Table 1.
Description of included studies.
| Maintenance phase | Acute Phase | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Study (year) | Total subjects/ Region |
Intervention | n | Age | %Female | Trial length (wks) |
Relapse definition | Dose(mg) | Discontinuation method | Acute phase length (wks) | Severity of acute phase (mean) | Scale | Continuation after remission (wks) | ||
| Stein et al. (1980) | 55 | Amitriptyline | (29) | 42.3 | (12.8) | 65.0 | 26 | NA | 100–150 | Flexible | abrupt | 8 | 25.1 | HAMD21 | 2 |
| North America | Placebo | (26) | 26 | ||||||||||||
| Doogan et al. (1992) | 295 | Sertraline | (185) | NA | NA | NA | 44 | CGI-S > = 4 | 50–200 | Flexible | abrupt | 8 | NA | NA | 0 |
| Europe | Placebo | (110) | 44 | ||||||||||||
| Montgomery et al. (1993a) | 135 | Paroxetine | (68) | 48.3 | (8.4) | 77.6 | 16/52 | (1) CGI-S > = 4, (2) deterioration of the CGI by 2 points or more, (3) patients meet DSM-III-R criteria for MDD of 2 weeks, (4) patients need antidepressant (5) depressive symptomatology was present for more than 7 days | 20–30 | Flexible | NA | 8 | 26.9 | HAMD21 | 0 |
| Europe | Placebo | (67) | 45.9 | (9.0) | 79.4 | 52 | |||||||||
| Montgomery et al. (1993b) | 147 | Citalopram | (105) | NA | NA | NA | 24 | MADRS score of 22 or more as a measure of the return of the symptoms of depression | 20/40 | Fixed | NA | 6 | NA | MADRS | 0 |
| Unclear | Placebo | (42) | 24 | ||||||||||||
| Robert et al. (1995) | 226 | Citalopram | (152) | 46.5 | NA | 68.9 | 24 | MADRS > = 25 and the clinical judgment of the investigator | 20/60 | Fixed | NA | 8 | NA | MADRS | 0 |
| Europe | Placebo | (74) | 49.5 | 73.0 | 24 | ||||||||||
| Stewart et al. (1997) | 32 | Imipramine | (17) | 38.0 | (7.0) | 66.0 | 26 | 2 consecutive weeks of a CGI -I > = 3 (compared with the pretreatment baseline). | 150–400 | Flexible | tapering | 12 | 13.0 | HAMD | 26 |
| North America | Placebo | (15) | 26 | ||||||||||||
| Keller et al. (1998) | 161 | Sertraline | (77) | 42.4 | (9.7) | 69.0 | 76 | (1) DSM-III-R criteria for MDD for at least 3 weeks, (2) CGI-S > = 4, CGI-I > = 3, (4) an increase in HAM-D to a score of 4 or more than maintenance phase baseline. | 50–200 | Flexible | tapering | 28 | 24.5 | HAMD24 | 16 |
| Unclear | Placebo | (84) | 40.8 | (9.0) | 62.0 | 76 | |||||||||
| Reimherr et al. (1998) | 395 | Fluoxetine | (299) | 40.5 | (10.5) | 80.2 | 14/38/50 | Met the criteria for MDD (even if all symptoms were classified as mild) for at least 2 weeks at any assessment during the double-blind phase or HAM-D > = 14 for 3 consecutive weeks | 20 | Fixed | NA | 12–14 | about 20 | HAMD | 0 |
| North America | Placebo | (96) | 40.2 | (10.5) | 65.9 | 50 | |||||||||
| Terra et al. (1998) | 204 | Fluvoxamine | (110) | 45.0 | (11.4) | 78.0 | 52 | Definition of DSM-III-R or suicide attempt | 100 | Fixed | tapering | 26 | 24.2 | HAMD21 | 19 |
| Europe | Placebo | (94) | 44.5 | (10.7) | 70.0 | 52 | |||||||||
| Feiger et al. (1999) | 131 | Nefazodone | (65) | 42.0 | NA | 71.0 | 36 | Either HAMD17 > = 18 for two consecutive visits or lack of efficacy | 100–600 | Flexible | NA | 16 | 24.2 | HAMD | 0 |
