pharmacokinetic pharmacodynamic integration and modelling ...€¦ · 1997). the latter group...

Pharmacokinetic–pharmacodynamic integration and modelling of

florfenicol in calves

P. SIDHU1

A. RASSOULI2

J. ILLAMBAS3

T. POTTER4

L. PELLIGAND

A. RYCROFT &

P. LEES

The Royal Veterinary College, Hawkshead

Campus, Hatfield, Hertfordshire, UK;1 Present address:Animal Disease Research

Centre, College of Veterinary Science, Guru

Angad Dev Veterinary and Animal Sciences

University, Ludhiana, Punjab, 141004,

India; 2 Present address:Department of

Pharmacology, Faculty of Veterinary Medi-

cine, University of Tehran, PO Box 14155-

6453, Tehran, Iran; 3 Present address:Pfiz-

er Animal Health, VMRD, Pfizer European

Service Center, Building 3, HogeWei 10,

B-1930, Zaventem, Belgium; 4 Present

address:Westpoint Veterinary Group, Dawes

Farm, Bognor Road, Warnham, West

Sussex, RH12 3SH, UK

Sidhu, P., Rassouli, A., Illambas, J., Potter, T., Pelligand, L., Rycroft, A., Lees P.

Pharmacokinetic–pharmacodynamic integration and modelling of florfenicol in

calves. J. vet. Pharmacol. Therap. 37, 231–242.

Florfenicol was administered subcutaneously to 10 calves at a dose of

40 mg/kg. Pharmacokinetic–pharmacodynamic (PK-PD) integration and

modelling of the data were undertaken using a tissue cage model, which

allowed comparison of microbial growth inhibition profiles in three fluids,

serum, exudate and transudate. Terminal half-lives were relatively long, so

that florfenicol concentrations were well maintained in all three fluids. Mini-

mum inhibitory concentration (MIC) and minimum bactericidal concentra-

tion were determined in vitro for six strains each of the calf pneumonia

pathogens, Mannhemia haemolytica and Pasteurella multocida. An PK-PD inte-

gration for three serum indices provided mean values for P. multocida and

M. haemolytica, respectively, of 12.6 and 10.4 for Cmax/MIC, 183 and 152 h

for AUC0–24 h/MIC and 78 and 76 h for T>MIC. Average florfenicol concen-

trations in serum exceeded 4 9 MIC and 1.5 9 MIC for the periods 0–24and 48–72 h, respectively. Ex vivo growth inhibition curves for M. haemolytica

and P. multocida demonstrated a rapid (with 8 h of exposure) and marked (6

log10 reduction in bacterial count or greater) killing response, suggesting a

concentration-dependent killing action. During 24-h incubation periods, inhi-

bition of growth to a bacteriostatic level or greater was maintained in serum

samples collected up to 96 h and in transudate and exudate samples har-

vested up to 120 h. Based on the sigmoidal Emax relationship, PK-PD model-

ling of the ex vivo time–kill data provided AUC0–24 h/MIC serum values for

three levels of growth inhibition, bacteriostatic, bactericidal and 4 log10decrease in bacterial count; mean values were, respectively, 8.2, 26.6 and

39.0 h for M. haemolytica and 7.6, 18.1 and 25.0 h for P. multocida. Similar

values were obtained for transudate and exudate. Based on pharmacokinetic

and PK-PD modelled data obtained in this study and scientific literature values

for MIC distributions, Monte Carlo simulations over 100 000 trials were

undertaken to predict once daily dosages of florfenicol required to provide

50% and 90% target attainment rates for three levels of growth inhibition,

namely, bacteriostasis, bactericidal action and 4 log10 reduction in bacterial

count.

(Paper received 31 May 2013; accepted for publication 22 October 2013)

Pritam Sidhu, Animal Disease Research Centre, College of Veterinary Science, Guru

Angad Dev Veterinary and Animal Sciences University, Ludhiana, Punjab141004,

India. E-mail: [email protected]

INTRODUCTION

In most countries, florfenicol has replaced for veterinary use

the first drug of this class, chloramphenicol, which was intro-

duced in 1947 as a broad spectrum antibiotic and classified as

a bacteriostat. Florfenicol is now the major and, in many coun-

tries, the only remaining drug of the fenicol class in veterinary

use. It is a fluorinated analogue of thiamphenicol with a simi-

lar mechanism of action to chloramphenicol and thiampheni-

col, involving inhibition of the 30S ribosomal subunit of

bacteria, leading to inhibition of protein synthesis (Cannon

et al., 1990).

© 2013 John Wiley & Sons Ltd 231

J. vet. Pharmacol. Therap. 37, 231--242. doi: 10.1111/jvp.12093.

Clinically, the main use for florfenicol is treatment for respi-

ratory tract infections in calves and young pigs by parenteral

administration (Hoar et al., 1998; Varma et al.,1998; Jim

et al.,1999; Aslan et al., 2002; De Haas et al., 2002; Catry

et al., 2008). In calves, the spectrum of activity includes the

pneumonia pathogens, Mannheimia haemolytica, Pasteurella

multocida and Histophilus somnus (Varma et al.,1998; De Haas

et al., 2002; Catry et al., 2008). The latter authors reported

no resistance to these species using Kirby–Bauer discs; of the

isolates tested, all were sensitive to florfenicol, with the excep-

tion of two of 39 P. multocida isolates, which were intermedi-

ate in sensitivity. Similarly, Priebe and Schwarz (2003)

reported high activity of florfenicol against these bovine respi-

ratory tract pathogens. Reports of minimum inhibitory con-

centration (MIC) for P. multocida and M. haemolytica

(MIC ≤ 2.0 lg/mL) indicate the relatively high potency of flor-

fenicol for these organisms (Priebe & Schwarz, 2003; Kehren-

berg et al., 2004; Shin et al., 2005; Dowling, 2006; Thiry

et al., 2011). However, Ayling et al. (2000) reported high

MIC50 and MIC90 values of 4 and 16 lg/mL, respectively,

against Mycoplasma bovis.

It is of interest that, whilst florfenicol has been in widespread

use in farm animals for 20 years and whilst resistance occur-

rence and mechanisms have been described (Kehrenberg et al.,

2004; Schwarz et al., 2004; Berge et al., 2005; Kehrenberg &

Schwarz, 2005), it has retained a high level of activity against

respiratory tract pathogens of calves (Catry et al., 2005; Shin

et al., 2005). Moreover, De Haas et al. (2002) demonstrated a

clear bactericidal action against bovine and porcine respiratory

tract pathogens, and they classified florfenicol as concentration

dependent in its killing action.

The pharmacokinetic profile of florfenicol has been described

in chickens (Shen et al., 2003; Anadon et al., 2008), turkeys

(Switala et al., 2007), pigs (Li et al., 2002; Liu et al., 2003),

dogs (Park et al., 2008), rabbits (Koc et al., 2009) and sheep

(Lane et al., 2007). Florfenicol pharmacokinetics has also been

reported in adult cattle (Bretzlaff et al., 1987; Varma et al.,

1994; Soback et al., 1995) and calves (Varma et al., 1986;

Adams et al., 1987; Lobell et al., 1994; De Craene et al.,

1997). The latter group reported an elimination half-life of

3.2 h in calves after intravenous dosing, similar to values of

2.7–3.2 h in adult cows (Bretzlaff et al., 1987; Lobell et al.,

1994; Soback et al., 1995). However, the rate of decline in

plasma concentration was much slower in calves administered

florfenicol intramuscularly or subcutaneously (Lobell et al.,

1994; Varma et al., 1998), as a consequence of flip-flop phar-

macokinetics. Similarly, Lacroix et al. (2011) reported a termi-

nal half-life of florfenicol of 39.6 h after subcutaneous dosing

of a combination product containing florfenicol and the nonste-

roidal anti-inflammatory drug (NSAID), flunixin, at a dosage of

the former of 40 mg/kg.

