Doi:10.1016/s1474-4422(10)70314-8

Fluoxetine for motor recovery after acute ischaemic stroke
(FLAME): a randomised placebo-controlled trial
François Chollet, Jean Tardy, Jean-François Albucher, Claire Thalamas, Emilie Berard, Catherine Lamy, Yannick Bejot, Sandrine Deltour, Assia Jaillard, Philippe Niclot, Benoit Guillon, Thierry Moulin, Philippe Marque, Jérémie Pariente, Catherine Arnaud, Isabelle Loubinoux Summary
Background Hemiplegia and hemiparesis are the most common defi cits caused by stroke. A few small clinical trials Published Online
suggest that fl uoxetine enhances motor recovery but its clinical effi
cacy is unknown. We therefore aimed to investigate DOI:10.1016/S1474-
whether fl uoxetine would enhance motor recovery if given soon after an ischaemic stroke to patients who have 4422(10)70314-8
motor defi cits.
Methods In this double-blind, placebo-controlled trial, patients from nine stroke centres in France who had ischaemic 4422(10)70326-4
stroke and hemiplegia or hemiparesis, had Fugl-Meyer motor scale (FMMS) scores of 55 or less, and were aged Neurology Department
(F Chollet MD, J Tardy MD, between 18 years and 85 years were eligible for inclusion. Patients were randomly assigned, using a computer
J-F Albucher MD, J Pariente MD) random-number generator, in a 1:1 ratio to fl uoxetine (20 mg once per day, orally) or placebo for 3 months starting and Clinical Epidemiology
5–10 days after the onset of stroke. All patients had physiotherapy. The primary outcome measure was the change on Unit (E Berard MD,
the FMMS between day 0 and day 90 after the start of the study drug. Participants, carers, and physicians assessing C Arnaud MD), Centre
Hospitalier Universitaire de
the outcome were masked to group assignment. Analysis was of all patients for whom data were available (full
Toulouse, Hôpital Purpan,
analysis set). This trial is registered with ClinicalTrials.gov, number NCT00657163.
Toulouse, France; Institut des
Sciences du Cerveau de
Findings 118 patients were randomly assigned to fl uoxetine (n=59) or placebo (n=59), and 113 were included in the Toulouse (INSERM, CNRS,
Université de Toulouse),
analysis (57 in the fl uoxetine group and 56 in the placebo group). Two patients died before day 90 and three withdrew Toulouse, France (F Chollet,
from the study. FMMS improvement at day 90 was signifi cantly greater in the fl uoxetine group (adjusted mean J Tardy, J-F Albucher, J Pariente,
34·0 points [95% CI 29·7–38·4]) than in the placebo group (24·3 points [19·9–28·7]; p=0·003). The main adverse P Marque MD,
events in the fl uoxetine and placebo groups were hyponatraemia (two [4%] vs two [4%]), transient digestive disorders I Loubinoux PhD); Université
de Toulouse, Université Paul
including nausea, diarrhoea, and abdominal pain (14 [25%] vs six [11%]), hepatic enzyme disorders (fi ve [9%] vs ten Sabatier INSERM, Imagerie
[18%]), psychiatric disorders (three [5%] vs four [7%]), insomnia (19 [33%] vs 20 [36%]), and partial seizure (one Cérébrale et Handicaps
[<1%] vs 0).
Neurologiques UMR 825,
Centre Hospitalier
Universitaire de Toulouse,
Interpretation In patients with ischaemic stroke and moderate to severe motor defi cit, the early prescription of Hôpital Purpan, Purpan,
fl uoxetine with physiotherapy enhanced motor recovery after 3 months. Modulation of spontaneous brain plasticity Toulouse, France
by drugs is a promising pathway for treatment of patients with ischaemic stroke and moderate to (F Chollet, J Tardy, J-F Albucher,
severe motor defi cit.
J Pariente, P Marque); Unité
Mixte de Recherche
INSERM U558/Université
Funding Public French National Programme for Clinical Research.
Toulouse III, Toulouse, France
(E Berard, C Arnaud); Clinical
and promote hippocampal neurogenesis.4–6 In clinical Investigation Centre INSERM
CIC-9302 and Department of
Thrombolysis with alteplase given within the fi rst few trials of amphetamine in patients with stroke, either no Clinical Pharmacology,
hours of an ischaemic stroke has long been the only positive eff ect was noted on the recovery of motor function Université de Toulouse,
treatment recognised to improve the spontaneous or the results were contradictory.7–11 The few small clinical Toulouse, France
recovery of neurological functions. However, we have trials of serotonin-reuptake inhibitors that have been (C Thalamas MD); Service de
Neurologie, Hôpital
learnt over the past decade, by use of neuroimaging and reported (table 1) all suggest that drugs of this type might Sainte-Anne, Paris, France
electrophysiological techniques, that spontaneous have a positive eff ect.13–16 Use of functional MRI in other (C Lamy MD); Service de
recovery of neurological functions is associated with a studies showed that single doses of fl uoxetine and Neurologie, Hôpital Général,
Dijon, France (Y Bejot MD);
large intracerebral reorganisation of the damaged paroxetine overactivated motor cortices compared with Service des Urgences
human brain.
