Ann Thorac Surg 1999;68:2123-2128
© 1999 The Society of Thoracic Surgeons
Original Articles: Cardiovascular
Hydrogen peroxide for prevention of bacterial growth on polymer biomaterials
Eckhard Alt, MDa,
Francesca Leipold, MDa,
Danica Milatovic, MDa,
Günter Lehmann, MDa,
Sybille Heinz, MDa,
Albert Schömig, MDa
a Medizinische Klinik, Klinikum rechts der Isar, Technische Universität München, München, Germany
Address reprint requests to Dr Alt, Klinikum rechts der Isar, Ismaninger Strasse 22, D-81675 München, Germany
e-mail: alt{at}med1.med.tu-muenchen.de
 |
Abstract
|
|---|
Background. Despite widespread use of potent antibiotics, infections of artificial implants and catheters are of increasing concern. We tested whether local treatment with 3% hydrogen peroxide (H2O2), long known as an inexpensive wound disinfectant, could prevent or reduce bacterial growth on polymer biomaterials.
Methods. Two-centimeter-long pieces of polyurethane and silicone tubing were contaminated with a standardized solution of Staphylococcus epidermidis (105/mL) and then rinsed and wiped with saline (0.9%) solution. Bacterial growth was assessed after incubation at 37°C for 24 hours. Bacterial colonies were compared for the following treatments: wiping only with saline; wiping with 1.5%, 2%, or 3% H2O2; pretreating biomaterials with 3% H2O2 and subsequent contamination for 2 and 4 hours without treatment after contamination; and contamination of tubings 1 month after pretreatment with 3% H2O2. The effect of 3% H2O2 was also assessed on contamination with Escherichia coli.
Results. Bacterial growth was reduced by more than 99% when the contaminated tubes were treated with 3% H2O2 compared with saline control (p < 0.001). Lower concentrations of H2O2 were less effective. The length of the contamination period had no influence on the effectiveness of H2O2 when used on polyurethane but did with silicone tubings. Pretreatment with H2O2 1 month before contamination still reduced bacterial growth rate by 90% on polyurethane and by 75% on silicone tubings. Comparable effects on bacterial growth rate were observed for staphylococci (-90%, p < 0.001) and escherichiae (-90%, p < 0.001).
Conclusions. Local treatment with 3% H2O2 significantly reduced bacterial growth on polymer biomaterials even for 1 month after treatment. This finding might influence clinical strategies of prevention of foreign body infection.
|
Introduction
|
|---|
Although the adjunctive use of potent antibiotic therapy has helped to reduce the risk of infection of temporary or long-term implants such as catheters, prostheses, heart valves, pacemakers, and defibrillators [1–3] infections of biomaterials still continue to be of concern. This holds especially true for elderly patients who often suffer additionally from concomitant diseases such as diabetes. The relatively high infection rate with implanted foreign bodies results from special properties of the bacteria, such as enhanced adherence to biomaterials, and to production of protective slimy substances as in the case of Staphylococcus epidermidis [4]. The effect of systemic antibiotic treatment for this primarily low-pathogenic bacterial species causing most implant-related infections [5] and especially late infections 24 months and later after primary infections, is limited [1, 6].
Therefore it is desirable to prevent infections by reducing the susceptibility to contaminations. Hydrogen peroxide (H2O2) has been used as a wound disinfectant for a long time. It is cheap, nontoxic to human tissue [7], and effective on gram-negative and gram-positive bacteria as well as fungi and yeasts [8]. The aim of the present study was to assess the efficacy of H2O2 in reducing and preventing bacterial growth on typical clinically used biomaterials such as polyurethane and silicone tubing.
|
Material and methods
|
|---|
Tested materials and assessment of bacterial growth
Polyurethane and silicone (Dow chemical) tubing, corresponding to materials commonly used for short- and long-term instrumentation of patients, were cut into samples of 2-cm length and sterilized thereafter with ethylene oxide.
Tubings were contaminated with 105 colony forming units (cfu)/mL of either Staphylococcus epidermidis or Escherichia coli, which were obtained from blood cultures of patients with catheter-related septicemia. Determinations of bacterial concentrations were carried out with a photometer [9]. Contaminations were done over 2 hours at 37°C in saline (0.9%) solution corresponding to bacterial growth conditions between 34°C and 37°C in humans [9] and a clinically relevant duration of contamination [10]. After contamination, all tubings were rinsed five times with 40 mL of saline to remove nonadherent bacteria. For assessment of bacterial growth, the semiquantitative blood agar roll method of Maki and associates [11] was used. Accordingly, all polymer biomaterial pieces were rolled onto the surface of nonselective sheep blood agar (Columbia Agar) dishes five times. Then the culture dishes were incubated at 37°C for 24 hours and colonies counted thereafter.
