UKPID MONOGRAPH ALUMINIUM SULPHATE SM Bradberry BSc MB MRCP ST Beer BSc JA Vale MD FRCP FRCPE FRCPG FFOM National Poisons Information Service (Birmingham Centre), West Midlands Poisons Unit, City Hospital NHS Trust, Dudley Road, Birmingham B18 7QH This monograph has been produced by staff of a National Poisons Information Service Centre in the United Kingdom. The work was commissioned and funded by the UK Departments of Health, and was designed as a source of detailed information for use by poisons information centres. Peer review group: Directors of the UK National Poisons Information Service. ALUMINIUM SULPHATE Toxbase summary Type of product Used in the paper industry, in water purification, for tanning leather, fireproofing and water proofing cloth and in antiperspirants. Toxicity Most cases occur in renal dialysis patients exposed to intravenous or intraperitoneal aluminium-containing dialysates. Mild gastrointestinal symptoms were reported in Camelford in 1988 after residents drank water to which aluminium sulphate had inadvertently been added in substantial amount, but there were no confirmed on-going adverse sequelae. Features Topical - Irritant to the skin and eyes. Ingestion Minor ingestions (dilute solutions): - Burning in the mouth and throat, mild gastrointestinal upset. Substantial ingestions: - Nausea, vomiting, diarrhoea, abdominal pain, rarely haemorrhagic gastritis, circulatory collapse and multi-organ failure. - Increased aluminium absorption and retention in bone is reported following acute ingestion without apparent adverse sequelae. - Chronic exposure to aluminium sulphate in drinking water may be involved in the pathogenesis of Alzheimer's disease though this remains a highly contentious issue. Inhalation - Potential pulmonary irritant Injection - Aluminium sulphate used in water purification is a source of aluminium toxicity in haemodialysis patients and may cause "dialysis dementia". Aluminium toxicity in these circumstances may contribute also to renal osteodystrophy and a microcytic anaemia. Management Topical - Irrigate with copious volumes of lukewarm water. Ingestion Minor ingestions (dilute solutions; mildly acidic or neutral): 1. Gastrointestinal decontamination is unnecessary. 2. Symptomatic and supportive measures only. Substantial ingestions (concentrated acid solutions): 1. Do not attempt gastric decontamination. 2. Secure cardiorespiratory stability. 3. Replace fluids and electrolytes if necessary. 4. Upper gastrointestinal endoscopy may be required. 5. Measure serum aluminium concentration in patients with clinical features. 6. Parenteral desferrioxamine may be considered if there is evidence of an increased aluminium body burden. Seek specialist advice from the NPIS. Injection - There is clinical evidence that desferrioxamine therapy can improve aluminium-induced encephalopathy, bone disease and anaemia in dialysis patients but seek specialist advice from the NPIS. References Day JP, Ackrill P. The chemistry of desferrioxamine chelation for aluminum overload in renal dialysis patients. Ther Drug Monit 1993; 15: 598-601. Eastwood JB, Levin GE, Pazianas M, Taylor AP, Denton J, Freemont AJ. Aluminium deposition in bone after contamination of drinking water supply. Lancet 1990; 336: 462-64. Lowermoor Incident Health Advisory Group; Department of Health. Water pollution at Lowermoor, North Cornwall. 2nd report. London: HMSO, 1991; 1-51. McLaughlin RS. Chemical burns of the human cornea. Am J Ophthalmol 1946; 29: 1355-62. Substance name Aluminium sulphate Origin of substance Naturally occurring, as the mineral alunogenite. (DOSE, 1992) Synonyms Alum Cake alum Aluminium sesquisulphate Dialuminium trisulphate (CSDS, 1989) Chemical group Compound of aluminium, a group III metal (CSDS, 1989) Reference numbers CAS 10043-01-3 (CSDS, 1989) RTECS BD1700000 (RTECS, 1995) UN NIF HAZCHEM CODE NIF Physicochemical properties Chemical structure Aluminium sulphate, Al2 (SO4)3 (DOSE, 1992) Molecular weight 342.15 (DOSE, 1992) Physical state at room temperature Solid (powder) (CSDS, 1989) Colour White (CSDS,1989) Odour Odourless (HSDB, 1995) Viscosity NA pH Aluminium sulphate has a natural pH of 2. (Anonymous, 1989) Solubility Soluble in water: 313 g/L at 0°C, 891 g/L at 100°C. (DOSE, 1992) Autoignition temperature NIF Chemical interactions Sulphuric acid is formed on hydrolysis of aluminium sulphate. On heating, toxic sulphur oxide and aluminium oxide fumes are released. (CSDS, 1989) Major products of combustion NIF Explosive limits NA Flammability May burn, but will not ignite. (HSDB, 1995) Boiling point The solid decomposes at 770°C. (HAZARDTEXT, 1995) Density 1.61 at 25°C (CSDS, 1989) Vapour pressure Essentially zero (HSDB, 1995) Relative vapour density NA Flash point NA Reactivity NIF Uses Aluminium sulphate is used in the paper industry, for tanning leather and as a mordant in dyeing. Other uses include water purification and sewage treatment, fireproofing and waterproofing cloth, and as an ingredient of antiperspirants. (CSDS, 1989; DOSE, 1992) Hazard/risk classification NIF INTRODUCTION Aluminium sulphate is relatively non toxic. In 1988, contamination of drinking water with aluminium sulphate led to a variety of acute symptoms (Lowermoor Incident Health Advisory Group, 1991) but there are no case reports of substantial ingestions in the last 30 years. Uraemic patients on long-term haemodialysis may have an increased aluminium load via intravenous exposure to high concentrations of aluminium sulphate in dialysis (usually tap) water (Ward et al, 1978; Alfrey, 1980), although this should now be avoidable by deionization or reverse osmosis of the water prior to use. MECHANISM OF TOXICITY There is experimental evidence that aluminium inhibits bone mineralization partly by the deposition of aluminium at the osteoid/calcified-bone boundary thereby directly inhibiting calcium influx, and partly by aluminium accumulation in the parathyroid glands with suppression of parathyroid hormone secretion (Visser and Van de Vyver, 1985; Berland et al, 1988; Firling et al, 1994). Proposed mechanisms of aluminium-induced neurotoxicity include free-radical damage via enhanced lipid peroxidation, impaired glucose metabolism, effects on signal transduction and protein modification and alterations in the axunal transport and phosphorylation state of neurofilaments (Birchall and Chappell, 1988; Exley and Birchall, 1992; Erasmus et al, 1993; Winship 1993; Haug et al, 1994; Joshi et al, 1994; Strong, 1994). TOXICOKINETICS Absorption In a healthy adult only approximately 15 µg of the average daily dietary aluminium intake of 3-5 mg is absorbed (Winship, 1992). The intestinal absorption of aluminium is enhanced by citrate which is found frequently in effervescent drug formulations. Main and Ward (1992) reported a reversible increase in the serum aluminium concentration from 67.5 to 499.5 µg/L in a patient on haemodialysis taking oral aluminium hydroxide when she was also given an effervescent analgesic containing sodium citrate. Conversely the bioavailability of aluminium in aqueous solution is greatly reduced by silica, such that the toxicity of aluminium-containing phosphate binders may be reduced significantly by the co-administration of dissolved silica (Birchall, 1993). In dialysis patients with aluminium overload, endogenous silicon may serve a protective role in limiting tissue aluminium uptake (Fahal et al, 1994). There may also be implications for domestic water supplies if a high silicic acid concentration evades any hazards posed by aluminium sulphate in water (Birchall, 1993). Distribution More than 90 