There are no major reports linking the pharmacokinetics

and pharmacodynamics of florfenicol in calves within a single

investigation. Therefore, the aims of this study were to: (i)

establish the serum concentration–time profile and to derive

pharmacokinetic data for florfenicol in calves after subcutane-

ous administration at the dose of 40 mg/kg, (ii) determine the

rate and extent of florfenicol penetration into and elimination

from carrageenan-inflamed (exudate) and noninflamed (transu-

date) fluids in a tissue cage model, (iii) investigate ex vivo time

–kill curves for florfenicol in samples of serum, exudate and

transudate collected from florfenicol-treated calves against clin-

ical isolates of each of two calf pathogens, M. haemolytica and

P. multocida, (iv) integrate and model the data as a basis for

determination of the magnitude and duration of action of flor-

fenicol at the 40 mg/kg dose and (v) use the data generated as

a basis for calculating daily dosage rates for three levels of

growth inhibition.

MATERIALS AND METHODS

Animals, surgical procedures and tissue cage model

The study was carried out in 10 healthy female Aberdeen

Angus calves weighing 145–204 kg (mean = 179 kg,

SD = 16.7) and aged 79–131 days (mean = 108 days,

SD = 15 days). Silicon tissue cages were implanted subcutane-

ously in the flank under general anaesthesia, as previously

described (Sidhu et al., 2003). A period of 35 days after sur-

gery was allowed for wound healing and to permit the growth

of granulation tissue into the cages, prior to dosing.

At zero time (immediately preceding the injection of florfeni-

col), 0.5 mL of 1% w/v sterile lambda carrageenan solution in

saline (Viscarin; FMC biopolymers, Philadelphia, PA, USA) was

injected into a single tissue cage. This was subsequently used

to harvest exudate. An unstimulated cage was used to collect

noninflammatory fluid (transudate).

Drug administration

Florfenicol was administered as Nuflor� (Schering Plough

Animal Health, Milton Keynes, Bucks, UK) subcutaneously at a

dose of 40 mg/kg over the rib cage with a maximum volume

of 10 mL per site. Florfenicol was administered immediately

before carrageenan injection into the tissue cage.

Sampling procedures

Blood samples (10 mL) were collected from a jugular vein into

vacutainers (Becton, Dickinson and Company, Oxford, Oxon,

UK) without anticoagulant at times of 15, 30 and 45 min and

1, 2, 3, 4, 6, 8, 10, 12, 24, 29, 33, 48, 54, 72, 80, 96 and

120 h after administration of florfenicol (Nuflor Injectable

Solution; Intervet Schering Plough Animal Health, Boxmeer,

The Netherlands). Samples were kept at room temperature for

30 min and then placed on ice for 30–60 min to facilitate clot

contraction. The supernatant serum was harvested after centri-

fugation (2000 g for 10 min at 4 °C) and aliquoted into four

polypropylene tubes for subsequent analyses or retention as

reserve samples. All samples were stored at �70 °C until either

analysed for florfenicol concentration or used for measurement

of ex vivo antibacterial activity.

© 2013 John Wiley & Sons Ltd

232 P. Sidhu et al.

Exudate and transudate samples (1.5 mL) were collected

before and at predetermined times (2, 4, 6, 8, 10, 12, 24, 29,

33, 48, 54, 72, 80, 96 and 120 h) after drug administration.

Samples were placed on ice for 30–60 min and then centri-

fuged at 2000 g for 10 min at 4 °C to remove leucocytes. Su-

pernatants were divided into three aliquots and stored at

�70 °C until analysed for florfenicol concentration, used for

measurement of ex vivo antibacterial activity or retained as a

reserve sample.

Analysis of florfenicol in serum, exudate and transudate

Serum, exudate and transudate samples were assayed for florfe-

nicol by a high-performance liquid chromatography (HPLC)

method with UV detection. The method was adapted from one

previously published (De Craene et al., 1997). Briefly, 250 lL of

0.1 M sodium phosphate buffer (pH = 7) and 10 lL of a

250 lg/mL stock solution of chloramphenicol (internal stan-

dard) were added to 250 lL of serum, exudate or transudate,

and the solutions vortexed for 30 sec. After mixing, 2.5 mL

ethyl acetate was added, and the solution was vortexed for

1 min. It was then sonicated for 5 min. After centrifugation at

3500 g for 10 min at 24 °C, 2 mL supernatant was evaporated

to dryness at 45 °C under nitrogen. The residue was reconsti-

tuted in a volume of 250 lL of water:acetonitrile (80:20 v/v),

vortexed for 30 sec and then sonicated for 3 min. A 20 lLaliquot of the reconstituted sample was injected onto the HPLC

column. The system comprised a 125 solvent module pump, a

Rheodyne manual injector (Rheodyne, Cotati, CA, USA) and a

System Gold 166 Detector at 224 nm. The column was an

Ultrasphere C18, 5 lm, 250 9 4.6 lm (Beckman Coulter Ltd.,

High Wycombe, Bucks, UK). The flow rate was 1.0 mL/min,

and the run-time was 8 min. The mobile phase was a methanol:

water mixture (50:50 v/v). All reagents, of HPLC grade, were

obtained from Sigma-Aldrich Chemicals (Poole, Dorset, UK).

Retention times for florfenicol and the internal standard were

approximately 4.5 and 6.0 min, respectively. The lower limit

of quantification (LLOQ) of florfenicol in serum, exudate and

transudate was 0.25 lg/mL. Samples of serum, exudate and

transudate samples collected from drug na€ıve animals were

spiked with florfenicol (a gift from Merck Animal Health) to

prepare standards in the range of 0.25 to 25.0 lg/mL. Stan-

dards were run with each assay. The analytical method was

validated for specificity (lack of interfering peaks), range of

detection, sensitivity as LLOQ and linearity of response as r2 of

the calibration curve. Accuracy and precision were determined

from interday and intraday variances of assays of spiked con-

centrations for drug in each fluid.

Florfenicol binding to serum protein

Serum protein binding of florfenicol was determined in tripli-

cate on each of nine pooled blood samples, harvested at prede-

termined times from the 10 calves used in the

pharmacokinetic study. For each sample, total concentration of

florfenicol was determined as described above. Samples were

then centrifuged at 4000 g for 20 min at 25 °C using an Am-

icon Ultra Centrifugal filter (Ultracel 10K; Millipore (UK) Lim-

ited, Watford, Hertfordshire, UK) and florfenicol concentration

redetermined on the ultrafiltrate.

Pharmacokinetic analyses

Florfenicol concentration–time data in serum, exudate and

transudate for individual calves were analysed by noncompart-

mental analysis (WIN-NONLIN; Pharsight Corporation, Moun-

tain View, CA, USA) using the statistical moment approach

(Yamaoka et al., 1978). The linear trapezoidal rule was used to

calculate areas under concentration–time curves (AUC) and

areas under the first moment curves (AUMC). The mean resi-

dence time (MRT) was determined as AUMC/AUC.

In vitro microbiology

Twenty isolates each of the calf pneumonia pathogens,

M. haemolytica and P. multocida, were supplied on swabs by

the United Kingdom Veterinary Laboratories Agency (VLA). All

isolates had been obtained at post-mortem from clinical cases of

calf pneumonia in various geographical regions of the United

Kingdom. Following receipt, they were stored at �70 °C in a

fluid of composition glycerol:milk:water in the ratio 20:10:70.

This was prepared by mixing 2 g Marvel milk powder (Premier

Foods, Dublin, Ireland) with 10 mL distilled water, and 4 g of

glycerol was added. Distilled water was added to provide a final

volume of 20 mL. To sterilize, the fluid was boiled for 5 sec,

left to cool for 12 h and then boiled again for a further 5 sec.

The solution was stored at �20 °C until used.

Two criteria were used to select six isolates of each organ-

ism for detailed study. First, each isolate was tested for abil-

ity to grow logarithmically in four fluids: Mueller–Hinton

Broth (MHB) and calf serum, exudate and transudate. Sec-

ond, each isolate was investigated for sensitivity to florfenicol

by an initial screen, involving disc diffusion test and mea-

surement of diameter of zone of growth inhibition conducted

using CLSI guidelines (data not shown). Sensitivity was then

confirmed by determination of MIC in MHB, using doubling

dilutions. This preselection procedure ensured that all isolates

could be used to determine MIC and minimum bactericidal

concentration (MBC) in both MHB and the three calf fluids.

