Double-blind comparison of fluoxetine and nortriptyline in the treatment of moderate to severe major depression

Journal of Clinical Pharmacy and Therapeutics (2003) 28, 379–384 Double-blind comparison of fluoxetine and nortriptylinein the treatment of moderate to severe major depression S. Akhondzadeh*  PhD, H. Faraji* MD, M. Sadeghi* MD, K. Afkham* MD,H. Fakhrzadehà MD and A. Kamalipour* BSc*Psychiatric Research Center, Roozbeh Psychiatric Hospital, Tehran University of Medical Sciences, Tehran,Iran,  Institute of Medicinal Plants, Tehran, Iran and àEndocrinology and Metabolism Research Center,Tehran University of Medical Sciences, Tehran, Iran Results: The results suggest that the efficacy of nortriptyline is superior to fluoxetine in this Background: Depression is an international pub- group of major depressed patients. No significant lic health problem. Impairment in social and differences were observed between dropout rates occupational functioning, increased comorbidity in the two groups but anti-cholinergic side-effects with other psychiatric and medical conditions, were significantly more frequent with nortriptyl- and an increased risk of mortality are a few of its ine than with fluoxetine but there was no signi- ficant difference in cardiovascular effects in impression that selective serotonin re-uptake inhibitors may not work as well as tricyclic anti- Conclusion: The results of the current study depressants in severe depression and/or melan- suggest that nortriptyline was more effective than cholia. On the contrary, there is a general belief fluoxetine in the treatment of moderate to severe that selective serotonin re-uptake inhibitors are depression. A larger study is warranted.
superior to the tricyclic anti-depressants inhaving fewer side-effects, particularly cardiovas- cular effects. The objective of this double-blind study was to compare the efficacy and safety offluoxetine and nortriptyline in patients withmoderate to severe major depression.
Methods: A total of 48 adult outpatients who met Depression, thought to result from biochemical the Diagnostic and Statistical Manual of Mental changes in the brain (1), is a common disease of Disorders (DSM IV), forth edition for major adulthood. This affective disorder afflicts about 5% depression, based on the structured clinical of the adult population in the United States. The interview for DSM IV participated in the trial.
mainstay for the treatment of depression is phar- Patients had a baseline Hamilton Rating Scale for macotherapy, usually in combination with some Depression score of at least 20. In this double- type of limited supportive psychotherapy (2, 3).
blind, single-center trial, patients were randomly Until the 1980s, the so-called Tricyclic Anti- assigned to receive nortriptyline 150 mg/day depressants (TCAs) had been the first-line drugs; (group 1) or fluoxetine 60 mg/day (group 2) for however, this situation has changed because of the 6-weeks. The outcome of the two groups was availability of newer second-generation anti- assessed using Hamilton Depression Rating depressants that have more favorable side-effect Scale, a side-effect checklist and a regular ECG profile and low toxicity associated with an over- The effectiveness of the newer anti-depressants Received 17 April 2003, Accepted 12 August 2003 in comparison with the more established agents is Correspondence: Dr Shahin Akhondzadeh, No. 29, 39th Street,Gisha Street, Tehran 14479, Iran. Tel.: +98-21-542222; fax: +98-21- frequently questioned. Although selective sero- tonin re-uptake inhibitors (SSRIs) show a high degree of pharmacologic selectivity, this does not For moderate to severe depression, nortriptyline seem to translate into improved clinical efficacy.
continues to be regarded as the most effective The clinical effectiveness of the tricyclic agents may treatment, particularly for elderly patients. Because be attributed to their dual inhibitory effects on both of its benign side-effect profile in both acute and noradrenergic and serotonergic systems (5, 6).
long-term treatment, nortriptyline is the most There is a commonly held belief among psychia- widely prescribed TCA today (12, 13). To our trists that newer anti-depressants, notably the knowledge, there is only one published controlled selective serotonin re-uptake inhibitors, are clinic- comparison of the efficacy and safety of fluoxetine ally somewhat less effective than the TCAs in the and nortriptyline in major depression (14). The treatment of severe depression. The limited num- studies that indicate superiority of SSRIs and in ber of clinical trials of SSRIs conducted in this particular fluoxetine over TCAs, have used older patient population has, however, generally failed reference TCAs such as imipramine or amitriptyl- to confirm this contention (7–9). Nevertheless, the ine although nortriptyline has a very different side- possibility of real differences in efficacy between TCAs and newer anti-depressants cannot be Against this background we carried out a rand- omized trial of nortriptyline and fluoxetine in a Recently, the FDA and members of the medical double-blind study to compare the efficacy and community have raised concerns about medica- safety of fluoxetine and nortriptyline and in par- tions associated with QTc prolongation. In normal ticular their cardiovascular effects in patients with cardiac conduction, the QTc interval is approxi- moderate to severe major depression.
