Hyperphosphatemia is extensively documented as an independent factor associated with increased morbidity in patients with chronic kidney disease (CKD) (
1 – ). Epidemiologic studies have demonstrated that elevated serum phosphate is associated with an increased relative risk of death in patients on hemodialysis ( 4 ). 5
The control of serum phosphate has therefore become a key therapeutic target in the management of patients receiving maintenance dialysis, and may become an increasingly important issue in patients with less severe impairment of their renal function. Standard thrice-weekly hemodialysis is unable to remove the ingested oral phosphate load adequately, despite the application of higher-efficiency dialysis regimens (
6 , ). Restricting oral phosphate ingestion alone can lead to a state of worsening malnutrition, with the associated negative effects on patient outcome ( 7 ). The control of hyperphosphatemia has come to rely on the use of oral phosphate-binding medication ingested with food, rendering at least some of the phosphate therein insoluble and unavailable for absorption. 8
Oral phosphate binders control serum phosphate to a certain extent, but serum phosphate remains uncontrolled in approximately half of the patients on current therapy (based largely on sevelamer hydrochloride and calcium) in France, Germany, Italy, Japan, Spain, the United Kingdom, and the United States (
). More recently, the mean serum phosphate of the maintenance hemodialysis patient population in the United States has been reported as 2.01 mmol/L, and 60% of patients have a serum phosphate level in excess of 1.78 mmol/L ( 9 ). 10
Treatment with current phosphate binders is associated with several side effects. Those that are calcium-based have an increased body of evidence suggesting that exposure to higher amounts of calcium is associated with progressive vascular calcification (
11 , ). It is also of note that vascular calcification progression, associated with the use of calcium-containing binders, is of significantly greater severity in conjunction with poorly controlled serum phosphate ( 12 ). This reinforces the critical requirement of high efficacy in phosphate-lowering therapies. In addition, the current balance of evidence appears to justify the position that an ideal therapy would limit calcium exposure. Therapy with some agents can be associated with a high pill burden and newer agents that could potentially reduce this would also be of benefit. 13
Fermagate (Ineos Healthcare, Warrington, United Kingdom) contains magnesium and ferric iron held in an insoluble hydrotalcite structure. The iron and magnesium are held in a tight crystalline-layered structure, with carbonate groups, which are exchanged for phosphate, lying between the layers (see
Figure 1). In vitro studies have shown that fermagate has high affinity for phosphate over a wide pH range ( ). The high efficacy of fermagate compared with sevelamer hydrochloride, aluminum hydroxide, lanthanum carbonate, and magnesium hydroxide has been demonstrated using an artificial “stomacher” and simulated breakfast meal, and confirmed in a simulated upper gastrointestinal tract model against sevelamer hydrochloride ( 14 ). A 7-d, placebo-controlled study in healthy volunteers, using fermagate doses of 1 and 2 g three times daily, revealed significant phosphate binding, with marked reductions in 24-h urinary phosphate excretion and a corresponding increase in fecal phosphate content ( 15 ). 16
The aim of this clinical phase II study was to examine the efficacy and safety of two doses of fermagate (1 and 2 g three times daily) in the treatment of hyperphosphatemia in chronic hemodialysis patients.
Materials and Methods
The primary objective of the study was to determine the efficacy (lowering of serum phosphate) of multiple oral doses of fermagate, compared with placebo, in the treatment of hyperphosphatemia in patients on stable hemodialysis. A secondary objective was to compare fermagate with placebo for its ability to lower cholesterol in patients on hemodialysis.
This was a randomized, double-blind, placebo-controlled, parallel-group study to investigate the safety and efficacy of multiple doses of fermagate (1 and 2 g three times daily) against placebo. It was conducted in five centers in the United Kingdom. Authorization to conduct the trial was provided by the U.K. regulatory authority, the Multicenter Research Ethics Committee (REC), and the local REC at each site before patient recruitment.