placebo in both healthy individuals and patients with Cérébro-Vasculaires, Hôpital
Various interventions, such as monoaminergic drugs, stroke, and use of transcranial magnetic stimulation Pitié Salpêtrière, Paris, France
have been shown to modulate brain plasticity after a showed that cortex overactivation was associated with (S Deltour MD); Centre
d'Investigation Clinique,
stroke and to reduce the residual neurological defi cit and drug-induced cortex hyperexcitability.12 Hôpital Michalon, Grenoble,
subsequent disability.1–3 Amphetamines have enhanced In the fl uoxetine in motor recovery of patients with France (A Jaillard MD); Centre
recovery in animal models of acute brain lesions, whereas acute ischaemic stroke (FLAME) trial, we aimed to test Hospitalier René Dubos,
neuroleptic drugs or benzodiazepines have reduced it.1–3 whether a 3-month treatment with fl uoxetine would Pontoise, France (P Niclot MD);
Service de Neurologie, Hôpital
Little evidence exists for the effi cacy of serotonin-reuptake enhance motor recovery when given early after an Nord Laennec, Nantes, France
inhibitors in studies of animals, but these inhibitors have ischaemic stroke to patients with moderate to severe (B Guillon MD); and Centre
an acute neuroprotective action on the ischaemic brain motor defi cits.
www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8
Dose, regimen, and
Number of
Trial design
Time of inclusion
Clinical outcome
Other outcome
Patients in
Main results
after stroke
Fluoxetine and Fluoxetine 20 mg once Parallel groups 1–6 months Graded neurological None 10·7% improvement per day for 90 days 20 mg (single dose) Finger tapping and 20–30% fi nger hyperactivation of dynamometer improvement 40 mg (single dose) 11·4% improvement with nine-hole-peg in nine-hole -peg test 10 mg once per day for Parallel groups Not reported 38·8% improvement Results of all trials showed positive eff ects on motor performance. Results of a randomised placebo-controlled trial by Gerdelat-Mas and colleagues16 in healthy individuals also confi rmed the modulation of cortical excitability induced by transcranial magnetic stimulation (TMS) with a single and chronic doses of paroxetine. HSS=hemispheric stroke scale. NIHSS=National Institutes of Health stroke scale.
Table 1: Reported prospective randomised placebo-controlled clinical trials of selective serotonin-reuptake inhibitors in motor recovery after ischaemic stroke
Hospitalier Universitaire de
Fluoxetine
Besançon, Besançon, France
Patients who had an acute ischaemic stroke within the Correspondence to: past 5–10 days that caused hemiparesis or hemiplegia Prof François Chollet, Neurology Department, HÔpital Purpan,
were prospectively enrolled from nine stroke units in Body-mass index (kg/m²) Place Baylac, 31059, France. Those who were aged between 18 years and Vascular risk factors 85 years and who had Fugl-Meyer motor scale (FMMS) scores of 55 or less at baseline were eligible for inclusion.
Patients were excluded if they had severe post-stroke disability (National Institutes of Health stroke scale [NIHSS] score >20), substantial premorbid disability, or a Previous cardiac disease pre-existing defi cit that could interfere with assessments— Atrial fi brillation ie, residual motor defi cit from a previous stroke, comprehension defi cits severe enough to prevent Stroke characteristics understanding of motor testing, or severe aphasia masking detection and assessment of depression. Carotid territory Patients were excluded if they were clinically diagnosed with depression or Montgomery Åsberg depression rating scale (MADRS) score of more than 19;17 taking Baseline stroke severity antidepressant drugs, monoamine oxidase inhibitors, neuroleptic drugs, or benzodiazepines during the month Upper limb FMMS score before inclusion; or due to underg carotid endarterectomy. Lower limb FMMS score Other exclusion criteria included pregnancy and other major diseases that would prevent follow-up. Enrolment, NIHSS motor component score follow-up, and clinical assessments were done in each Modifi ed Rankin scale score centre by the investigators.
The study was approved for all centres, according to the 3: moderate disability French law, by the Toulouse Ethics Committee. All 4: moderately severe disability patients provided written informed consent.
5: severe disability Intravenous thrombolysis Randomisation and masking
Randomisation was balanced by centre, with an allocation Time from stroke to treatment (days) sequence based on a block size of four, generated with a computer random-number generator by the pharmacist Data are number (%) or mean (SD). FMMS=Fugl-Meyer motor scale. at the coordinating centre (Hôpital Purpan, Toulouse). NIHSS=National Institutes of Health stroke scale. MADRS=Montgomery Åsberg depression rating scale.
