DRUG DISPOSITION Clin Pharmacokinet 1998 Apr; 34 (4): 281-302

0312-5963/98/0004-0281/$22.00/0

© Adis International Limited. All rights reserved.

Metabolism of the
Newer Antidepressants
An Overview of the Pharmacological and
Pharmacokinetic Implications
Silvio Caccia
Istituto di Ricerche Farmacologiche ‘Mario Negri’, Milan, Italy

Contents
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
1. Selective Serotonin Reuptake Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
2. Serotonin and Noradrenaline Reuptake Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
3. The Selective Noradrenaline Reuptake Inhibitor Reboxetine . . . . . . . . . . . . . . . . . . . . . . 288
4. The Monoamine Uptake Inhibitor and 5-HT 2 Receptor Antagonist Nefazodone . . . . . . . . . . . 290
5. Tianeptine, a Serotonin Reuptake Enhancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
6. The Noradrenergic and Specific Serotoninergic Agent Mirtazapine . . . . . . . . . . . . . . . . . 294
7. The Reversible Inhibitor of Monoamine Oxidase Moclobemide . . . . . . . . . . . . . . . . . . . . 294
8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296

Summary Several chemically unrelated agents has been developed and introduced in the
past decade, to supplement the earlier antidepressants. These include inhibitors
of the reuptake of serotonin [the selective serotonin reuptake inhibitors (SSRI)]
or noradrenaline (reboxetine) or both (milnacipran and venlafaxine), as well as
drugs with distinct neurochemical profiles such as mirtazapine, nefazodone,
moclobemide and tianeptine. Like the earlier drugs, these newer antidepressants
are almost totally biotransformed before excretion, except for milnacipran whose
clearance appears to be due equally to both urinary excretion and metabolism.
Sometimes – as in the case of moclobemide – up to 20 metabolites have been
identified in body fluids. In some cases, however, only a few metabolites have
been detected, and a substantial proportion of the dose remains unaccounted for
(e.g. fluoxetine and fluvoxamine).
Metabolism generally proceeds through sequential or parallel oxidative path-
ways. These may be affected to varying degrees by physiological and pathological
factors and those mediated by cytochrome P450 (CYP) 2D6 and CYP2C19
through genetic polymorphism. Some are influenced by chirality (e.g. the dealk-
ylation of citalopram and fluoxetine), although information on this aspect of
disposition is still lacking for other drugs existing as racemates (e.g. mirtazepine
and tianeptine) and milancipran, which is probably a mixture of 4 stereoisomers.
Others again are saturable within the therapeutic range of doses (e.g. some path-
ways of metabolism of fluoxetine, fluvoxamine, nefazodone, paroxetine and
venlafaxine). This may explain the individual variability with all these drugs
282 Caccia

which, from the pharmacokinetic point of view, is the same as with tricyclic
agents.
Our knowledge of the isoenzymes involved in the various oxidation pathways
and their relevance for potential drug interactions varies from a considerable
amount for most of the SSRI and nefazodone, to minimal for reboxetine and
tianeptine. This information is useful for predicting the pharmacokinetic interac-
tions mediated through inhibition of specific isoenzymes. This would be better
appreciated if the enzymatic mechanisms involved in the biotransformation of
the metabolite(s), and their role in drug interactions, were also known. This in-
formation is still lacking for some drugs, although metabolites may exhibit in
vitro inhibitory potencies of similar to (paroxetine and its M2 metabolite as in-
hibitors of CYP2D6) or even greater than that of the parent drug (norfluoxetine
is more potent than fluoxetine as an inhibitor of CYP3A3/4, and in view of the
longer half-life (t1⁄2) of the metabolite the potential for interactions may persist
for weeks after discontinuation of the parent drug).
While we do know something about the biological activity of the metabolites
of some of these drugs, we know very little about others. With few exceptions
this knowledge refers only to the major metabolite(s) and regards the main in
vitro effects as exerted by the parent drug. However, in vitro potency and selec-
tivity may not translate directly into in vivo, and either major or minor metabolites
may have characteristic in vitro and in vivo properties. For example, unlike the
parent drug some minor ring-opened metabolites of moclobemide have mono-
amine oxidase-B inhibitory activity in the rat, and the nefazodone metabolite
m-chlorophenyl-piperazine shows activity on 5-HT 2C receptors in rats and
humans.
Data on the brain-to-blood partition of metabolites compared with their parent
drug are available only in a few cases. They are still not known for the main
metabolites of fluvoxamine, milnacipran, mirtazapine, moclobemide, nefazo-
done, paroxetine, reboxetine and venlafaxine, despite the fact that total blood
concentrations do not always reflect the metabolite : parent drug ratio in brain.
Thus, in most cases, we do not really know what part hepatic metabolism plays
in the overall effect of the administered parent drug.

Supplementing tricyclic antidepressant deriva-
tives (TCA), monoamine oxidase inhibitors (MAOI)
and second generation antidepressants,[1-4] a whole
new generation of chemically and neuropharmaco-
logically unrelated antidepressant agents has been
introduced and marketed over the past decade, in-
cluding several of the drugs listed in table I. While
most of these drugs offer new approaches to the
treatment of depressive disorders, their therapeutic
equivalence is not yet apparent, although they are
considered generally safer and better tolerated than
the older drugs.[5,6]
The pharmacokinetics of the older antidepres-
sant drugs – the term ‘old’ obviously referring only
to the date of introduction – have been described in
a number of reviews, specifying the characteristics
governing the frequency of drug administration
and the time to reach steady state, possibility of
therapeutic drug monitoring, genetic polymor-
phism of metabolism and potential for drug-drug
interactions.[7-10] The importance of the formation
of metabolites and their possible contribution to the
final outcome has been underlined.[11-14] This is
because most of these drugs are almost entirely
eliminated from the body after biotransformation,
often yielding active metabolites. These may be-
have either similarly to or differently from the par-
ent drug as regards pharmacokinetics and pharma-

© Adis International Limited. All rights reserved. Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants 283

codynamics, sometimes resulting in a situation teristics but all are extensively biotransformed be-
which is no different in principle from polytherapy. fore excretion, initially through oxidative path-
Although exhaustive reviews have been pub- ways. However, metabolism apparently has a va-
lished on the comparative clinical pharmacokinet- riety of effects on the pharmacological activity of
ics of some of the newer antidepressants,[15-19] data the parent drug (table II). Thus, fluvoxamine is
on the in vivo biotransformation of others and the metabolised by oxidation, oxidative deamination
pharmacological significance of their metabolites and hydrolysis to many metabolites, 9 of which
are generally scarce and fragmentary. Here we re- have been identified and constitute 65% of the uri-
view the metabolism of the newer antidepressants nary excretion products. Only 2 have been tested
with emphasis on metabolites that may contribute for their effect on monoamine uptake mechanisms,
to primary and secondary pharmacological actions and showed no or only weak inhibition of serotonin
of the parent drug. uptake (i.e the carboxyl acid derivative) in compar-
ison with the parent drug.[20,28]
1. Selective Serotonin The specific isoenzymes involved in fluvoxam-
Reuptake Inhibitors ine metabolism have not been identified but in vitro
and in vivo the drug potently inhibits the metabo-
The 5 selective serotonin reuptake inhibitors
lism of substrates for CYP1A2.[18,29-32] The asso-
(SSRIs) currently available (citalopram, fluvox-
ciation between fluvoxamine metabolism and the
amine, fluoxetine, paroxetine and sertraline) differ
activity of CYP1A2 is supported by the lower
in chemical structure and pharmacokinetic charac-
fluvoxamine plasma concentrations in smokers
than nonsmokers,[33] smoking being known to in-
Table I. The newer antidepressant agents and some pharmacolog- duce the metabolism of drugs catalysed by this
ically related compounds isoenzyme. However, fluvoxamine also inhibits
Mechanism of action Drug the metabolism of typical substrates of CYP2C19,
Selective serotonin reuptake inhibition Citalopram
Fluoxetine
CYP3A4, CYP2D6 and CYP1A1.[18,29-31,34-37] Flu-
Fluvoxamine voxamine exhibits nonlinear kinetics which are not
Paroxetine due to Michaelis-Mentem saturation of a single
Sertraline
metabolic pathway, but to a complex metabolic
Serotonin and noradrenaline reuptake Duloxetine
inhibition Milnacipran picture involving multiple parallel pathways.
Venlafaxine The metabolic clearance of fluvoxamine is af-
Selective noradrenaline reuptake inhibition Reboxetine fected by gender (with higher plasma concentra-
Nisoxetine
Teniloxazine tions in women with depression than in men with
Tomoxetine depression)[38] and liver disease (cirrhosis). It does
Serotonin uptake and 5-HT2 receptor Nefazodone not appear to change in relation to age and renal
inhibition Trazodone
function after single doses.[17,18,20,28] However, all
Serotonin reuptake enhancement Tianeptine
5-HT1A receptor-partial agonism Buspirone these studies have focused on the parent drug.
Gepirone The main metabolites of paroxetine are also
Ipsapirone reported to have very weak in vitro inhibitory ac-
Selective cAMP phosphodiesterase Rolipram
inhibition
tivity on serotonin (and catecholamine) uptake
α2-Adrenergic auto- and hetero-receptor Mirtazapine mechanisms. These include an unstable catechol
blockade intermediate and its methylated and glucuronide-
Reversible and selective inhibition of Brofaromine and sulphate-conjugated derivatives.[21,29] Paroxet-
monoamine oxidase type A Cimoxatone
Moclobemide ine metabolism is partly under the pharmacoge-
Toloxatone netic control of CYP2D6[34,40-43] but another en-
Abbreviation: cAMP = cyclic adenosine monophosphate. zyme with lower affinity for the drug may also be

© Adis International Limited. All rights reserved. Clin Pharmacokinet 1998 Apr; 34 (4)
284 Caccia

Table II. Comparative in vitro potencies of selective uptake inhibitors and their main active metabolites, with mean metabolite-to-parent drug
ratios at steady state
Drug Main active Serotonin uptake Main enzyme inhibited Steady-state
metabolite inhibitorsa CYP2D6 CYP3A3/4 CYP1A2 CYP2C19 metabolic ratio
Citalopram +++ + 0.3-0.4b
Norcitalopram ++ ++
Fluoxetine +++ +++ + ++
Norfluoxetine ++/+++c +++ ++ ++ 0.6-1.7b
Fluvoxamine +++ + ++ +++ +++
Carboxy acid + NA NA NA NA NA
Paroxetine +++ +++
M2 metabolite + +++ NA
Sertraline +++ +/++
Norsertraline ++ +/++ 1.6-2.1b
a Only compared with parent drug.[20-26]
b Mean values in different populations.[18,27]
c The R-isomer is weaker than the parent drug isomers or its own S-isomer.
Abbreviation and symbols: NA = not available; +++ = substantial; ++ = moderate; + = weak.

involved.[42] This enzyme has not yet been identi-
fied but it may be inhibited by cimetidine, decline
in efficiency with age and be induced by some anti-
convulsants.[44]
At therapeutic doses paroxetine causes substa-
tial inhibition of CYP2D6 activity in vivo.[15-18,29,30,45]
One of its metabolites (designated M2) may con-
tribute to these actions, being a potent inhibitor of
this isoenzyme in vitro.[34] Paroxetine also inhibits
its own clearance on repeated administration, al-
though only in extensive, not in poor, metabolisers
of sparteine. The lack of effect on the clearance of
desipramine in poor metabolisers confirms that this
drug is a selective inhibitor of CYP2D6, which is
lacking from the liver of poor metabolisers.[42,43]
Clearance of the parent drug is influenced by aging,
severe renal impairment and hepatic disease, with
higher plasma concentrations and slower elimina-
tion than in healthy volunteers,[21,39] but the phar-
macokinetics of its main metabolites are not known
for any of these populations.
Citalopram, fluoxetine and sertraline are all
initially demethylated to their nor-derivatives
demethylcitalopram, norfluoxetine and norsertral-
ine, which are also SSRI (see table II). Unlike the
other nor-derivatives, demethylcitalopram is fur-
ther N-demethylated to di-demethylcitalopram,
which retains weak serotonin reuptake inhibiting
activity and reaches lower concentrations than the
parent drug and its main metabolite in human
plasma.[22] All the parent compounds and their nor-
derivatives are further biotransformed to a number
of metabolites, many of which remain to be identi-
fied.[22,46] Norcitalopram may amount to about
40% of the mean parent drug steady-state plasma
concentration (table II), with large interindividual
variability, and its elimination t1⁄2 (t1⁄2β) is more than
twice that of the parent drug.[17,18,47] However, it
has poor blood-brain barrier penetration.[22] In rats
given repeated citalopram 10 mg/kg interperitone-
ally twice daily for 14 days, the plasma concentra-
tions were comparable with those of the parent
drug but concentrations in the brain were less than
10% (within the first hours after administration),
suggesting that it was unlikely to contribute signif-
icantly to the pharmacological activity in this spe-
cies (Caccia S., unpublished results).
The formation of norcitalopram is at least partly
under the genetic control of CYP2C19, being sig-
nificantly lower in Caucasians who were poor me-
tabolisers of mephenytoin than in those who were
extensive metabolisers.[47,48] The subsequent step
to di-demethylcitalopram appears instead to be under
the genetic control of CYP2D6. Levomepromazine,

