April 20, 2007
Critical Observations on the Neurotoxicity of Silver
By Lansdown, A B G
Silver is a xenobiotic element with no recognized trace metal value in the human body. It is absorbed into the body through the lungs, gastrointestinal tract, mucus membranes of the urinogenital tract, and through the skin, mainly in the form of silver protein complexes. Although silver is metabolized throughout the soft tissues, available evidence from experimental animal studies and human clinical reports has failed to unequivocally establish that it enters tissues of the central nervous system or is a cause of neurotoxic damage. Argyria characterized by deposition of particles of silver sulfide or silver selenide is the principle contraindication for using silver in medical devices or occupationally. This presents discoloration of the skin but is not regarded as a health risk or manifestation of toxicity. No evidence is available to demonstrate the toxic risk of silver to the peripheral nervous system, although silver sulfide deposits have been identified in the region of cutaneous nerves. Transitory silver sulfide deposits seen in the tissues of the blood-brain and blood- CSF barriers are mostly lysosomally bound or deposited on basement membranes or collagen without toxic effect. Silver is mostly excreted from the body in the urine and feces. Further research is indicated to evaluate the role of metal binding proteins including metallothioneins as cytoprotectants for neurological tissue.Keywords Argyria, Blood Brain Barrier, Brain, Central Nervous System, Medical Devices, Silver
Silver is widely distributed in the earth's crust and is found in soil, fresh and sea water, and the air. It is readily absorbed into the human body with food and drink and through inhalation, but the low levels of silver commonly present in the bloodstream (
Experience has shown that a large proportion of the silver ion released from medical devices not required for antimicrobial action is disseminated into tissue fluids and exudates, where it combines with albumins and macroglobulins.7 These silverprotein complexes are absorbed into the systemic circulation to be deposited in key soft tissues, including the skin, liver, kidney, spleen, lungs, and brain.8,9 As a xenobiotic material, silver must be presumed to present a health risk to exposed persons under some circumstances. Unlike the well-documented neurotoxic metals including lead and mercury, silver does not appear to be a cumulative poison and is eliminated from the body through the urine and feces.1,10 Excretion of silver by these routes may be a measure of mean daily intake, but since this view is based largely on the clinical use of silver nitrate and silver sulfadiazine used in burn wound therapy, its true relevance in the metabolism of silver used in the wider context of medical devices is questionable.
Argyria is the most widely publicized clinical condition associated with silver accumulation in blood and soft tissues.11-13 It commonly occurs in individuals exposed to high levels of silver occupationally (metallurgy, photography, and mining industries), or consuming or inhaling silver hygiene products (including colloidal silver products) for long periods.14 Silver is absorbed into the body and deposited in the perivascular regions of the skin and other soft tissues as black granules of silver sulfide or silver selenide. The resulting slate grey discoloration of the skin occasionally associated with melanogenic changes, is semipermanent and cosmetically undesirable but is not known to be life-threatening.13 Fung and Bowen reported that up to 10% of silver salts ingested may be deposited in soft tissues, with highest concentrations in skin, liver, spleen, and adrenal glands, with lesser levels in muscle and brain.9 The clinical implications of argyria are discussed extensively in the literature and there is evidence that in severe cases, silver deposited in the central nervous system "may be a cause" of neuromuscular dysfunction and behavioral abnormality.15"'9 This has not been unequivocally substantiated clinically or experimentally.
William Roper in 1990 comprehensively reviewed the toxicology of silver for the U.S .Public Health Service and discussed clinical and experimental studies associating silver deposition in the brain and neurological tissue and its propensity to cause injury.20 He indicated that published reports available then were not only inconclusive but failed to demonstrate clear causal relationships between silver deposition and tissue damage. He recommended that specific behavioral tests might be beneficial in defining the neurotoxic hazards of silver exposure and the vulnerability of the central nervous system to injury. Surprisingly, although analytical procedures for silver have become increasingly more sensitive in recent years and accurate methods of xray microanalysis, histochemistry, and autometallography have been developed for visualising silver deposits in tissues,21,22 no such reports have been published as far as I am aware.13,23 Such an appraisal is highly relevant now in view of the greatly increased use of silver in medical devices, major technological advances in materials science, and the widespread applications of nanotechnology in medicine.24,25 The minute silver particles produced by nanotechnology in wound dressings (Acticoat) and in in-dwelling catheters permit considerably greater levels of ionization and biological reactivity than previously shown with silver metal foil or many soluble silver salts, and are expected to achieve higher levels of antimicrobial activity and bioreactivity within the human body.26 The present critical review was conducted to analyze experimental and clinical reports purporting to show neurotoxic effects attributable to silver absorption, and to identify circumstances under which they might occur in patients fitted with a medical device with silver antibiotic content or coating.
NEUROTOXICITY OF METALS
Metals differ greatly in their transport mechanisms in the human body and their accumulation in the central nervous system. Acknowledged neurotoxic metals differ greatly in their ability to penetrate into soft tissues and their mechanisms of action and pathogenicity, but share the characteristics listed in Table 1.27,28 At least nine metals are known to penetrate into neurological tissues,29 including the toxic metals lead, cadmium, and mercury, with no known trace metal value. Sodium, potassium, calcium, iron, copper, zinc, manganese, cobalt, and molybdenum perform essential physiological functions in the human body or serve as enzyme cofactors, transcriptional factors, or modulators of gene expression. This last group of metals may exert pathophysiological changes if present in supra-optimal quantities.2 The position of silver as a neurotoxic metal is equivocal and unclear at the moment.29 An early experimental study in tadpoles implicated silver nitrate as a cause of "white matter edema," with water-filled vacuoles developing between extracellular membrane surfaces of myelin lamelli,30-32 but this has not been observed in mammals.
General characteristics of neurotoxic materials
The blood-brain barriers (BBB) perform a central role in maintaining chemical homeostasis within the central nervous system.33,34 By modulating the uptake of nutrients and electrolytes from the circulation and regulating the egression of metabolites, they control brain chemistry and limit minor changes, which may be expressed in terms of learning difficulties, memory loss, and behavioral dysfunction.
This research illustrates how chemically induced defects in the BBB may be a cause of edema, aberrant brain development and neurodegeneration.27,28,31 Zheng demonstrated that BBB have special significance in regulating the uptake and neurotoxic action of metals and exhibit a limited capacity to metabolize certain lipophilic materials that influence carrier-mediated processes.33 The neurotoxicity of metals and other xenobiotic materials is largely determined by the protective efficiency of the BBB in different regions of the central nervous system.27,34 An understanding of the critical role of the BBB is essential in appreciating the putative neurotoxic action of silver and the i\ncreased vulnerability of certain areas of the brain to injury.
