Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Pharmacology and Therapeutics

journal homepage: www.elsevier.com/locate/pharmthera

Post-stroke depression: Mechanisms and pharmacological treatment

⁎
Roberto Federico Villa , Federica Ferrari, Antonio Moretti
Laboratory of Pharmacology and Molecular Medicine of Central Nervous System, Department of Biology and Biotechnology, University of Pavia, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: Depression, the most frequent psychiatric disorder following ischaemic stroke, negatively affects survivals'
Post-stroke depression functional outcome, response to rehabilitation and quality of life. Approximately, one-third of them are affected
Pathophysiology by post-stroke depression (PSD), making it a serious social and public health problem and anti-depressant
Clinical studies preventive and curative therapies worth investigating. However, a two-way association between depression and
Selective serotonin reuptake inhibitors
stroke has been also established: stroke increases the risk of PSD, but depression is an independent risk factor for
stroke.
The pathophysiology of PSD is presumably multifactorial, involving a combination of various ischaemia-
induced neurobiological dysfunctions in the context of psychosocial distress. The damage of frontal-basal ganglia
brainstem pathway suggested alterations of monoaminergic neurotransmitter systems. Several lines of evidence
point to a relationship between neuroinflammatory response to acute ischaemic stroke, stress activation of the
hypothalamic-pituitary-adrenal (HPA) axis and the impairment of adaptive response (neurogenesis) within a
background of altered energy metabolism (i.e. mitochondrial dysfunction).
The complexity of PSD mechanisms makes its biologically-based prevention and treatment a difficult task. So
far, especially the selective serotonin (5-hydroxytriptamine, 5-HT) reuptake inhibitors (SSRIs) have mainly
proved to be clinically active in preventing and treating PSD, although their effects have not been demonstrated
unequivocally and they may cause bleeding and intracerebral haemorrhage. Besides the primary pharmacolo-
gical activity of SSRIs (i.e. the inhibition of neuronal 5-HT reuptake) there is evidence supporting their pleio-
tropic mechanisms of action: anti-inflammatory and enhanced neurogenesis through the up-regulation of neu-
rotrophins, conceivably supported by the stimulation of mitochondrial energy metabolism.
In the future, novel developments might point at anti-cytokine modulators which can improve symptoms of
depression, especially in subjects affected by inflammation processes.
This review will address the various areas of epidemiology, pathophysiology, preventive and therapeutic
strategies for PSD. The activity of SSRIs in clinical trials, as well as their pharmacology, pharmacokinetics, safety
and mechanisms of action, will be examined in detail. A final section will deal with the effect of depression as
risk factor for stroke. The literature on PubMed from 1990 to 2017 was reviewed.

1. Introduction substantially increasing, especially in low- and middle-income coun-

tries (Feigin et al., 2014). Even more worrying, stroke is the main cause
Stroke is the third cause of death worldwide, with 16.9 million first- of chronic, severe adult disability, requiring long-term rehabilitation
ever cases, 5.9 million deaths per year and a projected increase to 23 procedures. Thus, despite the advancement in prevention and therapy
million cases (12 million deaths and 70 million survivors) in 2030. (Moretti, Ferrari, & Villa, 2015a, 2015b; Villa, Ferrari, & Moretti,
Stroke incidence and prevalence increase sharply with age and the 2017), stroke remains both a serious human problem for patients and
numbers of patients are expected to rise with the growing of the aged families and a dramatic public financial burden.
population, although the proportion of patients < 65 years is Among the consequences of stroke for survivors, post-stroke

Abbreviations: ADL, Activities of Daily Living; BDNF, Brain-Derived Neurotrophic Factor; CBF, Cerebral Blood Flow; CI, Confidence Interval; CNS, Central Nervous System; COX,
Cyclooxygenase; CSF, Cerebro-Spinal Fluid; DES, Depression-Executive Dysfunction Syndrome; DST cortisol, (overnight) Dexamethasone Suppression Test for cortisol; HPA axis,
Hypothalamic-Pituitary-Adrenal axis; HR, Hazard Ratio; 5-HT, 5-Hydroxytryptamine, Serotonin; 5-HTTLPR, Serotonin-Transporter-Linked Polymorphism Region; IL, Interleukin; MDD,
Major Depressive Disorders; OR, Odds Ratio; NSAIDs, Non-Steroidal Anti-Inflammatory Drugs; PSD, Post-Stroke Depression; RR, Relative Risk; SSRI, Selective Serotonin Reuptake
Inhibitors; TCA, Tricyclic Anti-depressants; TNF, Tumor Necrosis Factor; VEGF, Vascular Endothelial Growth Factor
⁎
Corresponding author at: University of Pavia, Department of Biology and Biotechnology, Laboratory of Pharmacology and Molecular Medicine of Central Nervous System, Via Ferrata,
9, 27100 Pavia, Italy.
E-mail address: robertofederico.villa@unipv.it (R.F. Villa).

https://doi.org/10.1016/j.pharmthera.2017.11.005

0163-7258/ © 2017 Elsevier Inc. All rights reserved.

Please cite this article as: Villa, R.F., Pharmacology and Therapeutics (2017), https://doi.org/10.1016/j.pharmthera.2017.11.005
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

depression (PSD) is the most frequent psychiatric problem. PSD is found prevalence rate of 19.3% of major depression (MDD) and 18.5%
strongly associated with further worsening of physical and cognitive of minor depression (MnD).
recovery, functional outcome and quality of life. Moreover, depression Recently, a thorough meta-analysis of 108 studies on mood dis-
negatively affects patients' ability to engage in rehabilitation therapies. orders observed 147 cases from 2 days to 7 years post-stroke and de-
However, this scenario is complicated by the bidirectional re- monstrated a 33.5% prevalence of any depressive disorder. MDD ac-
lationship between depression and stroke: stroke increases the risk of counted for 17.7%, MnD for 13.1%, dysthymia for 3.1%. Adjustment
PSD, but depression is an independent risk factor for stroke and stroke disorder was present in 6.9% of patients and anxiety in 9.8% (Mitchell
mortality. et al., 2017).
Despite this conceptual complexity, the term post-stroke depression
is commonly used to define any depression state present after stroke, 2.2. Natural history
regardless of the timing of symptoms onset.
Empirical evidence of pathophysiological factors associated with Longitudinal studies provided conflicting results as to the time
PSD suggests alterations in ascending monoamine pathways, excess of course of PSD. According to the meta-analyses by Ayerbe et al. (2013)
pro-inflammatory cytokines, dysfunction of the hypothalamic-pituitary- and Hackett and Pickles (2014) the prevalence rate was stable in the
adrenal (HPA) axis and alterations in neuroplasticity. first year but declined thereafter. However, the review by Werheid
The development of a rational therapy would require an integrated (2016) based on 10 prospective longitudinal studies, observed a bi-
view of these pathophysiological factors. Up to now, this topic is only phasic pattern, with a rise in depressive symptoms within the first six
speculative and the current therapy is mainly based on the monoamine months, a slight drop at about one year and a new increase within the
hypothesis of depression. Indeed, various antidepressant medications second year after stroke.
could be used for PSD, being selective serotonin reuptake inhibitors
(SSRIs) and tricyclic antidepressants (TCAs) the classes most widely 2.3. Outcome
studied.
This review will address the epidemiology, predictors, pathophy- PSD is associated with severe disability, anxiety, lower Quality of
siology, prevention and therapeutic strategies for PSD. We will focus on Life, speech and language dysfunction, anhedonia, feeling of despair,
the treatment with SSRIs, their pharmacology, pharmacokinetics, safety functional and cognitive impairment, greater dependency with regard
and mechanisms of action (reviews by Hayhow, Brockman, & to activities of daily living (ADL), lack of medication compliance
Starkstein, 2014; Loubinoux et al., 2012; Paolucci, 2017; Robinson & (Ayerbe, Ayis, Crichton, Wolfe, & Rudd, 2014; De Ryck et al., 2013).
Jorge, 2016; Towfighi et al., 2017; Williams, 2005). The term PSD is PSD and even more the depression-executive dysfunction syndrome
used for depression after ischaemic (not haemorrhagic) stroke, because (DES), i.e. the clinical entity defined as late-life depression plus execu-
ischaemic stroke has been the subject of the vast majority of the lit- tive dysfunctions, were associated with earlier recurrence of stroke:
erature. The role of depression as a risk factor for stroke will also be 8.15 years for PSD and 7.15 years for DES versus 9.63–9.75 years in
discussed. non-depressed and non-DES patients (Sibolt et al., 2013). Not surpris-
ingly, PSD was associated with increased costs of stroke hospitalization
2. Epidemiology (Husaini et al., 2013).
Increased mortality is the most dramatic clinical event following
Currently approximately half of stroke survivors suffer from per- PSD. In a meta-analysis of 13 studies Bartoli et al. (2013) reported an
sisting disability and need institutional care and rehabilitation (Bonita, OR of 1.46 (95% CI 0.76–2.80, NS) with a follow-up < 2 years, 1.21
Solomon, & Broad, 1997; Go et al., 2013). To further complicate the (95% CI 1.12–1.32) at 2–5 years and 1.37 (95% CI 0.95–1.97) for >
complex rehabilitation process, a substantial proportion of survivors 5 years. A recent study, lasting 10 years on all-cause of mortality in a
may develop PSD after stroke onset. large population of stroke patients stratified by age, documented a HR
of 1.56 (95% CI 0.49–5.02, NS) at the ages between 25 and 74 years,
2.1. Prevalence and incidence and 2.28 (95% CI 1.79–2.90) for patients aged 65–74 years (Razmara
et al., 2017). Increased mortality in PSD patients might be the result of
Approximately one third of survivors suffer from PSD, with a cu- cardiovascular mortality linked to decreased heart rate variability,
mulative incidence of 55%. The prevalence of PSD varied across studies caused by the alteration of autonomic functions (Robinson, Spalletta,
depending on population characteristics, diagnostic criteria, inclusion/ et al., 2008).
exclusion criteria (e.g. the presence/absence of aphasia), time after
stroke and the clinical setting where the patients were examined (acute 2.4. Predictors
or rehabilitation hospital settings, community, outpatient clinics) (re-
views by Paolucci, 2008; Robinson, 2003; Truelsen et al., 2006). The most frequent predictors cited in the vast (but somehow di-
In their meta-analysis of 43 studies Ayerbe, Ayis, Wolfe, and Rudd vergent) literature are female gender, a personal history of pre-stroke
(2013) reported a cumulative incidence of depression of 39–52% within depression, physical disability, anxiety, aphasia, stroke severity, cog-
5 years following stroke and a pooled prevalence of 29% at any time nitive and physical impairment, level of independence, dysphagia and
within 10 years. Interestingly, a significant fraction of those patients psychosocial factors like pre-stroke life events and lack of perceived
who became depressed early after the acute event recovered in sub- family and social support (De Rick et al., 2014; Hackett & Anderson,
sequent assessments. 2005; Kutlubaev & Hackett, 2014; Shi, Yang, Zeng, & Wu, 2017).
A meta-analysis of 61 studies by Hackett and Pickles (2014) re-
ported a pooled frequency of depression of 31% at any time up to five 3. Pathophysiology
years following stroke, consistently with results found in a 10-year
earlier review where the pooled frequency was 33% (Hackett & Pathophysiology of PSD is complex and multifactorial, resulting
Anderson, 2005). This emphasises the need for better evidenced-based from the combination of ischaemia-induced neurobiological dysfunc-
strategies of screening, prevention and therapy. Indeed, PSD remains tions and psychosocial distress. Indeed, the current evidence indicates
relatively under-diagnosed, under-treated and under-researched. the neurobiological factors (rather than the psychological response to
A limitation of most studies was the lack of diagnostic criteria for disability) as the main factors associated with PSD.
specific mood disorders, thus missing important clinical variables. A This heterogeneity may result from the subtype of depression
previous review by Robinson (2003) on pooled data of hospital studies (major, minor), the post-stroke onset (early or late), the cerebral area

