Pharmacokinetics of ceftiofur in the metacarpophalangeal joint after standing intravenous regional limb perfusion in horses

Abstract

This study investigated the pharmacokinetics of ceftiofur after intravenous regional limb perfusion (IVRLP). Six horses were involved in 3 IVRLP sessions. For each session, operators with varying clinical experience placed the tourniquet. A wide-rubber tourniquet was applied in the antebrachium as 2 g of ceftiofur in a total volume of 100 mL was injected into the cephalic vein. Plasma and metacarpophalangeal synovial fluid samples were obtained to evaluate perfusate leakage and synovial fluid concentrations of ceftiofur over 24 h. Overall, mean plasma concentrations were not significantly different before and after tourniquet removal. Mean synovial fluid ceftiofur concentrations were significantly higher 5 min and 8 h after tourniquet removal versus 24 h, after which values above the minimum inhibitory concentration (MIC) (1 μg/mL) were not detected. Concentrations above the MIC were detected in 72% and 50% of the horses at 5 min and 8 h, respectively. Overall, higher synovial fluid concentrations were obtained for the operator with the most recent clinical experience performing IVRLP.

Introduction

Orthopedic infections, either iatrogenic or secondary to a wound, are common in equids and may represent a therapeutic challenge, especially if a resistant bacterial strain is involved (1). In the last few decades, intravenous regional limb perfusion (IVRLP) has become a common procedure in equine medicine and provides a valuable and effective adjunctive treatment modality for distal limb infections (1,2). The technique is simple, cost-effective, well-tolerated by standing sedated horses, and often achieves a high concentration of the drug of choice at the area of interest with only a fraction of the recommended systemic dosage. This reduces the risk of systemic side effects or antimicrobial resistance (3).

The benefits associated with IVRLP have made this treatment modality the target of numerous studies to determine the most effective method. Research has focused on the injection site (4), the drug or combination of drugs used (5), dosage administered (6), total perfusate volume (7–9), characteristics of the tourniquet (type, width, pressure, location, number and duration of application) (10–15), the use of locoregional or general anesthesia (16), and the use of exsanguination techniques (11–17). Consequently, IVRLP effectiveness depends on all the aforementioned factors in addition to others, such as operator experience and horse variables (e.g., pathology, movement) for which knowledge is limited. Intravenous regional limb perfusion, therefore, has been described as inherently unreliable and difficult to standardize by some authors (1,7).

Most studies have focused on the use of aminoglycosides due to their concentration-dependent activity (4,8,10,12–14,16,17). However, some bacteria, such as Streptococcus equi subsp. zooepidemicus, have intrinsic resistance against aminoglycosides and aminoglycoside-IVRLP may be ineffective despite achieving in-vitro antimicrobial concentrations above the minimum inhibitory concentration (MIC) for common pathogens (6). A recent study reported a high prevalence (25%) of S. zooepidemicus in horses affected by joint sepsis in a North American hospital, which is higher than previously reported (18). Thus, the use of an antimicrobial such as ceftiofur to which S. zooepidemicus and other Gram-positive bacteria are susceptible would be of interest for IVRLP administration (18). A previous study demonstrated that concentrations of ceftiofur above the minimum inhibitory concentration (MIC) for common susceptible pathogens (1 μg/ mL) can be obtained in the radiocarpal joint for up to 24 h after IVRLP (19). Similarly, another study reported ceftiofur concentrations in the subcutaneous tissue and bone above the MIC for 24 h after IVRLP (20).

The objective of this study was to investigate the pharmacokinetics of ceftiofur concentration in the metacarpophalangeal joint and plasma following IVRLP. We hypothesized that ceftiofur concentration in the synovial fluid of the metacarpophalangeal joint remains above the MIC (1 μg/mL) for susceptible pathogens commonly isolated in septic arthritis for at least 24 h.

Materials and methods

Six healthy adult horses (2 geldings and 4 mares) of medium-sized mixed breeds with a mean age of 9.5 y (range: 4 to 14 y) and a mean weight of 513 kg (range: 440 to 586 kg) were included in the study. Upon enrollment, horses were not lame at the walk and had an unremarkable physical examination. The procedures were approved by the local Institutional Animal Care and Use Committee.