The BBB is a complex system comprising the interface between the blood and the brain, and that separating the blood and cerebrospinal fluid (CSF).28,34,35 Endothelial cells lining the extensive vascular network of the brain and subarachnoid space provide a major component of the BBB, whereas the blood-CSF barrier resides largely in the choroid plexus and in the ependymal cells lining the cavity of the CSF. Peripheral nerves have analogous barrier systems comprising the vascular network and connective tissues of the endoneurium and the perineurium surrounding nerves and nerve bundles, respectively. In humans, the BBB is established at birth but in the choroid plexus and circumventricular organs (median eminence, subfornical organ, area postrema and neurohypophysis) it becomes less well developed. The endothelial cells lining vascular channels are of a fenestrated type even though they maintain tight gap junctions.27 Variations in the permeability of the BBB according to age and the region of the brain implies that that certain areas of the brain are more vulnerable to metal-induced injury than others.34,35 In the rat, the structure of the BBB and its relationship to surrounding astrocytes has been investigated using a silver-protein complex as a marker.36 This has demonstrated that the subfornical organ is largely devoid of BBB and that micro-vessels are separated from surrounding astrocytes only by a basement membrane.
Classification of metal ions according to their toxic action on the choroid plexus
Early evidence of the protective role of the BBB in controlling the penetration of xenobiotc materials was provided by experiments in which the intravital dye trypan blue was injected intravenously into rabbits.37,38 The dye bound to plasma protein was not absorbed into the tissues of the brain but sequestered and bound lysosomally in endothelial cells. It stained other soft tissues but was excluded from the brain unless injected intracerebrally. These observations promoted the concept that the BBB was unique to the central nervous system and acted in the form of an "exclusionary interface" separating brain from blood. According to Rapoport, cerebral capillaries facilitate diffusion and regulate exchange of metabolites between blood and brain.34 The ependymal surfaces of the cerebral ventricles and the piaglial surfaces of the brain do not impede transfer of substances between the cerebrospinal fluid and the brain and do not constitute a subbarrier.
Zheng studied the morphology of the blood-CSF barrier, with particular reference to the role of the choroid plexus in modulating metal-induced neurotoxicities.29,33 The choroid plexus is a highly vascular villous structure extending from the ventricular surfaces of the brain into cerebral spinal fluid like coral fronds.39,40 Although it represents less than 5% of total brain weight,41 it has a proportionately high surface area permitting greater exposure to the circulating CSF. Compelling evidence illustrates the critical role of the choroid plexus in sequestering toxic heavy metals like lead and mercury, and may regulate the neurotoxic action of silver.29 The blood flow in the choroid plexus is high and exposes it to a greater influx of toxic materials and efflux of metabolites than elsewhere in the brain.34
Ependymal cells lining the CSF surface are densely packed with tight junctions providing a modest barrier to the transfer of metal ions. In contrast, the fenestrated endothelial cells lining choroidal capillaries are more porous or "leaky," thereby permitting greater exchange of solutes and metal ions between the blood and connective-tissue matrix.34,39 The ependymal cells regulate the production and composition of the CSF including the interchange of metal ions, but their mechanisms of action, possibly involving sodium and potassium ATP-ase pump mechanisms, are imperfectly understood. Transport through the BBB is limited to nonpolar substances and several nutrients for which special carrier-mediated pathways exist.27 Experimental studies have demonstrated the ability of the connective tissue of the choroid plexus to concentrate metal ions, including organic mercury, cadmium, arsenic, and lead, and regulate their penentration into the neural tissues to evoke pathological damage.42-44 Metal ions may be conveniently classified according to their specific action on the choroid plexus33 (Table 2).
SILVER AND THE BLOOD-BRAIN BARRIER
Silver-induced neurotoxicity is believed to be rare,33 even though some experimental studies in the rat claim that silver ions do penetrate the BBB and the blood-placental barrier to locate heterogeneously throughout the central nervous system.21,22,42,43 Predictive experimental studies conducted in animal models are expected to provide more accurate and reproducible information on the neurotropic action of silver than is possible with postmortem material obtained from patients dying with argyria or supposed silver intoxication. Neurological tissues autolyse readily after death and visualization of sites of silver deposition may be obscured. Electron microscopy has been widely used in examining the deposition of silver in the region of the brain and other tissues, but x-ray microanalysis and autometallography have been developed as a means of visualizing fine silver deposits in the choroid plexus, neurons, glial cells, and extraneural tissues of the BBBs.22,45-48 Analysis of silver in "the brain" using ^sup 111^Ag tracer studies, atomic absorption spectrometry, and neutron-activation analysis is insufficient to discriminate between silver deposited within tissues of the brain and that contained within tissues of the BBB.49
Early studies conducted in rats exposed chronically to silver nitrate in drinking water failed to provide evidence that silver passes the BBB to accumulate in neural tissues of any part of the central nervous system. Where silver nitrate was employed as an intravital dye to demonstrate the integrity of the BBB in the rat, silver was deposited preferentially in basal laminae and perivascular spaces of the choroid plexus, hypophysis, pineal body, area postrema, and subfornical organ.51-54. It could not be identified outside circumventricular areas or around cerebral capillaries in severely argyric rats. Although silver is readily metabolized from tissues like liver and kidney in humans,1 it exhibited a longer half-life in endothelial cells of the BBB site than in other soft tissues in the rat. Later more elaborate studies by Scott and Norman confirmed the inability of silver to cross the BBB and demonstrated fine electron-dense silver granules (10-15 nm diameter) in the basal laminae of arterioles of the parietal cortex and subcortical white matter.55 Accumulation in these sites "maximized" by 241 days and did not change in concentration or distribution up to 455 days after exposure. In an attempt to increase the vulnerability of the brain to silver, Scott and Norman induced surgical intracerebral stab wound injury.55 This had the effect of increasing silver accumulation in the laminae of small blood vessels, tissue fragments of the BBB, and associated macrophages, but deposits were not identified within the brain parenchyma. They confirm earlier studies demonstrating that silver protein complexes do not penetrate the gap junctions of cerebral endothelia, even though some silver might dissociate at cell membranes and penetrate cells by an undefined mechanism other than pinocytosis.36 Alternatively, silver ion bound strongly to collagen and glycoproteins of the BBB.56
Two Russian studies (unseen) cited by the Joint FAO/WHO Expert Committee on Food Additives (1977) may provide evidence of a direct toxic effect on silver in the brain.57,58 The first claimed to show decreased brain RNA and DNA and dystrophic changes in rats given 0.2% silver nitrate in drinking water for 12 months or 2.0% for 6 months, while in the second study histopathological changes were reported in neuronal, glial, and vascular tissues of the encephalon and medulla of rabbits dosed with 0.025 or 0.25 mg/kg silver (possibly by intravenous injection), but further details are not available.
