Steroids 67 (2002) 873–882

Study of natural and artificial corticosteroid phase IImetabolites in bovine urine using HPLC–MS/MS

Jean-Philippe Antignac∗, Bruno Le Bizec, Fabrice Monteau, François AndréLaboratoire d’Etude des Résidus et Contaminants dans les Aliments (LABERCA), Ecole Nationale Vétérinaire de Nantes,

BP 50707, 44307 Nantes Cedex 3, France

Received 30 January 2002; received in revised form 28 May 2002; accepted 5 June 2002

Abstract

Corticosteroid compounds are widely used therapeutically for their anti-inflammatory properties and sometimes as growth promotersin food producing animals. In the field of drug residue analysis, knowledge of the main metabolic pathways of target analytes improvesthe efficiency of the corresponding control. Thus, phase II metabolism of corticosteroids, for which very little literature is available,was investigated in cattle. An LC–MS/MS detection method was developed for five commercially available conjugated corticosteroids,permitting direct monitoring during the development of their separation on anion exchange SPE. This separation method is further appli-cable to other potential urinary conjugated corticosteroids. Because our purpose was not to identify all the existing corticosteroid phase IImetabolites, but to obtain their total relative proportions, enzymatic hydrolysis was optimized and performed on each separated fraction(glucuronides and sulfates). Finally, the phase II metabolic profiles of natural and artificial corticosteroids in bovine urine were studied andcompared. LC–MS/MS detection with negative electrospray ionization appeared efficient for both glucuronide and sulfate conjugated corti-costeroids, and quaternary ammonium stationary phase permitted their effective separation. The experimental design used for optimizationof the enzymatic hydrolysis with a purifiedHelix pomatiapreparation demonstrated optimal values for pH 5.2, temperature of 50◦Cand incubation duration of 4 h. Results on bovine urine samples collected on two animals before and after dexamethasone administrationshowed important differences regarding the proportion of total conjugated forms between endogenous cortisol, endogenous tetrahydrocor-tisol, and exogenous dexamethasone. This proportion appeared significantly higher for tetrahydrocortisol (40–65%) than cortisol (2–8%)or dexamethasone (4–27%). This innovative methodology demonstrates the suitability of anion exchange SPE and LC–MS/MS for thestudy of steroid hormones phase II metabolism, and appears promising to investigate metabolic profile differences linked to the hormoneadministration mode or origin, with direct application in the field of doping controls.© 2002 Elsevier Science Inc. All rights reserved.

Keywords:Corticosteroids; Metabolism; Urine; Glucuronide; Sulfate; Electrospray; LC–MS/MS

1. Introduction

1.1. Corticosteroids, growth promoters, and metabolism

Natural corticosteroids are hormones secreted by theadrenal cortex. The discovery of their anti-inflammatoryproperties has led to the chemical synthesis of more activeartificial corticosteroids that are used as therapeutic drugs.In veterinary medicine, this legal utilization is strictly reg-ulated, with withdrawal periods between treatment andslaughtering and maximal residue levels for some com-pounds. On the other hand, and especially at low concen-tration, glucocorticosteroids are known to increase weight

∗ Corresponding author. Tel.:+33-2-40-68-77-66;fax: +33-2-40-68-78-78.

E-mail address:laberca@vet-nantes.fr (J.-P. Antignac).

gain, to reduce the feed conversion ratio, and to have asynergetic effect with other molecules like�-agonists oranabolic steroids[1–3]. Thus, these compounds have beenused as growth promoters in cattle, whereas these prac-tices are forbidden in Europe. Analytical methods havebeen developed to control corticosteroid residues in edibletissues or urine samples using GC–MS, or more recently,LC–MS/MS [4–7], and all are based on detection of theparent drug. However, in the field of drug residue analysisand for other fundamental or practical purposes, it is nec-essary to know the main metabolic pathways of the targetanalyte. Indeed, this knowledge can improve the controlefficiency by focusing on specific and/or abundant phase Ias well as phase II metabolites. In this context, we decidedto specifically study the nature and proportions of differentconjugated corticosteroid phase II metabolites in bovineurine.

