· pdf filedankwoord vier jaar geleden ben ik binnengestapt in het laboratorium voor chemische...
TRANSCRIPT
Development of LC-MSnmethods for residue-analysis of veterinary
medicinal products
Nathalie Van Hoof
Thesis submitted in fulfilment of the requirements for the degree of Doctor in Veterinary Science (PhD)
Promoter: Prof. Dr. H. De Brabander
Ghent University, Faculty of Veterinary Medicine
Department of Veterinary Public Health and Food Safety
Laboratory of Chemical Analysis
Printed: www.dclsigns.be
Dankwoord
Vier jaar geleden ben ik binnengestapt in het Laboratorium voor Chemische Analyse, vol
enthousiasme en nieuwsgierigheid voor wat komen zou. Ik kreeg er de smaak te pakken van
het wetenschappelijk onderzoek en stond versteld van de mogelijkheden van LC-MS.
Vandaag is het resultaat er, een bundeling van het onderzoek dat ik de voorbije jaren heb
uitgevoerd. Ik ben dan ook verheugd dat ik mijn collega’s, vrienden en familie kan bedanken
die elk hun steentje hebben bijgedragen om dit alles te realiseren.
Eerst en vooral wil ik mijn promoter Prof. Hubert De Brabander bedanken voor het
vertrouwen dat hij me geschonken heeft en de vrijheid die hij me gegeven heeft om me te
ontplooien in zijn laboratorium. Vier jaar geleden kon ik beginnen in een laboratorium dat
reeds alle ingrediënten bevatte om mijn onderzoek tot een succes te maken. Zowel de nodige
apparatuur als de kennis in residue-analyse en massaspectrometrie, stonden van in het begin
voor mij ter beschikking.
Een andere voorwaarde om dit alles te kunnen realiseren was natuurlijk financiële steun. Ik
heb gedurende de afgelopen vier jaar gewerkt op twee projecten die gefinancierd werden door
het FOD Volksgezondheid, Voedselveiligheid en Milieu. Dankzij deze financiële steun kan ik
vandaag dit werk presenteren.
Natuurlijk wil ik de partner van deze projecten niet vergeten. Dirk Courtheyn en Mieke Van
de Wiele, bedankt voor de aangename samenwerking gedurende de afgelopen 4 jaar. Ik kon
steeds bij jullie terecht met vragen en verzoeken.
Ook andere personen die hebben meegewerkt aan deze projecten wil ik bedanken. Robert
Schilt (TNO, Nederland) voor de samenwerking rond zilpaterol, Bruno Le Bizec, Jean-
Philippe Antignac en Gaud Pinel (Laberca, Nantes, Frankrijk) voor de kans die ze me
gegeven hebben deel te nemen aan SARAF (School for Advanced Residue Analysis in Food)
en voor de goede samenwerking op verschillende vlakken, Dorine Bruneel (KaHo) voor de
synthese van norchloortestosteronacetaat en methyl-3-methyl-2-quinoxalinecarboxylaat-1,4-
dioxide, Walter Gillis voor de samenwerking gedurende de BolTest vergaderingen en de
onderzoeken die daaraan ontsproten zijn.
Printed: www.dclsigns.be
Tevens wil ik de leden van de begeleidings- en leescommissie bedanken voor het kritisch
lezen van mijn werk en voor hun waardevolle opmerkingen. Dit heeft ervoor gezorgd dat mijn
proefschrift een samenhangend geheel is geworden.
Gedurende de voorbije 4 jaar werd ik omringd door de meest fantastische collega’s van LCA,
sommigen onder hen zijn ondertussen vrienden geworden. Ze waren steeds een steun in
goede, maar ook in mindere momenten. Bedankt, Mieke, Wendy, Martine, Marleen, Els,
Lucie, Ann, Dirk, Sigrid, Sofie, Herlinde, Karolien, Antoine, Johan en Marie-José.
Een speciale dank voor Mieke en Wendy voor hun hulp, steun en luisterend oor. Zonder jullie
zou ik hier vandaag niet staan met de resultaten die ik nu heb.
Marleen, bedankt voor de practische hulp bij de layout van mijn proefschrift.
Ook de collega’s van LHT en Soetkin wil ik bedanken. Jullie maakten van de vakgroep een
aangename werkomgeving.
Tevens wil ik Soetkin en ook Davy bedanken voor de hulp bij mijn talloze
computerproblemen.
Davy, met wie ik het afgelopen jaar een bureau deelde, had steeds een luisterend oor en heeft
op gepaste tijdstippen gezorgd voor de nodige ontspanning!
Eén persoon heb ik nog niet vermeld. Sinds een jaar werkt ze niet meer op de vakgroep, maar
daarom is ze zeker niet vergeten! Katia heeft me alles geleerd wat ik momenteel weet van LC-
MS. Zij was de schakel die nodig was om mijn onderzoek te laten slagen. Dankzij haar kreeg
ik de smaak van wetenschappelijk onderzoek te pakken. Bedankt, dit doctoraat ligt hier
vandaag mede dankzij jou!
Heel speciaal wens ik mijn ouders te bedanken. Bedankt, mama en papa voor de kansen die
jullie me gaven. Jullie hebben me steeds gesteund in alle stappen die ik heb ondernomen in
mijn leven. Jullie zijn er altijd voor mij en staan steeds voor me klaar!
En tot slot, bedankt Joachim! Jij bent mijn steun en toeverlaat. Je stond steeds voor me klaar
om me te helpen de dingen te relativeren wanneer ik het even niet meer zag zitten!
Nu ben ik aan het einde gekomen, mijn doctoraat zit erop.
BEDANKT ALLEMAAL!
Printed: www.dclsigns.be
TABLE OF CONTENTS
GENERAL INTRODUCTION AND AIM OF THE STUDY.............................................. 1
CHAPTER 1 ANALYTICAL INTRODUCTION AND LEGISLATIVE ASPECTS ...... 5
Chapter 1.1 Liquid chromatography-tandem mass spectrometry ...................................... 7 1.1.1. The ‘analytical chain’ of LC-MS ................................................................................. 7 1.1.2. Separation techniques................................................................................................... 8 1.1.3. Interfacing techniques .................................................................................................. 9 1.1.4. Mass analysis.............................................................................................................. 14 1.1.5. Electron multiplier...................................................................................................... 18 1.1.6. References .................................................................................................................. 19
Chapter 1.2 European legislation ......................................................................................... 21 1.2.1. Council Regulation (EEC) No 2377/90 ..................................................................... 22 1.2.2. Council Directive 96/23/EC....................................................................................... 23 1.2.3. Commission Decision 2002/657/EC .......................................................................... 24 1.2.4. Flexible scope and secondary validation.................................................................... 30 1.2.5. Legislation in Belgium............................................................................................... 32 1.2.6 References..........................................................................................................................................33
CHAPTER 2 BETA-AGONISTS ......................................................................................... 35
Chapter 2.1 Introduction....................................................................................................... 37 2.1.1. Beta-agonists as growth promoters ............................................................................ 37 2.1.2. Food poisoning........................................................................................................... 38 2.1.3. Legislation.................................................................................................................. 38 2.1.4. References .................................................................................................................. 40
Chapter 2.2 Multi-residue LC-MSn method of beta-agonists in urine usine molecular imprinted polymers ................................................................................................................ 43
2.2.1. Analytical introduction............................................................................................... 43 2.2.1.1. Ion suppression.................................................................................................... 44 2.2.1.2. Molecular Imprinting .......................................................................................... 45
2.2.2. Method setup .............................................................................................................. 48 2.2.3.1. Reagents and chemicals ...................................................................................... 48 2.2.3.2. Instrumentation.................................................................................................... 50 2.2.3.3. Extraction and clean-up....................................................................................... 52
2.2.4. Results and discussion................................................................................................ 54 2.2.4.1. LC-MS methodn .................................................................................................. 54 2.2.4.2. Ion suppression.................................................................................................... 57 2.2.4.3. Qualitative validation .......................................................................................... 61
2.2.5. Conclusion.................................................................................................................. 62 2.2.6. References .................................................................................................................. 63
Chapter 2.3 Excretion profile of zilpaterol in calf urine and faeces.................................. 64 2.3.1. Introduction ................................................................................................................ 65
Printed: www.dclsigns.be
2.3.2. Method setup .............................................................................................................. 66 2.3.3. Experimental .............................................................................................................. 67
2.3.3.1. Reagents and chemicals ...................................................................................... 67 2.3.3.2. Animal experiment.............................................................................................. 67 2.3.3.3. Extraction and clean-up....................................................................................... 67 2.3.3.4. Instrumentation................................................................................................... 68
2.3.4. Results and discussion................................................................................................ 70 2.3.4.1. Chemical structure of zilpaterol .......................................................................... 70 2.3.4.2. LC-MS methods for the detection of beta-agonists in urine and faecesn ............ 72 2.3.4.3. Excretion profile.................................................................................................. 73 2.3.4.4. Phase I metabolites of zilpaterol ......................................................................... 74 2.3.4.5. Quantification...................................................................................................... 75
2.3.5. Conclusion.................................................................................................................. 77 2.3.6. References.........................................................................................................................................78
CHAPTER 3 VETERINARY DRUGS ................................................................................ 79
Chapter 3.1 Introduction....................................................................................................... 81 3.1.1. Classification of veterinary medicinal products......................................................... 82
3.1.1.1. Antibiotics and chemotherapeutics ..................................................................... 82 3.1.1.2. Anthelmintics ...................................................................................................... 88 3.1.1.3 Non-steroidal anti-inflammatory drugs................................................................ 89 3.1.1.4. Glucocorticosteroids............................................................................................ 89
3.1.2. Summary .................................................................................................................... 90 3.1.3. References .................................................................................................................. 92
Chapter 3.2 Introduction and semi-quantification of veterinary medicinal products in injection sites........................................................................................................................... 97
3.2.1. Introduction ................................................................................................................ 97 3.2.2. Experimental .............................................................................................................. 98
3.2.2.1. Reagents and chemicals ...................................................................................... 98 3.2.2.2. Extraction and clean-up procedure...................................................................... 98 3.2.2.3. Instrumentation.................................................................................................... 98 3.2.2.4. Some definitions.................................................................................................. 99
3.2.3. Different approaches ................................................................................................ 100 3.2.3.1. Infusion-MSn ..................................................................................................... 100 3.2.3.2. LC-MSn ............................................................................................................. 100 3.2.3.3. Proposed strategy .............................................................................................. 102
3.2.4. A practical example: sulfadimethoxine.................................................................... 105 3.2.4.1. Identification ..................................................................................................... 105 3.2.4.2. Quantification.................................................................................................... 108
3.2.5. Examples of identified analytes in routine analysis ................................................. 111 3.2.5.1. Identification of penicilline G-benzathine......................................................... 111 3.2.5.2. Interpretation of a florfenicol formulation ........................................................ 112 3.2.5.3. Identification of prednisolone ........................................................................... 115 3.2.5.4. Identification of tolfenamic acid ....................................................................... 117
3.2.6. Discussion ................................................................................................................ 118 3.2.7. Conclusion................................................................................................................ 123 3.2.8. References ................................................................................................................ 124
Printed: www.dclsigns.be
Chapter 3.3 Multi-residue LC-MS method for the detection of quinolones in muscle and bovine milk
n
............................................................................................................................ 125 3.3.1. Introduction .............................................................................................................. 125
3.3.1.1. Mechanism of action ......................................................................................... 128 3.3.1.2. Resistance.......................................................................................................... 128 3.3.1.3. Human health risks............................................................................................ 128 3.3.1.4. Legislation......................................................................................................... 129
3.3.2. Method setup ............................................................................................................ 131 3.3.3. Experimental ............................................................................................................ 131
3.3.3.1. Reagents and chemicals .................................................................................... 131 3.3.3.2. Instrumentation.................................................................................................. 132 3.3.3.3. Extraction and clean-up..................................................................................... 133
3.3.4. Results ...................................................................................................................... 133 3.3.4.1. LC-MS method2 ................................................................................................ 133 3.3.4.2. Validation of the LC-MS method for the detection of quinolones in bovine muscle
2
............................................................................................................................. 135 3.3.4.3. Secondary validation of the LC-MS method for the detection of quinolones in chicken, porcine and aquacultured products muscle
2
...................................................... 142 3.3.4.4. Validation of the LC-MS method for the detection of quinolones in bovine milk2
........................................................................................................................................ 144 3.3.5. Conclusion................................................................................................................ 149 3.3.6. References ................................................................................................................ 150
Chapter 3.4 Multi-residue LC-MS method for the detection of non-steroidal anti-inflammatory drugs in bovine muscle
n
................................................................................ 155 3.4.1. Introduction .............................................................................................................. 155
3.4.1.1. Mechanism of action ......................................................................................... 156 3.4.1.2. Side effects ........................................................................................................ 157 3.4.1.3. Classification..................................................................................................... 157 3.4.1.4. Legislation......................................................................................................... 159
3.4.2. Method setup ............................................................................................................ 160 3.4.3. Experimental ............................................................................................................ 160
3.4.3.1. Reagents and chemicals .................................................................................... 160 3.4.3.2. Instrumentation.................................................................................................. 161 3.4.3.3. Extraction and clean-up..................................................................................... 161
3.4.4. Results ...................................................................................................................... 162 3.4.4.1. Hydrolysis of acetylsalicylic acid ..................................................................... 162 3.4.4.2. LC-MS method2 ................................................................................................ 162 3.4.4.3. Mass spectrometric detection of salicylic acid.................................................. 165 3.4.4.4. Mass spectrometric detection of phenylbutazone ............................................. 165 3.4.4.5. Confirmation of salicylic acid, tolfenamic acid and ketoprofen ....................... 1663.4.4.6. Validation .......................................................................................................... 168
3.4.5. Conclusion................................................................................................................ 175 3.4.6. References....................................................................................................................................... 176
DISCUSSION ....................................................................................................................... 177
SUMMARY........................................................................................................................... 189
Printed: www.dclsigns.be
LIST OF ABBREVIATIONS
ADI Acceptable Daily Intake AL Action limit APCI Athmospheric pressure chemical ionization API Atmospheric pressure ionization BSE Bovine spongiform encephalopathy CAR Carprofen CCα Decision limit CCβ Detection capability cipro Ciprofloxacin COX Cyclooxygenase CSD Clean screen dau CV Coefficient of variation dano Danofloxacin dc Direct current di Difloxacin DLI Direct liquid introduction DOM Desoximethasone enro Enrofloxacin ESI Electrospray ionization EU European Union FAVV Federaal Agentschap voor de Voedselveiligheid FLD Fluorescence detection flum Flumequine FLX Flunixin FLX-d3 Flunixin-d3 FOD Federale overheidsdienst GC Gas chromatography Gyr-A Gyrase subunit A Gyr-B Gyrase subunit B H2O water HPLC High performance liquid chromatography HR High resolution IAC Immuno-affinity chromatography IP Identification point IS Internal standard KET Ketoprofen LC Liquid chromatography LC-MSn Liquid chromatography-tandem mass spectrometry LMCO Low mass cutoff LR Low resolution M Molarity marbo Marbofloxacin MeOH Methanol µg Microgram MIP Molecular imprinted polymers ml Milliliter
Printed: www.dclsigns.be
µl Microliter MLC Meloxicam MPA Medroxyprogesterone acetate MRL Maximum residue limit MRPL Minimum required performance limit MS Mass spectrometry NOEL No-observed-effect-level NRL National Reference Laboratory NSAIDs Non-steroidal anti-inflammatory drugs OTC Oxytetracycline oxo Oxolinic acid PB Particle beam PB Phenylbutazone Pen Penicillin G PFPA Pentafluoropropionic acid quin Quinine RIVM Rijksinstituut voor volksgezondheid en milieu RT Retention time SA Salicylic acid sara Sarafloxacin SCX Strong cation exchange SDT Sulfadimethoxine SIP Standard injection protocol SPE Solid phase extraction TIC Total ion current TMP Trimethoprim TOF Time-of-flight TOLF Tolfenamic acid TYL Tylosin UV Ultraviolet
Printed: www.dclsigns.be
GENERAL INTRODUCTION AND AIM OF THE STUDY
The world and particularly the European Union is becoming increasingly concerned about
human health. Food and environment, very often interrelated, are the main issues giving rise
to a growing concern. An increasing knowledge and awareness of the consumer has led to an
evolution in the demand for safe and healthy food. The confidence of the consumer has been
tested several times over the last few years. After the dioxin crisis, BSE, foot and mouth
disease, MPA crisis, … consumers have become very critical when it comes to their food.
Year by year the use of veterinary medicinal products in animal husbandry has increased.
These products are used for the treatment, control and prevention of animal diseases, but also
to improve feed intake and promote growth of animals. The administration of veterinary
medicinal products may result in the presence of residues of these drugs or their metabolites
in food from animal origin, and these residues may produce hazards for public health. In order
to reduce the likelihood of harmful levels of authorised drugs reaching the human food chain,
the European Union and many other countries have set Maximum Residue Limits. Below
these limits residues are assumed to be harmless to the consumer. Regulatory bodies need to
enforce and verify these requirements and check for the absence of residues of unauthorised
veterinary medicinal products. In Belgium the inspection and control of animal products is
regulated by the Federal Agency for the Safety of the Food Chain. Official (Federaal
Agentschap voor de Voedselveiligheid, FAVV) samples are taken at the slaughterhouse and
the farm to be analysed for unauthorised substances and registered veterinary drugs.
Laboratories play a key role in the control mechanism and have to ensure that regulations are
met. Therefore, the availability of selective and reliable analytical methods is required. The
criteria used for generating results, i.e. for identification and quantification of the analytes,
should encounter the strict international standards. To meet these requirements residue
laboratories working for the government have to be accredited. Every step, from the incoming
sample to the outgoing result, must be traceable. By maintaining this intense level of quality
control, by developing new methods and by using high technological equipment operated by
well trained personnel, laboratories will maintain their important position as part of the chain
of food quality assurance.
1 Printed: www.dclsigns.be
In the Laboratory of Chemical Analysis there are five ion trap mass spectrometers. Mass
spectrometry is necessary for the unequivocal confirmation of analytes. The LC-MS
department consists of one LCQClassic and two LCQDECA (ThermoFinnigan, San José,
California, USA), while the GC-MS department consists of one Polaris and one Polaris Q
(ThermoFinnigan).
The research described in this thesis was performed within the framework of two research
projects funded by FOD Public Health, Safety of the Food Chain and Environment. The
research projects were entitled ‘Identification and quantification of residues of “problem”
molecules in food products of animal origin’ (S-6044/S3) and ‘MSn “flexible” method
development regarding recent residue issues’ (S-6150). The different subjects of each project
are summarized in Table 1, together with the papers written in the context of these projects
and related publications. These subjects coincide with the demands of the Federal Agency for
the Safety of the Food Chain over the last four years for the control of veterinary medicinal
products, including recent growth promoters and veterinary drugs. The goal of these projects
was to develop fast, accurate and flexible MSn methods for the detection of residues of
veterinary medicinal products. These analytical methods will subsequently be implemented in
monitoring programs.
2 Printed: www.dclsigns.be
General introduction and aim of the study
Table 1 Content of the research projects S-6044/S3 and S-6150 and related publications
Anabolic steroids
Norchlorotestosterone acetate (S-6044/S3)
Norchlorotestosterone acetate: GC-MS2 analysis in kidney fat, urine and faeces and study of the metabolisation by
Neomysis integer
N. Van Hoof et al., Chromatographia (2004) 59, S85-S93
Metabolism study of a new anabolic steroid in bovine: preliminary data on 19-norchlorotestosterone acetate.
B. Le Bizec et al., Journal of Steroid Biochemistry and Molecular Biology, Accepted July 2005, in press Chlorodehydromethyltestosterone (incorporated in the multi-residue method for the detection of anabolic steroids)
(S-6150) Flugestone acetate (S-6150)
1-Testosterone (S-6150)
Methenolone acetate
Metabolism of methenolone acetate in a veal calf
N. Van Hoof et al., Veterinary Research Communications, Accepted, in press Beta-agonists
Zilpaterol (S-6044/S3)
Detection of zilpaterol (Zilmax®) in calf urine and faeces with liquid chromatography-tandem mass spectrometry
N. Van Hoof et al., Analytica Chimica Acta (2005) 529, 189-197
Multi-residue liquid chromatography-tandem mass spectrometry analysis of beta-agonists in urine using molecular
imprinted polymers
N. Van Hoof et al., Rapid Communications in Mass Spectrometry (2005) 19, 2801-2808
Flavonoids
Ipriflavone, methoxyisoflavone (S-6150)
Phytoestrogens content in commercial milk samples
N. Van Hoof et al., Proceedings Recent Advances in Food Analysis (2005) Prague
Study of the androgenic activity of ipriflavone by exposure of Neomysis integer
K. Verheyden et al., Proceedings Recent Advances in Food Analysis (2005) Prague Veterinary drugs
Veterinary drugs in injection sites (S-6044/S3 and S-6150)
Detecting veterinary residues in practice: the case of veterinary medicinal products
N. Van Hoof et al., In: Rapid and on-line instrumentation for food quality assurance (2003) ed. I.E. Tothill
Identification of "unknown analytes" in injection sites: a semi-quantitative interpretation
K. De Wasch et al., Analytica Chimica Acta (2003) 483, 387-399
Quinoxalines (methyl-3-methyl-2-quinoxalinecarboxylate-1,4-dioxide) (S-6044/S3)
Unknown quinoxalines, one of the dangers of black market products
N. Van Hoof et al., Proceedings EuroFoodChem XII (2003) Brugge Quinolones
Validation of a liquid chromatography-tandem mass spectrometric method for the quantification of eight quinolones
in bovine muscle, milk and aquacultured products
N. Van Hoof et al., Analytica Chimica Acta (2005) 529, 265-272
Malachite green (accredited method) (S-6150)
Non-steroidal anti-inflammatory drugs (S-6150)
Multi-residue liquid chromatography-tandem mass spectrometry method for the detection of non-steroidal anti-
inflammatory drugs in bovine muscle: optimisation of ion trap parameters
N. Van Hoof et al., Rapid Communications in Mass Spectrometry (2004) 18, 2823-2829
3 Printed: www.dclsigns.be
This thesis imparts only a fraction of the research that has been performed by the candidate
over the last four years, as reflected by the publications published in international ‘peer-
reviewed’ scientific journals (Table 1). The focus of this thesis is on LC-MSn method
development for both growth promoters and veterinary drugs. Therefore, no metabolisation
studies of anabolic steroids, performed with routine GC-MS methods, are included in this
thesis. Two main parts have been selected, one on beta-agonists and a second on veterinary
drugs. In both cases LC-MS was used as detection technique.
The first chapter is a theoretical introduction on the hyphenation of liquid chromatography
and mass spectrometry. It also includes relevant legislative aspects.
In chapter 2 the development of a multi-residue LC-MSn method for the detection of beta-
agonists in urine is discussed and the new beta-agonist zilpaterol is studied. The multi-residue
LC-MSn method compares two different clean-up steps to minimize ion suppression and
subsequently improve the detection of beta-agonists (chapter 2.2). The beta-agonist zilpaterol
is licensed for use as feed additive in Mexico and South-Africa, but its use, like other beta-
agonists, is prohibited in the European Union. Therefore, the excretion profile of zilpaterol
was studied in urine and faeces (chapter 2.3).
Chapter 3 consists of a multi-residue LC-MSn method for the detection of veterinary
medicinal products in injection sites and two specific LC-MSn confirmation methods for the
veterinary drugs, quinolones and non-steroidal anti-inflammatory drugs (NSAIDs). Injection
sites were collected at the slaughterhouse and analysed to give an overview of which
veterinary medicinal products are used nowadays. Therefore, a simple extraction and clean-up
is combined with a multi-residue LC-MSn identification and semi-quantification (chapter 3.2).
Based on the results obtained from the analysis of injection sites and on demand of the
Federal Agency for the Safety of the Food Chain, a quantitative confirmation method was
developed for quinolones (chapter 3.3) and NSAIDs (chapter 3.4).
4 Printed: www.dclsigns.be
Chapter 1.1
Liquid chromatography-tandem mass spectrometry
Mass spectrometry was developed as a technique used by physicists to determine the structure
of an atom at the beginning of the 20th century. In the 1940s, when mass spectrometry began
to be used for the identification and quantification of organic matter, commercially built
instruments began to appear. This led to wider uses in more diverse fields. The concept of
mass spectrometry is to form ions from a sample, to separate the ions based on their mass-to-
charge ratio and to measure the abundance of the ions.
Nowadays, mass spectrometers (MS) are used in combination with separation techniques such
as gas chromatography (GC) and liquid chromatography (LC). This combination of LC/GC
and MS provides both the separation of LC/GC and the identification strength of MS.
The successful hyphenation of liquid chromatography with mass spectrometry is one of the
most important analytical developments of the last decades. Investigation into the coupling of
LC and MS began early 1970s. In the first 20 years, LC-MS was facing interface problems.
The interface between LC and MS has always been a bottleneck to achieve an ideal LC-MS
system. The most difficult obstacle to overcome was maintaining the mass spectrometer high
vacuum in the presence of a liquid flow. This is because the vapour flow in LC-MS is much
greater than in GC-MS (1 ml min-1 of water becomes approximately 1250 ml min-1 of
vapour). A whole variety of interfaces has been developed [1].
Nowadays, LC-MS is used in routine experiments. LC is capable of providing routine
separations of compounds unsuitable for GC analysis (thermally labile and/or non-volatile
analytes, polar or ionic compounds and analytes with a high molecular mass) without the
necessity of preparing volatile derivatives.
1.1.1. The ‘analytical chain’ of LC-MS
The ‘analytical chain’ of LC-MS starts from a sample, through different stages, to an
analytical result (Fig. 1). The first stage is a separation. After a separation the analytes move
from the interface to the mass analyser [2]. The different possibilities for each technology are
summarised in Fig. 1 and the principle of the technologies used in the Lab of Chemical
Analysis will be discussed in the next paragraph.
7
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
SAMPLE
Sample pre-treatment
Separations
Reversed phase LC Normal phase LC Size exclusion LC Ion exchange LC Capillary electrophoresis
Ionisation/ interfaces
Electrospray ionisation Atmospheric pressure chemical ionisation Particle beam Moving belt Direct liquid introduction Thermospray Continuous flow fast atom bombardment
Mass analysis
Quadrupole Ion trap Time of flight Sector Fourier transform
Ion detection
Electron multiplier Photomultiplier
Data integration/interpretation
RESULT
Fig. 1 The ‘analytical chain’ of LC-MS and the alternatives for each technology (the
technologies used in the Laboratory of Chemical Analysis are indicated in bold)
1.1.2. Separation techniques
Chromatography is a physical separation method in which the components are selectively
distributed between two immiscible phases: a mobile phase flowing through a stationary
phased bed. The following types of LC techniques can be distinguished: adsorption (normal
phase and reversed phase), ion-exchange and size exclusion chromatography.
Nowadays reversed-phase chromatography is the most commonly used separation technique.
The reason is the broad application range; reversed-phase chromatography is able to handle
compounds of a diverse polarity and molecular mass. The retention of an analyte depends on
the partition of the analyte between the polar mobile phase and the non-polar stationary phase.
Reversed phase columns consist of a silica or polymeric carrier and a coating of long chain
saturated hydrocarbons or other non-polar functional groups. The most popular packing
8
Printed: www.dclsigns.be
Liquid chromatography-tandem mass spectrometry
material is octadecylsilane with an 18-carbon aliphatic chain. The covalent bonds between the
silica carrier and the aliphatic chain are thermostable and chemically stable, generally, in a pH
range between 2 and 8. Solvents most often used are water, methanol and acetonitrile. The
partitioning of an analyte between the mobile and stationary phase depends upon hydrophobic
interactions between the sample and the mobile phase. Small polar molecules elute more
rapidly than large apolar ones [2-3].
When LC is coupled to MS, some considerations are mandatory in the selection of the
solvents for the mobile phase. The commonly used solvents in reversed-phase
chromatography; water, methanol and acetonitrile, are also ideal for LC-MS. However, non-
volatile solvent additives which are frequently used for LC separations, are not compatible for
LC-MS analysis; phosphate buffers are included in this category. These solvents can
crystallize in the ion source and prevent the mass spectrometer from functioning properly.
Instead of these non-volatile buffers, volatile buffers, such as ammonium acetate and -
formate, can be used.
For electrospray ionisation (paragraph 1.1.3), the ionisation process occurs in the liquid phase.
The pH of the mobile phase will influence the ionisation. An acid pH (an excess of H+ ions)
can enhance the electrospray ionisation. Buffers, such as ammonium acetate and -formate, can
be used to adjust the pH or the pH of the mobile phase can be decreased by adding acetic acid
or formic acid to water or an organic solvent. The concentrations of these buffers or acids are
preferably not too high, since this will influence the ionisation process [4].
1.1.3. Interfacing techniques
The usefulness of any given interface and ionisation system will depend on the ability to
efficiently transport an analyte from solution into the gas phase and into the vacuum system,
where it arrives as a charged species. An important factor determining the choice of an
interface is its ability to maintain the structural integrity of the analyte through these
conversion processes.
Fig. 2 shows the conversion processes required for interfacing liquid chromatography with
mass spectrometry.
9
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
LC Conversion process MS
Liquid-phase Evaporation
Gas-phase
Atmospheric
Pressure reduction
High vacuum
Neutral (ionic) Ionization
Ionic
Fig. 2 Conversion processess
There are two types of interfaces. One group creates ions at atmospheric pressure that are
sampled through atmospheric pressure interfaces (API), the other group samples the analyte
into the vacuum system and ionises the analyte at reduced pressure [2-4, 6].
The first intention of the LC-MS development was to imitate GC-MS and therefore be able to
do electron ionisation for compounds that could not be analysed by GC-MS. To do so the
mobile phase had to be removed prior to ionisation. The moving belt interface is a good
example of this technique. It is based on the selective vaporisation of the elution solvent
before it enters the spectrometer source. Moving belt LC-MS has been replaced by particle
beam LC-MS, a technique which is mechanically more simple and capable of analysing a
similar range of compounds. Direct liquid introduction (DLI) relies on reducing the flow of
the liquid that is introduced into the interface in order to obtain a flow that can be directly
pumped into the source. Constriction in flow and blockages due to the small size of the
pinhole through which the jet of liquid is formed, led to the replacement of DLI interface by
thermospray interface. This interface not only reduces the amount of solvent entering the
mass spectrometer but also ionises analyte molecules including involatile samples. After 1992
thermospray interfaces were replaced by interfaces based on atmospheric pressure ionisation
(API). API interfaces are easier to use and better detection limits are obtained. Continuous
flow fast atom bombardment is also based on the principles of DLI, but its use has decreased
since the wide availability of electrospray interface [2-4,6].
10
Printed: www.dclsigns.be
Liquid chromatography-tandem mass spectrometry
In residue analysis the most common ionisation techniques are the API techniques
electrospray ionisation and atmospheric pressure chemical ionisation. These are soft
ionisation techniques in the sense that at least some analyte molecules are converted, intact,
into corresponding ions.
1.1.3.1. Atmospheric pressure ionisation
These interfaces ionise the sample, eluting from the LC column, at atmospheric pressure and
transfer the ions into the gas phase in order to analyse them in the mass spectrometer. The two
problems with these techniques are the transport of the ions from an atmospheric pressure
region into the vacuum region of the mass spectrometer and the cooling of the mixture of gas
and ions when expanding into the vacuum. The result is condensation of polar molecules such
as water on the analyte ions which will produce cluster ions with a mass exceeding the
capacity of the mass spectrometer. The first problem is solved by introducing optical lenses
that transmit the ions towards the mass analyser, while the gas is pumped away with high-
capacity pumps. The cooling and clustering problem is prevented by applying a high
temperature transfer tube or by applying a dry nitrogen counter-current or gas curtain or by
applying a tension at tube lens/skimmer height (source collision) [3,5].
Electrospray ionisation
Electrospray ionisation (ESI) is a process which occurs in the liquid phase, ions present in the
liquid phase are transferred into the gas phase. These gas-phase ions are sampled into the
vacuum system through a series of successive vacuum chambers. A prerequisite for gas phase
ion production is that the analyte exists in solution as an ion. Therefore, an adaptation of the
pH of the mobile phase is often required. ESI can also be used in the case of molecules
without any ionisable site through the formation of sodium [M+23]+, potassium [M+39]+ or
ammonium [M+18]+ adducts in positive ion mode or acetate [M-60]- adducts in negative ion
mode [5].
There are three major steps in the production of gas-phase ions by electrospray. The first one
is the production of charged droplets at the tip of the capillary. Subsequently these charged
droplets shrink by evaporation of the solvent which results in the formation of small but
highly charged droplets. The last step is the production of gas-phase ions out of these highly
charged droplets (Fig. 3).
11
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
Fig. 3 Electrospray ionisation process
A liquid flow (50 μl min-1 up to 1 ml min-1) enters a hypodermic needle which is held at a
high voltage (typically 6 kV). The liquid is expelled from the needle tip and spreads out as a
plume of charged droplets (the so called Taylor cone). These droplets are highly charged,
having either excess positive or negative charges. This first stage of the process takes place at
atmospheric pressure. Solvent evaporates rapidly from the small droplets. As a result, the
distance between the charges on the surface becomes smaller and smaller, leading to Coulomb
expulsion. This results in the formation of highly-charged microdroplets. Two theories
describe the formation of gas phase ions from these droplets. The ion evaporation model
states that the decrease in droplet size continues until the electric field at the droplet surface is
sufficiently powerful to desorb ions directly into the surrounding gas. This is due to the
repulsion between the escaping ion and the charges that remain in the droplet. The charge
residue model proposes that gas phase ions are produced by evaporation of the solvent from
small droplets that contain only one analyte. In the case of macro-ions the charged residue
model is more likely because these macro-ions will probably be unable to evaporate out of the
droplet. The gas-phase ions are drawn through the lens towards the conical nozzle and
skimmer sections of the API source, where pressure is reduced sufficiently for the free ions to
enter the mass analyser [2,4,5-7].
12
Printed: www.dclsigns.be
Liquid chromatography-tandem mass spectrometry
Electrospray ionisation can be used in both positive and negative ion mode and can analyse
virtually any ion in solution ranging from less than 500 to over 100 000 daltons. Due to the
gentleness of the electrospray process there is usually little or no fragmentation, so the mass
spectrum contains mainly the pseudo-molecular ion, [M+H]+ or [M-H]- in positive and
negative ion mode respectively. Attention needs to be made in the interpretation of ESI
spectra when the analysis has taken place in the presence of additives or contaminants, since
different adducts can be created when such ions are present in solution. Large
macromolecules, such as proteins, can be analysed because electrospray is capable of
attaching many charges to large molecules, generating a series of multiply-charged ions [2,6].
Atmospheric pressure chemical ionisation
In APCI the LC effluent is first evaporated, leaving analyte particles, which are then ionised
with a corona discharge needle. The chromatographic eluate is directly introduced into a
pneumatic nebulizer where it is converted into a thin fog by a high speed nitrogen beam.
Droplets are then displaced by the gas flow through a heated tube called a desolvation
chamber. The heat transferred to the spray droplets allows the vaporisation of the mobile
phase and of the sample in the gas flow. The temperature of this chamber is controlled which
makes the vaporisation conditions independent of the flow and of the nature of the mobile
phase. The temperature of the desolvation chamber is around 400-500 °C, so the vapour
temperature exceeds 100 °C and there is almost a complete evaporation of the mobile phase.
The hot gas and the compounds leaving this tube arrive in an area under atmospheric pressure.
At the end of the desolvation tube is a corona needle which is held at 2.5-3.0 kV. The high
voltage creates a corona discharge which produces ions by a combination of collisions and
charge transfer. First the corona discharge produces ions like N2+ or O2
+, which then collide
with mobile phase molecules producing reactant gas ions. So, the evaporated mobile phase
acts as the ionizing gas. Positive ions are formed through proton transfer and negative ions are
formed through electron transfer or proton loss [2,4,6-7].
APCI is fairly insensitive to the presence of corrosive or oxidizing gasses. The rapid
desolvation and vaporisation of the droplets reduce the thermal decomposition considerably
and thus preserve the molecular species. The ions produced at atmospheric pressure enter the
mass spectrometer and are then focussed towards the analyser.
13
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
APCI is used for the analysis of drugs and metabolites which have a molecular mass below
m/z 2000. Due to the ruggedness of APCI, it is less susceptible to minor changes in buffer
and/or buffer strength [2].
1.1.4. Mass analysis
Mass spectrometry is based on the measurement of mass-to-charge (m/z) ratios of ions. All
molecular ions are, in principle, accessible by mass spectrometry, making it a universal
method for chemical analysis. Its implementation requires suitable methods of ion generation,
ion analysis and ion detection. Next to the potential to yield molecular mass information,
mass spectrometry also gives structural information.
Mass analysis, the separation of ions according to their m/z ratio in space or time, can be
accomplished by several types of analysers: sector mass analysers (single focusing and double
focusing), quadrupole mass analysers, ion trap mass analysers, time-of-flight mass analysers
and Fourier-transform ion cyclotron resonance mass analysers [2,8]. Quadrupole and ion trap
mass spectrometers are the most commonly used detection techniques in residue analysis.
1.1.4.1. Quadrupole mass spectrometer
The quadrupole mass spectrometer consists of a linear array of four symmetrically arranged
rods (poles) to which voltages are supplied (Fig. 4). By applying precisely controlled voltages
to opposing set of poles, a ‘mass filter’ is created. Only ions with a selected mass-to-charge
ratio will pass the poles to be detected at a particular applied voltage [2,4,8].
Fig. 4 Quadrupole mass spectrometer
14
Printed: www.dclsigns.be
Liquid chromatography-tandem mass spectrometry
1.1.4.2. Ion trap mass spectrometer
In an ion trap mass spectrometer ions are trapped within a system of three electrodes with
hyberbolic surfaces, the central ring electrode and two adjacent end-cap electrodes. Ions are
subjected to a three dimensional electric field and oscillate in both the r- and z-directions of
the ion trap (Fig. 5). This electric field results from the application of the potential Φ to the
caps. The potential at any point in this field is given by
Φr,z=(U+Vcosωt)(r2-2z2 + 2z02 / r0
2 + 2z02)
r0 is the internal radius of the ring electrode
z0 is the closest distance from the center to the end-cap
U is the direct current (dc) potential
V is the rf potential applied between the ring and end-cap electrodes
ω is its angular frequency and
t is time
where the first part (U+Vcosωt) describes its temporal variation and the second part (r2-2z2 +
2z02 / r0
2 + 2z02) its spatial dependence.
Fig. 5 Schematic presentation of the ion trap (r0 and z0 represent its size) and the flow of the
ions
15
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
In an ion trap, ions are alternatively subjected to stabilizing and destabilizing forces and
oscillate in both the r- and z-directions. When the phase of rf signal is positive, an ion will be
accelerated from the center of the device. As the rf field changes sign, the same ion is
accelerated towards the center of the trap. Similar considerations apply with respect to an ion
displaced in the radial direction. The ions will be trapped in both the r- and z-direction.
Mathieu stability diagram
Ions of different masses are present together inside the trap and are expelled according to their
masses so as to obtain the mass spectrum. The stability (and instability) of the trajectory of an
ion within the electric field of the ion trap is determined by the Mathieu equations. The
mathematical analysis using these Mathieu equations allows us to locate areas wherein ions of
given masses have a stable trajectory. The Mathieu stability diagram is formed by joining the
stability diagrams in the z- and r-direction. The mathematical background of these diagrams
will not be discussed here. Many publications already described them [8,10]. The Mathieu
stability diagram is expressed in terms of the Mathieu coordinates az en qz (Fig. 6).
az = -2ar = -16zU / m(r02+2z0
2)ω2
qz = -2qr = 8zV / m(r02+2z0
2)ω2
Ions are stable in both the r and z direction if their Mathieu parameters fall within the shaded
area in this diagram (Fig. 6). So the ion trajectories in the ion trap will never exceed the
dimensions of the trap, z0 and r0.
Almost no dc potential is applied between the ring and end-cap electrodes, so the confining
field is purely oscillatory. Each ion species in the ion trap is associated with a qz value which
is calculated according to the Mathieu equation and which lies on the qz axis of the stability
diagram; ions of relatively high m/z ratio have qz values near the origin while ions of lower
m/z ratios have qz values which extend towards the stability boundary.
Optimum operation requires that the ions have favourable initial conditions, which is
achieved by using a helium buffer gas to remove kinetic energy from the ions and cause them
to occupy the central region of the trap. When too many ions are present in the trap,
Coulombic repulsions between the ions will affect their trajectories, this is called space
charging.
16
Printed: www.dclsigns.be
Liquid chromatography-tandem mass spectrometry
Fig. 6 Mathieu stability diagram for the ion trap mass spectrometer
The ion trap is a mass spectrometer that operates in two steps. During ionisation the ring
electrode is driven at an initial imposed rf voltage so that all ions in a given mass/charge
range are trapped within the imposed field. This initial rf voltage imposes a low-mass cut-off
(LMCO) value so that ions of lower m/z ratio are not stored in the trap. They exceed the
instability boundary and exit the trap. Mass selective ion ejection occurs by increasing the
amplitude V of the applied rf voltage so as to ‘move’ ions along the qz axis until they become
unstable at the boundary, where qz = 0.908. There they will exit the trap. Ions of increasing
m/z are ejected and detected as the rf voltage V is ramped.
Tandem mass spectrometry
Tandem mass spectrometry is the succession of two mass-selective operations. The objective
of the first mass-selective operation is to isolate an ion species designated as the precursor
17
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
ion. Isolation of a precursor ion involves ejecting all other ions from the trap. The second
operation determines the mass-to-charge ratios of the fragment or product ions by mass
selective ion ejection.
Isolation is achieved by ramping the rf amplitude until the LMCO is just below the m/z ratio
of the selected ion at which point ions of lower m/z ratios are ejected. The ejection of ions
with m/z ratios higher than the selected ion is achieved by applying a broadband waveform
which narrows the stability area to the selected ion. Once isolation of the selected ion is
completed, the rf amplitude is reduced to obtain a certain qz-value at which the selected ion,
the precursor ion is stable. Not only the precursor ion but also the product ions need to be
stable at this qz-value. By default the qz-value of a Thermo ion trap mass spectrometer is set to
0.25, corresponding to a certain LMCO value. By increasing the qz-value, also the LMCO
value will increase, so possible product ions with m/z ratios below this LMCO value will not
be stored. Subsequently, the precursor ion is excited, typically by applying a supplementary rf
voltage to the end caps. The product ions are recorded by scanning the rf voltage to perform a
second mass-analysis scan [2,4,8-10].
1.1.5. Electron multiplier
A detector is used to transform ions coming from the mass analysers into a measurable signal.
Ions exiting the ion trap reach the conversion dynode, which causes the emission of several
secondary particles. When positive ions strike the negative high-voltage conversion dynode,
the secondary particles are negative ions and electrons. When negative ions strike the positive
high-voltage conversion dynode, the secondary particles of interest are positive ions. These
secondary particles strike the cathode with sufficient energy to dislodge electrons. The
electrons pass further into the electron multiplier, striking walls and causing the emission of
more and more electrons (Fig. 7). This cascade results in a measurable current [4].
Fig. 7 The electron gain at each successive dynode
18
Printed: www.dclsigns.be
Liquid chromatography-tandem mass spectrometry
1.1.6. References
[1] W.M.A. Niessen (2003) Progress in liquid chromatography-mass spectrometry
instrumentation and its impact on high-throughput screening, Journal of Chromatography A
1000, 413-436
[2] R. Willoughby, E. Sheehan, S. Mitrovich (1998) What are your LC/MS alternatives?, In:
A global view of LC/MS; How to solve your most challenging analytical problems, Global
View Publishing, Pittsburgh, PA, USA, 51-99
[3] F.A. Mellon (1991) Liquid Chromatography/Mass Spectrometry, In: VG Monographs in
Mass Spectrometry, volume 2, No. 1
[4] K. De Wasch (2001) The use of liquid chromatography multiple stage mass spectrometry
(LC-MSn) for the determination of residues of growth promoters and veterinary drugs, thesis,
Ghent University, Faculty of Veterinary Medicine, 4-26
[5] E. De Hoffmann, J. Charette, V. Stroobant (1996) Ion sources, In: Mass Spectrometry,
Principles and applications, John Wiley & Sons, Chichester, UK, 9-38
[6] E. De Hoffmann, J. Charette, V. Stroobant (1996) Mass spectrometry-chromatogaphy
coupling, In: Mass Spectrometry, Principles and applications, John Wiley & Sons, Chichester,
UK, 99-113
[7] P. Fürst (2000) LC-MS – a powerful tool in residue analysis of veterinary drugs,
Proceedings of the EuroResidue IV conference, 8-10 May, Veldhoven, The Netherlands, 63-
72
[8] E. De Hoffmann, J. Charette, V. Stroobant (1996) Mass analysers, In: Mass Spectrometry,
Principles and applications, John Wiley & Sons, Chichester, UK, 39-59
[9] P.S.H. Wong and R.G. Cooks (1997) Ion trap mass spectrometry, Current Separations 16,
85-92
[10] R.E. March (1997) An introduction to quadrupole ion trap mass spectrometry, Journal of
Mass Spectrometry 32, 351-369
19
Printed: www.dclsigns.be
Chapter 1.2
European legislation
Pharmacological active substances, including veterinary medicinal products, are used for both
therapeutic and prophylactic purposes, but some of them are also applied as growth
promoters. This has led farmers to use some of these substances having a hormonal,
thyrostatic or adrenergic action, during fattening of livestock. The obvious economic reasons
do not take into account possible harmful effects for the consumer of the products derived
from the carcass of the slaughtered animal.
