Multiresidue method for the determination of avermectin and

moxidectin residues in the liver using HPLC with fluorescence

detection
B. Roudaut
CNEVA, Laboratoire des Medicaments Veterinaires, Javene, BP 203, 35302 Fougeres,
France
Received 6th July 1998, Accepted 4th August 1998
A method using high-performance liquid chromatography (HPLC) and fluorescence detection is presented for the
simultaneous determination of the antiparasitic agents avermectins (abamectin, ivermectin and doramectin) and
moxidectin in liver. Samples are extracted using acetonitrile and cleaned up using solid phase extraction on a C
18
column, followed by derivatization with trifluoroacetic anhydride. The samples are injected without further
cleanup into the HPLC column and any avermectins or moxidectin present are detected using fluorescence
detector set at 361 nm excitation and 465 nm emission wavelengths. The limits of quantification are below the
stipulated EU Maximum Residue Limit for each of the avermectins and moxidectin. Mean recoveries in bovine
liver ranged from 77.5 to 90.8%, with a RSD from 2.7 to 7.7%. The procedure provides a rapid, reliable and
sensitive method for determinating avermectin and moxidectin residues in liver of different species (cattle, sheep,
pigs). It is also applicable to muscle without modification.
Introduction
Avermectins and milbemycins are very potent antiparasitic
agents widely used in meat producing animals. These com-
pounds are highly effective at extremely low dose levels (0.2 to
0.5 mg kg
21
) against nematode and arthropod species in cattle,
sheep, pigs and horses. They are not mutagenic or carcinogenic
but may be embryotoxic in laboratory animals. Liver is the
target tissue for residue control and the unaltered parent drugs
were shown to be the major residue components in animal liver.
In the European Union, the Maximum Residue Limits (MRL) of
avermectins in liver have been fixed by Commission Regula-
tions from 15 to 100 mg kg
21
depending on the species
(abamectin: 20 mg kg
21
for the marker residue B
1a
in bovine
liver, doramectin: 50 mg kg
21
in ovine and porcine liver and 100
mg kg
21
in bovine liver, ivermectin: 15 mg kg
21
for the marker
residue H
2
B
1a
, in ovine, procine and horse liver and 100
mg kg
21
in bovine liver, moxidectin: 100 mg kg
21
in bovine and
ovine liver). A number of analytical methods for the determina-
tion of avermectin or ivermectin,
19
and moxidectin
1012
in
animal tissues or products have been described using HPLC and
fluorescence detection. But, no multiresidue methods have been
published.
Our objective was to develop a simple, sensitive, reliable and
multiresidue method for the simultaneous determination of
residues of three avermectins (abamectin, ivermectin and
doramectin) and one milbemycin (moxidectin) in liver (Fig. 1).
HPLC with fluorescence detection after the formation of
fluorescent derivatives of compounds of interest was used in the
analyses. The method developed was based on the extraction
and HPLC procedure used in our laboratory and in field
laboratories in France for monitoring ivermectin concentrations
in liver.
13
The cleanup step and the chromatographic conditions
were adapted for the two other avermectins and for moxidectin.
The fluorescent derivatives were obtained by a dehydrative
reaction, previously described for ivermectin,
14
with tri-
fluoracetic anhydride and N-methylimidazole as the catalyst in
acetonitrile.
Experimental
Materials
Acetonitrile was HPLC grade and was obtained from Merck
(Darmstadt, Germany). Triethylamine was for biochemical
uses, sodium sulfate and trifluoracetic anhydride were analyt-
ical grade and were also obtained from Merck. N-methylimida-
zole (99%) was purchased from Aldrich Chimie (Saint Quentin
Fallavier, France) and methanol from Prolabo (Paris, France).
Water was purified using a Alpha-Q-System (Millipore, Saint
Quentin Yvelines, France). Solid phase extraction cartridges
(C
18
, 100 mg, 1 cm
3
) were obtained from Waters (Saint
Quentin, France).
Standard solutions
Abamectin and ivermectin were provided by Merck Sharp and
Dohme (Bruxelles, Belgium), doramectin from Pfizer (Orsay,
France) and moxidectin from Cyanamid (Tours, France). Stock
solutions of 1 g l
21
were prepared by dissolving each reference
compound (taking into account the purity of the reference
standard: residue marker, H
2
B
1a
for ivermectin and B
1a
for
abamectin) in methanol and were stable for six months in the
freezer. An intermediate standard solution (10 mg ml
21
) was
prepared monthly by diluting each stock solution with methanol
and stored in the freezer. The working standards (2, 1, 0.5, 0.2,
0.15 and 0.075 mg ml
21
) were prepared monthly by dilution of
the intermediate standard in methanol and stored in the freezer.
Aliquots (0.1 ml) of the working standards were taken to
dryness under nitrogen, derivatized and injected onto the HPLC
as described below.
Presented at the Third International Symposium on Hormone and
Veterinary Drug Residue Analysis, Bruges, Belgium, June 25, 1998.
Analyst, 1998, 123, 25412544 2541
P
u
b
l
i
s
h
e
d