An IVRLP was performed 3 times on each horse with a 2-week washout period between IVRLPs. A randomly selected forelimb (left or right) received the IVRLP on the first session and forelimbs were alternated for the following sessions. Blood samples (5 mL) from the jugular vein were obtained 1 min before and after tourniquet release and synovial fluid samples (1 mL) were aseptically collected from the metacarpophalangeal joint at 3 time points (5 min, 8 h, and 24 h after tourniquet removal) using a palmaro-lateral approach through the collateral ligament of the lateral proximal sesamoid bone. Three blinded operators with varying clinical experience regarding IVRLP and tourniquet placement [a Board-certified surgeon in large animal who finished a residency 3 y earlier (Operator 1), a 3rd-year equine surgery resident with 2 previous internships (Operator 2), and an equine intern during the first months of his first internship (Operator 3)] participated in the study. A single operator was present at each study session. Operators were solely responsible for placing the tourniquet at each bi-weekly session and operator order was randomly selected at the beginning of the study. Venipuncture, perfusate administration, and sample collection were performed by an independent experienced investigator (AB for all procedures) to standardize the rest of the procedure. Before placing the tourniquet, each operator received an information session to follow detailed instructions regarding tourniquet placement.

After entering standing stocks, median, ulnar, and musculocutaneous perineural anesthesias were performed in the selected limb using 10 mL of 2% lidocaine hydrochloride (Lurocaine; Vétoquinol, Lavaltrie, Quebec) per site. Then, horses were sedated with detomidine hydrochloride (Domidine; Dechra Veterinary Products, Pointe-Claire, Quebec), 0.01 mg/kg body weight (BW), IV, and butorphanol tartrate (Dolorex; Merck Santé Animale, Kirkland, Quebec), 0.01 mg/kg BW, IV, administered in either jugular vein. Next, and 10 min after application of the local anesthetic, the blinded operator applied the tourniquet as instructed: a 10-cm wide elastic rubber tourniquet (Esmarch bandage) was placed 10 cm proximally to the accessory carpal bone. A 10-cm long gauze roll was placed medially and laterally over the antebrachium to minimize vascular leakage. The tourniquet was then placed as tightly as possible during 6 turns before securing the remainder of the Esmarch bandage in a quick-release fashion on the seventh pass around the limb (15). More specifically, the remainder of the Esmarch bandage was secured overlying the cephalic vein to add pressure at this level and minimize the risk of perfusate leakage. Last, an independent investigator (AB) placed a 21-gauge butterfly needle in the cephalic vein, distal to the tourniquet, prior to injection of the perfusate which took 1 to 2 min. The perfusate contained 2 g of ceftiofur (Excenel; Zoetis Canada, Kirkland, Quebec), 50 mg/mL, diluted in sterile saline to a final volume of 100 mL. The needle was removed at the end of the perfusion and a tight bandage was placed over the injection site to prevent leakage and hematoma formation. The tourniquet was released 20 min after perfusion. Xylazine hydrochloride (Rompun; Bayer, Mississauga, Ontario), 0.2 to 0.4 mg/kg BW, IV was administered during the procedure to increase the sedation plane and prevent movement associated with tourniquet discomfort, if deemed required by the independent investigator (AB).

For each IVRLP, horse movement was recorded and scored by an independent investigator (AB) as described by Plunkett et al (15). Movement scores were analyzed separately (first 10 min and second 10 min of tourniquet application) and globally (20-minute period). Two types of movements were evaluated: Type 1, or movements during which the hoof maintained contact with the ground, and Type 2, or movements during which the hoof was lifted off the ground. Cumulative Types 1 and 2 movements were calculated globally and during the two 10-minute periods and a scoring was assigned according to the following criteria: Score 0 — no movement observed; Score 1 — 1–5 Type-1 movements and no Type-2 movements; Score 2 — 6–10 Type-1 movements or 1–5 Type-2 movements; and Score 3 — > 10 Type-1 movements or > 5 Type-2 movements. After each session, horses were confined to a stall and a physical examination was performed twice daily to detect any signs of inflammation at the arthrocentesis or venipuncture sites, joint effusion, or lameness for at least 72 h after IVRLP.

