EUROPEAN JOURNAL OF DRUG METABOLISM AND PHARMACOKINETICS, 1983, Vol. 8, No 3, p. 269-280

Fluvoxamine maleate: metabolism.In man

H. OVERMARS, P.M. SCHERPENISSE, L.C. POSTDuphar - Research Laboratories Weesp, The Netherlands

Received for publication: June 3, 1982

Key words: fluvoxamine, antidepressant, 5-HT-uptake inhibitor, metabolism

SUMMARY

The metabolic fate of fluvoxamine maleate in man was investigated. The metabolites were isolated from the pooled urines ofhealthy volunteers who had ingested either 5 mg radioactive, or 100 mg non-radioactive fluvoxamine maleate as a single dose.The main isolation methods were solvent extraction, column and thin-layer chromatography.

Eleven metabolites were isolated; eight of these were carboxylic acids. Identification of nine metabolites was accomplishedby mass spectrometry supported by information from the UV spectra and the ionogenic properties. The main route of metabolicdegradation of fluvoxamine begins with oxidative elimination of the methoxyl group, another route with removal of the primaryamino group.

In view of the nature of the degradation pattern none of the metabolites is likely to possess psychotropic activity. For thetwo primary metabolites this has, in effect, been demonstrated.

INTRODUCTION

In a previous paper, De Bree et of. (I) reportedthe results of studies on the intestinal absorption andplasma kinetics of orally administered f1uvoxaminemaleate in healthy volunteers.

healthy volunteers who had ingested a single oraldose of 2 mg radioactive, or 100mg non-radioactivef1uvoxamine maleate.

MATERIALS AND METHODS

5-methoxy-4'-(trifluoromethyl)valerophenone(E)-0-(2-aminoethyl)oxime maleate (I: I)

Send reprint requests to : L.c. Post Duphar BV P.O.Box 2 1380 AA Weesp, The Netherlands.

The aim of the present work was to elucidate themetabolic pathway of f1uvoxamine in man. Thesource of metabolites for this study was the urine of

F.COC-CH -eH -CH -CH -O-CH3 II 22223

N,O-~-~-NH2

HCCOOH• II

HCCOOH

Source of metabolites

The urine pool consisted of the 8 hrs urine of tenhealthy volunteers, nine males, one female, each ofwhom had ingested 100mg f1uvoxamine maleate asa single dose (I). To this pool we added the urine offive healthy volunteers, collected during the first 8hrs after an oral 5 mg dose of [ 14C]f1uvoxaminemaleate (I). The total pool volume was 8.5 Iitres.

Fluvoxamine maleate, used as reference, waspharmaceutical grade material from our own re­sources. The compound found in the present studyto be the main metabolite (B; Fig. 4) of f1uvoxaminein man, was synthesized in our laboratories. Syn­thetic metabolite C 3 was an intermediate product inthe synthesis of f1uvoxamine. The solvents andreagents were of ultra-pure grade.

270 European Journal of Drug Metabolism and Pharmacokinetics. 1983, No 3

60:40:1575:15

100:170:30:170:30:1

90:150:50:1

Separation and isolation techniques

Measurement of radioactivity

Fluid samples were mixed with 10 ml scintillationliquid, consisting of toluene (3.6 I), Triton X-100(1.41), PPO (25 g) and POPOP (0.25 g). The radio­activity was measured in a Philips PW 4520 liquidscintillation counter, and the results were auto­matically corrected for quenching by the externalstandard channel-ratio technique.

Enzymatic hydrolysis of conjugated metabolites

The pooled urine was buffered to pH 5.0 andincubated for 72 h at 37DC under nitrogen with 1%v/v Sue d'Helix Pomatia (Industrie Biologique Fran­caise), stated to contain 310 ukat ~-g1ucuronidase

and 650 ukat sulphatase per ml,

Extraction of metabolites

Following the enzyme treatment the urine poolwas percolated over Amberlite XAO-2 (BOH) toextract the fluvoxamine metabolites. Before use theresin was washed with water, and then methanol,and stored as a slurry in water/methanol (1:1) at4DC.

A column of 60 X 4 cm was packed, and washedsuccessively with methanol (I I) and water (I I).Following percolation of the urine through thecolumn the latter was washed with 0.5 litre waterand most of the residual water was removed fromthe column in a stream of nitrogen. Then methanol(I I) was passed through the column to elute themetabolites. The eluate was concentrated to 20 mlby evaporation under reduced pressure at 30DC.