The marketing authorization of veterinary medicinal products is governed by Directive
2001/82/EC, as amended, on the Community code relating to veterinary medicinal products
[1] and by Regulation (EEC) No 2309/93, as amended, laying down Community procedures
for the authorization and supervision of medicinal products for human and veterinary use and
establishing a European Agency for the Evaluation of Medicinal Products [2]. Council
Regulation (EEC) No 2377/90, as amended, regulates the use of veterinary medicinal
products in foodstuff of animal origin by the establishment of maximum residue limits [3].
The prohibition of the use in stock farming of certain substances having a hormonal or
thyreostatic action and of beta-agonists is laid down in Council Directive 96/22/EC, as
amended [4]. Council Directive 96/23/EC regulates the residue control (monitoring and
surveillance) of veterinary drugs, growth-promoting agents and specific contaminants in live
animals and animal products [5]. Criteria for identification and confirmation, both for
qualitative and quantitative methods, were set out in Decision 93/256/EEC and 93/257/EEC
[6-7]. These criteria have been re-examined in order to take into account the developments in
scientific and technical knowledge. Technical guidelines and performance criteria for residue
control in the framework of Directive 96/23/EC are now described in Commission Decision
2002/657/EC [8-10].
Gradually an extensive network of analytical residue laboratories has been created since 1987
for the purpose of official control. This network consists of a hierarchical system of so-called
Routine and/or Field Laboratories, about 40 National Reference Laboratories and four
Community Reference Laboratories (located in Germany, The Netherlands, France and Italy)
[10-11].
21
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
1.2.1. Council Regulation (EEC) No 2377/90 [3]
Council Regulation (EEC) No 2377/90, as amended, establishes maximum residue limits
(MRL) for veterinary medicinal products in foodstuffs of animal origin. An MRL means the
maximum concentration of a residue resulting from the use of a veterinary medicinal product
(expressed in mg/kg of μg/kg on a fresh weight basis) which may be accepted by the
Community to be legally permitted or recognised as acceptable in food.
Maximum Residue Limits (MRL) are based on the determination of the Acceptable Daily Intake (ADI). The
ADI is an estimate of the residue, that can be ingested daily over a lifetime without a health risk to the consumer.
The ADI is determined following the evaluation of pharmacological and toxicological studies. The basis for the
calculation of the ADI is the no-observed-effect-level (NOEL) and the calculation includes an extremely large
safety factor. In addition, to derive MRLs from the ADI it is assumed that the average person consumes, on a
daily basis, 500 g of meat, 1.5 litres of milk and 100 g of eggs or egg products [12].
The Regulation has four annexes (which are updated on a regular base), which present the
following information [9-10]:
Annex I: pharmacologically active substances for which final MRLs have been established
Annex II: substances for which it is not considered necessary for the protection of public
health to establish MRL values. They are allowed to be used for the animal species identified
and according to the conditions established (e.g. route of administration)
Annex III: pharmacologically active substances for which provisional MRLs have been fixed.
Provisional MRLs are established when not all requirements for the establishment of a MRL
have been fully addressed. Once these issues have been satisfactory addressed, the substance
can be included in Annex I.
Annex IV: pharmacologically active substances for which no MRLs can be fixed. Residues of
these substances in foodstuffs of animal origin are a hazard for public health at whatever
limit. The administration of substances listed in Annex IV to food-producing species is
prohibited in the EU. Substances listed in Annex IV are: Aristolochia spp. and preparations
thereof, Chloramphenicol, Chloroform, Chlorpromazine, Colchicine, Dapsone,
Dimetridazole, Metronidazole, Nitrofurans (including furazolidone), Ronidazole.
22
Printed: www.dclsigns.be
European legislation
1.2.2. Council Directive 96/23/EC [5]
The directive comprises the residue control of food-producing animals as well as their
primary products like meat, milk, eggs and honey. This means that samples are taken from the
living animal on the producing farms as well as from the carcass in the slaughterhouse. The
Directive also establishes National Surveillance Schemes for monitoring of residues of
veterinary medicinal products and contaminants. Annex I of the Council Directive divides all
residues into Group A compounds, which comprise unauthorised substances and Group B
compounds which comprise all authorised veterinary medicinal products (Table 1) [10].
Table 1 List of substances and residues listed in Annex I of the Council Directive 96/23/EC
Group A, substances having anabolic effects and unauthorized substances
• Stilbenes, stilbene derivatives, and their salts and esters
• Antithyroid agents
• Steroids
• Resorcylic acid lactones including zeranol
• Beta-agonists
• Compounds included in Annex IV of Council Regulation (EEC) No 2377/90
Group B, veterinary drugs and contaminants
• Antibacterial substances, including sulphonamides and quinolones
• Other veterinary drugs
1. Anthelmintics
2. Anticoccidiostats, including nitroimidazoles
3. Carbamates and pyrethroids
4. Carbadox and olaquidox
5. Sedatives
6. Non-steroidal anti-inflammatory drugs
7. Other pharmacologically active substances
• Other substances and environmental contaminants
1. Organochlorine compounds including PCBs
2. Organophosphorus compounds
3. Chemical elements
4. Mycotoxins
5. Dyes
6. Others
23
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
1.2.3. Commission Decision 2002/657/EC [8]
In order to ensure the harmonized implementation of Directive 96/23/EC, performance
criteria for analytical residue methods are defined in Commission Decision 2002/657/EC. It is
necessary to ensure the quality and comparability of analytical results generated by
laboratories approved for official residue control. This should be achieved by applying
methods validated according to common procedures and performance criteria. For substances
for which no permitted limit has been established and for substances which are not
authorized, a minimum required performance limit (MRPL) should be provided to be able to
harmonize analytical methods [9,13].
The minimum required performance limit (MRPL) is the minimum content of an analyte in a sample which at
least has to be detected and confirmed. The MRPL is used as a parameter in order to harmonize the analytical
performance of methods for substances for which no permitted limits have been established. Consequently, a
MRPL has nothing to do with toxicity; it is established based on the characteristics of the available analytical
method.
The Commission Decision was published in August 2002. It is a revised version of the
repealed Decisions 93/256/EEC and 93/257/EEC and it takes recent technical and scientific
developments into account [10].
1.2.3.1. Performance criteria
Confirmatory methods for organic residues and contaminants must provide information on the
chemical structure of the analyte. Consequently methods based only on chromatographic
analysis without the use of spectrometric detection are not suitable on their own for use as a
confirmatory method.
Every confirmatory method requires a ‘standard injection protocol’ (SIP) to guarantee the
quality of detection and quantification. A SIP is a logical succession of standards, blanks and
samples. To start, a standard mixture of the analytes of interest is injected on column. This is
to check the performance of the instrument and the shift in retention time. If a shift is
observed, the mass spectrometric segments are adjusted. After the standard injection, mobile
phase is injected to assure that there is no carry-over. Afterwards the quality control samples,
24
Printed: www.dclsigns.be
European legislation
spiked and blank matrix, are acquired followed by the real samples and again a standard
mixture [11].
When a sample reveals a ‘suspected’ mass spectrum, quality criteria are necessary for the
qualification and quantification. Quality criteria for the identification of organic residues and
contaminants are based on the use of identification points (IP). The system of identification
points balances the identification power of the different analytical techniques and moreover
has the advantage that new techniques may easily be incorporated in the procedure (Table 2).
Table 2 Relationship between nature of MS information and IPs obtained
MS technique IP earned per ion
Low resolution mass spectrometry (LR) 1.0
LR-MSn precursor ion 1.0
LR-MSn transition products 1.5
High resolution mass spectrometry (HR) 2.0
HR-MSn precursor ion 2.0
HR-MSn transition products 2.5
The minimum number of IPs for unauthorised compounds (substances listed in group A of
Annex I of Directive 96/23/EC, Table 1) is set to four, for compounds with a MRL
(substances listed in group B of Annex I of Directive 96/23/EC, Table 1) a minimum of three
IPs is required for the confirmation of the compounds’ identity. LC-MSn precursor ions
account for 1 IP and LC-MSn product ions for 1.5 IP. The selected product ions should not
exclusively originate from the same part of the molecule. All product ions must have a signal
to noise ratio of at least 3:1. The relative intensities of the detected product ions expressed as
a percentage of the intensity of the most abundant ion must correspond within the tolerances
given in Table 3, to those of the standard analyte or spiked matrix at comparable
concentrations and measured under the same conditions.
25
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
Table 3 Maximum permitted tolerances for relative ion intensities
Relative ion intensity Tolerances (LC-MS, LC-MSn)
> 50 % ± 20 %
> 20 – 50 % ± 25 %
> 10 – 20% ± 30 %
≤ 10 % ± 50 %
The criteria for identification are based on precursor and product ions, which give structural
information. Retention time is another parameter that gives an indication of the identity of an
analyte, but it contains no structural information. Using LC-MSn as detection method, the
relative retention time of the analyte in the unknown sample must correspond to that of the
standard or spiked matrix and this within a tolerance range of 2.5 % [11].
Confirmation methods can be both qualitative or quantitative. Qualitative methods are used
for unauthorised substances (substances listed in group A of Annex I of Directive 96/23/EC,
Table 1) and non-compliant use of veterinary medicinal products. Qualitative methods
determine if a sample is compliant or non-compliant without quantification. The identification
of the analyte must be completed according to the described quality criteria. Quantitative
methods are required to detect veterinary drugs with an established MRL (substances listed in
group B of Annex I of Directive 96/23/EC, Table 1). The methods need to confirm if the
concentration of an analyte is below or above this limit [14].
Qualitative method means an analytical method which identifies a substance on the basis of its chemical,
biological or physical properties
Quantitative method means an analytical method which determines the amount or mass fraction of a substance
so that it may be expressed as a numerical value of appropriate units
26
Printed: www.dclsigns.be
European legislation
1.2.3.2. Validation
Before an analytical method can be applied for official control analyses, some performance
characteristics must be determined through a validation. The validation parameters required in
the validation of a qualitative confirmation method are: specificity/selectivity,
applicability/ruggedness/stability, detection capability (CCβ) and decision limit (CCα). The
validation parameters of a quantitative confirmation method are equal to the ones of a
qualitative validation plus recovery and precision (Table 4).
Table 4 The performance characteristics required for each analytical method
CCβ CCα Trueness/
Recovery
Precision Selectivity/
Specificity
Applicability/
Ruggedness/
Stability
S + - - - + + Qualitative
methods C + + - - + +
S + - - + + + Quantitative
methods C + + + + + +
S = screening method
C = confirmation method
Screening method means methods that are used to detect the presence of a substance or class of substances at
the level of interest. These methods have the capability for a high sample throughput and are used to sift large
numbers of samples for potential non-compliant results. They are specifically designed to avoid false compliant
results
Confirmatory method means methods that provide full or complementary information enabling the substance
to be unequivocally identified and if necessary quantified at the level of interest
27
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
Detection capability (CCβ) and decision limit (CCα)
Detection capability is the smallest content of the analyte that may be detected, identified and/or quantified in a
sample with an error probability of β. In the case of substances for which no permitted limit has been
established, the detection capability is the lowest concentration at which a method is able to detect truly
contaminated samples with a statistical certainty of 1-β. In the case of substances with an established permitted
limit, this means that the detection capability is the concentration at which the method is able to detect permitted
limit concentrations with a statistical certainty of 1-β.
β-error means the probability that the tested sample is truly non-compliant, even though a compliant
measurement has been obtained (false compliant decision).
The decision limit means the limit at and above which it can be concluded with an error probability of α that a
sample is non-compliant.
α-error means the probability that the tested sample is compliant, even though a non-compliant measurement
has been obtained (false non-compliant result).
The introduction of decision limit CCα has eliminated the problem of the calculation of the
method uncertainty, since CCα includes this uncertainty. In contrast to CCα, the detection
capability CCβ has no function as far as the assessment of sample conformity. It is a
parameter to estimate the proficiency of the method as regards its false-compliant rate (β-
error) [10].
Depending on political decisions, protection of producers, industrials or consumers, the
criteria to declare a sample compliant or non-compliant will differ [13]. The α-error (false
non-compliant result) includes a risk for the producer, while the β-error (false compliant
result) includes a risk for the consumer.
Trueness/recovery
Trueness of a quantitative method must be determined by repeated analyses of a certified
reference material. When such certified reference material is not available, trueness can be
assessed through recovery of spiked matrix. The recovery must fall within the ranges shown
in Table 5.
28
Printed: www.dclsigns.be
European legislation
Table 5 Minimum trueness of quantitative methods
Mass fraction Range
≤ 1 µg/kg - 50 % to + 20 %
> 1 µg/kg to 10 µg/kg - 30 % to + 10 %
≥ 10 µg/kg - 20 % to + 10 %
Precision
The coefficient of variation is the determining parameter for an estimation of the precision.
The inter-laboratory coefficient of variation for the repeated analysis of spiked samples, under
reproducibility conditions, must not exceed the level calculated by the Horwitz equation, CV
= 2(1-0.5logC) (%), where C is the mass fraction expressed as a power of 10 (Table 6). For mass
fractions lower than 100 µg kg-1 the application of the Horwitz equation gives unacceptable
high values. Therefore, the CV for concentration lower than 100 µg kg-1 need to be as low as
possible.
Table 6 Examples of CVs for quantitative methods at a range of analyte mass fractions
Mass fraction CV (%)
1 - 10 µg kg-1 *
100 µg kg-1 23
1000 µg kg-1 16
* Unacceptable high values for mass fractions lower than 100 µg kg-1
For analyses carried out under repeatability conditions, the intra-laboratory CV would be
between one half and two thirds of the values calculated by the Horwitz equation. For
analyses carried out under within-laboratory reproducibility conditions, the within-laboratory
CV shall not be greater than the inter-laboratory CV.
Specificity/Selectivity
The analytical method must be able to distinguish between the analyte of interest and closely
related substances and matrix interferences. This is the specificity of the method.
The selectivity of the analytical method is determined by the identification points of each
analyte.
29
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
Applicability/Ruggedness/Stability
Applicability and ruggedness can be easily tested when the analytical method is already in
use in routine analysis. Applicability is the observation of the consequences when minor
reasonable variations are introduced into the method. Such factors may include the analyst,
temperature during evaporation, pH values, as well as many other factors that can occur in the
laboratory.
The stability of the analyte during storage or analysis may give rise to significant deviations
in the outcome of the result of analysis. Therefore, monitoring of the storage conditions is
necessary.
1.2.4. Flexible scope and secondary validation
Laboratories working for the government have to be accredited to meet the strict international
requirements concerning identification and quantification of samples. In Belgium, the institute
responsible for the control of the accreditation of laboratories is BELAC.
The application scope of an accreditation is the description of the activities for which a
laboratory is accredited. There are different types of application scopes. A fixed scope
includes specific applications which are in accordance with the accreditation requirements.
Including a new application requires a preceding approval of BELAC. A flexible scope is
attributed to laboratories that already have proven their capability concerning the application
of new methods. As a consequence, these laboratories are allowed to extend or to change
analytical methods within their application scope of accreditation. No preceding approval of
BELAC is necessary [15].
Laboratories having a flexible scope will make a distinction between a total validation
(Commission Decision 2002/657/EC, paragraph 1.2.3.) and a secondary validation. A
secondary validation is sufficient in the case of an extension of a totally validated analytical
method. The extension is limited to analytes belonging to the same subgroup presented in
Table 1 or to matrices belonging to the same matrix-class (Table 7).
30
Printed: www.dclsigns.be
European legislation
Table 7 The classification of groups and subgroups of matrices
for the detection of residues and contaminants
Animal tissue and eggs
Tissue of animals living on the land (meat, kidney, liver,
thyreostat, other muscle tissue)
Tissue of fish and crustaceans
Eggs
Fat tissue
Urine, water and bile
Urine
Drinking and waste water
Bile
Plasma, blood, milk and milk products
Plasma
Blood
Milk and milk products
Faeces
Animal feed
Preparations such as syringes and pharmaceuticals
Hair and fur
Hair
Fur
Other matrices such as honey, retina
The performance characteristics required in a secondary validation of a qualitative method are
selectivity/specificity and detection capability CCβ. The parameters of a secondary validation
of a quantitative confirmation method are equal to the ones of a qualitative validation plus
recovery and precision.
The detection capability, recovery and precision will be determined for a limited number of
samples. Afterwards the detection capability and the intra-laboratory repeatability will be
expanded by analysing spiked blank matrices with each batch of samples [16].
31
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
1.2.5. Legislation in Belgium
Samples taken by official control services in the slaughterhouse or the farm need to be
analysed in official laboratories for unauthorised substances and for legally used veterinary
drugs. If in Belgium residues of veterinary medicinal products are detected in a concentration
higher than the MRL then the farm receives a R-status. This will implement that during 8
weeks there will be one analysis for every 10 slaughtered animals at the cost of the owner. If
residues are found of an unauthorised substance the consequences are more severe and the
farm will receive an H-status. An H-status will be implemented for 52 weeks. A sample from
one of every 10 animals slaughtered will be analysed at the cost of the owner [17].
32
Printed: www.dclsigns.be
European legislation
1.2.6 References
[1] Directive 2001/82/EC of the European parliament and of the Council of 6 November 2001
on the Community code relating to veterinary medicinal products (2001) Official Journal of
the European Communities, no. L 311
[2] Council Regulation (EEC) No 2309/93 of 22 July 1993 laying down Community
procedures for the authorization and supervision of medicinal products for human and
veterinary use and establishing a European Agency for the Evaluation of Medicinal Products
(1993) Official Journal of the European Communities, no. L 214
[3] Council Regulation (EEC) N° 2377/90 of 26 June 1990 laying down a Community
procedure for the establishment of maximum residue limits of veterinary medicinal products
in foodstuffs of animal origin (1990) Official Journal of the European Communities, no. L 67
[4] Council Directive 96/22/EC of 29 April 1996 concerning the prohibition on the use in
stock farming of certain substances having a hormonal or thyrostatic action and of beta-
agonists, and repealing Directives 81/602/EEC, 88/146/EEC and 88/299/EEC (1996) Official
Journal of the European Communities, no. L 125
[5] Council Directive 96/23/EC of 29 April 1996 on measures to monitor certain substances
and residues thereof in live animals and animal products and repealing Directives 85/358/EEC
and 86/469/EEC and Decision 89/187/EEC and 91/664/EEC (1996) Official Journal of the
European Communities, no. L 125
[6] Commission Decision 93/256/EEC of 14 April 1993 laying down the methods to be used
for detecting residues of substances having a hormonal or a thyreostatic action (1993) Official
Journal of the European Communities, no L 118
[7] Commission Decision 93/257/EEC of 15 April 1993 laying down the reference methods
and the list of national reference laboratories for detecting residues (1993) Official Journal of
the European Communities, no L 118
[8] Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive
96/23/EC concerning the performance of analytical methods and the interpretation of results
(2002) Official Journal of the European Communities, no. L 221
[9] C. Van Peteghem and E. Daeseleire (2003) Drug residue analysis in food and feed: state-
of-the-art for growth promoters, Proceedings Euro Food Chem XII: Strategies for safe food,
24-26 September, Brugge, Belgium, 379-385
[10] A.A.M. Stolker (2005) Determination of veterinary drugs and growth-promoting agents
in food producing animals, thesis, Faculty of Sciences, University of Amsterdam, 15-30
33
Printed: www.dclsigns.be
Analytical introduction and legislative aspects
[11] F. André, K.K.G. De Wasch, H.F. De Brabander, S.R. Impens, L.A.M. Stolker, L. van
Ginkel, R.W. Stephany, R. Schilt, D. Courtheyn, Y. Bonnaire, P. Fürst, P. Gowik, G.
Kennedy, T. Kuhn, JP. Moretain and M. Sauer (2001), Trends in the identification of organic
residues and contaminants: EC regulations under revision, Trends in Analytical Chemistry 20
(8), 435-445
[12] K. Grein (2000) The safe use of veterinary medicines and the need of residue
surveillance, Proceedings of the Euroresidue IV conference, 8-10 May, Veldhoven, The
Netherlands, 73-78
[13] JP. Antignac, B. Le Bizec, F. Monteau and F. André (2003) Validation of analytical
methods based on mass spectrometric detection to the ‘2002/657/EC’ European decision:
guideline and application, Analytica Chimica Acta 483, 325-334
[14] H.F. De Brabander, K. De Wasch, L. Okerman en P. Batjoens (1998) Moderne
analysemethodes voor additieven, contaminanten en residuen, Vlaams Diergeneeskundig
Tijdschrift 67, 96-105
[15] BELAC 2-101 Rev 1-2004 (2004) Toepassingsgebied van een accreditatie toegekend aan
een beproevingslaboratorium: leidraden voor de omschrijving en de beoordeling ervan
[16] BELAC 2-105 Rev 0-2004 (2004) Criteria waaraan de geaccrediteerde laboratoria
moeten beantwoorden die een flexibele scope aanvragen voor analyses ter uitvoering van de
richtlijn 96/23/EG overeenkomstig beschikking 2002/657/EG
[17] L. Okerman, K. De Wasch, H. De Brabander, R. Abrams, J. Van Hoof, M. Cornelis en L.
Laurier (1999) Oude en nieuwe opsporingstechnieken voor antibioticaresiduen in het kader
van de huidige Belgische en Europese wetgeving, Vlaams Diergeneeskundig Tijdschrift 68,
216-223
34
Printed: www.dclsigns.be
Chapter 2.1
Introduction
Beta-agonists have been derived from the endogenous catecholamine adrenaline (Fig. 1). By
increasing the bulkiness of the substituent on the N-atom and substitution of the catechol
hydroxyl groups, substances with greater beta-selectivity and less susceptibility to metabolic
degradation were obtained.
HO
HO
OH
HN
CH3
Fig. 1 Chemical structure of adrenaline (= epinephrine)
Beta-agonists bind to beta-2-receptors. Stimulation of the beta-2-receptors results in
relaxation of smooth muscle. As a consequence beta-agonists are used as bronchodilator for
the treatment of pulmonary diseases and are used for the treatment of premature labour. The
main side effects are muscle tremor and peripheral vasodilatation [1].
2.1.1. Beta-agonists as growth promoters
When beta-agonists are used as growth promoters in animal production, an excess of 5 to 10
times the recommended therapeutic dose (20-40 mg kg-1) is required in the diet of cattle to
increase the muscle/fat ratio. The carcass composition is improved due to an increase in
muscle mass and a breakdown of fat. Repartitioning effects have been shown for up to 70
days after withdrawal. Despite intensive investigation, the exact mode of action of beta-
agonists on muscle cells is still not completely clear [1-3]. Clenbuterol is an effective growth
promoter in ruminants. Treatment of young animals resulted in a 10-20 % higher growth rate
and better feed conversion. Fat depots were reduced by more than 10 % and muscle
percentage increased by 10-25 %. The response to beta-agonists varies with species as well as
type and dose of beta-agonist, length of treatment period, length of withdrawal period,
genotype, sex, growth phase and dietary composition [4]. The repartitioning effect has led
farmers to use beta-agonists during fattening of cattle.
37
Printed: www.dclsigns.be
Beta-agonists
2.1.2. Food poisoning
Since 1990, several incidences of acute food poisoning resulting from the consumption of
clenbuterol contaminated bovine meat and liver have been documented. In Spain 2 major
outbreaks with 367 cases were reported, in France 22 cases, in Italy 62 people developed food
poisoning symptoms, and in Portugal there was a report on 50 intoxicated patients [5-15]. In
1990 there were outbreaks of food poisoning in Spain caused by consumption of bovine liver.
This was the first time that pharmacotoxicological residues were found in slaughtered cattle
that caused acute food-borne intoxication in consumers [5]. The main symptoms found were
tremors and tachycardia, nervousness and general malaise. No fatalities have been reported
until now [5-15].
2.1.3. Legislation
The potential toxicological implications for man have urged the EU to prohibit beta-agonists
as growth promoting substances in cattle raised for human consumption. Therefore, the use of
beta-agonists as growth promoters is banned since 1996 in the European Union (Council
Directive 96/23/EC) [16]. Clenbuterol is the only member of this group of drugs licensed in
the EU for therapeutic use in food producing species. However, the addition of clenbuterol in
the diet of animals is not allowed without veterinary prescription for therapeutic purposes and
detailed records of administration are required. A maximum residue limit of 0.5 µg kg-1 for
clenbuterol in liver of cattle and horses and a minimum 15 days withdrawal period is
proposed by law to ensure that meat products for human consumption are virtually free from
residues of beta-agonists [17]. The recommended dosage of clenbuterol for treating
pulmonary diseases in cattle and horses is 0.8 µg/kg bodyweight for a period of 10-14 days
[1]. Although, beta-agonists have been prohibited for repartitioning purposes in the European
Union, other countries like USA, Mexico and South-Africa have licensed some of them at
growth promoting doses and there is also a strong black market for illegal use of beta-agonists
in cattle because a substantial financial profit can be attained.
38
Printed: www.dclsigns.be
Introduction
This chapter is divided into two different parts. In chapter 2.2 a multi-residu LC-MSn
analysis of beta-agonists in calf urine is discussed. Two different clean-up steps, Clean Screen
Dau and Molecular Imprinted Polymers, were evaluated with respect to their ability to
minimise ion suppression in liquid chromatography-tandem mass spectrometry. In chapter
2.3 the excretion profile of zilpaterol, a new beta-agonist, was studied in urine and faeces after
oral treatment of a male veal calf with therapeutic doses of Zilmax®.
39
Printed: www.dclsigns.be
Beta-agonists
2.1.4. References
[1] A. Koole (1998) Multi-residue analysis of growth promoters in food-producing animals,
University of Groningen, Faculty of Mathematics and Physics, 29-43
[2] K. De Wasch (2001) The use of liquid chromatography multiple stage mass spectrometry
(LC-MSn) for the determination of residues of growth promoters and veterinary drugs, thesis,
Ghent University, Faculty of Veterinary Medicine, 169-184
[3] A. Prezelj, A. Obreza, S. Pecar (2003) Abuse of clenbuterol and its detection, Current
Medicinal Chemistry 10, 281-290
[4] M. Lafontan, M. Berlan and M. Prud’hon (1988) Beta-adrenergic agonists. Mechanism of
action: lipid mobilization and anabolism, Reproduction, Nutrition, Development 28, 61-84
[5] J.F. Martinez-Navarro (1990) Food poisoning related to consumption of illicit beta-agonist
in liver, Lancet 336, 1311
[6] C. Pulce, D. Lamaison, G. Keck, C. Bostvironnois, J. Nicolas, J. Descotes (1991)
Collective human food poisonings by clenbuterol residues in veal liver, Veterinary and
Human Toxicology 33, 480-481
[7] L. Salleras, A. Dominguez, E. Mata, J.L. Taberner, I. Moro, P. Salva (1995)
Epidemiologic study of an outbreak of clenbuterol poisoning in Catalonia, Spain, Public
Health Report 110, 338-342
[8] S. Maistro, E. Chiesa, R. Angeletti, G. Brambilla (1995) Beta blockers to prevent
clenbuterol poisoning, Lancet 346, 180
[9] J. Bilbao-Garay, J.F. Hoyo-Jimenez, M. Lopez-Jimenez, M. Viruesa-Sebastian, J.
Perianes-Matesanz, P. Munoz-Moreno, J. Ruiz-Galiana (1997) Clenbuterol poisoning.
Clinical and analytical data on an outbreak in Mostoles, Madrid, Rev Clin Esp 197, 92-95
[10] G. Brambilla, A. Loizzo, L. Fontana, M. Strozzi, A. Guarino, V. Soprano (1997) Food
poisoning following consumption of clenbuterol-treated veal in Italy, Journal of the American
Medical Association 278, 635
[11] G. Brambilla, T. Cenci, F. Franconi, R. Franconi, R. Galarini, A. Marci, F. Rondoni, M.
Strozzi, A. Loizzo (2000) Clinical and pharmacological profile in a clenbuterol epidemic
poisoning of contaminated beef meat in Italy, Toxicology Letters 114, 47-53
[12] V. Sporano, L. Grasso, M. Esposito, G. Oliviero, G. Brambilla, A. Loizza (1998)
Clenbuterol residues in non-liver containing meat as a cause of collective food poisoning,
Veterinary and Human Toxicology 40, 141-143
[13] Z. Chodorowski, J. Sein Anand (1997) Acute poisoning with clenbuterol – a case report,
Przegl Lek 54, 763-764
40
Printed: www.dclsigns.be
Introduction
[14] F. Ramos, I. Silveira, J.M. Silvo, J. Barbosa, C. Cruz, J. Martins, C. Neves, C. Alves
(2004) Proposed guidelines for clenbuterol food poisoning, American Journal of Medecine
117, 362
[15] G.A. Mitchell, G. Dunnavan (1998) Illegal use of beta-adrenergic agonists in the United
States, Journal of Animal Science 76, 208-211
[16] Council Directive 96/22/EC of 29 April 1996 concerning the prohibition on the use in
stock farming of certain substances having a hormonal or thyrostatic action and of beta-
agonists, and repealing Directives 81/602/EEC, 88/146/EEC and 88/299/EEC (1996) Official
Journal of the European Communities, no. L 125
[17] K. De Wasch, H.F. De Brabander, D. Courtheyn (1998) LC-MS-MS to detect and
identify four beta-agonists and quantify clenbuterol in liver, The Analyst, 123, 2701-2706
41
Printed: www.dclsigns.be
Chapter 2.2
Multi-residue LC-MSn method of beta-agonists in urine using molecular
imprinted polymers
Adapted from:
N. Van Hoof, D. Courtheyn, JP. Antignac, M. Van de Wiele, S. Poelmans, H. Noppe and H. De
Brabander
Multi-residue liquid chromatography-tandem mass spectrometry analysis of beta-agonists in urine
using molecular imprinted polymers
Rapid Communications in Mass Spectrometry (2005) 19, 2801-2808
2.2.1. Analytical introduction
In the field of residue control, the fulfilment of precise analytical criteria is mandatory as
described in Commission Decision 2002/657/EC [1]. In particular, the requirements in terms
of unambiguous identification of the target analytes led to the widespread utilisation of mass
spectrometry as a confirmatory technique. While GC-MS instruments were historically the
more widely used for various classes of residues, LC-MS today appears as the method of
choice and the major actual investment for many laboratories, especially for the analysis of
polar compounds. However, after a first period of great enthusiasm shared by most end-users,
some problems related to these LC-MS techniques were reported. One main source of pitfalls
was the existence of matrix effects in general, and ion suppression phenomenon in particular.
Therefore, one should adopt a standard practice that acknowledges the necessity of improved
sample preparation before measurement in order to minimize problems of this kind [2-5].
Multi-residue mass spectrometric methods for the detection of beta-agonists, described in
literature are mainly based on mixed-phase solid phase extraction (SPE) [6-14]. These SPE
procedures have proven to be selective not only for beta-agonists, but also for other basic
drugs. Therefore they cannot provide the degree of selectivity needed for each beta-agonist. A
possible solution is Molecular Imprinted Polymers (MIP) for the sample clean-up of beta-
agonists. In this context, the aim of the work presented in this chapter was to test whether the
43
Printed: www.dclsigns.be
Beta-agonists
clean-up of beta-agonists using MIP columns could improve the overall method performance,
not only in terms of analyte recovery but also in terms of removal of interfering compounds
and reduction of the ion suppression phenomenon.
2.2.1.1. Ion suppression
Ion suppression is a problem occurring in the early stages of the ionization process. It can
occur when a coeluted compound suppresses the ionization of the sample molecules in the
MS source. Both endogenous substances (such as carbohydrates, amines, urea, lipids,
peptides, …) present in the sample and the final extract, as exogenous substances can
contribute to ion suppression. This phenomenon affects many aspects of the method
performances [5].
Ion suppression is the result of competition between non-volatile matrix components and
analyte ions in the charged droplet formation process or the droplet evaporation process. The
presence of non-volatile or less volatile solutes cause a change in the spray droplet solution
properties. The interfering compounds can increase the viscosity and the surface tension of
the droplet, and reduce the capability of the analytes to reach the gas phase. The co-
precipitation of the analytes with non-volatile material can also limit their transfer into the gas
phase [2-3,5,15-16]. The mass of individual analytes are also factors influencing ion
suppression. It has been shown that molecules with higher mass will suppress the signal of
smaller molecules and that polar analytes are more susceptible to suppression. Regarding the
ionization polarity, the negative ion mode is usually considered as more specific, and
consequently less subjected to ion suppression [5,15,17].
The importance of ion suppression on the reliability of LC-MSn has been shown in terms of
accuracy and precision. Ion suppression may lead to false compliant results due to the non-
detection of an existing analyte, the underestimation of the real concentration, or the non-
fulfillment of the identification criteria. If the internal standard is affected rather than the
analyte, ion suppression will lead to an overestimation of the analyte concentration with
increased risk of false non-compliant results for MRL compounds [5].
The two main techniques used to determine the degree of ion suppression of a LC-MSn
method are post-extraction addition and post-column infusion. The post-extraction addition
technique can be considered a static technique that only provides information about matrix
effects at the point of elution of the analyte of interest. A more dynamic technique for
determining ion suppression is post-column infusion [2].
44
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
To overcome ion suppression different actions can be taken. As mentioned before the mass
spectrometric conditions influence ion suppression; different ionization techniques, ionization
modes and equipment need to be evaluated. Another solution is the use of a stable isotope
internal standard, so that the ion suppression is identical for the analyte and internal standard.
But the only way to definitively overcome this problem is to improve the sample preparation
and purification in order to limit the presence of interfering compounds in the final extract
[2,5,15].
2.2.1.2. Molecular Imprinting
Clean-up of beta-agonists in biological samples
The determination of beta-agonists in biological samples is a difficult analytical task because
of the low concentrations of the drugs and the complexity of the matrices. Historically, liquid-
liquid extraction has been the preferred technique for clean-up of biological samples. These
extractions resulted in relatively clean extracts with good recoveries, but were also time
consuming and the solvents used have often involved environmental and health hazards. In
recent years, solid-phase extraction (SPE) and immuno-affinity chromatography (IAC) have
become the methods of choice. SPE is cheap, quite fast, gives good recoveries, but does not
provide the selectivity needed for very clean extracts. Selectivity is achieved by using IAC,
but this technique is expensive, often time-consuming and has to be performed under very
specific conditions to keep the affinity sites intact [9,18].
A possible solution is Molecular Imprinted Polymers (MIPs). MIP material can be used as a
sorbent in SPE, where it offers a highly selective binding of the analyte compared with silica-
based and other polymeric sorbents. The advantage of MIPs compared to IAC phases which
have naturally produced antibodies, is the superior stability. These polymers can withstand a
large pH range and organic solvents without losing their recognition properties. Furthermore,
they are faster and cheaper to produce and no animals are involved for antibody-antigen
production [18-20].
Theoretical background
Although molecular imprinted polymers have become only recently commercial available, the
concept of molecular imprinting has a long history. The methodology, as originally developed
during the 1980’s, involves three principal phases: pre-arrangement, polymerisation and
extraction of the template (Fig. 1).
45
Printed: www.dclsigns.be
Beta-agonists
Fig. 1 The three phases of MIP preparation
MIPs are cross-linked polymers prepared in the presence of a template molecule. The
template molecule may be a particular analyte or drug molecule, or an analogue of it.
Functional monomers interact with the template molecule during polymerisation and the
template is removed from the polymer afterwards. The cavities thus created in the polymer are
complementary to the template both in shape and in chemical positioning of functional
groups. These multiple interaction sites lead to cavities with highly selective binding affinity
[18-22].
MIP4SPE beta-agonists
The template, acidic monomers, difunctional acrylic cross-linker, initiator and porogen
(solvent to be used in conjugation with the polymerisation process) were mixed together. The
template used had the common structure of beta-agonists. After polymerisation, the polymer
was milled, sieved and washed extensively in several steps to minimise bleeding of the
template. The selective cavities formed in the polymer contain acidic functional monomers
that interact with the -OH and -NH groups of the different beta-agonists by forming hydrogen
bounds. This illustrates the importance of the template having these functional groups at the
same positions as the beta-agonists.
In addition to the highly selective interactions of beta-agonists with the imprinted cavities,
MIP sorbents also undergo non-selective adsorptions with both analytes and matrix
components. Non-selective interactions should be eliminated during the washing steps to
enhance the selectivity of the procedure.
46
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
The selective interaction between beta-agonists and the sterically-orientated carboxylic
moieties, occurs in acetonitrile allowing strong hydrogen bonding. By adding small amounts
of acetic acid to acetonitrile, non-selective binding is restrained. The presence of electron-rich
groups close to the binding sites will decrease the strength of the hydrogen bonds between the
analyte and polymer, leading to reduced recoveries [19-20].
MIP4SPE procedure
Conditioning consists of wetting the MIP using methanol and water and then adjusting the pH
to 6.7 so that the acidic monomers are in a negatively charged state for ionic bonding.
During sample loading, beta-agonists are non-selectively retained together with substances
from the urine matrix. In a water environment, not all interactions between the selective
cavities of MIP and beta-agonists are established; bonding not only occurs in the selective
cavities, but all over the polymer.
Using undiluted urine as a sample, the method suffered from low recoveries and variations
were too large. A probable explanation is that the high ionic strength of the calf urine reduces
the amount of beta-agonists adsorbed to the polymer during sample loading. Therefore,
various dilutions were tested. The lowest dilution which increased the recovery and reduced
the variation, was 1:1 urine/water.
After sample loading the MIP was washed with water to elute salts and matrix components
that were not bonded or absorbed. Interactions between the selective cavities of MIP and beta-
agonists take place in an acetonitrile environment. This was achieved by using a selective
wash of acetonitrile-acetic acid (99:1); the addition of a small amount of acetic acid to
acetonitrile restrained non-selective binding. When the water content on the polymer is too
high, there can be a loss of beta-agonists. Therefore a few minutes of vacuum was necessary
to semi-dry the MIP before the selective wash. Buffers of various concentrations and pH
values were tested for the elution of interferences that were ion and/or hydrogen bonded, but
not selectively bonded. No differences were observed between the pH values. The use of a 50
mM buffer resulted in cleaner extracts compared with buffers of lower concentrations. In
search of a solution that could break both hydrophobic and hydrogen bonds, acetonitrile was
mixed with water. When the water content was 40 % the last of the interferences was eluted.
For the elution of beta-agonists, methanol was mixed with acids. When the concentration of
the acid was too high, some beta-agonists were degraded during evaporation. As a
47
Printed: www.dclsigns.be
Beta-agonists
compromise between elution strength and degradation, methanol with 10 % acetic acid was
chosen for the elution of beta-agonists [18].
2.2.2. Method setup
Previous experiments revealed that sample clean-up using MIP extraction is well suited for
bioanalysis at trace levels and that the resulting methods can be robust with good precision
and accuracy [18, 20-21, 23]. Fiori et al. already compared two different clean-up steps
involving SPE using non-endcapped C18 and Molecular Imprinted Polymers. The mechanism
of C18 SPE columns is based on the hydrophobic behaviour of the columns, and therefore can
be used for clean-up of a wide range of compounds; better recoveries were observed using
MIP columns [23]. Since mixed phase SPE columns (e.g. Clean Screen Dau) are more
selective than C18 columns and are used more frequently nowadays for the routine analysis of
beta-agonists, it was useful to evaluate these two different clean-up steps, Clean Screen Dau
and Molecular Imprinted Polymers, with respect to their ability to minimise ion suppression
in liquid chromatography-tandem mass spectrometry.
2.2.3. Experimental
2.2.3.1. Reagents and chemicals
The beta-agonist standards salbutamol, clenbuterol, isoxsuprine, fenoterol and tulobuterol
were obtained from Sigma-Aldrich (St Louis, MO, USA), while cimaterol, mabuterol,
brombuterol, terbutaline, hydroxymethyl clenbuterol, cimbuterol, mapenterol, ractopamine
and clenproperol were from the EU Reference Laboratory for Residues of Veterinary Drugs
(Berlin, Germany) and zilpaterol was a gift from Intervet (Schwabenheim, Germany). The
internal standard, deuterated clenbuterol, was obtained from RIVM (Bilthoven, The
Netherlands) (Fig. 2). Clean screen dau columns were from UCT technologies (Bristol, PA,
USA), and MIP4SPE beta-agonist columns were from MIP technologies (Lund, Sweden). The
enzymatic deconjugation was performed with β-glucuronidase from Sigma-Aldrich (St Louis,
MO, USA). All chemicals used were of analytical grade from Merck (Darmstadt, Germany)
and Acros (Geel, Belgium).
Stock standard solutions of 1000 ng μl-1 were prepared in ethanol. For the preparation of
working solutions, methanol was used as diluent. All standard and working solutions were
stored at -20 °C.
48
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
OH
HN
R6
R1
R2
R3
R4
HR5
R1 R2 R3 R4 R5 R6
Cimaterol H CN NH2 H H CH(CH3)2
Salbutamol H CH2OH OH H H C(CH3)3
Terbutaline H OH H OH H C(CH3)3
Clenproperol H Cl NH2 Cl H CH(CH3)2
Ractopamine H H OH H H CH(CH3)-(CH2)2-PhOH
Clenbuterol H Cl NH2 Cl H C(CH3)3
Tulobuterol Cl H H H H C(CH3)3
Mabuterol H Cl NH2 CF3 H C(CH3)3
Brombuterol H Br NH2 Br H C(CH3)3
Isoxsuprine H H OH H CH3 CH(CH3)-CH2-O-Ph
Cimbuterol H CN NH2 H H C(CH3)3
Fenoterol H OH H OH H CH(CH3)-CH2-PhOH
Hydroxymethyl
clenbuterol
H Cl NH2 Cl H C(CH3)2-CH2-OH
Mapenterol H Cl NH2 CF3 H C(CH3)2CH2CH3
HN
N
O
OH
NH
CH3
CH3
zilpaterol
Fig. 2 Chemical structures of the beta-agonists considered
49
Printed: www.dclsigns.be
Beta-agonists
2.2.3.2. Instrumentation
The HPLC apparatus comprised of a P4000 quaternary pump and an AS3000 autosampler
(ThermoFinnigan, San José, CA, USA). Chromatographic separation was achieved using a
Alltima HP C18 column (5 µm, 150 x 2.1 mm, Alltech, Deerfield, Illinois, USA). The mobile
phase consisted of a mixture of methanol (A) and water with 5 mM pentafluoropropionic acid
(PFPA) (B). A linear gradient was run (20 % A for 5 min, increasing to 35 % A over 10 min,
and finally increasing to 100% A in the next 3 min) at a flow rate of 0.3 ml min-1.
LC-MSn detection used a LCQ Deca ion trap (ThermoFinnigan, San José, CA, USA) with an
electrospray ionisation (ESI) interface operating in positive ion mode. The instrument
parameters are summarised in Table 1. The precursor isolation width was set to 2 Da for each
beta-agonist. Each analyte was identified on the basis of at least two product ions present in
the MS2 or MS3 spectra (Table 2).
50
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
Table 1 Instrumental method for the detection of beta-agonists in urine samples
Segments Scan events Precursor ion → product ion
Mass range
Analyte
Segment 1 Scan event 1 262.0 → 244.0; 100-270 Zilpaterol
0 – 6.2 min Scan event 2 220.0 → 202.0; 100-230 Cimaterol
Scan event 3 240.0 → 222.0; 100-250 Salbutamol
Scan event 4 226.0 → 170.0; 100-230 Terbutaline
Scan event 5 234.0 → 216.0; 100-240 Cimbuterol
Segment 2 Scan event 1 263.0 → 245.0; 100-270 Clenproperol
6.2 – 13.5 min Scan event 2 302.0; 100-310 Ractopamine
Scan event 3 277.0 → 259.0; 100-280 Clenbuterol
Scan event 4 283.0; 100-290 Clenbuterol-d6 (I.S.)
Scan event 5 304.0; 100-310 Fenoterol
Scan event 6 293.0 → 275.0; 100-300 Hydroxymethyl
clenbuterol
Segment 3 Scan event 1 228.0; 100-230 Tulobuterol
13.5 – 22 min Scan event 2 311.0 → 293.0; 100-320 Mabuterol
Scan event 3 367.0 → 349.0; 100-370 Brombuterol
Scan event 4 302.0 → 284.0; 100-310 Isoxsuprine
Scan event 5 325.0; 100-330 Mapenterol
51
Printed: www.dclsigns.be
Beta-agonists
Table 2 The precursor and product ions (m/z) used for the evaluation of the beta-agonists
Analyte Precursor ion MS2 first generation
product ions
MS3 second generation
product ions
Zilpaterol 262 244 185 202
Cimaterol 220 202 160
Salbutamol 240 222 148 166
Terbutaline 226 170 152
Clenproperol 263 245 203
Tulobuterol 228 154 172 210
Ractopamine 302 164 284
Clenbuterol 277 259 203
Mabuterol 311 293 237
Brombuterol 367 349 293
Isoxsuprine 302 284 107 135 150 190
Hydroxymethyl
clenbuterol
293 275 203
Fenoterol 304 135 286
Cimbuterol 234 216 160
Mapenterol 325 237 307
2.2.3.3. Extraction and clean-up
Clean Screen Dau SPE: To blank calf urine samples (5 ml) a spike solution of beta-agonists
at their MRPL concentration (Table 3) and 1 µg l-1 clenbuterol-d6 (internal standard) was
added. After addition of 2.5 ml 0.2 M acetate buffer (pH 4.6) the pH of the mixture was
adjusted to 4.6 and the sample was centrifuged (8000 rpm, 15 min, 5 °C). The clean up was
performed using a 500 mg Clean Screen Dau (CSD) (mixed C8 and SCX) SPE column. The
column was conditioned with 2 ml methanol, 2 ml water and 2 ml 0.1 mol l-1 phosphate buffer
(pH 6). After application of the extract, the cartridge was washed first with 1 ml 1 mol l-1
acetic acid, vacuum dried, and subsequently washed with 6 ml methanol and vacuum dried
again. Elution used 6 ml ethylacetate containing concentrated ammonia (97:3). The eluate was
evaporated to dryness at 60 °C under a stream of nitrogen. The residue was reconstituted in 30
µl methanol and 90 µl 5 mM PFPA solution, before injecting 30 µl on the HPLC column.
