quantitation of 17 -nandrolone metabolites in boar and...

A

fwwadu5Eeqissf©

K

1

hpitmc

a

iT

0d

Analytica Chimica Acta 586 (2007) 184–195

Quantitation of 17�-nandrolone metabolites in boar and horse urine by gaschromatography–mass spectrometry

Meritxell Roig a,b, Jordi Segura a,b, Rosa Ventura a,b,∗a Unitat de Recerca en Farmacologia, Institut Municipal d’Investigacuo Medica (IMIM), Dr. Aiguader, 80, 08003 Barcelona, Spain

b Departament de Ciencies Experimentals i de la Salut, Universitat Pompeu Fabra, UPF, Barcelona, Spain

Received 19 June 2006; received in revised form 28 July 2006; accepted 3 August 2006Available online 24 August 2006

bstract

A method to quantify metabolites of 17�-nandrolone (17�N) in boar and horse urine has been optimized and validated. Metabolites excreted inree form were extracted at pH 9.5 with tert-butylmethylether. The aqueous phases were applied to Sep Pak C18 cartridges and conjugated steroidsere eluted with methanol. After evaporation to dryness, either enzymatic hydrolysis with �-glucuronidase from Escherichia coli or solvolysisith a mixture of ethylacetate:methanol:concentrated sulphuric acid were applied to the extract. Deconjugated steroids were then extracted at

lkaline pH with tert-butylmethylether. The dried organic extracts were derivatized with MSTFA:NH4I:2-mercaptoethanol to obtain the TMSerivatives, and were subjected to analysis by gas chromatography mass spectrometry (GC/MS). The procedure was validated in boar and horserine for the following metabolites: norandrosterone, noretiocholanolone, norepiandrosterone, 5�-estran-3�, 17�-diol, 5�-estran-3�, 17�-diol,�-estran-3�, 17�-diol, 17�-nandrolone, 17�N, 5(10)-estrene-3�, 17�-diol, 17�-estradiol and 17�-estradiol in the different metabolic fractions.xtraction recoveries were higher than 90% for all analytes in the free fraction, and better than 80% in the glucuronide and sulphate fractions,xcept for 17�-estradiol in the glucuronide fraction (74%), and 5�-estran-3�, 17�-diol and 17�N in the sulphate fraction (close to 70%). Limits ofuantitation ranged from 0.05 to 2.1 ng mL−1 in the free fraction, from 0.3 to 1.7 ng mL−1 in the glucuronide fraction, and from 0.2 to 2.6 ng mL−1

n the sulphate fraction. Intra- and inter-assay values for precision, measured as relative standard deviation, and accuracy, measured as relativetandard error, were below 15% for most of the analytes and below 25%, for the rest of analytes. The method was applied to the analysis of urineamples collected after administration of 17�N laureate to boars and horses, and its suitability for the quantitation of the metabolites in the threeractions has been demonstrated.

2006 Elsevier B.V. All rights reserved.

phy/m

at(iaibc

eywords: Boar; Horse; Urine; 17�-Nandrolone metabolites; Gas chromatogra

. Introduction

The use of anabolic agents is prohibited in sports and inorseracing, as well as in food producing animals in the Euro-ean Union (96/22/EC Directive) [1]. 17�-Nandrolone (17�N)s one of the most common anabolic steroids used. Detection ofhe illegal use of 17�N is difficult as a result of the extensive

etabolism and the possibility of interference with endogenous
ompounds, which are species dependent.
The metabolism of 17�N has been studied in humans [2]nd in different animal species [3–11]. Norandrosterone (NorA)

∗ Corresponding author at: Unitat de Recerca en Farmacologia, Institut Munic-pal d’Investigacuo Medica (IMIM), Dr. Aiguader, 80, 08003 Barcelona, Spain.el.: +34 93 2211009; fax: +34 93 2217196.

E-mail address: [email protected] (R. Ventura).

[aom5pth

003-2670/$ – see front matter © 2006 Elsevier B.V. All rights reserved.oi:10.1016/j.aca.2006.08.033

ass spectrometry

nd noretiocholanolone (NorE) excreted in the glucuronide frac-ion are the main metabolites in humans; norepiandrosteroneNorEpiA) is also excreted in the sulphate fraction [2]. In horses,somers of estranediol (mainly 5�-estrane-3�,17�-diol, aba,nd 5�-estrane-3�,17�-diol, abb) and the metabolite result-ng from epimerization in C17, 17�-nandrolone (17�N), haveeen detected as main metabolites, 17� isomers mainly in glu-uronide fraction and 17� isomers, mainly in sulphate fraction3–6]. In bovines, 17�N, aba and NorE have also been detecteds main metabolites in cows and calves [7–9]. 17�N, isomersf 3-hydroxyestran-17-one (NorA, NorE, NorEpiA) and iso-ers of estrane-3, 17-diol (5�-estrane-3�,17�-diol, bab, and

�-estrane-3�,17�-diol, abb), have been identified in miniatureigs after administration of 17�N laureate [10,11]. For most ofhe animal species, quantitative data on 17�N and its metabolitesas never been reported.

mailto:[email protected]

dx.doi.org/10.1016/j.aca.2006.08.033

imica

slor

bp(Teguctdnawnamood

aitf

2

2

bhha(G(wmwtWaS

uOw(b

2

nntR(5wUESsw

ef1ow

2

a

2

hms(bco5ad

2

taopwmho

M. Roig et al. / Analytica Ch

In recent years, different authors reported data to demon-trate the endogenous presence of 17�N or some of its metabo-ites in different species [10–20]. However, comprehensive datan the endogenous presence of all 17�N metabolites was noteported.

Detection of the consumption of anabolic agents is performedy identification of specific metabolites using gas chromatogra-hy or liquid chromatography coupled to mass spectrometryGC/MS, LC/MS) or to tandem mass spectrometry [21–23].hese techniques are effective to detect the administration ofxogenous compounds; however they are not able to distin-uish the exogenous origin of compounds normally present inrines. In order to detect consumption on the basis on con-entrations detected in urine and to define the time windowo obtain the maximum retrospectivity of consumption and toiscard endogenous production, a suitable target metaboliteeeds to be selected. A deep knowledge on the metabolismnd the elimination kinetics of the different metabolites asell as their concentrations endogenously present in urine iseeded in order to select the suitable target metabolite to benalyzed by GC/MS or LC/MS. For this reason, quantitativeethods able to quantify 17�N and the metabolites in samples

btained after administration of 17�N to different animal speciesr in samples obtained from non-treated animals need to beeveloped.

