Kidney International, Vol. 56, Suppl. 73 (1999), pp. S-42–S-45

Phosphate binders on iron basis: A new perspective?

OLAF HERGESELL and EBERHARD RITZ

Department of Internal Medicine, Ruperto Carola University, Heidelberg, Germany (FRG)

Phosphate binders on iron basis: A new perspective? Uremic weekly dialysis approximately 100 mmol per week arepatients on maintenance hemodialysis are in positive phosphate eliminated. This amount is considerably less than thebalance. This is mainly the result of the complex elimination cumulative net weekly adsorption of P in the anurickinetics of phosphate during dialysis. Removal of phosphate

patient [5], so there is a clear need to prevent intestinalis less than net dietary intake. Classical phosphate binders suchabsorption of P so as to achieve neutral P balance.as calcium carbonate, calcium acetate, and aluminum-based

compounds are limited by side effects (hypercalcemia) andoutright toxicity (aluminium). There have been numerous re-cent attempts to develop alternative phosphate binders, e.g., ELIMINATION KINETICS OF PHOSPHORUSpolyallylamine-hydrochloride (Renagel), lanthanum carbon- DURING DIALYSISate, and trivalent iron-containing compounds. The latter is

Although as a first approximation the elimination ofbased on old observations that iron salts may cause hyperphos-urea follows first-order reaction consistent with uniformphatemia and rickets in experimental animals and in patients.

This idea has recently been taken up again, and effective inhibi- distribution in a single compartment, elimination kineticstion of net intestinal phosphate uptake in non-uremic and ure- of P are more complex. Hou et al. [4] as well as previousmic rats has been shown using simple iron salts (citrate, chlo-

investigators [6, 7] found a rapid initial decrease of serumride, ammonium citrate) and complex compounds (cross-linkedP concentration during a dialysis session; as soon as thedextran and stabilized polynuclear iron hydroxide). In uremic

rats, the latter compound reduces urinary phosphate excretion normal range of serum P concentration was reached,as an indicator of reduced intestinal phosphate uptake and has however, the serum P concentration tended to remainalso been shown to be effective in subjects with preterminal constant independent of blood flow [7]. The major pro-renal failure. So far, no side effects or short-term toxicity has

portion of P was eliminated during the first hour. Con-been observed. The compound appears promising and deservesversely, upon termination of hemodialysis, a rapid re-further evaluation.bound occurred, and within 12 hr approximately 80%of predialysis serum P values were reached [6, 7]. Theexplanation is that P is sequestered in deep compart-Hyperphosphatemia continues to be a challenge forments; the transfer rate constants from the intracellularthe clinical nephrologist. Control of serum phosphate con-

centration is a central strategy in the treatment of second- into the extracellular space are small, so that the rate ofary hyperparathyroidism, but it is also important in view P exit from intracellular space can no longer keep paceof recently recognized cardiovascular hazards of hyper- with the rate of removal through dialysis and becomesphosphatemia [1]. rate-limiting. That redistribution of P is indeed the bot-

The mass balance of phosphorus (P) is precarious in tleneck of P elimination has been nicely illustrated bythe uremic patient on dialysis treatment. A normal West- recent observations [8] showing that more frequent longern diet contains approximately 800–1500 mg of P per dialysis sessions, i.e., 6 instead of 3 times per week, per-day [2], of which 50–70% are absorbed in the intestine. mit elimination of substantial amounts of P, so that,It is still uncertain whether the net P intestinal absorption paradoxically, even addition of P to the dialysate is neces-is less in the uremic patient compared to controls [3], sary in some patients to prevent P depletion [9].but if some intestinal malabsorption of P exists, it does

Classical phosphate bindersnot have a major impact on prevention. According toHou et al. [4], during a 4-hour hemodialysis session using Very early on, this dilemma forced clinical nephrolo-1.0 m2 cellulose acetate hollow fiber dialyzers, approxi- gists to use phosphate binders in order to sequester P inmately 33 mmol P are removed, so that with thrice the intestinal tract and to reduce entry of P into the

extracellular fluid space. Historically, calcium carbonateand subsequently calcium acetate had been used mostKey words: hyperphosphatemia, renal failure, phosphate binders, iron,

hemodialysis. frequently. They are undoubtedly effective, but morerecently, with more successful management of hyper- 1999 by the International Society of Nephrology

