Magnesium carbonate as a phosphorus binder: A prospective, controlled, crossover study

Kidney International, Vol. 49 (1996), pp. 163—167

Magnesium carbonate as a phosphorus binder: A prospective, controlled, crossover study JAMES A. DELMEZ, JACOBO KELBER, KATHRYN Y. NORWORD, KARLA S. GILES, and EDUARDO SLATOPOLSKY Renal Division and Chromalloy American Kidney Center, Washington University School of Medicine, Saint Louis, Missouri, USA

Magnesium carbonate as a phosphorus binder: A prospective, controlled, crossover study. The use of calcium carbonate (CaCO3) to bind phosphorus (P) in chronic hemodialysis patients has been a popular tactic in the past decade. Nonetheless, problems with hypercalcemia decrease its usefulness, particularly in patients treated with calcitriol. A P binder not containing calcium (Ca) would be of value in these circumstances. In short-term studies, we showed that magnesium carbonate (MgCO5) was well-tolerated and controlled P and Mg levels when given in conjunction with a dialysate Mg of 0.6 mg/dl. We, therefore, performed a prospective, randomized, crossover study to evaluate if the chronic use of MgCO3 would allow a reduction in the dose of CaCO3 and yet achieve acceptable levels of Ca, P, and Mg. We also assessed whether the lower dose of CaCO3 would facilitate the use of larger doses of calcitriol. The two phases were MgCO3 plus half the usual dose of CaCO3 and CaCO3 alone given in the usual dose. It was found that MgCO3 (dose, 465 52 mg/day elemental Mg) allowed a decrease in the amount of elemental Ca ingested from 2.9 0.4 to 1.2 0.2 glday (P < 0.0001). The Ca, P, Mg levels were

the same in the two phases. The maximum dose of iv. calcitriol without causing hypercalcemia was 1.5 0.3 jgItreatment during the MgCO3 phase and 0.8 0.3 g/treatment during the Ca phase (P < 0.02). If these studies are confirmed, the use of MgCO3 and a dialysate Mg of 0.6 mg/dl may be considered in selected patients who develop hypercalcemia during treatment with iv. calcitriol and CaCO3.

In chronic renal failure, control of serum phosphorus (P) concentrations is a decisive factor in the prevention and treatment of secondary hyperparathyroidism [1 1. Dietary P restriction may

tients with very high or low bone turnover 8, 9]. We have previously found that decreasing the calcium (Ca) concentration in the dialysate from 6.5 to 5 mg/dl allowed for excellent control of P with CaCO3 without hypercalcemia [10]. Nonetheless, significant problems with phosphorus binders persist, particularly in patients receiving calcitriol. Treatment with this sterol increases both Ca and P absorption by the intestine [11]. Hence, patients treated with calcitriol are at risk to develop hypercalcemia and hyperphosphatemia. Acute studies have shown that P is bound by half the amount of elemental Ca when given as calcium acetate compared to carbonate [12, 13]. However, no long-term studies have demonstrated that this results in less hypercalcemia [14—16]. A potent calcium-free P binder would be ideal. The P binding properties of Mg were extensively studied by Fine et al in acute studies involving normal subjects [17]. They found a dose-dependent decrease in P absorption from 75% with placebo to 28% with

77 mEq magnesium acetate. The use of a P binder containing magnesium (Mg) used in conjunction with a low or Mg-free dialysate may therefore be a possible strategy in chronic renal failure. In short-term studies, we have shown that a Mg-free dialysate was poorly tolerated in patients undergoing high-efficiency hemodialysis with one half developing severe muscle cramps [18]. A dialysate containing 0.6 mg/dl, however, was well

tolerated when used in conjunction with magnesium carbonate (MgCO3) to bind P. The purposes of the current study were: (1) be useful, but excessive limitations may lead to an inadequate to determine if the chronic use of MgCO3 in conjunction with a dietary protein intake and malnourishment. There are currently 0.6 mg/dl Mg dialysate would allow a reduction in the dose of no routine dialytic treatments or regimens that remove enough P CaCO3 and yet achieve acceptable levels of Ca, P and Mg; and (2) to maintain balance in the well-nourished patient [2, 3]. There- to assess if the lower intake of CaCO3 would result in less fore, compounds that bind P in the bowel have been used with hypercalcemia, allowing the administration of a higher dose of some success for over 25 years. The initial enthusiasm for P intravenous (i.v.) calcitriol. binders containing aluminum slowly faded when it was realized that the metal accumulated systemically causing osteomalacia, Methods dementia, myopathy, and anemia [4—6]. Although Clarkson, McDonald and de Wardener studied the effects of CaCO3 on P Subjects absorption in 1966, its widespread popularity as a P binder has The criteria for selection of subjects included treatment of renal been largely confined to the last decade [7]. Often large doses are failure with hemodialysis at our institution for more than six required and hypercalcemia is common, especially in those pamonths, parathyroid hormone (PTH) levels > 8.0 ng/ml, and good