2
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

involved (left versus right hemisphere: Wei et al., 2015) and the causal (iv) Pro-inflammatory cytokines stimulate the hypothalamic-pituitary-
factors (Provinciali & Coccia, 2002). adrenal axis to release glucocorticoid hormones (Szczudlik,
Major aetiopathological predictors of PSD were genetic factors, such Dziedzic, Bartus, Slowik, & Kieltyka, 2004). Stroke patients ex-
as polymorphisms of (i) serotonin (5-hydroxitriptamine, 5-HT) trans- hibited increased levels of serum adrenocorticotropic hormone
porter 5-HTTLPR, (ii) 5-HT2A receptor and of (iii) Brain-Derived (ACTH) and cortisol and impaired dexamethasone test (DST)
Neurotrophic Factor (BDNF) genotypes (Kim et al., 2012; Mak, Kong, (Åström, Olsson, & Asplund, 1993); these alterations were asso-
Mak, Sharma, & Ho, 2013; Queirazza & Cavanagh, 2014; meta-analysis ciated with worsening of functional and neurological post-stroke
by Zhao, Guo, Yang, Yang, & Meng, 2016). Various pathogenetic me- outcomes and higher mortality (review by Barugh, Gray, Shenkin,
chanisms were postulated for PSD, some of them reflecting those for MacLullich, & Mead, 2014). Dysregulation of HPA axis is also
primary depression (Penninx, Milaneschi, Lamers, & Vogelzangs, 2013), associated with increased risk of depression, at least in subgroups
as follows: of patients (Ferrari & Villa, 2017).
Glucocorticoids have several detrimental chronic effects on the
(i) in analogy to the monoaminergic hypothesis of depression, PSD Central Nervous System (CNS), including decreased neurogenesis
could be related to the interruption induced by the ischaemic le- and neuronal survival, and dysfunction of the neurotrophic
sion of the amine-containing axons ascending from the brainstem system, notably in the hippocampus. These findings have been
to the left cerebral cortex. This would lead to reduced synthesis of corroborated by numerous animal studies reporting the harmful
5-HT and of norepinephrine (NE) in limbic areas of frontal and effects of either pre- or post-stroke stress (Song & Wang, 2011; see
temporal lobes and basal ganglia (Robinson, Kubos, Starr, Rao, & also Table 2 in Kronenberg, Gertz, Heinz, & Endres, 2014). In
Price, 1984; Santos, Kövary, Gold, Bozikas, & Hof, 2009; Terroni agreement with these observations, the treatment with the glu-
et al., 2011; meta-analysis by Narushima, Kosier, & Robinson, cocorticoid receptor antagonist mifepristone reversed the stroke
2003). Imbalance of 5-HT neurotransmission was supported by the damage in a social stress paradigm in mice (Sugo et al., 2002). A
findings of lower CSF concentration of the metabolite 5-hydro- role of hypothalamic endocannabinoid CB1 receptor could also be
xyindolacetic acid (Bryer et al., 1992) and of up-regulated PET- played through the stress activation of HPA axis, as found in an
measured 5-HT2A receptors (Mayberg et al., 1988; Møller, experimental model of PSD (Wang et al., 2016). These results are
Andersen, & Gjedde, 2007). However, the monoamine hypothesis consistent with bidirectional relationships between HPA axis
alone is no more generally accepted. regulation, the levels of inflammatory cytokines (Robinson &
(ii) In the context of the enhanced glutamate-mediated excitotoxicity Jorge, 2016) and brain energy metabolism during aging (Villa,
hypothesis of depression (Sanacora, Treccani, & Popoli, 2012), the Ferrari, & Gorini, 2012).
involvement of glutamatergic system in PSD was suggested by the The HPA axis is also correlated with 5-HT system through its effect
increased glutamate plus glutamine/creatine ratio in the con- on 5-HT binding to 5-HTTLPR (as well as the binding of the other
trolateral prefrontal regions in the immediate post-stroke period, monoamines to their transporters) (Figlewicz, 1999). In turn, the
but not persisting after 4 months (Glodzik-Sobanska et al., 2006). 5-HTTLPR polymorphism is associated with biological stress re-
Moreover, a strong relationship was observed between PSD at activity through the higher and more prolonged level of cortisol
3 months and elevated levels of plasma glutamate and reduced (Gotlib, Joormann, Minor, & Hallmayer, 2008).
glutamate oxalacetate transaminase (GOT) during stroke (Cheng (v) Impairment of adaptive response of brain to ischaemia (neuronal
et al., 2014). plasticity). There is considerable evidence linking hippocampal
(iii) Extensive investigations suggest a role of neuroinflammatory re- neurogenesis, axon branching, dendritogenesis and synaptogen-
sponse, based on the activation of microglia and astrocytes and up- esis to chronic stress, affective disorders and anti-depressant ac-
regulation of pro-inflammatory cytokines (IL-1β, IL-6, IL-18 and tivity, even more so in PSD (Ferrari & Villa, 2017; Masi &
TNF-α). Cytokines are involved both in the inflammatory response Brovedani, 2011). As a matter of fact, the intrinsic, spontaneous
to acute ischaemic stroke (Ferrarese et al., 1999; Ferrari, neurogenesis that was experimentally demonstrated in the dentate
Devecchi, Pero, Gorini, & Villa, 2016; Spalletta et al., 2006), and gyrus of the hippocampus as a response to focal ischaemia was
in depression (Ferrari & Villa, 2017) and may have a role for de- impaired in rats exposed to chronic mild stress (Wang, Zhang,
veloping PSD (Yang et al., 2010, only for IL-18; Spalletta et al., Guo, Teng, & Chen, 2008).
2013; Li, Ling, et al., 2014). In support of this hypothesis, mi- In this framework, it has been demonstrated that BDNF plays a
croglia activation may contribute to the development of PSD crucial role in PSD and the action of anti-depressant drugs on
(Jawaid, Krajewska, Pawliczak, Kandra, & Schulz, 2016). More- neurogenesis is similar to that in non-stroke patients (Ferrari &
over, the serum proteome of PSD patients showed a significant Villa, 2017).
down-regulation of the complement, a known precursor to the Several studies found decreased levels of BDNF in depressed pa-
adaptive immune system through cytokine signalling (Nguyen tients and recovery after anti-depressant treatment. In PSD pa-
et al., 2016). However, a recent, systematic review on 37 studies tients, there was a strong relationship between low levels of BDNF
provided controversial results regarding the association between at admission and the development of PSD. Yang et al. (2011)
cytokines and the development of PSD (Pietra Pedroso, Rachid, & proved that serum BDNF at concentration of < 5.86 ng/ml on day
Teixeira, 2016; see also Hannestad, DellaGioia, & Bloch, 2011). one post-stroke was independently associated with PSD also
Should cytokines be really involved in the pathogenesis of PSD, 2 weeks with an OR of 28.9 (95% CI 8.01–104.9).
they could interact with every physiological domain relevant to Furthermore, Li, Zhao, et al. (2014) found:
PSD, including neurotransmitter metabolism, neuroendocrine (a) a highly significant negative correlation between the levels of
function and synaptic plasticity. BDNF and the severity of stroke defined by National Institutes
Cytokine activation may up-regulate the expression of the gene of Health Stroke Scale (NIHSS) score;
encoding indoleamine 2,3-dioxygenase (IDO) which metabolises (b) lower serum BDNF in patients with depression at admission
tryptophan to kynurenine, thereby diverting it from 5-HT syn- compared with those without depression,
thetic pathway route. This could result in decreasing 5-HT in the (c) a 11.5-fold increased risk of developing PSD at 3 months for a
frontal cortex and basal ganglia and increasing neurotoxic tryp- BDNF level of < 10.2 ng/ml, indicating that serum BDNF is a
tophan catabolites (Maes, Leonard, Myint, Kubera, & Verkerk, significant, biological predictor of PSD.
2011), and suggests a further mechanism for 5-HT involvement in Accordingly, experimental studies reported the decrease of
PSD. BDNF levels in the hippocampus of post-stroke rodents with

3
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

stress-induced models of depressive-like behavior (O'Keefe 4. Animal models

et al., 2014; Zhang, Wu, Song, & Li, 2012), whereas BDNF
over-expression alleviated this behavior (Chen et al., 2015). Experimentally, various rodent models of PSD have been widely
(vi) The stroke event also makes plausible the relationship between employed (reviews by Kim et al., 2015; Kronenberg et al., 2014). They
PSD pathogenesis and vascular factors, such as the white matter mainly included a combination of focal ischaemia induced by middle
lesions [identified by Magnetic Resonance Imaging (MRI) hyper- cerebral artery occlusion or microembolism, and chronic mild stressors
intensities] which could interrupt monoaminergic projections (e.g. spatial restraint stress, social isolation) followed by the assessment
from midbrain and brainstem. This observation would be similar of the lesion volume, histology, immunohistochemistry and by sensor-
to what has been reported for late life depression in the “Vascular imotor (e.g. rotarod test) and behavioural tests (elevated plus maze, the
Depression” hypothesis (reviews by Naarding & Beekman, 2011; shuttle box, forced swim test, sucrose consumption and so on).
Provinciali & Coccia, 2002; Taylor, Alzenstein, & Alexopoulos, Although they are not standardised and not always reproducible
2013). (for example, the lesion volume is of secondary importance for the
Of note, chronic cerebral hypoperfusion in rats induced depres- clinical post-stroke functional outcome), these models manifest a broad
sive-like behaviours accompanied by increased plasma levels of spectrum of depressive-like symptoms (otherwise not observed in post-
corticosterone and altered hippocampal glucocorticoid receptors. stroke rodents without chronic mild stressors), allowing the investiga-
This finding supports the clinical hypothesis that vascular de- tion on pathophysiological mechanisms and pharmacological treatment
pression can be induced by chronic hypoperfusion via a me- of PSD.
chanism involving the enhancement of HPA axis (Lee et al., 2016).
It is worth mentioning a meta-analysis of thirty-three clinical 5. Clinical studies
studies by Noonan, Carey, and Crewther (2013) aimed at in-
vestigating the biological markers of PSD. They provided the first The identification and treatment of neuropsychiatric complications
evidence of significantly reduced amygdala volume (OR − 0.45, in neurological diseases is crucial. In particular, the prevalence and
95% CI − 0.89 to − 0.02) in PSD (similarly to major depressive invalidating condition of PSD make a prompt diagnosis mandatory for
disorders) and confirmed significant association of PSD with de- appropriate prevention and therapy. A previous history of depression
creased dexamethasone test cortisol (OR 3.28, 95% CI 1.28–8.39), should deserve accurate screening for depression (Bowen, Knapp,
total absolute regional Cerebral Blood Flow (CBF; OR − 0.35, 95% Hoffman, & Lowe, 2005; Karamchandani et al., 2015; Towfighi et al.,
CI − 0.64 to − 0.06), and of serum level of BDNF (− 0.52, 95% CI 2017).
− 0.84 to − 0.21). Surprisingly, no significant associations with Diagnosis is currently based on clinical evaluation (Diagnostic and
PSD were found across a variety of inflammatory mediators (C- Statistical Manual of Mental Disorders, DSM-V) supported by standar-
reactive protein, cytokines, TNF-α), a result likely attributable to dised psychiatric scales for depression (Hamilton Rating Scale for
the small number of studies and the reduced time post-stroke. Depression, Beck Depression Inventory, Montgomery-Åsberg
Another limitation was the inclusion in some studies of patients Depression Rating Scale, Center for Epidemiological Studies Depression
with transient ischaemic attack, or silent stroke or non-depressed. Scale, Clinical Global Impression of the Nursing) (Berg, Lönnqvist,
(vii) In the context of the previous points, great potential interest arises Palomäki, & Kaste, 2009; Spalletta, Ripa, & Caltagirone, 2005), but also
from the possible involvement of mitochondrial energy metabolism a scale specifically devised for PSD, the Post-Stroke Depression Rating
in PSD pathogenesis. Scale (Quaranta, Marra, & Gainotti, 2008). It is beyond the ambit of the
Despite the lack of specific clinical or experimental studies in PSD, present review to go into details, however there is considerable diag-
two elements support this hypothesis. On the one hand, the high nostic variability because of: (i) the heterogeneity of symptoms because
energy demands and the large number of mitochondria make the of the various depression subtypes, (ii) the potentially overlapping
brain particularly susceptible to the ischaemia-induced impair- between symptoms of stroke and those of depression.
ment of aerobic metabolism (Villa, Gorini, Ferrari, & Hoyer, Numerous pharmacological studies aimed at preventing and
2013). Oxidative stress also plays a crucial role in ischaemic re- treating PSD (reviews by Flaster, Sharma, & Rao, 2013; Paolucci, 2017;
perfusion injury after stroke (Allen & Bayraktutan, 2009). Ramasubbu, 2011; Towfighi et al., 2017). Among the various hy-
On the other hand, neuroimaging studies in depressed (non- potheses on the pathophysiology of PSD, the neurotransmitter hy-
ischaemic) patients have reported modifications in the brain en- pothesis has attracted the greatest attention as a target for pharma-
ergy metabolism measured by the changes in CBF and Cerebral cotherapy. Indeed, the analogy between MDD and PSD has made the
Metabolic Rate of glucose (CMRglu) (Drevets, 2000; Harper et al., anti-depressant therapy a logical therapeutic approach. Among the
2017; Moretti, Gorini, & Villa, 2003). Evidence of oxidative stress various anti-depressant classes, selective serotonin reuptake inhibitors
in depression was also provided by biochemical, genetic and (SSRIs) are the more commonly used, because of better tolerability than
pharmacological studies (Ng, Berk, Dean, & Bush, 2008). In this other drugs (e.g. the tricyclics, TCAs).
context, Renshaw et al. (2001) showed that the agents that in-
crease the cerebral ATP availability may have antidepressant ef- 5.1. Prevention
fects.
Moreover, mitochondrial dysfunctions were observed in animal Meta-analyses and randomized clinical trials (RCTs) proved un-
models of depression (reviews by Bansal & Kuhad, 2016; Ferrari & ambiguous, though modest, prophylactic effects in non-depressed
Villa, 2017; Gardner & Boles, 2011; Rezin, Amboni, Zugno, stroke patients treated with SSRIs compared to the untreated ones
Quevedo, & Streck, 2009) leading to the proposal of a mitochon- (Table 1). The meta-analyses by Chen, Patel, Guo, and Zhan (2007) and
drial hypothesis of depression (Klinedinst & Regenold, 2015). Yi, Liu, and Zhai (2010) documented a larger reduction of the risk for
depression (but not its severity) for SSRIs and TCAs compared to pla-
In conclusion, an integrated view of the aetiopathology of PSD (si- cebo. Lower risk of PSD was also confirmed by the large meta-analysis
milarly to that recently proposed for depression by Ferrari & Villa, by Salter, Foley, Zhu, Jutai, and Teasell (2013). Particular emphasis has
2017) posits the inter-linking between monoamines, neuroinflamma- been placed on FLAME study which reported significantly improvement
tion, stress-activation (HPA axis), and neurogenesis as the possible in the primary outcome (Fugl-Meyer Motor Scale — FMMS, a measure
primum movens with energy metabolism as the common denominator. In of arm and leg motor function), as well as in the incidence of depression
this connection, a pivotal role could be played by BDNF (Burkhalter, and the activity of daily living (modified Ranking scale 0–2, not shown
Fiumelli, Allaman, Chatton, & Martin, 2003). in Table 1).

4
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

Table 1
The prevention of post-stroke depression.

