Abstract
Purpose:
Allograft infection remains the greatest challenge in orthopaedic reconstructive surgery especially methicillin-resistant Staphylococcus aureus (MRSA). This risk can be minimized with the use of antibiotic laden allograft (ALA) via iontophoresis. Ceftaroline fosamil (CF) is an advanced-generation cephalosporin, an alternative treatment for MRSA infections. Its antibacterial activity and safety profile are better than vancomycin. CF iontophoresed bone has not been used before. This study was conducted to establish the feasibility of creating a CF ALA and establish the prime conditions for its expenditure.
Method:
We created an iontophoresis cell; 3% CF was inserted within medullary segment of goat bone and sealed from external saline solution. The cell operated at the following voltages 30, 60 and 90 V and at the following durations 5, 10, 15, 20, 25 and 30 min. Information regarding optimal conditions for its application was then obtained. After which, correlation between voltages and time with CF concentration in the bone was analysed. A bioavailability test was also conducted to observe the optimal rate of CF elution from the graft.
Result:
The optimal condition for the impregnation process is 3% CF at 90 V for 10 min. Bone graft impregnated with CF at optimal conditions can elute above minimum inhibitory concentration of the CF against MRSA for 21 days.
Conclusion:
CF iontophoresis was found feasible for allograft impregnation. The technique is simple, inexpensive and reproducible clinically. Iontophoresis offers a novel solution to reduce the rate of perioperative infection in reconstructive surgery involving use of bone graft.
Keywords
Introduction
Allograft-related infection is a major complication of bone transplant procedure and often leads to revision or amputation in life-threatening situations. It is usually revealed in early post-operative period, with approximately 75% occurring within first 4 months. 1,2 Therefore, various methods were developed to prevent post-operative infection, including the development of antibiotic-impregnated allograft (antibiotic laden allograft (ALA)). Several techniques have been described in literature such as soaking, mixing, pressurization and iontophoresis to produce ALA. 3
Iontophoresis is a process of using electrical current to transport antibiotics with electrically charged properties into the bone. 4,5 It helps to administer high concentrations of antibiotic to the bone, which acquires the ability to elute antibiotics for 2 weeks. 4 Previously published works have used antibiotics such as gentamicin, flucloxacillin and vancomycin. 4,5 ALA is better than the standard allograft as it aids in the prevention of allograft-related infections by eluding high local concentration of antibiotics. 2
The most common cause of orthopaedic infection is Gram-positive microbes and the most devastating by far is methicillin-resistant Staphylococcus aureus (MRSA). 2 Currently, vancomycin is the gold standard for treatment of MRSA infection. 5 The use of vancomycin has increased in the past few decades in the field of orthopaedic oncology and arthroplasty as a mode of prophylaxis and therapeutic treatment. 6
Ceftaroline Fosamil® (CF) was recently introduced for the treatment of MRSA infections after vancomycin, telavancin and linezolid. 5 CF is an innovative broad-spectrum fifth-generation cephalosporin, which has potent activity against MRSA strains attributed to its high affinity for S. aureus. 7,8 CF exhibits antibacterial activity against MRSA that is equivalent to or better than that of vancomycin 9 –12 and has a better safety profile. 13,14 To date, CF has not been employed for iontophoresis of allograft or autograft. We wish to establish the sustainability of this procedure and equate the prime conditions for its expenditure.
Method
For this experimental study, goat bone was used instead of sheep bone, as fresh goat bones are easily available in Malaysia at a reasonable cost. Macrostructurally, goat’s bones represent human bones relatively closely. Summarized in Table 1 are the similarities of animal model to human bone. 15 Iontophoresis of human tibia bone was shown to be comparable with those of the sheep model in rate and uniformity of penetration, although the human tibia has thicker cortices with full penetration occurring at a longer duration. 4 For the aforementioned reasons, goat bone was used. Fresh goat tibia and femur were attained from the wet market.
Comparison of properties between animal and human bone.
a Most similar.
b Moderately similar.
c Least similar.
The study design was modified based on previous study that involved iontophoresis of bone with gentamicin and vancomycin. 4,5
The 20-mm long sections of bone were cut from diaphysis of goat tibia. Bone marrow was removed and intramedullary canal was brushed clean with water. The bone segment was then dried. One end of the bone segment was sealed with acrylic disc and the other end was extended with a 30-mm acrylic tube. Cyanoacrylate glue (DERMABOND™ ProPen Topical Skin Adhesive) was used to attach the acrylic disc and tube to the bone. DERMABOND ProPen is a highly purified 2-octyl cyanoacrylate monomer with an embedded technology of chemical initiator and a unique blend of plasticizers in the applicator tip to ensure consistent, reliable polymerization times and avoid clogging during application. 16 Thus, it is stronger, flexible, has high three-dimensional strength, it is less brittle after polymerization and has a shorter setting time.
