Topical Adjuvants Incompletely Remove Adherent Staphylococcus Aureus From Implant Materials

Abstract

Adjuvant treatments including Betadine, Dakin’s solution (sodium hypochlorite), or hydrogen peroxide (H₂O₂) have been attempted to eradicate prosthetic joint infection caused by biofilm or intracellular bacteria. The purpose of this study was to evaluate the in vitro abilities of chemical adjuvants to decrease Staphylococcus aureus (S. aureus) biofilm presence on orthopaedic implant grade materials, including titanium, stainless steel, and cobalt chrome. S. aureus biofilms were grown for 48 h and evaluated for baseline colony forming units/centimeter squared (CFU/cm²) and compared to treatments with Betadine, Dakin’s solution, H₂O₂, or 1% chlorine dioxide (ClO₂). Control discs (n = 18) across all metals had an average of 4.2 × 10⁷ CFU/cm². All treatments had statistically significant reductions in CFU/cm² when compared to respective control discs (p <0.05). For all metals combined, the most efficacious treatments were Betadine and H₂O₂, with an average 98% and 97% CFU/cm² reduction, respectively. There were no significant differences between reductions seen with Betadine and H₂O₂, but both groups had statistically greater reductions than Dakin’s solution and ClO₂. There was no change in antibiotic resistance patterns after treatment. Analysis of S. aureus biofilms demonstrated a statistically significant reduction in biofilm after a five-minute treatment with the modalities, with an average two log reduction in CFU/cm². Statement of clinical significance: While statistically significant reductions in CFU/cm² were accomplished with chemical adjuvant treatments, the overall concentration of bacteria never fell below 10⁵ CFU/cm², leading to questionable clinical significance. Further techniques to eradicate biofilm should be investigated.

Every year approximately one million primary joint arthroplasties are performed in the United States. By 2030, it has been estimated that primary total hip and knee arthroplasties will see a 137% and 601% increase in demand, respectively.¹ Of these procedures, up to 1.7% of primary total hip arthroplasties (THA) and 2% of primary total knee arthroplasties (TKA) will result in periprosthetic joint infection (PJI).^2–4 The increasing infection burden is a concern both medically and financially as PJIs have a 5-year mortality of 33% and an increased cost of care greater than $100,000.^5,6

When addressing biofilms, some of the major obstacles to treatment are their heterogeneity, slow metabolism, exopolysaccharide matrix (glycocalyx), inherent resistance to antibiotics, and strong defensive mechanisms against stress.⁷ The inherent fastidiousness of biofilm infections are particularly difficult for antibiotic applications and surgical revision techniques, such as one-stage exchange, two-stage exchange, and debridement and irrigation and retention of components (DAIR). The current gold standard, the two-stage exchange, is not only expensive, but it also has a lengthy course of treatment with a high morbidity. More recent studies have demonstrated that two-stage exchange has a failure rate close to 30%, with many patients never undergoing the second stage surgery.^8–10 The low cost and lower morbidity associated with DAIR make it an appealing surgical option to both the surgeon and the patient, especially in the setting of an acute infection.¹¹ DAIR has been shown to decrease biofilm mass; however, substantial amounts of residual bacteria and biofilm still remain, which may be a contributing factor in the unacceptably high failure rates with this treatment modality.^12,13

In an effort to improve biofilm and infection control strategies, some surgeons employ adjuvant treatments to improve their debridement which has come in the form of a chemical adjuvant with the off-label use of 10% povidone-iodine (Betadine), Dakin’s solution (sodium hypochlorite), and 3% hydrogen peroxide (H₂O₂).¹⁴ These adjuvants employ similar mechanisms in an effort to disrupt cell walls and molecular metabolism.¹⁵ Povidone-iodine solutions disrupt cell walls while Dakin’s solution also destroys cell walls and negatively affects the metabolism and enzyme activation of biofilms. Dakin’s solution has been found in previous studies to have effects on Methicillin-resistant Staphylococcus aureus (MRSA).¹⁶ Hydrogen peroxide overrides the catalase protection mechanism, subsequently oxidizing and destroying cell walls.¹⁷ One additional topical application of interest is 1% chlorine dioxide (Camelbak cleaning tabs, Camelbak, Petaluma, CA), which is not commonly used intra-operatively. However, chlorine dioxide has been shown to be efficacious against biofilm-related problems in water treatments.¹⁸ Chlorine dioxide directly kills the bacteria by disrupting transport across cell walls and effectively penetrates the glycocalyx. Chlorine dioxide was chosen as a treatment for this study instead of chlorhexidine, because unlike chlorine dioxide, chlorhexidine has been shown to be efficacious against PJI.¹⁹ The purpose of this investigation was to evaluate the efficacy of topical applications against Methicillin-sensitive S. aureus biofilms grown in vitro on metals commonly used in hip and knee arthroplasties, including cobalt chrome (CoCr), titanium alloy (Ti), and stainless steel (SS).