mately 420 ms. However, if repolarization isdelayed, prolongation of the QT interval results. Ingeneral, a QTc longer than 450 ms is of potential concern. Prolongation longer than 500 ms indicates an elevated risk of progression to tachyarrhythmias(torsade de points), which may be associated with This was a 6-week, parallel group, randomized trial symptoms such as palpitation, dizziness, light- headedness, and syncope; and progression to Psychiatric Hospital, Tehran, Iran during January ventricular fibrillation which can potentially cause The effects of anti-depressant drugs on the car- diac conduction system have been of long standingconcern after the observation that overdoses of A total of 48 adult outpatients aged 19–54 who met TCAs frequently lead to cardiac-related death. At the Diagnostic and Statistical Manual of Mental therapeutic doses, these agents have been shown to 1 Disorders, Forth edition (16) (DSM IV) for major increase PR, QRS, and QT intervals and to have depression based on the structured clinical inter- quinidine-like class IA anti-arrhythmic properties.
view for DSM IV participated in the trial. Patients All drugs with class I action, including the TCAs, have a baseline Hamilton Rating Scale for Depres- block fast sodium channels, and such activity more sion (HAM-D 17-item) (17) score of at least 20. The recently has been associated with increased mor- HAM-D, the most widely used physician-admin- tality in the context of underlying cardiovascular istrated rating scale for depression, summates disease. Subsequently, systematic studies of the 17 individual item scores to provide a total score TCAs in patients with and without cardiac disease indicative of the severity of depression. Participa- documented a number of cardiovascular effects, tion was precluded for patients with any other most notably increase in heart rate and orthostatic primary psychiatric disease, current or past history hypotension. As a class, the SSRIs are believed to of bipolar disorder, for patients requiring treatment have more benign cardiovascular safety profiles with psychoactive drugs like anxiolytic, Monoam- than do the TCAs because they do not block ine Oxidase Inhibitors (MAOIs) or tryptophan, fast sodium channels or prolong PR or QRS patients with organic brain disorders including epilepsy, patients with predominantly suicidal Ó 2003 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 28, 379–384 Fluoxetine vs. nortriptyline in the treatment of depression tendencies, and with any severe general disease. In addition, a one-way repeated measures ANOVA addition, pregnant and lactating women and two-tailed post hoc Tukey mean comparison test patients with alcohol or drug dependency were excluded. As depression is a serious and poten- Depression Rating Scale scores from baseline. To tially life-threatening condition and the partici- compare the reduction in the Hamilton Depression pants were outpatients, extensive safeguards were Rating Scale score at week 6 compared with baseline, needed. Patients were excluded if they posed a an unpaired two-sided Student’s t-test was used.
significant risk of suicide at any time during par- Results are presented as mean ± SEM differences and were considered significant with P £ 0Æ05. To com- informed consent, and the protocol satisfied the pare the baseline data and frequency of side-effects Ministry of Health and Medical education of Iran’s between the protocols, Fisher’s exact test was per- formed. A traditional ‘observed cases’ (OC, the Patients were randomly assigned to receive patients who completed the trial) analysis at 6 weeks tablet nortriptyline 150 mg/day (titrated-up over was the primary efficacy analysis. In addition, 4 weeks and were put in capsules to provide sim- intention to treat (ITT) analysis with last observation ilar appearance with fluoxetine) (group 1) or fluo- carried forward (LOCF) procedure was also per- formed. All results discussed are based on OC ana- (group 2) for the 6-week study. Eleven patients dropped out of the trial leaving 37 subjects whomet the DSM IV criteria for major depression to complete the trial. Patients were assessed by athird-year resident of psychiatry, using a stan- A total of 106 patients were screened for the study dardized protocol for the HAM-D, at baseline and and 48 were randomized to trial medication after 1, 2, 4, and 6 weeks after the medication (24 patients in each group). No significant differ- started. The principal measure of the outcome was ences were identified between patients randomly the 17-item HAM-D. The mean decrease in HAM-D assigned to the two groups with regard to basic score from baseline was used as the main outcome demographic data including age and gender measure of response of depression to treatment.
