Concurrent Determination of Four Fluoroquinolones in Catfish, Shrimp, and Salmon by Liquid Chromatography with Fluorescence Detection

Abstract

A liquid chromatographic (LC) method with fluorescence detection was developed for concurrent determination of 4 fluoroquinolones: ciprofloxacin (CIPRO), enrofloxacin (ENRO), sarafloxacin (SARA), and difloxacin (DIFLX) in catfish, shrimp, and salmon. The procedure consists of extraction from fish tissue with acidified ethanol, isolation and retention on a cation exchange solid-phase extraction column, elution with basic methanol, and LC analysis with fluorescence detection. LC is performed by isocratic elution with acetonitrile-2% acetic acid (16 + 84) mobile phase, and a PLRP-S polymer column with fluorescence detection, excitation 278 nm and emission 450 nm. A target level of 20 ppb for each of the 4 fluoroquinolones has been established for this method. Fortified and incurred fish sample results are based on a 5-point standard curve calculation (10-160 ppb). Overall percent recoveries (% relative standard deviation) from fortified catfish were 78 (10), 80 (11), 70 (9.4), and 78 (10); from fortified shrimp, 69 (5.9), 85 (4.9), 79 (5.9), and 90 (4.5); and from fortified salmon, 56 (15), 93 (5.6), 61 (11), and 87 (5.0) for CIPRO, ENRO, SARA, and DIFLX, respectively. Data from the analysis of fluoroquinolone-incurred catfish, shrimp, and salmon are presented.

Full text

DRUGS, COSMETICS, FORENSIC SCIENCES
Concurrent Determination of Four Fluoroquinolones in Catfish,
Shrimp, and Salmon by Liquid Chromatography with
Fluorescence Detection
JOSÉ E. ROYBAL,CALVIN C. WALKER,ALLEN P. PFENNING,SHERRI B. TURNIPSEED,JOSEPH M. STOREY,STEVE A. GONZALES,
and JEFFREY A. HURLBUT
U.S. Food and Drug Administration, Animal Drugs Research Center, Denver Federal Center, PO Box 25087, Denver, CO
80225-0087
A liquid chromatographic (LC) method with fluo-
rescence detection was developed for concurrent
determination of 4 fluoroquinolones: ciprofloxacin
(CIPRO), enrofloxacin (ENRO), sarafloxacin
(SARA), and difloxacin (DIFLX) in catfish, shrimp,
and salmon. The procedure consists of extraction
from fish tissue with acidified ethanol, isolation
and retention on a cation exchange solid-phase ex-
traction column, elution with basic methanol, and
LC analysis with fluorescence detection. LC is per-
formed by isocratic elution with acetonitrile–2%
acetic acid (16 + 84) mobile phase, and a PLRP-S
polymer column with fluorescence detection, exci-
tation 278 nm and emission 450 nm. A target level
of 20 ppb for each of the 4 fluoroquinolones has
been established for this method. Fortified and in-
curred fish sample results are based on a 5-point
standard curve calculation (10–160 ppb). Overall
percent recoveries (%relative standard deviation)
from fortified catfish were 78 (10), 80 (11), 70 (9.4),
and 78 (10); from fortified shrimp, 69 (5.9), 85 (4.9),
79 (5.9), and 90 (4.5); and from fortified salmon,
56 (15), 93 (5.6), 61 (11), and 87 (5.0) for CIPRO,
ENRO, SARA, and DIFLX, respectively. Data from
the analysis of fluoroquinolone-incurred catfish,
shrimp, and salmon are presented.
Fluoroquinolone (FQ) antibacterials have been very ef-
fective in combating various diseases in animal hus-
bandry and aquaculture. The addition of either fluorine
or a piperazino moiety, or both, to the basic quinolone back-
bone enhances overall antibacterial activity of the molecule.
The fluorine increases the activity against Gram-positive mi-
crobes (i.e., Clostridium, Staphylococcus, Streptococcus),
while the piperazino improves its effectiveness against
Gram-negative organisms (i.e., Escherichia coli, Pseudomo-
nas aeruginosa, Salmonella enteritidis; 1–3). These modifica-
tions make FQs very attractive for a vast number of maladies.
