Psychiatry-malaysia.org

Hum Psychopharmacol Clin Exp 2003; 18: 379–384.
Published online 9 May 2003 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/hup.490 Fluvoxamine versus fluoxetine in major depressive episode:a double-blind randomised comparison 1Hoˆpital Neurologique Pierre Wertheimer, F-69003 Lyon, France2University Hospital Maastricht, 6202 AZ Maastricht, The Netherlands A double-blind, multinational study was conducted to compare the efficacy and safety of fluvoxamine and fluoxetine inoutpatients with major depressive episode; 184 patients were randomised to fluvoxamine (100 mg/day) or fluoxetine(20 mg/day) for 6 weeks. Both drugs were effective and there were no statistically significant differences between themin the area under the curve of change from baseline in the Hamilton depression rating scale (HAMD) total score. However,the percentage of HAMD responders (!50% decrease in HAMD total score) at week 2, the clinical global improvementseverity of illness score at week 2 and the depression subscale of the irritability, depression and anxiety scale at weeks1, 2 and 4, all showed significant advantages for fluvoxamine. During the last 2 weeks, fluvoxamine was significantlymore effective in improving the HAMD sleep disturbance scale. Both drugs were well tolerated and there were no markeddifferences in their side effect profiles which were typical of SSRIs. Fluvoxamine and fluoxetine have similar efficacy andsafety profiles in the treatment of major depressive episode; the findings of this study indicate that fluvoxamine may have afaster onset of action with respect to resolution of depressive symptoms and result in a better improvement in sleep quality.
Copyright # 2003 John Wiley & Sons, Ltd.
key words — fluvoxamine; fluoxetine; selective serotonin reuptake inhibitor; depression with another. However, available evidence suggeststhat all SSRIs do not share the same profile (Leonard, Although tricyclic antidepressants are effective in the 1992; van Harten, 1993; Montgomery et al., 1994; treatment of depression, they are associated with a Wilde et al., 1993; DeVane, 1995; Hindmarch, number of adverse events (such as dry mouth, tremor, 1995; van den Berg, 1995; Price et al., 1996; Lane, low blood pressure and cardiovascular effects) that 1997; Waldinger et al., 1998). For example, there hinder compliance and may occasionally be danger- are considerable differences in their chemical struc- ous. Selective serotonin reuptake inhibitors (SSRIs) tures (van den Berg, 1995) and in their pharmacologi- generally show comparable efficacy to the tricyclic cal (Leonard, 1992) and pharmacokinetic (van Harten, antidepressants (Anderson and Tomenson, 1994) but 1993) behaviour. Findings from meta-analyses have have a more benign safety profile (Montgomery suggested there may be differences in efficacy between the SSRIs in certain patient subgroups, such Despite a wealth of studies comparing the SSRIs as those with severe depression (Anderson and with other classes of antidepressant and placebo, there Tomenson, 1994). There is also evidence that there have been few studies directly comparing one SSRI are differences in their safety profiles, with paroxetineand sertraline being associated with more sexual side-effects (Nemeroff et al., 1995; Lane, 1997; Waldingeret al., 1998) and fluoxetine being associated with * Correspondence to: Dr J. Dalery, Unite´ de PsychologieMe´dicale et de Psychiatrie de Liaison, Hoˆpital Neurologique more anxiety and nervousness (Aguglia et al., 1993).
Pierre Wertheimer, 59 Boulevard Pinel, F-69003 Lyon, France.
This study, the results have which have not pre- Tel: þ33 472 11 80 64. Fax: þ33 472 35 71 09.
viously been published in full, was designed to Contract/grant sponsor: Solvay Pharmaceuticals.
directly compare the efficacy and safety of two widely Copyright # 2003 John Wiley & Sons, Ltd.
prescribed SSRIs, fluvoxamine and fluoxetine, in a large, multicentre, randomised, double-blind study The primary efficacy parameter was the area under the in patients with a major depressive episode.
curve (AUC) of the change in 17-item HAMD totalscore from baseline (expressed as HAMD weeks).