The pharmacist assigned the participants to the trial groups in accordance with the randomisation list. Table 2: Demographic and baseline characteristics
Allocation was concealed by use of sequentially numbered www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8
opaque envelopes. Fluoxetine and placebo capsules were both prepared by the pharmacist for the study. The 118 patients randomly assigned capsules were distributed to each centre by the Toulouse Hospital Pharmacy. The placebo was identical to the active drug in appearance and packaging.
59 allocated to fluoxetine 59 allocated to placebo An independent organisation (Centre d'Investigation Clinique, Toulouse, France) was in charge of the logistic 1 died from respiratory distress 1 died from septic shock coordination of the study. Patients, carers, all site study after inhalation of food 2 withdrew (1 kidney tumour; staff , and investigators were masked to treat 1 withdrew (severe hypoxia) 1 pulmonary embolism) 57 analysed for primary endpoint in 56 analysed for primary endpoint in full-analysis set at day 90 full-analysis set at day 90 Patients were randomly allocated to fl uoxetine (20 mg once per day, orally) or placebo for 90 days. All patients, Figure 1: Trial profi le
irrespective of the treatment group, also received physiotherapy during the treatment period. Diff erence between p value
Physiotherapists were instructed not to design a specifi c groups (95% CI)
rehabilitation programme but to use the normal protocol for their centre. All participants received standard care, delivered by an organised inpatient 18·6 (9·2 to 27·9) stroke care team.
The primary outcome was the mean change in FMMS score18 between inclusion (day 0) and day 90. FMMS is 13·5 (6·2 to 20·8) an index that is widely used for assessment of motor recovery after stroke and has excellent intra-rater and inter-rater reliability and validity.19 The motor domain 5·1 (2·1 to 8·1) ranges from a score of 0 (fl accid hemiplegia) to 100 (normal movement), with 66 points for the upper limb Change from day 0 to day 90
and 34 points for the lower limb; each item is rated as not, partly, or fully performed. All motor assessments 14·5 (7·3 to 21·6) were made by a physiotherapist at day 0 (baseline) and then 30 days and 90 days after enrolment. Secondary Adjusted mean (95% CI) 34·0 (29·7 to 38·4) 24·3 (19·9 to 28·7) 9·8 (3·4 to 16·1) endpoints were NIHSS,20,21 modifi ed Rankin scale (mRS),22 and MADRS,23 with all scores measured at 12·4 (5·9 to 18·9) baseline, day 30, and day 90.
Adjusted mean (95% CI) 22·9 (18·6 to 27·1) 13·1 (8·9 to 17·4) 9·7 (3·6 to 15·9) Data were gathered for adverse events and deaths within 3 months after randomisation. Reports were 2·1 (–0·4 to 4·6) periodically examined by the FLAME executive Adjusted mean (95% CI) 12·8 (11·1 to 14·5) 9·5 (7·8 to 11·2) 3·3 (0·8 to 5·7) committee, which tried to assess whether the side- Mean was adjusted for age, history of stroke, and FMMS score at inclusion. *Mann-Whitney U test. †Linear eff ects could be related or not to the drug intake, but regression including treatment and centre as fi xed eff ects, and confounding factors (age, history of stroke, and the committee remained masked during the study. FMMS score at inclusion).
When depression occurred (according to clinical Table 3: Fugl-Meyer motor scale (FMMS) scores
assessment) during the 3 months of treatment, clinicians were instructed to continue the study treatment (fl uoxetine or placebo), to avoid use of any recovery of six functions or incomplete recovery of other antidepressant drugs and, if necessary, to give 12 functions, and is judged to be clinically relevant) open-label fl uoxetine (20 mg once a day) so that the 3 months after the stroke if 100 patients were enrolled, patient received either 20 mg (placebo group) or 40 mg assuming a spontaneous 30-point (SD 18) mean per day (fl uoxetine group). If a patient was given another increase at day 90.24 Because we expected frequency of antidepressant drug, the study treatment was stopped. patients lost to follow-up to be high, related to The blinding code was not broken. All patients were depression and complications after stroke, we planned followed up until day 90.
to enrol up to 168 patients. However, the loss to follow-up was lower than expected, leading us to stop the enrolment after 118 patients.
The trial was designed to have greater than 90% power Baseline characteristics, including disease description, to detect a 40% difference in the change in the FMMS were reported by use of descriptive statistics (eg, mean score (ie, 12 points, which corresponds to a full and SD for continuous variables, or frequency www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8
limb score, and lower limb score) over 90 days. We did a multiple linear regression analysis to control for centre and baseline factors that showed treatment group imbalance. The same approach was used to assess changes from day 0 to day 90 in MADRS scores (secondary endpoint). We also undertook post-hoc binomial regression analyses to test whether the treatment had an eff ect at day 90 on the mRS scores (probability of being an independent patient—ie, with a score of 0–2) and on NIHSS scores (probability of having a less severe outcome—ie, with scores of 0–5) compared with placebo. The analyses were adjusted for centre, age of patient, history of stroke, and mRS or NIHSS scores, accordingly, at baseline. All reported p values are two- sided. The statistical analyses were done by use of Time from stroke (days) STATA (version 11.0).