Column chromatography

This method was used for the preparative se­paration of the metabolites. Silica gel, 400 g (Merck,art. no. 7734), was packed into a 75cm X 4cm (in­ternal diameter) column as a slurry in chloroform.The concentrated XAO-2 eluate was mixed with 20 gof the same silica gel, and the solvent was evaporated.The mixture was then put on top of the column, andelution followed with the solvents chloroform-ethanol­ammonia 25% 60:40:2 (I I); 40:60:2 (I I); 20:80:2(I I),and methanol (I I). Fractions of 100 ml each werecollected.

Thin layer chromatography

Precoated silica gel 60 F 254 plates were used. Foranalytical separations they were of 0.25 mm layer­thickness (Merck, art. no. 5715); for preparativework we used 2 mm plates (Merck, art. no. 5717).The following solvent systems were used

I ethanol-chloroform-ammonia 25%2 hexane-diethyl ether3 diethyl ether4 diethyl ether-formic acid5 chloroform-methanol-formic acid6 chloroform-methanol-ammonia 25%7 dichloromethanol-methanol8 hexane-dioxane-formic acid

Samples were mostly applied as stripes. Thesolvent was allowed to run for about 15 em.

Radioactive bands were located on the driedplates by means of autoradiography on Agfa-GevaertStructurix 0-10 X-ray film. In preparative work theradioactive bands were scraped off the plates and thescrapings eluted with either diethyl ether on me­thanol, according to the polarity of the metabolites.

Identification methods

Derivatization

For gas chromatography-mass spectrometry, meta­bolites with alcoholic hydroxyl groups were silylatedwith trimethylsilyl imidazole (TSIM, Macherey andNagel), according to Horning et al. (2). The silylationreaction was allowed to proceed at room tempe­rature for half an hour. Carboxylic acids wereesterified with methanol containing 20% anhydroushydrogen chloride (3). This reagent was prepared bythe addition of I ml acetyl chloride to 4 ml ice­cooled anhydrous methanol, stirring continuously.The esterification reaction was complete after halfan hour at room temperature.

Ultraviolet spectrometry

Spectra of isolated metabolites dissolved in di­ethyl ether or methanol, according to the solubilityproperties of the compound involved, were recordedon a Beckman 2A spectrophotometer, in quartz cellsof 10 mm optical path.

H. Overmars et al.. Fluvoxamine metabolism in man 271

Mass spectrometry (MS); gas chromatography-massspectrometry (GC-MS)

Electron impact (EI) spectra were recorded on anAEI MS-30 double-beam mass spectrometer, underthe following cohditions.

ionisation potential 70 eVelectron current 300 J.tAacceleration voltage 4 kV

Samples were introduced with the direct insertionprobe.

For gas chromatography-mass spectrometry aPye 104 gas chromatograph was connected to theMS-30 system with a membrane separator as theinterface. A column of 100 em length and 3 mminternal diameter was packed with 3% SE 30 onChromosorb WHP; it was operated at 170°C.

Chemical ionization (CI) mass spectra were re­corded to gain supportive evidence on the molecularweight of certain metabolites. They were recordedon a Finnigan 3200 mass spectrometer connected toa Finnigan data system. The reagent gas was iso­butane. The samples were introduced with the directinsertion probe.

RESULTS

Isolation

The urine pool from the radioactive study con­tained about 23 umoles of f1uvoxamine metabolites.

urine pool

I. enzyme treatment2. extraction with XAD-2

We estimated the pool from the non radioactivestudy to contain 550-700 umoles of f1uvoxamineequivalents, on the assumption that the excretionprofile of the 100 mg dose used in the latter studywas similar to that of the 5 mg radioactive dose.This assumption is supported by the results of thecorresponding kinetic studies (I).

Scheme I is a summary of the purification andseparation steps leading to four fractions, the com­position of which is specified in Table I. Eachfraction was processed further as outlined in Schemes2-5.

Table I : Metabolite patterns of the XAD-2 eluate, andin the chromatographic fractions obtained ac­cording to Scheme I.Separation method: thin-layer chromatographyin ethanol-chloroform-ammonia (25%) 60:40:5.Percentages are of the radioactivity applied tothe chromatograms.

metabolite designation total

A B C+D E F+G % urnol

XAD-2 eluate 15 38 15 17 II 96

fraction I II II 2.65

II 4 32 36 8.71

III 7 27 34 8.30

IV 4 4 0.92

chromatographyon silica gel

discarded

fraction II; 36%

Scheme I : Fractionation of the fluvoxamine metabolites in human urine.The percentages are based on the radioactivity measurements; they are not corrected for incomplete recovery.