S-42

Hergesell and Ritz: Phosphate binders and iron S-43

parathyroidism and a tendency for PTH (and in parallel linked iron dextran, i.e., iron (III)oxide-hydroxide-modi-fied dextran [26], as an oral phosphate binder in the rat.bone turnover) to be lower, the risk of hypercalcemia on

oral calcium-containing phosphate binders has increased This was based on phosphate-binding studies in aqueoussolutions and whole blood [27], which had shown re-[10]. Although reducing the dialysate calcium concentra-

tions diminishes the risk of hypercalcemia, Argiles et markable P binding efficacy. Binding was pH-dependentand maximal at pH 2–3, but when passage of the phos-al. [11] showed that, in the long run, dialysate calcium

concentrations below 6 mg/dl (1.5 mm) are not safe: they phate binders through the intestine was simulated invitro by varying the pH over 6 hours, net binding capacitymay cause stimulation of the parathyroid gland and loss

of bone mineral, particularly if patients are noncompliant was still substantial and reduced only modestly com-pared to pH 2. Iron (III)oxide-hydroxide dextran beadswith the intake of calcium-containing phosphate binders,

so that the calcium balance becomes negative. were exposed to chick duodenum to assess whether Fewas mobilized from the compound. Enzymatic iron liber-Aluminum-containing phosphate binders expose the

uremic patient to the hazard of aluminum intoxication ation was very limited at best. In vivo studies with ratsshowed that, while plasma phosphate concentrations re-[12]. In an ideal world, aluminum-containing phosphate

binders should be avoided completely, but, unfortunately, mained virtually unchanged when rats were fed withphosphate binder-containing food, the amount of P ex-one is often forced to use some when the use of calcium-

containing P binders leads to hypercalcemia [10]. creted in the urine was lowered by a factor of 15. It wasconcluded that 1 g of this P binder will bind approxi-

Novel phosphate binders mately 0.65 mm of P. If one extrapolates to a 70 kg humanbeing, approximately 15 g/day of the binder would beThe dilemma of side effects (hypercalcemia with cal-

cium-containing P binders) or toxicity (aluminum-con- necessary. No animal died in the study by Spengler etal. [25], animals grew and toxicity studies, including nec-taining P binders) led to efforts to develop alternative P

binders. One compound, polyallylamine-hydrochloride ropsy after 8 weeks of exposure, failed to reveal anyevidence of subacute toxicity. Using an organic ferric(Renagelt) has been intensively investigated in experi-

mental and clinical [13–15] studies. This class of com- compound, Hsu et al. [28], possibly unaware of this pre-ceding study, fed ferric ammonium citrate and ferricpounds was originally developed to lower plasma lipids,

and this effect is still seen with Renagel as well. Polyal- chloride to normal and azotemic rats. They were on a1.02% P diet to which the ferric salts had been added.lylamine-hydrochloride was shown to lower predialytic

plasma P concentrations from 2.93 mm to 2.13 mm in a Urinary P excretion was consistently lowered. Since in-take was unchanged, obviously net intestinal absorptionshort-term study, and similar effectiveness was also noted

in long-term studies at a maximum daily dose of 5.5 g. had been lowered. The authors calculated that the bind-ing capacity was 88–181 mg of P per g of elemental iron.It is apparently somewhat difficult to lower P levels, as

desired, in the mid-normal range, at least when this P Two recent observations point to the potential use-fulness of these compounds in humans. We [29] exam-binder is used as monotherapy.

Another approach is the development of lanthanum- ined 13 patients with stable preterminal renal failure(median serum creatinine 5.4 mg/dl). The patients werebased compounds. Lanthanum-chloride-hydrate was stud-

ied in animal experiments [16], and little tissue uptake hyperphosphatemic (median fasting plasma P 2.2 mm)on stable dietary P intake. After two weeks on no oraland toxicity was noted. A clinical study using lanthanum

carbonate is currently underway, but the results are not P binders, 3 3 2.5 g stabilized polynuclear iron hydroxydewas given with meals for 4 weeks—a deliberately subop-available.

A third line of products concerns trivalent iron salts. timal dose. This dose caused a median decrease of fastingplasma P of 20%. Urinary P excretion was lowered evenThe available information on this interesting group of

compounds will be discussed below. more, i.e., by 37%. Apart from a certain laxative actionand black discoloration of feces, no side effects were

Fe-III based phosphate binders noted in this short-term study. Figure 1 shows the individ-ual changes in plasma phosphate and urinary P excretionSeveral anecdotal clinical observations [17, 18] docu-

ment that ferric compounds bind dietary phosphate and rate. Of note, no significant changes in serum iron andserum ferritin were noted. In a preliminary study, Changlower serum phosphate when given orally and even intra-

venously [19]. Iron salts have been shown to reduce bone et al. [30] investigated 32 dialyzed uremic patients whowere treated either with conventional phosphate bindersash in Guinea pigs [20], to retard growth and induce phos-

phate depletion in rats [21, 22], and to cause phosphate- (calcium carbonate and/or aluminum-containing com-pounds) or a non-ionic ferric polymaltose complex (Fer-deficiency rickets in chicken [23] as well as in swine [24].

