AJP in Advnce. Published My 2, 2011 (doi: 10.1176/ppi.jp.2011.10111645) Article Combining Medictions to Enhnce Depression Outcomes (CO-MED): Acute nd Long-Term Outcomes of Single-Blind Rndomized Study A. John Rush, M.D. Mdhukr H. Trivedi, M.D. Jonthn W. Stewrt, M.D. Andrew A. Nierenberg, M.D. Murizio Fv, M.D. Benji T. Kurin, M.D. Dine Wrden, Ph.D. Dvid W. Morris, Ph.D. Jmes F. Luther, M.A. Mustf M. Husin, M.D. In A. Cook, M.D. Richrd C. Shelton, M.D. Ir M. Lesser, M.D. Susn G. Kornstein, M.D. Stephen R. Wisniewski, Ph.D. Objective: Two ntidepressnt mediction combintions were compred with selective serotonin reuptke inhibitor monotherpy to determine whether either combintion produced higher remission rte in first-step cute-phse (12 weeks) nd long-term (7 months) tretment. Method: The single-blind, prospective, rndomized tril enrolled 665 outptients t six primry nd nine psychitric cre sites. Prticipnts hd t lest modertely severe nonpsychotic chronic nd/ or recurrent mjor depressive disorder. Escitloprm (up to 20 mg/dy) plus plcebo, sustined-relese bupropion (up to 400 mg/dy) plus escitloprm (up to 20 mg/dy), or extended-relese venlfxine (up to 300 mg/dy) plus mirtzpine (up to 45 mg/dy) ws delivered (1:1:1 rtio) by using mesurement-bsed cre. The primry outcome ws remission, defined s rtings of less thn 8 nd less thn 6 on the lst two consecutive pplictions of the 16-item Quick Inventory of Depressive Symptomtology Self-Report. Secondry outcomes included side effect burden, dverse events, qulity of life, functioning, nd ttrition. Results: Remission nd response rtes nd most secondry outcomes were not different mong tretment groups t 12 weeks. The remission rtes were 38.8% for escitloprm-plcebo, 38.9% for bupropion-escitloprm, nd 37.7% for venlfxine-mirtzpine, nd the response rtes were 51.6% 52.4%. The men number of worsening dverse events ws higher for venlfxine-mirtzpine (5.7) thn for escitloprm-plcebo (4.7). At 7 months, remission rtes (41.8% 46.6% ), response rtes (57.4% 59.4% ), nd most secondry outcomes were not significntly different. Conclusions: Neither mediction combintion outperformed monotherpy. The combintion of extended-relese venlfxine plus mirtzpine my hve greter risk of dverse events. (Am J Psychitry Rush et l.; AiA:1 13) Mjor depressive disorder is serious, disbling, life-shortening illness with high lifetime risk: 7% 12% for men nd 20% 25% for women (1). It is often recurrent, episodes frequently lst more thn 2 yers (i.e., re chronic) (2), nd interepisode recovery is often incomplete (3). Chronic episodes nd recurrent courses re ssocited with worse prognoses nd re more likely to need longerterm tretment (4 7). Remission is the im of tretment (8) becuse ptients whose depression hs remitted hve better functioning nd better prognosis thn those without remission (5). Antidepressnt medictions, when used s monotherpies in plcebo-controlled registrtion trils, typiclly result in 30% 35% remission rtes (8). Lower remission rtes (25% 30%) re reported for ptients with more chronic depressions (5, 6). Could remission rtes be incresed with combintion of two ntidepressnt medictions used together s ini- til tretment? Other brnches of medicine often employ combintion tretments t the outset of chronic illness, especilly for the more severely ill (9, 10). In depression tretment, when single ntidepressnt mediction is not effective (7, 11), second is often dded to the first, with some evidence for efficcy (12). It lso ppers tht some ntidepressnt medictions work for some ptients but not for others. A combintion of two medictions might therefore increse the spectrum of ptients who could benefit from the combintion (13). Furthermore, n unexpected synergy between medictions might produce rpid onset of benefit, so tht fewer ptients would drop out of tretment, which, in turn, might enhnce remission rtes. From phrmcologicl perspective, combintion might ffect wider rnge of neurotrnsmitter or neuromodultor systems, which would enhnce efficcy for some ptients (14 16). Finlly, clinicl experience nd few smll rndomized, short-term trils (13, 17, 18) sug- AJP in Advnce jp.psychitryonline.org 1 Copyright 2011 Americn Psychitric Assocition. All rights reserved.
Combining Antidepressnts gest tht some combintions cn be more effective thn monotherpy. On the other hnd, tretment guidelines do not recommend such n pproch s first tretment step, nd the risk of serious dverse events or intrusive side effects hs not been fully evluted. Thus, combining ntidepressnts s first-step tretment for depression needs proper evlution. The Combining Medictions to Enhnce Depression Outcomes (CO-MED) tril ws designed s proof-ofconcept study to determine whether either of two different ntidepressnt mediction combintions would produce higher remission rte t 12 weeks nd, secondrily, fter 7 months thn monotherpy with selective serotonin reuptke inhibitor (SSRI) s first-step tretment in outptients with chronic or recurrent mjor depression. We lso compred the tretment effects on ptient retention, side effect burden, nd qulity of life. Method Study Overview CO-MED ws 7-month single-blind, rndomized, plcebocontrolled tril tht compred the efficcy of ech of two mediction combintions with escitloprm plus plcebo in 1:1:1 rtio s first-step tretment, including cute-phse (12 weeks) nd long-term continution (totl 7 months) tretment. We plnned study group of 660 outptients with nonpsychotic mjor depression from six primry nd nine psychitric cre sites to llow detection of roughly 15% difference in remission rtes between ech combintion nd escitloprm-plcebo (with n expected remission rte of 35%). This difference ws viewed s sufficiently lrge to ffect prctice since it pproximtes the benefit of single ntidepressnt mediction over plcebo in successful ntidepressnt registrtion trils (8). Site Selection Clinicl sites were selected on the bsis of our prior experience nd their performnce in the Sequenced Tretment Alterntives to Relieve Depression tril to ensure 1) dequte ptient flow, 2) committed dministrtive support, 3) dequte minority representtion, nd 4) dequte representtion of both primry nd psychitric cre sites. Recruitment Potentil prticipnts were screened t ech clinicl site with ech site s stndrd procedure (vrible cross sites). Most sites used two to nine questions from the Ptient Helth Questionnire (19, 20). Ptients identified by screening sw their study clinicins nd clinicl reserch coordintor to determine study eligibility following written informed consent. Prticipnts Brod inclusion nd miniml exclusion criteri ensured resonbly representtive prticipnt group. The outptient enrollees were 18 75 yers old nd met the DSM-IV-TR (21) criteri for either recurrent or chronic (current episode lsting t lest 2 yers) mjor depression ccording to clinicl interview nd confirmed with DSM-IV-bsed symptom checklist completed by the clinicl