INTRODUCTION

Halibut (Hippoglossus hippoglossus) and turbot (Scophthalmus

maximus) are the two most important nonsalmonid fish species

now emerging in the Norwegian aquaculture industry. The

production in 1997 was 60 tons/270 000 fry of halibut and 40

tons/300 000 fry of turbot (J. Stoss, personal communication).

The farming of these two species has already encountered

problems with bacterial infections, such as vibrio (Vibrio

anguillarum) and infections with atypical Aeromonas salmonicida.

To be able to apply the right drug and the best dosage regimen

for successful treatment and minimal environmental hazards,

knowledge about the pharmacokinetic and pharmacodynamic

properties of the drug(s) in the actual species are vital.

Treatment with antimicrobial agents in flatfish has mainly

been conducted extrapolating the pharmacokinetic and phar-

macodynamic data known from salmonids (Atlantic salmon,

Salmo salar and rainbow trout, Oncorhynchus mykiss).

Knowledge about the pharmacokinetic properties of anti-

bacterial agents in flatfish, following different routes of

administration, are scarce. One commonly used antimicrobial

agent in flatfish is flumequine. Flumequine is a broad spectrum

synthetic antimicrobial agent belonging to the 4-quinolones and

has the properties of a weak acid.

The aim of this study was to investigate the pharmacokinetic

properties of flumequine in halibut and turbot.

MATERIALS AND METHODS

Formulation of test substance

There are no commercial formulations of flumequine available in

Norway for intravenous (i.v.) or bath administration. Flume-

quine was obtained from Sigma Chemical Co., St. Louis, MO. The

solution for i.v. administration was prepared by dissolving

flumequine in 1.0 M NaOH, regulating the pH to 10.3 with 6 M

HCl. Further lowering of the pH resulted in precipitation of

flumequine. The final volume was adjusted with 0.9% saline to a

concentration of 10 g/L. Flumequine for oral administration was

#1999 Blackwell Science Ltd 122

J. vet. Pharmacol. Therap. 22, 122±126, 1999. PHARMACOKINETICS

Single-dose pharmacokinetics of flumequine in halibut (Hippoglossus

hippoglossus) and turbot (Scophthalmus maximus)

Hansen, M. K., Horsberg, T. E. Single-dose pharmacokinetics of flumequine in

halibut (Hippoglossus hippoglossus) and turbot (Scophthalmus maximus). J. vet.

Pharmacol. Therap. 22, 122±126.

Flumequine was administered to halibut (Hippoglossus hippoglossus) and turbot

(Scophthalmus maximus) intravenously (i.v.) and orally (p.o.) at a dose of 10 mg/

kg bodyweight, and as a bath-treatment at a dose of 10 mg/L water for 2 h,

using identical experimental designs. The study was performed in seawater with

a salinity of 3% and a temperature of 10.3+0.48C (halibut) and 18.0+0.3 8C(turbot). Pharmacokinetic modelling of the data showed that flumequine had

quite similar pharmacokinetic properties in halibut and turbot. Following

intravenous administration, the volumes of distribution at steady state (Vss)

were 2.99 L/kg (halibut) and 3.75 L/kg (turbot). Plasma clearances (Cl) were

0.12 L/kg (halibut) and 0.17 L/h.kg (turbot) and the elimination half-lives

(t�lz) were calculated to be 32 h (halibut) and 34 h (turbot). Mean residence

times (MRT) were 25.1 h (halibut) and 22.2 h (turbot). Following oral

administration, the t�lz were 43 h (halibut) and 42 h (turbot). Maximal plasma

concentrations (tmax) were 1.4 mg/L (halibut) and 1.9 mg/L (turbot), and were

observed 7 h post administration in both species. The oral bioavailabilities (F)

were calculated to 56% (halibut) and 59% (turbot). Following bath adminis-

tration maximal plasma concentrations were 0.08 mg/L (halibut) and 0.14 mg/

L (turbot), and were observed 0 h (halibut) and 3 h (turbot) after the end of the

bath. The bioavailability in halibut following a 2-h bath treatment was 5%.

(Paper received 29 May 1998; accepted for publication 8 December 1998)

M. K. Hansen, Department of Pharmacology, Microbiology and Food Hygiene, Nor-

wegian College of Veterinary Medicine, PO Box 8146 Dep. N-0033 Oslo, Norway.