272 European Journal of Drug Metabolism and Pharmacokinetics, 1983, No 3

fraction Iin water

extractionwith CH2CI2; phl2

discarded

Scheme 2

aqueous; 3%metabolite B

TLC (7.3)

fraction IIin water

extractionwith CH2CI2; pH 2

TLC (7,2)

G; 25%

Schemes 2-5 : Separation and purification of the meta­bolites in each of the chromatographicfractions of Scheme I. Yields per step arepercentages of the original radioactivity asper fraction (see Table I). Numbers inparentheses bahind «TLC" refer to theelution systems.

added tofraction III

extractionwith 0.1 moU-1 NaOH

Schemel

aqueous; 7%metabolite B

added tofraction III

extractionwith CH2CI2; pH 2 discarded

H. Overmars et al., F1uvoxamine metabolism in man

fraction IIIin methanol

TLC (4)

273

Scheme 4

I. esterification2. extraction

with CH2CI2; pH 123. TLC (6,1)

fraction IVin water

I. esterification2. extraction wit

CH2CI2; pH3. TLC (2)

extractionwith 0.1 mol.l- I NaOH

extraction withethylacetate; pH 2

discarded

extraction withethylacetate; pH 2

Scheme 5

I. TLC (5)2. esterification3. TLC (7)

IdentificationSurvey of the data

Table II is a survey of certain molecular cha­racteristics of fluvoxamine and of its metabolites, as

established during or after their isolation. In ad­dition the share in the urinary radioactivity isrecorded. The quantities of the metabolites obtainedin the study with non-radioactive fluvoxamine wereestimated by assuming their moL extinction coeffi­cients to be equal to that of fluvoxamine, or the two

274 European Journal of Drug Metabolism and Pharmacokinetics. 1983, No 3

Table II : Molecular characteristics of fluvoxamine and of the metabolites isolated. Proportions of the metabolites in the urinepool.

percentage of ratioionogenic absorption urinary non-radioactive

designation and structure a R Fb behaviour maximum (nm) radioactivity to radioactive

fluvoxamine R-C-CH 2-CH rCH 2-CH 2-0CH 3 0.66 basic 256 c 0IIN

\C-CH 2-CH 2-NH 2

AI R-C-CH rCH 2-CH 2COOH 0.06 acidic 254 8 10IIN

\O-CH2-COOH

A 2 not identified 0.04 acidic 256 27

B R-C-CH 2-CH 2-CH 2COOH 0.14 amphoteric 254 d 35 35IIN

\C-CH rCH 2-NH 2

C, R-C-CH 2-CH 2-CH 2COOH 0.27 acidic 252 4 4IIN

\OH

C2 R-C-CH 2-CH 2-CH 2COOH 0.30 acidic 259 7 4IIN

\O-CHrCHrOH

CJ R-C-CH 2-CH 2-CH 2COOH 0.30 acidic 232 3 30II0

D R-C-CH 2-CH 2-CH 2COOH 0.37 acidic 259 8 63\IN

\O-CH rCH rNH-CO-eH J

E R-C-CH rCH 2-CH 2-CH rOCH J 0.43 acidic 254 10 8\I .N

\O-CH2-eOOH

F I R-C-CH rCH 2-CH 2-CH rOCH J 0.76 non-ionic 259 5 45II .N

\O-CH 2-CH 2-0H

F 2 not identified 0.76 non-ionic 259 5 2

G R-C-CH 2-CH rCH 2-eH 2-0CH , 0.82 non-ionic 248 5 10II .N

\OH

1I R = FJC

b solvent system: ethanol-chloroform-ammonia 25% 60:40:5c c = 11.500d c = 10,600

30 333

290

334

60

320

320

H. Overmars et al., Fluvoxamine metabolism in man

Table III : Mass spectrometric data of fluvoxamine and its metabolites.

derivatization a M+ M-F common ions

metabolite A 1 M 361 342

B M 333 b 313

CI M 289 270 272, 200, 198, 187, 172, 145

C2 M 333 314

D M 375 b 355

E M 347 328

F, none

TMS

F2 TMS258,226,200,198,187,172,145,71,45

G none

TMS 347 328

fluvoxamine none 299

metabolite C 3 M 274 255

a M = methylation; TMS =trimethylsilylationb protonated molecular ion

other characteristic ions

73

30, 276

243,242,215,188,173,145

275

CI

MH+)

276

other reference compounds. Table III contains asurvey of the characteristic ions in the mass spectraof fluvoxamine and its metabolites. Since meta­bolites A-E were esterified with methanol followingtheir isolation, their mass spectrometric data arethose of the methyl esters.