The aim of the present study was to optimize and to validateprocedure for the quantification of 17�N and its metabolites

n different animal species to help in the selection of the bestarget analyte to detect the administration of 17�N esters forraudulent purposes.

. Experimental

.1. Reagents and solvents

Methanol and ethyl acetate (both of HPLC grade), tert-uthylmethyleter, 25% ammonia, ammonium chloride, sodiumydroxide pellets, di-sodium hydrogen phosphate, sodiumydrogen phosphate, potassium carbonate, ammonium iodidend 2-mercaptoethanol (all analytical grade), and sulphuric acidextra pure grade) were purchased from Merck (Darmstadt,ermany). n-Hexane (HPLC grade) was supplied by Scharlau

Barcelona, Spain). �-Glucuronidase from Escherichia coli K12as supplied by Roche Diagnostics GmbH (Mannheim, Ger-any). N-methyl-N-trimethylsilyl-trifluoroacetamide (MSTFA)as purchased by Macherey-Nagel (Duren, Germany). Car-

ridges Sep Pak C18 Vac RC (500 mg) were supplied byaters (Milford, MA, USA). Milli Q water was obtained byMilli-Q purification system (Millipore Iberica, Barcelona,

pain).
The solid-phase extraction (SPE) was performed on a Vac-
um manifold (Biochem Diagnostics, Edgewoodm, NY, USA).rganic layers were evaporated to dryness under nitrogen streamith a Turbo-Vap LV evaporator from Zymark Corporation

Hopkinto, MA, USA). Bulk human blank urine was suppliedy BioRad (USA).

g1tcaw

Acta 586 (2007) 184–195 185

.2. Standards and reference material

Norandrosterone (NorA), noretiocholanolone (NorE), 19-oretiocholanolone glucuronide, methyltestosterone, 17�-androlone (17�N), norandrosterone-d4 (NorA-d4), testos-erone (T) and epitestosterone (epiT) were supplied by NARL,eference Materials (Cottesloe, Australia). Norepiandrosterone

NorEpiA), 17�-estradiol (a-E2), 5�-estran-3�, 17�-diol (abb),�-estran-3�, 17�-diol (bab), 5�-estran-3�, 17�-diol (aba)ere purchased by Steraloids Inc. (Newport Rhode Island,SA). Dehydroepiandrosterone-3-sulphate, 17�-estradiol (b-2), 17�-nandrolone (17�N) and estrone were supplied byigma (St. Louis, Mo, USA). 5(10)-Estren-3�, 17�-diol wasupplied by Hong Kong Jockey Club and norandrostenedioneas purchased by Research Plus Inc. (Manasquan, NJ, USA).Stock standard solutions (1 mg mL−1, in free base form) of

ach analyte were prepared by dissolving 10 mg of the free baseorm in 10 mL of methanol. Working solutions of 100, 10 and�g mL−1 were prepared by 1:10, 1:100 and 1:1000 dilutionsf the 1 mg mL−1 stock solutions with methanol. All solutionsere stored at −20 ◦C.

.3. Extraction of steroids from urine matrix

Three metabolic fractions were extracted: free, glucuronidend sulphate fraction.

.3.1. Free fraction extractionAliquots of urine samples (10 mL for boars and 5 mL for

orses) were added with a concentration of 20 ng mL−1 ofethyltestosterone and norandrosterone-d4, used as internal

tandards, and adjusted to pH 9.5 with NH4Cl/NH3 buffer100 �L). Free steroids were extracted with 5 mL of tert-utylmethylether by shaking at 40 mpm for 20 min. Afterentrifugation (3500 rpm, 5 min), organic layers were evap-rated to dryness under nitrogen stream in a water bath at0 ◦C. The extracts were kept in desiccators containing P2O5nd maintained under vacuum for at least 30 min beforeerivatization.

.3.2. Extraction of the glucuronide fractionThe small volume of organic solvent still present on top of

he aqueous phase was evaporated under nitrogen stream. Thequeous phase was applied to Sep Pak C18 cartridges previ-usly conditioned with methanol (2 mL) and water (2 mL). Tworotocols of cartridge washing and elution were compared: A,ashing with water (2 mL), drying for 2 min, and eluting withethanol (2 mL); and B, washing with water (2 mL) and n-

exane (5 mL), drying for 2 min, and eluting with a mixturef methanol/ethyl acetate (30:70, v/v) (5 mL).

The organic extracts were evaporated to dryness under nitro-en stream in a bath at 50 ◦C. Residues were reconstituted inmL of sodium phosphate buffer (0.2 M, pH 7), and subjected

o enzymatic hydrolysis with 50 �L of �-glucoronidase from E.oli and incubation at 55 ◦C for 1 h. Then, samples were madelkaline with 250 �L of 5% solution of K2CO3 and extractedith tert-butylmethyleter (5 mL) by shaking at 40 mpm for

186 M. Roig et al. / Analytica Chimica Acta 586 (2007) 184–195

Table 1Trimethylsilyl (TMS) derivatives of the analytes under study, absolute and relative retention times (RT and RRT) and diagnostic ions used for identification and forquantitation