MICs and MBCs of each of six selected isolates of both spe-

cies were then determined in MHB and calf fluids by a mic-

rodilution method based on CLSI methodology, but using five

overlapping sets of doubling dilutions for MIC, to improve

accuracy, as described by the study of Aliabadi and Lees

(2001, 2002). MBCs were determined on agar plates using

doubling dilutions.

Ex vivo microbiology

To establish ex vivo time–kill curves, aliquots of serum, transu-

date and exudate samples were pooled from the 10 florfenicol-

treated calves for specific time points, as indicated below:


Florfenicol in calves 233

Fluid Time Point (h)

Serum 0, 12, 24, 33, 48, 72, 96, 120

Exudate and transudate 0, 6, 12, 24, 48, 72, 96, 120

An aliquot from each pooled sample was analysed by HPLC

to determine the florfenicol concentration. Ex vivo bacterial

time–kill curves were established on a second aliquot of the

pooled serum, exudate and transudate samples, as described by

the study of Aliabadi and Lees (2001). Briefly, for each experi-

ment, four isolates each of M. haemolytica and P. multocida

were freshly grown from stocks previously stored at �70 °C. Astarter culture was prepared by inoculating 4 mL of MHB with

a few colonies of the strain to be tested and then incubating

overnight at 37 °C in an orbital shaking incubator to provide

an estimated growth of 1 9 108 colony forming units/mL

(cfu/mL). Fifty microlitre of this culture was diluted 1:50 in

freshly prepared prewarmed MHB and incubated statically for

1 h at 37 °C. The cells were centrifuged at 3100 g for 2 min.

The supernatant was discarded, and the cells resuspended in

50 lL phosphate-buffered saline. The counts of viable cells

were determined by serial dilution and spot-plate counts. Vol-

umes of 396 lL serum, exudate or transudate were prewarmed

in an incubator at 37 °C, and each was inoculated with 4 lLof the starter culture to give a final volume of 400 lL. The cul-

tures were then placed in an orbital shaking incubator at

37 °C. Forty microlitre of each culture was sampled, and the

viable count was determined by serial dilution and spot-plate

counts after 0, 1, 2, 4, 8 and 24 h incubation. The lowest

detectable count was 33 cfu/mL.

PK-PD integration

For PK-PD integration, four surrogates of bacteriological out-

come, T > MIC (time for which florfenicol concentration

exceeded MIC), maximum concentration (Cmax)/MIC ratio and

area under concentration–time curve over 24 h (AUC0–24 h)/

MIC ratio and AUC0–∞/MIC ratio were calculated for each fluid

(serum, exudate and transudate) and each organism, M. haem-

olytica and P. multocida. Results were expressed as

mean � SEM. In addition, a fifth PK-PD variable (partial areas

under in vivo concentration–time curves) was used to express

the ratio of average serum, exudate or transudate concentra-

tion (Cav) relative to MIC for three consecutive 24 h periods

(serum) or four consecutive 24 h periods (exudate and

transudate) after dose administration, which is 0–24, 24–48,

48–72 and 72–96 h.

PK-PD modelling

For PK-PD modelling, AUC0–24 h/MIC data obtained after 24 h

incubation from ex vivo bacterial growth inhibition curves for

four isolates of each species, M. haemolytica and P. multocida,

were modelled to the sigmoidal Emax equation (1):

E ¼ E0 þ Emax � XN

ECN50 � XN

ð1Þ

where Eo is the change in log10 cfu/mL of sample (in serum,

exudate or transudate) after 24 h incubation in the control

sample (no florfenicol) subtracted from the initial inoculum

log10 count, Emax is the maximum growth inhibition deter-

mined as the change in log10 cfu/mL over 24 h in samples

incubated with florfenicol between time 0 and 24 h, EC50 is

the AUC0–24 h/MIC value providing 50% of the maximum

antibacterial effect, X is the AUC0–24 h/MICe in the effect com-

partment (ex vivo site), and N is the Hill coefficient which

describes the slope of the AUC0–24 h/MIC effect curve. These

pharmacodynamic parameters were calculated using a nonlin-

ear regression programme (WIN-NONLIN 5.2; Pharsight Cor-

poration).

PK-PD modelling provided a predicted plot of log10 change

in cfu/mL vs. AUC0–24 h/MIC. From this plot, the antibacte-

rial effect of florfenicol was quantified for three levels of

growth inhibition: bacteriostatic action (no change from ini-

tial inoculum count, E = 0), bactericidal action (3 log10reduction in count, E = �3) and 4 log10 reduction in count

(E = �4).

Daily dosage prediction

Assuming pharmacokinetics linearity, predicted daily doses

were calculated from equation (2):

Dose ¼Cl� ðAUC0�24h

MICeÞ �MIC

F� fuð2Þ

where, Cl = clearance, F = bioavailability, fu = unbound frac-

tion, AUC0–24 h/MICe = experimentally determined ratio of

area under the serum concentration–time curve over 24 h

to MIC for experimental strains of M. haemolytica and

P. multocida, which is the breakpoint PK/PD index to be

achieved, MICl distribution of MICs from the scientific litera-

ture. The daily dose was computed using Monte Carlo simu-

lations in Oracle Crystal Ball (Oracle Corporation, Redwood

Shores, CA, USA) with the following data input: (i) the dis-

tribution of Cl/F obtained for 10 individual calves in the

pharmacokinetic study; (ii) serum AUC0–24 h/MICe ratios

obtained from PK-PD modelling for four strains each of

M. haemolytica and P. multocida; (iii) the distribution of MIClvalues for M. haemolytica (109 strains) and P. multocida (59

strains) considered separately (Thiry et al., 2011). For MICl,

values were corrected to allow for florfenicol protein binding

in serum, as the reported literature values were determined

in broth. The probabilities of distribution for daily doses were

run for 100 000 trials. The daily dose to achieve target

attainment rates of 50% and 90% for bacteriostatic, bacteri-

cidal and 4 log10 reduction in bacterial count was deter-

mined.


234 P. Sidhu et al.

Statistical analyses

Data are presented as mean � SEM or SD. Arithmetic, geomet-

ric and harmonic means were determined, as appropriate, for

each pharmacokinetic variable. Differences between fluids were

determined by ANOVA using the Prism software package (Graph

pad Software Inc., London, UK).

RESULTS

Analytical method validation

The analytical method for florfenicol was validated in bovine

serum, exudate and transudate samples. Selectivity was

assessed for each matrix on six blank samples from a matrix

pool from untreated calves; no interfering peaks were detected

at the retention times of florfenicol and the internal standard,

chloramphenicol. The method was linear over the calibration

range 0.25–25 lg/mL; the linearity of the standard curve was

r2 > 0.999 for all three fluids. The LLOQ of florfenicol in

serum, exudate and transudate was 0.25 lg/mL. The intra-

assay and inter-assay repeatability and reproducibility of the

method were evaluated using spiked concentrations. Intra-

assay and inter-assay coefficients of variation (CV%) were <8%and <5.6% for serum, <8.2% and <7.0% for exudate and

<5.9% and <7.6% for transudate. Percentage recoveries were

95.2 � 2.49% for serum, 98.0 � 4.01% for exudate and

94.6 � 3.69% for transudate.

Florfenicol concentrations in serum, exudate and transudate

The mean (� SEM) concentrations of florfenicol in serum up to

24 h after administration at a dose of 40 mg/kg are presented

in Fig. 1. Florfenicol concentrations (�SEM) in serum, exudate

and transudate up to 120 h after dosing are presented in

Fig. 2. Compartmental pharmacokinetic analysis indicated that

serum concentration–time data best fitted a two compartment

open model, for exudate and transudate the data best fitted a

one compartment model. Pharmacokinetic variables derived by

noncompartmental analysis are presented in Table 1 (serum)

and Table 2 (exudate and transudate).

Florfenicol was present in serum in detectable concentrations

at all sampling times in all calves between 15 min and 80 h.

