A Discussion of Silver as an Antimicrobial Agent: Alleviating the Confusion
Silver used as an antimicrobial agent has an impressive history. Silver (as well as copper) is used as a disinfectant in hospital and hotel water sanitization systems. It is used in the food industry in chicken farming and oyster cleaning to inhibit bacterial and fungal growth and in the space program to sterilize recycled water aboard the MIR space station and the NASA space shuttle.1 Although not employed in allopathic mainstream medicine, colloidal silver taken orally has been used as a complementary health aid, the overuse of which in some cases has lead to argyria (also referred to as “blue skin disease,” where the reduced silver is deposited in dermal cells. This explains how the term “blue blood” originated). Silver also is used as a preservative in cosmetics and toiletries and has been incorporated into plastics of various forms to protect against microbial contamination.
Silver has a long medicinal history, going as far back as ancient Greece and Rome, when silver coins dropped into water served as a disinfectant.2 As early as 1884, Crede, a German obstetrician, used a 1% silver nitrate solution to eliminate blindness in newborns caused by post-partum infection.3 In 1887, von Behring used the same compound to treat typhoid and anthrax. In 1964, Moyer4 first used silver in the burn arena; 4 years later, Fox5 brought silver sulfadiazine (SSD) to medicine. Silver-coated catheters have been used to stem biofilm formation and prevent infection.6
Over the last 3 to 5 years, an influx of “next generation” silver-based antimicrobial dressings has been introduced into the medical arena. A review of the literature, performed over the last 3 years using references going back 30+ years to the present (key words included resistance, strains, kill rate, dissociation, biofilm) along with time spent with clinicians, reveals specific areas of confusion with respect to silver as an antimicrobial agent — specifically, silver resistance, silver dissociation, silver concentration, strains of pathogens, rate of kill, and cytotoxicity of silver. These topics will be reviewed using available publications from the peer-reviewed literature, professional presentations, and independent data.
Although some authors downplay silver resistance, the majority of authors place a fair amount of emphasis on this issue.7-31 Ovington7 writes, “Although silver has been used effectively for centuries as an antimicrobial agent, we should not assume that bacterial resistance to silver will never become a problem. The genetic basis of bacterial resistance to silver has been identified in specific bacterial isolates and closely related gene sequences have been identified in other bacteria. With silver being used more extensively in both environmental and clinical applications — particularly in wound management — the threat of resistance is not unlikely and urges for prudent use of topical antimicrobial products of any type.” In another paper, Ovington32 theorizes that resistance to silver (an antiseptic) is possible, but is less common than resistance to antibacterial agents.
Based on their own and additional research,8,9,11,12,15,17,19,21,23-26,29,33 Gupta and Silver,8 in a literature review/commentary, both state that “silver resistance is important to monitor, because modern technology has developed a wide range of products that depend on silver as a key microcidal component.” According to Levy,9 “the widespread use of silver could result in more bacteria developing resistance, analogous to the emergence of antibiotic- and biocide-resistant bacteria.”
Historically, knowledge of silver resistance has been relatively limited in the medical arena. One reasonable explanation is the fact that silver nitrate and SSD have been the only options available in wound care. A 0.5% aqueous solution of silver nitrate (soaked gauze applied once every 4 hours) and a 1% SSD cream (1/16-inch thickness, applied QID or BID) deliver in excess of 3,000 parts per million (ppm) of silver ion (Ag+) over 24 hours. Most data (depending on the strain tested) show that this level results in a rapid bacterial kill rate (bactericidal in the extreme); thus, minimizing the chances for survival, mutation, and resistance. The newer “advanced” dressings, some of which deliver low levels of silver, have elicited much discussion. Li et al11 were able to develop silver-resistant pathogens by growing microbes in the lab and then introducing sublethal (ie, bacteriostatic) levels of silver using a “multiple step exposure protocol.” In the study, a strain of Escherichia coli was developed that was resistant to silver of a concentration greater than 1,000 ppm. Based on this study, it appears that the use of sublethal doses of silver could, in fact, result in resistant strains. However, it was found that it is not possible to develop resistance in a single step. If an adequate level of silver was delivered rapidly, the bacteria were quickly killed, decreasing the chances for mutation and resistance. Because silver offers numerous modes of antimicrobial activity, a series of mutations in a single generation that result in resistance to all of silver’s modes of action is highly unlikely. However, after multiple generations, resistance seems possible. Li et al11 and others11,12,17-21,23,24,26,29,30,31,33,34 have demonstrated that plasmids within the pathogens and the pathogens’ own DNA provide the means to achieve silver resistance.