© Adis International Limited. All rights reserved. Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
a potent inhibitor of this enzyme, raises the steady-
state plasma concentrations of norcitalopram dur-
ing citalopram administration (see Baumann[18] for a
review). Norcitalopram may mediate the mild inter-
action potential of citalopram with drugs metabo-
lised by the CYP2D6 isoenzyme, being one order
of magnitude more potent as an inhibitor than the
parent drug (see also table II).[17,30]
Using Japanese human liver microsomes and
recombinant human CYP isoforms, Kobayashi et
al.[49] found that CYP2C19 was only a minor form
and CYP3A4 the major isoenzyme involved in the
first step of citalopram metabolism. The discrep-
ancy between the in vivo and in vitro findings has
been tentatively explained by ethnic differences in
the frequency of individuals heterozygous for the
extensive metabolisers genotype. Citalopram,
however, is a racemic compound, whose disposi-
tion is affected by stereoselectivity, with the more
potent SSRI S-(+)-citalopram achieving lower
steady-state concentrations than the R-(–)-enantio-
mer,[50] with possible age-related changes in
stereoselective metabolism of the S-isomer.[51]
In vitro studies with human microsomes show
that both CYP2C19 and CYP3A4 are involved in
the preferential N-demethylation of S-(+)-citalo-
pram.[52] Fluvoxamine, possibly by inhibiting
these enzymes, preferentially raises the concentra-
tion of the S-isomer, changing the S/R ratio in
dextromethorphan/mephenytoin phenotyped pa-
tients given repeated doses of citalopram.[53]
CYP2D6 is also involved in this step but with the
opposite stereoselectivity.[52]
Fluoxetine and norfluoxetine are also chiral
compounds, existing as R-(–)- and S-(+)-isomers.
While there are no major differences in pharmaco-
logical potency between R- and S-fluoxetine,[23]
the S-isomer of norfluoxetine is consistently more
effective than the R-isomer as an inhibitor of sero-
tonin uptake in vitro and ex vivo in animals, and
is possibly responsible for the long-lasting seroto-
nin uptake inhibition in vivo.[24] Stereoselectivity
in fluoxetine demethylation and/or subsequent
pathways has been observed, the R-isomer being
© Adis International Limited. All rights reserved.
285
metabolised faster than S-fluoxetine in vitro and in
vivo in humans.[54,55]
Norfluoxetine retains the pharmacological
properties of fluoxetine, and depending on the
species, experimental test and conditions may con-
tribute to, or even account for, the actions seen with
the parent compound.[25,56-59] In patients with de-
pression it achieves steady-state concentrations
comparable with, or even higher than, those of the
parent compound, accounting for a substantial
part of the therapeutic and adverse effects of the
drug.[16-18,60] Its formation is partly under the genetic
control of CYP2D6, being higher in extensive met-
abolisers than in poor metabolisers of debrisoquin
and dextromethorphan.[61]
Fluoxetine interacts in vitro with known
CYP2D6 substrates.[34,40,62] In this respect the S-
isomers of fluoxetine and norfluoxetine have sim-
ilar potency but are about 5 times more potent than
the R-isomers in vitro in human microsomes.
Norfluoxetine, however, is more potent than fluo-
xetine as an inhibitor of CYP3A3/4 and, because
of its unusually long t 1⁄2β, the risk of drug-drug
interactions persists, for a long time after fluoxet-
ine therapy is discontinued. The pharmacokinetics
and clinically relevant interactions between fluo-
xetine and coadministered drugs have been pre-
sented in a number of recent reviews.[16-18,29-31,63-67]
Unlike fluoxetine, CYP2D6 does not seem to
play an important role in the demethylation of
sertraline,[61] which is possibly dependent on
CYP3A isoenzyme.[18,30,44] However, sertraline
does inhibit CYP2D6 in vitro, although less po-
tently than fluoxetine and paroxetine.[29,30,63-67]
Dose-dependent increases in plasma concentra-
tions of desipramine have been observed after its
addition to therapy (50 to 150 mg/day), although
the interaction is weaker than with fluoxetine and
paroxetine at the usually effective therapeutic
doses.[17,18,29,44,67-70] Sertraline 150mg as a single
or repeated dose also reduces the clearance of
imipramine.[70] Norsertraline is approximately
equipotent to sertraline with respect to inhibition
of CYP2D6 in vitro[30,44] but reaches higher
steady-state plasma concentrations and is elimi-
Clin Pharmacokinet 1998 Apr; 34 (4)
286
nated more slowly than the parent compound (t1⁄2β
3 to 5 days compared with 1 to 2 days for the parent
compound, with some gender and age differ-
ences).[17,27] Norsertraline possibly plays a deci-
sive role in the in vivo inhibitory effect of the parent
drug.[31,66,68,71] Sertraline and norsertraline are weaker
inhibitors of CYP3A3/4 isoenzymes, and evidence
suggests that clinically significant interactions
with CYP3A substrates are unlikely.[72]
As an in vitro inhibitor of neuronal serotonin
uptake, norsertraline is less potent than the parent
drug.[26] However, the in vivo activity of the nor-
metabolite and its contribution to the parent drug’s
effects is less clear. Thus, norsertraline was inac-
tive in the mouse Porsolt test, a model of depres-
sion.[26] At doses of 1 to 10 mg/kg the nor-metabolite
did not increase extracellular serotonin or inhibit
the spontaneous firing rate of dorsal raphe neurons
in rat, 2 parameters affected dose-dependently by
sertraline (1 to 10 mg/kg administered subcutane-
ously), and possibly other SSRI as a consequence
of the block of serotonin reuptake.[73] However, at
higher doses, norsertraline (32 mg/kg administered
intraperitoneally) lowers brain 5-hydroxyindole-
acetic acid (5-HIAA) in rats, as do all SSRIs.
Norsertraline dose-dependently antagonised p-
chloroamphetamine-induced serotonin depletion,
although less potently than the parent compound in
the rat [intraperitoneal the dose causing 50% antag-
onism (ED50), 20 and 1 mg/kg] but almost as po-
tently as sertraline in mice (ED50, 6 and 4 mg/kg);
in this species the metabolite brain concentrations
after sertraline were comparable with those after
norsertraline.[74]
The nor-metabolite fully retains the ability of
sertraline to cause anorexia in rodents, a property
shared by other SSRI and in general by agents that
increase serotonergic transmission.[75]
Taken together these findings suggest that
norsertraline may play a role in the pharmacologi-
cal actions of the parent drug, though this depends
on the in vivo effect studied and experimental con-
ditions and animal model considered. However, it
is still not clear to what extent norsertraline con-
tributes to the therapeutic action of sertraline.
© Adis International Limited. All rights reserved.
Caccia
Compared with young (between the ages of 18
and 65 years old) volunteers, exposure to both the
parent drug and its nor-metabolite is generally
higher in elderly (between the ages of 65 and 75
years old) individuals for citalopram, fluoxetine
and sertraline.[6,17-18,44] For sertraline there is also
an age-gender interaction, plasma concentrations
of the parent drug and its nor-derivative in young
men are lower than in young women or the el-
derly.[27]
For all drugs the t1⁄2β is doubled in patients with
hepatic cirrhosis compared with healthy volun-
teers, so these agents should be used cautiously in
patients with liver disease.[17,44,76] Citalopram
clearance is reduced and the t 1⁄2β is increased in
patients with renal impairment.[17] Although full
doses of fluoxetine and sertraline have been used
in renally impaired patients without marked phar-
macokinetic changes,[26,58] caution is still advis-
able because these studies have been conducted
after single doses (and fluoxetine at least shows
nonlinear behaviour with repeated dose adminis-
tration, which may be magnified by renal impair-
ment) and the multiple-dose pharmacokinetics of
their more polar metabolite(s) is poorly known in
this population.[44]
2. Serotonin and Noradrenaline
Reuptake Inhibitors
The recently introduced serotonin and nor-
adrenaline reuptake inhibitors, milnacipran and
venlafaxine, differ structurally from selective in-
hibitors of serotonin or noradrenaline uptake. From
the point of view of the putative mechanism of ac-
tion, these drugs resemble the tertiary amine tri-
cyclic antidepressant agents (TCA), since they af-
fect both serotonin and noradrenaline reuptake.
Milnacipran almost equipotently inhibits these
mechanisms without significantly affecting dopa-
mine transporters.[77] Venlafaxine inhibits synapto-
somal serotonin more than noradrenaline uptake
and slightly inhibits dopamine uptake.[78] A com-
parative review of the neuropharmacological
profile of these and pharmacologically related
compounds such as duloxetine (whose develop-
Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
Table III. Effects of liver and renal impairment on the disposition of milnacipran and venlafaxine. The dose of milnacipran and venlafaxine
was 50mg
Drug Main metabolic pathway
Milnacipran Glucuronide conjugation
Venlafaxine Oxidative O-demethylation
a Compared with age-matched healthy individuals.
b CLCR values ranged from 0.6 to 5.1 L/h.[76]
c CLCR ranged from 1.8 to 4.2 L/h for the patients with mild impairment to 0.6 to 1.8 L/h for those with moderate impairment.[81,82]
Abbreviations: CL = total body clearance; CLCR = creatinine clearance.
ment as an antidepressant has been withdrawn) has
been published recently.[79]
Milnacipran, a 1-aryl-2-(aminomethyl)cyclo-
propane-carboxylic acid derivative, shares some
structural similarities with the ‘old’ monoamine
oxidase (MAO) inhibitor tranylcypromine.[80] The
disposition of this drug, which very likely is a mix-
ture of stereoisomers, has largely been studied us-
ing the radiolabelled compound and seems to differ
from most other antidepressants because its clear-
ance is due equally to urinary excretion and meta-
bolism. About 50 to 60% of the dose is recovered
unchanged in urine and most of the remainder as
its glucuronide conjugate (approximately 20%), its
N-dealkylated derivative and the corresponding
glucuronide conjugate after single doses in healthy
volunteers.[76]
In plasma, most of the radioactivity results from
milnacipran and its conjugate, N-dealkylmilnaci-
pran reaching only very low concentrations. No
data are available for concentrations of milnaci-
pran and its norderivative in tissues, including the
brain, relative to blood. However, N-dealkylmil-
nacipran is reported to be inactive in vivo in animal
models predictive of antidepressant activity.[80]
Liver insufficiency (cirrhosis) does not change
the pharmacokinetics of milnacipran 50mg either
orally or intravenously). Renal impairment pro-
gressively reduces the total clearance, plasma con-
centrations of the parent drug and its glucuronide
conjugate rising 2- to 4-fold after a single oral dose
(table III). The mean t1⁄2β of milnacipran is longer
© Adis International Limited. All rights reserved.
287
Disease Outcomea
Hepatic No changes in disposition
Renalb 2- or 4-fold increase in parent drug and conjugate metabolite
levels (and t1⁄2)
Hepatic CL is reduced by 50%
Renalc 25 to 50% reduction in CL with increase in t1⁄2 for the drug
and its active metabolite
in patients with renal impairment (9 to 25 hours
compared with about 8 hours for healthy individu-
als) who, therefore, need some adjustment in the
dosage and frequency of administration.[76]
Slightly higher milnacipran plasma concentra-
tions (approximately 20%) and prolonged t1⁄2β
have been observed in elderly patients given a sin-
gle oral dose of milnacipran 50mg compared with
healthy individuals, this is probably a consequence
of some loss of renal function in the elderly. Be-
cause the differences are small the manufacturer
does not recommend dosage adjustment in these
cases, except for those with impaired renal func-
tion.[76] No information is available for its metabo-
lites in this population.
Levomepromazine (a CYP2D6 inhibitor) slightly
raises milnacipran steady-state concentrations (by
approximately 20%) without appreciably affecting
its metabolic profile whereas carbamazepine,
which is metabolised by CYP3A3/4, stimulates
milnacipran clearance, the plasma concentrations
of N-dealkyl milnacipran rising by 50%.[76]
Venlafaxine is a phenethylamine bicyclic de-
rivative, which exists as a racemic mixture of the
(R)- and (S)-enantiomers, both of which inhibit
monoamine uptake in rat synaptosomes.[78] How-
ever, (–)-venlafaxine inhibits reuptake of serotonin
and noradrenaline, whereas the (+)-isomer primar-
ily inhibits serotonin reuptake. The pharmaco-
kinetics of the single isomers have been studied
after oral doses of the racemate and were stereo-
selective in rats but not in dogs or humans.[83]
Clin Pharmacokinet 1998 Apr; 34 (4)
288
O-Demethylation of (racemic) venlafaxine to
O-desmethyl-venlafaxine is the predominant route
of metabolism in humans, with about 60% of the
dose appearing in the urine as such or its glucuron-
ide conjugate. Other minor routes are N-demethyl-
ation of the alkylamino side chain to N-desmethyl-
venlafaxine and the loss of both the O-methyl and
1 of the N-methyl groups to form N,O-desmethyl-
venlafaxine.[84] Among these and other metabolites
identified in animals, the O-demethyl derivative re-
tains pharmacological activity in vitro (monoamine
uptake blockade) and in vivo (reversal of reserpine
hypothermia, induction of pineal β-adrenergic sub-
sensitivity) comparable with the parent drug in
rats. The N-desmethyl-derivative is also active in
these in vivo tests but is a weaker inhibitor of sero-
tonin and noradrenaline uptake than venlafax-
ine.[85]
The O-demethylation of venlafaxine is partially
mediated by CYP2D6, whereas N-demethylation
is probably catalysed by CYP3A3/4 isoenzymes.[86,87]
Venlafaxine, however is a less potent inhibitor of
CYP2D6 enzymes in vitro than SSRIs.[88,89] Al-
though at the usual therapeutic doses venlafaxine
should not greatly affect the biotransformation of
drugs transformed by CYP2D6, these might affect
the O-demethylation of venlafaxine and/or one of
its isomers.
The parent drug and its O- and N-desmethyl de-
rivatives are weak inhibitors of CYP3A isoforms.[87-89]
Accordingly, venlafaxine has no appreciable ef-
fects on the pharmacokinetics of diazepam[90]
whose main routes of metabolism may partly in-
volve CYP3A3/4.[91] Venlafaxine and metabolites
are also weak inhibitors of CYP1A2 activity in
vitro.[87,89,92]
After single and repeated doses in young and
elderly volunteers, O-demethyl venlafaxine ap-
pears rapidly in plasma and far exceeds the plasma
concentrations of the parent drug. The mean t 1⁄2β of
the metabolite (9 to 11 hours) is about twice that of
the parent drug (2 to 7 hours in healthy individu-
als,[81] and a sustained release formulation is now
on the market in several countries).
© Adis International Limited. All rights reserved.
Caccia
This metabolite is formed more slowly, depend-
ing to some extent on renal function in renally im-
paired individuals and patients undergoing haemo-
dialysis. The apparent total clearance of both the
drug and its main metabolite is reduced by 55% in
end-stage renal failure, and the t1⁄2β is more than
doubled for both compounds.[82] Similar decreases
in clearance have been observed in patients with
cirrhosis (table III). All these patients required a
reduction in dosage of 50%.[81]
3. The Selective Noradrenaline
Reuptake Inhibitor Reboxetine
Reboxetine is a recently introduced drug. Chem-
ically it is (RS)-2-[(RS)-α-(2-ethoxyphenoxy)]-
benzylmorpholine methan-sulphonate (fig. 1) and
shares some chemical similarities with viloxazine,
and to a lesser extent fluoxetine and structurally
related compounds. It exists as a racemic mixture
of the S,S-(+)- and R,R(–)-enantiomers[93] both of
which inhibit noradrenaline uptake in vitro in rat
hypothalamic synaptosomes and prevent reser-
pine-induced blepharospasm in vivo in rodents, the
S,S- approximately 20 times as potently as the R,R-
isomer.[94,95]
After single oral doses of reboxetine in mice and
rats the mean (R,R) : (S,S) brain concentration ra-
tios [in terms of the peak plasma drug concentra-
tion (Cmax) and area under the concentration-time
curve (AUC)] are about 2 and 4 respectively, sug-
gesting that the pharmacological effects of reboxet-
ine in rodents are mainly due to the more active
(S,S)-enantiomer. The mean brain : plasma AUC
ratio for both isomers is higher in mice than in
rats,[95] confirming that differences may exist in the
brain uptake of centrally acting drugs.[96,97] In hu-
man plasma the (R,R) : (S,S) isomeric ratio is sim-
ilar (about 2) to that in mouse plasma after single
oral doses of each isomer or the racemate. In
healthy individuals the mean t1⁄2β of the 2 isomers
is comparable (about 13 hours after reboxetine
4mg or 2mg of each isomer).[98]
The metabolism of oral reboxetine and its re-
covery in excreta have been studied in humans and
animals using the radiolabelled drug. Reboxetine
Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
CI O
NH
CH3
Fluoxetine
H3 C O CH3
CI O
O O
NH
Reboxetine
Fig. 1. Chemical structures of some monoamine reuptake inhibitory antidepressants.
is extensively biotransformed before excretion in
all species with no, or only trace amounts of, un-
changed drug excreted with urine. The main met-
abolic routes seem to include hydroxylation of the
ethoxyphenoxy ring, oxidative O-dealkylation,
oxidation of the morpholine ring and glucuronide
conjugation of the phase I metabolites, with differ-
ences between species.[99]
Information is limited on the enzymatic mech-
anisms that govern these metabolic pathways in
humans which, however, do not seem to involve
CYP2D6 because quinidine, a potent inhibitor of this
enzyme, does not affect the disposition of reboxet-
ine isomers. Reboxetine 2mg twice daily given to
healthy volunteers for 5.5 days does not appear to
significantly alter the CYP3A3/4-mediated 6β-
hydroxylation of endogenous cortisol.[100] At the
same dose it does not change the pharmacokinetics
of lorazepam,[98] a short-acting benzodiazepine
mainly eliminated by conjugation. Reboxetine is
reported to have negligible activity as an inhibitor
of CYP 1A2, 2C9, 2E1 and 3A4 isoenzymes in vitro
in human microsomes.[98] The effects of its met-
© Adis International Limited. All rights reserved.
289
NH2
O= CH2-CH3
N
CH2-CH3
Milnacipran
CH3
N
HO
CH 3
Venlafaxine
abolites on the activity of these drug-metabolising
enzymes are still not known.
The total oral clearance of racemic reboxetine
is low and mainly because of nonrenal mechanisms;
it may be markedly decreased in very old people
(between 66 and 98 years old) compared with
young individuals (see also table IV).[98] Oral
reboxetine clearance is unaltered after multiple ad-
ministration in both young and elderly individuals.
However, exposure to reboxetine is higher and
much more variable in elderly patients with de-
pression than in younger volunteers, as assessed by
the steady-state AUC values. A reduction in the
total daily dosage from 8 to 6mg is recommended
for elderly patients.[98]
The mean plasma Cmax of reboxetine is similar
in healthy volunteers and in patients with alcoholic
liver disease after a single 4mg dose. The mean
AUC is slightly higher (24 to 36%, depending on
the degree of impairment) and mean is t1⁄2β longer
than normal (21 to 31 hours vs 12 to 14 hours) in
those patients with alcoholic liver disease.[98,100,101]
Systemic exposure to reboxetine (and renal
clearance) is progressively affected by renal im-
Clin Pharmacokinet 1998 Apr; 34 (4)
290 Caccia

