Antimicrobial impregnated materials (AIMs) are effective in the treatment and prevention of orthopedic infections in both human and veterinary patients.1–4 They may be particularly beneficial in cases of posttraumatic osteomyelitis (PTO) that are challenging due to unique obstacles present within the traumatized bone. Localized ischemia after trauma allows infectious agents to establish infection where systemic administration of antimicrobials may not reach sufficient concentrations to be therapeutic.5–7 Additionally, many cases of PTO demonstrate some degree of antimicrobial resistance,8–10 and the development of biofilms on orthopedic implants makes these bacteria more resistant to antimicrobials than their planktonic counterparts.11,12 The combination of these factors makes systemic administration of antimicrobials alone an inefficient—and in some cases, ineffective—treatment option.13
When implanted into a surgical wound, AIMs demonstrate sustained elution of the antimicrobial at higher local concentrations than would be safely achievable through systemic administration, while concurrent systemic concentrations are low. Thus, susceptibility is increased while the risks for systemic effects and toxicity are minimized.14,15 One study by Nelson and colleagues demonstrated that tobramycin concentrations in wound exudates from experimentally induced osteomyelitis in rabbits reached 11.9 mg/mL when treated with impregnated calcium sulfate (CaSO4) pellets. Serum concentrations reached a maximum of 5.87 µg/mL at 3 hours post-implantation but were undetectable the next day. With intravenous administration of tobramycin, the highest serum concentrations reached 7.82 µg/mL.6 Turner et al. reported similar findings in an experimental study using canine subjects. CaSO4 pellets maintained elevated local levels of tobramycin for at least 14 days while serum levels were elevated for less than a day after implantation.16 As these much higher concentrations at the site of the infection are achieved, organisms that were previously resistant to antimicrobial therapy may become susceptible.17 AIMs are also shown to have some degree of antibiofilm potential.7,18
Of all the bioabsorbable AIMs, CaSO4 beads tend to see the most clinical use. Previous studies have shown rapid initial elution of approximately 80% of the total aminoglycoside load in the first 48 hours, with bactericidal concentrations being maintained for at least 14 days.7,16 However, the specific elution characteristics vary depending on the chosen antimicrobial and its concentration.17 The beads take approximately 2 to 3 months to radiographically resorb.
CaSO4 beads are available from some compounding pharmacies or may be made in-house by veterinarians using CaSO4 purchased “over-the-counter.” However, without guidelines for bead creation and dosing, AIMs produced in-house exhibit a high degree of variability in terms of material purity, drug mass, and elution characteristics when used in clinical cases. A veterinary-specific commercial kit has recently become available containing a mixing guide, medical grade calcium sulfate hemihydrate (CaSO4·1/2 H2O) powder, sterile diluent, a bead mold, and mixing implements (Kerrier, LLC).
Most of the existing literature on AIMs focuses on comparing the elution of different antimicrobials from a given material19–24 or the elution characteristics of different materials.3,25–27 It is difficult to compare outcomes across studies and clinical reports, because the AIMs are created according to study-specific protocols or practitioner discretion, with different drug concentrations and dosing.28 The purpose of this study is to examine the effect that different bead configurations have on antimicrobial elution. Specifically, different bead sizes and changes in antimicrobial concentrations within those beads were evaluated in terms of their effect on elution of amikacin. We hypothesized that larger beads would demonstrate a lower maximum peak elution of amikacin while maintaining therapeutic concentrations for a longer period compared with smaller beads. Additionally, both high- and low-concentration beads would experience a rapid initial elution period followed by a slower elution period, but eluent concentrations would remain higher for the high-concentration beads throughout the elution period.
Materials and Methods
Bead preparation
Amikacin-impregnated CaSO4 beads were created under sterile conditions using commercially available kits (Kerrier LLC). In brief, 15 g of CaSO4·1/2 H2O powder was combined with either 500 mg or 1 g amikacin sulfate (250 mg/mL; Sagent Pharmaceuticals, Inc.), and a sterile saline diluent, as necessary, to have a total of 4 mL liquid in the mixture. The mixtures were then spread slowly over a mold to minimize air trapping. The filled molds were left to set at room temperature. Three bead sizes were created: 3 mm, 5 mm, and 7 mm (Figure 1). There is not currently a mold mat for 7 mm beads, so a custom mold for 7 mm beads was made in-house with proportional dimensions to the 3 mm and 5 mm beads in the Kerrier mat. A master mold for creating the 7 mm bead mold was created with a 3D printer (Original Prusa i3 MK3S, Prusa Research) using acrylonitrile butadiene styrene filament followed by acetone smoothing to achieve the proper cavity dimensions. The mold was filled with commercial all-purpose silicone (Momentive Performance Materials Inc.). Once cured, the silicone mold mat was removed, thoroughly rinsed with water, and sterilized using ethylene oxide (Supplementary Figure S1).