Current Therapy and Washout Periods
Each patient gave written informed consent before the performance of any study-specific procedures. After a successful screening procedure, eligible patients remained on their current phosphate binder treatment for 2 wk while serum phosphate levels were observed to establish treatment stability (current therapy period). If serum phosphate levels were <2 mmol/L at the end of this period, the patient's current phosphate binder treatment was stopped for 2 to 4 wk (washout period). During this period, serum phosphate levels were monitored and only when the serum phosphate reached 1.7 mmol/L or greater were patients entered into the study medication (treatment) period. Patients were withdrawn from the study if their serum phosphate was <1.7 mmol/L or exceeded 3 mmol/L on three successive observations within the first 2 wk of the washout period or if serum phosphate levels had not risen to ≥1.7 mmol/L after 4 wk.
At the time of entry into the treatment period, patients were randomized to one of the three treatment arms (placebo, fermagate 1 g, or fermagate 2 g) in a ratio of 1:1:1. Study treatment (capsules containing 250 mg fermagate or matched-appearance placebo) was taken orally, three times daily, for 21 consecutive days just before meals. After completion of the treatment period, patients reverted to their original phosphate binder or equivalent and were followed up for a period of 2 wk (follow-up period).
Figure 2 illustrates the study design. Compliance was monitored by returned pill counts, protocol deviations, and monitoring of dispensing logs. Patients were withdrawn if their serum phosphate exceeded 3 mmol/L on three successive observations.
The study was conducted according to the current International Conference on Harmonization (ICH) Good Clinical Practice (GCP) guidelines, any local guidelines, and the Declaration of Helsinki (Edinburgh, Scotland, 2000).
Study Population Selection
The study population included male or female patients aged 18 yr or over, on a stable hemodialysis regimen (three times per week) for at least 3 mo, who were unlikely to change their dialysis prescription during the study period. In addition, subjects were receiving a phosphate binder at a stable dose for at least 1 mo before screening and had to be willing to abstain from using any oral magnesium-, aluminum-, or iron-containing preparations other than the study medication during the trial. Any vitamin D supplementation was to remain unchanged during the study period.
Patients were excluded from the study if they had a history of hemochromatosis, serum ferritin concentration ≥1000 ng/ml, clinically significant gastrointestinal motility disorder, dysphagia or swallowing disorder, or a current hemoglobin level of <10 g/dl. Female patients of childbearing potential were required to use effective contraception.
Method of Assigning Patients to Treatment Groups and Blinding
A randomization schedule was generated, each treatment packed accordingly in a block size of six (two sets of each treatment per block) and sent to the site. A unique randomization number was indicated on the label of each drug pack and sequentially dispensed to new patients at each center. Investigator sites, patients, and others involved in the trial conduct were blinded to treatment assignment until after database lock.
Fermagate and placebo capsules were identical in appearance (dark red, size 0, hard, gelatin capsules). To achieve blinding all subjects received 24 capsules a day regardless of treatment arm; subjects in the fermagate 2-g arm received 24 fermagate capsules, those in the fermagate 1-g arm received 12 fermagate and 12 placebo capsules, and those in the placebo arm received 24 placebo capsules.
A subject's vitamin D supplementation was to remain unchanged during the study period. Calcium supplements were not given and patients were not on cinacalcet.
Blood samples for the measurement of hematology (including coagulation) and biochemistry safety and efficacy parameters were performed during a normal dialysis visit at screening three times per week from the start of the current therapy period to the end of the treatment period, and once per week during follow-up. Total cholesterol (to assess any fermagate binding), transferrin, ferritin, serum iron, and total iron binding capacity were measured at the last visit of the washout and treatment periods. Patient dialysate composition and dialysis prescription remained unchanged throughout the study.
Statistical and Analytical Plans
An analysis of covariance (ANCOVA) model was used to test the intention-to-treat (ITT) and other populations for differences in the change in serum phosphate from baseline (washout) to the treatment period, with mean phosphate level at baseline (washout) as a covariate. Estimates of the difference between treatments were presented with 95% confidence intervals for the treatment comparisons: fermagate 2 g
versus fermagate 1 g, fermagate 2 g versus placebo, and fermagate 1 g versus placebo.