Fig. 1. Arithmetic plot of mean � SEM florfenicol concentration in

serum over first 24 h after a single subcutaneous injection of florfenicol

at a dose of 40 mg/kg.

Fig. 2. Arithmetic plot of mean � SEM (n = 10) florfenicol concentra-

tion in serum (red), exudate (blue) and transudate (black) after subcu-

taneous injection at a dose of 40 mg/kg.

Table 1. Pharmacokinetic variables for florfenicol in serum (mean and

SEM, n = 10)

Variable (units) Geometric mean (unless stated) SEM

Cmax (lg/mL) 6.04 0.49

Tmax (h) 2.96* 0.15

T½ kz (h) 27.54† 2.26

AUC0–last (lg�h/mL) 154.1 5.44

AUC0–∞ (lg�h/mL) 175.1 16.97

AUMC0–last (lg�h2/mL) 3856 120.1

AUMC0–∞ (lg�h2/mL) 6308 571.9

MRT0–last (h) 25.02 0.49

CI/F (mL/kg/h) 228.5 8.67

*Arithmetic mean.

†Harmonic mean.

Table 2. Pharmacokinetic variables for florfenicol in exudate and tran-

sudate (mean and SEM, n = 10)

Variable (units)

Exudate Transudate

Geometric

mean

(unless

stated) SEM

Geometric

mean

(unless

stated) SEM

Cmax (lg/mL) 3.41 0.22 2.97 0.21

Tmax (h) 15.2* 1.96 15.4* 2.18

T½ kz (h) 27.80† 2.29 28.62† 1.56

AUC0–last (lg�h/mL) 190.2 10.12 143.8 8.40

AUC0–∞ (lg�h/mL) 206.9 10.66 152.8 8.84

AUMC0–last (lg�h2/mL) 7752 482.6 5453 511.5

AUMC0–∞ (lg�h2/mL) 9972 752.3 6861 633.9

MRT0–last (h) 40.76 1.25 37.92 1.72

CI/F (mL/kg/h) 195.7 10.37 261.8 12.99

*Arithmetic mean.

†Harmonic mean.



Mean values of Cmax and Tmax were 6.04 lg/mL and 3.0 h,

respectively, and mean terminal half-life was 27.5 h. In all

calves, Cmax was represented as a single peak. Mean residence

time (MRT) was 25.0 h.

In exudate and transudate, florfenicol was present in quanti-

fiable concentrations at all sampling times up to and including

120 h. Rate of penetration of florfenicol was similar for exu-

date and transudate, with Cmax being achieved at 15.2 h and

15.4 h, respectively. In all calves, Cmax for both fluids was

represented by a single peak. Mean ratio of Cmax serum: Cmax

exudate was 1.77:1, and the corresponding serum: transudate

ratio was 2.03:1. At 24 h and all times up to 120 h, florfeni-

col concentrations in tissue cage fluids were higher than serum

concentrations (Fig. 2). Slightly higher florfenicol concentra-

tions were attained in exudate compared with transudate,

although the difference in Cmax was not significant.

The mean AUC of florfenicol for exudate (190 lg�h/mL) was

somewhat higher than serum (154 lg�h/mL), whilst the mean

AUC value for transudate was slightly lower (144 lg�h/mL) but

differences in AUC values were not statistically significant. The

elimination of florfenicol from tissue cage fluids was similar to

serum terminal half-life. However, longer persistence of drug in

tissue cage fluids compared with serum (P < 0.01) was indicated

by MRT values of 40.8 h and 37.9 h for exudate and transu-

date, respectively, compared with 25.0 h in serum (P < 0.05).

For all three fluids and for all time points, CV for AUCs were

low to moderate (mean = 20.9%, range = 8.4 to 37.0%), signi-

fying limited interanimal variability.

Florfenicol binding to serum protein

For nine pooled serum samples, harvested from the pharmaco-

kinetic study, the mean � SD total concentration was

2.93 � 2.24 lg/mL and percentage free drug concentration

was 82.2 � 4.4.

In vitro MIC and MBC values

For the six strains each of M. haemolytica and P. multocida,

MICs and MBCs are presented in Table 3. All ranges were

narrow. Values in serum were slightly higher than in MHB for

both MIC and MBC. MBC:MIC ratios were equal to or less than

2:1, suggesting a relatively steep concentration–effect relation-

ship.

PK-PD integration

Pharmacokinetic variables established in vivo were integrated

with in vitro MICs. Mean serum values for the integrated phar-

macokinetic and pharmacodynamic variables, Cmax/MIC, AUC0

–24 h/MIC, AUC0–∞/MIC and T > MIC, are presented in Table 4.

For each, means were slightly higher for MHB than for serum,

corresponding to the lower MIC values in MHB for both organ-

isms. The peak serum concentrations were some 10–13 times

greater than MICs for the six tested strains of M. haemolytica

and P. multocida. Peak exudate and transudate concentrations

were 5–6 times MICs (data not shown).

Mean values of average concentrations (Cav) of florfenicol

over successive 24 h periods after dosing, relative to MICs, are

presented in Tables 5 (serum), 6 (exudate) and 7 (transudate).

Cav/MIC ratios exceeded 1:1 for both bacterial species and both

MHB and serum up to 48–72 h and up to 72–96 h for exu-

date and transudate, with one exception, a slightly lower ratio

of 0.91:1 for the period 72–96 h for M. haemolytica in calf

serum.

Ex vivo growth inhibition curves

Ex vivo growth inhibition curves in serum against four strains

each of P. multocida and M. haemolytica are illustrated in

Fig. 3. With high florfenicol concentrations, relative to MIC, a

greater than 6 log10 reduction in bacterial count was obtained

after 4 or 8 h incubation for M. haemolytica and P. multocida,

respectively. The rapid reduction in bacterial count to <33 cfu/

mL indicates a bactericidal action by a concentration-depen-

dent killing mechanism. For M. haemolytica only, slight

regrowth occurred between 8 and 24 h incubation at all con-

centrations investigated. For both organisms, serum samples

collected up to 96 h produced a 1 log10 or greater reduction

from the initial inoculum count after 24 h incubation.

Table 3. Ranges of MIC and MBC values (lg/mL) of florfenicol for

P. multocida and M. haemolytica in Mueller–Hinton broth and calf

serum (n = 6)

P. multocida M. haemolytica

MHB Serum MHB Serum

MIC 0.35–0.45 0.40–0.55 0.50–0.55 0.55–0.65MBC 0.70–1.1 0.85–1.3 0.88–1.0 0.95–1.3

Table 4. PK-PD integration for florfenicol in calf serum for P. multocida and M. haemolytica after subcutaneous dosing at 40 mg/kg (mean and

SEM, n = 10)

Variable (units)

P. multocida M. haemolytica

Serum MIC (0.48 lg/mL) MHB MIC (0.39 lg/mL) Serum MIC (0.58 lg/mL) MHB MIC (0.52 lg/mL)

Cmax/MIC 12.58 (0.49) 15.48 (0.49) 10.41 (0.49) 11.62 (0.49)

AUC0–24 h/MIC (h) 183.4 (5.44) 225.7 (5.44) 151.8 (5.44) 169.3 (5.44)

AUC0–∞/MIC (h) 364.7 (16.97) 448.8 (16.97) 303.9 (16.97) 338.9 (16.97)

T>MIC (h) 78.4 (0.81) 79.1 (0.78) 76.3 (1.20) 77.7 (0.94)


236 P. Sidhu et al.

Ex vivo growth inhibition curves in exudate (Fig. 4) and

transudate (Fig. 5) were similar to those obtained in serum.

For both organisms and both tissue cage fluids, a 5 log10reduction in count was obtained with high florfenicol concen-

trations, relative to MICs, after 8 h incubation with no appar-

ent regrowth after 24 h incubation. With all three fluids,

concentrations approximating to the in vitro determined MICs

produced a 2–3 log10 decrease in count after both 8 and 24 h

incubation ex vivo.