Relative size differences between bead sizes. From left to right: 3 mm, 5 mm, and 7 mm beads. The 7 mm bead was created through proportionally increasing the dimensions of the 5 mm bead; 7 mm bead contain minor surface imperfections secondary to the custom-made mold compared with the manufactured mat provided with the commercial kit. Minor surface imperfections (bubbling or cavitations) were occasionally present on the 3 mm and 5 mm beads.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Relative size differences between bead sizes. From left to right: 3 mm, 5 mm, and 7 mm beads. The 7 mm bead was created through proportionally increasing the dimensions of the 5 mm bead; 7 mm bead contain minor surface imperfections secondary to the custom-made mold compared with the manufactured mat provided with the commercial kit. Minor surface imperfections (bubbling or cavitations) were occasionally present on the 3 mm and 5 mm beads.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Relative size differences between bead sizes. From left to right: 3 mm, 5 mm, and 7 mm beads. The 7 mm bead was created through proportionally increasing the dimensions of the 5 mm bead; 7 mm bead contain minor surface imperfections secondary to the custom-made mold compared with the manufactured mat provided with the commercial kit. Minor surface imperfections (bubbling or cavitations) were occasionally present on the 3 mm and 5 mm beads.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
The beads were configured to meet specifications for a total of 6 experimental groups: low-concentration 3 mm beads (group 3L), high-concentration 3 mm beads (group 3H), low-concentration 5 mm beads (group 5L), high-concentration 5 mm beads (group 5H), low-concentration 7 mm beads (group 7L), and high-concentration 7 mm beads (group 7H). The subcategorization of low- and high-concentration is defined as 500 mg amikacin/15 g CaSO4·1/2 H2O and 1 g amikacin/15 g CaSO4·1/2 H2O, respectively. Control beads were made similarly with CaSO4·1/2 H2O and sterile saline diluent, but with no antimicrobial, and using the 5 mm mold. All bead configurations were formed and evaluated in triplicates made from different batches, as performed in previous studies.19,29
Elution testing
The in vitro elution study was modeled after previous studies.21–23 Based on the calculated amikacin concentration per bead for each group, sufficient beads were selected so that each test tube contained approximately 150 mg of amikacin (groups 3L = 135 beads, group 3H = 67 beads, group 5L = 37 beads, group 5H = 19 beads, group 7L = 11 beads, group 7H = 6 beads). The beads were placed in 6 mL of phosphate-buffered saline (PBS; pH 7.4) in individual 10 mL glass test tubes (Figure 2). The tubes were incubated at 37 °C with constant agitation. PBS sampling from each tube, including controls, occurred at the following time points: 1, 3, 6, 12, and 24 hours, and at 2, 4, 6, 9, 12, 16, 20, 24, and 28 days. With each sampling, 6 mL of the eluent solution was removed and replaced by 6 mL of fresh PBS. One-half milliliter aliquots of each eluent sample were transferred into polypropylene tubes and stored at −20 °C until assayed.
Beads in test tubes with 6 mL of phosphate buffered saline for each group. The total bead content of each tube contains approximately 150 mg of amikacin. From left to right: 3 mm low-concentration (3L), 3 mm high-concentration (3H), 5 mm low-concentration (5L), 5 mm high-concentration (5H), 7 mm low-concentration (7L), and 7 mm high-concentration (7H).
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Beads in test tubes with 6 mL of phosphate buffered saline for each group. The total bead content of each tube contains approximately 150 mg of amikacin. From left to right: 3 mm low-concentration (3L), 3 mm high-concentration (3H), 5 mm low-concentration (5L), 5 mm high-concentration (5H), 7 mm low-concentration (7L), and 7 mm high-concentration (7H).
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Beads in test tubes with 6 mL of phosphate buffered saline for each group. The total bead content of each tube contains approximately 150 mg of amikacin. From left to right: 3 mm low-concentration (3L), 3 mm high-concentration (3H), 5 mm low-concentration (5L), 5 mm high-concentration (5H), 7 mm low-concentration (7L), and 7 mm high-concentration (7H).
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Concentrations of amikacin from the eluent samples were determined using ultra-performance liquid chromatography with tandem mass spectrometry (Waters Acquity UPLC with Xevo TQD, Water Corporation) using a method modified from previous reports.30,31 Full method details are available elsewhere (Supplementary Appendix S1). One aliquot from each sample was randomly selected and defrosted. Twenty microliters from the aliquot were obtained for testing and diluted in water plus 0.01% formic acid as needed to bring concentrations within the calibration curve. For this study, the therapeutic concentration of amikacin was defined as 40 µg/mL.
Statistical methods
Replication of elution groups in triplicate, with samples obtained from each of the three test tubes in each experimental group, has been performed in similar studies.19,21,23,29 When normal probability plots showed that data followed a normal distribution, they were summarized as mean (standard deviation). When they were found to follow a non-normal distribution, they were summarized as median (range). Effects of amikacin concentration within the beads and bead size on the elution concentrations were assessed using 2-way repeated measures analysis of variance (ANOVA). Factors in the ANOVA model were amikacin concentration and bead size. The repeated measure was specified as time. Subsequently, amikacin concentrations were compared within bead sizes, and bead sizes were compared within amikacin concentrations. The duration of the therapeutic window was compared between bead sizes within each level of amikacin concentration using the Kruskal-Wallis test. The duration of the therapeutic window was compared between amikacin concentrations within each bead size using the Wilcoxon rank sum test. Statistical significance was set to P < .05. All analyses were performed using SAS version 9.4.
Results
Bead dimensions
The calculated bead surface area was 38.2 mm2, 96.6 mm2, and 203.4 mm2 for the 3 mm, 5 mm, and 7 mm sizes, respectively. This measurement was based on the gross dimensions of ideal beads with no grossly apparent surface defects and does not factor in microscopic differences that may be present on the bead surface, such as microscopic porosity. The calculated bead volume was 19.8 mm3, 79.9 mm3, and 243.7 mm3 for the 3 mm, 5 mm, and 7 mm sizes, respectively, and was also based on the same ideal gross bead dimensions. The bead surface area-to-volume ratios were 1.93, 1.21, and 0.84 for the 3 mm, 5 mm, and 7 mm sizes, respectively.
Peak concentrations
Peak elution concentrations occurred during the first hour of sampling for each group. At 1 hour, the peak mean amikacin concentrations were 20.5 (±1.62) mg/mL for group 3L, 27.4 (±11.7) mg/mL for group 3H, 13.1 (±0.86) mg/mL for group 5L, 14.0 (±0.60) mg/mL for group 5H, 8.85 (±1.46) mg/mL for group 7L, and 6.75 (±1.99) mg/mL for group 7H (Figures 3 and 4). Controlling for bead size, antimicrobial concentration within the beads significantly affected the peak eluent concentration for the 3 mm beads (P < .0001), but not for the 5 mm beads (P = .448) or 7 mm beads (P = .058). The highest elution concentrations were reached by the 3 mm beads in both the low- and high-concentration categories, followed by the 5 mm beads, then the 7 mm beads. Controlling for bead concentration, peak eluent concentrations were significantly different for all bead sizes compared (P < .0006 for the 5 mm and 7 mm beads at low concentration, P < .0001 for all other comparisons).