Figure 2: Adjusted mean Fugl-Meyer motor scale (FMMS) total scores at
This study is registered with ClinicalTrials.gov, days 0, 30, and 90
number NCT00657163.
Mean was adjusted for centre, age, history of stroke (at days 0, 30, and 90), and FMMS score at inclusion (at days 30 and 90). Error bars represent 95% CI.
Role of the funding source
The sponsor had no involvement in study design, data
collection, data analysis, data interpretation, or writing of NIHSS scores on day 90 the report. The principal investigator (FC) had full access Total score, mean (SD) to all the data in the study and had fi nal responsibility for Patients with score 0–5, adjusted mean (95% CI) the decision to submit the report for publication.
Motor scores, mean (SD) mRS scores on day 90 Patients with mRS score 0–2§ Between March 14, 2005, and June 9, 2009, 118 patients Patients with mRS score 0–2§, adjusted mean were prospectively enrolled. The two groups were well balanced in terms of baseline and demographic characteristics and stroke severity (table 2). However, Day 90, mean (SD) mean age was slightly higher and previous history of Day 90, median (IQR) stroke was more frequent in the fl uoxetine group than in Change from day 0 to day 90, mean (SD) the control group. FMMS score at inclusion was higher Change from day 0 to day 90, adjusted mean –0·1 (–2·1 to 1·9) 3·2 (1·1 to 5·3) in the fl uoxetine group than in the placebo group (table 2). NIHSS, mRS, and MADRS mean scores did not diff er in the two groups (table 2).
Data are number (%), unless otherwise indicated. Mean was adjusted for age, history of stroke, and Fugl-Meyer motor scale score at inclusion. NIHSS=National Institutes of Health stroke scale. mRS=modifi ed Rankin scale. Treatment compliance was similar in the two groups. MADRS=Montgomery Åsberg depression rating scale. *Student's t test. †Binomial regression including treatment and The mean cumulative dose at day 90 (defi ned as the sum centre as fi xed eff ects, and confounding factors (age, history of stroke, and NIHSS score at inclusion). ‡χ2 test. §None of of all doses taken during the treatment period) was the patients had an mRS score of 0. ¶Binomial regression including treatment and centre as fi xed eff ects, and confounding factors (age, history of stroke, and mRS score at inclusion). Mann-Whitney U test. **Linear regression 88·3 tablets (SD 11·6) in the fl uoxetine group and including treatment and centre as fi xed eff ects, and confounding factors (age, history of stroke, and Fugl-Meyer motor 87·4 tablets (12·3) in the placebo group (p=0·722). The scale score at inclusion).
mean dose intensity at day 90 (calculated as the overall Table 4: Secondary endpoints
cumulative dose divided by the number of days that the patient was receiving a dose) was 0·9 tablets (0·1) per day in both groups (p=0·854). The mean relative dose tabulations for categorical variables) for each group. intensity (dose intensity divided by the planned dose Homogeneity of these variables was assessed intensity) was 94·8% (12·6) in the fl uoxetine group descriptively.25 At day 90, FMMS, NIHSS, and MADRS and 94·0% (13·0) in the placebo group (p=0·854).
scores were compared in the treatment groups by use of Study treatment began a similar number of days after a Student's t test or Mann-Whitney U test when the stroke in the fl uoxetine and placebo groups. appropriate. Patients were classifi ed on the basis of 113 patients were included in the full-set analysis (fi gure 1). mRS scores as independent (scores 0–2) or non- Two patients died before day 90, and three withdrew from independent (scores 3–5), and the groups were compared follow-up (fi gure 1). Mean progression in FMMS total at day 90 by use of the χ² test. The primary analysis for score from baseline to day 90 was signifi cantly higher in effi cacy (full-analysis set) consisted of a comparison of the fl uoxetine group than in the placebo group after the change in FMMS distribution (total score, upper controlling for centre, age, history of stroke, and FMMS www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8
score at inclusion (table 3). The gain was signifi cant for both the upper and the lower limb scores. The adjusted mean FMMS total score was signifi cantly higher at day 90 in the fl uoxetine group than in the placebo group (fi gure 2).
3 months after the stroke, NIHSS total score did not signifi cantly diff er in the fl uoxetine and control groups, whereas the motor component score was signifi cantly higher in the fl uoxetine group (table 4). However, after controlling for age, history of stroke, and FMMS score at baseline, the probability of having a NIHSS score of 0–5 did not signifi cantly diff er between groups. Independence in activities of daily life, measured by use of mRS, improved during treatment in both groups, but at day 90 the proportion of independent patients (mRS scores 0, 1, or 2) adjusted for centre, age, history of stroke, and mRS score at baseline was signifi cantly higher in the fl uoxetine group than in the control group (table 4; fi gure 3).