Secondary efficacy variables comprised: the numberof 17-item HAMD responders (i.e. the number of This was a prospectively randomised, double-blind, patients with at least a 50% improvement in 17-item fixed-dose, parallel-group study conducted according HAMD total score); the clinical global impression to the principles of good clinical practice in 24 Eur- (CGI) (Guy, 1976) severity of illness and global improvement scores; the clinical anxiety scale(CAS) score; the irritability, depression and anxiety scale (IDAS) total score and subscores; the Beck’sscale for suicide ideation (SSI) score; sleep evaluation Outpatients of either sex, aged 18–70 years old, meet- and the 17-item HAMD total and subtotal scores.
ing DSM-III-R (American Psychiatric Association, The 17-item HAMD was determined at screening, 1987) criteria for major depressive episode and having baseline and after 1, 2, 4 and 6 weeks (or upon prema- a total score of 17 or more on the 17-item Hamilton ture termination) of treatment; all other efficacy vari- depression rating scale (HAMD) (Hamilton, 1967) ables were determined at all visits except screening.
were included in the study. Patients were excluded if Adverse events were documented at baseline and at they had: acute suicidal ideation or a serious suicide each subsequent visit; vital signs and weight were attempt in the previous 6 months; dementia; a history measured at baseline and at the final visit and a phy- of epilepsy or seizures; concurrent or recent (6 sical examination was conducted at screening and at months) alcoholism, other psychoactive substance abuse or drug-induced psychosis; were pregnant, lac-tating or of childbearing potential and not taking ade-quate contraceptive measures; or if they had clinically uncontrolled hepatic, renal, pulmonary, endocrine orcollagen disease. Also excluded were patients who Efficacy was assessed using the intent-to-treat (ITT) had previously failed SSRI therapy or who required efficacy sample (i.e. patients who received at least concomitant lithium, warfarin, hepatically metabo- one dose of study medication and provided at least lised antivitamin K agents, carbamazepine, theophyl- one valid post-baseline efficacy evaluation on study line, insulin or hypoglycaemic agents. Patients were medication) and the per-protocol (PP) sample (i.e.
required not to receive monoamine oxidase inhibitors subset of ITT patient sample who did not have any ser- or electroconvulsive therapy (ECT) in the 2 weeks ious deviation from the study protocol). Analyses of prior to the study. All patients provided written or secondary efficacy variables were performed on the verbally witnessed informed consent.
ITT efficacy sample only. All analyses were per-formed using visit-wise (observed cases; OC) andlast observation carried forward (LOCF) data. The Wilcoxon two-sample test was used to analyse both All patients entered a 7-day placebo run-in period that the primary and the secondary efficacy variables. All enabled previous antidepressant therapy to be with- tests were two-sided. Differences with p values drawn and identified any patients who were likely to 0.05 were considered to be statistically significant.
respond to placebo. Eligible patients were then rando-mised to receive oral fluvoxamine (100 mg/day) or fluoxetine (20 mg/day) for 6 weeks; fluvoxamine(Solvay Pharmaceuticals) was given at a dose of One-hundred and eighty-four patients were rando- 50 mg/day for the first 7 days of the study. Doses were mised to treatment, 90 to fluvoxamine (33 male, 57 as recommended by the respective manufacturers for female; mean age 42 years) and 94 to fluoxetine the treatment of major depressive episode.
(35 male, 59 female; mean age 42.1 years); there were Oxazepam or nitrazepam were permitted as neces- no significant differences between the two treatment sary for night-time sedation; no other psychothera- groups with respect to gender, race, age distribution, peutic treatments or ECT were permitted during the body weight or height. Four patients in the fluvoxa- mine group and three in the fluoxetine group had no Copyright # 2003 John Wiley & Sons, Ltd.