reserch coordintor. Eligible prticipnts hd to hve n index episode lsting t lest 2 months nd hd to score t lest 16 on the 17-Item Hmilton Depression Rting Scle (HAM-D) (22). Those with ny psychotic illness or bipolr disorder nd those in need of hospitliztion were ineligible. (For complete list of exclusion criteri, see http://clinicltrils.gov/ ct2/show/nct00590863.) The study protocol nd ll consent nd study procedures were pproved by the institutionl review bords t the ntionl coordinting center (University of Texs Southwestern Medicl Center t Dlls), the University of Pittsburgh dt coordinting center, nd ech prticipting regionl center nd relevnt clinicl site. Bseline Dt Sociodemogrphic nd illness fetures were recorded t bseline. The nxiety subscle of the HAM-D ws used to estblish the presence of nxious fetures t bseline (23). This nxiety/ somtiztion fctor, derived from fctor nlysis of the HAM-D conducted by Clery nd Guy (24), includes six items from the originl 17-item version: item 10 (nxiety, psychic), item 11 (nxiety, somtic), item 12 (somtic symptoms, gstrointestinl), item 13 (somtic symptoms, generl), item 15 (hypochondrisis), nd item 17 (insight). A HAM-D nxiety/somtiztion fctor score of 7 or higher indicted nxiety. The HAM-D dministered t bseline by reserch outcome ssessors (not locted t ny clinicl site) ws used to define nxious fetures. The self-report Psychitric Dignostic Screening Questionnire (25) ws used to estblish the presence of current xis I disorders. The Self-Administered Comorbidity Questionnire (26) estblished the presence, severity, nd functionl impct of rnge of common concurrent generl medicl conditions. Antidepressnt Tretment A 12-week study period ws chosen for the primry nlysis to provide sufficient time for mximl dosing (if needed) nd to llow most cses of depression tht could remit to do so (27). Tretment visits were plnned t bseline nd weeks 1, 2, 4, 6, 8, 10, 12, 16, 20, 24, nd 28. No wshout ws required, but clinicins could choose wshout period if they thought it to be cliniclly dvisble. Dose djustments were bsed on mesurement-bsed cre following n opertions mnul to provide personlly tilored but vigorous dosing (28). Dose djustments were bsed on the score on the 16-item Quick Inventory of Depressive Symptomtology Clinicin-Rted (QIDS-C) (29), which ws extrcted from the 30- item Inventory of Depressive Symptomtology Clinicin-Rted (IDS-C) (30), nd the score on the Frequency, Intensity, nd Burden of Side Effects Rting Scle (31) obtined t ech visit. Tretment ws rndomly ssigned, strtified by clinicl site ccording to web-bsed rndomiztion system (32). Rndom block sizes of three nd six were used to minimize the probbility of identifying the next tretment ssignment. Dosing schedules were bsed on prior reports (33 35). Doses were incresed only in the context of cceptble side effects. Prticipnts could exit the study if uncceptble or intolerble side effects occurred nd could not be resolved with dose reduction or mediction tretment of the side effects. Escitloprm plus plcebo (monotherpy). Escitloprm tretment ws begun t 10 mg/dy (one pill) nd could be incresed to 20 mg/dy (two pills) t 4 weeks if the score on the QIDS-C ws higher thn 5 (if side effects llowed). Pill plcebo ws strted t week 2, with the option to increse it to two pills t week 4 if the QIDS-C score ws higher thn 5 (if side effects llowed). Bupropion plus escitloprm. The dose of sustined-relese bupropion ws 150 mg/dy initilly nd ws incresed to 300 mg/dy t the week 1 visit. Escitloprm ws begun t 10 mg/ dy t the week 2 visit. At week 4, the bupropion dose ws rised to 400 mg/dy (200 mg b.i.d.) nd/or the escitloprm dose ws rised to 20 mg/dy if the score on the QIDS-C ws higher thn 5 (if side effects llowed). At week 6 nd beyond, doses could be 2 jp.psychitryonline.org AJP in Advnce
Rush, Trivedi, Stewrt, Et Al. Outcomes t 12 Weeks During the first 12 weeks, the prticipnts were in tretment for n verge of 10 weeks (Tble 4). Of the 665 princresed up to mximum of 400 mg/dy (200 mg b.i.d.) for bupropion nd 20 mg/dy for escitloprm if the QIDS-C score ws bove 5 (if side effects llowed). Venlfxine plus mirtzpine. Tretment with extendedrelese venlfxine ws begun t 37.5 mg/dy for 3 dys nd then rised to 75 mg/dy. At week 1, the dose ws rised to 150 mg/dy. At week 2, if the score on the QIDS-C ws bove 5, mirtzpine ws dded t dose of 15 mg/dy. At week 4, if the QIDS-C score ws bove 5, the venlfxine dose ws rised to 225 mg/dy nd/ or the mirtzpine dose ws incresed to 30 mg/dy. At week 6, if the QIDS-C score ws higher thn 5, the mirtzpine dose could be rised to 45 mg/dy, the mximum dose. At week 8, if the score ws bove 5, the venlfxine dose could be rised to 300 mg/dy, the mximum llowed. Mediction Blinding The first mediction given in ech tretment group ws open lbel (both prticipnt nd study personnel were unblinded), while ech second mediction ws given in single-blind fshion (prticipnt only) to ensure tht the prticipnts ll took two types of pills. Specificlly, in the escitloprm-plcebo cell, plcebo dministrtion ws blinded. For the bupropion-escitloprm combintion, escitloprm ws blinded. For venlfxine-mirtzpine, mirtzpine ws blinded. The prticipnts remined blinded to the second mediction throughout the 7-month study. The reserch coordintors nd physicins were not blinded, to mximize sfety nd fcilitte informed flexible dosing decisions. Concurrent Tretments Only protocol ntidepressnt medictions were llowed. Tretments with possible ntidepressnt effects were proscribed, s were nxiolytics, sedtive-hypnotics, nd depression-trgeted, empiriclly vlidted psychotherpies for depression. Other therpies (e.g., supportive, couples, occuptionl) were llowed, s were medictions for ny generl medicl condition. Given the inhibition of the 2D6 isoenzyme by sustined-relese bupropion, we lerted clinicins bout nonprotocol medictions (e.g., type 1C ntirrhythmics, bet-blockers) for which serum or dose djustments might be needed. Medictions to tret ntidepressnt mediction side effects were llowed; dministrtion ws bsed on clinicin judgment. Reserch Outcomes Outcome mesures were ssessed t bseline nd ll tretment visits. The primry outcome, symptom remission, ws bsed on the score on the 16-item Quick Inventory of Depressive Symptomtology Self-Report (QIDS-SR) t 12 weeks (29). The designtion of remission ws bsed on the lst two consecutive mesurements during the 12-week cute-phse tril to ensure tht single good week ws not flsely signling remission. At lest one of these rtings hd to be less thn 6, while the other hd to be less thn 8. If prticipnts exited before 12 weeks, their lst two consecutive scores were used to determine remission. Those without two postbseline mesures were considered not to hve remission. Physicins were dvised tht prticipnts could exit the study if they hd received mximlly tolerted dose(s) for 4 or more weeks by week 8 