Several FQs are now available, and many more are being de-
veloped for treatment of gastrointestinal and respiratory infec-
tions. Of these drugs, only 2 are currently approved in the
United States. Sarafloxacin (SARA) was approved in August
1995fortreatingpoultry(chickensandturkeys)againstE.coli
infections (4, 5). Enrofloxacin (ENRO) was approved in Oc-
tober 1996 to control mortality in chickens from E. coli infec-
tions, and in turkey for infections caused by Pasteurella
multocida (6). In Europe, ENRO is most commonly used to
treat infections in cattle (7). The major metabolite of ENRO is
reported to be ciprofloxacin (CIPRO), its de-ethylated prod-
uct (8–10). Difloxacin (DIFLX) is similar in structure to
SARA, differing only by a methyl group in the
7-(4-piperazinyl) position. It was reported to be superior in ac-
tivity against Gram-negative and -positive microbes com-
pared with norfloxacin (11). When evaluated against
Edwardsiella ictaluri and Aeromonas sobria, SARA and
DIFLX manifested lower minimum inhibitory concentration
(MIC) than did nalidixic acid, Romet-30
(ormetoprin/sulfadimethoxine), Terramycin (oxytetracy-
cline), ampicillin, spectinomycin, and doxycycline (12).
The work presented here concentrated on CIPRO, ENRO,
SARA, and DIFLX (Figure 1) because of their availability and
very effective broad-spectrum activity against many microbes.
Although none of the FQs have been approved in the United
States for use as aquaculture therapeutic agents, the potential for
their extra-label use is of concern. The interest indicated by the
aquaculture industry in these drugs and potential for the emer-
gence of drug-resistant bacteria through their use has created a
needforanalyticalmethodstomonitorforresiduesofthesedrugs
in both domestic and imported aquaculture products.
A literature search revealed few methods for the determi-
nation of these 4 fluoroquinolones (4FQs): CIPRO, ENRO,
SARA, and DIFLX. A liquid chromatographic (LC) method
was reported for determination of SARA residues in cat-
fish (13). The procedure, however, was for a single analyte
and had a long LC run time, which was unsuitable for regula-
tory purposes. Several LC methods for FQ residues in other
matrixes such as bovine, porcine, and poultry tissues as well
methods for plasma, urine, and eggs were reported (14–20).
These procedures used either ion-pairing agents, or a basic
solvent extraction and/or strong cation exchange solid-phase
ROYBAL ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002 1293
Received December 21, 2001. Accepted by JM May 10, 2002.
Corresponding author’s e-mail: jroybal@ora.fda.gov.

extraction (SPE) columns for isolation and cleanup. When
applied to fish tissue with multianalyte residues, these meth-
ods produced inadequate recovery and cleanup for our in-
tended needs. Recently, we developed 2 methods for the de-
termination of these 4FQs in milk (21) and catfish (22) by a
simple extraction procedure with LC–fluorescence detec-
tion. Our goal was to apply these techniques and to develop a
method that could analyze residues of these 4FQs in
aquaculture products, concurrently, in a timely manner suit-
able to regulatory application.
METHOD
Apparatus
(a)Liquid chromatograph.—Hewlett-Packard Model
HP1090 with HP Vectra 486 HP Chem Station
(Hewlett-Packard, Avondale, PA). Operating conditions: mo-
bile phase flow, 0.9 mL/min; column temperature, 60°C; col-
umn pressure, 2300–2600 psi; volume injected, 50 µL.
(b)Detector.—Fluorescence programmable detector,
Model HP1046A, excitation (ex) 278 nm and emission (em)
450 nm with 418 nm cut-off filter (Hewlett-Packard).
(c)LC column.—PLRP-S polymer, 5 µm, 100Å, 250 ×
4.6 mm id (P/N 1512-5500 and S/N 5 µm-PRS1-62B-89).
With guard column consisting of cartridge holder
(P/N 1310-0016) and PLRP-S cartridge (P/N 1612-1801) of
same packing (Polymer Laboratories, Amherst, MA) or
equivalent.
(d)Blender.—5-speed, pulsed Oster Model 54841
(BaxterScientificProducts,McGrawPark,IL)orequivalent.
(e)Homogenizer.—Tissuemizer, Model SDT1810 with
ModelSDT-18ENProbe(Tekmar,Cincinnati,OH)orequivalent.
(f)Food grinder.—Hobart, consisting of No. 12
Brite-metal chopping end (P/N 119-860-3), No. 12 stay-sharp
blade(P/N 290-339), 0.25 in. stay-sharp plates (P/N 16425-2),
and No. 12 stainless steel feed pan (P/N 120903) with plastic
feedstomper(P/NA-119922-1;HobartCorp.,Denver,CO)or
equivalent.
(g)Pipettors.—(1) Adjustable, 5 mL pipet, Cat.
No. 851350, with disposable polypropylene macrotips, 5 mL
capacity, Cat. No. 851357; (2) Calibra®digital micropipet,
10–100 µL capacity, Cat. No. 851164, with disposable poly-
propylene microtips, Cat. No. 851271; and (3) Calibra digital
micropipet,100–1000 µLcapacity,Cat.No.851168,withdis-
posable polypropylene microtips, Cat. No. 851276 (Wheaton
Science Products, Millville, NJ) or equivalent.
(h)Propylsulfonic acid (PRS)-SPE columns.—Disposable,
500 mg, BondElut LRC 10 cc (P/N 1211-3038; Varian Associ-
ates, Harbor City, CA). Do not substitute PRS-SPE column.