Evidence that silver penetrates the BBB and blood-placental barrier relies heavily on a comprehensive series of anatomical, histochemical, and electron-microscopical studies conducted in rats at the University of Aarhus in Denmark. Rungby and Danscher considered that the symptomatic effects of paralysis, loss of coordination, cerebella ataxia, convulsions, and electroencephalograph (EEG) changes seen in patients with severe argyria and chronic silver exposure are difficult to explain if silver had not penetrated the BBB.46 They administered silver nitrate or silver lactate to rats and mice orally or by intraperitoneal injection and employed autometallographic methods to demonstrate silver penetration of the blood-brain barrier and its deposition in all parts of the central nervous system.46,59-63 Animals given silver nitrate or silver lactate chronically in drinking water (0.01%), or injected intraperitoneally with silver lactate (3-55 mg for up to 13 months) or a colloidal silver preparation (Protagol (0.1-0.5 ml, 2-5 days) exhibited intracellular and extracellular silver sulfide deposition throughout the brain, dorsal root ganglia, enteric ganglia, peripheral nervous system, anterior pituitary gland, and neural retina of the eye.64 These distribution patterns were heterogeneous but particularly heavy in large motor neurons and protoplasmic astrocytes. The silver granules were bound specifically in secondary lysosomes.47 In keeping with earlier observations,55 tissues of the BBBs were heavily stained with silver sulfide deposits but much silver was located as extracellular deposits on basement membranes of cerebral blood vessels and on ela\stic fibers.43,47,48 The intensity of silver deposition in each case was proportional to the amount of silver administered and the duration of exposure, although subtle differences were evident between administration of silver nitrate and silver lactate. Importantly, silver deposits were transitory in these locations and declined when silver treatment was withdrawn. Macrophages engorged with silver deposits have been consistently reported in the BBB. Interestingly, when silver was injected into the lateral ventricles of the brain, it was absorbed into ependymal cells of the BBB, rather than locating in neurons or glial cells. Rungby and Danscher46 conceded that the paralysis reported in earlier studies in rats dosed with silver nitrate might be attributable to the toxic effects of silver accumulating in capillaries associated with the central nervous system,46,65 but gave no details.No specific neurobehavioral tests of the type promulgated by Roper20 were conducted.
Although Rungby and Danscher provided substantial evidence based upon electron microscopy, photochemical and autometallographic techniques that silver does cross the BBB to accumulate in specific locations in the central nervous system, they failed to associate this with frank neurological damage or behavioral changes.46,60,61 Thus, irrespective of route of administration (oral or intravenous), silver accumulation in lysosomal vacuoles occurs in a dose-related fashion in neurons and glial cells of the olfactory lobes, cerebral cortex, hippocampus, substantia innominata, and hypothalamus, but the thalamus, substantia nigra, and nuclei pontis seemed to be resistant. Neurons of the globus pallidus, brainstem, spinal cord and basal root ganglia, cerebellum (deep nuclei), and the trigeminal nerve also showed a strong tendency to concentrate silver. Silver deposits in the rat hippocampus and in the peripheral nervous system remained stable for at least 45 days.60 The brains of young postnatal animals may be more vulnerable to the toxic effects of silver, as suggested by a significant reduction the pyramidal cell layer of the hippocampus.61 This may be an indication of a cytostatic effect or other toxic effect of silver on developing hippocampal cells, but its implications on further development in the brain and behavior patterns are unclear.
Mean tissue silver concentration (ng/g wet wight) in mice given 0.03 mg/L silver nitrate to drink for 1 or 2 weeks67
Rungby and Danscher failed to identify frank toxic changes in neurological tissues in rats exposed by various routes to silver nitrate, silver lactate, or Protargol, but they did report that mice exposed chronically to very low levels (0.015%) of silver nitrate or silver lactate in drinking water became argyric and hypoactive in open-field behavioral studies.66 If this species-specific neurobehavioral change attributable to silver accumulation has a pathophysiological basis, the mechanism is unclear at present. More recent work by Pelkonen et al. reports silver accumulation in the cerebellum and the soleus muscles of young adult mice given 0.03 ml/ L silver nitrate to drink for 1-2 weeks, but failed to show changes in behavioral activity or disturbed health patterns.67 Some neurobehavioral studies would have been useful to explore the possible health implications of these findings. Their pooled data indicate that tissue concentrations of silver do not vary significantly within 2 weeks and that concentrations of silver in blood, cerebellum, soleus, and gastrocnemius muscles stabilize (Table 3).
An unpublished communication purporting to show permanent facial paralysis in human patients following postoperative mastoid surgery and application of silver nitrate to dehiscent facial nerves suggests that silver or nitrate ions may exert direct toxic damage on exposed nerves.68 Experimental studies conducted in Sprague- Dawley rats demonstrated that silver nitrate cautery for 1 second did evoke some axonal injury but allowed modest neuronal recovery. In contrast, exposures of 5 or 10 seconds resulted in a 50% axonal loss and impaired mobility in the 14 days after operation. It is expected that the corrosive action of the nitrate moiety is largely responsible for the observed changes.
Experimental models have been developed to study patterns of silver release and tissues accumulation from medicated catheters for in-dwelling use.69 Thus, an experimental study was designed to evaluate the release of silver ion from a silver-iontophoretic catheter in rabbits. In this model, a total of 0.4-2.5 mg was liberated within 12 weeks, but whereas blood silver increased to 1.0 g/L over 7 days, it declined to 0.28 g/L in 8-12 weeks. It is unclear from this study to what extent the silver released accumulated in the central nervous system or whether it evoked behavioral or other pathological changes.
Evaluation and clinical definition of neurotoxic risks associated with occupational or environmental exposure to silver are complicated by wide variations in patterns of exposure, quantitative analysis of silver in blood and tissues, and scientific detail presented. Clinical studies on silver nitrate and silver sulfadiazine in treating patients with severe burns injury provide fundamental information on silver absorption and tissue distribution, but accurate information on the accumulation or distribution in the central nervous system is still urgently required.1,12,70 Wan et al. critically examined methods available for quantifying silver in body tissues and fluids, and provided useful "baseline" or control silver levels in key tissues of patients with no known exposure to silver occupationally or therapeutically (Table 4).1 In their experience, flameless thermal atomic absorption spectrometry was far more accurate than older and more commonly reported techniques including spectrophotometry and flame atomic absorption.71-73 The levels of silver found in the cerebral gray matter of patients not knowingly exposed to silver and analyzed by high-resolution spectroscopy have been given as 0.029 g/ g.17 More recent analyses of patients dying in North America has shown the average silver content of tissues to be: skin 1.3 g/g (range 0.8-2.5 g/g), liver 0.7-1.0 g/g, adrenal
The uptake and accumulation of silver in tissues is illustrated by a cancer patient injected intravenously with 0.1 mg ^sup 111^Ag tracer.75 A large proportion of the silver absorbed into the circulation within the first 3 days (77%) was bound to plasma macroglobulins, 15% to albumin, and 8% to fibrinogen. These levels declined rapidly following injection and only 10% remained after 2 hours, thereafter declining to 2% for the next 20 days as silver was distributed to the soft tissues. Silver remained high in the skin and liver until the patient died 195 days later. The silver deposition in the brain of this patient is not known.