| North America | Placebo | (66) | 40.0 | 72.0 | 36 | ||||||||||
| Versiani et al. (1999) | 286 | Reboxetine | (145) | 42.3 | (12.2) | 67.4 | 46 | HAMD > = 18 and 50%>=increase | 4–8 | Flexible | NA | 6 | 29.6 | HAMD21 | 0 |
| Cross-Continental | Placebo | (111) | 43.4 | (11.6) | 79.3 | 46 | |||||||||
| Dekker et al. (2000) | 30 | Fluoxetine | (15) | 37.0 | (10.0) | 61.9 | 22 | HAM-D > = 14 | 20 | Fixed | NA | up to 16 | NA | NA | 0 |
| Europe | Placebo | (15) | 22 | ||||||||||||
| Rouillon et al. (2000) | 214 | Milnacipran | (104) | 44.6 | (10.0) | 68.2 | 48 | Major depressive episode according to DSM III-R criteria and a minimum score of 18 on HAM-D with the need to treat the recurrence | 100 | Fixed | NA | 26 | 25.1 | HAMD21 | 19 |
| Europe | Placebo | (110) | 46.1 | (10.2) | 66.3 | 48 | |||||||||
| Schmidt et al. (2000) | 311 | Fluoxetine | (189) | 42.0 | (11.2) | 63.9 | 25 | SCID-P MDD and CGI increase > =2 | 20 | Fixed | abrupt | 13 | NA | NA | 0 |
| North America | Placebo | (122) | 41.7 | (11.3) | 70.9 | 25 | |||||||||
| Dalery et al. (2001) | 185 | Tianeptine | (111) | 42.2 | NA | 64.9 | 66 | Either HAMD17 > = 15 (CGI > = 4) or clinical confirmation | 37.5 | Fixed | NA | 6 | 23.3 | HAMD17 | 0 |
| Europe | Placebo | (74) | 44.1 | 65.7 | 66 | ||||||||||
| Gilaberte et al. (2001) | 140 | Fluoxetine | (70) | 43.8 | NA | 78.6 | 52 | Meeting DSMIIR criteria for MDD and having HAMD17 score > = 18 and CGI score > = 4 | 20 | Fixed | NA | 32 | 24.0 | HAMD17 | 24 |
| Europe | Placebo | (70) | 44.4 | 78.6 | 52 | ||||||||||
| Hochstrasser et al. (2001) | 264 | Citalopram | (132) | 42.4 | (11.5) | 75.0 | 48 | MADRS > = 22 | 20/60 | Fixed | NA | 25 | 30.6 | MADRS | 16 |
| Europe | Placebo | (132) | 43.8 | (9.7) | 67.4 | 48 | |||||||||
| Thase et al. (2001) | 156 | Mirtazapine | (76) | 40.7 | (11.3) | 48.8 | 40 | clinical judgment | 30/45 | Fixed | abrupt | 12 | 22.7 | HAMD17 | 2,4 |
| North America | Placebo | (80) | 40.1 | (12.0) | 52.6 | 40 | |||||||||
| Klysner et al. (2002) | 121 | Citalopram | (60) | 75.0 | NA | 72.0 | 48 | MADRS > = 22 | 20/40 | Fixed | NA | 24 | 27.0 | MADRS | 16 |
| Europe | Placebo | (61) | 74.0 | NA | 82.0 | 48 | |||||||||
| Weihs et al. (2002) | 423 | Bupropion | (210) | 39.9 | NA | 64.0 | 44 | Determined by the investigator to be necessary for the treatment of depression | 300 | Fixed | NA | 8 | NA | NA | 0 |
| North America | Placebo | (213) | 39.4 | NA | 66.0 | 44 | |||||||||
| Wilson 2003 et al. (2003) | 113 | Sertraline | (56) | 76.8 | (7.0) | 75.4 | 48/100 | HAMD score of 13 or over as well as meeting DSM–III–R criteria for major depressive disorder as determined by a trained psychiatrist. | 50–150 | Flexible | NA | 8 | 20.4 | HAMD17 | 16–20 |
| Europe | Placebo | (57) | 76.6 | (6.6) | 66.1 | 100 | |||||||||
| Emslie et al. (2004) | 40 | Fluoxetine | (20) | 11.7 | (2.5) | 55.0 | 32 | CDRS-S score of >40 with a 2-week history of worsening of depressive symptoms or relapse in the opinion of the physician | 10–60 | Flexible | abrupt | 19 | 57.1 | CDRS-R | 0 |