(i)Reservoir.—75 mL Polypropylene reservoir with
20 µm frit (P/N 1213-1018; Varian).
(j)Column connection adaptors.—12,20 mL adaptor for
PRS-SPE, BondElut LRC extraction columns (h)
(P/N 1213-1003; Varian).
(k)Centrifuge.—IEC Model PR-7000M, refrigerated,
with temperature set at 4°C, with rotor No. 825A for 50 mL
centrifuge tubes and/or rotor No. 259 for 150 mL centrifuge
tubes (International Equipment Co., Needham Heights, MA)
or equivalent.
(l)Centrifuge tubes.—50 and 150 mL, Falcon Blue Max,
disposable, conical, graduated, polypropylene with cap (Cat.
Nos. 2070 and 2076, respectively; Becton Dickinson, Lincoln
Park, NJ) or equivalent.
(m)Test tube.—Disposable, 13 ×100 mm borosilicate
glass, culture tube (P/N 73500; Kimble Products, Vineland,
NJ) or equivalent.
(n)Nitrogen evaporator.—Twelve-position 50–55°C wa-
ter bath (P/N 11155; Organomation Associates, Inc., Berlin,
MA) or equivalent.
(o)Syringes.—Disposable plastic, latex-free, 1 mL (Cat.
No. 309602; Becton Dickinson).
(p)Pasteur pipet.—Disposable, glass, 5.75 in.
(q)Nylon syringe filter.—Whatman, GD/X disposable,
13 mm, 0.45 µm, nylon filter media with glass filter prefilter
in polypropylene housing (Cat. No. 6870-1304; Whatman,
Inc., Clifton, NJ) or equivalent.
1294 ROYBAL ET AL.:JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002
Figure 1. Structures of the 4 fluoroquinolones.

Reagents
(a)Solvents.—Distilled-in-glass, LC and UV spectrograde
methanol and acetonitrile (Burdick & Jackson Laboratories,
Inc., Muskegon, MI) or equivalent.
(b)Water.—Deionized, purified to 18.2 MΩ⋅cm
(mega-ohms) using Milli-Q Plus water-purification system
(Cat. No. ZD5211584; Millipore, Bedford, MA).
(c)Acetic acid.—ACS grade, glacial, aldehyde-free. Use
to prepare 1 and 2% aqueous solutions.
(d)Ammonium hydroxide, NH4OH.—30%, Baker
Instra-Analyzed Reagent (Cat. No. 9733-01; J.T. Baker, Inc.,
Phillipsburg, NJ).
(e)Absolute ethanol.—Ethyl alcohol, 200 proof, dehy-
drated alcohol, USP, Punctilious (Quantum Chemical Corp.,
Tuscola, IL).
(f)Mobilephase.—Acetonitrile–2%aceticacid(16+84).
(g)Extracting solution.—Absolute ethanol–water–acetic
acid (98+1+1).
(h)SPE equilibration solution.—Extracting solution (g)+
1% acetic acid (35 + 20).
(i)Eluting solution.—NH4OH–methanol (1 + 3).
(j)Reference standards.—(1)Ciprofloxacin
HCl.—858 µg ciprofloxacin/mg, Lot No. Pt238870K, gra-
ciously provided by Bayer AG (Leverkusen, Germany).
(2)Enrofloxacin.—99.0%, Std No. 46.03, Lot No. R-177-2,
graciously provided by Miles Agriculture Division (Shawnee
Mission, KS). (3)Sarafloxacin HCl.—88.5%, Lot
No. 23-336-CE, graciously provided by Abbott Laboratories
(Chemical and Agriculture Products Division, North Chicago,
IL). (4)Difloxacin HCl—90.2%, Lot No. 36-776-CE, gra-
ciously provided by Abbott Laboratories.
(k)Standard solutions.—(1)Stock standards,
200 mg/mL.—Accurately weigh amount of each of the FQ
reference standards (CIPRO, ENRO, SARA, and DIFLX)
equivalent to 10.0 ± 0.5 mg (as free base after correcting for
purity) into individual 50 mL volumetric flasks, dissolve in
25 mL methanol, sonicate for 5 min, and dilute to volume with
methanol. Store in refrigerator. Properly sealed and stoppered,
these solutions are stable 6 months. (2)LC mixed working
standard, 2000 ng/mL.—Combine 1.0 mL aliquot of each of
the above stock standards into 100 mL volumetric flask and
dilute to volume with mobile phase. Store in refrigerator. Sta-
ble for at least 3 months. (3)LC calibration stan-
dards.—Aliquot 100, 200, 400, 800, and 1600 µL, respec-
tively, of LC mixed working standard into 5 separate 10 mL
volumetric flasks, and bring to volume with mobile phase.
