nathalie van hoof - universiteit gent

Nathalie Van Hoof

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.


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!


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


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


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


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


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


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.


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


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).


CHAPTER 1

ANALYTICAL INTRODUCTION AND

LEGISLATIVE ASPECTS


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


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



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



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



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



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



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



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


http://elchem.kaist.ac.kr/vt/chem-ed/ms/graphics/quad-sch.gif


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



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



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



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



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


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



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



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



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



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



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



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



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



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



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



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



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



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


[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



[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


CHAPTER 2

BETA-AGONISTS


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


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


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


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


http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TF4-4DGMS6P-2&_user=794998&_coverDate=01%2F24%2F2005&_alid=276096934&_rdoc=1&_fmt=full&_orig=search&_qd=1&_cdi=5216&_sort=d&_docanchor=&view=c&_acct=C000043466&_version=1&_urlVersion=0&_userid=794998&md5=715e75070e6fbe31a54c97f9b9cce188&artImgPref=F#bbib1

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


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


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


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


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



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


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



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


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



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


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



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


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



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: 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: 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







14,8114,41mabuterol


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


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








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







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



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


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



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


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


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


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


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


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



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


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



2.2.6. References




[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


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


http://www.miptechnologies.se/

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


http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TF4-4DGMS6P-2&_user=794998&_coverDate=01%2F24%2F2005&_alid=276096934&_rdoc=1&_fmt=full&_orig=search&_qd=1&_cdi=5216&_sort=d&_docanchor=&view=c&_acct=C000043466&_version=1&_urlVersion=0&_userid=794998&md5=715e75070e6fbe31a54c97f9b9cce188&artImgPref=F#fig1

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


http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TF4-4DGMS6P-2&_user=794998&_coverDate=01%2F24%2F2005&_alid=276096934&_rdoc=1&_fmt=full&_orig=search&_qd=1&_cdi=5216&_sort=d&_docanchor=&view=c&_acct=C000043466&_version=1&_urlVersion=0&_userid=794998&md5=715e75070e6fbe31a54c97f9b9cce188&artImgPref=F#bib5


2.3.3. Experimental


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.


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


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.


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


http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TF4-4DGMS6P-2&_user=794998&_coverDate=01%2F24%2F2005&_alid=276096934&_rdoc=1&_fmt=full&_orig=search&_qd=1&_cdi=5216&_sort=d&_docanchor=&view=c&_acct=C000043466&_version=1&_urlVersion=0&_userid=794998&md5=715e75070e6fbe31a54c97f9b9cce188&artImgPref=F#tbl1



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


70.0 – 290.0

Ritodrine 150, 121


75.0 – 305.0

Ractopamine 164, 121

Scan event 2: 307.0 → 289.0

75.0 – 310.0

Ractopamine-d5 167, 121


90.0 – 350.0

Formoterol 149, 121


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


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








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




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: 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: 7.16E5m/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




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



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





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


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







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


Beta-agonists

2.3.6. References

[1] C.S. Stachel, W. Radeck, P. Gowik (2003) Zilpaterol – a new focus of concern in residue


[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








[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



CHAPTER 3

VETERINARY DRUGS


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].


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].


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].


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.


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.


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


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


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


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].


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).


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


Veterinary drugs

3.1.3. References



in foodstuffs of animal origin (1990), Official Journal of the European Communities, no. L 67







and 86/469/EEC and Decision 89/187/EEC and 91/664/EEC (1996), Official Journal of the


[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


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


http://www.bcfi-vet.be/nlplan.lasso

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


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


Identification

[38] D.M. Boothe (2001) The analgesic, antipyretic, anti-inflammatory drugs, In: Veterinary


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


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


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


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


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].


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


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.


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.



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


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



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


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).



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).


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.



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


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


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


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.



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


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


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)


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


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]



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


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



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)



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


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



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).


http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TF4-473M965-5&_user=794998&_coverDate=04%2F25%2F2003&_alid=276098404&_rdoc=1&_fmt=full&_orig=search&_qd=1&_cdi=5216&_sort=d&_docanchor=&view=c&_acct=C000043466&_version=1&_urlVersion=0&_userid=794998&md5=09ef6c82f8ce3e34c1ddf15d4f3cf75e&artImgPref=F#fig4


Veterinary drugs


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).


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




®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


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].


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



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


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


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


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


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.



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).


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



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).


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



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.