This line of observation lay dormant for a long time, rumt chewable tablet, Hausmann Laboratories, Switzer-land) containing 1.78 mmol elemental iron per tablet.but recently has been independently taken up by several

groups of investigators. Spengler et al. [25] studied cross- After wash-out of the P binders, plasma P was 2.64 and

Hergesell and Ritz: Phosphate binders and ironS-44

Fig. 1. Percent decrease of plasma phosphate(A) and urinary phosphate excretion (B) in 13patients with stable preterminal failure at days14 and 28, respectively.

2.53 mm in the two groups, which subsequently received the substance [29]. Further details are currently beingassessed in a randomized prospective controlled studyconventional P binders again or Ferrum tablets, respec-in dialyzed patients. A potential advantage of these com-tively. Ferrum was less effective than conventional Ppounds is the favorable relation between cost and effec-binders. A mean decrease of 0.38 mm P was seen aftertiveness.8 weeks compared with 0.85 mm in controls, but at least

Whether the combination of P binders and inhibitorsproof of principle is provided by this preliminary obser-of the intestinal Na,P-transporter [31] is a useful strategyvation. There was no significant change in ionized cal-will require further studies. Available substances, e.g.cium, ferritin, serum iron, or iron saturation.Niceritorol [31], are not free of side effects, but it isnot excluded that in the future, non-absorbable specific

SUMMARY inhibitors with no systemic side effects can be developed.Various Fe(III) compounds are effective and presum-

Reprint requests to Professor Eberhard Ritz, Department of Internalably safe P binders. Further information is required with Medicine, University of Heidelberg, Bergheimer Straße 56a, D-69115

Heidelberg, Germany.respect to absorption of iron (which may be a benefitin disguise in the age of erythropoietin treatment) and

REFERENCESpotential long-term side effects, e.g., binding of drugs,micronutrient binding. Nevertheless toxicity studies in 1. Amann K, Gross ML, London GM, Ritz E: Hyperphosphate-

mia—a silent killer of patients with renal failure? Nephrol Dialanimals are very reassuring. The palatability of these com-Transplant 14:2085–2087, 1999pounds is remarkable (as we noted in self experiments) 2. Schaefer K: Alternative phosphate binders: an update. NephrolDial Transplant 1:35–39, 1993and there were no complaints when patients received

Hergesell and Ritz: Phosphate binders and iron S-45

3. Wiegmann TB, Kaye M: Malabsorption of calcium and phosphate binder? A comparative study on the dietary phosphorus absorp-tion. Res Commun Mol Pathol Pharmacol 89:373–378, 1995in chronic renal failure: 32P and 45Ca studies in dialysis patients.

17. Liu SH, Chu HI, Hsu HC, Chao HC, Cheu SH: Calcium andClin Nephrol 34:35–41, 1990phosphorus metabolism in osteomalacia. J Clin Invest 20:255–271,4. Hou SH, Zhao J, Ellmann CF, Hu J, Griffin Z, Spiegel DM,1941Bourdeau JE: Calcium and phosphorus fluxes during hemodialysis

18. Liu SH, Chu HI: Studies of calcium and phosphorus metabolismwith low calcium dialysate. Am J Kidney Dis 18:217–224, 1991with special reference to pathogenesis and effects of dihydrotachy-5. Ritz E: Phosphate removal during dialysis—does the membranesterol (A.T.10) and iron. Medicine (Baltimore) 22:103–161, 1943matter? Clin Nephrol 42 (Suppl.):S57–S60, 1994

19. Okada M, Imamura K, IidAM, Fuchigami T, Omae T: Hypophos-6. Zucchelli P, Santoro A: Inorganic phosphate removal duringphatemia induced by intravenous administration of saccharateddifferent dialysis procedures. Int J Artif Organs 10:173–178, 1987iron oxide. Klin Wochenschr 61:99–102, 19837. Pogglitsch H, Petek W, Ziak E, Sterz F, Holzer H: Phosphorus