The primed bone was then placed in a beaker surrounded by metal mesh to form the anode. The beaker was then filled with phosphate-buffered saline (PBS; pH 7.3) solution up to the level of just below the open end of the acrylic tube. CF was filled into the medullary canal and a metal rod electrode (cathode) was placed centrally into the canal to form an iontophoresis cell with radial geometry. The bone served as a membrane between the two half cells (Figure 1). This allowed the experimental bone segment to be completely submerged in an electrolyte solution while restricting any ion exchange at the bone segment. PBS solution is a water-based salt buffer solution containing sodium phosphate and sodium chloride, commonly used in biological research. The osmolality and ion concentrations of the solutions were physiological compared to the human body and aided in the movement of charged antibiotic particles through the bone segment, as well as not inhibiting the function of the bone cell or affecting its survival.

Iontophoresis cell with complete electrical connection.
The PowerPac™ Basic power supply (Bio-Rad Laboratories Inc., Singapore) for gel electrophoresis was used as the electric source for the iontophoresis process. Double banana plug wire connection to double crocodile clip was utilized to connect the PowerPac to the iontophoresis cell.
It is rationalized that a higher concentration of antibiotic will possess higher quantity of molecules and thus produce a higher concentration gradient across the bone cortex, therefore improving antibiotic loading into the bone. This was reinforced by the vancomycin study that showed increasing concentration improved antibiotic loading into the bone. 5 Hence, no further investigation of varying concentration component was necessary to test the outcome with ceftaroline concentrations of less than 3%.
Qualitative test
Iontophoresis was carried out with a test solution of 3.4 ml of 3% CF with 60 V for 20 min. Following iontophoresis, the bone segments were washed in distilled water and dried. A central ring of 2-mm thickness was cut using a handheld saw. This ring was then placed on an agar plate impregnated with MRSA for qualitative analysis and compared to the negative control, which was bone segment soaked in PBS solution for 20 min, and positive control, which was bone segment soaked with only CF 3% for 20 min without iontophoresis. The zones of inhibition were observed and compared.
Quantitative test
The 20-mm long sections of goat femur bone segment were used for the quantitative study. The optimum concentration for antibiotic was chosen to be 3% ceftaroline as the powder was unable to dissolve fully in PBS solution if the concentration exceeded 3%. Therefore, a maximum concentration of 3% ceftaroline was constantly used at different combinations of voltages and time for iontophoresis. Various voltages (30, 60 and 90 V) and durations (5, 10, 15, 20, 25 and 30 min) were used to make up 18 different combinations of this experiment. The voltages and durations were chosen based on previous study. 4,5 Duration of more than 30 min might not be practical for intraoperative applications. Voltage higher than 90 V may generate more heat and was dangerous to handle clinically.
The iontophoresed bone segment was washed in distilled water and dried. Then, the bone segment was cut in the centre with a handheld saw. Bone dust was grounded from two cut ends. A magnet was passed over the powder to remove impurities from the filling, and five samples of 0.2-g bone powder were collected in different containers. This was followed by the addition of 2-ml PBS solution to 0.2 g of the powdered sample. The samples were then agitated using ultrasound for 30 min. After centrifuging, supernatant fluid was removed and further 2 ml of fluid was added to the agitation, and centrifuging process was repeated twice; the last sample (fourth) was left for 12 h before centrifuging. The specimens were refrigerated at 4°C during this period to prevent degradation. Finally, the supernatant solutions were analysed for CF concentration using high-performance liquid chromatography (HPLC). 17 Technically, the elution samples were diluted with mobile phase and injected into the HPLC system. Duplicated samples were tested and the resulting chromatogram was recorded. The concentration of the released antibiotics was correlated with the contact time with PBS.
Concentration of supernatants for each combination was recorded and calculated. Then, descriptive statistics of the concentration of CF infused into the bone were determined. Test of normality, Shapiro–Wilk test, was applied to determine the nature of distribution of the variables. Then, the first order of correlation between voltage or time duration with the CF concentrations of bone was analysed. If significant correlation was found, it was continued into the second-order correlation analysis between combined voltage and time duration with CF concentrations of bone.