dietary and medical compliance as assessed by the dietary, Received for publication June 14, 1995 and in revised form August 21, 1995 Accepted for publication August 21, 1995

© 1996 by the International Society of Nephrology

nursing, and medical staff. No patient was treated with calcitriol within the month preceding the start of the protocol. Twenty-nine patients agreed to participate in the study. Six subjects died during the two year study. The causes of death were an acute myocardial infarction in four and sepsis in two subjects. Three were found to

163

164

Delmez et a!: Magnesium carbonate as a phosphorus binder

be non-compliant when the CaCO3 tablets were counted (see alone. After four weeks of this washout phase, subjects entered below) during the run-in phase. Two subjects withdrew during the the other phase and were managed in a fashion identical to that run-in phase because of headaches and myalgias, respectively. A described above. Thus, the protocol was a randomized, prospecsevere intercurrent illness requiring a prolonged hospitalization tive crossover design. In both groups, the dialysate Ca was 5 mg/dl. When subjects forced a withdrawal from the study in two subjects. One subject transferred to a different dialysis unit. There was no relationship were treated with MgCO3, a dialysate Mg of 0.6 mg/dl was used. between any of the study phases and patient withdrawal or death. Thus, 15 subjects successfully completed the protocol. The age four years (range 27 to 77). Six were male and nine was 58 female. Twelve were black and three were white. The duration of dialysis was 75 11 months (range 21 to 175). The causes of renal failure were hypertension (7), diabetes (3), chronic glomerulone-

This dialysate solution was prepared by mixing one volume of 1.8

The Kt/V was maintained between 1.2 to 1.4 in all subjects with reprocessed Clirans 175 (Terumo, Tokyo, Japan) dialyzers. The blood flow was 450 mI/mm and dialysate flow 500 to 800 ml/min with Fresenius 2008 D (Seratronics, Concord, CA, USA) dialysis machines using bicarbonate dialysate.

done weekly to ensure compliance. Three day diet histories, done during the baseline phase, showed a dietary intake of 951 48, 438 46, and 180 12 mg/day for P, Ca, and Mg, respectively [19]. The diet histories were repeated in the two phases and did not change (data not shown). Alkaline phosphatase, albumin, and potassium levels were measured monthly.

mg/dl Mg dialysate (Naturalyte 9018, National Medical Care, Rockleigh, NJ, USA) with two volumes of 0 mg/dl Mg dialysate (Naturalyte 9032). During treatment with CaCO3 alone, the dialysate Mg was 1.8 mg/dl. All subjects received erythropoietin and oral iron. During the entire protocol, the laboratory data were reviewed phritis (2), polycystic kidney disease (1), obstruction (1), and weekly by the dietitians and nephrologists. Adjustments in the systemic lupus erythematosus (1). No subject had a parathyroidectomy. Written consent was obtained from all patients, and the binder regimen were made as necessary to attain the target Ca studies were approved by the Human Research Committee of and P levels. All binders were given with meals in proportion to the estimated average P intake with each meal. Pill counts were Washington University.

Protocol At the start of the run-in phase, subjects were switched from generic CaCO1 to a single CaCO3 formulation (Calci-MIxTM 1250 mg, R&D Laboratories, Inc., Marina del Rey, CA, USA) to insure uniform dissolution of the salt in binding dietary P in the subjects.