Categorical data were presented using frequency counts and percentages, and continuous data were presented using summary statistics. All statistical tests were two-sided.
Because this was the first study of fermagate in patients, no formal sample size calculation was carried out. Twenty patients per treatment arm was considered a reasonable number to detect differences between treatment arms using ANCOVA.
The key populations used for analysis are detailed below.
Safety population: All patients who received at least one dose of study drug (presentation of demographic data, mean serum phosphate levels by week, and all safety data).
ITT population: All patients in the Safety population who had at least one determination of serum phosphate on or after the second visit during week 1 of the treatment period (presentation of primary efficacy endpoint data only).
A completers/compliers population (all patients who were randomized, complied with the protocol regarding phosphate levels, received at least one dose of study medication, had phosphate data from at least one visit during the last week of the study medication period, and had not taken any phosphate binder medication during the washout and study medication period) was also analyzed, the results of which were consistent with those of the ITT, but are not presented because of the high dropout rate.
Primary Efficacy Variable
The primary efficacy variable was the change in serum phosphate during the treatment period against untreated baseline (washout). Untreated baseline serum phosphate was calculated as the mean of the last two measurements from the last week of the washout period.
The serum phosphate level from the treatment period was calculated as the mean of the last two measurements taken during the last week of treatment. If only one measurement was recorded, the mean of this measurement and the measurement carried forward from the previous visit was calculated. Change from mean phosphate level at baseline (washout) to the treatment period was calculated using the means described above.
Patient disposition is shown in
Figure 3. Ninety-three patients were entered into the study. Of these, 30 were considered to be screening failures and were not randomized to study treatment. Sixty-three patients were randomized to receive study treatment. Twenty-four patients withdrew from the treatment period [20 because of adverse events (AEs), 1 noncompliance, 1 withdrawn consent, 1 protocol violation and 1 serum phosphate out of range].
Across the study arms (
Table 1), there were no major variations in the demographic data that might have generated a bias in treatment outcomes. Most patients included in the study were male and Caucasian. The selection procedures resulted in a patient population likely to be representative of the wider hemodialysis population in the United Kingdom ( ). 17
Of the 63 patients randomized to treatment, 40.3% were recorded as having deviated from the protocol (53 deviations; 21 in the placebo arm and 16 in each of the fermagate 1- and 2-g arms). The population analysis definitions were adhered to in terms of identification of patients placed for each type of analysis. All randomized subjects had been receiving a phosphate binder at the start of the study. These were sevelamer-containing binders (27 subjects), calcium carbonate (
), Calcichew ( 22 ), calcium acetate ( 15 ), aluminum hydroxide ( 10 ), ADCAL-D3 ( 4 ), and magnesium carbonate ( 1 ). 1
The first patient, first visit, took place on February 10, 2004 and the last patient, last visit, took place on May 23, 2005.
Serum Phosphate Levels
Table 2 shows the mean serum phosphate values for each study period and by treatment for the ITT population. The mean current therapy serum phosphate level across all treatment groups was 1.53 mmol/L, which, in the absence of phosphate binder during the washout period, rose to a mean of 2.16 mmol/L at baseline (washout). Mean serum phosphate values fell during treatment with fermagate, decreasing by 0.46 and 0.70 mmol/L, respectively, in the 1- and 2-g arms; the placebo arm showed little change compared with baseline. The mean serum phosphate of the 2-g fermagate arm was below that of each treatment arm during the current therapy period.
Figure 4 shows the mean serum phosphate levels of each of the treatment arms by study visit. By the third visit of week 1 on treatment, the mean serum phosphate levels of the 1- and 2-g fermagate groups had decreased to below the upper Kidney Disease Outcomes Quality Initiative (K/DOQI) target of 1.78 mmol/L.
