Renal Osteodystrophy in CAPD

A.J. Hutchison, A.J. Freemont, G.A. Lumb, R. Gokal

At start of dialysis most patients have histological bone abnormalities. These can be divided into two group.l' high turnover and low turnover bone disease. Low turnover aplastic disease was previously attributed to aluminum accumulation but is now known to occur even in patients with less than 5% surface stainable aluminium. It is characterised by a mineralisation defect, thin osteoid seams, decreased numbers of o.l1eoclasts and osteoblasts and absent aluminium staining .

We have avoided aluminium containing phosphate binders (ACPBs) completely, with a combination of oral calcium carbonate and "low calcium" ( 1.25 mMoi/I) dialysis fluid. Phosphate control has been good (mean < 1.6 mMoi/I) and over the first twelve months serum PTH levels have fallen significantly. Transient asymptomatic episodes of hypercalcaemia have occurred but no patient required ACPBs .Bone biopsies at the start ofCAPD in 34 patients showed over 50% to have osteitis fibrosa (OF) but in five cases ( 15.6% ) the aplastic lesion was found without aluminium staining. In sevenfollow-up biopsies OF improved in 3 cases, osteomalacia improved in I , became aplastic in I, while aplastic bone worsened in I and changed to mild OF in I.

We conclude that the predominant bone lesion in our patients at start of CAPD is OF, but 15% already have aplastic bone. "Low calcium" dialysis fluid enables ACPBs to be avoided in the majority ofCAPD patients .

Key words

CAPO, low calcium dialysate, hyperparathyroidism, renal osteodystrophy, aplastic bone, alumimum


the Manchester Royal Infirmary and University of Manchester Medical School, Oxford Road, Manchester England.


Renal osteodystrophy is now recognised to be a common complication of end-stage renal failure and is known to have its origins early in the onset of renal impairment (1). The incidence of symptomatic skeletal disease increases in relation to the durati on of renal failure, but even at the start of dialysis the majority of patients already have histological bone abnormalities.

Histological features

The classification of renal osteodystrophy has recently been simplified with the recognition that there are essentially two groups of diseases:

1) High turnover bone diseases

Osteitis fibrosa cystica: Increased secretion of parathyroid hormone increases both the activity and numbers of osteoclasts and osteoblasts. This results in deposition of fibrous tissue in the marrow spaces (peritrabecular fibrosis), tunnelling resorption, and the formation of so-called 'woven bone' .Skeletal mass may diminish as the rate of resorption exceeds that of formation.

Mild hyperparathyroidism: Elevated parathyroid hormone levels increase bone turnover but peritrabecular fibrosis is minimal or absent.

2) Low turnover bone diseases

Osteomalacia: Defective mineralisation of bone results in a relative increase in the amount of osteoid (unmineralised bone matrix), with an increase in the number of osteoblasts. Osteitis fibrosa can also increase osteoid mass simply as a result of increased bone turnover, but bone biopsy with dual tetracycline labelling will reliably distinguish these diseases.

Aplastic bone disease: Previously thought to be a result of aluminium accumulation in bone, this disease has now been shown also to occur in the absence of stainable bone aluminium and is characterised by a greatly reduced rate of bone formation (2). A mineralisation defect, thin osteoid seams and reduced numbers of osteoblasts and osteoclasts.

Mixed bone disease: Osteomalacia and hyperparathyroidism often coexist in uraemia with variable rates of bone turnover. Reports from other units suggest that aplastic bone disease is much commoner than previously realised. Hercz et al. found this lesion in 80/175 (45.7%) bone biopsies from unselected haemo and peritoneal dialysis patients. Serum PTH levels in patients with aplastic disease were lower than in patients with normal or high bone turnover disease. In 36% of these patients, the aplastic lesion was found in the absence of significant aluminium accumulation (2) .

Now that the toxic effects of aluminium accumulation on bone and other organs are well recognised (3,4), aluminium containing phosphate binders (ACPB) have been replaced by calcium salts at first line therapy for hyperphosphataernia in most renal units. Unfortunately, hypercalcaemia has been found to be a common side effect of calcium carbonate (5,6, 7), limiting its use and raising concern about increased metastatic calcification. Martis et at. (8) have calculated on theoretical grounds that a dialysis fluid calcium concentration of 1.25 mmol/1 would allow the use of larger doses of oral calcium carbonate whilst reducing the risk of hypercalcaemia, and our own preliminary experience supports this (9). Such "low calcium" dialysis fluids are now being tested in other renal units in Europe and North America.

Avoidance of ACPB should result in a decline in the incidence of aluminium-releated bone disease, but it remains to be seen if the aplastic lesion will simply take its place.

The Manchester experience

At Manchester Royal Infirmary we have prospectively studied 29 patients (22 men and 7 women) aged 25 to 63 years old, for up to 12 months from the time they first started CAPD, using calcium carbonate as the sole phosphate binder and dialysis fluid with a calcium concentration of 1.25 mmol/1 (with sodium 132, lactate 40, magnesium 0.25 and chloride 102 mMo1/I). As patients began dialysis with the "low calcium" fluid, the dose of calcium carbonate (in the form of CalcichewR, 1.25 gm calcium carbonate, 500 mg elemental calcium) was increased to 2.5 gm taken twice daily with main meals. This amount was subsequently adjusted to achieve a serum phosphate level of 1.5 mmol/litre or less, and an ionised calcium level of 1.2 to 1.3 mmol/litre. The mean dosage of calcium carbonate required to achieve these aims was around 10 gm daily.