Authors Type of study Anti-depressants Time since Duration of Drug Diagnostic criteria Main resultsa
stroke, days treatment

Chen et al. (2007) Meta-analysis SSRIs (6 trials) < 14 to > 56 30 days to 1 year – RD: −0.12 (− 0–20 to − 0.04) WMD − 0.5
TCAs (2 trials) NS
RD: − 0.23 (− 0.44 to − 0.11) WMD − 0.5
NS
Yi et al. (2010) Meta-analysis Fluoxetine (6 < 7 to > 28 4 to 12 weeks HDRS–HARS OR: 0.25 (0.11–0.56)
trials) SSS WMD: − 3.97 (− 9.85 to 1.90) NS
WMD: − 4.72 (− 8.31 to.1.13)
Salter et al. (2013) Meta-analysis SSRIs (5 trials) 10 to < 180 90 days to 1 year HDRS–HARS; OR: 0.34 (0.22–0.53)
Others (2 trials) MADRS
Chollet et al. (2011) RCT (FLAME) Fluoxetine > 10 90 days FMMS Motor recovery: fluoxetine 34.0 (29.7–38.4)
MADRS points vs pl 24.3 (19.9–28.7)
incidence of PSD: fluoxetine 4% vs pl 29%
Robinson, Jorge, Moser, Multicentre RCT Escitalopram < 90 1 year HDRS–HARS pl vs escitalopram: HR 4.5 (2.4–8.2)
Acion and Solodkin incidence of PSD: escitalopram 8.5% vs pl
(2008) 22.4%; NNT 7.2
Jorge, Acion, Moser, Adams, RCT Escitalopram < 90 1 year RBANS Total score: escitalopram 10.0 vs pl 3–1,
and Robinson (2010) delayed memory score: escitalopram 13.4 vs
pl 7.2
Kim et al. (2017) Multicentre RCT Escitalopram < 21 90 days MADRS OR: 1.00 (0.56–1,80) NS
ITT OR: 1.01 (0.61–1.69) NS
Tsai et al. (2011) RCT Milnacipranb < 28 1 year HRSD Incidence of PSD: milnacipran 2.22% vs pl
15.22%

SSRIs = citalopram, escitalopram, fluoxetine, sertraline; TCAs = maprotyline, nortryptiline; others = mianserine, mirtazapine; trazodone.
HDRS/HARS = Hamilton Depression Rating Scale/Hamilton Anxiety Rating Scale; MADRS = Montgomery Asberg Depression Rating Scale; FMMS = Fugl-Meyer Motor Scale;
RBANS = Repeatable Battery for the Assessment of Neuropsychological Status; SSS = Scandinavian Neurological Stroke Scale (functional disability); NNT = number needed to be
treated (patients to be treated to prevent one new case of depression); ITT = intention-to-treat analysis; RD = rate difference (reduction in the occurrence rate of PSD); WMD = weighted
mean difference (reduction in severity of PSD symptoms at end point versus control = 1); OR = odds ratio; HD = hazard ratio; pl = placebo;
a
In parentheses the 95% CIs (confidence intervals); differences were statistically significant (P < 0.05) or highly significant (P ≤ 0.001) except when indicated (NS).
b
Serotonin-norepinephrine reuptake inhibitor (SNRI).

Interestingly, old post-stroke patients were more likely to receive a (41% before versus 20% after; OR 4.00, 95% CI 1.68–9.57;
preventive SSRI treatment if they had a history of pre-stroke depression P = 0.003);
(Ried et al., 2010). At variance with the above findings, a recent, large (ii) a significant prediction of earlier clinical recovery (OR 2.35, 95%
RCT did not observe any difference between escitalopram and placebo CI 1.15–4.81; P = 0.02);
treatment (Kim et al., 2017), thus casting doubts on the potentiality of (iii) a trend towards predicting improved motor recovery (OR 1.82,
this drug in preventing PSD. 95% CI 0.90–3.68; P = 0.095).

5.2. Treatment Remarkably, the above-mentioned strong associations remained

statistically significant even after adjusting for age and i.v. thrombolysis
Table 2 shows the results of three meta-analyses of a total of 74 [OR 3.87 (95% 1.60–9.33; P = 0.003) for point (i) and 2.38 (95% CI
trials with six SSRIs (citalopram, escitalopram, fluoxetine, fluvoxamine, 1.15–4.91; P = 0.019) for (ii), respectively].
patoxetine, sertraline) and the TCA nortryptiline compared with pla- Conversely, Mortensen, Larsson, Johnsen, and Andersen (2014) re-
cebo. Three single trials on SSRIs and nortryptiline (not reported in the ported unchanged stroke severity in patients receiving SSRIs before
meta-analyses) are also displayed. ischaemic stroke but increased stroke mortality in patients with hae-
The pooled analyses suggest that anti-depressants are more effective morrhagic stroke.
than placebo in reducing the symptoms in PSD patients, in particular As regards the aforementioned studies, the following final points
disability, neurological deficit, depression and death (the latter was deserve consideration:
only evaluated by Mead et al., 2013). However, there is no clear evi-
dence to support a beneficial effect of anti-depressants on ADL: they (i) SSRIs are the first-choice for the pharmacological prevention and
were improved according to Chen, Guo, Zhan, and Patel (2006) when therapy of PSD, although the evidence is not robust and there is no
assessed by Barthel Index, an ordinal scale to measure performances in definite proof that they are more efficacious than the TCAs;
ADL, but not significantly recovered by Xu et al. (2016), nor by Gao (ii) results should be interpreted with great caution due to the sub-
et al. (2017). stantial clinical and methodological heterogeneity among trials,
As regards the crucial issue of the appropriate treatment timing, most of which were small and presented various sources of bias;
Chen et al. (2006) found that the longer the duration, the greater is the (iii) with the notable exception of Mortensen et al. (2014), stroke
degree of improvement in depression symptoms, especially after subtypes (ischaemic or haemorrhagic) are rarely mentioned, a
3–4 months. This result was confirmed by Gao et al. (2017). crucial dearth given the anti-thrombotic activity of SSRIs;
Since depression is a risk factor for stroke, it was interesting to in- (iv) aside from Siepmann et al. (2015), thrombolysis appears to have
vestigate the relevance of pre-stroke versus post-stroke anti-depressant been used only in the prospective observational cohort study by
therapy on clinical outcome. To answer this question, Siepmann et al. Miedema et al. (2010) which found that pre-stroke anti-depressant
(2015) found that, compared to that after stroke, the treatment with therapy with SSRIs was associated with a trend towards un-
SSRIs (citalopram, escitalopram, fluoxetine) before stroke was asso- favourable outcome at 3 months after stroke (OR = 0.4, 95% CI
ciated with: 0.14–1.13; P = 0.08). However, this study suffered from two im-
portant limitations: the small number of SSRI-treated patients (5%)
(i) a significantly more favourable functional outcome at discharge and the lack of a group without thrombolysis, which prevented the

5
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

Table 2
The treatment of post-stroke depression.

Authors Type of study Anti-depressants Time since stroke Duration of drug Diagnostic criteria Main results (a)
treatment

Chen et al. (2006) Meta-analysis SSRIs (12 trials) 1–16 weeks HDRS WMD = − 5.5 (− 8.3to-2.7) after
treatment vs 0.5 (0.0–1.0) before
treatment
RD = 0.23 (0.03–0.43)
Significant benefit after 3–4 weeks of
treatment
Mead et al. (2013) Review and SSRIs (52 trials) Mostly 3 months Barthel Index, FIM SMD = 1.11 (0.71–1.51) (b)
meta-analysis HDRS–MADRS RR = 0.43 (0.24–0.77)
NIHSS SMD = −1.00 (−1.26–0.75)
Death RR = 0.76 (0.34–1.70)
Xu et al. (2016) Meta-analysis SSRI (7 trials) 6–26 weeks HDRS–MADRS SMD = −0.53 (−0.97 to − 0.09)
TCA (3 trials) SMD = −1.41 (−2.51 to − 0.31)
Acler, Robol, Fiaschi, & RCT Citalopram 10 days ≥ 4 months NIHSS (c) 2.3 points in the scale vs baseline 5 points
Manganotti (2009) HDRS (c) 2.3 points vs pl 3.5
6.6 points in the scale vs baseline 9 points
6.6 vs pl 8
Mikami et al. (2011) RCT Fluoxetine < 6 months 3 months Modified Ranking t vs pl: fluoxetine 2.14
Nortryptiline score (d) Nortryptiline 2.91
Recovery continued over 12 months
Gao et al. (2017) RCT Citalopram < 3 months 3 months HDRS–MES Differences between citalopram and pl on
3–6-months Bartel Index, FIM HDRS and MES
> 6 months < 3 months after stroke: NS;
difference at 3–6 months and > 6 months:
S
NS at any time

SSRIs = citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline; TCAs = nortryptiline.

(a) in parentheses the 95% CIs (confidence intervals); differences were statistically significant S (P < 0.05) or highly significant (P ≤ 0.001) except when indicated (NS).
HDRS = Hamilton Depression Rating Scale; MADRS = Montgomery-Asberg Depression Rating Scale; MES = Bech-Rafaelsen Melancholia Scale.
NIHSS = National Institute of Health Stroke Scale (measures various neurological functions); modified Ranking Score, Barthel Index and FIM (Functional Independence Measure) assess
the disability, dependence and activities of daily life; (b) in depressed people; (c) after 1-month treatment; (d) mixed model analysis.
SMD = standardised mean difference; WMD = weighted mean difference (reduction in severity of PSD symptoms at end point vs control = 1).
RD = rate difference (difference of response rate between treatment and placebo); RR = risk ratio; pl = placebo.

assessment of the possible interaction between SSRIs and tPA. possible association of SSRIs with haemorrhagic stroke and/or mor-
Another trial did not find any confounding effect of thrombolysis tality has been significantly debated.
on fluoxetine clinical activity, however without providing the data Most studies were on pre-stroke SSRI therapy and the risk of sub-
(Chollet et al., 2011). sequent haemorrhage. A systematic review of 112 RCTs on the cardio-
vascular adverse events (including stroke and myocardial infarction)
induced by antidepressant drugs in high-risk patients found no differ-
5.3. Safety
ence in ORs between SSRIs and placebo (Swenson, Doucette, &
Fergusson, 2006). By contrast, a meta-analysis of 16 studies demon-
Overall, SSRIs are well tolerated since their side effects are rela-
strated an increase of intracerebral haemorrhage (ICH, RR 1.42, 95% CI
tively uncommon, benign and often transient. Their main adverse
1.23–1.65), but not of the composite of ICH and subarachnoid hae-
outcomes among older patients include gastroenterological symptoms
morrhage (RR 1.05, 95% CI 0.84–1.30) (Hackam & Mrkobrada, 2012).
(to be treated with proton pump inhibitors), falls/fracture, epilepsy/
Another meta-analysis of 13 studies confirmed the association of SSRI
seizures and hyponatraemia, due to the syndrome of inappropriate anti-
with ICH (1.30, 95% CI 1.02–1.67), also when ischaemic stroke events
diuretic hormone secretion.
are considered (1.48, 95% CI 1.08–2.02) (Shin et al., 2014). Recently, a
The complex cerebro- and cardio-vascular effects of SSRIs are in-
meta-analysis of 22 observational studies reported RRs of 1.29 (95% CI
herent to the serotonergic activation resulting from 5-HT reuptake in-
0.92–1.81) for haemorrhagic stroke and 1.15 (95% CI 0.98–1.36) for
hibition (Hoirisch-Clapauch, Nardi, Gris, & Brenner, 2014; Mortensen &
ischaemic stroke (Biffi, Scotti, & Corrao, 2017).
Andersen, 2015).
As regards the comparison with other antidepressants, a previous
The block of 5-HT uptake does not only take place in neurons, but
study reported that SSRIs use in patients aged > 65 years significantly
also across the platelet plasma membrane mediated by the Serotonin
reduced the risk of subsequent stroke in comparison with TCAs (HR
Reuptake Transporter (SERT) SLC6A4 (review by Hoirisch-Clapauch
0.67, 95% CI 0.47–0.96) (Lee et al., 2013). Conversely, another study
et al., 2014). Since platelets are the only blood cells committed to 5-HT
found that, compared with TCAs, SSRIs (as well as other antidepressant
transport (due to their inability to synthesise it), the block of reuptake
drugs) were associated with significantly higher rates of mortality (HR
lowers their 5-HT content impairing aggregation (with anti-thrombotic
1.32, 95% CI 1.26–1.39) and stroke/TIA (HR 1.15, 95% CI 1.05–1.26).
effect) and potentially causing bleeding. This might reduce the risk of
As a possible explanation of this unexpected result, TCAs were pre-
myocardial infarction (Mortensen, Larsson, Johnsen, & Andersen, 2013;
scribed at lower doses than SSRIs (Coupland et al., 2011).
Sauer, Berlin, & Kimmel, 2003).
In conclusion, despite some controversy, there is evidence of the
The situation is less clear for stroke patients. On the one hand, potential risk of haemorrhagic stroke related to the use of SSRI.
bleeding could result in haemorrhagic stroke. On the other hand, 5-HT
Anyway, given the rarity of this event, absolute risks are likely to be low
also exerts a constriction effect on cerebral vasculature, which could (Hackam & Mrkobrada, 2012). Nevertheless, great caution should be
theoretically counteracts the anti-thrombotic activity of SSRIs and po-
exercised in slowly titrating the doses for elderly patients, especially if
tentially cause ischaemic stroke (reviews by Mortensen & Andersen, they also receive oral anticoagulant therapy.
2015; Shin, Oh, Eom, & Park, 2014). Therefore, in recent years, the