Hum Psychopharmacol Clin Exp 2003; 18: 379–384.
post-baseline assessment and were therefore excludedfrom the ITT efficacy sample (86 fluvoxamine patientsand 91 fluoxetine patients). A further four patients inthe fluvoxamine group and seven in the fluoxetinegroup had major deviations from the protocol andwere excluded from the PP sample (82 fluvoxaminepatients and 84 fluoxetine patients). Reasons forexclusion were failure to meet inclusion or exclusioncriteria (3 fluvoxamine, 3 fluoxetine patients), disal-lowed prior medication (1 fluvoxamine and 3 fluoxe-tine patients) and baseline HAMD less than 17(I fluoxetine patient). Sixteen patients in the fluvoxa-mine group and ten in the fluoxetine group withdrewfrom the study prematurely.
Figure 1. Percentage of 17-item HAMD responders (ITT sample;LOCF analysis) Fluvoxamine and fluoxetine both resulted in a signifi-cant improvement in depressive symptoms.
CGI severity of illness score. The overall severity ofillness was progressively reduced in both groups from Primary variable. The mean AUC of the change from ‘moderately to markedly ill’ to ‘borderline to mildly baseline in 17-item HAMD total score decreased ill’. However, the reduction in the mean score was progressively during the study in both groups. After 6 greater with fluvoxamine than with fluoxetine at all weeks of treatment, the mean AUC value was À189 assessments, the difference achieving statistical sig- HAMD weeks in the fluvoxamine group and À175 nificance ( p 0.05) at week 2 (Figure 2).
HAMD weeks in the fluoxetine group. HAMD scoreswere reduced from a mean of 22.3 and 22.2 at baseline CGI global improvement score. Again, a continuous in the fluvoxamine and fluoxetine groups respectively improvement was seen in both groups; at the end of to 10.0 and 11.3, respectively, at day 42. There were the study, the mean score was 2.0 (much improved) in no statistically or clinically significant differences the fluvoxamine group and 2.2 (minimally to much between the treatments at any point or in any of the improved) in the fluoxetine group. There were no significant differences between the treatments at anypoint.
Secondary variables. The secondary efficacy vari-ables confirmed the overall effectiveness of fluvox-amine and fluoxetine, but revealed some benefits offluvoxamine over fluoxetine. The data are presentedfor the LOCF analysis; similar results were obtainedfrom the OC analysis.
Percentage of 17-item HAMD responders. A patientwas considered to be a responder if they had areduction in 17-item HAMD total score of at least50% from baseline. Patients appeared to respondbetter to fluvoxamine than to fluoxetine, especiallyduring the early stages of treatment (Figure 1); after 2weeks of treatment, the percentage of patients whoresponded was significantly higher with fluvoxaminethan with fluoxetine (29% vs 16%; p 0.05). At theend of the study, the majority (approximately 60%) ofpatients in both groups were considered to be Figure 2. Mean CGI severity of illness score (ITT sample; LOCF Copyright # 2003 John Wiley & Sons, Ltd.
Hum Psychopharmacol Clin Exp 2003; 18: 379–384.
Mean change in IDAS total and subscores (ITT sample; ap 0.05 fluvoxamine versus fluoxetine.
Figure 3. Mean change in sleep evaluation score (ITT sample;LOCF analysis) CAS total score. The mean score decreased progres-sively in both groups. In the fluvoxamine group, themean score fell from 12.6 at baseline to 6.9 at the end (25% in the fluvoxamine group and 30% in the of the study; the corresponding reduction in the fluoxetine group) which was allowed in the protocol fluoxetine group was from 13.1 to 7.7. There were nosignificant differences between the treatments at any 17-item HAMD total and subtotal scores. The mean total score decreased progressively in both groups; atthe end of the study, the mean score had fallen from IDAS total score and subscores. The IDAS total and 22.3 to 10.0 in the fluvoxamine group and from 22.2 to subscores, which assess levels of irritability, depres- 11.3 in the fluoxetine group. There were no significant sion and anxiety, all fell progressively during the differences between the treatments at any point.
