Rational dosage regimens for cephalothin and cefazolin using pharmacokinetics and pharmacodynamics analysis in healthy horses

Abstract Background First‐generation cephalosporins have good activity against gram‐positive bacteria and are extensively used in horses. There are few reports of pharmacokinetics and pharmacodynamics (PK/PD) analysis of cephalosporins in horses. Objective To optimise the dosages of the two first‐generation cephalosporins cephalothin (CET) and cefazolin (CEZ) in horses using PK/PD concepts. Study design Experimental study with single administration. Methods Drug plasma concentrations following a single intravenous (i.v.) administration of 22 mg/kg bodyweight (bwt) CET in 12 horses and of 10 mg/kg bwt CEZ in six horses were measured using LC‐MS/MS. Data were modelled using a nonlinear mixed effect modelling followed by Monte Carlo simulations. Minimum inhibitory concentrations (MICs) against Streptococcus zooepidemicus and Staphylococcus aureus isolated from horses were determined by the microbroth dilution method. Results The percentages of CET and CEZ binding to serum proteins were 19.9% ± 8.4% and 15.2% ± 8.5% respectively. For both CET and CEZ, the MIC90 against S. zooepidemicus was 0.12 mg/L and against S. aureus was 0.5 mg/L. For CET, to achieve a probability of target attainment (PTA) of 90% for a PK/PD target of a free serum plasma concentration exceeding the MIC90 for 40% of the dosing interval, an empirical CET dosage regimen of 22 mg/kg bwt q8h and 22 mg/kg bwt q4h i.v. administration were required for S. zooepidemicus and S. aureus respectively. For CEZ, the corresponding dosage regimens were 10 mg/kg bwt q12h and 10 mg/kg bwt q8h. Main limitations Small sample size only in healthy horses. Conclusions For CET, more frequent administration than that currently recommended (22 mg/kg bwt q6–12h) is required to empirically control S. aureus infection in horses. For CEZ, less frequent administration compared to the dosage regimen currently proposed (10–22 mg/kg bwt q6h) could control S. zooepidemicus and S. aureus infections in horses.


| INTRODUC TI ON
Streptococcus equi subsp. zooepidemicus is a gram-positive, β-haemolytic coccus belonging to the Lancefield group C. It is an opportunistic pathogen for adult horses, and it can cause shipping fever and pneumonia that could be fatal. 1,2 In addition, Staphylococcus spp. and Streptococcus spp. are frequently isolated from cellulitis in horses. 3 These gram-positive bacteria are considered as common pathogens for respiratory and limb infection in horses. β-Lactams, including penicillin and cephalosporins, are effective against gram-positive bacteria. Penicillin is commonly used in horses; however, it may not be effective against staphylococci owing to the development of resistant strains. 4,5 The first-generation cephalosporin, cefazolin (CEZ), has been used in horses 6,7 with a default recommended dosage regimen of 10-22 mg/kg bwt q6-8h in textbook and 25mg/kg bwt q6h in Clinical and Laboratory Standards Institute (CLSI) standards. 8,9 However, the recommended dose has a 2-fold range in textbooks, and the rational dosage for each organism has not been determined. Cephalothin (CET) is also a first-generation cephalosporin that has been previously reported in horses 10,11 and is listed in equine textbooks. 12,13 CET is no longer sold in certain countries but is still available in South America, Europe, Australia and Asia as a human drug and an off-label prescription drug for horses. For CET, a dose of 22 mg/kg bodyweight (bwt) q6-12h is routinely but empirically used in horses in Japan, 14 but this regimen has not been rationally determined using pharmacokinetics/pharmacodynamics (PK/PD) analysis. Because of increasing bacterial resistance against third-generation cephalosporins used in livestock and companion animals and of its consequent risk to humans, 15,16 the dosing schedules for these two cephalosporins which can be used as first-line antimicrobials in horses must be updated. In this report, PK/PD analysis was conducted based on the pharmacokinetics of CET and CEZ and on their minimum inhibitory concentrations (MICs) against bacteria isolated from horses to optimise their dosage regimen.