Analyte Abbreviation Derivative RT RRT Diagnostic m/z m/z for quantitation

Norandrosterone-d4 NorA-d4 Bis–O-TMS 8.04 0.5816 409, 424, 319 409Norandrosterone NorA Bis–O-TMS 8.06 0.5834 405, 420, 315 4055�-Estrane-3�,17�-diol baa Bis–O-TMS 8.08 0.5846 242, 332, 407 2425(10)-Estrene-3�,17�-diol – Bis–O-TMS 8.58 0.6207 330, 240, 405 3305�-Estrane-3�,17�-diol aab Bis–O-TMS 8.63 0.6245 242, 332, 407 242Norepirandrosterone NorEpiA Bis–O-TMS 8.77 0.6349 405, 420, 315 4055�-Estrane-3�,17�-diol aba Bis–O-TMS 8.86 0.6411 242, 332, 407 242Norepietiocholanolone NorEpiE Bis–O-TMS 8.94 0.6470 405, 420, 315 405Noretiocholanolone NorE Bis–O-TMS 8.95 0.6479 405, 420, 315 4055�-Estrane-3�,17�-diol bab Bis–O-TMS 9.34 0.6757 242, 332, 407 2425�-Estrane-3�,17�-diol abb Bis–O-TMS 9.61 0.6953 242, 332, 407 2425(10)-Estrene-3�,17�-diol – Bis–O-TMS 10.13 0.7330 330, 240, 405 33017�-Nandrolone 17�N Bis–O-TMS 10.53 0.7618 418, 403, 182 418Norandrostenedione Noraed Bis–O-TMS 10.88 0.7871 416, 401, 311 416Epitestosterone EpiT Bis–O-TMS 11.16 0.8076 432, 417 43217�-Estradiol a-E2 Bis–O-TMS 11.23 0.8126 285, 416, 325 28517�-Nandrolone 17�N Bis–O-TMS 11.24 0.8135 418, 403, 182 418Estrone – Bis–O-TMS 11.60 0.8395 414, 399 41417�-Estradiol b-E2 Bis–O-TMS 11.88 0.8596 285, 416, 325 285Testosterone T Bis–O-TMS 12.19 0.8825 432, 417 432Methyltestosterone – Bis–O-TMS 13.818 1.0000 446, 301 446

Table 2Analytes studied in each fraction, calibration range, extraction recovery and limits of detection (LOD) and quantitation (LOQ)

Fraction Analyte Concentrationrangea (ng mL−1)

Extraction recovery (%)mean ± S.D. (n = 4)

LOD (ng mL−1) LOQ (ng mL−1)

Free NorA 2–50 100.05 ± 5.0 0.4 1.1NorE 2–50 99.1 ± 9.9 0.1 0.2NorEpiA 2–50 102.4 ± 14.5 0.2 0.3bab 2–50 102.9 ± 3.6 0.7 2.1aba – 101.2 ± 4.5 0.3 1.0abb 2–50 105.9 ± 2.7 0.3 1.017�N – 89.6 ± 12.4 0.02 0.0517�N 2–50 92.3 ± 10.1 0.3 0.95(10)-Estrene-3�,17�-diol – 99.4 ± 6.5 0.7 2.1a-E2 – 101.4 ± 7.4 0.2 0.5b-E2 – 97.3 ± 2.9 0.2 0.6

Glucuronide NorA 2–50 88.5 ± 2.5 0.2 0.7NorE 2–50 78.8 ± 5.9 0.2 0.7NorEpiA 2–50 83.0 ± 9.4 0.4 1.2bab 2–50 88.5 ± 6.4 0.4 1.2aba 5–1500 81.9 ± 10.8 0.5 1.5abb 1–100 81.7 ± 8.1 0.4 1.317�N 1–100 84.9 ± 10.6 0.2 0.517�N 2–50 84.9 ± 5.8 0.1 0.35(10)-Estrene-3�,17�-diol – 83.3 ± 7.5 0.2 0.4a-E2 – 73.6 ± 9.3 0.3 0.7b-E2 – 81.7 ± 8.1 0.6 1.7

Sulphate NorA 2–50 79.9 ± 3.0 0.2 0.7NorE 2–50 96.4 ± 13.2 0.7 2.0NorEpiA 2–50 (b)/2–500 (h) 81.5 ± 10.3 0.6 1.7bab 2–50 68.7 ± 12.3 0.3 1.1aba 2–500 84.3 ± 6.2 0.9 2.6abb 2–500 104.1 ± 19.3 0.3 0.917�N 2–50 (b)/5–2000 (h) 68.8 ± 10.4 0.1 0.25(10)-Estrene-3�,17�-diol 5–2000 83.7 ± 5.0 0.4 1.2a-E2 2–100 87.8 ± 9.4 0.1 0.3b-E2 2–100 100.9 ± 11.0 0.03 0.1

a b: boar; h: horse.

M. Roig et al. / Analytica Chimica Acta 586 (2007) 184–195 187

F one la

2w5mt

2

ffrstaf

nw(S

er7ttab

2

thec

ig. 1. GC/MS results of a sample obtained after administration of 17�-nandrol

0 min. After centrifugation (3500 rpm, 5 min), organic layersere evaporated to dryness under nitrogen stream in a bath at0 ◦C. The extracts were kept in a desiccator with P2O5 andaintained under vacuum for at least 30 min, until derivatiza-

ion.

.3.3. Extraction of the sulphate fractionFor analysis of sulphate fraction, free fraction was extracted

rom urine samples (aliquots of 10 mL for boars and 5 mLor horses), as explained before. The remaining organicesidue in the aqueous phase was evaporated under nitrogentream. The aqueous phase was applied to Sep Pak C18 car-ridges as described previously. Two protocols of washingnd elution were compared as described for the glucuronideraction.

The organic extracts were then evaporated to dryness under
itrogen stream in a bath at 50 ◦C. Residues were reconstitutedith 4 mL of a mixture of ethyl acetate:methanol:sulphuric acid
80:20:0.06, v/v/v) and they were incubated at 55 ◦C for 2 h.amples were then neutralized with NaOH 1M (60 �L) and

2

m

ureate to boar. (A) free fraction; (B) glucuronide fraction; (C) sulphate fraction.

vaporated to dryness under nitrogen stream. The residues wereeconstituted in 1 mL of sodium phosphate buffer (0.2 M, pH) and 250 �L of 5% K2CO3 solution, and then extracted withert-butylmethyleter (5 mL). The organic layers were evaporatedo dryness under nitrogen stream in a bath at 50 ◦C and kept indesiccator containing P2O5 under vacuum for at least 30 minefore derivatization.

.4. Derivatization

The dry extracts were derivatized with 50 �L of a mix-ure of MSTFA:NH4I:2-mercaptoethanol (1000:2:6, v/w/v), andeated at 60 ◦C for 20 min. After incubation, the derivatizedxtracts were transferred to injection vials and analyzed by gashromatography mass spectrometry.