study. As shown in Table 1, the reduction in the mean Fluvoxamine was associated with a greater reduc- total score was greater with fluvoxamine than with tion in anxiety/somatisation and cognitive disturbance fluoxetine at all time points. Mean changes in the subscores, although there were no statistically signifi- anxiety, irritability and depression subscores were cant differences between the groups. The improvement also greater with fluvoxamine at all time points; a in the retardation subscore was similar in each group.
statistically significant difference being achieved with However, in agreement with the sleep evaluation score, the depression subscore at weeks 1, 2 and 4.
fluvoxamine was more effective than fluoxetine atimproving the quality of sleep. The improvement in Beck’s SSI score. The mean score decreased progres- the sleep disturbance subtotal was statistically signifi- sively in both groups; at the end of the study, it had cantly greater with fluvoxamine than with fluoxetine at fallen from 4.7 to 1.6 in the fluvoxamine group and week 4 (À2.3 vs À1.6; p 0.05) and at the end of the from 5.4 to 2.1 in the fluoxetine group. There were no study (À2.4 vs À1.8; p 0.05) and was of borderline significant differences between the treatments at any significance ( p ¼ 0.052) at week 1 (Figure 4).
Sleep evaluation. There was a gradual improvement inmean sleep evaluation scores in both groups over the Fluvoxamine and fluoxetine were both well tolerated.
course of the study. As shown in Figure 3, the There was a similar incidence of adverse events in each improvement in the quality of sleep was more pro- group, the majority of which were mild or moderate.
nounced with fluvoxamine at weeks 4 and 6, although The most frequently reported treatment-emergent the differences did not achieve statistical significance.
signs and symptoms (TESS) in both groups were nau- Overall, slightly more than half of all patients (51% in sea (24% with fluvoxamine and 20% with fluoxetine) both groups) used a hypnotic drug during the study.
and headache (13% with fluvoxamine and 14% with The most frequently used hypnotic was oxazepam fluoxetine). No other TESS occurred in more than Copyright # 2003 John Wiley & Sons, Ltd.
Hum Psychopharmacol Clin Exp 2003; 18: 379–384.
Fluvoxamine also appeared to have a more benefi- cial effect on sleep quality, especially after 4 weeks oftreatment. The reduction in sleep disturbance was sig-nificantly greater with fluvoxamine than with fluoxe-tine from week 4 onwards according to the 17-itemHAMD sleep disturbance score. Moreover, fluvoxa-mine was clearly superior to fluoxetine on the sleepevaluation score from week 4 onwards. Patients trea-ted with fluvoxamine showed a progressive improve-ment in sleep quality throughout the study, whilstthose given fluoxetine showed little improvementafter the second week. These findings are consistentwith clinical experience suggesting that fluvoxaminecauses less sleep disturbance and is less activating thanfluoxetine (Baldessarini and Marsh, 1990; Rickels andSchweizer, 1990; Freeman, 1991; Aguglia et al., Figure 4. Mean change in 17-item HAMD sleep disturbance Both treatments improved anxiety symptoms, irrit- ability, suicidal ideation, retardation and cognitive 10% of patients in either group. Sexual dysfunction disturbance. A somewhat greater improvement in was reported by one patient in the fluvoxamine group anxiety, irritability and cognitive disturbance was (ejaculatory abnormality) and two in the fluoxetine observed with fluvoxamine, but these differences group (ejaculatory abnormality and impotence).
