Key words
bipolar disorder - switch to mania - antidepressants - trazodone - mirtazapine - agomelatine
Introduction
Antidepressant drugs have been widely prescribed in patients with bipolar disorder,
despite clinicians’ concern about antidepressant-induced manic switch and the recommendations
of practice guidelines. In a recent report on antidepressant use in bipolar disorder,
the task force of the International Society of Bipolar Disorders (ISBD) stated that
there is striking incongruity between the wide use of antidepressant drugs in bipolar
disorder and the weak evidence base for their efficacy and safety. As the frequency
and severity of antidepressant-induced manic switch appears to be greater in bipolar
I disorder than bipolar II, it is recommended that antidepressants should only be
prescribed to bipolar I patients as an adjunct to mood-stabilizing medications [1].
Although the true rate of antidepressant-induced manic switch in bipolar disorder
is a controversial issue, there is a consensus that various antidepressants have differing
potential for inducing such a switch. Most studies show that such a potential is higher
for tricyclic antidepressants (TCAs) and venlafaxine than for selective serotonin
reuptake inhibitors (SSRIs) and bupropion [2]
[3]
[4]
[5].
However, there is sparse evidence concerning the risk of switching with other antidepressants,
such as trazodone, mirtazapine, and agomelatine. These 3 drugs are of special interest
because they are used in bipolar patients not only for their antidepressant effect,
but also to promote sleep [6]. Disturbed sleep is a common problem in patients with bipolar disorder, even in
inter-episode periods [7]
[8]. For many years, sedative antidepressants were used in such patients as an alternative
to benzodiazepine and non-benzodiazepine hypnotics (Z-drugs), because sleep-promoting
antidepressants are not related to the risk of hypnotic dependence and can, therefore,
be used for a longer time. As the use of antidepressants in bipolar patients has been
discouraged in many treatment guidelines, we observe the tendency of these drugs to
be replaced by antipsychotics (especially quetiapine), not only in the treatment of
bipolar depression, but also as sleep-promoting drugs. Such an approach does not seem
to be justified, because evidence for the use of sedative antidepressants in the treatment
of insomnia is much greater than for sedative antipsychotics. Sedative antidepressants
may also be better tolerated [9]. Furthermore, the doses of antidepressants needed to promote sleep are much lower
than the doses needed to treat depression [6], so the risk of switching is probably very low.
In this article, we review published case reports describing the switch to mania or
hypomania during treatment with trazodone, mirtazapine, or agomelatine, both in monotherapy
and in combination with mood stabilizers, and used both in the therapeutic doses for
antidepressant effect and in the low doses recommended to promote sleep.
Methods
We performed a literature search in PubMed for articles published in the English language,
which were published up to September 2014, using the following search terms in the
title or abstract: the antidepressant name (trazodone, mirtazapine, agomelatine) AND
mania, hypomania, bipolar disorder, or phase shift. To ensure accuracy, the search
performed by the first author of this paper was independently repeated by 2 reviewers
(M.J., Ł.O.). Furthermore, references of selected papers and relevant review articles
were scanned by all authors in order to locate other reports.
Results
Trazodone
Trazodone is an antidepressant of the serotonin antagonist and reuptake inhibitor
(SARI) class. Low doses of trazodone (below 150 mg/day) only act via the most potent
binding properties, that is, the serotonergic 5HT2A receptors, whereas higher doses
recruit additional pharmacological action and result in the blockade of the serotonin
transporter (SERT). The ability of trazodone to block SERTs is 100-fold less potent
than its ability to block 5HT2A receptors. As both of these actions are considered
necessary for an antidepressant effect, the trazodone dose recommended for the treatment
of depression is usually above 150 mg/day [10], whereas low doses (25–150 mg/day) are usually sufficient for sleep-promoting effects
[6].
The development of switch to mania or hypomania associated with trazodone treatment
was described in 17 patients ([Table 1]).
Table 1 Reports of switching to mania or hypomania during treatment with trazodone.