.5. Gas chromatography–mass spectrometry conditions

Analyses were performed in a Hewlett-Packard 5890 II GCodel fitted with a HP 7673 A auto sampler and connected


Fg

Fig. 2. GC/MS results of a sample obtained after administration of 17�-nandr

ig. 3. Concentrations of 17�-nandrolone and metabolites in urine obtained afterlucuronide fraction; (C) sulphate fraction.

olone laureate to horse. (A) Glucuronide fraction; (B) sulphate fraction.

administration of 17�-nandrolone laureate to a boar. (A) Free fraction; (B)


Table 3Intra- and inter-assay precisions and accuracies obtained for the free fraction of boar urine

Analyte C (ng mL−1) Assay n Intra-assay Inter-assay

Estimated concentration(mean ± S.D.) (ng mL−1)

Precision(R.S.D.%)

Accuracy (relativeerror, %)


Precision(R.S.D.%)


NorA 4 1 3 3.8 ± 0.4 10.9 6.8 3.8 ± 0.4 9.6 7.72 3 4.4 ± 0.3 7.0 11.33 3 3.9 ± 0.2 6.3 4.9

10 1 3 9.7 ± 0.3 2.5 2.8 9.3 ± 0.8 8.8 2.72 3 9.9 ± 0.1 1.3 1.33 3 10.4 ± 0.4 4.3 3.9

NorE 4 1 3 2.9 ± 0.2 6.7 27.9 3.6 ± 0.7 6.3 20.42 3 3.1 ± 0.2 5.8 23.23 3 4.4 ± 0.1 2.3 10.1

10 1 3 12.3 ± 0.4 3.6 22.6 9.0 ± 2.5 27.4 25.02 3 7.2 ± 0.2 3.3 27.93 3 7.6 ± 0.5 6.3 24.5

NorEpiA 4 1 3 3.1 ± 0.3 9.5 23.1 3.6 ± 0.6 16.8 15.42 3 3.5 ± 0.2 6.2 13.83 3 4.4 ± 0.1 2.5 9.3

10 1 3 11.4 ± 0.1 0.6 14.5 9.1 ± 1.8 19.7 18.52 3 7.6 ± 0.2 2.6 23.73 3 8.3 ± 0.5 6.4 17.2

bab 4 1 3 3.6 ± 0.4 11.8 10.8 3.9 ± 0.4 9.8 4.22 3 4.1 ± 0.2 5.8 4.23 3 3.9 ± 0.4 9.2 6.2

10 1 3 11.4 ± 0.3 2.4 14.4 10.2 ± 1.0 10.2 8.32 3 9.6 ± 0.6 6.7 5.13 3 9.5 ± 0.4 4.2 5.3

abb 4 1 3 3.7 ± 0.4 10.0 8.5 3.9 ± 0.4 11.0 7.72 3 4.1 ± 0.2 4.1 4.33 3 3.8 ± 0.7 17.9 11.1

10 1 3 10.2 ± 0.5 4.6 3.0 10.2 ± 0.4 3.6 3.52 3 10.3 ± 0.2 1.5 3.33 3 9.5 ± 0.7 7.3 6.0

17�N 10 1 3 12.29 ± 0.2 2.0 22.9 10.0 ± 1.8 18.4 15.42 3 8.4 ± 0.8 9.1 16.53 3 9.3 ± 0.4 4.2 6.9

30 1 3 31.5 ± 1.3 4.0 5.0 31.0 ± 3.9 12.5 9.32 3 30.1 ± 2.5 8.3 6.4

16.6

tP11

paua2uiec

2

felr

tb

3 3 31.4 ± 7.1 22.5

o a HP 5970 mass-selective detector (Agilent Technologies,alo Alto, CA, USA). The separation was carried out in a00% methylsilicone fused silica capillary column [HP Ultra-1;7 m × 0.2 mm i.d.; film thickness, 0.11 �m].

Oven temperatures were programmed as follows: initial tem-erature 181 ◦C; increase at 3 ◦C min−1 to 230 ◦C, then increaset 40 ◦C min−1 to 310 ◦C, maintained for 3 min. Helium wassed as carrier gas with a flow rate of 0.8 mL min−1 (measuredt 180 ◦C). The injector and the detector were maintained at80 ◦C. The injection volume was 2 �L and split mode 10:1 was
sed. The mass spectrometer was operated in selected-ion mon-toring acquisition mode using three diagnostic ions to monitorach compound of interest. The diagnostic ions used for eachompound are listed in Table 1.
ia

u

.6. Validation study

The validation study was performed for the three fractions:ree, glucuronide and sulphate fractions. The following param-ters were evaluated: selectivity, heteroscedasticity, linearity,imits of detection (LOD) and quantification (LOQ), extractionecovery, and intra- and inter-assay precisions and accuracies.

The validation protocol included four different assays. Selec-ivity and specificity were studied by the analysis of differentlank urines from boars and horses. The presence of any interfer-
ng substance at the retention time of the compounds of interestnd the internal standards was verified.
Due to the difficulties in obtaining high volumes of blankrine samples from boars and horses, calibration curves prepared


Table 4Intra- and inter-assay precisions and accuracies obtained for the glucuronide fraction of boar urine

Analytes C (ng mL−1) Assay n Intra-assay Inter-assy


Precision(R.S.D.%)

Accurracy(relative error, %)


Precision(R.S.D.%)

Accurracy(relative error, %)

NorA 4 1 3 4.1 ± 0.2 5.0 3.5 4.3 ± 0.2 5.3 7.82 3 4.3 ± 0.1 1.8 6.23 3 4.6 ± 0.1 1.9 13.8

10 1 3 11.0 ± 0.4 3.8 9.6 10.2 ± 0.9 8.5 7.52 3 9.2 ± 0.1 1.1 8.03 3 10.5 ± 0.6 5.7 5.6

NorE 4 1 3 4.5 ± 0.1 1.2 13.7 4.1 ± 0.5 12.8 11.72 3 3.9 ± 0.6 14.6 10.73 2 3.8 ± 0.6 15.2 10.1