per cent of absorbed aluminium is bound to transferrin which does not cross the blood-brain barrier readily. The remaining ten per cent is associated with low molecular weight complexes, such as citrate, which can accumulate in brain tissue. In the body aluminium is stored mainly in bone and liver. Excretion Aluminium is excreted predominantly via the kidneys and therefore will accumulate in patients with renal failure (Alfrey, 1980). Preterm infants also have a limited ability to excrete aluminium and there are reports of accumulation in infants on long-term parenteral nutrition (Sedman et al, 1985). Tsou et al (1991) demonstrated significantly higher plasma aluminium concentrations (mean 37.2 µg/L) in normal infants receiving oral aluminium-containing antacids for a prolonged period compared to controls but this was not associated with adverse clinical effects. CLINICAL FEATURES: ACUTE EXPOSURE Dermal exposure Topical aluminium sulphate is irritant to the skin (Royal Society of Chemistry, 1989; Meditext, 1995). Tohani et al (1991) reported a macular, pruritic rash in association with the accidental addition of aluminium sulphate to a water supply. The skin reaction may have been due to the leaching of other metals (especially nickel) from the domestic plumbing system due to the acidic pH of the water caused by the high aluminium sulphate concentration. Ocular exposure Ocular exposure to aluminium sulphate may cause corneal burns (McLaughlin, 1946). Ingestion Acute aluminium sulphate ingestion causes primarily gastrointestinal upset though neuropsychological and musculoskeletal sequelae were also reported in those who drank water to which 20 tonnes of eight per cent aluminium sulphate had accidentally been added in Camelford, Cornwall in 1988 (see below). However, a group reviewing this incident concluded that although "early symptoms, such as gastrointestinal disturbances, rashes and mouth ulcers, could probably be attributed to the toxic effects of the incident....The research reported to us does not provide convincing evidence that harmful accumulation of aluminium has occurred, nor that there is greater prevalence of ill-health due to toxic effects of the water in the exposed population" (Lowermoor Incident Health Advisory Group, 1991). Gastrointestinal toxicity Following the Camelford incident a variety of acute symptoms were reported (Eastwood et al, 1990; Lowermoor Incident Health Advisory Group, 1991), mainly mild gastrointestinal disturbance and mouth ulcers (McMillan et al, 1993b). Water aluminium concentrations recorded at the time ranged from 30 to 620 mg/L (WHO recommended maximum aluminium concentration in drinking water is 200 µg/L) (Eastwood et al, 1990; WHO, 1993). The low pH of the water also led to leaching of other metals, such as copper, from the distribution pipes so that those washing in the water developed green-discoloration of their hair. Substantial ingestion of aluminium sulphate causes burning in the mouth and throat, gingival necrosis, nausea, vomiting, diarrhoea, abdominal pain and, in severe cases, haemorrhagic gastritis with circulatory collapse (Gosselin et al, 1984; Royal Society of Chemistry, 1989; Meditext, 1995). Spira (1933) reported dry mouth and throat, anorexia, nausea and vomiting, glossitis, stomatitis, gingivitis and hiccup (in addition to cutaneous and neurological symptoms) suspected to be caused by the ingestion of elemental aluminium plus "aluminized" tap water. There are, however, no case reports of substantial aluminium sulphate ingestion in the literature over at least the last 30 years. Dermal toxicity Spira (1933) believed that skin conditions including urticaria, telangiectasia, alopecia and palmar keratosis were caused by ingestion of elemental aluminium plus "aluminized" tap water but this was not confirmed. Neuropsychological toxicity McMillan et al (1993b) reported "possible impairment of memory" in those who drank aluminium sulphate-contaminated water in Camelford. Serial neuropsychological assessments of 10 individuals between eight and 26 months after the incident showed mild cognitive impairment but a causal link with aluminium exposure could not be verified and in no patient was there evidence of residual aluminium in bone or plasma McMillan et al, 1993b). Retrospective psychological testing on 39 children from schools in the contaminated area showed no significant differences compared to unexposed children (McMillan et al, 1993a). Muscle twitching, limb paraesthesiae, arthralgia, neuralgia, giddiness, depression and lassitude have been attributed to the ingestion of elemental aluminium and aluminium contaminated tap water (Spira, 1933) but this was speculative. Musculoskeletal toxicity Some individuals who drank the aluminium sulphate-contaminated water in Camelford later complained of joint and muscle pains and fatigue (Anonymous, 1991; McMillan et al, 1993b). Bone toxicity Bone biopsies some six months after the Camelford incident from two individuals who drank the contaminated water showed increased aluminium staining in bone formed at a time compatible with exposure but normal bone aluminium concentrations overall, suggesting increased aluminium absorption and retention following acute ingestion without apparent adverse sequelae (Eastwood et al, 1990). Injection Cumming et al (1982) reported four dialysis patients who developed anorexia, nausea, vomiting, abdominal pain, weight loss and malaise within three days of using a contaminated peritoneal dialysis fluid containing 620-1460 µg/L aluminium. CLINICAL FEATURES: CHRONIC EXPOSURE Inhalation Pulmonary toxicity No evidence of respiratory disease was found in 25 workers engaged in the manufacture of aluminium sulphate from aluminium hydroxide where the aluminium concentration in factory air was kept below the permissible limit of 3 mg/m3 (Elo and Uksila, 1977). Ingestion Neuropsychological toxicity Aluminium sulphate is widely used in water purification. This source has in the past contributed to aluminium toxicity in haemodialysis patients (see below) and may be involved in the pathogenesis of Alzheimer's disease via accumulation of aluminium in the brain but this remains a highly contentious issue (Ebrahim, 1989; Petit, 1989; Murray et al, 1991; Crapper McLachlan, 1994; Munoz, 1994). Martyn et al (1989) and Neri and Hewitt (1991) reported a geographical relationship between Alzheimer's disease and the concentration of aluminium in drinking water but Wood et al (1988) found no significant difference in mental test score between hip fracture patients living in high versus low water aluminium areas. Animal studies have demonstrated the ability of aluminium to induce the formation of neurofibrillary tangles (Klatzo et al, 1965), impair the learning ability of rats, and increase brain acetylcholinesterase activity in a similar way to that seen in Alzheimer's disease (Bilkei-Gorzó, 1993). Other workers have shown elevated aluminium concentrations in brain tissue from patients with Alzheimer's disease (Crapper et al, 1973) and laser microprobe studies have demonstrated aluminium accumulation in the neurofibrillary tangles of these patients (Good et al, 1992). Harrington et al (1994) described Alzheimer's-disease-like pathological changes in the brains of renal dialysis patients in association with aluminium accumulation without clinical evidence of dialysis encephalopathy (see below). Injection Neuropsychological toxicity 'Dialysis dementia' involves the accumulation of aluminium, mainly in the brain, in patients on haemodialysis where the dialysis water contains significant amounts of aluminium