer group, reaching statistical significance at 2 and 12 months. Table 2 depicts histomorphometric data from baseline and 12-month bone biopsies. At baseline, bone parameters were similar in both the calcium and the sevelamer groups, except for a significantly lower osteoblastic surface (Ob.S/BS) and a trend (p = 0.09) towards a lower osteoclastic surface (Oc.S/BS) in the calcium group (p = 0.09). Bone aluminum surface was similar in the calcium (21.1828.7%) and the sevelamer groups (27.6827.4%; p = 0.186). At 12 months, the calcium group presented higher trabecular bone volume (BV/TV) and trabecular number (Tb.N), and lower trabecular separation (Tb.Sp). The calcium group presented lower osteoid volume (OV/ BV), and lower osteoid (OS/BS) and osteoblastic surfaces (Ob.S/BS). There was no difference between the two groups concerning bone resorption parameters, fibrosis or aluminum surface. Comparing recordings at baseline and at 12 months within each group separately, a significant increase in trabecular bone volume (BV/TV) and trabecular thickness (Tb.Th) was observed in the calcium group, whereas

Table 2. Histomorphometric characteristics by treatment group

Baseline

Structure BV/TV, % Tb.Th, ␮m Tb.Sp, ␮m Tb.N, n/mm Formation OV/BV, % OS/BS, % Ob.S/BS, % BFR/BS, ␮m3/␮m2/day MLT, days Resorption ES/BS, % Oc.S/BS, % Fibrosis Fb.V/BV, %

12 months

Reference range

calcium (n = 27)

sevelamer (n = 37)

calcium (n = 27)

sevelamer (n = 37)

women

18.985.6 113.6822.1 555.18151.6 1.5580.31

17.487.2 112.3819.7 610.18254.2 1.5380.50

20.286.2c 128.1825.0c 535.58141.3 1.5580.34

17.186.3b 123.7827.4b, d 651.78216.3b 1.3780.34b, d

21.887.2 24.086.1 126.0828.8 127.9829.7 498.38195.9 420.68124.1 1.7680.52 1.8980.42

3.183.0 22.1813.8 4.184.5 0.03780.051 133.18141.1

4.684.5 28.0816.9 7.788.6a 0.04280.054 130.68122.0

2.582.0 20.8811.3 4.384.5 0.06580.078 138.48138.0

4.983.4b 31.4815.6b 8.686.8b 0.08181.053d 128.38131.1

1.5581.9 9.288.4 1.283.2 0.0780.03 23.782.7

2.982.7 16.1812.6 1.281.4 0.1380.07 21.382.3

5.284.8 0.8480.97

6.585.1 1.0780.77

7.185.4 0.7980.79

9.288.1d 1.381.13

2.382.4 0.0380.06

1.7581.21 0.0380.11

0.1380.25

0.6082.08

0.1580.21

0.4380.75

0

men

0

BFR = Bone formation rate; BV/TV = trabecular bone volume; ES/BS = eroded surface; MLT = mineralization lag time; Ob.S/ BS = osteoblast surface; Oc.S/BS = osteoclast surface; OS/BS = osteoid surface; OV/BV = osteoid volume; Tb.N = trabecular number; Tb.Th = trabecular thickness. Reference values from Dos Reis et al. [17] and Melsen and Mosekilde [18]. a p < 0.05 in com-

parisons between treatment groups at baseline; b p < 0.05 in comparisons between treatment groups at 12-month bone biopsy; c p < 0.05 in comparisons between baseline and 12-month values within the calcium group; d p < 0.05 in comparisons between baseline and 12-month values within the sevelamer group.

histomorphometric parameters between the calcium and the sevelamer groups comparing baseline to 12 months. Table 3 shows the bone histomorphometric characteristics by treatment group in the patients diagnosed as having high-turnover bone disease at baseline. In the calcium group patients displayed an increase in trabecular thickness (Tb.Th), while in the sevelamer group patients displayed an increase in trabecular thickness (Tb.Th), a reduction in trabecular number (Tb.N) and an increase in bone resorption, expressed by eroded surface (ES/ BS). Table 4 shows the bone histomorphometric characteristics by treatment group in patients diagnosed as having low-turnover bone disease at baseline. At baseline, the only significant difference between treatment groups was a lower osteoid surface (OS/BS) in the calcium group. At 12 months, patients of the calcium group presented significantly higher bone volume (BV/TV), trabecular thickness (Tb.Th) and lower trabecular separation (Tb.Sp), when compared to those of the sevelamer group. These patients also had lower osteoid volume (OV/BV), osteoid

(OS/BS) and osteoblastic surfaces (Ob.S/BS). There was a significant increase in bone formation rate (BFR/BS) and reduction of mineralization lag time (Mlt) in both groups. Additionally, there was an increase in trabecular bone volume (BV/TV) and trabecular thickness (Tb.Th) in the calcium group. There was no difference in the renal osteodystrophy pattern among treatment groups either at baseline or at the 12-month follow-up (fig. 3). A single patient presented normal bone histology at baseline and at 12 months. Prior to the beginning of the study, 95% of the patients were taking calcium-containing phosphate binders (calcium acetate or carbonate) to treat renal osteodystrophy, while no patient was getting either sevelamer or aluminum hydroxide. Sixty-seven percent of the patients of the calcium group and 41% of those of the sevelamer group had used vitamin D supplementation at any time before the study (p = 0.035). Figure 4 depicts the calcium score distribution by treatment group at baseline and at 12 months. Table 5 shows quantitative information regarding CAC – the me-

c278


Barreto et al.