Argyria and argyrosis are the commonest observations reported in patients exposed to silver occupationally, or to silver nitrate or silver sulfadiazine in burn wound care.8,9,11,14,76 Light and electron-microscopic examination has demonstrated electron-dense granules of 30-100 nm in the skin and other tissues; these granules are composed largely of silver sulfide with traces of selenium, mercury, titanium, and iron. The electron microscope x-ray analyzer is capable of detecting silver sulfide deposits in tissue at concentrations as low as 1 10^sup -14^g/m^sup 2^.76 In each case, the granules have been observed mostly within secondary lysosomes of the basal lamina of the epidermis, small dermal blood vessels, Schwann cells, basement membranes of eccrine glands, and dermal elastic and collagen fibers and not associated with pathological changes.14 Other clinical studies examining the chemical constitution of so-called "silver deposits" in brain, liver, and other tissues have confirmed these observations.15,19,77,78
Silver analysis in the tissuesof humans not knowingly exposed to siver occupationally or therapeutically1
The famous "blue man" of Barnum and Bailey's Circus in 1927 is possibly the earliest recorded evidence of silver in the brain, where an estimated total body silver content of 90-100 g was associated with 0.011% in "the brain."79 Silver deposits were mainly associated with connective tissues and macrophages. The reliability of these estimates might be questioned on account of the accuracy of the analytical procedures available at the time.
Environmental and Occupational Exposures
Occupational exposure to silver in refining, metal work, photography, and preparation of silver compounds for industry is commonly associated with argyria and argyrosis.11,14,76 Blood silver concentrations in these workers may be more than twice that seen in unexposed individuals (11 g/L) and associated with high urinary (5 g/ g) and fecal silver excretion (15 g/g).71 Invariably, reported studies have focused on the deposition of silver sulfide granules in the skin and eye, with rare reference to other tissues or report of neurological abnormalities. Occupational health studies have shown that the cornea is a sensitive indicator of silver exposure.
Moss et al. examined 30 employees in an industrial plant involved in the manufacture of silver nitrate and silver oxide and identified corneal and conjunctival pigmentation in 20, with the severity of the discoloration being directly related to duration of employment.80 Ten workers with impaired night vision attributable to the silver deposits failed to show electrophysiologic or psychophysiologic evidence of functional deficits. Although direct evidence of neurobehavioral changes has not been seen in workers exposed to silver occupationally, Rosenman et al. did observe that most of the 20 New York factory workers showing occupational argyrosis complained of headaches, tiredness, and nervou\sness.81 They emphasized the importance of monitoring silver in the work environment and regularly examining staff with slit lamp to assess health status. In a more recent case of occupational argyrosis, multifocal degenerative epithelial changes in the cornea were associated with a diffuse deposition of silver in the corneal stroma and Descemet's membrane and tissue debris.82
Silver Nitrate in Oral Hygiene
Silver nitrate and colloidal silver preparations have been used in the treatment of mucus membrane infections and infective rhinitis for many years.83 Although not legally available now in the United States and some other countries, colloidal silver is widely available in various forms for treating miscellaneous ailments. It is a common cause of argyria and has been implicated as a cause of neurological problems.84-87 In his Manual of Pharmacology, Sollemann83 listed recommendations for the use of silver nitrate and colloidal silver for nose and throat infections as 2-10% silver nitrate, 0.5-10% strong silver proteins (Protargol), 10-30% sprays of mild silver proteins (Argyrol). In practice, it is almost impossible to calculate the amount of silver consumed in long-term therapies, and blood silver levels are a poor guide to silver absorption in the chronic consumption or inhalation of over-the- counter silver products. Silver accumulates in the blood initially but rapidly declines as some is excreted in urine and feces and the balance is distributed to soft tissues throughout the body.
A recent case reported as the "silver man" concerned a 42-year- old patient with severe argyria resulting from chronic use of a silver protein-containing vasoconstrictor preparation (Coldargan, SigmaPharm, Vienna) for treating allergic rhinitis.88 He consumed 10- 20 ml weekly of Coldargen (drops containing 0.85 mg silver protein) and punch biopsies showed perivascular deposits characteristic of argyria in muscle, skin, and nerves but no other undesirable effects.
In contrast, argyria reported in a fatal case of a 72-year-old woman with carcinoma of stomach and uterus was associated with a deposition of silver sulfide in the basal lamina of her choroidal epithelium.89 This patient had consumed an unknown concentration of Argyrol in nose drops over 2-5 years. Although her tissues were badly autolysed, the authors claimed that silver sulfide granules (70-220 nm) were not membrane bound (lysosomal) and mostly associated with collagen fibrils and stroma of blood vessel walls. Silver granules were not contained within leptomeninges, ependymal cells, or subependymal regions or in the cells of the choroid plexus and minimal amounts present within the area postrema. Elsewhere, florid agyria reported in a 78-year-old woman following chronic administration of overthe-counter nasal drops was associated with widespread silver sulfide deposits in skin, liver, kidney, arteries, pituitary, and choroid plexus.16,90 The authors employed specialized scanning electron microscopy with energy-dispersive spectroscopy (x- ray microanalysis) (EDAX) to characterize the chemical composition of the deposits. A later analysis of this case suggested that silver deposits were predominantly in those parts of the brain having higher regional blood flow and possibly greater permeability to environmental chemicals.8,34,35
A case of myoclonic status eplepticus following repeated oral ingestion of colloidal silver in the form of a homemade "silver drink" was a cause for irreversible neurologic toxicity with poor prognosis.17 Myoclonic status epilepticus has not previously been associated with silver toxicity and this case deserves close attention. The 71-year-old male had used this homeopathic remedy containing a colloidal silver preparation for 4 months, along with an antiandrogen for treating prostatic cancer and various nutritional supplements. He developed paralysis with high levels of silver in blood and CSF, and markedly elevated silver excretion in urine. Plasmaphoresis resulted in a significant reduction in blood and CSF silver concentrations but no improvement in his neurological condition. He lapsed into a coma and his EEG revealed 14-to 18-Hz electropositive centralfrontal polyspikes during myoclonic jerks. He died 5.5 months after the onset of his seizures. Autopsy revealed that his brain was overtly normal, but showed evidence of diffuse Alzheimer type 2 astrocytosis and microglial activation. His nervous system was sampled extensively but showed no evidence of neuronal loss or focal pathology. High-resolution spectroscopy revealed elevated silver deposition in the grey matter of his cerebrum (0.068 g/g wet weight), estimated to be at least twice control levels. Silver deposits were not specifically associated with Alzheimer- related changes and their distribution in aspects of the choroid plexus and BBB is unclear. The spectrometry confirmed silver in the region of the cerebrum but did not exclude the possibility that the vast proportion of this silver was contained within the BBB and not neurological tissues.