Average eluent concentration of amikacin measured over time for each of the low concentration bead groups: 3 mm (red), 5 mm (blue), 7 mm (black), and control beads (green). Peak elution concentrations were measured in the first hour with a rapid decrease in eluent concentrations and a secondary prolonged elution phase with a smaller secondary peak occurring for the 5 mm and 7 mm beads groups. *Indicates significance for all groups at this time point. †Indicates that only group 3L was significantly different from the other groups at this time point. Groups 5L and 7L were not statistically different from each other.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average eluent concentration of amikacin measured over time for each of the low concentration bead groups: 3 mm (red), 5 mm (blue), 7 mm (black), and control beads (green). Peak elution concentrations were measured in the first hour with a rapid decrease in eluent concentrations and a secondary prolonged elution phase with a smaller secondary peak occurring for the 5 mm and 7 mm beads groups. *Indicates significance for all groups at this time point. †Indicates that only group 3L was significantly different from the other groups at this time point. Groups 5L and 7L were not statistically different from each other.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average eluent concentration of amikacin measured over time for each of the low concentration bead groups: 3 mm (red), 5 mm (blue), 7 mm (black), and control beads (green). Peak elution concentrations were measured in the first hour with a rapid decrease in eluent concentrations and a secondary prolonged elution phase with a smaller secondary peak occurring for the 5 mm and 7 mm beads groups. *Indicates significance for all groups at this time point. †Indicates that only group 3L was significantly different from the other groups at this time point. Groups 5L and 7L were not statistically different from each other.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average eluent concentration of amikacin measured over time for each of the high concentration bead groups: 3 mm (red), 5 mm (blue), 7 mm (black), and control beads (green). Peak elution concentrations were measured in the first hour with a rapid decrease in eluent concentrations and a secondary prolonged elution phase with a smaller secondary peak. *Indicates significance for all groups at this time point. ‡Indicates that only group 7H was significantly different from the other groups at this time point. Groups 3H and 5H were not statistically different from each other.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average eluent concentration of amikacin measured over time for each of the high concentration bead groups: 3 mm (red), 5 mm (blue), 7 mm (black), and control beads (green). Peak elution concentrations were measured in the first hour with a rapid decrease in eluent concentrations and a secondary prolonged elution phase with a smaller secondary peak. *Indicates significance for all groups at this time point. ‡Indicates that only group 7H was significantly different from the other groups at this time point. Groups 3H and 5H were not statistically different from each other.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average eluent concentration of amikacin measured over time for each of the high concentration bead groups: 3 mm (red), 5 mm (blue), 7 mm (black), and control beads (green). Peak elution concentrations were measured in the first hour with a rapid decrease in eluent concentrations and a secondary prolonged elution phase with a smaller secondary peak. *Indicates significance for all groups at this time point. ‡Indicates that only group 7H was significantly different from the other groups at this time point. Groups 3H and 5H were not statistically different from each other.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Cumulative elution
After the initial peak, there was a rapid decline in eluent concentration until approximately 6–12 hours, after which all groups, except for the group 3L, exhibited a smaller, secondary peak between 24 and 96 hours followed by a more gradual decline in eluent concentration before leveling off between 6 and 9 days (Figures 3 and 4). The beads with a higher amikacin concentration were observed to have lower troughs between these 2 peaks followed by higher secondary peaks when compared with beads of the same size with lower concentrations of amikacin, though these differences in concentration were not statistically significant. Mean cumulative elution at 24 hours for groups 3L, 3H, 5L, 5H, 7L, and 7H was 191.8 ± 18.4 mg, 201.5 ± 67.1 mg, 122.3 ± 4.9 mg, 123.2 ± 3.9 mg, 100.1 ± 11.0 mg, and 76.9 ± 10.9 mg, respectively. Given that approximately 150 mg of amikacin was incorporated into the beads for each experimental group, the mean percent cumulative elution at 24 hours was 128 ± 12.2%, 134 ± 44.7%, 82.2 ± 3.2%, 82.2 ± 2.6%, 66.7 ± 7.3%, and 51.3 ± 7.2%, of the total amikacin mass incorporated into each group, respectively. Mean cumulative total elution at 28 days for groups 3L, 3H, 5L, 5H, 7L, and 7H was 195.8 ± 19.0 mg, 212.0 ± 68.1 mg, 129.6 ± 2.6 mg, 145.3 ± 5.2 mg, 120.6 ± 10.0, and 130.2 ± 11.9 mg, respectively. These amount to 130.7 ± 12.7%, 141.4 ± 45.4%, 86.6 ± 1.6%, 96.9 ± 3.4%, 80.5 ± 6.7%, and 86.9 ± 8.0% of the total amikacin mass incorporated into each group, respectively (Figure 5).