During the current therapy period, most patients in each treatment group had a mean serum phosphate level within the upper K/DOQI serum phosphate limit. Receiving no phosphate binder in the washout period, most patients exceeded this upper limit in each treatment group. On treatment, most patients in both fermagate treatment arms had a mean serum phosphate that was below the upper K/DOQI limit, whereas most patients in the placebo arm remained at levels observed at baseline (
In a pairwise comparison, there was a statistically significant difference in serum phosphate levels in the 1- and 2-g fermagate arms in comparison with placebo; although, between the two fermagate treatment arms, the differences in serum phosphate did not reach significance (
Table 4). AEs
In the safety population, 56 (88.9%) patients reported a total of 242 AEs during the course of the study. The number of patients reporting AEs during each period of the study is shown in
Table 5. Between the two fermagate arms, there was a statistically higher number of patients who reported AEs in the 2-g arm ( P = 0.0205).
Overall, 161 of 242 AEs (67%) were categorized as mild in severity.
Table 6 shows the incidence of treatment-emergent AEs in ≥5% of the safety population, which were mainly related to gastrointestinal events. The incidence and severity of diarrhea, discolored feces, and dyspepsia appeared treatment-related. One patient had an arteriovenous occlusion considered by the investigator to be severe and possibly related to study medication. Events Leading to Withdrawal
Of the 63 patients in the safety population, 23 (36.5%) did not complete the study. Of these, 22 (95.7%) withdrew during the treatment period, and 1 (4.3%) withdrew during follow-up. Across all treatment arms, the most frequent reason for withdrawal during the treatment and follow-up periods was an AE, accounting for 20 (31.8%) patients (placebo, 6; 1-g arm, 1; 2-g arm, 13) (
Fifteen patients were withdrawn after gastrointestinal AEs (31 events). Of these, 23 events of diarrhea and/or discolored feces were reported by 11 patients (placebo, 1; 1-g arm, 1; 2-g arm, 9).
Table 7 shows these and other events by treatment arm. Serious AEs and Deaths
Ten patients in the safety population reported a total of ten serious AEs during the study (placebo, 5; 1-g arm, 1; 2-g arm, 4), three of which (placebo, 1 patient, rectal hemorrhage; 2-g arm, 2 patients chest pain and arteriovenous fistula occlusion) were considered possibly related to study treatment by the investigator and resulted in patient withdrawal.
Two patients died during the study period (cardiac arrest during the screening (pretreatment) phase and acute myocardial infarction in a patient randomized to placebo).
Serum Magnesium and Calcium Levels
Serum magnesium levels in the ITT population within each study period, together with the change from baseline during the treatment period, are shown in
Table 8. Serum magnesium levels across all treatment groups were similar during the current therapy and washout periods. However, after treatment, they increased significantly in the 1- and 2-g fermagate treatment arms compared with placebo. Although between the fermagate arms there was no statistical significance, the levels appeared dose related. The levels had stabilized by the second week of active treatment and returned to pretreatment baseline on discontinuation of therapy at the end of the study.
During the treatment period, there were 2 (4.8%) patients on fermagate treatment, with reports of hypermagnesemia, one in each arm, and one report of increased blood magnesium in the 2-g arm. Hypermagnesemia and increased blood magnesium were each reported by one patient during follow-up. Patient withdrawals due to hypermagnesemia or increased blood magnesium are shown in
Table 7 and levels in Table 8.
There was no change of note in mean serum calcium levels across study periods or treatment arms.
There were no episodes of hypercalcemia during the washout or treatment periods, although one patient in the fermagate 1-g arm was considered to be hypocalcemic during the treatment period.
Other Hematology and Biochemistry Parameters
There were no consistent trends in hematology or biochemical indices at either fermagate dose. Specifically, there were no significant changes in iron status or erythropoiesis-stimulating agent use. Little change in cholesterol was seen during treatment compared with baseline across any treatment arm. In the ITT population, the mean increase in total cholesterol in the fermagate 1- and 2-g arms was 0.16 mmol/L.
Efficacy for the control of hyperphosphatemia remains the most significant challenge for any new phosphate binder. In this 21-d treatment fixed-dose phase II study, fermagate use reduced serum phosphate to below the K/DOQI-mandated upper limit in most patients, with acceptable tolerability at doses likely to be used in a future long-term treatment study.