The distribution of the MADRS scores did not diff er signifi cantly between the fl uoxetine and control groups Percentage of patients at inclusion or at day 90, whereas the adjusted mean
change in MADRS scores between day 0 and day 90 was Figure 3: Distribution of modifi ed Rankin scale scores at day 90
signifi cantly lower in the fl uoxetine group than in the Data are number (%).
placebo group (table 4). Moreover, the frequency of
depression was signifi cantly higher in the placebo group
(17 [29%] patients) than in the fl uoxetine group (four [7%] After adjustment of our analysis for clinical depression diagnosed before day 90, we noted that FMMS change between day 0 and day 90 was still signifi cantly greater in Hepatic enzyme disorders the fl uoxetine group (adjusted mean 34·2 points [95% CI Psychiatric disorders 29·7–38·6]) than in the placebo group (24·2 [19·6–28·7];
p=0·004). In a sensitivity analysis of the subgroup of patients who were not given thrombolysis (n=36 in fl uoxetine group, n=40 in placebo), improvement in Data are number (%). *Five adverse events in four patients. †Six adverse events in FMMS was still signifi cantly higher in the fl uoxetine fi ve patients. ‡Four adverse events in three patients.
group (37·7 [32·0–43·3]) than in the placebo group
Table 5: Adverse events
(24·4 [19·1–29·7]; p=0·002).
Two patients died (one in each group; fi gure 1). The cause of death was related to their neurological disorder in the FMMS subscores for both the upper and the lower (septic shock, respiratory distress; fi gure 1). The main limb at day 90. By contrast, no eff ect was noted with adverse events were hyponatraemia, transient digestive NIHSS at day 90. However, NIHSS motor component disorders including nausea, diarrhoea, and abdominal score at day 90 was lower in the fl uoxetine group than in pain, hepatic enzyme disorders, psychiatric disorders, the placebo group, in agreement with the data for FMMS insomnia, and partial seizure (table 5). Two of the adverse scores. The mRS scores showed more independent events in the fl uoxetine group were serious patients (scores 0–2) in the fl uoxetine group than in the (one hyponatraemia and one partial seizure). Transient placebo group at day 90, which, when combined with the digestive disorders were more frequent in the fl uoxetine effi cacy of fl uoxetine, confi rms the major role of motor group (p=0·19). Treatment was not interrupted in patients function recovery in global recovery and return to with adverse events.
independent activities of everyday life.
Results from some early and more recent studies suggested that a tight coupling between physiotherapy We noted a positive eff ect on motor recovery in patients and drug therapy was necessary for benefi cial motor with acute ischaemic stroke who were treated with changes.7–11 In our study, all patients were admitted to a fl uoxetine for 3 months. This eff ect, assessed as a change dedicated stroke unit and were all included in the local in FMMS score between day 0 and day 90, was noticeable daily inpatient management. Some were given acute www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8
thrombolysis with intravenous alteplase within the fi rst neuroprotective eff ect in the post-ischaemic brain 4·5 h after the stroke, in accordance with the through its anti-infl ammatory eff ects,5 and it has For more on Recommandations
recommendations of the French health authorities. The improved ischaemia-induced spatial cognitive defi cits by de la Haute Autorité de Santé
number of patients given thrombolysis was not increasing hippocampal neurogenesis after stroke in 2009: Accident Vasculaire
signifi cantly diff erent in the fl uoxetine and the placebo rats.6 One hypothesis is that a primary function of the Cérébral: prise en charge
groups, and we did not note any confounding eff ect of brain serotoninergic system is to facilitate motor output,29 précoce 2009 see
thrombolysis in this study. Physiotherapy and which emphasises that the drug intake would be more rehabilitation was delivered to each patient during the effi cient when paired with training. In Aplysia, serotonin 3-month treatment. Because intercentre variability in the enhances short-term facilitation, storage of long-term rehabilitation interventions was high, we decided that memory in sensorimotor synapses, long-term facilitation, each centre would propose rehabilitation interventions and growth factor gene expression.30 As a monoamine, according to the current onsite procedures rather than serotonin might promote long-term potentiation, according to a unifi ed procedure. A possible centre eff ect optimise activity-dependent learning, and possibly was taken into account in the linear regression analysis.