This provides LC calibration standards in the concentration
range of 20–320 ppb (ng/mL) for each FQ. These concentra-
tions are equivalent to 2 g extracts from tissues containing
concentrations of 10–160 ppb (ng/g). Prepare daily with each
assay set and use to generate the 5-point standard curve.
(4)4FQ fortification standards.—Solution A, 8000 ng/mL,
standardmix.—Combine2.0mLaliquotofeachofthe4stock
standard solutions into 50 mL volumetric flask, and dilute to
volume with methanol. Solution B, 4000 ng/mL, standard
mix.—Aliquot 25.0 mL solution A to 50 mL volumetric flask,
and dilute to volume with methanol. Solution C, 2000 ng/mL,
standard mix.—Aliquot 25.0 mL solution B to 50 mL volu-
metric flask, and dilute to volume with methanol. Solution D,
1000 ng/mL, standard mix.—Aliquot 25.0 mL solution C to
50 mL volumetric flask, and dilute to volume with methanol.
Solution E, 500 ng/mL, standard mix.—Aliquot 25.0 mL solu-
tion D to 50 mL volumetric flask, and dilute to volume with
methanol. Aliquot 40 µL spiking solutions A–E to separate
2.0 g portions of tissue to yield tissue fortification levels of
160, 80, 40, 20, and 10 ppb. Store in refrigerator (<4°C). Sta-
ble for at least 3 months.
Sample Preparation
Immediately freeze all samples (catfish, shrimp, and
salmon) after collection for shipment. Upon receipt at analyz-
ing laboratory, place samples in freezer (–10 to –20°C) until
analysis.
Catfish (Channel, Ictalurus punctatus)
Thaw frozen catfish (muscle only, skin removed) at room
temperature until semi-frozen. Cut catfish fillets into 1–2 in.
cubes. Homogenize tissue in a blender, using pulse action.
Scrape tissue from walls and blades of blender with a labora-
ROYBAL ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002 1295
Table 1. Linear regression data (
y
=m
x
+b)and
square of the correlation coefficients (r2) for typical
fluoroquinolone (FQ) fluorescence calibration standard
curve
FQ m
a
(slope) b (
y
-intercept) r2
CIPRO 1.59 –0.63 0.9999
ENRO 2.67 –1.65 0.9999
SARA 1.03 –2.60 0.9999
DIFLX 1.95 –3.76 0.9999
a
Area counts per ppb.
Table 2. Average recovery (%) of 4 fluoroquinolones
from fortified catfish
Fortification level, ppb
a
Average recovery, %
b
CIPRO ENRO SARA DIFLX
10 86 80 62 71
20 77 79 70 73
40 77 79 60 82
80 78 81 78 81
160 75 81 80 82
Overall recovery, %(
n
= 25) 78 80 70 78
Overall RSD, %(
n
= 25) 11 11 15 12
a
Five replicates of mixed FQ each level.
b
Average of 5 individual analysis determinations (
n
= 5).

tory spatula, and blend again. Repeat once more or until entire
sample has texture of thick paste and appears uniform through-
out. Place homogenate in Whirl-Pak®bags for storage, and
store in freezer (–10 to –20°C).
Shrimp (Penaeid, Penaeus vannamei)
Place whole frozen shrimp samples in cold water to thaw.
When the shrimp are limber, remove heads, chitinous shell, and
body appendages. Place shrimp meat in blender, and homoge-
nize with a pulse action. Scrape tissue from walls and blades of
blender with laboratory spatula, and blend until contents appear
uniform. Place homogenate in Whirl-Pak bags for storage, and
store in freezer (–10 to –20°C).
Salmon (Atlantic, Salmo salar)
Eviscerate, de-scale, and discard the head, tail, and fins re-
moved from the salmon samples. Cut the main torso, with skin,
into1in.thick steaks, and then split the steaks in half. Grind and
homogenizethesteaksintheHobartfoodgrinder.Grindtheho-
mogenate once more by passing it through the food grinder.
Place the homogenate in Whirl-Pak bags for storage, and store
in freezer (–10 to –20°C).
Fortification
For recovery determinations, aliquot 40 µL 4FQ fortification
standard solutions A–E to separate 2.0 g portions of tissue to
yield fortification levels of 160, 80, 40, 20, and 10 ppb, respec-
tively, of each FQ: CIPRO, ENRO, SARA, and DIFLX. For un-
known samples, analyze one control and one sample fortified at
20 ppb with each set.
