UKPID MONOGRAPH COPPER 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. COPPER Toxbase summary Separate toxbase entries exists for: Copper carbonate Copper chloride Copper oxide Copper sulphate Type of product A transition metal used in electrical conductors, alloys (notably brass and bronze), cooking utensils, coins, corrosion resistant plumbing pipes, heating and building materials. Toxicity Copper toxicity primarily occurs following leaching of copper ions from copper pipes. No fatalities have been reported among otherwise healthy individuals following ingestion of water contaminated in this way. However, renal failure patients have died following copper intoxication via parenteral exposure to copper contaminated dialysis water. Childhood cirrhosis is a multifactorial disease which may become manifest by excess dietary copper intake (Mühlendahl and Lange, 1994). Inhalation of copper fumes may cause metal fume fever. Features Dermal - Molten copper will burn. - Leaching of copper from pipes in areas with acidic water has caused green hair, particularly in fair-haired individuals. - Copper contact dermatitis is recognized. Ocular - Copper foreign bodies can cause serious eye damage. - Copper deposited in the anterior vitreous may gradually dissolve over years causing green/brown discolouration of the lens, cornea and iris with impaired visual acuity. - Retinal haemorrhage and abscess formation may ensue if particles reach the posterior vitreous. - Open angle glaucoma is a rare complication of intraocular copper dissemination. Ingestion - Ingestion of copper-contaminated water causes nausea, vomiting, abdominal pain and diarrhoea. - Gastrointestinal mucosal burns with subsequent stricture formation has followed molten copper ingestion. - Fatal habitual copper coin ingestion is reported with extensive hepatic fibrosis at autopsy (Hasan et al, 1995). Inhalation - Copper fumes cause pulmonary tract irritation. Metal fume fever with flu-like symptoms, cough and dyspnoea is also reported, but is far less common than among workers exposed to zinc fume. Injection - Haemodialysis with copper contaminated water has caused "copper fever" with headache, fever, myalgia, nausea, vomiting, abdominal pain and orthostatic hypotension. Intravascular haemolysis, rhabdomyolysis and pancreatitis are recognized. Fatalities have occurred (Klein et al, 1972). Management Dermal 1. Irrigate with copious lukewarm water. 2. Consider the possibility of systemic copper uptake if there has been significant or repeated exposure to broken skin. 3. Copper irritant dermatitis and contact sensitivity are managed most effectively by discontinuing exposure. Ocular 1. Irrigate immediately with lukewarm water or preferably saline for at least 10 minutes. 2. Application of local anaesthetic may be required for pain relief and to overcome blepharospasm to allow thorough decontamination. 3. Ensure no particles remain lodged in the conjunctival recesses. 4. Corneal damage may be detected by the instillation of fluorescein. 5. Specialist ophthalmological advice is indicated if an intraocular copper foreign body is suspected. Ingestion 1. Symptomatic and supportive measures are usually all that are required following ingestion of copper-contaminated water. The World Health Organization guideline value for the copper content of drinking water is 2 mg/L (WHO, 1993). 2. The diagnosis can be confirmed by measurement of the copper concentration of the water supply. 3. Remove the copper source. 4. Check the full blood count and liver profile if chronic exposure is suspected. 5. Management of copper coin ingestion is as for other metal objects with serial X-rays to track transit through the gastrointestinal tract and a surgical opinion if signs or symptoms suggest obstruction. 6. The value of chelation therapy in copper poisoning is unproven. Discussion with an NPIS physician is recommended. Inhalation 1. Remove from exposure and administer supplemental oxygen by face-mask if there is evidence of respiratory distress. 2. Arrange a chest X-ray if there are abnormal findings on respiratory examination. 3. There typically are no long-term sequelae of copper fume fever. Injection 1. Take a sample of haemodialysis water for copper concentration determination. 2. Remove copper source from water supply. 3. Measure the whole blood copper concentration. 4. Treat gastrointestinal features symptomatically, replacing fluid losses as necessary. 5. Monitor biochemical and haematological profiles plus acid-base status. 6. Intravascular haemolysis is managed conventionally. References Bentur T, Koren G, McGuigan M, Spielberg SP. An unusual skin exposure to copper; clinical and pharmacokinetic evaluation. Clin Toxicol 1988; 26: 371-80. Gleason RP. Exposure to copper dust. Am Ind Hyg Assoc J 1968; 29: 461-2. Hasan N, Emery D, Baithun SI, Dodd S. Chronic copper intoxication due to ingestion of coins: a report of an unusual case. Hum Exp Toxicol 1995; 14: 500-2. Knobeloch L, Ziarnik M, Howard J, Theis B, Farmer D, Anderson H, Proctor M. Gastrointestinal upsets associated with ingestion of copper-contaminated water. Environ Health Perspect 1994; 102: 958-61. Klein Jr WJ, Metz EN, Price AR. Acute copper intoxication. A hazard of hemodialysis. Arch Intern Med 1972; 129: 578-82. Mülendahl KE, Lange H. Copper and childhood cirrhosis. Lancet 1994; 344: 1515-6. Shibuya S, Takase Y, Sharma N. Esophageal ulcer due to ingestion of melted copper. Dig Dis Sci 1992; 37: 1785-90. Spitalny KC, Brondum J, Vogt RL, Sargent HE, Kappel S. Drinking-water-induced copper intoxication in a Vermont family. Pediatrics 1984; 74: 1103-6. Terry RF. Excess copper in a local water supply. Med J Aust 1996; 165: 296. WHO/World Health Organization. Guidelines for drinking-water quality. 2nd ed. Vol 1. Recommendations. Geneva: World Health Organization, 1993. Wyllie J. Copper poisoning at a cocktail party. Am J Public Health 1957; 47: 617. Substance name Copper Origin of substance Copper occurs in pheophytin (an analogue of chlorophyll), haemocyanin, tyrosinase and caeruloplasmin. It may also be found in various ores such as cuprite, azurmalachite, malachite, tetrahedrite, chalcopyrite, covellite, azurite, bornite, antlerite, chalcocite and enargite (HAZARDTEXT, 1997; DOSE, 1993). Synonyms Allbri natural copper Anac 110 Arwood copper Bronze powder Cuprum Gold Bronze Kafar copper Raney copper (HAZARDTEXT, 1997) Chemical group A group 1B transition metal (d block) element. Reference numbers CAS 7440-50-8 (DOSE, 1993) RTECS GL5325000 (RTECS, 1997) UN 2811 (REPROTEXT, 1997) HAZCHEM NIF Physicochemical properties Chemical structure Cu Molecular weight 63.546 (HSDB, 1997) Physical state at room temperature Solid (HSDB, 1997) Colour Red-brown (MEDITEXT, 1997) Odour Odourless (HSDB, 1997) Viscosity NIF pH NIF Solubility Copper is slowly soluble in ammonia water, hot sulphuric acid, nitric acid and is very slightly soluble in hydrochloric acid and ammonium hydroxide. (HSDB, 1997; MEDITEXT, 1997) Autoignition temperature NIF Chemical interactions Zinc, aluminium or iron may cause metallic copper to precipitate from solution. Following explosion of copper azide, metallic copper and nitrogen are formed. Copper is attacked at a very slow rate by dilute sulphuric acid or cold hydrochloric acid when exposed to the atmosphere. It is rapidly attacked by acetic and other organic acids, hydrogen bromide, hot concentrated sulphuric acid, alkalies and dilute nitric acid. A brownish-red precipitate of copper ferrocyanide is produced in the presence of potassium ferrocyanide. Copper is incompatible with 1-bromo-2-propyne. Copper and brasses (down to 60 per cent copper) react readily in the presence of wet acetylene and ammonia to form explosive acetylides. When acetylene comes in contact with copper that has been heated to form a tarnish of copper oxide unstable acetylides are formed. In the presence of copper, ethylene oxide, ammonium nitrate, lead azide, acetylenic compounds and 3-bromopropyne form potentially explosive reactions. Copper forms an incandescent reaction with potassium dioxide. (MEDITEXT, 1997) Major products of combustion NIF Explosive limits NA Flammability Copper powder may ignite. (NIOSH, 1997) Boiling point 2595°C (DOSE, 1993) Density 8.94 at 20°C (DOSE, 1993) Vapour pressure 2666.4 at 1970°C (OHM/TADS, 1997) Relative vapour density NIF Flash point NA Reactivity Upon exposure to moist air, copper gradually develops a coating of green basic carbonate. On contact with water, liquid copper explodes. Light friction, heat and percussion may cause combinations of finely divided copper and bromates of magnesium, sodium, zinc, potassium, calcium and barium to explode. A readily explosive peroxide may form upon long standing. The reacting mixture of copper, platinum, iron, phosphorus or nickel may become incandescent when heated. (MEDITEXT, 1997) Uses In electrical conductors such as wire and switches. In applications where high electrical and thermal conductivity are needed. Copper whiskers are used in thermal and electrical composites. In alloys such as bronze and brasses. Other copper alloys include Monel metal and beryllium-copper. In electroplated coatings and undercoatings for products made from nickel, chromium, zinc, and in cooking utensils; also in corrosion-resistant plumbing pipes, heating, roofing and building construction materials. In industrial machinery and automobiles. In agricultural applications; copper and particularly copper sulphate are used in insecticides, fungicides, herbicides, and algicides. In intrauterine contraceptive devices. Miscellaneous uses including chemical and pharmaceutical applications, as a pollution control catalyst, in pigments, dyes, and anti-fouling paints, in works of art, coinage, fabrics, textiles, glass, ceramics, cement, nylon, paper products, printing, photocopying, pyrotechnics and wood preservatives, also in