Calcium, P, Mg, and PTH levels were measured weekly in this four- to eight-week period. The amounts of CaCO3 prescribed were adjusted weekly to attain target serum Ca levels between 9.5 and 10.5 mg!dl and P levels less than 6 mg/dl. If the target levels were attained for four consecutive weeks, subjects were randomized to a phase where they each received half the amount of CaCO3 required in the run-in phase plus MgCO3 (Mg phase) or a phase in which the amount of CaCO3 prescribed did not change (Ca phase). The initial dose of MgCO3 was 750 mg/day (214 mg elemental Mg), which was titrated weekly as necessary for four to eight weeks to attain the target P of less than 6 mg/dl. If the target Ca and P levels were achieved for four consecutive weeks, the patients in both groups entered the calcitriol phase where 2 j.ig was injected intravenously after each dialysis three times a week.

Measurements

The concentration of total Ca and Mg was determined by atomic absorption spectroscopy (Model 503; Perkin-Elmer, Norwalk, CT, USA). The normal range of plasma Mg in our labora-

tory is 1.6 to 2.6 mg/dl. Phosphorus was measured by the Technicon Autoanalyzer (Technicon Instruments, Tarrytown, NY, USA). Serum levels of PTH were assessed by the CH9 antibody that recognizes the intact, midregion, and C-terminal portion of the hormone. The normal range is 360 pg/ml. The method has previously been described in detail [20]. Statistics

The Wilcoxon signed rank test was used to evaluate paired nonparametric data. The t-test was applied to parametric data. Spearman rank was used to test correlations. For statistical analysis of three or more groups, a one-way repeated measures analysis of variance was used for parametric data and Friedman repeated measures analysis of variance on ranks for nonparamet-

Calcium, P, Mg and PTH levels were measured weekly. If nc data. The statistical software was SigmaStat (Jandel Scientific, hypercalcemia (Ca > 11 mg/dl) did not occur in this two week San Rafael, CA, USA). Results are reported as mean SEM. period, the dose of calcitriol was increased to 3 ig after each Results dialysis for two weeks. If the calcium levels remained 11 mgldl, the dose was increased to its maximum of 4 jg i.v. three times a The mean of 10 weekly Ca, P, and Mg levels were calculated for week. This dose was continued for 10 weeks if hypercalcemia did each subject in the Mg and Ca phases. There were no significant not occur. If hypercalcemia developed at any time, the calcitriol differences (Fig. 1). The mean total Ca was 10.2 0.2 mg/dl was held and restarted at a dose of 3 ag when the hypercalcemia during the both the Mg and Ca phases. Both values were resolved. If hypercalcemia recurred in the following 10 weeks, the statistically greater (P < 0.05) than the mean 10 weekly Ca levels calcitriol was held and restarted when the serum Ca decreased to of 9.7 0.2 mg/dl measured during the run-in control phase. The < 11 mg/dl at a dose of 2 gig. The process was repeated using mean P level was 5.7 0.2 mg/dl during the Mg phase and 5.2 doses of I jig, 0.5 tg, and 0 j.ig if Ca> 11 mg/dl was noted in the 0.2 mg/dl in the Ca phase. These values were not different but weekly measurements. Thus, the outcome was the highest calcit- were marginally higher than the levels of 4.9 0.2 mg/dl in the rio! dose that did not cause hypercalcemia over a period of 10 run-in phase (P = 0.06). The Mg levels were 2.9 0.2 and 3 0.1 weeks. mgldl in the Mg and Ca phases, respectively. These were not Upon successful completion of the 10 week period, calcitriol different from the run-in phase where the mean Mg levels were 3.1 was discontinued and all patients returned to the regimen used in 0.1 mg/dl. The control of P in the Mg phase occurred despite the run-in phase when the subjects were treated with CaCO5 the ingestion of less than the amount of Ca binders compared to

165

Intravenouscalcitriol supplement, tig/treatment

W

()

CaCO3.

Fig. 3. Maximum tolerated dose of i. v. calcitriol in the two phases. See text for the definition of maximal tolerated dose of intravenous calcitriol. P < 0.02.