Table IV. The clearance of reboxetine in young and elderly individuals

Age (years) Gender and number CLpo (L/h) CLR (L/h)
20-39a 54m 1.18-2.42 0.126-0.336
66-98 4m, 8f 0.39-0.63 0.006-0.054
a Data are from four separate studies; reboxetine dose ranged from 1.98 to 4.15mg, as free base.[98]
Abbreviations: CLpo = total oral clearance; CLR = renal clearance; f = female; m = male.

pairment, with mean AUC and renal clearance
dropping with the increasing severity of the dis-
ease. A lower daily dosage is recommended in
patients with severe renal impairment.[98]
Although no detailed pharmacokinetic studies
of the main reboxetine metabolites are available in
these populations, in the plasma of healthy volun-
teers unchanged reboxetine mostly appears to re-
flect the circulating radioactivity (70% in terms of
AUC).[101]
4. The Monoamine Uptake Inhibitor and
5-HT2 Receptor Antagonist Nefazodone
This phenoxyethyl triazoline phenylpiperazine
derivative (fig. 2) is structurally related to
trazodone. It has potent antagonistic effects on 5-
HT2A receptors and moderate serotonin and nor-
adrenaline reuptake inhibiting properties, which
together may modulate serotoninergic neurotrans-
mission through postsynaptic receptors, particu-
larly 5-HT1A receptors.[102]
Nefazodone too is extensively biotransformed
in the liver, with the formation of 4 main met-
abolites and several minor ones.[103] The main
metabolites are hydroxynefazodone, p-hydroxy-
nefazodone, a triazoledione derivative and m-chloro-
phenyl-piperazine (mCPP). This latter results from
cleavage of the phenyl-piperazine side chain and is
a metabolite common to trazodone[104] and struc-
turally related compounds.[105] Many other cen-
trally acting drugs containing an arylpiperazine
side chain, including the potential antidepressants
buspirone, ipsapirone and gepirone,[106,107] yield
1-arylpiperazines during biotransformation in hu-

CI C-NH-CH2-CH2-N O
N
N O

N
CH3

Mirtazapine Moclobemide

C 2H5
CH3
N
CI SO2 N
N N
OCH2CH2 (CH2)3 N N

O CH
CI
NH (CH2 )6 COONa

Nefazodone Tianeptine
Fig. 2. Chemical structures of some antidepressants with neurochemical mechanisms other than monoamine reuptake inhibition.