The mass spectrum of fluvoxamine is depicted inFig. I, that of the esterified main metabolite, B, inFig. 2, and that of C 3 in Fig. 3. The mass spectra ofall the metabolites and of fluvoxamine itself con­tained the ion m/z 145. This mass evidently re­presents the trifluoromethyl phenyl ion; hence thismoiety was retained unchanged in all the meta­bolites.

Two groups of metabolites can be distinguishedby having a number of ions in common (Table III).Of one group, encompassing the metabolites E-Gand fluvoxamine itself, the common ions m/z 256,226... probably originate in the fragmentation of thefollowing structure

The other group, comprising A )-0, is characteri­zed by m/z 272. The ions at m/z 200, 198, 187 and172 are also present in the spectra of E-G and offluvoxamine. In addition the spectra of the me­thylated metabolites A 1-0 often contained an ad­ditional ion (M-31)" probably due to the methoxylgroup.

The structural fragment inferred for the me­thylated metabolites A ,-0 is as follows :

o1\

-0\' PJ2-cH2-cH2-e-OCHFC \ C 3

3 _ 'N

Metabolite C 3 (Table III; Fig. 3) stands by itself,having only m/z 145 in common with fluvoxamineand the other metabolites.

All the spectra containing a molecular ion showedin addition a weak peak at M-19, evidently due toloss of one fluorine atom.

FcO'C-CH -CH -CH -CH -O-CH33 ,,2 2 2 2N

'O-CH -CH -NH2 2 :2

European Journal of Drug Metabolism and Pharmacokinetics. 1983, No 3

20"T"""--.,.------------------------------,

276

10O

908070

60

50.........'cR. 40<;»

30»+> 20-.-I{J} 10~Q) 0

+>s:: 100-.-I

Q)!':10

:> 80-.-I+> 70ellr-l 60Q)

~ 50<40302010

0

Fig. 1 : Mass spectrum of fluvoxamine

F:cO'C-CH -CH -CH -g-OCH3 II 2 2 2 3N

\O-CH.,..CH -NH

222

70

60

5040

302010

0~TYflmTrl.lflnrTTrlf'tJrnm'rl'r"tTT'iJiltTT1rrl'ir+rTTTTriIlTrT1'r+MJmT'f"rTTT_mrTT1nT1nTl"rrrr

100-r- --.

9080

»+>-.-I{J}

s:lQ)

+>s:: 40-.-I 1C0-r---,- ----: -=- ,;;.;:.,.:..-_......:,~

807060

50<40302010

0-Ly-nT"rPt-T'tfJt'ffTTt'rlt'fIrT"TTtJfTf!tTTT't1''I-nm-r'hTn'''''-rr'rTtrtTrrn-r't&t-rT't'Tn''rTTn-rI~rr1

Fig. 2 : Mass spectrum of methylated metabolite B

H. Overmars et al.. Fluvoxamine metabolism in man 277

Notes on individual metabolites

The chemical structures derived for the meta­bolites are depicted in Table II.

A I : This compound gave an acid reaction in theisolation procedure; hence the amino group waslacking. The absorption maximum at 254 nm, closeto that of fluvoxamine, suggests that the oxime ethergroup was still present. Combination of this in­formation with the mass spectral data of the methylester of A I, in Table III, leads us to the proposedstructure.

A 2 : The mass spectrum of the methyl ester ofthis acidic compound yielded no interpretable in­formation. There was not enough material to identifythis compound.

B : In the isolation procedure this compoundreacted amphoterically. The mass spectrum of itsmethyl ester (Table III) shows, besides the pro­tonated molecular ion and M-19, the set of fragmentions common to the metabolites A 1-0, and inaddition strong peaks at m/z 290 and 30. The latterions originate in, respectively, the loss of C 2-H4N

from the molecular ion and the formation of CH 2== +NH2. These phenomena are further evidence forthe presence of the primary amino group in meta­bolite B.