Broth elution assay
For elution study, five samples of iontophoresed bone segments with a 2-mm thickness were prepared. Weight of each bone segment was recorded. Each bone segment was immersed in a polypropylene tube with 20-ml PBS solution and shaken using a rotator at 37°C. The study lasted up to 42 days, where elution samples were collected at days 0, 1, 2, 3, 7, 10, 14, 21, 28 and 42. The supernatant was removed and replaced with 20-ml fresh PBS solution at each time point stated above. The concentrations of CF in elution samples were determined using HPLC.
Results
Qualitative test
Bacterial colonies grew up to the bone in the negative control group (not exposed to antibiotics). In contrast, a zone of inhibition was recorded around specimen iontophoresed with CF. The zone of inhibition around specimens which had been iontophoresed with CF for 20 min was larger compared to CF-soaked specimens for the same duration of time, as shown in Figure 2.

The CF action against MRSA culture in goat femur bone segment showing (a) a control specimen, (b) a 20-min ceftaroline 3% iontophoresed and (c) a 20-min ceftaroline 3% soaked. CF: ceftaroline fosamil; MRSA: Methicillin-resistant Staphylococcus aureus.
Quantitative test
Iontophoresis of bone with ceftaroline at varying voltages and time periods is presented in Table 2. The maximum concentration of ceftaroline per gram bone achievable by the iontophoresis cell was 52,387.05 μg/g and the minimum concentration was 3134.65 μg/g. Minimum inhibitory concentration (MIC) for CF is >2 μg/ml for MRSA. Even at the minimum concentration, our result exceeds more than 1500 times than that of MIC.
Concentration of ceftaroline in the bone when iontophoresed at varying voltages and durations.
SD: standard deviation.
There was no significant relationship between mean of CF concentration in the bone with voltages (r = 0.118, p = 0.641) and time (r = 0.072, p = 0.776). However, the duration of iontophoresis needed to achieve peak concentration at higher voltages was shorter. There was no significant difference at 30 V with increasing time (p = 1.00) and no significant difference at 60 V (p = 1.00); however, at 90 V, there was a significant difference with increasing time (p = −0.0000). Only 10 min was needed for iontophoresis at 90 V to achieve a peak concentration; it also produced the highest CF concentration in the bone among all the combinations tested (Figure 3). There was a significant difference at all the six different time points with increasing voltages: at 5 min (p = −0.00), 10 min (p = −0.0011), 15 min (p = −0.00), 20 min (p = 0.0033), 25 min (p = −0.0001) and 30 min (p = 0.00). Simply meaning, by increasing voltage, a higher concentration of antibiotic can be iontophoresed into bone. It was also observed that CF concentration in the bone continued to reduce after 10 min and achieved a steady plateau at 30 min. The CF molecules were driven into, through and out of the bone as time lapsed. Iontophoresis with 60 and 30 V achieved a peak of concentration later, at 15 and 25 min, respectively.

Graph showing ceftaroline concentration in the bone when iontophoresed at varying voltages and time.
Broth elution assay
The rate of elution over time for ceftaroline-iontophoresed (3% ceftaroline, 90 V and 10 min) bone segments is shown in Table 3 and Figure 4. There was a logarithmic decrease in the rate of elution over time. Antibiotic continued to elute for 21 days (day 21 = 183.47 μg/g) into PBS solution with the concentration levels above MIC of ceftaroline for MRSA.
Mean concentration of CF eluted over 42 days from iontophoresed bone segments (ceftaroline 3%, 90 V and 10 min).a
CF: ceftaroline fosamil; SD: standard deviation; MIC: minimum inhibitory concentration.
a MIC of ceftaroline: >2 μg/g.

Graph showing the CF elution over time. CF: ceftaroline fosamil.
Discussion
Infection remains the greatest challenge for orthopaedic surgery especially arthroplasty revision and limb sparing surgery, involving non-vascularised allograft. 5 Therefore, antibiotic prophylaxis has become the standard of care in these surgeries, either delivered systemically or locally. 18 Antibiotic-loaded bone cement or substitutes have well been established to deliver a higher local concentration of antibiotic into the avascular area without causing systemic toxicity. 19 Autograft and allograft bones are the most successful bone substitutes because none of the other bone substitutes can replace the unique fibrous and mineral composite nature of the bone. 20 Antibiotic-impregnated bone graft via iontophoresis has the potential to incorporate into host bone effectively while delivering high concentrations of local antibiotic. To date, only three other antibiotics have been used for iontophoresis (gentamicin, flucloxacillin and vancomycin). 4,5 However, vancomycin has well-known limitations such as poor activity against methicillin-susceptible Staphylococcus aureus (MSSA), uncertainty regarding the susceptibility of Heterogeneous Vancomycin-intermediate resistance Staphaureus (hVISA) strains, variations within tissue distribution, requirement of Therapeutic drug monitoring (TDM), increase in vancomycin MIC and resistance as well as significant reduction in vancomycin susceptibility. 21 Ceftaroline has fourfold higher in vitro activity against MSSA compared to vancomycin as monotherapy and similar to vancomycin plus aztreonam combination therapy in treating MRSA and MSSA infection without the need for TDM. 13,22 Making ceftaroline a better drug to be iontophoresed into the bone.
