Methods

Results

Table 1 and Table 2 depict MIC values of the LAs and antibiotics, respectively. Tables 3–5 depict FIC indices for the combination of LAs and antibiotics.

A total of 153 checkerboard assays were conducted with 64 unique combinations of LA-antibiotic plus bacterial isolates in each checkerboard.

Synergistic and antagonistic drug interactions, as indicated by FIC indices ≤ 0.5 or > 4, were established for 18% of checkerboards, representing a total of 70 combinations. Synergism (FIC index ≤ 0.5) was observed in 14 of 70 combinations and antagonism (FIC index > 4) in 56 of 70 combinations. No interactions (indifference) were observed for the majority of investigated antibiotic-LA combinations (82%) as indicated by FIC indices > 0.5 to ≤ 4.

For the 56 combinations displaying an antagonistic drug interaction, the majority (n = 52) were observed for gram-negative pathogens, especially E coli. Antagonism was mainly evident for amikacin or gentamicin in combination with either lidocaine, mepivacaine, or bupivacaine (Tables 3–5). The antagonistic drug interactions resulted in 2 to 3 2-fold increases of the antibiotics’ MIC value.

For the 14 combinations displaying a synergistic drug interaction, the majority were observed for ceftiofur combined with either lidocaine or mepivacaine (Tables 3–5). The synergistic drug interactions resulted in 3 to 5 2-fold decreases in the MIC value of the antibiotics. Two isolates (S aureus 31598 and 26714, both MRSA) with a ceftiofur MIC of 128 µg/mL prior to the addition of LAs decreased to a ceftiofur MIC of 4 to 16 µg/mL after the addition of lidocaine/mepivacaine.

Discussion

For the majority of investigated antibiotic-LA combinations (82%), there was no drug interaction against equine clinical isolates according to the determined FIC indices. This indicates that the antimicrobial activity of the antibiotic is not affected by the combination with an LA.

Nonetheless, in a subset of combinations, antagonism (n = 56) or synergism (n = 14) was observed. Whether or not these elucidated changes in antimicrobial activity translate into clinical relevance is uncertain. Other factors, like: MICs different from those detected in the present study, the presence of fibrin, pus or biofilm in the joint, and the tissue concentration obtained at the injection site, may also influence clinical efficacy of antibiotics.

With regard to the antibiotic tissue concentration of antibiotics at injection sites, multiple studies^5,6,17–20 have investigated antibiotic concentrations achieved in SF following IA administration. Common clinical IA amikacin doses to equine joints are 125 to 500 mg, which can result in maximal SF concentrations of 1,680 to 20,900 µg/mL.⁶ Since concentration-dependent drugs like aminoglycosides exert a maximum effect at 10 to 12 times the MIC,²¹ this high concentration in SF after IA administration theoretically results in pathogens with MICs as high as 140 to 1,742 µg/mL (1,680/12 to 20,900/12) being susceptible to amikacin (Table 6 depicts theoretical MIC limits in SF). In other words, even the lowest-used dose of amikacin is likely to be effective in killing bacteria with MICs ≤ 140 µg/mL, including all isolates belonging to the 7 genera of the present study and analyzed at our laboratory within the last 16 years (P Damborg, DVM, PhD, Vet Diagnostics, Department of Veterinary Sciences, University of Copenhagen, Denmark, unpublished data, 2007 through 2023).

Clinically reported amikacin doses used for RLPs vary from 500 to 2,500 mg,¹⁸ but most commonly 500 to 1,000 mg are used.²² In a study by Godfrey et al,¹⁷ the SF concentration after RLP with 1,000 mg diluted to a total volume of 60 mL for the distal limb was 459 µg/mL 30 minutes after injection while the tourniquet was still applied and 14 µg/mL after 24 hours. This means that in theory, bacteria with MICs as high as 38 µg/mL would be susceptible to amikacin, considering the limit of up to 12 times the MIC to reach a concentration-dependent effect (459/12 = 38).²¹ Within the past 16 years, 98% of equine isolates belonging to the included 7 genera and analyzed at our laboratory had MICs < 38 µg/mL (P Damborg, DVM, PhD, Vet Diagnostics, Department of Veterinary Sciences, University of Copenhagen, Denmark, unpublished data, 2007 through 2023) and were therefore likely susceptible to amikacin when administered as an RLP without LAs (Table 6 depicts theoretical MIC limits).