Average cumulative elution of amikacin for each group. The total amikacin eluted for each sampling point was calculated based on the product of the eluent concentration measured at that time and the total volume of phosphate buffered saline contained within each test tube (6 mL). Note that while all groups contained approximately 150 mg of amikacin, the cumulative elution did not equal 150 mg, indicated by the dotted line (.....), for any group.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average cumulative elution of amikacin for each group. The total amikacin eluted for each sampling point was calculated based on the product of the eluent concentration measured at that time and the total volume of phosphate buffered saline contained within each test tube (6 mL). Note that while all groups contained approximately 150 mg of amikacin, the cumulative elution did not equal 150 mg, indicated by the dotted line (.....), for any group.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average cumulative elution of amikacin for each group. The total amikacin eluted for each sampling point was calculated based on the product of the eluent concentration measured at that time and the total volume of phosphate buffered saline contained within each test tube (6 mL). Note that while all groups contained approximately 150 mg of amikacin, the cumulative elution did not equal 150 mg, indicated by the dotted line (.....), for any group.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Therapeutic duration
The last measured concentration above the therapeutic threshold was 47.0 ± 4.6 µg/mL at a median of 6 days (range 6–9 days) for group 3L, 51.0 ± 13.2 µg/mL at 6 days for all replicates in group 3H, 100.0 ± 42.0 µg/mL at a median of 6 days (range 4–6 days) for group 5L, 160.0 ± 32.0 µg/mL at 6 days for all replicates in group 5H, 94.7 ± 82.6 µg/mL at a median of 9 days (range 6–9 days) for group 7L, and 104.0 ± 20.0 µg/mL at 9 days for all replicates in group 7H (Figure 6). For the low-concentration beads, bead size did not significantly affect the therapeutic duration (3 mm vs 5 mm beads: P = .297, 3 mm vs 7 mm beads: P = .551, 5 mm vs 7 mm beads: P = .101). For the high-concentration beads, bead size significantly affected therapeutic duration. The 7 mm beads, when compared with 3 mm (P = .044) and 5 mm (P = .044) had a longer therapeutic duration. The therapeutic duration was not different between 3 mm and 5 mm beads (P = 1.000).
Average eluent concentration of amikacin measured over time for all groups. Logarithmic scaling in the y-axis is used to represent the therapeutic duration more clearly for each group. The therapeutic threshold (40 µg/mL) is also notated by the solid line (___). §Indicates the final supratherapeutic measurement for groups 3L, 3H, 5L, and 5H. ¶Indicates the final supratherapeutic measurement for groups 7L and 7H; however, this was only significant among the high-concentration bead group.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average eluent concentration of amikacin measured over time for all groups. Logarithmic scaling in the y-axis is used to represent the therapeutic duration more clearly for each group. The therapeutic threshold (40 µg/mL) is also notated by the solid line (___). §Indicates the final supratherapeutic measurement for groups 3L, 3H, 5L, and 5H. ¶Indicates the final supratherapeutic measurement for groups 7L and 7H; however, this was only significant among the high-concentration bead group.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
Average eluent concentration of amikacin measured over time for all groups. Logarithmic scaling in the y-axis is used to represent the therapeutic duration more clearly for each group. The therapeutic threshold (40 µg/mL) is also notated by the solid line (___). §Indicates the final supratherapeutic measurement for groups 3L, 3H, 5L, and 5H. ¶Indicates the final supratherapeutic measurement for groups 7L and 7H; however, this was only significant among the high-concentration bead group.
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0216
When comparing low- and high-concentration beads within a given bead size, a significant difference was not detected in therapeutic duration for any subcategory (P = .535). Generally, throughout the elution period, there was not a significant difference in the eluent concentration between the low- and high-concentration configurations. For the 3 mm bead groups, the eluent concentration was only significantly different at 1 hour (low: 20.5 mg/mL, high: 27.4 mg/mL, P < .0001) and 3 hours (low: 5.93 mg/mL, high: 3.08 mg/mL, P = .0113). A significant difference was not detected at any time point for the 5 mm beads between the low- and high-concentration groups. For the 7 mm bead groups, the only time point in which the eluent concentration was significantly different throughout the study period was at 4 days (low: 1.29 mg/mL, high: 4.18 mg/mL, P = .0101). Comparison of the median therapeutic duration between all 6 combinations of concentration and bead sizes tested showed a significant difference between the high-concentration, 7 mm beads (group 7H; 9 days for all 3 replicates) and the low-concentration, 5 mm beads (group 5L; 6 days, range 4–6 days; P = .016). The therapeutic duration between all other group comparisons was not significantly different.
Discussion
In this in vitro antimicrobial elution study, 3 sizes of CaSO4 beads were impregnated with amikacin sulfate at low- and high-concentrations. The eluent concentrations of amikacin peaked in the first hour of sampling and rapidly declined thereafter. Peak eluent concentrations reached between 169 and 684 times the therapeutic threshold designated for this study. Consistent with our first hypothesis, smaller beads containing higher concentrations generally reach higher peak eluent concentrations compared with larger beads containing lower concentrations. The therapeutic duration was also affected by bead size. However, this was only statistically significant for the higher-concentration beads, so we partially accept our first hypothesis. The duration of therapeutic amikacin elution for high-concentration beads was not different from lower-concentration beads of the same size. Therefore, we reject our second hypothesis.
The differences observed in elution between the bead sizes may be explained as a product of changes in the surface area-to-volume (SA:V) ratio. Antimicrobial diffusion into the surrounding aqueous environment may occur wherever the bead interfaces with its environment. Beads with a higher surface area relative to their volume will present more opportunity for their contained drug to be exposed to the environment and, therefore, elution would be expected to occur more rapidly than beads with a lower SA:V. Logic would then suggest that high SA:V beads reach higher initial peak eluent concentrations compared with lower SA:V beads during the same time period and would also completely elute the contained antimicrobial more rapidly. Calculation of the surface area and volume in this study was simplified by utilizing gross dimensions of the beads and do not take into account the microscopic texture and porosity of the beads, which may change as a result of antimicrobial choice and concentration, as well as the environmental circumstances of bead formation.21,27,32 All beads were created under sterile conditions in the same controlled local environment, so the microscopic topography of the beads may be safely assumed to be comparable among the beads with similar drug concentrations. Changes caused by differing concentrations of amikacin sulfate on the crystalline structure of the CaSO4 beads, while not directly characterized, may be reflected in the differences in elution.