o
n

0
1

J
a
n
u
a
r
y

1
9
9
8
.

D
o
w
n
l
o
a
d
e
d

o
n

1
9
/
0
8
/
2
0
1
4

1
7
:
1
0
:
5
5
.

View Article Online / Journal Homepage / Table of Contents for this issue
O
O
O
O O
O
H
3
CO
H
3
C
HO
H
H
3
C
O O
CH
3
O
OH H
OH
CH
3
H
CH
3
C
2
H
5
H
3
C
CH
3
CH
3
O
O
O
O O
O
H
3
CO
H
3
C
HO
H
H
3
C
O O
CH
3
O
OH H
OH
CH
3
H
CH
3
C
2
H
5
H
3
C
CH
3
CH
3
O
O
O
O O
O
H
3
CO
H
3
C
HO
H
H
3
C
O O
CH
3
O
OH H
OH
CH
3
H
CH
3
CH
3
CH
3
O
O
O O
CH
3
O
OH H
OH
H
CH
3
CH
3
CH
3
CH
3
H
N
CH
3
O
CH
3
CH
3
Abamectin B
1a
Ivermectin H
2
B
1a
Moxidectin Doramectin
Equipment
The HPLC system consisted of a P4000 pump (Thermo
Separation Products, Les Ulis, France), a Thermo Separation
Products AS300 autosampler, a Jasco 821-FP fluorescence
detector (Prolabo, Paris, France) and a Thermo Separation
Products SP4510 interface connected to a PC1000 data station
(Thermo Separation Products). The analytical column was used
in a Jones column heater (Touzart et Matignon, Les Ulis,
France) set at 35 C.
Analytical method
Extraction and clean-up. Liver (20 g) was homogenized in
a Moulinette mixer. For each sample, 0.1 ml of methanol and 5
ml of acetonitrile was added to 1 g of homogenized liver in a
tube. The tubes were stirred and placed in an ultrasonic bath for
10 min and centrifuged for 10 min at 2600g. Seven ml of water
were added to the acetonitrile extract (reduced to 3 ml) and the
mixture was applied to a C
18
cartridge which had previously
been conditioned by irrigation with acetonitrile (5 ml) followed
by acetonitrilewater (30 + 70, v/v) containing 0.1% triethyla-
mine (5 ml). The extract was loaded at a flow rate of about 0.3
ml min
21
. The tubes and the cartridges were washed with
acetonitrilewater (30 + 70, v/v) containing 0.1% triethylamine
(1 ml). The cartridges were washed with acetonitrilewater (70
+ 30, v/v, 1 ml) and dried for about 1 min. Samples were then
eluted with acetonitrilewater (90 + 10, v/v, 1 ml) into
centrifuge tubes containing 0.2 g of sodium sulfate. The tubes
were vortexed and centrifuged 10 min at 2600g at a temperature
near 0 C. The supernatant was transferred in a polypropylene
tube and evaporated to dryness with a stream of nitrogen.
Derivatization. The residue in all tubes (standards and
samples) was dissolved in 100 ml N-methylimidazoleacetoni-
trile (1 + 1, v/v) then in 150 ml trifluoracetic anhydride
acetonitrile (1 + 2, v/v). These solutions should be made the day
of use. After mixing, the solutions were transferred to
autosampler polypropylene vials and vials were kept protected
from sunlight until the injection. HPLC analyses were carried
out within 6 h to avoid the degradation of the fluorescent
derivatives.
HPLC analysis. The analyses were performed on a 5 mm
Merck Lichrospher 100RP-18 E column (125 3 4 id mm)
protected by a precolumn (packed with the same phase). The
mobile phase consisted of a mixture of acetonitrile and water
(94 + 6, v/v) and the flow rate was 1 ml min
21
. The injection
volume was 20 ml. The excitation and emission wavelengths
were set at 361 and 465 nm, respectively.
Validation of the analytical method
Linearity of the detector response was checked from a set of
eight working standards ranging in concentration from 30 to
800 mg l
21
. Calibration curves were prepared by plotting the
peak area vs. the analyte concentration and least-squares linear
regression analysis was used to determine the slope. The
linearity of the analytical procedure was also tested by using
fortified livers with known amounts of avermectins and
moxidectin to cover the concentration range 7.5200 mg kg
21
.
Calibration standards were prepared by adding 100 ml of
working solutions to 1 g of liver.
The limits of detection in the HPLC system using the
fluorescence detector were determined from representative
Fig. 1 Chemical structures of three avermectins and one milbemycin (moxidectin).
2542 Analyst, 1998, 123, 25412544
P
u
b
l
i
s
h
e
d