| MATERIAL S AND ME THODS
For the CET study, 12 healthy 3-5-year-old experimental Thoroughbred horses (six stallions and six mares) with bodyweights (bwts) of 410-530 kg were used. For the CEZ study, six healthy 3-5-year-old horses (four stallions and two mares) with bodyweights of 425-530 kg bwt that were also enrolled in the CET study were used. Horses were kept in individual stalls during the experiments and had ad libitum access to grass, hay and water.
For the six horses in both the CET and CEZ studies, a 2 × 2 crossover design was carried out with a 2 weeks washout period; the horses were randomly allocated to the two sequences. The doses of the cephalosporins, 22 mg/kg bwt CET and 10 mg/kg bwt CEZ, were determined based on previous reports. 6,11 CET (Coaxin injection 1 g) (Chemix Inc) was dissolved in 50 mL sterile physiological saline and CEZ (Cefazolin sodium injection 1 g) (Fujita Pharmaceutical Company) was dissolved in 30 mL sterile physiological saline for intravenous (i.v.) administration into the right jugular vein by a short bolus infusion (<30 seconds). The CET formulation was approved for humans, and that of CEZ for animals was approved in Japan.
Blood samples were collected at time 0 (prior to administration) and at 5, 10, 20, 30 and 45 minute and 1, 2, 3, 4, 6, 8 and 12 hour after administration. All blood samples were taken from the left jugular vein using a 14G catheter (Becton Dickinson Company), and 10 mL blood samples were collected in heparinised vacuum blood collection tubes (Terumo). The samples were immediately centrifuged at 1,500 g for 10 minute, and the separated plasma samples were stored at −20°C until analysis.

| Determination of plasma concentrations
Concentrations of CET, its active metabolite, deacetylcephalothin (DCET) and CEZ were measured. Quality control samples for calibration of the plasma analysis were prepared by adding standard CET (Toronto Research Chemicals Inc.), DCET (Toronto Research Chemicals Inc.) and CEZ (FUJIFILM Wako Pure Chemical Corporation) to blank horse plasma. To 20 μL of plasma was added 400 μL of acetonitrile and 20 μL of 1 μg/mL oxacillin sodium monohydrate (AdooQ BioScience) as an internal standard.
The sample was incubated for 5 min at room temperature and centrifuged at 10 000 g for 5 min. Fifty microlitres of supernatant was transferred to a new vial and diluted with 250 μL of water.
Five microlitres of the sample was injected into a liquid chromatography system (Nexera X2) (Shimadzu Corporation) connected to a mass spectrometer (QTRAP4500) (SCIEX Corporation). Highperformance liquid chromatography separation was performed on the column (ACQUITY UPLC BEH, 100 mm × 2.1 mm, 1.7 μm) (Waters Corporation) with a mixture of formic acid (0.1 vol%) and acetonitrile as the mobile phase. The final calibration curve had a coefficient of correlation (R 2 ) >0.995 over the concentration range of 0.1-300.0 µg/mL for CET and DCET and the range of 0.03-100.0 µg/mL for CEZ. The lower limit of quantitation (LOQ) was 0.1 µg/mL for CET and DCET and 0.03 µg/mL for CEZ. The recovery ratios in quality control samples were determined at currently proposed (10-22 mg/kg bwt q6h) could control S. zooepidemicus and S. aureus infections in horses.

K E Y W O R D S
horse, cefazolin, cephalothin, gram positive infection concentrations of 0.3, 5 and 240 µg/mL for CET and DCET and 0.09, 2 and 80 µg/mL for CEZ (five replicates each). Interday and intraday precision were assessed in quality control samples at concentrations of 0.1, 0.3, 5 and 240 µg/mL for CET and DCET and of 0.03, 0.09, 2 and 80 µg/mL for CEZ (five replicates each), and their coefficients of variation were <10% except for that of 0.03 µg/mL of CEZ, which was 14.8%. Accuracies for CET, DCET and CEZ were between 97.6% and 103.4%, 90% and 105.7%, and 86% and 103.2% respectively.

| Protein binding
The ultrafiltration method was used to separate free and bound drug for CET, DCET and CEZ; 200 µl samples were placed in a filter (Pierce™ Protein Concentrators PES, 10K MWCO) (Thermo Fisher Scientific) and centrifuged at 15 000 g for 5 minute at room temperature. Then the free drug concentration following ultrafiltration and the total drug concentration in samples not subjected to ultrafiltration were quantified using the same assay method as previously described. The plasma samples for assay were collected at 1, 2 and 3 hour after administration. The extent of protein binding and the free fraction were calculated by comparing the free and total drug concentrations. The average free fraction was used for the simulation of free plasma concentrations of CET and CEZ.