C)

0

N3

a

a

Serum concentration, mp/dj

0

N)

Delmez et al: Magnesium carbonate as a phosphorus binder

in Figure 3, the mean tolerated calcitriol dose was 1.5 0.3 xg/treatment during the Mg phase and 0.8 0.3 rg/treatment

MgCO3;

o

0

-s

o o

co 0 o 0 0

Oral calcium supplement, giday -s Ce

a cii 0 0 0 0

a

(N = 15) in the two phases. Symbols are: ()

Oral magnesium supplement, mg/day

Fig. 1. Mean SE of 10 weekly Ca, P, and Mg calculated from each subject

during the Ca phase (P < 0.02). There were no differences in the mean PTH levels in the Mg and Ca phases (47 8 ng/ml vs. 48 10 ng/ml, respectively). If, however, the percent difference in PTH levels in the Ca phase versus Mg phase was plotted against the absolute difference in calcitriol dose (xg, Mg phase — Ca phase), a highly significant reciprocal relationship was found (r = 0.1, 3.9 —0.8468, P < 0.0001). The albumin levels were 3.9

0.1, and 4.0

0.1 g/dl in the Mg, Ca, and run-in phase,

respectively. The alkaline phosphatase level in the run-in phase was 92 8 lU/liter. This value, although within the normal range (38 to 126 lU/liter), was greater than those of the Mg and Ca

phases, where they were 78 6 and 78 9 lU/liter, respectively Fig. 2. Mean SE dose of elemental Ca and Mg used to bind P in the two

phases. Symbols are: ) MgCO3;

()

CaCO3.

(P < 0.05). The mean potassium levels were 4.9 0.1 in the Mg phase and 4.9 0.1 mEq/liter in the Ca phase (P = NS). Discussion Several investigators have evaluated the use of Mg-containing P

the Ca phase. As shown in Figure 2, the amount of elemental Ca used was 1.2 0.2 glday during the Mg phase and 2.9 0.4 g!day binders in dialysis patients. Guillot et a! treated nine patients during the Ca phase (P < 0.0001). The elemental Ca ingested undergoing conventional hemodialysis with Mg(OH)2 for three to during the run-in phase was 3.0 0.4 g/day, a value not different five weeks [211. Using doses averaging 734 mg of elemental from the CaCO3 phase. The mean dose of elemental Mg pre- Mg/day and dialysate Mg concentrations of 1.2 to 1.8 mg/dl, the P scribed to bind phosphorus in the Mg phase was 465 52 mg/day levels fell from a control (no binders) value of 9.0 to 8.1 mg'dl. (range 214 to 858 mg/day). Despite the relatively large dose of The mean Mg levels were 4.32 mg/dl. Four of nine developed MgCO3, the binder was well tolerated. No subject developed diarrhea. Mactier et al could find no effect of Mg Trilisate on P gastrointestinal symptoms such as loose stools, diarrhea, or bloat- levels in either hemodialysis or peritoneal dialysis patients [22]. ing. As previously reported, the dialysate Mg concentration of 0.6 Eighteen patients undergoing conventional hemodialysis were mg/dl was also well accepted with no adverse clinical sequelae switched from Al(OH)3 to Mg(OH)2 by Oe et al [23]. The P levels rose from 4.3 to 6.1 mg/dl despite an average intake of 991 mg of [181. The total number of episodes of hypercalcemia per patient elemental Mg. The Mg level averaged 4.3 mg/dl while the patients while receiving any dose of calcitriol was 1.8 0.2 during the Mg received Mg(OH)2 and were treated with a dialysate devoid of phase and 2.7 0.3 during the Ca phase (P = 0.03). The overall Mg. The potassium levels were significantly higher when patients incidence of hypercalcemia was 8.9% with mean hypercalcemic received Mg(OH)2 compared to the control phase (5.7 0.3 vs. 0.4 mEq/liter). More favorable results were reported by values of 11.6 0.2 mg/dl (range 11.1 to 14.4 mg/dl). One patient 5.1 developed pruritus and nausea which lasted two days after the O'Donovan et al [24]. They switched 28 patients undergoing conventional hemodialysis from A!(OH)3 to MgCO3 in combinadiscontinuation of calcitriol. The lower dose of CaCO3 allowed for the administration of a tion with a Mg-free dialysate. Over the two-year study period, Ca, higher dose of calcitriol without causing hypercalcemia. As shown P, and Mg levels were well controlled and not different than those

