Aluminium–related osteomalacia: response to reverse osmosis water treatment

Aluminium–related osteomalacia: response to reverse osmosis water treatment

Kidney International, Vol. 32 (1987), pp. 96—101 Aluminium—related osteomalacia: Response to reverse osmosis water treatment GEORGE D. SMITH, ROBIN J...

545KB Sizes 1 Downloads 61 Views

Kidney International, Vol. 32 (1987), pp. 96—101

Aluminium—related osteomalacia: Response to reverse osmosis water treatment GEORGE D. SMITH, ROBIN J. WINNEY, ALEXANDER MCLEAN, and JAMES S. ROBSON Department of Pathology, University Medical School, and Medical Renal Unit, Royal Infirmary, Edinburgh, Scotland, United Kingdom

In this paper we described seven chronic-hemodialysis patients with aluminium related osteomalacia who were subsethereby inhibiting mineralization of osteoid. Because this form of quently provided with reverse osmosis facilities, but had no osteomalacia is vitamin D resistant, the condition has often been other alteration in management. After three years, four patients

Aluminium—related osteomalacia: response to reverse osmosis water treatment. It is generally accepted that aluminium induces osteomalacia in chronic hemodialysis patients by binding to the calcification front,

showed a marked improvement in bone mineralization, whereas

assumed to be irreversible, although promising results have been

in the other three patients the skeletal pathology was un-

achieved recently by using a chelating agent for removal of aluminium from the skeleton. In this paper we present four chronic hemodialysis patients with aluminium toxicity and histologic osteomalacia in whom

changed. These observations are discussed with reference to the clinical, biochemical and bone histomorphometric findings.

the mineralization defect greatly regressed after the use of reverse osmosis treated—water for dialysis, but without further treatment. In three other patients, also with aluminium toxicity and histologic osteomalacia, similarly treated, the histological severity of the osteomalacia remained static. Those patients in whom bone mineralization status improved developed hyperparathyroidism after reverse osmosis water— treatment, whereas the static patients remained euparathyroid. The results suggest that resolution of alumium related osteomalacia may occur with reduction in dialysis fluid aluminium, and that parathyroid hormone plays a role in the healing of aluminium related osteomalacia. The therapeutic implications are twofold: (1) attempts to remove all traces of hyperparathyroidism may be detrimental to the bone mineralization status; and (2) stimulation of the parathyroid glands by means of a mild reduction in dialysis fluid calcium may be of value in the management of those cases with persistent osteomalacia and low bone turnover.


Patients Seven patients (4 males, 3 females, aged 29 to 64 years) with advanced chronic renal failure of various etiologies were studied. These were all of the patients on dialysis in our unit who

had clinical or histological evidence of aluminium toxicity before the introduction of reverse osmosis (RO) water—treat-

ment, and who had continued on dialysis for a sufficient duration to enable assessment of the histological response to

the introduction of RO water treatment. The patients had received regular hemodialysis at home or in the hospital for between 11 and 78 months before acquisition of a reverse

Aluminium is now considered to be the prime factor in the pathogenesis of dialysis osteomalacia affecting patients with chronic renal failure [1—6]. The major soure of the aluminium has been dialysis fluid [7—10], prepared directly from the domestic water supply to which aluminium sulphate has been added as a flocculating agent to remove brown particulate matter. The aluminium binds to the calcification front [11—13]

osmosis (RO) unit, during which time all patients in group 1 and patient 1 in group 2 used softened water, and patients 2 and 3 in group 2 had no water treatment. The water supply to all patients was treated with aluminium sulphate. Aluminium—containing

phosphate binders were prescribed if required to maintain

where it appears to inhibit mineralization of osteoid, and because skeletal uptake of calcium is blocked, there is a

serum inorganic phosphate below 2 mmol/liter throughout the period of observation and were continued during the period of reverse osmosis water—treatment. All patients had histologic evidence of osteomalacia associated with heavy deposition of aluminium at calcified bone-osteoid interfaces. As a result of the histologic findings in iliac crest bone biopsies obtained after the use of reverse osmosis water—treatment for three years, the patients were divided into group 1, comprising patients now without histologic osteomalacia and one patient with much less

tendency to hypercalcemia [5, 14] and relative hypoparathyroid-

ism. Because of low bone turnover and morbidity due to aluminium related anemia [15] and neurotoxicity [16], it has

been assumed that the prognosis is poor, although recently improvement in bone mineralization status has been reported severe osteomalacia, and group 2 which consisted of three after removal of aluminium from the dialysis water by reverse patients with unchanged skeletal pathology. The groups were compared using the Mann—Whitney U-test to determine any osmosis [14, 17, 18].