sulphate (McDermott et al, 1978). This should now be avoidable by reverse osmosis or deionization of dialysis water prior to use. Peritoneal dialysis solutions and haemofiltration and plasma exchange substitution fluids also may contain excessive aluminium although the incidence of toxicity from these preparations is small (Mion, 1985; Mousson et al, 1989). Elliott et al (1978a) reported a significant correlation between serum aluminium concentrations and the concentrations of aluminium in the water supply in eight patients on home dialysis with cases of dialysis dementia confined to regions with high water aluminium concentrations (>350 µg/L). A larger study (Registration Committee of the European Dialysis and Transplant Association, 1980) of patients treated in 65 dialysis centres in Europe in 1976 and 1977 identified 150 cases of dialysis dementia with a clear association between the occurrence of dementia and dialysis with water which was not treated by deionization or reverse osmosis. Only 23 of 150 patients were still alive in 1978. Similar findings were reported by Davison et al (1982). Although the problem of aluminium contamination of dialysates has reduced in recent years, these patients may still accumulate aluminium via oral aluminium hydroxide given as a phosphate binder (Salusky et al, 1991). The contribution this makes to dialysis dementia is likely to be small (McDermott et al, 1978; Registration Committee of the European Dialysis and Transplant Association, 1980). It has been suggested (Hodge et al, 1981) that to ensure no aluminium uptake by dialysis patients, the dialysate aluminium concentration should not exceed 14 µg/L. This corresponds to a maximum aluminium concentration in the water supply of 5 µg/L since aluminium is present in significant quantity in the dialysate concentrate (Hodge et al, 1981). The WHO recommended maximum aluminium concentration in drinking water is 200 µg/L (WHO, 1993). Dialysis dementia is progressive and often fatal (Alfrey et al, 1976; Burks et al, 1976). In a review of 412 dialysis patients admitted to one renal unit since 1972, Garret et al (1988) described 38 cases. The mean time between onset of regular dialysis and development of symptoms was 40 months with speech difficulties, seizures and myoclonus the most common presenting features. Dyspraxia, involuntary movements, poor concentration, loss of short-term memory, confusion, depression and anxiety were also described and only nine patients were still alive at the conclusion of the study (Garret et al, 1988). The neurological consequences of aluminium intoxication may be exacerbated in patients with aluminium bone disease who sustain fractures, probably via increased bone aluminium mobilization (Davenport and Ahmad, 1988). There are reports of impaired cerebral function in haemodialysis patients who have an increased body burden of aluminium but no evidence of "dialysis dementia" (Altmann et al, 1989; Bolla et al, 1992). Altmann et al (1989) found significant abnormalities of psychomotor function in 27 long-term haemodialysis patients who had only mildly raised serum aluminium concentrations (mean 59 ± (SEM) 9 µg/L, normal < 10 µg/L). In another study, 23 haemodialysis patients accidentally exposed to aluminium for up to six months following failure of a reverse osmosis system had a mean serum aluminium concentration of 147.3 ± (SEM) 11.7 µg/L without apparent neuropsychiatric sequelae (Caramelo et al, 1995). Bone toxicity In patients with renal failure aluminium toxicity may contribute to renal osteodystrophy (Goyer et al, 1994). In a survey conducted by the European Dialysis and Transplant Association, 102 of 150 patients with dialysis dementia also had evidence of bone disease (Registration Committee of the European Dialysis and Transplant Association, 1980). Parkinson et al (1979) demonstrated a significant correlation (p = 0.01) between the mean aluminium content of the water supply and the incidence of fracturing dialysis osteodystrophy in 1293 patients undergoing intermittent haemodialysis and Ward et al (1978) reported significantly fewer cases of osteomalacia in patients maintained on regular haemodialysis in Newcastle when the dialysate water was deionised. Another study (Chan et al, 1990) found that the incidence of osteomalacic fractures in dialysis patients could not be explained by the aluminium concentration in dialysate alone and a significant contribution by oral aluminium hydroxide was suggested. Aluminium associated osteomalacic osteodystrophy is progressive and characterized by bone pain, a proximal myopathy and spontaneous fractures. In a review of skeletal surveys of 67 patients with end-stage renal failure Garrett et al (1986) found that moderate or severe fracturing osteodystrophy with greater than five fractures had a diagnostic specificity for aluminium intoxication of 100 per cent, provided trauma could be excluded. Investigations typically show normal or only slightly increased alkaline phosphatase activity, normal serum calcium and normal or slightly high serum phosphate concentrations with reduced circulating parathyroid hormone and increased bone and serum aluminium concentrations (Winship, 1992). It is resistant to treatment with vitamin D but improvement may follow desferrioxamine therapy as discussed below. Cardiovascular toxicity Elliott et al (1978b) proposed aluminium induced cardiotoxity (via inhibition of magnesium and ATP-dependent enzymes) as a contributing factor in the sudden death of five dialysis patients, four with dialysis encephalopathy and one non-encephalopathic patient whose serum aluminium concentration was 600 µg/L. Haemotoxicity Aluminium intoxication may exacerbate the microcytic hypochromic anaemia of chronic renal failure via impaired iron utilization (Caramelo et al, 1995). This effect is at least partly reversible with desferrioxamine therapy as discussed below. Dermal toxicity Brown et al (1992) suggested aluminium overload as the cause of a widespread pruritic nodular rash (prurigo nodularis) in three patients on maintenance haemodialysis. The serum aluminium concentration, measured in two patients, was normal but complete resolution occurred in all cases following weekly treatment for some three months with one gram intravenous desferrioxamine. MANAGEMENT Dermal exposure Remove all soiled clothing, wash the exposed area thoroughly with copious amounts of water and treat symptomatically. Ocular exposure The affected eye should be irrigated with lukewarm water for not less than 15-30 minutes and any visible particles removed. Installation of local anaesthetic is usually necessary to enable adequate decontamination. An ophthalmic opinion should be obtained. Ingestion The likelihood of adverse effects following aluminium sulphate ingestion depends on the pH of the solution. Although pure aluminium sulphate has a pH of 2 most cases will involve ingestion of dilute solutions causing no, or only mild, gastrointestinal upset. Ingestion of dilute solutions These cases will require only symptomatic and supportive measures. Gastrointestinal decontamination is unnecessary as gastrointestinal aluminium absorption is poor. Measurement of the serum aluminium concentration or the administration of desferrioxamine are not necessary. Ingestion of concentrated solutions Theoretically, ingestion of a concentrated aluminium sulphate solution will cause severe gastrointestinal corrosion and, if this is clinically suspected, gastric aspiration or lavage should not be attempted. Activated charcoal does not adsorb aluminium and the administration of ipecac is not advocated