M. K. HANSEN &

T. E. HORSBERG

Department of Pharmacology, Microbiology

and Food Hygiene, Norwegian College of

Veterinary Medicine, Oslo, Norway

Ahed

Bhed

Ched

Dhed

Ref marker

Fig marker

Table mar-

ker

Ref endRef start

Paper 191 Disc

mixed into a 4 : 3 emulsion of ordinary fishfeed: cod liver oil at a

concentration of 10 g/L (1 mL=1.0076 g, SD=0.0132 g) after

first being dissolved in 1.0 M NaOH. Flumequine for bath

administration was prepared by dissolving flumequine in 1 M

NaOH. The concentrations in the fish tanks were confirmed with

an HPLC-assay.

Test facilities and test fish

The study was conducted at Stolt Seafarm, éye, Norway. The

fish had been raised at the farm, and were held in fibreglass

tanks of 1000 L supplied with running seawater with a salinity

of & 3%. The halibut (Hippoglossus hippoglossus) were & 1-year-

old and weighed 89+15 g (mean+ SD). They were held in

running seawater at the optimal growth temperature of

10.3+0.4 8C. The turbot (Scophthalmus maximus) were & 8-

month-old and weighed 210+55 g (mean+ SD). They were

held in running seawater at the optimal growth temperature of

18.0+0.3 8C.The experimental design was identical for each of the

two species: 90 fish of each species were randomly allocated

into three groups of 30 each, one group for each route

of administration.

Intravenous administration

The fish to be administered flumequine by the i.v. route were

allocated into groups of six, and were administered flumequine

i.v. individually at a dose of 10 mg/kg into the caudal vein. The

i.v. injection was performed with the fish positioned on a damp

cloth, with the bottomside down during injection, after

anaesthetizing with metomidate (10 mg/L water) (HypnodilTM,

Janssen Pharmaceutica, Beerse, Belgium) and weighing. The

flumequine solution was slowly injected using a 1-mL disposable

syringe and a 0.5 6 25 mm needle (Terumo, Leuven, Belgium).

The position of the needle was confirmed by aspiration of blood

before, during and after the injection. Fish in which the needle

dislocated during the injection were discarded and replaced.

Oral administration

The fish-groups to be administered orally (p.o.) were also

allocated into small groups of six, from which the fish

were administered flumequine orally individually at a dose of

10 mg/kg through a stomach tube (Martinsen et al., 1993),

after being lightly sedated with metomidate (10 mg/L water)

before administration.

Bath administrations

The bath administrations were carried out in 1000 L fibreglass

tanks with 100 L static aerated seawater. Flumequine from the

stock solution was diluted with 1 M NaOH and added to the

seawater to a final concentration of 10 mg/L. The fish were kept

in the flumequine bath solution for 2 h, and then transferred to

flow-through water tanks.

Sampling

Blood samples from six fish were collected at 1, 3, 7, 12, 18, 24,

48, 96, 168 and 288 h post administration. In the group given

bath-treatment, the first bloodsample was collected at 0 h post

treatment. The fish were anaesthetized with metomidate (10 mg/

L water) before blood sampling by caudal venipuncture using a

0.5 6 25 mm needle and 1 mL syringe. The blood was sampled

caudal to the injection site in the intravenously administered

group. Each sample consisted of 100 mL blood. The samples were

frozen at ±808C until analysed.

No mortalities were recorded in the experimental fish during

or 1 month after the study.

Analytical procedures

The plasma samples were cleaned by solid-phase extraction on a

column of the Bond Elute type, size 1 mL, with C2 sorbent

material, according to a previously published method (Rasmus-

sen et al., 1989). The concentrations of flumequine in plasma

were determined by means of high-performance liquid chroma-

tography using a fluorescence detector operated at an excitation

wavelength of 325 nm and emission wavelength of 360 nm.

Oxolinic acid was added before cleanup/extraction, and used as

internal standard.

The lower limits of quantitation of the method were 10 mg/L,and it was linear over a tested range of 100±4000 mg/L. Thelinear correlation coefficient was 0.9999. The linearity of the

calibration curve was also tested on a residual plot, revealing no

bias. The recovery of flumequine was from 99% (4000 mg/L) to105% (400 mg/L). Recovery of 105% is probably the result of

some evaporation of organic solvent during elution with vacuum.