| Pharmacokinetic data analysis
Plasma pharmacokinetic analyses were conducted using a nonlinear mixed effect (NLME) model using commercially available software To report the interindividual variability as a coefficient of variation, Equation (2) was used for conversion of the variance terms (ω 2 ) into a coefficient of variation (CV%).
Shrinkage of the random effects (eta) toward the means was described as follows: where var(η r ) is the variance of the random effects. When the shrinkage for eta was >0.3, it was considered that the data were not able to robustly estimate this random component. It was impossible to estimate this between-subject variability for all structural parameters (non-identifiability) and a random component was added only for V, CL, and V2 for CET and V, CL and V3 for CEZ. The residual model was an additive plus a multiplicative (proportional) model of the form.
with ε1, the multiplicative error term having a mean of 0 and a variance noted σ1. and ε2, the additive error term having a mean of 0 and a variance noted σ2.
The additive sigma was reported as its standard deviation noted with the same units as plasma concentration (µg/mL) and the multiplicative sigma was reported as coefficient of variation. For the present fitting, the precision of the parameters was estimated using the bootstrap tool

| Minimum inhibitory concentrations
The MICs of CET, DCET, and CEZ were obtained using customised commercial panels (Eiken Chemical Co., Ltd.) against 98 strains of S. zooepidemicus, 51 strains of Staphylococcus aureus (without methicillin-resistant S. aureus; MRSA), 54 strains of Escherichia coli, and 26 strains of Klebsiella spp. isolated from infected horses according to the CLSI standards. 17 The horses studied were training Thoroughbred racing horses, located in two training facilities (Ritto and Miho training centres) and were sampled for infectious diseases including pneumonia and cellulitis. The MICs in the test were in the range of 0.03-4.0 mg/L. MRSA were excluded in this study because MICs of first-generation cephalosporins against these MRSA strains were extremely high, cephalosporins being considered ineffective. 18

| D ISCUSS I ON
The pharmacokinetic parameters for CET and CEZ in this were different from those previously reported. 6,7,10 These differences are correlated with those in the LOQ of analytical technics. The LOQ in previous studies was higher than that in our study. When decreasing the LOQ, a supplementary phase is often observed, leading to a decrease in plasma clearance and an increase in Vss, and a new terminal half-life that can be significantly longer than that previously reported, 19 as observed in the current experiment. In our study, the volumes of distribution for CET and CEZ were similar. However, CET had a higher clearance and a shorter half-life than CEZ, thereby indicating a faster metabolic elimination. These differences were also reported in humans. 20 CET is eliminated either directly via renal clearance or as DCET, a hepatic metabolite. Meanwhile, CEZ, which is not metabolised, is only eliminated via renal clearance. 20 The extent of protein binding was important to establish a rational dosage regimen because only the free drug concentration is microbiologically active. 21 For both CET and CEZ, our results indicated low protein binding, close to previously reported values in horses. 6,10 The protein binding of CET was reported as 75% and that of CEZ as 52%-85% in humans 22,23 and of CEZ as 36.2% in dogs. 24 These results indicate that there are wide differences in drug plasma protein binding among species, influencing the calculation of the dosage regimen in PK/PD analysis and prohibiting simple interspecific extrapolations as to the doses that should be administered to the horse.
The extent of protein binding of CET in humans was decreased at high total concentrations and was considered as saturable. 23 In the present experiment, there were no apparent differences in CET and CEZ protein binding for plasma collected at different times, and we To establish an empirical dosage regimen (ie without resorting to an AST), the dose should cover a priori at least 90% of the horse's population for the reported MIC 90 , as can be determined by MCS of a meta-population of 5000 horses from a population PK model. 28,30  and can be severe. 31,32 Using data generated in this study, we can therefore propose a lower CBP, and therefore a lower dosage, for CEZ for treatment of S. horses. An in vitro study indicated synergy or a partial synergistic effect for the combination of CET and DCET. 36  There is currently no CBP reported for CET by CLSI or other regulatory agencies, as CET as an active pharmaceutical ingredient is   This is allowed because the NLME modelling is an appropriate tool to merge unbalanced data obtained with analytical techniques having different performances, or in a variety of observational conditions that can be formally introduced in the modelling process.

CO N FLI C T O F I NTE R E S T S
No competing interests have been declared.

E TH I C A L A N I M A L R E S E A RCH
All experiments were approved by the Animal Care and Use Committee of Equine Research Institute, Japan Racing Association.

I N FO R M E D CO N S E NT
Not applicable.

DATA ACCE SS I B I LIT Y S TATE M E NT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/evj.13406.