differences in mean plasma—aluminium levels, serum biochemistry and bone histomorphometric indices either before or after Received for publication April 1, 1985 and in revised form May 8 and November 21, 1986

the use of reverse osmosis. This non-parametric statistical model was employed because of the small sample size and

© 1987 by the International Society of Nephrology

unknown distribution types [19].



Aluminium—related osteomalacia

Table 1. Clinical and biochemical details of patients before RO water treatment Serum/plasma biochem istry


Calcium mmol/liter

Phosphate mmol/liter


Group 1 1 (JD) 2 (EW) 3 (iF) 4 (DM) Group 2

24 42 47 36

2.70 2.80 3.00 2.65

1.90 1.50 1.52 1.90

I (AD) 2 (MH)

59 64

1.77 1.50

3(RS) Normal range


2.99 2.60 2.55 2.16—2.62




PC < 0.03




Dialysis fluid aluminium ,amol/liter

PTH pg/liter


80 78 96 95

0.86 NAb 2.40

12.8 13.4 5.1



2.64 0.52




28 47 27

99 60 75


8.1 12.4 18.4



0.57 0.80








— NDd

Months on dialysis

15 11


— <0.03

a Alkaline phosphatase b Not available C


Statistical comparisons are for group 1 vs. group 2, by the U-test of Mann—Whitney Not significant

C These patients were commenced on reverse osmosis water treatment prior to measurement of dialysis fluid aluminium but mean water aluminium for both patients was 7.1 mol/liter

linked to a microcomputer, a method broadly similar to that Plasma and dialysis fluid aluminium and serum levels of employed by Malluche et al [24], the following histomorphometric variables were measured for each biopsy: relative calcium, inorganic phosphate and alkaline phosphatase (AP) osteoid volume (OV) [25], total osteoid surface (OS) [25], the had been measured monthly in each patient prior to the use of reverse osmosis and during the period of reverse osmosis proportion of OS bearing a calcification front (CF) [22], active water—treatment and were expressed as mean pre-RO values. osteoblastic surface (AOS) [251 and active resorption surface Aluminium concentrations were estimated by means of electro- (ARS) [26]. The width of the thickest osteoid seam was exthermal atomic—absorption spectrophotometry as described pressed as the maximum number of bright osteoid lamellae visualized under polarised light. previously [20]. Serum calcium, inorganic phosphate and alka- (MNOL) The values obtained for OV, OS, CF and MNOL were line phosphatase were assayed by standard laboratory techniques. Serum PTH was measured on a few occasions only, employed as indicators of the presence and severity of around the times of the bone biopsies, by a double antibody osteomalacia. A diagnosis of osteomalacia was made in those radioimmunoassay using rabbit anti-PTH serum directed pre- biopsy specimens with MNOL >4 and CF <60% [22]. The values obtained for AOS and ARS were regarded as indices of dominantly against the N-terminal sequence of PTH. the degree of secondary hyperparathyroidism. Biochemistry

Bone biopsies


Iliac crest bone biopsy—specimens were obtained from each patient shortly before acquisition of a reverse osmosis unit and after three years of water treatment. Pre-RO bone biopsies were performed using a trephine bone—marrow needle providing a