52
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
Molecular Imprinted Polymer SPE: To blank calf urine samples (5 ml) a spike solution of
beta-agonists at their MRPL concentration (Table 3) and 1 µg l-1 clenbuterol-d6 was added.
The urine samples were first 1:1 diluted with water and centrifuged at 9000 rpm for 10
minutes. The clean-up was performed using a 25 mg MIP4SPE (beta-agonist) SPE column.
The column was conditioned with 1 ml methanol, 1 ml water and 1 ml 25 mM ammonium
acetate buffer (pH 6.7). After application of the extract, the cartridge was washed with 1 ml
water and vacuum dried, and subsequently with 1 ml 1 % acetic acid in acetonitrile, 1 ml 50
mM ammonium acetate buffer and 1 ml 60 % acetonitrile in water. Elution used 2 x 1 ml 10
% acetic acid in methanol, applying gentle vacuum between the two fractions. The flow rate
was 0.5 ml min-1, except for the analyte elution a lower flow rate was applied. The eluate was
evaporated to dryness at 60 °C under a stream of nitrogen. The residue was reconstituted in 30
µl methanol and 90 µl 5 mM PFPA, before injecting 30 µl on the HPLC column.
Table 3 MRPL values for beta-agonists in urine proposed by the EU Reference Laboratory for
Residues of Veterinary Drugs (Berlin).
Analyte Proposed MRPL (µg l-1)
Zilpaterol 1
Cimaterol 3
Salbutamol 3
Terbutaline 3
Clenproperol 3
Tulobuterol 1
Ractopamine 3
Clenbuterol 1
Mabuterol 1
Brombuterol 1
Isoxsuprine 3
Cimbuterol 3
Mapenterol 1
Fenoterol 3
Hydroxymethyl clenbuterol 1
53
Printed: www.dclsigns.be
Beta-agonists
2.2.4. Results and discussion
2.2.4.1. LC-MSn method
A multi-residue LC-MSn method was developed for the qualitative analysis of fifteen beta-
agonists (Fig. 2) in urine. The beta-agonists were spiked into blank calf urine at their MRPL
concentrations. Fig. 3 shows the extracted ion chromatograms for the beta-agonists after
clean-up with CSD columns and without enzymatic hydrolysis. All the beta-agonists could be
detected at the MRPL level, but the signals for zilpaterol and terbutaline were weak and
subjected to significant interferences (low signal-to-noise ratio). Fig. 4 shows the extracted
ion chromatograms for the different beta-agonists after clean-up with MIP columns and
without hydrolysis. All the beta-agonists could be detected at the MRPL level according to
the 2002/657/EC decision criteria [9]. Recoveries for the different beta-agonists using MIP
clean-up are in the range 40 – 70 %, except for zilpaterol, salbutamol and terbutaline which
have recoveries below 40 %.
The beta-agonists could also be detected with signals of the same order-of-magnitude as in
Figs. 3 and 4 after hydrolysis with glucuronidase at 50 °C for 2 hours. The aim of this work
was to compare the effectiveness of the clean-up using CSD with that using MIP SPE, with
respect to removal of interfering compounds and reduction of ion suppression. Hydrolysis did
not seem to interfere with the analyte signals, there was no suppression or enhancement of the
signals as the result of this hydrolysis (these data are not presented here). Therefore, the
subsequent experiments concerning ion suppression were performed without hydrolysis. Of
course, for real samples from animals dosed with beta-agonists, in which the analytes could
be conjugated and for which quantitative analyses are required, the effect of enzymatic
hydrolysis would have to be examined in detail.
54
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
RT: 2,39 - 16,31 SM: 7B
3 4 5 6 7 8 9 10 11 12 13 14 15 16Time (min)
0
50
1000
50
1000
50
1000
50
1000
50
1000
50
1000
50
1000
50
1002,46
2,57 3,702,67 4,10 4,33 6,47 7,016,10 7,38 8,015,26 8,38 8,85
3,18
6,27
3,45 3,67 4,45 5,23 6,77 8,147,34 8,502,963,75
2,523,01 4,80 5,15 6,36 8,395,56 8,12 8,846,97
3,59 3,62
3,793,98 4,78 5,30 5,67 6,24 6,612,84 7,16 8,20 8,84
10,42
10,72 13,5511,02 12,09 13,03 13,869,9, 562311,21
12,3627 12,789, 11,0110,38 13,8712,40
13,7012,7210, 10,36 969,42 13,9712,129,1515,12
15,7714,45
NL: 8,86E5m/z= 243,5-244,5 F: + c ESI Full ms2 262,00@24,00 [ 100,00-270,00] MS 050107s06
NL: 1,63E6m/z= 221,5-222,5 F: + c ESI Full ms2 240,00@23,00 [ 100,00-250,00] MS 050107s06
NL: 2,87E5m/z= 169,5-170,5 F: + c ESI Full ms2 226,00@30,00 [ 100,00-250,00] MS 050107s06
NL: 6,99E6m/z= 244,5-245,5 F: + c ESI Full ms2 263,00@22,00 [ 100,00-270,00] MS 050107s06
NL: 2,34E6m/z= 164,5-165,5+283,5-284,5 F: + c ESI Full ms2 302,00@25,00 [ 100,00-310,00] MS 050107s06
NL: 3,94E6m/z= 258,5-259,5 F: + c ESI Full ms2 277,00@23,00 [ 100,00-280,00] MS 050107s06
NL: 8,82E6m/z= 153,5-154,5+171,5-172,5+209,5-210,5 F: + c ESI Full ms2 228,00@27,00 [ 100,00-230,00] MS 050107s06
zilpaterolNL: 4,71E6m/z= 201,5-202,5 F: + c ESI Full ms2 220,00@24,00 [ 100,00-230,00] MS 050107s06cimaterol
salbutamol
terbutaline
clenproperol
ractopamine
clenbuterol
tulobuterol
RT: 4,19 - 21,28 SM: 7B
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21Time (min)
0
50
1000
50
1000
50
1000
50
1000
50
1000
50
1000
50
10016,02
21,0620,8517,28 17,71 18,73 19,61
NL: 5,80E6m/z= 292,5-293,5 F: + c ESI Full ms2 311,00@23,00 [ 100,00-320,00] MS 050107s06
NL: 1,17E7m/z= 283,5-284,5 F: + c ESI Full ms2 302,00@30,00 [ 100,00-310,00] MS 050107s06
NL: 1,15E7m/z= 215,5-216,5 F: + c ESI Full ms2 234,00@24,00 [ 100,00-240,00] MS 050216s07
NL: 7,42E6m/z= 134,5-135,5+285,5-286,5 F: + c ESI Full ms2 304,00@28,00 [ 100,00-320,00] MS 050216s07
NL: 3,23E6m/z= 274,5-275,5 F: + c ESI Full ms2 293,00@23,00 [ 100,00-300,00] MS 050216s07
NL: 5,57E7m/z= 236,5-237,5+306,5-307,5 F: + c ESI Full ms2 325,00@24,00 [ 100,00-330,00] MS 050216s07
14,8114,41mabuterol
NL: 2,55E6m/z= 348,5-349,5 F: + c ESI Full ms2 367,00@21,00 [ 100,00-370,00] MS 050107s06
15,20
15,73 21,1316,51 20,4819,1917,5014,74 18,47brombuterol 17,43
20,9120,0618,16 18,8016,7315,8215,2614,52isoxsuprine
5,28
13,95
11,3410,057,135,81
cimbuterol 7,26 8,534,28 9,60 13,0812,44
10,10
10,40
12,936,57 10,639,64 11,887,026,154,28 9,18 13,44 14,624,85 8,377,42fenoterol
10,26
10,5610,099,71 11,08 11,95 13,9812,875,36 7,21 8,784,27 14,467,765,73 6,45hydroxymethyl clenbuterol
20,55
2720,17,32 19,0416,00 18,09
mapenterol 15,35
Fig. 3 Ion chromatograms of the different beta-agonists at MRPL concentration in calf urine
using CSD clean-up
55
Printed: www.dclsigns.be
Beta-agonists
RT: 2,34 - 16,71 SM: 7B
3 4 5 6 7 8 9 10 11 12 13 14 15 16Time (min)
0
50
1000
50
1000
50
1000
50
1000
50
1000
50
1000
50
1000
50
1003,85
7,372,72 7,485,335,06 5,62 8,886,27
3,383,29
3,98 6,074,96 5,36 6,74 7,957,642,45 8,29 8,913,92
3,20 5,12 7,266,542,46 5,61 8,404,63 7,72 8,763,78 3,85
4,43 4,89 8,025,54 5,95 8,547,657,13 8,902,46 3,3710,52
11,13 11,55 12,92 13,44 14,059, 9,9, 33 940511,32
12,4437 13,3967639, 10,9,9,10 14,0012,48
13,04 13,7713364110 11,10,9,9, 11,9815,22
15,9914,41
NL: 2,26E5m/z= 243,5-244,5 F: + c ESI Full ms2 262,00@24,00 [ 100,00-270,00] MS 050107s05
NL: 1,38E6m/z= 221,5-222,5 F: + c ESI Full ms2 240,00@23,00 [ 100,00-250,00] MS 050107s05
NL: 4,43E5m/z= 169,5-170,5 F: + c ESI Full ms2 226,00@30,00 [ 100,00-250,00] MS 050107s05
NL: 1,21E7m/z= 244,5-245,5 F: + c ESI Full ms2 263,00@22,00 [ 100,00-270,00] MS 050107s05
NL: 6,11E6m/z= 163,5-164,5+283,5-284,5 F: + c ESI Full ms2 302,00@25,00 [ 100,00-310,00] MS 050107s05
NL: 7,09E6m/z= 258,5-259,5 F: + c ESI Full ms2 277,00@23,00 [ 100,00-280,00] MS 050107s05
NL: 1,89E7m/z= 153,5-154,5+171,5-172,5+209,5-210,5 F: + c ESI Full ms2 228,00@27,00 [ 100,00-230,00] MS 050107s05
zilpaterol NL: 6,73E6m/z= 201,5-202,5 F: + c ESI Full ms2 220,00@24,00 [ 100,00-230,00] MS 050107s05
cimaterol
salbutamol
terbutaline
clenproperol
ractopamine
clenbuterol
tulobuterol
RT: 3,98 - 21,17 SM: 7B
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21Time (min)
0
50
1000
50
1000
50
1000
50
1000
50
1000
50
1000
50
10016,14
16,71 17,40 18,18 18,93 21,0719,6815,6214,9815,32
16,55 16,90 17,71 20,9218,96 19,36
NL: 9,46E6m/z= 292,5-293,5 F: + c ESI Full ms2 311,00@23,00 [ 100,00-320,00] MS 050107s05
NL: 2,04E7m/z= 283,5-284,5 F: + c ESI Full ms2 302,00@30,00 [ 100,00-310,00] MS 050107s05
NL: 2,58E7m/z= 215,5-216,5 F: + c ESI Full ms2 234,00@24,00 [ 100,00-240,00] MS 050228s21
NL: 8,98E5m/z= 134,5-135,5+285,5-286,5 F: + c ESI Full ms2 304,00@28,00 [ 100,00-320,00] MS 050228s21
NL: 9,01E6m/z= 274,5-275,5 F: + c ESI Full ms2 293,00@23,00 [ 100,00-300,00] MS 050228s21
NL: 1,08E8m/z= 236,5-237,5+306,5-307,5 F: + c ESI Full ms2 325,00@24,00 [ 100,00-330,00] MS 050228s21
mabuterol NL: 7,66E6m/z= 348,5-349,5 F: + c ESI Full ms2 367,00@21,00 [ 100,00-370,00] MS 050107s05
14,45brombuterol
17,63
18,46
isoxsuprine 14,54 16,01 20,9414,94 16,99 19,05 20,17
5,18
5,78
cimbuterol 6,43 6,93 13,707,49 8,224,23 8,90 10,639,96 11,49 13,24 14,5512,44
6,74
8,70 11,2310,2810,036,34 14,248,257,49 13,625,66 12,504,42
fenoterol 14,58
10,21
9,398,09 11,23 12,95 13,656,47 6,92 14,9812,029,075,174,40 6,0720,44
hydroxymethyl clenbuterol
19,8515,44 18,8018,4216,8015,05
mapenterol
Fig. 4 Ion chromatograms of the different beta-agonists at MRPL concentration in calf urine
using MIP clean-up
56
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
2.2.4.2. Ion suppression
The main analytical problems encountered in LC-MSn arise from matrix effects, and in
particular involve ion suppression. This phenomenon affects many aspects of the method
performance such as detection capability, repeatability and accuracy. The cause of ion
suppression is a change in the spray droplet solution properties arising from the presence of
co-eluting non-volatile or less volatile solutes. Polar compounds such as beta-agonists seem to
be particularly susceptible to ion suppression. The positive ionization mode is usually
considered as less specific, and consequently more subjected to ion suppression [5, 15].
The typical experimental system used to evaluate ion suppression in LC-MSn is depicted in
Fig. 5 [15]. Either clean mobile phase or real samples are injected into the LC system. A
standard solution containing the analyte of interest is continuously infused through a T-
coupling system, mixed with the LC eluate, and passed into the mass spectrometer interface.
The resulting signal recorded by the mass spectrometer is the net result of these two solutions.
Because the analyte is introduced into the mass detector at a constant rate, a constant ESI
response should ideally be observed. This is the case when pure mobile phase is injected into
the LC. When blank urine is injected into the LC system, the resulting total-ion-current
increases due to the new material arriving in the interface, and the product signal of the
analyte decreases in certain retention time regions as a result of the negative influence of
interfering compounds eluting at these retention times [5].
HPLC mobile phase
blank urine
Column ESI
interface
Mass
spectrometer
Syringe pump standard solution
Fig. 5 Postcolumn infusion system
This experiment was performed for the beta-agonists zilpaterol, cimaterol, salbutamol and
terbutaline. These beta-agonists elute around the same retention time, the signals for zilpaterol
and terbutaline were weak after clean-up with CSD and the recoveries were low after clean-up
with MIP. First the standard solution and a spiked urine sample at the MRPL concentration
were injected to obtain the retention time of the analytes. Subsequently, pure mobile phase
was injected while the analyte was continuously infused. Finally, to evaluate ion suppression
57
Printed: www.dclsigns.be
Beta-agonists
blank urine was injected while the analyte was infused. Fig. 6 shows the data obtained by the
injection of blank urine extracts obtained after clean-up with CSD and after clean-up with
MIP while continuously infusing zilpaterol and cimaterol. After clean-up with CSD no
significant suppression was observed for the product signal of cimaterol near its expected
retention time (RT = 3.1 min). However, severe ion suppression appeared for zilpaterol (RT =
3.6 min), i.e., in the time window in which zilpaterol elutes there was a serious decrease of the
zilpaterol signal due to the interfering compounds that also eluted in this retention time
window. In contrast, after clean-up with MIP there was no significant suppression of the
signals for either zilpaterol or cimaterol in the time windows in which each analyte elutes.
Similar MIP results were obtained for salbutamol and terbutaline. However, after clean-up
with CSD there was severe suppression of the signal for terbutaline near its expected retention
time; no significant suppression was observed for the signal for salbutamol at its retention
time.
This experiment shows that CSD sample clean-up could lead to underestimation of the
concentrations of some beta-agonists and could lead to a potential risk of false compliant
results. A possible solution to overcome false compliant results is the use of an adequate
internal standard, preferably an isotope-labelled internal standard, in order to correct for the
ion suppression effect [5,15]. Of course, this is only possible when the ion signal is not
suppressed completely. Since different beta-agonists are suppressed, different adequate
internal standards are necessary which can in turn lead to analytical problems concerning
sensitivity of the multi-residue method. Also, purchase of multiple isotope-labeled internal
standards can be expensive for a purely qualitative method.
58
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
RT: 0,16 - 8,92 SM: 7B
0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5Time (min)
0
100000
200000
300000
400000
500000
600000
700000
800000
900000
1000000
1100000
1200000
1300000
1400000
1500000
1600000
1700000
1800000
1900000
2000000
2100000
2200000
2300000
2400000
2500000
2600000
2700000
2800000
2900000
3000000
3100000
3200000
Inte
nsity
2,77 2,80
2,59
4,233,11
4,294,133,233,30
2,42 6,994,04 4,784,46 4,96 5,765,09 5,87 7,086,635,29 8,676,59 7,14 8,593,983,88 7,93 7,99
3,72 7,80
0,21
1,18
0,72
2,091,96
NL: 3,25E6m/z= 243,5-244,5 F: + c ESI Full ms2 262,00@24,00 [ 100,00-270,00] MS 050217s07
RT: 0,00 - 7,25 SM: 7B
0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0Time (min)
0
500000
1000000
1500000
2000000
2500000
3000000
3500000
4000000
4500000
5000000
5500000
6000000
6500000
7000000
7500000
8000000
8500000
9000000
9500000
10000000
Inte
nsity
1,42
4,546,064,50
4,59 5,820,22 3,24
3,21 5,013,09 4,380,64 1,261,14 4,24
5,06 5,696,906,86
6,816,32 6,77
6,985,48
3,39 3,962,85
2,43
2,31
2,17
NL: 1,00E7m/z= 201,5-202,5 F: + c ESI Full ms2 220,00@24,00 [ 100,00-230,00] MS 050217s09
Clean Screen Dau
zilpaterol (1) cimaterol (2)
RT: 0,00 - 9,09 SM: 7B
0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 8,5 9,0Time (min)
0
100000
200000
300000
400000
500000
600000
700000
800000
900000
1000000
1100000
1200000
1300000
1400000
1500000
1600000
1700000
1800000
1900000
2000000
2100000
2200000
2300000
2400000
2500000
2600000
2700000
2800000
2900000
3000000
3100000
3200000
Inte
nsity
2,892,761,20
2,725,53 8,445,64 8,496,954,170,14 0,24 5,744,40 6,905,894,48 7,413,970,81 6,30 7,614,63 5,00
3,11 3,47
1,682,39
2,34
NL: 3,25E6m/z= 243,5-244,5 F: + c ESI Full ms2 262,00@24,00 [ 100,00-270,00] MS 050211s06
RT: 0,21 - 7,09 SM: 7B
0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0Time (min)
0
500000
1000000
1500000
2000000
2500000
3000000
3500000
4000000
4500000
5000000
5500000
6000000
6500000
7000000
7500000
8000000
8500000
9000000
9500000
10000000
Inte
nsity
4,29
4,573,433,36
0,273,29
1,17 1,21 3,230,49 3,571,073,04
2,93 6,475,984,92 6,383,972,82 5,914,99
5,42 5,46 6,57
5,32 5,51
6,94
2,55
1,82 1,88
1,99
NL: 1,00E7m/z= 201,5-202,5 F: + c ESI Full ms2 220,00@24,00 [ 100,00-230,00] MS 050211s08
Molecular Imprinted Polymers
zilpaterol (3) cimaterol (4)
Fig. 6 MS/MS signals for zilpaterol (1&3) and cimaterol (2&4) detected using the apparatus
shown in Figure 5, using continuous infusion of zilpaterol or cimaterol and LC injection of
blank urine samples after clean-up using either CSD (1&2) or MIP (3&4).
The percentages of ion suppression for the different beta-agonists are reported in Table 4 as
percentages of the expected signal. They were calculated by analysing five post-extraction
spiked samples and five pure standards, and calculating the ratio between the two values. The
concentrations of the beta-agonists added to the blank urine samples were equal to those
present in the standard solution. If the signal is not suppressed, the percentage of the expected
signal is 100 %. The values obtained experimentally (Table 4) indicate that clean-up using
59
Printed: www.dclsigns.be
Beta-agonists
MIP columns is more selective than that using CSD columns for many beta-agonists
(zilpaterol, terbutaline, ractopamine, clenbuterol, brombuterol, isoxsuprine, cimbuterol,
mapenterol and hydroxymethyl clenbuterol). Only the beta-agonists clenproperol and
fenoterol gave a higher percentage of the expected signal after clean-up with CSD, but even in
those cases the percentages after clean-up with MIP are satisfactory (≥ 75 %).
Based on the information received from MIP Technologies, the manufacturers of the MIP
columns, there should be no ion suppression from the template; the template bleeding level is
normally just a few ng/ml. However, the manufacturer did not reveal the nature of the
template. A blank water sample was processed by the MIP method and analysed in full scan
mode to check for template bleeding, but no clear chromatographic peak or signal was
obtained.
Table 4 Percentage of the expected signal as an indicator of the percentage of ion suppression
Analyte % of the expected signal
CSD MIP
Zilpaterol 49 ± 0.63 89 ± 2.40
Cimaterol 112 ± 0.50 111 ± 9.65
Salbutamol 107 ± 0.37 109 ± 4.70
Terbutaline 47 ± 1.18 88 ± 2.18
Clenproperol 123 ± 0.72 75 ± 1.39
Tulobuterol 136 ± 0.80 101 ± 4.59
Ractopamine 40 ± 0.20 80 ± 3.38
Clenbuterol 56 ± 0.34 76 ± 1.62
Mabuterol 85 ± 0.61 107 ± 3.15
Brombuterol 46 ± 1.25 107 ± 4.04
Isoxsuprine 61 ± 1.00 104 ± 1.65
Cimbuterol 55 ± 0.17 102 ± 3.75
Mapenterol 51 ± 0.14 72 ± 0.58
Fenoterol 100 ± 0.95 77 ± 1.29
Hydroxymethyl clenbuterol 53 ± 0.14 81 ± 2.07
These experiments indicate that sample clean-up using MIP columns is more selective than
that using CSD columns. As a result, sample clean-up can influence the repeatability of a
method in routine analysis when different samples are run within the same session since the
60
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
co-eluting interferences are not necessarily reproducible. Consequently, ion suppression
experiments should be performed during method development to prevent problems regarding
false compliant results and problems regarding the repeatability.
2.2.4.3. Qualitative validation
The multi-residue method using MIP SPE columns presented in this paper is only a
qualitative method. The following qualitative validation parameters were tested: specificity,
selectivity, decision limit (CCα) and detection capability (CCβ).
The specificity of the method was demonstrated by LC-MS2 and LC-MS3 analyses of blank
urine (at least 20 blank urine samples were analysed); no interferences were observed in
analysis of these blank samples and in analysis of urine spiked with the different beta-
agonists.
The minimum number of identification points (IP) for beta-agonists is set to four. Table 2
shows the MS2 and MS3 product ions needed for the identification of each beta-agonist. Most
beta-agonists only have one MS2 product ion (so 2.5 IPs are earned), therefore MS3
fragmentation is necessary to obtain enough identification points.
The CCβ of each beta-agonist is equal to or lower than the MRPL concentrations, i.e., 1 µg
kg-1 for clenbuterol, brombuterol, hydroxymethylclenbuterol, mabuterol, mapenterol,
tulobuterol and zilpaterol, and 3 µg kg-1 for cimaterol, cimbuterol, clenproperol, isoxsuprine,
fenoterol, ractopamine, salbutamol and terbutaline. The CCα was calculated by subtracting
1.64 times the maximum standard deviation of the CCβ-value. For the calculation of CCα, the
maximum standard deviation was derived from the maximum coefficient of variation of 25 %
(CCα ≤ 0.59 µg kg-1 when CCβ ≤ 1 µg kg-1 and CCα ≤ 1,77 µg kg-1 when CCβ ≤ 3 µg kg-1).
Additional experiments will be necessary to obtain the standard deviation for each beta-
agonist.
For purposes of quantification, the clean-up needs to be optimised for the different beta-
agonists in order to be able to obtain reproducible results; also hydrolysis of the spiked and
real urine samples is necessary, and more than one internal standard should be added to the
method since some of the beta-agonists have rather different chemical structures.
61
Printed: www.dclsigns.be
Beta-agonists
2.2.5. Conclusion
A multi-residue method was developed for the detection of 15 beta-agonists in urine. Two
different SPE clean-up steps were evaluated, using either Clean Screen Dau or Molecular
Imprinted Polymers. Ion suppression experiments revealed that CSD sample clean-up could
lead to false compliant results for some beta-agonists; the percentages of the expected signal
actually observed show that there is less suppression of the signals when urine is pretreated
with MIP columns, i.e., clean-up using MIP columns is more selective than that using CSD
columns.
A qualitative validation was performed using MIP clean-up; at this point only a qualitative
determination of the different beta-agonists is possible. Before quantification can be done,
suitable internal standards need to be added to the method and the clean-up needs to be
optimised to obtain reproducible results.
This study has shown that molecular imprinted polymers are very promising for sample clean-
up for beta-agonists, but further research is necessary before they can be incorporated into
fully validated quantitative assays.
62
Printed: www.dclsigns.be
Multi-residue LC-MSn method of beta-agonists in urine using molecular imprinted polymers
2.2.6. References
[1] Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive
96/23/EC concerning the performance of analytical methods and the interpretation of results
(2002) Official Journal of the European Communities, no. L 221
[2] P.J. Taylor (2005) Matrix effect: The Achilles heel of quantitative high-performance
liquid chromatography-electrospray-tandem mass spectrometry, Clinical Biochemistry 38,
328-334
[3] B.K. Choi, D.M. Hercules and A.I. Gusev (2001) Effect of liquid chromatography
separation of complex matrices on liquid chromatography-tandem mass spectrometry signal
suppression, Journal of Chromatography A 907, 337-342
[4] M.D. Nelson and J.W. Dolan (2002) Ion suppression in LC-MS-MS – a case study, LCGC
North America 20
[5] J.P. Antignac, K. De Wasch, F. Monteau, H. De Brabander, F. Andre, B. Le Bizec (2005)
The ion suppression phenomenon in liquid chromatography-mass spectrometry and its
consequences in the field of residue analysis, Analytica Chimica Acta 529, 129-136
[6] M.P. Montrade, B. Le Bizec, F. Monteau, B. Siliart, F. André (1993) Multi-residue
analysis for beta-agonistic drugs in urine of meat-producing animals by gas-chromatography
mass-spectrometry, Analytica Chimica Acta 275, 253-268
[7] S. Collins, M. Okeeffe, M.R. Smyth (1994) Multi-residue analysis for beta-agonists in
urine and liver samples using mixed-phase columns with determination by
radioimmunoassay, Analyst 119, 2671-2674
[8] F. Ramos, M.C. Banobre, M.D. Castilho, M.I.N. Silveira (1999) Solid phase extraction
(SPE) for multi-residue analysis of beta(2)-agonists in bovine urine, Journal of Liquid
Chromatography and Related Technologies 22, 2307-2320
[9] F.J. dos Ramos (2000) Beta(2)-agonist extraction procedures for chromatographic
analysis, Journal of Chromatography A 880, 69-83
[10] C.S. Stachel, W. Radeck, P. Gowik (2003) Zilpaterol - a new focus of concern in residue
analysis, Analytica Chimica Acta 493, 63-67
[11] L.D. Williams, M.I. Churchwell, D.R. Doerge (2004) Multiresidue confirmation of beta-
agonists in bovine retina and liver using LC-ES/MS/MS, Journal of Chromatography B 813,
35-45
[12] A.A.M. Stolker, U.A.T. Brinkman (2005) Analytical strategies for residue analysis of
veterinary drugs and growth-promoting agents in food-producing animals – a review, Journal
of Chromatography A 1067, 15-53
63
Printed: www.dclsigns.be
Beta-agonists
[13] N. Van Hoof, R. Schilt, E. van der Vlis, P. Boshuis, M. van Baak, A. Draaijer, K. De
Wasch, M. Van de Wiele, J. Van Hende, D. Courtheyn, H. De Brabander (2005) Detection of
zilpaterol (Zilmax®) in calf urine and faeces with liquid chromatography-tandem mass
spectrometry, Analytica Chimica Acta 529, 189-197
[14] J. Blanca, P. Munoz, M. Morgado, N. Mendez, A. Aranda, T. Reuvers, H. Hooghuis
(2005) Determination of clenbuterol, ractopamine and zilpaterol in liver and urine by liquid
chromatography tandem mass spectrometry, Analytica Chimica Acta 529, 199-205
[15] T.M. Annesley (2003) Ion suppression in mass spectrometry, Clinical Chemistry 49,
1041-1044
[16] I. Fu, E.J. Woolf and B.K. Matuszewski (1998) Effect of the sample matrix on the
determination of indinavir in human urine by HPLC with turbo ion spray tandem mass
spectrometric detection, Journal of Pharmaceutical and Biomedical Analysis 18, 347-357
[17] P. Fürst (2000) LC-MS - a powerful tool in residue analysis of veterinary drugs,
Proceedings EuroResidue IV, Veldhoven, The Netherlands, 63-72
[18] A. Blomgren, C. Berggren, A. Holmberg, F. Larsson, B. Sellergren, K. Ensing (2002)
Extraction of clenbuterol from calf urines using a molecularly imprinted polymer followed by
quantitation by high-performance liquid chromatography with UV detection, Journal of
Chromatography A 975, 157-164
[19] www.miptechnologies.se
[20] C. Widstrand, F. Larsson , M. Fiori, C. Civitareale, S. Mirante, G. Brambilla (2004)
Evaluation of MISPE for the multi-residue extraction of B-agonists from calves urine, Journal
of Chromatography B 804, 85-91
[21] P.R. Kootstra, C.J.P.F. Kuijpers, K.L. Wubs, D. van Doorn, S.S. Sterk, L.A. van Ginkel,
R.W. Stephany (2005) The analysis of beta-agonists in bovine muscle using molecular
imprinted polymers with ion trap LCMS screening, Analytica Chimica Acta 529, 75-81
[22] M.P. Davies, V. De Biasi, D. Perrett (2004) Approaches to the rational design of
moleculary imprinted polymers, Analytica Chimica Acta 504, 7-14
[23] M. Fiori, C. Civitareale, S. Mirante, E. Magaro, G. Brambilla (2005) Evaluation of two
different clean-up steps, to minimise ion suppression phenomena in ion trap liquid
chromatography-tandem mass spectrometry for the multi-residue analysis of beta agonists in
calves urine, Analytica Chimica Acta 529, 207-210
64
Printed: www.dclsigns.be
Chapter 2.3
Excretion profile of zilpaterol in calf urine and faeces
Adapted from:
N. Van Hoof, R. Schilt, E. Van der Vlis, P. Boshuis, M. Van Baak, A. Draaijer, K. De Wasch, M. Van
de Wiele, J. Van Hende, D. Courtheyn and H. De Brabander
Detection of zilpaterol (Zilmax®) in calf urine and faeces with liquid chromatography–tandem mass
spectrometry
Analytica Chimica Acta (2005) 529, 189-197
2.3.1. Introduction
Zilpaterol (Fig. 2, Multi-residue LC-MSn method for the detection of beta-agonists in urine
using molecular imprinted polymers, paragraph 2.2.3.1) is a new powerful beta-agonist
developed as growth promoter for cattle. Zilmax® has been licensed as feed additive in
Mexico and South Africa. Its chemical structure is different from the well-known N-alkyl
beta-agonists (such as clenbuterol) as well as the di-aromatic beta-agonists (such as
ractopamine and isoxsuprine). Zilpaterol is capable of redirecting the cellular metabolism in
favour of protein synthesis. It is more effective than ractopamine, but only one-tenth effective
as clenbuterol [1]. Plascencia et al. (1999) performed an experiment with 140 steers on the
influence of zilpaterol on growth performance and carcass characteristics. The diet of the
steers was supplemented with 6 mg/kg zilpaterol (added as Zilmax®) and this during the last
42 days of the feeding period. Zilpaterol supplementation had a beneficial effect on the
growth performance, enhancing weight gain and feed efficiency. In addition, zilpaterol also
improved the carcass leanness [2].
65
Printed: www.dclsigns.be
Beta-agonists
Stachel et al. (2003) studied the residual behavior of zilpaterol by analyzing different matrices
after various withdrawal periods [1]. Previously, the manufacturer of Zilmax®, Hoechst
Roussel Vet, also studied the concentrations of zilpaterol in muscle, liver and kidney in a 50
day study on cattle. This study showed that a maximum concentration was achieved between
day 10 and day 30; after day 30 there was no additional accumulation in the different tissues.
Zilpaterol was mainly retained in liver and kidney and in a lower extent in muscle and fat. At
the end of the treatment, there was a fast decline in the concentration of zilpaterol recovered
from the different tissues. After 24 hours, the concentration of zilpaterol was halved and after
48 hours 80 % of the residues were eliminated [3].
Stachel et al. (2003) studied the behavior of zilpaterol in two animal species, porcine and
bovine species. In porcine tissues, liver and kidney, almost equal concentrations were found
after 1 day of withdrawal; after 4 days more than 90 % of zilpaterol was eliminated. In cattle
higher concentrations were found in both liver and kidney. These concentrations were in the
same order of magnitude as the values presented by the manufacturer of Zilmax®. The
concentration of zilpaterol in bovine and porcine muscle was much lower than the
concentrations in liver and kidney. Besides these tissues, also urine was analysed for the
presence of zilpaterol during a 14 day treatment. A constant increase of zilpaterol
concentration was observed for the first three days of treatment. Very high concentrations of
zilpaterol could be detected in urine. However, the concentration dropped to low values after
5 days of withdrawal [1].
2.3.2. Method setup
In this study, a LC–MS3 confirmatory method was developed for urine that was able to
identify simultaneously zilpaterol, ractopamine, isoxsuprine and other di-aromatic beta-
agonists. For faeces, an LC–MS2 method was optimised for detection of zilpaterol and
cimaterol (used as internal standard). To study the excretion profile, a male veal calf was
orally treated with therapeutic (for growth-promoting purposes) daily doses of Zilmax®
during 2 weeks [2]. During this period urine and faeces samples were collected. Without a
withdrawal period, the animal was sacrificed.
This study was performed in co-operation with TNO, Nutrition and Food Research, Product
Group Hormones and Veterinary Drugs, Department of Residue Analysis, The Netherlands.
66
Printed: www.dclsigns.be
Excretion profile of zilpaterol in calf urine and faeces
2.3.3. Experimental
2.3.3.1. Reagents and chemicals
Chemicals and solvents were obtained from Merck (Darmstadt, Germany) and Biosolve
(Volkenswaard, The Netherlands). The enzymatic deconjugation was performed with Helix
Pomatia juice (β-glucuronidase > 100,000 FU ml−1 and sulphatase > 1000,000 FU ml−1) from
Bioserpa (Marlborough, MA). Standards and internal standards were obtained from Sigma (St
Louis, MO) or RIVM (Bilthoven, The Netherlands). Zilmax® and zilpaterol were gifts from
Intervet (Schwabenheim, Germany).
2.3.3.2. Animal experiment
For the animal experiment, a male veal calf (3-4 months, ± 162 kg) was orally treated with
recommended (for growth-promoting purposes) daily doses of Zilmax® during 2 weeks. The
dose given was 0.15 mg zilpaterol per kg bodyweight per day, which is equal to 3.13 mg
Zilmax® per kg per day. During this period, urine and faeces samples were taken. Without a
withdrawal period, the animal was sacrificed.
2.3.3.3. Extraction and clean-up
To the urine samples (0.5-5 ml) isoxsuprine-d5 and ractopamine-d5 were added as internal
standards (5 ng ml−1). The urine was hydrolysed with Helix Pomatia at 37 °C for 16 h. After
adjustment of the pH to 9.6, the analytes were extracted with 10 ml isobutanol. After
centrifugation (10 min, 2000 rpm, 4 °C) and evaporation under nitrogen, the residue was
dissolved in 2 ml of phosphate buffer (pH 6). The clean up was carried out using a 130 mg
BondElut Certify (mixed C8 and SCX) SPE column (Varian Inc.). The column was
conditioned with methanol, water and 0.1 mol l−1 phosphate buffer (pH 6). The columns were
washed subsequently with 1 mol l−1 acetic acid and methanol. Elution was carried out using 3
ml of ethylacetate containing ammonia (0.57 mol l−1). Following evaporation of the solvents,
the residue was dissolved in 150 μl of methanol:water (5:95, v/v) with 10 mmol l−1
ammonium acetate and 50 μl was injected on the column.
To 1 g faeces cimaterol was added as internal standard at a level of 100 ng g−1. After addition
of 40 ml of hydrochloric acid 2 mol l−1, the sample was shaken for 15 min. After
centrifugation (15 min, 3600 rpm, 5 °C), 20 ml of the extract was decanted in a new
centrifuge tube. One millilitre of carbonate buffer (10 %, pH 9.8) was added and the pH was
67
Printed: www.dclsigns.be
Beta-agonists
adjusted to 9.8 using sodium hydroxide (32 %, 5 N). The extract was shaken for 1 min and
after centrifugation (15 min, 3600 rpm, 5 °C) 9 ml of the upper layer was applied on a Chem
Elut column (Varian Inc.). Elution was carried out using 40 ml of diethyl ether. A volume of
500 μl of pentafluoropropionic acid (PFPA) (0.3 mol l−1) was added to the tube. The sample
was placed in an ultrasonic bath for 10 min. After centrifugation for 15 min, the lower layer
of PFPA containing the analyte was formed. Approximately 400 μl of the drop was brought
into a vial for injection into the LC–MS system.
2.3.3.4. Instrumentation
For urine samples, chromatographic separation was achieved using an Inertsil ODS C18
column (3 μm, 3.0 × 100 mm, Varian Inc.). To separate the different compounds, a linear
gradient was used, using a mixture of water and methanol with ammonium acetate (Table 1).
The flow rate was 0.6 ml min−1. The mass spectrometer was operated in MS3-mode operating
in five segments. Each analyte was evaluated based on the product ions present in the MS3
mass spectra (Table 2).
Table 1 Mobile phase and gradient used to separate the non-N-alkyl beta-agonists
Time (min) Methanol/water (5:95, v/v) +
10 mM ammonium acetate
Methanol/water (80:20, v/v) +
30 mM ammonium acetate
0 100 0
1 100 0
15 0 100
20 0 100
22 100 0
25 100 0
68
Printed: www.dclsigns.be
Excretion profile of zilpaterol in calf urine and faeces
Table 2 Instrument method for the detection of beta-agonists in urine samples and the product
ions used for the evaluation of the beta-agonists
Segments Scan event Precursor ion →
product ion
Mass range
Analyte Second
transition
product ions
Segment 1 Scan event 1: 262.0 → 244.0
65.0 – 265.0
Zilpaterol 202, 185
Segment 2 Scan event 1: 288.0 → 270.0
70.0 – 290.0
Ritodrine 150, 121
Segment 3 Scan event 1: 302.0 → 284.0
75.0 – 305.0
Ractopamine 164, 121
Scan event 2: 307.0 → 289.0
75.0 – 310.0
Ractopamine-d5 167, 121
Segment 4 Scan event 1: 345.0 → 327.0
90.0 – 350.0
Formoterol 149, 121
Segment 5 Scan event 1: 302.0 → 284.0
75.0 – 305.0
Isoxsuprine 190, 150
Scan event 2: 307.0 → 289.0
75.0 – 310.0
Isoxsuprine-d5 190, 150
For faeces samples chromatographic separation was achieved using an Alltima C18 column (5
μm, 3.2 x 250 mm, Alltech Associates). The mobile phase consisted of a mixture of
pentafluoropropionic acid 10 mM (87%) and acetonitrile (13%). This mobile phase was
pumped at a rate of 0.5 ml min−1 for 4 min. The mass spectrometer was operated in MS2-
mode.
In both experiments, a 1100 series quaternary pump and an autosampler from Hewlett-
Packard (Palo Alto, CA, USA) were used. The MS detector was a ThermoFinnigan LCQ ion
trap mass spectrometer (San José, CA, USA) equipped with an atmospheric pressure chemical
ionisation (APCI) interface in positive ion mode.
69
Printed: www.dclsigns.be
Beta-agonists
2.3.4. Results and discussion
2.3.4.1. Chemical structure of zilpaterol
In evaluating the chemical structure of zilpaterol, the LC–MSn mass spectra in APCI positive
mode were recorded. In MS-full scan, the pseudo-molecular ion with m/z 262 appeared but
also a fragment ion with m/z 244 (Fig. 1). This fragment was due to the loss of water (Fig. 4).
MS2-full scan of the pseudo-molecular ion only showed the product ion with m/z 244 (Fig. 2).
Fragmentation of this product ion gave rise to two fragments, one with m/z 202 and the other
with m/z 185 (Fig. 3). The fragment ion with m/z 202 was due to the loss of CH3CCH3, a
subsequent loss of NH3 led to the fragment ion with m/z 185 (Fig. 4).
zilpaterol#185-217 RT: 0.60-0.71 AV: 33 NL: 3.65E8T: + c Full ms [ 50.00-300.00]
60 80 100 120 140 160 180 200 220 240 260 280 300m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
262.1
244.2
263.1
202.3 228.3 245.3203.3 278.9187.3164.8 227.3 256.4240.3149.3 175.2 299.1141.2117.3105.176.970.9 96.162.7
Fig. 1 MS-full scan of zilpaterol
zilpaterol#274-290 RT: 1.03-1.14 AV: 17 NL: 3.29E8T: + c Full ms2 [email protected] [ 70.00-300.00]
80 100 120 140 160 180 200 220 240 260 280 300m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e Ab
unda
nce
244.2
262.1203.1 244.9243.2 267.0 279.4 290.3185.7137.0 156.1125.9 225.1216.976.6 107.1 163.097.1
Fig. 2 MS2-full scan of zilpaterol
70
Printed: www.dclsigns.be
Excretion profile of zilpaterol in calf urine and faeces
zilpaterol#428-440 RT: 2.38-2.48 AV: 13 NL: 2.04E8T: + c Full ms3 [email protected] [email protected] [ 70.00-300.00]
80 100 120 140 160 180 200 220 240 260 280 300m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
202.2
185.2
187.3244.2227.2
202.9201.4173.2157.497.7 232.0 245.4 288.781.0 277.2107.7 142.1 266.6
Fig. 3 MS3-full scan of zilpaterol
OH
NH
NO
NH2
+
OH
NH
NO
NH2
+
NH
NO
NH2
+
NH
NO
NH3+
NH
N+
O
m/ z 262. 1556m/ z 244. 1450
MS2
m/ z 185. 0715 m/ z 202. 0980
Fig. 4 Fragmentation of the beta-agonist zilpaterol
The use of beta-agonists as growth promoter is forbidden in the European Union [7]. Beta-
agonists are therefore group A substances. The minimum number of identification points (IP)
for such forbidden compounds is set to four. LC-MSn precursor ions earn 1 IP and LC-MSn
product ions earn 1.5 IP [8]. MS2-full scan of the pseudo-molecular ion only showed one ion
with m/z 244 (Fig. 2), so 2.5 IP (one precursor ion and one product ion) were earned.
Moreover, the fragmentation in MS2-full scan is not specific (loss of water). To create more
specificity and to get enough identification points, MS3-full scan of the product ion was
checked. The mass spectrum of MS3-full scan of the product ion with m/z 244 showed the
ions with m/z 185 and m/z 202 (Fig. 3). This led to 5.5 IP (one precursor ion, one product ion
71
Printed: www.dclsigns.be
Beta-agonists
and 2 second transition product ions). Therefore, zilpaterol and the other di-aromatic beta-
agonists were identified in urine samples using LC-MS3 analysis.
2.3.4.2. LC-MSn methods for the detection of beta-agonists in urine and faeces
The standards zilpaterol, ritodrine, ractopamine, formoterol, isoxsuprine and the two internal
standards ractopamine-d5 and isoxsuprine-d5 were spiked to blank calf urine in a
concentration of 1 μg l−1. Fig. 5 shows the ion chromatograms of the different beta-agonists.
All the beta-agonists could be detected at a level of 1 μg l−1 with exception of formoterol
which was only detectable at 5 μg l−1. The two chromatographic peaks of isoxsuprine-d5 were
two possible isomers of the molecule. Levels of zilpaterol, ritodrine and ractopamine were
calculated using ractopamine-d5 as internal standard. Levels of formoterol and isoxsuprine
were calculated using isoxsuprine-d5.
RT: 6.00 - 14.00 SM: 5B
6 7 8 9 10 11 12 13 14Time (min)
0
50
1000
50
1000
50
1000
50
1000
50
100
Rel
ativ
e Ab
unda
nce 0
50
1000
50
1000
50
100RT: 9.65
RT: 10.43RT: 9.25RT: 13.16
RT: 12.74 RT: 13.72
RT: 6.83
RT: 6.92RT: 6.43
RT: 6.85
RT: 7.07RT: 6.54RT: 8.37
RT: 9.09RT: 7.82RT: 7.54RT: 9.71
RT: 10.22RT: 9.30RT: 11.18
RT: 11.70RT: 10.83
RT: 13.24
RT: 12.78
NL: 5.92E6m/z= 120.5-121.5+166.5-167.5 F: + c APCI Full ms3 [email protected] [email protected] [ 75.00-500.00] MS qj130404
NL: 1.25E6m/z= 149.5-150.5+189.5-190.5 F: + c APCI Full ms3 [email protected] [email protected] [ 75.00-500.00] MS qj130404
NL: 3.64E6m/z= 201.5-202.5 F: + c APCI Full ms3 [email protected] [email protected] [ 65.00-500.00] MS qj130404
NL: 5.91E5m/z= 184.5-185.5 F: + c APCI Full ms3 [email protected] [email protected] [ 65.00-500.00] MS qj130404
NL: 7.75E5m/z= 120.5-121.5+149.5-150.5 F: + c APCI Full ms3 [email protected] [email protected] [ 70.00-500.00] MS qj130404
NL: 9.26E5m/z= 120.5-121.5+163.5-164.5 F: + c APCI Full ms3 [email protected] [email protected] [ 75.00-500.00] MS qj130404
NL: 7.10E4m/z= 120.5-121.5+148.5-149.5 F: + c APCI Full ms3 [email protected] [email protected] [ 90.00-500.00] MS qj130404
NL: 7.16E5m/z= 106.5-107.5+133.5-134.5+149.5-150.5+189.5-190.5 F: + c APCI Full ms3 [email protected] [email protected] [ 75.00-500.00] MS qj130404
isoxsuprine
formoterol
ractopamine
ritodrine
zilpaterol
zilpaterol
isoxsuprine-d5
ractopamine-d5
Fig. 5 Ion chromatograms of ractopamine-d5, isoxsuprine-d5, zilpaterol (m/z 202), zilpaterol
(m/z 185), ritodrine, ractopamine, formoterol and isoxsuprine of a spiked calf urine sample
(1μg l-1)
For faeces a LC-MS2 method was optimised for the detection of zilpaterol. The goal was to
study the excretion profile of zilpaterol in faeces samples and not to develop a confirmatory
method for beta-agonists in faeces. Therefore, MS2 fragmentation of zilpaterol was enough
72
Printed: www.dclsigns.be
Excretion profile of zilpaterol in calf urine and faeces
since the identity of the compound in the faeces samples was well known in this experiment.