Study
|
Patient (age/gender/diagnosis)
|
Dose (mg/day)
|
Concomitant treatment
|
Time to switch onset (days)
|
Affective symptoms
|
Actions taken
|
Escobar et al.1980 [12]
|
middle-aged depressed man
|
400
|
none
|
14
|
mania
|
no data
|
Warren & Bick 1984 [17]
|
57-year-old woman, major depression
|
150
|
none
|
7
|
mania
|
discontinuation of trazodone, treatment with lithium and thiothixene
|
Warren & Bick 1984 [17]
|
25-year-old depressed woman
|
400
|
alprazolam 4 mg/day
|
28
|
mania
|
discontinuation of trazodone, treatment with chlorpromazine
|
Arana & Kaplan 1985 [11]
|
66-year-old man, recurrent depression
|
100
|
8 days earlier discontinuation of desipramine 75 mg/day due to switch to mania
|
4
|
mania
|
discontinuation of trazodone
|
Arana & Kaplan 1985 [11]
|
47-year-old depressed woman
|
400
|
2 weeks earlier discontinuation of desipramine 125 mg/day due to switch to mania
|
17
|
mania
|
discontinuation of trazodone
|
Theilman & Christenbury 1986 [16]
|
33-year-old woman, major depression
|
300
|
none
|
21 after discontinuation of trazodone
|
hypomania
|
treatment with thioridazine and lithium
|
Knobler et al. 1986 [14]
|
82-year-old man, bipolar disorder
|
100
|
after 10 days of treatment with maprotiline,
|
14
|
mania
|
discontinuation of trazodone, treatment with haloperidol
|
Knobler et al. 1986 [14]
|
35-year-old woman, bipolar disorder
|
300
|
after 6 weeks of treatment with chlorimipramine, up to 250 mg/day, followed by treatment
with imipramine, up to 300 mg/day, for another 6 weeks, 5 days of wash out
|
5
|
mania
|
discontinuation of trazodone, treatment with haloperidol
|
Knobler et al. 1986 [14]
|
84 year-old woman, major depression
|
300
|
none
|
56
|
mania
|
trazodone dose reduction to 150 mg/day
|
Zmitek 1987 [18]
|
58-year-old woman, major depression
|
150
|
none
|
7
|
mania
|
discontinuation of trazodone, treatment with haloperidol
|
Lennhoff 1987 [15]
|
53-year-old man, major depression
|
150
|
alcohol detoxification 17 days earlier, treatment doxepin 150 mg/day for 10 days,
3 days of wash out
|
4
|
hypomania
|
trazodone decreased to 50 mg/day, lithium carbonate 900 mg/day
|
Jabeen & Fisher 1991 [13]
|
61-year-old man, severe depression
|
100
|
after 7 days of benzodiazepine withdrawal from chlordiazepoxide (7.5 mg/day) and temazepam
(20 mg/day)
|
4
|
hypomania
|
discontinuation of trazodone, treatment with lithium, trazodone restarted at 200 mg/day,
benzodiazepines were gradually withdrawn
|
Jabeen & Fisher 1991 [13]
|
24-year-old woman, postnatal depression
|
150
|
none
|
4
|
hypomania
|
reduction of trazodone dose to 100 mg/day, after 7 days again increased to 150 mg/day
|
Jabeen & Fisher 1991 [13]
|
70-year-old depressed man
|
100
|
none
|
28
|
transient hypomania
|
trazodone dose increase to 200 mg/day
|
Horiguchi & Sai 2001 [19]
|
55-year-old man, major depression
|
50
|
fluvoxamine 50 mg/day added to stable treatment with trazodone and sulpiride 150 mg/day
|
|
hypomania
|
cessation of fluvoxamine
|
Hilleret et al. 2001 [20]
|
50-year-old man, bipolar disorder, right hemiplegia from birth with persisting motor
deficiency
|
50
|
venlafaxine 300 mg, CPAP
|
4 weeks after start of CPAP
|
hypomania, mixed episode
|
continuation CPAP, cessation of venlafaxine and trazodone, 6 days interruption of
pharmacological treatment, treatment with valproic acid and levomepromazine
|
Rachid et al. 2006 [21]
|
39-year-old woman, refractory major depression
|
200
|
paroxetine 40 mg/day, opipramol 150 mg/day, amitriptyline 50 mg/day, prazepam 100 mg/day,
methadone 7.5 mg/day rTMS
|
second week of rTMS treatment, depressive mixed episode
|
mixed episode
|
rTMS treatment was discontinued, treatment with valproic acid
|
Not all details given in reports, rTMS=repetitive transcranial magnetic stimulation,
CPAP=continuous positive airway pressure
9 of those patients were treated in monotherapy with the antidepressant dose (≥150 mg/day)
of trazodone [11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]. In only 4 cases, monotherapy with a low dose of trazodone (100 mg/day) was found