(Table 1). Thirty-seven patients completed the trial.
In the nortriptyline and fluoxetine group thenumber of dropouts were 4, and 7, respectively.
Although the number of dropouts in the fluoxetine group was higher than the nortriptyline group, this throughout the study and were assessed using a was not significant (P = 0Æ49) (Fig. 1).
checklist administered by a resident of psychiatryon days 7, 14, 28, and 42 (Table 2). All the clinical assessments, the ECG recordings and the assess-ment of changes in blood pressure were made The mean ± SEM scores of two groups of patients blind with respect to the medication received and are shown in Fig. 2. There were no significant dif- were assessed on days 7, 14, 28, and 42. Systolic ferences between the two groups in week 0 (base- and diastolic hypertension was considered as line) on the Hamilton Depression rating scale A two-way repeated measures ANOVA (time–treat- ment interaction) was used. The two groups as abetween-subjects factor (group) and the five meas- urements during treatment as the within-subjects factor (time) were considered. This was done for Hamilton Depression Rating Scale scores. In Ó 2003 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 28, 379–384 from week 1, on the Hamilton Depression ratingscale scores. The difference between the two treat- ments was not significant at the endpoint (OCanalysis) (week 6) (t = 1Æ79, d.f. = 35, P = 0Æ08) but was significant in the ITT analysis (t = 2Æ58,d.f. = 46, P = 0Æ01). The changes at the endpoint (mean ± SD) and )16Æ82 ± 11Æ08 for nortriptyline and fluoxetine, respectively. A significant differ- ence was not observed on the change of scores of the Hamilton Depression rating scale at week 6 compared with baseline in the two groups in theOC analysis (t = 1Æ09, d.f. = 35, P = 0Æ27) whereas in ITT analysis the difference was significant (t = 2Æ01,d.f. = 46, P = 0Æ04).
Nortriptyline 150 mg/day
Table 2. Number of patients with side-effects.

Fluoxeting 60 mg/day
Hamilton Depression Scores
Trial (weeks)
**P < 0Æ01 and ***P < 0Æ001.
between the two protocols was not quite significant as indicated by the effect of group, the between- subjects factor (F = 2Æ83, d.f. = 1, P = 0Æ09) but was significant in the ITT analysis (F = 6Æ10, d.f. = 1, P = 0Æ01). The behavior of the two treatments was homogeneous across the time (groups-by-time interaction, Greenhouse–Geisser correction; F = 0Æ93, d.f. = 2Æ24, P = 0Æ40). In addition, a one-way effect of both treatments on the Hamilton Depres- sion rating scale scores (P < 0Æ0001). In both groups post hoc comparisons showed a significant change Ó 2003 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 28, 379–384 Fluoxetine vs. nortriptyline in the treatment of depression tempered by the small sample size. However, even if nortriptyline is more effective than fluoxetine for Twenty-eight side-effects were observed over the this patient group, other factors such as anti- trial. The difference between the nortriptyline and cholinergic effects in particular for elderly patients fluoxetine in the frequency of side-effects was not must be considered in the selection of an anti- significant except for blurred vision, constipation, dry mouth, headache, nausea, and appetite chan-ges (Table 2). In the nortriptyline group nine patients had hypertension. Seven of them haddiastolic hypertension and two patients had both This study formed part of the postgraduate thesis systolic and diastolic hypertension. In fluoxetine of Dr Faraji. The Authors wish to thank Dr Mar- group eight patients had hypertension. Six of them had diastolic hypertension and two patients hadboth systolic and diastolic hypertension. In the nortriptyline group the only ECG abnormality wasbecause of sinus arrhythmias whereas in the fluo- 1. Baker GB, Coutts RT, Greenshaw AJ (2000) Neuro- xetine group four cases of ECG abnormalities were chemical and metabolic aspects of antidepressants: because of QT and QRS interval prolongation and an overview. Journal of Psychiatry and Neuroscience, 25,481–496.
T inversion. The difference between the nortript- 2. Greenberg P, Stiglin LE, Finkelstein S (1993) yline and fluoxetine in the frequency of QTc pro- Depression: a neglected major illness. Journal of longation was not significant (P = 0Æ48).