6
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

Only few studies addressed the association between SSRI treatment of efficacy, as well as drug-drug interactions, and side or toxic effects
stroke patients and adverse effects in PSD. A retrospective study on old (Mandrioli, Mercolini, Saracino, & Raggi, 2012). This is particular
veterans examined the impact of pre-stroke and/or post-stroke SSRI evident for fluoxetine, due to the long half-life of the drug (1–4 days),
treatment on post-stroke mortality at 1 year. Reduced risk of mortality and of its active metabolite norfluoxetine (7–15 days), and to the fact
was found only in depressed patients receiving a SSRI after stroke (HR that fluoxetine is both a substrate and a strong inhibitor of CYP2D6.
0.57, 95% CI 0.25–1.32), or both before and after stroke (HR 0.31, 95%
CI 0.11–0.86) (Ried et al., 2011).
6.2. Pharmacology
Another study on post-stroke SSRI treatment documented a slight
increase of mortality (HR 1.13, 95% CI 1.00–1.28, NS), possibly asso-
Despite the differences in chemical structure, SSRIs share the same
ciated with intracranial bleeding (HR 1.14, 95% CI 0.62–2.12, NS), and
mechanism of action, i.e. the block of neuronal transport of serotonin
enhancement of major bleeding (e.g. gastrointestinal) (HR 1.33, 95% CI
(5-HTT). Table 3 also shows the in vitro SSRI potency (Ki) to displace the
1.14–1.55). Treatment was not significantly associated with reduced
specific ligand to monoamine transporter, which measures SSRI affi-
risk of recurrent stroke, but that of myocardial infarct significantly
nities. The rank order of potencies for 5-HT is: paroxetine > sertra-
declined (HR 0.65, 95% CI 0.42–1.00) (Mortensen et al., 2013).
line > fluoxetine > citalopram > fluvoxamine. The potencies for
The favourable results of the FLAME trial on motor recovery after
noradrenaline and dopamine reuptake are much lower, thus showing
stroke (see Prevention) raised interest, albeit with some controversy, for
the high selectivity of SSRIs for 5-HT (escitalopram has the highest
the possibility of routine, early use of antidepressants in stroke patients
degree of selectivity).
(Chollet, 2012; Marshall, 2012; Selim & Molina, 2012). In this context,
Besides the in vitro potency of a drug, the clinical effect depends on a
a large study on SSRI treatment at a median time of 5 days after ad-
number of other factors, primarily the concentration reached at the site
mission was associated with a substantially lower all-cause mortality at
of action, i.e. its pharmacokinetics. Protein-binding is particularly im-
30 days (OR 0.28, 95% CI 0.17–0.43). Interestingly, stratification ac-
portant, since the amount of a drug in the extracellular fluid of brain
cording to stroke severity revealed a different treatment effect on
(i.e. CSF) tends to be equivalent to that unbound to plasma proteins, the
mortality: for patients with very severe stroke the OR was 0.08 (95% CI
“free fraction” that represents the active drug concentration.
0.04–0.17), but 0.43 (95% CI 0.21–0.92) in those with moderate stroke
It is generally accepted that the action of anti-depressant drugs
and 0.76 (95% CI 0.24–2.37) in those with mild stroke (Mortensen,
(SSRIs and TCAs) on major depressive disease results from a complex
Johnsen, Larsson, & Andersen, 2015). The Authors speculated that, in
chain of events triggered by monoamine reuptake inhibition. As regards
addition to the antidepressant effect, SSRIs may act through a neuro-
SSRIs, the acute rise of 5-HT in the raphe nuclei of the brainstem ac-
protective or anti-platelet effects by protecting against recurrent stroke.
tivates inhibitory somato-dendritic auto-receptors, thereby reducing
serotonin rise in terminal fields and 5-HT firing. Over time (weeks), the
6. Pharmacokinetics and pharmacology of SSRIs drug administration leads to down-regulation and desensitisation of
inhibitory auto-receptors, and to the increased synthesis and release of
Chemically, most SSRIs are aryl- or aryloxy-alkylamines; two 5-HT, resulting in enhanced neurotransmission in terminal fields and in
(fluoxetine, citalopram) are racemates and the active S-(+)enantiomer the activation of intracellular signal transduction pathways (second,
of citalopram was developed as escitalopram. third messengers, gene expression). This, however, is only part of the
story: indeed, a fundamental role in chronic anti-depressant effects is
played by adaptive events stemming from the modulation of neuronal
6.1. Pharmacokinetics
plasticity (Ferrari & Villa, 2017), a concept that can be also applied to
the treatment of PSD.
As displayed on Table 3, SSRIs are generally well absorbed and
extensively bound to plasma proteins. Moreover, they reach steady-
state concentration in the blood within 1–2 weeks. SSRIs are metabo- 7. Could neuroprotection and brain plasticity play a role in the
lised by various cytochrome P-450 (CYP) isoenzymes, thus making action of SSRIs on PSD?
more likely the interactions with other drugs and the potential influ-
ence on their own metabolism if administered at high doses. The assay SSRIs appear to promote recovery from PSD through a host of
of plasma levels is therefore very useful for monitoring the clinical pleiotropic mechanisms beyond their primary pharmacological activity,

Table 3
Pharmacological parameters of six SSRIs.

Pharmacological parameters Fluoxetine Citalopram Escitaloprama Fluvoxamine Paroxetine Sertraline

Oral bioavailability % > 80 80 ± 13 80 ± 13 53 Dose-dependent > 44

t1/2 in hours 53 ± 41 33 ± 4 27–32 18 17 ± 3 23
Protein binding % 94 80 55 77 95 98
Cytochrome P-450 2D6, 2C9 3A4, 2C19 3A4, 2C19 1A2, 2D6, 2D6 2D6
Isoenzymes, CYP 3A4, 2C9
Active metabolites Yes Yes No No No No
Concentrations in CSF in nM 26 110 31 – 7 –
Ki in nMb
5HTT 0.81 9.6 2.5 2.22 0.125 0.29
NAT 244 5029 6514 1300 40 417
DAT 3600 > 100,000 > 100,000 9100 500 25
5-HT/NE 301 523 2605 586 320 1423

References: Goodman & Gilman's, The pharmacological basis of therapeutics (In Brunton LL, Lazo JS, Parker KL, Eds). 11th Edition.
Lenox RH and Alan Frazer. Mechanism of action of antidepressants and mood stabilizers. In Davis KL, Charney D, Coyle JT, Nemeroff C (Eds) Neuropsychopharmacology — The fifth
generation of progress. Lippincott Williams and Wilkins, Philadelphia, 2002, pp. 1139–1163. McGraw Hill, New York 2006; Mandrioli et al. (2012); Nikisch et al. (2004); Owens, Knight,
and Nemeroff (2001); Pastoor and Gobburu (2014); Paulzen et al. (2016).
a
S-(+) enantiomer of citalopram.
b
Ki = potency at human transporters for serotonin (5-HT), norepinephrine (NE), dopamine (DA).

7
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

i.e. the inhibition of 5-HT re-uptake. Results in normal animals (i.e. non-ischaemic, non-depressed) were
Experimentally, a systematic review and meta-analysis of 23 studies inconclusive: some studies found that SSRIs enhanced oxidative phos-
(McCann et al., 2014) demonstrated that – despite some methodological phorylation and displayed anti-oxidant effect, whereas others reported
bias – SSRIs improve infarct volume and neurobehavioural outcome in an uncoupling effect (Adzic, Brkic, Bulajic, Mitic, & Radojcic, 2016).
animal models of ischaemic stroke (see also the reviews by Burns & These conflicting results may be due to:
Greenberg, 2010; Kronenberg et al., 2014; Siepmann et al., 2015).
SSRIs administered following brain ischaemia acted through various (i) the use of drug concentrations (10 mM and above) in most in vitro
mechanisms: experiments that were substantially higher than those found in
patients' CSF (30–40 nM) active in the in vitro binding affinity to 5-
(i) neuroprotection, as shown by: HT transporter (0.1–10 nM);
(a) the anti-inflammatory activity by fluoxetine in post-ischaemic (ii) biases in experimental ex vivo study design which (with the partial
rat brain (Lim et al., 2009); exception of Filipovic, Costina, Peric, Stanisavljevic, & Findeisein,
(b) the BDNF-associated protection by escitalopram against neu- 2017) did not include analysis at subcellular level in this way
ronal damage, oxidative stress and microglial activation in CA1 overlooking the bioenergetic macro-heterogeneity of brain areas
hippocampal area of ischaemic gerbils (Lee et al., 2011); and the micro-heterogeneity of different subcellular compartments
(c) the Vascular Endothelial Growth Factor (VEGF)-associated en- (pre-synaptic nerve endings and post-synaptic terminals) (Ferrari &
hancement by escitalopram of proliferation and differentiation Villa, 2017);
of new cells and angiogenesis in hypo-perfused rat brain area
(Ma, Lu, Hu, & Yao, 2015); At this latter regards, we investigated the effects of in vivo treatment
(d) the improvement by fluoxetine and sertraline in CBF auto- with the anti-depressants desipramine and fluoxetine on brain energy
regulation and vascular tone independently of Nitric Oxide metabolism of rat frontal cerebral cortex (Villa, Ferrari, Gorini,
Synthase (NOS)-related pathways in post-ischaemic mice brain Brunello, & Tascedda, 2016) and hippocampus (Villa et al., 2017) by
(Shin et al., 2009). assaying the catalytic activities of regulatory enzymes of energy-
(ii) stimulation of the spontaneous neurogenesis is proved by the following yielding metabolic pathways and taking into account the micro-het-
results: erogeneity of rat brain mitochondria populations. Energy metabolism
(a) chronic fluoxetine treatment enhanced the survival of newborn was stimulated by desipramine and even more so by fluoxetine in non-
neurons in the ischaemic hippocampus, and corrected spatial synaptic mitochondria of frontal cerebral cortex and hippocampus, but
memory impairment in mice (Li et al., 2009). However, there inhibited in intra-synaptic mitochondria of frontal cerebral cortex and
was no effect on motor recovery in rats when ischaemia was unchanged in the hippocampus.
induced by intracortical and striatal injection of endothelin-1 The recent proteomic study by Filipovic et al. (2017) confirmed that
(Sun et al., 2016; Windle & Corbett, 2005). fluoxetine chronic treatment increased the energy metabolism towards
(b) chronic citalopram increased the number of new neurons and the citric acid cycle and oxidative phosphorylation in rat hippocampal
microvessels in the peri-infarct region in middle cerebral artery non-synaptic mitochondria.
occlusion stroke model in mice through the increase in BDNF These findings give new insights into the molecular mechanisms of
expression, and this was associated with better sensorimotor anti-depressants (Ferrari & Villa, 2017). It remains to ascertain whether
recovery (Espinera, Ogle, Gu, & Wei, 2013). Moreover, the these considerations could be also applied to the effects of SSRIs in
chronic co-administration at different doses of citalopram and animal models of depression and of PSD.
WAY-100635 (an antagonist of 5-HT1A receptor that has been Putting together the pathophysiological mechanisms described in
prominently implicated in the modulation of mood) potentiated Section 3 and the pharmacological findings described above, we can
the effects of citalopram on post-stroke hippocampal neuro- speculate that:
genesis and behavioural tests, possibly through reinforcing
serotoninergic neurotransmission (Wang, Zhang, Guo, Sui, & (i) impaired neurogenesis might be an important part of the patho-
Sun, 2010). genesis of PSD, similarly to the major depressive disorder;
In a relevant countercheck study on SSRI activity, deletion of 5- (ii) BDNF expression in the hippocampus impacts on neuroplasticity,
HT1A receptor in knockout (KO) mice resulted in the blockade synaptic structure and function;
of both neurogenic and behavioural effects of fluoxetine, sug- (iii) low BDNF levels result in decreased neurogenesis, contributing to
gesting that these two effects may be causally related (Santarelli the development of PDS, and
et al., 2003). (iv) SSRI and other anti-depressants may correct the neurogenesis im-
A potential mechanism for the stimulation of neurogenesis in pairment by enhancing BDNF
the dentate gyrus of the hippocampus is the up-regulation of (v) SSRIs may exert anti-inflammatory activity.
neurotrophins and growth factors, in particular BDNF and
VEGF that modulate post-ischaemic progenitor proliferation 8. Other antidepressants
(Espinera et al., 2013; Schmidt & Duman, 2007). As a matter of
facts, the hippocampal level of BDNF correlate with mood im- Some tricyclic antidepressants (TCAs), (e.g. amitriptyline, clomi-
provement after anti-depressant treatment in patients (Chen, pramine, venlafaxine) are relatively selective inhibitors of serotonin
Dowlatshahi, MacQueen, Wang, & Young, 2001) and in rodents transport. Others (e.g. nortryptiline, maprotiline, mianserin, viloxazine)
(Xu, Steven Richardson, & Li, 2003). are more active on norepinephrine. TCAs were extensively used for
decades as antidepressant therapy. Significant benefit was also reported
Clinically, the possible effects of antidepressants on inflammation for PSD patients (review by Xu et al., 2016).
was not investigated in PSD patients. However, SSRIs (but not other However, treatment with TCAs is associated with higher adverse
antidepressants) display anti-inflammatory effects in non-stroke de- effects and discontinuation rate than SSRIs, particularly in older pa-
pressed patients (meta-analyses by Hannestad et al., 2011; Vogelzangs tients. Their main side effects are related to the potent anti-muscarinic
et al., 2012). activity (causing dry mouth, dizziness, tachycardia, blurred vision),
To our knowledge, there are no data on the effects of SSRIs (or other anti-H1 activity (sedation), anti-α1-adrenergic activity (postural hypo-
anti-depressants) on mitochondrial energy metabolism in PSD patients or tension). They may also have direct cardiac depressing effects and
in experimental models. prolong the conduction time with potential arrhythmias.

8
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

Serotonin-norepinephrine reuptake inhibitors (SNRIs: venlafaxine, with the improvement in physical illness. The Authors concluded that
desvenlafaxine, duloxetine, milnacipran) differ from TCAs for not being anti-cytokine treatment could be particularly beneficial in therapy-re-
antagonists of the above-mentioned receptors. Therefore, they lack sistant depressed patients with elevated inflammation, who are free of
TCAs' main adverse effects. physical illness (Kappelmann, Lewis, Dantzer, Jones, & Khandaker,
The presence of dual action should result in higher efficacy in 2016).
comparison with drugs acting only on one neurotransmitter: compared However, this strategy raises two issues: the need for the mono-
to SSRIs, SNRIs increase norepinephrine concentration in neuronal sy- clonal antibodies to penetrate the blood-brain-barrier and the com-
napsis, and this might be useful in patients with symptoms of hy- plexity of the TNF-α signalling pathways and receptors, which make its
poactivation. On the other hand, this could lead to the enhancement of function both neuroprotective and neurodegenerative, calling into
the cerebrovascular risk by increasing blood pressure. question the long-term validity of anti-TNF treatment (Baune, 2017).
Actually, meta-analyses did not report substantial differences be- Moreover, the possible adverse effects are still inadequately known.
tween SSRIs and SNRIs regarding the antidepressant efficacy (Cipriani As a whole, the evidence of the anti-inflammatory approach for the
et al., 2009; Papakostas, Thase, Fava, Nelson, & Shelton, 2007) and the therapy of depression is still limited and dependent on a number of
risk of ischaemic stroke or intracranial haemorrhage (Lee et al., 2016). factors that deserve careful investigations, including the patients' in-
There are very few studies on SNRIs in PSD. A RCT reported ben- flammatory and immune status, the drug mechanisms of action and
eficial effects of milnacipran in preventing PSD (Tsai et al., 2011: see long-term side effects.
Table 1). An open single-blind trial found that duloxetine prevented
PSD by 16% and improved rehabilitation, cognitive function and 9.2. Targeting sphingomyelinase
quality of life (Zhang et al., 2013). Compared to SSRIs, an advantage of
the dual action of SNRIs is the improvement in chronic, painful physical Sphingolipids are constituents of biological membranes that play a
symptoms, which could be particularly beneficial in depressed patients role in cell signalling and are involved in physiological (e.g. cell growth)
(Gaynor et al., 2011). and pathophysiological (e.g. stress, inflammation) processes. The lyso-
Finally, mirtazapine is a tetracyclic atypical antidepressant that somal acid sphingomyelinase (sphingomyelin phosphodiesterase,
enhances noradrenergic and serotoninergic neurotransmission by in- aSMase) catalyses the formation of ceramide (N-acetyl-sphingosine).
creasing the release of the monoamines via the antagonism of post-sy- This second messenger has a role in cell arrest and cell death in re-
naptic serotonin receptors (resulting in gradual down-regulation of 5- sponse to stress and appears to cause the damage of mitochondrial
HT2α receptors) and the reduced activity of α2-adrenergic auto- Electron Transfer Chain in cerebral ischaemia/reperfusion (Yu,
receptors. Mirtazapine was efficacious in preventing and treating PSD Novgorodov, Chudakova, Zhu, & Bielawska, 2007).
(Niedermaier, Bohrer, Schulte, Schlattmann, & Heuser, 2004), however Following previous studies showing that some TCAs functionally
it induces sedation by antagonising H1-histaminergic receptors. inhibited aSMase by inducing its proteolytic degradation (Kölzer,
Werth, & Sandhoff, 2004), the group of Gulbins et al. (2013, 2015)
9. Perspectives demonstrated that:

9.1. Anti-inflammatory and anti-cytokine drugs (i) ceramide had a key role in experimental depression;
(ii) amytriptiline and fluoxetine reduced hippocampal aSMase both in
To our knowledge specific strategies for PSD therapy based on non- vitro and ex vivo, increased neuronal proliferation, maturation and
steroidal anti-inflammatory drugs (NSAIDs) were not explored. survival and improved behavior in models of stress-induced de-
In depressive non-stroke patients, the possible therapeutic activity of pression;
various NSAIDs was investigated either when given alone or as ad- (iii) a genetic deficiency of aSMase mimicked the effects of the anti-
junctive to anti-depressants (review by Baune, 2017). So far, the overall depressants and prevented their further effects.
results are inconsistent probably due to methodological discrepancies
and heterogeneity of the tested drugs. A further reason for complication These findings may open the way to the development of specific and
in these trials is the possible co-presence in depressive patients of in- direct inhibitors of aSMase potentially faster and more potent than the
flammatory state (e.g. osteoarthritis) and the lack of appropriate patient available ones. So far, this topic is still in its infancy. It might be also
stratification and accurate selection. growing due to aSMase involvement in the pathophysiology of other
A meta-analysis of four studies reported the beneficial effects of CNS diseases (Alzheimer's, Ischaemia, Niemman-Pick) (Adada, Luberto,
adjunctive treatment of depressed patients with celecoxib (a COX-2 & Canals, 2016; Kornhuber, Tripal, Gulbins, & Muehlbacher, 2013).
selective NSAID) versus placebo on the Hamilton Rating Scale for
Depression (weighted mean difference = 3.26, 95% CI 1.81–4.71), re- 10. Depression as a risk factor for stroke
mission rate (OR = 6.58, 95% CI 2.55–17.00), response rate (OR 6.49,
95% CI 2.89–14.55) and tolerability (OR 0.45, 95% CI 0.18–1.13, non- Considerable evidence suggests that depression is independently
significant) (Na, Lee, Lee, Cho, & Jung, 2014). These results should be associated with stroke morbidity and mortality.
evaluated in the light of the known functions of COX-2 on the brain Depression was one of the ten modifiable risk factors for stroke in
(both beneficial and detrimental) in contrast to COX-1 which is pre- the INTERSTROKE study in 22 countries with an OR of 1.35 (95% CI
dominantly pro-inflammatory (Choi, Aid, & Bosetti, 2009). In principle, 1.10–1.66) (O'Donnell et al., 2010). Not surprisingly, depression is also
although inhibiting COX-2 should increase neuroinflammation, actu- among the risk factors associated with coronary heart disease.
ally, the concept of anti-depressant efficacy based on the specificity for In a large meta-analysis of 28 prospective cohort studies, the
COX enzymes is still a matter of debate. pooled-adjusted HR were 1.45 (95% CI, 1.29–1.63) for total stroke and
A more recent, alternative strategy for anti-depressant therapy is 1.25 (95% CI, 1.11–1.40) for ischaemic stroke (Pan, Sun, Okereke,
that of anti-cytokine modulators, including monoclonal antibodies and Rexrode, & Hu, 2011).
cytokine inhibitors. Another meta-analysis of 17 prospective studies reported a relative
A meta-analysis of 16 RCTs demonstrated that anti-cytokine therapy risk (RR) of 1.34 (95% CI 1.17–1.54) with no gender effect. Restricting
(mainly anti-TNF-α drugs) either alone or in combination with an anti- the analysis to studies with > 100 cases yielded an RR of 1.25 (95% CI
rheumatic drug (metothrexate) elicited a robust, favourable effect on 1.16–1.35) (Dong, Zhang, Tong, & Qin, 2012).
depressive symptoms, when compared to placebo. This effect was as- Results (adjusted for covariates) of subsequent longitudinal studies
sociated with the severity of depressive symptoms at baseline, but not corroborated those aforementioned. In the PRIME study on a

9
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

population of middle-aged men followed for 10 years, depressive consequence of stroke. The high prevalence and incidence of these
symptoms were associated with stroke (HR 1.96, 95% CI 1.21–3.19), two disorders in the general population and the medical significance
even more with ischaemic stroke (HR 2.48, 95% CI 1.45–4.25) (Majed make their association a serious human, social and public health
et al., 2012). Moreover, in a large study on subjects of different ages, problem.
those with stable high depressive symptoms had significantly higher 2. The complexity of this bidirectional association involving common
incident stroke hazard (HR 2.14, 95% CI 1.69–2.71) than those with pathophysiology makes its dissecting a difficult task. This is parti-
low or no depressive symptoms (HR 1.08, 95% CI 0.81–1.44) (Gilsanz cularly relevant both at diagnostic and therapeutic levels.
et al., 2015). 3. Despite numerous clinical and experimental studies, the pathophy-
A recent study on half million Chinese adults showed a significant siological mechanisms of PSD remain far from clear. The available
dose-response association between the number of symptoms and stroke evidence supports the reciprocal modulation of neurotransmitter
risk. Compared with patients having 0 to 2 symptoms (in a 7-point systems, neuroinflammation, neuroendocrine activation, neuronal
scale), those with 7 symptoms had an adjusted HR of 1.47 (95% CI plasticity, vascular factors, underlining the energy metabolism as a
1.04–2.08) (Sun et al., 2016). common denominator.
However, there are also some discrepancies. For example, Surtees 4. The uncertainty regarding the biological basis of PSD has made it
et al. (2008) noted a significant association between psychological difficult to devise rational therapies for prevention and treatment.
distress (but not major depressive disorder) and stroke. By analogy to the therapy for depression, the neurotransmitter
systems (serotonin, noradrenaline) are still the main pharmacolo-
10.1. Effect of gender gical target for PSD. SSRIs are preferred over TCAs for lower side
effects, although the possible SSRI-induced haemorrhagic compli-
No gender effect was found by Dong et al. (2012). However, an cations are still debated. The clinical efficacy of other anti-
increased stroke risk in women (compared with men) with a previous depressants has not been proven yet, although SNRIs may be useful
diagnosis of depression was shown by Jackson and Mishra (2013), Pan for pain-related symptoms. As a whole, the effects of these therapies
et al. (2011) and Seifert et al. (2012). By contrast, stronger association were not demonstrated unequivocally and were limited. In this
in men compared to women was reported in the Rotterdam study by context, both the 2009 Guidelines of the European Stroke
Bos et al. (2008) and by Hamano et al. (2015). Organisation (ESO) for stroke rehabilitation (Quinn et al., 2009),
and those of the American Heart Association/American Stroke As-
10.2. Effect of age sociation (AHA/ASA) (Towfighi et al., 2017; Winstein et al., 2016)
recommend the pharmacological treatment of PSD with SSRIs or
The Framingham study proved that the risk for developing a stroke TCAs, in particular for patients in rehabilitation settings.
was significantly elevated in subjects aged < 65 years compared to 5. Future developments should tackle other, more specific areas of the
those > 65 years (Salaycik et al., 2007). Similar results were reported pathophysiology of PSD. Two promising examples of this trend
by Pan et al. (2011) and by Seifert et al. (2012) who found a nearly could be the research on anti-inflammatory drugs and on the in-
threefold increased risk in subjects aged 55–64 years, but not in those hibitors of sphingomyelinase.
≥ 65 years.
Conflicts of interest statement
10.3. Mechanisms
The Authors declare no conflicts of interest.
Although the link between depression and stroke is broadly co-
herent (with the notable exception of the gender effect), the inter- Funding
pretation of the underlying mechanism(s) remains speculative.
One hypothesis points to a role of inflammation as a mediator be- This Research on Clinical Pharmacology of Cerebral Ischaemia was
tween the two phenomena. Depression was indeed associated with in- supported by a Grant of the “Enrico ed Enrica Sovena Foundation” –
creased inflammatory process (review by Ferrari & Villa, 2017). Ad- Rome – for the Fellowship of Dr. Federica Ferrari.
ditionally, non-steroidal anti-inflammatory drugs (NSAIDs) showed
antidepressant activity (Köhler, Krogh, Mors, & Benros, 2016). In turn Acknowledgments
inflammation can influence the susceptibility of patients to stroke and
its outcome (Iadecola & Anrather, 2011; McColl, Allan, & Rothwell, The Authors thank Dr. Federica Busatti-Bei for editing the article.
2009).
In this connection, depression may lead to stroke through the en- References
hancement of platelet aggregation and reactivity (Musselman et al.,
1996), and the development of major stroke risk factors, such as hy- Acler, M., Robol, E., Fiaschi, A., & Manganotti, P. (2009). A double blind placebo RCT to
pertension, diabetes, obesity, dyslipidaemia (Meng, Chen, Yang, Zheng, investigate the effects of serotoninergic modulation on brain excitability and motor
recovery in stroke patients. Journal of Neurology, 256, 1152–1158.
& Hui, 2012) and the cluster of metabolic syndrome (Pan et al., 2012). Adada, M., Luberto, C., & Canals, D. (2016). Inhibitors of the sphingomyeline cycle:
Moreover, the alternative hypothesis of reverse causation posits that Sphingomyelin synthases and sphingomyelinases. Chemistry and Physics of Lipids, 197,
depression may be a prodromal symptom of vascular origin generated 45–59.
Adzic, M., Brkic, Z., Bulajic, S., Mitic, M., & Radojcic, M. B. (2016). Antidepressant action
by subclinical changes in the brain before the event (Brunner et al., on mitochondrial dysfunction in psychiatric disorders. Drug Development Research, 77,
2014) (see Pathophysiology for reviews on vascular depression). While 400–406.
there is cogent evidence of the association between depression and Allen, C. L., & Bayraktutan, U. (2009). Oxidative stress and its role in the pathogenesis of
ischaemic stroke. International Journal of Stroke, 4, 461–470.
stroke, it remains to be established whether this effect is direct or in-
Åström, M., Olsson, T., & Asplund, K. (1993). Different linkage of depression to hy-
direct. percortisolism early versus late after stroke. A 3-year longitudinal study. Stroke, 24,
52–57.
Ayerbe, L., Ayis, S., Crichton, S., Wolfe, C. D. A., & Rudd, A. G. (2014). The long-term
11. Conclusions
outcomes of depression up to 10 years after stroke; the South London Stroke Register.
Journal of Neurology, Neurosurgery, and Psychiatry, 85, 514–521.
1. A two-way association between depression and stroke has been Ayerbe, L., Ayis, S., Wolfe, C. D., & Rudd, A. G. (2013). Natural history, predictors and
clearly established. On the one hand depression is an independent outcomes of depression after stroke: Systematic review and meta-analysis. The British
Journal of Psychiatry, 202, 14–21.
risk factor for stroke. On the other hand, depression is a harmful