The chemical ionization mass spectrum confirmsthat the molecular weight was 332. The chemicalstructure inferred for metabolite B was confirmed bycomparison of its UV and mass spectral data withthose of the synthetic compound (Fig. 2).

C I : This compound was an acid. Its Ama x , at252 nm, was lower than that of fluvoxamine ormetabolite B, indicating that in C I the aminoethylgroup is absent. According to the data in Table III,the methyl ester of C belongs to the group of meta­bolites containing the same arylalkyl group as A I

and B. Its molecular weight of 289 was confirmed bychemical ionization mass spectrometry.

C2 : According to the data concerning its methylester, in Table III, this compound had the samearylalkanecarboxylic group as A" Band C I. Its ab­sorption maximum was at 259 nm, suggesting thatthe oxime ether function was intact. The acidiccharacter of C 2 indicated that the amino group was

FcO'C-CH -CH -CH - ~-()CH .3 II 2 2 2 3o

40'" I 1'1 .IrI i I~~ I I ,'', i '~'f"'~'f"'1-rI,'r,Tj''r,-il l't,'T"T..,..,.-r-.-,-,r-r"1...rrrt't"TTrrTTTT~,.,n"1rT"'r_r'm/z

10090 *S

SO70GOSO40

302010

0

Q)

>--

>..p·rlCIls::Q)

.p 80.~ 100..,- ---,

908070605040302010o

Fig. 3 : Mass spectrum of methylated metabolite C3

FcO

'C

-CH

--e

H-C

H-e

H-O

-CH

31

12

22

23

N

'O-C

H-C

H-N

H2

22

IV -..l

00 .... ~ ~ "'"~ ... .g ~ ::l ~ s:: ~ ~ ~ to;,

~ ~ ~ f I:l ::l l:o.

~ I:l ~ I:l i. ~ ~~.

@FcO

'C

-CH

-eH

-CH

-g-O

H3

II2

22

-N

'OH

FcO

'C

-CH

-eH

-CH

J-O

H3

II2

22

-a

FcO

'C

-CH

-CH

-CH

-CH

-O-C

H3

II2

22

23

N

'OH

®

FC

O'

C-C

H-e

H-C

H-C

H-O

-CH

3II

22

22

3N, O

-CH

-C-O

H2

II a

®

FC

O'

C-C

H-e

H-C

H-C

H-O

-CH

31

12

22

23

N

'O-C

H-C

H-O

H2

2

/@FLUV

OIAM

INE

/

FC

O'

C-C

H-C

H-C

H-g

-OH

3II

22

2N, O

-CH

2-CH

2-OH

® @FC

O'

C-C

H-e

H-C

H-~

-OH

3II

22

2N, O

-CH

-CH

-NH

22

2

FC

O'

C-C

H-e

H-C

H-g

-OH

3II

22

2N

'O-C

H-C

H-N

H-C

-CH

22

II3

o@

)

FC

O'

C-C

H-e

H-C

H-C

H-O

-CH

('3

II2

22

23

N

'O-C

H-C

H-N

H-C

-CH

22

II3

o

eFC

O'

C-C

H-C

H-C

H-g

-OH

3II

22

2N, O

-CH

-C-O

H2

II a

@

Fig

.4

:P

ropo

sed

met

abol

icpa

thw

ays

offl

uvox

amin

ein

man

.

H. Overmars et al.. Fluvoxamine metabolism in man 279

absent. The molecular weight was 333, confirmed bychemical ionization.

C3 : The mass spectrum of the methyl ester ofthis metabolite was different from those of flu­voxamine or the other metabolites, in all respectsexcept for m/z 145 of the trifluormethyl phenylmoiety. The Ama x 232 nm, much lower than that offluvoxamine, indicates hydrolysis of the oxime etherto the corresponding ketone. The mass spectrum ofthe methyl ester of C 3 (Table III) contained, besidesthe molecular ion at m/z 274, ions at m/z 243, 242and 2I5, evidently resulting from expulsion of CH 30,CH 30H and COOCH 3 from the molecular ion.This is a characteristic feature of methyl esters. Theevidence quoted in favour of the ketone function inC 3 is supported by m/z 173, evidently representingthe trifluoromethylphenone ion. The mass spectrumof the synthetic compound (Fig. 3) confirmed theproposed structure.