Chromatography

The HPLC system used consisted of a Perkin Elmer (Norwalk, CT)

LC 250 pump connected to a Waters (Milford, MA) wisp 710 B

autoinjector and a Perkin Elmer LC 240 fluorescence detector. A

150 6 4.6 mm 5 mm PLRP-S analytic column with a 5.0 6 3.0

mm 36 mm PLRP-S precolumn was used. The integrator was the

Analytic workstation, Omega-2, V2.60, Perkin Elmer. The

system was operated at room temperature (228C) with mobile

phase containing 0.001 M H3PO4: tetrahydrofuran: acetonitrile

(13 : 3 : 4). The flow was 0.7 mL/min.

Pharmacokinetic analysis

Pharmacokinetic evaluation was performed using the computer

program WIN-NONLIN, version 1.1 (Statistical consultants Inc.,

Lexington, KY), in a least square nonlinear regression analysis.

Standard pharmacokinetic parameters were calculated accord-

ing to a noncompartment model. In the i.v. group, the intercept

with the y-axis was calculated by back-extrapolation of the

curve, using the first two data points. In all groups, the curve

was extrapolated to infinity using the lambda-z calculated from

the depletion data, using the algorithm of Dunne (1985).

#1999 Blackwell Science Ltd, J. vet. Pharmacol. Therap. 22, 122±126

Pharmacokinetics of flumequine in halibut and turbot 123

Paper 191 Disc

The bioavailabilities of the oral preparations were calculated

comparing the area under the concentration time curve

(AUC)i.v.(0-?) and AUCp.o.(0-?) or AUCbath(0-?), extrapolated to

infinity using the algorithm of Dunne (1985).

RESULTS

The estimated elimination half-life (t�lz) was 34 h (halibut) and

32 h (turbot). The observed volume of distribution (Vss) was 2.99

L/kg (halibut) and 3.75 L/kg (turbot), and the total plasma

clearance (ClT) was, respectively, 0.12 L/h.kg and 0.17 L/h.kg

for halibut and turbot.

Oral administration gave maximal plasma concentrations

(Cmax) of 1.4 mg/L (halibut) and 1.9 mg/L (turbot). The time-

point when maximal plasma concentrations were observed

(tmax) was 7 h post administration. Elimination half-life after oral

administration was calculated to be 43 h (halibut) and 42 h

(turbot). Bioavailability (F) was 56% (halibut) and 59% (turbot).

After bath-administration, Cmax of 0.083 mg/L in halibut and

0.134 mg/L in turbot was detected at 0 h (halibut) and 3 h

(turbot) post treatment (= tmax). The pharmacokinetic para-

meters are listed in Table 1, and the plasma concentration vs.

time of flumequine for halibut is shown in Fig. 1 and for turbot in

Fig. 2.

DISCUSSION

In pharmacokinetic studies in humans and domestic animals,

each individual is normally blood sampled throughout the whole

study period. This experimental protocol is very difficult to use in

studies with fish. In the current study, only small experimental

fish were available (halibut 89 g, turbot 210 g). Frequent blood-

sampling of each fish was considered impossible. This is the

reason why we chose a design in which each fish was blood-

sampled only twice. The small size of the fish also made it difficult

and time consuming to perform intravenous administration.

Anaesthesia was necessary to obtain a good result, when

sampling blood and administering drug by the intravenous

route. The dose of 10 mg/kg for intravenous administration was

chosen by extrapolation of data from pharmacokinetic studies of

flumequine in Atlantic salmon (Rogstad et al., 1993; Elema et al.,

1995). The same dose (and concentration of formulation) was

chosen for the orally administered fish as for the i.v.

administered fish to ensure the comparison between the different

methods of administration was as accurate as possible, and also

for practical reasons regarding preparation and administration.

The amount of cod liver oil used was adjusted to make the

suspension suitable for administration through a stomach tube.

Fish administered flumequine orally were only lightly sedated to

reduce the risk of regurgitation.

The bath administrations were carried out with a dosage of 10

mg/L flumequine in the water. This is the same dosage as used in

the groups given i.v. and oral administration, but with the

amount of flumequine related to the amount of water instead of

bodyweight. The same dosage was chosen to ensure the

comparison between the different methods of administration

was as accurate as possible.