3. Judd L (1995) Mood disorders in the general popu- lation represent an important and worldwide publichealth problem. International Journal of Clinical Depression is an international public health issue Psychopharmacology, 10 (Suppl), 5–10.
with impairment in social and occupational func- 4. Mendlewich J (2001) Optimizing antidepressant use tioning, increased psychiatric and medical comor- in clinical practice: towards criteria for antide- bidity, and an increased risk of mortality among pressant selection. The Br. Journal of Psychiatry, depressed individuals as consequences (2, 3, 18).
There is an impression among psychiatrists that 5. Siegfried K (1997) Efficacy of antidepressants in the treatment of severe depression: The place of mit- SSRIs may not work as well as TCAs in severe razapine. Journal of Clinical Psychopharmacology, depression and/or melancholia (5) and a general belief that SSRIs are superior to the TCAs regarding 6. Gordon P (2001) New and old antidepressants: all equal in the eyes of the lore? British Journal of adverse effects (15, 19, 20). The present study suggests that nortriptyline is more effective than 7. Hirscfeld RMA (1999) Efficacy of SSRIs and newer fluoxetine in this group of major depressed antidepressants in severe depression: comparison patients. No significant differences were observed with TCAs. Journal of Clinical Psychiatry, 60, 326–335.
between dropout rates in the two groups but anti- 8. Anderson I (2000) Selective serotonin reuptake cholinergic side-effects were significantly more inhibitors versus tricyclic antidepressants: a meta common with nortriptyline than fluoxetine but analysis of efficacy and tolerability. Journal of Affective there was no significant difference in cardiovascu- 9. Bech P, Cialdella P, Haugh M, Birkett MA, Hours A, lar effects in particular QTc prolongation. Indeed, Boissel JP, Tollefson GD (2000) Mata-analysis of QTc prolongation was observed only in the fluo- randomized controlled trials of fluoxetine v. placebo xetine group in this study. However, the difference and tricyclic antidepressants in the short-term treat- was not significant (P = 0Æ48). In conclusion, the ment of major depression. British Journal of Psychiatry, results of the current study provide evidence in favor of an efficacy advantage nortriptyline over 10. Isaac M (1999) Where are we going with SSRIs? fluoxetine in the treatment of moderate to severe European Neuropsychopharmacology, 9 (Suppl), S101– depression. However, this conclusion should be Ó 2003 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 28, 379–384 11. Roose SR (2000) Considerations for the use of anti- in the treatment of major depressive disorder in pri- depressants in patients with cardiovascular disease.
mary care: A randomized comparison of fluoxetine American Heart Journal, 140, S84–S88.
and a tricyclic antidepressant. American Journal of 12. Navarro V, Gasto C, Torres X, Marcos T, Pintor L (2001) Citalopram versus nortriptyline on late-life 16. American Psychiatric Association (1994) Diagnostic depression: a 12-week randomized single-blind and Statistical Manual of Mental Disorders, 4th edn.
study. Acta Psychiatricia Scandinavica, 103, 435–440.
(DSM-IV). Washington DC: American Psychiatric 13. Gasto C, Navarro V, Marcos T, Porella MJ, Torra M, Rodanmilas M (2003) Single-blind comparison of 17. Hamilton M (1960) A rating scale for depression.
venlafaxine and nortriptyline in elderly major Journal of Neurology, Neurosurgery and Psychiatry, 23, depression. Journal of Clinical Psychopharmacology, 23, 18. British Association for Psychopharmacology (1993) 14. Joyce PR, Mulder RT, Luty SE, Sullivan PF, McKenzie Guidelines for treating depressive illness with anti- JM, Abbott RM, Stevens IF (2002) Patterns and pre- depressants. Journal of Psychopharmacology, 7, 19–23.
dictors of remission, response and recovery in major 19. Richelson E (1994) Pharmacology of antidepressants- depression treated with fluoxetine or nortriptyline.
characteristic of the ideal drug. Mayo Clinic Proceed- Australian and News land Journal of Psychiatry, 36, 384– 20. Demyttenaere K (1997) Compliance during treatment 15. Thompson C, Peveler RC, Stephenson D, McKendrick with antidepressants. Journal of Affective Disorders, 43, J (2000) Compliance with antidepressant medication Ó 2003 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 28, 379–384

Source: http://emri.tums.ac.ir/upfiles/54221228.pdf