10 1 3 10.4 ± 0.5 4.7 4.4 9.4 ± 1.1 11.9 9.92 3 8.5 ± 0.4 4.3 15.4

NorEpiA 4 1 3 4.6 ± 0.3 6.0 16.3 4.3 ± 0.6 13.8 15.22 3 4.7 ± 0.2 5.1 18.53 3 3.6 ± 0.1 3.5 10.6

10 1 3 9.9 ± 0.1 1.4 0.9 9.4 ± 0.8 9.1 6.72 3 9.5 ± 0.7 7.4 6.23 3 8.7 ± 1.1 12.7 12.9

bab 4 1 3 4.6 ± 0.1 2.0 14.6 4.0 ± 0.5 13.0 11.02 3 3.6 ± 0.3 7.4 10.13 2 3.7 ± 0.2 6.6 6.9

10 1 3 9.8 ± 0.2 2.2 1.8 10.0 ± 1.6 16.2 11.12 3 11.7 ± 1.6 14.1 16.73 3 8.5 ± 0.5 6.2 14.7

17�N 10 1 3 10.6 ± 0.4 3.3 6.5 10.4 ± 0.8 7.5 7.52 3 9.9 ± 1.1 11.4 8.83 3 10.6 ± 0.7 6.1 7.1

6.55.35.7

wocCac

to

F(

30 1 3 31.9 ± 1.8 5.71 3 28.6 ± 1.6 5.42 3 30.8 ± 2.1 6.7

ith human blank urine were used. To demonstrate the abilityf the calibration curve to predict values in animal urine, quality
ontrol samples were prepared in drug-free boar or horse urine.alibration samples were prepared daily by adding appropri-te volumes of stock solutions to drug-free human urine. Fivealibration levels were prepared covering the expected concen-
padt

ig. 4. Concentrations of 17�-nandrolone and metabolites in urine obtained after adB) sulphate fraction.

30.4 ± 2.2 7.1 5.8

ration range for each compound (see Table 2). In the first assayf validation, the calibration curves were prepared in quadru-
licate. Peak area ratios between the selected ion of the analytend the selected ion of the ISTD (m/z 409 for Norandrosterone-4 (NorA-d4) or m/z 446 for methyltestosterone) were plot-ed against the concentration. The Dixon’s test (α = 5%) was
ministration of 17�-nandrolone laureate to a horse. (A) Glucuronide fraction;


Table 5Intra- and inter-assay precisions and accuracies obtained for the glucuronide fraction of horse urine



Precision(R.S.D.%)



Precision(R.S.D.%)


aba 50 1 3 49.9 ± 7.0 14.1 10.7 57.4 ± 10.1 17.6 18.92 3 64.5 ± 9.4 14.6 29.03 2 58.0 ± 10.7 18.4 16.1

500 1 3 443.8 ± 23.6 5.3 11.2 497.4 ± 52.4 10.5 8.32 3 511.8 ± 18.5 3.6 4.23 3 536.5 ± 56.1 10.5 10.4

abb 15 1 3 13.9 ± 3.5 25.4 15.6 14.4 ± 2.2 15.4 10.22 3 13.9 ± 1.16 8.3 7.23 2 15.9 ± 0.6 3.9 6.5

40 1 3 40.8 ± 1.8 4.4 3.4 39.1 ± 2.8 7.1 6.02 3 36.2 ± 1.3 3.2 9.63 3 40.4 ± 2.5 6.2 5.1

17�N 15 1 3 13.8 ± 1.5 10.8 9.3 14.7 ± 1.7 11.7 9.22 3 16.1 ± 1.9 11.9 11.33 2 14.1 ± 0.7 4.8 5.8

40 1 3 36.6 ± 2.2 6.1 8.5 38.3 ± 2.3 6.1 5.52 3 38.9 ± 2.7 7.0 4.93 3 39.3 ± 1.6 4.0 3.1

17�N 6 1 3 4.9 ± 0.8 16.4 18.0 5.1 ± 0.6 11.0 15.52 3 5.3 ± 0.7 9.2 12.13 2 5.0 ± 0.4 9.0 16.8

15 1 3 13.4 ± 0.7 5.0 10.8 13.8 ± 1.0 7.5 8.57.47.3

atrwrdma

oca

caottr

apcrsr

cwt

2

t1Sdowcts

iwUds

2 3 13.9 ± 0.7 5.43 3 14.2 ± 1.6 11.5

pplied to detect outliers in the replicates at each concentra-ion level. The behavior of the variance over the calibrationange (homoscedasticity/heteroscedasticity of the procedure)as evaluated by applying the Levene test (α = 5%). In the

est of validation assays, the calibration curve was prepared inuplicate. To demonstrate the goodness of fit using the linearodel, and F test (α = 5%) was applied to compare the variance

ttributable to lack of fit with that due to random error.The LOD and LOQ were defined as 3.3 and 10 times the value

f noise, respectively. The standard deviation of the estimatedoncentration values for the lowest calibration point for eachnalyte was used as a measure of the noise.

Extraction recoveries of each analyte from urine were cal-ulated by comparison of peak areas of compounds obtainedfter analysis of spiked urine samples spiked with 20 ng mL−1

f each compound (n = 4) with the mean of those obtained whenhe standards were added to extracted blank urines and subjectedo the derivatization process (representing 100% of extractionecovery).

Intra-assay precision and accuracy were determined by thenalysis at the same day of quality control urine samples, pre-ared in drug-free boar or horse urine and containing different
oncentration of each analyte. Inter-assay precision and accu-acy were calculated by the analysis of the quality controlamples in three different days. Precision was expressed as theelative standard deviation (R.S.D.) of the control samples con-
3

t

entration calculated using the calibration curve and accuracyas expressed as the relative error (%) of the estimated concen-

ration.

.6.1. Samples from animalsAdministration studies of 17�-nandrolone laureate (17�NL)

o boars and horses were performed. Two doses of 2 mg kg−1 of7�NL in arachnid oil (Erton fortificante®. Lab. Ovejero, Leon,pain) were administered intramuscularly in two consecutiveays to healthy male boars. The boars were around 10 weeksld and their weights were in the range from 20.8 to 24.6 kghen the study started. The animals were housed in metabolic

ages during the study. Urine samples were collected daily upo 15 days after the administration of the drug. Samples weretored at −20 ◦C until analysis.