Fermagate binds phosphate with a high degree of specificity
in vitro. This is related to the charge issues involved in exchange of carbonate groups for phosphate and the stearic inhibition of the rigid crystalline structure in the binding of other potential ligands.
The 1- and 2-g three-times-daily treatments were associated with significant phosphate binding over placebo, with 52 and 81% of patients, respectively, in the ITT population being below the K/DOQI upper limit for serum phosphate. These reductions were achieved despite study subjects being randomized to a dose of either 1 or 2 g three times daily rather than titrated to an optimal effective dose. It is not possible, however, to interpret data from the 2-g arm definitively because of the high dropout rate. A rapid response to treatment is evident with fermagate, with mean serum phosphate values in both fermagate treatment groups being below the upper K/DOQI limit within 1 wk.
Detailed dialysis-specific data were not collected in this study. Although this was a limitation of the design, differences in small solute clearance have only a modest effect on phosphate clearance and serum phosphate levels and are unlikely to have been a significant confounder. No other indices of poor dialysis (hemoglobin, serum albumin, etc.) varied between the treatment groups in a manner consistent with there being a significant difference in dialysis adequacy after randomization. All study subjects were on a standard hemodialysis regimen (4 h three times weekly). The phosphate intake of patients in this study is likely to have been in excess of 1000 mg/d (
The general tolerability profile is not dissimilar to studies of other phosphate binders. The reported events were predominantly gastrointestinal in origin. Several reports of fecal discoloration could possibly result from the iron content of the product itself.
Tolerability issues were evident in the 2-g three-times-daily arm. This dose is likely to be in excess of that required for phosphate reduction in most patients, although further studies are needed to elucidate the best balance between efficacy and tolerability.
A formal analysis of the efficacy and tolerability of fermagate as compared with previous phosphate binder therapy would have little scientific validity, because over time previous binders would have been selected and titrated to an optimally effective and tolerable dose in each subject.
Magnesium-containing medications have been in use as phosphate binders in hemodialysis patients for several years. The maximum tested 2-g three-times-daily (equivalent to 6 g/d) dose of fermagate contains 960 mg of elemental magnesium.
In vitro data suggest that approximately 20% of magnesium (around 190 mg at the top dose) might be liberated into the digestive tract under low pH conditions.
A dose-dependent increase in serum magnesium was observed with fermagate treatment. The mean levels of serum magnesium plateau with no individual excursions at the 2-g three-times-daily dose beyond 2.015 mmol/L. It is expected, although not determined in this study, that reduction in serum magnesium would occur during dialysis, returning elevated levels to normal or near normal. Hypermagnesemia is common in patients on hemodialysis (
) and sustained hypermagnesemia in a high proportion of peritoneal dialysis patients has also been described ( 19 ). Hypermagnesemia at levels up to 1.5 mmol/L is only rarely associated with clinical effects that are usually mild in severity ( 20 21 – ). Neuromuscular toxicity, although not evident in this study, may become apparent at levels of 2 to 3 mmol/L ( 26 ). 27
There was no evidence of an effect of treatment in this study on hematology or biochemical parameters other than serum phosphate and magnesium. Specifically, there was no effect on cholesterol or iron status, the latter group of tests indicating no apparent contribution from the iron component of the compound.
Given the size and quantities of capsules required each day and the potential for noncompliance, satisfactory improvements in mean serum phosphate levels were achieved. Mean serum phosphate was markedly lower in the fermagate 2-g arm as compared with the fermagate 1-g arm; the fact that the mean serum magnesium was, however, only slightly higher may be explained by poorer compliance at the higher dose, or the high dropout rate leading to a reduced cumulative intake of magnesium. Future fermagate formulations will address the pill count, with higher unit doses expected to reduce the daily pill burden for patients.
In conclusion, fermagate is a promising agent for the management of hyperphosphatemia in chronic hemodialysis patients. In the 1-g arm of this study, it offered a high level of efficacy with acceptable tolerability. Further studies are ongoing to define optimal doses for the management of serum phosphate.