© Adis International Limited. All rights reserved. Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
mans and/or animals. These metabolites concentrate
in brain tissues contributing to and, or accounting
for, a substantial part of the central effects of the
parent compound.[105,107]
The metabolism of nefazodone is at least partly
mediated by CYP3A4.[102,108,109] The drug and its
metabolites hydroxynefazodone and p-hydroxy-
nefazodone inhibit this isoenzyme in vitro[108] and
impair the in vivo clearance of concomitantly ad-
ministered carbamazepine, cyclosporin and triazolo-
benzodiazepines (i.e alprazolam and triazolam)
which depend on 3A4 activity for their metabo-
lism.[109] Nefazodone also impairs its clearance by
increasing the single dose and during repeated ad-
ministration (see table V).[109]
The pharmacokinetics of nefazodone and
hydroxynefadozone do not differ in extensive or
poor metabolisers of dextromethorphan, and in
vitro nefazodone and its main metabolite are weak
inhibitors of dextromethorphan 2D6-mediated O-
demethylation. However, the steady-state plasma
Cmax and AUC of mCPP were substantially higher
in poor metabolisers than in extensive metabolisers
(table V), and its t1⁄2β increases from 6.1 hours in
extensive metabolisers to 16.4 hours in poor meta-
bolisers. This suggests that the subsequent metabo-
lism of mCPP, which involves ring hydroxylation
before conjugation and urinary excretion, is medi-
ated by CYP2D6.[110] Therefore, drugs that inhibit
this enzyme might reduce the clearance and raise
the concentrations of mCPP. This was observed
when nefazodone was co-administered with fluox-
Table V. Plasma area under the concentration-time curve (AUC) of nefazodone and its active metabolites after single and repeated doses
in extensive (EM) and poor metabolisers (PM) of dextromethorphan[110]
Compound Dosage
Nefazodone Single dose
Steady state
Hydroxy-nefazodone Single dose
Steady state
m-Chlorophenyl-piperazine Single dose
Steady state
a Plasma AUC after a single dose (200mg) and over the dosage interval (AUC0-12h) at steady state (200mg twice daily). The metabolite-
to-parent drug ratios are shown in parentheses.
b Statistically significant.[110]
© Adis International Limited. All rights reserved.
291
etine, the persistence of the inhibitory effect pos-
sibly being related to the long half-life of fluoxet-
ine, and particularly norfluoxetine.[111] Likewise
the addition of the CYP2D6 inhibitor thioridazine
40 mg/day to trazodone 150 to 300 mg/day raises
the plasma concentrations of mCPP by about
50%.[112]
Hydroxynefazodone shows affinity for 5-HT2A
receptors and serotonin uptake sites comparable
with nefazodone in rat cortex. The triazoledione
derivative is much less active than the parent com-
pound – or even inactive – in these in vitro systems,
whereas no information is available about p-
hydroxynefazodone.[102]
mCPP displays potency similar to nefazodone
for serotonin reuptake inhibition and low affinity
for 5-HT2A receptors.[102] However, this well
known serotoninergic agent causes a number of
behavioural and endocrinological changes in ani-
mals and humans. Using selective agonists, partial
antagonists and antagonists for serotonin receptor
subtypes and families, some pharmacological ef-
fects of mCPP, such as the increase in plasma
prolactin and other hormones, oral dyskinesias and
hyperthermia, reduction of food intake and anxiety-
like behaviours, have been attributed to its action
on post-synaptic 5-HT2C receptors, where it acts as
a partial agonist.[113,114]
In humans mCPP serves as an interesting probe
of serotoninergic mechanisms that may modulate
anxiety. Given acutely to healthy individuals and
patients with panic or obsessive-compulsive disor-
Plasma AUC (μmol/L • h)
EM PM
2.60 ± 0.77 2.45 ± 0.99
13.57 ± 3.64 15.85 ± 1.77
0.72 ± 0.30 (0.28) 0.80 ± 0.44 (0.31)
4.53 ± 1.04 (0.33) 5.61 ± 2.71 (0.35)
2.09 ± 1.77 (0.80) 8.34 ± 1.43 (4.00)b
1.97 ± 1.98 (0.15) 8.66 ± 3.04 (0.55)b
Clin Pharmacokinet 1998 Apr; 34 (4)
292
ders it produces manifestations of anxiety and
changes in many physiological functions, although
with repeated administration tolerance rapidly
arises to these effects.[115]
In healthy volunteers the hydroxynefazodone :
nefazodone ratio in plasma (metabolic ratio on a
g/L basis) averages 0.3 to 0.4 regardless of the dose
and schedule of treatment, supporting pharmaco-
logical findings that this metabolite may contribute
to the final outcome.[102,109,116]
Plasma concentrations of nefazodone and hydr-
oxynefazodone, in terms of Cmax and AUC, are
higher in the elderly than in younger volunteers
after single and repeated doses, and in elderly
women than in elderly men after repeated doses,
but the metabolic ratio remains relatively constant,
although with wide interindividual variability.[116-118]
The mean Cmax and AUC of nefazodone and hydr-
oxynefazodone are also significantly higher and
the t1⁄2 longer in patients with liver impairment than
in healthy individuals after 50 to 100mg twice
daily, these patients need a lower daily nefazodone
dosage.[117,119]
The mCPP : nefazodone ratio tends to decrease
with higher doses and longer duration of treatment,
indicating that the formation of this metabolite is
saturable.[102-117] At steady-state, exposure to mCPP
is comparable for males and females and young and
elderly individuals (extensive metabolisers) with
the metabolite : parent drug ratios averaging 0.1 : 0.2
in all groups, on a molar basis.[116] Exposure to
mCPP and the metabolic ratios are not altered sig-
nificantly by renal impairment[102,117-118] though they
tend to increase in patients with biopsy-confirmed
hepatic cirrhosis, compared with healthy individ-
uals.[119] As mentioned, steady-state Cmax and AUC
of mCPP and the mean metabolic ratio are mark-
edly higher in poor metabolisers than extensive
metabolisers (see table V).[110]
Although exposure to mCPP appears to be low
in most patients, there is no information on the
metabolite : parent drug ratio at the site of action,
at least in preclinical use. However, considerable
individual differences in exposure to nefazodone,
but particularly to its metabolite mCPP, are appar-
© Adis International Limited. All rights reserved.
Caccia
ent in all populations. Thus in different individuals
there may be a different interplay among the neuro-
chemical mechanisms characteristic of each com-
pound, with possible differences in antidepressant
activity and adverse effects.
5. Tianeptine, a Serotonin
Reuptake Enhancer
Unlike antidepressants which facilitate post-
synaptic effects of central neurotransmitters by
blocking their reuptake (i.e. the drugs mentioned
above) or by inhibiting their catabolism (i.e. MOAIs),
tianeptine increases presynaptic serotonin reuptake,
thereby reducing synaptic availability of the neuro-
transmitter.[120,121] This original neurochemical ac-
tion has been observed ex vivo in rat cortical and
hippocampal synaptosomes and in rat and human
platelets after single and repeated doses of tianept-
ine. Possible explanations of how this action is
involved in its antidepressant activity were reviewed
by Ansseau[122] and, more recently, by Wilde and
Benfield.[123]
Chemically it is a modified tricylic [i.e. the so-
dium salt of 3-[(3-chloro-6-methyl-5,5-dioxo-6,11-
dihydrodibenzo[c,f(1,2-thiazepin)]11-yl)amino]-
heptanoic acid (see fig. 2)], sharing some similari-
ties with amineptine whose structure also contains
a long aminoheptanoic acid side chain. Metabo-
lically too it is similar to amineptine, being rapidly
and extensively biotransformed in humans and an-
imals, mainly by β-oxidation to shorter chain de-
rivatives which appear in plasma, tissues and urine.
With tianeptine, the pentanoic acid derivative
(MC5) is the major metabolite which is further ox-
idised to a propionic acid (MC3) and a lactam de-
rivative (LMC5). A minor route includes N-dealk-
ylation of the heptanoic acid side chain to a number
of intermediate compounds and subsequent modi-
fication of the remaining tricyclic moiety.[124]
Tianeptine is a racemate and the antidepressant-
like properties associated with its effect on sero-
tonin disposition appear to depend on the (–)-
isomer.[125] However, there are no studies in the
literature on the pharmacokinetics and metabolism
Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants 293

of the 2 isomers after administration with either the
racemate or the single isomers.
Tianeptine metabolism may include the forma-
tion of reactive metabolite(s) which bind covalently
to liver microsomal proteins both in vitro and in
vivo, in the mouse, hamster and rat. This activation
also occurs in vitro in human microsomes and ap-
pears to be mediated by the glucorticoid-inducible
CYP3A3 but not by CYP2D6 or CYP1A1 isoen-
zymes. The reactive metabolite(s), however, have
no direct hepatotoxic potential in the hamster at
less than a very high, sublethal dose of tianept-
ine.[126]
The main oxidised metabolite MC5 is reported
to retain antidepressant activity,[123] although de-
tails on the test, species and the potency relative to
the parent drug are not available. In rats given
single and repeated neuropharmacologically effec-
tive doses of tianeptine, the metabolite reaches
lower plasma Cmax but higher AUC than the parent
compound because its elimination is slower. The
same is true for the brain concentrations, though
brain uptake is slightly lower for the metabolite
than the parent drug.[127] Thus the metabolite
should play a decisive role in the central effects of
tianeptine, if its activity is at least comparable.
After single oral doses of tianeptine in healthy
volunteers (12.5mg) the metabolite MC5 reaches
mean plasma Cmax and AUC slightly lower than the
parent drug, with a mean metabolite : parent drug
ratio averaging 0.6 to 0.8 (on a mg/L basis because
molecular weights are similar). This ratio remains
essentially unchanged after repeated administra-
tion although the plasma AUC of both compounds
are slightly higher than after a single dose. After
the usual therapeutic doses in adults with depres-
sion (12.5mg 3 times daily) steady-state plasma
AUC of both compounds are again higher than af-
ter a single dose but the active metabolite (t 1⁄2 14
hours) accumulates more than the parent drug (t1⁄2
6.3 hours), so the steady-state metabolic ratio rises
compared with after acute administration[123] (see
table VI).
Compared with young volunteers, exposure to
both tianeptine and M5 is higher in the elderly with
or without depression. The clearance of tianeptine
is reduced (by approximately 40% after intra-
venous administration) and the t1⁄2β prolonged
(from 4.7 to 9 hours). A reduction in the daily dose

Table VI. Mean plasma area under the concentration-time curve (AUC) of tianeptine and its active M5 metabolite in different populations.
The elderly individuals were aged from 73 to 86 years.[128] The renally impaired patient group comprised 12 patients with CLR <1.14 L/h, 6
patients requiring chronic haemodialysis.[130] The mean metabolite-to-parent drug ratio (on a mg/L basis) is shown in parentheses
Individual Dosage (mg) AUC (mg/L • h)
tianeptine MC5
Healthy adults 12.5 po 0.7 ± 0.3 0.5 ± 0.3 (0.8)
12.5 po 0.7 ± 0.3 0.4 ± 0.1 (0.6)
12.5 tid × 4 days 1.1 0.8 ± 0.2 (0.7)
Adults with depression 12.5 po 1.0a 0.6a (0.6)
12.5 tid × 1 month 1.3a,b 1.4a,b (1.1)
Elderly 12.5 po 1.3 ± 0.2a 1.1 ± 0.3a (0.9)
12.5 IV 1.5 ± 0.3a
1.2 ± 0.6a (0.9)
Elderly person with depression 12.5 1.1 ± 0.5 1.2 ±0.4 (1.0)
12.5 × 17 days 1.3 ± 0.5b 1.1 ± 0.8 (0.8)
Patient with renal impairment 12.5 po 0.8 ± 0.3 1.1 ± 0.6c (1.4)
Patient with cirrhosis 12.5 po 0.8 ± 0.3c 0.4 ± 0.1c (0.4)
Patient with chronic alcoholism 12.5 po 0.5 ± 0.1c 0.2 ± 0.1c (0.4)
[123,128]
a Not compared with healthy individuals.
b Steady-state AUC was significantly greater than after single dose.[129]
c Significantly different from healthy volunteers.[130,131]
Abbreviations: IV = intravenously; po = orally; tid = 3 times daily.

