Studies on anabolic steroids. VII Analysis of urinary metabolites of formebolone in man by gas chromatography—mass spectrometry

Analyrrcu Chrmrcu Acre, 247 (1991) 211-221 Elsevter Scrence Publishers B.V., Amsterdam

211

Studies on anabolic steroids. VII

Analysis of urinary metabolites of formebolone in man by gas chromatography-mass spectrometry

Robert Masse *, Honggang Bi and Ping Du iitstltur iWlonai~a2 ia R&iercie .!hent&ue, IvN~RS&mt& Lhverslt.&dti &%ec, .T43 Hymus illid:, Yointe-C?hrre,

Quebec H9R IG6 (Canada)

(Received 20th August 1990)

Ahstraet

The metabolism of formebolone in man IS described.. After Sep-Pak Crs extraction, the free, conjugated and acuhc steroids were separated After enzymatrc hydrolysis of the conjugates, the neutral sterotds were identified by GC-MS Two metabolites identified as 2-hydroxymethyl-llcY-l7~-~ydroxy-l7~-methyl~drost-l,4-diene-3-one and lla-hydroxymethyltestosterone were isolated from both the free and conjugated sterord fractions. An acidic metabolite identified as llcr,l7~-dihydroxy-l7a-methyl-3-oxo-androsta-l,4,-diene-2-c~boxyhc acrd was rsolated from the free steroid fraction. Thrs acrdrc steroid was stable when denvatrzed as the methyl ester-TMS ether dmvative but its TMS ester-TMS ether denvatrve progressively degraded to yield the 17-mono and 11,17-dr-0-TMS ether derivatives of lla-hydroxymethandienone. The GC-MS features of formebolone metabobtes are presented and its routes of brotransformatton are discussed.

Keywords. Gas chromatography; Mass spectrometry; Formebolone metabolism; Sterords, anabohc

The problem of misuse of anabolic steroids in sports and their ban by the International Olympic Committee [l] have fostered the development of a new approach for the determination of these compounds and their urinary metabolites in humans [2]. This methodology is based on the use of solid supports for the isolation of free and conjugated steroids, enzymatic hydrolysis of steroid conjugates, derivatization of the neutral steroids and selective determination of the resulting derivatives by high-resolution gas chromatography-mass spectrometry (HRGC-MS). A thorough knowledge of the biotransformation routes and urinary excretion profiles of the metabolites of anabolic steroids is an essential prerequisite to their GC- MS monitoring. The unique structural features in many of these substances can significantly alter

0003-2670/91/$03 50 0 1991 - Elsevrer Science Pubhshers B.V.

their biotransformation routes compared to those of testosterone, after which anabolic steroids were originally synthesized. In that context, a research program on the metabolism of anabolic steroids in man has been undertaken in this laboratory [3-61.

This paper deals with the GC-MS of the urinary metabolites of formebolone (2-formyl-17a-methylandrosta-l,Cdiene-lla-17p-dihydroxy-3-one) [7]. The anabolic properties of this steroid have been investigated in humans [8,9] and its biotransformation was investigated in the rat [lo]. To our best knowledge, the metabolism of formebolone in man and the determination of its urinary metabolites by GC-MS have not been reported in the literature. In the present paper, a detailed study of the urinary excretion profile of formebolone metabolites in man is described. Three metabolites

212 R MASSI? ET AL

were characterized by GC-MS and their identities were assessed either by chemical synthesis or by comparison with authentic reference steroids. Tri-

methylsilylation of formebolone-2-carboxylic acid metabolite afforded two major by-products which were characterized by GC-MS.

EXPERIMENTAL

Sterords, chemrcals and reagents Formebolone was obtained from the LPB In-

stituto Farmaceutico (Milan, Italy). The llcw-17/3-

dihydroxy-17a-methyl-androst-4-en-3-one (lla- hydroxymethyltestosterone) and lla-17P-dihydroxy-17cr-methyl-androst-l,Cdiene-3-one(lla-hydroxymethandienone) and 5a-androstan-17-one

were purchased from Steraloids (Wilton, NH). Inorganic salts (J.T. Baker) were of analytical

grade. All the solvents were of HPLC grade (Caledon Laboratories, Georgetown, Ontario) and

were used as provided. Sodium chlorite (tech. 80%) dithioerythritol, trimethyliodosilane (TMIS), tri- methylchlorosilane (TMCS), sulfamic acid (99 + W) and l-methyl-3-nitro-1-nitrosoguanidine (MNNG) diazomethane kits were supplied by Aldrich; Sep- Pak C,, cartridges were purchased from Waters

Associates, N-Methyl-N-(trimethylsilyl)trifluoro- acetamide (MSTFA) was purchased from Regis (Morton Grove, IL). [ N,O] . bis-( d,)-trimethyl- silylacetamide (d,-BSA) and d,-trimethylchloro- silane (d,-TMCS) were obtained from MSD Iso- tope, Pointe-Claire.