The observed antagonistic drug interaction for the combination of amikacin with LAs resulted in a 2-to-3 2-fold increase of the MIC of specific combinations. These antagonistic drug interactions are without clinical implications for the isolates included in this study as a 2-to-3 2-fold increase in the MIC would still be below the MIC limits (see Table 6 for theoretical MIC limits). The maximum amikacin MIC value of isolates of the present study was 8 µg/mL (Table 2), and other studies^23,24 have reported common equine pathogens to have amikacin MICs ≤ 16 µg/mL. Such Isolates with MICs of 16 µg/mL (as reported in other studies) would in the presence of an antagonistic drug interaction equivalent to the drug interaction displayed in the present study still be susceptible to the lowest IA amikacin dose used. Nonetheless, the opposite applies for RLPs as the increased MIC (a 2-to-3 2-fold increase of MIC of 64 to 128 µg/mL) is higher than the RLP MIC limit of SF. Therefore, caution is warranted when combining amikacin with LAs for RLPs for the treatment of bacteria with MICs > 8 µg/mL since potential antagonistic drug interactions result in MIC increases above concentrations achieved in SF. This underlines the importance of susceptibility testing prior to using amikacin-LA combinations for RLPs. Nonetheless, septic arthritis in horses needs prompt intervention for horses to survive. Therefore, aggressive treatment, including antibiotic administration, is commonly initiated based on clinical manifestations and SF analysis before culture results are available. In these cases, the addition of LAs to amikacin for RLPs should ideally be avoided, and pain should be managed alternatively (eg, through sedation and/or perineural analgesia).

The finding that the displayed antagonism after the combination of amikacin with LAs in our study (for isolates with an amikacin MIC of ≤ 8 µg/mL) is unlikely to have clinical implications is supported by an equine ex vivo study.⁹ Colbath et al⁹ found that the antimicrobial activity of amikacin was unaffected when SF, after RLP with 1,000 mg of amikacin combined with 500 mg of mepivacaine, was investigated against E coli and S aureus. The finding of an unaltered clinical antimicrobial activity for this RLP administration is likely caused by administration of amikacin in excess combined with low MICs of included isolates. This is corroborated by findings of the same study⁹ revealing a median amikacin concentration in SF of 104 µg/mL (IQR, 26.0 to 269.0) and an amikacin MIC of their 2 included reference strains of 0.5 to 4 µg/mL.^25,26

A common clinical dose of gentamicin to a medium-sized joint (carpal joints, metacarpophalangeal and metatarsophalangeal joints) is 150 mg,⁴ whereas a common dose for regional perfusion of the distal limb is 500 mg in dilution.^19,22 Immediately after the administration of 150 mg to the radiocarpal joint, a mean peak SF concentration of 1,828 µg/mL⁴ was measured, and the mean SF concentration of gentamicin in the metacarpophalangeal joint was 125.9 µg/mL 30 minutes after the administration of a 500-mg distal limb regional perfusion.¹⁹ The high concentration in SF after IA and RLP administration theoretically results in the killing of pathogens with MICs as high as 150 µg/mL (1,828/12 = 152) and 11 µg/mL (125.9/12 = 11) upon IA and RLP administration of gentamicin, respectively.

In our study, the observed antagonistic drug interactions for combinations of gentamicin with LAs resulted in a 2-to-3 2-fold MIC increase for specific isolates. The presented antagonistic drug interactions are unlikely to be clinically relevant for isolates included in the present study as a 2-to-3 2-fold gentamicin MIC increase would still render isolates either susceptible or resistant to gentamicin, equivalent to results prior to combining antibiotics with LAs, albeit many common clinical equine pathogens that are considered to be susceptible to gentamicin have gentamicin MICs higher than those included in our study.²⁷ In such cases, the presence of antagonistic drug interactions may have clinical implications. Isolates with gentamicin MICs of up to 16 µg/mL would, in the presence of equivalent drug antagonism as displayed in the present study, still be susceptible to common IA administration of gentamicin. On the other hand, for RLP, only isolates with gentamicin MICs < 2 µg/mL would still be susceptible to gentamicin as an increased MIC (3 2-fold increases of 2 = 16 µg/mL) exceeds theoretical SF MIC limits (Table 6).

