The Journal of Bone and Joint Surgery American Volume

Electrically generated silver ions have been shown previously to be a potent antibacterial agent with an exceptionally broad spectrum as mdicated by in vitro testing. The present study reports on clinical experience using electrically generated silver ions as adjunctive treatment in the management of chronic osteomyelitis. Fourteen patients had fifteen treatment attempts: thirteen for chronic osteomyelitis of the tibia, one for acute and chronic pyarthrosis and osteomyelitis of the knee, and one for a chronically draining sinus after total hip replacement. Wound d#{233}bridement, silver ion iontophoresis, and subsequent wound care (usually provided by the patient) resulted in control of the infection in twelve of the fifteen treatment attempts and in healing of the non-union after follow-up ranging from three to thirty-six months. The other three attempts led to two partial and one complete failure. In the treatment of chronic osteornyelitis, many antibiotics have been used locally and systemically to control the infecting organisms, but none of the agents used either singly or in combination has been consistently effective43 71#{176}1t 14. For the past three years we have been evaluating the possibility of using silver ion iontophoresis as a bactericide in infected tissue. The bactericidal properties of silver have been known for a long time’ , but as currently applied its clinical utility * Supported by the Veterans Administration Research Service, Project No. 09814-7718-01 , and by the Ritter Company, Division of Sybron , Incorporated. 1 Veterans Administration Hospital. Irving Avenue and University Place, Syracuse, New York 13210. is limited. The diffusion of silver ions from a metallic surface such as silver foil is negligible due to its low solubility in aqueous solutions. Silver compounds that dissociate easily, such as silver nitrate, are locally sclerosing, and if used in large amounts the absorbed nitrates or other cations may be toxic. Most silver compounds that are non-toxic, such as the silver proteinates or silver sulfadiazine, are only sparingly dissociabl& . Recent studies indicated that the bactericidal properties of silver sulfadiazine are at least in part due to the attached silver” , although the action of silver on bacteria is not known. There is some evidence that it interferes with cell-membrane function, cellular DNA, and the respiratory chain of enzymestuid. Free silver ions coming from a silver anode, because of their small size, can penetrate to some extent any structure that has an aqueous component even when the structure is avascular. The penetration is the result of both diffusion and ionic migration along a voltage gradient. Since silver ions combine with many proteins to form relatively insoluble compounds, silver ions must be present in the tissues in some excess to exert their full bactericidal effect. With ions being released continuously from the anode, the necessary excess of ions in the tissues is assured. At the same time, silver is minimally toxic” and the amount administered by iontophoresis is far below the level necessary to produce a detectable body burden. In our previous study of electrical stimulation of bone growth , in which we used pure silver electrodes as cathodes implanted in bone to produce cellular stimulation, no localized necrosis was evidenO . Based on this experience, it seemed possible that with the same silver electrode we could deliver silver ions for their bactericidal effect and electric current for fracture 872 R. 0. BECKER AND J. A. SPADARO THE JOURNAL OF BONE AND JOINT SURGERY healing in the treatment of non-unions complicated by infection. At the beginning of our study, we did not know what organisms would be susceptible, what the local cellular toxicity might be, and whether the ions would be efficient in a clinical situation. We attempted to study each of these problems. The effects of electrically generated silver ions on a fairly large number of microorganisms were evaluated. While some minor differences in susceptibility were noted, all of the organisms that we tested were sensitive to the electrically generated silver ion, including some that were resistant to all known antibiotics2’2 . The minimum inhibitory concentrations for silver ions were determined for a variety of bacteria. These were compared with the minimum inhibitory concentrations of antibiotics and found to be at least an order of magnitude lower on either a weight or a molar basis2 . We also obtained some evidence that electrically generated silver ions are fungicidalt. The depth of penetration into bone of silver ions electrically generated at the anode and moving along a voltage gradient is not known. In agar gels containing proteins, chlorides, and bacterial nutrients, we could not demonstrate penetration of more than one centimeter with any of the variations in electrical current or potential, shape of the electrode, or composition of the gel that we 12 , For the purposes of this study, we assumed that the penetration into bone was similar and at the most one centimeter. The mechanism of the action of silver ions on bacteria has not been explored fully. In the case of gram-positive cocci , electron micrographs of bacteria treated in vitro with silver anodes showed that incomplete septa formed in the cells as well as dense, enlarged mesosomes. There was also some separation of the plasma I , In Escherichia co/i, the induction of f3-galactosidase was almost completely inhibited by the silver anode . It would therefore appear that the cell membrane is the primary site of growth-inhibiting action on Eseherichia co/i, Staphylococcus aureus , and si milar organ isms . There was no indication that the action was merely the result of gross protein precipitation, since this would affect all cells, not specifically bacteria. Toxicity for mammalian cells was also evaluated in our laboratory (unpublished observations) using cultures of mouse fibroblasts grown in the presence of anodes of pure silver wire which were subsequently energized at the levels used clinically. After two hours the fibroblasts in the immediate vicinity of the energized wire became rounded, but they were not lysed and their cell membranes remained intact as shown by the trypan-blue exclusion test. On subculture, these fibroblasts reproduced and their progeny had normal morphology and showed normal replication. The chronic nature and variable behavior of osseous infections and the difficulties in isolating the effects of the different parts of the treatment program, such as meticulous wound care and the like, are well known. Despite the fact that such factors were operative in the clinical study to be described, our preliminary results have been so gratifying that we believe that it is worth while to report on them at this time.