Variable

Table 3. Histomorphometric characteristics by treatment group in patients diagnosed with high-turnover state at baseline

Variable

Structure BV/TV, % Tb.Th, ␮m Tb.Sp, ␮m Tb.N, n/mm Formation OV/BV, % OS/BS, % Ob.S/BS, % BFR/BS, ␮m3/␮m2/day Mlt, days Resorption ES/BS, % Oc.S/BS, % Fibrosis Fb.V/BV, %

Baseline

12 months

calcium (n = 9)

sevelamer (n = 17)

calcium (n = 9)

sevelamer (n = 17)

16.483.7 104.6814 555.98135.7 1.5580.29

19.387.7 114.3820.9 543.08245.7 1.6880.53

18.585.5 117.8821.3a 537.28106.3 1.5380.22

19.787.0 138.2830b 634.68285.0 1.4280.40b

5.783.7 34.7810.1 9.183.8 0.0880.07 95.48129.1

7.185.4 37.1815.6 13.1810.0 0.0780.06 91.6888.7

4.082.4 29.3810.2 8.185.1 0.0680.08 166.28140.8

5.583.8 34.3815.0 10.886.6 0.0980.12 166.28144.8

7.885.0 1.3580.97

9.085.9 1.4580.88

9.086.0 1.2280.96

13.189.5b 1.6581.16

0.3480.36

1.2582.99

0.2580.28

0.8180.98

a p < 0.05 in comparisons between baseline and 12-month values within the calcium group; b p < 0.05 in comparisons between baseline and 12-month values within the sevelamer group. BFR = Bone formation rate; BV/TV = trabecular bone volume; ES/BS = eroded surface; MLT = mineralization lag time; Ob.S/BS = osteoblast surface; Oc.S/BS = osteoclast surface; OS/BS = osteoid surface; OV/BV = osteoid volume; Tb.N = trabecular number; Tb.Th = trabecular thickness.

100 90 80 70 60 50 40 30 20 10 0

High turnover

Low turnover 1

1

17

19

9

10

20

18

17

16

12 months

Baseline

Baseline

Sevelamer

12 months

Calcium

Fig. 3. Bone remodeling by treatment group. Columns represent percentage of patients diagnosed as having normal bone, lowturnover status or high-turnover status at baseline and at 12 months; bone biopsies were divided by treatment group. The appointed values represent the absolute number of patients in each category. At baseline, 1 patient from the calcium group did not have a bone biopsy. At 12 months, 3 patients of the calcium group and 4 patients of the sevelamer group refused to have bone biopsies.


c279



Normal

Patients (%)

dian CAC score was similar in both treatment groups at baseline and at 12 months. At baseline, 53% of the patients of the calcium group and 66% of the sevelamer group had a CAC score 130 AU (p = 0.286). After 1 year, only 1 patient with CAC !30 AU at baseline presented CAC progression. When the patients with baseline CAC !30 AU were excluded from the analysis in order to evaluate CAC progression, there were still no differences in baseline and 12-month CAC between both treatment groups (table 6). After 1 year, there were significant changes in CAC scores within both treatment groups, but there was no significant difference in absolute and relative progression of CAC between groups. In order to investigate if the choice of calcium dialysate concentration had an effect on CAC progression for each treatment group, patients of the calcium and the sevelamer groups were subdivided according to calcium dialysate concentration [i.e. patients that had been on 2.5 mEq/l dialysate for at least 3 months during the study (2.5 vs. 3.5 dialysate)]. There was no significant difference in CAC scores at baseline (p = 0.565) and 12 months (p = 0.551) between the 4 resulting subgroups [calcium + 2.5 dialysate (n =

Table 4. Histomorphometric characteristics by treatment group in patients diagnosed with low turnover state

at baseline Variable

Structure BV/TV, % Tb.Th, ␮m Tb.Sp, ␮m Tb.N, n/mm Formation OV/BV, % OS/BS, % Ob.S/BS, % BFR/BS, ␮m3/␮m2/day MLT, days Resorption ES/BS, % Oc.S/BS, % Fibrosis Fb.V/BV, %

Baseline

12 months

calcium (n = 17) sevelamer (n = 20)

calcium (n = 17) sevelamer (n = 20)