Biomaterial pretreament and experiments
For each of the four experiments described below, bacterial growth rate for three different samples was determined: one sample was left untreated and served as control, one was wiped with a sterile gauze swab soaked with 0.9% saline solution (NaCl), and the third one was first wiped with saline and then treated with H2O2 of various concentrations. Each experiment was done with nine pieces of biomaterial.
In the first experiment, the effect of H2O2 on polyurethane was compared with that on silicone according to the above-mentioned pretreatment procedures. Bacterial contamination of the samples was carried out using S epidermidis as described above.
The second experiment was done to determine the lowest effective concentration of H2O2 that impairs bacterial growth. After a 2-hour contamination period of the biomaterial with S epidermidis, the tubings were wiped with 1.5%, 2%, and 3% hydrogen peroxide.
The third series of experiments was done to study the influence of the sequence of disinfection and contamination and, additionally, of the contamination duration (2 hours versus 4 hours, beginning immediately after wiping the test tubing with 3% hydrogen peroxide) as well as of the persistence of disinfectant effects. To assess a prophylactic effect of 3% hydrogen peroxide, the interval between wiping and contamination was continuously prolonged to 1 month.
The fourth experiment studied bacterial growth on silicone after contamination with gram-negative and gram-positive bacteria with regard to the respective pretreatment forms. Accordingly, silicone tubings were contaminated with either S epidermidis or Escherichia coli colonies for 2 hours. In addition, the effect of pretreatment before contamination, which was done after 1 month, was assessed.
Statistical analysis
Values are expressed as mean ± standard deviation. When control samples were contaminated, quantification of bacterial colonies was stopped at 500 cfu because the exact number was no longer discernible at that density. Statistical analyses were performed using Student’s t test for paired or unpaired groups as appropriate. A p value less than 0.05 was considered statistically significant.
|
Results
|
|---|
Different test materials
If the samples were wiped with saline after contamination, 3.8 cfu per dish were found (Table 1). Bacterial growth on untreated control samples was more than 500 cfu. Additional wiping of the tubing with 3% hydrogen peroxide reduced bacterial growth to 0.2 cfu per dish, equivalent to a reduction of 99.96% compared with untreated controls or by 95% compared with saline treatment (p < 0.001).
View this table:
[in this window]
[in a new window]
|
Table 1. Comparison of Biomaterials With Respect to Growth of Staphylococcus epidermidis After Contamination With 105 Colony-Forming Unitsa
|
|
Analogously, on control silicone tubings, more than 500 cfu per dish were found. After treatment with saline, a reduction of bacterial growth from more than 500 to 68.2 cfu was observed. A residual bacterial growth of 4.3 cfu per dish was determined after wiping the samples additionally with 3% hydrogen peroxide (Fig 1), representing a reduction by 99.14% compared with controls and by 94% compared with saline treatment alone (p < 0.001).
View larger version (25K):
[in this window]
[in a new window]
|
Fig 1. Comparison of three different concentrations of H2O2 (abscissa) on bacterial growth (ordinate) of 105 colony-forming units (cfu)/mL. Columns represent mean ± standard deviation.
|
|
Impact of various concentrations of hydrogen peroxide
Starting with unimpaired growth of S epidermidis on control polyurethane pieces yielding more than 500 cfu per dish, wiping with various concentrations of hydrogen peroxide diminished bacterial growth significantly (p < 0.001) to 5.7 cfu and to 6.8 cfu per dish, with 1.5% and 2% H2O2, respectively. Use of 3% hydrogen peroxide resulted in the lowest bacterial growth rate (0.2 cfu; p < 0.01 and p < 0.001 versus treatment with 1.5% and 2% hydrogen peroxide, respectively), rendering the 3% concentration the most effective one (Fig 2).