© Adis International Limited. All rights reserved. Clin Pharmacokinet 1998 Apr; 34 (4)
294
(to 12.5mg twice daily) has been recommended in
very old patients.[123,128,129]
Chronic renal failure does not appear to greatly
affect the pharmacokinetics of the parent drug but
the plasma AUC (see also table VI) and t1⁄2β of the
metabolite (14.2 vs 4.9 hours) are higher than in
healthy individuals after a single 12.5mg dose of
tianeptine. Reducing the dosage to 12.5mg twice a
day has been recommended for these patients
too.[130] Dosage reduction is not likely to be neces-
sary in patients with combined depression and al-
tered liver function since no great difference (pa-
tients with alcoholism and cirrhosis) or even a
slight reductions (alcoholism without cirrhosis)
have been observed in tianeptine and M5 plasma
AUC compared with individuals with normal liver
function.[131]
6. The Noradrenergic and Specific
Serotoninergic Agent Mirtazapine
This is the only antidepressant that increases
noradrenergic and serotoninergic (5-HT1-medi-
ated-) neurotransmission through a blockade of
central α2-adrenergic auto- and hetero-recep-
tors.[132,133] Chemically, mirtazapine is 1,2,3,4,
10,14b-hexahydro-2-methyl-pyrazino[2,1-a]-pyrido-
[2,3-c]benzazepin(Z)-2 butenedioate (fig. 2), the
6-aza-derivative of mianserin and is a racemate of
the R- and S-isomers which show quantitative and
qualitative differences in their in vivo and in vitro
α2-adenoceptors antagonist properties. The com-
parative pharmacological profile of the racemate
and its enantiomers has been described by de Boer
et al.[134] and reviewed by Davis and Wilde.[135]
Information is minimal about the disposition of
racemic mirtazapine. The drug undergoes an exten-
sive first pass metabolism and clearance is mostly
via hepatic metabolism, urinary excretion account-
ing for only about 4% of the administered dose.[132]
The main metabolic pathways in humans are N-
demethylation, N-oxidation and 8-hydroxylation
followed by conjugation.[136,137] A preliminary in
vitro study using human liver microsomes and cells
expressing different enzymes suggests that CYP2D6
(and to a lesser extent CYP1A2) mainly catalyses
© Adis International Limited. All rights reserved.
Caccia
the 8-hydroxylation of mirtazapine, while the other
2 oxidative processes are catalysed by CYP3A4.
Although mean oral clearance and t1⁄2β of racemic
mirtazapine (and t1⁄2 of its demethylated metabo-
lite) are very similar in the poor and extensive me-
tabolisers of debrisoquine after a single oral dose
of mirtazapine (7.5 and 15mg in poor and extensive
metabolisers, respectively) the possible stereo-
selective dependendence of mirtazapine and nor-
mirtazepine metabolism on this polymorphic en-
zyme cannot be excluded.[66,137] Mirtazapine does
not appear to alter the pharmacokinetics of desipra-
mine (CYP2D6) and diazepam (CYP2C19 and
CYP3A3/4).[132,135]
The demethylated metabolite retains pharmaco-
logical activity (3 to 4 times less active than mirta-
zapine)[132] although, again, details on the test and
species are not available. Its plasma AUC is ap-
proximately 50% that of the parent drug. The
plasma t1⁄2β of the parent drug (mean 22 hours,
range from 13 to 34 hours in young healthy indi-
viduals)[135] and the demethylated metabolite (about
19 hours)[137] are similar after single oral doses.
Although the Cmax and t1⁄2β of mirtazapine in el-
derly patients with cirrhosis are only slightly
higher than in matched controls, the hepatic clear-
ance is approximately 33% lower and dosage ad-
justment may be necessary in individuals with
altered liver functions.[135]
7. The Reversible Inhibitors of
Monoamine Oxidase Moclobemide
While the older monoamine oxidase inhibitor
nonselectively and irreversibly inhibit both MAO-
A and MAO-B isoenzymes, the newer generation
preferentially (and reversibly) affects the MAO-A
subtype, the enzyme system primarily responsible
for deamination of those amines believed to be im-
plicated in the aetiology of depression. This new
class includes brofaromine, cimoxatone and mo-
clobemide. The neuropharmacological profiles of
cimoxatone[138] and moclobemide[139,140] have
been reviewed. Only moclobemide is considered
here. Information is minimal on the metabolism of
cimoxatone which, however, is reported to be
Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
Table VII. Some pharmacokinetic parameters of moclobemide and its metabolites Ro 12-8095 and Ro 12-5637 in healthy individuals and
hepatically and renally impaired patients[142-145]
Participants and dose (mg) Parameter
Healthy individuals (100) CLpo (L/h)
Cmax (mg/L)
Patients with cirrhosis (100) CLpo (L/h)
Cmax (mg/L)
Patients with renal impairment (100) CLpo (L/h)
Cmax (mg/L)c
Healthy individuals (100 tid for 15 days) CLpo (L/h)
Cmin (mg/L)
a Significantly different from healthy individuals.
b First day of dose administration (n = 12).
c In parentheses, the coefficient of variation (n = 13, with varying degrees of renal impairment, CLR 0 to 2.4 L/h).
Abbreviations: CLpo = oral clearance; Cmax = maximum plasma drug concentrations: Cmin = minimum steady-state concentration; tid = 3 times
daily.
extensively biotransformed to its active O-demeth-
ylated derivative.[138] The development of brofaro-
mine has been stopped.[141]
Moclobemide is a benzamide derivative con-
taining a morpholine ring as its main structural
characteristic (p-chloro-N-[2-morpholinoethy]
benzamide) [fig. 2]. It has intermediate-to-high
clearance and is subject to a substantial first pass
metabolism, resulting in low and variable oral bio-
availability. Clearance is essentially metabolic and
decreases as the dose increases and after repeated
administration (table VII), possibly because of
auto-inhibition or metabolite-induced inhibition of
moclobemide metabolism. Clearance is also re-
duced by cimetidine, an inhibitor of drug metabo-
lising systems. An extensive review of the clinical
pharmacokinetics of this drug has recently been
published.[142]
Moclobemide undergoes a complex metabolic
fate initially involving morpholine carbon and
nitrogen oxidation, deamination and aromatic
hydroxylation, with quantitative differences be-
tween species. [146,147] Of the identified metabo-
lites the N-oxide derivative and a ring-opened
metabolite retain modest MAO-A inhibitory activ-
ity in vitro.[146] In the rat the N-oxide is equipotent
to the parent drug as an inhibitor of brain and liver
MAO-A.[140,148] A more active, but as yet uniden-
tified, metabolite has been suggested. [140,149]
© Adis International Limited. All rights reserved.
295
Moclobemide Ro 12-8095 Ro 12-5637
83.6 ± 41.0
0.58 ± 0.23 0.49 ± 0.15 0.10 ± 0.03
20.2 ± 17.8a
1.61 ± 0.68a O.35 ± 0.16 0.14 ± 0.08
77.6 ± 45.0
0.59 (47) 0.58 (26) 0.19 (36)a
33.8 ± 7.72b
0.22 ± 0.12 0.23 ± 0.09 0.07 (27)c
Two minor ring opened metabolites have MAO-
B inhibitory activity in rat liver while the primary
amine Ro 16-6491 is a potent inhibitor of rat brain
MAO-B in vitro, possibly accounting for the ex vivo
peripherical MAO-B inhibition by the parent drug.
The amine derivative, however, is undetectable in
human plasma after a single dose of moclobemide
100mg.[147]
Morpholine C-oxidation to the lactam deriva-
tive Ro 12-8095 is quantitatively prominent in
humans and yields a number of metabolic forms
with carbamide, tertiary and secondary amine
structure.[146] This main pathway appears to in-
volve CYP2C19, as the oral clearance of moclo-
bemide is lower and t1⁄2 longer in poor metabolisers
than in extensive metabolisers of S-mephenytoin,
although the difference is less marked after re-
peated administration.[142,150]
CYP2D6 does not appear to play a significant
role in the metabolism of moclobemide. The parent
drug and its main metabolites are weaker inhibitors
of CYP2D6 in vitro, but there is evidence that
moclobemide inhibits this isoenzyme in vivo[151]
although the clinical relevance of these findings is
not clear.[142]
The lactam metabolite appears rapidly in
plasma, reaching concentrations comparable with
those of moclobemide in healthy volunteers[143]
but it is unlikely to contribute to the effects of the
Clin Pharmacokinet 1998 Apr; 34 (4)
296
parent drug since – as mentioned – it is essentially
inactive.[140,142] The N-oxide metabolite Ro 12-5637
is also present in blood but at lower concentrations
than the parent drug and its lactam derivative.[143]
The single and multiple dose pharmacokinetics of
this and other metabolites in humans, and their dis-
position compared with the parent drug in animals,
have not yet been fully characterised.
Information is also limited on how various
physiological and pathological factors affecting
drug elimination influence the pharmacokinetic
behaviour of metabolites. The pharmacokinetics of
moclobemide are not significantly affected by age
and renal function, but the tmax and Cmax of its N-
oxide derivative are significantly higher in patients
with moderate or severe renal impairment than in
healthy volunteers.[144,145] Chronic liver disease in-
creases the mean Cmax of the N-oxide and reduces
that of the lactam derivative, besides markedly re-
ducing clearance and prolonging the t1⁄2β (3.9 vs 1.7
hours) of intravenous moclobemide 90mg com-
pared with average values in healthy individu-
als.[139,142,145] The oral bioavailability and Cmax of
moclobemide are also significantly higher in these
patients; mean Cmax of the 2 metabolites are less
affected while tmax was longer and far more vari-
able than in healthy individuals.[142,145]
8. Conclusions
Like most of the older drugs in the same ther-
apeutic class, these newer antidepressants are
extensively biotransformed before excretion.
Sometimes, as in the case of moclobemide,[142] nu-
merous metabolites have been identified in the
blood and urine of animals and humans, making it
hard to specify what role the administered drug
plays in the overall outcome. In some cases, how-
ever, only the main metabolites have been identi-
fied, and a substantial proportion of the dose re-
mains unaccounted for.
Most of these drugs are primarily dependent on
1 or more phase I oxidative pathways for their
metabolism, except milnacipran whose clearance
appears due equally to renal excretion and glucu-
ronide conjugation. However, a dealkylated meta-
© Adis International Limited. All rights reserved.
Caccia
bolite whose formation has not yet been well
characterised, but appears to be carbamazepine-
inducible, has been identified in the body fluids of
volunteers and patients taking milnacipran.[76]
The primary metabolites, often pharmacologi-
cally active, are generally substrates for further
phase I processes. Thus most of these drugs un-
dergo various parallel and/or sequential oxidation
steps before conjugation and excretion as polar me-
tabolites. These are usually the rate-limiting steps,
as they are influenced to varying degrees by
recognised endogenous and exogenous factors af-
fecting drug disposition, and at least those medi-
ated by CYP2D6 and CYP2C19 isoenzymes by
genetic polymorphism.[17,18,66]
Some pathways proceed through saturable
mechanisms as in the case of fluvoxamine,[152]