In addition, no deuterated internal standard was used. For faeces a level of 1 μg kg−1 zilpaterol
could be detected.
Both methods were developed in two different laboratories (in the Netherlands and in
Belgium), therefore different chromatographic conditions and different internal standards
were used to analyse zilpaterol. For urine, zilpaterol was incorporated in the qualitative multi-
residue method for the detection of di-aromatic beta-agonists. For faeces no attempts were
made to incorporate zilpaterol in the existing method for the detection of beta-agonists and
optimise this method, since the goal was to study the excretion profile of zilpaterol in faeces
and not to develop a qualitative multi-residue method for beta-agonists in faeces. Cimaterol
was chosen as internal standard because of its good response and good reproducibility. Since
both methods were developed independent of each other, no attempts were made to extract
and analyse di-aromatic beta-agonists from faeces and cimaterol from urine.
2.3.4.3. Excretion profile
A male calf was orally treated with 0.15 mg zilpaterol per kg bodyweight per day (3.13 mg
Zilmax®). Each day urine and faeces samples were collected and the calf was sacrificed after
14 days.
Fig. 6 shows the excretion profile of zilpaterol in urine and faeces. The levels of zilpaterol in
the urine samples were relatively high. Already after 2 days the concentration of zilpaterol
exceeded 1000 μg l−1. A steady-state concentration of about 1200 μg l−1 was quickly reached.
Also in faeces, a steady-state concentration of 83 μg kg−1 was quickly reached (first
measurement was already 71 μg kg−1 on day 2). A minimum value of 49 μg kg−1 was detected
on day 8, after 5 days a maximum value of 126 μg kg−1 was reached. It could be concluded
that zilpaterol was mainly excreted via urine.
73
Printed: www.dclsigns.be
Beta-agonists
Excretion profile of zilpaterol
0
500
1000
1500
2000
2500
3000
0 2 4 6 8 10 12 14
day
conc
. (µg
l-1, µ
g kg
-1)
conc zilp (µg kg-1) in faeces conc zilp (µg l-1) in urine
treatment
Fig. 6 Excretion profile of zilpaterol in urine and faeces
As the animal was sacrificed after the last treatment, no data were available for the final
elimination of zilpaterol. Based on the results, it could be concluded that zilpaterol could be
easily detected in farm samples during the application of zilpaterol as feed additive.
2.3.4.4. Phase I metabolites of zilpaterol
To study the presence of any co-extracted metabolites of zilpaterol in urine, also MS-full scan
analysis of a number of urine samples (days 2, 4, 8 and 12) was performed. The multi-residue
method of di-aromatic beta-agonists in urine was used to analyse possible co-extracted
metabolites of zilpaterol. In each sample, a de-isopropyl metabolite was found. Fig. 7 shows
the chromatograms of zilpaterol and its de-isopropyl metabolite and Fig. 8 shows MS-full
scan of de-isopropyl zilpaterol. The pseudo-molecular ion of de-isopropyl zilpaterol was m/z
220.
74
Printed: www.dclsigns.be
Excretion profile of zilpaterol in calf urine and faeces
RT: 0.00 - 12.00 SM: 5G
0 1 2 3 4 5 6 7 8 9 10 11 12Time (min)
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e Ab
unda
nce
RT: 7.36AA: 59657384353
RT: 10.96AA: 150733110
RT: 2.53AA: 2657249435
NL: 5.11E9m/z= 261.5-262.5 F: + c APCI Full ms [ 50.00-600.00] MS ICIS qj130409
NL: 1.83E8m/z= 219.5-220.5 F: + c APCI Full ms [ 50.00-600.00] MS ICIS qj130409
Fig. 7 Chromatograms of zilpaterol and its de-isopropyl metabolite
qj130409 #488-553 RT: 2.43-2.72 AV: 66 NL: 1.19E8T: + c APCI Full ms [ 50.00-600.00]
100 110 120 130 140 150 160 170 180 190 200 210 220 230m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e Ab
unda
nce
219.9202.2
150.1
152.0
212.0194.1203.2
148.2 220.9200.2185.2
138.0 218.4 228.0121.1 182.1146.2 180.1166.2 195.1153.1 210.0186.2130.1 226.1127.1114.2111.2
zilpaterol
de-isopropyl
zilpaterol
Fig. 8 MS full scan of the de-isopropyl metabolite of zilpaterol
The amount of de-isopropyl zilpaterol was calculated compared to the concentration of
zilpaterol in each sample. The ratio de-isopropyl zilpaterol/zilpaterol ranged between 2.45 %
and 5.66 %.
The phase II metabolism of zilpaterol was not investigated since previous studies revealed
that other beta-agonists were mainly excreted as conjugates [9], and therefore, the addition of
glucuronidase/sulphatase to urine samples is necessary for the multi-residue detection of beta-
agonists.
2.3.4.5. Quantification
Although the LC-MS3 method for the detection of beta-agonists in urine is primarily a
qualitative method, some quantitative data were examined. Three series, each containing two
75
Printed: www.dclsigns.be
Beta-agonists
blank urine samples spiked at a concentration of 1 μg l−1 were analysed at three different days.
Table 3 shows the calculated concentrations of zilpaterol, ritodrine, ractopamine, formoterol
and isoxsuprine.
Table 3 Quantitative data for zilpaterol, ritodrine, ractopamine, formoterol and isoxsuprine
analysed with the LC-MS3 method for beta-agonists in urine
Concentration (µg l-1)
zilpaterol ritodrine ractopamine formoterol isoxsuprine
Day 1 0.92 0.92 1.12 0.72 0.97
1.04 0.95 1.09 0.48 0.82
Day 2 0.90 0.82 1.02 1.02 1.07
1.07 1.19 1.01 1.02 1.03
Day 3 1.00 0.99 1.04 1.09 0.94
1.06 1.05 1.06 0.82 1.04
Average 1.00 0.99 1.05 0.86 0.98
Stdev 0.07 0.13 0.04 0.23 0.09
CV % 7.1 13.1 3.8 27 9.5
The detection limit (CCβ) is the smallest content of a compound that may be detected and
identified with an error probability of β. Fig. 5 shows that the CCβ of zilpaterol, ritodrine,
ractopamine and isoxsuprine was lower or equal to 1 μg kg−1. For formoterol, the CCβ was
lower or equal to 5 μg kg−1.
For samples spiked at a concentration of 1 μg l−1, the accuracy should range from 50 % to 120
% [8]. The accuracies of all beta-agonists lay within this acceptable range, except for one
analysis of formoterol. Fig. 5 shows that formoterol is not well detectable at 1 μg l−1, so
quantification for formoterol should be done at 5 μg l−1, like already mentioned above. The
precision of this method was evaluated by the coefficient of variation (CV), which should not
exceed the level calculated by the Horwitz equation [8]. For mass fractions lower than 100 μg
l−1, the application of the Horwitz equation gave unacceptable high values. Therefore, the CV
should be as low as possible; 23 % (CV at 100 μg l−1 = 23 %) was taken as a guideline. All
the coefficients of variation were lower than 14 % except for formoterol (27 %). So these
quantitative data for the LC–MS3 method for the detection of beta-agonists in urine were very
promising, even though this method was primarily a qualitative method.
76
Printed: www.dclsigns.be
Excretion profile of zilpaterol in calf urine and faeces
2.3.5. Conclusion
A LC-MS3 confirmatory method was developed that was able to simultaneously identify
zilpaterol, ractopamine, isoxsuprine and other di-aromatic beta-agonists in calf urine at a level
of 1 μg l−1. For faeces, a LC-MS2 method was optimised for the detection of zilpaterol in this
experiment.
When Zilmax® was administered orally to a male veal calf, the detection of zilpaterol in urine
and faeces could be easily achieved. Zilpaterol was mainly excreted via urine.
The method described for urine samples is used in routine control since 2001 within the
framework of self-control of veal calves in The Netherlands in order to extend the scope of
beta-agonists screened. So far no positive samples were found in this exclusive approach.
77
Printed: www.dclsigns.be
Beta-agonists
2.3.6. References
[1] C.S. Stachel, W. Radeck, P. Gowik (2003) Zilpaterol – a new focus of concern in residue
analysis, Analytica Chimica Acta 493, 63-67
[2] A. Plascencia, N. Torrentera, R.A. Zinn (1999) Influence of the beta-agonist, zilpaterol, on
growth performance and carcass characteristics of feedlot steers, Proceedings, Western
Section, American Society of Animal Science 50, 331-33
[3] Guia Technica Zilmax®, Hoechst Roussel Vet
[4] B. Bocca, C. Cartoni, M. Di Mattia (2003) Feed additives in animal nutrition:
Quantification of a new adrenergic drug by hyphenated techniques, Journal of Separation
Science 26, 363-368
[5] B. Bocca, M. Di Mattia, C. Cartoni, M. Fiori, M. Felli, B. Neri, G. Brambilla (2003)
Extraction, clean-up and gas chromatography-mass-spectrometry characterization of
zilpaterol as feed additive in fattening cattle, Journal of Chromatography B 783, 141-149
[6] B. Bocca, M. Fiori, C. Cartoni, G. Brambilla (2003) Simultaneous determination of
zilpaterol and other beta-agonists in calf eye by gas chromatography/tandem mass
spectrometry, Journal of AOAC International 86, 8-14
[7] Council Directive 96/23/EC of 29 April 1996 on measures to monitor certain substances
and residues thereof in live animals and animal products and repealing Directives 85/358/EEC
and 86/469/EEC and Decision 89/187/EEC and 91/664/EEC (1996) Official Journal of the
European Communities, no. L 125
[8] Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive
96/23/EC concerning the performance of analytical methods and the interpretation of results
(2002) Official Journal of the European Communities, no. L 221
[9] M.J. Sauer, M. Dave, B.G. Lake (1999) Beta(2)-agonist abuse in food producing animals:
use of in vitro liver preparations to assess biotransformation and potential target residues for
surveillance, Xenobiotica 29, 483-497
78
Printed: www.dclsigns.be
Chapter 3.1
Introduction
A wide range of veterinary medicinal products is administered legitimately to farm animals to
treat outbreaks of diseases or prevent diseases from spreading when modern intensive farming
practices are used. In order to reduce the likelihood of harmful levels of these veterinary drugs
reaching the human food chain, the European Union and many other countries have set
Maximum Residue Limits (MRL) [1]. Besides regulated veterinary drugs, there are also
veterinary medicinal products which are used illegally with the intention to promote growth.
Although the use of growth promoters is forbidden in the European Union, farmers still use
these compounds during the fattening of cattle [2-3].
Regulatory bodies are required to enforce and verify the requirements set by the European
Union. Therefore, official samples taken at the slaughterhouse or the farm are analysed for
unauthorised substances (substances listed in group A of Annex I of Directive 96/23/EC) but
also for registered veterinary medicinal products (substances listed in group B of Annex I of
Directive 96/23/EC) legally or illegally administrated. Laboratories testing these food
products and farm samples, have to ensure that the regulations are met [3-4].
Screening methods for veterinary medicinal products are developed to give an indication if
there is some form of drug residue present in a sample. They are classified as either
microbiological methods or group specific methods, such as immunological tests and receptor
tests. Microbiological tests are considered as multi-residue screening tests, while
immunological tests are more specific and can detect one substance or a group of related
chemicals [5-10]. Finally, the results of ‘suspect’ samples need to be confirmed. In this stage
an identification of an analyte can be combined with a quantification. Most techniques
comprise a chromatographic separation and a detection technique. Liquid chromatography
(LC) is often combined with ultraviolet detection (UV), fluorescence detection (FLD) and
mass spectrometry (MS) [11]. Gas chromatography (GC) can be combined with electron
capture detection, infrared detection, flame ionisation detection and mass spectrometry [12].
Confirmation methods can be both qualitative and quantitative. Quantitative methods identify
and quantify veterinary drugs that are permitted in some matrices below a maximum
concentration. Qualitative methods are used for unauthorised substances [13].
81 Printed: www.dclsigns.be
Veterinary drugs
3.1.1. Classification of veterinary medicinal products
In this part, the different classes of veterinary medicinal products (= veterinary drugs and
growth-promoting agents) which contain compounds that were recovered in samples analysed
in the laboratory of Chemical Analysis, are discussed [14-15]. For each class there is a brief
description of their characteristics and methods of detection.
3.1.1.1. Antibiotics and chemotherapeutics
Incorrect use of antibiotics in veterinary practice may lead to residues in edible tissues. These
residues may have a direct toxic effect on consumers, e.g. allergic reactions in hypersensitive
individuals. The European Union has set MRLs for several antibiotics in tissues, milk and
eggs [1].
Sulfonamides and trimethoprim
Sulfonamides (Fig. 1) are synthetic antibacterial agents with a wide spectrum encompassing
most gram-positive and many gram-negative organisms. Residues in food are of concern
because of the potential carcinogenic nature of sulfonamides.
H2N S
O
NH
O
R
Fig. 1 General structure of sulfonamides
Most sulfonamide formulations are supplied as combination products with
diaminopyrimidines, such as trimethoprim. These combinations act synergistically on specific
targets in bacterial DNA synthesis; they interfere with the production of folic acid, a precursor
in bacterial DNA synthesis [16].
Sulfonamides are analysed by a combination of LC with UV detection or, when more
selectivity is necessary, MS detection, most often with electrospray ionisation. Also the more
traditional LC-FLD method is reported, but derivatisation is necessary for fluorescence
detection [17-21].
82 Printed: www.dclsigns.be
Identification
Beta-lactam antibiotics
Penicillins and cephalosporins (Fig. 2) are still commonly used beta-lactam antibiotics, active
against mainly gram-positive bacteria. However, during recent years much progress has been
made in the development of new beta-lactams which are active against both gram-positive and
gram-negative bacteria (e.g. amoxicillin).
N
S
COR2
R3O
NH
C
O
R1
N
O
NH
C
O
R1S
COR2
CH3
CH3
Fig. 2 General structure of penicillins and cephalosporins
Beta-lactam antibiotics interfere with the enzyme transpeptidase which is involved in the
synthesis of the peptidoglycan cell wall. The difference in susceptibility between gram-
positive and gram-negative bacteria depends on the relative amount of peptidoglycan present
(gram-positive bacteria possess far more) and on the ability of the drugs to penetrate the outer
cell membrane of gram-negative bacteria [22-24].
Due to the unstable chemical structure of beta-lactams, these compounds are sensitive to heat
and alcohols. Therefore, precautions have to be taken during sample preparation.
The conventional detection techniques are LC-UV and sometimes after derivatisation LC-
FLD. Interfering matrix compounds often complicate these techniques. The use of LC-MS
can solve these selectivity problems. Electrospray ionisation is the interface of choice [17-
21,25-26].
83 Printed: www.dclsigns.be
Veterinary drugs
Tetracyclines
Tetracyclines are broad-spectrum antibiotics. By far the most commonly used tetracyclines in
veterinary practice are oxytetracycline (Fig. 3) and doxycycline.
OH O OHOH
O
CONH2
OH
N(CH3)2OHCH3HO
Fig. 3 Chemical structure of oxytetracycline
Tetracyclines inhibit bacterial protein synthesis by blocking the attachment of the transfer
RNA-amino acid to the messenger RNA (Fig.4) [27-28].
Fig. 4 Inhibition of protein synthesis by antibiotics [29]
After administration of tetracyclines, bound residues of the antibiotic will be found in bones
of slaughtered animals even months after treatment.
Due to the presence of two keto-groups, tetracyclines chelate to metal ions. They can also
interact with the silanol groups during LC separation, causing tailing of the chromatographic
peaks. The detection of tetracyclines is possible using LC-UV, LC-FLD and LC-MS.
84 Printed: www.dclsigns.be
Identification
Chelating agents are used to eliminate the problem of peak tailing. However, the presence of
non-volatile agents prevents the use of ESI-MS for detection because of the rapid
contamination of the mass spectrometer. Therefore, volatile buffers or acid solutions should
be used for LC-MS [17-21,25-26,30].
Quinolones
Quinolones and fluoroquinolones are a group of relatively new highly-potent, synthetic
antibacterial compounds, derived from 3-quinolonecarboxylic acid. Like the sulfonamides,
the quinolones are synthetic chemicals with antibacterial activity. They are active against
gram-negative bacteria and some gram-positive bacteria.
Quinolones inhibit two enzymes of DNA metabolism in bacteria, thereby inhibiting normal
bacterial DNA synthesis [31].
Since most quinolones show native fluorescence, LC-FLD is the technique traditionally used
for routine residue analysis. Also LC-UV is used for the determination of quinolones. Due to
the different types of substituents on the core structure, quinolones have rather different
physical properties. As a consequence, most analytical methods have been designed for the
determination of individual or two/three quinolones. LC-MS methods allow the multi-residue
determination of quinolones [17-20,25].
The characteristics will be explained more thoroughly in chapter 3.2.
Macrolides
Macrolide antibiotics are widely used in veterinary medicine to treat respiratory diseases or as
feed additives to promote growth. Macrolides are mainly active against gram-positive
bacteria, with some activity against gram-negative organisms. Erythromycin (Fig. 5), tylosin
and tilmicosin have found the most clinical applications of the macrolide class in veterinary
medicine.
85 Printed: www.dclsigns.be
Veterinary drugs
O
CH3
O
CH3
O
O
OCH3
CH3
OH
CH3
O
CH3
OH
H3C
OH
OH
H3C
CH3H3C
O O
HO
CH3
N
H3C
CH3
Fig. 5 Chemical structure of erythromycin
Macrolide antibiotics inhibit protein synthesis by inhibition of translocation (Fig. 4) [32-33].
Traditionally, UV absorbance is used for detection. However, erythromycin and some other
macrolides lack a suitable chromophore. Therefore, instead of the non-selective UV detection,
MS is preferred [17-21,25-26,34].
Aminoglycosides
Aminoglycoside antibiotics (e.g. spectinomycin, Fig. 6) are mainly active against gram-
negative bacteria. They are the drug of choice for the treatment of serious gram-negative
infections in animals. Aminoglycosides are commonly synergistic with beta-lactam
antibiotics.
O
O OHN
H3C
OH
CH3
OOH
NHH3C
HO
Fig. 6 Chemical structure of spectinomycin
86 Printed: www.dclsigns.be
Identification
Aminoglycosides inhibit protein synthesis by causing a misreading of messenger RNA
information (Fig. 4) [35].
For identification and quantification of aminoglycosides, LC-FLD and LC-ESI-MS are used
[17-21,25-26,36].
Lincosamides
Lincosamides are active against gram-positive bacteria. The major lincosamide is lincomycin
(Fig. 7). The mechanism of action is similar to the one of macrolide antibiotics. Lincomycin is
often used in combination with spectinomycin. This combination acts synergistically on
specific targets in bacterial protein synthesis [32].
NH3C
H3C
OCH
CHHO
CH3
OOH
OH
OH
SCH3
NH
Fig. 7 Chemical structure of lincomycin
Florfenicol and analogues
Florfenicol (Fig. 8) is a broad-spectrum antibiotic. It inhibits bacterial protein synthesis by
interfering with the formation of peptide bonds between amino acids (transpeptidation) (Fig.
4;its action is comparable to chloramphenicol) [32].
OH
HN
FO
Cl
Cl
SH3C
OO Fig. 8 Chemical structure of florfenicol
87 Printed: www.dclsigns.be
Veterinary drugs
GC in combination with mass spectrometry provides excellent analyte detectability of
florfenicol, but the main drawback of GC-MS is the need for derivatisation in order to
improve the chromatographic properties. More recently, LC-MS procedures are developed as
a substitute for GC-MS [17-18, 20, 25].
3.1.1.2. Anthelmintics
Anthelmintics (e.g. ivermectine, Fig. 9) are used both therapeutically and prophylactically to
control internal worm parasites and have, therefore, become an integral part of the animal
producing industry.
O
OCH3
OH
CH3 O
O
OCH3
CH3 O
CH3
O
OH
CH3
OH
OO
CH3
O
O
CH3
CH3
R
Component B R=CH
Fig. 9 Chemical structure of ivermectine
LC methods with UV or FLD detection are most commonly used. However, when a large
number of anthelmintics have to be simultaneously detected, selectivity problems can occur.
These problems can be solved by the use of LC-MS techniques [17].
1a 2CH3Component B R=CH1b 3
88 Printed: www.dclsigns.be
Identification
3.1.1.3 Non-steroidal anti-inflammatory drugs
Non-steroidal anti-inflammatory drugs (NSAIDs) are widely used and are often the initial
therapy for common inflammation. NSAIDs act by inhibiting the body’s ability to synthesise
prostaglandins [37-38].
LC-MS is the main detection technique for the analysis of NSAIDs, especially for multi-
residue methods [17,19].
Their characteristics will be explained more thoroughly in chapter 3.3.
3.1.1.4. Glucocorticosteroids
The major application of glucocorticosteroids (e.g. dexamethasone, Fig. 10) is in the
treatment of inflammatory and immunological disorders. In large doses glucocorticosteroids
cause reduced growth rates. However, low doses of glucocorticosteroids result in improved
feed intake, increased live weight gain, reduced feed conversion ratio, reduced nitrogen
retention and increased water retention [39-40].
O
CH3
F
HOCH3
CH3
OH
OOH
Fig. 10 Chemical structure of dexamethasone
For a long time, GC-MS was the method of choice for the detection of corticosteroids. This
technique requires derivatisation or oxidation of the analytes. Today, LC-APCI-MS is widely
used as detection technique for corticosteroids [17,19,40].
89 Printed: www.dclsigns.be
Veterinary drugs
3.1.2. Summary
From the beginning of 2001, injection sites have been collected at the slaughterhouse and
analysed in the laboratory of chemical analysis for the presence of legally and illegally used
veterinary medicinal products (chapter 3.2). Based on these results an overview could be
given which products are used frequently in practice and subsequently, the approach for
screening can be altered. The veterinary medicinal products which were recovered from these
samples, are summarised in Table 1 together with their classification.
Based on the results obtained after the analysis of these injection sites and on demand of the
Federal Agency for the Safety of the Food Chain a quantitative confirmation method was
developed for quinolones (chapter 3.3) and non-steroidal anti-inflammatory drugs (chapter
3.4).
90 Printed: www.dclsigns.be
Identification
Table 1 Analytes detected in injection sites since 2001 and their classification
Analyte Classification
Sulfadimethoxine sulphonamides
Sulfadoxine sulphonamides
Amoxicillin β-lactam antibiotics
Penicillin G β-lactam antibiotics
Oxytetracycline tetracyclines
Tetracycline tetracyclines
Enrofloxacin quinolones
Erythromycin macrolides
Tilmicosin macrolides
Tylosin macrolides
Spectinomycin aminoglycosides
Lincomycin lincosamides
Florfenicol florfenicol and analogues
Doramectin anthelmintics
Ivermectin anthelmintics
Levamisole anthelmintics
Flunixin NSAIDs
Meloxicam NSAIDs
Tolfenamic acid NSAIDs
Phenylbutazone NSAIDs
Dexametasone corticosteroids
Methylprednisolone corticosteroids
Prednisolone corticosteroids
91 Printed: www.dclsigns.be
Veterinary drugs
3.1.3. References
[1] Council Regulation (EEC) N° 2377/90 of 26 June 1990 laying down a Community
procedure for the establishment of maximum residue limits of veterinary medicinal products
in foodstuffs of animal origin (1990), Official Journal of the European Communities, no. L 67
[2] Council Directive 96/23/EC of 29 April 1996 on measures to monitor certain substances
and residues thereof in live animals and animal products and repealing Directives 85/358/EEC
and 86/469/EEC and Decision 89/187/EEC and 91/664/EEC (1996) Official Journal of the
European Communities, no. L 125
[3] Council Directive 96/23/EC of 29 April 1996 on measures to monitor certain substances
and residues thereof in live animals and animal products and repealing Directives 85/358/EEC
and 86/469/EEC and Decision 89/187/EEC and 91/664/EEC (1996), Official Journal of the
European Communities, no. L 125
[4] K. Grein (2000) The safe use of veterinary medicines and the need of residue surveillance,
Proceedings of the Euroresidue IV conference, 8-10 May, Veldhoven, The Netherlands, 73-78
[5] K. De Wasch, L. Okerman, S. Croubels, H. De Brabander, J. Van Hoof, P. De Backer
(1998) Detection of residues of tetracycline antibiotics in pork and chicken meet: correlation
between results of screening and confirmatory tests, The Analyst 123, 2737-2741.
[6] W. Haasnoot and R. Schilt (2000) Immunochemical and receptor technologies, In:
Residue analysis in food – principles ans applications, ed. M. O’Keeffe, Harwood Academic
Publishers, Singapore,107-144.
[7] L. Myllyniemi, A.L. Nuotio, E. Lindfors, R. Rannikko, A. Niemi, C. Bäckman (2001) A
microbiological six-plate method for the identification of certain antibiotic groups in incurred
kidney and muscle samples, The Analyst 126, 641-646.
[8] L. Okerman, K. De Wasch, J. Van Hoof (1998) Detection of antibiotics in muscle tissue
with microbiological inhibition tests: effects of the matrix, The Analyst 123, 2361-2365.
[9] L. Okerman, K. De Wasch, H. De Brabander, R. Abrams, J. Van Hoof, M. Cornelis, L.
Laurier (1999) Oude en nieuwe opsporingstechnieken voor antibioticaresiduen in het kader
van de huidige Belgische en Europese wetgeving, Vlaams Diergeneeskundig Tijdschrift 68,
216-223.
[10] L. Okerman, S. Croubels, S. De Baere, J. Van Hoof, P. De Backer, H.F. De Brabander
(2001) Inhibition tests for detection and presumptive identification of tetracyclines, beta-
lactam antibiotics and quinolones in poultry meat, Food Additives and Contaminants 18, 385-
393
92 Printed: www.dclsigns.be
Identification
[11] F.A. Mellon (1991) Liquid chromatography/ mass spectrometry, In: VG Monographs in
mass spectrometry, volume 2, No. 1
[12] M.E. Rose (1990) Modern practice of gas chromatography/mass spectrometry, In: VG
Monographs in mass spectrometry, volume 1, No. 1
[13] H.F. De Brabander, K. De Wasch, L. Okerman, P. Batjoens (1998) Moderne
analysemethodes voor additieven, contaminanten en residuen, Vlaams Diergeneeskundig
Tijdschrift 67, 96-105.
[14] K.N. Woodward and G. Shearer (1995) Antibiotic use in animal production in the
European Union – Regulation and current methods for residue detetcion, In: Chemical
Analysis for antibiotics used in agriculture, In: Chemical analysis for antibiotics used in
agriculture, ed. H. Oka, H. Nakazawa, K.I. Harada and J.D. Macneil, AOAC International,
Arlington, 47-76.
[15] www.bcfi-vet.be/nlplan.lasso
[16] J.W. Spoo and J.E. Riviere (2001) Sulfonamides, In: Veterinary Pharmacology and
Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press, Ames, 796-817
[17] A.A.M. Stolker, U.A.Th. Brinkman (2005) Analytical strategies for residue analysis of
veterinary drugs and growth-promoting agents in food-producing animals – a review, Journal
of Chromatography A 1067, 15
[18] A. Di Corcia and M. Nazzari (2002) Liquid chromatography-mass spectrometric methods
for analyzing antibiotics and antibacterial agents in animal food products, Journal of
Chromatography A 975, 53-89
[19] G. Balizs and A. Hewitt (2003) Determination of veterinary drug residues by liquid
chromatography and tandem mass spectrometry, Analytica Chimica Acta 492, 105-131
[20] A. Gentili, D. Peret and S. Marchese (2005) Liquid chromatography-tandem mass
spectrometry for performing confirmatory analysis of veterinary drugs in animal-food
products, TRAC-Trends in Analytical Chemistry 24, 704-733
[21] B. Shaikh and W.A. Moats (1993) Liquid-chromatography analysis of antibacterial drug
residues in food-products of animal origin, Journal of Chromatography 643, 369-378
[22] S.L. Vaden and J.E. Riviere (2001) Penicillins and related β-lactam antibiotics, In:
Veterinary Pharmacology and Therapeutics (8th edition), ed. H.R. Adams, Iowa State
University Press, Ames, 818-827
[23] J.O. Boison (1995) chemical analysis of β-lactam antibiotics, In: Chemical analysis for
antibiotics used in agriculture, ed. H. Oka, H. Nakazawa, K.I. Harada and J.D. Macneil,
AOAC International, Arlington, 235-306
93 Printed: www.dclsigns.be
Veterinary drugs
[24] J.F. Prescott and J.Desmond Baggot (1988) Beta-lactam antibiotics: penicillins,
cephalosporins, and newer antibiotics, In: Antimicrobial Therapy in Veterinary Medicine, ed.
J.F. Prescott and J. Desmond Baggot, Blackwell Scientific Publications Ltd., Oxford, 71-109
[25] F.J. Schenck and P.S. Callery (1998) Chromatographic methods of analysis of antibiotics
in milk, Journal of Chromatography A 812, 99-109
[26] D.R. Bobbitt and K.W. Ng (1992) Chromatographic analysis of antibiotic materials in
food, Journal of Chromatography 624, 153-170
[27] J.E. Riviere and J.W. Spoo (2001) Tetracycline antibiotics, In: Veterinary Pharmacology
and Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press, Ames, 828-840
[28] H. Oka and J. Patterson (1995) Chemical analysis of tetracycline antibiotics, In:
Chemical analysis for antibiotics used in agriculture, ed. H. Oka, H. Nakazawa, K.I. Harada
and J.D. Macneil, AOAC International, Arlington, 333-406
[29] www.elmhurst.edu/~chm/vchembook/654antibiotic.html
[30] H. Oka, Y. Ito and H. Matsumoto (2000) Chromatographic analysis of tetracycline
antibiotics in foods, Journal of Chromatography A 882, 109-133
[31] M.G. Papich and J.E. Riviere (2001) Fluoroquinolone antimicrobial drugs, In Veterinary
Pharmacology and Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press,
Ames, 898-912
[32] M.G. Papich and J.E. Riviere (2001) Chloramphenicol and derivatives, macrolides,
lincosamides, and miscellaneous antimicrobials, In: Veterinary Pharmacology and
Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press, Ames, 868-897
[33] M. Horie (1995) Chemical analysis of macrolide antibiotics, In: Chemical analysis for
antibiotics used in agriculture, ed. H. Oka, H. Nakazawa, K.I. Harada and J.D. Macneil,
AOAC
[34] I. Kanfer, M.F. Skinner and R.B. Walker (1998) Analysis of macrolide antibiotics,
Journal of Chromatography A 812, 255-286
[35] J.E. Riviere and J.W. Spoo (2001) Aminoglycoside antibiotics, In: Veterinary
Pharmacology and Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press,
Ames, 841-867
[36] N. Isoherranen and S. Soback (1999) Chromatographic methods for analysis of
aminoglycoside antibiotics, Journal of AOAC International 82, 1017-1045
[37] K. Baert (2003) Pharmacokinetics and Pharmacodynamics of Non-Steroidal Anti-
Inflammatory Drugs in Birds, thesis, Ghent University, Faculty of Veterinary Medecine, 3-18
94 Printed: www.dclsigns.be
Identification
[38] D.M. Boothe (2001) The analgesic, antipyretic, anti-inflammatory drugs, In: Veterinary
Pharmacology and Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press,
Ames, 433-451
[39] D. Courtheyn, B. Le Bizec, G. Brambilla, H.F. De Brabander, E. Cobbaert, M. Van de
Wiele, J. Vercammen, K. De Wasch (2002) Recent development in the use and abuse of
growth promoters, Analytica Chimica Acta 473, 71-82
[40] O. Van den hauwe (2005) Identification and confirmation of synthetic glucocorticoid
residues in biological matrices by liquid chromatography combined with tandem mass
spectrometry, thesis, Ghent University, Faculty of Pharmaceutical Sciences, 5-35
95 Printed: www.dclsigns.be
Chapter 3.2
Identification and semi-quantification of veterinary medicinal products in
injection sites
Adapted from:
N. Van Hoof, K. De Wasch, S. Poelmans and H.F. De Brabander
Detecting veterinary residues in practice: the case of veterinary medicinal products
In: Rapid and on-line instrumentation for food quality assurance (2003), ed. I.E. Tothill, Woodhead
Publishing Limited, Cambridge, UK, 91-115
And
K. De Wasch, N. Van Hoof, S. Poelmans, L. Okerman, D. Courtheyn, A. Ermens, M. Cornelis and
H.F. De Brabander
Identification of "unknown analytes" in injection sites: a semi-quantitative interpretation
Analytica Chimica Acta (2003) 483, 387-399
3.2.1. Introduction
From the beginning of 2001, injection sites have been sampled at the slaughterhouse for
identification of legally and illegally used veterinary medicinal products. In analysing these
samples, an overview could be given of what is frequently used nowadays in practice.
Subsequently, the approach for screening can be altered.
A wide range of different groups of veterinary medicinal products is used in practice. Since
every group requires a specific extraction and detection procedure, it has become too
expensive to check every sample for a whole batch of different groups. Therefore, an
alternative approach is proposed in which a simple extraction and clean-up is combined with a
multi-residue LC-MSn identification and semi-quantification. The generic MS method allows
the detection of unknown veterinary medicinal products and tandem mass spectrometry is
necessary for the identification and quantification. Another important aspect to consider is
that injection sites very often contain high concentrations of the administered product.
Injection sites are considered as meat by inspection services and therefore the MRL for meat
applies, especially because of the possible consumption of an injection site. Because of the
97 Printed: www.dclsigns.be
Veterinary drugs
high concentrations there is no demand for the registered veterinary drugs to be quantified in
the concentration range of the MRL. A different quantification approach will be used.
3.2.2. Experimental
3.2.2.1. Reagents and chemicals
Standards were obtained from Sigma (St. Louis, MO) and the injectable solutions from the
Clinical Department of the Faculty of Veterinary Medicine (Ghent, Belgium). The injectable
solutions were used for identification purposes. The internal standard desoximethasone
(DOM) was obtained from Sigma (St. Louis, MO).
Stock solutions of 1000 ng µl-1 were prepared in ethanol and stored at 4 °C. For the
preparation of working solutions methanol was used.
3.2.2.2. Extraction and clean-up procedure
The injection site is sampled by cutting at least 8 g suspect material and transferring it to a
double bag of a stomacher. Methanol is added at a ratio of 2.5 ml g-1 and also 1500 µg kg-1
desoximethasone (internal standard) is added. The mixture is extracted with a stomacher
during at least one minute. The next day, the mixture is filtered. This primary extract is
prepared in a laboratory room separated from the laboratory for residue analysis to avoid any
contamination. Afterwards 5 ml of the extract is evaporated under nitrogen to 2 ml or less.
The clean-up was performed using an Isolute C18 cartridge (500 mg) (IST International, Mid
Glamorgan, UK). The columns were conditioned with 2 x 2 ml methanol followed by 2 x 2
ml ultrapure water. Ultrapure water was added to the extract till 3 ml. After application of this
extract, the cartridge was rinsed with 2 x 2 ml methanol/water (40:60). The veterinary
medicinal products were eluted with 2 ml methanol/water (70:30) and 2 ml methanol. The
eluate was evaporated to dryness and the residue was reconstituted in 50 µl methanol and
subsequently 100 µl 0.4 % acetic acid in methanol/water (60:40), before injecting 30 µl onto
the HPLC column [1].
3.2.2.3. Instrumentation
Chromatographic separation was achieved using a Symmetry C18 column (5 µm, 150 x
2.1 mm, Waters, Milford). The mobile phase consisted of a mixture of methanol (A) and 1 %
acetic acid in water (B). The flow rate was 0.3 ml min−1. A linear gradient was used. Twenty
98 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
percent of A was maintained for 7 min and increased to 100% A in 10 min (maintained for
7 min). In between samples there was an equilibration time of 10 min at the initial conditions.
The LC apparatus comprised of a TSP P4000 pump and a model AS3000 autosampler
(ThermoFinnigan, San José, CA, USA). The MS detector was a LCQ Classic and a LCQ
Deca ion trap mass spectrometer (ThermoFinnigan, San José, CA, USA) equipped with an
electrospray ionisation (ESI) interface. For each sample an acquisition was made in positive
and negative ion mode to obtain complementary information.
3.2.2.4. Some definitions
Unknown: an analyte which is identified in a non-target analysis, for which no specific
extraction or confirmation procedure is used or developed, of which there is no information of
the group of veterinary drugs or growth promoters to which it belongs.
Suspect ion, during infusion: an ion with a signal-to-noise (s/n) ratio > 3 that was not present
in the previously infused mobile phase or methanol.
Suspect ion, injection on column: a species that generates a chromatographic peak in the total
ion current with s/n > 3, or a chromatographic peak of a specific ion trace with s/n > 3.
Layout: option in the software (Xcalibur 1.2) in which mass traces of pseudo-molecular ions
of injectable solutions or standards are combined in a window. A layout can be added
depending on the knowledge of analytes at that time [2].
MSn acquisition: MS1, MS2 3, MS , … MSn fragmentation of pseudo-molecular ions.
Fragmentation in MSn is performed until the spectrum becomes ‘unstable’.
Scan event: a mass spectrometer scan that is defined by selecting the required and optional
scan event settings. Required settings are scan power, ion polarity and scan mode. Multiple
scan events can be defined for each segment of time [2].
Injectable solution: a registered veterinary medicinal product used in veterinary practice of
which the concentration of the active component is known.
Specific method: a method containing specific MSn parameters of the identified analyte and
which contains three scan events: MS-full scan 100-1000, MS2-full scan of the identified
analyte, MS2-full scan of DOM.
99 Printed: www.dclsigns.be
Veterinary drugs
3.2.3. Different approaches
Since the beginning of 2001 different injectable or standard solutions of registered veterinary
medicinal products were collected. These solutions were subjected to infusion-MSn and LC-
MSn. The collected data will function as a database for the identification of unknown analytes
present in an injection site. Injectable solutions are not the active compounds but the drugs as
used in veterinary practice. Additional impurities can therefore obscure the chromatogram and
the spectrum, but this can also be expected in injection sites.
For the identification of unknown analytes two approaches can be used depending on the
availability of the instrument. A first approach is infusion MSn, a second approach is LC-MSn.
3.2.3.1. Infusion-MSn
A first approach is infusion-MSn. Mobile phase is pumped at 0.3 ml min−1 and mixed with the
extract that is connected via a T-piece and pumped at 5 µl min−1. The first acquisition is
always the infusion of a blank (methanol) to obtain the background ions. Background ions are
not taken into consideration for further fragmentation when analysing the sample unless the
intensity of the background ions in the sample would be considerably higher than during
acquisition of methanol. Full scan MSn data in positive and negative ion mode are acquired of
the suspect ions.
3.2.3.2. LC-MSn
In addition to the infusion approach a default gradient is used in MS full scan in positive and
negative ion mode. The advantage is that LC-MSn is automated and data can be acquired
overnight while infusion MSn is an online interpretation.
Ion traces of ‘known’ compounds, collected injectable and standard solutions of registered
veterinary medicinal products, are examined by applying a layout. A layout is an option in the
software in which mass traces of pseudo-molecular ions are combined in a window. A layout
can be added depending on the knowledge of analytes at that time [2]. The layouts used in this
application are given in Tables 1 and 2.
100 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
Table 1 Layout of ion traces in positive ion mode
Layout name MS-ions Analyte VMP-pos1 TIC 291 trimethoprim 311 sulfadoxine/sulfadimethoxine 615-308 neomycin 360-316 enrofloxacin 358 danofloxacin 255 ketoprofen 461 oxytetracycline VMP-pos2 407 lincomycin 429 lincomycin 333-351-365 spectinomycin 478 gentamycin C1 464 gentamycin C2 450 gentamycin C1a 322 gentamycin VMP-pos3 335 penicillin G 237 procaine 241 benzathine 279 flunixin 297 flunixin 435 tilmicosin 869.5 tilmicosin 921 doramectin VMP-pos4 521-543 beclomethasone-dipropionic acid 407-429 Flugestone acetate 734 erythromycin 445 tetracycline 916-948 tylosin 321-339-357 desoximethasone (MS2) (I.S.) 366 amoxicillin 205 levamisole VMP-pos5 897 ivermectin 313 tetrahydrogestrinone 221 xylazine
101 Printed: www.dclsigns.be
Veterinary drugs
Table 2 Layout of ion traces in negative ion mode
Layout name MS-ions Analyte VMP-neg1 TIC 309 sulfadoxine 309 sulfadimethoxine 333 penicillin G 405 lincomycin 465 lincomycin 350 meloxicam 673 neomycin 253 ketoprofen VMP-neg2 459 oxytetracycline 336 florfenicol 356 florfenicol 392 florfenicol 331 spectinomycin 349 spectinomycin 251 flunixin 295 flunixin VMP-neg3 321-381 chloramphenicol 897 doramectin 957 doramectin 216-260 tolfenamic acid 355-375 desoximethasone (MS2) (I.S.) 342-380-758 clorsulon 307 phenylbutazone 137 salicylic acid VMP-negcost 379-469 flumethasone 329-419 prednisolone 343-433 methylprednisolone 413-493 triamcinolone acetonide
355-435 desoximethasone (I.S.)/
fluorometholone 361-451 dexamethasone / betamethasone 429-465-525 clobetasol propionic acid
3.2.3.3. Proposed strategy
Electrospray ionization (ESI) and atmospheric pressure chemical ionisation (APCI) are both
soft ionisation techniques but ESI is preferred since fragmentation of the pseudo-molecular
ion in full scan MS is not as intense as when using APCI. Fragmentation in MS-full scan can
mask the presence of the pseudo-molecular ion which is the direct link with the molecular
mass of the analyte of interest.
Possible molecular masses are calculated from [M+H]+ or [M-H]- + - ions, Na or Ac adducts.
There are different identification strategies. It is possible to derive the molecular mass of the
102 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
unknown analyte by complementary data from the positive and negative pseudo-molecular
ions ([M+H]+ and [M-H]- ions). Sometimes there appears an adduct ion in MS-full scan
whether or not in the presence of the pseudo-molecular ion. Adducts are formed by reaction
between the analyte and the solvent used in the mobile phase. In this method acetic acid was
added to the mobile phase and therefore acetate-adducts can be formed in negative ion mode.
Some analytes, depending on their functional groups, cannot form positive or negative ions or
adducts. The absence of ions in one ion mode also gives structural information about the
analyte. In the presence of a second compound (a veterinary drug or a chemical product)
different combinations can be formed leading to mass spectra in which these different
combinations can be recognized. Different masses will appear in the spectrum at different
retention times. This will be demonstrated with the example of Penicillin G-benzathine
(paragraph 3.2.5.1).
Using the collected data of the different injectable or standard solutions and the database of
the Merck-index the identity of the analyte can be elucidated. All possible compounds from
the Merck database are filtered based on their therapeutic category or intended use.
Identification
When a sample reveals a ‘suspected’ mass spectrum, these MSn data will be compared with
the MSn data of the standard or injectable solution and the identity is confirmed based on
comparison of the mass spectra. Criteria for the identification are based on identification
points (IP). The minimum number of IPs for unauthorised compounds is set to four, for
compounds with a MRL a minimum of three IPs is required for the confirmation of the
compounds’ identity [3].
The identity of the analyte can be reported if the substance is a unauthorised substance. In that
case quantification is not mandatory.
Quantification
If the identified analyte has a MRL and the standard or injectable solution is available, the
concentration must be estimated. A quantitative validation normally consists of determining
the required validation parameters at three levels: 0.5 MRL, MRL, 1.5 MRL [3]. This
validation approach is elaborative and time consuming. For highly concentrated injection
sites, an alternative approach is proposed. The alternative validation consists of a comparison
103 Printed: www.dclsigns.be
Veterinary drugs
of the analyte concentration in the sample with the spike at MRL and 10 times MRL
concentration. The alternative approach is performed as a mini-validation.
A mini-validation consists of three blank matrices fortified with the MRL concentration of the
analyte, three blank matrices fortified with 10 times MRL concentration (10MRL) of the
analyte and one blank matrix. This approach allows the analyst to meet the needs and
requirements of the customer awaiting the results. The analysis is accurate, fast and the total
cost of this approach is minimized in comparison with a traditional validation and analysis.
Identification can be performed within 48 hours and an extra 24 hours are necessary for the
quantification. The choice between a traditional validation and an alternative validation
depends on the analytical purpose and method.
Before an injection site can be reported as non-compliant, there are some conditions that
needs to be fulfilled.
If the area ratio of the spike at 10MRL is ≥ 3 times the area ratio of the spike at MRL
concentration AND the area ratio of the sample is ≥ 4 times the area ratio of the spike at MRL
concentration, the sample is reported as non-compliant with a concentration higher than the
MRL.
If the spike at MRL concentration has no response in MS2-full scan (positive or negative ion
mode) and the spike at 10MRL has a positive response, the concentration of the sample can be
reported as higher than the MRL if its area ratio is ≥ 1.5 times the area ratio of 10MRL.
If the spike at MRL and 10MRL have no response in MS2-full scan (positive or negative ion
mode), this approach is considered as a reductio ad absurdum (reduction to absurdity or
contradiction). If the analyte is identified in an injection site, the concentration can be
reported as higher than the MRL without any reasonable doubt.