facilitate relearning after stroke in human beings.3 Patients were enrolled, allocated to treatment, and fi rst Few clinical trials with serotonin-reuptake inhibitors given treatment between days 5 and 10 after the stroke to have been reported (table 1; panel). They have all exclude those who had early complications that would included small numbers of patients; all have results have compromised follow-up. Depressed patients and that suggest a positive eff ect on recovery after stroke. In those with comprehension disturbances related to an early trial, fl uoxetine and maprotiline were tested profound aphasia or cognitive impairment were also against placebo for 3 months in patients with hemiplegic excluded. Patients had an initial severe motor defi cit as stroke enrolled 1–6 months after the stroke.12 The assessed with the FMMS score (table 1), probably more patients in the fl uoxetine group (n=16) had a better severe than in patients in other trials: for example, in a outcome than did those in the maprotiline or placebo trial of dexamfetamine that included 71 patients, mean groups.12 Acler and colleagues15 confi rmed this fi nding FMMS score at inclusion was 26·9 in the dexamfetamine in ten patients in the active-treatment group versus ten group and 30·0 in the placebo group.10 in the placebo group. In a double-blind placebo- The treatment was well tolerated. Occurrence of controlled study, Pariente and colleagues,13 by combining depression during the 3 months was signifi cantly lower clinical motor testing and functional MRI motor in the fl uoxetine group than in the placebo group, suggesting that fl uoxetine when given early after the Panel: Research in context
stroke can prevent depression. These results are in accord with those reported in other studies, showing the benefi cial eff ect of serotonin-reuptake inhibitors on the We searched the Cochrane Central Register of Controlled occurrence of depression after stroke.26–28 An eff ect of Trials (CENTRAL) on the Cochrane Library (issue 1, 2010), fl uoxetine on mood is likely, as shown by the signifi cant Medline (1985–2010), and Embase (1985–2010). We diff erence in the change in MADRS score between the included placebo-controlled, randomised clinical trials in two groups. However, we do not think that fl uoxetine which the eff ects of selective serotonin-reuptake inhibitors acted only through antidepressant mechanisms in this on motor recovery after stroke were assessed. We also study. In a previous study,13 a single dose of fl uoxetine selected key articles and illustrative reviews of studies in improved hand motor function and increased activity in animals, selective serotonin-reuptake inhibitors, and the motor cortex compared with placebo in patients depression after stroke, and found a large number of recovering from stroke, showing a specifi c motor eff ect, references for clinical trials of amphetamines and dopamine whereas a mood eff ect is unlikely after a single dose. in motor recovery after stroke.
However, a fl uoxetine-mediated attention eff ect cannot be excluded in our patients.
The few previous clinical trials of selective serotonin-reuptake Studies in animals show that the rate and extent of inhibitors after stroke12–15 were small and their results have functional recovery after brain injury can be modulated suggested a positive eff ect on motor recovery. The results of by the eff ects of drugs on neurotransmitters in the the FLAME trial extend these fi ndings by showing in a larger CNS.1–3 For example, infusion of norepinephrine hastens group of patients with moderate to severe hemiplegia after recovery in rats with brain lesions, whereas antagonists ischaemic stroke that early treatment with fl uoxetine delay the recovery process.1 Further evidence for enhances motor recovery and reduces the number of modulation of recovery has come from studies showing dependent patients. The positive eff ect of the drug on motor drug-induced physiological or structural changes in the function of recovering patients suggests that the neuronal, brain that might be relevant to recovery. By contrast, non-vascular-targeted action of selective serotonin-reuptake little evidence exists to suggest that serotonin-reuptake inhibitors provides a new pathway that should be explored inhibitors induce motor recovery after focal ischaemia in further in the treatment of acute ischaemic stroke.
rats.4 However, fl uoxetine was recognised to have a www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8
assessment in patients recovering from post-stroke (mRS, which scores the capacity of patients to return to hemiplegia (n=8), showed that a single dose (20 mg) of home daily life, rather than FMMS, which is a pure fl uoxetine improved hand motor function and was neurological analytical scale of motor function), and the correlated with an overactivation of motor cortices on duration of treatment and permanence of the eff ects functional MRI.13 In a subsequent double-blind, will have to be addressed. Selective serotonin-reuptake placebo-controlled trial in healthy individuals, inhibitors are not a uniform category of drugs, and transcranial magnetic stimulation has shown that the further basic science and pharmacology studies will also intake of a single dose of the serotonin-reuptake be needed to increase understanding of their inhibitor paroxetine was associated with a mechanisms of action.
hyperexcitability of the primary motor cortex, whereas Fluoxetine is a well tolerated drug that no longer has a chronic intake was associated with hypoexcitability of patent, and therefore its cost is reasonable. Acute brain motor cortices.15 Serotonin-reuptake inhibitors deocclusion of brain arteries is already a successfully increase interneuron-facilitating activity in the primary validated approach to treatment, and modulation of motor cortex. In a double-blind, placebo-controlled spontaneous brain plasticity by external agents is
crossover trial, Zittel and colleagues14 investigated the undoubtedly another promising pathway for patients
eff ects of a single dose (40 mg) of citalopram in eight with stroke.