Urme samples and extraction of metabobtes Blank urine samples were collected during a

period of 24 h before the administration of an oral

and single 40 mg dose of formebolone to a healthy male volunteer. Urine samples were then collected for the next 7 days and were frozen in the dark at - 20 o C immediately after collection.

The extraction of steroid metabolites from urinary samples (3-5-ml aliquots), their fractiona- tion mto free and comugated fractions and the

hydrolysis of the conjugated steroids were done as previously described [2,3].

For the extraction of acidic metabolites, lo-ml aliquots of urine were passed through a Sep-Pak Cl, cartridge (prewashed with 5 ml of methanol and 5 ml of water). The cartridge was then washed with 2 ml of hexane and steroids were eluted with 5 ml of methanol. The solvent was evaporated to dryness under a stream of nitrogen at 40” C and

the residue was dissolved in 1 ml of water, the pH of which was set to 1 with 10% (v/v) hydrochlonc acid. The acidic compounds were then extracted

with two 5-ml portions of diethyl ether. After evaporation of the solvent, the residue was de- rivatized and analyzed by GC-MS as described

below.

Preparation of denvatwes TMS enol-TMS ether derwatwes. The steroidal

extract was transferred to a 0.3-ml Reactivial and 10 ng ~1~’ 5a-androstan-17-one (external standard) was added. After evaporation of the solvent at 40” C under a nitrogen stream, 0.5-l mg of dithioerythritol was added and the vial stoppered under nitrogen. After the addition of MSTFA/ TMIS (100: 2, v/v), the mixture was heated at 70°C for 30 min, and 1 ~1 was injected into the

gas chromatograph. TMS ether and d,-TMS ether denvatwes. The

urinary extract was treated with MSTFA/ pyridine/TMCS (25 : 25 : 2.5, v/v) as described above and the resulting mixture was heated at 70 o C for 30 mm. The sample was then cooled to room temperature and the solvent removed at 60” C under a nitrogen stream. The dry residue was dissolved in 50 ~1 of hexane and 1 ~1 was injected mto the gas chromatograph. The corresponding perdeuterated d,-TMS derivatives were prepared by using d,-BSA/pyridine/d,-TMCS (25 : 25 : 2.5, v/v) under the conditions mentioned

above. TMS ester-TMS ether and d,-TMS ester-d,-

TMS ether derwatwes. These mixed derivatives of the acidic metabolite 3 were prepared as follows: the urinary extract was dissolved in pyridine/ MSTFA/TMCS (9: 1 :O.l, v/v) and heated at 70 o C for 30 min; the solution was then evaporated to dryness and the residue dissolved in 50 ~1 of

hexane. The d,-TMS derivative was prepared by using pyridine/d,-BSA/d,-TMCS (9 : 1 : 0.1, v/v)

STUDIES ON ANABOLIC STEROIDS 213

under the same expertmental conditions. Note that when the relattve amount of MSTFA was in-

creased, the degradation of the TMS ester-TMS ether derivattve of 3 proceeded at a faster rate.

0-Methyloxlme TMS ether (MO-TMS) and MO-d,-TMS. The dry steroidal extract was dissolved in 50 ~1 of a 5% (w/v) solution of methox- ylamine hydrochlonde in pyridine and heated at

70 o C for 30 min. After evaporation of the solvent, the trimethylstlyl and d,-TMS ether derivatives were prepared by adding 25 ~1 of MSTFA/TMCS (25 : 2.5, v/v) or d,-BSA/d,-TMCS (25 : 2.5, v/v). The resulting mixture was heated at 70°C for 30 min. The solvent was evaporated at 60 o C under a

stream of nitrogen and hexane (50 ~1) was added to the dry residue. An aliquot (1 ~1) of the resulting soluttons was injected for GC-MS.