| North America | Placebo | (20) | 13.5 | (2.4) | 45.0 | 32 | |||||||||
| Montgomery et al. (2004) | 235 | Venlafaxine | (112) | 43.5 | (11.2) | 67.0 | 52 | CGI-S > = 4 | 100–200 | Flexible | tapering | 26 | 25.2 | HAMD21 | 19 |
| Cross-Continental | Placebo | (123) | 43.8 | (11.0) | 71.0 | 52 | |||||||||
| Rapaport et al. (2004) | 274 | Escitalopram | (181) | 41.8 | (11.9) | 62.4 | 36 | MADRS > = 22 | 10/20 | Fixed | NA | 8 | 14.9 | MADRS | 0 |
| North America | Placebo | (93) | 42.9 | (11.6) | 60.2 | 36 | |||||||||
| Simon et al. (2004) | 318 | Venlafaxine | (161) | 41.0 | NA | 62.0 | 26 | Either a combination of the reappearance of MDD (DSM-IV criteria) and a CGI-S > = 4, two consecutive CGI-S > = 4, or a final CGI-S > = 4 | 75/150/225 | Fixed | tapering | 8 | 24.5 | HAMD21 | 0 |
| North America | Placebo | (157) | 43.0 | NA | 66.0 | 26 | |||||||||
| Fava et al. (2006) | 278 | Duloxetine | (136) | 44.8 | (11.9) | 77.5 | 26 | CGI-S increase of > =2 points compared with the randomization visit and meeting the MINI criteria for MDD for two consecutive visits | 60 | Fixed | tapering | 12 | 23.7 | HAMD17 | 0 |
| Cross-Continental | Placebo | (142) | 45.7 | (12.7) | 67.6 | 26 | |||||||||
| McGrath et al. (2006) | 262 | Fluoxetine | (131) | 38.2 | (10.9) | 55.3 | 26/52 | Using CGI-I but definition is not clear | 40/60 | Fixed | NA | 12 | 17.7 | HAMD17 | 0 |
| North America | Placebo | (131) | 52 | ||||||||||||
| Gorwood et al. (2007) | 305 | Escitalopram | (152) | 72.0 | NA | 79.0 | 24 | MADRS > = 22 or lack of efficacy by investigator | 10/20 | Fixed | tapering | 12 | 31.1 | MADRS | 0 |
| Europe | Placebo | (153) | 73.0 | NA | 78.0 | 24 | |||||||||
| Kocsis et al. (2007) | 267 | Venlafaxine | (132) | 42.6 | NA | 67.0 | 26/52 | HAMD > 12 and 2 consecutive visit, HAMD reduction <50% from acute phase | 75–300 | Flexible | tapering | 36 | 22.5 | NA | 26 |
| North America | Placebo | (135) | 42.0 | NA | 69.0 | 52 | |||||||||
| Cheung et al. (2008) | 22 | Sertraline | (13) | 16.3 | NA | 78.0 | 52 | Clinical judgment | 25–200 | Fixed | tapering | 36 | 20.7 | HAMD17 | 24 |
| North America | Placebo | (9) | 15.2 | NA | 77.0 | 52 | |||||||||
| Dobson et al. (2008) | 49 | Paroxetine | (28) | 38.9 | (10.0) | 78.2 | 52 | HAMD > = 14 or Psychiatric status rating > = 5 for two successive weeks. | 10–50 | Flexible | tapering | 16 | 20.9 | HAMD17 | 0 |
| North America | Placebo | (21) | 52 | ||||||||||||
| Emslie et al. (2008) | 102 | Fluoxetine | (50) | 11.5 | (2.8) | 36.3 | 24 | Either a one-time CDRS-R score > = 40 with worsening of depressive symptoms for at least 2 weeks, or a clinician determination that there was significant clinical deterioration suggesting that full relapse would be likely without altering treatment | 10–40 | Fixed | abrupt | 12 | 57.6 | CDRS-R | 0 |
| North America | Placebo | (52) | 24 | ||||||||||||