10
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

Bansal, Y., & Kuhad, A. (2016). Mitochondrial dysfunction in depression. Current Ferrari, F., & Villa, R. F. (2017). The neurobiology of depression: An integrated overview
Neuropharmacology, 14, 610–618. from biological theories to clinical evidence. Molecular Neurobiology, 54, 4847–4865.
Bartoli, F., Lillia, N., Lax, A., Crocamo, C., Mantero, V., Carrà, G., et al. (2013). Depression Figlewicz, D. P. (1999). Endocrine regulation of neurotransmitter transporters. Epilepsy
after stroke and risk of mortality: A systematic review and meta-analysis. Stroke Research, 37, 203–210.
Research and Treatment, 2013, 862978. Filipovic, D., Costina, V., Peric, I., Stanisavljevic, A., & Findeisein, P. (2017). Chronic
Barugh, A. J., Gray, P., Shenkin, S. D., MacLullich, A. M., & Mead, G. E. (2014). Cortisol fluoxetine treatment directs energy metabolism towards the citric acid and oxidative
levels and the severity and outcomes of acute stroke: A systematic review. Journal of phosphorylation in rat hippocampal nonsynaptic mitochondria. Brain Research, 1659,
Neurology, 261, 533–545. 41–54.
Baune, B. T. (2017). Are non-steroidal anti-inflammatory drugs clinically suitable for the Flaster, M., Sharma, A., & Rao, M. (2013). Poststroke depression: A review emphasizing
treatment of symptoms in depression-associated inflammation? Current Topics in the role of prophylactic treatment and synergy with treatment for motor recovery.
Behavioral Neurosciences, 31, 303–319. Topics in Stroke Rehabilitation, 20, 139–150.
Berg, A., Lönnqvist, J., Palomäki, H., & Kaste, M. (2009). Assessment of depression after Gao, J., Lin, M., Zhao, J., Bi, S., Ni, Z., & Shang, X. (2017). Different intervention for post-
stroke: A comparison of different screening instruments. Stroke, 40, 523–529. ischaemic stroke depression in different time periods: A single-blind randomized
Biffi, A., Scotti, L., & Corrao, G. (2017). Use of antidepressants and the risk of cardio- controlled trial with stratification by time after stroke. Clinical Rehabilitation, 31,
vascular and cerebrovascular disease: A meta-analysis of observational studies. 71–81.
European Journal of Clinical Pharmacology, 73, 487–497. Gardner, A., & Boles, R. G. (2011). Beyond the serotonin hypothesis: Mitochondria, in-
Bonita, R., Solomon, N., & Broad, J. B. (1997). Prevalence of stroke and stroke-related flammation and neurodegeneration in major depression and affective spectrum dis-
disability. Estimates from the Auckland stroke studies. Stroke, 28, 1898–1902. orders. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 35, 730–743.
Bos, M. J., Linden, T., Koudstaal, P. J., Hofman, A., Skoog, I., Breteler, M. M. B., et al. Gaynor, P. J., Gopal, M., Zheng, W., Martinez, J. M., Robinson, M. J., & Marangell, L. B.
(2008). Depressive symptoms and the risk of stroke: The Rotterdam study. Journal of (2011). A randomized placebo-controlled trial of duloxetine in patients with major
Neurology, Neurosurgery, and Psychiatry, 79, 997–1001. depressive disorders and associated painful physical symptoms. Current Medical
Bowen, A., Knapp, P., Hoffman, A., & Lowe, D. (2005). Psychological services for people Research and Opinion, 27, 1849–1858.
with stroke: Compliance with the U.K. National Clinical Guidelines. Clinical Gilsanz, P., Walter, S., Tchetgen Tchetgen, E. J., Patton, K. K., Moon, J. R., Capistrant, B.
Rehabilitation, 19, 323–330. D., et al. (2015). Changes in depressive symptoms and incidence of first stroke among
Brunner, E. J., Shipley, M. J., Britton, A. R., Stansfeld, S. A., Heuschmann, P. U., Rudd, A. middle-aged and older US adults. Journal of the American Heart Association, 4 (pii
G., et al. (2014). Depressive disorders, coronary heart disease, and stroke: Dose-re- e001923).
sponse and reverse causation effects in the Whitehall II cohort study. European Glodzik-Sobanska, L., Slowik, A., McHugh, P., Sobiecka, B., Kozub, J., Rich, K. E., et al.
Journal of Preventive Cardiology, 21, 340–346. (2006). Single voxel proton magnetic resonance spectroscopy in post-stroke depres-
Bryer, J. B., Starkstein, S. E., Votypka, V., Parikh, R. M., Price, T. R., & Robinson, R. G. sion. Psychiatry Research, 148, 111–120.
(1992). Reduction of CSF monoamine metabolite in poststroke depression: A pre- Go, A. S., Mozaffarian, D., Roger, V. L., Benjamin, E. J., Berry, J. D., Blaha, M. J., et al.
liminary report. The Journal of Neuropsychiatry and Clinical Neurosciences, 4, 440–442. (2013). Heart disease and stroke statistics-2014 update. A report from the American
Burkhalter, J., Fiumelli, H., Allaman, I., Chatton, J.-Y., & Martin, J.-L. (2003). Brain- Heart Association. Circulation, 129, e28–e292.
derived neurotrophic factor stimulates energy metabolism in developing cortical Gotlib, I. H., Joormann, J., Minor, K. L., & Hallmayer, J. (2008). HPA-Axis reactivity: A
neurons. The Journal of Neuroscience, 23, 8212–8220. mechanism underlying the association among 5-HTTLPR, stress, and depression.
Burns, M. M., & Greenberg, D. A. (2010). Antidepressants in the treatment of stroke. Biological Psychiatry, 63, 847–851.
Expert Review of Neurotherapeutics, 10, 1237–1241. Gulbins, E., Palmada, M., Reichel, M., Lüth, A., Böhmer, C., Amato, D., et al. (2013). Acid
Chen, B., Dowlatshahi, D., MacQueen, G. M., Wang, J. F., & Young, L. T. (2001). Increased sphingomyelinase-ceramide system mediates effects of antidepressant drugs. Nature
hippocampal BDNF immunoreactivity in subjects treated with antidepressant medi- Medicine, 19, 934–938.
cation. Biological Psychiatry, 50, 260–265. Gulbins, E., Walter, S., Becker, K. A., Halmer, R., Liu, Y., Reichel, M., et al. (2015). A
Chen, Y., Guo, J., Zhan, S., & Patel, N. C. (2006). Treatment effects of antidepressants in central role for the acid sphingomyelinase/ceramide system in neurogenesis and
patients with post-stroke depression: A meta-analysis. The Annals of Pharmacotherapy, major depression. Journal of Neurochemistry, 134, 183–192.
40, 2115–2122. Hackam, D. G., & Mrkobrada, M. (2012). Selective serotonin reuptake inhibitors and
Chen, Y., Patel, N. C., Guo, J. J., & Zhan, S. (2007). Antidepressant prophylaxis for brain hemorrhage. A meta-analysis. Neurology, 79, 1862–1865.
poststroke depression: A meta-analysis. International Clinical Psychopharmacology, 22, Hackett, M. L., & Anderson, C. S. (2005). Predictors of depression after stroke: A sys-
159–166. tematic review of observational studies. Stroke, 36, 2296–2301.
Chen, H.-H., Zhang, N., Li, W.-Y., Fang, M.-R., Zhang, H., Fang, Y. S., et al. (2015). Hackett, M. L., & Pickles, K. (2014). Part I: Frequency of depression after stroke: An
Overexpression of brain-derived neurotrophic factor in the hippocampus protects updated systematic review and meta-analysis of observational studies. International
against post-stroke depression. Neural Regeneration Research, 10, 1427–1432. Journal of Stroke, 9, 1017–1025.
Cheng, S. Y., Zhao, Y. D., Li, J., Chen, X. Y., Wang, R. D., & Zeng, J. W. (2014). Plasma Hamano, T., Li, X., Lönn, S. L., Nabika, T., Shiwaku, K., Sundquist, J., et al. (2015).
levels of glutamate during stroke is associated with development of post-stroke de- Depression, stroke and gender: Evidence of a stronger association in men. Journal of
pression. Psychoneuroendocrinology, 47, 126–135. Neurology, Neurosurgery, and Psychiatry, 86, 319–323.
Choi, S. H., Aid, S., & Bosetti, F. (2009). The distinct roles of cyclooxygenase-1 and -2 in Hannestad, J., DellaGioia, N., & Bloch, M. (2011). The effect of antidepressant medication
neuroinflammation: Implications for translational research. Trends in Pharmacological treatment on serum levels of inflammatory cytokines: A meta-analysis.
Sciences, 30, 174–181. Neuropsychopharmacology, 36, 2452–2459.
Chollet, F. (2012). Selective serotonin reuptake inhibitors may be helpful in most patients Harper, D. G., Jensen, J. E., Ravichandran, C., Perlis, R. H., Fava, M., Renshaw, P. F., et al.
with stroke. Stroke, 43, 3150–3152. (2017). Tissue type-specific bioenergetic abnormalities in adults with major depres-
Chollet, F., Tardy, J., Albucher, J.-F., Thalamas, C., Berard, E., Lamy, C., et al. (2011). sion. Neuropsychopharmacology, 42, 876–885.
Fluoxetine for motor recovery after acute ischaemic stroke (FLAME): A randomised Hayhow, B. D., Brockman, S., & Starkstein, S. E. (2014). Post-stroke depression. In T. A.
placebo-controlled trial. Lancet Neurology, 10, 123–130. Schweizer, & R. L. Macdonald (Eds.). The behavioral consequences of stroke (pp. 227–
Cipriani, A., Furukawa, T. A., Salanti, G., Geddes, J. R., Higgins, J. P., Churchill, R., et al. 240). New York: Springer Science.
(2009). Comparative efficacy and acceptability of 12 new-generation anti- Hoirisch-Clapauch, S., Nardi, A.-E., Gris, J.-C., & Brenner, B. (2014). Are the antiplatelet
depressants: A multiple-treatments meta-analysis. Lancet, 373, 746–758. and profibrinolytic properties of selective serotonin-reuptake inhibitors relevant to
Coupland, C., Dhiman, P., Morriss, R., Arthur, A., Barton, G., & Hippisley-Cox, J. (2011). their brain effects? Thrombosis Research, 134, 11–16.
Antidepressant use and the risk of adverse outcomes in older people: Population Husaini, B., Levine, R., Sharp, L., Cain, V., Novotny, M., Hull, P., et al. (2013). Depression
based cohort study. BMJ, 343, d4551. increases stroke hospitalization cost: An analysis of 17,010 stroke patients in 2008 by
De Rick, A., Brouns, R., Geurden, M., Elseviers, M., De Deyn, P. D., & Engelborghs, S. race and gender. Stroke Research and Treatment, 2013, 846732.
(2014). Risk factors for poststroke depression: Identification of inconsistencies based Iadecola, C., & Anrather, J. (2011). The immunology of stroke: From mechanisms to
on a systematic review. Journal of Geriatric Psychiatry and Neurology, 27, 147–158. translation. Nature Medicine, 17, 796–808.
De Ryck, A., Brouns, R., Fransen, E., Geurden, M., Van Gestel, G., Wilssens, I., et al. Jackson, C. A., & Mishra, G. D. (2013). Depression and risk of stroke in midaged women:
(2013). A prospective study on the prevalence and risk factors of post-stroke de- A prospective longitudinal study. Stroke, 44, 1555–1560.
pression. Cerebrovascular Diseases Extra, 3, 1–13. Jawaid, A., Krajewska, J., Pawliczak, F., Kandra, V., & Schulz, P. E. (2016). A macro role
Dong, J.-Y., Zhang, Y.-H., Tong, J., & Qin, L.-Q. (2012). Depression and the risk of stroke. for microglia in post-stroke depression. Journal of the American Geriatrics Society, 64,
A meta-analysis of prospective studies. Stroke, 43, 32–37. 459–461.
Drevets, W. C. (2000). Neuroimaging studies of mood disorders. Biological Psychiatry, 48, Jorge, R. E., Acion, L., Moser, D., Adams, H. P., Jr., & Robinson, R. G. (2010).
813–829. Escitalopram and enhancement of cognitive recovery following stroke. Archives of
Espinera, A. R., Ogle, M. E., Gu, X., & Wei, L. (2013). Citalopram enhances neurovascular General Psychiatry, 67, 187–196.
regeneration and sensorimotor functional recovery after ischemic stroke in mice. Kappelmann, N., Lewis, G., Dantzer, R., Jones, P. B., & Khandaker, G. M. (2016).
Neuroscience, 247, 1–11. Antidepressant activity of anti-cytokine treatment: A systematic review and meta-
Feigin, V. L., Forouzanfar, M. H., Krishnamurthi, R., Mensah, G. A., Connor, M., Bennett, analysis of clinical trials of chronic inflammatory conditions. Molecular Psychiatry.
D. A., et al. (2014). Global and regional burden of stroke during 1990–2010: Findings http://dx.doi.org/10.1038/mp.2016.167.
from the Global Burden of Disease Study 2010. Lancet, 383, 245–255. Karamchandani, R. R., Vahidi, F., Baigur, S., Vu, K. Y. T., Choi, H. A., Hamilton, R. K.,
Ferrarese, C., Mascarucci, P., Zoia, C., Cavarretta, R., Frigo, M., Begni, B., et al. (1999). et al. (2015). Early depression screening is feasible in hospitalized stroke patients.
Increased cytokine release from peripheral blood cells after acute stroke. Journal of PLoS One, 10, e0128246.
Cerebral Blood Flow and Metabolism, 19, 1004–1009. Kim, Y. R., Kim, H. N., Pak, M. E., Ahn, S. M., Hong, K. H., Shin, H. K., et al. (2015).
Ferrari, F., Devecchi, E., Pero, G., Gorini, A., & Villa, R. F. (2016). Bioenergetic char- Studies on the animal model of post-stroke depression and application of anti-
acterization of lymphocytes from stroke patients. Cerebrovascular Diseases, 41, S4. psychotic aripiprazole. Behavioural Brain Research, 287, 294–303.