o
n

0
1

J
a
n
u
a
r
y

1
9
9
8
.

D
o
w
n
l
o
a
d
e
d

o
n

1
9
/
0
8
/
2
0
1
4

1
7
:
1
0
:
5
5
.

View Article Online
blank samples. It was equal to three times signal-to-noise, as
recommended in Decision 93/256/EEC.
15
Recovery of the analytical method. To determine the
recovery of the procedure, a series of 1 g samples of bovine liver
was fortified at three levels (15, 20 and 100 mg kg
21
). All
recoveries were determined by comparing the peak areas
obtained from fortified samples with the peak areas resulting
from direct injection of standards carried out through the
derivatization procedure.
Precision of the analytical method. To determine the
intraday and interday repeatability, liver samples were fortified
at three levels (15, 20 and 100 mg kg
21
). They were randomly
coded and analyzed by a technician on three days. For the
intralaboratory reproducibility study, three technicians using
three different LC systems received eight samples (two controls
and six positives) and the standard operating procedure of the
method. Repeatability and reproducibility were calculated as
described in ISO 5725.
Results and discussion
Typical chromatograms of standard solutions, control or
fortified livers to a level of 100 mg kg
21
each of moxidectin,
abamectin, doramectin and ivermectin are shown in Fig. 2. The
retention times of the residue markers moxidectin, abamectin
B
1a
, doramectin and ivermectin H
2
B
1a
were respectively about
6.5, 11, 14 and 19 min, whereas the abamectin B
1b
and the
ivermectin H
2
B
1b
were 9.5 and 16 min. Avermectins and
Table 1 RSD of intraday repeatability (RSD
r
), RSD of interday
repeatability (RSD
R
) and recoveries from fortified bovine livers.
a
Analyte
Theorical
concentration/
mg kg
21
RSD
r
(%)
RSD
R
(%)
Recovery (%)
(mean s)
Moxidectin 15 4.84 7.94 84.3 2.9
20 3.80 5.53 84.2 4.0
100 4.05 4.05 79.7 6.5
Abamectin 15 9.40 9.40 87.5 2.7
20 8.11 8.59 79.3 7.7
100 5.67 5.67 79.0 3.5
Doramectin 15 4.36 4.36 88.8 3.2
20 4.76 4.76 89.2 4.8
100 4.32 6.02 77.5 2.7
Ivermectin 15 9.34 9.61 90.8 3.6
20 5.48 5.48 89.7 5.6
100 4.82 4.82 78.8 3.3
a
Six replicates were conducted on three days for each concentration.
Fig. 2 Chromatograms of (I) a control liver, (II) avermectins and moxidectin standards and (III) a fortified liver (100 mg kg
21
).
Table 2 RSD of repeatability (RSD
r
) and reproducibility (RSD
R
) for
fortified bovine livers by three analysts using different liquid chromatog-
raphy systems
System Moxidectin Abamectin Doramectin Ivermectin
S1 14.8 13.9 13.9 14.3
13.8 13.2 13.3 13.4
13.3 13.4 13.0 12.9
13.7 13.1 13.8 13.5
S2 14.9 15.0 15.2 14.8
14.7 14.8 15.3 14.6
14.3 14.0 14.8 14.0