ammunition, flameproofing, fuel additives and as insulation for liquid fuels. (HAZARDTEXT, 1997) Hazard/risk classification NIF INTRODUCTION Copper plays an important role as a co-factor in several metalloproteins, including cytochrome oxidase and superoxide dismutase and is essential for the utilization of iron and haemoglobin formation. The daily copper intake among the general population is 1-2 mg/day with over 90 per cent in food (IPCS, 1996a). The richest food sources of copper are shellfish, 'organ' meats, seeds, nuts and grains where it is bound to specific proteins. Copper tends to exist in the cupric Cu(II) state in biological systems including water although it may also be found as Cu(I) (Linder and Hazegh-Azam, 1996). Menkes disease (Kinky hair syndrome) is an X-linked inherited copper deficiency which manifests in the first six months of life with poor growth, severe learning difficulties and hair abnormality. Copper deficiency may be seen also as "Swayback" in lambs and calves born to sheep and cows grazing on copper deficient pastures. Wilson's disease is an inborn error of metabolism inherited as an autosomal recessive trait whereby there is reduced biliary copper excretion associated with decreased or absent circulating caeruloplasmin (Schilsky, 1996). The disease is characterized by excessive accumulation of copper in the liver, brain, kidneys and cornea. Basal ganglia degeneration and cirrhosis are the principle complications. Most of the literature regarding copper poisoning relates to acute ingestion or chronic inhalation of copper salts, notably copper sulphate. Elemental copper may be a source of toxicity when leached from copper piping into water supplies or inhaled occupationally as dust or fumes (almost invariably with copper oxides). Copper fragments may cause severe penetrating eye injury. Separate UKPID monographs are available for the following: Copper carbonate Copper chloride Copper oxide Copper sulphate EPIDEMIOLOGY Significant copper contamination of domestic water supplies is rare and only occurs where water is soft and acidic. A "minor epidemic" of green hair occurred among college students following the introduction of fluoride to the town water supply (Cooper and Goodman, 1975). Acidic water flowing through copper pipes plus frequent hair washing were likely aetiological factors. Several cases of severe copper poisoning have occurred among patients undergoing haemodialysis with copper contaminated water (Klein et al, 1972; Lyle et al, 1976) although copper components are no longer permitted in haemodialysis systems. Metal fume fever following occupational inhalation of copper fume is recognized (Cohen, 1979) but not widely reported. It is a considerably less significant problem than among zinc workers. MECHANISM OF TOXICITY Cellular damage and cell death may result from excess copper accumulation. This is likely when copper-metallothionein binding and copper clearance from the cell are blocked. Metallothionein is a cysteine rich, low molecular weight (6500 Da) metal-binding protein which is important in heavy metal detoxification, metal ion storage, and in the regulation of normal cellular Cu(II) (and Zn(II)) metabolism. It is also thought to be a free radical scavenger, playing a protective role in oxidative stress. Metallothionein is found in both intra- and extracellular compartments. It is known to bind zinc, cadmium, copper, mercury and silver (in increasing order of affinity) and its gene transcription is greatly enhanced upon exposure of cells to these metals. High metallothionein concentrations are also induced in the liver by physical and chemical stress, infection and glucocorticoids. It is proposed that free Cu(I) binds to intracellular sulphydryl groups and inactivates enzymes such as glucose-6-phosphate dehydrogenase and glutathione reductase (Dash, 1989). In addition copper may interact with oxygen species (e.g. superoxide anions and hydrogen peroxide) and catalyze the production of reactive toxic hydroxyl radicals. Copper(II) ions can oxidize haem iron to form methaemoglobin. TOXICOKINETICS Available toxicokinetic data are derived from studies using water soluble divalent copper salts (usually the acetate, chloride or sulphate). Absorption and distribution Strickland et al (1972) suggested a mean copper absorption of 57 per cent (range 40 to 70 per cent) following oral administration of 0.4 - 4.5 mg copper (as copper acetate) to four volunteers. An early human study suggested a maximum blood copper concentration was reached some two hours after oral copper chloride administration (1.5 - 12 mg copper) (Earl et al, 1954). Copper transport across the intestinal mucosa following ingestion is facilitated by cytosolic metallothionein. In blood, copper is initially albumin-bound and transported via the hepatic portal circulation to the liver where it is incorporated into caeruloplasmin (an alpha globulin synthesized in hepatic microsomes) (Britton, 1996). Ninety-eight per cent of copper in the systemic circulation is caeruloplasmin-bound. Copper is distributed to all tissues with the highest concentrations in liver, heart, brain, kidneys and muscle. Intracellular copper is predominantly metallothionein-bound. Kurisaki et al (1988) reported copper in the lungs, liver, kidney, blood, bile and stomach (33.7, 35.1, 41.4, 13.8, 2.8, and 2988 µg/g wet weight respectively) following ingestion of some 10 g copper sulphate in a 58 year-old male. Although copper in the liver and kidneys was metallothionein bound, pulmonary copper was not, probably because copper had entered the lung via aspiration. Copper ions can penetrate the erythrocyte membrane. Following acute copper sulphate ingestion this occurs quite rapidly as indicated by the markedly higher whole blood than serum copper concentration within the first few hours of poisoning (Singh and Singh, 1968). In a series of 40 cases of acute copper sulphate ingestion Singh and Singh (1968) noted that haemolysis (secondary to erythrocyte copper uptake) occurred typically 12-24 hours post poisoning, suggesting that red cell copper accumulation is maximal around this time. Studies among vineyard sprayers provide evidence of haematogenous dissemination of inhaled copper (II) ions (Villar, 1974; Pimentel and Menezes, 1977). Copper ions can also be absorbed through the skin giving rise to systemic effects (Holtzman et al, 1966; Pande and Gupta, 1969). Copper can cross the placenta. Excretion Caeruloplasmin renders free copper innocuous with subsequent excretion via a lysosome-to-bile pathway. This process is essential to normal copper homeostasis and provides a protective mechanism in acute copper poisoning. An impaired or overloaded biliary copper excretion system results in hepatic copper accumulation, as occurs in patients with Wilson's disease and in copper poisoning. Renal copper elimination is normally low (Tauxe et al (1966) retrieved less than one per cent of an injected copper acetate dose in the urine over 72 hours) but will increase in acute copper poisoning. For example, a child who ingested 3 grams copper sulphate had increased urine copper concentrations (maximum 3.0 mg/L) for three weeks post poisoning (Walsh et al, 1977). In a series of 40 cases of acute copper sulphate ingestion increased whole blood copper concentrations were noted up to ten days post poisoning with values returning to normal over 17 hours to seven days (Singh and Singh, 1968). The whole-body half-life of copper has been estimated as approximately four weeks (Strickland et al, 1972). CLINICAL FEATURES: ACUTE EXPOSURE Dermal exposure Elemental copper is not acutely toxic to intact skin. Numerous shrapnel fragments penetrated the skin of a 55 year-old chemist following the explosion of a copper azide solution, yielding metallic copper and nitrogen (Bentur et al, 1988). Skull and hand X-rays revealed the presence of multiple soft tissue foreign bodies believed to be copper-covered glass, metallic copper and copper azide debris. The patient was otherwise asymptomatic with normal full blood count, kidney and liver function tests. Skin copper absorption was estimated as 7.7 mg. Serum copper concentration 12 hours post exposure was 0.83 mg/L, peaking at 1.24 mg/L four days post exposure (normal 0.7-1.4 mg/L). The copper serum half life was estimated at 167.4 days. Chelation therapy was not administered and the patient remained asymptomatic nine weeks later except for traumatic hand motor dysfunction. Ocular exposure Intraocular copper foreign bodies from, for example, exploding brass cartridges or war-time mines may cause serious injury. The extent of penetration of the particles determines the nature of subsequent damage. The smaller the particle and the further from the retina the copper foreign body is located, the less damage will result (Grant and Schuman, 1993). If copper particles penetrate only as far as the anterior vitreous they may take years to dissolve with gradual dissemination of copper to the lens, cornea and iris producing green-brown discolouration (chalcosis) and impaired visual acuity. Removal of the copper source even years later may allow reversal of these effects (Grant and Schuman, 1993). Copper particles which penetrate as far as the retina may lead to haemorrhage, abscess formation and connective tissue encapsulation. The vitreous body is destroyed and retinal detachment may ensue. Subsequently copper may escape its encapsulation and continue to do damage. Oxidation products of metallic copper are believed to mediate these effects (Grant and Schuman, 1993). Open angle glaucoma is a rare complication of intraocular copper dissemination from foreign bodies (Grant and Schuman, 1993). Ingestion Gastrointestinal toxicity Abdominal pain, diarrhoea, nausea and vomiting have been reported following ingestion of copper contaminated water (Spitalny et al, 1984; Knobeloch et al, 1994). Contamination