to be associated with switch to mania or hypomania [11]
[13]
[14]. All of these 4 cases were older patients (over 60 years old), so they could have
a slower elimination rate and, therefore, higher trazodone plasma concentrations compared
to the younger patients. Furthermore, in 3 of those 4 patients, other factors were
present that could increase the risk of switch to mania. 2 patients were treated with
other antidepressants before the start of the trazodone treatment, and one patient
obtained trazodone treatment during detoxification from benzodiazepines. In only 2
patients, a very low dose of trazodone (50 mg/day) was described as being related
to switch to mania or a mixed depressive state. In both cases, treatment with trazodone
was combined with other antidepressants, and in one patient additionally with a non-pharmacological
intervention (continuous positive airway pressure, CPAP) [19]
[20] A non-pharmacological intervention (repetitive transcranial magnetic stimulation,
rTMS) added to combined treatment with trazodone (200 mg/day), paroxetine, opipramol,
amitriptyline, prazepam and methadone induced a depressive mixed episode in a further
patient [21]. In one patient, a switch to hypomania was observed 3 weeks after the discontinuation
of trazodone treatment [16]. Additionally, in 2 depressed patients with probable bipolar disorder, who did not
respond to trazodone, a switch to mania was described after trazodone was abruptly
replaced with imipramine [22]. There are no published cases describing a switch to mania or hypomania in patients
who were treated with a mood stabilizer and trazodone as an antidepressant or a sleep-promoting
treatment.
The average time until the onset of mania/hypomania after the start of treatment with
trazodone was 14.7 days (median 7 days). 9 patients were described as switching to
mania, 6 to hypomania, one to hypomania followed by a mixed episode, and one to a
mixed depressive episode.
Mirtazapine
Mirtazapine is a noradrenergic and specific serotonergic antidepressant (NaSSA). Its
mechanism of action involves noradrenergic alpha2-auto- and hetero-receptor blockade,
enhancing noradrenaline (NE) and serotonin (5-HT) release, as well as blockade of
serotonergic 5-HT2 and 5-HT3 and histaminergic H1 receptors [23]. The recommended dose of mirtazapine for the treatment of depression is 30–45 mg/day
because, in this dose range, mirtazapine acts as an alpha2 antagonist and a 5-HT and
NE disinhibitor. However, for the sleep-promoting effect, only the antagonistic action
of mirtazapine on histaminergic H1 receptors is necessary. For this effect, low doses
of mirtazapine of 3.75–15 mg/day are usually sufficient [6].
Development of a switch to mania or hypomania associated with mirtazapine treatment
was described in 10 patients ([Table 2]).
Table 2 Reports of switching to mania or hypomania during treatment with mirtazapine.
Study
|
Patient (age/gender/diagnosis)
|
Dose (mg/day)
|
Concomitant treatment
|
Time to switch onset (days)
|
Affective symptoms
|
Actions taken
|
Soutullo et al. 1998 [32]
|
45-year-old, woman, MDE
|
15
|
sertraline 250 mg/day
|
4
|
hypomania
|
discontinuation of mirtazapine
|
De Leon et al., 1999 [30]
|
68-year-old, man, poststroke depression
|
15
|
after 10 days of wash out from paroxetine 60 mg/day
|
3
|
hypomania
|
mirtazapine was discontinued, treatment with carbamazepine and haloperidol
|
MacCall & Callender 1999 [33]
|
65-year-old, woman, MDE
|
30
|
none
|
2, after mirtazapine discontinuation after 35 days of treatment
|
hypomania
|
symptoms remitted after 6 weeks
|
Bhanji et al. 2002 [29]
|
43-year-old woman, MDE
|
60
|
after washout form venlafaxine 225 mg/day, topiramate 400 mg/day
|
7, after dose increase to 60 mg/day
|
mania
|
mirtazapine dose reduced to 45 mg/day, lithium carbonate 600 mg/day, topiramate 400 mg/day
|
Ng 2002 [27]
|
48-year-old depressed woman
|
30
|
fluoxetine 40 mg/day, alprazolam 1 mg/day
|
35
|
mania
|
discontinuation of fluoxetine, dose mirtazapine reduced to 15 mg/day and discontinued,
valproate 2 000 mg/day and clonazepam 0.5 mg/day at bedtime