More tangible evidence of neurotoxicity resulting from silver nitrate administration was reported in a 59-year-old woman using self-adminstered drops for ulcers of her tongue.15 She developed cutaneous argyria and a manic-depressive psychosis but died 6 years later from a ruptured aortic aneurism. At autopsy silver deposits were identified in skin, mucus membranes, and in many aspects of her central nervous system, notably leptomeninges, choroid plexus, basal ganglia, hypothalamus, substantia nigra, and cerebellum. The silver deposits were lysosomally bound and located specifically within intraparenchymal regions and not in neurons or glial cells. Progressive glial cells changes and cellular gliosis were evident in many areas of the brain. In a similar way, generalized argyrosis, was reported in a 52-year-old man treated with 35 mg of an unidentified silver preparation for 18 years (estimated total intake of 35 g silver).91 This patient died of cardiac failure, but dense silver sulfide deposits were observed in blood vessels, kidney, liver, and choroid plexus at postmortem. Westhofen and Schafer considered that silver exhibits a strong predilection for membrane and neuronal structures in severe cases of argyria with neurological involvement, but that silver sulfide deposition advanced the "progression of clinical disease."92 They used light and electron microscopy to demonstrate silver sulfide granules in the perineurium of peripheral nerves of a severely argyric patient following chronic self-administration of an unidentified silver product. Symptoms of progressive taste and smell disorders, vertigo, and hypesthesia were confirmed by chemosensitivity tests and electrophysiological investigations. Blood and brain silver levels in this patient were not given and it is unclear whether the symptoms (other than argyria) receded following withdrawal of silver therapy. [Silver- induced alterations in zinc metabolism and metallothionein induction may underlie changes in smell and taste perception.93]
SILVER IN WOUND CARE
Silver nitrate (0.5 or 1.0%) was probably the first antibacterial agent adopted in human and veterinary medicine. In Sollemann's Manual of Pharmacology,83 inorganic salts of silver, notably the nitrate, were recorded as being astringent, caustic, and antiseptic but with their local action easily controlled by their precipitation with proteins at the site of application. The cutaneous irritancy seen is directly proportional to concentration, duration of exposure, and the sensitivity of the skin at sites of exposure. The literature is replete with case studies of the use of silver nitrate in treating neonatal eye disease, abrasion of warts, ulcers, and excessive granulations, and in the cauterization of chronic catarrhal infections.8,9 Additionally, silver nitrate has been used as an abortifacient and urethral sterilant. The literature shows considerable inconsistency in the exposure of skin and mucus membranes to silver nitrate, and the extent to which the silver ion absorbed into the circulation is metabolized to internal organs. The toxicology of silver nitrate has been reviewed and limited evidence provided to show that when used under clinical conditions for burn wound antisepsis or wound abridement, it is a potential cause of neurological damage.8 Silver nitrate ionizes readily in the presence of moisture and light energy; the nitrate anion is acidic and largely responsible for the corrosive and toxic effects ascribed to the parent compound.94 Strong silver nitrate solutions are still used to cauterize or remove calluses, warts, and excessive granulations, but application is normally acute and levels of silver ion penetrating to the circulation are exceedingly low. In a fatal case of a 60-year-old man exposed to silver nitrate dressings 8 h daily for 30 days, argyria developed and skin silver levels of 2800 mg/kg and plasma silver of 0.12 mg/L were recorded, but no silver was seen in his brain.95 A similar situation was seen in an 18-year- old man receiving silver nitrate for only 6 days; his plasma silver was 0.12 mg/L and skin silver 1250 mg/L. Neither patient was reported as showing neurological or behavioral change.
Greater risk of argyria and deposition of silver in the central nervous system are anticipated where silver nitrate is used to cauterize the cervix following surgical biopsy or as a means of inducing abortion. The extent to which this practice is performed today is not known, but at concentrations of greater than 5% silver nitrate is highly astringent and irritant to mucosal membranes.96 Silver ion released in the presence of urethra fluids actively binds to cell surfaces and proteins in tissue exudates, but some will be available for diffusion to the peripheral circulation. This diffusion will be promoted through local inflammation and cellular damage through th\e acidity of the nitrate ion. Free silver ion can be expected to precipitate locally in the form of an innocuous argyria. A forensic case is recorded of a German woman given a highly corrosive 7% solution of silver nitrate to induce abortion.65 The woman died with extreme trauma and circulatory failure within 3 hours and at autopsy, silver was found widely distributed throughout her body, including her brain, but her tissues were heavily congested.
Silver sulfadiazine is a white microcrystalline powder with low solubility in water. As a 1% formulation in amphiphilic cream it readily ionized in burn wounds to release Ag+ for antimicrobial purposes.97,98, It is appreciably less corrosive to the skin than silver nitrate although local irritancy is reported.99,100 The reservoir of silver ions accumulating in the wound allows a prolonged release of silver for protein binding and absorption into the circulation. Up to 10% of topically applied silver sulfadiazinde is absorbed from deep partial-thickness burn wounds, particularly in the region of high vascularity. Blood concentrations in patients with greater than 60% total body area burns may rise to 300 g/L, with urinary excretion in the range 100-400 g/L.70 Although early reports emphasise the low risks of silver sulfadiazine toxicity in routine wound care,13 its more extensive use and incorporation in medical devices for long-term implantation have indicated greater caution than at one time considered, including the risk of neurological damage.