of water supplies may occur when water remains stagnant in a copper main or following corrosion of copper plumbing materials. Leaching of copper from plumbing materials is only significant when the water is extremely soft and acidic. Copper contaminated water is usually first detected by blue-green staining of laundry, sinks and tubs (at copper concentrations above 1 mg/L) or an unpleasant smell and bitter taste (at copper concentrations greater than 5 mg/L) (IPCS, 1996a). The International Programme on Chemical Safety has suggested copper(II) ion concentrations of some 30 mg/L are typically required before acute gastrointestinal upset ensues although this "may vary with the binding and chemical form of copper present" (IPCS, 1996b). There are several reports where adverse gastrointestinal effects have followed the consumption of water containing less than 10 mg/L copper. In a study in Wisconsin during 1992-3 (Knobeloch et al, 1994), five separate cases of gastrointestinal upset caused by copper-contaminated drinking water were investigated. Children were thought to be more susceptible to the effects of ingested copper in these circumstances possibly due to a higher copper exposure in relation to body weight or perhaps a greater sensitivity to the irritant effects of copper. The authors concluded that water copper concentrations greater than 1.3 mg/L (the "federal action limit" in Wisconsin) commonly caused mild gastrointestinal features (Knobeloch et al, 1994). Three members of a family experienced recurrent abdominal pain and vomiting between five and 20 minutes after ingestion of copper contaminated water in beverages or with food over a period of some 12 months. Analysis of a sample of the water source revealed a copper concentration of 7.8 mg/L. Hair and nail copper concentrations of one of the children were elevated (1200 µg/g and 100 µg/g respectively; normal range 11-53 µg/g). All symptoms resolved when the family ceased drinking the water (Spitalny et al, 1984). Persistent vomiting and diarrhoea resulting in weight loss and dehydration in a six week-old female were attributed to copper contaminated drinking water used to dilute her milk (Knobeloch et al, 1994). Analysis of water supplies from her home revealed copper concentrations of 0.16-7.8 mg/L. Simultaneous elevated methaemoglobin concentrations were attributed to nitrate contamination of the same water supply. Her symptoms resolved when bottled water was used. Similar symptoms were reported in a group of 15 nurses who experienced nausea, vomiting, abdominal pain and diarrhoea within an hour of ingesting alcohol which had been refrigerated in a copper-lined flask at a cocktail party (Wyllie, 1957). Copper intake was estimated to vary between 5.3 and 32 mg. Oesophageal and stomach ulcers in association with epigastric pain were described in a 49 year-old foundry worker one week following the accidental ingestion of a small amount of melted copper (Shibuya et al, 1992). Ulcer scars were noted on the tongue and several teeth were "completely burnt out". Forty days post ingestion the patient could swallow only liquids due to ulcer-induced oesophageal stricture confirmed by barium meal. Computed tomography showed circular oesophageal thickening progressing to occlusion. These injuries necessitated total gastrectomy and thoracic oesophagus removal some four months later. Histological findings showed thickened and densely fibrosed oesophageal and gastric walls with deep ulceration extending to muscle. The authors concluded that the copper primarily burned the oesophageal mucosa and was cooled by the gastric juice at the lesser curvature of the stomach. Terry (1996) noted elevated serum copper and caeruloplasmin concentrations in two females who had consumed copper-contaminated water (concentration not stated). These abnormalities persisted despite removal of the contaminating source. Further investigation revealed they both were receiving hormone replacement therapy which caused elevated caeruloplasmin, and hence total (but not free), copper concentrations. Pulmonary toxicity Dyspnoea which resolved following three days oxygen therapy was reported in a worker following melted copper ingestion (Shibuya et al, 1992). Neurotoxicity Headaches and dizziness in association with gastrointestinal upset were reported in a group of nurses within an hour of ingesting an alcoholic beverage which had been refrigerated in a copper-lined flask (Wyllie, 1957). Estimated copper intake varied from 5.3-32 mg. Recurrent headaches were also described in association with gastrointestinal features in five adults and four children after drinking copper-contaminated water (Knobeloch et al, 1994). Inhalation Occupational exposure to copper fumes may cause upper respiratory tract irritation and sometimes symptoms of metal fume fever. These complaints are, however, typically encountered in those working in the copper industry for prolonged periods (see Chronic exposure). CLINICAL FEATURES: CHRONIC EXPOSURE Dermal exposure Parish (1975) noted that cases of green hair among copper workers have been reported since 1882. In all cases it generally has been accepted that copper staining was "from without and not within the hair" (Parish, 1975). Cooper and Goodman (1975) reported a "minor epidemic" of green hair in girls from a state college following the introduction of fluoride to the town water supply. Low pH water leaching copper from piping was a possible cause of the discolouration. The authors noted that individuals with blonde hair were affected mainly, although green discolouration may not be as apparent in dark haired individuals. In addition, Goldsmith and Holmes (1975) noted that, independent of any copper effect, artificial hair bleaching may lead to green hair discolouration on exposure to chlorinated water. A hair copper concentration of 466 mg/kg (normal 4-128 mg/kg) was measured in a six year-old boy who developed green hair following repeated bathing in a swimming pool (Lampe et al, 1977). In another case of "green hair" in a five year-old girl (Lampe et al, 1977), water analysis revealed a copper concentration of only 0.9 mg/L. Hair returned to normal following daily alternate washing with acidic and basic shampoos. The source of small amounts of copper in the swimming pool water was likely to be algicide residue or copper leached from pipes. Nordlund et al (1977) described two nursing students who acquired green discolouration to blonde hair following a four to six week stay in a university dormitory where they washed their hair daily. Analysis of one student's hair revealed a copper concentration of 1042 ppm (normal 17-38 ppm) and the copper concentration of the dormitory water source was noted to range from 0.41 to 4 ppm (normal 0.25 ppm). Interference from a faulty electrical connection to the copper water pipes was thought to have resulted in increased amounts of ionized copper in the supply. Despite its widespread use, the sensitizing potential of copper has been described as "extremely low" (Walton, 1983a). In patch tests among 354 eczema patients, six tested positive to copper sulphate (5 per cent solution) and 39 to nickel sulphate (2.5 per cent solution) (Walton, 1983a). All patients positive to copper sulphate were also nickel sulphate positive. None of the subjects positive to copper sulphate were occupationally exposed to copper or had a history of atopy; all were females with hand eczema. The authors postulated nickel- and copper-containing coins as the source of exposure. Interpretation of these results was complicated by the possibility that patients were sensitive to the nickel sulphate trace (0.01 per cent) in the copper sulphate test solution. The author subsequently demonstrated (Walton, 1983b) that the six copper sensitive patients were patch test negative to nickel sulphate (0.01 per cent), suggesting true copper sensitivity. Further evidence of true copper allergy was presented by Van Joost et al (1988) who described two females patch test positive to copper (as sulphate 5 per cent) and nickel (as sulphate 2.5 per cent) in whom the possibility of nickel contamination of the copper test solution was largely excluded by the observation that 11 "control" nickel sensitive patients each gave no positive reaction to the copper solution. Epstein (1955) described combined nickel/copper sensitivity in 38 per cent of 32 patients patch tested, and emphasized that many nickel-containing alloys also contain copper. The author suggested the frequency of cross-sensitivity reactions, the close chemical relationship between copper and nickel (in adjacent positions in the transition metal series of the periodic table) and evidence for a true cross-sensitivity between nickel and cobalt as reasons to assume a true cross-sensitivity between copper and nickel rather than a coincidental occurrence of separate sensitivities (Epstein, 1955). In 30 patients known to be contact sensitive to nickel but patch-test negative to copper, the severity of patch test reaction to a copper/nickel mixture was greater (p<0.001) than to nickel alone, suggesting copper ions somehow enhanced the sensitivity reaction to nickel (Santucci et al, 1993). The authors proposed that the presence of copper ions facilitated the formation of nickel protein complexes in the skin although the precise mechanism remains obscure. In a study by Karlberg et al (1983), 13 of 1190 eczema patients showed a patch test reaction to two per cent copper sulphate. However, these patients had concomitant reactions to other known contact allergens and serial dilution tests with copper sulphate provided no confirmed cases of copper sulphate contact sensitivity. The authors recommended a serial dilution test in cases of suspected copper allergy to eliminate the possibility of an irritant effect and confirm whether true copper sensitivity is present (Karlberg