Methyl ester-TMS ether derwatwes. The methyl ester derivatives of the acidic metabolites were prepared by using a 1-methyl-3-nitro-l-nitrosoguanidine (MNNG) diazomethane kit. The steroid-containing residue was dissolved in 2 ml of

ether and the solution was placed in the outside tube of the glass apparatus. Then, 50 mg of MNNG was placed in the inside tube along with

0.5 ml of water. The lower part of the apparatus was immerged in an ice bath and ca. 0.5 ml of 5 M potassium hydroxide was slowly added dropwise with a syringe through a silicone rt._bber septum, until the reaction was complete. The reaction mixture was left for 10 n-tin at room temperature and the solvent was evaporated to dryness. The residue was then reacted with a mixture of MSTFA/ TMCS as described above.

Gas chromatography-mass spectrometty

The steroid extracts were analyzed by using a Hewlett-Packard HP-5970 mass-selective detector

linked to a HP-5890 gas chromatograph equipped with a HP-5 (cross-linked 5% phenylmethylsih- cone) fused-silica capillary column (25 m X 0.2 mm t.d., 0.33 pm film thrckness). The injections were made m the splitless mode (30-s delay before the splitter was opened) with helium as carrier gas

at 0.8 ml min- ‘. The oven temperature was main-

tained at 100” C for 1 min and programmed at

16°C rnin-’ to 220 o C and then 3.8” C mm’ to 300°C which was mamtamed for 10 min.

Synthesis of 17cx-methyl-l I a,1 7p-drhydroxy-an-

drost-l,4-dlene-3-one-2-carboxyhc acrd

Thts compound (3) was synthesized by a method adapted from that reported by Bal et al. [ll]. To a solution of formebolone (104 mg, 0.3 mmol) in 4 ml of acetone was added 232 mg (2.4 mmol) of sulfamic acid dissolved in 3 ml of water. The resulting solution was stirred at room temperature while a solution of sodium chlorite (216 mg, 2.4 mmol) in 1 ml of water was added dropwise over a penod of 2 min. The resulting yellowish mixture was then stirred for 5 h. The solvent was partly evaporated under vacuum at room temperature until a gelatmous mass was obtained. Then 3 ml

of water was added and the pH was adjusted to 9.5 with an aqueous 2 M KHCO,/K,CO, (1: 1, w/w). The resulting solution was extracted three times with lo-ml portions of diethyl ether to re- move non-acidic compounds. The aqueous phase was acidified to pH 2 with 2 M hydrochloric acid and extracted as above. The combined organic phases were successively washed twice with 2 ml

of water, dried over anhydrous sodium sulfate and evaporated to dryness under vacuum. The resulting amorphous white solid was recrystalhzed from diethyl ether to gave 25.7 mg (23.7% yield) of pure 3 as white crystalline needles (m.p. 86-87 o C; v,,,,, 3510, 1745, 1725, 1660, 1650, 1580 and 1460 cm-‘). ‘H-NMR and 13C-NMR spectra were re- corded on Bruker WH-400 and Varian VXA-300

spectrometers, respectively, and m (methylsulfox- ide)-d,. Chermcal shifts and coupling constants were obtained from a first-order analysis of the ‘H

homonuclear chemical shift-correlated (COSY) spectra and 13C-‘H chemical shift-correlated spectra. NMR data were consistent with the struc-

ture of compound 3.

RESULTS AND DISCUSSION

Identlflcatlon of metabohtes

Given the structural features of formebolone, it was expected that in vivo btotransformattons would be directed by A-ring substituents, particu- larly at the formyl group. Two metabolites were isolated from the free fraction (Fig. 1A) as well as

from the conjugated fraction (not shown). Selected

214

ion-monitoring GC-MS showed that 44.2% of compound 2 and 85.5% of compound 5 were excreted in the free steroid fraction. These data suggested that 2 was probably more hydrophilic than 5.