Clinicians should be cognizant of the existence of silver-resistant pathogens and that a potential mechanism of promoting resistance has been demonstrated. Silver-resistant bacterial strains have been isolated from both clinical and environmental sources. Examples include strains of Acinetobacter baumannii, E. coli, Enterobacter cloacae, Klebsiella pneumoniae, Pseudomonas aeruginosa, Pseudomonas stutzeri, Citrobacter freundii, and Salmonella typhimurium.15,17,21,23,24
Silver Dissociation (Release)
A recent in vitro study by Ovington32 assessed various silver technologies. Some of the information contradicts other publications, further fueling the silver debate. For instance, according to the manufacturer and published literature on a methacrylate gel sheet impregnated with silver chloride (Silvasorb, Medline Industries Inc., Mundelein, Ill) and a hydrofiber dressing impregnated with silver (Aquacel Ag, E.R. Squibb and Sons, Ickenham, Uxbridge, UK) (two of the dressings evaluated35-36), these dressings typically deliver silver in the range of 1 to 2 ppm, not the 40- to 60-ppm and 120-ppm range, respectively, stated in the Ovington article. However, the silver release data for the nanocrystalline silver-coated dressing (Acticoat, Smith & Nephew, Largo, Fla) seems to be consistent with other publications (50 to 100 ppm, released in de-ionized, purified, sterile water).
The literature contains other inconsistencies with respect to silver release and the resulting antimicrobial efficacy. Because the level of Ag+ has been found to be directly proportional to the rate of kill in several in vitro studies6,37,38 and not proportional in others,32,39 the accuracy of some published data is called into question. The lack of detail on how some tests are run and the established chemical principles of solubility also make some data difficult to interpret. Basic scientific principles and established physical laws help clarify the information available.
For example, in in vitro silver dissociation testing, the extraction media used can have a significant impact on the data obtained. It has been shown that a hydrofiber dressing containing the silver compound Ag/CMC releases 0.8 ppm of silver in water and in saline34 but releases 85 ppm of silver in thiosulfate.40 The difference is more than 100-fold. Nanocrystalline silver releases 50 ppm34 or 70 ppm in water (depending on who is doing the testing), 0.8 ppm in saline (according to one analysis),35 and 640,000 ppm in thiosulfate.40 The differences are staggering. Serum (ie, simulated wound fluid) also alters the rate of silver release.
Although these results can largely be explained via Le Chatelier’s principles of chemical equilibrium and solubility products,41 to the lay person, such data can be confusing. Applying Le Chatelier’s principles to silver compounds, silver compounds (in solution) exist in a state of chemical equilibrium between the silver compound and its dissociation products. This is exemplified by the following chemical equation for silver chloride (AgCl), a common silver compound:
AgCl ⇔ Ag+ + Cl-
If Ag+ is removed from the right side of the equation, silver chloride dissolves, providing an additional silver ion to replace the one removed. In other words, the chemical equilibrium shifts to the right. Organic and inorganic material in a given extraction media will bind with Ag+, removing it from the right side of the equation. This causes a shift to the right. So when the solution is tested for Ag+ content at the end of the experiment, the level of Ag+ detected will depend on the amount of organic and inorganic matter found in the extraction media. In a nutshell, organic and inorganic matter (present in the extraction media) can shift the position of chemical equilibrium, increasing the amount of silver released during a given test. However, this phenomena can create a disconnect between silver dissolution data and antimicrobial efficacy data because organic and inorganic matter can bind (deactivate) free Ag+, negating its antimicrobial attributes. Given this information, the best test media might be purified/de-ionized water from which inorganic and organic matter for the most part have been removed. However, it is not the amount of silver delivered that is important — rather, it is the availability and antimicrobial activity of that silver.