et al, 1983). Chronic, low grade gingivitis was reported in an individual with a copper-containing dental prosthesis. The gingivitis resolved after replacement of the prosthesis with a non copper-containing device (Trachtenberg, 1972). Urticarial hypersensitivity to copper-containing dental cement has also been described (Reid, 1968) An urticarial rash associated with angioedema and joint pain occurred in a 24 year-old woman one month after insertion of a copper-containing intrauterine contraceptive device (IUD) (Barkoff, 1976). She required treatment with adrenaline, systemic steroids and antihistamines. She showed a positive scratch test reaction to copper sulphate (one per cent solution) but was patch test negative to copper. All symptoms resolved when the IUD was removed. In conclusion, available evidence regarding copper contact sensitivity suggests that while a true copper contact allergy exists, cross sensitivity between nickel and copper contributes to many cases. Copper also may cause an irritant dermatitis or a generalized type 1 hypersensitivity response. Keratinization of the hands and soles of the feet have been reported following chronic topical exposure to metallic copper but without reference to original case data (Sittig, 1985). Ingestion Gastrointestinal toxicity A 15 month-old infant who presented with failure to thrive and diarrhoea was found to have a serum copper concentration increased to 2.9 mg/L which was attributed to the consumption of contaminated domestic water (copper concentration 0.8 mg/L) for three months. Resolution of symptoms and substantial weight gain accompanied removal from exposure. The child also received a five week course of oral d-penicillamine 75 mg tds which was associated initially with an increased urine copper concentration (Salmon and Wright, 1971). The water copper concentration reported in this case is well below the provisional guideline value for drinking water of 2 mg/L (WHO, 1993) which is defined as "the concentration of a constituent that does not result in any significant risk to health ........ over a lifetime of consumption" (IPCS, 1996b). This child must have had "abnormal sensitivity to the metal" as the authors emphasized for copper intoxication to occur. Wilson's disease was excluded by serum caeruloplasmin assay and liver biopsy but another defect of copper metabolism is possible. A 46 year-old man who habitually ingested coins for several years presented with a three day history of abdominal distension and vomiting (Hasan et al, 1995). At laparotomy 700 coins were removed from a markedly dilated stomach. The patient "became very unwell" post operatively and died. At autopsy the gastric mucosa was inflamed and oedematous but most pathological abnormalities were noted in the liver (see below). Hepatotoxicity Indian childhood cirrhosis (ICC) is a frequently fatal disease affecting children (mean age 18 months) in rural areas of India and is caused by massive hepatic copper accumulation (Pandit and Bhave, 1983; Pandit and Bhave, 1996). A high dietary copper intake, due to copper leaching into milk from brass cooking vessels, is the most important aetiological factor (Pandit and Bhave, 1983; Pandit and Bhave, 1996). The milk protein casein has been shown to avidly bind copper and serve as an effective metal ion carrier from brass to the infant (O'Neill and Tanner, 1989). Fortunately the previously high incidence of ICC, accounting for 10 per cent paediatric mortality in some Indian hospitals (Pandit and Bhave, 1983), has been reduced dramatically by an effective health education campaign aimed at maximizing breast feeding and avoiding the use of copper-containing cooking utensils (Pandit and Bhave, 1996). ICC is now rare in India. Although a high dietary copper intake is undoubtedly the main cause of ICC, the observed male preponderance and familial occurrence suggests an inherited predisposition (Pandit and Bhave, 1996). Further support for a genetic component in at least some cases of paediatric copper-induced cirrhosis comes from reports of an ICC-like condition among children in Western countries who have had a high, but not massive, dietary copper intake (Mühlendahl and Lange, 1994). A large-scale epidemiological survey in Massachusetts in 1993 concluded that a moderately increased domestic water copper concentration alone does not cause liver disease (Scheinberg and Sternlieb, 1994). In this study none of the 135 deaths occurring between 1969 and 1991 in children under six years-old in three towns with the highest tapwater copper concentration (8.5-8.8 mg/L) of any "medium size" USA town, were attributed to any form of liver disease. In conclusion it appears that idiopathic childhood cirrhosis is a multifactorial disease which requires increased copper ingestion superimposed on an inherited defect of copper metabolism to be manifest fully. A 15 month-old infant fed with copper-contaminated water (copper concentration 0.8 mg/L) for three months developed transiently increased liver enzyme activities in association with features of gastrointestinal and neurological toxicity (Salmon and Wright, 1971). As discussed above this child must have been predisposed to copper toxicity since the water copper concentration was not particularly high. Walker-Smith and Blomfield (1973) described a 14 month-old infant who died six weeks after presenting with clinical and histopathological features of cirrhosis. The child had been bottle fed with copper-contaminated water from an acidic private supply running through domestic copper pipes (water copper concentration 6.75 mg/L). A slightly low plasma caeruloplasmin concentration and raised urinary copper excretion were consistent with Wilson's disease although the acute presentation at such a young age suggested concomitant abnormally high copper exposure was aetiologically significant. Some six litres of ascites were drained during operation and a further litre at autopsy from a 46 year-old male with a history of ingestion of at least 700 coins (Hasan et al, 1995). Extensive liver fibrosis was evident at autopsy. Histological findings included intracanalicular bile plugs, bile ductule proliferation and lymphocyte infiltration. Copper associated protein deposits were identified throughout the hepatic parenchyma. Neurotoxicity Hypotonia, photophobia and "behaviour change" were noted in a young child who presented with failure to thrive and was found to have an increased serum copper concentration (2.9 mg/L) attributed to chronic copper intoxication from a domestic water supply (Salmon and Wright, 1971). Inhalation Although upper respiratory tract irritation and metal fume fever are cited as "common complaints" among copper workers (Cohen, 1979) original case data are scarce. Metal fume fever is associated more typically with zinc oxide inhalation (see Zinc monograph). Occupational exposure to dusts and fumes of copper and copper salts have been reported to cause nasal mucosal congestion and occasionally nasal septum perforation but no original case data have been identified (Scheinberg, 1983). Employees complained of head stuffiness, "common cold" symptoms and "sensations of chills or warmth" a few weeks after commencing copper plate polishing (Gleason, 1968). Analysis of settled dust revealed "major" amounts of copper and "minor" amounts of aluminium; air copper concentrations ranged from 0.03 - 0.12 mg/m3. Following dust control via exhaust ventilation air copper concentrations were reduced to less than 0.008 mg/m3 and symptoms resolved. A seven year study of 494 long term (21 years ± (SEM) 1) copper refinery workers in Canada revealed no increased prevalence of chronic obstructive pulmonary disease or small airways dysfunction (Ostiguy et al, 1995). The plant workers were exposed to dusts of copper, selenium, silver, lead, arsenic and "other trace metals" at concentrations below the threshold limit value (TLV), suggesting TLV enforcement was likely to have prevented the development of pulmonary disorders. The authors concluded low concentration chronic exposure to foundry fumes and metal dusts does not necessarily cause a significant reduction in forced vital capacity and respiratory dysfunction. Oral toxicity Superficial green staining of the teeth was reported in a 21 year-old brass foundry worker after 10 months exposure to fumes containing approximately 75 per cent copper (Donoghue and Ferguson, 1996). The authors suggested the staining was attributed to copper adherence from the brass fume and its subsequent conversion to copper carbonate. Neurotoxicity A recent population based case control study among 144 Parkinsonian patients and 464 controls in Detroit identified a significant (p<0.05) association between Parkinson's disease and more than 20 years occupational copper exposure (Gorell et al, 1997). Chronic manganese exposure was independently significantly associated with this disorder. The neurological hazards of chronic inhalational metal exposure require further clarification. Injection There are several case reports describing copper-intoxication following haemodialysis with copper contaminated water; fatalities have been described. Copper contamination occurs typically when copper piping is used in the dialysate supply system (Lyle et al, 1976). A faulty mains water deionizer which feeds acidic water to the dialysate-making machine is an important cause of increased copper leaching (Manzler and Schreiner, 1970). "Copper fever" typically presents as an acute illness although failure to identify the cause frequently leads to acute-on-chronic (Manzler and Schreiner, 1970) or chronic (Lyle et al, 1976) copper exposure. Copper components are no longer permitted in haemodialysis systems. The clinical picture is likened by some authors to metal fume fever (Lyle et al, 1976). Headache, myalgia, rigors, fever and gastrointestinal upset typically