Concentrations of ceftiofur and desfuroylceftiofur-related metabolites, in the form of the active metabolite desfuroylceftiofur acetamide (DCA), were measured for each plasma and synovial fluid sample by UV chromatography detection in the research pharmacokinetic platform of the Centre Hospitalier Université de Montréal. Ceftiofur and DCA-related metabolites were extracted from plasma (200 μL) and synovial fluid (100 μL) following a reduction step. The sample was vortexed for 5 s and incubated at 50°C for 30 min. The reduced sample was loaded and centrifuged. Then, 1 mL of the solution was added to the cartridge and allowed to flow by gravity on the column. Afterwards, another washing step was done and the DCA was removed from the cartridge. The eluate was evaporated, re-suspended, and transferred to an injection vial for analysis. A gradient mobile phase was used with a Thermo Scientific Hypersil Gold aQ analytical column (Thermo Scientific Hypersil Gold aQ analytical column; ThermoFisher Scientific, Waltham, Massachusetts, USA) operating at 40°C. The flow rate was fixed at 200 μL/min and both compounds eluted at 4.6 min. The mass spectrometer (MS) was interfaced with the ultrahigh performance liquid chromatography system using a pneumatic assisted heated electrospray ion source. Mass spectrometer detection was performed in a positive ion mode, using selected reaction monitoring (MSR). Standard solutions of DCA and DCA-D₃ were infused into the MS to optimize the MS/MSR parameters. For analytical qualification, ceftiofur standard stock solution was prepared in dimethyl sulfoxide (DMSO). The internal standard stock solution was prepared at 1 mg/mL of ceftiofur-D₃ in DMSO. Calibration standards were prepared by fortifying blank horse plasma or blank horse synovial fluid with the standard working solutions at 2% to enable concentrations spanning an analytical range 0.020 to 9.872 μg/mL for plasma and 0.005 to 9.872 μg/mL for synovial fluid. The method was linear using a linear regression weighted 1/× analysis in both matrices.

A mixed linear model with the operator as a fixed variable and the horse as a random variable was used to measure the operator effect on total movement, combined with a Tukey post-hoc analysis. A paired Student’s t-test statistical analysis was used for analysis of movement (first 10-minute versus second 10-minute period). A mixed linear model was used to analyze plasma values, with the operator and time as fixed variables, total movement score as cofactor, and horse as random variable. To analyze synovial fluid values, the concentration values were transformed with the logarithm of base 10 to normalize the distributions. A mixed linear model was used with operator and time as fixed variables, total motion score as cofactor and horse as random variable. Comparison of mean plasma and synovial fluid concentrations obtained by the different operators at each time and the mean concentrations between 2 times for each operator was performed. Geometric means for ceftiofur synovial concentrations and 95% confidence intervals (CI) were calculated with an inverse logarithm transformation. A Benjamini-Hochberg sequential method was performed to adjust the alpha or significance level for each of the multiple comparisons to control the false discovery rate. Thus, the significance level varied among comparisons according to the number of comparisons performed.

Lastly, a test Cochran-Mantel Haenszel for repeated measurements was performed to determine any significant difference between the plasma values obtained and the synovial fluid values. Data were analyzed using a commercially available statistical software (SAS, version 9.3; Cary, North Carolina, USA). The level of statistical significance was set at P < 0.05.

Results

All procedures were tolerated well by the horses and no complications were noted during IVRLP or after the procedure. Venipuncture was performed at first attempt without complications in all horses. Minimal peri-cephalic vein swelling was occasionally seen and resolved within 24 to 48 h.

Xylazine administration was used in 3 horses by Operators 1 and 2 to increase the sedative plane. Horses seemed subjectively more sedated after the initial sedation (detomidine and butorphanol) and occasionally knuckled and moved for Operator 3; only 1 horse in this group required additional sedation.