If one of these conditions is not fulfilled, the sample can be analysed with a specific
confirmation method in our laboratory or transferred to the National Reference Laboratory
(NRL). This will be decided in consultation with the Federal Agency for the Safety of the
Food Chain (FAVV).
104 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
3.2.4. A practical example: sulfadimethoxine
Sulfadimethoxine (Fig. 1) is a sulfonamide chemotherapeutic. Sulfonamides are antibacterial
agents widely used in veterinary practice to prevent infections in livestock. They have also
been used in animal feeds to promote growth and to treat diseases. Residues are often found in
meat and milk products where they enter the human food chain. Sulfadimethoxine has a MRL
of 100 μg kg-1 in muscle tissue [4]. Because injection sites are considered as meat by
inspection services, the MRL for meat applies.
H2N
SNH
OO
N
N
OCH3
OCH3
Fig. 1 Chemical structure of sulfadimethoxine
3.2.4.1. Identification
Injection of an extract of an injection site revealed in MS-full scan an intensive negative ion
with m/z 309 and an intensive positive ion with m/z 311 (Fig. 2). An analyte with molecular
mass 310 can be expected from the complementary information of the positive and negative
ion spectra. Using the Merck index a search is performed in the molecular mass range 309-
311. Different possibilities were found: mepazine (tranquilliser), methoprene
(ectoparasiticide), sulfadoxine (antibacterial) and sulfadimethoxine (antibacterial).
105 Printed: www.dclsigns.be
Veterinary drugs
020524s17 #1229-1242 RT: 15,14-15,29 AV: 7 SB: 189 1,46-2,64, 3,15-6,46 NL: 1,25E9T: - c ESI Full ms [ 100,00-1000,00]
100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e Ab
unda
nce
309,3
310,2
595,9 640,9311,2480,0390,9 927,9154,2 230,3 972,8722,9677,9 882,8307,4 342,9171,2 466,9139,2 754,1594,9 798,8 837,4526,0
020527s20 #2318-2346 RT: 15,22-15,35 AV: 9 SB: 98 14,01-14,88, 15,97-16,65 NL: 6,20E8T: + c ESI Full ms [ 100,00-1000,00]
100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e Ab
unda
nce
311,1
642,6
288,2
659,0
333,0156,0
619,8 664,7 929,1 968,1734,1245,1 482,0412,7334,1 598,2 718,4468,6 756,4108,0 802,9157,2 890,1218,1 522,3 816,0547,3
Fig. 2 MS-full scan in negative and positive ion mode
Mass spectra of the injectable and standard solutions of some of the compounds present in the
Merck database, revealed that both sulfadoxine and sulfadimethoxine (MM 310.33) produced
the same pseudo-molecular ions as the ‘unknown’ sample. There are no differences between
the two components in MS-full scan, but the ion ratios in MS2-full scan in positive ion mode
are different. Therefore a second injection of the extract was performed in positive ion mode,
both in MS-full scan and MS2-full scan (Fig. 3). Also the standard solutions of sulfadoxine
and sulfadimethoxine were injected (Fig. 4). The different mass spectra suggest that the
veterinary drug present in the sample is probably sulfadimethoxine.
106 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
107
020527s20 #2322-2349 RT: 15,24-15,37 AV: 9 SB: 187 12,94-14,77, 16,63-17,87 NL: 3,94E8F: + c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
100 120 140 160 180 200 220 240 260 280 300 320m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e Ab
unda
nce
156,0
245,1
217,9
246,1108,0
157,0
311,0249,9218,9 235,7155,1108,9 293,0126,9 172,2 251,0141,0 236,3200,1 215,3157,7 266,4 282,1
Fig. 3 MS2-full scan in positive mode of the ion m/z 311
020912s21 #1955-1978 RT: 13,12-13,22 AV: 8 SB: 77 11,78-12,36, 13,47-14,43 NL: 5,98E5F: + c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
100 120 140 160 180 200 220 240 260 280 300 320m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
155,9
108,1245,1
156,9
292,8109,1 250,7 315,5282,4 309,6217,7141,0 211,2 267,7237,0171,1 185,1 263,8154,8 220,2196,6157,4
020912s25 #2278-2283 RT: 15,27-15,29 AV: 2 SB: 187 12,91-15,00, 17,45-19,14 NL: 1,02E6F: + c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
100 120 140 160 180 200 220 240 260 280 300 320m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
156,0
245,1
218,0
246,1
311,0156,9
108,0218,9155,3 296,7249,9235,7 266,5 284,5 309,2109,0 274,8163,1 311,6141,0
Fig. 4 MS2-full scan in positive mode of sulfadoxine and sulfadimethoxine
However, different quality criteria need to be checked for qualification and quantification.
First the number of identification points (IP), secondly the signal to noise ratio of the product
ions and finally the relative intensity of the product ions [3]. Because sulfadimethoxine is a
veterinary drug with a MRL, three identification points are required for the identification. The
precursor ion earns 1 IP (m/z 311) and each product ion earns 1.5 IP (m/z 156, 218 and 245),
so 5.5 IP are earned. The signal to noise ratio of the different product ions, which needs to be
at least three, is more than three. The permitted margins for the relative intensities of the
product ions are summarized in Table 3. MS2-full scan of the standard sulfadimethoxine was
used to calculate these margins. The relative intensities of the product ions of the analyte
Printed: www.dclsigns.be
Veterinary drugs
present in the injection site are all ranged within the permitted margins. So, it can be
concluded that the injection site contained the veterinary drug sulfadimethoxine.
Table 3 Comparison between the relative intensities of the product ions of the standard
sulfadimethoxine and the relative intensities of a sample
Product ions
(m/z)
Relative intensity
sulfadimethoxine
Permitted range Relative intensity
sample
156 100 80 - 100 100 OK
218 41 30.8 - 51.3 45 OK
245 91 72.8 - 109.2 94 OK
3.2.4.2. Quantification
Because sulfadimethoxine (SDT) is a veterinary drug with a MRL of 100 μg kg-1,
quantification is required. This is performed by comparing the analyte concentration in the
sample (Fig. 5) with the spike at MRL and 10MRL concentration (Fig. 6).
The area ratio is calculated by dividing the area of sulfadimethoxine (SDT) by the area of the
internal standard desoximethasone (DOM). The area ratio of the sample is 19.3 the one of the
spike at MRL concentration is 0.086 and the one at 10 times MRL concentration is 0.92. So,
the area ratio of the spike at 10 times MRL concentration is 10.7 (≥ 3) times the area ratio of
the spike at MRL concentration AND the area ratio of the sample is 220 (≥ 4) times the area
ratio of the spike at MRL concentration. In conclusion, the sample is non-compliant with a
concentration higher than the MRL.
108 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
D:\Doctoraat\...\data\020527s20 28-5-2002 4:02:12 K509K
RT: 10,07 - 24,94 SM: 7B
12 14 16 18 20 22 24Time (min)
0
20
40
60
80
100
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
0
20
40
60
80
100
RT: 15,27MA: 9007275092SN: 148
15,8116,63 18,35 20,2014,02 21,3313,7611,52 23,66
RT: 15,27MA: 14471516783SN: 179
16,23 17,13 19,17 20,59 21,86 24,8113,7913,0411,73RT: 18,77MA: 750852605SN: 319
14,8819,4414,50 15,7312,38 17,73 20,75 24,7922,66
NL: 6,90E8m/z= 310,5-311,5 F: + c ESI Full ms [ 100,00-1000,00] MS 020527s20
020527s20 #2321-2352 RT: 15,22-15,38 AV: 11 SB: 187 12,94-14,77, 16,63-17,87 NL: 3,84E8F: + c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
NL: 1,11E9m/z= 107,5-108,5+155,5-156,5+217,5-218,5+244,5-245,5 F: + c ESI Full ms2 311,00@35,00 [ 100,00-320,00] MS 020527s20
NL: 7,53E7m/z= 320,5-321,5+338,5-339,5+356,5-357,5 F: + c ESI Full ms2 377,00@30,00 [ 100,00-385,00] MS 020527s20
100 150 200 250 300m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e A
bund
ance
MS2156,0
245,1
217,9
246,1
SDT MS
108,0157,0
311,0219,0 249,9155,1108,9 293,0172,2 200,1 266,4
020527s20#2321-2351 RT: 15,22-15,38 AV: 11SB: 187 12,94-14,77, 16,63-17,87F:m/z Intensity Relative 156,0 384123402,6 100,00 245,1 345359339,8 89,91 217,9 163262796,2 42,50 246,1 49544441,5 12,90 108,0 39683005,2 10,33 157,0 25743545,3 6,70 311,0 12457942,5 3,24 219,0 10777219,4 2,81 249,9 10069636,4 2,62 235,7 9535393,5 2,48
SDT MS2+ c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
2DOM MS
Fig. 5 Chromatographic and mass spectrometric data of the extract of an injection site (SDT =
sulfadimethoxine and DOM = desoximethasone)
109 Printed: www.dclsigns.be
Veterinary drugs
D:\Doctoraat\...\data\020527s29 28-5-2002 16:41:17 S580B
RT: 10,07 - 24,94 SM: 7B
12 14 16 18 20 22 24Time (min)
0
20
40
60
80
100
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
0
20
40
60
80
10019,42
17,21
17,3416,3913,83 18,30
12,83 22,0811,12 21,5124,36
22,40
RT: 15,33MA: 49805647SN: 554
15,98 17,25 18,7614,6210,96 12,60 23,5220,47 21,71RT: 18,85MA: 578054896SN: 200
14,96 19,5813,5810,48 16,7011,59 21,06 22,49
NL: 1,83E7m/z= 310,5-311,5 F: + c ESI Full ms [ 100,00-1000,00] MS 020527s29
Spike at MRL 020527s29 #2361-2404 : 15,23-15,42 : 14 SB: 443 11,52-13,65, 16,05-21,45 NL: 1,31E6F:
RT AV+ c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
NL: 3,89E6m/z= 107,5-108,5+155,5-156,5+217,5-218,5+244,5-245,5 F: + c ESI Full ms2 311,00@35,00 [ 100,00-320,00] MS 020527s29
NL: 5,68E7m/z= 320,5-321,5+338,5-339,5+356,5-357,5 F: + c ESI Full ms2 377,00@30,00 [ 100,00-385,00] MS 020527s29
100 150 200 250 300m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e A
bund
ance
156,0
245,1
217,9
309,2
291,2246,1
MS2
SDT MS
108,2157,0 311,1249,8235,7 289,2155,3126,9 171,8 194,8
020527s29#2363-2402 RT: 15,23-15,42 AV: 14SB 351 11,44-13,94, 16,63-20,08F:m/z Intensity Relative 156,0 1307402,1 100,00 245,1 1147143,5 87,74 217,9 480647,5 36,76 309,2 376500,5 28,80 291,2 194523,5 14,88 246,1 139796,9 10,69 108,2 125930,6 9,63 292,2 118436,7 9,06 302,0 102127,9 7,81 301,0 77033,1 5,89
SDT MS2 :+ c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
2DOM MS
D:\Doctoraat\...\data\020527s30 28-5-2002 17:18:06 S580C
RT: 10,07 - 24,94 SM: 7B
12 14 16 18 20 22 24Time (min)
0
20
40
60
80
100
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
0
20
40
60
80
10015,37
19,54
19,35
19,6918,4317,39
16,8813,31 19,9013,15 20,78
23,09
RT: 15,36MA: 505821740SN: 3990
16,24 18,4914,48 19,4713,0611,76 24,2421,78RT: 18,87MA: 550687778SN: 149
19,18
Spike at 10 times MRL
NL: 3,40E7m/z= 310,5-311,5 F: + c ESI Full ms [ 100,00-1000,00] MS 020527s30
020527s30 #2459-2495 : 15,26-15,43 : 13 SB: 533 11,36-14,26, 16,30-22,65 NL: 1,38E7F:
RT AV+ c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
14,9613,6011,25 20,0716,55 17,46 24,7223,28
NL: 4,06E7m/z= 107,5-108,5+155,5-156,5+217,5-218,5+244,5-245,5 F: + c ESI Full ms2 311,00@35,00 [ 100,00-320,00] MS 020527s30
NL: 4,99E7m/z= 320,5-321,5+338,5-339,5+356,5-357,5 F: + c ESI Full ms2 377,00@30,00 [ 100,00-385,00] MS 020527s30
100 150 200 250 300m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e A
bund
ance
156,0
245,1
218,0
246,0
MS2
SDT MS
108,0157,0
309,2291,2155,1 218,9 249,9109,1 201,6157,8 178,0 278,9
SDT MS2 020527s30#2459-2504 RT: 15,26-15,47 AV: 16SB 544 11,23-13,93, 16,63-23,77F:m/z Intensity Relative 156,0 12265833,8 100,00 245,1 11466612,8 93,48 218,0 4466156,9 36,41 246,0 1292496,0 10,54 108,0 1186452,1 9,67 157,0 753990,3 6,15 309,2 468434,2 3,82 291,3 408216,2 3,33 310,8 356913,3 2,91 155,1 315792,1 2,57
:+ c ESI Full ms2 311,00@35,00 [ 100,00-320,00]
2DOM MS
Fig. 6 Chromatographic and mass spectrometric data of a spike at MRL and a spike at 10
times MRL concentration of sulfadimethoxine (SDT)
110 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
3.2.5. Examples of identified analytes in routine analysis
3.2.5.1. Identification of penicilline G-benzathine
An extract of an injection site was directly infused into the mass spectrometer through a T-
piece. In positive ion mode major ions, m/z 241, 575 and 909, with a large signal-to-noise
ratio were observed. In negative ion mode ions with m/z 333, 573, 907 were acquired (Fig. 7).
Since electrospray is a soft ionisation technique the presence of a pseudo-molecular ion
([M+H]+ or [M-H]−) or an adduct can be expected. In this example two molecular masses
(908 and 574) can be derived from the positive and negative ions (positive ions: 909−1=908,
575−1=574) (negative ions: 907+1=908, 573+1=574). An analyte with molecular mass 240
can be protonated and give the positive ion with m/z 241. In a similar way the negative ion
with m/z 333 indicates an analyte with molecular mass 334.
D:\Doctoraat\...\data\010321s02 21-3-2001 9:55:25 K010125AESI-infusie
010321s02 #94-108 RT: 1,58-1,69 AV: 5 NL: 3,02E9F: + c Full ms [ 50,00-2000,00]
200 400 600 800 1000 1200 1400 1600 1800 2000m/z
0
10
20
30
40
50
60
70
80
90
100575,2
241,1
909,1577,2
Rel
ativ
e Ab
unda
nce
931,2242,1 416,2 597,2134,0 666,0566,6 1131,1837,2750,1 1265,1 1483,8977,4 1527,5 1877,31412,2 1747,51624,5 1936,1
010321s02 #989-1006 RT: 12,58-12,80 AV: 13 NL: 3,94E8F: - c Full ms [ 50,00-2000,00]
200 400 600 800 1000 1200 1400 1600 1800 2000m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e Ab
unda
nce
907,4
908,4573,4
333,11481,0
1147,3 1241,3365,2 1483,1910,3609,3495,5 1149,3 1816,0118,8 812,6669,1 1722,31430,4 1503,1307,3178,6 1272,7 1976,3941,0 1615,01083,4
Fig. 7 MS-full scan in positive and negative ion mode
The above mentioned data contain enough information to perform a targeted search. The
Merck Index is used as a starting point. A search is performed in the molecular mass range
111 Printed: www.dclsigns.be
Veterinary drugs
906-910. Three possibilities were examined: metocurine iodide, penicillin G-benzathine,
platonin. Because of the predominant presence of three ions, penicillin G-benzathine is of
most interest.
The molecular mass of penicillin G is 333.4. Penicillin G (Fig. 8), because of the presence of
carboxylgroups, shows a tendency to form negative ions. The negative ion with m/z 333 is an
indication of the presence of penicillin G.
NH
O
N
S
O
CH3
CH3
COOH Fig. 8 Chemical structure of penicillin G
Benzathine (Fig. 9) (MW=240.35) is a diamine that will preferentially become protonated.
This explains the presence of m/z 241 in positive ion mode. Penicillin G-benzathine contains
two penicillin G molecules and one benzathine molecule. Fragmentation and loss of one
penicillin G fragment gives the positive ion with m/z 575. An extra confirmation is the
presence of a sodium adduct (+ 23), the ion with m/z 931. MS2 fragmentation of the penicillin
G fragment (m/z 333 in negative ion mode) corresponds to the standard that was already
acquired in a different application.
NH
HN
Fig. 9 Chemical structure of benzathine
3.2.5.2. Interpretation of a florfenicol formulation (Nuflor®, Schering-Plough Animal Health)
Nuflor is an injectable solution formulation with the active analyte florfenicol (Fig. 10). The
average molecular mass is 358.21, the exact mass is 357 and the empirical formula is
C12H14Cl2FNO4S. Each chlorine atom occurs as two stable isotopes 35Cl and 37Cl, with an
112 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
abundance of 75.77 % and 24.23 %, respectively. Working with the exact mass, the expected
positive pseudo-molecular ion has a m/z of 358 and the expected negative pseudo-molecular
ion has a m/z of 356.
OH
HN
FO
Cl
Cl
SH3C
OO Fig. 10 Chemical structure of florfenicol
After infusion of a 100 ng µl−1 solution, florfenicol could only be detected in the negative ion
mode (Fig. 11). A distinct m/z 356 (35Cl) ion was observed combined with isotopic peaks m/z
358 (35 37 37Cl Cl) and m/z 360 ( Cl). Also a chlorinated adduct with m/z 392 (m/z 394 and m/z
396) was observed. Fragmentation of the adduct ions produced the original ions of florfenicol.
010511s40 #205-273 RT: 3,75-4,20 AV: 33 SB: 40 1,73-2,85, 3,34-4,52 NL: 3,53E6F: - c Full ms [ 150,00-1000,00]
300 310 320 330 340 350 360 370 380 390 400 410 420 430 440 450 460 470 480 490 500m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100 356,1
358,1
Rel
ativ
e A
bund
ance
392,0
394,1
446,0396,0360,1 415,8
448,0417,8
454,0336,2 470,0438,0398,1 497,9473,8418,9 457,0361,1 381,0 415,1322,2 478,9329,1 355,3 373,0339,2 425,0309,4 319,0
-1Fig. 11 MS-full scan of Nuflor 100 ng µl in negative ion mode
When the infusion concentration was lowered to 10 ng µl−1 chlorinated adducts dominated the
spectrum. The pseudo-molecular ion was reduced to a background ion (Fig. 12).
113 Printed: www.dclsigns.be
Veterinary drugs
010511s39 #354-369 RT: 5,31-5,45 AV: 13 NL: 4,40E7F: - c Full ms [ 240,00-900,00]
300 310 320 330 340 350 360 370 380 390 400 410 420 430 440 450 460 470 480 490m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100415,9
392,1
417,9 446,0
448,0
Rel
ativ
e Ab
unda
nce
470,0
396,1
356,2472,0
358,2419,9
450,0415,1341,2315,2453,9313,1 456,0397,0351,0 473,1359,2 420,9319,0 337,1305,1 373,0329,0 497,7432,6411,1381,1 462,9444,0 478,8
-1Fig. 12 MS-full scan of Nuflor 10 ng µl in negative ion mode
In positive ion mode the spectrum was dominated by ion clusters with a mass difference of
44. These clusters can be attributed to fragment ions of poly(ethylene glycol). The positive
ion spectrum could therefore not be used for further information (Fig. 13).
010511s40 #79-102 RT: 1,45-1,79 AV: 24 NL: 2,31E7F: + c Full ms [ 150,00-1000,00]
150 200 250 300 350 400 450 500 550 600m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100344,2
300,2
256,2437,4
388,3
481,4
393,4Rel
ativ
e A
bund
ance
432,3
525,4
349,4
305,4 340,1261,3212,1 375,1 456,0327,3
239,1221,0 476,2 569,4521,3283,2 438,4
457,9 482,5199,0 278,8181,8 409,2 549,3473,9 497,3355,3 597,2223,0 321,2 585,1168,0 415,2 557,0
499,6
-1Fig. 13 MS-full scan of Nuflor 100 ng µl in positive ion mode
114 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
®If Nuflor would be present in an injection site it would be very hard to determine the
presence of florfenicol because of the interference of poly(ethylene glycol) and the formation
of adduct ions. Therefore it is better to know the mass spectral data of the commercially
available veterinary medicinal product and not the pure standard. It is also important to infuse
a low and high (10 and 100 ng µl−1) concentration of the analyte because of the difference in
adduct formation.
3.2.5.3. Identification of prednisolone
Injection of the extract of an injection site showed in MS-full scan in negative ion mode the
ions with m/z 419 and 359 and in positive ion mode the ion with m/z 361 (Fig. 14). The
molecular mass 360 can be derived from the positive and negative ions (positive ions: 361-
1=360) (negative ions: 359+1=360, 419-60 (acetate-adduct)=359).
020403s11 #1508-1535 RT: 17,88-18,11 AV: 9 SB: 84 8,69-9,96, 12,16-13,89 NL: 4,57E8T: - c ESI Full ms [ 100,00-1000,00]
100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e A
bund
ance
419,2
329,4 420,2
778,9330,4 421,2 719,2359,2313,4295,5 501,0187,2 858,0582,7532,9172,4 673,0599,4 798,9 893,0 958,5125,3 251,3 989,2
020403s06 #3535-3565 RT: 17,86-17,99 AV: 8 SB: 116 16,26-17,60, 18,29-19,23 NL: 1,81E8T: + c ESI Full ms [ 100,00-1000,00]
100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000m/z
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e A
bund
ance
361,0
478,4742,9
720,8
744,0758,9710,7
486,9550,4362,0343,1
678,8
398,9 760,8 878,3542,9442,6 831,4304,3 925,1250,7 578,7506,7 624,4400,7 766,4661,4185,1 991,9104,6 223,1 949,9130,1
Fig. 14 MS-full scan in negative and positive ions mode
115 Printed: www.dclsigns.be
Veterinary drugs
MS2-full scan in positive ion mode of the ion with m/z 361 revealed a specific mass spectrum.
A large number of product ions was observed decreasing in intensity with decreasing m/z
(Fig. 15). This fragmentation pattern is typical for glucocorticosteroids [1].
020403s06 #3534-3567 RT: 17,85-18,00 AV: 9 SB: 155 16,24-17,63, 18,30-19,91 NL: 2,63E7F: + c ESI Full ms2 361,00@35,00 [ 100,00-361,00]
100 120 140 160 180 200 220 240 260 280 300 320 340 360m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100343,0
325,0307,1
289,1279,1
265,1147,2 297,1223,1181,0 263,1
Rel
ativ
e A
bund
ance
144,9 281,1163,1 277,1 342,4199,0 248,9237,3 360,4211,2 312,9182,6120,9 159,1132,9 351,2106,8
Fig. 15 MS2-full scan in positive ion mode of the ion with m/z 361
Using the Merck index a search was performed in the molecular mass range 359-361. Two
glucocorticosteroids were found: cortisone and prednisolone. MS2 fragmentation of the
extract corresponds to the standard of prednisolone (Fig. 16) (MW 360.44) that was already
acquired in a different application.
O
CH3
HOCH3 OH
OHO
Fig. 16 Chemical structure of prednisolone
116 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
3.2.5.4. Identification of tolfenamic acid
MS-full scan in negative mode of an extract of an injection side showed the intense ion with
m/z 260. No complementary positive ion was detected. In negative ion mode the distinct m/z
260 ion was combined with the isotopic peak witk m/z 262, in a ratio of 3:1 (Fig. 17). This
indicated the presence of one chlorine atom in the analyte.
D:\Doctoraat\...\data\020115s22 15-1-2002 14:43:32 K020034Aeluens KGRS/ ESI
020115s22 #2-29 RT: 0,04-0,77 AV: 28 NL: 3,37E8F: - c Full ms [ 80,00-2000,00]
200 400 600 800 1000 1200 1400 1600 1800 2000m/z
0
20
40
60
80
100260,2
262,2
Rel
ativ
e Ab
unda
nce
543,2
342,1216,3 372,2 826,0547,2 625,1118,9 1855,7498,6 1953,71757,8783,2 1556,41468,41258,6910,2 1414,51182,2 1715,21022,1
020115s22 #2-29 RT: 0,04-0,77 AV: 28 NL: 3,37E8F: - c Full ms [ 80,00-2000,00]
205 210 215 220 225 230 235 240 245 250 255 260 265 270 275 280 285 290 295 300m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
260,2
262,2
216,3 263,2218,3203,1 206,2 225,2214,3 281,5259,5 293,3 297,3279,3265,5 284,2255,5243,0 287,2273,4239,3227,2 250,4230,9 246,4
020115s22 #360-402 RT: 5,12-6,18 AV: 43 NL: 6,76E7F: + c Full ms [ 90,00-2000,00]
200 400 600 800 1000 1200 1400 1600 1800 2000m/z
0
20
40
60
80
100
Rel
ativ
e A
bund
ance
518,2
869,3542,2
780,4725,4 870,4
781,4 959,3 1095,6278,6 502,2426,2250,6 1140,2 1561,91296,6617,1 892,5681,9 1410,2164,6 1709,9 1834,21573,4224,4 1896,6372,2
Fig. 17 MS-full scan in negative and positive ion mode
After searching the Merck Index, tolfenamic acid (Fig. 18) was the most probable candidate.
The identity was confirmed after injection of the injectable solution Tolfedine®.
HN
COOH CH3
Cl
Fig. 18 Chemical structure of tolfenamic acid
117 Printed: www.dclsigns.be
Veterinary drugs
3.2.6. Discussion
Table 4 gives an overview of the veterinary medicinal products that were detected in non-
compliant injection sites in a concentration higher than the MRL for muscle tissue.
In 2002, 348 injection sites were analysed, 101 of them (29.0 %) were reported non-
compliant. The identity of the analytes is given in Table 4; 7.5 % of the identified analytes
were beta-lactam antibiotics (penicillin G-procaine and penicillin G-benzathine), 5.8 % were
NSAIDs (flunixin, tolfenamic acid and meloxicam), especially flunixin (4.6 %) and 4.6 %
were tetracyclines (oxytetracycline). Other analytes were not as frequently detected.
In 2003, a comparable amount of injection sites were analysed, namely 333 samples. Hundred
and five injection sites (31.5 %) were reported non-compliant; 6.9 % of the identified analytes
belonged to the group of NSAIDs, especially flunixin (3.0 %) and tolfenamic acid (3.3 %),
6.6 % were beta-lactam antibiotics and 3.3 % were tetracyclines.
In 2004, the number of analysed injection sites was significantly lower than in 2002 and 2003,
namely 201 injection sites. Fifty one samples (25.4 %) were non-compliant. The veterinary
groups that were recovered the most were, NSAIDs (6.0 %), tetracyclines (5.5 %) and
macrolides (4.0 %), especially tylosin (3.5 %).
The number of injection sites analysed this year, until 30 September, was even lower than in
2004. In 2005, 169 injection sites were analysed and 41 of them (24.3 %) were reported non-
compliant; 6.0 % of the identified analytes were beta-lactam antibiotics (penicillin G-procain
and amoxicillin), 6.0 % were NSAIDs and 4.8 % were macrolides (erythromycin, tilmicosin
and tylosin).
Other analytes which were identified at a lower percentage during these years, are classified
among the following groups of veterinary medicinal products: sulfonamides, quinolones,
glucocorticosteroids, pyrimidines, florfenicol, lincosamides, aminoglycosides and
anthelmintics.
118 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
Table 4 Overview of the veterinary medicinal products detected in non-compliant injection
sites in a concentration higher than the MRL
Analyte 2002 2003 2004 2005 n = 348 % n = 333 % n = 201 % n = 169 %
7.5 6.3 5.4 Penicillin G 26 21 2 1.0 9 Flunixin 16 4.6 3.0 3.5 3.0 10 7 5 Oxytetracycline 16 4.6 3.3 5.5 3.6 11 11 6 Sulfadimethoxine 6 1.7 9 2.7 4 2.0 2 1.2 Enrofloxacin 5 1.4 1 0.3 1 0.5 Erythromycin 4 1.1 4 1.2 1 0.5 1 0.6 Prednisolone 4 1.1 4 1.2 1 0.5 1 0.6 Tilmicosin 4 1.1 6 1.8 3 1.8 Trimethoprim 4 1.1 9 2.7 5 2.5 3 1.8 Tylosin 4 1.1 6 1.8 7 3.5 4 2.4 Dexamethasone 3 0.9 4 1.2 5 2.5 Florfenicol 3 0.9 1 0.5
3.3 Tolfenamic acid 3 0.9 11 4 2.0 4 2.4 Lincomycin 2 0.6 2 0.6 Meloxicam 1 0.3 1 0.3 Clorsulon 2 0.6 1 0.5 Amoxicillin 1 0.3 1 0.6 Methylprednisolone 1 0.3 Phenylbutazone 1 0.3 1 0.5 1 0.6 Spectinomycin 1 0.3 Ivermectine 1 0.5 Levamisole 1 0.6 TOTAL = 29.0 31.5 25.4 24.3 101 105 51 41
In this study (2002 untill 2005) beta-lactam antibiotics (penicillin G), NSAIDs (flunixin) and
tetracyclines (oxytetracycline) are the most commonly detected veterinary drugs in injection
sites. Also sulfadimethoxine, trimethoprim, tylosin and tolfenamic acid are frequently
detected (Fig. 19).
119 Printed: www.dclsigns.be
Veterinary drugs
0,0
1,0
2,0
3,0
4,0
5,0
6,0
7,0
8,0
perc
enta
ge (%
)
PEN G FLX OTC SDT TMP TYL TOLF
Injection sites 2002 - 2005
2002 2003 2004 2005
Fig. 19 Overview of the most frequently detected veterinary medicinal products in injection
sites, namely penicillin G (PEN G), flunixin (FLX), oxytetracycline (OTC), sulfadimethoxine
(SDT), trimethoprim (TMP), tylosin (TYL) and tolfenamic acid (TOLF)
In 2004 almost no penicillin was detected in the analysed injection sites, although penicillin G
was the most detected veterinary drug during the other years. In 2002 and 2003, penicillin G
was detected as penicillin G-benzathine and penicillin G-procain. At the end of 2003, there
was a withdrawal of penicillin G-benzathine from the market. The data recovered from the
analysis of injection sites clearly indicate this. In 2004 and 2005 no penicillin G-benzathin
was detected in any injection site.
The percentage of injection sites in which flunixin was detected was quite constant during the
four years. This observation could also be made for sulfadimethoxine and oxytetracycline.
In 2002, the percentage of injection sites containing trimethoprim and tolfenamic acid was
lower compared to the percentage of the following years. These two analytes were detected
for the first time in 2002, but before a sample can be reported as non-compliant, the standard
or injectable solution of the veterinary drug is necessary for identification and quantification.
During the following years (2003-2005) the percentage of injection sites in which
trimethoprim and tolfenamic acid were detected, was quite constant.
The percentage of injection sites containing tylosin is higher in 2004 and 2005. The reason
can be an increased use of tylosin during these two last years or these high percentages can
120 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
also be due to the lower number of analysed injection sites. The number of injection sites
containing tylosin is not increased during 2004 and 2005 (Table 4), it is the number of
analysed injection sites which is decreased. However, it is recommended to follow up these
observations during the next years.
In 2004 the Federal Agency for the Safety of the Food Chain has optimised its strategy
concerning the detection of veterinary medicinal products in injection sites. Until then, the
farm received a R-status when the residue of a veterinary drug was detected in the injection
site in a concentration higher than the MRL. However, since the injection site was removed
from the carcass, there was no longer a danger of possible consumption of the injection site
and its residues. Therefore, next to the injection site also some muscle tissue is collected at the
slaughterhouse. If the injection site contains a veterinary drug, both the injection site and the
muscle tissue will be analysed. Only the results of the muscle tissue will have legal and
financial consequences. If residues of a forbidden substance are found in the injection site, the
farm will receive an H-status. The corresponding muscle tissue does not need to be analysed.
In Table 5 the identity of the analytes present in non-compliant injection sites in 2004 and
2005 is summarized together with the percentage of the corresponding muscle tissue in which
the analyte was identified in a concentration higher than the MRL.
A problem encountered with this new strategy is the necessity of quantitative methods for the
detection of different groups of veterinary medicinal products. To develop and to use very
specific confirmation methods takes time and is too expensive considering the limited number
of analyses a year. Therefore, the semi-quantitative multi-residue approach applied for
injection sites will also be used for muscle tissue, unless a specific, quantitative confirmation
method is available in the laboratory. In this context and in co-operation with the Federal
Agency for the Safety of the Food Chain, a specific LC-MSn method was developed for the
group of quinolones (chapter 3.3) and the group of non-steroidal anti-inflammatory drugs
(chapter 3.4).
121 Printed: www.dclsigns.be
Veterinary drugs
Table 5 Overview of the veterinary medicinal products detected in non-compliant injection
sites in 2004 and 2005 and the percentage of the corresponding muscle tissue containing
concentrations of the veterinary medicinal product higher than the MRL
Analyte 2004 2005 Injection sites Muscle tissue Injection sites Muscle tissue
n = 201 % n % n = 123 % n % Penicillin G 2 1.0 1 50.0 9 5.4 1 11.1 Flunixin 7 3.5 5 3.0 1 20.0 Oxytetracycline 11 5.5 4 36.4 6 3.6 1 16.7 Sulfadimethoxine 4 2.0 2 1.2 Enrofloxacin 1 0.5 Erythromycin 1 0.5 1 0.6 Prednisolone 1 0.5 1 0.6 Trimethoprim 5 2.5 1 20.0 3 1.8 1 33.3 Tylosin 7 3.5 1 14.3 4 2.4 1 25.0 Dexamethasone 5 2.5 Florfenicol 1 0.5 1 100
Tolfenamic acid 4 2.0 4 2.4 Clorsulon 1 0.5 1 100 Phenylbutazone 1 0.5 1 0.6 1 100 Ivermectine 1 0.5 1 100 Tilmicosin 3 1.8 Amoxicillin 1 0.6 Levamisole 1 0.6
122 Printed: www.dclsigns.be
Identification and quantification of veterinary medicinal products in injection sites
3.2.7. Conclusion
In co-operation with the inspection services it was possible to screen a large number of
injection sites for the presence of a variety of veterinary medicinal products. Since we were
working with official samples, a fast and correct way of identifying the analyte and reporting
concentrations was mandatory.
A wide range of different groups of veterinary medicinal products are used in practice. Since
every group requires a specific extraction and detection procedure, it has become too
expensive and too time consuming to switch to different specific applications for only one
sample. Therefore, a multi-residue generic approach was developed. No specific method
development for the extraction, clean up and confirmation was needed. Because of the high
concentration of veterinary drugs in injection sites, there is no need for quantification of the
registered veterinary drugs in the concentration range of the MRL. An alternative validation is
used comparing the analyte concentration in the sample with the spike at MRL and 10 times
MRL concentration. The alternative approach is performed as a mini-validation. Identification
is based on the collected data of the different injectable and standard solutions of registered
veterinary products and the database of the Merck-index. Extraction and identification can be
performed within 48 hours. If the identified compound also needs to be quantified an extra 24
hours are necessary before the result can be reported.
This illustrates the advantage of using infusion-MSn or LC-MSn with a default gradient as a
fast screening and confirmation technique for highly concentrated samples. This multi-residue
method allows the detection of known compounds, but also unknown, new products can be
detected through MS screening. Structure elucidation is possible by applying tandem mass
spectrometry. This is in contrast with confirmatory MSn methods which are able to detect low
concentration of a selected group of veterinary medicinal products, but neglect changes in the
chemical structure of a compound due to the selectivity of the method.
123 Printed: www.dclsigns.be
Veterinary drugs
3.2.8. References
[1] K. De Wasch, H. De Brabander, D. Courtheyn, C. Van Peteghem (1998) Identification of
corticosteroids in injection sites and cocktails by MSn, The Analyst 123, 2415-2422
[2] ThermoElectron, Xcalibur 1.2
[3] Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive
96/23/EC concerning the performance of analytical methods and the interpretation of results
(2002) Official Journal of the European Communities, no. L 221
[4] Anonymous (2003) Informal consolidated version of the Annexes I to IV of Council
Regulation n° 2377/90, The European Agency for the Evaluation of Medicinal Products, 22
July 2003
124 Printed: www.dclsigns.be
Chapter 3.3 n Multi-residue LC-MS method for the detection of quinolones in muscle and
bovine milk
Adapted from:
N. Van Hoof, K. De Wasch, L. Okerman, W. Reybroeck, S. Poelmans, H. Noppe and H. De
Brabander
Validation of a liquid chromatography-tandem mass spectrometric method for the quantification of
eight quinolones in bovine muscle, milk and aquacultured products
Analytica Chimica Acta (2005) 529, 265-272
And extended with extra validation data
3.3.1. Introduction
The use of fluoroquinolones in veterinary medicine has increased tremendously in the last ten
years. Fluoroquinolones are synthetic antibacterial agents. The first agent of this family,
introduced in veterinary medicine was enrofloxacin. The advantages of the fluoroquinolones
are that they are rapid bactericidal agents against a wide variety of clinically important
bacterial organisms. Fluoroquinolones are potent, well-tolerated by animals, and can be
administered by a variety of routes.
Nalidixic acid was the first quinolone developed in the early 1960s and was used for treatment
of urinary tract infections in humans. It had a limited activity against gram-negative bacteria.
Second generation quinolones clearly showed improved antibacterial activity as well as
pharmacokinetic properties. Some of these quinolones, such as flumequine and oxolinic acid
are still used in veterinary medicine. These original quinolones have modest activity against
Enterobacteriaceae and some other gram-negative bacteria. A significant breakthrough was
achieved by the introduction of fluorinated derivatives, called fluoroquinolones, during the
1980s. These fluoroquinolones extended the spectrum of activity to include Pseudomonas
aeruginosa and some gram-positive bacteria, and they have a substantially increased activity
against gram-negative bacteria. Emerging resistance to fluoroquinolones has led to the
125 Printed: www.dclsigns.be
Veterinary drugs
development of newer agents through the addition of a methoxy side chain [1-3]. These drugs,
such as grepafloxacin, trovafloxacin and premafloxacin, have increased activity against gram-
positive cocci and anaerobic bacteria and may offer advantages to treat certain infections [1].
Quinolones are used both in human and veterinary medicine. In veterinary medicine, they are
used for the treatment of pulmonary, urinary and digestive tract infections [4]. No data are
available on the use of the newest generation of fluoroquinolones in veterinary medicine . The
general structure of quinolones consists of a 1-substituted-1,4-dihydro-4-oxopyridine-3-
carboxylic moiety combined with an aromatic or heteroaromatic ring (Fig. 1). The carboxylic
acid at position 3 and the ketone group at position 4 are necessary for DNA gyrase inhibition
(mechanism of action), whereas substitutions at position 1 and 7 influence the potency and
biological spectrum of activity of the drugs [3,5-6].
126 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
HN
N N
COOH
O
F
N
N
H3C
N
COOH
O
F
enrofloxacin ciprofloxacin
HN
N N
COOH
O
F
F
N
N N
COOH
O
F
F
H3C
sarafloxacin difloxacin
N
N N
COOH
O
F
H3CNO
CH3
N
N N
COOH
O
F
H3C
danofloxacin marbofloxacin
N
COOH
O
F
CH3
N
COOH
O
CH3
O
O
flumequine oxolinic acid
Fig. 1 Chemical structure of some quinolones
127 Printed: www.dclsigns.be
Veterinary drugs
3.3.1.1. Mechanism of action
Fluoroquinolones inhibit two enzymes of the DNA metabolism in bacteria, topoisomerase II
(DNA gyrase) and topoisomerase IV. DNA topoisomerases are responsible for separating the
strands of duplex bacterial DNA, inserting another strand of DNA, and then resealing the
originally separated strands. In gram-negative organisms DNA gyrase is the primary target,
whereas in gram-positive bacteria topoisomerase IV was found to be most affected [1-3,7].
Models to explain the activity of quinolones at the target site only exist for DNA gyrase. The
DNA gyrase is composed of two subunit pairs (gyr-A and gyr-B). The gyr-A subunits initially
bind to the double stranded DNA helix. Both DNA strands are cleaved and the 5’-ends of the
DNA chain are thereby bound covalently to the gyr-A subunits. When the DNA is present as
single strands, quinolone molecules interact with the gyrase-DNA complex. This stabilizes
the intermediate stage of this reaction step. In this way the normal bacterial DNA synthesis is
inhibited, leading to irreversible DNA damage and, finally, cell death [2].
Mammals are resistent to the killing effects of quinolone antimicrobials because
topoisomerase II in mammalian cells is not inhibited until the drug concentration reaches 100-
1000 µg ml-1. Bacteria are inhibited by concentrations less than 0.1-10 µg ml-1 [1].
3.3.1.2. Resistance
Bacterial resistance is not a targeted process. It occurs by chance and becomes more frequent
when selective pressure on the bacterial flora is present. Resistance to quinolones primarily
occurs through decreased binding affinity of these drugs to bacterial topoisomerase II or IV
caused by chromosomal mutation (single step type) (1), alterations in bacterial cell wall
penetration, changing the cell walls to exhibit fewer passageways (2), or through expression
of multidrug resistent membrane-associated efflux pumps, resulting in decreased intracellular
drug concentrations (3).
Bacteria resistant to one fluoroquinolone agent are generally cross-resistant to other
fluoroquinolones. This cross resistance includes fluoroquinolones used in animals and those
available for human use [1,2,7-8].
3.3.1.3. Human health risks
A variety of antimicrobials are used in livestock production. Their use inevitably leads to
selection of resistant bacteria in the ecosystem. The use of quinolones in livestock is an area
of particular concern because of the significance of this group of antimicrobials for the
128 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
treatment of a broad range of infections in humans including gastrointestinal infections by
zoonotic bacteria transmitted to humans via the food chain. Transfer of fluoroquinolone
resistance from animals to people has been suggested to occur for Campylobacter species and
Salmonella typhimurium.
Campylobacter species are the most common cause of bacterial gastrointestinal infections in
humans throughout the world. Contaminated food is the usual source of human infections;
therefore the risk of fluoroquinolone-resistant strains in the food chain has raised concerns
that the treatment of human infections will be comprised. However, most cases of
Campylobacter enteritis do not require antimicrobial treatment and transmission of resistant
strains to humans is infrequent.
Resistance to fluoroquinolones in Campylobacter has clearly increased over the past decade
in many parts of the world, and this period coincides with the introduction of
fluoroquinolones in poultry. Links between quinolone use in animals and the occurrence of
problems in infectious disease treatment in humans remains to be elucidated. Nevertheless,
the continuous use of fluoroquinolones in livestock is a public health risk because it can
potentially lead to resistant mutants being passed on to humans through the food chain [1, 9-
15].
3.3.1.4. Legislation
The administration of quinolones to animals used for human consumption can generate
residues in food products. These residues represent a potential hazard for the consumer. The
European Union has set Maximum Residue Limits (MRL) for quinolones with the aim of
minimising the risk to human health associated with the consumption of quinolone residues.
Table 1 summarizes the MRL concentrations in the animal species bovine, porcine, poultry
and fish and this for the tissues muscle and milk [16].
129 Printed: www.dclsigns.be
Veterinary drugs
Table 1 The MRLs for quinolones
Pharmacologically active substance Animal species MRL Target tissue
bovine 100 μg kg-1 muscle Enrofloxacin (enrofloxacin +
ciprofloxacin) 100 μg kg-1 milk
porcine 100 μg kg-1 muscle
poultry 100 μg kg-1 muscle
all food producing
species
100 μg kg-1 muscle
Sarafloxacin Salmonidae 30 μg kg-1 muscle
Danofloxacin bovine 200 μg kg-1 muscle
30 μg kg-1 milk
chicken 200 μg kg-1 muscle
all food producing
species
100 μg kg-1 muscle
Oxolinic acid bovine 100 μg kg-1 muscle
porcine 100 μg kg-1 muscle
chicken 100 μg kg-1 muscle
fin fish 100 μg kg-1 muscle
Flumequine bovine 200 μg kg-1 muscle
50 μg kg-1 milk
porcine 200 μg kg-1 muscle
poultry 400 μg kg-1 muscle
fin fish 600 μg kg-1 muscle
Salmonidae 600 μg kg-1 muscle
all food producing
species
200 μg kg-1 muscle
Difloxacin bovine 400 μg kg-1 muscle
porcine 400 μg kg-1 muscle
poultry 300 μg kg-1 muscle
all food producing
species
300 μg kg-1 muscle
Marbofloxacin bovine 150 μg kg-1 muscle
75 μg kg-1 milk
porcine 150 μg kg-1 muscle
130 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
3.3.2. Method setup
For the determination of quinolones in biological matrices several spectroscopic techniques,
such as ultraviolet (UV), fluorescence or mass spectrometry (MS) are used in combination
with liquid chromatography (LC). Earlier methods used UV almost exclusively [17-20], but
more recent systems use fluorescence detection [17-18,21-36]. These procedures are,
however, restricted to a limited number of quinolones. Since several years LC with MS
detection has been used for confirmatory analysis because this detection method is more
sensitive, selective and allows rapid and multi-residue determination in complex matrices and
gives structural information [4,10,17,37-43].
In this work a LC-ESI-MSn multi-residue method was developed allowing the detection of
eight quinolones: enrofloxacin, ciprofloxacin, sarafloxacin, danofloxacin, oxolinic acid,
flumequine, difloxacin and marbofloxacin. Therefore, a simple and rapid extraction and
clean-up method was developed for the matrices bovine/porcine/chicken muscle, muscle of
aquacultured products and bovine milk. All quinolones were analysed in a single
chromatographic run at MRL level. An ion trap mass spectrometer was used as identification
as well as quantification method instead of the more commonly used quadrupole mass
spectrometer [4,10,37-39,42-43].