under any circumstances. Measures to secure cardiorespiratory stability are mandatory and clinical suspicion of oesophageal or gastric corrosion may require endoscopic examination. Corticosteriods are not indicated and their administration may mask signs of abdominal perforation. The serum aluminium concentration should be measured in patients with clinical features and parenteral desferrioxamine therapy should be considered if there is evidence of an increased aluminium body burden; such as increase is only likely to occur after chronic aluminium ingestion. Inhalation The patient should be removed from exposure, and those with evidence of respiratory distress should receive high flow oxygen by face-mask. Corticosteriods may be employed if laryngeal or pulmonary oedema are present and, theoretically endotracheal intubation or tracheostomy may be required although there are no such case reports in the literature. Injection Most cases of aluminium intoxication following aluminium sulphate exposure occur in renal dialysis patients exposed to either intravenous or intraperitoneal aluminium-containing dialysates. The oral administration of aluminium containing phosphate binders may exacerbate aluminium accumulation in these circumstances. Prevention The prevention of aluminium intoxication in renal dialysis patients requires: 1. Monitoring of the water aluminium concentration in the supply used to prepare the dialysate. The maximum EC admissible aluminium concentration in domestic water is 0.2 mg/L (Eastwood et al, 1990) but Hodge et al (1981) suggested a maximum water aluminium concentration of 5 µg/L as the 'safe' limit for use in haemodialysis. 2. Judicious use of aluminium-containing phosphate binding agents with the use of alternative preparations, such as calcium carbonate, where possible. 3. Monitoring the total body aluminium load. There is some disagreement regarding the role of plasma aluminium concentrations in the estimation of total body aluminium burden (Gilli et al, 1983; Seyfert et al, 1987; D'Haese et al, 1990) but it is likely to be a useful measurement provided it is undertaken by a qualified laboratory. 4. The definitive investigation for aluminium osteodystrophy is a bone biopsy (Seyfert et al, 1987). The desferrioxamine infusion test is a less invasive method of assessing body aluminium although its interpretation requires clarification (see 'Medical Surveillance'). Antidotes There is some evidence that parenteral desferrioxamine therapy slows the rate of cognitive deterioration in patients with Alzheimer's disease (Crapper McLachlan et al, 1991; Crapper McLachlan et al, 1993) but further studies are required. Desferrioxamine (deferoxamine) Desferrioxamine forms a stable complex with aluminium and in animal studies it mobilises aluminium primarily from bone with subsequent urinary elimination of the chelate (Gómez et al, 1994; Yokel 1994). It is absorbed poorly from the gastrointestinal tract and parenteral therapy is necessary. Theoretically 100 mg desferrioxamine can bind 4.1mg aluminium (Winship, 1993). The desferrioxamine chelate is dialyzable and all published clinical studies of aluminium chelation using desferrioxamine involve patients with renal failure undergoing haemodialysis or, less commonly, peritoneal dialysis (O'Brien et al, 1987) or haemofiltration (Sulkova et al, 1991). Following intravenous desferrioxamine administration the concentrations of protein (mainly transferrin)-bound and erythrocyte aluminium remain relatively constant as chelatable aluminium is mobilized from bone (Day and Ackrill, 1993). The aluminium-desferrioxamine chelate concentration reaches a maximum 12-24 hours post infusion (Day and Ackrill, 1993) producing a rise in the total plasma aluminium concentration which persists until the next dialysis. The administration of desferrioxamine shortly before dialysis will reduce this effect (Douthat et al, 1994). In the longer term, a fall in the erythrocyte aluminium concentration is observed as red cells are formed from bone marrow with a lower aluminium load (Day and Ackrill, 1993). In chronic renal failure patients treated with 1 gram intravenous desferrioxamine, Sulkova et al (1991) observed a mean 41 per cent decrease in the serum aluminium concentration during 28 five hour haemodialyses (aluminium clearance 28 mL/min during each dialysis) compared to a mean 66 per cent reduction in the serum aluminium concentration during 36 sessions of haemofiltration (volume exchange 60 per cent of body weight and calculated aluminium clearance 42 mmol/L in the 60th minute of each filtration). The authors concluded haemofiltration is superior to haemodialysis in enhancing aluminium elimination using desferrioxamine. In another study O'Brien et al (1987) calculated aluminium clearance rates in a 32 year-old male with aluminium osteomalacia following a change from haemodialysis to chronic ambulatory peritoneal dialysis (CAPD). CAPD plus intravenous desferrioxamine (six grams once a week) gave an aluminium clearance of 4.2 mL/min compared to a clearance of 3.1 mL/min when the same cumulative dose of desferrioxamine was given into the peritoneal cavity and an aluminium clearance of 2.5 mL/min with CAPD alone. Indications for desferrioxamine therapy There is clinical evidence that desferrioxamine therapy can improve aluminium-induced encephalopathy, bone disease and anaemia in dialysis patients (Day and Ackrill, 1993). Its use should be considered, therefore, in these patients in the following circumstances: 1. When features are present compatible with dialysis encephalopathy in the absence of an alternative neurological diagnosis. 2. Where a desferrioxamine infusion test indicates an increased body aluminium load. As discussed below (see Medical Surveillance) there are some problems with the interpretation of this test. Milliner et al (1984) found that a 200 µg/L increase above baseline in the plasma aluminium concentration following 40 mg/kg intravenous desferrioxamine provided 50 per cent specificity for a diagnosis of aluminium osteodystrophy with the diagnostic specificity improving to 71 per cent with an increase above baseline in the plasma aluminium concentration of 300 µg/L. 3. Where there is clinical evidence of aluminium-related bone disease. This is usually associated with a 'positive' desferrioxamine infusion test. 4. Possibly in the presence of severe, transfusion-dependant anaemia even in the absence of characteristic clinical or analytical features of aluminium overload (Praga et al, 1987). 