Total mean movement score (0 to 20 min) varied significantly among operators (P = 0.014). Total mean movement score was significantly higher for Operator 3 (P < 0.001) compared to Operators 1 or 2 (P = 0.04 and P = 0.02, respectively) (Table 1). The combined movement score (sum of all operators) was significantly higher for the second 10 min of tourniquet placement (P = 0.03) when compared to the first 10 min (Table 1). Despite that, movement did not have a significant effect when included in the linear mixed model to analyze plasma and synovial fluid results (P = 0.68 and 0.23, respectively).

Table 1.

Mean movement score [standard deviation], {median} and (range) during the first and the second 10-minute intervals during cephalic intravenous regional limb perfusion with ceftiofur in 6 horses for each operator and for all operators combined.

	Operator 1		Operator 2		Operator 3
	Surgeon		Resident		Intern		All operators
Interval (min)	0 to 10	10 to 20	0 to 10	10 to 20	0 to 10	10 to 20	0 to 10	10 to 20
Movement score	0 [0.0]	1.2 [1.3]	0.3 [0.8]	1.2 [0.8]	1.7 [0.5]	1.5 [0.5]	0.7 [0.9]	1.3 [0.9]*
	{0}	{1}	{0}	{1}	{2}	{1,5}	{0}	{1}
	(0 to 0)	(0 to 3)	(0 to 2)	(0 to 2)	(1 to 2)	(1 to 2)	(0 to 2)	(0 to 3)

^*Movement score was higher (P = 0.03) at 10 to 20 min versus 0 to 10 min.

For plasma values, mean ceftiofur concentrations were not significantly different between times and operators (before or after removing the tourniquet for each operator) (P > 0.3 after adjustments) or between operators (difference in concentration before and after removing the tourniquet for each operator) (P ≥ 0.013 after adjustments) (Table 2).

Table 2.

Mean plasma concentrations for ceftiofur (results in μg/mL) [standard error], {median} and (95% confidence interval) obtained from the jugular vein for each operator 1 min before and 1 min after tourniquet release in cephalic intravenous regional limb perfusion with ceftiofur in 6 horses.

	Operator 1	Operator 2	Operator 3
	Surgeon	Resident	Intern	All operators
1 min before tourniquet release	12.80 [1.43]	13.20 [1.38]	13.46 [1.52]	13.15 [0.85]
	{14.15}	{13.32}	{13.12}	{13.41}
	(9.57 to 16.03)	(10.07 to 16.32)	(10.02 to 16.90)	(11.21 to 15.11)
1 min after tourniquet release	16.89 [1.43]	18.65 [1.38]	16.62 [1.66]	17.52 [0.85]
	{17.96}	{18.90}	{18.28}	{18.20}
	(13.66 to 20.12)	(15.52 to 21.77)	(12.88 to 20.38)	(15.57 to 19.47)

For synovial fluid samples, there were no significant differences in the mean ceftiofur concentration between operators (at 5 min, 8 h, or 24 h after tourniquet release) (P > 0.072 after adjustments). However, mean ceftiofur values were higher at 5 min (P < 0.0008 after adjustments) and 8 h (P < 0.0026 after adjustments) than 24 h for each operator. However, there were no significant differences between 5 min and 8 h (P > 0.014 after adjustments) (Table 3).

Table 3.

Geometric mean (inverse logarithmic transformation), {median} and lower and upper bounds for 95% confidence interval of synovial fluid concentrations for ceftiofur (results in μg/mL) in the metacarpophalangeal joint for each operator at 5 min, 8 h, and 24 h after tourniquet release in cephalic intravenous regional limb perfusion with ceftiofur in 6 horses.

	Operator 1	Operator 2	Operator 3
Time	Surgeon	Resident	Intern	All operators
+5 min	1.25*	4.65*	5.65*	3.22
	{0.84}	{12.53}	{6.78}	{2.02}
	(0.42 to 3.65)	(1.65 to 13.10)	(1.83 to 17.47)	(1.81 to 5.75)
+8 h	0.98#	1.12#	1.35#	1.14
	{1.00}	{1.34}	{1.01}	{0.96}
	(0.34 to 2.83)	(0.40 to 3.16)	(0.44 to 4.18)	(0.64 to 2.026)
+24 h	0.15	0.13	0.22	0.16
	{0.15}	{0.14}	{0.17}	{0.16}
	(0.05 to 0.43)	(0.05 to 0.37)	(0.07 to 0.67)	(0.090 to 0.29)

^*P < 0.00008 and

^#P < 0.0026 after adjustments. Significant values when 5-minute and 8-hour values where independently compared to 24 h values.