Before the introduction of reverse osmosis

phosphate—buffered formaldehyde and embedded in methacryl-

phosphate and alkaline phosphatase. Serum PTH tended to be

The patients of group 1 were younger and had been on core of bone 15 x 3 mm; post-RO transiliac bone biopsy dialysis for longer than those of group 2 (Table 1). There was no specimens (8 mm diameter) were obtained by the Meunier difference between the groups with regard to plasma and needle. All biopsy specimens were fixed in 4% neutral dialysis fluid aluminium or serum levels of calcium, inorganic ate; from each biopsy specimen thin (5 m) undecalcified higher in group 1. Dialysis fluid aluminium was high in four sections were cut on a Jung K sledge microtome. Three patients, and likely to have been high in a further two patients representative sections from each biopsy specimen were on the basis of very high water—aluminium. In one patient stained with Goidner's technique [21] for quantitative histology, and from each of the post-RO biopsy specimens, the immedi-

(patient 3 of group 1) dialysis fluid aluminium was not high, but

he had been on dialysis for four years and dialysis fluid

ately adjacent sections were stained with toluidine blue for aluminium measurements were available for only a few months estimation of the extent of surface osteoid bearing a calcifica- prior to the use of reverse osmosis. Fracturing osteodystrophy tion front [22]. The "Aluminon" method [23] was utilized for was present in two patients in group 1, (fractured scapula and detection of aluminium. ribs in patients 1 and 2) and one patient in group 2 (fractured By means of a semiautomatic method using the Graphic neck of femur and pubic ramus in patient 1). Patient 1 in group Digitising System 1 equipment (Graphic Information Services 2 also had dialysis encephalopathy and all patients except Ltd) which comprised a digitizing board with electronic cursor patient 4 in group 1 had a microcytic anemia.

Smith et a!

98 Relative osteoid volume %


pre-R.0. post-R.0.

pre-R.0. post-R.0.


Total osteoid surface %



front % post-R.0.

pre-R.0. post-R.0.





60 9

• o




• 10

• •










Fig. 1. Bone histomorphometry indicating the severity of osteomalacia and its response to RO water treatment. Symbols are: (•) Group 1 patients, (0) group 2 patients. P statistical comparisons are from group 1 vs. group 2 by the Mann—Whitney U-test. Bars indicate the upper limit of the normal range for relative osteoid volume, maximum number of osteoid lamellae (MNOL) and total osteoid surface. The lower limit of the normal range for the extent of calcification fronts along total osteoid surfaces is indicated by the bar at 60%.




• •



20 0

0 o


0 NS

0 o







Active resorption

osteoblastic surface %

surface %


























. 5

. .
















Histologically, all seven patients had moderately severe osteomalacia characterized by an increase in MNOL and OV, there being no differences in these parameters between the two groups (Fig. 1). A calcification front could not be detected by toluidine blue in any of the biopsies. Whereas the mineralization deficit was extensive in group 1 (OS > 48%), the patients in group 2 had a more focal form of osteomalacia (OS < 32%). In all cases, strongly positive Aluminon staining was demonstrated along most of the length of the calcified bone—osteoid interface. As well as osteomalacia, all seven patients had mild


Fig. 2. Bone histomorphometry indicating the degree of hyperparathyroidism before and after RD water treatment. Symbols are: (•) Group 1 patients, (0) group 2 patients. P statistical comparisons are from group 1 vs. group 2 by the Mann—Whitney U-test. Bars indicate the upper limit of the normal ranges.

osteitis fibrosa (hyperparathyroid bone disease), expressed as a mild increase in active resorption surface with deep Howship's

lacunae and mild focal fibrosis not exceeding grade 2 [27]; active osteoblastic surfaces were increased in only two patients. The two groups did not differ in the degree of hyperparathyroidism (Fig. 2).

After the introduction of reverse osmosis With reverse osmosis water—treatment there was a significant

reduction in both plasma and dialysis fluid aluminium in both


Aluminium—related osteomalacia

Table 2. Clinical a nd biochemical details of patients after RO water treatment Serum/plasma biochemistry Patients Group 1 I (JD) 2 (EW) 3 (JF) 4 (DM) Group 2 I (AD) 2 (MH) 3 (RS) Normal range