5. In the presence of an increased "baseline" serum aluminium concentration. D'Haese et al (1990) suggested that a serum aluminium concentration in excess of 60 µg/L reliably indicated aluminium overload. Desferrioxamine and dialysis encephalopathy McCarthy et al (1990) treated 28 dialysis patients suffering from aluminium toxicity with long-term (mean 11.0 months) intravenous desferrioxamine, initially at a mean dose of 41.7 mg/kg body weight once weekly, increasing to a maximum dose of 60 mg/kg as tolerated. After five to seven months of treatment serum aluminium concentrations decreased from a mean of 401 µg/L to 245 µg/L. Four patients, who had advanced dementia before treatment, died during the study period. With desferrioxamine treatment seven of 28 patients showed neurological improvement and 25 patients showed improved or stable muscle strength and overall functional capacity. The authors concluded that whilst long-term desferrioxamine therapy can be important in the treatment of patients with significant aluminium exposure it should be employed only when symptoms demand treatment and when patients can be monitored regularly for desferrioxamine toxicity (McCarthy et al, 1990). A 44 year-old man treated with home dialysis (with deionized water) for four years and oral aluminium hydroxide 2850 mg daily for four months, developed severe short-term memory loss and myoclonic jerks (Arze et al, 1981). The serum aluminium concentration was 84 µg/L. Symptomatic improvement followed the additional use of a reverse osmosis unit with reduction in the dialysate aluminium concentration from 6-109 µg/L to 1-12 µg/L but no aluminium was eliminated during dialysis. Two to six grams intravenous desferrioxamine once weekly for six months removed 259mg aluminium with considerable improvement on psychometric testing and no myoclonus at the end of this time. In another report (Payton et al, 1984) a patient with aluminium encephalopathy on CAPD improved substantially following two months treatment with intraperitoneal desferrioxamine (500-750 mg added to each two litre bag of dialysate). There was a significant (p<0.001) increase in the post-dialysis dialysate aluminium concentration during desferrioxamine treatment with an initial increase (from 189 µg/L to 224 µg/L) then progressive decrease (to 27 µg/L) in the serum aluminium concentration. Desferrioxamine and aluminium bone disease In 32 patients desferrioxamine therapy controlled progression of dialysis-associated aluminium osteodystrophy with bone scans in 21 cases reverting to normal or showing a pattern typical of hyperparathyroidism (Botella et al, 1984). Charhon et al (1986) reported similar findings with intravenous desferrioxamine (3-6 grams once a week for 5-11 months) leading to dramatic clinical improvement in the bone disease of three haemodialysed patients. Desferrioxamine and dialysis-associated anaemia A three month course of desferrioxamine (30 mg/kg iv three times a week) significantly improved the microcytic anaemia of 15 haemodialysis patients who had only modestly raised serum aluminium concentrations (5-125 µg/L) and no neurological symptoms of aluminium toxicity (Altmann et al, 1988). Praga et al (1987) reported significant reduction in the transfusion requirement of seven anaemic haemodialysis patients following desferrioxamine therapy (two grams intravenously after each haemodialysis session for six months) even though none had either clinical or analytical characteristic features of aluminium intoxication. Treatment protocol for desferrioxamine In patients with an indication for desferrioxamine therapy 40-80 mg/kg should be administered, usually intravenously, once a week. The dose can be reduced to 20-60 mg/kg (as indicated by response and adverse effects) if treatment is to be continued for several months (Domingo 1989) . Canavese et al (1989) have suggested the therapeutic effectiveness of desferrioxamine may be exhausted after some two years therapy even if aluminium bone deposits persist after this time. Adverse effects of desferrioxamine Side-effects of long-term treatment with desferrioxamine include hypotension, gastrointestinal upset, porphyria cutanea tarda-like lesions, transient visual disturbances (McCarthy et al, 1990), posterior cataracts, ototoxicity (Domingo, 1989) and an increased potential for septicaema, especially Yersinia sepsis (Boyce et al, 1985). Some dialysis patients with aluminium encephalopathy develop worsening of neurological symptoms within hours of desferrioxamine treatment which may be due to desferrioxamine alone or in combination with a rising plasma aluminium concentration (McCauley and Sorkin, 1989). There are several reports of desferrioxamine-associated systemic fungal infection (mucormycosis) in dialysis patients (Goodill and Abuelo, 1987; Windus et al, 1987) and an international registry of this potentially fatal complication has been established (Boelaert et al, 1991) although a causal link between desferrioxamine and fungal infection in these patients has not been confirmed (Vlasveld and van Asbeck, 1991). Desferrioxamine and charcoal haemoperfusion Chang and Barre (1983) compared aluminium clearance by desferrioxamine plus charcoal haemoperfusion with desferrioxamine plus haemodialysis in 17 patients with chronic renal failure who were stable on standard haemodialysis. Neither method enhanced aluminium clearance without desferrioxamine but forty-eight hours after intravenous desferrioxamine charcoal haemoperfusion produced more effective aluminium clearance (mean 65.3 ± (SD) 11.2 mL/min; n=6) than haemodialysis (mean 44.6 ± (SD) 13.7 mL/min; n=4). The authors proposed haemoperfusion plus desferrioxamine as a effective method of rapid aluminium elimination in intoxicated patients to be used in series with haemodialysis in patients with renal failure. Other chelating agents The practical problems of desferrioxamine administration and its side effects have prompted a search for an alternative aluminium chelator. Uncontrolled clinical studies with d-penicillamine and dimercaprol in dialysis encephalopathy were unsuccessful (Yokel, 1994) and although in animal studies parenteral citric acid is effective (Domingo et al, 1988), evidence in man that oral citrate enhances gastrointestinal aluminium absorption means the problems of parenteral administration persist. Although as yet there is no confirmed alternative to desferrioxamine (Domingo, 1989; Main and Ward, 1992; Yokel, 1994), in a recent clinical trial Kontoghiorghes et al (1994) demonstrated that the administration of oral 1,2-dimethyl-3-hydroxypyrid-4-one in a dose of 40-60 mg/kg to six haemodialysis patients resulted in rapid aluminium mobilization. The plasma aluminium concentration peaked at one hour post chelation therapy and returned to baseline in most cases within seven hours. The aluminium chelate was readily dialysable during both haemodialysis and continuous ambulatory peritoneal dialysis. Haemoperfusion, haemodialysis and haemofiltration Chang and Barre (1983) demonstrated that haemoperfusion and haemodialysis only enhance aluminium elimination in the presence of desferrioxamine and that under these circumstances haemoperfusion is more effective than haemodialysis (see above). Protein-bound aluminium is not dialyzable (Day and Ackrill, 1993). Sulkova et al (1991) also observed no aluminium elimination during four sessions of haemodialysis in patients with known aluminium accumulation who were not pre-treated with intravenous desferrioxamine. The same authors reported a 15 per cent fall in the mean serum aluminium concentration during four haemofiltrations without desferrioxamine therapy. AT-RISK-GROUPS Chronic renal failure patients As discussed above, patients with chronic renal failure are at increased risk of aluminium toxicity from "tapwater" dialysate and possibly also oral aluminium-containing phosphate binders. Infants Sedman et al (1985) found significantly increased plasma (p<0.0001), urine (p<0.01) and bone (p<0.0001) aluminium concentrations in 18-23 premature infants receiving parenteral nutrition. Aluminium accumulation in these circumstances reflects a combination of aluminium contamination of the intravenous fluid and impaired renal aluminium excretion. Ten infants with normal renal function who had received oral aluminium containing antacids for at least one week had significantly higher plasma aluminium concentrations (mean 37.2 ± (SEM) 7.13 µg/L compared to controls (4.13 ± 0.66 µg/L) (p<0.005) without signs of toxicity (Tsou et al, 1991). The aluminium content of cow's milk and soy milk are considerably higher (10-20 and 100 fold respectively) than human breast milk which has an aluminium