Ceftiofur concentration in the metacarpophalangeal joint was above the MIC in 72% (13/18) of the horses 5 min after tourniquet removal (geometric mean synovial fluid concentration: 3.22 μg/mL; median: 2.02 μg/mL) and remained elevated above MIC at 8 h in 50% (9/18) (geometric mean synovial fluid concentration: 1.14 μg/mL; median: 0.96 μg/mL), but did not remain above the MIC at 24 h (geometric mean synovial fluid concentration: 0.16 ± 0.11 μg/mL; median: 0.16 μg/ mL) (Table 3).

The prevalence of synovial fluid samples with a concentration above the MIC was higher (P < 0.02) when the plasma ceftiofur concentration (before and after tourniquet release) was > 3 μg/ mL. A ceftiofur concentration 10 times above the MIC (10 μg/mL) was observed in only 39% (7/18) of the cases 5 min after tourniquet release (Table 4).

Table 4.

Number (and percentage) of horses with ceftiofur synovial fluid concentrations > 1 μg/mL (minimum inhibitory concentration for common pathogens) and > 10 μg/mL in the metacarpophalangeal joint at 5 min, 8 h, and 24 h after tourniquet release in cephalic intravenous regional limb perfusion with ceftiofur in 6 horses.

		Operator 1	Operator 2	Operator 3
	Time	Surgeon	Resident	Intern	All operators
Number with concentrations > 1 μg/mL	+5 min	3 (50%)	5 (83%)	5 (83%)	13 (72%)
	+8 h	2 (33%)	4 (67%)	3 (50%)	9 (50%)
	+24 h	0	0	0	0
Number with concentrations > 10 μg/mL	+5 min	1 (17%)	4 (67%)	3 (50%)	7 (39%)
	+8 h	0	0	0	0
	+24 h	0	0	0	0

Discussion

Recent literature highlights that some bacteria have intrinsic resistance against aminoglycosides and, therefore, aminoglycoside-IVRLP cannot be used to satisfactorily manage such bacterial infections (6). Finding antimicrobial alternatives for IVRLP is essential in circumstances in which bacterial culture and sensitivity reveal aminoglycoside resistance. In a recent study at our institution, 25% of isolates from septic synovitis were S. zooepidemicus, which are intrinsically resistant to aminoglycosides but commonly susceptible to cephalosporines (18). Intravenous regional limb perfusion with ceftiofur has been rarely studied experimentally in horses (19,20); however, a good response to IVRLP with time-dependent antimicrobials has been reported in clinical cases (21).

Ceftiofur is a time-dependent broad-spectrum 3rd-generation cephalosporin that is tolerated well in horses at the high dose used in this study (~4 mg/kg BW) and has good activity against a large variety of bacterial pathogens (19,22). Time-dependent antimicrobial agents are not commonly used for regional limb perfusion because traditionally, their anti-bacterial activity is related to the interval that they remain above the MIC, which may hamper clinical applicability (1). In contrast, β-lactams exceeding 8 to 9 times the MIC may result in an increased efficacy of bacterial killing, which could be obtained after IVRLP (20). In the current study, median ceftiofur values did not routinely achieve 8 to 9 times the MIC except at the 5-min timepoint with Operator 2 (12.53 μg/mL), suggesting that recent contemporaneous experience in the application of the tourniquet may influence the synovial fluid concentrations achieved during cephalic IVRLP.