11
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

Kim, J. S., Lee, E.-J., Chang, D.-I., Park, J.-H., Ahn, S. H., Cha, J. K. et al. for the EMOTION hypertension incidence: A meta-analysis of prospective cohort studies. Journal of
Investigators (2017). Efficacy of early administration of escitalopram on depressive Hypertension, 30, 842–851.
and emotional symptoms and neurological dysfunction after stroke: A multicentre, Miedema, I., Horvath, K. M., Uyttenboogaart, M., Koopman, K., Lahr, M. M. H., De
double-blind, randomised, placebo-controlled study. The Lancet Psychiatry, 4, 33–41. Keyser, J., et al. (2010). Effect of selective serotonin re-uptake inhibitors (SSRIs) on
Kim, J. M., Steward, R., Bae, K. Y., Kim, S. W., Kang, H. J., Shin, I. S., et al. (2012). functional outcome in patients with acute ischemic stroke treated with tPA. Journal of
Serotoninergic and BDNF genes and risk of depression after stroke. Journal of Affective the Neurological Sciences, 293, 65–67.
Disorders, 136, 833–840. Mikami, K., Jorge, R. E., Adams, H. P., Jr., Davis, P. H., Leira, E. C., Jang, M., et al. (2011).
Klinedinst, N. J., & Regenold, W. T. (2015). A mitochondrial bioenergetic basis of de- Effect of antidepressants on the course of disability following stroke. The American
pression. Journal of Bioenergetics and Biomembranes, 47, 155–171. Journal of Geriatric Psychiatry, 19, 1007–1015.
Köhler, O., Krogh, J., Mors, O., & Benros, M. E. (2016). Inflammation in depression and Mitchell, A. J., Sheth, B., Gill, J., Yadegarfar, M., Stubbs, B., et al. (2017). Prevalence and
the potential for anti-inflammatory treatment. Current Neuropharmacology, 14, predictors of post-stoke mood disorders: A meta-analysis and meta-regression of
732–742. depression, anxiety and adjustment disorder. General Hospital Psychiatry, 47, 48–60.
Kölzer, M., Werth, N., & Sandhoff, K. (2004). Interactions of acid sphingomyelinase and Møller, M., Andersen, G., & Gjedde, A. (2007). Serotonin 5HT1A receptor availability and
lipid bilayers in the presence of the tricyclic antidepressant desipramine. FEBS Letters, pathological crying after stroke. Acta Neurologica Scandinavica, 116, 83.90.
559, 96–98. Moretti, A., Ferrari, F., & Villa, R. F. (2015a). Pharmacological therapy of acute ischaemic
Kornhuber, J., Tripal, P., Gulbins, E., & Muehlbacher, M. (2013). Functional inhibitors of stroke: Achievements and problems. Pharmacology & Therapeutics, 153, 79–89.
sphingomyelinase (FIASMAs). Handbook of Experimental Pharmacology, 215, Moretti, A., Ferrari, F., & Villa, R. F. (2015b). Neuroprotection for ischaemic stroke:
169–186. Current status and challenges. Pharmacology & Therapeutics, 146, 23–34.
Kronenberg, G., Gertz, K., Heinz, A., & Endres, M. (2014). Of mice and men: Modelling Moretti, A., Gorini, A., & Villa, R. F. (2003). Affective disorders, antidepressant drugs and
post-stroke depression experimentally. British Journal of Pharmacology, 171, brain metabolism. Molecular Psychiatry, 8, 773–785.
4673–4689. Mortensen, J. K., & Andersen, G. (2015). Safety of selective serotonin reuptake inhibitor
Kutlubaev, M. A., & Hackett, M. L. (2014). Part II: Predictors of depression after stroke treatment in recovering stroke patients. Expert Opinion on Drug Safety, 14, 911–919.
and impact of depression on stroke outcome: An updated systematic review of ob- Mortensen, J. K., Johnsen, S. P., Larsson, H., & Andersen, G. (2015). Early antidepressant
servational studies. International Journal of Stroke, 9, 1026–1036. treatment and all-cause 30-day mortality in patients with ischemic stroke.
Lee, Y.-C., Lin, C.-H., Lin, M.-S., Lin, J.-W., Chang, C.-H., & Lai, M.-S. (2013). Effects of Cerebrovascular Diseases, 40, 81–90.
selective serotonin reuptake inhibitors versus tricyclic antidepressants on cere- Mortensen, J. K., Larsson, H., Johnsen, S. P., & Andersen, G. (2013). Post stroke use of
brovascular events. Journal of Clinical Psychopharmacology, 33, 782–789. selective serotonin reuptake inhibitors and clinical outcome among patients with
Lee, Y.-C., Lin, C.-H., Lin, M.-S., Lu, Y., Chang, C.-H., & Lin, J.-W. (2016). Comparison of ischemic stroke. A nationwide propensity score-matched follow-up study. Stroke, 44,
the effects of serotonin-norepinephrine reuptake inhibitors versus selective serotonin 420–426.
reuptake inhibitors on cerebrovascular events. The Journal of Clinical Psychiatry, 77, Mortensen, J. K., Larsson, H., Johnsen, S. P., & Andersen, G. (2014). Impact of prestroke
e1–e7. selective serotonin reuptake inhibitor treatment on stroke severity and mortality.
Lee, C. H., Park, J. H., Yoo, K.-Y., Choi, J. H., Hwang, I. K., Ryu, P. D., et al. (2011). Pre- Stroke, 45, 2121–2123.
and post-treatments with escitalopram protect against experimental ischemic neu- Musselman, D. L., Tomer, A., Manatunga, A. K., Knight, B. T., Porter, M. R., Kasev, S.,
ronal damage via regulation of BDNF expression and oxidative stress. Experimental et al. (1996). Exaggerated platelet reactivity in major depression. The American
Neurology, 229, 450–459. Journal of Psychiatry, 153, 1313–1317.
Li, W. L., Cai, H. H., Wang, B., Chen, L., Zhou, Q. G., Luo, C. X., et al. (2009). Chronic Na, K. S., Lee, K. J., Lee, J. S., Cho, Y. S., & Jung, H. Y. (2014). Efficacy of adjunctive
fluoxetine treatment improves ischemia-induced spatial cognitive deficits through celecoxib treatment for patients with major depressive disorders: A meta-analysis.
increasing hippocampal neurogenesis after stroke. Journal of Neuroscience Research, Progress in Neuro-Psychopharmacology & Biological Psychiatry, 48, 79–85.
87, 112–122. Naarding, P., & Beekman, A. T. F. (2011). Vascular depression: Where do we go from
Li, W., Ling, S., Yang, Y., Hu, Z., Davies, H., & Fang, M. (2014). Systematic hypothesis for now? Expert Review of Neurotherapeutics, 11, 77–83.
post-stroke depression caused inflammation and neurotransmission and resultant on Narushima, K., Kosier, J. T., & Robinson, R. G. (2003). A reappraisal of poststroke de-
possible treatments. Neuroendocrinology Letters, 35, 104–109. pression, intra-and inter-hemispheric lesion location using meta-analysis. The Journal
Li, J., Zhao, Y.-D., Zeng, J.-W., Chen, X.-Y., Wang, R.-D., & Cheng, S. Y. (2014). Serum of Neuropsychiatry and Clinical Neurosciences, 15, 422–430.
Brain-derived neurotrophic factor levels in post-stroke depression. Journal of Affective Ng, F., Berk, M., Dean, O., & Bush, A. L. (2008). Oxidative stress in psychiatric disorders:
Disorders, 168, 373–379. Evidence base and therapeutic implications. The International Journal of
Lim, C. M., Kim, S. W., Park, J. Y., Kim, C., Yoon, S. H., & Lee, J. K. (2009). Fluoxetine Neuropsychopharmacology, 11, 851–876.
affords robust neuroprotection in the postischemic brain via its anti-inflammatory Nguyen, V. A., Carey, L. M., Giummarra, L., Faou, P., Cooke, I., Howells, D. W., et al.
effect. Journal of Neuroscience Research, 87, 1037–1045. (2016). A pathway proteomic profile of ischemic stroke survivors reveals innate
Loubinoux, I., Kronenberg, G., Endres, M., Schumann-Bard, P., Freret, T., Filipkowski, R. immune dysfunction in association with mild symptoms of depression — A pilot
K., et al. (2012). Post-stroke depression: Mechanisms, translation and therapy. study. Frontiers in Neurology, 7, 85.
Journal of Cellular and Molecular Medicine, 16, 1961–1969. Niedermaier, N., Bohrer, E., Schulte, K., Schlattmann, P., & Heuser, I. (2004). Prevention
Ma, L., Lu, Z.-N., Hu, P., & Yao, C.-J. (2015). Neuroprotective effect of escitalopram and treatment of poststroke depression with mirtazapine in patients with acute
oxalate in rats with chronic hypoperfusion. Journal of Huazhong University of Science stroke. The Journal of Clinical Psychiatry, 65, 1619–1623.
and Technology. Medical Sciences, 35, 514–518. Nikisch, G., Mathé, A. A., Czernik, A., Eap, C. B., Jiménez-Vasquez, P., Brawand-Amey,
Maes, M., Leonard, B. E., Myint, A. M., Kubera, M., & Verkerk, R. (2011). The new ‘5-HT’ M., et al. (2004). Steroselective metabolism of citalopram in plasma and cere-
hypothesis of depression: Cell-mediated immune activation induces indoleamine 2,3- brospinal fluid of depressive patients: Relationship with 5-HIAA in CSF and clinical
dioxygenase, which leads to lower plasma tryptophan and an increased synthesis of response. Journal of Clinical Psychopharmacology, 24, 283–290.
detrimental tryptophan catabolites (TRYCATs), both of which contribute to the onset Noonan, K., Carey, L. M., & Crewther, S. G. (2013). Meta-analyses indicate association
of depression. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 35, between neuroendocrine activation, deactivation in neurotrophic and neuroimaging
702–721. markers in depression after stroke. Journal of Stroke and Cerebrovascular Diseases, 22,
Majed, B., Arveiler, D., Bingham, A., Ferrieres, J., Ruidavets, J.-B., Montaye, M., et al. e124–e135.
(2012). Depressive symptoms, a time-dependent risk factors for coronary heart dis- O'Donnell, M. J., Xavier, D., Liu, L., Zhang, H., Chin, S. L., Rao-Melacini, P., et al. (2010).
ease and stroke in middle-aged men. The PRIME study. Stroke, 43, 1761–1767. Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries (the
Mak, K. K., Kong, W. Y., Mak, A., Sharma, V. K., & Ho, R. C. M. (2013). Polymorphisms of INTERSTROKE study): A case-control study. Lancet, 376, 112–123.
the serotonin transporter gene and post-stroke depression: A meta-analysis. Journal of O'Keefe, L. M., Doran, S. J., Mwilambwe-Tshilobo, L., Conti, L. H., Venna, V. R., &
Neurology, Neurosurgery, and Psychiatry, 84, 322–328. McCullough, L. D. (2014). Social isolation after stroke leads to depressive-like be-
Mandrioli, R., Mercolini, L., Saracino, M. A., & Raggi, M. A. (2012). Selective serotonin havior and decreased BDNF levels in mice. Behavioural Brain Research, 260, 162–170.
reuptake inhibitors (SSRIs): Therapeutic drug monitoring and pharmacological in- Owens, M. J., Knight, D. L., & Nemeroff, C. B. (2001). Second-generation SSRIs: Human
teractions. Current Medicinal Chemistry, 19, 1846–1863. monoamine transporter binding profile of escitalopram and R-fluoxetine. Biological
Marshall, R. S. (2012). Should every patient with stroke be on selective serotonin re- Psychiatry, 50, 345–350.
uptake inhibitors? No. Stroke, 43, 3152–3153. Pan, A., Keum, N., Okereke, O. I., Sun, Q., Kiwimaki, M., Rubin, R. R., et al. (2012).
Masi, G., & Brovedani, P. (2011). The hippocampus, neurotrophic factors and depression. Bidirectional association between depression and metabolic syndrome: A systematic
CNS Drugs, 25, 913–931. review and meta-analysis of epidemiological studies. Diabetes Care, 35, 1171–1180.
Mayberg, H. S., Robinson, R. G., Wong, D. F., Parikh, R., Bolduc, P., Starkstein, S. E., et al. Pan, A., Okereke, O., Sun, Q., Logroscino, G., Manson, J. E., Willett, W. C., Ascherio, A.,
(1988). PET imaging of cortical S2 serotonin receptors after stroke: Lateralized et al. (2011). Depression and incident stroke in women. Stroke, 52, 2770–2775.
changes and relationship to depression. The American Journal of Psychiatry, 145, Pan, A., Sun, Q., Okereke, O. I., Rexrode, K. M., & Hu, F. B. (2011). Depression and risk of
937–943. stroke morbidity and mortality. A meta-analysis and systematic review. JAMA, 306,
McCann, S. K., Irvine, C., Mead, G. E., Sena, E. S., Currie, G. L., Egan, K. E., et al. (2014). 1241–1249.
Efficacy of antidepressants in animal models of ischemic stroke. A systematic review Paolucci, S. (2008). Epidemiology and treatment of post-stroke depression.
and meta-analysis. Stroke, 45, 3055–3063. Neuropsychiatric Disease and Treatment, 4, 145–154.
McColl, B. W., Allan, S. M., & Rothwell, N. J. (2009). Systemic infection, inflammation Paolucci, S. (2017). Advances in antidepressants for treating post-stroke depression.
and acute ischemic stroke. Neuroscience, 158, 1049–1061. Expert Opinion on Pharmacotherapy, 18, 1011–1017.
Mead, G. E., Hsieh, C.-F., Lee, R., Kutlubaev, M., Claxton, A., Hankey, G. J., et al. (2013). Papakostas, G. I., Thase, M. E., Fava, M., Nelson, J. C., & Shelton, R. C. (2007). Are
Selective serotonin reuptake inhibitors for stroke recovery — A systematic review antidepressant drugs that combine serotoninergic and noradrenergic mechanisms of
and meta-analysis. Stroke, 44, 844–850. action more effective than the selective serotonin reuptake inhibitors in treating
Meng, L., Chen, D., Yang, Y., Zheng, Y., & Hui, R. (2012). Depression increases the risk of major depressive disorders? A meta analysis of newer agents. Biological Psychiatry, 62,