Metabolite 2. The molecular ion at m/z 562 (m/z 589, d,-TMS) in the mass spectrum of the

1 Ion 143 _ 00 amu _

2.SES

I

R MASSI? ET AL

TMS ether derivative of compound 2 (Fig. 2A) was indicative of the presence of three hydroxyl groups. Structurally informative ions at m/z 222 (m/z 231, d,-TMS) and 235 (m/z 244, d,-TMS) arising from B-ring cleavages provided strong evi- dence for the presence of a 2-hydroxymethyl group. As evidenced by d,-TMS labelling, the ion

A

. .-_*+---=Y-,! . ( , 20 30 32 34 36 38 40

Ion 143.00 amu. 2.SE.5: 11

2. BE57 -

: 2 l.SES:

TJ 5 I.OES:

ii S.0E4:

0-. A -. .., ., . __ I

28 30 32 34 36 38 40

I. 6E57 Ion 143.00 amu. c

1.4ES:

1.2ES: u ; 1.0ES:

,” 8.0E4: c , 6.0E4

: 4.0E4

. I . *. I . :.-. ,~--v, ?\--_. , 28 30 32 34 36 38 40

Time (man 1

Fig. 1 Reconstructed ton chromatograms from GC-MS: (A) free steroids analyzed as the TMS ether denvatlve; (B) combmed free

and acldlc steroid fractions analyzed as the TMS ester-TMS ether denvatlves; (C) a blank unnary extract denvatwd v&h MSTFA Unne sample was collected 6 h after formebolone admmistratlon. The structures of the compounds correspondmg to the labelled

peaks are gwen m Ftg 7. See expenmental for analytical condltlons.


at m/z 207 (m/z 213, d,-TMS) is apparently tion of a TMS group on the 3-keto function, were formed by elimination of a CH, radical from the observed at m/z 281 (m/z 299, d,-TMS) and 294 TMS group of the fragment ion of m/z 222. Two (m/z 312, d,-TMS). The formation of such ions fragment ions, originating from the same B-ring in the mass spectra of the TMS derivatives of cleavages and accompanied by long-range migra- steroids bearing a”4-3-keto, el-3-keto and A’-3-

0.0E4 s e

7.0E4

6.0E4

:5.0E4

: a4.0E4

: ;3.0E4

2.0E4

10000

Mass/Charge

t-lass/Charge

200 300 400 500 600 Mass/Charge

Fig. 2. Electron-impact mass spectra of metabohte 2 as the TMS ether (A), MO-d9-TMS (B), and MO-TMS (C) denvatlves.

216 R MASSI? ET AL

hydroxy groups has been reported [6]. Long-range trimethylsilyl group migration under electron impact has previously been reported in hydroxylated a4-3-keto cholestene [12]. The proposed structural assignment for compound 2 was ascertained from the mass spectral features of its MO-TMS derivative (Fig. 2B). The molecular ion at m/z 591 (m/z 618, d,-TMS) and ion fragments at m/z 561 (m/z 587, d,-TMS), 470 (m/z) 488, d,-TMS), 380 (m/z 389, d,-TMS) and 143 (m/z 152, d,- TMS) were consistent with the proposed structure. Interestingly, d,-TMS labelling (Fig. 2C) showed that the ion at m/z 290 originates from two different fragmentation pathways. As illustrated m Figs. 2 (B,C), the successive elimination of a methoxy radical and three molecules of trimethyl- silanol gave rise to an ion at m/z 290 which was

not shifted in the mass spectrum of the d,-TMS derivative. Conversely, specific cleavage of the Cio-Cl1 and C&i4 bonds gave also rise to a fragment ion of m/z 290 comprising A- and B-rings which was shifted to m/z 299 in the d,-TMS derivative. On the basis of the mass spectral properties of the TMS ether and MO-TMS derivatives, compound 2 was assigned the structure 2-hydroxymethyl-lla,l7b-dihydroxy-17a- methylandrosta-1,4-diene-3-one.

Metabolzte 3. This metabolite was isolated from the unconjugated and acidic steroid fraction (Fig. 1B). This suggested that it probably arose from oxidation of the 2-formyl group to the corresponding carboxylic acid. This hypothesis was supported by the preparation of the corresponding TMS ester-TMS ether and methyl ester-TMS

u 70

: 60 zl c 50 ,

Q 40

30

-

I

A OTMS

3iI0 Mars/Charge

0 70

: 60 m

7) c 50 3

: 40

30

OTMS

143

428

200 300 Mass/Chcrqe

Fig 3. Electron-Impact mass spectra of metabobte 3 as the TMS ester-TMS ether (A) and methyl ester-TMS ether (B) denvatwes.

STUDIES ON ANABOLIC STEROIDS

ether derivatives, the mass spectra of which are consistent with the proposed structure (Fig. 3). Furthermore, these mass spectral features were identical to those of the corresponding derivatives of 11~17&dihydroxy-17a-methyl-3-oxo-androsta- 1,4-diene-2-carboxylic acid synthesized by selective oxidation of the 2-formyl group of formebolone.