3.3.3. Experimental
3.3.3.1. Reagents and chemicals
The quinolone standards, enrofloxacin and ciprofloxacin were obtained from ICN
Biomedicals (Irvine, CA, USA) while flumequine and oxolinic acid were from Sigma–
Aldrich (St. Louis, MO, USA), marbofloxacin from Vetoquinol (Aartselaar, Belgium),
danofloxacin and sarafloxacin from DVK-CLO (Melle, Belgium) and difloxacin from
Laboratory of Hygiene and Technology, Department of Veterinary Public Health and Food
Safety (Ghent, Belgium). The internal standard quinine was obtained from ICN Biomedicals
(Irvine, CA, USA). All chemicals used were of analytical grade from Merck (Darmstadt,
Germany) and Acros (Geel, Belgium).
Stock standard solutions of 1000 ng μl−1 were prepared in ethanol for enrofloxacin,
danofloxacin, difloxacin and marbofloxacin; in HPLC-water for ciprofloxacin and in 0.1 M
NaOH for flumequine, oxolinic acid and sarafloxacin. For the preparation of working
solutions HPLC-water was used. All standard and working solutions were stored at -20 °C.
131 Printed: www.dclsigns.be
Veterinary drugs
3.3.3.2. Instrumentation
The HPLC apparatus comprised of a 1100 series quaternary pump and an autosampler of
Hewlett Packard (Palo Alto, CA, USA). Chromatographic separation was achieved using a
Symmetry C18 column (5 μm, 150 mm × 2.1 mm, Waters, Milford, USA). The mobile phase
consisted of a mixture of methanol with 0.1 % trifluoroacetic acid (A) and water with 0.1 %
trifluoroacetic acid (B). A linear gradient was run (20 % A for 5 min and increasing to 100 %
in the next 10 min) at a flow rate of 0.3 ml min-1.
LC-MS2 detection was carried out with a ThermoFinnigan LCQ Deca ion trap with
electrospray ionisation (ESI) interface in positive ion mode (San José, CA, USA). The MS
detector was operated in three time segments each divided in different scan events, so the
quinolones were separated both chromatographically and massspectrometrically. Each analyte
was evaluated based on the product ions present in the mass spectra (Table 2).
Table 2 Instrument parameters of the LC-MS2 method for the detection of quinolones
Segment Scan event Analyt Precursor ion → product ions Mass range
Segment 1 Scan event 1 Quinine = IS 325 → 184,198,253,264,307 100 - 330
Scan event 2 Marbofloxacin 363 → 276,320,245 200 - 370
Segment 2 Scan event 1 Enrofloxacin 360 → 316,342 200 - 370
Scan event 2 Ciprofloxacin 332 → 288,314 200 - 340
Scan event 3 Sarafloxacin 386 → 342,368 200 - 390
Scan event 4 Danofloxacin 358 → 314,340 200 - 365
Scan event 5 Difloxacin 400 → 356,382 200 - 410
Segment 3 Scan event 1 Flumequine 262 → 244 Oxolinic acid 200 - 270
Scan event 2 276 Flumequine 200 - 500 Oxolinic acid
132 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
3.3.3.3. Extraction and clean-up
Muscle tissue
To an amount of 2 g of minced muscle tissue 100 μg kg−1 quinine was added as internal
standard. The quinolones were extracted from the muscle tissue using 20 ml ultrapure water.
After mixing and centrifugation (5 min, 5500 rpm) only 10 ml supernatant was used for
further clean-up. The clean-up was carried out using an Isolute 500 mg C18 SPE Cartridge
(IST International, Mid Glamorgan, UK). The columns were conditioned with 2 ml MeOH
and 4 ml water. After application of the extract, the cartridge was rinsed with 2 ml
MeOH/water (20:80), 2 ml hexane and vacuum dried. The quinolones were eluted from the
column with 3 ml 1 % trifluoroacetic acid in acetonitrile. The eluate was evaporated to
dryness at 45 °C under a stream of nitrogen. The residues were reconstituted in 30 μl
methanol with 0.1 % trifluoroacetic acid and 120 μl water with 0.1 % trifluoroacetic acid
before injecting 15 μl on the HPLC column.
Bovine Milk −1To an amount of 2 ml milk 100 μg kg quinine was added as internal standard. To precipitate
the proteins present in the milk, 2.5 ml trichloroacetic acid (20 % in methanol) was added.
After mixing and centrifugation (10 min, 5500 rpm) the quinolones were extracted from the
supernatant using 10 ml ultrapure water. The entire supernatant was used for further clean-up
after mixing and centrifugation (10 min, 5500 rpm). The clean-up was analogous to the one
described for muscle tissue.
3.3.4. Results
3.3.4.1. LC-MS2 method
Since most quinolones are fluorescent, liquid chromatography with fluorescence detection is
mainly used as determination method for routine residue analysis [17-18,21-36]. Fluorescence
depends strongly on the pH of the medium. The highest fluorescence is obtained at a pH value
ranging from 2.5 to 4.5, whereas the anionic species do not generally show native
fluorescence. Marbofloxacin has a poor native fluorescence and therefore has almost
exclusively been determined with UV detection [17-18]. In this paper the more sensitive and
selective detection method mass spectrometry was chosen. Eight different quinolones, in
which marbofloxacin, could be determined with this detection method in a single
chromatographic run. Most mass spectrometry methods for the identification and
133 Printed: www.dclsigns.be
Veterinary drugs
134
quantification of quinolones used a quadrupole mass spectrometer that monitors specific
transitions (precursor ion-product ion) of each quinolone. In this paper an ion trap mass
spectrometer was used as identification as well as quantification method. So full scan MS2-
mass spectra of each quinolone were recorded.
The standards marbofloxacin, enrofloxacin, ciprofloxacin, sarafloxacin, danofloxacin,
difloxacin, oxolinic acid, flumequine and the internal standard quinine were spiked to blank
muscle tissue (bovine/chicken/porcine muscle and muscle of aquacultured products) and
blank bovine milk at the MRL concentration of each quinolone. Fig. 2 shows the extracted ion
chromatograms and the MS2-mass spectra of the different quinolones in bovine muscle.
Comparable chromatograms and MS2-mass spectra were obtained for the matrices chicken
muscle, porcine muscle, muscle of aquacultured products and bovine milk. Fig. 2 shows all
quinolones at their MRL concentration.
RT: 0,00 - 20,00 SM : 7B
0 5 10 15 20Time (min)
0
20
40
60
80
1000
20
40
60
80
1000
20
40
60
80
1000
20
40
60
80
1000
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
0
20
40
60
80
1000
20
40
60
80
1000
20
40
60
80
100 5,42
3,755,34
3,85 7,799,82
11,649,329,62
11,6510,65
9,09
9,97
10,75 11,819,3510,26
11,589,1612,86
14,40
12,38 15,41 18,57
NL: 1,03E7m/z= 183,5-184,5+197,5-198,5+252,5-253,5+263,5-264,5 F: + c ESI Full ms2 325,00@37,00 [ 100,00-330,00] M S 030619s25
NL: 3,51E5m/z= 275,5-276,5+319,5-320,5+344,5-345,5 F: + c ESI Full ms2 363,00@30,00 [ 200,00-370,00] M S 030619s25
NL: 5,05E6m/z= 315,5-316,5+341,5-342,5 F: + c ESI Full ms2 360,00@30,00 [ 200,00-370,00] M S 030619s25
NL: 1,51E6m/z= 287,5-288,5+313,5-314,5 F: + c ESI Full ms2 332,00@30,00 [ 200,00-340,00] M S 030619s25
NL: 1,02E6m/z= 313,5-314,5+339,5-340,5 F: + c ESI Full ms2 358,00@30,00 [ 200,00-365,00] M S 030619s25
NL: 4,42E6m/z= 355,5-356,5+381,5-382,5 F: + c ESI Full ms2 400,00@30,00 [ 200,00-410,00] M S 030619s25
: 1,21E7/z= 243,5-244,5 F: + c ESI Full s2 262,00@30,00 [ 0,00-270,00] M S 030619s25
030619s25 # 51-65 : 4,94-5,83 : 8 : 22 1,29-3,95, 7,00-12,30 NL: 4,59E6F:
NL: 4,41E5m/z= 341,5-342,5+367,5-368,5 F: + c ESI Full ms2 386,00@35,00 [ 200,00-390,00] M S 030619s25
NLmm7
RT AV SB+ c ESI Full ms2 325,00@37,00 [ 100,00-330,00]
100 150 200 250 300m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
307,1
184,1264,1253,1
198,1
279,1160,1
110,0 226,1174,1 210,1 290,1252,1134,0 138,1 308,0
030619s25 # 284-325 : 10,41-10,74 : 8 : 11 3,65-9,23, 12,22-17,51 NL: 3,10E5F:
RT AV SB+ c ESI Full ms2 386,00@35,00 [ 200,00-390,00]
200 250 300 350m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
386,0
342,1
368,1
343,1322,2 369,1299,1255,3 289,2216,8 237,3
030619s25 # 48-61 RT: 4,64-5,46 : 7 : 26 0,89-4,05, 7,00-14,21 NL: 2,59E5F:
AV SB+ c ESI Full ms2 363,00@30,00 [ 200,00-370,00]
200 220 240 260 280 300 320 340 360m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
319,9
363,0
345,1275,9
319,1276,9233,0 335,1 362,2299,4 364,1266,1253,9213,1
030619s25 # 203-237 : 9,72-9,97 7 17 1,68-9,13, 11,64-17,90 NL: 1,23E6F:
RT AV: SB:+ c ESI Full ms2 358,00@30,00 [ 200,00-365,00]
200 220 240 260 280 300 320 340 360m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
358,0
314,1
340,1
216,1283,2 291,2257,2 271,1219,0 360,1245,1 353,2326,8211,0
030619s25 # 194-225 : 9,65-9,90 : 7 : 36 2,96-8,73, 10,36-16,48 NL: 4,03E6F:
RT AV SB+ c ESI Full ms2 360,00@30,00 [ 200,00-370,00]
200 220 240 260 280 300 320 340 360m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
316,1
360,0
342,1
359,1315,1 341,1245,1 295,2237,0 287,9256,9202,0
030619s25 # 237-284 : 10,01-10,39 : 10 : 2,24E6F:
RT AV NL+ c ESI Full ms2 400,00@30,00 [ 200,00-410,00]
200 250 300 350 400m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
400,0356,1
382,1
357,1401,0
299,0 381,1 402,0340,8285,9 308,0267,9202,7 239,8
030619s25 # 166-209 9,38-9,74 9 : 25 2,17-8,18, 10,81-16,36 NL: 9,02E5F:
RT: AV: SB+ c ESI Full ms2 332,00@30,00 [ 200,00-340,00]
200 220 240 260 280 300 320 340m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
332,0
288,1
314,1
268,1245,1 333,2286,0 304,1204,1 223,2 231,1 315,0
030619s25 # 470-491 12,61-13,06 22 : 113 3,55-11,44, 14,43-18,65 NL: 8,65E6F:
RT: AV: SB+ c ESI Full ms2 262,00@30,00 [ 70,00-270,00]
200 210 220 230 240 250 260 270m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
262,0
244,1
263,0245,1 261,2216,0 234,1202,8 252,2243,2230,1220,0211,9
Fig. 2 Ion chromatograms and MS2-mass spectra of quinine (quin) (I.S.), marbofloxacin
(marbo), enrofloxacin (enro), ciprofloxacin (cipro), sarafloxacin (sara), danofloxacin (dano),
oxolinic acid (oxo) and flumequine (flum) in bovine muscle
quin quin
marbo
marbo enro
cipro
enro
cipro
sara
sara
dano
di
dano
oxo
di
flum
oxo and flum
Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
23.3.4.2. Validation of the LC-MS method for the detection of quinolones in bovine muscle
The different quinolones are registered in the European Union for use in bovine species, but
they have a MRL (Table 1) [16]. Therefore, a quantitative confirmation method is required.
Oxolinic acid and sarafloxacin are an exception. Oxolinic acid is an Annex III compound with
a provisional MRL of 100 µg kg-1. This provisional MRL will expire on 1.1.2006 [44].
Sarafloxacin, on the contrary, is not registered for use in bovine species [16].
Specificity 2The specificity of the method could be demonstrated by LC-MS analysis of blank muscle
tissue. No interferences were observed after analysis of these blank samples and after analysis
of spiked bovine muscle with the eight quinolones.
Selectivity
Quinolones are veterinary drugs with a MRL, so the minimum number of identification points
(IP) is set to three. LC-MSn precursor ions earn 1 IP and LC-MSn product ions earn 1.5 IP
[44]. MS2-full scan of the pseudo-molecular ion of the quinolones enrofloxacin, ciprofloxacin,
sarafloxacin, danofloxacin and difloxacin each showed two product ions; a loss of 18 due to
the loss of water and a loss of 44 due to the loss of the carboxylic acid group (Fig. 3). 030619S19 #206-240 RT: 9,64-9,89 AV: 7 NL: 3,81E6F: + c ESI Full ms2 360,00@30,00 [ 200,00-370,00]
200 220 240 260 280 300 320 340 360m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
316,14
360,04
342,12
359,15315,12 341,14245,30 266,09 288,03 361,34317,20200,34 274,10 342,85225,13 299,27258,19
Fig. 3 MS2-full scan of enrofloxacin spiked to blank bovine muscle at a concentration of 100
µg kg-1
Fragmentation of the pseudo-molecular ion m/z 262 of the quinolones oxolinic acid and
flumequine only showed the product ion m/z 244, due to the loss of water (Fig. 4). So 2.5 IP
135 Printed: www.dclsigns.be
Veterinary drugs
were earned. Therefore the ion with m/z 276 in MS-full scan was also used as a precursor ion,
so 3.5 IP were earned. The ion with m/z 276 is an adduct ion of the pseudo-molecular ion
with m/z 262. A mass of 14 was added to the pseudo-molecular ion. The origin of this adduct
ion is unclear. The addition of mass 14 has not yet been mentioned in the literature. MS2-full
scan of the ion with m/z 276 showed the ion with m/z 262, so after fragmentation of the
adduct ion the pseudo-molecular ion was revealed. Hence, the ion with m/z 276 is clearly an
adduct ion and not an impurity since fragmentation of this ion revealed the same product ions
as fragmentation of the pseudo-molecular ion. If a sample contains flumequine or oxolinic
acid MS3-full scan of the ion m/z 262 will be obtained in an extra run for the confirmation of
these quinolones (Fig. 5).
030619S19 #488-497 RT: 12,80-12,93 AV: 10 NL: 1,30E7F: + c ESI Full ms2 262,00@30,00 [ 70,00-270,00]
80 100 120 140 160 180 200 220 240 260m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
261,96
244,11
262,97
245,09243,4192,74 110,08 120,75 203,07137,83 215,99189,20155,98 161,97 263,81
Fig. 4 MS2-full scan of oxolinic acid and flumequine spiked to blank bovine muscle at a
concentration of 100 µg kg-1040622s02 #234-317 RT: 2,89-4,39 AV: 84 NL: 2,60E6F:
040622s01 #243-270 RT: 2,72-3,21 AV: 28 NL: 1,12E6F:+ c Full ms3 262,00@30,00 244,00@39,00 [ 65,00-250,00] + c Full ms3 262,00@30,00 244,00@39,00 [ 65,00-250,00]
80 100 120 140 160 180 200 220 240m/z
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
220,0
238,1
202,3
234,0
176,2216,3
200,3 244,3174,3148,4 229,0158,1 215,1188,4146,3130,2117,378,9 107,597,8
80 100 120 140 160 180 200 220 240m/z
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
234,1
216,3
200,2
158,3
188,3248,1176,2170,3142,3 148,2 244,3229,3201,2186,2130,3118,3
198,4 242,3115,4 226,5160,2149,2 204,0103,5 184,491,4 132,3
Fig. 5 MS3-full scan of flumequine (left) and oxolinic acid (right)
136 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
MS2-fragmentation of the ion with m/z 363 of the quinolone marbofloxacin had a typical
MS2-mass spectrum with three product ions, m/z 276, 320 and 345 (Fig. 6).
030619S19 #48-63 RT: 4,64-5,58 AV: 8 NL: 2,40E5F: + c ESI Full ms2 363,00@30,00 [ 200,00-370,00]
200 220 240 260 280 300 320 340 360m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Rel
ativ
e A
bund
ance
319,86
363,01
345,04
275,94
362,17276,89 344,05232,83 335,11 364,15318,54246,76 302,03268,70 348,51218,82 250,12212,94 286,90
Fig. 6 MS2-full scan of marbofloxacin spiked to blank bovine muscle at a concentration of
150 µg kg-1
In the MS2-mass spectra of all the quinolones the pseudo-molecular ion was still clearly
present. There was no improvement by increasing the collision energy. This phenomenon
could not be explained.
In Table 2 the precursor ions and product ions of each quinolone are summarised. The
different quinolones can be identified according to the criteria of Commision Decision
2002/657/EEC by their MS2 and MS3 +-full scan spectra of the pseudo-molecular ion [M+H]
[45].
Calibration curves
The chromatographic peak areas, used for the quantification were calculated from the
extracted ion chromatograms of the most abundant product ions. These product ions are
shown in Table 2.
The calibration curves obtained for spiked bovine muscle were linear in the concentration
range 0.5 times MRL to 1.5 times MRL for the eight quinolones. The coefficients of
correlation ranged between 0.91 and 0.99.
137 Printed: www.dclsigns.be
Veterinary drugs
Recovery
Since no Certified Reference Material is available, the recovery is determined by experiments
using fortified blank bovine muscle tissue. For samples spiked at a concentration above 10 μg
kg−1, the recovery should range from 80 to 110 % [45]. Eighteen samples of blank bovine
muscle were fortified at 0.5, 1 and 1.5 times the MRL. The mean recovery from the six results
at each level was calculated. All these samples had a recovery within the permitted range.
Table 3 summarises the recovery at each level for the different quinolones.
Precision
The coefficient of variation (CV) for the repeated analysis of fortified blank bovine muscle,
should not exceed the level calculated by the Horwitz equation [45]. For mass fractions lower
than 100 μg kg−1 the application of the Horwitz equation gives unacceptable high values.
Therefore the CV for concentrations lower than 100 μg kg−1 should be as low as possible. In
that case 23 % is taken as a guideline (CV at 100 μg kg−1 = 23 %). Eighteen samples of blank
bovine muscle were fortified at 0.5, 1 and 1.5 times the MRL. Table 3 summarise the CV at
each level and the overall CV for each quinolone. These coefficients of variation were lower
than the permitted CV.
To calculate the intra-laboratory coefficient of variation, these analyses were repeated on two
other occasions by the same analyst, under repeatability conditions. The overall mean
concentration, standard deviation and coefficient of variation of these fifty four fortified blank
samples, were calculated. The intra-laboratory coefficient of variation would typically be
between one half and two third of the CV calculated by the Horwitz equation. The intra-
laboratory CV for all quinolones was within the permitted range except for oxolinic acid,
which had a CV of 18.1 % which is slightly higher than the maximum tolerance of 15.3 %.
To calculate the within-laboratory coefficient of variation, these analyses were repeated on
another occasion under reproducibility conditions (by a different analyst). The coefficient of
variation of this set of samples was calculated. The within-laboratory coefficient of variation
should not exceed the coefficient of variation at 0.5 times the MRL concentration. The within-
laboratory CV of all quinolones, except flumequine, was lower than the CV at 0.5 MRL. The
within-laboratory CV of flumequine were slighltly higher than the CV at 0.5 MRL.
In Table 4 the intra-laboratory and the within-laboratory coefficients of variation are
summarized.
138 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
Table 3 The validation parameters recovery and coefficient of variation (CV) at the levels 0.5,
1 and 1.5 times the MRL for the different quinolones in bovine muscle
Marbofloxacin Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%)
75 105 10.6 150 95 5.0 225 102 1.9
Overall CV (%) = 7.7 Enrofloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 50 100 12.0
100 100 3.3 150 100 3.2
Overall CV (%) = 7.0 Ciprofloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 50 99 9.3
100 101 7.0 150 100 4.0
Overall CV (%) = 6.7 Sarafloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 15 99 9.0 30 101 6.6 45 100 6.8
Overall CV (%) = 7.1 Danofloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 100 98 10.5 200 102 7.2 300 99 4.0
Overall CV (%) = 7.4 Difloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 200 95 11.5 400 105 8.0 600 98 5.0
Overall CV (%) = 9.0 Oxolinic acid
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 50 91 39.7
100 109 2.4 150 97 3.7
Overall CV (%) = 21.4 Flumequine
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 100 100 5.6 200 100 5.9 300 100 3.6
Overall CV (%) = 4.8
139 Printed: www.dclsigns.be
Veterinary drugs
Table 4 The validation parameters coefficient of variation (CV), decision limit (CCα) and
detection capability (CCβ) for the different quinolones
marbo enro cipro sara dano di oxo flum
Intra-laboratory CV (%) 12.0 11.9 12.5 11.4 10.7 9.3 18.1 11.2
Within-laboratory CV
(%)
9.2 7.4 9.2 7.7 10.2 7.6 10.8 6.4
CCα (µg kg-1) 178.8 119.1 120.7 35.5 234.6 459.8 128.0 236.9
CCβ (µg kg-1) 208.3 138.6 140.4 41.1 269.7 520.7 157.3 273.6
Decision limit (CCα)
The decision limit is the limit at and above which it can be concluded with an error
probability of α that a sample is non-compliant. The data used to calculate the intra-laboratory
coefficient of variation, are applied to determine the decision limit CCα. The corresponding
mean concentration at the MRL level plus 1.64 times the standard deviation equals the
decision limit (α = 5 %). In Table 4 the CCα for each quinolone is summarized.
Detection capability (CCβ)
In the case of substances with a maximum residue limit (MRL), the detection capability is the
concentration at which the method is able to detect MRL concentrations with a statistical
certainty of 1-β. The detection capability CCβ was calculated as the decision limit CCα plus
1.64 times the corresponding standard deviation (β = 5%), supposing that σCCα equals σMRL. In
Table 4 the CC β for each quinolone is summarized.
Applicability and ruggedness
Applicability and ruggedness can best be tested when the analytical method is used for routine
analysis. Applicability is the observation of the consequences when minor reasonable
variations are introduced into the method. Such factors may include the analyst, temperature
during evaporation, pH values, as well as many other factors that can occur in the laboratory.
Applicability and ruggedness will be tested by control spiked samples or by participating in
performance studies. With each batch of samples a control spiked sample will be analysed and
once a year unknown control spiked samples are analysed.
140 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
Stability
During one year the stability of a working solution containing the different quinolones, was
evaluated. In the beginning a stock standard solution was prepared of the different quinolones
and of the internal standard quinine. From then on, a working solution was prepared every
three months, T0 was the initial working solution and T1, T2, … were the working solutions
prepared after three, six, … months, respectively. Every time the new prepared working
solution (T1, T2, …) and the initial working solution (T0), were analysed and statistically
compared (ANOVA test). After one year the results were evaluated. No significant
degradation of the different quinolones was observed. Therefore, the period for which the
working solution can be used was set to one year. The working solution should be stored at -
20 °C during this period. Every year a new working solution will be prepared.
The multi-residue LC-MS2 method for the detection of quinolones in bovine muscle was
validated according to the criteria of Commission Decision 2002/657/EEC [45].
3.3.4.3. Secondary validation of the LC-MS2 method for the detection of quinolones in
chicken, porcine and aquacultured products muscle
The different quinolones are registered in the European Union for use in poultry, porcine and
aquaculture species. Therefore a quantitative confirmation method is required. Exceptions are
marbofloxacin and sarafloxacin which are not registered for use in poultry; sarafloxacin is not
registered for use in porcine species and marbofloxacin is not registered for use in aquaculture
species [16]. Since a complete validation was performed for the matrix bovine muscle, a
secondary validation for the matrices belonging to the same matrix-class is sufficient [46].
The calibration curve of flumequine in shrimp muscle was not linear in the concentration
range 0.5 to 2 MRL. The MRL in aquacultured products was 600 μg kg-1 [16]. This high
concentration can cause space charging in the ion trap. A possible consequence is a non-linear
calibration curve. Therefore, samples of aquacultured products containing flumequine need to
be diluted before quantification.
Specificity and selectivity
The specificity and selectivity was demonstrated in the validation of the LC-MSn method for
the detection of quinolones in bovine muscle.
141 Printed: www.dclsigns.be
Veterinary drugs
Recovery
The recovery was determined using fortified blank muscle tissue. Blank chicken/pork/shrimp
muscle was fortified at 0.5, 1 and 2 times MRL. The recovery at the MRL level was
calculated. All these samples had a recovery within the permitted range. Table 5, 6 and 7
summarise the recovery for the different quinolones in chicken muscle, porcine muscle and
shrimp muscle, respectively.
Table 5 The validation parameters recovery, coefficient of variation (CV) and detection
capability (CCβ) for the different quinolones in chicken muscle
Chicken muscle marbo enro cipro sara dano di oxo flum -1Spiked conc. (µg kg ) 150 100 100 30 200 300 100 400
Recovery (%) 106 104 104 100 109 101 102 106
Overall CV (%) 14.6 11.0 11.3 14.5 15.9 11.5 14.2 9.0 -1 220 136 138 44 317 410 146 517 CCβ (µg kg )
Table 6 The validation parameters recovery, coefficient of variation (CV) and detection
capability (CCβ) for the different quinolones in porcine muscle
Porcine muscle marbo enro cipro sara dano di oxo flum -1Spiked conc. (µg kg ) 150 100 100 30 100 400 100 200
Recovery (%) 107 103 108 102 106 101 103 95
Overall CV (%) 12.0 8.7 14.4 9;3 11.1 5.7 8.1 8.2 -1 206 128 145 39 133 473 126 256 CCβ (µg kg )
Table 7 The validation parameters recovery, coefficient of variation (CV) and detection
capability (CCβ) for the different quinolones in shrimp muscle
Shrimp muscle marbo enro cipro sara dano di oxo flum -1Spiked conc. (µg kg ) 150 100 100 30 100 300 300 600
Recovery (%) 101 95 99 97 93 85 102 -
Overall CV (%) 10.3 14.2 7.5 11.0 14.3 18.0 11.0 - -1 200 148 124 41 148 505 404 - CCβ (µg kg )
142 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
Precision
The coefficient of variation (CV) was determined by the repeated analysis of fortified blank
muscle tissue. Blank chicken/pork/shrimp muscle was fortified at 0.5, 1 and 2 times the MRL.
The overall CV was calculated. Table 5, 6 and 7 summarise the coefficient of variation for
each quinolone in chicken muscle, porcine muscle and shrimp muscle, respectively. The
overall coefficient of variation was lower than the permitted CV.
The intra-laboratory repeatability will be expanded by analysing spiked blank samples at the
MRL concentration with each batch of samples.
Detection capability (CCβ)
The data used to calculate the coefficient of variation, are applied to determine the detection
capability CCβ. The corresponding concentration at the MRL level plus 3.28 times the
standard deviation equals the detection capability (β = 5 %) (CCβ = corresponding
concentration at MRL + 1.64 x standard deviation + 1.64 x standard deviation). In Table 5, 6
and 7 the CC β for each quinlone is summarized.
The detection capability will be expanded by analysing spiked blank samples at the MRL
concentration with each batch of samples.
23.3.4.4. Validation of the LC-MS method for the detection of quinolones in bovine milk
The different quinolones are registered in the European Union for use in animals producing
milk for human consumption. Therefore, a quantitative confirmation method is required.
However, the quinolones oxolinic acid and difloxacin are an exception; they are prohibited for
use in milk producing animals. In addition, sarafloxacin is not registerd for use in bovine
species [16].
For the matrix bovine milk, the validation is not yet completed. Until now the intra-laboratory
and within-laboratory coefficients of variation were not determined. The analyses for the
calculation of the precision were performed on one occasion and needs to be repeated on two
other occasions under repeatability conditions (intra-laboratory CV) and two times under
reproducibility conditions (within-laboratory CV).
143 Printed: www.dclsigns.be
Veterinary drugs
Specificity
The specificity of the method could be demonstrated by LC-MS2 analysis of blank bovine
milk. No interferences were observed after analysis of these blank samples and after analysis
of spiked bovine milk with the eight quinolones.
Selectivity
In Table 2 the precursor ions and product ions of each quinolone are summarised. The
different quinolones can be identified according to the criteria of Commision Decision
2002/657/EEC by their MS2 and MS3 +-full scan spectra of the pseudo-molecular ion [M+H]
[45].
Calibration curves
The calibration curves obtained for spiked bovine milk were linear in the concentration range
0.5 times MRL to 2 times MRL for the eight quinolones. The coefficients of correlation were
higher than 0.94.
Recovery
The recovery was determined using fortified blank bovine milk samples. The recovery should
range from 80 to 110% [45]. Thirty samples of blank bovine milk were fortified at 0.5, 1 and
2 times MRL. The mean recovery of the ten replicates at each level was calculated. All these
samples had a recovery within the permitted range. Table 8 summarises the recovery at each
level for the different quinolones.
Precision
The coefficient of variation (CV) for the repeated analysis of fortified blank bovine milk,
should not exceed the level calculated by the Horwitz equation [45]. For mass fractions lower
than 100 μg kg−1, a CV of 23 % was taken as a guideline. Thirty samples of blank bovine milk
were fortified at 0.5, 1 and 2 times MRL. Table 8 summarises the CV at each level and the
overall CV for each quinolone. The coefficients of variation were lower than the permitted
CV.
144 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
Table 8 The validation parameters recovery and coefficient of variation (CV) at the levels 0.5,
1 and 2 times the MRL for the different quinolones in bovine milk
Marbofloxacin Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%)
40 109 10.1 75 92 8.2
150 101 7.2
Overall CV (%) = 13.2 Enrofloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 50 103 3.9
100 97 7.1 150 100 6.2
Overall CV (%) = 6.0 Ciprofloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 50 100 5.5
100 100 5.3 150 100 7.2
Overall CV (%) = 5.9 Sarafloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 25 108 3.8 50 93 2.6
100 101 3.7
Overall CV (%) = 6.9 Danofloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 15 106 17.4 30 95 3.8 60 101 4.2
Overall CV (%) = 11.6 Difloxacin
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 25 121 2.9 50 83 2.9
100 103 5.5
Overall CV (%) = 15.8 Oxolinic acid
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 25 108 4.6 50 94 7.3
100 101 8.7
Overall CV (%) = 8.8 Flumequine
Spiked conc. (µg kg-1) Recovery (%) Coefficient of variation (%) 25 103 6.3 50 97 5.4
100 99 10.2
Overall CV (%) = 7.9
145 Printed: www.dclsigns.be
Veterinary drugs
Decision limit (CCα)
The data used to calculate the coefficient of variation, are applied to determine the decision
limit CCα. The corresponding concentration at the MRL level plus 1.64 times the standard
deviation equals the decision limit (α = 5 %). In Table 9 CCα for each quinolone is
summarized.
Table 9 The validation parameters decision limit (CCα) and detection capability (CCβ) for the
different quinolones
marbo enro cipro sara dano di oxo flum
CCα (µg kg-1) 95 110 110 56 36 65 58 57
CCβ (µg kg-1) 111 120 119 62 42 78 65 63
Detection capability (CCβ)
The detection capability CCβ was calculated as the decision limit CCα plus 1.64 times the
corresponding standard deviation (β = 5%), supposing that σCCα equals σMRL. In Table 9 the
CC β for each quinolone is summarized.
Applicability and ruggedness
Applicability and ruggedness will be tested by control spiked samples or by participating in
performance studies. With each batch of samples a control spiked sample will be analysed and
once a year unknown control spiked samples are analysed.
146 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
3.3.5. Conclusion 2A LC-ESI-MS multi-residue method was developed to simultaneously analyse eight
quinolones in bovine/chicken/pork muscle, muscle of aquacultured products and bovine milk.
A simple and rapid extraction and clean-up method was used for the different matrices and
ion trap mass spectrometry was used as identification as well as quantification technique. All
quinolones were detectable at and below their MRL concentration.
The multi-residue method for the detection of quinolones in bovine muscle was validated
according the criteria of Commission Decision 2002/657/EEC. Quantification was possible in
the concentration range 0.5 times MRL to 1.5 times MRL (linear calibration curves). A
secondary validation was performed for the matrices chicken muscle, pork muscle and muscle
of aquacultured products. For the matrix bovine milk, the validation is not yet completed.
Until now the intra-laboratory and within-laboratory coefficients of variation were not
determined.
147 Printed: www.dclsigns.be
Veterinary drugs
3.3.6. References
[1] M.G. Papich and J.E. Riviere (2001) Fluoroquinolone antimicrobial drugs, In Veterinary
Pharmacology and Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press,
Ames, 898-912
[2] J.M. Blondeau (2004) Fluoroquinolones: mechanism of action, classification, and
development of resistance, Survey of Ophthalmology 49, supplement 2, S73-S78
[3] V.T. Andriole (2005) The quinolones: past, present and future, CID 41, supplement 2,
S113-S119
[4] B. Delépine and D. Hurtaud-Pessel (2000) Determination of ten quinolones in chicken
muscle by liquid chromatography/APCI/MS-MS, Proceedings of the Euroresidue IV, 8-10
May, Veldhoven, The Netherlands, 350–355.
[5] C.K. Holtzapple, S.A. Buckley, L.H. Stanker (2001) Determination of fluoroquinolones in
serum using an on-line clean-up column coupled to high-performance immunoaffinity-
reversed-phase liquid chromatography, Journal of Chromatography B 754, 1-9
[6] M.Q. Zhang, A. Haemers (1991) Quinolone antimicrobial agents-structure-activity-
relationships, Pharmazie 46, 687-700
[7] D.T. Bearden, L.H. Danziger (2001) Mechanism of action and resistance to quinolones,
Pharmacotherapy 21, 224S-232S
[8] G.A. Jacoby (2005) Mechanisms of resistance to quinolones, CID 41, supplement 2, S120-
S126
[9] D. Barron, E. Jiménez-Lozano, S. Bailac, J. Barbosa (2003) Simultaneous determination
of flumequine and oxolinic acid in chicken tissues by solid phase extraction and capillary
electrophoresis, Analytica Chimica Acta 477, 21-27
[10] B. Toussaint, G. Bordin, A. Janosi, A.R. Rodriguez (2002) Validation of a liquid
chromatography-tandem spectrometry method for the simultaneous quantification of 11
(fluoro)quinolone antibiotics in swine kidney, Journal of Chromatography A 976, 195-206
[11] M. Lipsitch, R.S. Singer, B.R. Levin (2002) Antibiotics in agriculture: when is it time to
close the barn door?, PNAS 99 (9), 5752-5754
[12] A.E. Van den Bogaard, E.E. Stobberingh (1999) Antibiotic usage in animals; impact on
bacterial resistance and public health, Drugs 58 (4), 589-607
[13] J. Engberg, F.M. Aarestrup, D.E. Taylor, P. Gerner-Smidt , I. Nachamkin (2001)
Quinolone and Macrolide Resistance in Campylobacter jejuni and C. coli: resistance
mechanisms and trends in human isolates, Emerging Infectious Diseases 7 (1), 24-33
148 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
[14] World Health Organisation (1998) Use of quinolones in food animals and potential
impact on human health, Report of a WHO Meeting, Geneva, Switzerland, 2-5 June 1998,
WHO/EMC/ZDI/98.10
[15] J. Turnidge (2004) Antibiotic use in animals – prejudices, perceptions and realities,
Journal of Antimicrobial Chemotherapy 53, 26-27
[16] Anonymous (2003) Informal consolidated version of the Annexes I to IV of Council
Regulation n° 2377/90, The European Agency for the Evaluation of Medicinal Products, 22
July 2003
[17] J.A. Hernandez-Arteseros, J. Barbosa, R. Compaño, M.D. Prat (2001) Analysis of
quinolone residues in edible animal products, Journal of Chromatography A 945, 1-24
[18] M.D. Marazuela, M.C. Moreno-Bondi (2004) Multiresidue determination of
fluoroquinolones in milk by column liquid chromatography with fluorescence and ultraviolet
absorbance detection, Journal of Chromatography A 1034, 25-32
[19] S. Bailac, O. Ballesteros, E. Jimenez-Lozano, D. Barron, V. Sanz-Nebot, A. Navalon,
J.L. Vilchez, J. Barbosa (2004) Determination of quinolones in chicken tissues by liquid
chromatography with ultraviolet absorbance detection, Journal of Chromatograph A 1029,
145-151
[20] C. Maraschiello, E. Cusido, M. Abellan, J. Vilageliu (2000) Validation of an analytical
procedure for the determination of the fluoroquinolone ofloxacin in chicken tissues, Journal of
Chromatography B 754, 311-318
[21] B. Roudaut, J.C. Yorke (2002) High-performance liquid chromatographic method with
fluorescence detection for the screening and quantification of oxolinic acid, flumequine and
sarafloxacin in fish, Journal of Chromatography B 780, 481-485
[22] M. Ramos, A. Aranda, E. Garcia, T. Reuvers, H. Hooghuis (2003) Simple and sensitive
determination of five quinolones in food by liquid chromatography with fluorescence
detection, Journal of Chromatography B 789, 373-381
[23] I. Pecorelli, R. Galarina, R. Bibi, Al. Floridi, E. Casciarri, A. Floridi (2003)
Simultaneous determination of 13 quinolones from feeds using accelerated solvent extraction
and liquid chromatography, Analytica Chimica Acta 483, 81-89
[24] H. Pouliquen, M.L. Morvan (2002) Determination of residues of oxolinic acid and
flumequine in freeze-dried salmon muscle and skin by HPLC with fluorescence detection,
Food Additives and Contaminants 19, 223-231
149 Printed: www.dclsigns.be
Veterinary drugs
[25] J.C. Yorke, P. Froc (2000) Quantitation of nine quinolones in chicken tissues by high-
performance liquid chromatography with fluorescence detection, Journal of Chromatography
A 882, 63-77
[26] S.M. Plakas, K.R. El-Said, F.A. Bencsath, S.M. Musser, C.C. Walker (1999)
Determination of flumequine in channel catfish by liquid chromatography with fluorescence
detection, Journal of AOAC International 82, 614-619
[27] O.R. Idowu, J.O. Peggins (2004) Simple, rapid determination of enrofloxacin and
ciprofloxacin in bovine milk and plasma by high-performance liquid chromatography with
fluorescence detection, Journal of Pharmaceutical and Biomedical Analysis 35, 143-153
[28] J.H. Shim, J.Y. Shen, M.R. Kim, C.J. Lee and I.S. Kim (2003) Determination of the
fluoroquinolone enrofloxacin in edible chicken muscle by supercritical fluid extraction and
liquid chromatography with fluorescence detection, Journal of Agricultural and Food
Chemistry 51, 7528–7532
[29] J.E. Roybal, C.C. Walker, A.P. Pfenning, S.B. Turnipseed, J.M. Storey, S.A. Gonzales,
J.A. Hurlbut (2002) Concurrent determination of four fluoroquinolones in catfish, shrimp, and
salmon by liquid chromatography with fluorescence detection, Journal of AOAC International
85, 1293-1301
[30] M.A. Garcia, C. Solans, E. Hernandez, M. Puig, M.A. Bregante (2001) Simultaneous
determination of enrofloxacin and its primary metabolite, ciprofloxacin, in chicken tissues,
Chromatographia 54, 191-194
[31] M.J. Schneider and D.J. Donoghue (2000) Multiresidue determination of
fluoroquinolones in eggs, Journal of AOAC International 83, 1306–1312
[32] J.E. Roybal, A.P. Pfenning, S.B. Turnipseed, C.C. Walker and J.A. Hurlbut (1997)
Determination of four fluoroquinolones in milk by liquid chromatography, Journal of AOAC
International 80, 982–987.
[33] M.D. Marazuela and M.C. Moreno-Bondi (2004) Multiresidue determination of
fluoroquinolones in milk by column liquid chromatography with fluorescence and ultraviolet
absorbance detection, Journal of Chromatography A 1034, 25-32
[34] C. Ho, D. W.M. Sin, H.P.O. Tang, L.P.K. Chung, S.M.P. Siu (2004) Determination and
on-line clean-up of (fluor)quinolones in bovine milk using column-switching liquid
chromatography fluorescence detection, Journal of Chromatography A 1061, 123-131
150 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of quinolones in muscle tissue and bovine milk
[35] A.L. Cinquina, P. Roberti, L. Giannetti, F. Longo, R. Draisci, A. Fagiolo, N.R. Brizioli
(2003) Determination of enrofloxacin and its metabolite ciprofloxacin in goat milk by high-
performance liquid chromatography with diode-array detection, optimization and validation,
Journal of Chromatography A 987, 221-226
[36] P.G. Gigosos, P.R. Revesado, O. Cadahia, C.A. Fente, B.I. Vazquez, C.M. Franco, A.
Cepeda (2000) Determination of quinolones in animal tissues and eggs by high-performance
liquid chromatography with photodiode-array detection, Journal of Chromatography A 871,
31-36
[37] L. Johnston, L. Mackay, M. Croft (2002) Determination of quinolones and
fluoroquinolones in fish tissue and seafood by high-performance liquid chromatography with
electrospray ionisation tandem mass spectrometric detection, Journal of Chromatography A
982, 97-109
[38] B. Delépine, D. Hurtaud-Pessel, P. Sanders (1998) Simultaneous determination of six
quinolones in pig muscle by liquid chromatography-atmospheric pressure chemical ionisation
mass spectrometry, The Analyst 123, 2743-2747
[39] G. van Vyncht, A. Janosi, G. Bordin, B. Toussaint, G. Maghuin-Rogister, E. De Pauw,
A.R. Rodriguez (2002) Multiresidue determination of (fluoro)quinolone antibiotics in swine
kidney using liquid chromatography-tandem mass spectrometry, Journal of Chromatography
A 952, 121-129
[40] S.B. Turnipseed, J.E. Roybal, A.P. Pfenning, P.J. Kijak (2003) Use of ion-trap liquid
chromatography-mass spectrometry to screen and confirm drug residues in aquacultured
products, Analytica Chimica Acta 483, 373-386
[41] M.J. Schneider, D.J. Donoghue (2002) Multiresidue analysis of fluoroquinolone
antibiotics in chicken tissue using liquid chromatography-fluorescence-multiple mass
spectrometry, Journal of Chromatography B 780, 83-92.
[42] B. Toussaint, M. Chedin, G. Bordin, A.R. Rodriguez (2005) Determination of
(fluoro)quinolone antibiotic residues in pig kidney using liquid chromatography-tandem mass
spectrometry I. Laboratory-validated method, Journal of Chromatography A 1088, 32-39
[43] B. Toussaint, M. Chedin, U. Vincent, G. Bordin, A.R. Rodriguez (2005) Determination
of (fluoro)quinolone antibiotic residues in pig kidney using liquid chromatography-tandem
mass spectrometry Part II: Intercomparison exercise, Journal of Chromatography A 1088, 40-
48
[44] www.emea.eu.int
151 Printed: www.dclsigns.be
Veterinary drugs
[45] Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive
96/23/EC concerning the performance of analytical methods and the interpretation of results
(2002) Official Journal of the European Communities, no. L 221
[46] BELAC 2-105 Rev 0-2004, Criteria waaraan de geacrediteerde laboratoria moeten
beantwoorden die een flexibele scope aanvragen voor analyses ter uitvoering van de richtlijn
96/23/EG overeenkomstig beschikking 2002/657/EG.
152 Printed: www.dclsigns.be
Chapter 3.4 nMuti-residue LC-MS method for the detection of non-steroidal anti-
inflammatory drugs in bovine muscle
Adapted from:
N. Van Hoof, K. De Wasch, S. Poelmans, H. Noppe and H. De Brabander
Multi-residue liquid chromatography-tandem mass spectrometry method for the detection of non-
steroidal anti-inflammatory drugs in bovine muscle: optimisation of ion trap parameters
Rapid Communications in Mass Spectrometry (2004) 18, 2823-2829
And extended with validation data
3.4.1. Introduction
As long as 2500 years ago, Hippocrates recommended willow bark to relieve the pain of
childbirth and to reduce fever. These medicinal extracts of barks contained salicylates. The
origin of the group of salicylates lies in the naturally occurring compound salicin that can be
found in a number of different plants. The presence of salicin has been documented to occur
in willow and poplar species (Salicaceae), wintergreen, birch and a variety of rose. Salicylic
acid is also found naturally in many herbs, vegetables and fruits [1]. Acetylsalicylic acid, the
active ingredient of aspirin, was synthesised by Bayer in 1899. Since that time a number of
new anti-inflammatory compounds have been developed. Non-steroidal anti-inflammatory
drugs (NSAIDs) are used routinely in veterinary practice since the early 1970s. They are often
the initial therapy for inflammation disorders of several animal species. They are commonly
prescribed for musculoskeletal pain, coliform mastitis, pulmonary diseases and enteritis. The
use of NSAIDs in veterinary medicine has evolved in a way similar to that in human
medicine. In recent years the treatment of pain in animals has become an important issue,
even in food-producing animals.
153 Printed: www.dclsigns.be
Veterinary drugs
3.4.1.1. Mechanism of action
NSAIDs act by inhibiting the body’s ability to synthesize prostaglandins. Prostaglandins are a
family of hormone-like chemicals some of which are made in response to cell injury. The
generally accepted mechanism of action of NSAIDs is the inhibition of cyclooxygenase
(COX), an enzyme that converts arachidonic acid into different prostaglandins (Fig. 1).
Arachidonic acid is released into the cell from damaged cell membranes. Inside the cell, it
serves as a substrate for the COX enzyme which generates prostaglandins. Cyclooxygenase 1
(COX 1) synthesises constitutive prostaglandins, which are constantly present and impart a
variety of normal physiological effects: stomach mucus production, kidney water retention
and platelet formation. Cyclooxygenase 2 (COX 2) catalyses the formation of inducible
prostaglandins, which are important in the process of inflammation. COX 1 is stimulated
continuously and COX 2 is stimulated only as part of an immune response. NSAIDs work by
temporarily blocking the attachment site of arachidonic acid on the COX enzyme, preventing
the enzyme from converting arachidonic acid to prostaglandin. The exception is aspirin which
irreversibly acetylates cyclooxygenase.