0039-128X/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.PII: S0039-128X(02)00048-X

874 J.-P. Antignac et al. / Steroids 67 (2002) 873–882

1.2. Present analytical state

The phase II metabolism of drugs consists of conjugationreactions between parent molecule (or phase I metabolites)and sulfuric acid, glucuronic acid, or other highly polargroups, such as glycosides, phosphates, etc. The result ofthese reactions is the synthesis of hydrophilic compoundsfor faster elimination. In the case of corticosteroids, veryfew studies have been devoted to the nature and relativeproportions of the different conjugated metabolites excretedin urine or feces. For endogenous cortisol in humans meanvalues of 95% for glucuronide conjugates, 4% for sulfateconjugates, and 1% for unconjugated forms are usuallyreported[8,9]. Nevertheless, it was thought that the pres-ence of halogenous atoms on artificial compounds modifiedthis repartition in favor of biologically active unconjugatedforms. Moreover, no information is available on the differ-ent potential positions for conjugation (C3, C11, C17), andon the existence of multiple conjugated compounds. Froman analytical point of view, several authors proposed detec-tion methods for steroid conjugates using GC–MS[10,11]or LC–MS [12,13], but no study was specifically devotedto corticosteroid conjugates. The few papers describing aseparation method for glucuronide and sulfate forms us-ing anion exchange chromatography on Sephadex[14] orquaternary ammonium[15] stationary phases only focusedon steroid conjugates. For the deconjugation of phase IImetabolites, enzymatic hydrolysis is the technique morecurrently used, andHelix pomatiajuice, which is character-ized by glucuronidase and arylsulfatase activities, remainsthe preparation of choice. This enzyme preparation wasused for corticosteroids, but widely different experimentalconditions were described by different authors[16–18].

1.3. Present work objectives

The general purpose of this study was to investigate thephase II metabolism of corticosteroids, mainly in order toimprove the efficiency and the specificity of their control incattle. An HPLC–MS/MS detection method was developedfor five commercially available glucuronide and sulfate cor-ticosteroid conjugates. This direct monitoring was used todevelop an efficient separation method for these compounds

Table 1Nature and MS/MS detection parameters of investigated free and conjugated corticosteroids

Analyte MRM transition Cone voltage (V) Collision energy (V)

Fluorometholone (external standard) 435→ 355 20 20Dexamethasone 451→ 361 20 25Cortisol 421→ 331 20 25Cortisone 419→ 329 15 20Cortisol 21-glucuronide 537→ 537 80 30Cortisol 21-sulfate 441→ 441 80 30Cortisone 21-sulfate 439→ 439 80 30Tetrahydrocortisone 3-glucuronide 539→ 539 80 30Deoxycorticosterone 21-glucoside 551→ 551 30 5

based on anion exchange SPE that was further supposed tobe applicable for other potential urinary conjugated corti-costeroids. Because our purpose was not to identify all theexisting corticosteroid phase II metabolites, but to obtaintheir total relative proportions, enzymatic hydrolysis wasoptimized using the experimental design methodology andperformed on each separated fraction (glucuronides andsulfates). Finally, this analytical development was appliedto determine and compare the phase II metabolic profile ofnatural and artificial corticosteroids in bovine urine.

2. Experimental

2.1. Reagents and chemicals

Analytical grade methanol, cyclohexane, glacial aceticacid, and sulfuric acid 2 M were provided by solventsdocumentation synthesis (SDS; Peypin, France). Sodiumacetate and sodium carbonate were purchased by Merck(Darmstadt, Germany). Formic acid, triethylamine, as wellas standard reference corticosteroids and conjugated cor-ticosteroids (Table 1) were from Sigma (St. Louis, MO).Standard solutions were prepared at 1 mg/ml in methanoland stored at−20◦C in darkness.

2.2. Liquid chromatography

An Alliance® 2690 HPLC pump with an automatic in-jector was used (Waters®, Milford, MA). Reversed phaseliquid chromatography was performed on octadecyl graftedsilica Nucleosil® C18AB (50 mm×2 mm, 5 mm) stationaryphase (Macherey-Nagel®, Düren, Germany) with a guardcolumn (Nucleosil® C18AB, 10 mm×2 mm, 5 mm). Elutionsolvents were methanol (A) and 0.5% (v/v) acetic acid inwater (B). Mobile phase composition (A:B, v/v) was 40:60at 0 min, 90:10 at 10 min, and 40:60 from 20 to 30 min. Theflow rate was 220 ml/min, and 10 ml were injected.