The short duration (2 h) of the bath administration was

chosen because bath administration of long duration, with static

aerated water, is difficult to carry out in a commercial

aquaculture setting.

A noncompartment model was used in the pharmacokinetic

calculations. Other ways of modelling the data were also tested.

The data from the i.v. administered group could be interpreted

using a 1, 2 or 3 compartmental model. Based on the minimal

Akaike's information criterion estimation (Yamaoka et al.,

1978), in which all data were weighted to produce the best

curve fit during the elimination phase, the model which

described the data best was a 3 compartment model. In the 3

compartment model the t�g was estimated to be 50 h in halibut

and 40 h in turbot. In turbot, the last data point (288 h) had to

be omitted for the dataset to be interpreted by the model. The

volume of distribution (Vss) was calculated to be 3.01 L/kg

(halibut) and 2.34 L/kg (turbot). Total body clearance was 0.11

L/h.kg (halibut) and 0.17 L/h.kg (turbot). As expected, t�g was

longer than t�lz. The distribution volume (Vss) and clearance

124 M. K. Hansen & T. E. Horsberg

#1999 Blackwell Science Ltd, J. vet. Pharmacol. Therap. 22, 122±126

Table 1. Pharmacokinetic parameters in

halibut and turbot held in running

seawater at 108C and 188C, respectively,following a single dose of 10 mg

flumequine/kg bodyweight intravenously

or orally, or 10 mg flumequine/L in a

seawater bath for 2 h. The parameters

were calculated by a noncompartmental

model

Pharmacokinetic i.v. administration p.o. administration bath administration

Parameter

Halibut Turbot Halibut Turbot Halibut Turbot

ÐÐÐÐÐÐÐÐÐ ÐÐÐÐÐÐÐÐÐ ÐÐÐÐÐÐÐÐÐ

AUC(0-?) (mg.h/L) 83.7 59.3 47.5 35.8 4.5 *

t�lz (h) 32 34 43 42

MRT (h) 25.1 22.2

Vss (L/kg) 2.99 3.75

Cl (L/h.kg) 0.12 0.17

Cmax (mg/L) 1.4 1.9 0.08 0.13

tmax (h) 7 7 0 3

F (%) 56 59 5 *

*(AUC) could not be extrapolated to infinity. AUC (0-?), area under plasma concentration time curve

extrapolated to infinity; t�lz, elimination half life during the elimination phase; MRT, mean residence

time; Vss, volume of distribution at steady state; Cl, plasma clearance; Cmax, maximum plasma

concentration; tmax, time of peak plasma concentration; F, bioavailability.

Paper 191 Disc

are nearly the same in both the noncompartment and the 3

compartment model.

There are few pharmacokinetic studies on halibut and turbot.

Data comparing different ways of administration of flumequine

(or other quinolones) in the two species are not published.

Our results from the i.v. administered group showed that

flumequine had quite similar pharmacokinetic properties in

halibut and turbot.

The elimination half-lives after i.v. administration were

estimated to be 32 and 34 h for halibut and turbot, respectively.

Distribution volume (Vss) was estimated to be 2.99 L/kg (halibut)

and 3.75 L/kg (turbot). This is a large distribution volume,

indicating that flumequine was well distributed to the tissues in

both halibut and turbot. Total body clearance was 0.12 L/h.kg

(halibut) and 0.17 L/h.kg (turbot). Similar results have been

reported for Atlantic salmon (Salmo salar). The studies on

Atlantic salmon were performed at 5±10 8C (Rogstad et al.,

1993; Elema et al., 1995; Martinsen & Horsberg, 1995).

Pharmacokinetic data from oral administration of flumequine

in halibut or turbot are not available. Cmax after oral adminis-

tration in this study was achieved in 7 h at 1.4 mg/L (halibut)

and 1.9 mg/L (turbot). This is a little higher than what has been

found in Atlantic salmon, where Rogstad et al. (1993), Elema et

al. (1995) and Martinsen & Horsberg (1995) reported a Cmax

of 1.42±2.46 mg/L after an oral dose of 25 mg/kg in sea water

(5±10 8C).The bioavailability was calculated to be 56% (halibut) and 59%

(turbot). This is higher than reported from the above mentioned

studies in Atlantic salmon, in which the bioavailability was

calculated to 44.7±46%. The elimination half-life in the orally

administered group was 43 h (halibut) and 42 h (turbot). This is,

as expected, longer than t�lz calculated after i.v. administration

owing to the influence of the absorption process.