S3 13.4 12.2 12.9
13.1 11.9 12.8 12.1
13.7 12.6 12.8 12.6
13.2 12.2 12.0 12.2
Global mean 13.97 13.30 13.69 13.48
RSD
r
(%) 3.26 3.18 2.88 3.50
RSD
R
(%) 5.56 9.49 10.43 8.41
Analyst, 1998, 123, 25412544 2543
P
u
b
l
i
s
h
e
d

o
n

0
1

J
a
n
u
a
r
y

1
9
9
8
.

D
o
w
n
l
o
a
d
e
d

o
n

1
9
/
0
8
/
2
0
1
4

1
7
:
1
0
:
5
5
.

View Article Online
moxidectin were chromatographically well resolved under
isocratic conditions. The lack of interferences in the separation
suggests a high specificity of the chromatographic method and
a good selectivity of the extraction procedure. It must be
mentioned that the presence of trace of water should be avoided
before injection. So, the use of polypropylene tubes and vials for
the derivatization step allows the avoidance of the hydrolysis of
the trifluoroacetylated derivatives (4B-hydroxyl) which leads to
the formation of more polar products
16
and to the presence of
interfering peaks on the chromatogram.
The linearity of the analytical procedure was tested by using
bovine liver fortified at 7.5, 15, 20, 50, 100 and 200 mg kg
21
in
eight duplicates at each concentration. All correlation coeffi-
cients of the standard curve linear regression equations
exceeded 0.993, demonstrating good linearity of the derivatiza-
tion reaction and detection system.
The limits of quantification were validated at 7.5 mg kg
21
and
the limits of detection were estimated to be 0.8, 1.5, 2.5 and 2.5
mg kg
21
, respectively, for moxidectin, abamectin B
1a
, dor-
amectin and ivermectin H
2
B
1a
.
The mean recoveries of avermectins and moxidectin in
bovine liver are given in Table 1. The good and reproducible
recoveries obtained for all the analytes at different fortification
levels (77.5 to 90.8%), clearly show that the method is
applicable to determine avermectin and moxidectin residues in
liver. The reproducibility of the method (Table 2) expressed by
the RSD was < 10.5% for the four analytes and the RSD of
repeatability was < 4%. The method was shown to work well
for porcine and sheep liver at the MRL level fixed for
moxidectin, doramectin and ivermectin (Table 3). However,
recoveries were lower in sheep liver than in swine liver.
The applicability of the method to muscle was also tested.
The linearity of the procedure was checked with fortified
muscles at 7.5, 15, 20, 50, 100 and 200 mg kg
21
and the
repeatability at 10 kg
21
; the MRL for doramectin in bovine
muscle (Table 4). In addition to testing the efficiency of the
method to determine avermectins in real samples, we analysed
tissues from a dosed steer with doramectin. The results obtained
were respectively, 48.6 12.1 and 172.7 10.7 mg kg
21
for
muscle and liver. This multiresidue method is being applied in
France in the National Surveillance Scheme of antiparasitic
agents in animal tissues.
Conclusion
The sample preparation and extraction procedure is simple,