|
Prost & Abraham 2004 [28]
|
35-year-old woman, major depression
|
45
|
paroxetine 50 mg/day
|
56
|
hypomania
|
mirtazapine discontinued
|
Goyal & Sinha 2008 [31]
|
15-year-old girl, severe depressive episode
|
22.5
|
none
|
14
|
mania
|
lithium 900 mg/day after discontinuing mirtazapine, haloperidiol 5 mg/day
|
Liu et al. 2009 [26]
|
66-year-old women with late life depression
|
30
|
switch from fluoxetine 20 mg/day without a period of tapering-off, washout or cross-titration
|
3
|
mania
|
mirtazapine discontinued, treatment with valproate and low-dose risperidone
|
Habermeyer et al. 2010 [25]
|
51-year-old patient, major depression
|
45
|
none
|
5 after dose increase to 45 mg/day
|
mixed episode
|
mirtazapine discontinued, treatment with olanzapine and lithium
|
Basavraj et al. 2011 [24]
|
42-year-old woman, recurrent depressive disorder
|
30
|
after tapering down escitalopram 20 mg/day
|
14
|
mania
|
discontinuation of mirtazapine, treatment with lithium 900 mg/day, quetiapine 100 mg/day
|
MDE – major depressive episode
3 of those patients were treated in monotherapy with the antidepressant dose (≥30 mg/day)
of mirtazapine [24]
[25]
[26]. One of those patients may have been at higher risk of switching, because she was
an older patient and the treatment with mirtazapine was started without a period of
washout from fluoxetine (20 mg/day) [26]. In 2 further patients, treatment with antidepressant dose of mirtazapine was combined
with treatment using SSRIs [27]
[28]. In one patient hypomania was observed after a dose of mirtazapine was increased
to 60 mg/day [29]. Mirtazapine doses lower than 30 mg/day were found to be associated with switch
to mania or hypomania in only 3 cases [30]
[31]
[32]. In the first of those patients, mirtazapine (15 mg/day) was added to a high dose
(250 mg/day) of sertraline [32]. The second patient was an older patient, a 68-year-old woman with organic, post-stroke
depression [30], and the third patient was a young patient, a 15-year-old girl [31]. In one older depressed patient, a switch to hypomania was observed 2 days after
mirtazapine discontinuation, following 35 days of treatment [33]. There are no published cases that describe a switch to mania or hypomania in patients
who were treated with a mood stabilizer and mirtazapine either in antidepressant or
in a sleep-promoting dose.
The average time until the onset of mania/hypomania after the start of the treatment
or a dose increase of mirtazapine was 15.7 days (median 7 days). 5 patients were described
as switching to mania, 4 to hypomania, and one to a mixed depressive episode.
Agomelatine
Agomelatine is an antidepressant with a novel non-monoaminergic mechanism of action.
It is a melatonin agonist at both melatonin receptors MT1 and MT2 and a serotonin
antagonist at 5-HT2C receptors. The recommended dose of agomelatine to treat depression
is 25–50 mg/day [34]. So far, it is unknown whether lower doses of agomelatine can be used to have only
a sleep-promoting effect.
Recently, a switch to hypomania has been described in a 52-year-old female patient
with the diagnosis of a recurrent major depressive disorder and comorbid panic disorder.
The patient was treated with paroxetine 20 mg/day and trazodone 100 mg/day. As no
improvement was observed after 1 month, the patient was admitted to hospital and the
paroxetine treatment was shifted to agomelatine 25 mg/day, while 100 mg/day trazodone
was maintained. After 4 days, the symptoms of hypomania occurred, and the agomelatine
was discontinued 2 days later. One week after stopping the agomelatine treatment,
the hypomanic symptoms remitted [35]. In a study that evaluated the efficacy of agomelatine in patients with bipolar
I disorder, who were experiencing a major depressive episode during treatment with
lithium or valpromide, there were 3 of 13 agomelatine and lithium treated patients
who experienced manic or hypomanic episodes during an optional observation period
of up to 46 weeks Only one of those cases was rated as treatment-related [36].