Argyria is occasionally observed in patients treated with silver sulfadiazine for severe burns of 60% or greater total body area. In a patient with end-stage renal disease, argyria was associated with a marked elevation of blood silver and deterioration in his mental state.101 Blood silver levels of 291 g/L were associated with greatly raised brain silver (617.3 ng/g, cerebrum; 823.7 ng/g cerebellum wet weight). Hemodialysis, hemofiltration, and plasma exchange were effective in reducing blood silver, but the patient died. Although this case might implicate silver per se as a neurotoxin, the information presented fails to demonstrate silver within neurological tissues or its association with neurodegenerative changes. The study does not preclude infection or immunosupression as a possible cause of fatality. Flammacerium (Solvay Pharmaceuticals) (containing 1 % silver sulfadiazine and 2.2% cerium nitrate) was introduced to alleviate problems of immunosuppression attributable to products forming in burn wounds as a result of thermal energy.102
Sustained Silver-Release Wound Dressings
Numerous sustained silver-release wound dressings have been developed in recent years. The silver content of these dressings ranges from
SILVER IN MEDICAL DEVICES
Medical devices including catheters, bone cements, orthopedic fixation pins, and cardiac prostheses and valves are notoriously prone to bacterial adhesion, colonization, and biofilm formation. Recent advances in silver nanotechnology, materials science, and ion beam silver coating techniques have been increasingly employed in an attempt to engineer out these risks of infection and improve patient comfort and survival.106,107 Intraurethral catheters are of particular interest; in the United States alone physicians implant more than 5 million catheters annually.
Adverse effects attributable to silver including argyria are exceedingly rare. Absorption of silver from catheters implanted following prostatectomy and surgery is not well documented despite recent advances in biotechnology and catheter design.107-110 The hydrophilic coatings impregnated with silver metal (including nanocrystalline forms), silver oxide, and silver sulfadiazine on inner and outer surfaces of catheters release free silver ion in the presence of urethral fluids and exudates for antimicrobial purposes. The concentration of silver required for antimicrobial action in intraurethrine catheters has been estimated to be about 10-9.111 This exceedingly low concentration is achieved by novel technology in which nanocrystalline silver is distributed in polyurethane at 0.8% in a hygroscopic matrix of 450 cm^sup 2^/polyurethane. An Erlanger polyurethane catheter developed to reduce the risks of infection with a mass ratio of 0.6% silver released a total of 7 g/ L silver into physiological saline within 26 hours (0.1 g/L for 30- cm catheter).112 Other estimates suggest that concentrations of silver released from acute care catheters coated with a polymer containing 1012~13 silver nanoparticles ranged from 250 to 350 ng/ cm^sup 2^/day over 10 days. These levels are deemed suitable for antimicrobial action but unlikely to significantly influence blood silver levels or neurophysiological activity. The Cochrane Incontinence Group and the Cochrane Renal Group evaluated the management of silver coated catheters for short-term use.113 They emphasized the cost-benefit ratio of the new silver technology in controlling bacterial adhesions and biofilm formation as a cause of fatality but have failed to recognize the potential toxic implications of the silver released. Cymet questioned whether silver "alloy" catheters might increase the inherent risks of systemic argyria and risks of silver toxicity,114 but no satisfactory responses have been received. An unseen Russian study did report urethral argyria following use of silver nitrate,115 but further details including blood silver levels are not available. Cymet emphasized that argyremia could be accurately monitored and that this information is clinically important, if merely to ascertain whether levels were consistent with permitted limits in the United States and elsewhere.114 Saint conducted 5 randomized trials employing silvered catheters for short-term use but failed to observe evidence of local argyria,116 but acknowledged that the risk did exist in long term urethral drainage with the consequence of silver deposition in internal organs.
Clinical studies with hemodialysis and intravascular catheters have similarly failed to produce evidence of silver toxicity or brain involvement even though silver ion released directly into the circulation would be expected to lead to increased plasmabound silver and greater tissue deposits. Tobin and Bambauer reviewed clinical studies designed to assess the efficacy and biocompatibility of silver-coated dialysis catheters.117 They noted that blood silver levels increase from a mean of 1.3 to 6.9 g/L for acute catheters and from 3.4 to 19.6 g/L for long-term catheters, in each case with plasma levels returning to normal on removal of the catheters. No data was provided for tissue silver deposition or evidence of toxic side effects. Maki et al. evaluated triple-lumen catheters (ARROWgard) designed for intravenous insertion.118 They contained 0.70 mg silver sulfadiazine and evoked a mild local erythema at insertion points and plasma silver levels of 45-73 ng/ mL in the 12 patients tested. These incredibly low concentrations are of minimal toxicological significance.
Mechanical heart valves containing silver have been associated with greater hazard than in-dwelling catheters. Thus, St. Jude Medical applied a silver coating to the sewing cuff of its Silzone range of heart valves with the objective of reducing risks of infective endocarditis. Over 30,000 of these valves were distributed after 1997, but the valve was withdrawn through thrombo-embolytic complications.119 In one such case, a St. Jude Medical Silzone valve was implanted into a 72-year-old woman suffering from mitral valve disease. Her fatality was attributed to chronic inflammatory disease but the implication of silver in this case is unclear. Experimental studies in sheep implanted with the Silzone valve showed plasma silver of 40 ppb within 10 days and mean brain silver of 4.32 0.28 g/g dry tissue weight after up to 20 days.117 Liver concentrations were a lot higher at 16.75 5.18 g/g, but changes were not reported in other tissues.
Acupuncture needles should be included among the medical devices containing silver. Their use has been associated with macular or widespread argyric changes with occasional neurological involvement.121 This "Hari" therapy conducted in Japan for many centuries for relief of fatigue and headache involves long-term intracutaneous insertion of silver-gold needles. A 21-year-old Japanese woman given this Hari therapy over 2 years to relieve asthma developed a profound macular argyria and chrysiasis on her neck, face, and chest.77 Minute silver-gold particles (20-60 nm) were deposited along the outer edge of basement membranes of blood vessel and sweat glands, and in lines around but not in nerve fibers. Small amounts of silver-gold were evident also in basement membrane collagen associated with myelinated and nonmyelinated nerves but nerve damage was not reported. Silver acupuncture needles are used in treating for sterility and general fatigue conditions, but have argyric implications.78 In a particularly severe case of a woman using up to 2500 needles over 13 years, macular argyria with irregularly shaped silver sulfide granules of 40-500 nm diameter distributed mainly in extracellular dermal sites and around nerve fibers and elastic tissues but overt nerological changes were not recorded.\Suggestions that argyria developing through implantation of acupuncture needles might interfere with tissue function have not been substantiated.8 Although discolorations of the face and body have been associated with acupuncture needles containing up to 69% silver, brain involvement has not been reported.78,121 Blood silver levels are not known in these cases but are expected to be well below toxic range of 50-500 mg/kg body weight that has been associated with abnormal encephalographic changes and brain scan findings.122
Antismoking remedies containing silver are included in medical devices, although there is no evidence that they present a neurotoxic hazard. An example documented is of a healthy 47-year- old female patient who showed profound argyria following excessive oral dosage of silver acetate as an antismoking remedy for 6 months.123 Silver-protein complexes are readily absorbed through the buccal mucosa, and in this patient analyses using neutron activation showed her total body silver burden to be 6.4 g, of which only 1.8% was retained within the blood. Silver absorption and retention analyzed by radioactive trace administration showed that after an initial decrease, 18% silver tracer remained in the body for up to 30 weeks, but the amount deposited in her brain is not known. She remained in overt good health throughout the observation.