Calcium mmol/liter

Phosphate mmol/liter

27 45 50 39

2.33 2.83

1.62 1.80 1.64



62 67 66

2.67 2.30 2.38









1.60 1.23


AP units/liter 125 138

239 180

PTH pg/liter

Aluminium pinol/liter

Dialysis fluid aluminium pinol/liter


6.3 3.4 5.0 8.3

0.33 0.6 0.54

2.30 2.70 2.60


0.34 0.58 0.47

<100 <0.03

<0.60 <0.03

60 64

6.3 7.3

5.5 —


0.6 0.4 0.4 —


O Alkaline phosphatase b Statistical comparisons are for group 1 vs. group 2, by the U-test of Mann-Whitney C

Not significant

groups, and there was again no difference between the groups. However, the plasma aluminium remained significantly elevated in most patients. The serum calcium was similar in both groups, but the serum phosphate in the patients of group 1 was significantly greater than in those of group 2. The patients of group 1 exhibited biochemical evidence of hyperparathyroidism with elevated serum levels of alkaline phosphatase and PTH whereas those in group 2 did not; the differences were significant (Table 2). Healing of fractures occurred in all patients with fracturing bone disease before the introduction of reverse osmosis and no

study. While the mineralization defect eventually resolved in one of the patients of Boyce et a! [141 with aluminium related osteomalacia after removal from exposure to aluminium, one of

patient developed new fractures. This was accompanied by resolution of microcytosis and improvement in anemia, but patient 1 in group 2 subsequently died of dialysis encephalopathy.

repsonse could not be distinguished on the basis of bone

Histologically, three of the patients in group 1 no longer had

the two patients with fracturing osteomalacia reported by Brown et al [181 failed to respond to RO water purification alone, although was later successfully treated by desferrioxamine. Our results indicate that, while a good clinical response with healing of fractures may occur following the introduction of RO water treatment, this is not always accompanied by histological

resolution of osteomalacia. Patients with a good histological histology or routine serum biochemistry before the use of RO water treatment. Both groups had been exposed to high levels of aluminium in the dialysis fluid with very high plasma— aluminium levels which were similar in both groups. Following

osteomalacia (MNOL <5, CF >60%) and the fourth patient the introduction of RO water treatment the dialysis fluid showed substantial improvement in bone mineralization, with aluminium was maintained at a low level in both groups and reduction in OV from 21.2% to 11.4% and reduction of MNOL there was a significant fall in plasma aluminium. However, the from 9 to 6 (Fig. 1). By contrast, the degree of osteomalacia plasma aluminium remained significantly elevated in most paaffecting those patients of group 2 was unchanged with respect tients and was higher than we would expect to result from the to the previous biopsies. In six patients, weakly—positive continuation of small doses of aluminium—containing phosphate Aluminon staining was distributed focally along the length of binding agents. This response confirms our previous observathe calcified bone—osteoid interface and occasionally at cement tion that the plasma aluminium falls only slowly following lines within calcified bone; in the seventh patient (from group 1) reduction in exposure to aluminium, presumably as a result of aluminium could not be shown histochemically. After reverse poor clearance during hemodialysis and relatively slow uptake osmosis water—treatment the patients in group 1 had more by tissues [291. We did not assess the bone apposition rate in our patients advanced osteitis fibrosa reflected histomorphometrically by significantly—increased active resorption and active osteoblastic using double tetracycline labelling, but there was a marked rise surfaces, whereas in group 2 the degree of osteitis fibrosa in the percentage of osteoid seams bearing a calcification front in all patients with RO water treatment, the rise being signifiremained mild (Fig. 2).

cantly higher in patients in group 1. This, together with the striking reduction in indices of osteoid bone volume in the Although the studies of McClure et al [26, 281 indicate that patients in group 1, despite increased bone turnover, is strong introduction of RO water treatment was associated with an indirect evidence that there was an improvement in bone Discussion

improvement in bone mineralization status, they do not specifically refer to the progress of individual patients. Leather et al [17] reported that fractures in their renal osteodystrophy cases usually healed following the use of a deionizer to prepare the dialysis fluid, but they did not include bone histology in their

mineralization. The main distinguishing feature of the patients with a good

histological response to RO water treatment was that they tended to have higher serum PTH levels. It is interesting to note that they were younger and had been on hemodialysis for longer


Smith et a!