Extraction
Accurately weigh 2.0 ± 0.2 g blended sample tissue into
50 mL polypropylene conical tube. Fortify control sample tis-
sue by adding 40 µL 4FQ fortification standard solution D to
surface of tissue. Wait at least 5 min; then add 18 mL extracting
solution, and homogenize with tissuemizer at high speed for
20 s. Rinse sides of probe with 3–4 mL ethanol in the 50 mL
centrifuge tube. Cap the tube and centrifuge at 3000 rpm
(1870 rcf) for 5 min at 4°C. Decant supernatant into 150 mL
centrifuge tube. Repeat extraction by adding another 18 mL ex-
tracting solution to the tube containing the sample pellet. Cap
the tube and mix on a Vortex mixer vigorously, to break up and
mix pellet, for 20 s. Centrifuge capped tube at 3000 rpm
(1870 rcf) for 5 min at 4°C. Decant supernatant into 150 mL
centrifuge tube containing the first supernatant. Add 20 mL 1%
acetic acid to combined extracts; then cap and mix by swirling.
Centrifuge at 3000 rpm (2420 rcf) for 5 min at 4°C.
Prepare PRS-SPE column by placing it on a vacuum mani-
fold and conditioning it with 2 mL MeOH, 4 mL equilibration
solution with the full vacuum on. Stop the flow, leave
15–20 mm equilibration solution above the PRS-SPE column
bed. Do not allow column to dry. Using BondElut adaptor, at-
tach 75 mL reservoir with 20 µm pore frit to PRS-SPE column
on vacuum manifold. Decant entire sample extract contents
from 150 mL centrifuge tube into 75 mL reservoir attached to
PRS-SPE column. With the aid of full vacuum, pass entire
75 mL sample extract through the PRS-SPE column at
1–2 drops/s. After the entire extract has passed through the
PRS-SPE column (this usually takes ca 45–60 min), discon-
nect the reservoir. Sequentially wash PRS-SPE column with
2 mL MeOH, 5 mL water, and 2 mL MeOH. Remove excess
MeOH by vacuum aspiration. Aspirate for 30 s after last
MeOH wash has just entered the column bed. Elute FQs from
PRS-SPE column with 2.5 mL NH4OH–MEOH (1 + 3) into
disposable test tube. Evaporate to dryness by using nitrogen
flow in water bath at 50–55°C. Dissolve remaining residue in
1.0 mL mobile phase, and mix on a Vortex mixer 20 s. Using
Pasteur pipet, transfer reconstituted sample to 1 mL dispos-
able syringe with nylon syringe filter, and filter into LC vial
for LC analysis.
Liquid Chromatography
All LC injections are 50 µL in volume. Inject mobile phase
first, then all 5 mixed LC calibration standards solutions A–E,
and then the set of sample extracts. For samples containing
1296 ROYBAL ET AL.:JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002
Table 3. Average recovery (%) of 4 fluoroquinolones
from fortified shrimp
Fortification level, ppb
a
Average recovery, %
b
CIPRO ENRO SARA DIFLX
10 70 85 82 90
20 66 81 76 88
40 69 85 79 91
80 71 88 80 92
160 70 85 79 91
Overall recovery, %(
n
= 25) 69 85 79 90
Overall RSD, %(
n
= 25) 5.9 4.9 5.9 4.5
a
Five replicates of mixed FQ each level.
b
Average of 5 individual analysis determinations (
n
= 5).
Table 4. Average recovery (%) of 4 fluoroquinolones
from fortified salmon
Fortification level, ppb
a
Average recovery, %
b
CIPRO ENRO SARA DIFLX
10 45 90 61 86
20 50 91 57 85
40 56 92 61 86
80 60 94 60 86
160 67 98 68 91
Overall recovery, %(
n
= 25) 56 93 61 87
Overall RSD, %(
n
= 25) 15 5.6 11 5.0
a
Five replicates of mixed FQ each level.
b
Average of 5 individual analysis determinations (
n
= 5).

ROYBAL ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002 1297
Table 5. Fluoroquinolone (FQ) found in incurred catfish
Catfish FQ dosed
a
Days post-dosing CIPRO, ppb ENRO, ppb SARA, ppb DIFLX, ppb
1A ENRO 6 17 270
b
00
1B ENRO 6 <10 (LOQ = 10) 41, 52, 44
c
00
2 ENRO 6 29 340 (est.)
d
00
3 ENRO 6 34 404 (est.)
d
00
4A ENRO 6 23 280
b
00
4B ENRO 6 <10 (LOQ = 10) 52, 31, 54
c
00
5 SARA 6 0 0 62 0
6 SARA 6 0 0 15, 17
e
0
7 SARA 6 0 0 32, 29
e
0
8 SARA 6 0 0 17, 15
e
0
9 SARA 6 0 0 <10 (LOQ = 10) 0
10 DIFLX 10 0 0 0 29, 28
e
11 DIFLX 10 0 0 0 29, 26
e
12 DIFLX 10 0 0 0 30, 38
e
a
Single oral dose=5mgdrug/kg body weight.
b
Value outside LC calibration curve. Sample re-analyzed,
see
footnote
c
.
c
3/0.4 g portions of incurred catfish tissue blended with 1.6 g control catfish tissue (1 + 5 dilution) and re-analyzed.
d
Value outside LC calibration curve. Sample not re-analyzed.
e
Analyzed in duplicate.