NSAIDs block both COX enzymes. The benefit of an NSAID comes from its COX 2 blocking
action. Drugs with the greatest specificity to COX 1 are the drugs with the greatest side
effects. Therefore a COX 1/COX 2 ratio of less than 1 is desirable. Salicylates, flunixin,
phenylbutazone and tolfenamic acid have a high ratio (preferential COX 1 inhibitors), while
meloxicam and carprofen are COX 2 selective drugs with a ratio equal to or less than 1 [2-3].
Membrane phospholipids
Corticosteroids Phospholipase A2
Arachidonic acid
NSAIDs COX 1 COX 2 5-lipoxygenase
Constitutive
prostaglandins
Inducible
prostaglandins
Leucotrienes
Fig. 1 Site of action of NSAIDs on the arachidonic acid metabolism pathway
154 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
3.4.1.2. Side effects
The major toxicities affect the gastro-intestinal, hematopoietic and renal systems. Other
effects associated with use of NSAIDs include hepatotoxicity, aseptic meningitis, diarrhea,
and central nervous system depression. Gastrointestinal erosions and ulcerations are the most
common and serious side effects of NSAIDs. There are two components to NSAID-induced
ulceration. First there is a local acid effect of the dissolved drug and secondly there is a
restriction by NSAIDs of the self-protection mechanism induced by COX 1 prostaglandins. It
is primarily through this mechanism, not a local acid effect, that NSAIDs cause stomach
ulcers. All NSAIDs are able to impair platelet activity. Renal toxicities include renal
vasoconstriction and renal insufficiency [2-3]. In addition, combination of several NSAIDs
can be fatal.
3.4.1.3. Classification
NSAIDs may be structurally classified as carboxylic acids or enolic acids. The carboxylic
acid derivatives include salicylates (acetylsalicylic acid), acetic acids, propionic acids
(ketoprofen, carprofen), anthranilic acids, aminonicotinic acids (flunixin) and fenamates
(tolfenamic acid), while the enolic acids include pyrazolones (phenylbutazone) and oxicams
(meloxicam) (Fig. 2) [2-4].
155 Printed: www.dclsigns.be
Veterinary drugs
COOH
O CH3
O
O
COOH
CH3
acetylsalicylic acid ketoprofen
HN
CH3
Cl
COOH
NHN
COOH
CH3
CF3
tolfenamic acid flunixin
HN
Cl
CH3
COOH
SN
NH
S
NOH O
O O
CH3
CH3
carprofen meloxicam
N
NO
OH3C
phenylbutazone
Fig. 2 Chemical structure of acetylsalicylic acid, ketoprofen, flunixin, tolfenamic acid,
carprofen, phenylbutazone and meloxicam
156 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
3.4.1.4. Legislation
In food-producing animals the use of drugs is restricted to registered products for which a
Maximum Residue Limit (MRL) is established. NSAIDs which have an MRL (Annex I) are
carprofen (bovine, equine), vedaprofen (equine), flunixin (bovine, porcine, equine),
tolfenamic acid (bovine, porcine) and meloxicam (bovine, porcine, equine) (Table 1). Two
NSAIDs do not have a MRL (Annex II), namely, ketoprofen and salicylates [5]. However,
drugs that are not registered in Belgium cannot be legally administered to food-producing
animals. Therefore acetylsalicylic acid can not be used in Belgium since it is not registered for
use in bovine species. For acetylsalicylic acid a Minimum Required Performance Limit
(MRPL) of 40 µg kg-1 for bovine muscle is used in Belgium. Ketoprofen is licensed for
bovine species in EU, but a withdrawal period of 4 days needs to be respected (only trace
levels of residues are detected at the injection sites 96 hours after treatment). Phenylbutazone
is unauthorised for use in bovine species in the European Union.
Table 1 Maximum Residue Limits set for NSAIDs in bovine muscle
Analyte Marker residue MRL (µg kg-1) Target tissue
Carprofen carprofen 500 muscle
Flunixin flunixin 20 muscle
Tolfenamic acid tolfenamic acid 50 muscle
Meloxicam meloxicam 20 muscle
157 Printed: www.dclsigns.be
Veterinary drugs
3.4.2. Method setup
Since 2002 injection sites from slaughtered animals were analysed for the presence of
veterinary medicinal products (chapter 3.1). In 2002, 29.0 % of the injection sites were non-
compliant. In 5.8 % of them, NSAIDs (flunixin, tolfenamic acid, meloxicam and
phenylbutazone) were detected. In 2003, 6.9 % of the 31.5 % non-compliant injection sites
contained NSAIDs. In 2004, there were 6.0 % (of the 25.4 %) non-compliant injection sites
containing NSAIDs. And this year, 2005, 27.4 % of the injection sites were non-compliant of
which 7.3 % were NSAIDs. In conclusion, NSAIDs were each year the most detected
veterinary drugs in injection sites, next to beta-lactam antibiotics and tetracyclines (Table 4,
chapter 3.2.6).
Literature data for analysis of NSAIDs indicate extraction and clean-up procedures for the
determination of one or two compounds [4, 8-20], with mass spectrometry as the main
detection technique. No literature data were found on multi-residue methods in bovine muscle
for structurally different compounds. In this study a LC-MSn multi-residue method was
developed for bovine muscle to identify salicylic acid, phenylbutazone, flunixin, tolfenamic
acid, meloxicam and ketoprofen.
3.4.3. Experimental
3.4.3.1. Reagents and chemicals
The NSAID standards, phenylbutazone, tolfenamic acid and ketoprofen were obtained from
Sigma-Aldrich (St Louis, MO, USA), while salicylic acid was from Acros (Geel, Belgium),
meloxicam from ICN Biomedicals (Irvine, CA, USA) and flunixin was a generous gift from
the Department of Pharmacology, Pharmacy and Toxicology (Merelbeke, Belgium). The
internal standard flunixin-d3 was obtained from Witega Laboratorien (Berlin, Germany). All
chemicals used were of analytical grade from Merck (Darmstadt, Germany) and Acros (Geel,
Belgium). -1Stock standard solutions of 1000 ng μl were prepared in ethanol. For the preparation of
working solutions 0.4 % acetic acid in MeOH/H2O (60:40) was used. All standard and
working solutions were stored at -20 °C.
158 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
3.4.3.2. Instrumentation
The HPLC apparatus comprised a 1100 series quaternary pump and an autosampler (Hewlett
Packard, Palo Alto, CA, USA). Chromatographic separation was achieved using an Alltima
HP C18 column (5 µm, 150 x 2.1 mm, Alltech, Deerfield, Illinois, USA). The mobile phase
consisted of a mixture of methanol (A) and water with 0.1 % acetic acid (B). A linear gradient
was run (60 % A for 9 min, increasing to 100 % in the next 4 min) at a flow rate of 0.3 ml
min-1.
LC-MSn detection used a LCQ Deca ion trap (ThermoFinnigan, San José, CA, USA) with an
electrospray ionisation (ESI) interface, in both negative and positive ion modes. Each analyte
was evaluated based on the product ions present in the MS2 and MS3 spectra (Table 2).
Table 2 The precursor and product ions (m/z) used for the evaluation of different NSAIDs and
the internal standard flunixin-d3 in bovine muscle and bovine milk
Analyte Detection
mode
Precursor ion MS2 first
transition
product ions
MS3 second
transition
product ions
Salicylic acid negative 137 93 65
Phenylbutazone negative 307 279 131
Flunixin negative 295 251 231
Tolfenamic acid negative 260 216 180
Meloxicam negative 350 146 210 286
Flunixin-d3 (IS) negative 435 355 375
Ketoprofen positive 255 209 105 131 194
Flunixin-d3 (IS) positive 377 321 339 357
3.4.3.3. Extraction and clean-up
To a 2 g aliquot of minced muscle tissue, 50 µg kg-1 flunixin-d3 was added as internal
standard. The NSAIDs were extracted from the muscle tissue using 10 ml acetonitrile. After
mixing and centrifugation (5 min, 5500 rpm) the supernatant was evaporated to dryness at 60
°C under a stream of nitrogen. The clean-up was performed using an Oasis HLB column (60
mg, 3 cc) (Waters, Milford, USA). The columns were conditioned with 1 ml methanol and 1
ml ultrapure water. The residue was reconstituted in 100 µl methanol and 900 µl ultrapure
water. After application of this extract, the cartridge was rinsed with 1 ml MeOH/H O (5:95) 2
159 Printed: www.dclsigns.be
Veterinary drugs
and vacuum dried. The NSAIDs were eluted from the column with 1 ml methanol and 1 ml
10% acetic acid in hexane. The eluate was evaporated to dryness at 60 °C under a stream of
nitrogen. The residue was reconstituted in 50 µl methanol and subsequently 100 µl 0.4 %
acetic acid in MeOH/H O (60:40), before injecting 30 µl on the HPLC column. 2
3.4.4. Results
3.4.4.1. Hydrolysis of acetylsalicylic acid
Acetylsalicylic acid is rapidly hydrolysed to salicylic acid by aryl esterases, a group of
enzymes that are widely distributed in blood, plasma, liver, kidney and certain tissues (Fig. 3).
Therefore the detection of acetylsalicylic acid was performed by detection of the marker-
residue salicylic acid [21].
COOHO CH3
O
COOHOHaryl esterases
Fig. 3 Enzymatic hydrolysis of acetylsalicylic acid to salicylic acid
3.4.4.2. LC-MS2 method
The different NSAIDs were detected using a LC-MS2 method in both negative and positive
ion mode. The NSAIDs salicylic acid, phenylbutazone, flunixin, tolfenamic acid and
meloxicam were detected in negative ion mode, while ketoprofen was detected in positive ion
mode. A combination of the two polarity modes was not ideal since polarity change
(alternating detection in negative and positive ion modes) led to a loss in sensitivity and a loss
of information. Therefore two acquisitions were necessary to detect all NSAIDs although the
extraction and clean-up were identical. The instrument parameters of the MS2 fragmentation,
collision energy and activation q are summarised in Table 3. The isolation width was set to 2
Da for each NSAID. Salicylic acid had an activation q different from the default value 0.25 as
discussed in the next paragraph.
160 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
Table 3 Instrument parameters (collision energy and activation q) of the LC-MS2 method for
the detection of NSAIDs
Analyte Collision energy (%) Activation q
Salicylic acid 47 0.35
Phenylbutazone 49 0.25
Flunixin 40 0.25
Tolfenamic acid 46 0.25
Meloxicam 35 0.25
Ketoprofen 25 0.25
The standards of salicylic acid, phenylbutazone, flunixin, tolfenamic acid, meloxicam,
ketoprofen and the internal standard flunixin-d3, were spiked into blank bovine muscle at
concentrations listed in Table 4. For flunixin, tolfenamic acid and meloxicam the spike
concentration was the MRL, for acetylsalicylic acid (spiked as salicylic acid) the MRPL; and
for phenylbutazone and ketoprofen an internal action limit.
Table 4 Concentrations at which the NSAIDs were spiked into blank bovine muscle
Analyte Spiked concentration (µg kg-1)
Salicylic acid 40 (MRPL acetylsalicylic acid)
Phenylbutazone 100 (internal AL)
Flunixin 20 (MRL)
Tolfenamic acid 50 (MRL)
Meloxicam 20 (MRL)
Ketoprofen 20 (internal AL)
Figs. 4 and 5 show the extracted ion chromatograms and the MS2-mass spectra for the
different NSAIDs.
161 Printed: www.dclsigns.be
Veterinary drugs
162
D:\Doctoraat\...\NSAIDs\050901s07 1-9-2005 16:48:00 Q296Balltima NSAID
RT: 0,00 - 20,00 SM: 7B
0 5 10 15 20Time (min)
0
20
40
60
80
1000
20
40
60
80
1000
20
40
60
80
1000
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
0
20
40
60
80
1000
20
40
60
80
1002,54
3,97 5,521,564,62
3,602,069,70
7,918,48
9,197,4010,81 12,15
9,62
8,717,4117,14
19,4313,78 15,0612,62
NL: 2,27E6m/z= 92,5-93,5 F: - c ESI Full ms2 137,00@35,00 [ 50,00-140,00] MS 050901s07
NL: 7,53E6m/z= 145,5-146,5+209,5-210,5+285,5-286,5 F: - c ESI Full ms2 350,00@35,00 [ 100,00-355,00] MS 050901s07
NL: 9,67E6m/z= 230,5-231,5+250,5-251,5 F: - c ESI Full ms2 295,00@40,00 [ 80,00-300,00] MS 050901s07
NL: 5,78E5m/z= 130,5-131,5+278,5-279,5 F: - c ESI Full ms2 307,00@49,00 [ 80,00-310,00] MS 050901s07
NL: 2,20E6m/z= 215,5-216,5 F: - c ESI Full ms2 260,00@46,00 [ 100,00-270,00] MS 050901s07
NL: 5,17E7m/z= 233,5-234,5+253,5-254,5 F: - c ESI Full ms2 298,00@35,00 [ 80,00-300,00] MS 050901s07
50 60 70 80 90 100 110 120 130 140m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
93,1
92,3 137,0109,393,9 107,1 119,679,964,8 122,0 135,1
100 120 140 160 180 200 220 240 260 280 300 320 340m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
285,9
209,9
146,1202,0192,1 331,1217,1 250,3 285,2168,0 314,1235,1144,1 349,1267,1 302,9112,9
80 100 120 140 160 180 200 220 240 260 280 300m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
251,1
231,1 251,9211,1 250,4 277,1181,2 192,0113,3 172,1127,699,1 284,7151,1 219,5
80 100 120 140 160 180 200 220 240 260 280 300m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
279,0
131,0188,0 214,1 289,1168,3 235,1 249,9 278,391,9 132,0 160,1 291,8128,8 204,1118,9 217,3171,1
80 100 120 140 160 180 200 220 240 260 280 300m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
254,2
234,2252,3 254,9233,3213,2 280,2183,1 193,2157,1139,9 171,1
100 120 140 160 180 200 220 240 260m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
216,0
214,3200,2160,2 243,2112,0 180,3123,2 187,0 217,6149,1 227,9
Fig. 4 Extracted ion chromatograms and MS2 spectra of [M-H]- ions of salicylic acid (SA),
meloxicam (MLC), flunixin (FLX), phenylbutazone (PB), flunixin-d3 (FLX-d3) (I.S.) and
tolfenamic acid (TOLF) in bovine muscle
D:\Doctoraat\...\NSAIDs\050112s36 13-1-2005 7:09:33 Q10Calltima NSAID
RT: 0,00 - 13,50 SM: 7B
0 2 4 6 8 10 12Time (min)
0
20
40
60
80
1000
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
4,45
2,04 3,02 5,28 7,90 8,777,24 11,239,540,15 13,018,71
2,31 10,484,363,65 11,565,59 8,016,570,53
NL: 2,48E6m/z= 208,5-209,5 F: + c ESI Full ms2 255,00@25,00 [ 100,00-260,00] MS 050112s36
NL: 2,76E7m/z= 281,5-282,5 F: + c ESI Full ms2 300,00@35,00 [ 80,00-310,00] MS 050112s36
050112s36 # 344-377 4,26-4,56 12 : 150 1,10-3,26, 5,10-8,63 NL: 2,09E6F:
RT: AV: SB+ c ESI Full ms2 255,00@25,00 [ 100,00-260,00]
100 120 140 160 180 200 220 240 260m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
209,0
237,1255,1223,0105,1 195,0177,1 208,1 240,8212,8163,0 187,2172,1129,1 158,9137,0 143,1112,7 125,0
050112s36 # 665-710 : 8,51-9,08 15 : 177 2,75-6,61, 10,11-13,43 NL: 2,18E7F:
RT AV: SB+ c ESI Full ms2 300,00@35,00 [ 80,00-310,00]
80 100 120 140 160 180 200 220 240 260 280 300m/z
0
20
40
60
80
100
Rel
ativ
e Ab
unda
nce
282,4
280,1 283,4259,1 300,1273,1241,0135,0 184,1155,0 216,9207,196,1 232,8160,2 197,1129,1117,2
Fig. 5 Extracted ion chromatograms and MS2 spectra of [M+H]+ ions of ketoprofen (KET)
and flunixin-d3 (FLX-d3) (I.S.) in bovine muscle
SA
SA
MLC
FLX
MLC
FLX-d3
FLX
PB
FLX-d3
TOLF
PB
TOLF
KET
FLX-d3
KET
FLX-d3
Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
Carprofen, which has an MRL specified for bovine muscle, is not yet incorporated in the
present multi-residue method. There was no demand from the Federal Agency for the Safety
of the Food Chain for the detection of carprofen, and this NSAID has not yet been detected
during the monitoring of injection sites from 2002 till 2005 (chapter 3.1) [6-7].
3.4.4.3. Mass spectrometric detection of salicylic acid
The detection of salicylic acid using MS-full scan was rather poor, and its fragmentation (loss
of the carboxylic acid group as CO2) was not reproducible. Therefore attempts were made to
derivatise salicylic acid, but this compromised the multi-residue method. So another solution
was proposed based on MS2.
Once isolation of the selected precursor ion in an ion trap is completed, the rf amplitude is
reduced to obtain a certain qz-value. Not only the precursor ion but also the product ions to be
monitored need to be within the stability region at this qz-value. By default the qz-value of a
Thermo ion trap mass spectrometer is set to 0.25, corresponding to a certain low mass cutoff
(LMCO) value. By increasing the qz-value, the LMCO value will increase, so possible
product ions with m/z ratios below this LMCO value will not be stored.
In negative ion mode salicylic acid produced a [M-H]- ion with m/z 137. During isolation the
rf amplitude was ramped in order to eject all ions with m/z < 137. Subsequently a broadband
waveform was applied to eject ions with m/z > 137. At that point the [M-H]- ion of salicylic
acid was isolated in the ion trap and the rf amplitude was reduced again to position this ion
within the stability region along the qz-axis. At the default qz-value of 0.25, the precursor ion
with m/z 137 was not stable and not every scan contained the product ion with m/z 93,
although both the precursor ion and the product ion were present within the stability region.
Therefore the qz-value for m/z 137 was increased. At a qz-value of 0.35 the [M-H]- ion of
salicylic acid was stable, and a good and reproducible detection of the product ion with m/z
93 was obtained. This qz-value corresponds to an LMCO value of m/z 50, low enough to
detect the product ion [22-24].
3.4.4.4. Mass spectrometric detection of phenylbutazone
Detection of phenylbutazone required adaption of another parameter, the maximum ion
injection time, the time for which ions are allowed to accumulate in the mass analyser when
automatic gain control is on. Automatic gain control serves to maintain the optimum quantity
of ions for each scan to avoid space charge effects. By default the maximum ion injection
163 Printed: www.dclsigns.be
Veterinary drugs
time is set by the manufacturer to 200 ms. Too low a value can result in loss of sensitivity
because the mass analyser traps fewer than the optimum number of ions, but too high a value
can give insufficient data points across the chromatographic peak. The latter is only the case
when the number of microscans is already set to its lowest value. Each microscan is one
complete mass analysis cycle, i.e., ion injection and storage followed by scan out of ions,
followed by ion detection; a number of microscans is summed to produce one scan.The ion
injection time and the number of microscans affect the scan time.
Since the method discussed in this paper is a LC-MS2 multi-residue method, MS parameters
had to be set within several LC time segments in order to be able to detect all the NSAIDs. In
this case three time segments were used in the instrumental method, each time segment
corresponding to analysis of different sets of compounds (Table 5). Time segment 2 analysed
phenylbutazone, flunixin and the internal standard flunixin-d3. The partitioning of the total
scan time within LC time segment 2 among three compounds meant that there was a decrease
in the number of scans available to detect each analyte. Therefore the number of microscans
was set to 1 to increase the number of scans recorded across the chromatographic peak of
each analyte. However, at the default maximum ion injection time of 200 ms, there were still
not enough scans obtained across the chromatographic peak for phenylbutazone to obtain a
well-defined and intense chromatographic peak. Therefore the maximum ion injection time
was lowered to 50 ms.
Table 5 NSAIDs analysed in each LC time segment
Time segment Analyte
Segment 1 Salicylic acid
Meloxicam
Segment 2 Phenylbutazone
Flunixin-d3 (IS)
Flunixin
Segment 3 Tolfenamic acid
3.4.4.5. Confirmation of salicylic acid, tolfenamic acid and ketoprofen
Salicylic acid, tolfenamic acid and ketoprofen had one product ion in the MS2-full scan
spectra of their [M-H]- and [M+H]+ ions (Figs. 4 and 5), so 2.5 identification points (IP) were
earned (1 precursor ion and 1 product ion). To create more selectivity and to achieve enough
164 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
identification points [26], full scan MS3 spectra of the product ions were investigated. Fig. 6
shows the MS3-mass spectrum for tolfenamic acid spiked into blank bovine muscle at a
concentration of 50 µg kg-1, containing a second transition product ion at m/z 180, so 4 IPs
were earned (1 precursor ion, 1 product ion and 1 second transition product ion). Also the
MS3-mass spectrum for ketoprofen spiked at a concentration of 20 µg kg-1 is shown in Fig. 6.
This mass spectrum contains three second transition product ions, m/z 105, 131 and 194. Fig.
6 also shows the MS3 --mass spectrum of ([M-H] → m/z 93) for salicylic acid spiked into
blank bovine muscle at a concentration of 40 µg kg-1. At the default qz-value of 0.25 the
product ion with m/z 93 was not stable and no signal was received in MS3-full scan. The qz-
value for m/z 93 was increased to 0.35. Again the product ion with m/z 93 was not stable, but
a second transition product ion with m/z 65 was revealed in MS3-full scan. However, not
every scan contained this second transition product ion. At a qz-value of 0.45 the product ion
with m/z 93 was stable and a good and stable detection of the second transition product ion
with m/z 65 was obtained.
So, tolfenamic acid, ketoprofen and salicylic acid can be confirmed according to the criteria
of Commision Decision 2002/657/EEC by their MS3 full scan spectra of the [M-H]- and
[M+H]+ ions via their first transition product ion [25].
165 Printed: www.dclsigns.be
Veterinary drugs
040624s31 #1656-1661 RT: 18,00-18,06 AV: 3 NL: 6,90E3F: - c ESI Full ms3 260,00@46,00 216,00@46,00 [ 100,00-270,00]
100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270m/z
0
20
40
60
80
100180,2
216,1
Rel
ativ
e A
bund
ance
tolfenamic acid
050112s36 #350-380 RT: 4,34-4,58 AV: 10 SB: 174 1,10-3,26, 5,04-9,56 NL: 2,69E5F: + c ESI Full ms3 255,00@25,00 209,00@45,00 [ 100,00-260,00]
100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260m/z
0
20
40
60
80
100194,2105,1
131,1181,0 195,0
Rel
ativ
e A
bund
ance
191,3 211,4179,3165,9152,8122,0 138,0105,8
ketoprofen
050901s07 #198-259 RT: 2,24-2,90 AV: 13 SB: 98 0,31-2,23, 3,57-10,63 NL: 1,11E5F: - c ESI Full ms3 137,00@35,00 93,00@55,00 [ 50,00-150,00]
50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125 130 135 140 145 150m/z
0
20
40
60
80
10065,0
74,866,9
Rel
ativ
e A
bund
ance
salicylic acid
Fig. 6 MS3 mass spectra of tolfenamic acid, ketoprofen and salicylic acid
3.4.4.6. Validation
The NSAIDs flunixin, tolfenamic acid and meloxicam are registered for use in the European
Union, but they have a MRL (Table 1) [5]. Therefore, a quantitative confirmation method is
required. Salicylic acid and phenylbutazone require a qualitative confirmation method since
acetylsalicylic acid is not registered for use in bovine in Belgium and phenylbutazone is
unauthorised in the European Union.
Before the validation characteristics are determined, the efficiency of the extraction with
acetonitrile (see extraction and clean-up) needs to be evaluated. Methods described in the
literature for the determination of NSAIDs in edible bovine tissues involve an initial acid
hydrolysis prior to extraction with acetonitrile or ethyl acetate [8,16].
166 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
Evaluation of the extraction
Two samples that were previously analysed for the presence of NSAIDs contained flunixin in
a concentration higher than the MRL (20 µg kg-1). These two samples were used to evaluate
the extraction of NSAIDs in bovine muscle. Each sample was analysed 5 times with and 5
times without hydrolysis. In the case of acid hydrolysis, prior to extraction with acetonitrile,
there was an addition of 2.5 ml 0.2 M acetate buffer pH 4.6 and 50 µl glucuronidase to
minced bovine muscle. Hydrolysis was performed at 50 °C for 2 hours. Afterwards, 7.5 ml
acetonitrile was added for the extraction of flunixin from the muscle tissue. After mixing and
centrifugation, the supernatant was evaporated to 2.5 ml at 60°C. The clean-up was analogues
to the one described for bovine muscle (see extraction and clean-up).
The area ratio’s (area flunixin / area flunixin-d3) of the replicates are shown in Table 6.
Table 6 The area ratio’s of two samples containing the NSAID flunixin after extraction with
and without hydrolysis
Sample 1 Sample 2
Without hydrolysis With hydrolysis Without hydrolysis With hydrolysis
2.53 2.11 1.27 1.83
1.99 1.88 1.58 2.48
2.31 2.04 1.65 1.38
1.93 1.99 1.65 1.54
2.35 1.77 1.55 1.37
The Dunnett test (ANOVA) authorized the comparison of extraction with acid hydrolysis and
extraction without hydrolysis of these two samples containing the NSAID flunixin. The
signification degree p for the data of sample 1 was 0.074043 ( > 0.05) and the signification
degree for the data of sample 2 was 0.434891 ( > 0.05); so, there is no significant difference
between the two extraction methods concerning the NSAID flunixin. Therefore, the method
consisting of an extraction with acetonitrile followed by solid phase extraction will be used
for the quantitative analysis of flunixin, tolfenamic acid and meloxicam.
167 Printed: www.dclsigns.be
Veterinary drugs
Validation characteristics
The validation parameters of a qualitative confirmation method are: specificity/selectivity,
applicability/ruggedness/stability, detection capability (CCβ) and decision limit (CCα). The
validation parameters of a quantitative confirmation method are equal to the ones of a
qualitative validation plus recovery and precision (Table 4, Chapter 1.2.3) [25].
Specificity 2The specificity of the method could be demonstrated by LC–MS analysis of blank bovine
muscle. No interferences were observed after analysis of these blank samples and after
analysis of spiked matrices with the different NSAIDs.
Selectivity
Flunixin, tolfenamic acid and meloxicam are veterinary drugs with a MRL, so the minimum
number of identification points is set to three. Salicylic acid and phenylbutazone are
unauthorized compounds which require four identification points. In Table 2 the precursor ion
and the product ions of each NSAID are summarized.
The different NSAIDs can be identified according to the criteria of Commision Decision
2002/657/EEC by their MS2 3- and MS -full scan spectra of the pseudo-molecular ion [M-H]-.
Recovery
Since no Certified Reference Material is available, the recovery is determined by experiments
using fortified blank bovine muscle tissue. For samples spiked at a concentration above 10 μg
kg−1, the recovery should range from 80 to 110 %. Eighteen samples of blank bovine muscle
were fortified at 0.5, 1 and 1.5 times MRL. The mean recovery from the six results at each
level was calculated and this for the NSAIDs flunixin, tolfenamic acid and meloxicam. All
these samples had a recovery within the permitted range. Table 7, 8 and 9 summarise the
recovery at each level for flunixin, tolfenamic acid and meloxicam, respectively.
168 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
Table 7 The validation parameters recovery and coefficient of variation (CV) at the levels 0.5,
1 and 1.5 times MRL for the NSAID flunixin
Flunixin Spiked concentration Recovery (%) Coefficient of variation (%)
10 101 3.6 20 99 9.7 30 100 2.4
Overall CV (%) = 5.8
Table 8 The validation parameters recovery and coefficient of variation (CV) at the levels 0.5,
1 and 1.5 times MRL for the NSAID tolfenamic acid
Tolfenamic acid Spiked concentration Recovery (%) Coefficient of variation (%)
25 100 7.1 50 92 9.0 75 97 17.0
Overall CV (%) = 11.7
Table 9 The validation parameters recovery and coefficient of variation (CV) at the levels 0.5,
1 and 1.5 times MRL for the NSAID meloxicam
Meloxicam Spiked concentration Recovery (%) Coefficient of variation (%)
10 86 18.3 20 112 8.6 30 95 11.2
Overall CV (%) = 16.3
Precision
The coefficient of variation (CV) for the repeated analysis of fortified blank bovine muscle
tissue, should not exceed the level calculated by the Horwitz equation. For mass fractions
lower than 100 μg kg−1 the application of the Horwitz equation gives unacceptable high
values. Therefore the CV for concentrations lower than 100 μg kg−1 should be as low as
possible. In that case 23 % is taken as a guideline (CV at 100 μg kg−1 = 23 %). Eighteen
samples of blank bovine muscle were fortified at 0.5, 1 and 1.5 times MRL. Table 7, 8 and 9
summarise the CV at each level and the overall CV for flunixin, tolfenamic acid and
meloxicam, respectively. The coefficient of variation was lower than the permitted CV.
169 Printed: www.dclsigns.be
Veterinary drugs
To calculate the intra-laboratory coefficient of variation, these analyses were repeated on two
other occasions by the same analyst, under repeatability conditions. The overall mean
concentration, standard deviation and coefficient of variation of these fifty four fortified blank
samples, were calculated. The intra-laboratory coefficient of variation would typically be
between one half and two third of the CV calculated by the Horwitz equation. Since the mass
fractions are lower than 100 μg kg−1 the intra-laboratory coefficient of variation should be as
low as possible (taken 23 % as a guideline). The intra-laboratory CVs for flunixin, tolfenamic
acid and meloxicam were lower than 23 %.
To calculate the within-laboratory coefficient of variation, these analyses were repeated on
one other occasion under reproducibility conditions (by a different analyst). The coefficient of
variation of this set of samples was calculated. The within-laboratory coefficient of variation
should not exceed the overall coefficient of variation. The within-laboratory CVs of flunixin
and meloxicam were lower than the overall CVs and the within-laboratory CV of tolfenamic
acid was just slightly higher than the overall CV.
In Table 10 the different coefficients of variation are summarized.
Table 10 The validation parameters coefficient of variation (CV), decision limit (CCα) and
detection capability (CCβ) for the NSAIDs flunixin, tolfenamic acid and meloxicam
flunixin tolfenamic acid meloxicam
Intra-laboratory CV (%) 4.74 10.69 14.54
Within-laboratory CV (%) 4.47 12.80 3.95
CCα (µg kg-1) 21.56 57.58 24.40
CCβ (µg kg-1) 23.11 66.16 29.10
Decision limit (CCα)
The decision limit is the limit at and above which it can be concluded with an error
probability of α that a sample is non-compliant. For the NSAIDs flunixin, tolfenamic acid and
meloxicam, the data used to calculate the intra-laboratory coefficient of variation, are applied
to determine the decision limit CCα. The corresponding mean concentration at the MRL level
plus 1.64 times the standard deviation equals the decision limit (α = 5 %). In Table 10 CCα
for flunixin, tolfenamic acid and meloxicam is summarized.
170 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
For the NSAIDs salicylic acid and phenylbutazone, which require only a qualitative
validation, the decision limit CCα was derived from the detection capability CCβ and a
maximum coefficient of variation of 23 %.
Detection capability (CCβ)
Detection capability is the smallest content of the analyte that may be detected, identified
and/or quantified in a sample with an error probability of β. In the case of substances with a
maximum residue limit (MRL), the detection capability is the concentration at which the
method is able to detect MRL concentrations with a statistical certainty of 1-β. For the
NSAIDs flunixin, tolfenamic acid and meloxicam, CCβ was calculated as the decision limit
CCα plus 1.64 times the corresponding standard deviation (β = 5%), supposing that σCCα
equals σMRL. In Table 10 CCβ for flunixin, tolfenamic acid and meloxicam is summarized.
In the case of unauthorized substances, the detection capability is the lowest concentration at
which a method is able to detect truly contaminated samples with a statistical error of 1-β.
Twenty two blank bovine muscle samples were fortified with salicylic acid at 20 µg kg-1. In
100 % of these spiked samples salicylic acid was identified. So the detection capability CCβ
for salicylic acid is lower or equal to 20 µg kg-1. Also for phenylbutazone this experiment was
performed; twenty two blank bovine muscle samples were fortified at 50 µg kg-1. In 100 % of
these spiked samples phenylbutazone was identified. So the detection capability CCβ for
phenylbutazone is lower or equal to 50 µg kg-1. Using these data, the decision limit CCα
could be calculated. From a maximum coefficient of variation of 23 %, the corresponding
standard deviation was derived. The decision limit CCα was calculated as the detection
capability CCβ minus 1.64 times the standard deviation. In Table 11 the CCα and CCβ for
salicylic acid and phenylbutazone are summarized.
171 Printed: www.dclsigns.be
Veterinary drugs
Table 11 The validation parameters decision limit (CCα) and detection capability (CCβ) for
the NSAIDs salicylic acid and phenylbutazone
salicylic acid phenylbutazone
CCα (µg kg-1) ≤ 12.46 ≤ 31.14
CCβ (µg kg-1) ≤ 20 ≤ 50
Applicability/ruggedness
Applicability and ruggedness can best be tested when the analytical method is used for routine
analysis. Applicability is the observation of the consequences when minor reasonable
variations are introduced into the method. Such factors may include the analyst, temperature
during evaporation, pH values, as well as many other factors that can occur in the laboratory.
Applicability and ruggedness will be tested by control spiked samples or by participating in
performance studies. With each batch of samples a control spiked sample will be analysed and
once a year unknown control spiked samples are analysed.
Stability
Stability tests were started at the beginning of 2005. After one year the results will be
evaluated. At the beginning of 2005 a stock standard solution was prepared of the different
NSAIDs and of the internal standard flunixin-d3. From then on, a working solution was
prepared every three months, T0 was the initial working solution and T1, T2, … were the
working solutions prepared after three, six, … months, respectively. Every time the new
prepared working solution (T1, T2, …) and the initial working solution (T0) were analysed
and statistically compared (ANOVA test). After one year the results will be evaluated and the
period for which the working solution can be used, will be determined. Until now, no
significant degradation of the different NSAIDs was observed. The working solutions were
stored at -20 °C during the entire period of the stability study.
The multi-residue liquid chromatography-tandem mass spectrometry method for the detection
of NSAIDs in bovine muscle was validated according to the criteria of Commission Decision
2002/657/EEC [25].
172 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
3.4.5. Conclusion 2A LC-MS multi-residue method was developed to identify salicylic acid, phenylbutazone,
flunixin, tolfenamic acid, meloxicam and ketoprofen in bovine muscle. Ketoprofen was
detected in positive ion mode, while the other NSAIDs were detected in negative ion mode, so
two acquisitions were necessary to detect all NSAIDs. In addition, for the confirmation of
salicylic acid, tolfenamic acid and ketoprofen, full scan MS3-mass spectra of the [M-H]- or
[M+H]+ ions via their first transition product ion, were necessary. The ion trap parameters
activation q and maximum ion injection time, needed to be adapted for optimal detection of
salicylic acid and phenylbutazone, respectively.
This multi-residue method is a quantitative confirmation method for the NSAIDs , flunixin,
tolfenamic acid and meloxicam, and a qualitative method for the unauthorized NSAIDs
salicylic acid and phenylbutazone. The method was validated according to the criteria of
Commission Decision 2002/657/EEC.
173 Printed: www.dclsigns.be
Veterinary drugs
3.4.6. References
[1] J.R. Lawrence, R. Peter, G.J. Baxter, J. Robson, A.B. Graham, J.R. Paterson (2003)
Urinary excretion of salicyluric and salicylic acids by non-vegetarians, vegetarians, and
patients taking low dose aspirin, Journal of Clinical Pathology 56, 651-653
[2] K. Baert (2003) Pharmacokinetics and Pharmacodynamics of Non-Steroidal Anti-
Inflammatory Drugs in Birds, thesis, Ghent University, Faculty of Veterinary Medecine, 3-18
[3] D.M. Boothe (2001) The analgesic, antipyretic, anti-inflammatory drugs, In: Veterinary
Pharmacology and Therapeutics (8th edition), ed. H.R. Adams, Iowa State University Press,
Ames, 433-451
[4] E. Daeseleire, L. Mortier, H. De Ruyck, N. Geerts (2003) Determination of flunixin and
ketoprofen in milk by liquid chromatography-tandem mass spectrometry, Analytica Chimica
Acta 488, 25-34
[5] Anonymous (2003) Informal consolidated version of the Annexes I to IV of Council
Regulation n° 2377/90, The European Agency for the Evaluation of Medicinal Products, 22
July 2003
[6] K. De Wasch, N. Van Hoof ,S. Poelmans,L. Okerman ,D. Courtheyn,A. Ermens ,M.
Cornelis ,H.F. De Brabander (2003) Identification of “unknown analytes” in injection sites : a
semi-quantitative interpretation, Analytica Chimica Acta 483, 387-399
[7] N. Van Hoof, K. De Wasch, S. Poelmans, H.F. De Brabander, In: Rapid and on-line
instrumentation for food quality assurance, ed. I.E. Tothill, Woodhead Publishing Limited,
Cambridge, England, 91-115
[8] P.L. Boner, D.D.W. Liu, W.F. Feely, R.A. Robinson, J. Wu (2003) Determination of
flunixin in edible bovine tissues using liquid chromatography coupled with tandem mass
spectrometry, Journal of Agricultural and Food Chemistry 51, 7555-7559
[9] P.L. Boner, D.D.W. Liu, W.F. Feely, M.J. Wisocky, J. Wu (2003) Determination and
confirmation of 5-hydroxyflunixin in raw bovine milk using liquid chromatography tandem
mass spectrometry, Journal of Agricultural and Food Chemistry 51, 3753-3759
[10] P. van Eeno , F.T. Delbeke, K. Roels, K.Baert (2003) Detection and disposition of
tolmetin in the horse, Journal of Pharmaceutical and Biomedical Analysis 31, 723-730
[11] H.S. Rupp, D.C. Holland, R.K. Munns, S.B. Turnipseed, A.R. Long (1995)
Determination of flunixin in milk by liquid chromatography with confirmation by gas
chromatography-mass spectrometry and selected ion monitoring, Journal of AOAC
International 78, 959-967
174 Printed: www.dclsigns.be
Multi-residue LC-MSn method for the detection of NSAIDs in bovine muscle
[12] S.M.R. Stanley, N.A. Owens, J.P. Rodgers (1995) Detection of flunixin in equine urine
using high-performance liquid chromatography with particle-beam and atmospheric-pressure-
ionization mass spectrometry after solid-phase extraction, Journal of chromatography B-
Biomedical Applications 667, 95-103
[13] A.K. Singh, Y. Jang, U. Mishra, K. Granley (1991) Simultaneous analysis of flunixin,
naproxen,ethacrynic acid,indomethacin,phenylbutazone,mefenamic acid and thiosalicylic acid
in plasma and urine by high-performance liquid chromatography and gas chromatography-
mass spectrometry, Journal of chromatography B- Biomedical Applications 106, 351-361
[14] V. Hormazabal , M. Yndestad (2001) Simultaneous determination of chloramphenicol
and ketoprofen in meat and milk and chloramphenicol in egg, honey, and urine using liquid
chromatography-mass spectrometry, Journal of Liquid Chromatography & Related
Technologies 24, 2477-2486
[15] M.E. Abdel-Hamid, L. Novotny, H. Hamza (2001) Determination of diclofenac sodium,
flufenamic acid, indomethacin and ketoprofen by LC-APCI-MS, Journal of Pharmaceutical
and Biomedical Analysis 24, 587-594
[16] P.A. Asea, J.R. Patterson, G.O. Korsrud, P.M. Dowling, J.O. Boison (2001)
Determination of flunixin residues in bovine muscle tissue by liquid chromatography with
UV detection, Journal of AOAC International 84, 659-665
[17] J.L.Wiesner, A.D. de Jager, F.C.W. Sutherland, H.K.L. Hundt, K.J. Swart, A.F. Hundt,
J.Els (2003) Sensitive and rapid liquid chromatography-tandem mass spectrometry method
for the determination of meloxicam in human plasma, Journal of Chromatography B 785,
115-121
[18] M.I.G. Martin, C.I.S. Gonzales, A.J. Hernandez, M.D.G Cachan, M.J.C. de Cabo, A.L.G.
Cuadrado (2002) Determination by high-performance liquid chromatography of
phenylbutazone in samples of plasma from fighting bulls, Journal of chromatography B 769,
119-126
[19] Y. Luo, J.A. Rudy, C.E. Uboh, L.R. Soma, F. Guan, J.M. Enright, D.S. Tsang (2004)
Quantification and confirmation of flunixin in equine plasma by liquid chromatography-
quadrupole time-of-flight tandem mass spectrometry, Journal of chromatography B 801, 173-
184
[20] J.Y. Kim, S.J. Kim, K.J. Paeng, B.C. Chung (2001) Measurement of ketoprofen in horse
urine using gas chromatography-mass spectrometry, Journal of Veterinary Pharmacology and
Therapeutics 24, 315-319
175 Printed: www.dclsigns.be
Veterinary drugs
[21] S.K. Bakar, S. Niazi (1983) Stability of aspirin in different media, Journal of
Pharmaceutical sciences 72, 1024-1026
[22] E. De Hoffmann, J. Charette, V. Stroobant (1996) Mass analysers, In: Mass
Spectrometry, Principles and applications, John Wiley & Sons, Chichester, UK, 39-59
[23] R.E. March (1997) An Introduction to Quadrupole Ion Trap Mass Spectrometry, Journal
of Mass Spectrometry 32, 351-369
[24] P.S.H. Wong, R.G. Cooks (1997) Ion Trap Mass Spectrometry, Current Separations 16,
85
[25] Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive
96/23/EC concerning the performance of analytical methods and the interpretation of results
(2002), Official Journal of the European Communities, no. L 221
176 Printed: www.dclsigns.be
DISCUSSION
Consumers throughout the world are becoming more conscious about the importance of
healthy food. The confidence of the consumer has been tested a couple of times over the last
few years. After several meat scandals, consumers have become very critical when it comes to
their food. From these crises, it can be concluded that no approach, no agency, no government
whatever will be able to protect the population completely. However, it is the duty of
inspection services and scientists to learn from these situations and to work together in order
to avoid the next one.
Due to the increasing pressure of rentability demands, farmers are pushed towards a more
intensive production and consequently towards the use of veterinary medicinal products. The
administration of veterinary medicinal products may result in the presence of residues of these
substances or their metabolites in food from animal origin, and these residues may produce
hazards for public health. A wide range of veterinary medicinal products is administered
legitimately to farm animals to treat outbreaks of diseases or prevent diseases from spreading.
In order to reduce the likelihood of harmful levels of these veterinary drugs reaching the
human food chain, the European Union and many other countries have set Maximum Residue
Limits (MRL). Besides regulated veterinary drugs, there are also veterinary medicinal
products which are used illegally with the intention to improve the feed intake and promote
growth of animals. Although the use of growth promoters is forbidden in the European Union
from January 2006 onwards, some farmers still use these compounds during fattening of
cattle. Enormous profits have stimulated the black market to continue with illegal practises, in
spite of severe punishments.
Regulatory bodies are required to enforce and verify the requirements set by the European
Union. Therefore, official samples taken at the slaughterhouse or the farm are analysed for
unauthorised substances and registered veterinary drugs. Laboratories which are part of food
quality assurance must test these food products and farm samples to ensure that regulations
are met.
177 Printed: www.dclsigns.be
Liquid chromatography-tandem mass spectrometry
The successful combination of liquid chromatography with mass spectrometry is one of the
most important developments of the last decades. In the beginning, most of the attention was
given to solving interface problems and building new technologies. Nowadays, LC-MS is
used in routine experiments in residue analysis. The integration of separation, ionisation and
detection has enabled LC-MS to become a practical problem-solving tool. While GC-MS
instruments were historically the more widely used for various classes of residues, LC-MS
today appears as the method of choice and the major actual investment for many laboratories.
Because of its flexibility, a lot of new applications in residue analysis are developed using
LC-MS in favor of GC-MS. LC is capable of providing routine separations of compounds
unsuitable for GC analysis. Using LC-MS it is easy to switch between different applications,
different mobile phases, columns and interfaces. This flexibility allows research labs to give a
fast response to the ever changing environment of legal and illegal veterinary medicine.
LC-MSn was the method of choice in this thesis and in residue analysis in general. Looking at
the possibilities and the price of a LC-MSn apparatus, LC-MSn is the best buy considering the
broad range of applications in residue analysis. LC-MSn has the capability to identify and
quantify veterinary medicinal products at low concentrations and due to tandem mass
spectrometry structure elucidation of unknowns and metabolites is possible. So, LC-MSn
covers the needs of residue analysis, it is a complete solution and an indispensable tool in
residue analysis.
To develop analytical methods in residue analysis, the availability of high technological
equipment operated by well trained personnel is necessary. Besides knowledge on the
substance of interest, it is as important to know the instrument you work with. Tandem mass
spectrometry is the succession of two mass-selective operations. The objective of the first
operation is to isolate an ion species. The second operation determines the m/z ratios of the
fragment ions. Once the isolation of the selected ion is completed, the applied rf amplitude is
reduced again to obtain a stable precursor ion at a certain qz-value. At this qz-value not only
the precursor ion but also the product ions need to be stable to obtain a MS2-full scan mass
spectrum in which each scan contains the different product ions. Some substances,
particularly small molecules, may not be stable at the default activation q of 0.25. Therefore,
the qz-value needs to be adapted for such compounds to obtain more stability for both the
precursor ion as the product ions. So, it is important to know and check the possibilities MS
offers to guarantee accurate and precise results.