occur at the time of dialysis, improving over several hours when dialysis is stopped. Severe cases may be complicated by delayed onset haemolytic anaemia as discussed below. Neurotoxicity Dialysis associated headaches, fatigue, "chills" and myalgia resolved in a nine year-old girl on haemodialysis following removal of a five metre copper pipe between the softener and dialysate supply system (Lyle et al, 1976). A 53 year-old male was admitted in coma, responding only to painful stimuli, two days following his 219th dialysis (Klein et al, 1972). At the time of dialysis he had developed headache and diarrhoea, progressing to a stuporous state 36 hours later, with haematological abnormalities (see below). The patient died seven days after onset of symptoms despite exchange transfusion, haemodialysis and peritoneal dialysis. Gastrointestinal toxicity Abdominal pain, nausea, vomiting and diarrhoea typically occur as early features during haemodialysis with copper contaminated dialysate and persist for up to 24 hours (Klein et al, 1972). Increased serum amylase activity has been observed in dialysis-associated copper toxicity (Klein et al, 1972). Autopsy findings in fatal cases include necrotizing haemorrhagic pancreatitis (Klein et al, 1972). Musculoskeletal toxicity Myalgia is a typical feature of dialysis-associated "copper fever". A nine year-old female often experienced facial, back and limb pain during or shortly after dialysis (Lyle et al, 1976). Severe myalgia may be associated with copper-induced muscle damage and the development of rhabdomyolysis with free serum myoglobin and elevated creatine phosphokinase activity (Klein et al, 1972). Haemotoxicity Manzler and Schreiner (1970) described intravascular haemolysis as a delayed complication of dialysis-associated copper intoxication. In a typical case "chills", nausea, vomiting, abdominal pain and diarrhoea occurred at the time of dialysis, improving over three hours but the patient re-presented 18 hours later with recurrence of gastrointestinal symptoms, profound postural hypotension and intravascular haemolysis requiring a four unit blood transfusion. In a similar case leukocytosis, haemolysis with reticulocytosis, a decrease in haematocrit (17 per cent), and the presence of free serum haemoglobin/myoglobin were observed in association with other systemic features of copper poisoning following haemodialysis. The serum copper concentration rose from 2.0 mg/L before dialysis to 26.6 mg/L post dialysis. The patient died seven days later (Klein et al, 1972). Cardiovascular toxicity Several authors have observed orthostatic hypotension typically associated with reflex tachycardia in haemodialysis patients exposed to copper- contaminated dialysate (Manzler and Schreiner, 1970; Klein et al, 1972). Hypotension and cardiac arrhythmias were described in a patient four days following haemodialysis with copper-contaminated dialysate. The patient did not respond to therapy and died on the fifth hospital day (Klein et al, 1972). Metabolic disturbances Metabolic acidosis is described in patients with haemodialysis-associated copper poisoning (Klein et al, 1972). Genitourinary toxicity Acute and chronic orchitis were reported at autopsy in two patients who died following haemodialysis-induced copper poisoning. It is likely however that this was a complication of chronic haemodialysis and/or uraemia rather than copper intoxication (Klein et al, 1972). MANAGEMENT Dermal exposure Following acute exposure irrigate the affected area with lukewarm water. Particular care is required if copper has been in prolonged contact with broken skin since systemic copper uptake is then possible. Copper contact sensitivity or irritant dermatitis are managed most effectively by discontinuing exposure. Ocular exposure Irrigate immediately with lukewarm water or preferably saline for at least 10 minutes. A local anaesthetic may be indicated for pain relief and to overcome blepharospasm. Ensure removal of any particles lodged in the conjunctival recesses. The instillation of fluorescein allows detection of corneal damage. Copper foreign bodies pose a serious hazard to the eye, and if suspected, an ophthalmological opinion should be obtained promptly. The non-magnetic properties of copper complicate ocular removal. Surgical removal often is necessary. Highly specialized techniques have been established to accurately determine the extent of damage and the precise intra-ocular location of copper foreign bodies (Grant and Schuman, 1993). Ingestion Elemental copper is radiopaque allowing easy localization following ingestion. In many circumstances, however, exposure is to ionized copper. Effective management of copper ingestion primarily involves appropriate symptomatic and supportive care. Gastrointestinal decontamination is most unlikely to favourably influence outcome and should only be considered within the first hour after a potentially life threatening ingestion. Liver function assessment is important, especially in children, if chronic excess copper ingestion is suspected. In these circumstances the serum copper concentration is likely to be increased. The role of chelating agents is discussed below. Gastrointestinal burns following molten copper ingestion are managed as a thermal burn with early endoscopy and a surgical opinion if severe burns are present. Inhalation Metal fume fever is managed most effectively by removal from exposure. Other symptomatic and supportive measures should be instituted according to the patient's condition. There typically are no permanent radiological abnormalities although transient ill-defined opacities on chest X-ray are recognized. Injection Suspected copper toxicity in haemodialysis patients requires confirmation by determination of the copper concentration in the water supply. Management is essentially supportive following removal of the copper source. Whole blood copper concentrations give some indication of the body copper burden. Haematological, biochemical and immunological profiles should be monitored. Intravascular haemolysis is managed conventionally. The use of chelating agents is limited since they primarily serve to enhance renal copper elimination. Antidotes Animal studies The application of dimercaprol-containing ointments or the injection of aqueous dimercaprol into the eyes of animals with experimentally induced penetrating copper injury was of no benefit (Grant and Schuman, 1993). d-Penicillamine, triethylenetetramine dihydrochloride (trien) and DMPS each administered in a dose of 50 µmol/kg intraperitoneally daily for five days were the most effective chelating agents in increasing copper excretion in the urine (p<0.01) in copper-poisoned rats fed a high copper diet for 20 days prior to chelation (Planas-Bohne, 1979). Faecal copper excretion was unaffected. Other workers have demonstrated enhanced renal copper elimination following parenteral DMPS and DMSA (Maehashi et al, 1983). Rana and Kumar (1983) suggested oral sodium calciumedetate (1g/kg daily for ten days) could limit histopathological renal damage in rats fed oral copper sulphate 0.1 g/kg daily for 20 days prior to chelation therapy. Protection against copper-induced hepatic and renal lesions was observed also in mice administered intraperitoneal DMPS 132 mg/kg 20 minutes after intraperitoneal copper sulphate 10 mg/kg (approximately the LD50) (Mitchell et al, 1982). DMPS was the most effective antidote in protecting against copper-induced mortality in copper sulphate-intoxicated mice (10 mg/kg intraperitoneally, LD50 8.7 mg/kg) administered intraperitoneal antidotes 20 minutes post dosing at a 10:1 molar ratio antidote: copper sulphate. Mice were observed for two weeks or until death. The survival ratio following DMPS was 25/30, compared to 7/30, 5/15, 4/15, 3/15, 3/15 for d-penicillamine, triethylene-tetramine, sodium calciumedetate, DMSA and dimercaprol respectively (p<0.0001 for DMPS compared to all chelating agents except triethylenetetramine, p<0.0005) (Jones et al, 1980). Henderson et al (1985) investigated the effect of single and repeated doses of chelating agents on copper toxicity. Copper intoxicated mice (10-130 mg/kg subcutaneously) were given single doses of dimercaprol 10 mg/kg or N-acetylcysteine 200 mg/kg, 30 minutes post dosing. With a single dose of chelating agent, the calculated LD50 (± SE) was significantly (p<0.05) increased from 54.7 ± 10 mg/kg in control mice to 95.2 ± 22 mg/kg and 87 ± 14 mg/kg in mice treated with dimercaprol or NAC respectively. The chelating agents were even more effective (p< 0.05) in copper-poisoned mice (40-170 mg/kg subcutaneously) treated with repeated doses of chelating agent: dimercaprol 10 mg/kg, N-acetylcysteine 200 mg/kg or d-penicillamine 50 mg/kg every hour for five hours, with calculated LD50 values of 60.5 ± 12 mg/kg, 150.3 ± 35 mg/kg, 139.4 ± 8 mg/kg and 91.4 ± 16 mg/kg for controls, dimercaprol, NAC and d-penicillamine treated mice respectively. d-Penicillamine, 52 mg/kg daily for six days, significantly (p<0.05) enhanced urinary copper excretion in four copper-poisoned sheep (given 20 mg/kg copper sulphate intraruminally daily for 35 days) (Botha et al, 1993). Under the same conditions triethylenetetramine failed to increase urinary copper excretion although the authors suggested this might have been related to specific features of ruminant metabolism. There is some evidence that polyamines structurally related to triethylenetetramine (e.g. 2,3,2-tetramine) have a more potent cupruretic action (Borthwick et al, 1980) but experience with these agents is limited (Twedt et al, 1988). Diethyldithiocarbamate (DDC) chelates copper but the lipophilic chelate accumulates in tissues, especially the brain (Iwata et al, 1970; Jasim et al, 1985), suggesting it may be an unsuitable antidote in copper poisoning. It has been suggested that DDC modifies the permeability of cell membranes