There has been a vast increase in the last 5 years in the number of available silver-containing dressings. Their use has seen a corresponding rise in the number of publications referring to silver and its potential benefits. It is important that recognition is also given to the potential pitfalls of use, particularly in relation to toxicity. These factors have been recently reviewed. This article will explore what happens once the body absorbs silver, discuss the relevance of the carrier dressing to efficacy of silver, and review the clinical relevance of microbial kill time.

       In recent years, use of silver in medical healthcare devices has seen a vast increase. This increase has been largely dominated by wound dressings. Silver sulfadiazine, which has been available for approximately 40 years, provides broad-spectrum antimicrobial activity and has been widely used, particularly as a topical cream to manage burn infection. In the last 5 years, the number of available silver-containing dressings has increased. These dressings are used primarily on chronic wounds and in clinical practice are regularly applied for periods of time up to and in excess of 4 weeks. As silver dressings may be used in clinical situations other than as a temporary option, it is important that potential toxicity be considered, particularly in relation to the type and amount of silver, and as regards the risk of selecting for bacterial resistance. These factors have been recently reviewed.¹
       This article will consider what happens once the body absorbs silver and discuss the levels of silver required to exert a toxic effect on bacteria. In addition, it will explore the relevance of the carrier dressing to efficacy of silver and the clinical relevance of microbial kill time. It is preferable that any testimony made in respect of silver should be clinically relevant. In order to do this, data should be drawn from clinical (in-vivo) studies wherever possible. Unfortunately, few such studies exist, so much data will be drawn from available in-vitro, ex-vivo, and animal studies.
       Metabolic fate of topical silver. Silver, as a component of wound dressings, antibiotic cream, and first-aid plasters, comes into contact with intact skin and breached skin on an increasingly regular basis. The penchant for silver as an antimicrobial has seen it incorporated into simple adhesive dressings for minor cuts and abrasions and is no longer reserved for “serious” wound management. Potential repercussions associated with these applications need to be acknowledged and explored. Apart from increasing the risk of contact dermatitis and selecting for resistance, there will be concerns about possible systemic and cutaneous toxicity. The interaction of metallic silver with intact skin does not cause any detectable increase in blood levels and is not of great toxicological interest. However, the recent increase in the use of silver-based wound treatments raises some concerns about the systemic effects of silver and warrants a toxicological review. Several factors influence the capacity of a metal to produce either local or systemic toxic effects. These factors include 1) the degree of absorption as influenced by solubility of the metal or its compounds, 2) the ability to bind to biological sites, and 3) the degree to which the metal complexes are sequestered, metabolized, and ultimately excreted.
       Silver is applied to open, dermal wounds in the form of inorganic silver salts (eg, silver nitrate), metallic silver, or as organic compounds, such as silver sulfadiazine (SSD).² The chemical nature of the applied silver will influence its absorption, distribution, and metabolism. Studies on “background” or environmental silver levels in human tissues have been conducted.^3,5,6 Normal silver concentration is very low. Concentrations of silver in blood, urine, liver, and kidney of subjects without industrial or medicinal exposure are < 2.3 µg/L, 2 µg/day, 0.05 µg/g wet tissue, and 0.05 µg/g wet tissue, respectively.³ Reference values quoted by Guys and St. Thomas’ Hospital Toxicology Laboratory (London, UK) are < 3 nmol/L (approx. 0.32 µg/L) for blood and < 8 nmol/L (~ 0.86 µg/L) for urine.⁴ Most studies on the metabolic fate of topically applied silver have been in patients with burns treated with SSD cream.