PK-PD modelling of ex vivo growth inhibition data

Data for PK-PD modelling of the ex vivo time–kill data are pre-

sented in Table 8. Three levels of growth inhibition, no change

from initial inoculum count and 3 log10 and 4 log10 reductions

in bacterial count were determined. Mean values for each level

of growth inhibition were broadly similar for each of the three

fluids, serum, exudate and transudate, and, overall, the means

were similar for the two pathogens.

Predicted daily dosages

Based on the distributions of Cl/F in the pharmacokinetic

study (n = 10), AUC0–24 h/MIC ratios for three levels of bac-

terial kill derived from PK/PD modelling (n = 4 for each

organism) and distributions of MIC reported in the scientific

literature (Thiry et al., 2011) predicted once daily dose

requirements are presented in Table 9. Ninety% target

attainment rate for P. multocidawas achieved with predicted

amounts (dosages) of 1.23 mg/kg (bacteriostasis), 2.87 mg/

kg (bactericidal action) and 4.18 mg/kg (4 log10 reduction

in bacterial count). Comparative values for M. haemolytica

were 3.09, 9.55 and 14.06 mg/kg. A worst-case scenario

analysis (highest clearance, lowest protein binding, highest

AUC0–24 h/MIC and highest MIC values amongst reported

distributions) yielded doses of 5.92 and 18.93 mg/kg per

day for eradication of P. multocida and M. haemolytica,

respectively.

Sensitivity analysis for M. haemolytica and P. multocida

revealed that 70.1% and 21.4%, respectively, of the variability

in the calculated dose related to the AUC0–24 h/MICe distribu-

tion. For subsequent calculations, the variance attributable to

the breakpoint distribution was removed by considering the

highest possible breakpoint value for both species for the Monte

Carlo simulation. Then, for M. haemolytica, the remaining vari-

ability in the predicted dose was related predominantly to the

Cl/F distribution (71.6%), with 16.4% and 12.0% attributable

0 4 8 12 16 20 24101102103104105106107108109

101010111012

0 µg/mL<LOQ (120 h)0.259 µg/mL (96 h)0.553 µg/mL (72 h)1.127 µg/mL (48 h)2.819 µg/mL (24 h)4.156 µg/mL (12 h)LOD

MIC = 0.538 µg/mL

M.haemolyticaFlorfenicol Serum

Time (h)

0 4 8 12 16 20 24Time (h)

CFU

/mL

101102103104105106107108109

101010111012

CFU

/mL

0 µg/mL4.156 µg/mL (12 h)2.819 µg/mL (24 h)1.127 µg/mL (48 h)0.553 µg/mL (72 h)0.259 µg/mL (96 h)<LOQ (120 h)LOD

MIC = 0.475 µg/mLP.multocidaFlorfenicol serum

(a)

(b)

Fig. 3. Ex vivo growth inhibition for florfenicol in serum against (a)

M. haemolytica and (b) P. multocida (mean, n = 4). SEM bars not shown

for clarity.

Table 5. Average serum florfenicol concentration/MIC ratios for three 24 h periods after subcutaneous dosing at 40 mg/kg (n = 10)

Time period after dosing (h)

Cav/MIC

0–24 24–48 48–72

P. multocida Value based on mean MIC in serum (0.48 lg/mL) 7.65 3.58 1.85

Value based on mean MIC in broth (0.39 lg/mL) 9.41 4.41 2.28

M. haemolytica Value based on mean MIC in serum (0.58 lg/mL) 6.33 2.97 1.53

Value based on mean MIC in broth (0.52 lg/mL) 7.06 3.31 1.71

Table 6. Average exudate florfenicol concentration/MIC ratios for four

24 h periods after subcutaneous dosing at 40 mg/kg (n = 10)


Cav/MIC

0–24 24–48 48–72 72–96

P. multocida 5.15 5.19 3.21 1.71

M. haemolytica 4.26 4.29 2.66 1.41

Table 7. Average transudate florfenicol concentration/MIC ratios for

four 24 h periods after subcutaneous dosing at 40 mg/kg (n = 10)


Cav/MIC

0–24 24–48 48–72 72–96

P. multocida 4.08 3.73 2.00 1.10

M. haemolytica 3.38 3.09 1.66 0.91



to MIC and free fraction distributions, respectively. Sensitivity

analysis for P. multocida indicated that 69% of the variability

in the calculated dose was attributable to the MICl distribution,

26.8% was due to Cl/F distribution and 4.1% was due to distri-

bution of the fraction of florfenicol unbound.

DISCUSSION

The excellent responses in clinical use of florfenicol in bovine

respiratory disease (Varma et al., 1998; Catry et al., 2008) can

probably be ascribed to several factors. (i) Resistance of the

principal bacterial pathogens, M. haemolytica and P. multocida,

seems not to be widespread, despite clinical use for some

20 years. Shin et al. (2005) reported on the sensitivity of cattle

and pig pathogens; all Actinobacillus pleuropneumoniae, P. multo-

cida, M. haemolytica and 98.6% of isolates of Bordetella bronchi-

septica were sensitive to florfenicol, with MIC90 values ≤1 lg/mL.

Lizarazo et al. (2006) also reported MIC50 and MIC90 values of

0.5 and 1.0 lg/mL and no resistance in 132 isolates of

P. multocida harvested from cases of porcine pneumonia over

the period 2003–2004. (ii) Recent studies have demonstrated

that florfenicol exerts a protective effect in acute lung injury

induced by lipopolysaccharide (Zhang et al., 2009), and it sig-

nificantly increases survival time in murine endotoxaemia

(Zhang et al., 2008). These beneficial effects were attributed to

modulation of early cytokine responses in shock by blocking

the NF-jB pathway. Florfenicol therefore possesses significant

host animal immunomodulatory properties, which, in terms of

bacterial kill, may be additive to or synergistic with its direct

actions on bacterial protein synthesis. (iii) In clinical use, the

drug is administered intramuscularly at a dosage of 20 mg/kg

once daily or subcutaneously at a dose of 40 mg/kg as a single

dose. For both parenteral routes (and the two dosages), the ter-

minal half-life is much longer than the elimination half-life

determined after intravenous dosing. Administration by the

two nonvascular parenteral routes, therefore, in the commer-

cially available ‘depot’ formulation results in flip-flop pharma-

cokinetics (Varma et al., 1998). The prolonged terminal half-

life, especially after subcutaneous dosing at the 40 mg/kg dose,

of 27.5 h in serum in this study thus ensures that concentra-

tions in serum are maintained in excess of MICs of key patho-

gens for up to 3 days or longer. These high concentrations are

likely to ensure a high level of efficacy and also minimize

opportunities for the emergence of resistance.

Most standard texts describe drugs of the fenicol group,

including florfenicol, as bacteriostatic agents. However, the

studies of Varma et al. (1998), De Haas et al. (2002) and the

data reported in the present investigation indicate that this is

not so for the growth inhibiting action of florfenicol against iso-

lates of M. haemolytica and P. multocida harvested from clinical

cases of calf pneumonia. The ex vivo growth inhibition curves

were characteristic of a concentration-dependent killing action;

a high level of kill was attained within 4–8 h of exposure, and

the concentration–response relationship was steep. Moreover, a

bactericidal level of action (3 log10 reduction in bacterial count)

was well maintained in all three fluids, serum, exudate and

transudate. Furthermore, based on modelled ex vivo serum

AUC0–24 h/MIC ratios, the average florfenicol concentrations

0 4 8 12 16 20 24101102103104105106107108109

1010

0 µg/mL0.46 µg/mL (2 h)3.03 µg/mL (12 h)2.48 µg/mL (24 h)1.80 µg/mL (48 h)0.81 µg/mL (72 h)0.48 µg/mL (96 h)0.22 µg/mL (120 h)

Serum MIC = 0.054 µg/mL

LOD

M.haemolyticaFlorfenicol transudate

Time (h)

0 4 8 12 16 20 24Time (h)

CFU

/mL

101102103104105106107108109

1010

CFU

/mL

0 µg/mL0.46 µg/mL (2 h)3.03 µg/mL (12 h)2.48 µg/mL (24 h)1.80 µg/mL (48 h)0.81 µg/mL (72 h)0.48 µg/mL (96 h)


0.22 µg/mL (120 h)