Table 6. Fluoroquinolone (FQ) found in incurred shrimp
Shrimp sample FQ/dose level, ppm Time post-dosing, h CIPRO, ppb ENRO, ppb SARA, ppb DIFLX, ppb
1 ENRO/0.2 8 0000
2 ENRO/0.2 33 0000
3 ENRO/2.0 9 0 2.7, 2.7
a
00
4 ENRO/2.0 33 0000
5 ENRO/10 9.5 0 6.7, 6.6
a
00
6 ENRO/10 34 0 3.2, 3.2
a
00
7 SARA/0.2 2 0000
8 SARA/0.2 14 0000
9 SARA/2.0 2 0000
10 SARA/2.0 15 0000
11 SARA/10 2 0 0 6.5, 6.1
a
0
12 SARA/10 16 0000
13 DIFLX/0.2 9 0000
14 DIFLX/0.2 33 0000
15 DIFLX/2.0 10 0000
16 DIFLX/2.0 33 0000
17 DIFLX/10 11 0 0 0 8.0, 8.1
a
18 DIFLX/10 35 0000
a
Value outside (below) LC calibration curve. Value reported is estimate only.

>160 ppb, dilute sample extracts with mobile phase, so that
the peak response of the analyte lies between the highest and
lowest points of the standard curve. Follow with an injection
of the 40 ppb standard to verify instrument performance. If
peak response for 40 ppb standard has varied by >10% of the
initial peak response or retention time, wash the chromato-
graphic column for 30 min with a mixture of acetonitrile–6%
acetic acid (1 + 1), equilibrate the column with mobile phase
for 10 min, and re-inject all of the samples and LC calibration
standards solutions. The square of the correlation coefficient
(r2) of the standard curve should be >0.995. Calculations for
quantitation are based on this standard curve.
Results and Discussion
Dosing of catfish, shrimp, and salmon was initiated for the
purpose of generating drug-incurred residues for method trial
validation study and was not intended for depletion studies.
Data from a typical calibration standard curve are reported
in Table 1. Tables 2–4 are recoveries of the 4FQs from each
matrix: catfish, shrimp, and salmon, respectively. The results
demonstrate the performance of the developed method in all
3 matrixes. Recoveries of CIPRO and SARA, the 2 metabo-
lites, were slightly lower in salmon; however, the overall rela-
tive standard deviations (RSDs) for all 4FQs in all matrixes
were <15%, showing good precision at levels of 160 ppb or
less. All analyses of incurred samples included a control and a
fortified sample with each set run.
Channel catfish were administered a single oral dose of
5 mg drug/kg body weight of either ENRO, SARA, or DIFLX
to produce muscle samples containing incurred residues over
the range of 10–160 ng/g. Catfish dosed with ENRO and
SARA were sacrificed at 6 days post-dosing, and those dosed
with DIFLX were sacrificed at 10 days post-dosing.
The primary residue extracted and analyzed by this method
for incurred catfish muscle was the parent drug for ENRO,
SARA, and DIFLX. Although CIPRO is used mainly in hu-
man medicine, it was included in the study because it is a
majormetaboliteofENROinseveralspeciesandcouldpoten-
tially be used in aquaculture. Additionally, CIPRO may be
present in veterinary proprietary ENRO products used in for-
eign countries and could result in CIPRO residues in imported
foods of animal origin (23). However, the chromatograms of
the ENRO-incurred catfish muscle extracts exhibited only a
small peak at the retention time of CIPRO. SARA is reported
as the N-desmethyl metabolite of DIFLX in humans (15).
However,theonlyresidue detected in DIFLX-incurred catfish
muscle extracts was the parent DIFLX. Table 5 shows the re-
sults from the analysis of these incursions.
Incurred shrimp were dosed with ENRO, SARA, and
DIFLX. Three feeds were prepared for each drug at 3 levels:
0.2, 2.0, and 10 ppm. Penaeid shrimp (P. vannamei), 12–18 g
each, were fed medicated feed. Each drug was administered,
individually, at a total rate of 10% body weight in feed, at
3 feedings/day of equivalent size. Medicated feeding was con-
tinued for 14 days, and was followed by nonmedicated feed-
ing during the collection period. Harvesting of tissue occurred
at the times indicated in Table 6 after final dosing. The abdo-
men was removed from the thorax, and the tails, including
midgut and hindgut, were frozen immediately to –20°C for
shipment.