^5,6 In SSD cream-treated burn patients, plasma concentrations may be as great as 50 µg/L within 6 hours of treatment, dependent upon the area that was treated, and can reach a maximum of 310 µg/L. Silver in urine may reach a maximum of 400 µg/day. After absorption, silver has been found in various tissues. Silver concentrations in a burn patient who died of renal failure after 8 days of treatment were 970 µg/g, 14 µg/g, and 0.2 µg/g wet tissue in the cornea, liver, and kidney, respectively.³ This equates to liver concentration that is 280 times background. Renal toxicity from silver has been reported after topical application, leading the authors to conclude that topical SSD should not be used for long periods on extensive wounds.⁷ Lansdown and Williams⁸ acknowledge that the use of topical silver in burns and chronic ulcers can lead to systemic absorption and subsequent deposits in the organs; however, they conclude that the risk is low. Studies on the fate of SSD using radiolabeled silver (Ag¹¹⁰) showed that label accumulation occurred in superficial layers and in a short time period after exposure (2–8 hours)—clearance was complete in 28 days.⁹ This indicates that silver binds superficially and has low absorption from single application. Few reported data on systemic absorption of silver from sources other than topical SSD exist; thus, it must be assumed that under normal conditions of use, the modern silver-containing wound dressings are safe in this respect.¹⁰ Chen et al¹¹ have reported increased serum and urine silver levels and mild hepatic dysfunction in patients with burns treated with a silver dressing but concluded that it was safe on small to medium partial-thickness burns. However, Trop et al¹² reported silver-related hepatotoxicity and argyria-like symptoms in a case of a silver-coated dressing used on a child with 30% total body surface area burns. Elevated plasma and urine silver levels of 107 µg/kg (~ 104 µg/L) and 28 µg/kg (~ 27 µg /L) were measured, as were elevated liver enzymes (AST, ALT, and GGT). These abnormalities eventually resolved after cessation of silver treatment.
       The limits of exposure to silver in industrial conditions have been reviewed and documented.¹³ Silver compounds and metal ionize, largely to the monovalent cation Ag+; there are other cations, but these are very reactive and short-lived. Ag+ has a high affinity for thiol groups and binds to reduced glutathione¹⁴; it also binds to the amino-, sulphydryl, carboxyl, and phosphate groups of nucleic acids.¹⁵ Reduced glutathione is important in erythrocyte function and in the elimination of organic peroxides,¹⁴ and any chemical alteration to nucleic acids is likely to result in transcription errors.
       In-vitro, ex-vivo, and animal evaluation of silver toxicity. The cytotoxicity of topically applied agents, such as antimicrobials, has been evaluated using skin cells in vitro,^16–19 reconstituted human epithelium (RHE),²⁰ and in grafted skin substitutes.²¹ The risk-benefit of cytotoxicity versus antimicrobial activity for agents applied topically to wounds has been discussed.²² The toxicity of silver in cells and tissues has been assessed using silver nitrate, silver sulfadiazine, and silver dressings. In a study on cultured human dermal fibroblasts, silver nitrate exposure to various quantities of fetal calf serum (resembling physiological conditions) produced cytotoxic effects at 8.2 nmol/L after 8 and 24 hours.¹⁸ In similar experiments using monolayer cultures of 3T3 fibroblasts and keratinocytes from surgical discards, silver from silver nitrate and from a silver dressing was found to be cytotoxic (lethal to silver nitrate at 50 x 10^-4% after 3 hours as well as the dressing). The same silver dressing was evaluated for cytotoxicity using cultured skin substitutes in vitro and in vivo after grafting.²¹ Results showed the dressing was cytotoxic in vitro within 1 day, but the in-vivo dressing was not after 1 week. These results suggest that i cytotoxicity is more sensitive than in vivo, as there is no means of reducing toxicity via blood circulation, tissue reservoir, metabolism, or by dilution effects. It is justifiable to conclude that in-vivo or ex-vivo studies are more likely to reflect the possible toxicity in routine clinical use.