View larger version (44K):
[in this window]
[in a new window]
|
Fig 2. Number of colony-forming units (cfu) (ordinate) at 2 hours (left column of each pair) and at 4 hours (right column of each pair) after contamination with 105 cfu/mL and pretreatment (abscissa) with either 3% H2O2 (left pair of columns), saline (middle pair of columns), or no pretreatment (right pair of columns). p* indicates p < 0.001.
|
|
Importance of contamination duration
If the contamination duration was prolonged from 2 hours to 4 hours, there was no effect of saline on either biomaterial tested. In contrast, additional wiping of the tubing with 3% hydrogen peroxide was always effective, irrespective of the duration of contamination and the material tested. Specifically, a residual bacterial growth of 7% and 12% (p < 0.001 versus saline) compared with controls was observed at 2 hours and 4 hours on polyurethane samples (Fig 3), which amounted to 22% and 39% (p < 0.001 versus saline) of the bacteria found on silicone tubings.
View larger version (14K):
[in this window]
[in a new window]
|
Fig 3. Bacterial growth (ordinate) after contamination of 2 cm polyurethane (A) and silicone tubings (B) with 105 colony-forming units (cfu)/mL and subsequent incubation for 2 hours at 37°C after different periods of time (abscissa) between wiping biomaterials with 3% H2O2 (solid lines), saline (dotted lines), or nothing (dashed lines) and bacterial contamination.
|
|
Relevance of interval between tubing treatment and contamination
Irrespective of the materials tested, saline’s protective effect regarding staphylococcal growth was not discernible for contaminations longer than 24 hours after treatment (Table 2). In contrast, prophylactic treatment with 3% hydrogen peroxide diminished staphylococcal proliferation specifically for both materials and constantly for 1 month. For the various samples of polyurethane, reductions of 85% to 91% (p < 0.001 versus saline pretreatment) were observed throughout this series of experiments. A range between 70% and 80% (p < 0.001 versus saline) was found for pretreatment of silicone samples (Fig 4).
View larger version (33K):
[in this window]
[in a new window]
|
Fig 4. Bacterial growth (ordinate) after wiping silicone electrode pieces with 3% H2O2 (left pair of columns), saline (middle pair of columns), or no treatment (right pair of columns) at 4 weeks before contamination with either 105 colony-forming units (cfu) of Escherichia coli (checkered columns) or Staphylococcus epidermidis (dotted columns).
|
|
Impact of hydrogen peroxide on growth of different bacterial species
Contamination of silicone tubings 1 month after pretreatment with saline alone led to growth of E coli of 107 cfu. Pretreatment with 3% hydrogen peroxide, however, reduced the growth rate to 13 cfu per dish (p < 0.001 versus saline pretreatment) (Fig 5). Treated and untreated samples of silicone and polyurethane tubings were studied by scanning electron microscopy (AMR 1200; Fa. Leitz, Wetzlar, Germany). The surface was smooth; no cracks or surface alterations were found as a potential indicator of damage.
View larger version (75K):
[in this window]
[in a new window]
|
Fig 5. Surfaces of silicone electrodes that had been wiped with saline (A) or saline plus 3% H2O2 (B) before contamination with Staphylococcus epidermidis. (Original magnification, x6000.)
|
|
Clinical experience in using hydrogen peroxide
From July 1992 to July 1998, 990 patients who received a pacemaker (n = 840) or defibrillator (n = 150) implantation, respectively, had been treated with H2O2 during the implant procedure at our institute. After placement of the leads, the remaining parts of the electrode had been wiped once with 3% H2O2 and rinsed with saline thereafter before the leads were connected to the implant device and before the device was inserted into the preformed pocket. Of those 990 patients, only 2 (one pacemaker, one defibrillator) had an infection (0.2% infection rate). There were no signs of delayed wound healing or other indications of local irritation after the treatment regimen.
This infection rate compares favorably to a 1.83% infection rate (p < 0.05) for the period before 1992 in which no H2O2 was used; duration of operation, distribution of devices, and patient and operation-associated variables were not different between patients treated with and without H2O2.
|
Comment
|
|---|
Bacteremia from intravascular devices has continued to be a major source of morbidity and even death in hospitalized patients. Despite the use of potent antibiotics, device-related bacteremia still presents a clinically relevant problem. The therapy of choice is surgical removal of the infected foreign body with all its associated risks and morbidity [12]. Consequently, prevention of foreign body infection is of major concern.
Extent of bacterial growth reduction, importance of disinfectant concentration, and potential of damaging biomaterials
Wiping polymer biomaterial tubings with hydrogen peroxide after bacterial contamination led to reductions of bacterial growth of more than 99%. This is a remarkable finding, because a comparable degree of reduction of bacterial growth was found in a study using concentrations of hydrogen peroxide in vapor form of as high as 30% for sterilization after contamination [13]. In the present study, reductions of at least 70% were found when contamination occurred as late as 1 month after pretreatment with H2O2.