patients with chronic stroke; dexterity was signifi cantly Contributors
improved.14 Our trial is the largest in which the eff ects FC was the principal investigator, participated in the study design and
of serotonin-reuptake inhibitors and stroke recovery data analysis, and wrote the report. JFA and JT participated in data
were investigated, and the largest in the specialty of collection, data analysis, and correction of the report. CA and EB
participated in the statistical analysis and data analysis. CT contributed monoamines and stroke recovery. Several trials with to the study design and data analysis, and liaised with the funding amphetamine or amphetamine-like drugs have been source. CL, YB, SD, AJ, BG, TM, and PN participated in data reported.3,7–11,31 The results of a few studies suggested collection, reading the report, and giving scientifi c advice, and contributed to discussions. PM and JP contributed to the design of the amphetamine and norepinephrine were effi cacious, but study and correction of the report. IL gave scientifi c advice, and this positive eff ect was not confi rmed in further larger participated in reading and correction of the report.
trials.3,11,31 Notably, most of the trials included only a few Participating centres (number of patients)
patients (from eight to 71) and the dose (amphetamine Toulouse Purpan (57) F Chollet, J Tardy, J F Albucher, I Loubinoux, 10 mg once per day) and regimen (1–17 days) remain J Pariente, P Marque, A Damène, M Durrieux; Toulouse Rangueil (16) questionable because no rationale was proposed for V Larrue; Dijon (15) M Giroud, Y Bejot; Paris Ste Anne (9) J L Mas, C Lamy, D Calvet, V Domigo; Paris Pitié Salpêtrière (8) Y Samson, dose and duration of treatment. Results of other S Deltour, A Leger; Grenoble (7) M Hommel, A Jaillard, O Detante; small trials of drugs including levodopa were Pontoise (3) P Niclot, S Descombes, J Servant; Nantes (2) B Guillon, contradictory and no defi nite conclusions could be J Rome; Besançon (1) T Moulin, P Decavel.
drawn from them.7,32 Confl icts of interest
Although the results of the FLAME trial show the We declare that we have no confl icts of interest.
cacy of fl uoxetine in motor recovery of patients with Acknowledgments
ischaemic stroke, we must draw attention to some This work was supported by a grant from the Clinical Research Hospital limitations of the study. First, the number of patients Program from the French Ministry of Health (PHRC 2003, to FC). We thank the Centre d'Investigation Clinique de Toulouse (O Rascol, F Puisset included was small. Those who were included were [pharmacist], S Bernard) for the global coordination of the trial and
selected for motor defi cit and did not represent the C Cristini for valuable expertise in statistics. This work was sponsored by
general population of stroke patients, as shown by the the University Hospital of Toulouse for regulatory and ethics submission.
inclusion and exclusion criteria and the clinical References
characteristics at inclusion (table 2). Second, treatment 1
Schallert T, Jones TA, Weaver MS, Shapiro LE, Crippens D, Fulton R. Pharmacologic and anatomic considerations in recovery was stopped after 90 days and we have no idea of the of function. Phys Med Rehabil 1992; 6: 375–93.
long-term development of patients' motor function and 2 Liepert J. Pharmacotherapy in restorative neurology. whether the treatment eff ect persisted in the months Curr Opin Neurol 2008; 21: 639–43.
after treatment was stopped. Third, a potential random 3 Loubinoux I, Chollet, F. Neuropharmacology in stroke recovery. In: Cramer SC, Nudo RJ, eds. Brain repair after stroke. Cambridge: error cannot be eliminated completely and our results Cambridge University Press, 2010: 183–93.
remain to be confi rmed, even though error is unlikely 4 Windle V, Corbett D. Fluoxetine and recovery of motor function because of the low p value associated with the change in after focal ischemia in rats. Brain Res 2005; 1044: 25–32.
Lim CM, Kim SW, Park JY, Kim C, Yoon SH, Lee JK. Fluoxetine the FMMS score at day 90 (p=0·003). However, the aff ords robust neuroprotection in the postischemic brain via its eff ect of fl uoxetine seems to be strong and clinically anti-infl ammatory eff ect. Neurosci Res 2009; 87: 1037–45.
relevant, and the data show a global coherence (ie, motor Li WL, Cai HH, Wang B, et al. Chronic fl uoxetine treatment improvement in FMMS and motor items of NIHSS). improves ischemia-induced spatial cognitive defi cits through increasing hippocampal neurogenesis after stroke. J Neurosci Res Future studies should include a larger number of 2009; 87: 112–22.
patients whose characteristics are more similar to those Sonde L, Lökk J. Eff ects of amphetamine and/or L-dopa and of the general population of patients with stroke; the physiotherapy after stroke: a blinded randomized study.
Acta Neurol Scand 2007; 115: 55–59.
primary outcome criteria should be more functional www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8
Treig T, Werner C, Sachse M, Hesse S. No benefi t from 20 Brott T, Adams HP Jr, Olinger CP, et al. Measurements of acute D-amphetamine when added to physiotherapy after stroke: a cerebral infarction: a clinical examination scale. Stroke 1989; randomized, placebo-controlled study. Clin Rehabil 2003; 17: 590–99.