In the present study, 3 mm beads had the highest SA:V and 7 mm beads had the lowest. Consistent with our hypothesis, 3 mm beads having the highest SA:V reached the highest peak concentrations, followed respectively by 5 mm and 7 mm beads as the SA:V decreased. The only size category that demonstrated a significantly longer therapeutic duration than the other bead sizes was the 7 mm bead at the higher antimicrobial concentration. The low-concentration 7 mm beads did demonstrate a longer therapeutic window than the smaller low-concentration beads when the raw data was first observed, but it did not reach statistical significance. We can only partially accept our first hypothesis because while larger bead sizes reached lower peak concentrations compared with small beads, bead size only affected the therapeutic duration among the high-concentration beads. It may be tempting based on this result to conclude that the combination of larger bead size and higher concentration is required to prolong the therapeutic window and that bead size alone does not produce this effect. However, it should be noted that when all 6 experimental groups were compared, combining the contributions of both bead size and antimicrobial concentration, there was no significant difference between the groups. It is possible that alterations to our methods may have aided in making differences that exist between the groups more apparent. For example, more frequent sampling throughout the study period also may have more precisely distinguished the end of the therapeutic window of each bead size for both antimicrobial concentrations. The use of lower volumes of PBS per test tube or sampling lower volumes at a time may have prolonged the elution window, making the differences between groups more apparent. Additionally, by having only 3 replicates for each of the 6 groups, despite low variability between those replicates, it is difficult to detect differences in therapeutic duration between the bead sizes with confidence, introducing the possibility of type II error.
Interestingly, antimicrobial concentration within the beads only affected the peak concentration reached among the 3 mm beads, but not among the other bead sizes, and it did not affect the therapeutic duration. With few exceptions where a difference was detected, the high-concentration beads did not maintain a higher eluent concentration than the low-concentration beads throughout the elution period. Therefore, we reject our second hypothesis. Given that each test tube contained approximately 150 mg of amikacin, the number of beads in each test tube differed for each group. Considering the effect that differing bead counts might have on SA:V and elution, the product of SA:V and the number of beads per tube might be considered to describe the total SA:V for each group. This product is approximately 261, 129, 45, 23, 9, and 5 for groups 3L, 3H, 5L, 5H, 7L, and 7H, respectively. If the previously discussed effect of SA:V on elution were to hold true, the lower-concentration beads (higher SA:V) might be expected to elute antimicrobial more rapidly than their higher-concentration (lower SA:V) counterparts, as the SA:V product is higher for low-concentration beads. However, this effect seems to be confounded by the low-concentration beads having a higher CaSO4·1/2 H2O matrix-to-antimicrobial ratio. For the low-concentration beads, this doubles the amount of CaSO4·1/2 H2O that the surrounding fluid must penetrate to reach the antimicrobial, slowing release of the drug from the beads.
The mean cumulative elution of amikacin from each of the groups did not equal 150 mg for any group. For the 3 mm beads, it exceeded 150 mg, while both 5 mm and 7 mm beads did not reach 150 mg, despite the elution rate approaching zero for all groups (Figure 5). The 5 mm beads most closely approximated 150 mg elution of the total drug content. Previous elution studies have observed similar phenomena.19,21,22,29 The total amount of antimicrobial for each group approximated as closely as possible, but did not exactly match, 150 mg per test tube. This is because, with defects within the surface of individual beads, each bead cannot be guaranteed to contain the expected amount of amikacin. However, this variability is likely too small to account for all of the variability in cumulative elution observed. It may be reasoned that some of the amikacin was retained within the larger beads where the PBS could not sufficiently diffuse into the bead, preventing the drug from diffusing out into the surrounding fluid. The calculated release of >150 mg of the antimicrobial with the 3 mm beads suggests that another factor may account for these discrepancies. The dilution process during sample analysis to bring sample concentrations within the calibration curve would have differed between the bead sizes and may have introduced artifactual variability to the measurements.21 The primary advantage of CaSO4 beads over polymethylmethacrylate (PMMA), which historically was the primary AIM used clinically, is that CaSO4 beads are bioabsorbable.16,17,21,27,28 As the CaSO4 erodes and is absorbed by the body, any retained antimicrobial will be completely released, negating the retention of the drug deep within the matrix and prolonging elution of the drug until complete absorption of the bead occurs.17,21 In contrast to implantation into living tissue, minimal resorption of the beads within the test tube was observed throughout the study period.
The ideal dosing of different antimicrobials with AIMs has yet to be determined. One strategy is to implant sufficient beads such that the total amount of drug delivered is less than or equal to the recommended daily systemic dose of a given antimicrobial.4,16,33 This ensures that the total delivered drug will be unlikely to cause toxicity, regardless of the rate of release from the AIM. However, Turner et al. demonstrated safety in canine subjects receiving 36 mg/kg of tobramycin using CaSO4 pellets as the delivery material.16 This equated to approximately 2.5 times the daily recommended dose34 and remained above bactericidal concentrations for at least 14 days, yet the tobramycin was not detectable within the serum 24 hours after implantation.16 Caution in dosing should be exercised, however, as there have been reports in humans of renal toxicity when high doses of aminoglycosides were administered using AIMs, and a rapid rate of antimicrobial release was implicated.35,36 Toxicity of the local tissues has also been reported in horses with regional perfusion of high concentrations of gentamicin37 and human osteoblastic cell lines have demonstrated toxicity at extremely high doses of tobramycin.38
Compared with other antimicrobial classes, there are several advantages of aminoglycosides that make them favorable for use in AIMs. Historically, PMMA beads saw the most clinical use for this purpose and aminoglycosides were not denatured by the highly exothermic polymerization process.15,17,39,40 The small chemical structure and hydrophilic nature of aminoglycosides allow them to diffuse more readily and completely into a surrounding aqueous environment than larger or more hydrophobic antimicrobials.17,41 The relatively broad spectrum of aminoglycoside activity is improved at the extremely high local concentrations achieved with AIMs.17 Aminoglycoside antimicrobial should reach concentrations 8–10 times greater than the bacterial breakpoint MIC for maximum effectiveness.21,22,42 Thus, the therapeutic concentration of amikacin for this study was defined as 40 µg/mL, which represents 10 times the breakpoint MIC of amikacin (4 µg/mL) as published by the Clinical Laboratory Standards Institute (CLSI) for bacteria commonly involved in post-traumatic osteomyelitis in dogs and cats22,43 such as Staphylococcus spp., Escherichia coli, and Pseudomonas aeruginosa.