inorganic phosphate were not involved in the genesis of postRO hyperparathyroidism, the presence of raised serum, and therefore extracellular fluid levels would augment the calcium— hyperplasia, and the parathyroid glands of the responders, phosphate ion product in osteoid and counteract any possible exposed to abnormal mineral biochemistry for longer periods of localized phosphate depleting effect of aluminium [38, 39]. Two main points emerge from this study. First, aluminium time than the nonresponders, have had more opportunity to develop hyperplasia. After the introduction of RO units, those related osteomalacia may persist in chronic hemodialysis papatients with persistent osteomalacia remained biochemically tients with low bone turnover despite a large reduction in euparathyroid and only had mild osteitis fibrosa, whereas the plasma aluminium. Although it is good policy to attempt to keep responders developed biochemical and histologic evidence of serum inorganic phosphate within reasonable limits to prevent hyperparathyroidism consistent with removal of aluminium— metastatic calcification in soft tissue [40], we consider it possimediated suppression of PTH release from already hyperplastic bly detrimental to the bone mineralization status to manipulate glands. Furthermore, the results suggest that PTH could be serum levels of calcium and inorganic phosphate or to perform involved in the restoration of mineralization in "paralyzed" parathyroidectomy in order to remove all traces of hyperparaosteoid seams, possible through the genesis of matrix vesicles thyroidism. Although it is generally accepted that osteomalacia by PTH-activated osteoblasts. Although it is likely that matrix is the crippling form of renal osteodystrophy, in our experience vesicles are not involved in mineralization of lamellar bone mild to moderate osteitis fibrosa is rarely symptomatic and under normal circumstances [31], Anderson et al [321 identified therefore is clinically acceptable. Secondly, aluminium related osteomalacia will respond to matrix vesicles and associated tiny clusters of apatite—like material in osteoid seams subjacent to active osteoblasts in six RO water treatment alone, but a good histological response out of seven patients with renal osteodystrophy. While calcium occurs only in the presence of hyperparathyroidism. In those hydroxyapatite crystal formation at the calcification front ap- cases with persistent osteomalacia and low bone turnover, a pears to be blocked by aluminium [11—13], possibly acting as a mild reduction in dialysis fluid calcium with the aim of stimucrystal poison [33], matrix vesicles derived from osteoblasts lating the parathyroid glands may offer a useful adjunct or reactivated by high PTH levels could serve to nucleate a new alternative to desferrioxamine in the management of this discalcification front in the osteoid seam superficial to the tressing disease.

than the nonresponders, While aluminium appears to suppress parathyroid hormone release [301 there is currently no evidence that the element can inhibit the development of parathyroid

aluminium—saturated calcified bone—osteoid interface, with sub-

sequent mineralization of the seam. This mechanism of bypass-

Reprint requests to Dr. G.D. Smith, Department of Pathology,

ing the aluminium barrier is not available to euparathyroid Stobhill General Hospital, Glasgow G21 3UW, United Kingdom. patients whose inactive osteoblasts were unable to provide References matrix vesicles. While the similar aluminium staining on our two groups after RO water treatment may simply reflect quantitative limitations of the staining method, this might also be explained by PTH-induced mineralization in a new calcification

front. This hypothesis is consistent with the observation by Cournot—Witmer et al [12] that moderate to severe hyperparathyroidism appeared to protect against osteomalacia despite the

HA, MCCARTHY JH, HERRINGTON J: Bone aluminium in haemodialysis patients and in rats injected with aluminium chloride: Relationship to impaired bone mineralisation. J Clin Pathol


32:832—844, 1979 2. KANIS JA: Osteomalacia and chronic renal failure. J Clin Pathol 34:1295—1307, 1981 3. DRUEKE T: Dialysis

fact that the bone aluminium levels in hemodialysis patients with osteitis fibrosa did not differ significantly from those with osteomalacia. Unfortunately we did not have bone aluminium measurements in our patients and so were unable to substantiate this observation. There is strong circumstantial evidence for

a role of relative hypoparathyroidism in the pathogenesis of aluminium related osteomalacia [6, 12, 34, 351, and the occurrence of osteomalacia after parathyroidectomy further supports the role of PTH in maintaining bone mineralization in uremia [36]. Although it has been implied that the protective effect of high serum PTH levels is mediated via high bone turnover with dilution of the bone surface aluminium—concentration [3], we

consider this hypothesis unlikely because the two groups of patients in our study did not differ in plasma aluminium levels or, by assumption, in aluminium levels at the bone surface. In addition, bone aluminium staining after RO water treatment was similar in both groups.