The present data show the optimal concentration of hydrogen peroxide as 3% with respect to disinfectant properties. No untoward human health effects are to be expected at this concentration [14]. With respect to biodegradation of synthetic materials, surface and mechanical qualities as well as molecular weight changes of polyurethane were reported only with concentrations as high as 25% at temperatures of 100°C over 24 hours [15]. In contrast, sterilization of polyurethane and silicone with 30% hydrogen peroxide for 30 minutes did not reveal any changes such that, at the concentrations used in the present study, no damage to either material tested is assumed [16].
Tissue-damaging potential of hydrogen peroxide
Hydrogen peroxide is a disinfectant that has a function in various aspects of clinical practice without causing damage in concentrations of up to 3% [9, 14]. At these concentrations, which were used in the present in vitro study as well as in the initial clinical pilot phase application in conjunction with pacemaker implants, there is no potential of increasing mutagenicity, teratogenicity, carcinogenesis, toxicity, or allergy [14]. On the other hand, high concentrations of hydrogen peroxide not only damage myocardium [9] but also inhibit the enzyme glucose oxidase irreversibly [13] such that its concentration should be kept to a minimum.
Impact of tested material, duration of disinfectant effects, bacterial species differences, and duration of the contamination period
Silicone is a hydrophobic material, whereas polyurethane, which has gained widespread use in conjunction with electrodes and catheters [5], is comparably more hydrophilic and smooth. These differences might be relevant during the initial process of bacterial adhesion to synthetic biomaterials, which appears more facilitated with materials that are hydrophobic and less smooth [4, 17].
In addition, negatively charged electrodes have been shown to diminish bacterial adhesion [18]. Both mechanisms could have influenced the present study. Moist wiping with saline instantaneously reduced subsequent bacterial growth. This effect was not apparent when a longer period of time elapsed between treatment and contamination. Use of hydrogen peroxide, a disinfectant with oxidizing potential, is assumed to render materials more hydrophilic and more negatively charged. These conditional changes of the surfaces are considered one mechanism of the prophylactic action of the disinfectant. According to a study on silicone contact lenses [19], a high degree of penetration of hydrogen peroxide into silicone was found despite silicone’s chemical inertia. As in the present study, residues of the disinfectant inside the synthetic material might explain the observed duration of effectiveness. In addition, changes of the surface characteristics affecting the bacterial adhesion properties are considered equally important.
Obviously 3% hydrogen peroxide pretreatment of the silicone tubings reduced growth of S epidermidis to a lesser extent than growth of E coli. This might be due to the species’ capability to adhere rather than the disinfectant’s specificity for different kinds of bacteria because E coli has a lower degree of adherence than S epidermidis [20]. In addition, silicone appears to be a preferred synthetic material for adherence of the latter bacterial species [20].
An influence of the type of biomaterial became evident when the contamination period was prolonged. Accordingly, the above mentioned affinity of staphylococci for silicone, together with the material’s hydrophobicity, despite its capability of hydrogen peroxide penetration appear to best explain the higher degree of impairment of bacterial growth of polyurethane compared with silicone, irrespective of the duration of the contamination period.
Clinical implications
Patients who have treatment with foreign bodies, such as various types of catheters, electrodes, pacemakers, defibrillators, or orthopedic and surgical implants, are at an increased risk of local bacterial infections or bacteremia, known to occur up to 24 months after implantation [1–3, 19, 21, 22]. Despite application of potent antibiotics there often remains an infection rate of up to 10%, pointing to a susceptibility of the sterilized materials to bacterial contamination during implantation or, alternatively, to bacterial pathomechanisms, such as enhanced adherence to biomaterials due to production of slimy substances as in the case of S epidermidis, which reduces the antibiotic’s effectiveness [2–4]. Our results imply that disinfection of biomaterials with 3% hydrogen peroxide not only reduces bacterial growth, but can also provide effective prophylaxis because of long-lasting antibacterial effects in vitro. Randomized, prospective clinical studies are required to substantiate the clinical importance of our in vitro findings and of our nonrandomized clinical findings showing an effective reduction in infection rate from 1.83% without treatment to 0.2% with H2O2 treatment.