concentration of 5-20 µg/L and this may contribute to aluminium intoxication in premature infants with renal failure (Bishop et al, 1989). Freundlich et al (1985) reported two infants with congenital uraemia and aluminium toxicity where the source of excess aluminium was believed to be a powdered milk formulation. MEDICAL SURVEILLANCE Aluminium toxicity should be considered in those exposed occupationally to aluminium dust who develop respiratory or neuropsychiatric symptoms and in patients with renal failure who may be at risk of aluminium retention. Useful indicators of exposure include the 24 hour urine aluminium excretion (normal range < 15 µg/24 hours) and the blood aluminium concentration (normal range < 10 µg/L). Aluminium is evenly distributed between plasma and blood cells so that plasma and whole blood aluminium concentrations have similar value in assessing toxicity (van der Voet and de Wolff, 1985). In 71 dialysis patients D'Haese et al (1990) demonstrated that a serum aluminium concentration of 60 µg/L or greater identified aluminium-related bone disease with 82 per cent sensitivity and 86 per cent specificity. However, Gilli et al (1983) suggested serum aluminium concentrations were unlikely to reflect total aluminium accumulation in uraemic patients and Seyfert et al (1987) suggested plasma aluminium concentrations were not reliable in the diagnosis of aluminium-related bone disease. This author emphasised the importance of bone biopsy as the definitive investigation for aluminium osteomalacia and advocated the 'desferrioxamine test' as a useful diagnostic tool. Desferrioxamine infusion test The desferrioxamine infusion test involves the intravenous administration, to renal failure patients with suspected aluminium toxicity, of a standard desferrioxamine dose (usually 20-40 mg/kg) and comparison of the peak post-infusion plasma aluminium concentration with the "baseline" concentration, as an indication of the total body aluminium burden (Day and Ackrill, 1993). Ackrill et al (1980) suggested that an increase in the serum aluminium concentration of at least 200 µg/L was required following a desferrioxamine test dose if a significant amount of aluminium was to be removed by desferrioxamine during dialysis. The interpretation of this test has not yet been standardized and its predictive value in the assessment of aluminium bone disease requires clarification. Furthermore, the investigation is not without risk. Ravelli et al (1990) reported acute visual disturbances in 13 of 15 patients given a standard desferrioxamine test with persisting symptoms in four patients some six months later. It recently has been suggested that a lower test dose of desferrioxamine can reduce toxicity without loss of diagnostic efficacy (Yaqoob et al, 1991). Hair analysis (Wilhelm et al, 1989) and individual spot urine concentrations (Gitelman et al, 1995) are poor indicators of aluminium exposure. Furthermore, the kinetics of urine aluminium excretion varies depending on the form of aluminium involved (Pierre et al, 1995). Estimation of the aluminium content of cerebrospinal fluid may be important in the investigation of aluminium-related dementia (Sjögren et al, 1994). OCCUPATIONAL DATA Occupational exposure standard Aluminium salts, soluble: Long-term exposure limit (8 hour TWA reference period 2 mg/m3 (Health and Safety Executive, 1995). OTHER TOXICOLOGICAL DATA Carcinogenicity There is no evidence that aluminium salts are carcinogenic in man (Leonard and Gerber, 1988). Reprotoxicity Pregnant rats given oral aluminium hydroxide 192-768 mg/kg/day on gestational days 6-15 showed no maternal or foetal developmental toxicity (Gomez et al, 1990). There was no evidence of adverse effects in pregnancy following the accidental contamination of drinking water with aluminium sulphate in Camelford, Cornwall in 1988 (Golding et al, 1991). Genotoxicity Bacillus subtilis H17 (rec+) M45 (rec-) negative DNA damage. In vitro human lymphocyte cells (72 hr) 20 µg/mL induced chromosomal aberrations in cells from male and female subjects, while the frequency of translocations and dicentrics was low. Oral rat (prolonged exposure) induced dose-dependent inhibition of dividing cells and increased chromosomal aberrations, not influenced by duration of exposure (DOSE, 1992). Fish toxicity LC50 goldfish (12-96 hr) 100 mg/L (DOSE, 1992). EEC Directive on Drinking Water Quality 80/778/EEC Aluminium: Guide level 0.05 mg/L, maximum admissible concentration 0.2 mg/L. Sulphate: Guide level 25 mg/L, maximum admissible concentration 250 mg/L (DOSE, 1992). AUTHORS SM Bradberry BSc MB MRCP ST Beer BSc JA Vale MD FRCP FRCPE FRCPG FFOM National Poisons Information Service (Birmingham Centre), West Midlands Poisons Unit, City Hospital NHS Trust, Dudley Road, Birmingham B18 7QH UK This monograph was produced by the staff of the Birmingham Centre of the National Poisons Information Service in the United Kingdom. The work was commissioned and funded by the UK Departments of Health, and was designed as a source of detailed information for use by poisons information centres. Date of last revision 16/7/96 REFERENCES Ackrill P, Ralston AJ, Day JP, Hodge KC. Successful removal of aluminium from patient with dialysis encephalopathy. Lancet 1980; 2: 692-3. Alfrey AC. Aluminum metabolism in uremia. Neurotoxicology 1980; 1: 43-53. Alfrey AC, Le Gendre GR, Kaehny WD. The dialysis encephalopathy syndrome. N Engl J Med 1976; 294: 184-8. Altmann P, Plowman D, Marsh F, Cunningham J. Aluminium chelation therapy in dialysis patients: evidence for inhibition of haemoglobin synthesis by low levels of aluminium. Lancet 1988; 1: 1012-15. Altmann P, Hamon C, Blair J, Dhanesha U, Cunningham J, Marsh F. Disturbance of cerebral function by aluminium in haemodialysis patients without overt aluminium toxicity. Lancet 1989; 2: 7-12. Anonymous. Aluminium and Alzheimers disease. Lancet 1989; 1: 82-3. Anonymous. Camelford revisited. Lancet 1991; 338: 1200. Arze RS, Parkinson IS, Cartlidge NEF, Britton P, Ward MK. Reversal of aluminium dialysis encephalopathy after desferrioxamine treatment. Lancet 1981; 2: 1116. Berland Y, Charbit M, Henry JF, Toga M, Cano JP, Olmer M. Aluminium overload of parathyroid glands in haemodialysed patients with hyperparathyroidism: effect on bone remodelling. Nephrol Dial Tansplant 1988; 3: 417-22. Bilkei-Gorzó A. Neurotoxic effect of enteral aluminium. Food Chem Toxicol 1993; 31: 357-61. Birchall JD. Dissolved silica and bioavailability of aluminium. Lancet 1993; 342: 299. Birchall JD, Chappell JS. Aluminium, chemical physiology, and Alzheimer's disease. Lancet 1988; 2: 1008-10. Bishop N, McGraw M, Ward N. Aluminium in infant formulas. Lancet 1989;1: 490-1. Boelaert JR, Fenves AZ, Coburn JW. Deferoxamine therapy and mucormycosis in dialysis patients: report of an international registry. Am J Kidney Dis 1991; 18: 660-7. Bolla KI, Briefel G, Spector D, Schwartz BS, Wieler L, Herron J, Gimenez L. Neurocognitive effects of aluminum. Arch Neurol 1992; 49: 1021-6. Botella J, Gallego JL, Fernadez-Fernandez J, Sanz-Guajardo D, de-Miguel A, Ramos J, Franco P, Enriques R, Sanz-Moreno C. The bone scan in patients with aluminium-associated bone disease. Proc Eur Dial Transplant Assoc Eur Ren Assoc 1984; 21: 403-9. Boyce N, Wood C, Holdsworth S, Thomson NM, Atkins RC. Life-threatening sepsis complicating heavy metal chelation therapy with desferrioxamine. Aust N Z J Med 1985; 15: 654-5. Brown MA, George CRP, Dunstan CR, Kalowski S, Corrigan AB. Prurigo nodularis and aluminium overload in maintenance haemodialysis. Lancet 1992; 340: 48. Burks JS, Alfrey AC, Huddlestone J, Norenberg