166

Delmez et al: Magnesium carbonate as a phosphorus binder

in the control phase. The amount of elemental Mg used varied between 155 to 465 mg/day, doses similar to this study. Diarrhea was mild and transient. Similar data were reported by Moriniere et at [25]. They also reported severe hyperkalemia as high as 8 mEq/liter in many patients, the etiology of which was unclear. More recently, this same group performed a controlled study where patients were either treated with CaCO3 plus Mg(OH)2 as needed or Mg(OH)2 alone and 1-alpha-hydroxyvitamin D3 [26]. The latter group was not superior to the former in suppressing PTH secretion due to uncontrolled hyperphosphatemia causing a reduction in the dose of 1-alpha-vitamin D3. It is noteworthy that P levels did not increase despite decreasing

mineralization improved in six patients after the dialysate Mg was

decreased to 0.6 mg/dl [29]. However, Moririiere, Vinatier and Westeel noted no significant changes in bone histomorphology after 8 to 20 months of treatment with Mg(OH)2, during which

time the Mg levels averaged 3.6 mg/dl [25]. Mg may have beneficial vascular effects. Meema, Oreopoulos and Rapaport, for example, found that patients on peritoneal dialysis with high levels of serum Mg had less evidence of vascular calcification than those with lower levels [30]. They postulated that Mg may inhibit the calcification process.

Although the PTH levels were lower in both the Mg and Ca phases than the control run-in phase, they were not different

the dose of elemental Ca by art average of 85 mEq/day while prescribing an average of 39 mEq/day of Mg. This would imply that MgCO3 is a more potent P binder than CaCO3. In vitro studies, however, suggest that Mg, when studied in the form of Mg(OH)2, is less effective that CaCO3 [13]. Acute studies in normal subjects also suggest that CaCO3 is a more effective P binder than magnesium acetate [13, 17]. There are a number of

between the two experimental phases. A likely reason is that PTH

It is also conceivable that Mg and Ca salts may act synergistically to bind P, although this has not been studied. Finally, there may be differences in P binding between the different salts of Mg. To

compared to the Ca phase. Although we found the strategy of using MgCO3 to bind P in

levels were not a primary end-point in our study design. Nonetheless, the suppression of PTH was dose-dependent. This seemingly inconsistency was due to the lower levels of PTH in the Ca phase seen in those patients whose calcitriol dose was the same in

both phases. Cantley et al have clearly shown that calcitriol suppresses PTH secretion in a dose-dependent manner in vitro possible explanations for this apparent discrepancy. The P binding [31]. Nonetheless, patients in the Mg phase tolerated a higher properties of Mg and Ca salts may change when given chronically. dose of calcitriol without the development of hypercalcemia conjunction with a low Mg dialysate and decreased dose of CaCO3

our knowledge, there has been no direct comparison of the

to be well tolerated and efficacious in this study, further investiefficacy or tolerance of various Mg salts in the setting of renal gations are warranted before it can be recommended. These failure. Finally, the similar P levels in the two groups may not would need to include an evaluation of bone histomorphology. represent comparable gastrointestinal absorption of P. Magne- Nonetheless, its use may be considered in selected hemodialysis sium and other salts can also affect P levels by altering skeletal and patients with severe hyperparathyroidism whose treatment with calcitriol is limited by the development of hypercalcemia and/or soft tissue deposition and/or resorption. In the current study, we reasoned that MgCO3 would he a hyperphosphatemia. useful adjunct to CaCO3 in binding P and that lowering the amount of CaCO3 ingested would allow for the administration of Acknowledgments higher doses of calcitriol, This was found to be correct. The This study was supported in part, by U.S. Public Health Service dialysate Mg concentration of 0.6 mg/dl was well tolerated and did not cause the muscle cramps we previously observed with patients undergoing high efficiency hemodialysis using a dialysate devoid

of Mg. In addition, the use of MgCO3 as a P binder was well

tolerated with no gastrointestinal side effects. This is in contrast to

the studies previously cited. Possible reasons for the tolerance

NTADDK grants DK49240, DK-09976, DK-07126 and RR-00036. Por-

tions of this work were presented at the 26th Annual Meeting of the American Society of Nephrology in Boston, Massachusetts, November 14—17, 1993 and were published in abstract form (J Am Soc Nephrol 4:698A, 1993). The authors thank Ms. Katie Fox of National Medical Care who kindly provided the dialysate, and Mr. Tom Wagner and Mr. Richard Ganz from Abbott Laboratories, for supplying the calcitriol for this study.

include the use of CO3 instead of OH2 as the anion, the Reprint requests to James A. Delmez, M.D., 660 South Euclid Ave., Box concomitant administration of CaCO3 and/or ferrous sulfate, and the use of smaller amounts of elemental Mg to bind P. We found 8129, St. Louis, Missouri 63110-1093, USA. no differences in the Ca levels in the two phases despite adminReferences istering twice the amount of calcitriol in the Mg phase. Presumably this was due to the lower dose of CaCO3 used for P binding. 1. DELMEZ JA, SLATOPOLSKY E: Hyperphosphatemia: Its consequences