2.3. Mass spectrometry

Data were acquired using a QuattroLC® triple quadrupoleanalyzer (Micromass®, Manchester, UK) operating in the

J.-P. Antignac et al. / Steroids 67 (2002) 873–882 875

negative electrospray ionization mode. Nitrogen was usedas the nebulization and desolvation gas at flow rates of 90and 600 l/h, respectively. Source and desolvation tempera-tures were 120 and 350◦C, respectively. The potential ap-plied on the capillary was 3.5 kV, and the sampling conevoltage was optimized for each molecule. MS/MS experi-ments were performed in the multiple reaction monitoring(MRM) mode using argon as the collision gas at a pressureof 4.0× 10−4 mbar and a collision energy adapted for eachcompound (Table 1).

2.4. LC–MS/MS identification of conjugatedcorticosteroids

Each of the conjugated corticosteroid standard solu-tions (100 ng/ml in methanol/0.5% acetic acid in water,50:50, v/v) was directly introduced in the electrospraysource (10 ml/min). The main ions produced in MS andMS/MS were identified in both positive and negative ion-ization mode. The more interesting diagnostic ions werethen selected, and all mass spectrometer parameters wereoptimized to improve their intensity. Then, the MRM ac-quisition method was written, and the previously describedHPLC separation conditions were tested on a 0.1 ng/mlstandard mixture.

Fig. 1. (ESI+, MS) spectra of cortisol 21-glucuronide (a) and cortisol 21-sulfate (b).

2.5. Separation and purification of free, glucuronide,and sulfate corticosteroids

Urine samples (10 ml) were buffered with 2 ml of 2 Macetate buffer (pH 5.2) and applied onto C18 SPE car-tridges (2 g solid phase, SDS) previously activated with 5 mlmethanol and 5 ml water. After washing with 5 ml water and5 ml cyclohexane, free and conjugated forms were elutedwith 5 ml methanol. The organic phase was then appliedonto quaternary ammonium SPE cartridges (1 g solid phase,SDS) previously activated with 5 ml methanol, 5 ml water,20 ml 0.5 M acetic acid in water (to replace initial chlorinecounter-ions by acetate groups), 20 ml water (to eliminateresidual acetic acid), and 5 ml methanol. As only the glu-curonic and sulfate forms were retained, the deposed samplewas collected, and additional 5 ml methanol were appliedto elute residual non-retained free and glycoside forms. Ina second step, 10 ml formic acid (0.5 M) in methanol wereused to specifically elute the glucuronide forms. In a thirdstep, 10 ml triethylamine sulfate (0.5 M) in water (extempo-raneously prepared solution by titration of triethylamine bysulfuric acid to pH 5) were used to elute the sulfate forms.This sulfate fraction was applied onto a C18 SPE cartridgein order to wash the viscous and non-volatile triethylaminesulfate with 10 ml water before eluting the sulfate forms

876 J.-P. Antignac et al. / Steroids 67 (2002) 873–882

with 10 ml methanol. Finally, all fractions were evaporatedto dryness.

2.6. Enzymatic hydrolysis of conjugated corticosteroids

Assuming that the previously described process permit-ted the separation of all potential urinary free+ glycoside,glucuronide, and sulfate corticosteroids, our purpose wasnot to identify these compounds exhaustively, but to ob-tain their relative proportions. Thus, enzymatic hydrolysiswas performed on both the glucuronide and sulfate fractionsafter their separation in order to detect the correspondingtotal deconjugated form. Enzymatic hydrolysis conditionswere optimized using the experimental design statisticalmethod. Fixed parameters were the nature of the enzymaticpreparation (Helix pomatia, Sigma) and the added quantity(10,000 units, i.e. 400�l at 25 units/�l). The optimized pa-rameters were the pH, the temperature (T) and duration ofincubation (D). The sample used for these experiments wasan incurred bovine urine sample containing dexamethasone.For each assay, urine samples (10 ml) were transferred intoa 50 ml tube, and 2 ml acetate buffer (2 M, pH adjusted toX)as well as the enzyme were added. The incubation was thencarried out atY◦C for Zh. The studied domain (i.e. the dif-ferent values forX, Y, andZ) are presented inTable 2. After

Fig. 2. (ESI−, MS) spectra of cortisol 21-glucuronide (a) and cortisol 21-sulfate (b).