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




[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


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


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


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].


http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TF4-4D75MHH-4&_user=794998&_coverDate=01%2F24%2F2005&_alid=276099367&_rdoc=1&_fmt=full&_orig=search&_qd=1&_cdi=5216&_sort=d&_docanchor=&view=c&_acct=C000043466&_version=1&_urlVersion=0&_userid=794998&md5=57deb991879ca636672cf2d7d0d6372e&artImgPref=F#bib1


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


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



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].


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


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



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



all food producing

species

300 μg kg-1 muscle

Marbofloxacin bovine 150 μg kg-1 muscle

75 μg kg-1 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


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.






Veterinary drugs


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




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


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




: 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:


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:


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:


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:


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



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



http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6TF4-4D75MHH-4&_user=794998&_coverDate=01%2F24%2F2005&_alid=276099367&_rdoc=1&_fmt=full&_orig=search&_qd=1&_cdi=5216&_sort=d&_docanchor=&view=c&_acct=C000043466&_version=1&_urlVersion=0&_userid=794998&md5=57deb991879ca636672cf2d7d0d6372e&artImgPref=F#fig3

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)



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.


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.



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


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


Overall CV (%) = 7.4 Difloxacin


Overall CV (%) = 9.0 Oxolinic acid


100 109 2.4 150 97 3.7

Overall CV (%) = 21.4 Flumequine


Overall CV (%) = 4.8


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.



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.


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 )


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 )


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 )



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.


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).


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.




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


100 97 7.1 150 100 6.2

Overall CV (%) = 6.0 Ciprofloxacin


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


Overall CV (%) = 11.6 Difloxacin


100 103 5.5

Overall CV (%) = 15.8 Oxolinic acid


100 101 8.7

Overall CV (%) = 8.8 Flumequine


100 99 10.2



Veterinary drugs


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


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






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.


Veterinary drugs

3.3.6. References

[1] M.G. Papich and J.E. Riviere (2001) Fluoroquinolone antimicrobial drugs, In Veterinary


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



[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



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


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



[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


Veterinary drugs




[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.


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.


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


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].


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



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


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


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.




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


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


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.



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.


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: 2,20E6m/z= 215,5-216,5 F: - c ESI Full ms2 260,00@46,00 [ 100,00-270,00] 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



050112s36 # 344-377 4,26-4,56 12 : 150 1,10-3,26, 5,10-8,63 NL: 2,09E6F:


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



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


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



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].


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].



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.


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.




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



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



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


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.


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


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.



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 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.


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.




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].



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.


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


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



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



[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


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



(2002), Official Journal of the European Communities, no. L 221


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.



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.


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


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


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.


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


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


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.


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



in foodstuffs of animal origin (1990), Official Journal of the European Communities, no. L 67



and 86/469/EEC and Decision 89/187/EEC and 91/664/EEC (1996), Official Journal of the




(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


[11] C.S. Stachel, W. Radeck, P. Gowik (2003) Zilpaterol – a new focus of concern in residue


[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



http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Shelver+WL%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Smith+DJ%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Kim+HJ%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Smith%2DSchalkwijk+MJ%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Uncini+Manganelli+RE%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Camangi+F%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Tomei+PE%22%5BAuthor%5D

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


http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Haas+KB%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Reichling+J%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Saller+R%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Hikino+H%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Nabetani+S%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Nomoto+K%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Arai+T%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Takemoto+T%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Bathori+M%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Slama+K%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Koudela+K%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Tenora+J%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Mathova+A%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Tsitsimpikou+C%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Tsamis+GD%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Siskos+PA%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Spyridaki+MH%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Georgakopoulos+CG%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Capuco+AV%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Dahl+GE%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Wood+DL%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Moallem+U%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Erdman+RE%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Dohoo+IR%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Leslie+K%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22DesCoteaux+L%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Fredeen+A%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Dowling+P%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Preston+A%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Shewfelt+W%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Baldi+A%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Etherton+TD%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Bauman+DE%22%5BAuthor%5D

[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


http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Pinel+G%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Andre+F%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Le+Bizec+B%22%5BAuthor%5D

http://www.bodybuilding.com/

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Thurman+EM%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Ferrer+I%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Fernandez%2DAlba+AR%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Ferrer+I%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Garcia%2DReyes+JF%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Mezcua+M%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Thurman+EM%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Fernanndez%2DAlba+AR%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Nielen+MW%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22van+Bennekom+EO%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Heskamp+HH%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22van+Rhijn+JH%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Bovee+TF%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Hoogenboom+LR%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Stolker+AL%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Niesing+W%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Fuchs+R%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Vreeken+RJ%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Niessen+WM%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&term=%22Brinkman+UA%22%5BAuthor%5D

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


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


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


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


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


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.


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.


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


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


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


nathalie van hoof - universiteit gent

Documents