       The oligodynamic nature of silver. Bacteria in wounds, notably chronic wounds, exist as both planktonic and sessile organisms. The latter are attached to a surface (eg, biofilm form) that is postulated, but not yet confirmed, as a feature of chronic wounds.²³ Bacteria behave differently in each of these 2 forms. This behavior becomes relevant to bioburden control measures when the 2 forms exist contemporaneously in the wound. Bacteria in planktonic form are freely available to topical antimicrobial agents, whereas in biofilms, bacteria are less susceptible.²⁴ The antimicrobial activity of silver has been known for many years, and numerous publications report its action against a wide variety of organisms in vitro.²⁵ It is generally accepted that silver is active as the monovalent cation Ag+ and that this species is active at low concentrations (parts per billion [ppb] or µg/L, to parts per million [ppm] or mg/L) in aqueous solutions.²⁶ The term oligodynamic, meaning active in small quantities, has been used in this context in much research over the last century.^27–31 In a review directed at SARS (severe acute respiratory syndrome), Rentz³⁰ referred to the work of von Näegeli²⁷ who found Ag+ to be an active biocide at concentrations between 9.2 x 10-⁹ and 5.5 x 10-⁶ M, (ie, 9.2 ppb and 5.5 ppm). Rentz³⁰ cited a 1987 study by Cliver³² that reported Ag+ was active at 250 ppb in 2 hours. The efficacy of silver ion disinfection has been illustrated by the following calculation:
       At a concentration of 10⁴ cells/mL and 50 ppb (4.7 x 10-⁷ mol/L) metal ions, there are approximately 2.8 x 10¹⁰ metal ions per cell.^25,32
       This calculation represents a typical bacterial concentration in wound exudate and a “low” level of silver dissolution from a silver-containing dressing. However, exudate will have an influence on silver ion activity by virtue of its anion content (eg, Cl-), which bind the Ag+ ion. The effects of protein on the binding and bioavailability of silver ions have been investigated using an oral bacterium, Porphyromonas gingivalis³³; the frequent presence of silver in the mouth from dental amalgam³⁴ is known to select for resistance and casts some doubt on the validity of these observations.³⁵ According to Bechert et al,²⁹ “the oligodynamic activity of silver ions is not reduced by pre-incubation with albumin and fibrinogen.” Currently, not much information is available that relates to the effects of silver on wound clinical isolates in the presence of common anions and protein (ie, an exudate equivalent environment). However, Bowler et al³⁶ addressed this situation and challenged a silver dressing with clinical isolates tested in a simulated wound fluid. Their findings suggest that the silver-containing dressing is likely to provide a barrier to infection.
       The authors are unaware of any published studies on the mutant selection window (MSW) and mutant prevention concentration (MPC) for silver.^37,38 The MSW has been developed using antibiotics, so its value in antiseptic studies is a matter for conjecture. The principles are nevertheless intriguing and further research is necessary. Mutant selection window (ie, 2 concentrations of antimicrobials, usually antibiotics), at the lower level, blocks the majority of susceptible bacteria growth. The upper limit of the window is the concentration of antimicrobials that blocks the growth of the least susceptible bacteria. Resistance is rarely expected to develop when drug concentrations are kept above the upper boundary of the MSW. This expectation led to the upper boundary being designated as the MPC.
       The results of such studies would be highly desirable before making assertions on the likelihood of selection for resistance through the use of dressings delivering different amounts of silver in everyday clinical practice.³⁹
       While claims have been made that a rapid kill rate is essential in order to avoid resistance and biofilm formation,⁴⁰ there is no supporting MSW data. It is known that silver has the capacity to disrupt the biofilm matrix at a low dose (50 ppb).⁴¹ It could, therefore, be argued that such claims are made more for commercial advantage than for the advancement of clinical intervention.
       Silver ions at low ppb concentrations are effective antibacterial agents against most planktonic bacteria.²⁶ This level (50 ppb) has also been found effective in destabilizing biofilms of Staphylococcus epidermidis in vitro.⁴¹ In an in-vivo study using a biofilm-forming Staphylococcus epidermidis, Illingworth et al⁴² showed that a silver-coated heart valve cuff exerted bactericidal activity. This implies that the (probable) biofilm formation on the cuff does not protect from silver ions. Thus, it is reasonable to conclude that in the case of silver ions, oligodynamic equates to bactericidal activity at nanomolar (ppb) concentrations.