Strong evidence implicating silver as a cause of neurological toxicity and behavioral changes derives from use of a silver as an antimicrobial agent in arthroplasty cement. Bone cement containing an unknown quantity of silver was used to anchor a Christiansen prosthesis in a 78-year-old woman.18,19 Five years after insertion of the prosthesis, the patient became unstable and exhibited muscle weakness in her left leg. Electromyography revealed no activity in those muscles innervated by her left tibial and femoral nerves and a total paralysis of her quadriceps muscle. This was related to exceptionally high levels of silver in her hip joint fluid (956 nmol/ L), blood (58 nmol/L), and biopsies of acetabulum. Biopsies of soft tissue revealed granules characteristic of argyria in the region of elastic fibers and in numerous macrophages, but not in peripheral nerves. The right leg was entirely normal. The prosthesis was removed and blood silver levels declined to 15 nmol/L within 12 months and conducting activity was restored fully to her tibial muscle and partially to her femoral nerve. Motor activity was also improved in her quadriceps muscle. The patient was closely monitored over 10 yearss by which time the paralysis had receded and the patient was able to walk unaided. Although transitory electrophysiological changes were observed several months after removal of the prosthesis, neurotoxic action of silver was not substantiated.
All medical devices, wound dressings, and medicaments containing silver or coated with a silver complex to achieve antimicrobial action release silver ion in the presence of moisture, body fluids, and exudates.103 A proportion of this biologically active silver will be absorbed into the body in a carrier-mediated process; it enters the bloodstream and is deposited in soft tissues.1,12,71 Clinical studies in patients with severe argyria or with purported silver "poisoning" have consistently failed fully establish the ability of silver to pass the blood-brain barrier or evoke irreversible damage at any site in the central nervous system. Silver does not satisfy the criteria for neurotoxins set out in Table 1, although several equivocal issues remain. Not least of these is whether and to what extent silver is able to penetrate the blood-brain barriers in any part of the central nervous system to precipitate in the form of argyria within neurons or glial cell populations to evoke pathological or behavioral changes. Evidence reviewed shows that silver is absorbed into the circulation following inhalation, ingestion, and through use of various medicated devices, but levels or argyremia tend to fall rapidly as the silver is disseminated throughout the body. Prolonged exposure to silver occupationally or therapeutically can lead to manifestations of argyria and argyrosis with lysosomally bound deposits of silver sulfide or silver selenide occurring in liver, kidney, vascular tissues, and in connective tissues of the bloodbrain barrier and skin. Silver ion exhibits a strong binding to sulfhydryl (-SH) moieties in collagens of connective tissues and basement membranes and has frequently been observed in the region of peripheral nerves (myelinated and nonmyelinated) but not within neurological tissues.
Considerable evidence points to the efficiency of the bloodbrain barriers in various parts of the central system in accumulating silver, thereby mitigating whatever toxic influence the metal might have on nervous tissue. This wider survey of silver provides further evidence in support of Zheng's hypothesis that silver should be classified alongside iron, zinc, and gold as a "sequestered choroid plexus toxicant," namely, a metal that is sequestered by the tissue and not associated with pathological changes or pathophysiological consequences, including changes in the blood-CSF barrier.29,33
Evidence that silver can penentrate the blood-brain barriers to cause transitory physiological change is provided largely by experimental studies in rodents dosed with high levels of ionizable silver compound in drinking water or through intravenous injection. Thus Rungby and Danscher have demonstrated silver throughout the central nervous system but have failed to identify frank pathological changes.46,59,62 They based their observations entirely on autometallographic demonstration of silver in tissues. This technique employs histochemical methods in which silver impregnation is used as an aid to identifying neuropathological changes.124,125 It relies on reduction of silver residues in tissue sections by potassium cyanide and development of metal deposits using a developer containing silver as in photographic processing. The validity of the method might be questioned, not only in the suitability of a silver-containing reagent to demonstrate silver deposits in a tissue, but in the resistance of the two principal silver precipitates seen in argyria, that is, silver sulfide and silver selenide, to reduction with potassium cyanide. The observations are entirely contradictory to the results of several earlier studies using a similar experimental model, and in which silver was contained within the BBB and acted in the form of an intravital dye like trypan blue.50-55
Lack of frank evidence for the neurotoxic action of silver may in part be explained by the deposition of inert precipitates of silver sulfide or silver selenide in lysosomal vesicles. Electron microscopy has demonstrated the membrane-bound deposits in endothelial cells of the BBB, especially the choroid plexus in cases of generalized argyria.15,59,92 Previous studies have comprehensively established that lysosomally bound silver in liver and kidney (14 g Ag/g wet tissue weight) is not associated with functional abnormalities and that there is no correlation between silver levels in the tissues and circulation.12 Wang et al. used 20- 50% silver sulfadiazine to treat patients with severe burns and reported argyremia of >300 g/L, yet reported normal tissue function.1,12 Early alterations in liver metabolizing enzymes reported in burn wound patients treated with silver sulfadiazine normalized and were not associated with changes in long-term exposure.126,127
Silver absorbed into soft tissues induces and binds the cysteine- rich metallothioneins, which exibit a regulator and cytoprotective role.128,129 In the skin at least, metallothioneins play an instrumental role in the metabolism of silver in normal and damaged tissues and possibly contribute to the action of silver in wound repair.130,131 Metallothioneins are present in all living cells and have a unique structure relating to their ability to sequester and bind metals like zinc and silver. Four isoforms of metallothionein have been identified so far, including MT-I and MT-II, which are expressed in a variety of tissues including the brain, MT-III, which is predominantly located in the brain, and MT-IV, which occurs in squamous epithelia like the tongue. Metallothioneins modulate three main processes in mammalian cells:
1. Release of mediators (e.g., hydroxyl radical, nitric oxide).
3. Binding and exchange of heavy metals (zinc, cadmium, copper, silver, etc.).129,112
Much remains to be learned regarding the primary function of MTs, including their regulatory role in tissue repair and regeneration. They are present to a greater or lesser extent in all regions of the brain and spinal cord and have been implicated in nerve regeneration and Alzheimer's disease and other neurological conditions.133 Experimental evidence suggests that MT-III expression is altered in brain injury and that regeneration of peripheral nerves is advanced in MT-III knock-out mice.134 Metallothioneins, notably MT-III, are predominantly expressed in zinc-containing neurons of the hippocampus but absent from glial elements.133 Although silver displaces zinc from zinc-metallothionein complexes in the skin, it is unclear at the moment whether it can displace zinc bound in MT- III complexes in the hippocampus or other regions of the brain leading to increased extracellular zinc as required for RNA and DNA synthetases and other essential enzymes. If, as suggested by experimental studies in the rat,46 silver does pass the blood-brain barrier to be sequestered by neurons and glial cells, it should be assumed that it will induce MT synthesis in neural tissues and initiate certain stress symptoms and pathophysiological processes associated with excess zinc.93 At the moment, it is unclear whether silver induces MT-III or to what extent the pro\tein acts as a cytoprotectant in the central nervous system.