P.multocidaFlorfenicol transudate

LOD

(a)

(b)

Fig. 5. Ex vivo growth inhibition for florfenicol in transudate against (a)

M. haemolytica and (b) P. multocida (n = 4). SEM bars not shown for clarity.

0101102103104105106107108109

1010

4 8 12 16 20 24

0 µg/mL0.93 µg/mL (2 h)3.13 µg/mL (12 h)2.79 µg/mL (24 h)1.34 µg/mL (48 h)0.97 µg/mL (72 h)0.32 µg/mL (96 h)0.30 µg/mL (120 h)


LOD

M.haemolyticaFlorfenicol exudate

Time (h)

0 4 8 12 16 20 24Time (h)

CFU

/mL

101102103104105106107108109

1010

CFU

/mL

0 µg/mL0.93 µg/mL (2 h)3.13 µg/mL (12 h)2.79 µg/mL (24 h)1.34 µg/mL (48 h)0.97 µg/mL (72 h)0.32 µg/mL (96 h)


0.30 µg/mL (120 h)LOD

P.multocidaFlorfenicol exudate

(a)

(b)

Fig. 4. Ex vivo growth inhibition for florfenicol in exudate against (a)

M. haemolytica and (b) P. multocida (n = 4). SEM bars not shown for clarity.


238 P. Sidhu et al.

over the 24-h incubation period, expressed as multiples of MIC,

were 0.34, 1.11 and 1.63 for M. haemolytica count reductions

of 0, 3 log10 and 4 log10. Corresponding values of P. multocida

were 0.32, 0.75 and 1.04, respectively. Thus, concentrations

similar to or slightly greater than in vitro MICs reduced

bacterial counts by 3 log10 cfu/mL. High levels of bacterial kill

in time–kill studies were previously reported for florfenicol by

De Haas et al. (2002) for single isolates of M. haemolytica,

P. multocida and H. somnus at MIC concentrations.

In interpreting data from the tissue cage model, it should be

noted that the rate and extent of penetration of any drug into

intracaveal fluids depend in part on the serum concentration–

time profile, as this provides the driving concentration for

passage into exudate and transudate. However, penetration

depends also on whether the granulation tissue within the case

has been stimulated with a mild irritant, such as carrageenan

and, for several drugs of the NSAID class in several species,

penetration rate, and extent into exudate exceeds penetration

into transudate (Lees et al., 2004). A trend for greater penetra-

tion into exudate of florfenicol was noted in this study (AUC

exudate:transudate ratio = 1.31:1), but this difference was not

statistically significant. The lack of significance may be due to

low percentage binding of florfenicol to serum proteins,

because a factor favouring penetration of drugs such as NSA-

IDs is the very high degree of protein binding (Lees et al.,

2004). In addition, a third factor influencing drug penetration

into tissue cages is model dependency, varying with the shape

of the tissue cage and, in particular, the surface area:volume

ratio. Therefore, drug penetration into subcutaneous tissue

cages cannot be equated to penetration into other tissues,

whether inflamed or noninflamed, under clinical circum-

stances. In fact, tissue cages have been described, from a phar-

macokinetic perspective, as a deep (as opposed to superficial)

peripheral compartment (Clarke, 1989; Clarke et al., 1989).

The fact that the terminal half-life of florfenicol was similar for

both tissue cage fluids and plasma, whereas MRT was longer

for the tissue cage fluids than for plasma, is explained by the

fact that pseudo-steady state has been achieved; when the ter-

minal half-life in serum is very long (as here), the drug exit

rate from the tissue cage is no longer the limiting step for elim-

ination from the tissue cage.

The principal value of tissue cages is, therefore, not to study

extravascular drug distribution profiles per se but to provide a

simple, ethically acceptable technique for undertaking PK-PD

modelling. The latter technique enables determination of the

whole sweep of concentration antimicrobial–effect relationship

for three biological fluids (serum, exudate and transudate). As

in previous investigations with fluoroquinolones in farm ani-

mal species (Aliabadi & Lees, 2001, 2002; Aliabadi et al.,

2003), mean AUC0–24 h/MIC ratios for three levels of growth

inhibition were numerically broadly similar for each of the flu-

ids. This similarity for the three fluids gives confidence in the

Table 8. PK-PD modelling of ex vivo florfenicol data (mean and SD, n = 4 for each organism) after subcutaneous administration at a dose of

40 mg/kg

Organism Parameter (units)

Serum Exudate Transudate

Mean SD Mean SD Mean SD

M. haemolytica Log E0 (cfu/mL) 2.38 0.67 0.89 0.35 1.66 0.64

Log Emax (cfu/mL) �6.38 0.65 �5.22 0.48 �5.28 0.56

Log Emax – Log E0 (cfu/mL) �8.76 1.14 -6.11 0.25 �6.94 0.60

AUC24 h/MIC for bacteriostatic action (h) 8.16 3.22 6.96 4.12 6.39 1.42

AUC24 h/MIC for bactericidal action (h) 26.63 8.33 17.88 4.25 18.64 2.22

AUC24 h/MIC for 4 log10 reduction in count (h) 39.04 12.10 26.78 11.41 27.84 5.19

P. multocida Log E0 (cfu/mL) 2.34 0.32 1.18 0.41 2.06 0.39

Log Emax (cfu/mL) �5.96 0.24 �4.99 0.57 0.58 0.29

Log Emax – Log E0 (cfu/mL) �8.31 0.22 �6.18 0.93 �7.81 0.44

AUC24 h/MIC for bacteriostatic action (h) 7.62 1.50 9.13 5.53 2.35 1.08

AUC24 h/MIC for bactericidal action (h) 18.06 3.14 17.34 2.16 17.86 3.09

AUC24 h/MIC for 4 log10 reduction in count (h) 25.04 5.78 24.14 3.90 44.66 14.41

Table 9. Predicted once daily dosages based on PK-PD modelling of ex

vivo florfenicol data in serum

Target attainment rate

90% 50%

Predicted daily doses for P. multocida

Bacteriostatic 1.23 0.82

Bactericidal 2.87 1.94

4 log10 reduction in count 4.18 2.63

Predicted daily doses for M. haemolytica

Bacteriostatic 3.09 1.83

Bactericidal 9.55 6.36

4 log10 reduction in count 14.06 9.32

Dosages calculated using equation 2: (i) with Cl/F range of 163–272 mL/kg/h (n = 10); (ii) AUC0–24 h/MICe distribution ranges (n = 4)

of 5.8–9.4 h (P. multocida) and 4.8–11.0 h (M. haemolytica) for bacte-

riostasis, 15.2–21.3 h (P. multocida) and 17.2–33.8 h (M. haemolytica)

for bactericidal action, and 19.4–32.0 h (P. multocida) and 25.6–51.2 h (M. haemolytica) for 4 log10 reduction in count; and (iii) MICldistributions as reported by Thiry et al. (2011) of 0.25–0.50 lg/mL

(n = 59) for P. multocida and 0.50 to 1.00 lg/mL (n = 109) for

M. haemolytica. For MICl, values were corrected by 18% to allow for

florfenicol protein binding in serum, as the reported literature values

were determined in broth.



methodology for predicting effective antimicrobial drug concen-

trations in vivo. The present findings thus suggest that similar

numerical values for the three levels of growth inhibition con-

sidered are likely to apply for florfenicol for most extracellular

and transcellular fluids. The findings therefore lend support to

the PK-PD modelling approach for estimating an effective in

vivo dosage regimen.

The present findings extend previous data from our labora-

tory in three respects. (i) The findings indicate that PK-PD mod-

elling can be extended to florfenicol and previous studies in our

laboratory have been restricted to drugs of the fluoroquinolone

class in a range of farm animal species. (ii) Our previous studies

investigated only one bacterial species, M. haemolytica, whereas

the present study provides information on a second clinically

important calf pathogen, P. multocida. (iii) Our previous investi-

gations used a single isolate of M. haemolytica and therefore

provided no information on interstrain variability in levels of

the surrogate marker AUC24 h/MIC required for each level of

bacterial growth inhibition. By using four to six strains of each

bacterial species, the present investigation provided new data

on the variability in response between those strains. Thus, CV%

AUC0–24 h/MIC values for bacteriostatic, bactericidal and 4

log10 count reduction responses were 39, 31 and 31, respec-

tively, for M. haemolytica. Corresponding values for P. multocida

were somewhat lower, 20, 17 and 23, respectively, indicating

lesser variability between isolates for this species. This might

ensure good consistency between animals in drug concentra-

tions achieving therapeutic efficacy.