AI129 patients (plus five others who subsequently did not enter the study) underwent a bone biopsy at start of dialysis, and agreed to two further annual biopsies. The predominant lesion in the initial 34 biopsies were osteitis fibrosa (Figure I) , but in fi ve cases (15.6%) aplastic bone was found. None of the five aplastic biopsies showed any staining for aluminium (solochrome azurine stain) so that this can be excluded as a causative factor (10). Our definition of non-aluminium related aplastic bone comprises a mineralisation defect, reduced numbers of osteoclasts and osteoblasts, thin osteoid seams and negative staining for aluminium. Correlation of biopsy histology with serum PTH (intact molecule assay -Nichols " Allegro@") taken at the time of the biopsy (Figure 2) showed the aplastic group to have a lower mean value than the osteitis fibrosa group, although at the limit of statistical significance (p = 0.05).

Mean serum ionised calcium at start of dialysis was 1.14 :t 0.17 mmol/l, rising to 1.24 :t 0.12 mmol/1 after 6 months and 1.25 :t 0.07 after 12 months. Serum phosphate control was good, with a mean value of around 1.50 mMol/1 throughout the twelve months (Figure 3). Serum aluminium levels were all within the normal range (0-10 ug/l) at start of dialysis (mean = 5.8 :t 3.1 fl,g/l) and remained so, emphasising the fact that aluminium containing phosphate binders are the main source of aluminium load for dialysis patients today. Most importantly, mean serum PTH (intact molecule assay) fell from 385 pg/ml at start of dialysis to 120 pg/ml at six months and 28 pg/ml at 12 months (Figure 4). Episodes of hypercalcaemia have occurred in some patients, partly as a result of our deliberate attempt to maintain serum ionised calcium above 1.2 mMol/l. These results correlate well with the findings of Hercz et al (4), and Slatopolsky et at. (II).

Repeat bone biopsies are currently available from seven patients. In three of these the initial biopsy showed OF, and all of these have improved significantly. Two biopsies showed OM initially, one has improved and one has become aplastic. The other two biopsies showed aplastic bone and of these one has worsened and the other has changed to mild OF.


The predominant bone lesion seen in our patients at start of CAPD therapy is osteitis fibrosa which appears to improve on subsequent annual biopsies. Aplastic bone without aluminium staining occurred in 5/34 ( 15.6% ) patients and further repeat biopsies and PTH levels (which appear to correlate) are awaited with interest.

Lowering the calcium concentration of dialysis fluid to 1.25 mmol/l allows larger doses of calcium carbonate to be used, serum phosphate control to be improved and serum parathyroid hormone levels to be reduced. A few patients may require a lower calcium dialysate -perhaps 0.75 mMol/l, but failure to comply with large doses of calcium carbonate might then worsen bone disease.

The use of "low calcium" dialysis fluid with oral calcium carbonate enables aluminium containing phosphate binders to be avoided in the majority of CAPD patients.


Dr. Hutchison is supported by a Baxter Healthcare research grant.


  1. Avioli L V. The renal osteodystropies. In: Brenner BM, Rector FC, eds. The kidney, 1986.
  2. Hercz G, Pei Y, Manuel A, et at. Aplastic osteodystrophy without aluminium staining in dialysis patients. Kidney International, 1990; 37: 449.
  3. de Broe ME, D'Hasse PC, EIseviers MM et at. Aluminium and end-stage renal failure. In: Davidson AM Ed. Nephrology: Proceedings of the Xth ICN. 1988; 1086-1116.
  4. Andreoli SP, Briggs ill, Junor BJR. Aluminium intoxication from aluminium-containing phosphate binders in children with azotemia not undergoing dialysis. N Eng J Med 1984; 310: 1079-84.
  5. Salusky IB , Coburn JW, Foley J et at. Effects of oral calcium carbonate on control of serum phosphorus and changes in plasma aluminum levels after discontinuation of aluminum-containing gels in children receiving dialysis. J Pediat 1986; 108: 767 70.
  6. Slatopolsky E, Weerts C, Lopez-Hilker S et at. Calcium carbonate as a phosphate binder in patients with chronic renal failure undergoing dialysis. N EngJMed1986;315: 157-61.
  7. Stein HD, Yudis M, Sirota RA. Calcium carbonate as a phosphate binder. N EngJ Med 1987; 316: 109-10.
  8. Martis L, Serkes KD, Nolph KD. CaC03 as a phosphate binder: Is there a need to adjust peritoneal dialysate calcium concentrations. ..? Perit Dial Int 1989; 9: 325-28.
  9. Hutchison AJ, Gokal R. Low calcium dialysate and high dose oral calcium carbonate in CAPD. Nephrol Dial Trans (in Press).
  10. Denton J, Freemont AJ, Ball J. Detection and distribution of aluminium in bone. J Clin Patho11984; 37: 136-43.
  11. Slatopolsky E, Weerts C, Norwood K et at. Longterm effects of calcium carbonate and 2.5 mEq/liter calcium dialysate on mineral metabolism. Kidney International 1989; 36: 897 -903.
Correspondence to:
Alastair J. Hutchison, Renal Unit, Manchester Royal Infirmary, Oxford Road, Manchester MI3 9WL, United Kingdom.