A common IA dose of ceftiofur for medium-sized joints is 150 mg, and a previous study⁵ revealed that the maximum IA concentration of ceftiofur was 5,825.08 µg/mL 15 minutes after administration of this dose to the radiocarpal joint. At 24 hours after administration, the concentration was 4.94 µg/mL.⁵ This concentration is above the MIC of the majority (91%) of isolates belonging to the included 7 genera analyzed at our laboratory during a 16-year period (P Damborg, DVM, PhD, Vet Diagnostics, Department of Veterinary Sciences, University of Copenhagen, Denmark, unpublished data, 2007 through 2023). Because the concentration remained above this MIC for the entire dosing interval of 24 hours, IA administration of this time-dependent β-lactam alone should be successful in treating clinical joint infections from common equine ceftiofur-susceptible isolates at this dose.²¹

Only 1 isolate (E coli 26638) displayed an antagonistic drug interaction when ceftiofur was combined with lidocaine. The MIC of that specific isolate increased 2 2-fold dilutions. While this increase did not result in any clinically relevant findings of the present study as the increased MIC (= 2 µg/mL) remained below the MIC limit of ceftiofur (Table 6), potential clinical implications for other isolates cannot be ruled out. Interestingly, ceftiofur did, as the only antibiotic of the 3 included, reveal promising synergistic effects when combined with lidocaine and mepivacaine. The observed synergistic drug interactions for combinations of ceftiofur and lidocaine/mepivacaine resulted in a 3-to-5 2-fold reduction of MICs. These observed decreases of MICs are likely to be clinically relevant, particularly for the 2 MRSA strains. For these 2 isolates, the combination of ceftiofur and lidocaine and mepivacaine consistently reduced MICs with 3 to 5 2-fold dilutions to final concentrations either below MIC limits (4 µg/mL) or in the near clinical range (8 to 16 µg/mL). Although concentrations of 8 to 16 µg/mL are above the MIC limit, the IA ceftiofur combination with mepivacaine/lidocaine might also be effective against these MRSA as the theoretical MIC limit is based on achieved concentrations of ceftiofur in the SF of joints (4.94 µg/mL) 24 hours after IA injection. Based on the fact that ceftiofur only needs to exceed the MIC for approximately 50% of the 24-hour dosing interval to be effective and that the maximal measured concentration is 5,825.08 µg/mL⁵ after a standard IA ceftiofur dose, concentrations are likely to be higher than 16 µg/mL for 50% of the dosing interval. Nonetheless, elaborate pharmacokinetic studies are needed to firmly conclude this.

Concentrations of ceftiofur in SF after RLP with a full IV dose of 2 g ceftiofur revealed mean SF concentrations of 393 ± 103 µg/mL 30 minutes after injection and at 1.62 µg/mL after 24 hours.⁵ The same study revealed that ceftiofur remained above 2 µg/mL for 8 hours after RLP. While obtained concentrations in SF after RLP are higher than those achieved after IV administration, 8 hours is < 50% of the time during the dosing interval (once daily), which is less than that necessary for a time-dependent antibiotic to have clinical success.²¹ Consequently, independently of our results, ceftiofur is less likely to have clinical success in the treatment of septic arthritis when administered as RLP compared with an IA injection.

The main limitation of our study is inherent to its in vitro nature as the theoretical calculations provided need to be confirmed in vivo for final clinical conclusions. Further, due to economic and logistic reasons, we investigated a limited subset of equine isolates within 7 genera that most often cause septic arthritis in horses. Even though the inclusion of additional isolates would have increased the validity of our results, we ensured that the included isolates were representative of the MIC ranges observed in other studies. Some of the included isolates originated from infections other than joint infections. This was a necessity due to our limited strain database comprising only a few isolates from joints. Nevertheless, this is unlikely to have affected results negatively in view of the aforementioned representativeness of MICs.

In conclusion, our in vitro results suggest that commonly used antibiotics for the treatment of septic arthritis in horses can generally be combined with LAs for pain management during IA injections without compromising the clinical antimicrobial activity of the antibiotic. This information is pivotal for clinicians working in equine practice as our results show that these drugs can mostly be combined without compromising the antimicrobial activity of the antibiotic agent. For RLPs, caution is warranted when using aminoglycosides as drug antagonism was observed when combining LAs with gentamicin and amikacin in this in vitro study. Therefore, we suggest avoiding adding LAs to aminoglycosides for RLPs in the treatment of orthopedic infections caused by bacteria with unknown antimicrobial susceptibility.

The combination of ceftiofur with lidocaine and mepivacaine shows promising results by decreasing the MIC dramatically for the tested MRSA strains. The combination of ceftiofur with lidocaine or mepivacaine for IA injections could potentially be a treatment option for confirmed MRSA septic arthritis cases. However, the described synergistic interaction alongside thorough pharmacokinetics of IA ceftiofur should be further investigated before making a firm conclusion on its potential clinical efficacy.

Disclosures

The authors have nothing to disclose. No AI-assisted technologies were used in the composition of this manuscript.

Funding

The authors have nothing to disclose.

ORCID

D. M. T. Adler https://orcid.org/0000-0001-5042-0810