The florfenicol dosage, product formulation and administra-

tion route used in this investigation were designed by the

manufacturer to achieve, for single-dose administration, a

therapeutic effect over at least 72 h against most or all clini-

cal isolates of M. haemolytica and P. multocida. PK-PD integra-

tion of data indicated that this objective was readily achieved

against the 12 isolates (six strains of each bacterial species)

investigated. Thus, for both MHB and calf serum, Cmax/MIC

exceeded 10:1, AUC0–24 h/MIC exceeded 150 h, AUC0–∞/MIC

exceeded 300 h and T>MIC was greater than 75 h. As the

MIC values for strains of both organisms used in this investi-

gation were typical of those reported for field isolates (Priebe

& Schwarz, 2003; Kehrenberg et al., 2004; Shin et al., 2005;

Dowling, 2006; Lizarazo et al., 2006; Zhang et al., 2009),

these integrated variables are relevant to prediction of the effi-

cacy of florfenicol in clinical use in calf pneumonia. More-

over, Cav values over successive 24 h periods exceeded the

MIC for serum for the three periods up to 72 h and for tissue

cage fluids for all periods up to 96 h, except for a single

value for transudate. The anticipated effectiveness of florfeni-

col over these extended periods, based on integrating in vivo

pharmacokinetic with in vitro pharmacodynamic data, was

confirmed by the ex vivo growth inhibition curves on three

counts: (i) with the high florfenicol concentrations achieved

in vivo reductions in bacterial count of 5 log10 cfu/mL or

greater were obtained; (ii) the speed of bacterial killing was

rapid, indicating a concentration-dependent killing action; and

(iii) high levels of growth inhibition were maintained for up

to 72 h in serum and up to 96 h in exudate and transudate.

The product used in this study was a depot/slow release for-

mulation of florfenicol, for which the manufacturer’s recom-

mended dose is 40 mg/kg subcutaneously as a single dose (the

dose and route used in the pharmacokinetic part of this study)

and 20 mg/kg administered intramuscularly twice with an

interval of 48 h. However, growth inhibition produced by

M. haemolytica and P. multocida was established ex vivo (in the

pharmacodynamic part of this study) over a 24-h incubation

period. Therefore, from these experimental data, it was possible

for 24-h dosage regimens to be calculated for each bacterial

species. The findings were thus used to design daily dosage

schedules for the drug substance rather than the drug product.

Based on Monte Carlo simulations providing a 4 log10 reduc-

tion in bacterial count and literature reported MIC distribu-

tions, the estimated amounts (daily dose) for a 90% target

attainment rate were 14.1 mg/kg for M. haemolytica and

4.2 mg/kg for P. multocida.

ACKNOWLEDGMENTS

This study was supported by a generous grant from the

Department for the Environment, Food and Rural Affairs

(United Kingdom). Merck Animal Health supplied a sample of

pure florfenicol for HPLC and microbiological analyses. Prof.

P. Sidhu was sponsored by a UKIERI fellowship.

REFERENCES

Adams, P.E., Varma, K.J., Powers, T.E. & Lamendola, J.F. (1987) Tissue

concentrations and pharmacokinetics of florfenicol in male veal

calves given repeated doses. American Journal of Veterinary Research,

48, 1725–1732.Aliabadi, F.S. & Lees, P. (2001) Pharmacokinetics and pharmacody-

namics of danofloxacin in serum and tissue fluids of goats following

intravenous and intramuscular administration. American Journal of

Veterinary Research, 62, 1979–1989.Aliabadi, F.S. & Lees, P. (2002) Pharmacokinetics and pharmacoki-

netic/pharmacodynamic integration of marbofloxacin in calf serum,

exudate and transudate. Journal of Veterinary Pharmacology and Thera-

peutics, 25, 161–174.Aliabadi, F.S., Landoni, M.F. & Lees, P. (2003) Pharmacokinetics (PK),

pharmacodynamics (PD) and PK-PD integration of danofloxacin in

sheep biological fluids. Antimicrobial Agents and Chemotherapy, 47,

626–635.Anadon, A., Martinez, M.A., Martinez, M., Rios, A., Cabellero, V., Ares,

I. & Martinez-Laranga, R.M. (2008) Plasma and tissue depletion of

florfenicol and florfenicolamine in chickens. Journal of Agriculture

Food Chemistry, 56, 11049–11056.Aslan, V., Maden, M., Erganis, O., Birdane, F.M. & Corlu, M. (2002)

Clinical efficacy of florfenicol in the treatment of calf respiratory tract

infections. Veterinary Quarterly, 24, 35–39.

Ayling, R.D., Baker, S.E., Peek, M.L., Simon, A.J. & Nicholas, R.A.J.

(2000) Comparison of in vitro activity of danofloxacin, florfenicol,

oxytetracycline, spectinomycin and tilmicosin against recent field iso-

lates of Mycoplasma bovis. Veterinary Record, 146, 745–747.


240 P. Sidhu et al.

Berge, A.C.B., Epperson, W.B. & Pritchard, R.H. (2005) Assessing the

effect of a single dose florfenicol treatment in feedlot cattle on the

antimicrobial resistance patterns in faecal Escherichia coli. Veterinary

Research, 36, 723–734.Bretzlaff, K.N., Neff-Davis, C.A., Ott, R.S., Koritz, G.D., Gustafsson, B.K.

& Davis, L.E. (1987) Florfenicol in non-lactating dairy cows: phar-

macokinetics, binding to plasma proteins and effects on phagocytosis

by blood neutrophils. Journal of Veterinary Pharmacology and Therapeu-

tics, 10, 233–240.

Cannon, M., Harford, S. & Davies, J. (1990) A comparative study on

the inhibitory actions of chloramphenicol, thiamphenicol and some

fluorinated analogs. Journal of Antimicrobial Chemotherapy, 26, 307–317.

Catry, B., Haesebrouck, F., De Vliegher, S., Feyen, B., Vanrobaeys, M.,

Opsomer, G., Schwarz, S. & De Kruif, A. (2005) Variability in

acquired resistance of Pasteurella and Mannheimia isolates from the

nasopharynx of calves with particular reference to different herd

types. Microbial Drug Resistance, 11, 387–394.Catry, B., Duchateau, L., Van De Ven, J., Laevens, H., Opsomer, G.,

Haesebrouck, F. & De Kruif, A. (2008) Efficacy of metaphylactic flor-

fenicol therapy during natural outbreaks of bovine respiratory dis-

ease. Journal of Veterinary Pharmacology and Therapeutics, 31, 479–487.

Clarke, C.R. (1989) Tissue-chamber modeling systems-applications in

veterinary medicine. Journal of Veterinary Pharmacology and Therapeu-

tics, 12, 349–368.

Clarke, C.R., Short, C.R., Bourne, D.W. & Usenik, E.A. (1989) Subcuta-

neously implanted tissue chambers – a pharmacokinetic study. Jour-

nal of Veterinary Pharmacology and Therapeutics, 12, 312–321.De Craene, B.A., Deprez, P., D’Haese, E., Nelis, H.J., Van den Bossche,

W. & De Leenheer, P. (1997) Pharmacokinetics of florfenicol in cere-

brospinal fluid and plasma of calves. Antimicrobial Agents and Chemo-

therapy, 41, 1991–1995.De Haas, V., Bonnier, M., Gicquel, M., Etore, F. & Shuster, D. (2002)

Florfenicol: a time- or concentration-dependent antibiotic? New

advances in calf disease management. XXII World Buiatrics Con-

gress, IV, 17–25.Dowling, P. (2006) Chloramphenicol, thiamphenicol and florfenicol. In

Antimicrobial Therapy in Veterinary Medicine, 4th edn. Eds Giguere, S.,

Prescott, J.F., Baggot, J.D., Walker, R.D. & Dowling, P.M., pp. 241–

248. Blackwell Publishing, Oxford.

Hoar, B.R., Jelinski, M.D., Ribble, C.S., Janzen, E.D. & Johnson, J.C.