did not achieve statistical significance. Again, this is There were four patients with serious adverse consistent with the propensity of fluoxetine to cause events. One (anxiety and alcohol intoxication) occur- activation (Baldessarini and Marsh, 1990; Rickels red in the fluvoxamine group and was considered and Schweizer, 1990; Aguglia et al., 1993; Lane unrelated to treatment and three occurred in the fluoxetine group (anorexia and stomach complaints There are a number of clear differences between that were considered highly probably related to treat- fluvoxamine and fluoxetine. They have completely ment in one patient, a suicide attempt that was consi- different chemical structures and pharmacokinetic dered possibly related to treatment in another and a and pharmacological profiles. For example, fluoxetine suicide attempt that was considered unrelated to treat- has a considerably longer half-life than fluvoxamine (2–4 days vs 17–22 h) (Preskorn, 1997) and is the only There were no clinically significant changes in vital SSRI with a metabolite that has significant clinical signs, body weight or physical condition in either activity (Lane et al., 1995). In addition, fluoxetine reduces dopamine synthesis (Baldessarini and March,1990) whilst fluvoxamine has no effect on dopaminer-gic neurotransmission.
However, there have been very few studies directly Fluvoxamine (100 mg/day) and fluoxetine (20 mg/day) comparing the efficacy of one SSRI with another. This both resulted in a progressive, clinically significant direct, double-blind comparison of fluvoxamine and improvement in 17-item HAMD total score over the fluoxetine in patients with a major depressive episode course of treatment. However, analysis of secondary was therefore designed to provide data on potential efficacy parameters indicated that fluvoxamine may areas of differentiation between the efficacy of the have some advantage over fluoxetine during the early two treatments. Thus, a wide range of variables were phase of treatment: the percentage of 17-item HAMD selected in order to assess all the characteristic symp- responders, the CGI severity of illness score and the toms of major depressive episode including depres- IDAS depression subscore all showed a significantly sion, sleep disturbance, irritability, anxiety, cognitive greater improvement with fluvoxamine after 2 weeks disturbance, retardation and suicidal ideation.
of treatment. However, whilst the differences between A number of previous studies have suggested that the two treatments were maintained after 4 weeks of the SSRIs may show some differences in their safety treatment, they remained statistically significant only profiles. Thus, sertraline (Nemeroff et al., 1995) and paroxetine (Waldinger et al., 1998) appear to be Copyright # 2003 John Wiley & Sons, Ltd.
Hum Psychopharmacol Clin Exp 2003; 18: 379–384.
associated with more sexual side-effects than fluvox- Freeman CP. 1991. Fluvoxamine: clinical trials and clinical use.
amine, whilst fluoxetine (Aguglia et al., 1993; Stokes, J Psychiatr Neurosci 16(Suppl. 1): 19–25.
Guy W. 1976. ECDEU Assessment Manual for Psychopharmacol- 1993; DeVane, 1995) has been reported to cause more ogy. National Institute of Mental Health, US Department of anxiety and nervousness than the other SSRIs. Simi- Health, Education and Welfare: Washington, DC.
larly, in accordance with its pharmacological profile, Hamilton M. 1967. Development of a rating scale for primary paroxetine (Dunbar, 1989; Kiev and Feigher, 1997) depressive illness. Br J Soc Clin Psychol 6: 278–296.
Hindmarch I. 1995. The behavioural toxicity of the selective sero- tends to be associated with a higher incidence of tonin reuptake inhibitors. Int Clin Psychopharmacol 9(Suppl. 4): anticholinergic effects (such as sedation, sweating and dry mouth) than the other SSRIs. However, Kiev A, Feiger A. 1997. A double-blind comparison of fluvoxamine neither fluvoxamine nor fluoxetine have adverse car- and paroxetine in the treatment of depressed outpatients. J Clin diac side-effects, as measured using both ECG and Lane RM. 1997. A critical review of selective serotonin reuptake echocardiography, a more direct reflection of cardiac inhibitor-related sexual dysfunction: incidence, possible aetiol- functioning (Strik et al., 1998). In the current study, ogy and implications for management. J Psychopharmacol 11: fluvoxamine and fluoxetine were equally well toler- ated and had very similar safety profiles. Nausea Lane R, Baldwin D, Preskorn S. 1995. The SSRIs: advantages, dis- advantages and differences. J Psychopharmacol 9(Suppl.): 163– and headache were the most common adverse events and were reported by similar proportions of patients Leonard BE. 1992. Pharmacological differences of serotonin reuptake inhibitors and possible clinical relevance. Drugs In conclusion, fluvoxamine 100 mg/day is at least as effective and safe as fluoxetine 20 mg/day in the Montgomery SA, Henry J, McDonald G, et al. 1994. Selective ser- otonin reuptake inhibitors: meta-analysis of discontinuation treatment of major depressive episode and may be rates. Int Clin Psychopharmacol 9: 47–53.