Discussion
The described cases suggest that trazodone, mirtazapine, and agomelatine may induce
manic symptoms in some patients. It happens most frequently at an early stage of treatment,
which has been already noticed before [37]. The risk of a switch to mania or hypomania during treatment with trazodone or mirtazapine
seems to be related, first of all, to the doses recommended for antidepressant treatment,
administered without concomitant mood-stabilizer therapy. Low doses of these antidepressants,
which can be used for their hypnotic or sedative effects, were observed to cause a
switch to mania or hypomania only in patients with other risk factors for switching,
such as treatment with other antidepressants, organic origin of the depressive symptoms,
and being of an older or very young age.
There is no evidence for trazodone and mirtazapine, and only sparse evidence for agomelatine
[36], claiming that treatment with these antidepressants is related to an increased risk
of switch to mania when administered in combination with a mood stabilizer. This last
finding of our review is supported by results of a recent study based on Swedish national
registries, which examined the risk of antidepressant-induced manic switch in patients
with bipolar disorder who were being treated either with antidepressant monotherapy
or with an antidepressant in conjunction with a mood stabilizer. Although an increased
risk of manic switch was found among patients with bipolar disorder on antidepressant
monotherapy, the risk of manic episodes was not increased in the short or long term
for patients treated with an antidepressant and a concurrent mood stabilizer [38].
To interpret the findings from the case reports reviewed by us, we have to take into
account that switching to mania or hypomania related to treatment with trazodone,
mirtazapine, or agomelatine may be more common, but underreported. The reason may
be, for example, the relatively mild intensity of the hypomanic symptoms. In many
described case reports, the described symptoms did not require any other action than
stopping or reducing the dose of the administered antidepressant. Some authors even
claimed that, in their experience, trazodone-induced hypomanic symptoms may be transient
and that immediate withdrawal of trazodone may not always be necessary [13]. In some cases, switching to mania or hypomania may also be considered as resulting
from illness progression [39] or stressful life events [40].
However, at least for mirtazapine, the low risk of a switch to mania was also found
in studies that analyzed data from clinical trials. In a review of the initial clinical
trials, manic symptoms associated with mirtazapine treatment were described in 3 (0.25%)
of 1 299 patients [41]. In a study that analyzed data from 2 clinical trials in patients with rapid-cycling
bipolar disorder, the rate of treatment-derived mania/hypomania was 30.1% for SSRIs
(highest for fluoxetine 42.1%, lowest for fluvoxamine 0%), 35.7% for bupropion, 30.6%
for venlafaxine, 18.8% for nefazodone (a drug with similar pharmacological action
to trazodone), and there were no cases of switching observed for mirtazapine [42].
All of these data suggest that low doses of trazodone and mirtazapine are safe in
patients with bipolar disorder and should still be considered as important alternatives
to hypnotics when long-term pharmacological treatment of insomnia is necessary. We
believe that such a use of sleep-promoting antidepressants can even decrease the risk
of switching, because the effective treatment of insomnia in bipolar patients improves
the course of the disorder and quality of life. It also seems that trazodone, mirtazapine,
and agomelatine can be used safely in antidepressant doses when combined with a mood
stabilizer, especially in patients with bipolar II disorder.
Further work should clarify whether doxepin, the only sedative antidepressant approved
by the FDA for the treatment of insomnia characterized by difficulty with sleep maintenance,
can also be used to treat insomnia in bipolar disorder. Although TCAs are regarded
as antidepressants with the highest risk of a switch to mania [43], there is no evidence to claim that low doses (3 and 6 mg) of doxepin, which show
only antihistaminergic effects [6]
[44], are related to treatment-emergent mania or hypomania. Another interesting tricyclic
antidepressant worthy of consideration as sleep-promoting drug is trimipramine. Trimipramine
is regarded as an atypical antidepressant with antipsychotic and sedative properties.
It has been studied in monotherapy in delusional depression [45] and in low dose to treat primary insomnia [46]. In some countries it is frequently used to induce sleep, also as adjuvant therapy
to other drugs [47].
It is also noteworthy that the sleep-promoting effect may not be the most relevant
feature for estimating the switch risk and choice of an antidepressant. In patients
with bipolar disorder without sleep problems a dopaminergic antidepressant bupropion
may be a reasonable first-line treatment. Bupropion has lower rates of manic switch
than tricyclic and tetracyclic antidepressants and norepinephrine-serotonin reuptake
inhibitors [1] and is especially useful, e. g., in bipolar disorder comorbid with attention-deficit/hyperactivity
disorder [48].