12
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

1217–1227. inhibitors and risk of stroke: A systematic review and meta-analysis. Journal of
Pastoor, D., & Gobburu, J. (2014). Clinical pharmacology review of escitalopram for the Neurology, 261, 686–695.
treatment of depression. Expert Opinion on Drug Metabolism & Toxicology, 10, Sibolt, G., Curtze, S., Melkas, S., Pohjasvaara, T., Kaste, M., Karhunen, P. J., et al. (2013).
121–128. Post-stroke depression and depression-executive dysfunction syndrome are associated
Paulzen, M., Lammertz, S. E., Gründer, G., Veselinovic, T., Hiemke, C., & Tauber, S. C. with recurrence of ischaemic stroke. Cerebrovascular Diseases, 36, 336–343.
(2016). Measuring citalopram in blood and central nervous system: Revealing a Siepmann, T., Kepplinger, J., Zerna, C., Schatz, U., Penzlin, A. I., Pallesen, L. P., et al.
distribution pattern that differs from other antidepressants. International Clinical (2015). The effects of pretreatment versus de novo treatment with selective serotonin
Psychopharmacology, 31, 119–126. reuptake inhibitors on short-term outcome after acute ischemic stroke. Journal of
Penninx, B. W. H., Milaneschi, Y., Lamers, F., & Vogelzangs, N. (2013). Understanding the Stroke and Cerebrovascular Diseases, 24, 1886–1892.
somatic consequence of depression: Biological mechanisms and the role of depression Siepmann, T., Penzlin, A. I., Kepplinger, J., Illigens, B. M.-W., Weidner, K., Reichmann,
symptoms profile. BMC Medicine, 11, 129. H., et al. (2015). Selective serotonin reuptake inhibitors to improve outcome in acute
Pietra Pedroso, V. S., Rachid, M. A., & Teixeira, A. L. (2016). Biomarkers of post-stroke ischemic stroke: Possible mechanisms and clinical evidence. Brain and Behavior: A
depression. Current Neurovascular Research, 13, 163–173. Cognitive Neuroscience Perspective, 5, e00373.
Provinciali, L., & Coccia, M. (2002). Post-stroke and vascular depression: A critical re- Song, C., & Wang, H. (2011). Cytokines mediated inflammation and decreased neuro-
view. Neurological Sciences, 22, 417–428. genesis in animal models of depression. Progress in Neuro-Psychopharmacology and
Quaranta, D., Marra, C., & Gainotti, G. (2008). Mood disorders after stroke: Diagnostic Biological Psychiatry, 35, 760–768.
validation of the poststroke depression rating scale. Cerebrovascular Diseases, 26, Spalletta, G., Bossù, P., Ciaramella, A., Bria, P., Caltagirone, C., & Robinson, R. G. (2006).
237–243. The etiology of poststroke depression: A review of the literature and a new hypothesis
Queirazza, F., & Cavanagh, J. (2014). Poststroke depression and 5-HTTLPR. Journal of involving inflammatory cytokines. Molecular Psychiatry, 11, 984–991.
Neurology, Neurosurgery, and Psychiatry, 85, 241–243. Spalletta, G., Cravello, L., Imperiale, F., Salani, F., Bossù, P., Picchetto, L., et al. (2013).
Quinn, T. J., Paolucci, S., Sunnerhagen, K. S., Silvenius, J., Walker, M. F., Toni, D., et al. Neuropsychiatric symptoms and interleukin-6 levels in acute stroke. The Journal of
(2009). Evidence-based stroke rehabilitation: An expanded guidance document from Neuropsychiatry and Clinical Neurosciences, 25, 255–263.
the European Stroke Organisation (ESO) guidelines for the management of ischaemic Spalletta, G., Ripa, A., & Caltagirone, C. (2005). Symptom profile of DSM-IV major and
stroke and transient ischaemic attack 2008. Journal of Rehabilitation Medicine, 41, minor depressive disorders in first-ever stroke patients. The American Journal of
99–111. Geriatric Psychiatry, 13, 108–115.
Ramasubbu, R. (2011). Therapy for prevention of post-stroke depression. Expert Opinion Sugo, N., Hurn, P. D., Morahan, M. B., Hattori, K., Traystman, R. J., & DeVries, A. C.
on Pharmacotherapy, 12, 2177–2187. (2002). Social stress exacerbates focal cerebral ischemia in mice. Stroke, 33,
Razmara, A., Valle, N., Markovic, D., Sanossian, N., Ovbiagele, B., et al. (2017). 1660–1664.
Depression is associated with a high risk of death among stroke survivors. Journal of Sun, J., Ma, H., Yu, C., Lv, J., Guo, Y., Bian, Z., et al. (2016). Association of major de-
Stroke and Cerebrovascular Diseases. http://dx.doi.org/10.1016/jstrokecerebrovasdis. pressive episodes with stroke risk in a prospective study of 0.5 million Chinese adults.
2017.07.006. Stroke, 47, 2203–2208.
Renshaw, P. F., Parow, A. M., Hirashima, F., Ke, Y., Moore, C. M., Frederick, B. B., et al. Sun, X., Zhou, Z., Liu, T., Zhao, M., Zhao, S., Xiao, T., et al. (2016). Fluoxetine enhances
(2001). Multinuclear magnetic resonance spectroscopy studies of brain purines in neurogenesis in aged rats with cortical infarcts, but this is not reflected in a beha-
major depression. The American Journal of Psychiatry, 158, 2048–2055. vioral recovery. Journal of Molecular Neuroscience, 58, 233–242.
Rezin, G. T., Amboni, G., Zugno, A. L., Quevedo, J., & Streck, E. L. (2009). Mitochondrial Surtees, P. G., Wainwright, N. W., Luben, R. N., Wareham, N. J., Bingham, S. A., & Khaw,
dysfunction and psychiatric disorders. Neurochemical Research, 34, 1021–1029. K. T. (2008). Psychological distress, major depressive disorder, and risk of stroke.
Ried, L. D., Jia, H., Cameon, R., Feng, H., Wang, X., & Tueth, M. (2010). Does prestroke Neurology, 70, 788–794.
depression impact poststroke depression and treatment? The American Journal of Swenson, J. R., Doucette, S., & Fergusson, D. (2006). Adverse cardiovascular events in
Geriatric Psychiatry, 18, 624–633. antidepressant trials involving high-risk patients: A systematic review of randomized
Ried, L. D., Jia, H., Feng, H., Cameon, R., Wang, X., Tueth, M., et al. (2011). Selective trials. Canadian Journal of Psychiatry, 51, 923–929.
serotonin reuptake inhibitor treatment and depression are associated with poststroke Szczudlik, A., Dziedzic, T., Bartus, S., Slowik, A., & Kieltyka, A. (2004). Serum inter-
mortality. The Annals of Pharmacotherapy, 45, 888–897. leukin-6 predicts cortisol release in acute stroke patients. Journal of Endocrinological
Robinson, R. G. (2003). Poststroke depression: Prevalence, diagnosis, treatment, and Investigation, 27, 37–41.
disease progression. Biological Psychiatry, 54, 376–387. Taylor, W. D., Alzenstein, H. J., & Alexopoulos, G. S. (2013). The vascular depression
Robinson, R. G., & Jorge, R. E. (2016). Post-stroke depression: A review. The American hypothesis: Mechanisms linking vascular disease with depression. Molecular
Journal of Psychiatry, 173, 221–231. Psychiatry, 18, 963–974.
Robinson, R. G., Jorge, R. E., Moser, D. J., Acion, L., & Solodkin, A. (2008). Escitalopram Terroni, L., Amaro, E., Jr., Iosifescu, D., Tinone, G., Sato, J. R., Leite, C. C., et al. (2011).
and problem-solving therapy for prevention of poststroke depression. JAMA, 299, Stroke lesion in cortical neural circuits and post-stroke incidence of major depressive
2391–2400. episode: A 4-month prospective study. The World Journal of Biological Psychiatry, 12,
Robinson, R. G., Kubos, K. L., Starr, R. B., Rao, K., & Price, T. R. (1984). Mood disorders in 539–548.
stroke patients. Importance of location of lesions. Brain, 107, 81–93. Towfighi, A., Ovbiagele, B., El Husseini, N., Hackett, M., Jorge, R. E., et al. (2017).
Robinson, R. G., Spalletta, G., Jorge, R. E., Bassi, A., Colivicchi, F., Ripa, A., et al. (2008). Poststroke depression. A scientific statement for healthcare professionals from the
Decreased heart rate variability is associated with poststroke depression. The American Heart Association/American Stroke Association. Stroke, 48, e30–e43.
American Journal of Geriatric Psychiatry, 16, 867–873. Truelsen, T., Piechowski-Jóźwiak, B., Bonita, R., Mathers, C., Bogousslavsky, J., &
Salaycik, K. J., Kelly-Hayes, M., Beiser, A., Nguyen, A.-H., Brady, S. M., Kase, C. S., et al. Boysen, G. (2006). Stroke incidence and prevalence in Europe: A review of available
(2007). Depressive symptoms and risk of stroke. The Framingham study. Stroke, 38, data. European Journal of Neurology, 13, 581–598.
16–21. Tsai, C. S., Wu, C. L., Chou, S. Y., Tsang, H. Y., Hyng, T. H., & Su, J. A. (2011). Prevention
Salter, K. L., Foley, N. C., Zhu, L., Jutai, J. W., & Teasell, R. W. (2013). Prevention of post- of poststroke depression with milnacipran in patients with acute ischemic stroke: A
stroke depression: Does prophylactic pharmacotherapy work? Journal of Stroke and double-blind randomized placebo-controlled trial. International Clinical
Cerebrovascular Diseases, 22, 1243–1251. Psychopharmacology, 26, 263–267.
Sanacora, G., Treccani, G., & Popoli, M. (2012). Towards a glutamate hypothesis of de- Villa, R. F., Ferrari, F., & Gorini, A. (2012). Energy metabolism of rat cerebral cortex,
pression: An emerging frontier of neuropsychopharmacology of mood disorders. hypothalamus and hypophysis during ageing. Neuroscience, 227, 55–66.
Neuropharmacology, 62, 63–77. Villa, R. F., Ferrari, F., Gorini, A., Bagini, L., Brunello, N., & Tascedda, F. (2017).
Santarelli, L., Saxe, M., Gross, C., Surget, A., Battaglia, F., Dulawa, S., et al. (2003). Mitochondrial energy metabolism of rat hippocampus after desipramine and fluox-
Requirement of hippocampal neurogenesis for the behavioral effects of anti- etine treatments. Neuropharmacology, 121, 30–38.
depressants. Science, 301, 805–809. Villa, R. F., Ferrari, F., Gorini, A., Brunello, N., & Tascedda, F. (2016). Effect of desi-
Santos, M., Kövary, E., Gold, G., Bozikas, V. P., & Hof, P. R. (2009). The neuroanatomical pramine and fluoxetine on energy metabolism of cerebral mitochondria.
model of post-stroke depression: Towards a change of focus? Journal of the Neuroscience, 330, 326–334.
Neurological Sciences, 283, 158–162. Villa, R. F., Ferrari, F., & Moretti, A. (2017). Effects of neuroprotectants before and after
Sauer, W. H., Berlin, J. A., & Kimmel, S. E. (2003). Effect of antidepressants and their stroke: Statins and anti-hypertensives. In P. Lapchak, & J. Zhang (Eds.).
relative affinity for the serotonin transporter on the risk of myocardial infarction. Neuroprotective therapy for stroke and ischemic disease (pp. 349–399). New York:
Circulation, 108, 32–36. Springer.
Schmidt, H. D., & Duman, R. S. (2007). The role of neurotrophic factors in adult hippo- Villa, R. F., Gorini, A., Ferrari, F., & Hoyer, S. (2013). Energy metabolism of cerebral
campal neurogenesis, antidepressant treatments and animal models of depressive-like mitochondria during aging, ischemia and post-ischemic recovery assessed by func-
behavior. Behavioural Pharmacology, 18, 391–418. tional proteomics of enzymes. Neurochemistry International, 63, 765–781.
Seifert, C. L., Poppert, H., Sander, D., Feurer, R., Etgen, T., Ander, K.-H., et al. (2012). Vogelzangs, N., Dulvis, H. E., Beekman, A. T. F., Kluft, C., Neuteboom, J., Hoogendijk, W.,
Depressive symptoms and the risk of ischemic stroke in the elderly—Influence of age et al. (2012). Association of depressive disorders, depression characteristics and an-
and sex. PLoS One, 7, e50803. tidepressant medication with inflammation. Translational Psychiatry, 2, e79.
Selim, M. H., & Molina, C. A. (2012). Poststroke treatment with selective serotonin re- Wang, S., Sun, H., Liu, S., Wang, T., Guan, J., & Jia, J. (2016). Role of hypothalamic
uptake inhibitors. A journey from sadness to motor recovery. Stroke, 43, 3154–4155. cannabinoid receptors in post-stroke depression in rats. Brain Research Bulletin, 121,
Shi, Y., Yang, D., Zeng, Y., & Wu, W. (2017). Risk factors for post-stroke depression: A 91–97.
meta-analysis. Frontiers in Aging Neuroscience, 9, 18. Wang, S.-H., Zhang, Z.-J., Guo, Y.-J., Sui, Y.-X., & Sun, Y. (2010). Involvement of ser-
Shin, T. K., Kang, M. S., Lee, H. Y., Seo, M. S., Kim, S. G., Kim, C. D., et al. (2009). otonin neurotransmission in hippocampal neurogenesis and behavioral responses in a
Fluoxetine and sertraline attenuate postischemic brain injury in mice. The Korean rat model of post-stroke depression. Pharmacology, Biochemistry, and Behavior, 95,
Journal of Physiology and Pharmacology, 13, 257–263. 129–137.
Shin, D., Oh, Y. H., Eom, C.-S., & Park, S. M. (2014). Use of selective serotonin reuptake Wang, S.-H., Zhang, Z.-J., Guo, Y.-J., Teng, G.-J., & Chen, B.-A. (2008). Hippocampal

13
R.F. Villa et al. Pharmacology and Therapeutics xxx (xxxx) xxx–xxx

neurogenesis and behavioural studies on adult ischemic rat response to chronic mild may indicate the development of PSD in patients with acute ischemic stroke.
stress. Behavioural Brain Research, 189, 9–16. International Journal of Geriatric Psychiatry, 26, 495–502.
Wei, N., Yong, W., Li, X., Zhou, Y., Deng, M., Zhu, H., et al. (2015). Post-stroke depression Yang, L., Zhang, Z., Sun, D., Xu, Z., Zhang, X., & Li, L. (2010). The serum interleukin-18 is
and lesion location: A systematic review. Journal of Neurology, 262, 81–90. a potential marker for development of post-stroke depression. Neurological Research,
Werheid, K. (2016). A two-phase pathogenetic model of depression after stroke. 32, 340–346.
Gerontology, 62, 33–39. Yi, Z. M., Liu, F., & Zhai, S. D. (2010). Fluoxetine for the prophylaxis of post-stroke de-
Williams, L. S. (2005). Depression and stroke: Cause or consequence? Seminars in pression in patients with stroke: A meta-analysis. International Journal of Clinical
Neurology, 25, 396–409. Practice, 64, 1310–1317.
Windle, Y., & Corbett, D. (2005). Fluoxetine and recovery of motor function after focal Yu, J., Novgorodov, S. A., Chudakova, D., Zhu, H., & Bielawska, A. (2007). JNK3 signaling
ischemia in rats. Brain Research, 1044, 25–32. pathway activates ceramide synthase leading to mitochondrial dysfunction. The
Winstein, C. J., Stein, J., Arena, R., Bates, B., Cherney, L. R., Cramer, S. C., et al. (2016). Journal of Biological Chemistry, 282, 25940–25949.
Guidelines for adult stroke rehabilitation and recovery: A guideline for healthcare Zhang, L. S., Hu, X. Y., Yao, L. Y., Geng, Y., Wei, L. L., Zhang, J. H., et al. (2013).
professionals from the American Heart Association/American Stroke Association. Prophylactic effects of duloxetine on post-stroke depression symptoms: An open
Stroke, 47, e98–e169. single-blind trial. European Neurology, 69, 336–343.
Xu, H., Steven Richardson, J., & Li, X. M. (2003). Dose-related effects of chronic anti- Zhang, Z. H., Wu, L. N., Song, J. G., & Li, W. Q. (2012). Correlation between cognitive
depressants on neuroprotective proteins BDNF, Bcl-2 and Cu/Zn-SOD in rat hippo- impairment and brain-derived neurotrophic factor expression in the hippocampus of
campus. Neuropsychopharmacology, 28, 53–62. post-stroke depression rats. Molecular Medicine Reports, 6, 889–893.
Xu, X.-M., Zou, D.-Z., Shen, L.-Y., Liu, Y., Zhou, X.-Y., Pu, J. C., et al. (2016). Efficacy and Zhao, Q., Guo, Y., Yang, D., Yang, T., & Meng, X. (2016). Serotonin transporter gene 5-
feasibility of antidepressant treatment in patients with post-stroke depression. HTTLPR polymorphism as a protective factor against the progression of post-stroke
Medicine, 95(45), e5349. depression. Molecular Neurobiology, 53, 1699–1705.
Yang, L., Zhang, Z., Sun, D., Xu, Z., Yuan, Y., Zhang, X., et al. (2011). Low serum BDNF

14