superior in terms of onset of action and improvement Nemeroff CB, Ninan PT, Ballenger J, et al. 1995. Double-blind mul- ticenter comparison of fluvoxamine versus sertraline in the treat-ment of depressed outpatients. Depression 3: 163–169.
Preskorn SH. 1997. Clinically relevant pharmacology of selective serotonin reuptake inhibitors. Clin Pharmacokinet 32(Suppl. 1): The study was supported by a grant from Solvay Phar- Price JS, Waller PC, Wood SM, et al. 1996. A comparison of the postmarketing safety of four selective serotonin reuptake inhibi- tors including the investigation of symptoms occurring on with-drawal. Br J Clin Pharmacol 42: 757–763.
Rickels K, Schweizer E. 1990. Clinical overview of serotonin reup- take inhibitors. J Clin Psychiatry 51(Suppl. B): 9–12.
Stokes PE. 1993. Fluoxetine: a five-year review. Clin Ther 15: Aguglia E, Casacchia M, Cassano GB, et al. 1993. Double-blind study of the efficacy and safety of sertraline versus fluoxetine Strik JJMH, Honig A, Lousberg R, et al. 1998. Cardiac side-effects of in major depression. Int Clin Psychopharmacol 8: 197–202.
two selective serotonin reuptake inhibitors in middle-aged American Psychiatric Association. 1987. Diagnostic and Statistical and elderly depressed patients. Int Clin Psychopharmacol 13: 1–5.
Manual of Mental Disorders 3rd revised edn (DSM-III-R).
van den Berg SJ. 1995. Comparing SSRIs: from chemistry to clin- American Psychiatric Association: Washington, DC.
ical choice. Hum Psychopharmacol 10: S199–S209.
Anderson IM, Tomenson BM. 1994. The efficacy of selective sero- Van Harten J. 1993. Clinical pharmacokinetics of selective seroto- tonin reuptake inhibitors in depression: a meta-analysis of studies nin reptake inhibitors. Clin Pharmacokinet 24: 203–220.
against tricyclic antidepressants. J Psychopharmacol 8: 238–249.
Waldinger MD, Hengeveld MW, Zwinderman AH, et al. 1998. The Baldessarini RJ, Marsh E. 1990. Fluoxetine and side effects. Arch effect of SSRI antidepressants on ejaculation: a double-blind, randomised, placebo-controlled study with fluoxetine, fluvoxa- DeVane CL. 1995. Comparative safety and tolerability of selective mine, paroxetine and sertraline. J Clin Psychopharmacol 18: serotonin reuptake inhibitors. Hum Psychopharmacol 10(Suppl.
Wilde MI, Plosker GL, Benfield P. 1993. Fluvoxamine: an updated Dunbar GC. 1989. An interim overview of the safety and tolerability review of its pharmacology, and therapeutic use in depressive ill- of paroxetine. Acta Psychiat Scand 80(Suppl. 350): 135–137.
Copyright # 2003 John Wiley & Sons, Ltd.
Hum Psychopharmacol Clin Exp 2003; 18: 379–384.

Source: http://www.psychiatry-malaysia.org/file_dir/180824807244f300668f7f9.pdf