Shrimp sample No. 17 (DIFLX-incurred at 10 ppm, at 11 h
post-dosing) was re-analyzed, and the residue was calculated
by adjusting the LC calibration curve to cover a range from
1 to 40 ppb. The average incurred residues (n= 4) were SARA
3 ppb (5.0% RSD), and DIFLX 7 ppb (2.3% RSD). The initial
analysis as reported in Table 6 shows DIFLX at 8 ppb (esti-
mated) and no SARA residue was detected. This finding dem-
onstrates that the method can detect and analyze residue levels
at <10 ppb if the need arises.
Incurred salmon was obtained by gastric intubation, with
each dosing based on the weight of the individual fish, to pro-
1298 ROYBAL ET AL.:JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002
Table 7. Fluoroquinolone (FQ) found in incurred salmon
Salmon FQ administered Dose level, mg
drug/kg body weight CIPRO,
ppb (RSD, %)
a
ENRO,
ppb (RSD, %)
a
SARA,
ppb (RSD, %)
a
DIFLX,
ppb (RSD, %)
a
100 ENRO 5 10 (14%) 1772 (15%)0 0
102 ENRO 0.5 0 95 (1.4%)0 0
104 ENRO 0.05 0 10 (3.4%)0 0
106 SARA 5 0 0 30 (7.3%)0
108 SARA 0.5 0 0 4.8, est. (2.1%)
b
2.9, est. (3.3%)
b
110 SARA 0.05 0 0 0 1134 (3.6%)
111 SARA 0.05 0 0 0 5.9, est. (1.4%)
b
112 DIFLX 5 0 0 8.0, est. (4.1%)
b
1535 (2.0%)
114 DIFLX 0.5 0 1.3, est. (8.9%)
b
0 126 (4.2%)
116 DIFLX 0.05 0 0 0 28 (1.9%)
a
Average of 4 replicates and RSD of those replicates.
b
Value outside (below) LC calibration curve. Value reported is estimate only.

ROYBAL ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002 1299
Figure 2. Typical liquid chromatogram examples of
standard 4FQs, control, and FQ-incurred catfish tissue
samples. All 50 mL injections with LC conditions specified
in method. (I) 20 ppb each 4FQ LC working standard:
1, CIPRO; 2, ENRO; 3, SARA; and 4, DIFLX. (II) Control
catfish, 2 g through method to final volume of 1.0 mL.
(III) ENRO-incurred catfish: 1, CIPRO, trace (<10 ppb); 2,
ENRO, ~45 ppb. (IV) SARA-incurred catfish: 3, SARA,
~30 ppb. (V) DIFLX-incurred catfish: 4, DIFLX, ~28 ppb.
Figure 3. Typical liquid chromatogram examples of
standard 4FQs, control, and FQ-incurred shrimp tissue
samples. All 50 mL injections with LC conditions
specified in method. (I) 20 ppb each 4FQ LC working
standard: 1, CIPRO; 2, ENRO; 3, SARA; and 4, DIFLX.
(II) Control shrimp, 2 g through method to final volume
of 1.0 mL. (III) ENRO-incurred: 2, ENRO, trace (~7 ppb).
(IV) SARA-incurred: 3, SARA, trace (~6 ppb). (V) DIFLX-
incurred: 3, SARA, trace (~ 3 ppb); and 4, DIFLX, trace
(~8 ppb).

vide the desired quantity of drug. The fish were held in a sepa-
ratetankwiththewatertemperatureat9°C.After18h,thefish
were anesthetized with tricaine methane sulfonate (MS-222)
at a dose of 43 mg/mL in a 25 L vessel. They were eviscerated
andfrozenat20°Cforshipmentindryice.Thesalmonwasin-
curred in duplicates at 3 levels for each of 3 FQ: ENRO,
SARA, and DIFLX. One salmon at each level for each FQ in-
cursion was prepared and used for analysis. Table 7 presents
the analyses of the FQ-incurred salmon samples. As men-
tioned previously, the aim of the incursion was to generate
FQ-incurred tissues at levels appropriate for a validation
study. However, the data from samples 108 and 114 indicate
possible contamination with FQs that were not intentionally
giventothesalmonduringintubation.ScientistsfromtheUni-
versity of British Columbia, where the fish were incurred, re-
ported that after dosing, the salmon were left to recover in a
single tank. Although no regurgitation was observed, there is a
possibility of cross-contamination with the fish occupying the
same holding tank. Data from salmon samples 110 and 111 are
reported as found by LC determination and confirmed by
LC/mass spectrometry (MS). We have no knowledge of the in-
cursionprocesstoexplainwhatmayhavehappenedtothesefish.
Figures 2–4 represent typical chromatograms of control
and incurred samples from the 3 matrixes analyzed, catfish,
shrimp, and salmon, respectively. All residues of FQs in cat-
fish were confirmed by electrospray/liquid chromatogra-
phy/mass spectrometry (ES/LC/MS) and have been previ-
ously reported (24). The residues found in the shrimp and
salmon were also confirmed by ES/LC/MS and that data will
be published.