fluoxetine, nefazodone, paroxetine and venlafaxine.
The clinical significance of this finding has been
reviewed elsewhere.[16,17,30,44] Others may be influ-
enced by chirality (e.g. citalopram and demethyl-
citalopram and fluoxetine N-dealkylation) when
the drug is a racemic mixture of enantiomers (as
are mirtazapine, reboxetine and tianeptine, al-
though few or no data are yet available on this as-
pect of their disposition). This may explain the in-
tra- and inter-population variability with most of
these drugs which from the pharmacokinetic point
of view is approximately the same as for TCAs.
The main isoenzymes responsible for the meta-
bolism of the older antidepressants, and lipophilic
drugs in general (CYP 3A3/4, 2D6, 1A2, 2C), are
involved to varying degrees in the oxidation of
these newer agents. They also mediate the metabo-
lic interactions between some of these agents and
other drugs metabolised through the same en-
zyme(s). There are full reviews[16-19,29-31,63-65]
dealing with the in vitro potency on selected en-
zymes and the likelihood of clinically important
drug-drug interactions. Metabolites have been less
studied but where information is available it ap-
pears that these may retain in vitro inhibitory po-
tencies similar to (e.g. paroxetine and its M2 meta-
bolites as inhibitors of 2D6) or even greater than
the parent drug on specific isoenzymes (for instance
Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
fluoxetine and norfluoxetine are equally active in
inhibiting 2D6 and 2C9 but the metabolite is more
potent as an inhibitor of 3A3/4).[15,18] Hence the
need to identify and characterise all the pathways
in drug metabolism and the enzymes involved, and
any factors which by altering their activity could
potentially modify the therapeutic response. This
information is lacking for some of these drugs.
As regards metabolites’ pharmacological activ-
ity, knowledge is clearly inadequate for many of
these drugs and in most cases there is no real
information about their role in the beneficial
and/or adverse effects of the parent drug. With few
exceptions, only the main metabolites have been
investigated, generally with regard to their in vitro
activity on monoaminergic mechanism(s), which
are of course important in mediating the anti-
depressant action of the parent compound. How-
ever, in vitro potency and selectivity may translate
poorly into in vivo pharmacological responses. The
MAO-A inhibitor potency of moclobemide N-oxide
in vitro in rat brain, for example, is several times
lower than that of moclobemide but the parent drug
and its metabolite have comparable potency and
selectivity ex vivo in rats.[148] Despite the fact that
norfluoxetine and norsertraline are less potent than
their parent drugs in inhibiting serotonin uptake in
vitro, the metabolites are at least as active in inhib-
iting food intake in animals, probably playing a
role in the hypophagic action of these compounds
in some species,[57,75] possibly through mecha-
nisms other than inhibition of serotonin uptake.
The phenyl-piperazine derivative of nefazodone,
mCPP, shows 5-HT2C-like agonist properties in rat
and humans,[114] further showing that metabolites
may have certain characteristics in vitro but others
in vivo, and that their activity must, therefore, be
evaluated in a broad sense, not only for the anti-
depressant-like effects of the parent drug. Unlike
moclobemide, some of its minor ring-opened me-
tabolites have MAO-B inhibitory activity in the
rat,[140,146,147] illustrating the need to extend phar-
macological and pharmacokinetic studies to all
metabolites for reliable extrapolation of pharmaco-
logical findings to the clinical situation.
© Adis International Limited. All rights reserved.
297
Further advances in this field require addi-
tional information on the pharmacological activity
of the metabolites after direct administration to an-
imals, particularly after repeated administration
which more closely mimics the clinical situation.
It would be useful to know their plasma protein
binding and pharmacokinetic behaviour with par-
ticular regard to the brain-to-blood partition and
concentrations reached at the site of action com-
pared with the parent drug. This is because the
blood-brain distribution of the metabolites may
differ from the original molecule, and as a con-
sequence the metabolite-to-parent drug ratio in
blood does not always reflect that at the site of
action.
This picture has been described for many cen-
trally acting drugs and their metabolites,[11,105] and
detailed information about the relationship be-
tween the parent drug and metabolite concentra-
tions in blood and brain is essential for under-
standing the contribution of each active species to
any given pharmacological or toxic effect. This
information needs to be extended, when appropri-
ate, to phase II metabolites as some of these too
may enter the brain and have direct pharmacolog-
ical or toxic effects.[153]
While animal studies illustrate the importance
of measuring drugs and metabolites at the precise
target site, they highlight the hazard of simply ex-
trapolating across species because of differences in
brain drug uptake.[96,97,154] Although studies in
different species may improve predictivity, defini-
tive evidence of metabolite activity, and how it
contributes to the overall outcome, can only come
from human studies. These are still far away, but
non-invasive techniques are increasingly being
used for evaluating the pharmacodynamic response
to drug therapy and to measure drug concentrations
in tissues.[155-157] As these technologies become
more refined, the relationships between drug dose,
blood and tissue concentrations and affinity for
specific neurotransmitter systems might be quanti-
fied in humans.
Clin Pharmacokinet 1998 Apr; 34 (4)
298
References
1. Coccaro EF, Siever LJ. Second generation antidepressants: a
comparative review. J Clin Pharmacol 1985; 25: 241-60
2. Blackwell B, Simon JS. Second-generation antidepressants.
Drugs Today 1986; 22: 611-33
3. Hollister LE. Current antidepressants. Annu Rev Pharmacol
Toxicol 1993; 26: 23-37
4. Rudorfer MV, Potter WZ. Antidepressants: a comparative re-
view of the clinical pharmacology and therapeutic use of the
‘newer’ versus the ‘older’ drugs. Drugs 1989; 37: 713-38
5. Möller H-J, Volz H-P. Drug treatment of depression in the
1990s: an overview of achievements and future possibilities.
Drugs 1996; 52: 625-6
6. Skerritt U, Evans R, Montgomery SA. Selective serotonin
reuptake inhibits in older patients: a tolerability perspective.
Drugs Aging 1997; 10: 209-18
7. Amsterdam J, Brunswick D, Mendels J. The clinical application
of tricyclic antidepressant pharmacokinetics and plasma lev-
els. Am J Psychiatry 1980; 137: 653-62
8. Guthrie S, Lane EA, Linnoila M. Monitoring of plasma drug
concentrations in clinical psychopharmacology. In: Meltzer
HY, editor. Psychopharmacology: the third generation of
progress. New York: Raven Press, 1987: 1323-9
9. Preskorn SH, Dorey RC, Jerkovich GS. Therapeutic drug
monitoring of tricyclic antidepressants. Clin Chem 1988;
34: 822-8
10. Orsulak PJ. Therapeutic monitoring of antidepressant drugs:
guidelines updated. Ther Drug Monit 1989; 11: 497-507
11. Caccia S, Garattini S. Formation of active metabolites of psy-
chotropic drugs: an updated review of their significance. Clin
Pharmacokinet 1990; 18: 434-59
12. Potter WZ, Manji HK. Antidepressants, metabolites, and appar-
ent drug resistance. Clin Neuropharmacol 1990; 13 Suppl. 1:
S45-53
13. Nordin C, Bertilsson L. Active hydroxymetabolites of antide-
pressants: emphasis on E-10-hydroxy-nortriptyline. Clin
Pharmacokinet 1995; 28: 26-40
14. Caccia S, Garattini S. Pharmacokinetic and pharmacodynamic
significance of antidepressant drug metabolites. Pharmacol
Res 1992; 26: 317-29
15. Preskorn SH. Pharmacokinetics of antidepressants: why and
how they are relevant to treatment. J Clin Psychiatry 1993; 54
Suppl.: 14-34
16. DeVane CL. Pharmacokinetics of the newer antidepressants:
Clinical relevance. Am J Med 1994; 97 Suppl. 6A: 13S-23S
17. Goodnick PJ. Pharmacokinetic optimisation of therapy with
newer antidepressants. Clin Pharmacokinet 1994; 27: 307-30
18. Baumann P. Pharmacokinetic-pharmacodynamic relationship
of the selective serotonin reuptake inhibitors. Clin Phar-
macokinet 1996; 31: 444-69
19. Rudorfer MV, Potter WZ. The role of metabolites of antidepres-
sants in the treatment of depression. CNS Drugs 1997; 7:
273-312
20. Claassen V. Review of the animal pharmacology and pharma-
cokinetics of fluvoxamine. Br J Clin Pharmacol 1983; 15:
349S-55S
21. Dechant KL, Clissold SP. Paroxetine: a review of its pharmaco-
dynamic and pharmacokinetic properties, and therapeutic po-
tential in depressive illness. Drugs 1991; 41: 225-53
22. Milne CJ, Goa KL. Citalopram: a review of its pharmacody-
namic and pharmacokinetic properties, and therapeutic poten-
tial in depressive illness. Drugs 1991; 41: 450-77
© Adis International Limited. All rights reserved.
Caccia
23. Benfield P, Heel RC, Lewis SP. Fluoxetine: a review of its
pharmacodynamic and pharmacokinetic properties, and
therapeutic efficacy in depressive illness. Drugs 1986; 32:
481-508
24. Wong DT, Bymaster FP, Reid LR, et al. Norfluoxetine en-
antiomers as inhibitors of serotonin uptake in rat brain. Neu-
ropsychopharmacology 1993; 8: 337-44
25. Fuller, RW, Wong DT. Serotonin re-uptake blockers in vitro and
in vivo. J Clin Psychopharmacol 1987; 7: 365-435
26. Murdoch D, McTavish D. Sertraline: a review of its pharmaco-
dynamic and pharmacokinetic properties, and therapeutic po-
tential in depression and obsessive-compulsive disorder.
Drugs 1992; 44: 604-24
27. Ronfeld RA, Tremaine LM, Wilner KD. Pharmacokinetics of
sertraline and its N-demethyl metabolite in elderly and young
male and female volunteers. Clin Pharmacokinet 1997; 32:
Suppl. 1: 22-30
28. Benfield P, Ward A. Fluvoxamine: a review of its pharmacody-
namic and pharmacokinetic properties, and therapeutic effi-
cacy in depressive illness. Drugs 1986; 32: 313-34
29. Preskorn SH, Magnus RD. Inhibition of hepatic P-450 isoen-
zymes by serotonin selective reuptake inhibitors: in vitro and
in vivo findings and their implications for patient care. Psy-
chopharmacol Bull 1994; 30: 251-9
30. Lane RM. Pharmacokinetic drug interaction potential of selec-
tive serotonin reuptake inhibitors. Int Clin Psychopharmacol
1996; 11 Suppl. 5: 31-61
31. Nemeroff CB, DeVane CL, Pollock BG. Newer antidepressants
and the cytochrome P450 system. Am J Psychiatry 1996; 153:
311-20
32. Becquemont L, Ragueneau I, Le Bot MA, et al. Influence of the
CYP1A2 inhibitor fluvoxamine on tacrine pharmacokinetics
in humans. Clin Pharmacol Ther 1997; 61: 619-27
33. Spigset O, Carleborg L, Hedenmalm K, et al. Effect of cigarette
smoking on fluvoxamine pharmacokinetics in humans. Clin
Pharmacol Ther 1995; 58: 399-403
34. Crewe HK, Lennard MS, Tucker GT, et al. The effect of selec-
tive serotonin re-uptake inhibitors on cytochrome P4502D6
(CYP2D6) activity in human liver microsomes. Br J Clin
Pharmacol 1992; 34: 262-5