Dressing Association with the Wound Bed

       Sibbald⁴³ highlighted the important relationship of effective wound bed preparation and the management of wound infection through use of antimicrobial agents. He stated that the selection of any product should account for microbial sensitivity, low allergenicity, and low cellular toxicity and should not be a systemic agent. These important considerations should not be disputed. Irrespective of the type of antimicrobial silver used in any medical device (eg, salts or metallic), the form of silver delivered to the wound should remain consistent (ie, Ag+) and not change irrespective of the carrier dressing. However, it is generally recognized that silver efficacy is influenced by the amount of silver and its availability, which are dependent on the chosen product. However, Parsons et al⁴⁴ stated that efficacy is unrelated to the total amount of silver in the dressing. One additional factor that impinges on antiseptic efficacy and is not related to the form of silver used or the dosage but should not be overlooked is the ability of the carrier dressing to conform to the wound bed. High conformability helps ensure that areas of noncontact between the dressing and the wound bed are minimized thus reducing the formation of voids (dead space) where bacteria may flourish. Dead space has been identified as an impediment to successful wound healing and efforts should be made to avoid their occurrence.⁴⁵ A dressing that gels in contact with wound fluid by internally binding water is more likely to achieve high conformability with the wound bed than one that does not conform and is relatively inflexible. Fibrous dressings maintain an excellent absorptive capacity yet can be removed from the wound atraumatically while remaining intact and retaining the additional benefit of avoiding dead space formation in the wound.^46,47 Additionally, the ability of a dressing to maintain a high tensile strength while binding water would seem to be advantageous.
       The value of fibrous dressings in wound management has been enhanced by the silver ion incorporation, resulting in an absorptive, antimicrobial dressing.⁴⁸ Such dressings are known to achieve and maintain intimate contact with the wound bed—this is deemed advantageous in wound healing because these dressings avoid creating dead space. Avoiding dead space creation at the wound bed/dressing interface can be demonstrated through in-vitro methods.⁴⁹ However, the clinical benefit obtained through the appropriate use of such dressings is not in the absolute bactericidal impact (ie, achieving sterility) but in the reduction of bacterial bioburden to a level where the host immune response can respond and regain control.
       Topical antimicrobials have an important role to play in managing wound bioburden,⁵⁰ and product selection should take into account not just those issues related to antimicrobial activity, such as sensitivity, allergenicity, and cellular toxicity, but should consider the physical relationship of the carrier vehicle (dressing) to the wound bed. High conformability will help ensure the effectiveness of the antimicrobial component at the dressing/wound bed interface.⁴⁹ Not only may the delivery of silver ions to the wound bed be enhanced through the close proximity of the dressing but wound bed bioburden may also be reduced through bacterial sequestration. Walker et al⁵¹ demonstrated the value of sequestration in managing bacterial pathogens using scanning electron microscopy. Their investigations showed that as the fibrous dressing became hydrated and formed a cohesive gel, bacteria absorbed into the dressing matrix were immobilized and retained within the gel structure. This dressing property immobilizes the bacteria and compliments the bactericidal activity of the silver ions by reducing the wound bed bioburden through the mechanism of sequestration.
       An issue related to silver and antimicrobial efficacy that has acquired a degree of attention is time to kill. Diametrically opposite views may be found in the literature regarding the relevance of time to kill in the control of wound pathogens.^40,44 A recent publication listed dressing products with their comparative silver content⁵² and should not be interpreted as though a higher level of silver leads to a shorter kill time or that silver efficacy is primarily time/dose related. In a review of silver biocides in dressings, Silver et al⁵³ stated that conclusions as to one product being more effective than another during in-vitro tests have little bearing on the efficacy of these products in human medicine. Choice of an apposite antibacterial dressing should be based on the clinical results (impact) and not on any single laboratory limitation.^44,53 Although the comparative tabular approach is helpful (it provides a hierarchy of silver content by product), the problem arises that much supporting evidence is usually generated following in-vitro testing, and the clinical relevance of such data requires exploration. It also should be put into context of resistance selection and the MSW. Some authors have published data related to a single dressing type extolling the apparent benefits of rapid time to kill. So this is not viewed as an aspect of dressing performance where products are differentiated for marketing purposes, it must be established whether or not time to kill due to topically applied antimicrobials is clinically relevant. The latter may be applicable where such studies use type cultures and not clinical isolates. Of equal or perhaps greater importance is that the test model must represent clinical use (eg, saline, serum, exudate, dead cells).
       Furthermore, the balance between antimicrobial dressings and systemic antibiotics also needs to be established—when is a topical antimicrobial dressing adequate and when should topical treatment be supplemented with systemic antibiotics?