After RO water purification, the group in which the bone mineralization status improved had higher serum levels of


osteomalacia and aluminium intoxication.

Nephron 26:207—210, 1980 HODSMAN AB, SHERRARD DH, ALFREY AC, OTT S, BRICKMAN AS, MILLER NL, COBURN JW: Bone aluminium and histomorpho-

metric features of renal osteodystrophy. J Clin

Endocrinol Metab

54:539—546, 1982 5.



AM: Dialysate aluminium concentration and renal bone disease. Kidney mt 21:411—415, 1982 ROBERTSON JA, FELSENFELD AJ, HAYGOOD CC, WILSON P,

CLARKE C, LLACH F: Animal model of aluminium—induced osteomalacia: Role of chronic renal failure. Kidney mt 23:327—335, 1983

KAEHNY WD, ALFRED AC, HOLMAN RE: Aluminium transfer during hemodialysis. Kidney mt 12:361—365, 1977 8. PARKINSON IS, WARD MK, FEaST TD, FAWCETT RWP, KERR DNS: Fracturing dialysis osteodystrophy and dialysis encephalopathy. An epidemiological survey. Lancet i:406—409, 1979 7.

9. ELLIOTT HL, DRYBURGH F, FELL GS, SABET S, MACDOUGALL Al: Aluminium toxicity during regular haemodialysis. Br Med

J i: 1101—

1103, 1978 10.

PLATTS MM, GOODE GC, HIsLOP JS: Comparison of the domestic water supply


and the incidence of fractures and encephalopathy in

on home dialysis. Br Med J

ii:657—660, 1977


inorganic phosphate than the group with static bone pathology. Despite there being no difference in serum calcium between the


groups it is possible that the hyperphosphatemia may have contributed to parathyroid overactivity [37]. However, even if

Scan Electron Microsc 3:329—337, 1981 12. COURNOT—WITMER G, ZINORAFF J, PLACHOT JJ, ESCAIG

of aluminium in human cancellous bone in renal osteodystrophy.



Aluminium—related osteomalacia

P, BOURDON R, DRUEKE T, BALSAN S: Aluminium localization in

27. ELLIS HA, PEART KM: Azotaemic renal osteodystrophy: A qualitative study on iliac bone. J Clin Pathol 26:83—101, 1973

bone from hemodialyzed patients: Relationship to matrix mineral-


ization. Kidney In! 20:375—385, 1981 13. OTT SM, MALONEY NA, COBURN JW, ALFREY AC, SHERRARD DJ:

in bone histoquantitative parameters and histochemical staining reactions for aluminium in a group of patients with chronic renal


The presence of bone aluminium deposition in renal osteodystrophy and its relation to the response to calcitriol therapy. N Engi J Med 307:709—713, 1982

14. BOYCE BF, FELL GS, ELDER HY, JUNOR BJ, ELLIOTT HL, BEASTALL G, FOGELMAN I, BOYLE IT: Hypercalcaemic osteomalacia due to aluminium toxicity. Lancet ii: 1009—1012, 1982 15. ELLIOTT HL, MACDOUGALL AT: Aluminium studies in dialysis

encephalopathy. Proc EDTA

failure following a reduction in the aluminium concentration of the haemodialysis fluid. J Clin Pathol 37:743—747, 1984 29. WINNEY RJ, C0wIE JF, ROBSON JS: What is the value of plasmalserum aluminium in patients with chronic renal failure? Clin Nephrol 24:S2—S8, 1985 30. MORRISSEY J, ROTHSTEIN M, MAYOR G, SLATOP0LSKI E: Suppres-