(1998) A comparison of the clinical field efficacy and safety of florfe-

nicol and tilmicosin for the treatment of undifferentiated bovine

respiratory disease in cattle in Western Canada. Canadian Veterinary

Journal, 39, 161–166.

Jim, G.K., Booker, C.W., Guichon, P.T., Schunicht, O.C., Wildman, B.K.,

Johnson, J.C. & Lockwood, P.W. (1999) A comparison of florfenicol

and tilmicosin for the treatment of undifferentiated fever in feedlot

calves in Western Canada. Canadian Veterinary Journal, 40, 179–184.

Kehrenberg, C. & Schwarz, S. (2005) Plasmid-borne florfenicol resis-

tance in Pasteurella multocida. Journal of Antimicrobial Chemotherapy,

55, 773–775.Kehrenberg, C., Mumme, J., Wallmann, J., Verspohl, J., Tegeler, R.,

Kuhn, T. & Schwarz, S. (2004) Monitoring of florfenicol susceptibility

among bovine and porcine respiratory tract pathogens collected in

Germany during the years 2002 and 2003. Journal of Antimicrobial

Chemotherapy, 54, 572–574.

Koc, F., Ozturk, M., Kadioglu, Y., Dogan, E., Yanmaz, L.E. & Okumus,

Z. (2009) Pharmacokinetics of florfenicol after intravenous and

intramuscular administration in New Zealand White rabbits.

Research in Veterinary Science, 87, 102–105.

Lacroix, M.Z., Gayrard, V., Picard-Hagen, N. & Toutain, P.L. (2011)

Comparative bioavailability between two routes of administration of

florfenicol and flunixin in cattle. R�evue de M�edecineV�eterinaire, 162,

321–324.Lane, V.M., Villarroel, A., Wetzlich, S.E., Clifford, A., Taylor, I. & Craig-

mill, A.L. (2007) Tissue residues of florfenicol in sheep. Journal of

Veterinary Pharmacology and Therapeutics, 31, 178–180.

Lees, P., Giraudel, J., Landoni, M.F. & Toutain, P.L. (2004) PK-PD

integration and PK-PD modeling of nonsteroidal anti-inflamma-

tory drugs: principles and applications in veterinary pharmacol-

ogy. Journal of Veterinary Pharmacology and Therapeutics, 27, 491–

502.

Li, J.Z., Fung, K.F., Chen, Z.L., Zeng, Z.L. & Zhang, J. (2002) Tissue

pharmacokinetics of florfenicol in pigs experimentally infected with

Actinobacilluspleuropneumoniae. European Journal of Drug Metabolism

and Pharmacokinetics, 27, 265–271.Liu, J., Fung, K.F., Chen, Z., Zeng, Z. & Zhang, J. (2003) Pharmacoki-

netics of florfenicol in healthy pigs and in pigs experimentally

infected with Actinobacillus pleuropneumoniae. Antimicrobial Agents

and Chemotherapy, 47, 820–823.Lizarazo, Y.A.V., Ferri, E.F.R., Martin de la Fuente, A.J. & Martin,

C.B.G. (2006) Evaluation of changes in antimicrobial susceptibility

patterns of Pasteurella multocida subsp. multocida isolates from pigs in

Spain in 1987–1988 and 2003–2004. American Journal of Veterinary

Research, 67, 663–668.

Lobell, R.D., Varma, K.J., Johnson, J.C., Sams, R.A., Gerken, D.F. &

Ashcraft, S.M. (1994) Pharmacokinetics of florfenicol following intra-

venous and intramuscular doses to cattle. Journal of Veterinary Phar-

macology and Therapeutics, 17, 253–258.Park, B.K., Lim, J.H., Kim, M.S., Hwang, Y.H. & Yun, H.I. (2008) Phar-

macokinetics of florfenicol and its metabolite, florfenicol amine, in

dogs. Research in Veterinary Science, 84, 85–89.

Priebe, S. & Schwarz, S. (2003) In vitro activities of florfenicol against

bovine and porcine respiratory tract pathogens. Antimicrobial Agents

and Chemotherapy, 47, 2703–2705.Schwarz, S., Kehrenberg, C., Doublet, B. & Cloeckaert, A. (2004)

Molecular basis of bacterial resistance to chloramphenicol and florfe-

nicol. FEMS Microbiology Reviews, 28, 519–542.

Shen, J., Hu, D., Wu, X. & Coats, J.R. (2003) Bioavailability and phar-

macokinetics of florfenicol in broiler chickens. Journal of Veterinary

Pharmacology and Therapeutics, 26, 337–341.Shin, S.J., Kang, S.G., Nabin, R., Kang, M.L. & Yoo, S.H. (2005) Evalu-

ation of the antimicrobial activity of florfenicol against bacteria iso-

lated from bovine and porcine respiratory disease. Veterinary

Microbiology, 106, 73–77.Sidhu, P., Aliabadi, S.F., Andrews, M. & Lees, P. (2003) Tissue cham-

ber model of acute inflammation in farm animal species. Research in

Veterinary Science, 74, 67–77.

Soback, S., Paape, M.J., Filep, R. & Varma, K.J. (1995) Florfenicol phar-

macokinetics in lactating cows after intravenous, intramuscular and

intramammary administration. Journal of Veterinary Pharmacology

and Therapeutics, 18, 413–417.

Switala, M., Hrynyk, R., Smutkiewicz, A., Jaworski, K., Pawlowski,

P., Okoniewski, P., Grabowski, T. & Dewbowy, T. (2007) Pharma-

cokinetics of florfenicol, thiamphenicol and chloramphenicol in

turkeys. Journal of Veterinary Pharmacology and Therapeutics, 30,

145–150.Thiry, J., Rubion, S., Sarasola, P., Bonnier, M., Hartmann, M. & de

Haas, V. (2011) Efficacy and safety of a new 450 mg/ml florfenicol

formulation administered intramuscularly in the treatment of

bacterial bovine respiratory disease. Veterinary Record, 169,

526–530.



Varma, K.J., Adams, P.E., Powers, T.E., Powers, J.D. & Lamendola, J.F.

(1986) Pharmacokinetics of florfenicol in veal calves. Journal of Vet-

erinary Pharmacology and Therapeutics, 9, 412–425.

Varma, K.J., Walker, R.D., Sams, R.A., Holland, R.E. & Trenti, F.

(1994) Pharmacokinetics and distribution of florfenicol in bronchial

secretions and tissue cage fluid in cattle. Proceedings 18th World Bui-

atrics Congress, 54, 7–550.

Varma, K.J., Lockwood, P.W., Cosgrove, S.B. & Rogers, E.R. (1998)

Pharmacology, safety and clinical efficacy of Nuflor (florfenicol) fol-

lowing subcutaneous administration to cattle. Schering-Plough Animal

Health Nuflor�: New therapeutic applications. Proceedings of a

Symposium held in conjunction with the XX World Buitarics Con-

gress, pp. 13–19, Sydney, Australia.Yamaoka, K., Nakagawa, T. & Uno, T. (1978) Statistical moments in

pharmacokinetics. Journal of Pharmacokinetics and Biopharmaceutics, 6,

547–558.

Zhang, X., Song, Y., Ci, X., An, N., Fan, J., Cui, J. & Deng, X. (2008)

Effects of florfenicol on early cytokine responses and survival in mur-

ine endotoxaemia. International Immunopharmacology, 8, 982–988.Zhang, X., Song, K., Xiong, H., Li, H., Chu, X. & Deng, X. (2009) Protec-

tive effect of florfenicol on acute lung injury induced by lipopolysac-

charide in mice. International Immunopharmacology, 9, 1525–1529.


242 P. Sidhu et al.

pharmacokinetic pharmacodynamic integration and modelling ...€¦ · 1997). the latter group...

Documents