METHODS

Inoculum and Bioreactor Preparation

A single clinical isolate of Methicillin-sensitive S. aureus (MSSA) was used for all biofilm production. Following propagation on 5% sheep’s blood agar plates (Remel, Inc., Lenexa, KS), standard inocula for each experiment was created by suspending MSSA in 50 ml brain heart infusion broth (BHI) (Becton, Dickinson and Company, Sparks, MD) with subsequent incubation for 18 h at 37°C on a rocker platform. The optical density of the bacterial solution was measured and adjusted with BHI broth to standardize the inoculum to a McFarland standard of 2 (~3 × 10⁸ CFU/ml). A fresh inoculum was created for each trial.

Tested biomaterials consisted of steam sterilized, surgical grade CoCr (1 cm diameter 0.1 cm thickness), Ti (0.8 cm diameter and 0.1 cm thickness), and SS discs (0.8 cm diameter and 0.1 cm thickness((Small Parts, Inc., Miami Lakes, FL, wet jet cut by Wilson Works, Morgantown, WV). The discs were aseptically transferred to 12-well flat bottom cell culture plates (Corning Inc., Corning, NY) with one disc per well. A 1:1 ratio of 1 ml S. aureus inoculum and 1 ml fresh BHI was subsequently aliquoted into each well. The 12-well plates were allowed to incubate for 48 h at 37°C on a rocker platform with 2 ml fresh media exchanged every 24 h.

Quantifying Colony Forming Units (CFU/cm²)

Mature biofilm was established on the discs after 48 h of incubation. After 48 h, control discs (no treatment, n =18) were washed to remove non-adherent planktonic bacteria and BHI media present. These discs were then sonicated (Model 575HTA, CREST Ultrasonics, Road Trenton, NJ) for 10 min at 50 Hz (120 V, 4 A) and vortexed for 2 min, then serially diluted and plated via a drop-plate method on 5% blood sheep agar plates (Remel, Inc.) in order to obtain colony forming units (CFU). Plates were incubated at 37°C for 24 h. After the 24 h growth period, colonies were counted and converted to CFU/cm².

Treatment of S. Aureus Biofilm

Discs not utilized as controls underwent treatment with one of four chemical adjuvants following 48 h incubation (Time-point T₀): Dakin’s Solution, 10% Betadine, 3% hydrogen peroxide, and 1% Camelbak^™ chlorine dioxide cleaning tablets. BHI broth was removed and each disc was saturated in 1 ml of a treatment solution for 5 min. Following the 5-min period, all discs were washed three times with 1 ml of phosphate buffered saline to remove the treatment and non-adherent planktonic bacteria. Adherent bacteria were then quantified on (6/12) discs via sonication to determine colony forming units (CFU/cm²). The other six discs were aseptically transferred to a new 12-well plate, immersed in 2 ml BHI, and allowed to incubate for another 24 h (T₂₄) at 37°C on a rocker platform in order to determine the longer-term efficacy of the treatments. After the additional 24 h, the same CFU biofilm quantification methods were employed.

Scanning Electron Microscopy

Samples were fixed with 10% formalin for 2 h. Samples were gold-paladium sputtered (Denton Desk V Sputter and Carbon Coater) and imaged with Hitachi S-4700 Scanning Electron Microscope (SEM) (Hitachi, Schaumburg, IL).

Antibiotic Resistance

After adjuvant treatments, all unique colony morphotypes were identified to genus or species level by matrix-assisted laser desorption ionization time of flight mass spectrometry (Vitek MS, Biomériuex, Durham, NC) using manufacturer-recommended protocols. Antimicrobial susceptibility testing (AST) was performed by a routine automated method (Vitek2 GN81 and GP67 cards, Biomériuex) and/or by manual Kirby–Bauer disk diffusion on Mueller Hinton media with breakpoints and interpretive criteria derived from Clinical Laboratory Standards Institute (CLSI) recommendations (M-100 S26). Standard quality control testing for all identification and AST procedures were verified as acceptable.

Statistical Analysis

Statistical analysis consisted of two-way analysis of variance (ANOVA) and Tukey honest significant difference (HSD) analysis to compare CFU count reduction between metals, treatments, treatment, and control, and time points. Statistical significance was set at p <0.05.