Table 2Experimental design parameters and studied domain for enzymatic hy-drolysis optimization

Factor * −1 0 +1 *

X (pH) 3 4 5 6 7Y (incubation temperature in◦C) 30 40 50 60 70Z (incubation duration in h) 0.5 2 4 6 16

*: ‘Star points’ levels;−1/+1: low/high levels; 0: central level.

centrifugation (2000× g for 10 min), the supernatant wascollected and purified using C18 SPE, alkaline liquid/liquidclean up with 10% Na2CO3 and SiOH SPE, according to apreviously published method[19]. Finally, purified extractswere reconstituted in 50 ml of a 1 ng/ml fluorometholone(external standard) solution in mobile phase (methanol/0.5%acetic acid in water, 40:60, v/v).

3. Results and discussion

3.1. LC–MS/MS identification of conjugated corticosteroids

In positive electrospray ionization (ESI+), the base peakobserved for the tested glucuronide and glycoside forms

J.-P. Antignac et al. / Steroids 67 (2002) 873–882 877

corresponded to a sodium adduct, [M + Na]+, or [M −H + 2Na]+. For the tested sulfate forms (potassium salts),the base peak in ESI+ corresponded to the ion [M − K +2Na]+; the ions [M + Na]+ and [M + K]+ being alsopresent (Fig. 1). For all tested conjugated corticosteroids,the different sodium adducts appeared very stable; thus, noMS/MS fragmentation were observed even for high colli-sion energy or gas pressure. In negative electrospray ioniza-tion (ESI−), the base peak observed in MS corresponded tothe pseudo-molecular ion [M − H]− for glucuronide formsand to the ion [M − K]− for sulfate forms (Fig. 2). Likein ESI+, these ions did not fragment in the collision cell.For deoxycorticosterone 21-glucoside, the same scheme al-ready observed for free corticosteroids was retrieved[20]with an abundant [M+ − acetate]− adduct ion leading tothe pseudo-molecular ion [M − H]− after collision. In con-clusion, the negative electrospray ionization mode was pre-ferred because of its efficiency for all free and conjugatedforms. Because of the high stability of the base peak ob-served for glucuronide and sulfate forms, only one MRMtransition was recorded using this unique ion both as pre-cursor and fragment. Nevertheless, a good specificity wasachieved because of the atypically applied high cone poten-tial (80 V) and collision energy (30 V) that permitted frag-mentation of interferences and a decrease of the noise.Fig. 3

Fig. 3. MRM ion chromatograms of the standard five investigated corticosteroid conjugates (1 ng injected).

shows the ion chromatograms obtained for 1 ng injectedof each corticosteroid conjugate in standard solution. Theseparation was found satisfying and demonstrated the appli-cability of the same elution gradient developed for uncon-jugated forms.

3.2. Separation and purification of free, glucuronide,and sulfate corticosteroids

When applied in methanol on the anion exchange SPEcartridge, the tested free and glycoside forms directly passedthrough the column, while glucuronide and sulfate formswere retained. For the specific elution of glucuronides,methanolic solutions of acetic, sulfuric, or formic acid(0.1–10%, v/v) were tested. Only a minimum of 10 ml of aformic acid solution (0.5 M) provided complete and selec-tive elution of glucuronide conjugates. Then, the stronglyretained sulfate conjugates were only efficiently eluted with10 ml of a triethylamine sulfate (TEAS; 0.5 M) solution.Fig. 4 shows the relative eluotropic strengths of methanol,formic acid, and TEAS on the anion exchange SPE columnregarding the retention of cortisol, cortisol 21-glucuronide,and cortisol 21-sulfate. Thus, the successive application ofthese solutions permitted the sequential elution of thesecompounds.

878J.-P.

An

tign

ac

et

al./S

tero

ids

67

(20

02

)8

73

–8

82

J.-P. Antignac et al. / Steroids 67 (2002) 873–882 879

3.3. Enzymatic hydrolysis of conjugated corticosteroids

Statistical analysis of the realized experimental designshowed that only the quadratic effect of pH appeared to bestatistically significant (3.00 > 2.45;P < 0.05), but the twofollowing effects also corresponded to the quadratic effectsof the temperature and duration of incubation. The corre-sponding response surfaces are presented inFig. 5. Theseresults showed a clear optimum for pH 5.2 and incubationtemperature 50◦C. As far as the incubation duration wasconcerned, a large efficiency range appeared between 3 and16 h, and the choice was made for a value of 4 h to permit thecomplete analysis of a 12-sample batch per day. These op-timal enzymatic hydrolysis conditions were very similar tothose determined by another focusing study on steroids[21].A last experiment was performed in order to confirm the ef-ficiency of these experimental conditions for corticosteroidsother than dexamethasone. The commercially available con-jugated corticosteroids were only natural compounds. Thus,we decided to use an incurred urine sample in which the

Fig. 5. Response surfaces obtained after statistical analysis of the experimental design assay.

production of endogenous cortisol and cortisone was com-pletely inhibited after administration of the artificial corti-costeroid dexamethasone. This sample was then spiked with50 ng of cortisol 21-glucuronide and cortisone 21-sulfate andanalyzed after deconjugation.Fig. 6 presents the resultingion chromatograms of cortisol and cortisone, showing thehydrolysis efficiency for both glucuronide and sulfate forms.