35. Skjelbo E, Brosen K. Inhibitors of imipramine metabolism by
human liver microsomes. Br J Clin Pharmacol 1992; 34:
256-61
36. von Moltke LL, Greenblatt DJ, Court MH, et al. Inhibition of
alprazolam and desipramine hydroxylation by paroxetine and
fluvoxamine: comparison with other selective serotonin
reuptake inhibitor antidepressants. J Clin Psychopharmacol
1995; 15: 125-31
37. Jeppesen U, Rasmussen BB, Brosen K. Fluvoxamine inhibits
the CYP2C19-catalyzed bioactivation of chloroguanide. Clin
Pharmacol Ther 1997; 62: 279-86
38. Hartter S, Wetzel H, Hammes E, et al. Inhibition of antidepres-
sant demethylation and hydroxylation by fluvoxamine in de-
pressed patients. Psychopharmacology 1993; 110: 302-8
39. Kaye CM, Haddock RE, Langley PF, et al. A review of the
metabolism and pharmacokinetics of paroxetine in man. Acta
Psychiatr Scand 1989; 80 Suppl. 350: 60-75
40. Crewe S, Lennard MS, Tucker GT, et al. The effect of paroxetine
and other specific serotonin re-uptake inhibitors on cyto-
chrome P450IID6 activity in human liver microsomes. Br J
Clin Pharmacol 1991; 32: 658P-9P
Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
41. Bloomer JC, Woods FR, Haddock RE, et al. The role of cyto-
chrome P4502D6 in the metabolism of paroxetine by human
liver microsomes. Br J Clin Pharmacol 1992; 33: 521-3
42. Sindrup SH, Brøsen K, Gram LF, et al. The relationship be-
tween paroxetine and the sparteine oxidation polymorphism.
Clin Pharmacol Ther 1992; 51: 278-87
43. Brøsen K, Hansen JG, Nielsen KK, et al. Inhibition by paroxet-
ine of desipramine metabolism in extensive but not in poor
metabolizers of sparteine. Eur J Clin Pharmacol 1993; 44:
349-55
44. Preskorn SH. Clinically relevant pharmacology of selective se-
rotonin reuptake inhibitors: an overview with emphasis on
pharmacokinetics and effects on oxidative drug metabolism.
Clin Pharmacokinet 1997; 32 Suppl. 1: 1-21
45. Ozdemir V, Naranjo CA, Herrmann N, et al. paroxetine poten-
tiates the central nervous system side effects of perphenazine:
contribution of cytocrome P4502D6 inhibition in vivo. Clin
Pharmacol Ther 1997; 62: 334-47
46. Tremaine LM, Welch WM, Ronfeld RA. Metabolism and dis-
position of the 5-hydroxytryptamine uptake blocker sertraline
in the rat and dog. Drug Metab Dispos 1989; 17: 542-50
47. Sindrup SH, Brøsen K, Hansen MGJ, et al. Pharmacokinetics
of citalopram in relation to the sparteine and the mephenytoin
oxidation polymorphisms. Ther Drug Monit 1993; 15: 11-7
48. Baumann P, Nil R, Souche A, et al. A double-blind placebo-con-
trolled study of citalopram with and without lithium in the
treatment of therapy-resistant depressive patients: a clinical,
pharmacokinetic, and pharmacogenetic investigation. J Clin
Psychopharmacol 1996; 16: 307-14
49. Kobayashi K, Chiba K, Yagi T, et al. Identification of cyto-
chrome P450 isoforms involved in citalopram N-demethyla-
tion by human liver microsomes. J Pharmacol Exp Ther 1997;
280: 927-33
50. Rochat B, Amey M, Baumann P. Analysis of enantiomers of
citalopram and its demethylated metabolites in plasma of de-
pressive patients using chiral reverse-phase liquid chroma-
tography. Ther Drug Monit 1995; 17: 273-9
51. Foglia JP, Pollock BG, Kirshner MA, et al. Plasma levels of
citalopram enantiomers and metabolites in elderly patients.
Psychopharmacol Bull 1997; 33: 109-12
52. Rochat B, Amey M, Gillet M, et al. Identification of three cy-
tochrome P450 isozymes involved in N-demethylation of
citalopram enantiomers in human liver microsomes. Pharma-
cogenetics 1997; 7: 1-10
53. Bondolfi G, Chautems C, Rochat B, et al. Non-response to
citalopram in depressive patients: pharmacokinetic and clin-
ical consequences of a fluvoxamine augmentation. Psycho-
pharmacology 1996; 128: 421-5
54. Torok-Both GA, Baker GB, Coutts RT, et al. Simultaneous de-
termination of fluoxetine and norfluoxetine enantiomers in
biological samples by gas chromatography with electron-cap-
ture detection. J Chromatogr B 1992; 579: 99-106
55. Fjordside L, Jeppesen U, Eap CB et al. The stereoselective
metabolism of fluoxetine in poor and extensive metabo-
lisers of sparteine [abstract]. Eur J Clin Pharmacol 1997;
Suppl. I: A127
56. Fuller RW, Snoddy HD, Perry KW. Importance of duration of
drug action in the antagonism of p-chloroamphetamine deple-
tion of brain serotonin-comparison of fluoxetine and
chlorimipramine. Biochem Pharmacol 1978; 27: 193-8
© Adis International Limited. All rights reserved.
299
57. Anelli M, Bizzi A, Caccia S, et al. Anorectic activity of fluoxet-
ine and norfluoxetine in mice, rats and guinea-pigs. J Pharm
Pharmacol 1992; 44: 696-8
58. Caccia S, Fracasso C, Garattini S, et al. Effects of short- and
long-term administration of fluoxetine on the monoamine
content of rat brain. Neuropharmacology 1992; 31: 343-7
59. Caccia S, Anelli M, Codegoni AM, et al. The effect of single
and repeated anorectic doses of 5-hydroxytryptamine uptake
inhibitors on indole levels in rat brain. Br J Pharmacol 1993;
110: 355-9
60. Altamura AC, Moro AR, Percudani M. Clinical pharmacokinet-
ics of fluoxetine. Clin Pharmacokinet 1994; 26: 201-14
61. Hamelin BA, Turgeon J, Vallée F et al. The disposition of
fluoxetine but not sertraline is altered in poor metabolizers of
debrisoquin. Clin Pharmacol Ther 1996; 60: 512-21
62. Otton SV, WU D, Joffe RT, et al. Inhibition by fluoxetine of
cytochrome P4562D6 activity. Clin Pharmacol Ther 1993;
53: 401-9
63. Harvey A, Preskorn S. Cytochrome P450 enzymes: interpreta-
tion of their interactions with selective serotonin reuptake
inhibitors. Part II. J Clin Psychopharmacol 1996; 16: 345-55
64. Brosen K. Are pharmacokinetic drug interactions with the
SSRIS and issue? Int Clin psychopharmacol 1996; 11 Suppl.
1: 23-7
65. Meyer UA, Amrein R, Balant LP et al. Antidepressants and
drug-metabolizing enzymes: expert group report. Acta Psy-
chiatr Scand 1996; 93: 71-9
66. Bertilsson L, Dahl M-L, Tybring G. Pharmacogenetics of anti-
depressants: clinical aspects. Acta Psychiatr Scand Suppl
1997; 391: 14-21
67. Sproule BA, Naranjo CA, Bremner KE, et al. Selective seroto-
nin reuptake inhibitors and CNS drug interactions. Clin Phar-
macokinet 1997; 33: 454-71
68. Zussman BD, Davie CC, Fowles SE, et al. Sertraline, like other
SSRIs, is a significant inhibitor of desipramine metabolism
in vivo. Br J Clin Pharmacol 1995; 39: 550P
69. Alderman J, Preskorn SH, Greenblatt DJ, et al. desipramine
pharmacokinetics when coadministered with paroxetine or
sertraline in extensive metabolizers. J Clin Psychopharmacol
1997; 17: 284-91
70. Kurtz DL, Bergstrom RF, Goldberg MJ, Cerimele BJ. The effect
of sertraline on the pharmacokinetics of desipramine and
imipramine. Clin Pharmacol Ther 1997; 62: 145-56
71. Carson SW. Pharmacokinetic and pharmacodynamic drug in-
teractions with polypharmacotherapy of treatment-resistant
affective and obsessive-compulsive disorders. Psychopharmacol
Bull 1996; 32: 555-68
72. Preskorn SH, Alderman J, Greenblatt DJ, et al. Sertraline does
not inhibit cytochrome P4503A-mediated drug metabolism in
vivo. Psychopharmacol Bull 1997; 33: 659
73. Sprouse J, Clarke T, Reynolds L, et al. Comparison of the ef-
fects of sertraline and its metabolite desmethylsertraline on
blockade of central 5-HT reuptake in vivo. Neuropsycho-
pharmacology 1996; 14: 225-31
74. Fuller RW, Hemrick-Luecke SK, Littlefield ES, et al. Compar-
ison of desmethylsertraline with sertraline as a monoamine
uptake inhibitor in vivo. Prog Neuropsychopharmacol 1995;
19: 135-49
75. Garattini S, Bizzi A, Caccia S, et al. Progress report on the
anorectic effects of dexfenfluramine, fluoxetine and
sertraline, Int J Obes Relat Metab Disord 1992; 16 Suppl.
3: S43-50
Clin Pharmacokinet 1998 Apr; 34 (4)
300
76. Puozzo C, Leonard BE. Pharmacokinetics of milnacipran in
comparison with other antidepressants. Int Clin Psy-
chopharmacol 1996; 11 Suppl. 4: 15-27
77. Briley M, Prost JF, Moret C. Preclinical pharmacology of
milnacipran. Int Clin Psychopharmacol 1996; 11 Suppl. 4:
10-4
78. Muth EA, Haskins JT, Moyer JA, et al. Antidepressant biochem-
ical profile of the novel bicyclic compound Wy-45,030, an
ethyl cyclohexanol derivative. Biochem Pharmacol 1986; 35:
4493-7
79. Artigas F. Selective serotonin/noradrenaline reuptake inhibitors
(SNRIs): pharmacology and therapeutic potential in the treat-
ment of depressive disorders. CNS Drugs 1995; 4: 79-89
80. Bonnaud B, Cousse H, Mouzin G, et al. 1-Aryl-2-(aminoethyl)
cyclopropanecarboxylic acid derivatives: a new series of po-
tential antidepressants. J Med Chem 1987; 30: 318-25
81. Morton WA, Sonne SC, Verga MA. Venlafaxine: a structurally
unique and novel antidepressant. Ann Pharmacother 1995; 29:
387-95
82. Troy SM, Schultz RW, Parker VD, et al. The effect of renal disease
on the disposition of venlafaxine. Clin Pharmacol Ther 1994;
56: 14-21
83. Wang CP, Howell SR, Scatina J, et al. The disposition of
venlafaxine enantiomers in dogs, rats, and humans receiving
venlafaxine. Chirality 1992; 4: 84-90
84. Howell SR, Husbands GEM, Scatina JA, et al. Metabolic dis-
position of 14C-venlafaxine in mouse, rat, dog, rhesus monkey
and man. Xenobiotica 1993; 23: 349-59
85. Muth EA, Moyer JA, Haskins JT, et al. Biochemical, neuro-
physiological, and behavioral effects of Wy-45,233 and other
identified metabolites of the antidepressant venlafaxine. Drug
Dev Res 1991; 23: 191-9
86. Otton SV, Ball SE, Cheung SW, et al. Venlafaxine oxidation in
vitro is catalysed by CYP2D6. Br J Clin Pharmacol 1996; 41:
149-56
87. von Moltke LL, Greenblatt DJ, Duan SX, et al. Phenacetin O-
deethylation by human liver microsomes in vitro: inhibition
by chemical probes, SSRI antidepressants, nefazodone and
venlafaxine. Psychopharmacology 1996; 128: 398-407
88. Otton SV, Ball SE, Cheung SW, et al. Comparative inhibition
of the polymorphic enzyme CYP2D6 by venlafaxine (VF)
and other 5HT uptake inhibitors. Clin Pharmacol Ther 1994;
55: 141
89. Ball SE, Ahern D, Scatina J, et al. Venlafaxine: in vitro inhibi-
tion of CYP2D6 dependent imipramine and desipramine me-
tabolism. Comparative studies with selected SSRIs, and
effects on human hepatic CYP3A4, CYP2C9 and CYPIA2.
Br J Clin Pharmacol 1997; 43: 619-26
90. Troy SM, Lucki I, Peirgies AA, et al. Pharmacokinetic and phar-
macodynamic evaluation of the potential drug interaction be-
tween venlafaxine and diazepam. J Clin Pharmacol 1995; 35:
410-9
91. Ketter TA, Flockhart DA, Post RM, et al. The emerging role of
cytochrome P450 3A in psychopharmacology. J Clin Psy-
chopharmacol 1995; 15: 387-98
92. Ereshefsky L. Drug-drug interactions involving antidepres-
sants: focus on venlafaxine. J Clin Psychopharmacol 1996; 16
Suppl. 2: 37S-50S
93. Melloni P, Della Torre A, Lazzari E, et al. Configuration studies
on 2[α-ethoxyphenoxy][benzyl]morpholine FCE 20124. Tet-
rahedron 1985; 41: 1393-9
© Adis International Limited. All rights reserved.
Caccia
94. Melloni P, Carniel G, Della Torre A, et al. Potential anti-
depressant agents. α-aryloxy-benzyl derivatives of ethanol-
amine and morpholine. Eur J Med Chem Chim Ther 1984;
19: 235-42
95. Strolin Benedetti M, Frigerio E, Tocchetti P, et al. Stereoselec-
tive and species-dependent kinetics of reboxetine in mouse