without obtining t lest 30% reduction from the bseline score on the QIDS-C. They could enter continution tretment (weeks 12 28) if they hd received n cceptble benefit (defined s QIDS-C score of 9 or less by week 12) or if they reched score of 10 13 nd the clinicin nd prticipnt judged the benefit to be substntil enough to recommend tretment continution. Thus, virtully ll prticipnts entering the continution phse hd t lest 40% reduction in the QIDS-C score. If prticipnt exited the study t ny time, study exit form ws completed. Clinicl reserch coordintors ttempted to contct ll prticipnts who did not come for finl exit visit. Secondry outcomes included ttrition, nxiety s reflected in the score on the nxiety subscle of the IDS-C (30), functioning s mesured by the Work nd Socil Adjustment Scle (36), qulity of life s mesured by the Qulity of Life Inventory (37), side effect burden s mesured by the Frequency, Intensity, nd Burden of Side Effects Rting Scle (31), nd specific side effects s mesured by the Systemtic Assessment for Tretment Emergent Events Systemtic Inquiry (38). Mnic symptoms were ssessed by using the Altmn Self-Rting Mni Scle (39), nd cognitive nd executive dysfunction ws ssessed by mens of the Cognitive nd Physicl Functioning Questionnire (40). Sttisticl Anlyses Descriptive sttistics, including mesures of centrl tendency nd dispersion, were computed for continuous dt. Frequency distributions were estimted for ctegoricl dt. Outcome nlyses were conducted with the full group, on the bsis of the intention-to-tret principle. Ech combintion therpy ws compred to the monotherpy. To control for the overll type I error rte, type I error rte of 0.025 ws plnned for the comprison of ech combintion tretment with monotherpy. The nlytic pproch for the two comprisons ws identicl. A chi-squre test ws used to compre the remission rtes cross the tretment groups. Fisher s exct test ws used when the expected cell frequencies were less thn 5. For binry outcomes (e.g., remission), bivrite logistic regression models were fit to estimte the effect of tretment on outcome. Multivrible logistic regression models were then fit to control for the effect of regionl center nd bseline chrcteristics tht were not blnced cross tretment groups. A similr pproch ws used for discrete outcomes with more thn two levels, except polytomous logistic regression model ws used. For continuous outcomes, t test ws used to compre the mens when distributions were norml, nd the nonprmetric Kruskl-Wllis test ws used when distributions were nonnorml. Liner regression models were used to compre the mens fter controlling for regionl center nd bseline chrcteristics not blnced cross tretment groups. A generl liner model with negtive binomil distribution nd log link ws estimted for outcomes with severely nonnorml distributions (the lst number of worsening dverse events indicted by the Systemtic Assessment for Tretment Emergent Effects nd the score on the IDS-C nxiety subscle). Results Bseline Chrcteristics Figure 1 contins chrt specifying how the study group ws formed. From Mrch 2008 through Februry 2009, the study enrolled 665 prticipnts. They were modertely to severely ill, s indicted by men HAM-D score of 23.8 (SD=4.8) nd score on the QIDS-SR of 15.5 (SD=4.3). About hlf of the prticipnts were unemployed, nd twothirds were femle (Tble 1). Over three-qurters hd recurrent mjor depression (Tble 2). More thn one-hlf of the prticipnts were in chronic current mjor depressive episode. About one-third hd both recurrent nd chronic depression. Almost one in 10 hd mde prior suicide ttempt, nd for over 40% the illness hd begun before ge 18. Most (75%) hd nxious fetures. Concurrent comorbid xis I nd xis III disorders were common (Tble 3). AJP in Advnce jp.psychitryonline.org 3
Combining Antidepressnts FIGURE 1. Recruitment nd Tretment of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Offered consent (N=832) Screened (N=731) Declined consent (N=101) Rndomiztion (N=7) Mediction combintions (N=12) Too much time (N=26) Insufficient compenstion (N=2) Fer of side effects (N=7) Other (N=47) Ineligible (N=66; some were excluded for multiple resons) Psychotic mjor depression (N=36) Screening score on 17-item HAM-D <16 (N=26) Excluded medictions (N=18) Psychotic history (N=13) Excluded tretment (N=5) Off-site tretment (N=5) Psychotic symptoms (N=5) Seizures (N=5) Pregnncy (N=2) Tking study mediction (N=2) Eting disorder (N=1) Substnce use disorder (N=1) Intolernt of study mediction (N=1) Hospitliztion (N=1) Tking nticonvulsnt (N=1) Rndomly ssigned to tretment (N=665) Escitloprm + plcebo (monotherpy) Acute Phse: Enrolled (N=224) Exited (N=43) Completed (N=181) Sustined-relese bupropion + escitloprm Acute Phse: Enrolled (N=221) Exited (N=59) Completed (N=162) Extended-relese venlfxine + mirtzpine Acute Phse: Enrolled (N=220) Exited (N=54) Completed (N=166) Continution Phse: Declined (N=12) Enrolled (N=169) Exited (N=18) Completed (N=151) Continution Phse: Declined (N=11) Enrolled (N=151) Exited (N=18) Completed (N=133) Continution Phse: Declined (N=7) Enrolled (N=159) Exited (N=22) Completed (N=137) TABLE 1. Bseline Sociodemogrphic Chrcteristics of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Chrcteristic Totl (N=665) Monotherpy: Escitloprm Plus Plcebo (N=224) Ptient Group Sustined-Relese Bupropion Plus Escitloprm (N=221) Extended-Relese Venlfxine Plus Mirtzpine (N=220) Comprison With Monotherpy b Bupropion Plus Escitloprm Venlfxine Plus Mirtzpine N % N % N % N % p p Sex 0.43 <0.05 c Mle 213 32.0 81 36.2 72 32.6 60 27.3 Femle 452 68.0 143 63.8 149 67.4 160 72.7 Rce 0.90 0.84 White 431 67.0 147 67.7 142 67.0 142 66.4 Blck 174 27.1 56 25.8 58 27.4 60 28.0 Other 38 5.9 14 6.5 12 5.7 12 5.6 Hispnic 101 15.2 37 16.5 36 16.3 28 12.7 0.95 0.26 Employed 331 49.8 99 44.2 119 53.8 113 51.4 <0.05 c 0.14 Men SD Men SD Men SD Men SD p p Age (yers) 42.7 13.0 43.6 13.1 42.4 13.5 42.1 12.4 0.34 0.22 Eduction (yers) 13.8 3.0 13.8 3.2 13.8 2.6 13.7 3.1 0.85 0.82 Monthly household income (dollrs) 2,678 5,353 2,449 3,696 2,828 5,037 2,759 6,832 0.81 0.28 For some vribles, dt were not vilble for ll subjects. b Chi-squre nlysis for ctegoricl dt nd t tests for continuous dt. c Significntly different from monotherpy. 4 jp.psychitryonline.org AJP in Advnce