and the blood brain barrier to copper (Allain and Krari, 1993). Clinical studies Wilson's disease Wilson's disease, characterized by decreased biliary copper excretion traditionally has been treated with d-penicillamine which serves to increase urinary copper elimination (Scheinberg et al, 1987). Adverse reactions to d-penicillamine are not uncommon and frequently are immunologically rather than toxicologically-induced including nephrotic syndrome, systemic lupus erythematosus (Walshe, 1982), white cell dyscrasias, thrombocytopenia, haemolytic anaemia (Walshe, 1982) and urticaria (Walshe, 1968). Anorexia, nausea and vomiting are described (Walshe, 1968). In animal studies penicillamine induces hepatic metallothionein (Heilmaier et al, 1986) which may disrupt the body distribution of other trace elements. Adverse effects occur in up to 10 per cent of patients receiving penicillamine and may necessitate treatment withdrawal (Walshe, 1982). Thus, in recent years, alternative agents have been investigated. Sunderman et al (1963) advocated parenteral and/or oral DDC in the management of Wilson's disease but evidence that this antidote enhances cerebral copper uptake limits its usefulness (see above). Walshe (1982) demonstrated increased urine copper elimination, symptomatic improvement and resolution of basal-ganglia abnormalities on CT brain scan among 20 patients with Wilson's disease treated with triethylenetetramine. These authors suggested triethylenetetramine as an effective drug for the treatment and maintenance of patients with Wilson's disease at all stages of the illness. Others concur with this view (Dubois et al, 1990; Morita et al, 1992) although there are potential hazards of triethylenetetramine therapy, notably sideroblastic anaemia (Perry et al, 1996). Although zinc sulphate has been utilized as alternative therapy to penicillamine in patients with Wilson's disease (Hoogenraad and Van den Hamer, 1983; Van Caillie-Bertrand et al, 1985; Veen et al, 1991), this treatment is unsuitable for acute copper poisoning as the mechanism of benefit is reduced gastrointestinal copper absorption. DMPS 200 mg bd increased urine copper elimination in a patient with Wilson's disease (Walshe, 1985). Acute poisoning There are no controlled data regarding the use of any chelating agent in acute copper poisoning. In severely poisoned patients the presence of acute renal failure often limits the potential for antidotes which enhance urinary copper elimination. d-Penicillamine, the standard therapy for Wilson's disease, has been utilized in copper poisoning (Holtzman et al, 1966; Jantsch et al, 1984/85; Hantson et al, 1996) but without confirmed evidence of enhanced urinary copper excretion. Intramuscular dimercaprol (Fairbanks, 1967; Jantsch et al, 1984/85; Schwartz and Schmidt, 1986; Hantson et al, 1996) and intravenous sodium calciumedetate (Holleran, 1981; Agarwal et al, 1975) have also been employed but again without confirmed benefit. A five year-old child with copper intoxication following repeated application of copper sulphate crystals to skin burns received a 12 day course of d-penicillamine 250 mg qds (Holtzman et al, 1966). Six hour urine copper excretion on the first day of chelation was 1000 µg, with a maximum value of 2000 µg/6h some 24 hours later. No pre- or post-chelation copper excretion data were given. Jantsch et al (1984/85) advocated the use of chelation therapy with dimercaprol and d-penicillamine following their experience with a patient who survived the alleged ingestion of 250 g copper sulphate. A single intramuscular dimercaprol dose 4 mg/kg was administered within the first ten hours (time not specified) followed by oral d-penicillamine 250 mg qds for at least seven days. The only 24 hour urine copper excretion measured "after initiation of chelation therapy" was 8160 µg (time not specified) with no pre- or post-chelation data presented. This case was unusual in that despite massive copper sulphate ingestion the patient developed no features of severe gastrointestinal irritation (save initial vomiting), no haemolysis or oliguria. Walsh et al (1977) administered intramuscular dimercaprol 2.5 g/kg (?2.5 mg/kg) plus 12.5 g/kg (?12.5 mg/kg) "edetic acid" four hourly to an 18 month-old child, commencing five hours after ingestion of 3 g copper sulphate. The urine copper concentration from a two hour collection was 500 µg/L on the second day, increasing to 3000 µg/L on day 12. The chelating agent was then switched to d-penicillamine 250 mg daily for one month with a gradual fall in urine copper excretion. Unfortunately urine volumes were not stated and no pre-chelation measurements were possible. Hantson et al (1996) recently treated an 86 year-old woman with acute copper sulphate poisoning with intramuscular dimercaprol 4 mg/kg qds and oral d-penicillamine 250 mg qds, both commenced within four hours of poisoning. Urine copper elimination was not enhanced and chelation was discontinued after 48 hours following onset of renal failure. These authors concluded that "available clinical and toxicokinetic data do not support the benefits of chelation in addition to supportive therapy" in acute copper (and zinc) sulphate poisoning. Alkaline diuresis Muthusethupathi et al (1988) advocated forced alkaline diuresis in copper sulphate poisoning. In 103 copper sulphate-poisoned patients in whom gastric lavage followed by forced alkaline diuresis were instituted immediately, the incidence of renal failure was claimed to be substantially lower (14.6 per cent) than in other similar series. However, no copper excretion data were reported, and it is possible that prompt fluid resuscitation with correction of hypovolaemia played an important role in patient recovery (Muthusethupathi et al, 1988). Haemodialysis Haemodialysis for five hours in a 41 year-old female failed to remove copper when instituted 12 hours after the ingestion of 280 mL dissolved copper sulphate (Agarwal et al, 1975). The patient had already undergone gastric lavage, had received intravenous sodium calciumedetate (1g) and a blood transfusion but died on the sixth hospital day after developing septicaemia, hepatic and renal failure. Peritoneal dialysis Cole and Lirenman (1978) reported a two year old child who had ingested some 30 mL super-saturated copper sulphate solution and underwent peritoneal dialysis for the management of renal failure. Copper extraction into the dialysate was enhanced markedly by the addition of salt-poor albumin 25 g/L. Over a 40 hour dialysis period (between 17 and 57 hours post ingestion) 0.7 mg copper was removed in 17 litres dialysate compared to 9.1 mg copper removed in 24 litres during dialysis with added albumin between 57 and 117 hours. The authors advocated albumin-enriched peritoneal dialysis in the management of copper poisoning complicated by acute renal failure. It should be noted, however, that the child consumed at least 2.7 g copper so that the amount removed by dialysis, even with albumin, was small. Enhancing elimination: Conclusions and recommendations 1. There are no controlled clinical data regarding the use of chelating agents in copper poisoning. 2. Animal data suggest DMPS may be the most effective antidote in copper poisoning, though DMPS was administered within 20 minutes of copper dosing in these studies. DMPS has a more favourable adverse effect profile than dimercaprol and d-penicillamine although these are alternatives if DMPS is not available. DMPS usually is given orally or parenterally in a dose of 30 mg/kg body weight per day. Side effects are infrequent but have included allergic skin reactions, nausea and vertigo (Aposhian, 1983). Discussion of individual cases with an NPIS physician is recommended. 3. There is insufficient evidence to advocate alkaline diuresis in the management of acute copper poisoning. 4. The role of haemodialysis and peritoneal dialysis is limited to the management of renal failure. Management of copper and caeruloplasmin concentrations in biological fluids Although whole blood copper concentrations correlate well with the severity of poisoning following acute ingestion, they should always be interpreted in conjunction with the clinical features. Serum copper concentrations are less useful in acute intoxications (Chuttani et al, 1965). In 20 patients who ingested copper sulphate, mean (± SD) whole blood copper concentrations were markedly lower (2.9 ± 1.3 mg/L) in those with only gastrointestinal symptoms compared to those who developed jaundice, renal failure or shock (mean whole blood copper 8.0 ± 4.0 mg/L). The number of patients in each group was not stated. Among 65 cases of acute copper sulphate poisoning, Wahal et al (1976) observed that although patients who developed complications had higher whole blood, red cell and plasma copper concentrations than uncomplicated cases, the difference was not statistically significant (p>0.05). No correlation was found between plasma copper concentrations and prognosis. However, whole blood copper concentrations greater than 1.2 mg/L were associated generally with the development of complications. The four fatalities reported, who were admitted within 6-8 hours of ingestion, had whole blood concentrations of at least 2.1 mg/L. Serum caeruloplasmin concentration estimation has been suggested as a useful prognostic indicator in cases of acute copper sulphate poisoning. Wahal et al (1978) observed significantly higher (p<0.001) serum caeruloplasmin concentrations in uncomplicated cases of copper sulphate poisoning than in those with complications (gastrointestinal haemorrhage, jaundice, renal impairment, delirium or coma). Values less than 35 mg/dL within 24 hours of poisoning or less than 44 mg/dL beyond 72 hours post ingestion were associated with the development of complications. Increased urine copper excretion (preferably as a 24 hour collection) will be present