Molecular structure of fermagate.
Schematic of study design.
Summary of patient disposition in the study, including all patients screened (see ”Statistical and Analytical Plans” for definition of populations).
Mean serum phosphate values during the current therapy, washout, study, and follow-up periods for the ITT population: primary efficacy variable. (Serum phosphate level during treatment with 1- and 2-g fermagate three times daily was significantly lower than with either washout or in comparison with placebo-treated patients.)
Summary of patient characteristics (safety population) who entered the treatment period.
Mean serum phosphate values for the ITTa population during each study period: primary efficacy variableb
Proportions of patients below the upper K/DOQIa recommended value (<1.78 mmol/L) for serum phosphate in the ITT populationb
Pairwise comparison of change in mean serum phosphate level from baseline (washout) to during treatment (ITT population)
Number and proportion of patients with AEsa by each study period (safety population)
Incidence of treatment-emergent AEs occurring in ≥5% of subjects in any treatment group (safety population; treatment and follow-up periods)a
Treatment-emergent AEs leading to patient withdrawal in the safety population
Predialysis serum magnesium values (mmol/L) during the study and change from baseline during treatment (ITT population)
A.J.H. has undertaken consultancy work for INEOS Healthcare and Shire. M.W. undertakes speaking work, has received honoraria from Amgen, and has undertaken consultancy work for INEOS Healthcare. P.P. has undertaken consultancy work for INEOS Healthcare and has received honoraria from Amgen. C.W.M. and A.J.T. are employees of INEOS Healthcare. G.W. has no conflicts of interest to declare.
Published online ahead of print. Publication date available at
. www.cjasn.org References
1. Gipstein RM, Coburn JW, Adams DA, Lee DB, Parsa KP, Sellers A, Suki WN, Massry SG: Calciphylaxis in man. A syndrome of tissue necrosis and vascular calcification in 11 patients with chronic renal failure. Arch Intern Med136 :1273– 1280,1976
2. Goodman WG, Goldin J, Kuizon BD, Yoon C, Gales B, Sider D, Wang Y, Chung J, Emerick A, Greaser L, Elashoff RM, Salusky IB: Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis. N Engl J Med342 :1478– 1483,2000
3. Janigan DT, Hirsch DJ, Klassen GA, MacDonald AS: Calcified subcutaneous arterioles with infarcts of the subcutis and skin (“calciphylaxis”) in chronic renal failure. Am J Kidney Dis35 :588– 597,2000
4. London GM, Guerin AP, Marchais SJ, Metivier F, Pannier B, Adda H: Arterial media calcification in end-stage renal disease: Impact on all-cause and cardiovascular mortality. Nephrol Dial Transplant18 :1731– 1740,2003
5. Block GA, Klassen PS, Lazarus JM, Ofsthun N, Lowrie EG, Chertow GM: Mineral metabolism, mortality, and morbidity in maintenance hemodialysis. J Am Soc Nephrol15 :2208– 2218,2004
6. DeSoi CA, Umans JG: Phosphate kinetics during high-flux hemodialysis. J Am Soc Nephrol4 :1214– 1218,1993
7. Pohlmeier R, Vienken J: Phosphate removal and hemodialysis conditions. Kidney Int Suppl78 :S190– S194,2001
8. Shinaberger CS, Kilpatrick RD, Regidor DL, McAllister CJ, Greenland S, Kopple JD, Kalantar-Zadeh K: Longitudinal associations between dietary protein intake and survival in hemodialysis patients. Am J Kidney Dis48 :37– 49,2006
9. Young EW, Akiba T, Albert JM, McCarthy JT, Kerr PG, Mendelssohn DC, Jadoul M: Magnitude and impact of abnormal mineral metabolism in hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J Kidney Dis44 :34– 38,2004
10. Qunibi WY, Nolan CR: Treatment of hyperphosphatemia in patients with chronic kidney disease on maintenance hemodialysis: Results of the CARE study. Kidney Int SupplS33– S38.2004
11. Block GA, Spiegel DM, Ehrlich J, Mehta R, Lindbergh J, Dreisbach A, Raggi P: Effects of sevelamer and calcium on coronary artery calcification in patients new to hemodialysis. Kidney Int68 :1815– 1824,2005
12. Chertow GM, Burke SK, Raggi P: Sevelamer attenuates the progression of coronary and aortic calcification in hemodialysis patients. Kidney Int62 :245– 252,2002
13. Braun J, Asmus HG, Holzer H, Brunkhorst R, Krause R, Schulz W, Neumayer HH, Raggi P, Bommer J: Long-term comparison of a calcium-free phosphate binder and calcium carbonate–phosphorus metabolism and cardiovascular calcification. Clin Nephrol62 :104– 115,2004
14. Roberts NB, Sargazi M, Rhodes NP, Toft AJ: Comparison of a new phosphate binder, a magnesium iron hydroxy carbonate compound (MFHC), with other phosphate binders. EDTA Abstract W421,2003 .