20. Cox G, Dodds M, Wigman H, Murphy F: The effects of highkinetics during hemodialysis and hemofiltration. Proc EDTA-ERA doses of aluminium and iron on phosphorus metabolism. J Biol21:461–468, 1984 Chem 92:xi–xii, 1931

8. Mucsi I, Hercz G, Uldall R, Ouwendyk M, Francoeur R, Pier- 21. Brock J, Diamond L: Rickets in rats by iron feeding. J Pediatratos A: Control of serum phosphate without any phosphate bind- 4:442–453, 1934ers in patients treated with nocturnal hemodialysis. Kidney Int 22. Rehm P, Winters J: The effect of ferric chloride utilization of53:1399–1404, 1998 calcium and phosphorus in the animal body. J Nutr 19:213–222, 1940

9. Mucsi I, Hercz G: Control of serum phosphate in patients with 23. Deobald H, Elvehjem C: The effect of feeding high amounts ofrenal failure—new approaches. Nephrol Dial Transplant 13:2457– soluble iron and aluminium salts. Am J Physiol 111:118–123, 19352460, 1998 24. Furugouri K: Effect of elevated dietary levels of iron stores in

10. Ritz E, Passlick-Deetjen J, Lippert J: What is the appropriate liver, some blood constituents and phosphorus deficiency in youngdialysate calcium concentration for the dialysis patient? Nephrol swine. J Animal Sci 34:573–577, 1972

25. Spengler K, Follmann H, Boos KS, Seidel D, von der Haar F,Dial Transplant 11 (Suppl. 3):91–95, 1996Elsner R, Maywald F: cross-linked iron dextran is an efficient11. Argiles A, Kerr PG, Canaud B, Flavier JL, Mion C: Calciumoral phosphate binder in the rat. Nephrol Dial Transplant 11:808–kinetics and the long-term effects of lowering dialyste calcium812, 1996concentration. Kidney Int 43:630–640, 1993

26. Boos KS, Seidel D, Rauh A, Spengler K, Henke G: Verfahren12. Wills MR, Savory J: Aluminium poisoning: dialysis encephalopa-zur selektiven Elimination von anorganischem Phosphat aus Flus-thy, osteomalacia and anemia. Lancet 2:29–34, 1983sigkeiten mittels polynuclearen Metalloxidhydroxid-modifizierten13. Chertow GM, Burke SK, Lazarus JM, Stenzel KH, WomboltAdsorptionsmaterialien. German Pat Appl P41 39 442.2, 1992D, Goldberg D, Bonventre JV, Slatopolsky E: Poly[allylamine

27. Spengler K, Follmann H, Boos KS, Seidel D, Maywald F: Char-Hydrochloride] (RenaGel): a noncalcemic phosphate binder foracterisation and extracorporeal application of a new phosphate-the treatment of hyperphosphatemia in chronic renal failure. Ambinding agent. Eur J Clin Chem Clin Biochem 32:733–739, 1994J Kidney Dis 29:66–71, 1997

28. Hsu CS, Patel SR, Young EW: New phosphate binding agents:14. Slatopolsky EA, Burke SK, Dillon MA and the RenaGel ferric compounds. J Am Soc Nephrol 10:1274–1280, 1999Study Group: RenaGel, a nonabsorbed calcium and aluminium- 29. Hergesell O, Ritz E: Stabilized polynuclear iron hydroxide is anfree phosphate binder, lowers serum phophorus and parathyroid efficient oral phosphate binder in uraemic patients. Nephrol Dialhormone. Kidney Int 55:299–307, 1999 Transplant 14:863–867, 1999

15. Bleyer AJ, Burke SK, Dillon M, Garrett B, Kant KS, Lynch 30. Chang JM, Hwang SJ, Tsai JC, Chen HC, Guh JY, Lai YH:D, Rahman SN, Schoenfeld P, Teitelbaum I, Zeig S, Slatopolsky Effect of ferric polymaltose complex as a phosphate binder inE: A comparison of the calcium-free phosphate binder sevelamer hemodialysis patients. Nephrol Dial Transplant 14:1045–1047, 1999hydrochloride with calcium acetate in the treatment of hyperphos- 31. Shimada T, Okuda S, Motomura K, Onoyama K, Fujishima M:phatemia in hemodialysis patients. Am J Kidney Dis 33:694–701, Niceritorol decreases serum phosphate levels in hemodialysis pa-1999 tients—A randomized, double-blind, crossover study. J Am Soc

Nephrol 7:52, 199616. Graff L, Burnel D: a possible non-luminium oral phosphate