Rush, Trivedi, Stewrt, Et Al. TABLE 2. Bseline Clinicl Chrcteristics of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Chrcteristic All (N=665) Monotherpy: Escitloprm Plus Plcebo (N=224) Ptient Group Sustined-Relese Bupropion Plus Escitloprm (N=221) Extended-Relese Venlfxine Plus Mirtzpine (N=220) Comprison With Monotherpy b Bupropion Plus Escitloprm Venlfxine Plus Mirtzpine N % N % N % N % p p First episode before ge 18 296 44.6 96 43.0 95 43.0 105 47.9 0.99 0.31 Recurrent depression 517 78.0 171 76.7 174 78.7 172 78.5 0.61 0.64 Ever ttempted suicide 59 9.2 14 6.5 23 10.7 22 10.3 0.13 0.16 Abused before ge 18 Emotionlly 261 39.3 94 42.2 88 39.8 79 35.9 0.62 0.18 Physiclly 131 19.7 45 20.2 42 19.0 44 20.0 0.76 0.97 Sexully 145 21.9 43 19.3 50 22.6 52 23.7 0.39 0.26 Chronic depression (index episode durtion 2 yers) 368 55.5 121 54.3 121 54.8 126 57.5 0.92 0.49 Chronic or recurrent depression 0.81 0.53 Chronic only 146 22.0 52 23.3 47 21.3 47 21.5 Recurrent only 295 44.5 102 45.7 100 45.2 93 42.5 Both 222 33.5 69 30.9 74 33.5 79 36.1 Anxious fetures (HAM-D) 497 74.7 156 69.6 177 80.1 164 74.5 0.02 c 0.25 Atypicl fetures (IDS-C) 103 15.5 33 14.7 38 17.2 32 14.5 0.48 0.96 Melncholic fetures (IDS-C) 124 20.5 42 20.5 36 18.0 46 23.0 0.53 0.54 Men SD Men SD Men SD Men SD p p Age t first episode (yers) 24.0 14.1 24.4 14.4 23.9 13.7 23.7 14.2 0.85 0.53 Yers since first episode 18.7 13.6 19.3 14.4 18.5 13.4 18.4 13.1 0.69 0.76 Index episode durtion (months) 61.7 104.8 66.4 114.4 58.1 100.8 60.6 98.5 0.91 0.77 Scores on clinicl rtings HAM-D 23.8 4.8 23.4 4.9 23.8 4.6 24.3 5.0 0.34 <0.05 c IDS-C 38.0 9.1 37.0 8.8 37.8 9.2 39.3 9.3 0.39 0.02 c QIDS-C 15.8 3.4 15.6 3.4 15.7 3.5 16.1 3.5 0.72 0.13 QIDS-SR 15.5 4.3 15.2 4.0 15.3 4.6 15.9 4.2 0.77 0.10 Altmn Self-Rting Mni Scle (39) 1.5 2.3 1.6 2.4 1.6 2.2 1.3 2.2 0.79 0.21 Cognitive nd Physicl Functioning Questionnire (40) 27.6 5.9 27.4 5.7 27.7 6.1 27.8 5.8 0.62 0.39 Qulity of Life Inventory (37) 1.2 1.9 1.2 1.9 1.1 1.9 1.3 1.9 0.85 0.41 Work nd Socil Adjustment Scle (36) 26.9 8.8 26.2 8.8 26.7 9.2 27.9 8.4 0.50 0.04 c For some vribles, dt were not vilble for ll subjects. HAM-D, 17-item Hmilton Depression Rting Scle (22). IDS-C, 30-item Inventory of Depressive Symptomtology Clinicin-Rted (30). QIDS-C, 16-item Quick Inventory of Depressive Symptomtology Clinicin-Rted (29). QIDS-SR, 16-item Quick Inventory of Depressive Symptomtology Self-Report (29). b Chi-squre nlysis for ctegoricl dt nd t tests for continuous dt. c Significntly different from monotherpy. ticipnts, 86.0% (N=572) completed t lest 4 weeks of tretment. Overll, 78.3% (N=521) completed week 8 nd 72.2% (N=480) completed t lest 12 weeks of tretment. For the escitloprm-plcebo group, the escitloprm dose ws close to the mximum trget dose of 20 mg/ dy. For bupropion-escitloprm, the comprble men exit escitloprm dose during cute tretment ws significntly lower, t 12.5 mg/dy (SD=8.3) (c 2 =31.15, df=1, p<0.0001). Also of note, while the venlfxine dose ws close to 200 mg/dy by 12 weeks, the men mirtzpine dose ws only 20.0 mg/dy (SD=15.7) (Tble 4). As shown in Tble 5 nd Figure 2, the tretment groups did not differ in either remission or response rtes, nor did they differ in the percentge of chnge in QIDS-SR score (bseline to exit or 12 weeks) or in effects on qulity of life. The venlfxine-mirtzpine combintion ws ssocited with more side effect burden thn escitloprm-plcebo. Ptients tking venlfxine-mirtzpine hd more dverse symptoms (er ches, blurred vision, irritbility, etc.) present t bseline tht becme worse during tretment, s mesured by the Systemtic Assessment for Tretment Emergent Events (men number of effects=5.7, SD=5.8), thn the monotherpy group (Tble 5). Outcomes t 7 Months Overll, while 72.2% of the 665 prticipnts (N=480) completed t lest 12 weeks of tretment, 65.6% completed 16 weeks or more, 61.4% completed 20 weeks, 55.6% AJP in Advnce jp.psychitryonline.org 5
Combining Antidepressnts TABLE 3. Bseline Axis I nd III Disorders of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Ptient Group Comprison With Monotherpy b Illness Vrible All (N=665) Monotherpy: Escitloprm Plus Plcebo (N=224) Sustined-Relese Bupropion Plus Escitloprm (N=221) Extended-Relese Venlfxine Plus Mirtzpine (N=220) Bupropion Plus Escitloprm Venlfxine Plus Mirtzpine N % N % N % N % p p Comorbid xis I disorders c Agorphobi 69 10.4 20 8.9 28 12.7 21 9.5 0.21 0.83 Alcohol buse 67 10.1 23 10.3 24 10.9 20 9.1 0.83 0.68 Bulimi 78 11.7 27 12.1 22 10.0 29 13.2 0.48 0.73 Drug buse 35 5.3 15 6.7 12 5.4 8 3.6 0.58 0.15 Generlized nxiety 131 19.7 39 17.4 43 19.5 49 22.3 0.58 0.20 Hypochondrisis 29 4.4 9 4.0 12 5.4 8 3.6 0.49 0.84 Obsessive-compulsive 79 11.9 27 12.1 25 11.3 27 12.3 0.81 0.95 Pnic 65 9.8 16 7.1 25 11.3 24 10.9 0.13 0.17 Posttrumtic stress 81 12.2 29 12.9 32 14.5 20 9.1 0.64 0.20 Socil phobi 178 26.8 60 26.8 59 26.7 59 26.8 0.99 1.00 Somtoform 21 3.2 7 3.1 7 3.2 7 3.2 0.98 0.98 Number of comorbid xis I disorders 0.27 0.66 0 296 44.6 107 47.8 85 38.6 104 47.3 1 159 23.9 51 22.8 67 30.5 41 18.6 2 92 13.9 27 12.1 29 13.2 36 16.4 3 50 7.5 16 7.1 19 8.6 15 6.8 4 67 10.1 23 10.3 20 9.1 24 10.9 Number of comorbid xis III disorders d 0.25 0.86 0 161 24.2 55 24.7 59 26.7 47 21.4 1 198 29.8 66 29.6 67 30.3 65 29.5 2 154 23.2 54 24.2 43 19.5 57 25.9 3 77 11.6 20 9.0 32 14.5 25 11.4 4 74 11.1 28 12.6 20 9.0 26 11.8 Men SD Men SD Men SD Men SD p p Axis III comorbidity score d 3.4 3.5 3.5 3.7 3.0 3.1 3.6 3.8 0.41 0.56 For some vribles, dt were not vilble for ll subjects. b Chi-squre nlysis for ctegoricl dt nd t tests for continuous dt. c From the Psychitric Dignostic Screening Questionnire (25). d From the Self-Administered Comorbidity Questionnire (26). completed 24 weeks, nd 58.0% completed 28 weeks. Attrition rtes over the 7-month period did not differ mong tretment groups. Averge drug doses were bsiclly unchnged from 12 weeks to 7 months of tretment, regrdless of tretment group (Tble 4). At 7 months (or study exit, if erlier), the three groups were not different in terms of remission rte (rnge: 41.8% 46.6%), response rte (rnge: 57.4% 59.4%), or ttrition rte. Nor did the groups differ in the percentge of chnge in QIDS-SR (bseline to exit or 7 months), qulity of life, or work nd socil djustment. Tble 6 compres the side effect frequency, intensity, nd burden in the escitloprm monotherpy group nd ech of the combintion groups. Overll, there were modestly more side effects with escitloprm-bupropion thn with escitloprm-plcebo in both the 12-week nd 7-month comprisons. On the other hnd, the venlfxine-mirtzpine group hd greter side effect frequency nd intensity t 12 weeks nd greter side effect frequency, intensity, nd burden t 7 months s compred to escitloprm-plcebo. Discussion The study hs four key findings: 1) remission nd response rtes were not different t 12 weeks, 2) remission nd response rtes were not different t 7 months, 3) the effects of the three tretments on qulity of life nd on work nd socil djustment were not different, nd 4) extendedrelese venlfxine plus mirtzpine ws ssocited with greter side effect burden t 12 weeks nd 7 months thn escitloprm plus plcebo nd higher number of worsening dverse events thn escitloprm plus plcebo t 7 months. We found no clinicl dvntge over escitloprm-plcebo from either combintion of ntidepressnt medictions in terms of either remission or response rtes t either 12 weeks or 7 months. The remission rtes pproximted those expected on the bsis of monotherpy 6 jp.psychitryonline.org AJP in Advnce