18.486.5 116.8824.3 557.08167.5 1.5580.34

15.886.4 110.5818.9 667.28253.3 1.480.44

20.886.6c 132.8826.2c 540.48161.7 1.5480.40

14.984.8b 111.4817.8b 666.18140.4b 1.3380.29

1.781.4 15.1810.7 1.582.0 0.01580.016 159.58147.5

2.581.8 20.2814.0a 3.0582.5 0.01780.016 163.78138.1

1.681.3 16.289.5 2.382.7 0.06280.078c 130.58140.3c

4.383.0b 28.9815.9b 6.786.6b 0.07180.084d 96.08112.0d

3.884.4 0.5680.92

4.382.9 0.7580.48a

6.285.2 0.5480.61

0.0280.41

0.0580.08

0.0980.14

5.984.9 0.9881.0 0.180.12

BFR = Bone formation rate; BV/TV = trabecular bone volume; ES/BS = eroded surface; MLT = mineralization lag time; Ob.S/BS = osteoblast surface; Oc.S/BS = osteoclast surface; OS/BS = osteoid surface; OV/BV = osteoid volume; Tb.N = trabecular number; Tb.Th = trabecular thickness.a p < 0.05 in comparisons between treatment groups at baseline; b p < 0.05 in comparisons between treatment groups at 12-month bone biopsy; c p < 0.05 in comparisons between baseline and 12-month values within the calcium group; d p < 0.05 in comparisons between baseline and 12-month values within the sevelamer group.

Discussion

This study compared the treatment of hyperphosphatemia in 101 hemodialysis patients that were randomized to receive either sevelamer or calcium acetate. Both groups presented equivalent serum phosphorus control, and no difference in the development or progression of CAC was observed between groups. In contrast to these findings, previous studies have shown an association between CAC and the use of calcium-based phosphate binders in hemodialysis patients. Goodman et al. [4] observed that the daily intake of calcium-containing phosphate binders was associated with CAC development, and later on, Chertow et al. [8] demonstrated that patients receiving calcium-based phosphate binder treatment had greater progression of CAC c280


compared to those receiving sevelamer. In the latter study, vitamin D supplements were given to all participants in a fixed dose, and were not adjusted by calcium or PTH levels. Therefore, Fournier et al. [19] reasoned that, as a result, PTH levels may have fallen below the target range in the calcium group. The PTH oversuppression may have led to lower bone turnover, lower bone density and a decrease in the uptake of supplemental calcium in the bone, resulting in a higher risk of vascular calcification. To prevent PTH oversuppression, we adjusted calcitriol and dialysate prescriptions on a monthly basis, according to biochemical endpoints (phosphorus, calcium and PTH) and baseline bone biopsy diagnosis. Moreover, unlike former reports [20–22], the prevalence of vitamin D deficiency was relatively low (8%) in this Brazilian population of chronic kidney disease patients, and there was less active vitamin D supplementation on follow-up. Earlier experimental studies reported that vitamin D derivatives may promote vascular calcification by inducing hypercalcemia [23] or by direct effects on vascular smooth muscle cells [24] in uremic animals. In the present study, both bone remodeling and the progression of vascular Barreto et al.


16), sevelamer + 2.5 dialysate (n = 15), calcium + 3.5 dialysate (n = 14), sevelamer + 3.5 dialysate (n = 26)]. Absolute (p = 0.598) and relative (p = 0.546) CAC progression was also similar between subgroups.

Table 5. Absolute and relative increase in CAC score in all pa-

tients 7,000

Calcium (n = 30)

Sevelamer (n = 41)

p value1

Baseline Median 12-month Median Absolute increase Median Relative increase Median

67581,267 50 85781,559 57 1828333 4 558127 3.5

5078814 192 6468973 293 1398240 43 45881 45

0.384

p value2

0.002

Calcium score (AU)

5,000 3,000 2,000 500 500 250 0

a

Baseline

12 months

0.526 0.526

30

0 Baseline

12 months

Fig. 4. Calcium score distribution by treatment group at baseline

and at the 12-month follow-up. Patients with baseline calcium score !30 AU were excluded from the analysis. a Calcium acetate. b Sevelamer.

calcification could have been affected by the absence of sustained hypercalcemia. Former experimental studies suggested that, in the progression of vascular calcification, adequate phosphate control may be more important than the choice of phosphate binder. Phan et al. [25] evaluated the effect of sevelamer and calcium carbonate [26] on vascular calcification in Apo-E-deficient mice, and found that the treatment with both phosphate binders attenuated intima and media calcification. Therefore, we may conclude that calcium was not a progression factor in these studies. The association between CAC and disturbances of mineral metabolism in chronic kidney disease might supersede the effect of specific phosphate binders on CAC progression. Concerning the role of secondary hyperPhosphate Binders, Bone Remodeling and Coronary Calcification in HD

Calcium (n = 16)

Sevelamer (n = 27)

p value1

Baseline Median 12-month Median Absolute increase Median Relative increase Median

1,26381,521 775 1,60281851 935 3398397 149 52878 35

7678902 336 97681,062 505 2088272 162 48878 24

0.782

p value2

0.002

0.642 0.514 0.940