3.4. Comparison of natural and artificial corticosteroidphase II metabolism

The developed methodology was applied to determinethe relative proportions of free, glucuronide, and sulfatecorticosteroids in bovine urine. This study was realizedon two animals (cows, “Normande” specie, 400–500 kg),receiving 40 mg of dexamethasone acetate (Animal 1) ordexamethasone phosphate (Animal 2) by intra-muscularinjection. Dexamethasone (1 day after treatment) as wellas endogenous cortisol and tetrahydrocortisol (before treat-ment) were analyzed. The results (Fig. 7) first indicated

880J.-P.

An

tign

ac

et

al./S

tero

ids

67

(20

02

)8

73

–8

82

J.-P. Antignac et al. / Steroids 67 (2002) 873–882 881

Fig. 7. Relative proportions (%) of free, glucuronide, and sulfate forms for endogenous cortisol, tetrahydrocortisol, and exogenous dexamethasone inbovine urine samples of two animals treated with 40 mg of dexamethasone acetate (Animal 1) and dexamethasone phosphate (Animal 2).

a good correlation between the two animals concerningthe metabolic profiles of the target analytes (proportionsof free, glucuronide, and sulfate forms), especially forendogenous corticosteroids. Secondly, a strong differenceappeared regarding the total proportion of conjugated formsbetween endogenous tetrahydrocortisol (40–65%) and en-dogenous cortisol (2–8%) or exogenous dexamethasone(4–27%). For dexamethasone, this observation could beexplained by the exogenous administration mode that canlead to a distinct metabolic pathway compared to the cor-ticosurrenal endogenous production. Another hypothesiscould be the influence of some chemical structural differ-ences of artificial corticosteroids, particularly the presenceof halogenous atoms, which can reduce or inhibit somemetabolic enzymatic reactions. For tetrahydrocortisol, theA ring reduction leading to an hydroxyl group in position 3probably favors the conjugation reactions, and could explainthe high amount of glucuronide forms observed for thiscompound.

4. Conclusion

This work first demonstrated the suitability of LC–MS/MSfor the direct detection of free and phase II corticosteroid

metabolites through the negative electrospray ionizationmode. Secondly, this direct analysis permitted the devel-opment of an efficient separation method for free, glu-curonide, and sulfate forms. Then, enzymatic hydrolysisof conjugated forms was optimized using an experimentaldesign method and taking into account the matrix effectsand different conjugated corticosteroids. The proposed an-alytical methodology appeared innovative and promising,even when there were some limits, such as the lack ofcommercially available conjugated corticosteroids reducingthe testing and validating possibilities. Finally, the presentwork demonstrated the usefulness of anion exchange SPEand LC–MS/MS for the study of steroid hormones phase IImetabolism, and appears promising to investigate metabolicprofile differences linked to the hormone administrationmode or origin, with direct application in the fields of clin-ical biochemistry or doping controls. In the near future, weplan to use this methodology to discriminate endogenousforms from exogenous residues.

References

[1] Istasse L, De Haan V, Van Eenaeme C, Buts B, Baldwin P, Gielen M,et al. Effect of dexamethasone injections on performances in a pair ofmonozygotic cattle twins. J Anim Physiol Anim Nutr 1989;62:150–8.

882 J.-P. Antignac et al. / Steroids 67 (2002) 873–882

[2] Rijckaert MLJ, Vlemmix HPJ. The growth promoting effect ofglucocorticosteroids. Department of Chemical Engineering,Eindhoven University of Technology, 1992.

[3] Duchatel JP, Beduin JM, Jauniaux T, Coignoul F, Vindevogel H.Premières observations sur l’utilisation des glucocorticoıdes commeagent dopant chez le pigeon voyageur. Ann Méd Vét 1993;137:557–64.