The observation of transient paralysis in the woman treated with silver-containing bone cement to anchor a Christianson prosthesis is difficult to explain in terms of silver toxicity.18,19 Despite the electrophysiological evidence of functional change, reduced conduction was not associated with silver deposits in nerves. The high concentrations of silver present in her hip joint fluid were probably strongly bound to proteins and hence inert. Silver deposition was reported in connective tissues around the hip joint and in the perineurium but the physiological implications of this are not known. Silver is not lipophilic and is not known to penetrate myelin sheaths.29 Further neurophysiological studies on the effects of extracellular silver on nerve conduction may shed some light on the clinical significance of this case.
The present review complements my earlier observations that silver is without serious toxic risk to any organ system in the human body. However, recommendations for safe reference values are extremely difficult to make since silver deposits will accumulate temporarily in most soft tissue, to be eliminated eventually via the kidney or liver without toxic implications.1,12 Silver sulfide/ selanide deposits in the skin and eye may be long-lasting but have not been associated with toxicity or ill health.13,105 Silver deposits in the eye may obscure night vision, but this has not been associated with neurological changes.80,81 A small number of people experience delayed hypersensitivity to silver, and they should avoid silver ingestion, inhalation, and use of silver-containing medical devices.
Permissible exposure limits and threshold limit values set by the American Conference of Governmental Industrial Hygienists (ACGIH),135,136 the National Institute for Occupational Safety and Health (NIOSH), and the European Commission recommended threshold limit values (TLV) of 0.01 mg/m^sup 3^ for metallic silver (including nanoparticulate forms) and silver compounds. According to Drake and Hazelwood the AGCIH recognized the different outcomes from exposure to soluble and/or insoluble silver compounds, rationalizing that soluble compounds are more likely to cause argyria and associated effects than does the "dust or fume of metallic silver."135,136 As such the recommended TLV for metallic silver exposure is set at 0.01 mg/m^sup 3^ and that for soluble silver compounds is 0.1 mg/m^sup 3^.136
I appreciate the constructive advice from Professor C. Kennard, Deputy Principal, Imperial College Faculty of Medicine, London, and express thanks to Professor A. Dayan, formerly director of DHSS Department of Toxicology, St. Bartholomew's Hospital Medical School, London.
1. Wan, A.T., Conyers, RAJ., Coombs, C.J., and Masterton, J.P. (1991). Determination of silver in blood, urine and tissues of volunteers and burn patients. Clin. Chem. 37:1683.
2. Lansdown, A.B.G. (1995). Physiological and toxicological changes in the skin resulting from the action and interaction of metal ions. CRC Crit. Rev. Toxicol. 25:397.
3. Lansdown, A.B.G. (2002). Silver 2: Toxicity in mammals and how its products aid wound repair. J. Wound Care 11:173.
4. White, R.J. (1999). A historical review of silver in wound management. Br. J. Nurs.S3-S8.
5. Klasen, H.J. (2000). A historical review of the history of silver in burns. Burns 26:117.
6. Hambidge, A. (2001). Reviewing efficacy of alternative water treatment techniques. Health Estate 55:23.
7. Lansdown, A.B.G., Williams, A., Chandler, S., and Benfield, S. (2005). Silver absorption and antibacterial efficacy of silver dressings. J. Wound Care 14/4:155.
8. Humphreys, S.D.M., and Routledge, P.A. (1998). The toxicology of silver nitrate. Adverse Drug React. Toxicol. Rev. 17:115.
9. Fung, M.C., and Bowen, D.L. (1996). Silver products for medical indications: risk-benefit assessment. Clin. Toxicol. 34:119.
10. Boosalis, M.G., McCall, J.T., Ahrenholtz, D.H. et al. (1982). Serum and urinary silver levels in thermal injury patients. Surgery 101:40.
11. Pariser, R.J. (1978). Generalised argyria: clinicopathologic features and histochemical studies. Arch. Dermatol. 114:373.
12. Coombs, C.J., Wan, A.T., Masterton, J.P., et al. (1992). Do burn patients have a silver lining? Burns 18:179.
13. Lansdown, A.B.G., and Williams, A. (2004). How safe is silver in wound care? J. Wound Care 13:131.
14. Bleehan, S.S., Gold, D.J., Harrington, C.I., et al. (1981). Occupational argyria: Light and electron microscopic studies and X- ray microanalysis. Br. J. Dermatol. 104:19.
15. Deitl, H.W., Anzil, A.P., and Mehraein, P. (1984). Brain involvement in generalised argyria. Clin. Neuropathol. 3:32.
16. Landas, S., Bonsib, S.M., Ellerbroek, R., and Fischer, J. (1986). Argyria: Microanalytic-morphologic correlation using paraffin embedded tissue. Ultrastruct. Pathol. 10:129.
17. Mirsattari, S.M., Hammond, A.R., Sharpe, M.D., et al. (2004). Myoclonic status epilepticus following repeated injection of colloidal silver. Neurology 62:1408.
18. Vik, H., Andersen, K.J., Juhlshamn, K., and Todnem, K. (1985). Neuropathy caused by silver absorption from arthroplasty cement. Lancet 1:872.
19. Sudmann, E., Vik, H., Rait, M., et al. (1994). Systemic and local silver accumulation after total hip replacement using silver- impregnated bone cement. Med. Prog. Technol. 20:179-184.
20. Roper, W.L. (1990). Toxicological profile for silver. Agency for Toxic Substances and Disease Registry, U.S. Public Health Service, Atlanta, GA.
21. Danscher, G., and Rungby, J. (1986). Differentiation of histochemically visualised mercury and silver. Histochemistry 18:109.
22. Stoltenberg, M., Juhl, S., Poulsen, E.H., and Ernst, E. (1994). Autometallographic detection of silver in hypothalamic neurons in rats exposed to silver nitrate. J. Appl. Tox