A second set of urine samples was obtained in an admin-stration study to horses. A single dose of 375 mg of 17�NLas administered intramuscularly to two healthy gelded horses.rine samples were collected before the administration of therug and up to 9 days after administration. Urine samples weretored at −20 ◦C until analysis.

. Results and discussion

In this study, an analytical method has been optimized forhe quantitation of 17�-nandrolone (17�N), its metabolites and


Table 6Intra- and inter-assay precisions and accuracies obtained for the sulphate fraction of boar urine



Precision(R.S.D.%)



Precision(R.S.D.%)


NorA 4 1 3 3.9 ± 0.1 3.2 2.9 4.3 ± 0.3 7.7 9.92 3 4.6 ± 0.2 3.6 15.83 3 4.5 ± 0.2 3.9 11.2

10 1 3 9.5 ± 0.6 6.4 5.5 9.9 ± 0.6 5.7 4.32 3 10.4 ± 0.2 2.2 3.73 3 9.9 ± 0.5 5.3 3.6

NorE 4 1 3 4.07 ± 0.3 8.3 7.1 4.0 ± 0.5 12.4 8.72 3 3.9 ± 0.6 15.4 10.93 3 3.7 ± 0.2 4.7 8.2

10 1 3 9.4 ± 1.6 17.0 10.0 9.2 ± 1.8 19.3 18.52 2 6.9 ± 0.5 6.8 31.2

NorEpiA 4 1 3 4.2 ± 0.4 10.9 10.1 4.0 ± 0.4 8.8 7.52 3 3.7 ± 0.2 5.4 7.23 3 4.1 ± 0.3 6.3 5.5

10 1 3 10.4 ± 1.6 15.5 13.4 9.7 ± 1.1 10.5 8.72 3 9.6 ± 0.7 7.9 6.93 3 9.9 ± 0.8 8.3 5.9

bab 4 1 3 4.4 ± 0.5 12.4 9.9 4.0 ± 0.5 11.4 7.92 3 3.9 ± 0.4 9.2 6.23 3 3.7 ± 0.1 3.3 7.7

10 1 3 11.2 ± 1.7 14.8 16.9 10.7 ± 1.1 10.3 8.72 3 10.1 ± 0.3 2.6 1.83 3 10.7 ± 0.9 8.7 7.3

17�N 10 1 3 10.8 ± 2.6 24.4 22.3 9.9 ± 1.6 16.2 6.52 3 9.0 ± 0.5 5.7 9.73 3 9.9 ± 0.8 8.5 6.4

30 1 3 32.0 ± 0.8 2.4 6.6 31.4 ± 1.7 5.5 12.38.84.0

owesto

eatfttsfiTwwstc

ea[oe

psaabilft1a

2 3 31.4 ± 2.9 9.13 3 30.7 ± 1.3 4.3

ther endogenous steroids in boar and horse urine. The methodill be useful for metabolic studies of 17�N and to determine the

ndogenous production of some of the compounds by analysingamples obtained from non-treated animals, in order to definehe suitable target analytes to detect exogenous administrationf 17�N for fraudulent purposes.

It has been described that 17�N and its metabolites arexcreted in urine in free form and conjugated with glucuroniccid or sulphate, depending on the animal species [2–11]. Forhis reason, the developed method was able to separate the threeractions. The method consist of a first liquid–liquid extractiono recover free steroids present in urine and a solid-phase extrac-ion of the remaining aqueous phase containing the conjugatedteroids, followed by an enzymatic hydrolysis or solvolysis and,nally, a liquid–liquid extraction of the deconjugated steroids.wo different solvents (n-pentane and tert-butylmethylether)ere compared for the liquid–liquid extractions. The extraction
ith n-pentane was found to be more selective and it provided
imilar extraction yields than tert-butylmethylether for most ofhe compounds, except for the most polar compounds (e.g., thoseontaining two alcohol groups, 5�-estran-3�, 17�-diol, bab; 5�-

f[(

stran-3�, 17�-diol, abb and 5�-estran-3�, 17�-diol, aba), ingreement with results obtained with other steroid metabolites21]. Due to the relevance of these compounds in the metabolismf 17�N, tert-butylmethylether was selected for liquid–liquidxtractions.

The effect of the different clean-up procedures on the solid-hase extraction of the glucuronide and sulphate conjugates wastudied for three target analytes: norandrosterone (NorA), babnd 17�N. Similar selectivity was obtained in both proceduresnd the extraction recoveries for the target analytes were slightlyetter for the extraction A and, therefore, it was selected to val-date the analytical methodology. The extraction recoveries areisted in Table 2. As can be observed, they were higher than 90%or all the analytes in the free fraction, and better than 80% forhe analytes in the glucuronide and sulphate fraction, except for7�-estradiol (a-E2) (74%) in the glucuronide fraction and babnd 17�N in the sulphate fraction, that were close to 70%.

�-Gucuronidase from E. coli has been conventionally usedor the hydrolysis of steroids conjugated with glucuronic acid21]. The hydrolysis efficiency was estimated in 91.7 ± 7.5%n = 4) for noretiocholanolone glucuronide. Solvolysis was used


Table 7Intra- and inter-assay precisions and accuracies obtained for the sulphate fraction of horse urine



Precision(R.S.D.%)



Precision(R.S.D.%)


aba 25 1 3 26.4 ± 0.7 2.6 5.7 24.3 ± 2.4 9.8 7.32 3 23.3 ± 1.9 8.3 8.13 3 23.1 ± 2.8 12.3 8.2

150 1 3 142.1 ± 16.9 11.9 9.0 146.1 ± 10.5 7.2 5.42 3 147.3 ± 7.8 5.3 3.43 3 148.8 ± 7.7 5.2 3.9

abb 25 1 3 23.3 ± 2.6 11.0 9.5 24.0 ± 2.2 9.0 7.12 3 24.5 ± 1.6 6.4 4.33 3 24.2 ± 2.9 12.1 7.8

150 1 3 134.2 ± 14.1 10.5 10.5 137.2 ± 11.3 8.3 8.52 3 137.5 ± 15.3 11.1 8.33 3 139.8 ± 7.7 5.5 6.8