| Goodwin et al. (2009) | 339 | Agomelatine | (165) | 43.1 | (10.3) | 76.4 | 24 | HAMD17 > = 16, any withdrawal for lack of efficacy [clinical judgment based on HAMD and CGI], [suicide or suicide attempt] | 25/50 | Fixed | abrupt | 8/10 | 27.0 | HAMD17 | 0 |
| Cross-Continental | Placebo | (174) | 43.4 | (10.9) | 72.1 | 24 | |||||||||
| Perahia et al. (2009) | 288 | Duloxetine | (146) | 48.0 | (12.3) | 74.6 | 52 | (1) They had a CGI-S score≧4 and met DSM-IV criteria for MDD for at least 2 weeks; (2) they had 3 consecutive visits that met re-emergence criteria or 10 total re-emergence visits; or (3) they discontinued the study with a reason of “lack of efficacy” | 60 | Fixed | tapering | 34 | 23.1 | HAMD17 | 24 |
| Cross-Continental | Placebo | (142) | 47.1 | (12.8) | 68.5 | 52 | |||||||||
| Rickels et al. (2010) | 375 | Desvenlafaxine | (190) | 42.8 | (11.8) | 68.0 | 24 | HAM-D17 > = 16 or CGI-I > = 6 at any office visit during the DB treatment phase or as withdrawal from the study because of an unsatisfactory response to treatment as determined by the investigator | 200/400 | Fixed | tapering | 12 | 24.2 | HAMD17 | 0 |
| Cross-Continental | Placebo | (185) | 42.7 | (12.3) | 67.0 | 24 | |||||||||
| Segal et al. (2010) | 58 | Various | (28) | 41.9 | (11.6) | 67.0 | 78 | relapse of DSM-IV MDD episode | Flexible | tapering | 19.4 | HAMD17 | 28 | ||
| North America | Placebo | (30) | 45.8 | (11.4) | 71.0 | 78 | |||||||||
| Boulenger et al. (2012) | 400 | Vortioxetine | (206) | 45.1 | (12.1) | 62.5 | 64 | MADRS total score > = 22 or an insufficient therapeutic response | 5/10 | Fixed | abrupt | 12 | 32.3 | MADRS | 0 |
| Cross-Continental | Placebo | (194) | 44.8 | (12.4) | 63.7 | 64 | |||||||||
| Goodwin et al. (2013) | 367 | Agomelatine | (187) | 46.0 | (10.1) | 79.4 | 24/42 | HAMD17 > = 16, any withdrawal for lack of efficacy [clinical judgment based on HAMD and CGI], [suicide or suicide attempt] | 25 | Fixed | NA | 8 | 26.3 | HAMD17 | 0 |
| Europe | Placebo | (180) | 45.3 | (10.5) | 76.5 | 42 | |||||||||
| Rosenthal et al. (2013) | 548 | Desvenlafaxine | (272) | 45.3 | (13.0) | 71.7 | 26 | HAMD17 total score >16, discontinuation for unsatisfactory response, hospitalization for depression, suicide attempt, or suicide (any 1 or more) | 50 | Fixed | tapering | 20 | 24.2 | HAMD17 | 12 |
| Cross-Continental | Placebo | (276) | 46.6 | (13.0) | 71.0 | 26 | |||||||||
| Shiovitz et al. (2014) | 348 | Levomilnacipran | (235) | 44.7 | (12.7) | 54.5 | 24 | (1) MADRS > = 22 at two consecutive visits; (2) increase > =2 in the CGI-I relative to the double-blind baseline score at two consecutive visits; (3) discontinuation from the score of 4 or greater | 40/80/120 | Fixed | tapering | 12 | 30.7 | MADRS | 0 |
| North America | Placebo | (113) | 42.6 | (12.0) | 59.7 | 24 | |||||||||
CGI Clinical Global Impression Scale (I improved; S severity), DSM APA Diagnostic & Statistical Manual, Dx diagnosis, HAMD Hamilton Depression Rating Scale (17-, 21-, or 24-item versions), MADRS Montgomergy-Åsberg Depression Rating Scale, MDD major depressive disorder, NA information not available, wks weeks, yr years.