D : This metabolite also had acidic properties.The oxime ether function was probably intact, sinceAmax was 259 nm. The mass spectrum of the methylester shows that 0 contained the same arylalkylmoiety as metabolites A I, B, C 1 and C 2. Besides,the spectrum shows a quasi-molecular ion at m/z375, and several additional masses which providedthe clue to the structure of the oxime ether sidechain. High resolution mass spectrometry revealedthe fragment-ions m/z 60, 86 and 102 to belong to,respectively, C 2H6NO, C 4HgNO and C 4H gN0 2. Inview of the origin of these ions, the following stru­ctures present themselves in the same order :

HO+=CH-CH rNH 2, +CHrCH rNH-CO-CH 3and

HO +=CH-CH 2-NH-CO-CH 3

Furthermore, high resolution mass spectrometryconfirmed the formula C I7H 22N 20 4F3 for the pro­tonated methyl ester.

E : This metabolite also was acidic in character.The mass spectrum of its methyl ester (Table III)had a number of fragment-ions in common withfluvoxamine, and with the group of metaboliteshaving the methyl ether side chain retained intact.Hence the metabolic change in metabolite E was inthe oxime ether side chain. The Amax at 254 nmindicates that the oxime ether function itself wasretained. The molecular weight of 347 for the methylester then leads to the structure for E shown inTable II.

F I : This compound showed no ionogenic pro­perties. Its absorption maximum was at 259 nm in-

dicating that the oxime ether function was intact. Itsmass spectrum had the same fragment-ions in com­mon with fluvoxamine as had metabolite E. Themass spectrum of the silylated compound showedtwo additional fragment-ions at m/z 103 and 117;these correspond to CH 2=OSi(CH 3) 3 andCH rCH rOSi(CH 3) 3 and indicate that F containeda hydroxyl group. The quasi-molecular ion at m/z320 in the chemical ionization spectrum means thatthe molecular weight is 319.

F2 : This metabolite also was non-ionogenic. Itsmass spectrum was quite similar to that of F z- butthe intensity of the peaks was not sufficient topermit elucidation of its structure.

G : This was another non-ionogenic metabolite.Its mass spectrum contained the fragment-ions com­mon to fluvoxamine and the metabolites E, F andF 2' The mass spectrum of silylated G, showed inaddition a molecular ion m/z 347. The chemicalionization spectrum of G itself showed a quasi­molecular ion at m/z 276. The structure of G asinferred from these data (Table II) was confirmed bycomparison with the synthetic oxime.

DISCUSSION

We isolated eleven, and identified nine fluvox­amine metabolites from the urine of healthy vo­lunteers. These metabolites accounted for 85% of theradioactivity in the urine pool. The available data donot permit a precise quantitative account of themetabolites following the 100-mg dose of non­radioactive fluvoxamine maleate. Our estimates re­garding the non-radioactive material were based onthe assumption of equal molar absorption coefficientsfor fluvoxamine and its metabolites. This assumptionwas tested with only three of the metabolites, namelyB, C 3, and G, which had been synthesized.

The results in Table II reveal a large variety inthe ratios of unlabelled to labelled. Yet no tendencyemerges towards a systematic difference related todose level.

Our present results provide no information aboutthe occurrence of conjugated fluvoxamine meta­bolites; in unpublished work we have found evidenceto the effect that less than 10% of the urinary radio­activity was in conjugated form.

Fig. 4 shows the metabolic pathway of flu­voxamine in man as it emerges from the presentwork. It is of interest that none of the metabolites islikely to possess psychotropic activity; for the twoprimary metabolites (B and F I; Fig. 4) this wasactually demonstrated (F.J. Hillen, unpublished).

280 European Journal of Drug Metabolism and Pharmacokinetics, 1983, No 3

REFERENCES

I. De Bree, H., van der Schoot, J.B. and Post, L.c.(1983): Fluvoxamine maleate: Disposition in man.Europ. J. Drug. Metab. Pharmacok. 8,175-179.

2. Horning, M.G., Moss, A.M., and Horning, E.C. (1967):

A new method for the separation of the catecholaminesby gas-liquid chromatography. Biochim. Biophys. Acta148, 597-600.

3. Knapp, D.R. Handbook of Analytical DerivatizationReactions. John Wiley and Sons, New York 1979, p.152.