That the P levels were similar in the two phases attests to the efficacy of using MgCO3 as a phosphorus binder. The similar values of Mg in the two phases reflected, in part, the greater removal of Mg with the dialysate lower in Mg. We have previously

reported a net removal of 306 69 mg/treatment using a dialysate Mg of 0.6 mg/dl versus 56 50 mg/treatment with a concentration of 1.8 mg/dl. The net Mg balances in the two phases, however, was

beyond the scope of the study.

Most investigators have shown a good correlation between bone and serum Mg. It is not surprising that bone Mg is elevated in renal failure. The consequences of this are controversial. It has

been suggested that Mg may inhibit the formation of appatite ciystals and thereby inhibit bone formation in vitro [27, 28]. In support of this contention is the finding by Gonella et at that bone

and treatment in patients with chronic renal disease. Am J Kidney Dis 19:303—317, 1992 2. DELMEZ JA, SLATOPOLSKY E, MARTIN KJ, GEARING BN, HARTER HR:

Minerals, vitamin D and parathyroid hormone in continuous ambulatory peritoneal dialysis. Kidney mt 21:862—867, 1982

3. Hou SH, Zuo J, ELLMAN CF, Hu J, GRIFFIN Z, SPIEGEL DM, BOURDEAU JE: Calcium and phosphorus fluxes during hemodialysis with low calcium dialysate. Am J Kidney Dis 18:217—224, 1991 4. ELLIOTT HL, DRYBURGI-I F, FELL GS, SABET S, MACDOUGALL Al: Aluminum toxicity during regular haemodialysis. Br Med J 1978:1101— 1103, 1978 5. PARKINSON IS, WARD MK, FEEST TD, FAWCETT RWP, KERR DNS:

Fracturing dialysis osteodystrophy and dialysis encephalopathy. An epidemiological survey. Lancet i:406—409, 1979 6. PLATTS MM, GOODE GC, HISLOP JS: Comparison of the domestic

water supply and the incidence of fractures and encephalopathy in patients on home dialysis. Br Med J ii:657—660, 1977

Delmez et al: Magnesium carbonate as a phosphorus binder

167

7. CLARKSON EM, MCDONALD SJ, DE WARDENER HE: The effect of a

20. HRUSKA KA, KOPELMAN R, RUTHERFORD WE, KLAHR 5, SIToPoL-

high intake of calcium carbonate in normal subjects and patients with chronic renal failure. Clin Sci (Land) 30:425—438, 1966

SKY E: Metabolism of immunoreactive parathyroid hormone in the dog: The role of the kidney and the effects of chronic renal disease. J

8. PEI Y, HERCZ G, GREENWOOD C, SHERRARD D, SEGRE G, MANUEL A,

SAIPH00 C, FENTON S: Non-invasive prediction of aluminum bone disease in hemo- and peritoneal dialysis patients. Kidney mt 41:1374— 1382, 1992 9. MERIC F, YAP P, BIA MJ: Etiology of hypercalcemia in hemodialysis patients on calcium carbonate therapy. Am J Kidney Dis 16:459—464, 1990 10. SLATOPOLSKY E, WEERTS C, NORWOOD K, GILES K, FRYER P, FINCH

J, WINDUS D, DELMEZ J: Long term effects of calcium carbonate and

2.5 mEq/L calcium dialysate on mineral metabolism. Kidney mt

Clin Invest 56:39—48, 1975

21. 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. Nephron 30:114—117, 1982 22. MACTIER RA, LEUNG ACT, HENDERSON IS, DoaBlE JW: Control of

hyperphosphatemia in dialysis patients: Comparison of aluminum hydroxide, calcium carbonate, and magnesium trilisate. Dial Transplant 16:599—601, 1987 23. OE PL, Lirs P, VAN DER MEULEN J, t)E VRIES PMJM, VAN BRONSWIJK

ii. BRICKMAN AS, HARTENBOWER DL, NORMAN AW, COBURN JW:

H: Long-term use of magnesium hydroxide as a phosphate binder in patients on hemodialysis. Clin Nephrol 28:180—185, 1987