Acknowledgments
We thank Steven M. Plakas (U.S. Food and Drug Adminis-
tration, Dauphin Island, AL) for his valuable and expert assis-
tance in preparing incurred catfish for this work. Appreciation
is acknowledged to Rodney Williams (University of Arizona)
for providing all shrimp samples, control and incurred, for this
project. Special thanks to K.M. McErlane (University of British
Columbia) for furnishing and preparing all control and incurred
salmon for this study.
References
(1) Jain, R., & Jain, C.L. (1992) LC/GC 10, 707–708
(2) Anadón, A., Martinez-Larrañaga, M.R., Daiz, M.J., Velez,
C., & Bringas, P. (1990) Ann. Rech. Vét. 21 (suppl. 1),
137s–144s
(3) Kaartinen, L., Salonen, M., Älli, L., & Pyörälä, S. (1995) J.
Vet. Pharmacol. Ther. 18, 357–362
(4) Year in Review (1995) Food Chemical News, Washington,
DC, pp 55–57
(5) Fed. Regist. (1996) 61, 56892–56893
(6) Fed. Regist. (1995) 60, 50097–50098
(7) Hammer, P., & Heeschen, W. (1995) Milchwissenschaft 50,
513–514
(8) Hormazabal, V., & Yndestad, M. (1994) J. Liq. Chromatogr.
17, 3775–3782
1300 ROYBAL ET AL.:JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002
Figure 4. Typical liquid chromatogram examples of
standard 4FQs, control, and FQ-incurred salmon tissue
samples. All 50 mL injections with LC conditions
specified in method. (I) 20 ppb each 4FQ LC working
standard: 1, CIPRO; 2, ENRO; 3, SARA; and 4, DIFLX.
(II) Control salmon, 2 g through method to final volume
of 1.0 mL. (III) ENRO-incurred: 2, ENRO, ~10 ppb.
(IV) SARA-incurred: 3, SARA, ~30 ppb.
(V) DIFLX-incurred: 4, DIFLX, ~ 28 ppb.

(9) Tyczkowska, K.L., Voyksner, R.D., Anderson, K.L., &
Papich, M.G. (1994) J. Chromatogr. B. Biomed. Appl. 658,
341–348
(10) Tyczkowska, K., Hedeen, K.M., Aucoin, D.P., & Aronson,
A.L. (1989) J. Chromatogr. 493, 337–346
(11) Granneman, G.R., Snyder, K.M., & Shu, V.S. (1986)
Antimicrob. Agents Chemother. 30, 689–693
(12) Wilson, J.C., & MacMillan, J.R. (1989) J. Aquat. Anim.
Health 1, 222–226
(13) Meinertz, J.R., Dawson, V.K., & Gingerich, W.H. (1994) J.
AOAC Int. 77, 871–875
(14) Tarbin, J.A., Tyler, D.J., & Shearer, G. (1992) Food Addit.
Contam. 9, 345–350
(15) Granneman, G.R., & Sennello, L.T. (1987) J. Chromatogr.
413, 199–206
(16) Horie, M., Saito, K., Nose, N., & Nakazawa, H. (1993)
Shokuhin Eiseigaku Zasshi 34, 289–293
(17) Horie, M., Saito, K., Nose, N., & Nakazawa, H. (1994) J.
Chromatogr. B. Biomed. Appl. 653, 69–76
(18) Posyniak, A., Zmudzki, J., Semeniuk, S., Neidzielska, J., &
Ellis, R. (1999) Biomed. Chromatogr.13, 279–285
(19) Gigosos, P.G., Revesado, P.R., Cadahia, O., Fente, C.A.,
Vazquez, B.I., Franco, C.M., & Cepeda, A. (2000) J.
Chromatogr. A 871, 31–36
(20) Yorke, J.C., & Froc, P. (2000) J. Chromatogr. A 882, 63–77
(21) Roybal, J.E., Pfenning, A.P., Turnipseed, S.B., Walker, C.C.,
& Hurlbut, J.A. (1997) J. AOAC Int. 80, 982–987
(22) Walker, C.C., Storey, J.M., Roybal, J.E., Pfenning, A.P.,
Plakas, S.M., & Turnipseed, S.B. (1996) FDA/ORA/DFS
Laboratory Information Bulletin 12, LIB No. 4046
(23) Sumano, L.H., Gomez, R.B., Gracia, M.I., & Ruiz-Ramirez,
L. (1994) Vet. Hum. Toxicol.36, 477–479
(24) Turnipseed, S.T., Walker, C.C., Roybal, J.E., Pfenning, A.P.,
& Hurlbut, J.A. (1998) J. AOAC Int. 81, 554–562
ROYBAL ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 6, 2002 1301