The results of the analysis after bath administration showed

that flumequine was poorly absorbed from seawater. This is in

accordance with results obtained earlier, showing decreasing

flumequine absorption in Atlantic salmon with increasing pH

and Ca2+ in the water (O'Grady et al., 1988). Cmax after 2 h bath

treatment with 10 mg/L in our study was only 0.083 mg/L

(halibut) and 0.14 mg/L (turbot). Because of the poor absorp-

tion, flumequine concentrations in plasma very soon reached

values close to the detection limit, reducing accuracy. This is

probably the reason for the apparent small rise in flumequine

concentration from 96 h onwards seen in turbot, which is most

likely an artefact. This made it difficult to interpret the turbot

bath data using the pharmacokinetic model.

Pharmacokinetics of flumequine in halibut and turbot 125

#1999 Blackwell Science Ltd, J. vet. Pharmacol. Therap. 22, 122±126

Fig. 1. Mean (+ SEM) plasma concentration

profiles of flumequine in halibut (n=6) after a

single 10 mg/kg dose administered by the

intravenous (i.v.) and oral routes, and a single

10 mg/L dose administered as a bath.

Fig. 2. Mean (+ SEM) plasma concentration

profiles of flumequine in turbot (n=6) after a

single 10 mg/kg dose administered by the

intravenous (i.v.) and oral routes, and a single

10 mg/L dose administered as a bath.

Paper 191 Disc

Samuelsen & Lunestad (1996) studied bath treatment of

halibut with flumequine and oxolinic acid in 3±5 g halibut. They

found a t�lz of 10 h for halibut in muscle tissue after bath

administration of flumequine. This is much shorter than the t�lz

of 34 h in our study. Samuelsen & Lunestad (1996) used fish of

3±5 g whereas our fish had an average weight of 89 g. This, and

the different tissues sampled, are probably the most important

reasons for the differing results. Although flumequine shows a

poor degree of absorption, bath treatment of halibut and turbot is

possible as shown by Samuelsen & Lunestad (1996) and

Samuelsen (1997). To obtain adequate levels of flumequine in

the fish, a high dose has to be administered over a long period of

time. Samuelsen & Lunestad (1996) found a Cmax of 14.2 mg/L

in muscle after a 72 h bath in 150 mg/L flumequine.

Few studies of minimum inhibitory concentration (MIC) of

flumequine for susceptible strains of pathogenic bacteria for

halibut and turbot have been conducted. Samuelsen & Lunestad

(1996) reported MICs for flumequine of two pathogenic vibrios

isolated from halibut to be 0.06 mg/L (Vibrio anguillarum O1 (HI

11341) and 0.015 mg/L (V. anguillarum O2 (HI 11347)). MICs

of susceptible strains of bacteria pathogenic to Atlantic salmon

(Aeromonas salmonicida, Vibrio salmonicida, Vibrio anguillarum,

Yersinia ruckeri) have been reported to range from 0.005 to 0.5

mg/L (for most strains 5 0.1 mg/L) (Barnes et al., 1990;

Martinsen et al., 1992).

V. anguillarum isolated from turbot and atypical A. salmonicida

isolated from halibut have shown large inhibition zones against

flumequine in plate-agar diffusion tests. It is therefore likely that

MICs for these bacteria are in the same range as the bacteria

pathogenic to Atlantic salmon.

Blaser et al. (1987) reported bacterial regrowth in vitro unless

the peak concentration/MIC ratio exceeded 8 : 1 for the

fluoroquinolone enoxacin. If using the same ratio evaluating

the plasma concentrations of flumequine in our study, after oral

administration of 10 mg/kg flumequine, both halibut and turbot

exceed this peak concentration/MIC ratio for most susceptible

strains. The peak concentration and favourable bioavailability

(56% halibut, 59% turbot) makes this flumequine suitable for

oral treatment against bacterial infections in halibut and turbot.

The plasma concentrations from bath administration in 10 mg/L

flumequine for 2 h did not exceed the MICs. Both the dose and

the length of time in the bath must be significantly increased to

achieve a successful bath treatment (peak concentration/MIC

ratio 4 8 for a sufficient time).