in any moderate or severe case of copper sulphate poisoning. The main value of this measurement is to monitor the effect of chelation therapy. AT RISK GROUPS Infants are at increased risk of excess copper accumulation during the first three months of life since their hepatic copper stores are significantly higher than in adults. This is due to the presence of fetal copper-binding protein which enables the fetal liver to accumulate sufficient copper to maintain body stores despite the relatively lower copper content of breast milk (Walker-Smith and Blomfield, 1973). MEDICAL SURVEILLANCE Close attention to personal hygiene and the use of appropriate protective equipment are of primary importance among those occupationally exposed to copper. Twenty-four hour urine copper excretion is a useful screening procedure if copper intoxication is suspected but the source of exposure is unclear. However, when collected in an occupational setting great care must be taken to avoid sample contamination. Serum or whole blood copper concentrations may be useful if exogenous copper contamination of urine samples is suspected (Cohen, 1979). It should be remembered that impaired biliary copper excretion from any cause will lead to increased renal copper elimination. Pre-employment screening for Wilson's disease may be indicated in those occupationally exposed to copper. Normal copper concentrations in biological fluids Plasma/serum: 0.7 - 1.3 mg/L (Weatherall et al, 1996). Whole blood: 1.6 - 2.7 mg/L (Chuttani et al, 1965). Urine: Less than 60 µg/24h (Weatherall et al, 1996). OCCUPATIONAL DATA Occupational exposure standard Copper: Long-term exposure limit (8 hour TWA reference period) fume 0.2 mg/m3; dusts and mists 1 mg/m3 (Health and Safety Executive, 1997). OTHER TOXICOLOGICAL DATA Carcinogenicity There is no conclusive evidence that copper is carcinogenic in humans (Aaseth and Norseth, 1986). Enterline et al (1995) studied the incidence of cancer in a cohort of 2802 copper smelters employed for at least one year during the period 1940-64. A significant (p<0.01) increase in deaths from respiratory cancer (SMR 209.7) was noted during the follow-up period 1941-86 but were attributed to cumulative arsenic rather than copper exposure during employment. Chewing copper-containing Areca nuts, common in the Orient, has been associated with oral submucous fibrosis and an increased risk of oral cancer (World Health Organization, 1984). In three volunteers saliva copper concentrations were significantly (p<0.001) increased during nut chewing compared to control values (Trivedy et al, 1997). The authors suggested copper released from the nuts induced lysyl oxidase activity, upregulating collagen synthesis and facilitating its cross- linking, resulting in connective tissue accumulation. Reprotoxicity In a controlled study Barash et al (1990) investigated the teratogenic potential of copper releasing intrauterine contraceptive devices (IUD) on the developing human embryo. No malformations or copper deposits were observed in the organs/placentae of copper IUD-exposed embryos (n=11) examined between seven and 12 weeks gestation. The results from the small study suggest that copper releasing IUDs have no observed negative effects on the developing embryo. Genotoxicity Copper induced sister chromatid exchanges in human peripheral lymphocytes (DOSE, 1993). Fish toxicity Chronic, partial chronic and early life stage toxicity tests were carried out on bluegill sunfish, bluntnose minnow, king salmon, fathead minnow and brook trout. The study duration was 30-60 days post-hatch. Under hard water conditions for fat head minnow the lowest observed effect concentration (LOEC) - no observed effect concentration (NOEC) was 33-15 µg/L and for the bluntnose minnow 18-4 µg/L. The reproduction part of the life cycle gave the most sensitive responses. In a chronic study (30-60 days post-hatch) with fathead minnow and bluegill sunfish, in soft water conditions LOEC-NOEC range was 40-11 µg/L, fry survival was the most sensitive response. A partial chronic study (30-60 days post-hatch) the LOEC-NOEC for brook trout was 17-9 µg/L, fry growth and survival were the most sensitive responses. LC50 (96 hr) Oreochromis niloticus 1.06 mg/L. LC50 (96 hr) rainbow trout 0.253 mg/L. LC50 (48 hr) larvae of flat fish Paralichthys olivaceus 0.36 mg/L (Cu2+). The fertilized eggs of Cyprinus carpio (108 hr) were exposed to 10, 50, 70 and 100 ppb copper. Survival of developing eggs, hatchlings, hatching and hatchability percentage decreased with increasing concentration. Deformities observed were formation of blisters, curved tail, stunted growth, circulatory failure, enlargement of the pericardial sac, deformed head region, underdeveloped fins and deformed vertebral column. Rainbow trout exposed to a number of combinations of copper, water hardness and pH showed reduced growth rate during the first 10 days, followed by partial or complete recovery. The lethal concentration of copper to rainbow trout was not affected by alkalinity at 10-50 ppm in soft water, however, the toxicity doubled by the same alkalinity change in hard water. Synergism between pH value and copper toxicity was observed (DOSE, 1993). EC Directive on Drinking water quality 80/778/EEC EC advisory level for drinking water, 100 µg/L at source of supply; 3000 µg/L after standing in piping for 12 hours (DOSE, 1993). WHO Guidelines for Drinking Water quality Guideline value for drinking water 2 mg/L (WHO, 1993). 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 28/1/98 REFERENCES Aaseth J, Norseth T. Copper. In: Friberg L, Nordberg GF, Vouk V, eds. Handbook on the toxicology of metals. Vol 2. 2nd ed. Amsterdam: Elsevier, 1986; 233-54. Agarwal BN, Bray SH, Bercz P, Plotzker R, Labovitz E. Ineffectiveness of hemodialysis in copper sulphate poisoning. Nephron 1975; 15: 74-7. Allain P, Krari N. Diethyldithiocarbamate and brain copper. Res Commun Chem Pathol Pharmacol 1993; 80: 105-12. Aposhian HV. DMSA and DMPS - water soluble antidotes for heavy metal poisoning. Ann Rev Pharmacol Toxicol 1983; 23: 193-215. Barash A, Shoham (Schwartz) Z, Borenstein R, Nebel L. Development of human embryos in the presence of a copper intrauterine device. Gynecol Obstet Invest 1990; 29: 203-6. Barkoff JR. Urticaria secondary to copper intrauterine device. Int J Dermatol 1976; 15: 594-5. Bentur T, Koren G, McGuigan M, Spielberg SP. An unusual skin exposure to copper; clinical and pharmacokinetic evaluation. Clin Toxicol 1988; 26: 371-80. Borthwick TR, Benson GD, Schugar HJ. Copper chelating agents. A comparison of cupruretic responses to various tetramines and D-penicillamine. J Lab Clin Med 1980; 95: 575-80. Botha CJ, Naude TW, Swan GE, Dauth J, Dreyer MJ, Williams MC. The cupruretic effect of two chelators following copper loading in sheep. Vet Hum Toxicol 1993; 35: 409-13. Britton RS. Metal-induced hepatotoxicity. Semin Liver Dis 1996; 16: 3-12. Chuttani HK, Gupta PS, Gulati S, Gupta DN. Acute copper sulfate poisoning. Am J Med 1965; 39: 849-54. Cohen SR. Environmental and occupational exposure to copper. In: Nriagu JO, ed. Copper in the environment. Part II: Health effects. New York: John Wiley & Sons, Inc, 1979; 1-16. Cole DEC, Lirenman DS. Role of albumin-enriched peritoneal dialysate in acute copper poisoning. J Pediatr 1978; 92: 955-7. Cooper R, Goodman J. Green Hair. N Engl J Med 1975; 292: 483-4. Dash SC. Copper sulphate poisoning and acute renal failure. Int J Artif Organs 1989; 12: 610. Donoghue AM, Ferguson MM. Superficial copper staining of the teeth in a brass foundry worker. Occup Med 1996; 46: 233-4. DOSE/Dictionary of substances and their effects. Vol 2. Cambridge: Royal Society of Chemistry, 1993. Dubois RS, Rodgerson DO, Hambidge KM. Treatment of Wilson's disease with triethylene tetramine hydrochloride (Trientine). J Pediatr Gastroenterol Nutr 1990; 10: 77-81. Earl CJ, Moulton MJ, Selverstone B. Metabolism of copper in Wilson's disease and in normal subjects. Am J Med 1954; 17: 205-13. Enterline PE, Day R, Marsh GM. Cancers related to exposure to arsenic at a copper smelter. Occup Environ Med 1995; 52: 28-32. Epstein S. Cross-sensitivity between nickel and copper. With remarks on cross-sensitivity between nickel, cobalt and chromates. J Invest Dermatol 1955; 25: 269-74. Fairbanks VF. Copper sulfate-induced hemolytic anemia. Inhibition of glucose-6-phosphate dehydrogenase and other possible etiologic mechanisms. Arch Intern Med 1967; 120: 428-32. Gleason RP. Exposure to copper dust. Am Ind Hyg Assoc J 1968; 29: 461-2. Goldsmith LA, Holmes LB. Green Hair. N Engl J Med 1975; 292: 484. Gorell JM, Johnson CC, Rybicki BA, Peterson EL, Kortsha GX, Brown GG, Richardson RJ. Occupational exposures to metals as risk factors for Parkinson's disease. Neurology 1997; 48: 650-8. Grant WM, Schuman JS. Toxicology of the eye, 4th ed. Springfield, Illinois: Charles C Thomas, 1993. Hantson P, Lievens M, Mahieu P. Accidental ingestion of a zinc and copper sulfate preparation. Clin Toxicol 1996; 34: 725-30. Hasan N, Emery D, Baithun SI, Dodd S. Chronic copper intoxication due to ingestion of coins: a report of an unusual case. Hum Exp Toxicol 1995; 14: 500-2. Hazardtext. In: Tomes plus. Environmental Health and Safety Series 1. Vol 33. Colorado: Micromedex, Inc., 1997. Heilmaier HE, Jiang JL, Greim H, Schramel P, Summer KH. d-Penicillamine induces rat hepatic metallothionein. Toxicology 1986; 42: 23-31. Henderson P, Hale TW, Shum S, Habersang RW. N-acetylcysteine therapy of acute heavy metal poisoning in mice. Vet Hum Toxicol 1985; 27: 522-5. Holleran RS. Copper sulfate overdose. J Emerg Nurs 1981; 7: 136-7. Holtzman NA, Elliott DA, Heller RH. Copper intoxication. Report of a case with observations on ceruloplasmin. N Engl J Med 1966; 275: 347-52. Hoogenraad TU, Van den Hamer CJA. 3 years of continuous oral zinc therapy in 4 patients with Wilson's disease. Acta Neurol Scand 1983; 67: 356-64. HSDB/Hazardous Substances Data Bank. In: Tomes plus. Environmental Health and Safety Series 1. Vol 33. National Library of Medicine, 1997. IPCS/International Programme on Chemical Safety. Copper: Essentiality and toxicity. IPCS News 1996a; 10: 4-5. IPCS/International Programme on Chemical Safety. Guidelines for drinking-water quality. 