NorEpiA 25 1 3 21.7 ± 2.7 12.4 13.2 23.0 ± 2.1 9.2 8.32 3 24.4 ± 0.7 2.8 2.73 3 22.8 ± 2.1 9.4 8.9

150 1 3 131.2 ± 12.3 9.4 12.5 141.6 ± 11.8 8.4 7.62 3 148.3 ± 9.6 6.4 5.03 3 145.2 ± 8.1 5.6 5.2

17�N 50 1 3 54.4 ± 7.6 14.0 13.8 52.4 ± 8.2 15.7 13.12 3 47.7 ± 8.2 17.1 12.53 3 55.1 ± 9.8 17.9 13.1

800 1 3 688.7 ± 151.1 21.9 13.9 667.3 ± 96.1 14.4 16.62 3 683.9 ± 59.8 8.7 14.53 3 629.22 ± 84.9 13.5 21.3

5(10)-Estren-3�,17�-diol

50 1 3 53.6 ± 2.4 4.5 7.3 48.6 ± 8.8 18.1 14.72 3 39.2 ± 2.4 6.1 21.73 3 53.1 ± 9.9 18.6 15.1

800 1 3 679.7 ± 80.6 11.9 15.0 703.6 ± 52.3 7.4 12.12 3 716.7 ± 39.7 5.5 10.43 3 714.3 ± 39.7 5.6 10.7

a-E2 10 1 3 10.2 ± 0.5 5.1 4.5 10.6 ± 0.7 6.5 7.22 3 10.9 ± 0.7 6.1 8.73 3 10.7 ± 0.9 8.1 8.3

50 1 3 51.0 ± 5.9 11.6 8.6 49.4 ± 4.0 8.1 6.12 3 50.3 ± 2.8 5.5 3.83 3 47.0 ± 2.7 5.8 6.0

b-E2 10 1 3 9.4 ± 1.2 13.2 8.3 9.3 ± 1.0 11.1 8.12 3 9.6 ± 0.6 6.6 4.23 3 8.8 ± 1.3 14.9 11.8

50 1 3 52.6 ± 4.1 7.7 5.6 54.6 ± 2.9 5.3 9.32 3 55.9 ± 1.0 1.9 11.9

1

tat[Tf1y

tce

3 3 55.2 ± 2.6 4.6

o deconjugate sulphate metabolise. The use of enzymes such asryl-sulphatase from Helix pomatia has proven to be not effec-ive for the hydrolysis of sulphate metabolites in 17�-position23], which are important in the metabolism of 17�N in animals.
he solvolysis efficiency was estimated in 90.7% ± 5.4 (n = 4)
or dehydroepiandrosterone-3-sulphate. No pure standards of7� steroid sulphate was available to estimate the hydrolysisield for this type of conjugates.

op

a

0.4

The derivatives formed for each analyte, with the retentionimes (RT and RRT), and the diagnostic ions used for identifi-ation and quantitation are listed in Table 1. In Figs. 1 and 2,xamples of chromatograms obtained after the administration
f 17�N laureate (17�NL) to boars and horses, respectively, areresented.
The analytes validated of each fraction (free, glucuronidend sulphate fraction) and the concentration ranges studied are

1 imica

ltr(rhsd

rt2

euwmpdipbfniRdIb

eTats1fr

tahl

ae2ftp13

aCsd

wwf

s1ftiBa

imth

A

(0KvlRe

R

[

[[

[

[

94 M. Roig et al. / Analytica Ch

isted in Table 2. The procedure was found to be heteroscedas-ic for all analytes studied in all three fractions, so peak areaatios between the target analytes and the internal standardISTD) were subjected to a proportional weighted least-squareegression analysis. Determination coefficients (r2) were alwaysigher than to 0.99. The comparison of variances was found notignificant (α = 5%), indicating an adequate adjustment of theata to the proposed lineal model.

The estimated LOD and LOQ are also listed in Table 2. LOQanged from 0.05 to 2.1 ng mL−1 in the free fraction, from 0.3o 1.7 ng mL−1 in the glucuronide fraction, and from 0.2 to.6 ng mL−1 in the sulphate fraction.

The selectivity/specificity were evaluated after analysis ofight different blank urines from boars and five different blankrines from horses. In general terms, no matrix interferencesere detected at the RT of the analytes and the ISTD for the ionsonitored in any of the fractions studied. Only an endogenous

eak appeared close to the RT of norepiandrosterone (NorEpiA)erivative in the chromatogram at m/z 420, which was not theon chosen for quantification of this compound. This endogenouseak was present in the free fraction of some of the samples fromoars and in the glucuronide fraction of some of the samplesrom boars and horses. It is worth to notice that its presence didot preclude the quantitation of NorEpiA. On the other hand,n the sulphate fraction of some urines from boars, peaks at theT of NorA and norandrosteron-d4 (NorA-d4) derivatives wereetected in the chromatograms of the ion used for quantitation.n the samples containing the interfering peak, NorA could note quantified.

Results of the precision and accuracy studies for the differ-nt fractions of urines from boars and horses are presented inables 3–7. These results are in agreement with the internation-lly accepted ranges for these parameters [24–27]. Regardinghe free fraction of urine from boar (Table 3) intra-assay preci-ions were below 12% for all analytes, except for one assay of7�N, and intra-assay accuracy values were below 25% exceptor norethicholanolone (NorE). Inter-assay precision and accu-acy were below to 20%, except for NorE.

The intra- and inter-assay precisions of the glucuronide frac-ion of boar urines (Table 4) were below 15.2%, while theccuracies were below 17%. For the glucuronide fraction oforse urine (Table 5), precisions and accuracies were in generalower than 15%.

Regarding sulphate fraction of boar urine (Table 6), intra-ssay precision values were below 17% for most of the analytes,xcept for 17�N. Intra-assay accuracy values were below to5%, except for one validation day of NorE. Inter-assay valuesor precision and accuracy were in the range below to 20% for allarget analytes. In horse urine (Table 7), intra- and inter-assayrecision and accuracies for the sulphate fraction were below5% except for some experiments of 17�N and 5(10)-estren-�, 17�-diol.