Rush, Trivedi, Stewrt, Et Al. TABLE 4. Tretment Chrcteristics t 12 Weeks nd 7 Months of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Chrcteristic All (N=665) Monotherpy: Escitloprm Plus Plcebo (N=224) Ptient Group b Sustined-Relese Bupropion Plus Escitloprm (N=221) Extended-Relese Venlfxine Plus Mirtzpine (N=220) Comprison With Monotherpy c Bupropion Plus Escitloprm Venlfxine Plus Mirtzpine 12 weeks N % N % N % N % p p Weeks in tretment <4 93 14.0 30 13.5 33 14.9 30 13.7 0.66 0.94 <8 144 21.7 43 19.3 55 24.9 46 21.0 0.16 0.66 Men SD Men SD Men SD Men SD p p Weeks in tretment 9.9 3.9 10.0 3.9 9.6 4.0 10.0 3.8 0.13 0.88 Number of postbseline visits 5.3 2.2 5.4 2.1 5.1 2.2 5.3 2.2 0.21 0.88 Mximum open dose (mg/dy) 17.6 4.5 324.0 80.4 207.6 69.2 Lst open dose (mg/dy) 16.8 5.3 287.7 121.2 192.3 82.2 Mximum blinded dose (mg/dy) d 1.4 0.7 14.0 7.2 25.3 32.0 Lst blinded dose (mg/dy) d 1.3 0.7 12.5 8.3 20.0 15.7 7 months N % N % N % N % p p Weeks in tretment <12 185 27.8 56 25.1 72 32.6 57 26.0 0.09 0.83 Men SD Men SD Men SD Men SD p p Weeks in tretment 19.9 10.5 20.5 10.3 19.1 10.8 20.1 10.4 0.38 0.77 Number of postbseline visits 7.7 3.7 7.9 3.6 7.4 3.8 7.7 3.7 0.26 0.81 Mximum open dose (mg/dy) 17.9 4.4 328.5 81.7 217.3 73.3 Lst open dose (mg/dy) 15.6 6.9 271.0 136.8 177.6 94.0 Mximum blinded dose (mg/dy) d 1.5 0.7 14.2 7.3 26.7 32.2 Lst blinded dose (mg/dy) d 0.7 0.9 11.5 8.6 18.0 16.4 For some vribles, dt were not vilble for ll subjects. b Medictions listed first re referred to s open, while those listed second re referred to s blinded. c Chi-squre nlysis for ctegoricl dt nd t tests for continuous dt. d For the subgroup receiving escitloprm plus plcebo, the unit of mesurement is pills. studies of chronic depression (5, 6). Both combintion tretments hd more side effects (in terms of frequency, intensity, or burden) thn escitloprm-plcebo in both the cute nd continution phses. Attrition rtes, however, were not different cross the three tretment groups in either phse of tretment. The venlfxine-mirtzpine group ws t prticulrly greter risk for side effects. In fct, it hd significntly greter worsening of side effects thn escitloprm-plcebo despite the fct tht the mirtzpine dose ws not high (bout 20 mg/dy). Prior reports hve suggested tht the response to either mediction combintion would exceed the effects of monotherpy. An open study of 49 ptients (41) given escitloprm (up to 40 mg/dy) plus sustined-relese bupropion (400 450 mg/dy) found 63% remission rte t week 8. Blier et l. (18) compred mirtzpine, proxetine, nd the combintion in 6-week double-blind, rndomized, controlled tril conducted t two reserch clinics with cliniclly referred ptients nd symptomtic volunteers (N=61). Remission rtes t 6 weeks were 19% for mirtzpine, 26% for proxetine, nd 43% for the combintion. Most of these ptients hd melncholic symptom fetures nd either hd nonrecurrent depression or hd n index episode lsting less thn 1 yer. Drug doses included up to 45 mg/dy of mirtzpine nd proxetine mounts tht could exceed 30 mg/dy (verge finl or exit doses not reported). In recent, lrger 6-week double-blind cute rndomized, controlled tril, Blier et l. (13) compred fluoxetine (20 mg/dy) with mirtzpine (30 mg/dy) in combintion with fluoxetine (20 mg/dy), extended-relese venlfxine (225 mg/dy), or sustined-relese bupropion (150 mg/dy). Ech combintion hd remission rte (46% 58%) tht exceeded tht of fluoxetine lone (25%). In the study, 76% of the prticipnts met melncholi criteri, 63% hd recurrent mjor depression, nd 61% hd n index episode longer thn 1 yer. It is interesting tht the response rtes were not significntly different (54% for fluoxetine, 68% for mirtzpine-fluoxetine, 73% for mirtzpine-venlfxine, 65% for mirtzpine-bupropion). Of note, this 6-week study my hve been too brief to llow the full benefit of fluoxetine to be expressed (42). There re severl other possible explntions for why our findings differ from those of Blier et l. (13). The studies differ in terms of length of tretment, primry outcome, nd scles used to ssess outcomes. Our results re not ccounted for by either differentil ttrition cross the three tretments or bseline differences. On the other AJP in Advnce jp.psychitryonline.org 7
Combining Antidepressnts TABLE 5. Outcomes t 12 Weeks nd 7 Months of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Chrcteristic All (N=665) Monotherpy: Escitloprm Plus Plcebo (N=224) Ptient Group Sustined-Relese Bupropion Plus Escitloprm (N=221) Extended-Relese Venlfxine Plus Mirtzpine (N=220) Comprison With Monotherpy b Bupropion Plus Escitloprm Venlfxine Plus Mirtzpine 12 weeks N % N % N % N % p p Erly termintion 182 27.4 55 24.6 70 31.7 57 25.9 0.10 0.75 Remission c 256 38.5 87 38.8 86 38.9 83 37.7 0.99 0.81 Lst QIDS-SR score 5 242 36.6 81 36.2 82 37.4 79 36.2 0.78 0.99 Reduction in QIDS-SR score 50% 334 51.9 113 51.8 111 51.6 110 52.4 0.97 0.91 Mximum side effect burden d 0.07 <0.0001 e No impirment 128 20.2 46 21.6 44 21.0 38 18.1 Miniml/mild 276 43.6 110 51.6 90 42.9 76 36.2 Moderte/mrked 167 26.4 48 22.5 55 26.2 64 30.5 Severe/intolerble 62 9.8 9 4.2 21 10.0 32 15.2 Lst side effect burden d 0.66 0.64 No impirment 344 54.7 117 55.5 118 56.5 109 52.2 Miniml/mild 215 34.2 74 35.1 69 33.0 72 34.4 Moderte/mrked 53 8.4 16 7.6 14 6.7 23 11.0 Severe/intolerble 17 2.7 4 1.9 8 3.8 5 2.4 At lest one serious dverse event 27 4.1 8 3.6 7 3.2 12 5.5 0.82 0.34 At lest one psychitric serious dverse event 7 1.1 1 0.4 1 0.5 5 2.3 1.00 0.12 Men SD Men SD Men SD Men SD p p Lst QIDS-SR score 8.1 5.4 7.9 5.2 8.1 5.3 8.4 5.7 0.74 0.34 Percent chnge in QIDS-SR score 45.6 35.1 46.5 35.2 44.6 34.6 45.8 35.8 0.59 0.85 Score on IDS-C nxiety subscle 2.6 2.1 2.4 2.0 2.6 2.1 2.8 2.3 0.28 0.10 Lst Qulity of Life Inventory (37) score 0.2 2.3 0.1 2.4 0.3 2.1 0.1 2.4 0.57 0.92 Lst Work nd Socil Adjustment Scle (36) score 14.9 12.3 14.9 11.9 13.9 11.9 15.9 13.0 0.43 0.53 Lst number of symptom worsenings f 5.1 5.1 4.7 4.9 5.0 4.4 5.7 5.8 0.12 0.04 e 7 months N % N % N % N % p p Erly termintion 244 36.7 78 34.8 84 38.0 82 37.3 0.49 0.60 Remission c 298 44.8 103 46.0 103 46.6 92 41.8 0.90 0.38 Lst QIDS-SR score 5 292 44.4 101 45.3 101 46.3 90 41.5 0.83 0.42 Reduction in QIDS-SR score 50% 374 58.4 129 59.4 125 58.4 120 57.4 0.83 0.68 Mximum side effect burden d 0.14 <0.0001 e No impirment 115 18.2 43 20.2 41 19.5 31 14.8 Miniml/mild 269 42.5 107 50.2 88 41.9 74 35.2 Moderte/mrked 184 29.1 52 24.4 60 28.6 72 34.3 Severe/intolerble 65 10.3 11 5.2 21 10.0 33 15.7 Lst side effect burden d 0.77 0.02 e No impirment 374 59.3 135 63.7 128 61.0 111 53.1 Miniml/mild 184 29.2 60 28.3 60 28.6 64 30.6 Moderte/mrked 59 9.4 13 6.1 15 7.1 31 14.8 Severe/intolerble 14 2.2 4 1.9 7 3.3 3 1.4 At lest one serious dverse event 46 6.9 16 7.1 13 5.9 17 7.7 0.60 0.82 (continued) hnd, our prticipnts differed from those studied by Blier et l. Neither prticipnt group ws tretment resistnt. However, prticipnts in our study were required to hve chronic nd/or recurrent depression. In fct, there were fr more chroniclly ill prticipnts in our study thn in the one by Blier et l. In ddition, 62% 85% of their prticipnts hd melncholic fetures (compred to only 20% in this study). Some reports (17, 43) suggest better effic- cy, i.e., drug-plcebo differences, for inptients (who re more likely to hve melncholic fetures) with combintion medictions or dul-ction medictions. In ddition, met-nlysis by Perry (44) reveled tht broder-ction gents (e.g., tricyclic ntidepressnts) hve fr greter efficcy thn SSRIs in melncholic depression. To evlute the potentil impct of chronicity on outcome, we renlyzed the dt. Chronicity ws ssocited 8 jp.psychitryonline.org AJP in Advnce