This study was performed as a single-dose pharmacokinetic

study, and the results show that flumequine is not suitable for

single-dose administration in the treatment of bacterial diseases.

Although plasma concentrations higher than 8 6 MIC may be

obtained, the flumequine level in plasma will fall below this

concentration too soon for adequate treatment. In order to obtain a

high bioavailability and to minimize the pollution of the environ-

ment with flumequine, the oral route of administration should be

chosen when treating halibut and turbot, whenever possible.

REFERENCES

Barnes, A.C., Lewin, C.S., Hastings, T.S. & Amyes, S.G.B. (1990) In vitro

activities of 4-Quinolones against the fish pathogen Aeromonas

salmonicida. Antimicrobial Agents and Chemotherapy, 34, 1819±1820.

Blaser, J., Stone, B.B., Groner, M.C. & Zinner, S.H. (1987) Comparative

study with enoxacin and netilmicin in a pharmacodynamic model to

determine importance of ratio of antibiotic peak concentration to MIC

for bactericidal activity and emergence of resistance. Antimicrobial

Agents and Chemotherapy, 5, 1059±1064.

Dunne, A. (1985) JANA: a new iterative polyexponential curve strip-

ping program. Computerized Methods and Programs in Biomedicine, 20,

269±275.

Elema, M.O., Hoff, K.A. & Kristensen, H.G. (1995) Bioavailability of

flumequine after oral administration to Atlantic salmon (Salmo salar L).

Aquaculture, 136, 209±219.

Martinsen, B. & Horsberg, T.E. (1995) Comparative single-dose

pharmacokinetics of 4 quinolones, oxolinic acid, flumequine, sara-

floxacin, and enrofloxacin, in Atlantic salmon (Salmo salar) held in

seawater at 10 degrees C. Antimicrobial Agents and Chemotherapy, 39,

1059±1064.

Martinsen, B., Horsberg, T.E., Sohlberg, S. & Burke, M. (1993) Single-

dose kinetic study of sarafloxacin after intravenous and oral

administration of different formulations to Atlantic salmon (Salmo

salar) held in sea-water at 8.5 degrees C. Aquaculture, 118, 37±47.

Martinsen, B., Oppegaard, H., Wichstrom, R. & Myhr, E. (1992)

Temperature-dependent in vitro antimicrobial activity of 4, 4-

quinolones and oxytetracycline against bacteria pathogenic to fish.

Antimicrobial Agents and Chemotherapy, 36, 1738±1743.

O'Grady, P., Moloney, M. & Smith, P.R. (1988) Bath administration of

the quinoline antibiotic flumequine to brown trout Salmo trutta and

Atlantic salmon S. Salar. Diseases of Aquatic Organisms, 4, 27±33.

Rasmussen, K.E., Tennesen, F., Thanh, H.H., Rogstad, A. & Aanesrud, A.

(1989) Solid-phase extraction and high-performance liquid chromato-

graphic determination of flumequine and oxolinic acid in salmon plasma.

Journal of Chromatography Biomedical Applications, 496, 355±364.

Rogstad, A., Ellingsen, O.F. & Syvertsen, C. (1993) Pharmacokinetics

and bioavailability of flumequine and oxolinic acid after various routes

of administration to Atlantic salmon in seawater. Aquaculture, 110,

207±220.

Samuelsen, O.B. (1997) Efficacy of bath-administered flumequine and

oxolinic acid in the treatment of vibriosis in small Atlantic halibut.

Journal of Aquatic Animal Health, 9, 127±131.

Samuelsen, O.B. & Lunestad, B.T. (1996) Bath treatment, an alternative

method for the administration of the quinolones flumequine and

oxolinic acid to halibut Hippoglossus hippoglossus, and in vitro

antibacterial activity of the drugs against some Vibrio sp. Diseases of

Aquatic Organisms, 27, 13±18.

Yamaoka, K., Nakagawa, T. & Uno, T. (1978) Application of Akaike's

information criterion (AIC) in the evaluation of linear pharmacokinetic

equations. Journal of Pharmacokinetics and Biopharmaceutics, 6, 165±175.

126 M. K. Hansen & T. E. Horsberg

#1999 Blackwell Science Ltd, J. vet. Pharmacol. Therap. 22, 122±126

Paper 191 Disc