Compared with other materials, CaSO4 has several advantages, leading to it becoming a popular choice for local antimicrobial delivery. Its bioabsorbable nature compared with PMMA is one reason for this. Additionally, bone cements (like PMMA), are poorly hydrophilic and likely have less porosity to allow for aqueous penetration, decreasing drug release. The exothermic polymerization of PMMA makes it suitable as an AIM only for heat-stable drugs, while the low curing temperature of CaSO4 allows it to carry a wider range of antimicrobials.27,44 Calcium phosphate ceramics, like hydroxyapatite, offer many benefits as an antimicrobial carrier over CaSO4 including improved elution characteristics for many antimicrobials, and its close resemblance to the inorganic matrix of bone.2,17 Expense is the primary impediment to its widespread use for local antimicrobial delivery.7 Wound drainage is a concern when using CaSO4 as an AIM. CaSO4 purchased over-the-counter for this use is produced from naturally occurring gypsum. Gypsum processing into CaSO4 powder creates a product that contains residual by-products that may stimulate an inflammatory response, increasing fluid production at the site, diluting the antimicrobial, and increasing drainage.44,45 The high purity, medical-grade CaSO4 available in commercial kits contains little, if any, of these by-products.46,47 In the authors’ clinical experience, minimal wound drainage is observed when using commercial kits.
Further in vivo studies are needed to more completely characterize the effects of AIM configuration and dosing on antimicrobial elution. Understanding such effects will allow practitioners to utilize AIMs more uniformly so that outcomes may become more consistent and predictable.
This study has several limitations. Firstly, the elution characteristics of an in vitro study cannot be expected to directly represent what will be observed in clinical patients. This has proven problematic for many in vitro elution studies and has led in some cases to the conclusion that some biomaterials or biomaterial/antimicrobial combinations are not suitable for clinical use, while clinical reports in human and veterinary patients have demonstrated their utility.13,17,48 The total therapeutic duration for the different groups reported in the present study ranged from 6–9 days. Other in vitro studies have reported both longer or shorter therapeutic durations, but all of these likely underestimate the true therapeutic duration in vivo.16,48,49 Thus, the authors suspect that the true therapeutic durations of the beads in this study used in vivo would be longer, and it is possible that more pronounced differences would be observed in elution between bead sizes and antimicrobial concentration per bead. To reduce animal use and morbidity, the authors chose to perform in vitro testing first to identify patterns in concentration and bead size combinations that may be the most clinically applicable and allow the use of a smaller number of animal groups when moving to in vivo testing.
Another limitation is the use of only one antimicrobial, amikacin. Antimicrobial elution characteristics vary based largely on the chemical properties of the drug itself, with the relative water solubility of each drug acting as a major determinant. Amikacin was chosen because aminoglycosides are frequently selected for use with AIMs due to their hydrophilicity, broad-spectrum, good bone penetration, and long-term heat stability. Because the focus of the study was to evaluate the effects of the configurations of the delivery vehicle itself, testing the elution characteristics of different antimicrobials was beyond the scope of this study and additional studies are necessary to determine if these same findings are reflected with different antimicrobial choices. This study is also limited in its estimation of antimicrobial effectivity by referencing the CLSI standards for breakpoint concentrations of certain bacteria commonly isolated from osteomyelitis cases rather than using direct in vivo or in vitro testing against various live cultures. While some in vitro studies have proceeded to perform susceptibility testing with their elution samples on live cultures, others have used these published standards as an estimate and point of reference. The authors chose the latter approach, as proving antimicrobial susceptibility will be more clinically applicable when demonstrated in future in vivo testing.
In this in vitro study, we found that amikacin-impregnated CaSO4 beads eluted the antimicrobial reaching peak concentrations much higher than what would typically be required to be therapeutic. Peak concentrations were highest when the beads were smaller, but all bead sizes reached peak concentrations that were orders of magnitude higher than the therapeutic threshold. The therapeutic duration was prolonged with larger beads, though this was only statistically significant at the higher concentration. Antimicrobial concentration within the beads did not have a significant effect on elution. Based on these findings and our clinical experience, the authors recommend that the largest beads that may be reasonably used in a given clinical situation should be selected if a longer therapeutic window is desired. While the authors also advocate for a higher concentration of antimicrobials per bead, there does not appear to be a specific advantage to this practice aside from reducing the total number of beads that will be implanted into the site of interest and may be desirable if the available space for bead implantation is limited.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org
Acknowledgments
The authors thank the Veterinary Memorial Fund, Virginia-Maryland College of Veterinary Medicine for financial support of this project and Kerrier, LLC for providing Absorbable Bead Kits for this project.
The authors declare that there were no conflicts of interest.
References
- 1.↑
Wenke J, Owens B, Svoboda S, Brooks D. Effectiveness of commercially-available antibiotic-impregnated implants. J Bone Jt Surg. 2006;88(8):1102–1104. doi:10.1302/0301-620x.75b1.8380599
- 2.↑
Korkusuz F, Uchida A, Shinto Y, Araki N, Inoue K, Ono K. Experimental implant-related osteomyelitis treated by antibiotic-calcium hydroxyapatite ceramic composites. J Bone Jt Surg. 1993;75(1):111–114. doi:10.1302/0301-620x.75b1.8380599
- 3.↑
Anderson BH, Ethell MT. Modes of local drug delivery to the musculoskeletal system. Vet Clin North Am Equine Pract. 1999;15(3):603–622. doi:10.1016/S0749-0739(17)30135-9
- 4.↑
Ham K, Griffon D, Seddighi M, Johnson AL. Clinical application of tobramycin-impregnated calcium sulfate beads in six dogs (2002–2004). J Am Anim Hosp Assoc. 2008;44(6):320–326. doi:10.5326/0440320
- 5.↑
Lew DP, Waldvogel FA. Osteomyelitis. The Lancet. 2004;364(9431):369–379. doi:10.1016/S0140-6736(04)16727-5
- 6.↑
Nelson CL, McLaren SG, Skinner RA, Smeltzer MS, Thomas JR, Olsen KM. The treatment of experimental osteomyelitis by surgical debridement and the implantation of calcium sulfate tobramycin pellets. J Orthop Res. 2002;20(4):643–647. doi:10.1016/S0736-0266(01)00133-4
- 7.↑
Santschi EM, McGarvey L. In vitro elution of gentamicin from plaster of Paris beads. Vet Surg. 2003;32(2):128–133. doi:10.1053/jvet.2003.50010
- 8.↑
Turk R, Singh A, Weese JS. Prospective surgical site infection surveillance in dogs. Vet Surg. 2015;44(1):2–8. doi:10.1111/j.1532-950X.2014.12267.x
- 9.