The TMS ester-TMS ether derivative (Fig. 3A) exhibited a molecular ion at m/z 576 (m/z 603, d,-TMS) and fragment ions at m/z 220 (m/z 229, d,-TMS), 235 (m/z 244, d,-TMS) and 249 (m/z 258, d,-TMS) which were analogous to those formed by identical cleavage of B-ring bonds in the mass spectrum of the TMS ether derivative of compound 2 (Fig. 2A). The mass spectrum of the methyl ester-TMS ether derivative of compound 3

217

(Fig. 3B) was also consistent with the proposed structure. Interestingly, the formation of fragment ions of m/z 220, 235 and 249 suggested that intramolecular transesterification of the methyl ester function, most probably by the lla-OTMS group, occurs on electron-impact ionization prior to cleavage of the B-ring. This hypothesis was supported by the shift of these fragment ions to m/z 229, 244 and 258 as in the mass spectrum of the d,-TMS ester derivative (Fig. 3A). Other fragmentation routes gave rise to ions at m/z 339 (m/z 339, d,-TMS; M+‘- TMSOH - TMSO’) and 396 (m/z 405, d,-TMS; M+‘- MeOH + TMSOH) and were consistent with the proposed structure.

Metabollte 5. A prominent peak eluting at 28.3 min was observed in the selected ion chromato-

1 .2E4 316 + -, A I OTMS

143 TMSO,, --

10000 // I?

Me

-H : d? 143 8000 ----*

:, 0 ’

z1 :

6000 s

I? 4000 1 372

267

200: u,LiuJ;J2J63;63*;r_ 1 Tyi’ _ JY$’

200 300 400 Mats/Charge

Tz-------- B

2.0E5 OTMS M+’

1. BEST TMSO,, ‘// Me

1.4E5: I.y/c;L::J

534

--

1.6E5:

: 5 1.2E5: \ \ 5 z1 I: , 1.0E5:

TMSO

; B.OE4:

6 0E4: 389

4.0E4- 223 339 445 143 208

2.0E4 310 t 519 4 0 d’ s ..‘, ” ’ ’ .‘V

1,.1,...,.,L 1 w Ia i I” I : 1 ‘-

200 300 400 500 Mass/Charge

Fig. 4 Electron-Impact mass spectra of metabobte 5 as the TMS ether (A) and TMS enol-TMS ether (B) denvatwes.

218 R MA& ET AL

gram (Fig. 1A) obtained from repetitive scanning GC-MS of the free steroid excreted after administration of formebolone. The mass spectrum of its TMS ether derivative (Fig. 4A) showed a molecular ion at m/z 462 (m/z 480, +TMS) and fragment ions at m/z 143 (m/z 152, d,-TMS), m/z 405 (m/z 423, d,-TMS; M+‘- 57) and m/z 392 (m/z 410, d,-TMS; M+‘- 70) which are char- acteristic of D-ring cleavage and rearrangements of 17&hydroxy-17a-methyl steroids [13-U]. The molecular ion at m/z 462 and fragment ions at m/z 372 and 282 were suggestive of a a4-3-keto steroid with no substituent at the C, position. This hypothesis was supported by the formation of the corresponding TMS enol-TMS ether derivative, the mass spectrum of which showed a molecular ion at m/z 534 (m/z 552, d,-TMS- enol-d,-TMS ether mixed derivative, Fig. 4B) and fragment ions at m/z 445 [M+‘- TMSO]+, 444 [M - TMSOH]+’ and 389 (M+‘- 145). The latter ion comprises the A-, B- and C-rings and probably arises from D-ring fragmentation with con- comitant elimination of a hydrogen radical from the C-ring [13,16]. Compound 5 was identified as llcu,l7/3-dihydroxy-l7c~-methylandrost-4-en-3-one. Its identity was confirmed by comparison of the GC-MS properties of the TMS ether and MO- TMS denvatives of an authentic standard.