RESULTS

Statistical Reductions for All Treatments

At both T₀ and T₂₄, a statistically significant reduction (p <0.05) in CFUs was observed for all treatments of all metals as compared to their respective controls (Fig. 1). All data points for each treatment were averaged to give a comprehensive representation of the efficacies of the treatments. For all metals combined, at each time point, the most efficacious treatments were 10% Betadine and H₂O₂. At T₀, the average log reduction for 10% Betadine and H₂O₂ was 1.7 (98%) and 1.4 (97%), respectively. At T₂₄, the average log reduction for 10% Betadine and H₂O₂ was 1.2 (89%) and 0.9 (84%), respectively. However, under the same two-way ANOVA and Tukey HSD analysis, there was no significant difference between 10% Betadine and H₂O₂, but both were significantly different compared to Dakin’s solution and ClO₂ (Fig. 1). Further, internal controls designed to detect residual biofilm following sterilization indicated no viable biofilm growth across all metal surfaces (data not shown).

Figure 1.

A comparison of the average CFU/cm² at the time of each treatment (T₀) and 24 h post-treatment (T₂₄), averaged across all metals (titanium alloy, stainless steel, and cobalt chromium).

Treatment Comparison by Metal

The CoCr discs demonstrated a statistically higher bacterial adherence at all time points when compared to both Ti and SS. Thus, further analysis was performed to examine relative treatment effects in reference to the control group for each metal; analysis was organized by metal and treatment. When comparing the control to treatments for Ti, the largest decrease in CFUs at T₀ was a 2.4 log (99.6%) reduction observed with Dakin’s solution and a 1.2 log (94.4%) reduction at T₂₄ with H₂O₂ (Table 1, Fig. 2A). The remaining adherent bacteria was 1.53 × 10⁵ CFU/cm² and 1.63 × 10⁶ CFU/cm², respectively. When comparing the control to treatments for CoCr, the largest reduction was observed with 10% Betadine, in which a 1.9 log (98.7%) reduction was seen at T₀ and a 1.4 log (95.8%) reduction was seen at T₂₄ (Table 1, Fig. 2B). The remaining adherent bacteria was 1.07 × 10⁶ CFU/cm² and 1.74 × 10⁶ CFU/cm², respectively. For SS, the largest reduction in CFUs for treatments as compared to the control was a 2.8 log (99.8%) reduction at T₀ observed with Dakin’s solution and a 0.7 log (80.9%) reduction at T₂₄ with Dakin’s solution (Table 1, Fig. 2C). The remaining adherent bacteria was 7.96 × 10⁴ CFU/cm² and 1.99 × 10⁶ CFU/cm², respectively.

Table 1.

Percent Reduction of CFUs With Initial Treatment Application and Growth Recovery at 24 h

Treatment and Metal	% Reduction from Control at T0	% Reduction from Control at T24	Growth Recovery at T24 After Treatment (log10)
Betadine	98%	89%	0.22
CoCr	98.7%	95.8%	0.21%
SS	97.6%	78.1%	0.30
Ti	97.2%	94.3%	0.14
Dakin’s	91%	71%	0.97
CoCr	74.3	48.2	0.01
SS	99.8	80.9	1.40
Ti	99.6	83.0	1.51
H2O2	97%	84%	0.24
CoCr	95.7%	85.2%	0.24
SS	98.3%	73.9%	0.54
Ti	95.7%	94.4%	0.05b
ClO2	86%	30%	0.23
CoCr	78.5%	64.0%	0.07b
SS	84.4%	54.0%a	0.34
Ti	94.9%	80.5%	0.42

^aDenotes an increase in CFUs compared to control.

^bDenotes a negative growth recovery.

Figure 2.

Percent CFU reduction at initial application of treatment (T₀) for biofilms cultured on Titanium alloy (A), cobalt chrome (B), and stainless steel (C) as compared to saline control (^*p <0.05).

Persistent Effect

There was a trend toward growth recovery with an increase in CFUs when comparing T₀ to T₂₄. No statistically significant persistent effect was found when comparing T₀ to T₂₄ across all metals for all topical applications. However, there was a statistically significant increase seen between T₀ and T₂₄ for Dakin’s solution with titanium and stainless steel.

Antibiotic Resistance Patterns

Across both time points, the antibiotic resistance profiles for S. aureus before and after each treatment for each metal remained the same and showed no changes in resistance profiles.