|
References
|
|---|
-
Jara F.M., Toledo-Pereyra L., Lewis J.W., Jr, Magilligan D.J. The infected pacemaker pocket. J Thorac Cardiovasc Surg 1979;78:298-300.[Abstract]
-
Bluhm G., Jacobson B., Julander I., Levander-Lindgren M., Olin C. Antibiotic prophylaxis in pacemaker surgery—a prospective study. Scand J Thorac Cardiovasc Surg 1984;18:227-234.[Medline]
-
Pfeiffer D., Jung W., Fehske W., Korte T., Manz M., Moosdorf R. Complications of pacemaker-defibrillator devices. Am Heart J 1994;127:1073-1080.[Medline]
-
Peters G., Saborowski F., Locci R., Pulverer G. Adherence and growth of coagulase-negative staphylococci on surfaces of intravenous catheters. J Infect Dis 1982;146:479-482.[Medline]
-
Hafferjee A.A., Moodley J., Pillay K., Singh B., Thomson S., Bhamjee A. Evaluation of a new hydrocolloid occlusive dressing for central catheters used in total parenteral nutrition. South Afr J Surg 1991;28:142-146.
-
Archer G.L., Armstrong B.C. Alteration in staphylococcal flora in cardiac surgery patients receiving antibiotic prophylaxis. J Infect Dis 1983;147:642-649.[Medline]
-
Tschesche R., Korte F., Bethge J. Über die Rolle des Wasserstoffperoxyds bei der Inhibition nach Dold, 3. Z Naturforsch 1951;6b:22-24.
-
Yosphe-Purer Y., Eylan E. Disinfection of water by hydrogen peroxide. Health Lab Sci 1968;5:233-238.[Medline]
-
Balows A., Hausler W.J., Herrmann K.L., Isenberg H.D., Shadomy H.J. Manual of clinical microbiology, 5th ed. Washington DC: American Society for Microbiology, 1991.
-
Bluhm G., Nordlander R., Ransjö U. Antibiotic prophylaxis in pacemaker surgery. PACE 1986;9:720-725.
-
Maki D.G., Faris A., Wadstrom T., Hjerten S. A semiquantitative culture method for identifying intravenous catheter related infection. N Engl J Med 1977;296:1305-1309.[Abstract]
-
Alt E., Neuzner J., Binner L., et al. Three-year experience with a stylet for lead extraction. PACE 1996;19:18-25.
-
Tse P.H. H202. Biotechnol Bioeng 1987;29:705-708.[Medline]
-
Richardson M.L., Gangolli S. Hydrogen oxide. In: Richardson M.L., Gangolli S., eds. . The dictionary of substances and their effects. Oxford: The Royal Society of Chemistry, 1994:877-881.
-
Meijis G.F., McCarthy S.J., Rizzardo E., Chen Y.-C., Chatelier R.C. Degradation of medical-grade polyurethane elastomers, the effect of hydrogen peroxide in vitro. J Biomed Mater Res 1993;27:345-356.[Medline]
-
Ikarashi Y., Tsuchiya T., Nakamura A. Cytotoxicity of medical materials sterilized with vapour-phase hydrogen peroxide. Biomaterials 1995;16:177-183.[Medline]
-
Hogt A.H., Dankert J., De Vries J.A., Feijen J. Adhesion of coagulase-negative staphylococci to biomaterials. J Gen Microbiol 1983;129:2959-2968.[Abstract/Free Full Text]
-
Liu W.K., Tebbs S.E., Byrne P.O., Elliott T.S. The effects of electrical current on bacteria colonising intravascular catheters. J Infect 1993;27:261-269.[Medline]
-
Harris M.G., Torres J., Tracewell L. pH and H2O2 Concentration of hydrogen peroxide disinfection systems. Am J Optom Physiol Optics 1988;65:527-535.[Medline]
-
Lopez-Lopez G., Pascual A., Perea E.J. Effect of plastic catheter material on bacterial adherence and viability. J Med Microbiol 1991;34:349-353.[Abstract/Free Full Text]
-
Fry D.E., Fry R.V., Borzotta A.P. Nosocomial blood-borne infection secondary to intravascular devices. Am J Surg 1994;167:268-272.[Medline]
-
Gristina A.G., Dobbins J.J., Giammara B., Lewis J.C., DeVries W.C. Biomaterial-centered sepsis and the total artificial heart. JAMA 1988;259:870-874.[Abstract/Free Full Text]
Accepted for publication May 1, 1999.
Top
Abstract
Introduction