Measured geometric mean ceftiofur values in the synovial fluid of the metacarpophalangeal joint were significantly higher and above the MIC at 5 min (3.22 μg/mL) and 8 h (1.14 μg/ mL) compared to 24 h (0.16 μg/mL) for all operators. However, ceftiofur synovial concentrations were above the MIC (1 μg/mL) in only 72% and 50% of the horses 5 min and 8 h after tourniquet removal, respectively. This variability in drug concentration among horses was consistent with that in similar IVRLP studies (10,15,16,19). In contrast, values at 24 h were consistently below the MIC for all horses; therefore, daily ceftiofur IVRLP may not achieve therapeutic results in the metacarpophalangeal joint with the protocol used. The pharmacokinetics of ceftiofur after IVRLP in another study demonstrated a rapid decrease towards the MIC (1 μg/mL) in the first 8 to 12 h, whereas the curve flattened thereafter until 24 h (19). Extrapolating from that study, results of the current study suggest that ceftiofur IVRLP should be repeated at least every 8 h to always be above the MIC for the metacarpophalangeal joint. Nevertheless, it could be repeated every 14 to 16 h if there is a requirement to be above the MIC for only 40 to 50% of the dosing interval to be therapeutically effective as suggested in a study in humans (19).

Our ceftiofur synovial concentrations were significantly lower than those reported in the radiocarpal joint following IVRLP in another study (19). However, different volumes of perfusate were used (100 versus 60 mL) and different synovial structures were targeted. Furthermore, there were differences in types of tourniquets (Esmarch bandage versus pneumatic) and application times (20 versus 30 min). Lastly, horses were not anesthetized during IVRLP in the current study (19). In general, higher synovial antimicrobial concentrations are obtained in studies using pneumatic tourniquets above the carpus, radiocarpal joint samples or performed under general anesthesia. In contrast, studies that use an Esmarch bandage above the carpus, sample the metacarpophalangeal joint or are performed in standing horses tend to result in lower synovial antimicrobial concentrations (10,14,16). However, 1 study reported more than double the antimicrobial synovial concentration in the metacarpophalangeal joint after cephalic IVRLP when 100 versus 60 mL of perfusate was used (9). Thus, there are multiple factors that may have affected the results of the current study and make comparisons with other studies difficult.

In the current study, there were no significant differences in the mean ceftiofur plasma concentration values between times for each operator or between operators. Besides achievement of high synovial fluid concentrations following IVRLP, a significant difference in plasma values before and after tourniquet release could also be considered as a surrogate for effective tourniquet placement. In fact, all horses with an increase in plasma concentrations > 3 μg/mL after tourniquet release reached a synovial fluid concentration above the MIC for common pathogens involved in synovial sepsis. Perhaps applying a tourniquet in a way that minimizes systemic leakage would yield higher synovial concentrations. Operator 2 was the operator with most recent hands-on experience placing tourniquets and a trend was seen for this operator to obtain higher values of plasma ceftiofur concentrations after tourniquet release (18.65 ± 1.38 versus 13.20 ± 1.38 μg/mL) (P = 0.013 after adjustments; adjusted cut-off value indicating significance was P < 0.006). Additionally, Operator 2 obtained synovial fluid ceftiofur values above the MIC in a higher proportion of horses (Table 4). However, these results should be interpreted with caution, as only 1 operator was used to represent each category of experience, and experience in tourniquet placement is difficult to be objectively assessed. Information on the impact of the operator responsible for placing the tourniquet during IVRLP is lacking and a study with a more representative number of individuals in each category group is warranted.

Inadequate tourniquet placement has been suggested as a cause for failure to achieve high antimicrobial concentration in target tissues (10). The use of wide rubber tourniquets (Esmarch bandage) is clinically widespread but has been described as inherently unreliable and requires close attention during placement (1,7). Its efficacy depends on the technique used during application, the force applied during placement, and patient variables (15,23). As aforementioned, the high variability of ceftiofur concentrations detected in plasma and synovial fluid samples is comparable to the results of previous studies and has been attributed to type of tourniquet used, differences in drug metabolism, differences in synovial vascularization of an individual joint, atypical vascularization patterns, blood contamination of synovial fluid samples, and movement (10,15,16,19). In addition, another explanation for our inconsistent results among operators could be the use of an inadequate tourniquet application protocol. In a recent study using a protocol similar to ours, mean subtourniquet pressures of 163 mm Hg above the systolic blood pressure (15), which is well above the suggested target pressure to be obtained during IVRLP (1). Lastly, it has been suggested that an excessively large volume may result in an intravascular pressure that exceeds the effective sealing capacity of the tourniquet and therefore, leakage of the perfusate under-neath the tourniquet may occur (24). A suboptimally placed tourniquet combined with the large volume of perfusate used in the current study could have acted synergistically and led to leakage of perfusate in several instances. Large perfusate volumes (100 mL) may not be indicated if proper tourniquet placement cannot be ensured, or operator experience is limited.