Loading allograft with antibiotic does not alter the rate of healing, seen in several animal-based studies that compared allograft with and without antibiotic could not determine any statistically significant difference between ragiographical healing, rate of graft incorporation, histopathological or immunohistochemical parameters. 23 Another study that looked into histology found that necrotic trabecular bone (allograft) was gradually replaced by viable new bone. Despite the concern of iontophoresis causing bone necrosis, these studies proved that ALA is similar to non-ALA and that necrotic bone via creeping substitution is replaced by new bone. 24 Two studies that made use of iontophoresed segmental allografts at 90 V for 20 min showed excellent incorporation rates; 12 patients undergoing two-stage revision for infection at mean follow-up of 47 months, none of the patients had become reinfected and one patient required further bone grafting due to non-union thus retention of 100%. 25 In second study, 28 of the 34 allografts were retained after limb salvage surgery at various sites. The ones requiring revision were due to haematogenous infection. 26 Both iontophoresed and non-iontophorosed allografts have similar properties.
The qualitative study clearly exhibited that iontophoresed bone was more effective against MRSA than antibiotic-soaked bone. From literature, it was noted that different antibiotics required different optimal concentration, voltage and duration of iontophoresis depending on molecular size, for example, gentamicin 1% (molecular weight = 477.6 g/mol) and flucloxacillin 1% (molecular weight = 453 g/mol) iontophoresis with 90 V produced maximum antibiotic loading at 5 min, while vancomycin 5% (molecular weight = 1485.73 g/mol) iontophoresis with 60 V at 20 min contributed to the highest amounts of antibiotic concentration in the bone but optimal condition of vancomycin after analysis was vancomycin 5% at 110 V for 30 min. 4,5 The size of molecules of CF (molecular weight = 762 g/mol) 22 is in between these two molecules; thus, maximum CF loading was between 5 min and 30 min. This was confirmed by the results of our study that concluded the best combination to be, 10 min at 90 V, which was faster than vancomycin at similar conditions. This is probably due to the smaller molecular size of CF and the lack of aggregation of CF in the solution form.
The iontophoresis cell in this study was capable of producing a minimum concentration of 3134.65 μg/g and a maximum concentration of 52,387.05 μg/g. The amount of concentrations achieved by CF iontophoresis is comparable to gentamicin (180 μg/g), flucloxacillin (34 μg/g) and vancomycin (maximum 2240 μg/g; minimum 309 μg/g). 4,5 CF (MIC50/90, 0.25/1 μg/ml) is active against 97.9% of MRSA isolates at ≤1 μg/ml; the highest MIC of CF was >2 μg/ml for MRSA. 27 Even at minimum concentration, the results obtained exceeded the effective MIC of CF by more than 1500 times. With this promising in vitro result, it is possible to eradicate early allograft or autograft infection. The duration of bioavailability of different antibiotics is different. A study on vancomycin found that the highest concentration of antibiotic eluted from the allograft was within the first 2 weeks. 5 From our study, it can be observed that the elution of ceftaroline from iontophoresed bone segments lasted for more than 21 days above MIC. Hence, the use of ceftaroline for local prophylaxis in preventing early post-operative infections is effective up to 21 days. This is probably due to the different molecular properties of different antibiotics and its response towards iontophoresis. With the growing trend of using extracorporeal treatment such as heating, freezing, irradiation or pasteurization to devitalize tumour-infiltrated segment of primary bone malignancy before surgical reimplantation in limb salvage therapy, 28 the application of antibiotic iontophoresis procedure after sterilization and before reimplantation will ensure lower infection rates of the implanted graft. However, further investigations need to be carried out before clinical use.
Conclusion
Iontophoresis is a feasible method to deliver high concentrations of ceftaroline into cortical bone graft. We recommend iontophoresis of 3% CF at 90 V for 10 min to be optimal. The iontophoresis cell is reproducible and easy to be applied in a clinical setting for both autograft and allograft.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