and rat. Chirality 1995; 7: 285-9
96. Renshaw PF, Guimaraes AR, Fava M, et al. Accumulation of
fluoxetine and norfluoxetine in human brain during therapeu-
tic administration. Am J Psychiatry 1992; 149: 1592-4
97. Bergami A, Fracasso C, Garattini S, et al. Brain uptake and
acute indole-depleting effect of dexfenfluramine in squirrel
monkeys after a high-dose regimen. Pharm Sci 1995; 1: 45-8
98. Dostert P, Benedetti MS, Poggesi I. Review of the pharmacoki-
netics and metabolism of reboxetine, a selective noradrena-
line reuptake inhibitor. Eur Neuropsychopharmacol 1997; 7
Suppl. 1: S23-35
99. Cocchiara G, Battaglia R, Pevarello P, et al. Comparison of the
disposition and of the metabolic patterns of reboxetine, a new
antidepressant, in the rat, dog, monkey and man. Eur J Drug
Metab Pharmacokinet 1991; 16: 231-9
100. Pellizzoni C, Poggesi I, Jørgensen NP, et al. Pharmacokinetics
of reboxetine in healthy volunteers: single against repeated
oral doses and lack of enzymatic alterations. Biopharm Drug
Dispos 1996; 17: 623-33
101. Edwards DMF, Pellizzoni C, Breuel HP, et al. Pharmacokinetics
of reboxetine in healthy volunteers: single oral doses, linearity
and plasma protein binding. Biopharm Drug Dispos 1995; 16:
443-60
102. Davis R, Whittington R, Bryson HM. Nefazodone: a review of
its pharmacology and clinical efficacy in the management of
major depression. Drugs 1997; 53: 608-36
103. Mayol RF, Cole CA, Luke GM, et al. Characterization of the
metabolites of the antidepressant drug nefazodone in human
urine and plasma. Drug Metab Dispos 1994; 22: 304-11
104. Caccia S, Fong MH, Garattini S, et al. Plasma concentrations
of trazodone and 1-(3-chlorophenyl)-piperazine in man after
a single oral dose of trazodone. J Pharm Pharmacol 1982; 34:
605-6
105. Caccia S, Garattini S. 1-Arylpiperazine as active metabolites of
(4-substituted aryl-1-piperazinyl)alkyl heterocyclic drugs.
Acta Pharm Jugosl 1990; 40: 441-60
106. Caccia S, Vigano GL, Mingardi G, et al. Clinical pharmacoki-
netics of oral buspirone in patients with impaired renal func-
tion. Clin Pharmacokinet 1988; 14: 171-7
107. Bianchi G, Caccia S, Della Vedova F, et al. The alpha2-adreno-
ceptor antagonist activity of ipsapirone and gepirone is medi-
ated by their common metabolite 1-(2-pyrimidinyl)-
piperazine (PmP). Eur J Pharmacol 1988; 151: 365
108. Schmider J, Greenblatt DJ, von Moltkr LL, et al. Inhibition of
cytochrome P450 by nefazodone in vitro: studies of dextro-
methorphan O- and N-demethylation. Br J Clin Pharmacol
1996; 41: 339-43
109. Greene DS, Barbhaiya RH. Clinical pharmacokinetics of
nefazodone. Clin Pharmacokinet 1997; 33: 260-75
110. Barbhaiya RH, Buch AB, Greene DS. Single and multiple dose
pharmacokinetics of nefazodone in subjects classified as ex-
tensive and poor metabolizers of dextromethorphan. Br J Clin
Pharmacol 1996; 42: 573-81
111. Marino MR, Langenbacher KM, Uderman HD. Interaction of
nefazodone (N) and fluoxetine (F). Clin Pharmacol Ther
1996; 59: 180
Clin Pharmacokinet 1998 Apr; 34 (4)
Metabolism of the Newer Antidepressants
112. Yasui N, Otani K, Kaneko S, et al. Inhibition of trazodone me-
tabolism of thioridazine in humans. Ther Drug Monit 1995;
17: 333-5
113. Hamik A, Peroutka SJ. 1-(m-chlorophenyl)piperazine (mCPP)
interactions with neurotransmitter receptors in the human
brain. Biol Psychiatry 1989; 25: 569-75
114. Murphy DL, Lesch KP, Aulakh CS, et al. Serotonin-selective
arylpiperazines with neuroendocrine, behavioral, tempera-
ture, and cardiovascular effects in humans. Pharmacol Rev
1991; 43: 527-50
115. Benjamin J, Greenberg BD, Murphy DL. Daily administration
of m-chlorophenylpiperazine to healthy human volunteers
rapidly attenuates many of its behavioral, hormonal, cardio-
vascular and temperature effects. Psychopharmacology 1996;
127: 140-9
116. Barbhaiya RH, Buch AB, Greene DS. A study of the effect of
age and gender on the pharmacokinetics of nefazodone after
single and multiple doses. J Clin Psychopharmacol 1996; 16:
19-25
117. Barbhaiya RH, Shukla UA, Greene DS. Single-dose pharmaco-
kinetics of nefazodone in healthy young and elderly subjects
and in subjects with renal or hepatic impairment. Eur J Clin
Pharmacol 1995; 49: 221-8
118. Barbhaiya RH, Brady ME, Shukla UA, et al. Steady-state phar-
macokinetics of nefazodone in subjects with normal and im-
paired renal function. Eur J Clin Pharmacol 1995; 49: 229-35
119. Barbhaiya RH, Shukla UA, Natarajan CS, et al. Single- and
multiple-dose pharmacokinetics of nefazodone in patients
with hepatic cirrhosis. Clin Pharmacol Ther 1995; 58: 390-8
120. Mocaër E, Rettori MC, Kamoun A. Pharmacological antide-
pressive effects and tianeptine-induced 5-HT uptake increase.
Clin Neuropharmacol 1988; 11 Suppl. 2: S32-42
121. Mennini T, Garattini S. Neurobiology of tianeptine: a new phar-
maceutic agent. Presse Med 1991; 20: 1823-7
122. Ansseau M. The paradox of tianeptine. Eur Psychiatry 1993; 8
Suppl. 2: 89S-93S
123. Wilde MI, Benfield P. Tianeptine: a review of its pharmacody-
namic and pharmacokinetic properties, and therapeutic effi-
cacy in depression and coexisting anxiety and depression.
Drugs 1995; 49: 411-39
124. Grislain L, Gele P, Bertrand M, et al. The metabolic pathways
of tianeptine, a new antidepressant, in healthy volunteers.
Drug Metab Dispos 1990; 18: 804-8
125. Oluyomi AO, Datla KP, Curzon G. Effects of the (+) and
(–)-enantiomers of the antidepressant drug tianeptine on 5-
HTP-induced behaviour. Neuropharmacology 1997; 36:
383-7
126. Larrey D, Tinel M, Letteron P, et al. Metabolic activation of the
new tricyclic antidepressant tianeptine by human liver cyto-
chrome P450. Biochem Pharmacol 1990; 40: 545-50
127. Couet W, Girault J, Latrille F, et al. Kinetic profiles of tianeptine
and its MC5 metabolite in plasma, blood and brain after single
and chronic intraperitoneal administration in the rat. Eur J
Drug Metab Pharmacokinet 1990; 15: 69-74
128. Carlhant D, Le Garrec J, Guedes Y, et al. Pharmacokinetics and
bioavailability of tianeptine in the elderly. Drug Invest 1990;
2: 167-72
129. Demotes-Mainard F, Galley P, Manciet G, et al. Pharmacoki-
netics of the antidepressant tianeptine at steady state in the
elderly. J Clin Pharmacol 1991; 31: 174-8
© Adis International Limited. All rights reserved.
301
130. Salvadori C, Merdjan H, Brouard R, et al. Tianeptine and its
main metabolite. Disposition in chronic renal failure and
haemodialysis. Fundam Clin Pharmacol 1990; 4: 663-71
131. Royer RJ, Royer-Morrot MJ, Paille F, et al. Tianeptine and its
main metabolite. Pharmacokinetics in chronic alcoholism and
cirrhosis. Clin Pharmacokinet 1989; 16: 186-91
132. Sitsen JMA, Zivkov M. Mirtazapine: clinical profile. CNS
Drugs 1995; 4 Suppl. 1: 39-48
133. Nutt D. Mirtazepine: pharmacology in relation to adverse ef-
fects. Acta Psychiatr Scand 1997; 96 Suppl. 391: 31-7
134. de Boer Th, Maura G, Raiteri M, et al. Neurochemical and
autonomic pharmacological profiles of the 6-aza-analogue of
mianserin, ORG 3770 and its enantiomers. Neuropharmacol-
ogy 1988; 27: 399-408
135. Davis R, Wilde MI. Mirtazapine: a review of its pharmacology
and therapeutic potential in the management of major depres-
sion. CNS Drugs 1996; 389-402
136. Sandker GW, Vos RME, Delbressine LPC, et al. Metabolism
of three pharmacologically active drugs in isolated human
and rat hepatocytes: analysis of interspecies variability and
comparison with metabolism in vivo. Xenobiotica 1994;
24: 143-55
137. Dahl M-L, Voortman G, Alm C, et al. In vitro and in vivo studies
on the disposition of mirtazepine in humans. Clin Drug Invest
1997; 13: 37-46
138. Poirier MF, Olié JP, Loo H, et al. Activité antidépressive,
caractéristiques pharmacocinétiques et proprétés biochimi-
ques de la cimoxatone, un nouvel IMAO-A réversible.
Encéphale 1983; IX: 331-43
139. Fitton A, Faulds D, Goa KL. Moclobemide: a review of its
pharmacological properties and therapeutic use in depressive
illness. Drugs 1992; 43: 561-96
140. Fulton B, Benfield P. Moclobemide: an update of its pharma-
cological properties and therapeutic use. Drugs 1996; 52:
450-74
141. Volz HP, Gleiter CH, Struck M, et al. Brofaromine: insight into
the nature of drug development. CNS Drugs 1995; 3: 1-8
142. Mayersohn M, Guentert TW. Clinical pharmacokinetics of the
monoamine oxidase-A inhibitor moclobemide. Clin Phar-
macokinet 1995; 29: 292-32
143. Schoerlin MP, Mayersohn M, Korn A, et al. Disposition kinetics
of moclobemide, a monoamine oxidase-A enzyme inhibitor:
single and multiple dosing in normal subjects. Clin Phar-
macol Ther 1987; 42: 395-404
144. Schoerlin MP, Horber FF, Frey FJ, et al. Disposition kinet-
ics of moclobemide, a new MAO-A inhibitor, in subjects
with impaired renal function. J Clin Pharmacol 1990; 30:
272-84
145. Stoeckel K, Pfefen JP, Mayersohn M, et al. Absorption and
disposition of moclobemide in patients with advanced age or
reduced liver or kidney function. Acta Psychiatr Scand 1990;
Suppl. 360: 94-97
146. Jauch R, Griesser E, Oesterhelt G, et al. Biotransformation of
moclobemide in humans. Acta Psychiatr Scand 1990; Suppl.
360: 87-90
147. Schoerlin MP, Da Prada M. Species-specific biotransformation
of moclobemide: a comparative study in rats and humans.
Acta Psychiatr Scand 1990; Suppl. 360: 108-10
148. Da Prada M, Kettler R, Keller HH, et al. Neurochemical profile
of moclobemide, a short-acting and reversible inhibitor of
monoamine oxidase type A. J Pharmacol Exp Ther 1989; 248:
400-8
Clin Pharmacokinet 1998 Apr; 34 (4)
302
149. Fritze J, Laux G, Sofic E, et al. Plasma moclobemide and me-
tabolites: lack of correlation with clinical response and bio-
genic amines. Psychopharmacology 1989; 99: 252-6
150. Guentert TW, Gram LF, Grange S, et al. Reversible interaction
of moclobemide with CYP2D6 and CYP2C9. Clin Pharmacol
Ther 1994; 55: PI-57
151. Gram LF, Brøsen K, and the Danish University Antidepressant
Group. Moclobemide treatment causes a substantial rise in the
sparteine metabolic ratio. Br J Clin Pharmacol 1993; 35:
649-52
152. Spigset O, Granberg K, Hägg S, et al. Non-linear fluvoxamine
disposition. Br J Clin Pharmacol 1998; 45: 257-63
153. Sperker B, Backman JT, Kroemer HK. The role of β-glucuro-
nidase in drug disposition and drug targeting in humans. Clin
Pharmacokinet 1997; 33: 18-31
154. Campbell B, Dard-Brunelle B, Caccia S. The use of pharmaco-
kinetics in the assessment of dexfenfluramine safety. In:
© Adis International Limited. All rights reserved.
Caccia
Nicolaidis S, editor. Obesity management and ReduxTM. San
Diego: Academic Press, 1997: 65-80
155. Matzke GR, St Peter WL. Clinical pharmacokinetics 1990. Clin
Pharmacokinet 1990; 18: 1-19
156. Bartels, Albert K. Detection of Psychoative drugs using 19F MR
spectroscopy. J Neural Transm Gen Sect 1995; 99: 1-6
157. Kennedy SH, Javanmard M, Vaccarino FJ. A review of func-
tional neuroimaging in mood disorders: positron emission
tomography and depression. Can J Psychiatry 1997; 42:
467-75
Correspondence and reprints: Dr Silvio Caccia, Istituto di
Ricerche Farmacologiche ‘Mario Negri’, via Eritrea 62,
20157 Milano, Italy.
E-mail: caccia@irfmn. mnegri.it
Clin Pharmacokinet 1998 Apr; 34 (4)