15. Newton M, Toft A, Roberts NB, Minkeus M: Study of the phosphate binding of Alpharen an iron magenesium hyroxycarbonate in a dynamic model of the gastrointestinal tract. EDTA Abstract SA P049,2007 .
16. Peers E, Chapman A, Brown C, Rhodes NP, Toft AJ: Effect of a new anion exchange compound on phosphate levels in man. EDTA Abstract SP 260,2004 .
17. U.K. Renal Registry Report 2007. Bristol, The Renal Association,2007
18. Calvo MS, Park YK: Changing phosphorus content of the U.S. diet: Potential for adverse effects on bone. J Nutr126 :1168S– 1180S,1996
19. Saha H, Harmoinen A, Pietila K, Morsky P, Pasternack A: Measurement of serum ionized versus total levels of magnesium and calcium in hemodialysis patients. Clin Nephrol46 :326– 331,1996
20. Navarro JF, Mora C, Macia M, Garcia J: Serum magnesium concentration is an independent predictor of parathyroid hormone levels in peritoneal dialysis patients. Perit Dial Int19 :455– 461,1999
21. Delmez JA, Kelber J, Norword KY, Giles KS, Slatopolsky E: Magnesium carbonate as a phosphorus binder: A prospective, controlled, crossover study. Kidney Int49 :163– 167,1996
22. Guillot AP, Hood VL, Runge CF, Gennari FJ: The use of magnesium-containing phosphate binders in patients with end-stage renal disease on maintenance hemodialysis. Nephron30 :114– 117,1982
23. Moriniere P, Vinatier I, Westeel PF Cohemsolal M, Belbrik S, Abdulmassih Z, Hocine C, Marie A, Leflon P, Roche D: Magnesium hydroxide as a complementary aluminium-free phosphate binder to moderate doses of oral calcium in uraemic patients on chronic haemodialysis: Lack of deleterious effect on bone mineralisation. Nephrol Dial Transplant3 :651– 656,1988
24. O'Donovan R, Baldwin D, Hammer M, Moniz C, Parsons V: Substitution of aluminum salts by magnesium salts in control of dialysis hyperphosphataemia. Lancet1 :880– 882,1986
25. Oe PL, Lips P, van der Meulen J, de Vries PM, van Bronswijk H, Donker AJ: Long-term use of magnesium hydroxide as a phosphate binder in patients on hemodialysis. Clin Nephrol28 :180– 185,1987
26. Parsons V, Baldwin D, Moniz C, Marsden J, Ball E, Rifkin I: Successful control of hyperparathyroidism in patients on continuous ambulatory peritoneal dialysis using magnesium carbonate and calcium carbonate as phosphate binders. Nephron63 :379– 383,1993
27. Navarro JF, Mora-Fernandez C: Magnesium in Chronic Renal Failure. In: New Perspectives in Magnesium Research: Nutrition and Health, edited by Nishizawa Y, Morii H, and Durlach J, London, Springer-Verlag London Ltd.,2007