Rush, Trivedi, Stewrt, Et Al. TABLE 5. Outcomes t 12 Weeks nd 7 Months of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions (continued) Chrcteristic All (N=665) Monotherpy: Escitloprm Plus Plcebo (N=224) Ptient Group Sustined-Relese Bupropion Plus Escitloprm (N=221) Extended-Relese Venlfxine Plus Mirtzpine (N=220) Comprison With Monotherpy b Bupropion Plus Escitloprm Venlfxine Plus Mirtzpine 7 months (continued) Men SD Men SD Men SD Men SD p p Lst QIDS-SR score 7.6 5.6 7.3 5.4 7.3 5.4 8.1 5.9 0.98 0.26 Percent chnge in QIDS-SR score 49.5 36.0 50.9 34.5 49.8 37.0 47.8 36.7 0.98 0.49 Score on IDS-C nxiety subscle 2.5 2.1 2.4 2.1 2.6 2.1 2.5 2.2 0.19 0.34 Lst Qulity of Life Inventory (37) score 0.5 2.4 0.4 2.6 0.6 2.1 0.4 2.4 0.27 0.87 Lst Work nd Socil Adjustment Scle (36) score 13.8 12.5 13.5 12.0 13.0 12.2 15.0 13.2 0.62 0.31 Lst number of worsening dverse events f 4.9 5.3 4.7 5.4 4.8 5.2 5.4 5.3 0.47 <0.05 e For some vribles, dt were not vilble for ll subjects. IDS-C, 30-item Inventory of Depressive Symptomtology Clinicin-Rted (30). QIDS-SR, 16-item Quick Inventory of Depressive Symptomtology Self-Report (29). b Chi-squre nlysis for ctegoricl dt nd t tests for continuous dt. c Defined s t lest one of the lst two consecutive QIDS-SR scores 5 nd the other 7. d From the Frequency, Intensity, nd Burden of Side Effects Rting Scle (31). e Significntly different from monotherpy. f From the Systemtic Assessment for Tretment Emergent Effects Systemtic Inquiry (38). FIGURE 2. Rtes of Remission nd Response for Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Percent 70 60 50 40 30 20 10 0 12 Weeks (Acute Phse) 7 Months (Continution Phse) Monotherpy: escitloprm + plcebo (N=224) Remission Response Sustined-relese bupropion + escitloprm (N=221) Remission Response Extended-relese venlfxine + mirtzpine (N=220) Remission Response Remission ws defined s scores of less thn 8 nd less thn 6 on the 16-item Quick Inventory of Depressive Symptomtology Self-Report (QIDS-SR) (29) t the lst two consecutive ssessments. Response ws defined s reduction of t lest 50% in QIDS-SR score. with lower remission rtes cross ll three tretment cells. Specificlly, in the bupropion-escitloprm group, 37.2% of the ptients with episodes of chronic depression hd remissions, wheres the remission rte ws 41.0% for nonchronic depression. Anlogous rtes were 35.5% versus 43.1% for escitloprm-plcebo nd 34.9% versus 41.9% for venlfxine-mirtzpine. We conducted similr nlysis to compre ptients with nd without melncholic fetures from the current study. Melncholic fetures were ssocited with more xis I comorbidity, greter symptom severity, nd more suicidl plns nd thoughts. However, remission rtes rnged from 30.0% to 39.1% for those with melncholic fetures nd from 37.5% to 39.5% for those without. There were no differences cross mediction groups. Thus, neither the difference in the proportion with chronic illness nor the difference in melncholi seems to ccount for why our results differ from those of Blier et l. (13). While we enrolled the kinds of ptients (i.e., those with chronic or recurrent depression) for whom most clinicins would be likely to consider ntidepressnt mediction combintions (45), we cnnot rule out potentil impct on outcomes from one or more unknown bseline fetures. This group with chronic nd/or recurrent depression hd high rtes of self-reported emotionl, sexul, or physicl buse before ge 18, nd high proportion hd nxious fetures. While these fetures could hve reduced the overll benefit of ny one tretment, they would be unlikely to obscure differences between tretment cells, given their proportionl distribution cross the cells. Perhps the differences between the present study nd the results reported by Blier et l. (13) re due to the specific ntidepressnt medictions nd doses tht we used nd to AJP in Advnce jp.psychitryonline.org 9
Combining Antidepressnts TABLE 6. Odds Rtios nd Bet Coeffi cients From Regression Models of Outcomes t 12 Weeks nd 7 Months of Depressed Ptients in Comprison of Antidepressnt Monotherpy With Two Antidepressnt Combintions Comprison with Monotherpy: Escitloprm Plus Plcebo (N=224) Sustined-Relese Bupropion Plus Escitloprm (N=221) Extended-Relese Venlfxine Plus Mirtzpine (N=220) Outcome Undjusted Adjusted b Undjusted Adjusted c 12 weeks Odds Rtio p Odds Rtio p Odds Rtio p Odds Rtio p Erly termintion 1.42 0.10 1.46 0.09 1.08 0.75 1.00 1.00 Side effects d Mximum frequency 1.51 0.02 e 1.42 0.06 2.12 <0.0001 e 2.05 <0.0001 e Mximum intensity 1.82 <0.001 e 1.73 0.003 e 1.97 0.0002 e 1.86 0.0008 e Mximum burden 1.37 0.09 1.28 0.18 1.96 0.0002 e 1.87 0.0008 e Lst frequency 1.25 0.23 1.12 0.57 1.70 <0.004 e 1.62 <0.02 e Lst intensity 1.31 0.15 1.19 0.36 1.68 <0.005 e 1.58 <0.02 e Lst burden 0.99 0.96 0.90 0.61 1.19 0.36 1.11 0.61 At lest one serious dverse event f 0.88 0.82 1.56 0.35 At lest one psychitric serious dverse event f 1.01 1.00 5.19 0.14 Lst QIDS-SR score 5 1.06 0.78 1.08 0.72 1.00 0.99 1.14 0.54 Reduction in QIDS-SR score 50% 0.99 0.97 0.96 0.85 1.02 0.91 1.03 0.87 Lst Work nd Socil Adjustment Scle (36) score g 0.93 0.67 0.98 0.93 1.15 0.43 0.95 0.80 b p b p b p b p Mximum number of worsening dverse events h 0.07 0.26 0.08 0.19 0.11 0.06 0.11 0.06 Lst number of worsening dverse events h 0.07 0.46 0.06 0.54 0.20 <0.04 e 0.14 0.14 Lst QIDS-SR score 0.17 0.74 0.13 0.80 0.50 0.34 0.02 0.97 Percent chnge in QIDS-SR score 1.82 0.59 2.44 0.47 0.66 0.85 0.22 0.95 Score on IDS-C nxiety subscle 0.10 0.25 0.06 0.44 0.16 0.06 0.09 0.24 Lst Qulity of Life Inventory (37) score 0.13 0.57 0.12 0.60 0.03 0.92 0.14 0.54 7 months Odds Rtio p Odds Rtio p Odds Rtio p Odds Rtio p Erly termintion 1.15 0.49 1.15 0.49 1.11 0.60 1.07 0.75 Side effects d Mximum frequency 1.53 <0.02 e 1.44 <0.05 e 2.31 <0.0001 e 2.20 <0.0001 e Mximum intensity 1.79 <0.002 e 1.67 <0.006 e 2.26 <0.0001 e 2.12 <0.0001 e Mximum burden 1.34 0.11 1.26 0.22 2.15 <0.0001 e 2.02 0.0002 e Lst frequency 1.40 0.08 1.32 0.16 1.80 <0.002 e 1.76 <0.004 e Lst intensity 1.53 <0.03 e 1.48 <0.05 e 1.94 0.0004 