The method was applied to the analysis of samples obtained
fter administration of 17�N laureate to boars and to horses.oncentrations for each metabolite obtained in boar are pre-
ented in Fig. 3. NorA, NorE, NorEpiA, bab and abb wereetected in the free fraction. NorA, NorE, NorEpiA and 17�N

[[

[

Acta 586 (2007) 184–195

ere detected in the glucuronide fraction, and NorA and 17�Nere detected in the sulphate fraction. Metabolites were detected

or up to 15 days.The concentrations obtained in horses in the glucuronide and

ulphate fractions are presented in Fig. 4. aba, abb, NorEpiA,7�N and 17�-nandrolone (17�N) were detected in glucuronideractions, and aba, abb, NorEpiA and 17�N were detected inhe sulphate fraction. The most important metabolites were aban the glucuronide fraction and 17�N in the sulphate fraction.efore administration of 17�NL (pre-dose samples), no targetnalytes were detected in any urine sample.

In summary, a method to quantify 17�N and its metabolitesn horse and boar urine has been developed and validated. The

ethod has proven to be suitable for the quantitative determina-ion of 17�N metabolites after administration of 17�N esters toorses and boars.

cknowledgements

Financial support from Ministerio de Educacion y CienciaSpain) is gratefully acknowledged (project number AGL2001-439-C02-01). The authors would like to thank of Dr. Pierceavanagh from Trinity College of University of Dublin for pro-iding samples obtained after administration of 17�-nandroloneaureate to horses. The technical support of J.C. Gonzalez and R.amırez and helpful discussions with Dr. J. Marcos are acknowl-dged.

eferences

[1] Council Directive 96/22/EC of 29 April 1996 Concerning the Prohibition onthe Use in Stockfarming of Certain Substances Having a Hormonal or Thy-rostatic Action and of �-Agonists, and Repealing, Directives 81/602/EEC,88/146/EEC and 88/299/EEC, OJ L125 23/5/1996, p. 3.

[2] W. Schanzer, Clin. Chem. 42 (1996) 1001.[3] E. Houghton, G.A. Oxley, M.S. Moss, S. Evans, Biomed. Mass Spectrom.

5 (1978) 170.[4] E. Houghton, M.C. Dumasia, Xenobiotica 10 (1980) 381.[5] M.C. Dumasia, E. Houghton, Xenobiotica 14 (1984) 647.[6] E. Houghton, M.C. Dumasia, P. Teale, M.S. Moss, J. Chromatogr. 383

(1986) 1.[7] E. Benoit, J.L. Guyot, D. Courtot, P. Delatour, Ann. Rech. Vet. 20 (1989)

485.[8] E. Daeseleire, A. de Guesquiere, C. Van Peteghem, Anal. Chim. Acta 272

(1993) 95.[9] T.P. Samuels, A. Nedderman, M.A. Seymour, E. Houghton, Analyst 123

(1998) 2401.10] G. Debruyckere, C. Van Peteghem, H.F. De Brabander, G. Debackere, Vet.

Q. 12 (1990) 246.11] G. Debruyckere, C. Van Peteghem, J. Chromatogr. 564 (1991) 393.12] P. Hemmersback, A.H. Hagensen, J. Misund, in: W. Schanzer, H. Geyer, A.

Gotzmann, U. Mareck-Engelke (Eds.), Recent Advances in Doping Anal-ysis, vol. 8, Sport und Buch Strau�, Editio Sport, Koln, 2000, p. 141.

13] B. Le Bizec, F. Monteau, I. Gaudin, F. Andre, J. Chromatogr. B 723 (1999)157.

14] E. Houghton, J. Copsey, M.C. Dumasia, P.E. Haywood, M.S. Moss, P.Teale, Biomed. Mass Spectrom. 11 (1984) 96.

15] M.C. Dumasia, E. Houghton, M. Jackiw, J. Endocrinol. 120 (1989) 223.16] A.S. Clouet, B. Le Bizec, M.P. Montrade, F. Monteau, F. Andre, Analyst

122 (1997) 471.17] S. Sterk, H. Herbold, M. Bokland, H. van Rossum, L. van Ginkel, R.

Stephany, Analyst 123 (1998) 2633.

imica

[[

[

[

[

[[

[

M. Roig et al. / Analytica Ch

18] A.F. Rizzo, E. Alitupa, T. Hirvi, S. Berg, Anal. Chim. Acta 275 (1993) 135.19] V. Marcos, E. Perogordo, P. Espinosa, H. Hooghuis, Alimentaria 11 (2002)

11.20] S. Poelmans, K. De Wasch, H. Noppe, N. Van Hoof, S. Van Cruchten, B.

Le Bizec, Y. Deceunink, S. Sterk, H.J. van Rossum, M.K. Hoffman, H.F.de Brabander, Food Addit. Contam. 22 (2005) 808.

21] W. Schaenzer, in: S.B. Karch (Ed.), Drug of Abuse Handbook, CRC Press,Boca Raton, 1998, p. 671.

22] J. Segura, R. Ventura, M. Gonzalez, C. Jimenez, in: M.J. Bogusz(Ed.), Handbook of Analytical Separations, Elsevier, Amsterdam, 2000,p. 531.

[

[

Acta 586 (2007) 184–195 195

23] P. Teale, E. Houghton, Biomed. Mass Spectrom. 20 (1991) 109.24] V.P. Shah, K.K. Midha, S. Dighe, I.J. McGilveray, J.P. Skelly, A. Yacobi,

T. Layloff, C.T. Wiswanathan, C.E. Cook, R.D. McDowall, K.A. Pittman,S. Spector, J. Pharm. Sci. 81 (1992) 309.

25] C. Hartmann, D.L. Massart, R.D. McDowall, J. Pharm. Biomed. Anal. 17(1998) 193.

26] C. Hartmann, D.L. Massart, R.D. McDowall, J. Pharm. Biomed. Anal. 12(1994) 1337.

27] Guidance for Industry; Bioanalytical Method Validation, Center for DrugEvaluation and Research (CDER), U.S. Food and Drug Administration,Rockville, 2001.

quantitation of 17 -nandrolone metabolites in boar and...

Documents