DISCUSSION

In this in vitro model comparing adjuvant treatment modalities in the elimination of S. aureus biofilm from orthopaedic metals, statistically significant reductions were found with the various topical applications. Immediately following treatment, large numbers of colony forming units remained on the biomaterials, with the largest reduction being a 2.8 log reduction from 10⁷ to 10⁴ with Dakin’s solution treated SS. Following an additional 24 h of incubation after treatment, there were still quantifiable CFUs (Figs. 1 and 2A–C). The results from the 24 h additional incubation further support that these treatments did not demonstrate a sustained effect as there was growth recovery in all cases with the exception of ClO₂ treated SS. The threshold of biofilm for a clinically meaningful reduction is unknown, but prior evidence has shown that CFU levels must be reduced below the minimum infectious dose of S. aureus (10⁴ CFU/gram of tissue) for complex extremity wounds.²⁰ However, it is known that the threshold is substantially lower with hardware present and, as demonstrated, there was an average of 5.79 × 10⁶ CFU/cm² left after treatment. These data are qualitatively supported by S. aureus clusters visible on SEM at T₀ (Fig. 3). Further, antibiotic resistance profiles did not change before or after treatment. This factor is important as we focus on antibiotic stewardship in efforts to prevent antibiotic resistance patterns. Though the results may not be clinically significant, it is yet to be determined if the demonstrated reductions could have an additive effect with other treatments. Given these results, other topical applications should be carefully reconsidered in the future, especially with the rising commercialization of new antimicrobial surfactants and clinical products developed for biofilm reduction.

Figure 3.

SEM of Titanium alloy disc treated with H₂O₂ at T₀.

The limited ability to eradicate biofilms may lead to the high failure rates demonstrated in the literature when utilizing the DAIR method. Literature has shown failure rates ranging from 54 to 84%.^21–24 A study of 30 patients with S. aureus infected prosthetic joints treated with DAIR showed a 54% (16/30) failure rate within one year and a 69% (21/30) failure rate within 2 years.²³ Hartman et al.²⁴ retrospectively reviewed 33 infected total knee arthroplasties treated by irrigation and debridement (I&D) in which there was a 61% (20/33) reinfection rate. Bradbury et al.²² retrospectively reviewed 19 cases of MRSA-infected knee arthroplasties treated with I&D in which 84% (16/19) failed within a two-year follow-up period. Azzam et al.²¹ retrospectively reviewed hip and knee arthroplasty infections treated with I&D in which 56% (58/104) failed within an average follow-up of 5.7 years. Fehring et al.¹² showed that reinfection could occur within 10 days and also reported a 63% (54/86) reinfection rate within 90 days postoperatively. The inability to completely eliminate infection will lead to continued physical, financial, and emotional burdens to patients.²⁵

Limitations of this study consist of the clinical translatability of the results, including biofilm maturation and heterogeneity, biofilm quantification methods, the size, and shape of the coupons as compared to prosthetics, the duration of treatment exposure, and the cytotoxicity of these topical adjuvants to healthy cells. Although it has been shown that mature biofilm can be established in vitro following 48 h incubation, clinical presentations of biofilm-related joint infection generally consist of long-established biofilm.^26–30 However, the clinical application of adjuvant treatments as described here would be considered in the setting of acute infections. Further, this study involved infection with clinical isolates of MSSA, but many biofilms are heterogeneous in nature, a characteristic that poses one of the many challenges in treating biofilm-related PJI.³¹ Based on a previously conducted pilot study, adjuvant treatments administered for 1, 3, and 5 min all showed nearly equal efficacy. Further, biofilm quantification using CFU/cm² may be an underrepresentation of residual contamination following treatment due to persisters in mature biofilm that are difficult to detect. Despite this, CFU/cm² is a currently accepted measure of viable bacteria within biofilm, but likely represents an underestimation of the biofouling present.³² Additionally, the treatment exposure time was set at 5 min as this time point was thought to be a clinically feasible amount of time in the operating room. The treatments may not have had enough time for full efficacy; therefore, other time points (i.e., 15 or 20 min) may be worth investigating in the future.^33,34 It will also be beneficial to evaluate the additive effects of such modalities or additional treatment strategies to aid in the removal of bacterial biofilm. Lastly, while these topical adjuvants reduce bacterial loads, they also have cytotoxic effects to healthy tissue. Brown et al.³⁵ describe the use of dilute betadine when used as a prophylactic measure in the primary arthroplasty setting. This technique balanced the cytotoxic effects of the beta-dine with the beneficial effects of the wash. In infected tissue, the potential cytotoxic effects on local tissue have to be weighed against the benefits from the full strength delivery of these adjuvant washes. Despite these detrimental effects to healthy tissue, in current clinical practice, these undiluted adjuvants are used intraoperatively as a more aggressive treatment approach is necessary to eradicate infection and prevent further sepsis. Multiple studies support the use of these topical applications, despite their cytotoxicity.^33,35,36

In conclusion, the use of adjuvant chemicals led to an average two log reduction in CFU/cm² from orthopaedic implant materials without affecting the resistance patterns of the organism. However, the treatments left quantifiable amounts of biofilm adherent to orthopaedic metals and are likely not sufficient to treat PJI with planned retention of components in isolation.