[4] Courtheyn D, Vercammen J, De Brabander H, Vandenreyt I, BatjoensP, Vanoosthuyze K, et al. Determination of dexamethasone in urineand faeces of treated cattle with negative chemical ionization-massspectrometry. Analyst 1994;119:2557–64.

[5] Courtheyn D, Vercammen J, Logghe M, Seghers H, De Wash K,De Brabander H. Determination of betamethasone and triamcinoloneacetonide by GC–MS–NCI in excreta of treated animals anddevelopment of a fast oxidation procedure for derivatization ofcorticosteroids. Analyst 1998;123(12):2409–14.

[6] Negrioli J. Etude analytique des corticosteroıdes utilisés dansl’espèce bovine—Nouvelle dérivation utilisant la réaction avecle N,N-dimethylformamide diméthylacétal. Thèse de doctorat del’Université de Nantes. p. 283.

[7] De Wasch K, De Brabander H, Courteyn D, Van Peteghem C.Detection of corticosteroids in injection sites and cocktails by MSn.Analyst 1998;123:2415–22.

[8] Passingham BJ, Barton RN. Application of high-performance liquidchromatography to the measurement of cortisol secretion rate. JChromatogr 1987;416:25–35.

[9] Kasuya Y, Shibasaki H, Furuta T. The use of deuterium-labeledcortisol for in vivo evaluation of renal 11�-HSD activity in man:urinary excretion of cortisol, cortisone and their A-ring reducedmetabolites. Steroids 2000;65:89–97.

[10] Choi MH, Kim KR, Chung BC. Simultaneous determination ofurinary androgen glucuronides by high temperature gas chromato-graphy-mass spectrometry with selected ion monitoring. Steroids2000;65:54–9.

[11] Spielgelhalder B, Röhle G, Siekmann L, Breuer H. Mass-spectrometry of steroid glucuronides. J Steroid Biochem 1976;7:749–56.

[12] Bean KA, Henion JD. Direct determination of anabolic steroidconjugates in human urine by combined high-performance liquid

chromatography and tandem mass spectrometry. J Chromatogr B1997;690:65–75.

[13] Kuuranne T, Vahermo M, Leinonen A, Kostiainen R. Electrospray andatmospheric pressure chemical ionization tandem mass spectrometricbehavior of eight anabolic steroid glucuronides. J Am Soc MassSpectrom 2000;11:722–30.

[14] Jänne O, Vihko R, Sjövall J, Sjövall K. Determination of steroidmono- and disulfates in human plasma. Clin Chim Acta 1969;23:405–12.

[15] Vanluchene E, Vandekerckove D. Group fractionation of free andconjugated steroids by means of disposable silica-based anion-exchange columns. J Chromatogr 1988;456:175–82.

[16] Calvarese S, Rubini P, Urbani G, Ferri N, Ramazza V,Zucchi M. Experimental administration of 19-nortestosterone anddexamethasone in cattle: elimination of the two drugs in differentbiological matrices. Analyst 1994;119:2611–5.

[17] Rizea Savu S, Silvestro L, Haag A, Sörgel F. A confirmatoryHPLC–MS/MS method for ten synthetic corticosteroids in bovineurines. J Mass Spectrom 1996;31:1351–63.

[18] Yergey AL, Teffera Y, Esteban NV, Abramson FP. Directdetermination of human urinary cortisol metabolites by HPLC/CRIMS. Steroids 1995;60:295–8.

[19] Antignac JP, Le Bizec B, Monteau F, Poulain F, André F.Multi-residue extraction–purification procedure for corticosteroids inbiological samples for efficient control of their misuse in livestockproduction. J Chromatogr B 2001;757:11–9;Antignac JP, Le Bizec B, Monteau F, Poulain F, André F.Multi-residue extraction–purification procedure for corticosteroids inbiological samples for efficient control of their misuse in livestockproduction. Rapid Commun Mass Spectrom 1999;14:33–9.

[20] Antignac JP, Le Bizec B, Monteau F, Poulain F, André F.Collision-induced dissociation of corticosteroids in electrospraytandem mass spectrometry and development of a screening method byhigh performance liquid chromatography/tandem mass spectrometry.Rapid Commun Mass Spectrom 1999;14:33–9.

[21] Le Bizec B. Utilisation de la spectrométrie de masse pour le controlede l’usage des stéroıdes anabolisants en élevage. Application à l’étudedu métabolisme de la 4-chlorotestostérone acétate chez le bovin.Thèse de doctorat de chimie, Université de Nantes, 1996. p. 346.