e 1.99 0.0005 e Lst burden 1.15 0.48 1.09 0.69 1.63 <0.02 e 1.61 0.02 e At lest one serious dverse event 0.81 0.60 1.09 0.82 At lest one psychitric serious dverse event 0.60 0.50 1.44 0.54 Lst QIDS-SR score 5 1.04 0.83 1.02 0.93 0.86 0.42 0.98 0.95 Reduction in QIDS-SR score 50% 0.96 0.83 0.92 0.69 0.92 0.68 0.99 0.97 Lst Work nd Socil Adjustment Scle (36) score g 0.96 0.81 1.05 0.79 1.25 0.20 1.06 0.74 b p b p b p b p Mximum number of worsening dverse events h 0.05 0.37 0.06 0.26 0.10 0.10 0.10 0.09 Lst number of worsening dverse events h 0.03 0.76 0.04 0.73 0.14 0.19 0.13 0.22 Lst QIDS-SR score 0.04 0.94 0.05 0.93 0.76 0.16 0.22 0.66 Percent chnge in QIDS-SR score 1.14 0.74 2.09 0.54 3.17 0.36 2.12 0.53 Score on IDS-C nxiety subscle 0.10 0.27 0.06 0.49 0.07 0.41 0.02 0.77 Lst Qulity of Life Inventory (37) score 0.26 0.27 0.24 0.28 0.04 0.87 0.11 0.64 IDS-C, 30-item Inventory of Depressive Symptomtology Clinicin-Rted (30). QIDS-C, 16-item Quick Inventory of Depressive Symptomtology Clinicin-Rted (29). QIDS-SR, 16-item Quick Inventory of Depressive Symptomtology Self-Report (29). b Adjusted for regionl center, employment, nd nxious fetures. c Adjusted for regionl center, sex, bseline score on IDS-C, nd bseline score on Work nd Socil Adjustment Scle. d From the Frequency, Intensity, nd Burden of Side Effects Rting Scle (31). e Significnt odds (ctegoricl mesures) or bet (continuous mesures) for the combintion. f Adjusted models re unestimble. g An extremely nonnorml distribution required binning. h From the Systemtic Assessment for Tretment Emergent Events Systemtic Inquiry (38). 10 jp.psychitryonline.org AJP in Advnce
Rush, Trivedi, Stewrt, Et Al. the doses tht were dministered. There is evidence for the efficcy of venlfxine plus mirtzpine (13, 35) nd bupropion plus escitloprm (16, 34). Crpenter et l. (46) found tht mirtzpine, when used s n djunct to previously ineffective SSRIs lone, ws more effective thn plcebo in treting depression. In fct, Blier et l. (13) used the venlfxine-mirtzpine combintion nd Stewrt et l. (41) used bupropion plus escitloprm in their trils, lthough the doses were higher in both of them. The rtionle for higher venlfxine dose is tht the effect on the norepinephrine system (17, 47) is only relized t doses of t lest 225 mg/dy. Mirtzpine hs ntgonistic effects tht re modest t 15 mg/dy nd more clerly evident t 30 mg/ dy. Thus, s suggested by Blier (personl communiction), the doses in CO-MED my hve been insufficient in lrge enough proportion of prticipnts to preclude the benefit otherwise vilble from the combintion. In n ttempt to evlute tht notion, we identified the 86 prticipnts who reched 225 mg/dy of venlfxine nd 30 mg/dy of mirtzpine t ny time during tretment. Their remission rte ws 33.7% t 12 weeks nd 41.9% t 7 months. These results do not suggest tht underdosing ws the cuse of the poor performnce of this combintion. It remins n unnswered question whether these lrger doses (if they re required to chieve n dvntge for ntidepressnt combintions) re chievble in prctice with more representtive ptients who hve chronic nd/ or recurrent mjor depressive disorder nd more concurrent xis I nd III disorders. This study hd severl limittions. While lrger thn mny studies, the study group my not be representtive of the universe of outptients with chronic nd/or recurrent mjor depression. As noted, the doses used my not hve been sufficient to relize the full potentil vlue of combintion ntidepressnt medictions. The results for the continution tretment phse re limited by the fct tht the subjects were not rerndomized or strtified by level of improvement following the cute phse. Finlly, the clinicins were not blind to tretment, nd structured interview ws not used to estblish xis I dignoses. In summry, in outptients with chronic nd/or recurrent mjor depressive disorder, there ppers to be no dvntge to either mediction combintion over escitloprm lone s first-step tretment for nonresistnt depression. Some combintions my incur risk of higher side effect burden. This conclusion is conditioned on the doses employed. Received Nov. 16, 2010; revision received Feb. 10, 2011; ccepted Feb. 14, 2011 (doi: 10.1176 /ppi.jp.2011.101116 4 5). From the Duke Ntionl University of Singpore Grdute Medicl School, Singpore; the Deprtment of Psychitry, University of Texs Southwestern Medicl Center t Dlls; the Epidemiology Dt Center, Grdute School of Public Helth, University of Pittsburgh; the Depression Clinicl nd Reserch Progrm, Msschusetts G enerl Hospitl, Boston; New York Stte Psychitric Institute nd the Deprtment of Psychitry, College of Physicins nd Surgeons, Columbi University, New York; the Depression Reserch nd Clinic Progrm, Semel Insti- tute for Neuroscience nd Humn Behvior, UCLA; the Deprtment of Psychitry, Vnderbilt University, Nshville, Tenn.; the Deprtment of Psychitry, Hrbor-UCLA Medicl Center nd the Los Angeles Biomedicl Reserch Institute, Torrnce, Clif.; nd the Deprtment of Psychitry, Virgini Commonwelth University, Richmond. Address correspondence nd reprint requests to Dr. Trivedi, Deprtment of Psychitry, University of Texs Southwestern Medicl Center t Dlls, 5323 Hrry Hines Blvd., Dlls, TX 75390-9119; mdhukr.trivedi@utsouthwestern.edu (e-mil). Dr. Rush hs received consulting fees from Advnced Neuromodultion Systems, Best Prctice Project Mngement, Brin Resource, Otsuk, nd the University of Michign; he hs received consultnt/ speker fees from Forest; he hs received consultnt fees from nd owns stock in Pfizer; he hs received uthor roylties from G uilford Publictions, Helthcre Technology Systems, nd the University of Texs Southwestern Medicl Center; nd he hs received reserch support from the Ntionl Institute of Mentl Helth. Dr. Trivedi hs received reserch support from the Agency for Helthcre Reserch nd Q ulity, Corcept, Cyberonics, Merck, NARSAD, the Ntionl Institute of Mentl Helth, the Ntionl Institute on Drug Abuse, Novrtis, Phrmci & Upjohn, Predix (EPIX), Solvy, nd Trgcept; he hs received consulting nd speker fees from Abbott, Abdi Ibrhim, Akzo (Orgnon), AstrZenec, Bristol-Myers Squibb, Cephlon, Eli Lilly, Evotec, Fbre Krmer, Forest, GlxoSmithKline, Jnssen, Johnson & Johnson, Mede Johnson, Medtronic, Neuronetics, Otsuk, Prke-Dvis, Pfizer, Seprcor, Shire Development, VntgePoint, nd W yeth. Dr. Fv hs received reserch support from Abbott, Alkermes, Aspect Medicl Systems, AstrZenec, BioReserch, BrinCells, Bristol-Myers Squibb, CeNeRx BioPhrm, Cephlon, Clinicl Trils Solutions, Clintr, Covidien, Eli Lilly, EnVivo, Euthymics Bioscience, Forest, G neden Biotech, G lxosmithkline, Icon Clinicl Reserch, i3 Innovus/Ingenix, Johnson & Johnson, Lichtwer, Lorex, NARSAD, the Ntionl Center for Complementry nd Alterntive Medicine, the Ntionl Institute on Drug Abuse, NIMH, Novrtis, Orgnon, Pmlb, Pfizer, Phrmvite, Photother, RCT Logic, Roche, Snofi-Aventis, Shire, Solvy, Synthelbo, nd