Nazarali A, Singh A, Moens NM, et al. Association between methicillin-resistant Staphylococcus pseudintermedius carriage and the development of surgical site infections following tibial plateau leveling osteotomy in dogs. J Am Vet Med Assoc. 2015;247(8):909–916. doi:10.2460/javma.247.8.909
- 10.↑
Nicoll C, Singh A, Weese JS. Economic impact of tibial plateau leveling osteotomy surgical site infection in dogs. Vet Surg. 2014;43(8):899–902. doi:10.1111/j.1532-950X.2014.12175.x
- 11.↑
Clutterbuck AL, Cochrane CA, Dolman J, Percival SL. Evaluating antibiotics for use in medicine using a poloxamer biofilm model. Ann Clin Microbiol Antimicrob. 2007;6(1):2. doi:10.1186/1476-0711-6-2
- 12.↑
Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of persistent infections. Science. 1999;284(5418):1318–1322. doi:10.1126/science.284.5418.1318
- 13.↑
El-Husseiny M, Patel S, MacFarlane R, Haddad F. Biodegradable antibiotic delivery systems. J Bone Jt Surg. 2011;93(2):151–157. doi:10.1302/0301-620x.93b2.24933
- 14.↑
Robinson D. Osteomyelitis and implant-associated infections. In: Johnston SA, Tobias KM, eds. Veterinary Surgery Small Animal. 2 ed. Elsevier; 2017:2185–2207.
- 15.↑
Gogia JS, Meehan JP, Di Cesare PE, Jamali AA. Local antibiotic therapy in osteomyelitis. Thieme Medical Publishers; 2009:100. doi:10.1055/s-0029-1214162
- 16.↑
Turner TM, Urban RM, Hall DJ, Chye PC, Segreti J, Gitelis S. Local and systemic levels of tobramycin delivered from calcium sulfate bone graft substitute pellets. Clin Orthop Relat Res. 2005;437:97–104. doi:10.1097/01.blo.0000175127.37343.0d
- 17.↑
Streppa HK, Singer MJ, Budsberg SC. Applications of local antimicrobial delivery systems in veterinary medicine. J Am Vet Med Assoc. 2001;219(1):40–48. doi:10.2460/javma.2001.219.40
- 18.↑
Moley JP, McGrath MS, Granger JF, Stoodley P, Dusane DH. Reduction in Pseudomonas aeruginosa and Staphylococcus aureus biofilms from implant materials in a diffusion dominated environment. J Orthop Res. 2018;36(11):3081–3085. doi:10.1002/jor.24074
- 19.↑
Ethell MT, Bennett RA, Brown MP, Merritt K, Davidson JS, Tran T. In vitro elution of gentamicin, amikacin, and ceftiofur from polymethylmethacrylate and hydroxyapatite cement. Vet Surg. 2000;29(5):375–382. doi:10.1053/jvet.2000.7535
- 20.
Wichelhaus T, Dingeldein E, Rauschmann M, et al. Elution characteristics of vancomycin, teicoplanin, gentamicin and clindamycin from calcium sulphate beads. J Antimicrob Chemother. 2001;48(1):117–119. doi:10.1093/jac/48.1.117
- 21.↑
Phillips H, Boothe DM, Bennett RA. Elution of clindamycin and enrofloxacin from calcium sulfate hemihydrate beads in vitro. Vet Surg. 2015;44(8):1003–1011. doi:10.1111/vsu.12415
- 22.↑
Phillips H, Boothe DM, Shofer F, Davidson JS, Bennett RA. In vitro elution studies of amikacin and cefazolin from polymethylmethacrylate. Vet Surg. 2007;36(3):272–278. doi:10.1111/j.1532-950X.2007.00262.x
- 23.↑
Atilla A, Boothe HW, Tollett M, et al. In vitro elution of amikacin and vancomycin from impregnated plaster of Paris beads. Vet Surg. 2010;39(6):715–721. doi:10.1111/j.1532-950X.2009.00632.x
- 24.↑
Laycock PA, Cooper JJ, Howlin RP, Delury C, Aiken S, Stoodley P. In vitro efficacy of antibiotics released from calcium sulfate bone void filler beads. Materials. 2018;11(11):2265. doi:10.3390/ma11112265
- 25.↑
Nelson CL, Griffin FM, Harrison BH, Cooper RE. In vitro elution characteristics of commercially and noncommercially prepared antibiotic PMMA beads. Clin Orthop Relat Res. 1992;Nov.(284):303–309.
- 26.