178 Printed: www.dclsigns.be
Discussion
After a first period of great enthusiasm shared by most end-users, some problems related to
LC-MS techniques started to be reported. One main source of pitfalls was the existence of
matrix effects in general, and ion suppression phenomenon in particular. Ion suppression is a
problem occurring in the early stages of the ionization process. It can occur when a coeluted
compound suppresses the ionization of the sample molecules in the MS source. This
phenomenon affects many aspects of the method performance such as identification, detection
capability and repeatability. Ion suppression could lead to false compliant results due to the
non-detection of an existing analyte, the underestimation of the real concentration, or the non-
fulfilment of the identification criteria. To overcome ion suppression different actions can be
taken. The chromatographic and mass spectrometric conditions can be adapted. However, the
only way to definitively overcome this problem is to improve the sample preparation and
purification in order to limit the presence of interfering compounds in the final extract.
Consequently, to prevent problems regarding false compliant results and repeatability, one
should adopt a standard practice that acknowledges the necessity of improved sample
preparation [1-4].
Legislation
Besides the availability of modern analytical instrumentation and trained personnel, a
European or even worldwide legislation and harmonization concerning residue control should
be mandatory. Through this regulation the quality and comparability of analytical results
generated by laboratories approved for official residue control can be ensured.
The three most important EU documents on the control of veterinary medicinal products are
Council Regulation (EEC) No 2377/90, Council Directive 96/23/EC and Commission
Decision 2002/657/EC. Council Regulation EC N° 2377/90 establishes maximum residue
limits for veterinary drugs in foodstuffs of animal origin [5]. Council Directive 96/23/EC
establishes National Surveillance Schemes for monitoring of residues of veterinary medicinal
products and contaminants [6]. In order to ensure the harmonized implementation of Directive
96/23/EC, performance criteria for analytical residue methods are defined in Commission
Decision 2002/657/EC [7]. A parameter that was added to harmonize the analytical
performance of methods for substances, for which no permitted limit has been established, is
the minimum required performance limit (MRPL). A MRPL is based on the characteristics of
the available analytical methods.
Harmonisation, through legislation, is necessary to guarantee the uniformity of analytical
results generated by laboratories. On the other hand, it would be even more interesting that
179 Printed: www.dclsigns.be
laboratories approved for official residue control join forces to develop appropriate methods
which are subsequently used by different laboratories. In that way there would be no
competition between the laboratories and the quality of the results is guaranteed.
A problem which is not encountered by legislation, is the lack of appropriate standards of new
veterinary medicinal products. Such standards are necessary to develop an accredited method
which can subsequently be used in residue control. Without these standards the identification
of new veterinary medicinal products in control samples is impossible. It would be ideal that
standards are available via national and community reference laboratories.
Veterinary medicinal products
The number of reported non-compliant samples for unauthorized substances in the last years
was limited, despite of the intense control of illegal growth promoters within the European
Union. Based on this information, it could be concluded that the illegal use of growth
promoters has decreased. Sampling of preparations such as syringes and pharmaceuticals,
however, indicates that there is a shift to esters and analogues of known compounds, products
licensed in other countries and even new substances. Consequently, the molecule escapes the
regulatory control completely. Adding or deleting a group in a molecule may not change very
much the action wanted, but confirmation methods are by-passed. This is illustrated with
three examples that have been studied in the laboratory of Chemical Analysis (not all of the
examples are incorporated in this thesis). A first example of a new growth promoter is the
anabolic steroid norchlorotestosterone acetate. It resembles the known chlorotestosterone
acetate but lacks the methyl-group at position C19 [8-10]. Another example is the beta-
agonist zilpaterol. Zilmax® is licensed for use as feed additive in Mexico and South-Africa.
Its chemical structure is different from the well-known beta-agonists and therefore it could not
be detected with the conventional screening and confirmation methods. Four years ago,
different European laboratories started with the development of analytical methods which are
able to detect both zilpaterol as well as other beta-agonists [11-17]. These methods will be
used both for the control of export products to the EU as well as for the control of misuse of
zilpaterol in the European Union. A third example is methyl-3-methyl-2-
quinoxalinecarboxylate-1,4-dioxide (MMQCD), a quinoxaline. Carbadox and olaquindox,
compounds with growth promoting activity, were banned in 1998 for use as feed additive
because of their carcinogenic and mutagenic activity. With this ban, there is a potential risk
for the use of unknown quinoxalines. MMQCD is structurally related to carbadox and
olaquidox and has been recovered in feed from both Spain and Italy. So, MMQCD seems to
180 Printed: www.dclsigns.be
Discussion
be present on the black market in many countries, while no methods are available for the
control in feed or food [18-20].
Syringes and pharmaceuticals, confiscated on farms, are analysed in the Laboratory of
Chemical Analysis to screen for know veterinary medicinal products and unknowns.
Applying tandem mass spectrometry, the structure of unknowns can be elucidated.
Subsequently, animal experiments will be performed, if the product is available on the
market, to look for possible metabolites in urine and faeces and to study the excretion profile
of the parent compound and possible metabolites to get an idea about the concentrations that
can be expected after administration of the veterinary medicinal product. This is important
information for the set-up of monitoring plans and analytical methods.
Not only the use of growth promoters is condemned by the consumer, also the attitude
towards the intense use of veterinary drugs, such as antibiotics, is not positive. The European
Union as well as many other countries have set Maximum Residue Limits in order to reduce
the likelihood of harmful levels of authorized drugs reaching the human food chain. Below
these limits residues are assumed to be harmless to the consumer. Some substances do not
need a MRL because their use is not considered harmful for the public health. The use of
veterinary drugs is allowed exclusively for the animal species identified and according to the
conditions established (e.g. route of administration). According to the Belgian legislation,
drugs that are not registered in Belgium cannot be legally administered to food-producing
animals.
Besides identification, also quantification is important for these substances. Quantification
compares the concentration of the analyte in the sample with the MRL value prescribed by
law. If the MRL concentration is exceeded, the farmer probably did not respect the
withdrawal period or the applied dose was higher than the prescribed one.
The most recent class of veterinary drugs is the fluoroquinolones. They were introduced
during the 1980s in human medicine and their use in veterinary medicine has increased
tremendously in the last ten years. This class of drugs illustrates that drugs which are widely
used in human medicine and which are available at convenient prices, find their way to
veterinary medicine. One of the most commonly used therapeutic drug classes world-wide are
the NSAIDs. They are often the initial therapy for inflammation disorders. In recent years the
treatment of pain in animals has become an important issue, even in food-producing animals.
Some NSAIDs are not authorised for use in bovine species in Belgium. Phenylbutazone,
although prohibited by law, is still used by farmers because of its activity and its low cost.
181 Printed: www.dclsigns.be
Salicylic acid is an Annex II compound, but it is not registered in Belgium for use in bovine
species. In livestock breeding, salicylates are routinely used to condition animals just after
transport to reduce the effects of stress. Salicylates are available on the market at convenient
price.
In the Laboratory of Chemical Analysis a multi-residue confirmatory method was developed
for both quinolones and NSAIDs. These methods were developed on request of the FAVV
because of the lack of such multi-residue confirmatory methods and the need for the detection
of these veterinary medicinal products in different matrices.
Future trends in residue analysis and conclusion
Laboratories involved in residue control, should couple routine analysis with research in the
same area. Due to the continuous occurrence of new veterinary medicinal products, it is not
enough anymore to detect known substances. We must be aware of the inventiveness of
people operating on the black market and the possible use of new or not-registered veterinary
medicinal products.
The use of traditional growth promoters has decreased in the last few years, but some farmers
are still using compounds with hormonal, thyrostatic or adrenergic action during fattening of
livestock. Enormous profits have stimulated the black market to continue their research on
and production of growth promoting agents. Being aware of the enhanced selectivity of
residue analysis, the stress of the development of growth promoters is on esters and analogues
of (natural) compounds.
Besides new or altered substances, also combinations of products are applied. These products
may have a similar action or a combination of different actions is possible to create a
synergistic effect. Each compound is added at very low individual doses.
Another strategy used today to recover new veterinary medicinal products is based on
naturally occurring compounds. Phytotherapy, the use of herbal medicinal products, knows an
increasing popularity in veterinary medicine as an alternative to antibiotics [21-24]. For the
moment, however, there is no regulation concerning the use of plant extracts in veterinary
medicine. Besides phytotherapy, promising new growth promoters are searched in naturally
occurring substances. Examples are ecdysteroids which have been identified in plants and in
invertebrates. As early as 1963, it was found that ecdysterone enhanced the rate of protein
synthesis in mammalian tissue [25]; in 1969 the anabolic effect was confirmed in mice [26-
29]. Another example is the use of growth hormone, somatotropin. Since the development of
182 Printed: www.dclsigns.be
Discussion
recombinant bovine somatotropin, it has been widely used to stimulate milk production in
cows [30-36]. A final example is ipriflavone. Ipriflavone is derived from the natural occurring
isoflavone daidzein. According to body-builder websites, ipriflavone should increase the
testosterone production [37-38]. The effect of ipriflavone has not yet been investigated in
bovine species. The use of naturally occurring substances or derivatives of these compounds
gives promising results as alternative for veterinary drugs and for growth promotion. This
source of new compounds is still under investigation and will give new opportunities to the
farmers. From analytical point, there is the difficulty to discriminate between a natural source
and therapeutic or anabolic treatment.
Residue analysis of veterinary medicinal products should be focusing on two domains.
Selective MSn confirmatory methods should be developed to detect low concentrations of
unauthorized compounds and to quantify veterinary drugs at or below their MRL
concentration. Mass spectrometry is necessary for the unambiguous identification of analytes.
Tandem mass spectrometry will increase the selectivity of the method. The consequence of
increasing the selectivity is that minor changes are neglected. Therefore, screening methods
are necessary for the structure elucidation of unknown, new compounds. Multi-residue
generic MS methods must be developed to detect a wide range of veterinary medicinal
products. Using a soft ionization technique the MS full scan mass spectrum will provide the
molecular mass and fragmentation will give extra structural information. Besides LC-MSn
there are other techniques which can provide additional information to the LC-MSn data. One
of these techniques which is extremely powerful and relatively new in residue analysis is
TOF-MS, where a mass spectrometer is combined with a time-of-flight (TOF) instrument.
This technology is very promising in structure elucidation of unknown compounds due to the
accurate mass measurement. Using these accurate masses together with the fragmentation
capability of mass spectrometry, the chemical structure of the unknown compound can be
found out. This technology is becoming more popular in residue analysis due to its screening
possibilities [39-43]. However, the cost of TOF-MS makes it a technology which is not and
will not be widespread due to the economical impossibility of many laboratories concerned in
residue analysis to buy this apparatus. So, LC-MSn remains the standard method of choice in
residue analysis and TOF-MS is an optional technology.
Regulatory control bodies and laboratories will need to find ways to cope with the ever
changing environment of legal and illegal veterinary medicine. These changes make method
183 Printed: www.dclsigns.be
development in residue analysis to a challenge for the laboratories. The use of mass
spectrometry and the evolution in this technology will enable scientists to cope with these
challenges.
During the past four years LC-MSn methods have been developed and optimized for the
detection of unauthorized substances and veterinary drugs. These analytical methods were
subsequently implemented in monitoring programs. As an individual it is only possible to
make a small contribution to the quality assurance of food products of animal origin and
therefore it is important that scientists work together to restore the confidence of the consumer
and protect its health.
184 Printed: www.dclsigns.be
Discussion
References
[1] P.J. Taylor (2005) Matrix effect: The Achilles heel of quantitative high-performance
liquid chromatography-electrospray-tandem mass spectrometry, Clinical Biochemistry 38,
328-334
[2] B.K. Choi, D.M. Hercules and A.I. Gusev (2001) Effect of liquid chromatography
separation of complex matrices on liquid chromatography-tandem mass spectrometry signal
suppression, Journal of Chromatography A 907, 337-342
[3] J.P. Antignac, K. De Wasch, F. Monteau, H. De Brabander, F. Andre, B. Le Bizec (2005)
The ion suppression phenomenon in liquid chromatography-mass spectrometry and its
consequences in the field of residue analysis, Analytica Chimica Acta 529, 129-136
[4] T.M. Annesley (2003) Ion suppression in mass spectrometry, Clinical Chemistry 49,
1041-1044
[5] Council Regulation (EEC) N° 2377/90 of 26 June 1990 laying down a Community
procedure for the establishment of maximum residue limits of veterinary medicinal products
in foodstuffs of animal origin (1990), Official Journal of the European Communities, no. L 67
[6] Council Directive 96/23/EC of 29 April 1996 on measures to monitor certain substances
and residues thereof in live animals and animal products and repealing Directives 85/358/EEC
and 86/469/EEC and Decision 89/187/EEC and 91/664/EEC (1996), Official Journal of the
European Communities, no. L 125
[7] Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive
96/23/EC concerning the performance of analytical methods and the interpretation of results
(2002), Official Journal of the European Communities, no. L 221
[8] D. Courtheyn, B. Le Bizec, G. Brambilla, H.F. De Brabander, E. Cobbaert, M. Van de
Wiele, J. Vercammen, K. De Wasch (2002) Recent development in the use and abuse of
growth promoters, Analytica Chimica Acta 473, 71-82
[9] N. Van Hoof, K. De Wasch, S. Poelmans, D. Bruneel, S. Spruyt, H. Noppe, C. Janssen, D.
Courtheyn and H. De Brabander (2004) Norchlorotestosterone acetate: GC-MS2 analysis in
kidney fat, urine and faeces and study of the metabolisation by Neomysis integer,
Chromatographia 59, 85-93 [10] B. Le Bizec, N. Van Hoof, D. Courtheyn, I. Gaudin, M. Van de Wiele, E. Bichon, H. De
Brabander and F. André (2005) Metabolism study of a new anabolic steroid in bovine:
preliminary data on 19-norchlorotestosterone acetate, Journal of Steroid Biochemistry and
Molecular Biology (accepted July 2005), in press
185 Printed: www.dclsigns.be
[11] C.S. Stachel, W. Radeck, P. Gowik (2003) Zilpaterol – a new focus of concern in residue
analysis, Analytica Chimica Acta 493, 63-67
[12] B. Bocca, C. Cartoni, M. Di Mattia (2003) Feed additives in animal nutrition:
Quantification of a new adrenergic drug by hyphenated techniques, Journal of Separation
Science 26, 363-368
[13] B. Bocca, M. Di Mattia, C. Cartoni, M. Fiori, M. Felli, B. Neri, G. Brambilla (2003)
Extraction, clean-up and gas chromatography-mass-spectrometry characterization of
zilpaterol as feed additive in fattening cattle, Journal of Chromatography B 783, 141-149
[14] B. Bocca, M. Fiori, C. Cartoni, G. Brambilla (2003) Simultaneous determination of
zilpaterol and other beta agonists in calf eye by gas chromatography/tandem mass
spectrometry, Journal of AOAC International 86, 8-14
[15] N. Van Hoof, R. Schilt, E. Van der Vlis, P. Boshuis, M. Van Baak, A. Draaijer, K. De
Wasch, M. Van de Wiele, J. Van Hende, D. Courtheyn and H. De Brabander (2005) Detection
of zilpaterol (Zilmax®) in calf urine and faeces with liquid chromatography–tandem mass
spectrometry, Analytica Chimica Acta 529, 189-197
[16] W.L. Shelver, D.J. Smith (2004) Enzyme-linked immunosorbent assay development for
the beta-adrenergic agonist zilpaterol, Journal of Agricultural and Food Chemistry 52, 2159-
2166
[17] W.L. Shelver, H.J. Kim, Q.X. Li (2005) Development of a monoclonal antibody-based
enzyme-linked immuosorbent assay for the beta-adrenergic agonist zilpaterol, Journal of
Agricultural and Food Chemistry 53, 3273-3280
[18] K. De Wasch, H. De Brabander, K. Haustraete, A. Fontaine, B. Kestelyn, N. De Kimpe,
P. Batjoens, M. Cornelis (1997) Illegally used quinoxalines in Europe, Proceedings 7the
European Association for Veterinary Pharmacology & Toxicology (EAVPT), Madrid, Spanje
[19] C. Testa, G. Chessa, A. Orru, G. Calaresu, A. Pulina (1996) HPLC method for the
determination of methyl 3-methylquinoxaline-2-carboxylate-1,4-di-N-oxide in animal feed,
Proceedings Euroresidue III, Veldhoven, The Netherlands, 943-947
[20] N. Van Hoof, K. De Wasch, M. De Moor, D. Bruneel, D. Courtheyn, S. Poelmans, H.
Noppe, H.F. De Brabander (2003) Unknown quinoxalines, one of the dangers of black market
products, Proceedings EuroFoodChem XII, Brugge, Belgium
[21] M.J. Smith-Schalkwijk (1999) Veterinary phytotherapy: an overview, Canadian
Veterinary Journal 40, 891-892
[22] R.E. Uncini Manganelli, F. Camangi, P.E. Tomei (2001) Curing animals with plants:
traditional usage in Tuscany (Italy), Journal of Ethnopharmacoly 78, 171-191
186 Printed: www.dclsigns.be
Discussion
[23] K.B. Haas (2000) Animal therapy over the ages: 5. American botanical medicine,
Veterinary Heritage 23, 43-45
[24] J. Reichling, R. Saller (2001) Herbal remedies in veterinary phytotherapy, Schweizer
Archiv für Tierheilkunde 143, 395-403
[25] W.J. Burdette, R.L. Coda (1963) Effect fo ecdysone on incorporation of C-14-leucine
into hepatic protein in vitro, Proceedings of the Society for Experimental Biology and
Medecine 11, 216-217
[26] H. Hikino, S. Nabetani, K. Nomoto, T. Arai, T. Takemoto (1969) Effect of long term
administration of insect-metamorphosing substances on higher animals, Yakugaku Zasshi 89,
235-240
[27] M. Bathori (2002) Phytoecdysteroids effects on mammalians, isolation and analysis,
Mini Rev Med Chem. 2, 285-293
[28] K. Slama, K. Koudela, J. Tenora, A. Mathova (1996) Insect hormones in vertebrates:
anabolic effects of 20-hydroxyecdysone in Japanese quail, Experientia 52, 702-706
[29] C. Tsitsimpikou, G.D. Tsamis, P.A. Siskos, M.H. Spyridaki, C.G. Georgakopoulos
(2001) Study of excretion of ecdysterone in human urine, Rapid Communication Mass
Spectrometry 15, 1796-1801
[30] A.V. Capuco, G.E. Dahl, D.L. Wood, U. Moallem, R.E. Erdman (2004) Effect of bovine
somatotropin and rumen-undegradable protein on mammary growth of prepubertal dairy
heifers and subsequent milk production, Journal of Dairy Science 87, 3762-3769
[31] I.R. Dohoo, K. Leslie, L. DesCoteaux, A. Fredeen, P. Dowling, A. Preston, W. Shewfelt
(2003) A meta-analysis review of the effects of recombinant bovine somatotropin. 1.
Methodology and effects on production, Canadian Journal of Veterinary Research 67, 241-
251
[32] A. Baldi (1999) Manipulation of milk production and quality by use of somatotropin in
dairy ruminants other than cow, Domestic Animal Endocrinology 17, 131-137
[33] T.D. Etherton, D.E. Bauman (1998) Biology of somatotropin in growth and lactation of
domestic animals, Physiolical Reviews 78, 745-761
[34] K. Sejrsen, N. Oksbjerg, M. Vetsergaard, M.T. Sorensen (1995) Growth hormone and
related peptides as growth promoters, Proceedings of the Scientific Conference on Growth
Promotion in Meat Production, Brussels, Belgium, 87-119
[35] N. Rochut, B. Le Bizec, F. Monteau, F. André (2000) ESI-MS for the measurement of
bovine and porcine somatotropins, Analusis 28, 22-26
187 Printed: www.dclsigns.be
[36] G. Pinel, F. Andre, B. Le Bizec (2004) Discrimination of recombinant and pituitary-
derived bovine and porcine growth hormones by peptide mass mapping, Journal of
Agricultural and Food Chemistry 52, 407-414
[37] www.bodybuilding.com
[38] K. Verheyden, N. Van Hoof, S. Poelmans, H. Noppe, C.R. Janssen and H.F. De
Brabander (2005) Study of the androgenic activity of ipriflavone by exposure of Neomysis
integer, Proceedings Recent Advances in Food Analysis, Prague, Czech Republic
[39] E.M. Thurman, I. Ferrer, A.R. Fernandez-Alba (2005) Matching unknown empirical
formulas to chemical structure using LC/MS TOF accurate mass and database searching:
example of unknown pesticides on tomato skins, Journal of Chromatogry A 1067, 127-134
[40] A.A.M. Stolker, W. Niesing, E.A. Hogendoorn, F.M. Versteegh, R. Fuchs, U.A.T.
Brinkman (2004) Liquid chromatography with triple-quadrupole or quadrupole-time-of-flight
mass spectrometry for screening and confrmation of residues of pharmaceuticals in water,
Analytical and Bioanalytical Chemistry 378, 955-963
[41] I. Ferrer, J.F. Garcia-Reyes, M. Mezcua, E.M. Thurman, A.R. Fernanndez-Alba (2005)
Multi-residue pesticide analysis in fruits and vegetables by liquid chromatography-time-of-
flight mass spectrometry, Journal of Chromatography A 1082, 81-90
[42] M.W. Nielen, E.O. van Bennekom, H.H. Heskamp, J.H. van Rhijn, T.F. Bovee, L.R.
Hoogenboom (2004) Bioassay-directed identification of estrogen residues in urine by liquid
chromatography electrospray quadrupole time-of-flight mass spectrometry, Analytical
Chemistry 76, 6600-6608
[43] A.L. Stolker, W. Niesing, R. Fuchs, R.J. Vreeken, W.M. Niessen, U.A. Brinkman (2004)
Liquid chromatography with triple-quadrupole and quadrupole-time-of-flight mass
spectrometry for the determination of micro-constituents - a comparison, Analytical and
Bioanalytical Chemistry 378, 1754-1761
188 Printed: www.dclsigns.be
SUMMARY
In recent years the number of non-compliant samples containing illegal growth promoters has
decreased. However, there are indications of a shift to esters and analogues of known
compounds, products licensed in other countries and even new substances. These structural
changes can elude confirmatory tests. An example of such a compound is zilpaterol.
Zilpaterol is a beta-agonist of the third generation with a chemical structure which is different
from the well known beta-agonists. Therefore, a multi-residue LC-MSn method must be
developed which is able to detect both the known beta-agonists as well as the new beta-
agonist zilpaterol. Besides growth promoters, also the use of veterinary drugs needs to be
controlled. In this work confirmatory methods were developed for two important classes of
therapeutic drugs, quinolones and non-steroidal anti-inflammatory drugs (NSAIDs).
Fluoroquinolones are a group of relatively new highly-potent synthetic antibiotic compounds
and NSAIDs are probably one of the most commonly used therapeutic drug classes.
The goal of this work was to develop and optimize LC-MSn methods for the detection of
residues of beta-agonists and veterinary drugs. The thesis imparts only a fraction of the
research that has been done over the last four years in the framework of two FOD research
projects, S-6044/S3 and S-6150. The developed analytical methods were subsequently
implemented in monitoring programs.
Chapter 1 is a theoretical introduction on the hyphenation of liquid chromatography and
mass spectrometry. The successful combination of LC with MS is one of the most important
developments of the last decades. LC is capable of providing routine separations of
compounds unsuitable for GC analysis without the necessity of preparing volatile derivatives.
However, the interface between LC and MS has always been a bottleneck to achieve an ideal
LC-MS system. This chapter gives an overview of the development of LC-MS interfaces and
a more detailed description of the most commonly used interfaces in residue analysis,
electrospray ionisation and atmospheric pressure chemical ionisation interface. There are
different types of mass spectrometers, but only ion trap mass spectrometry is discussed in this
chapter because the Laboratory of Chemical Analysis exclusively has ion trap mass
spectrometers.
Besides an introduction on LC-MS, the first chapter also includes relevant legislative aspects.
Council Directive 96/23/EC, as amended, comprises the residue control of food-producing
189
Printed: www.dclsigns.be
animals as well as their primary products and it divides all residues into Group A compounds,
which comprise unauthorized substances and Group B compounds which comprise all
authorized veterinary medicinal products. In order to ensure the harmonized implementation
of Directive 96/23/EC, performance criteria for analytical residue methods are defined in
Commission Decision 2002/657/EC.
In chapter 2 the new beta-agonist zilpaterol is studied and the development of a multi-residue
LC-MSn method for the detection of beta-agonists in urine is discussed.
Chapter 2.1 evaluates two different clean-up steps, Clean Screen Dau (mixed-phase C8 and
SCX) and Molecular Imprinted Polymers, with respect to their ability to minimise ion
suppression in LC-MSn. Ion suppression will influence the expected analytical results by
affecting method performances such as identification criteria, detection capability and
repeatability. Ion suppression experiments revealed that CSD sample clean-up could lead to
false compliant results for some beta-agonists. The percentages of the expected signal actually
observed show that there is less suppression of the signals when urine is pretreated with MIP
columns, i.e., clean-up using MIP columns is more selective for most beta-agonists than that
using CSD columns. This study has shown that molecular imprinted polymers are very
promising for sample clean-up for beta-agonists in the prevention of problems regarding false
compliant results and repeatability problems.
In chapter 2.2 the beta-agonist zilpaterol is studied. This study was performed in corporation
with TNO in the Netherlands. Zilpaterol is a new beta-agonist developed as growth promoter
for cattle. Zilmax® has been licensed as feed additive in Mexico and South Africa. In this
chapter the excretion profile of zilpaterol in urine and faeces was studied after oral treatment
of a male veal calf with therapeutic doses of Zilmax®. The detection of zilpaterol in urine and
faeces could be easily achieved. The levels of zilpaterol in the urine samples were relatively
high. Already after 2 days the concentration of zilpaterol exceeded 1000 μg l−1. A steady-state
concentration of about 1200 μg l−1 was quickly reached. Also in faeces, a steady-state
concentration of 83 μg kg−1 was quickly reached (first measurement was already 71 μg kg−1
on day 2). It could be concluded that zilpaterol can easily be detected during a treatment with
Zilmax®. As the animal was sacrificed after the last treatment, no data are available for the
final elimination of zilpaterol.
Chapter 3 is dedicated to LC-MSn method development for the detection of registered
veterinary drugs. This chapter consists of two main parts: the detection of veterinary
190
Printed: www.dclsigns.be
Summary
medicinal products in injection sites and the development of specific LC-MSn confirmation
methods for the veterinary drugs, quinolones and non-steroidal anti-inflammatory drugs.
In chapter 3.1 a multi-residue LC-MSn method was developed for the identification and
semi-quantification of veterinary medicinal products in injection sites. From the beginning of
2001, injection sites have been sampled at the slaughterhouse for identification of legally and
illegally used veterinary medicinal products. In analysing these samples, an overview could
be given of the products which are frequently used nowadays in practice and the approach for
screening can be altered. Since, it has become too expensive and too time consuming to check
every sample with specific analytical methods for a whole batch of different groups of
veterinary medicinal products, an alternative approach is proposed. A simple extraction and
clean-up is combined with a multi-residue LC-MSn identification and/or semi-quantification.
Because of the high concentrations of veterinary drugs in injection sites, there is no demand
for quantification in the concentration range of the MRL. An alternative validation is used
comparing the analyte concentration in the sample with the spike at MRL and 10 times MRL
concentration. This application illustrates the advantage of using LC-MSn with a default
gradient as a fast screening and confirmation technique for highly concentrated samples.
Based on the results obtained from the analysis of injection sites and on demand of the
Federal Agency for the Safety of the Food Chain, a quantitative confirmation method was
created for quinolones and NSAIDs.
Chapter 3.2 describes the LC-MS2 multi-residue method developed to simultaneously
analyse eight quinolones in muscle tissue and bovine milk. A simple and rapid extraction and
clean-up method was applied for the different matrices and ion trap mass spectrometry was
used as identification as well as quantification technique. All quinolones were detectable at
and below their MRL concentration. The multi-residue method for the detection of quinolones
in muscle tissue and bovine milk was validated according the criteria of Commission
Decision 2002/657/EEC. However, the validation for bovine milk is not yet completed.
Chapter 3.3 discusses the development of a LC-MSn multi-residue method to identify
salicylic acid, phenylbutazone, flunixin, tolfenamic acid, meloxicam and ketoprofen in bovine
muscle. The ion trap parameters “activation q” and “maximum ion injection time”, needed to
be adapted for optimal detection of salicylic acid and phenylbutazone, respectively. This
multi-residue method is a quantitative confirmation method for the NSAIDs , flunixin,
tolfenamic acid and meloxicam, and a qualitative method for the unauthorized NSAIDs
salicylic acid and phenylbutazone. The method was validated according to the criteria of
Commission Decision 2002/657/EEC.
191
Printed: www.dclsigns.be
SAMENVATTING
Het aantal niet-conforme stalen waarin illegale groeibevorderaars werden teruggevonden is de
laatste jaren gedaald. Nochtans zijn er aanwijzingen van een verschuiving in de richting van
esters en analogen van gekende componenten, producten die toegelaten zijn in andere landen
en zelfs nieuwe producten. Bevestigingsmethoden worden omzeild door deze structurele
veranderingen. Een voorbeeld is de component zilpaterol. Zilpaterol is een beta-agonist van
de derde generatie met een chemische structuur die verschilt van de gekende beta-agonisten.
Daarom is het noodzakelijk om een multi-residu LC-MSn methode te ontwikkelen die in staat
is zowel de gekende beta-agonisten als de nieuwe beta-agonist zilpaterol te detecteren.
Naast groeibevorderaars is het eveneens noodzakelijk om het gebruik van diergeneesmiddelen
te controleren. In dit werk zijn bevestigingsmethoden ontwikkeld voor twee belangrijke
klasses van therapeutische geneesmiddelen, namelijk quinolones en niet-steroïdale anti-
inflammatoire geneesmiddelen (NSAIDs). Fluoroquinolones zijn relatief nieuwe, krachtige
synthetische antibiotica en NSAIDs zijn waarschijnlijk één van de meest gebruikte
therapeutische geneesmiddelen.
Het doel van dit werk is de ontwikkeling en de optimalisatie van LC-MSn methoden voor de
detectie van residuen van beta-agonisten en diergeneesmiddelen. Het proefschrift omvat
slechts een fractie van het onderzoek dat gedurende de laatste 4 jaar werd uitgevoerd in het
kader van twee FOD onderzoeksprojecten, S-6044/S3 en S-6150. De ontwikkelde analytische
methoden zullen vervolgens worden geïmplementeerd in controleprogramma’s.
Hoofdstuk 1 is een theoretische introductie over vloeistofchromatografie (LC) gekoppeld aan
massaspectrometrie (MS). De succesvolle combinatie van LC met MS is één van de meest
belangrijke ontwikkelingen van de laatste decennia. LC is in staat om componenten te
scheiden die niet geschikt zijn voor analyse met gaschromatografie en dit zonder de noodzaak
om vluchtige derivaten te creëren. Echter de interface tussen LC en MS is steeds een knelpunt
geweest om een ideaal LC-MS systeem te verkrijgen. Dit hoofdstuk geeft een overzicht van
de ontwikkelingen van LC-MS interfaces en geeft eveneens een gedetailleerde beschrijving
van de meest gebruikte interfaces in residu-analyse, namelijk electrospray ionisatie en
193Printed: www.dclsigns.be
atmosferische-druk-chemische ionisatie. Er zijn verschillende types massaspectrometers, maar
enkel ion trap massaspectrometrie zal in dit hoofdstuk beschreven worden aangezien het
Laboratorium voor Chemische Analyse uitsluitend ion trap massaspectrometers heeft.
Er wordt niet enkel een introductie over LC-MS gegeven in hoofdstuk 1, maar er worden
eveneens relevante aspecten van de wetgeving besproken. In Richtlijn 96/23/EG wordt de
residucontrole van voedselproducerende dieren en hun primaire producten besproken en
worden de residuen verdeeld in Groep A componenten, niet-toegelaten producten en groep B
componenten, alle toegelaten diergeneesmiddelen. Om een geharmoniseerde implementatie
van Richtlijn 96/23/EG te verzekeren, zijn de prestatieparameters voor de analytische
residumethoden gedefinieerd in Beschikking 2002/657/EG.
In hoofdstuk 2 wordt de nieuwe beta-agonist zilpaterol bestudeerd en de ontwikkeling van
een multi-residu LC-MSn methode voor de detectie van beta-agonisten in urine wordt
besproken.
Hoofdstuk 2.1 vergelijkt twee verschillende opzuiveringstechnieken, namelijk clean screen
dau (CSD) en MIP polymeren (molecular imprinted polymers), omwille van hun
mogelijkheid om de onderdrukking van de ionisatie in LC-MSn te minimaliseren. Zulke
onderdrukking zal de verwachte analytische resultaten beïnvloeden door prestatieparameters,
zoals identificatiecriteria, detectielimiet en herhaalbaarheid, te beïnvloeden. Experimenten om
de onderukking van de ionisatie te onderzoeken, tonen aan dat de opzuivering van stalen met
CSD tot vals conforme resultaten kan leiden voor sommige beta-agonisten. De percentages
van de verwachte signalen die werkelijk geobserveerd worden, tonen aan dat er minder
onderdrukking is van de signalen wanneer de urine werd voorbehandeld met MIP kolommen;
dus opzuivering met MIP kolommen is selectiever voor de meeste beta-agonisten dan het
gebruik van CSD kolommen. Deze studie heeft aangetoond dat MIP polymeren veelbelovend
zijn voor de opzuivering van beta-agonisten om problemen van vals conforme resultaten en
herhaalbaarheid te voorkomen.
In hoodfstuk 2.2 werd de beta-agonist zilpaterol bestudeerd. Zilpaterol is een nieuwe beta-
agonist die ontwikkeld werd als groeibevorderaar voor runderen. Zilmax® is toegelaten als
voederadditief in Mexico en Zuid-Afrika. In dit hoofdstuk werd het excretieprofiel van
zilpaterol in urine en faeces bestudeerd na orale behandeling van een mannelijk kalf met een
therapeutische dosis aan Zilmax®. Zilpaterol kon gemakkelijk gedetecteerd worden in urine
en faeces. Het gehalte aan zilpaterol in urine was relatief hoog. Reeds na 2 dagen werd een
concentratie van 1000 µg l-1 overschreden. Een steady-state concentratie van 1200 µg l-1 werd
194 Printed: www.dclsigns.be
Samenvatting
snel bereikt. Ook in faeces werd een steady-state concentratie van 83 µg kg-1 snel bereikt ( de
eerste meting op dag 2 was reeds 71 µg kg-1). Er kan dus besloten worden dat zilpaterol
gemakkelijk gedetecteerd kan worden gedurende een behandeling met Zilmax®. Aangezien
het kalf werd geslacht na de laatste behandeling zijn er geen gegevens beschikbaar van de
eliminatie van zilpaterol na het stopzetten van de behandeling.
In hoofdstuk 3 wordt de LC-MSn methode-ontwikkeling behandeld van geregistreerde
diergeneesmiddelen. Dit hoofdstuk is verdeeld in twee delen: de detectie van producten met
een diergeneeskundige werking in spuitplaatsen en de ontwikkeling van specifieke LC-MSn
bevestigingsmethoden voor de diergeneesmiddelen, quinolones en niet-steroïdale anti-
inflammatoire geneesmiddelen.
Hoodstuk 3.1 beschrijft de ontwikkeling van een multi-residu LC-MSn methode voor de
identificatie en semi-kwantificatie van producten met een diergeneeskundige werking. In
2001 werd gestart met de staalname van spuitplaatsen genomen in het slachthuis voor de
identificatie van legaal en illegaal gebruikte producten met een diergeneeskundige werking.
Door deze stalen te analyseren kon een overzicht gemaakt worden van de producten die
tegenwoordig gebruikt worden in de praktijk en op basis van deze resultaten kan de aanpak
voor de screening van stalen aangepast worden. Aangezien het te duur en te tijdrovend
geworden is om elk staal te controleren aan de hand van specifieke methoden voor
verschillende groepen van diergeneesmiddelen, werd een alternatieve aanpak voorgesteld.
Een eenvoudige extractie en opzuivering werden gecombineerd met een multi-residu LC-MSn
identificatie en/of semi-kwantificatie. Door de hoge concentratie aan diergeneesmiddelen in
spuitplaatsen, is er geen nood aan kwantificatie in het concentratiegebied van de MRL. Een
alternatieve validatie wordt toegepast die de concentratie van het analyt in het staal vergelijkt
met een spike op MRL en 10 keer MRL concentratie. Deze toepassing toont de voordelen van
een generische LC-MSn methode als een snelle screenings- en bevestigingstechniek voor sterk
geconcentreerde stalen.
Gebaseerd op de resultaten van de analyses van spuitplaatsen en op vraag van het Federaal
Agentschap voor Voedselveiligheid werd een kwantitatieve bevestigingsmethode ontwikkeld
voor quinolones en NSAIDs.
Hoofdstuk 3.2 beschrijft de ontwikkeling van een LC-MS2 multi-residu methode voor de
gelijktijdige analyse van 8 quinolones in vlees en rundermelk. Er werd gebruik gemaakt van
een eenvoudige en snelle extractie- en opzuiveringsmethode voor de verschillende matrices en
ion trap massaspectrometrie werd gebruikt als identificatie en kwantificatie techniek. De
195 Printed: www.dclsigns.be
quinolones waren detecteerbaar tot op en onder de MRL concentratie. De multi-residu
methode voor de detectie van quinolones in vlees werd gevalideerd volgens de criteria
beschreven in Beschikking 2002/657/EG. Voor de matrix rundermelk is de validatie nog niet
volledig afgewerkt.
Hoofdstuk 3.3 handelt over de ontwikkeling van een LC-MS2 multi-residu methode voor de
identificatie van salicylzuur, fenylbutazone, flunixine, tolfenamzuur, meloxicam en
ketoprofen in rundervlees. De ion trap parameters ‘activatie q’ en ‘maximale ion injectie tijd’
moesten aangepast worden om een optimale detectie van salicylzuur en fenylbutazon,
respectievelijk te bekomen. Deze multi-residu methode is een kwantitatieve
bevestigingsmethode voor de NSAIDs flunixine, tolfenamzuur en meloxicam, en een
kwalitatieve methode voor de niet-toegelaten NSAIDs fenylbutazone en salicylzuur. De
methode werd gevalideerd volgens de criteria van Beschikking 2002/657/EG.
196 Printed: www.dclsigns.be
CURRICULUM VITAE
Nathalie Van Hoof werd op 9 december 1978 geboren te Duffel. Na het behalen van het
diploma hoger secundair onderwijs aan het Vita et Pax college te Schoten (Latijn-Wiskunde),
begon zij in 1996 met de studie Bio-ingenieur aan de Universiteit Antwerpen (kandidaat bio-
ingenieur) en vervolgens aan de Universiteit Gent en behaalde het diploma Bio-ingenieur in
de scheikunde in 2001.
Daarna trad zij in dienst als wetenschappelijk medewerker bij de vakgroep Veterinaire
Volksgezondheid en Voedselveiligheid, afdeling Chemische Analyse. Zij werkte op twee
FOD projecten, van 2002 tot eind 2003 op het project getiteld ‘Identificatie en kwantificatie
van residuen van probleemmoleculen in voedingswaren van dierlijke oorsprong’ en van 2004
tot op heden op het project getiteld ‘MSn flexibele methodeontwikkeling in de actuele
residuproblematiek’. Daarnaast heeft zij zich ingewerkt in de routine residucontroles die
worden uitgevoerd in het laboratorium.
In 2005 behaalde zij het getuigschrift van de doctoraatsopleiding in de diergeneeskundige
wetenschappen. Nathalie Van Hoof is auteur of mede-auteur van 18 publicaties in nationale
en internationale tijdschriften. Zij nam actief deel aan verschillende internationale congressen.
197 Printed: www.dclsigns.be
Wetenschappelijke publicaties
N. Van Hoof, K. De Wasch, S. Poelmans, D. Bruneel, S. Spruyt, H. Noppe, C. Janssen, D.
Courtheyn, H. De Brabander (2004), Norchlorotestosterone acetate: an alternative metabolism
study and GC-MS² analysis in kidney fat, urine and faeces, Chromatographia Supplement, 59,
85-93
N. Van Hoof, K. De Wasch, S. Poelmans, H. Noppe and H. De Brabander (2004), Multi-
residue liquid chromatography/tandem mass spectrometry method for the detection of non-
steroidal anti-inflammatory drugs in bovine muscle: optimisation of ion trap parameters,
Rapid Communications in Mass Spectrometry, 18, 2823-2829
N. Van Hoof, K. De Wasch, L. Okerman, W. Reybroeck, S. Poelmans, H. Noppe, H. De
Brabander (2005), Validation of a liquid chromatography-tandem mass spectrometric method
for the quantification of eight quinolones in bovine muscle, milk and aquacultured products,
Analytica Chimica Acta, 529, 265-272
N. Van Hoof, R. Schilt, E. Van der Vlis, P. Boshuis, M. Van Baak, A. Draaijer, K. De
Wasch, M. Van de Wiele, J. Van Hende, D. Courtheyn, H. De Brabander (2005), Detection of
zilpaterol (Zilmax®) in calf urine and faeces with liquid chromatography-tandem mass
spectrometry, Analytica Chimica Acta, 529, 189-197
N. Van Hoof, D. Courtheyn, JP. Antignac, M. Van de Wiele, S. Poelmans, H. Noppe, H. De
Brabander (2005), Multi-residue liquid chromatography/tandem mass spectrometric analysis
of beta-agonists in urine using molecular imprinted polymers, Rapid Communications in
Mass Spectrometry, 19, 2801-2808
N. Van Hoof, D. Courtheyn, W. Gillis, J. Van Hende, C. Van Peteghem, M. Van de Wiele, S.
Poelmans, H. Noppe, E. Cobbaert, P. Vanthemse, H.F. De Brabander (2005), Metabolism of
methenolone acetate in a veal calf, Accepted, Veterinary Research Communications
198 Printed: www.dclsigns.be
Curriculum Vitae
N. Van Hoof, K. De Wasch, M. De Moor, D. Bruneel, D. Courtheyn, S. Poelmans, H. Noppe
and H.F. De Brabander (2003), Unknown quinoxalines, one of the dangers of black market
products, Proceedings Euro Food Chem XII, Brugge, Belgium
N. Van Hoof, K. De Wasch, S. Poelmans and H.F. De Brabander (2003), Detecting
veterinary drug residues, 91-115, In: Rapid and on-line instrumentation for food quality
assurance, I.E. Tothill, Woodhead Publishing Limited, Cambridge, England
K. De Wasch, N. Van Hoof, S. Poelmans, L. Okerman, D. Courtheyn, A. Ermens, M.
Cornelis, H.F. De Brabander (2003), Identification of “unknown analytes” in injection sites: a
semi-quantitative interpretation, Analytica Chimica Acta, 483, 387-399
B. Le Bizec, N. Van Hoof, D. Courtheyn, I. Gaudin, M. Van De Wiele, E. Bichon, H. De
Brabander, F. André (2005), New anabolic steroid illegally used in cattle – structure
elucidation of 19-norchlorotestosterone acetatemetabolites in bovine urine, Accepted (July
2005) Journal of Steroid Biochemistry & Molecular Biology
K. De Wasch, S. Poelmans, T. Verslycke, C. Janssen, N. Van Hoof and H.F. De Brabander
(2002), Alternative to vertebrate animal experiments in the study of metabolism of illegal
growth promoters and veterinary drugs, Analytica Chimica Acta, 473, 59-69
H. De Brabander, S. Poelmans, R. Schilt, R.Stephany, B. LeBizec, R. Draisci, S. Sterk, L.
Van Ginkel, N. Van Hoof, A. Macri, K. De Wasch (2004), Presence and metabolism of the
anabolic steroid boldenone in various animal species: a review, Food Additives and
contaminants, 21(6), 515-525
S. Poelmans, K. De Wasch, D. Courtheyn, N. Van Hoof, H. Noppe, C. Janssen, H.F. De
Brabander (2005), The study of some new anabolic drugs by metabolism experiments with
Neomysis integer, Analytica Chimica Acta, 529, 311-316
H. Noppe, K. De Wasch, S. Poelmans, N. Van Hoof, T. Verslycke, C.R. Janssen, H.F. De
Brabander (2005), Development and validation of an analytical method for detection of
estrogens in water, Analytical Bioanalytical Chemistry, 382, 91-98
199 Printed: www.dclsigns.be
H. Noppe, S. Poelmans, K. Verheyden, H.F. De Brabander, N. Van Hoof (2005), Actuele
mogelijkheden en uitdagingen in de residuanalyse, Vlaams Diergeneeskundig Tijdschrift, 74,
340-346
S. Poelmans, K. De Wasch, H. Noppe, N. Van Hoof, S. Van Cruchten, B. Le Bizec, Y.
Deceuninck, S. Sterk, H.J. Van Rossum, M.K. Hoffman, H.F. De Brabander (2005), The
endogenous occurrence of some anabolic steroids in swine matrices, Food Additives and
Contaminants, 22, 808-815
S. Poelmans, K. De Wasch, H. Noppe, N. Van Hoof, M. Vandewiele, D. Courtheyn, W.
Gillis, P. Vanthemse, H.F. De Brabander (2005), Androstadienetrione, a boldenone-like
component, detected in cattle faeces with GC-MSn, Accepted (June 2005) Food Additives
and Contaminants
H. Noppe, K. Arijs, K. De Wasch, S. Van Cruchten, S. Poelmans, D. Courtheyn, E. Cobbaert,
W. Gillis, P. Vanthemse, H. De Brabander, C. Janssen, N. Van Hoof (2005), Biological and
chemical approaches for the detection and identification of illegal estrogens in water based
solutions, Accepted, Veterinary Research Communications
200 Printed: www.dclsigns.be