Degradation of compound 3 on trimethylsdylatlon Decarboxylation of /3-keto acids in acidic or

basic conditions is a well known reaction in the chemistry of organic compounds. The characteri- zation of 17~methyl-17P-hydroxy-5~androstan-3- one (mestanolone), a urinary metabolite of oxymetholone [17] in man, was reported lately [6]. The formation of this latter metabolite was ra- tionalized according to a metabolic pathway in- volving the oxidation of the 2-formyl group of oxymetholone to the corresponding P-keto acid which was then readily decarboxylated to mestanolone. No trace of the intermediate 2- carboxylic-3-keto acid was detected in urine after administration of oxymetholone, probably because of the great ability of this steroid to undergo decarboxylation. Interestingly, its homologous fl- keto acid 3 produced from formebolone metabolism was much more stable, most likely because of

the presence of a ~~~~ function in this metabolite. The methyl ester-TMS ether derivative was highly stable and no by-product was detected by GC-MS over a period of 24 h after derivatization. The corresponding TMS ester-TMS ether derivative was kept in hexane after evaporation of the solvent and was apparently stable under the GC-MS conditions used as shown in Fig. 1B. Surprisingly, upon prolonged standing at room temperature, this derivative progressively degraded to give derivatives 6 and 7 as shown in Fig. 5. These by- products were identified as the 11,17-di-0-TMS (M+ = 460) and 11-hydroxy-17-0-TMS (M+ = 388) derivatives of lla-hydroxymethandienone. Their identity was confirmed by comparison with the authentic steroid and their structures are illustrated in Fig. 6. GC-MS monitoring of the degradation reaction of 3 (Fig. 5) showed that compound 6 was the first by-product resulting from decarboxylation of the TMS ester function. After prolonged standing at room temperature, compound 7 was formed. By using llar-hydroxymethandienone as substrate, it was shown that derivative 6 was stable for more than 24 h under the experimental conditions used, which demon- strates that derivative 7 was not produced by degradation of derivative 6. This strongly suggests that the loss of the TMS group at C,, in derivative 3 occurs because of the presence of a TMS ester group at C,. The loss of the C,,-TMS group appears to be concerted with the elimination of the carboxyl group because no trace of a by-product that could have resulted from the selective elimination of the C,,-TMS group, without loss of the COOTMS group was detected either m the urinary extracts or in the derivatization mixture of authentic 3. Examination of a molecular model of 3 (Fig. 6) shows a favorable coordination between the TMS group at C,, and the carboxyl function at C, that could lead to the formation of derivative 7.

Interestingly, the degradation of the TMS ester derivative of 3 proceeded at a slower rate when derivation was carried out with a urinary extract, compared to the degradation rate observed with authentic 3. This suggests that derivative 3 is stabilized by interaction with some urinary com- ponents, thus decreasing its intrinsic ability to


.4 Ion 143-00 amu.

PL JC 1

26 28 30 32 34 36 38 40

Ion 143.00 amu.

?L, 26 20 30 32 34 36 38 40

c

26 2s 30 32 34 36 38 40 Time (m\n )

Fig. 5 Reconstructed ion chromatograms from GC-MS of a urine sample collected 6 h after formebolone admnustratlon. Free and acidic sterord- fractions were IsoiatedA and-dknvatized- with a mxture of&l~~A and- r&l- In pyndine. .me mxture was evaporated- to dryness and the residue dissolved m hexane. The progressive degradation of the TMS denvative 3 and the formatlon of the resuitmg liy-products 6and T were momtored 1.U Ii (A), K.5 Ii (Rj and W.U Ii (CT after d&vatlzation was initiated1 %e expenmentai for detals.

undergo degradation to derivatives 6 and 7. The mechanistic aspects of these reactions were not further investigated.

Formebolone metabolic routes The biotransformation routes of formebolone 1

are summarized in Fig. 7 In contrast to several reports on 17c+methyl anabolic steroids [2- 4,15,18], no metabolic resulting from hydroxyl-

ation of formebolone were detected in urine. Metabolism was solely directed at the A-ring, where the reactive formyl group was reduced to the corresponding 2-hydroxymethyl metabolite 2 which was isolated in the free and conjugated steroid fractions. Conversely, oxidation of the 2- formyl group gave the P-keto acid 3. The occurrence of llcu-hydroxymethyltestosterone 5 in urine

samples collected after formebolone ingestion sug-

1 Fig 6. Degradation route of the TMS ester-TMS ether denva-

twe of compound 3 on standing at room temperature m

hexane

gested that compound 3 was probably metabolized to the intermediate /3-keto acid 4 (not detected in this study) by regioselective reduction of the a’-function. Then, this acidic steroid would undergo decarboxylation to give compound 5. An alternative metabolic route could involve decarboxylation of 3 to give lla-hydroxymethandienone which would then undergo reduction of its al-function to yield 5. However, this metabolic route seems unlikely because no trace of llcw-hydroxymethandienone was detected in urine. In-

R MASSk ET AL

stead, this steroid was shown to be a degradation product of the TMS ester-TMS ether derivative 3 of the carboxylic acid 3 and not a compound resulting from the biotransformation of formebolone. Finally, no metabolites that could have resulted from epimerization at the C,,-position were detected. This observation was in accordance with the fact that no sulpho-conjugate of the metabolites reported in this study were isolated from urine [2,3,6].