With regard to inconsistencies with respect to antimicrobial efficacy found in the literature, considering strains of microbes may help resolve some of the confusion. A recent in vitro study32 and other articles present data demonstrating that low levels of silver have rapid bactericidal activity; however, this view is in contrast to other data sources and much of the peer-reviewed literature. To understand this disconnect, it might be helpful to think in terms of strains. All pathogens have multiple strains42-44 and different strains can require different levels of silver for kill. For example, in 1992 there were more than 80 known strains of MRSA and bactericidal activity against all strains was achieved only after 60.5 ppm silver was applied.45 Knowing which strains are evaluated in a given comparison is important because some strains can be killed with very low levels of silver. However, these may not be the strains the wound’s own defenses are having difficulty controlling. Other strains require significantly higher levels of silver for kill. As the strain often is not mentioned in published data, determining the relevance of some data is difficult. Testing performed recently by an independent source (Hoosier Microbiological Laboratories, Muncie, Ind) on five different strains of Pseudomonas demonstrated that higher levels of silver result in rapid, bactericidal activity on all or most (five with nanocrystalline silver, four with SSD) strains tested; whereas lower levels of silver had an effect on only one of the strains tested (at times, no effect). This finding is in agreement with much of the published literature on the subject.46-52 Thinking in terms of strains might explain why the majority of silver dressings have the same or similar claims with respect to antimicrobial efficacy (ie, kill methicillin-resistant S. aureus [MRSA], vancomycin-resistant Enterococci [VRE], Pseudomonas, E. coli, and others). A dressing delivering lower levels of silver may show rapid bactericidal activity toward some strains (ones sensitive to lower levels of silver), yet not have the same effect on more robust strains.
Silver levels. Some authors believe very low levels of silver (ie, 1 ppm) provide sufficient bactericidal efficacy, while others support the view that a higher level of silver is required. For instance, for silver concentrations in vitro data showed >3 log reductions (bactericidal); whereas, for silver concentrations >36 ppm, 67.9% of the data showed log reductions >3 (bactericidal).46-51
One manufacturer-sponsored open-label, multicenter, non-comparative, non-controlled clinical study52 may provide insight into the importance of silver levels in a clinical setting. The study explored the effect of adding silver to a well-known absorbent dressing (composed of hydrofibers) to make a silver version. The authors demonstrated that adding silver to the dressing does not have a negative effect on the dressing’s absorptive capability or safety. However, the study failed to definitively demonstrate that adding silver (in the form of silver chloride) provides additional efficacy with respect to managing bioburden and assisting in healing over the non-silver version. Silver chloride provides a low level of free Ag+. This, then, begs the question: Is the level of silver provided by silver chloride (approximately 1 ppm of silver) adequate?
A factor that supports the need for a higher level of silver is the effect of organic and inorganic materials in the wound fluid that complex (or bind) ionic silver and prevent it from acting as an antimicrobial agent. This is why it is commonly believed that the silver source must provide sustained release of adequate levels of ionic silver over time. The silver ion is extremely reactive — not only does silver react with microbes, but it also reacts with a variety of anions (eg, chloride, sulfates, phosphates, carbonates, and acetates) found in biological fluids to form relatively insoluble complexes or precipitates. This complexation phenomenon is believed to necessitate the delivery of an adequate concentration of silver for antimicrobial activity to be observed. In addition, Ag+ binds to proteins and nucleic acids, further reducing the amount of available silver which, in turn, reduces the antimicrobial efficacy of the silver delivered. This argument is strongly supported in the literature46, 49,50,51,53,54 from which it can be summarized that 86% of the organisms reported in the literature have MIC values for silver of >1 ppm and MIC values increase from 80 to 5,000 times when organic materials are present.
Rate of Kill (Important or Not?)
Some authors argue that speed is unimportant and controlling pathogens within minutes versus hours or even days holds little real value.32,39 Others believe a rapid kill is a measure of an antimicrobial agent’s efficacy. In one in vitro study,7 nanocrystalline silver’s speedy antimicrobial activity in comparison to other forms of silver is mentioned several times. This indicates that the author is using speed as a measure of efficacy. A later in vitro study by the same author38 once again stresses the speed of antimicrobial activity and concludes that: “The nanocrystalline silver was found to achieve a more rapid delivery of silver and a significantly faster reduction of bacteria than the other agents......but the real advantage of nanocrystalline silver is in the speed of delivery to the bacterial cell.” This premise is supported in other literature as well: “The kill rate is directly proportional to Ag+ concentration, typically acting at multiple targets. The higher the silver-ion concentration, the higher the antimicrobial efficacy.”37 The authors of another in vitro study conclude that “Bactericidal activity can be measured with the time-kill kinetic method, whereby the kill rate of a given antimicrobial agent is determined. During this assay, the time course of antimicrobial activity is plotted by determining the kill rate of bacteria when exposed to antimicrobial agent(s). The bacteria killing rates tend to depend on the class of the antibiotic and its concentration. The time-kill kinetic method has been widely used to evaluate and compare new antimicrobial agents.”55
Bacteria reproduce rapidly. For instance, E.coli reproduces about every 20 minutes56 (depending on the strain) correlating to ~1021 colonies over 24 hours; with each new generation, mutations are possible. Over time mutations can lead to resistance.8,11,57 The longer a given microbe is allowed to live and multiply in the presence of a toxic agent, the greater the chances for selection for resistance to that agent.8,11,57 If the opportunity to limit the number of potential mutations exists, why not take it? On the other hand, some authors note that the frequency of dressing changes negates the necessity for a quick-acting silver dressing.32 Rate of kill, however, should not be reduced to a dressing change argument because guarding against resistance is the true benefit of a rapid kill rate. Hindering biofilm formation (under appropriate conditions, biofilms can form rapidly)58 and sepsis59 are believed to be potential additional benefits of a quick-acting agent.