Liu SJ, Ueng SWN, Chan EC, et al. In vitro elution of vancomycin from biodegradable beads. J Biomed Mater Res A. 1999;48(5):613–620. doi:10.1002/(SICI)1097-4636(1999)48:5<613::AID-JBM4>3.0.CO;2-%23
- 27.↑
Udomkusonsri P, Kaewmokul S, Arthitvong S, Phaochoosak N. Elution profiles of cefazolin from PMMA and calcium sulfate beads prepared from commercial cefazolin formulations. J Vet Med Sci. 2012;74(3):301–305. doi:10.1292/jvms.11-0095
- 28.↑
Peterson LC, Kim SE, Lewis DD, Johnson MD, Ferrigno CR. Calcium sulfate antibiotic-impregnated bead implantation for deep surgical site infection associated with orthopedic surgery in small animals. Vet Surg. 2021;50(4):748–757. doi:10.1111/vsu.13570
- 29.↑
Tulipan RJ, Phillips H, Garrett LD, Dirikolu L, Mitchell MA. Elution of platinum from carboplatin-impregnated calcium sulfate hemihydrate beads in vitro. Am J Vet Res. 2016;77(11):1252–1257. doi:10.2460/ajvr.77.11.1252
- 30.↑
Díez C, Guillarme D, Spörri AS, et al. Aminoglycoside analysis in food of animal origin with a zwitterionic stationary phase and liquid chromatography–tandem mass spectrometry. Anal Chim Acta. 2015;882:127–139. doi:10.1016/j.aca.2015.03.050
- 31.↑
Chan K, Wang W, Ledesma KR, Yin T, Tam VH. A robust LC–MS/MS method for amikacin: application to cellular uptake and pharmacokinetic studies. Bioanalysis. 2020;12(7):445–454. doi:10.4155/bio-2020-0007
- 32.↑
Aiken SS, Cooper JJ, Florance H, Robinson MT, Michell S. Local release of antibiotics for surgical site infection management using high-purity calcium sulfate: an in vitro elution study. Surg Infect. 2015;16(1):54–61. doi:10.1089/sur.2013.162
- 33.↑
Thomas DB, Brooks DE, Bice TG, DeJong ES, Lonergan KT, Wenke JC. Tobramycin-impregnated calcium sulfate prevents infection in contaminated wounds. Clin Orthop Relat Res. 2005;441:366–371. doi:10.1097/01.blo.0000181144.01306.b0
- 34.↑
Sykes JE, Greene CE. Infectious diseases of the dog and cat. 4 ed. Elsevier Health Sciences; 2013.
- 35.↑
Swieringa AJ, Tulp NJ. Toxic serum gentamicin levels after the use of gentamicin-loaded sponges in infected total hip arthroplasty. Acta Orthop. 2005;76(1):75–77. doi:10.1080/00016470510030355
- 36.↑
van Raaij TM, Visser LE, Vulto AG, Verhaar JA. Acute renal failure after local gentamicin treatment in an infected total knee arthroplasty. J Arthroplasty. 2002;17(7):948–950. doi:10.1054/arth.2002.34525
- 37.↑
Santschi E, Adams S, Murphey E. How to perform equine digital intravascular perfusion. Presented at Proceedings of the forty-fourth annual meeting of the American Association of Equine Practitioners; 1998; Baltimore, MD.
- 38.↑
Miclau T, Edin ML, Lester GE, Lindsey RW, Dahners LE. Bone toxicity of locally applied aminoglycosides. J Orthop Trauma. 1995;9(5):401–406. doi:10.1097/00005131-199505000-00007
- 39.↑
Balsamo LH, Whiddon DR, Simpson RB. Does antibiotic elution from PMMA beads deteriorate after 1-year shelf storage? Clin Orthop Relat Res. 2007;462:195–199. doi:10.1097/BLO.0b013e318074bb1d
- 40.↑
Shaw J, Gary J, Baker A, et al. Effects of sterilization techniques on bioactivity of polymethyl methacrylate antibiotic beads containing vancomycin and tobramycin. J Orthop Trauma. 2020;34(4):e109–e113. doi:10.1053/jvet.2003.50010
- 41.↑
National Center for Biotechnology Information. PubChem Compound Summary for CID 37768, Amikacin. (Accessed September 10, 2021). https://pubchem.ncbi.nlm.nih.gov/compound/Amikacin
- 42.↑
Levison ME, Levison JH. Pharmacokinetics and pharmacodynamics of antibacterial agents. Infect Dis Clin North Am. 2009;23(4):791–815. doi:10.1016/j.idc.2009.06.008
- 43.↑
Boothe D. Antimicrobial therapy: what diagnostics don’t tell you. Presented at Proceedings of the North American Veterinary Conference; January 2004; Orlando, FL.
- 44.↑
McPherson E, Dipane M, Sherif S. Dissolvable antibiotic beads in treatment of periprosthetic joint infection and revision arthroplasty-the use of synthetic pure calcium sulfate (Stimulan®) impregnated with vancomycin & tobramycin. Reconstr Rev. 2013;3(1). doi:10.15438/rr.v3i1.27
- 45.↑
Lum ZC, Pereira GC. Local bio-absorbable antibiotic delivery in calcium sulfate beads in hip and knee arthroplasty. J Orthop. 2018;15(2):676–678. doi:10.1016/j.jor.2018.05.001
- 46.↑
Medical Grade Calcium Sulfate Vs. Commercial Plaster of Paris. Kerrier Local Antibiotic Delivery. Updated February 26, 2022. (Accessed August 2, 2022). https://www.kerrier.com/comparisons-polymethyl-methacrylate-or-pmma/
- 47.↑
Lee GH, Khoury JG, Bell J-E, Buckwalter JA. Adverse reactions to osteoset bone graft substitute: the incidence in a consecutive series. Iowa Orthop J. 2002;22:35.
- 48.↑
Ipsen T, Jørgensen PS, Damholt V, Tørholm C. Gentamicin-collagen sponge for local applications: 10 cases of chronic osteomyelitis followed for 1 year. Acta Orthop Scand. 1991;62(6):592–594. doi:10.3109/17453679108994503
- 49.↑
Sørensen TS, Sørensen lI, Merser S. Rapid release of gentamicin from collagen sponge: in vitro comparison with plastic beads. Acta Orthop Scand. 1990;61(4):353–356. doi:10.3109/17453679008993535