Platz T, Kim IH, Engel U, et al. Amphetamine fails to facilitate 21 Goldstein LB, Bertels C, Davis JN. Interrater reliability of the NIH motor performance and to enhance motor recovery among stroke stroke scale. Arch Neurol 1989; 46: 660–62.
patients with mild arm paresis: interim analysis and termination of 22 Rankin J. Cerebral vascular accidents in patients over the age of 60. a double blind, randomised, placebo-controlled trial. II. Prognosis. Scott Med J 1957; 2: 200–15.
Restor Neurol Neurosci 2005; 23: 271–80.
23 Montgomery SA, Asberg M. MADRS Scale. Brit J Psychiatry 1979; 10 Gladstone DJ, Danells CJ, Armesto A, et al. Subacute Therapy with Amphetamine and Rehabilitation for Stroke Study Investigators. 24 Duncan PW, Goldstein LB, Horner RD, Landsman PB, Samsa GP, Physiotherapy coupled with dextroamphetamine for rehabilitation Matchar DB. Similar motor recovery of upper and lower extremities after hemiparetic stroke: a randomized, double-blind, after stroke. Stroke 1994; 25: 1181–88.
placebo-controlled trial. Stroke 2006; 37: 179–85.
25 Altman DG, Dore CJ. Randomisation and baseline comparisons in 11 Walker Batson D, Smith P, Curtis S, et al. Amphetamine paired clinical trials. Lancet 1990 335: 149–53.
with physical therapy accelerates motor recovery after stroke. 26 Fruehwald S, Gatterbauer E, Rehak P, Baumhackl U. Early Further evidence. Stroke 1995; 26: 2254–59.
fl uoxetine treatment of post-stroke depression—a three-month 12 Dam M, Tonin P, De Boni A, et al. Eff ects of fl uoxetine and double-blind placebo-controlled study with an open-label long-term maprotiline on functional recovery in poststroke hemiplegic follow up. J Neurol 2003; 250: 347–51.
patients undergoing rehabilitation therapy. Stroke 1996; 27: 1211–14.
27 Robinson RG, Jorge RE, Moser DJ, et al. Escitalopram and 13 Pariente J, Loubinoux I, Carel C, et al. Fluoxetine modulates motor problem-solving therapy for prevention of poststroke depression: performance and cerebral activation of patients recovering from a randomized controlled trial. JAMA 2008; 299: 2391–400.
stroke. Ann Neurol 2001; 50: 718–29.
28 Anderson CS, Hackett ML, House AO. Interventions for preventing 14 Zittel S, Weiller C, Liepert J. Citalopram improves dexterity in depression after stroke. Cochrane Database Syst Rev 2004; chronic stroke patients. Neurorehabil Neural Repair 2008; 22: 311–14.
2: CD003689.
15 Acler M, Robol E, Fiaschi A, Manganotti P. A double blind placebo 29 Jacobs BL, Fornal CA. Serotonin and motor activity. RCT to investigate the eff ects of serotonergic modulation on brain Curr Opin Neurobiol 1997; 7: 820–25.
excitability and motor recovery in stroke patients. J Neurol 2009; 30 Barbas D, DesGroseillers L, Castellucci VF, Carew TJ, Marinesco S. Multiple serotonergic mechanisms contributing to sensitization in 16 Gerdelat-Mas A, Loubinoux I, Tombari D, Rascol O, Chollet F, aplysia: evidence of diverse serotonin receptor subtypes. Learn Mem Simonetta-Moreau M. Chronic administration of selective serotonin 2003; 10: 373–86.
reuptake inhibitor (SSRI) paroxetine modulates human motor 31 Tardy J, Pariente J, Leger A, et al. Methylphenidate modulates cortex excitability in healthy subjects. Neuroimage 2005; 27: 314–22.
cerebral post stroke reorganization. Neuroimage 2006; 33: 913–22.
17 Wiart L, Petit H, Joseph PA, Mazaux JM, Barat M. Fluoxetine in 32 Scheidtmann K, Fries W, Muller F, Koenig E. Eff ect of levodopa in early post-stroke depression: a double-blind placebo-controlled combination with physiotherapy on functional motor recovery after study. Stroke 2000; 31: 1829–32.
stroke: a prospective, randomised, double-blind study. Lancet 2001; 18 Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Stegling S. The post stroke hemiplegic patient. Scand J Rehabil Med 1975; 7: 13–31.
19 Gladstone DJ, Danells CJ, Black SE. The Fugl Meyer assessment of motor recovery after stroke: a critical review of its measurement
properties. Neurorehabil Neural Repair 2002; 16: 232–40.
www.thelancet.com/neurology Published online January 10, 2011 DOI:10.1016/S1474-4422(10)70314-8

Source: http://tonic.inserm.fr/medias/fichier/cholletlancetneurol2011_1473933379535-pdf?INLINE=FALSE