Effects on Viable Cells
Some authors have suggested that the higher levels of silver delivered by antimicrobial agents such as nanocrystalline silver and silver nitrate are harmful to viable cells — silver nitrate contains the nitrate ion, a strong oxidizing agent and known to be cytotoxic. These concerns are based on a small amount of in vitro data36,60,61 that do not seem to support what is typically seen in the clinic. Making the leap from the lab to the clinic is difficult without actually performing clinical studies. However, in a controlled clinical study involving 17 patients, Innes et al62 showed that donor sites treated with nanocrystalline silver took longer to heal than sites treated with a foam dressing alone (control). It is important to stress that during the study, despite continued discussions with the manufacturer on how to wet and apply the silver dressing, the clinicians continued to apply them saturated (they did not squeeze out excess fluid) and continued to keep them saturated for the duration of the study. In three separate studies, Honari et al,63 Perlov et al,64 and Hanna et al65 demonstrated the clinical benefits of nanocrystalline silver used properly (ie, moistened if wound bed is dry) on donor sites. The clinical study by Honari and colleagues involved a comparison of SSD to nanocrystalline silver. In this retrospective study, 201 donor sites on 20 patients with methicillin-resistant S. aureus (MRSA, 10 in each group) were evaluated over a period of 8 months.63 The nanocrystalline silver dressing proved to be more clinically advantageous than the SSD. No negative effects on healing rates were noted. In addition, a substantial cost savings was noted using nanocrystalline silver dressings when compared to standard treatment. In a pilot study, Perlov et al,64 evaluated the nanocrystalline silver dressing in “difficult” donor sites to determine the incidence of infection and complications to healing. Eight patients with a total body surface area (TBSA) burn of 19% to 35% were evaluated. The results of the study indicated that donor sites in these problematic areas benefited from the test dressing and no delay in the rate of healing was found. In a prospective clinical study, Hanna et al65 compared nanocrystalline silver (Acticoat) to Xeroform (Tyco Health Care/Kendall, Mansfield, Mass) on 10 patients for donor site healing and patient satisfaction. A comparatively faster healing rate was found in all patients for the nanocrystalline silver, with nine out of 10 patients giving it higher satisfaction ratings. Although an in-depth statistical analysis was not performed, these studies demonstrate that in a clinical setting, nanocrystalline silver is an effective antimicrobial barrier and antimicrobial agent and does not impede the rate of healing.
A more “common sense” way of looking at the effects of silver levels on viable cells is the fact that SSD delivers >30 times more silver than nanocrystalline silver (and much more than some dressings on the market) and wounds have been healing with SSD for decades. Also, nanocrystalline silver has been found to help heal a wide variety of wounds and its positive wound management aspects have been shown in numerous controlled clinical studies and case series.63-73 If this level of silver were cytotoxic, such favorable clinical results would not be expected.
Silver has been used as an antimicrobial agent for centuries. Historically, SSD and silver nitrate have been the only choices in the medical arena but in recent years numerous silver-based antimicrobial agents have become available to the clinician. Relevant literature and practicing clinicians point to much confusion regarding the use of silver as an antimicrobial agent — confusion created by manufacturers and authors voicing different viewpoints. By addressing six particularly relevant factors regarding the properties of this intriguing element using data and opinions published in the peer-reviewed literature, independently generated data, and sound scientific principles, it is hoped that clinicians can start to work through some the confusion and become confident in their use of silver dressings. The ultimate goal is to provide the best level of patient care possible and to ensure that silver (an effective antimicrobial agent) is just as efficacious in future years as it is today.