2nd ed. Vol 2. Health criteria and other supporting information. Geneva: World Health Organization, 1996b. Iwata H, Watanabe K, Miichi H, Matsui Y. Accumulation of copper in the central nervous system on prolonged administration of sodium diethyldithiocarbamate to rats. Pharmacol Res Commun 1970; 2: 213-20. Jantsch W, Kulig K, Rumack BH. Massive copper sulfate ingestion resulting in hepatotoxicity. Clin Toxicol 1984/85; 22: 585-588. Jasim S, Danielsson BRG, Tjälve H, Dencker L. Distribution of 64Cu in foetal and adult tissues in mice: Influence of sodium diethyldithiocarbamate treatment. Acta Pharmacol Toxicol 1985; 57: 262-70. Jones MM, Basinger MA, Tarka MP. The relative effectiveness of some chelating agents in acute copper intoxication in the mouse. Res Commun Chem Pathol Pharmacol 1980; 27: 571-7. Karlberg A-T, Boman A, Wahlberg JE. Copper - a rare sensitizer. Contact Dermatitis 1983; 9: 134-9. Klein Jr WJ, Metz EN, Price AR. Acute copper intoxication. A hazard of hemodialysis. Arch Intern Med 1972; 129: 578-82. Knobeloch L, Ziarnik M, Howard J, Theis B, Farmer D, Anderson H, Proctor M. Gastrointestinal upsets associated with ingestion of copper-contaminated water. Environ Health Perspect 1994; 102: 958-61. Kurisaki E, Kuroda Y, Sato M. Copper-binding protein in acute copper poisoning. Forensic Sci Int 1988; 38: 3-11. Lampe R, Henderson AL, Hansen GH. Green hair. J Am Med Assoc 1977; 237: 2092. Linder MC, Hazegh-Azam M. Copper biochemistry and molecular biology. Am J Clin Nutr 1996; 63: 797-811. Lyle WH, Payton JE, Hui M. Haemodialysis and copper fever. Lancet 1976; 1: 1324-5. Maehashi H, Yamaguchi Y, Tsutsumi S. Arsenic and copper excretion after treatment of arsenic poisoning in rats with heavy-metal antagonists. Dev Sci Pract Toxicol 1983; 11: 325-8. Manzler AD, Schreiner AW. Copper-induced acute hemolytic anemia. A new complication of hemodialysis. Ann Intern Med 1970; 73: 409-12. Meditext. In: Tomes plus. Environmental Health and Safety Series 1. Vol 33. Colorado: Micromedex, Inc., 1997. Mitchell WM, Basinger MA, Jones MM. Antagonism of acute copper(II)-induced renal lesions by sodium 2,3-dimercaptopropanesulfonate. John Hopkins Med J 1982; 151: 283-5. Morita J, Yoshino M, Watari H, Yoshida I, Motohiro T, Yamashita F, Okano Y, Hashimoto T. Wilson's disease treatment by triethylene tetramine dihydrochloride (Trientine, 2HCI): Long-term observations. Dev Pharmacol Ther 1992; 19: 6-9. Mühlendahl KE, Lange H. Copper and childhood cirrhosis. Lancet 1994; 344: 1515-6. Muthusethupathi MA, Selvaraj AM, Shivakumar S, Chenthil S. Forced alkaline diuresis in acute copper sulphate poisoning. J Assoc Physicians India 1988; 36: 459-60. NIOSH/NIOSH Pocket Guide. In: Tomes plus. Environmental Health and Safety Series 1. Vol 33. National Institute for Occupational Safety and Health (NIOSH), 1997. Nordlund JJ, Hartley C, Fister J. On the cause of green hair. Arch Dermatol 1977; 113: 1700. OHM/TADS-Oil and hazardous Materials/Technical Assistance Data System. In: Tomes plus. Environmental Health and Safety Series 1. Vol 33. United States Environmental Protection Agency, 1997. O'Neill NC, Tanner MS. Uptake of copper from brass vessels by bovine milk and its relevance to Indian childhood cirrhosis. J Paediatr Gastroenterol Nutr 1989; 9: 167-72. Ostiguy G, Vaillancourt C, Bégin R. Respiratory health of workers exposed to metal dusts and foundry fumes in a copper refinery. Occup Environ Med 1995; 52: 204-10. Pande RS, Gupta YN. Thrombocytopenic purpura following copper sulphate therapy. J Indian Med Assoc 1969; 52: 227-8. Pandit AN, Bhave SA. Copper and Indian childhood cirrhosis. Indian Pediatr 1983; 20: 893-9. Pandit A, Bhave S. Present interpretation of the role of copper in Indian childhood cirrhosis, Am J Clin Nutr 1996; 63: 830S-5S. Parish LC. Green Hair. N Engl J Med 1975; 292: 483. Perry AR, Pagliuca A, Fitzsimons EJ, Mufti GJ, Williams R. Acquired sideroblastic anaemia induced by a copper-chelating agent. Int J Hematol 1996; 64: 69-72. Pimentel JC, Menezes AP. Liver disease in vineyard sprayers. Gastroenterology 1977; 72: 275-83. Planas-Bohne F. Influence of several chelating agents on the excretion and organ concentration of copper in the rat. Toxicol Appl Pharmacol 1979; 50: 337-45. Rana SVS, Kumar A. The protective effects of EDTA against copper poisoning in rats with special reference to the kidney. Int J Tissue React 1983; 2: 187-92. Reid DJ. Allergic reaction to copper cement. Br Dent J 1968; 125: 92. Reprotext. In: Tomes plus. Environmental Health and Safety Series 1. Vol 33. Colorado: Micromedex, Inc., 1997. RTECS/Registry of Toxic Effects of Chemical Substances. In: Tomes plus. Environmental Health and Safety Series I. Vol 33. National Institute for Occupational Safety and Health (NIOSH), 1997. Salmon MA, Wright T. Chronic copper poisoning presenting as pink disease. Arch Dis Child 1971; 46: 108-10. Santucci B, Cannistraci C, Cristaudo A, Picardo M. Interaction of metals in nickel-sensitive patients. Contact Dermatitis 1993; 29: 251-3. Scheinberg HI. Copper, alloys and compounds. In: Parmeggiani L, ed. Encylcopaedia of Occupational Health and Safety. Vol 1. 3rd ed. Geneva: International Labour Organisation, 1983; 546-8. Scheinberg IH, Sternlieb I, Schilsky M, Stockert RJ. Penicillamine may detoxify copper in Wilson's disease. Lancet 1987; 2: 95. Scheinberg IH, Sternlieb I. Is non-Indian childhood cirrhosis caused by excess dietary copper? Lancet 1994; 344: 1002-4. Schilsky ML. Wilson disease: Genetic basis of copper toxicity and natural history. Semin Liver Dis 1996; 16: 83-95. Schwartz E, Schmidt E. Refractory shock secondary to copper sulfate ingestion. Ann Emerg Med 1986; 15: 952-4. Shibuya S, Takase Y, Sharma N. Esophageal ulcer due to ingestion of melted copper. Dig Dis Sci 1992; 37: 1785-90. Singh MM, Singh GS. Biochemical changes in blood in cases of acute copper sulphate poisoning. J Indian Med Assoc 1968; 50: 549-54. Sittig M. Copper and compounds. In: Handbook of toxic and hazardous chemicals and carcinogens. 2nd ed. New Jersey: Noyes Publications, 1985; 256-9. Spitalny KC, Brondum J, Vogt RL, Sargent HE, Kappel S. Drinking-water-induced copper intoxication in a Vermont family. Pediatrics 1984; 74: 1103-6. Strickland GT, Beckner WM, Leu ML. Absorption of copper in homozygotes and heterozygotes for Wilson's disease and controls: isotope tracer studies with 67Cu and 64Cu. Clin Sci 1972; 43: 617-25. Sunderman FW Jr , White JC, Sunderman FW, Lucyszyn GW. Metabolic balance studies in hepatolenticular degeneration treated with diethyldithiocarbamate. Am J Med 1963; 34: 875-87. Tauxe WN, Goldstein NP, Randall RV, Gross JB. Radiocopper studies in patients with Wilson's disease and their relatives. Am J Med 1966; 41: 375-80. Terry RF. Excess copper in a local water supply. Med J Aust 1996; 165: 296. Trachtenberg DI. Allergic response to copper - its possible gingival implications. J Periodontol 1972; 43: 705-7. Trivedy C, Baldwin D, Warnakulasuriya S, Johnson N, Peters T. Copper content in Areca catechu (betel nut) products and oral submucous fibrosis. Lancet 1997; 349: 1447. Twedt DC, Hunsaker HA, Allen KGD. Use of 2,3,2-tetramine as a hepatic copper chelating agent for treatment of copper hepatotoxicosis in Bedlington Terriers. J Am Vet Med Assoc 1988; 192: 52-56. Van Caillie-Bertrand M, Degenhart HJ, Visser HKA, Sinaasappel M, Bouquet J. Oral zinc sulphate for Wilson's disease. Arch Dis Child 1985; 60: 656-9. Van Joost Th, Habets JMW, Stolz E, Naafs B. The meaning of positive patch tests to copper sulphate in nickel allergy. Contact Dermatitis 1988; 18: 101-2. Veen C, Van Den Hamer CJA, De Leeuw PW. Zinc sulphate therapy for Wilson's disease after acute deterioration during treatment with low-dose D-penicillamine. J Intern Med 1991; 229: 549-52. Villar TG. Vineyard sprayer's lung. Clinical aspects. Am Rev Respir Dis 1974; 110: 545-55. Wahal PK, Mehrotra MP, Kishore B, Patney NL, Mital VP, Hazra DK, Raizada MN, Tiwari SR. Study of whole blood, red cell and plasma copper levels in acute copper sulphate poisoning and their relationship with complications and prognosis. J Assoc Physicians India 1976; 24: 153-8. Wahal PK, Mehrotra MP, Kishore B, Goyal SP, Gupta MC, Patney NL, Raizada SN, Singh R. A study of serum ceruloplasmin levels in acute copper sulphate poisoning. J Assoc Physicians India 1978; 26: 983-7. Walker-Smith J, Blomfield J. Wilson's disease or chronic copper poisoning? Arch Dis Child 1973; 48: 476-9. Walsh FM, Crosson FJ, Bayley M, McReynolds J, Pearson BJ. Acute copper intoxication. Pathophysiology and therapy with a case report. Am J Dis Child 1977; 131: 149-51. Walshe JM. Toxic reactions to penicillamine in patients with Wilson's disease. Postgrad Med J 1968; PSuppl: 6-8. Walshe JM. Treatment of Wilson's disease with trientine (triethylene tetramine) dihydrochloride. Lancet 1982; 1: 643-7. Walshe JM. Unithiol in Wilson's disease. Br Med J 1985; 290: 673-4. Walton S. Investigation into patch testing with copper sulphate. Contact Dermatitis 1983a; 9: 89-90. Walton S. Patch testing with copper sulphate. Contact Dermatitis 1983b; 9: 337. Weatherall DJ, Ledingham JGG, Warrell DA, eds. Oxford Textbook of Medicine. 3rd ed. New York: Oxford University Press Inc., 1996. WHO/World Health Organization. Guidelines for drinking-water quality. 2nd ed. Vol 1. Recommendations. Geneva: World Health Organization, 1993. World Health Organization. Control of oral cancer in developing countries. Bull World Health Organ 1984; 62: 817-30. Wyllie J. Copper poisoning at a cocktail party. Am J Public Health 1957; 47: 617.