In conclusion, it has been shown that formebolone is metabolized in man mainly by oxidation reduction reactions of its 2-formyl group. The 2-carboxylic-3-keto metabolite was further transformed by decarboxylation and reduction of its al-function to give lla-hydroxymethyltestosterone. No metabolite, the formation of whtch could have resulted from reduction of the A~-

and/or 3-keto group, epimerization at Cl7 or hydroxylation at other positions of the steroid was detected in this study. Unchanged formebolone was not detected in urine, thus indicating exten- sive first-pass metabolism and/or excretion in bile and feces and poor absorption of the steroid. These observations are in accordance with those of De Marchi et al. who reported similar findings about absorption and excretion of formebolone in rats [19]. The data presented above also indicated

2 -

- - - Fig 7 Proposed metabolic routes accounting for the occurrence of urmary metabohtes 2, 3 and 5 of formebolone

STUDIES ON ANABOLIC STEROIDS

that the methyl ester derivative should preferably be prepared to detect its acidic metabolite 3 so as to prevent the formation of the by-products which resulted from the degradation of the TMS ester denvative.

We thank the Natural Science and Engineering Research Counctl of Canada and the Sport Medi- cme Council of Canada for financial support. The authors are grateful to Dr. George Just, McGill University, for useful discussion about the synthesis of compound 3.

REFERENCES

Report of International Olympic Comnuttee Medical Com-

nusslon, Insbruck, 1974

R. Ma&, C Ayotte and R. Dugal, J Chromatogr , 489

(1989) 23, and references cited therem.

R Mass&, H B1, C. Ayotte and R Dugal, Blamed Environ.

Mass Spectrom., 18 (1989) 429.

R Mass& C Ayotte, H B1 and R. Dugal, J Chromatogr ,

497 (1989) 17

D Goudreault and R Mass& J Steroid Blochem., 37

(1990) 137

221

6 R. Mass& H. Bl, C Ayotte, P. Du, H G6hnas and R.

Dugal, J. Chromatogr., 562 (1991) 323.

7 L. Canonica, G. Jonmu, F Pehzzom and C. Scolastlco,

Gazz Chlm. Ital., 95 (1965) 138

8 A Vaccare and E Llvml, Mmerva Medxa, 62 (1971) 846

9 D Gel11 and E Vlgnatl, J Int. Med. Res., 4 (1976) 96

10 G De Mar&, G. d’Atn and C. Scolastlco, Arznelm. For-

sch. (Drug Res ), 26 (1976) 537

11 B S. Bal, WE. Chdders Jr., and H.W Pmmck, Tetra-

hedron, 37 (1981) 2091

12 S.J. Gaskell, A.G Snuth and C.J W Brooks, Blamed. Mass

Spectrom , 2 (1975) 148

13 D C Mammato and G.A. Eadon, J. Org Chem , 40 (1975)

1784

14 H W Durbeck and I Buker, Blamed Mass Spectrom., 7

(1980) 437

15 C J.W Brooks, A.R Thawley, P Rocher, B.S MIddleditch,

G M. Anthony and W.G Stdlwell, J Chromatogr SCI, 9

(1971) 35

16 L TBk6s, G Jones and C. DJerassl, J Am. Chem Sot , 90

(1968) 5465.

17 H.J. Rmgold, E. Batres, 0 Halpem and E. Necoechea, J

Am. Chem. Sot., 81 (1959) 427

18 H.W. Durbeck, I. Buker, B. Scheulen and B Tehn, J

Cbromatogr. SCI., 21 (1982) 405.

19 G. De Mar&, G d’Atn and C Scolastlco, Arznelm For-

sch. (Drug Res.), 23 (1973) 1583

Documents

Studies on anabolic steroids. VII Analysis of urinary metabolites of formebolone in man by gas chromatography—mass spectrometry