MD, Lewin E. A fatal encephalopathy in chronic haemodialysis patients. Lancet 1976; 1: 764-8. Canavese C, Pacitti A, Segoloni G, Thea A, D'Amicone M, Stratta P, Rossi P, Sabbioni E, Pietra R, Constantini S, Giordano R, Ballanti P, Mocetti P, Luda E, Bugiani O. Clinical and histological results of long-term management of aluminium overloading in uraemic patients with desferrioxamine. J Trace Elem Electrolytes Health Dis 1989; 3: 17-23. Caramelo CA, Cannata JB, Rodeles MR, Martín JLF, Mosquera JR, Monzú B, Outeiriño J, Blum G, Andrea C, Farré AJL, Acuña G, Casado S, Hernando L. Mechanisms of aluminum-induced microcytosis: lessons from accidental aluminum intoxication. Kidney Int 1995; 47: 164-8. Chan MK, Varghese Z, Li MK, Wong WS, Li CS. Newcastle bone disease in Hong Kong: a study of aluminium associated osteomalacia. Int J Artif Organs 1990; 13: 162-8. Chang TMS, Barre P. Effect of desferrioxamine on removal of aluminium and iron by coated charcoal haemoperfusion and haemodialysis. Lancet 1983; 2: 1051-3. Charhon SA, Chavassieux P, Chapuy MC, Accominotti M, Traeger J, Meunier PJ. Traitement par la desferrioxamine de l'ostéomalacie par intoxication à l'aluminium. [Desferrioxamine treatment of aluminium-induced osteomalacia]. Presse Med 1986; 15: 55-9. Crapper DR, Krishnan SS, Dalton AJ. Brain aluminum distribution in Alzheimer's disease and experimental neurofibrillary degeneration. Science 1973; 180: 511-3. Crapper McLachlan DR. Aluminum and Alzheimer's disease. Response. Can Med Assoc J 1994; 151: 268-9. Crapper McLachlan DR, Dalton AJ, Kruck TPA, Bell MY, Smith WL, Kalow W, Andrews DF. Intramuscular desferrioxamine in patients with Alzheimer's disease. Lancet 1991; 337: 1304-8. Crapper McLachlan DR, Smith WL, Kruck TP. Desferrioxamine and Alzheimer's disease: video home behavior assessment of clinical course and measures of brain aluminum. Ther Drug Monit 1993; 15: 602-7. CSDS/Chemical Safety Data Sheets. Vol 2. Cambridge: Royal Society of Chemistry, 1989. Cumming AD, Simpson G, Bell D, Cowie J, Winney RJ. Acute aluminium intoxication in patients on continuous ambulatory peritoneal dialysis. Lancet 1982; 1: 103-4. Davenport A, Ahmad R. Grand mal fitting in a patient on regular haemodialysis treatment and receiving desferrioxamine therapy. Scand J Urol Nephrol 1988; 22: 159-60. Davison AM, Walker GS, Oli H, Lewins AM. Water supply aluminium concentration, dialysis dementia, and effect of reverse-osmosis water treatment. Lancet 1982; 2: 785-7. Day JP, Ackrill P. The chemistry of desferrioxamine chelation for aluminum overload in renal dialysis patients. Ther Drug Monit 1993; 15: 598-601. D'Haese PC, Clement JP, Elseviers MM, Lamberts LV, Van de Vyver FL, Visser WJ, De Broe ME. Value of serum aluminium monitoring in dialysis patients: a multicentre study. Nephrol Dial Transplant 1990; 5: 45-53. Domingo JL. The use of chelating agents in the treatment of aluminum overload. Clin Toxicol 1989; 27: 355-67. Domingo JL, Gómez M, Llobet JM, Corbella J. Comparative effects of several chelating agents on the toxicity, distribution and excretion of aluminium. Hum Toxicol 1988; 7: 259-62. DOSE/Dictionary of substances and their effects. Vol 1. Cambridge: Royal Society of Chemistry, 1992. Douthat WG, Acuña Aguerre G, Fernández Martín JL, Mouzo R, Cannata Andía JB. Treatment of aluminium intoxication: a new scheme for desferrioxamine administration. Nephrol Dial Transplant 1994; 9: 1431-4. Eastwood JB, Levin GE, Pazianas M, Taylor AP, Denton J, Freemont AJ. Aluminium deposition in bone after contamination of drinking water supply. Lancet 1990; 336: 462-4. Ebrahim S. Aluminium and Alzheimer's disease. Lancet: 1989; 1: 267. Elliott HL, Dryburgh F, Fells GS, Sabet S, MacDougall AI. Aluminium toxicity during regular haemodialysis. Br Med J 1978a; 1: 1101-3. Elliott HL, MacDougall AI, Fell GS. Aluminium toxicity syndrome. Lancet 1978b; 1: 1203. Elo R, Uksila E. Influence of aluminium sulphate on lungs in industry. In: VIIIth World Congress on the prevention of Occupational Accidents and Diseases, Bucharest, May 17-21 1977: Ministry of Labour of Romania in Co-operation with the International Social Security Association (ISSA) and the International Labour Office (ILO), 1977; 316-8. Erasmus RT, Savory J, Wills MR, Herman MM. Aluminum neurotoxicity in experimental animals. Ther Drug Monit 1993; 15: 588-92. Exley C, Birchall JD. The cellular toxicity of aluminium. J Theor Biol 1992; 159: 83-98. Fahal IH, Yaqoob M, Williams PS, Ahmad R, Roberts NB, Bell GM. Does silicon protect against aluminium toxicity in dialysis patients. Lancet 1994; 343: 122-3. Firling CE, Severson AR, Hill TA. Aluminum effects on blood chemistry and long bone development in the chick embryo. Arch Toxicol 1994; 68: 541-7. Freundlich M, Zilleruelo G, Abitbol C, Strauss J, Faugere MC, Malluche HH. Infant formula as a cause of aluminium toxicity in neonatal uraemia. Lancet 1985; 2: 527-9. Garrett P, McWade M, O'Callaghan J. Radiological assessment of aluminium-related bone disease. Clin Radiol 1986; 37: 63-70. Garrett PJ, Mulcahy D, Carmody M, O'Dwyer WF. Aluminium encephalopathy: Clinical and Immunological features. Q J Med 1988; 69: 775-83. Gilli P, Malacarne F, Fagioli F. Is serum aluminium monitoring useful in evaluating aluminium intoxication? Lancet 1983; 1: 656. Gitelman HJ, Alderman FR, Kurs-Lasky M, Rockette HE. Serum and urinary aluminium levels of workers in the aluminium industry. Ann Occup Hyg 1995; 39: 181-91. Golding J, Rowland A, Greenwood R, Lunt P. Aluminium sulphate in water in north Cornwall and outcome of pregnancy. Br Med J 1991; 302: 1175-7. Gomez M, Bosque MA, Domingo JL, Llobet JM. Evaluation of the maternal and developmental toxicity of aluminum from high doses of aluminum hydroxide in rats. Vet Hum Toxicol 1990; 32: 545-8. Gómez M, Domingo JL, del Castillo D, Llobet JM, Corbella J. Comparative aluminium mobilizing actions of several chelators in aluminium-loaded uraemic rats. Hum Exp Toxicol 1994; 13: 135-9. Good PF, Perl DP, Bierer LM, Schmeidler J. Selective accumulation of aluminum and iron in the neurofibrillary tangles of Alzheimer's disease: a laser microprobe (LAMMA) study. Ann Neurol 1992; 31: 286-92. Goodill JJ, Abuelo JG. Mucormycosis - a new risk of deferoxamine therapy in dialysis patients with aluminum or iron overload? N Engl J Med 1987; 31: 54. Gosselin RE, Smith RP, Hodge HC. Clinical Toxicology of commercial products. 5th Ed. Baltimore: Williams and Wilkins, 1984. Goyer RA, Epstein S, Bhattacharyya M, Korach KS, Pounds J. Environmental risk factors for osteoporosis. Environ Health Perspect 1994; 102: 390-4. Harrington CR, Wischik CM, McArthur FK, Taylor GA, Edwardson JA, Candy JM. Alzheimer's-disease-like changes in tau protein processing: association with aluminium accumulation in brains of renal dialysis patients. Lancet 1994; 343: 993-7. Haug A, Shi B, Vitorello V. Aluminum interaction with phosphoinositide-associated signal transduction. Arch Toxicol 1994; 68: 1-7. Hazardtext. In: Tomes plus. Environmental Health and Safety Series I. Vol 26. Colorado: Micromedex, Inc., 1995. Health and Safety Executive. EH40/95. Occupational exposure limits 1995. Sudbury: Health and Safety Executive, 1995. Hodge KC, Day JP, O'Hara M, Ackrill P, Ralston AJ. Critical concentrations of aluminium in water used for dialysis. Lancet 1981; 2: 802-3. HSDB/Hazardous Substances Data Bank. In: Tomes plus. Environmental Health and Safety Series I. Vol 26. National Library of Medicine, 1995. Joshi JG, Dhar M, Clauberg M, Chauthaiwale V. Iron and aluminum homeostasis in neural disorders. Environ Health Perspect 1994; 102: 207-13. Klatzo I, Wisniewski H, Streicher E. Experimental production of neurofibrillary degeneration. I. Light microscopic observations. J Neuropathol Exp Neurol 1965; 24: 187-99. Kontoghiorghes GJ, Barr J, Baillod RA. Studies of aluminium mobilization in renal dialysis patients using the oral chelator 1,2-dimethyl-3-hydroxypyrid-4-one. Arzneimittelforschung 1994; 44: 522-6. Leonard A, Gerber GB. Mutagenicity, carcinogenicity and teratogenicity of aluminium. Mutat Res 1988; 196: 247-57. Lowermoor Incident Health Advisory Group; Department of Health. Water pollution at Lowermoor, North Cornwall. 