Action of 1-hydroxyvitamin D3 and 1,25-dihydroxyvitamin D1 on

24. O'DONOVAN R, BALDWIN D, HAMMER M, MoNlz C, PARSONS V:

mineral metabolism in man. Am J Clin Nutr 30:1064—1069, 1977 12. MAI ML, EMME'I-F M, SHEIKH MS, SANTA ANA CA, SCHILLER L, FOROTRAN JS: Calcium acetate, an effective phosphorus binder in patients with renal failure. Kidney mt 36:690—695, 1989

Substitution of aluminium salts by magnesium salts in control of

36:897—903, 1989

13. SImIKII MS, MAGUIRE JA, EMMEI-r M, SANTA ANA CA, NICAR MJ,

SCHILLER LR, FORDTRAN iS: Reduction of dietaiy phosphorus ab-

dialysis hyperphosphataemia. Lancet 1:880—882, 1986 25. MORINIERE P, VINATIER I, WESTEEL PF: Magnesium hydroxide as a

complementary aluminum-free phosphate binder to moderate doses of oral calcium in uraemic patients on chronic haemodialysis. Nephrol Dial Transplant 3:651—656, 1988

sorption by phosphorus binders: A theoretical in vitro and in vivo

26. MORNIERE P, MAUROUARD C, BOUDAILLIEZ B, WESTEEL P, ACHARD

study. J Clin Invest 83:66—73, 1989 14. SCHAEFER K, SCHEER J, ASMUS G, UMLAUF E, HAGEMANN J, VON

J, Borrr F, EL ESPER N, COMPAGNON M, MAUREL G, BOUILLON R,

HERRATH D: The treatment of uraemic hyperphosphataemia with calcium acetate and calcium carbonate: A comparative study. Nephrol Dial Transplant 6:170—175, 1991 15. MORINIERE P, DJERAD M, BOUDAILLIEZ B, EL ESPER N, BoIrrE F, WESTEEL PF, COMPAGNON M, BRAZIER M, ACHARD JM, Fou RNIER A:

Control of predialytic hyperphosphatemia by oral calcium acetate and calcium carbonate. Nephron 60:6—11, 1992 16. DELMEZ JA, TINDJRA CA, WINDUS DW, NoRw000 KY, GILLS KS, NIGHSWANDER TL, SLATOPOLSKY E: Calcium acetate as a phosphorus

binder in hemodialysis patients. JAm Soc Nephrol 3:96—102, 1992 17. FINE KD, SANTA ANA CA, PORTER JL, FORDTRAN JS: Intestinal absorption of magnesium from food and supplements. J Clin Invest 88:396—402, 1991

18. KELBER J, SLATOPOLSKY E, DELMEZ J: Acute effects of different concentrations of dialysate magnesium during high-efficiency dialysis. AmJKidneyDis 24:453—460, 1994 19. ADAMS CF: Nutrititive Value of American Foods in Common Units. Washington, D.C., Agricultural Research Service, United States Department of Agriculture, 1975, Handbook No. 456

PAMPHILE R, FOURNIER A: Prevention of hyperparathyroidism in patients on maintenance dialysis by intravenous 1-aipha-hydroxyvita-

mm D3 in association with Mg(OH)2 as sole phosphate binder. Nephron 60:154—163, 1992

27. BACHRA BN, FISCHER HRA: The effect of some inhibitors on the nucleation and crystal growth of apatite. Calcif Tissue Res 3:269—279, 1969

28. BOSKEY AL, POSHNER AS: Effect of magnesium on lipid-induced calcification: An in vitro model of bone mineralization. Calcif Tissue Int 32:139—143, 1980 29. GONELLA M, BALLANTI P, ROCCA CD, CALABRESE G, PRATESI G, VAGELLI G, MAZZOTFA A, BONUCCI E: Improved bone morphology by

normalizing serum magnesium in chronically hemodialyzed patients.

Miner Electrol Metab 14:240—245, 1988 30. MEEMA HE, OREOPOULOS DG, RAPOPORT A: Serum magnesium level and arterial calcification in end-stage renal disease. Kidney mt 32:388— 394, 1987 31. CANTLEY LK, RUSSELL J, LETFIERI D, SHERWOOD LM: 1,25-dihy-

droxyvitamin D3 suppresses parathyroid hormone secretion from parathyroid cells in tissue culture. Endocrinology 117:2114—2119, 1985