sion of parathyroid hormone secretion by aluminium. Kidney bit

15:157—163, 1978

23:699—704, 1983


alopathy syndrome. Possible aluminium intoxication. N Engi J Med

Electron microscopic observations of bone tissue prepared by

294:184—188, 1976

ultracryomicrotomy. J Ultrastruc



Effect of water deionisers on "fracturing osteodystrophy" and

dialysis encephalopathy in Plymouth. Nephron 29:80—84, 1981 18. BROWN DJ, DAWBORN JK, HAM KN, XIPELL JM: Treatment of

dialysis osteomalacia with desferrioxamine. Lance! ii:343—345,

osteomalacic bone, in Bone Workshop,

induced osteomalacia: Inhibition of hydroxyapatite formation and


Blackwell Scientific Publications, 1971, pp. 394—407 20. WINNEY RJ, COWIE JF, CUMMING AD, SHORT AIK, SMITH GD,

growth. Calcif Tissue In! 36:439—441, 1984 34. KRAUT JA, SHINABERGER JH, SINGER FR, SHERRARD DJ, SAXTON

J, MILLER JH, KUROKAWA K, COBURN JW: Parathyroid gland responsiveness to acute hypocalcemia in dialysis osteomalacia.

Epidemiology of aluminium toxicity in a 'low incidence' area. Conir Nephrol 38:47—58, 1984 ROBSON JS:

21. GOLDNER J: A modification of the Masson trichrome technique for routine laboratory purposes. Am J Pathol 14:237—243, 1938

22. ELLIS HA: Metabolic bone disease, in Recent advances in histopathology (vol 12), edited by ANTHONY PP, MACSWEEN RNM. Edinburgh, Churchill Livingstone, 1981, pp. 185—202 23. PEARSE AGE: Histochemistry, Theoretical and Applied (vol 2). Edinburgh, Churchill Livingstone, 1972, pp. 1407—1408 MALLUCHE HH, SHERMAN D, MANAKA R, MASSEY SG: Compar-


Histomorphometry. 3rd International Workshop, edited by JEE WSS, PARFITT AM. Metab Bone Dis Rel Res 2(suppl):449—451, 1980

25. REVELL PA: Histomorphometry of bone. J Clin Pathol 36:1323— 1331, 1983 26. MCCLURE J, FAZZALARI NL, FASSETT RG, PUG5LEY DS: Bone

histo-quantitative findings and histochemical staining reactions for aluminium in chronic renal failure patients treated with haemodialysis fluids containing high and low concentrations of aluminium. J Pathol 36:1281—1287, 1983

Kidney Int 23:725—730, 1983 ANDRESS D, FELSENFELD AJ, V0IGT5 A, LLACH F: Parathyroid

hormone response to hypocalcemia in hemodialysis patients with osteomalacia. Kidney Int 24:364—370, 1983 36. FEL5ENFELD AJ, HAURESON JM, GUTMAN RA, WELLS SA,

DREZNER MK: Osteomalacia after parathyroidectomy in patients with uremia. Ann Intern Med 96:34—39, 1982 37. SLATOPOLSKY E, RUTHERFORD E, HRUSKA K, MARTIN K, KLAHR

5: How important is phosphate in the pathogenesis of renal

ison between different histomorphometric methods, in Bone


Histomorphometry. 3rd International

edited by JEE WSS, PARFITT AM. Metab Bone Dis Rel

Res 2(suppl):79—86, 1980 33. BLUMENTHAL NC, POSNER AS: In vitro model of aluminium—

19. ARMITAGE P: Statistical methods in medical research. Oxford,


Res 59:185—206, 1977



osteodystrophy? Arch Intern Med 138:848—852, 1978 PIERIDES AM, WARD MK, KERR DNS: Haemodialysis encephalopathy: Possible role of phosphate depletion. Lance! i:1234—1235,


RK, BAILLOD RN, MOORHEAD JF: Persistent hypophosphataemia

and osteomalacia in dialysis patients not on oral phosphate—binders: response to dihydrotachysterol therapy. Lancet ii:439—42, 1976 40. PARFITT AM: Soft—tissue calcification in uremia. Arch Intern Med 124:544—556,