2nd report. London: HMSO, 1991; 1-51. Main J, Ward MK. Potentiation of aluminium absorption by effervescent analgesic tablets in a haemodialysis patient. Br Med J 1992; 304: 1686. Martyn CN, Osmond C, Edwardson JA, Barker DJP, Harris EC, Lacey RF. Geographical relation between Alzheimer's disease and aluminium in drinking water. Lancet 1989; 1: 59-62. McCarthy JT, Milliner DS, Johnson WJ. Clinical experience with desferrioxamine in dialysis patients with aluminium toxicity. Q J Med 1990; 275: 257-76. McCauley J, Sorkin MI. Exacerbation of aluminium encephalopathy after treatment with desferrioxamine. Nephrol Dial Transplant 1989; 4: 110-4. McDermott JR, Smith AI, Ward MK, Parkinson IS, Kerr DNS. Brain-aluminium concentration in dialysis encephalopathy. Lancet 1978; 1: 901-4. McLaughlin RS. Chemical burns of the human cornea. Am J Ophthalmol 1946; 29: 1355-62. McMillan TM, Dunn G, Colwill SJ. Psychological testing on schoolchildren before and after pollution of drinking water in North Cornwall. J Child Psychol Psychiatry 1993a; 34: 1449-59. McMillan TM, Freemont AJ, Herxheimer A, Denton J, Taylor AP, Pazianas M, Cummin ARC, Eastwood JB. Camelford water poisoning accident: serial neuropsychological assessments and further observations on bone aluminium. Hum Exp Toxicol 1993b; 12: 37-42. Meditext. In: Tomes plus. Environmental Health and Safety Series I. Vol 26. Colorado: Micromedex, Inc., 1995. Milliner DS, Nebeker HG, Ott SM, Andress DL, Sherrard DJ, Alfrey AC, Slatopolsky EA, Coburn JW. Use of the deferoxamine infusion test in the diagnosis of aluminum- related osteodystrophy. Ann Internal Med 1984; 101: 775-80. Mion C. Aluminium in continous ambulatory peritoneal dialysis and post dilutional hemofiltration. Clin Nephrol 1985; 24: 88-93. Mousson C, Charhon SA, Ammar M, Accominotti M, Rifle G. Aluminium bone deposits in normal renal function patients after long-term treatment by plasma exchange. Int J Artif Organs 1989; 12: 664-7. Munoz DG. Aluminum and Alzheimer's disease. Can Med Assoc J 1994; 151: 268. Murray JC, Tanner CM, Sprague SM. Aluminum neurotoxicity: a re-evaluation. Clin Neuropharmacol 1991; 14: 179-85. Neri LC, Hewitt D. Alumiunium, Alzheimer's disease, and drinking water. Lancet 1991; 338: 390. O'Brien AAJ, McParland C, Keogh JAB. The use of intravenous and intraperitoneal desferrioxamine in aluminium osteomalacia. Nephrol Dial Transplant 1987; 2: 117-9. Parkinson IS, Ward MK, Feest TG, Fawcett RWP, Kerr DNS. Fracturing dialysis osteodystrophy and dialysis encephalopathy. An epidemiological survey. Lancet 1979; 1: 406-9. Payton CD, Junor BJR, Fell GS. Successful treatment of aluminium encephalopathy by intraperitoneal desferrioxamine. Lancet 1984; 3: 1132-3. Petit TL. Issues in aluminum neurotoxicology. Comments Toxicol 1989; 3: 225-38. Pierre F, Baruthio F, Diebold F, Biette P. Effect of different exposure compounds on urinary kinetics of aluminium and fluoride in industrially exposed workers. Occup Environ Med 1995; 52: 396-403. Praga M, Andrés A, de la Serna J, Ruilope LM, Nieto J, Estenoz J, Millet VG, Arnaiz F, Rodicio JL. Improvement of anaemia with desferrioxamine in haemodialysis patients. Nephrol Dial Transplant 1987; 2: 243-7. Ravelli M, Scaroni P, Mombelloni S, Movilli E, Feller P, Apostoli P, De-Maria G, Valotti C, Sciuto G, Maiorca R. Acute visual disorders in patients on regular dialysis given desferrioxamine as a test. Nephrol Dial Transplant 1990; 5: 945-9. Registration Committee of the European Dialysis and Transplant Association. Dialysis dementia in Europe. Lancet 1980; 2: 190-2. Royal Society of Chemistry. Aluminium sulphate. In: Allen R, ed. Chemical safety data sheets. Volume 2: main group metals and their compounds. Cambridge: Royal Society of Chemistry, 1989; 34-6. RTECS/Registry of Toxic Effects of Chemical Substances. In: Tomes plus. Environmental Health and Safety Series I. Vol 26. National Institute for Occupational Safety and Health (NIOSH), 1995. Salusky IB, Foley J, Nelson P, Goodman WG. Aluminum accumulation during treatment with aluminum hydroxide and dialysis in children and young adults with chronic renal disease. N Engl J Med 1991; 324: 527-31. Sedman AB, Klein GL, Merritt RJ, Miller NL, Weber KO, Gill WL, Anand H, Alfrey AC. Evidence of aluminum loading in infants receiving intravenous therapy. N Engl J Med 1985; 312: 1337-43. Seyfert UT, Didion K, Alvert FW. Aluminium -related bone disease and haemodialysis. Lancet 1987; 2: 456. Sjögren B, Ljunggren KG, Almkvist O, Frech W, Basun H. Aluminosis and dementia. Lancet 1994; 344: 1154. Spira L. The clinical aspect of chronic poisoning by aluminium and its alloys. London: John Bale, Sons and Danielsson Ltd, 1933. Strong MJ. Aluminum neurotoxicity: an experimental approach to the induction of neurofilamentous inclusions. J Neurol Sci 1994; 124: 20-6. Sulkova S, Laurincova Z, Valek A. Haemofiltration or haemodialysis in aluminium elimination? Nephrol Dial Transplant 1991; 6: 3-5. Tohani VK, McCann R, Fox M, Fulton R. Skin rash associated with accidental addition of excess aluminium sulphate to the water supply. Ulster Med J 1991; 60: 108-10. Tsou VM, Young RM, Hart MH, Vanderhoof JA. Elevated plasma aluminum levels in normal infants receiving antacids containing aluminum. Pediatrics 1991; 87: 148-51. van der Voet GB and de Wolff FA. Distribution of aluminium between plasma and erythrocytes. Human Toxicol 1985; 4: 643-8. Visser WJ, Van de Vyver FL. Aluminium-induced osteomalacia in severe chronic renal failure (SCRF). Clin Nephrol 1985; 24: 30-6. Vlasveld LT, van Asbeck BS. Treatment with deferoxamine: a real risk factor for mucormycosis? Nephron 1991; 57: 487-8. Ward MK, Feest TG, Ellis HA, Parkinson IS, Kerr DNS, Herrington J, Goode GL. Osteomalacic dialysis osteodystrophy: evidence for a water-borne aetiological agent, probably aluminium. Lancet 1978; 1: 841-5. Wilhelm M, Passlick J, Busch T, Szydlik M, Ohnesorge FK. Scalp hair as an indicator of aluminium exposure: comparison to bone and plasma. Human Toxicol 1989; 8: 5-9. Windus DW, Stokes TJ, Julian BA, Fenves AZ. Fatal Rhizopus infections in hemodialysis patients receiving deferoxamine. Ann Intern Med 1987; 107: 678-80. Winship KA. Toxicity of aluminium: a historical review, Part 1. Adverse Drug React Toxicol Rev 1992; 11: 123-41. Winship KA. Toxicity of aluminium: a historical review, Part 2. Adverse Drug React Toxicol Rev 1993; 12: 177-211. Wood DJ, Cooper C, Stevens J, Edwardson J. Bone mass and dementia in hip fracture patients from areas with different aluminium concentrations in water supplies. Age Ageing 1988; 17: 415-9. World Health Organisation. Guidelines for drinking-water quality. Vol 1. 2nd ed. Geneva: WHO, 1993. Yaqoob M, Ahmad R, Roberts N, Helliwell T. Low-dose desferrioxamine test for the diagnosis of aluminium-related bone disease in patients on regular haemodialysis. Nephrol Dial Transplant 1991; 6: 484-6. Yokel RA. Aluminum chelation: chemistry, clinical, and experimental studies and the search for alternatives to desferrioxamine. J Toxicol Environ Health 1994; 41: 131-74.