Sudden weight changes, lifting the limb off the ground or pawing can double intravascular pressure distally to the tourniquet and produce inadvertent leakage of the perfusate under the tourniquet (10,12,13,25). We attempted to minimize horse movement with a standardized sedation protocol and administration of perineural anesthesia to the treated limb before IVRLP. Perineural anesthesia is the most effective way to reduce discomfort from tourniquet application to standing sedated horses without affecting regional pharmacokinetic parameters of antimicrobials (16). In the current study, movement varied significantly between operators (P = 0.014), with the last operator (Operator 3) having higher scores. Greater amounts of movement with this operator could be hypothesized as less cooperation of horses in the last phase of the study. However, horses seemed more relaxed at this point of the study and movement was subjectively associated to wobbliness and limb knuckling due to a higher susceptibility of horses to the same initial dose of sedatives rather than movement associated with discomfort. The latter occurred in Groups 1 and 2 and horses required additional administration of sedatives. In addition, there was no significant effect of movement when synovial fluid or plasma concentrations were analyzed. Movement scores were significantly higher during the second 10-minute period, suggesting that horses tend to move more despite local perineural anesthesia when the sedation started wearing off 10 to 15 min after administration. The discomfort that some horses experience, the need for additional sedation, and that peak antimicrobial concentrations are often reached 10 to 20 min after IVRLP justify the use of shorter tourniquet application times in standing procedures (13,14).

The present study was limited to 6 horses and an a posteriori power calculation revealed that the study was underpowered (≥ 10 horses needed) to detect significant results for some variables. Furthermore, information about ceftiofur synovial concentration values between 8 and 24 h was not available, making it impossible to accurately calculate the pharmacokinetics of ceftiofur during that period. The design of the current study did not allow individual variations on tourniquet placement experience to be objectively evaluated. Additionally, only 1 individual representing each category group was included and therefore, results are not representative for all individuals within those categories. Also, movement may have affected the synovial fluid concentrations and we were unable to statistically detect it. As in other studies, sedation and perineural anesthesia with lidocaine did not abolish movement completely (16). It is possible that the interval between anesthetic administration and tourniquet application was insufficient; however, loss of skin sensitivity and absence of nociceptive responses were not tested. Additionally, a recent study highlighted that mepivacaine is superior to lidocaine in abolishing lameness after inducing hoof lameness (26). Lastly, we used horses free of synovial sepsis and our findings may not reflect the conditions present in the face of infection. One study using IVRLP with amikacin demonstrated earlier and higher maximal concentration of antimicrobial in joints with synovitis compared to normal joints, whereas septic environments may decrease both antimicrobial delivery and efficacy (12). In contrast, there are some indications that the typically protein-bound metabolite DCA passively accumulates at sites of infection due to the influx of inflammatory proteins (19,20).

In conclusion, ceftiofur should be used in IVRLP based on current antimicrobial use guidelines and, therefore, should be reserved for synovial samples positive for S. zooepidemicus or for bacteria that are susceptible based on sensitivity testing. Results of this study suggest that IVRLP with ceftiofur should be repeated at least every 8 h. This timing will ensure values above the MIC for susceptible pathogens involved in synovial sepsis of the metacarpophalangeal joint until further studies are conducted to determine the interval required to maintain concentrations above the MIC for ceftiofur-IVRLP to be clinically effective. Additionally, operator experience regarding tourniquet placement may play an important role in IVRLP efficacy and further investigations are warranted to determine the importance of this factor.