rapid and quantitative. The technique uses a small sample size,
has a minimum number of steps and requires minimum amounts
of solvents. No extensive sample clean-up procedure is
required. The method described is sensitive, precise and
accurate. It meets the requirements of methods used for
monitoring drug residues in animal tissues (Decision
93/256/EEC) and a technician can easily analyse 16 samples a
day.
Acknowledgements
The author sincerely thanks Marie-Pierre Fourmond for techni-
cal assistance and the following laboratory members: Maurice
Garnier, and Jacqueline Manceau for intralaboratory validation
of the method and Jean-Pierre Abjean for helpful discussion
about methodology.
References
1 P. C. Tway, J. S. Wood and G. V. Downing, J. Agric. Food Chem.,
1981, 29, 1059.
2 I. Nordlander and H. Johnson, Food Addit. Contam., 1990, 7, 79.
3 J. Markus and J. Sherma, J. Assoc. Off. Anal. Chem., 1992, 75,
757.
4 F. J. Schenck, S. A. Barker and A. R. Long, J. Assoc. Off. Anal.
Chem., 1992, 75, 655.
5 D. G. Kennedy, A. Cannavan, S. A. Hewitt, D. A. Rice and W. J.
Blanchflower, Food Addit. Contam., 1993, 10, 579.
6 C. D. C. Salisbury, J. Assoc. Off. Anal. Chem., 1993, 76, 1149.
7 J. M. Degroodt, B. Wyhowski De Bukanski and S. Srebrnik, J. Liq.
Chromatogr., 1994, 17, 1419.
8 E. Iosifidou, P. Shearan and M. OKeeffe, Analyst, 1994, 119,
2227.
9 L. D. Payne, M. B. Hicks and T. A. Wehner, J. Agric. Food Chem.,
1995, 43, 1233.
10 A. Khunachak, A. R. Dacunha and S. J. Stout, J. Assoc. Off. Anal.
Chem., 1994, 42, 381.
11 M. Alvinerie, J. F. Stura, M. Badri and P. Galtier, J. Chromatogr.,
1995, 674, 119.
12 M. Alvinerie, J. F. Sutra, D. Capela, P. Galtier, A. Fernandez-Suarez,
E. Horne and M. OKeeffe, Analyst, 1996, 121, 1469.
13 B. Roudaut, 1994, SOP P06/08, unpublished work.
14 P. DeMontigny, J. S. K. Shim and J. V. Pivnichy, J. Pharm. Biomed.
Anal., 1990, 8, 507.
15 Commission Decision 93/256/EEC, 14 April 1993, Off. J. Eur.
Comm., 1993, L118, 64.
16 D. W. Fink, P. deMontigny and J.-S. Kim Shim, Analyst, 1996, 121,
1533.
Paper 8/05218B
Table 3 Recoveries for avermectins and milbemycin in porcine and ovine livers fortified at the MRL levels mg kg
21
(n = 5)
Swine Sheep
Analyte Level/mg kg
21
Mean
recovery
(%) s (%) RSD (%) Level/mg kg
21
Mean
recovery
(%) s (%) RSD (%)
Moxidectin 100 66.9 2.36 3.53
Doramectin 50 77.8 3.64 4.67 50 68.2 2.13 3.13
Ivermectin 15 77.9 3.94 5.06 15 64.9 2.16 3.32
Table 4 Recoveries for avermectins and milbemycin in bovine muscles
fortified at 10 mg kg
21
(n = 5)
Analyte
Mean
recovery (%) s (%) RSD (%)
Moxidectin 77.8 7.94 6.1
Abamectin 85.0 9.40 8.3
Doramectin 84.3 4.36 9.9
Ivermectin 89.9 9.61 10.9
2544 Analyst, 1998, 123, 25412544
P
u
b
l
i
s
h
e
d

o
n

0
1

J
a
n
u
a
r
y

1
9
9
8
.

D
o
w
n
l
o
a
d
e
d

o
n

1
9
/
0
8
/
2
0
1
4

1
7
:
1
0
:
5
5
.

View Article Online