W yeth; he hs served s dviser or consultnt to Abbott, Affectis, Alkermes, Amrin, Aspect Medicl Systems, AstrZenec, Auspex, Byer, Best Prctice Project Mngement, BioMrin, Biovil, BrinCells, Bristol-Myers Squibb, CeNeRx BioPhrm, Cephlon, Clinicl Trils Solutions, CNS Response, Compellis, Cypress, DignoSerch Life Sciences, Dinippon Sumitomo, DOV, Edgemont, Eisi, Eli Lilly, ephrmsolutions, EPIX, Euthymics Bioscience, Fbre-Krmer, Forest, G enomind, G lxosmithkline, G runenthl, i3 Innovus/Ingenix, Jnssen, Jzz, Johnson & Johnson, Knoll, Lbophrm, Lorex, Lundbeck, MedAvnte, Merck, MSI Methyltion Sciences, Nurex, Neuronetics, NextW ve, Novrtis, Nutrition 21, Orexigen, Orgnon, Otsuk, Pmlb, Pfizer, PhrmStr, Phrmvite, PhrmoRx, Precision Humn Biolbortory, Prex, PsychoG enics, Psylin Neurosciences, Puretech Ventures, RCT Logic, Rexhn, Ridge Dignostics, Roche, Snofi-Aventis, Schering-Plough, Seprcor, Servier, Solvy, Somxon, Somerset, Sunovion, Synthelbo, Tked, Tl Medicl, Tetrgenex, Trnscept, TrnsForm, nd Vnd; he hs received speking or publishing fees from Admed, Advnced Meeting Prtners, the Americn Psychitric Assocition, the Americn Society of Clinicl Psychophrmcology, AstrZenec, Belvoir Medi Group, Boehringer Ingelheim, Bristol-Myers Squibb, Cephlon, CME Institute/Physicins Postgrdute Press, Eli Lilly, Forest, GlxoSmithKline, Imedex, MGH Psychitry Acdemy/ Primedi, MG H Psychitry Acdemy/Reed Elsevier, Novrtis, Orgnon, Pfizer, PhrmStr, United BioSource, nd W yeth; he owns stock in Compellis; he hs ptent for SPCD, ptent ppliction for combintion of zpirones nd bupropion in mjor depressive disorder, nd ptent for reserch nd licensing of SPCD with RCT Logic nd Lippincott, W illims & W ilkins; nd he receives copyright roylties for the MG H CPFQ, SFI, ATRQ, DESS, nd SAFER. Dr. Kurin hs received reserch support from Evotec, Pfizer, nd Trgcept. Dr. W rden hs owned stock in Bristol-Myers Squibb nd Pfizer in the pst 5 yers. Dr. Husin hs received reserch support from Cyberonics, Mgstim, Neuronetics, NIH/NIMH, the Stnley Foundtion, nd St. Jude Medicl; he hs lso served on dvisory bords for AstrZenec, BMS, Forest, nd Novrtis. Dr. Cook hs served s n dviser nd consultnt for Ascend Medi, Bristol-Myers Squibb, Cyberonics, Jnssen, NeuroSigm, nd the U.S. Deprtments of Defense nd of Justice; he hs served on the spekers bureus for Bristol-Myers Squibb, Neuronetics, nd W yeth/pfizer; he hs received reserch support from Aspect Medicl Systems/Covidien, Cyberonics, Eli Lilly, Neuronetics, AJP in Advnce jp.psychitryonline.org 11
Combining Antidepressnts NIH, Novrtis, Pfizer, nd Seprcor; his ptents on biomedicl devices re ssigned to the Regents of the University of Cliforni. Dr. Shelton hs received grnt/reserch support from Abbott, AstrZenec, Eli Lilly, G lxosmithkline, Jnssen, Pmlb, Pfizer, Snofi, nd W yeth; he hs been pid consultnt to Evotec, Jnssen, nd Sierr Neurophrmceuticls; nd he hs served on spekers bureus for Abbott, Bristol-Myers Squibb, Eli Lilly, G lxosmithkline, Jnssen, nd W yeth. Dr. Lesser hs received reserch support from Aspect Medicl Systems nd NIMH. Dr. Kornstein hs received grnts/reserch support from Boehringer-Ingelheim, Bristol-Myers Squibb, Eli Lilly, Forest, the Ntionl Institute of Mentl Helth, Novrtis, Pfizer, Rexhn, Tked, nd W yeth; she hs served on dvisory bords for Bristol- Myers Squibb, Dey, Eli Lilly, Forest, Pfizer, PG x Helth, Tked, nd W yeth; nd she hs received book roylties from G uilford Press. Dr. W isniewski reports finncil reltionships with Cyberonics (2005 2009), ImRx Therpeutics (2006), Bristol-Myers Squibb (2007 2008), Orgnon (2007), Cse-W estern University (2007), Singpore Clinicl Reserch Institute (2009), Dey (2010), nd Venebio (2010). The other uthors report no finncil reltionships with commercil interests. Funded by NIMH under contrct N01 MH-90003 to the University of Texs Southwestern Medicl Center t Dlls (principl investigtors, A.J. Rush nd M.H. Trivedi). Forest Phrmceuticls, G lxosmithkline, Orgnon, nd W yeth Phrmceuticls provided medictions for this tril t no cost. The content of this publiction does not necessrily reflect the views or policies of the U.S. Deprtment of Helth nd Humn Services, nor does mention of trde nmes, commercil products, or orgniztions imply endorsement by the U.S. government. NIMH hd no role in the drfting or review of the mnuscript or in the collection or nlysis of the dt. The uthors thnk the clinicl stff t ech clinicl site for their ssistnce with this project; ll of the study prticipnts; Eric Nestler, M.D., Ph.D., nd Crol A. Tmming, M.D., for dministrtive support; nd Jon Kilner, M.S., M.A., for editoril support. References 1. Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikngs KR, Rush AJ, Wlters EE, Wng PS: The epidemiology of mjor depressive disorder: results from the Ntionl Comorbidity Survey Repliction (NCS-R). JAMA 2003; 289:3095 3105 2. Depression Guideline Pnel: Clinicl Prctice Guideline, number 5: Depression in Primry Cre, Volume 1: Detection nd Dignosis: AHCPR Publiction 93-0550. Rockville, Md, Agency for Helth Cre Policy nd Reserch, 1993 3. Keller MB, Klein DN, Hirschfeld RMA, Kocsis JH, McCullough JP, Miller I, First MB, Holzer CP III, Keitner GI, Mrin DB, She T: Results of the DSM-IV mood disorders field tril. Am J Psychitry 1995; 152:843 849 4. Rush AJ, Kremer HC, Sckeim HA, Fv M, Trivedi MH, Frnk E, Ninn PT, Thse ME, Gelenberg AJ, Kupfer DJ, Regier DA, Rosenbum JF, Ry O, Schtzberg AF: Report by the ACNP Tsk Force on Response nd Remission in Mjor Depressive Disorder. Neuropsychophrmcology 2006; 31:1841 1853 5. Keller MB, Gelenberg AJ, Hirschfeld RM, Rush AJ, Thse ME, Kocsis JH, Mrkowitz JC, Fwcett JA, Korn LM, Klein DN, Russell JM, Kornstein SG, McCullough JP, Dvis SM, Hrrison W M: The tretment of chronic depression, prt 2: double-blind, rndomized tril of sertrline nd imiprmine. J Clin Psychitry 1998; 59:598 607 6. Keller MB, McCullough JP, Klein DN, Arnow B, Dunner DL, Gelenberg AJ, Mrkowitz JC, Nemeroff CB, Russell JM, Thse ME, Trivedi MH, Zjeck J: A comprison of nefzodone, the cognitive behviorl-nlysis system of psychotherpy, nd their combintion for the tretment of chronic depression. N Engl J Med 2000; 342:1462 1470 7. Mischoulon D, Nierenberg AA, Kizilbsh L, Rosenbum JF, Fv M: Strtegies for mnging depression refrctory to selective serotonin reuptke inhibitor tretment: survey of clinicins. Cn J Psychitry 2000; 45:476 481 8. Depression Guideline Pnel: Clinicl Prctice Guideline, number 5: Depression in Primry Cre: Volume 2. Tretment of Mjor Depression: AHCPR Publiction 93-0551. Rockville, Md, Agency for Helth Cre Policy nd Reserch, 1993 9. Ptel A, ADVANCE Collbortive Group, McMhon S, Chlmers J, Nel B, Woodwrd M, Billot L, Hrrp S, Poulter N, Mrre M, Cooper M, Glsziou P, Grobbee DE, Hmet P, Heller S, Liu LS, Mnci G, Mogensen CE, Pn CY, Rodgers A, Willims B: Effects of fixed combintion of perindopril nd indpmide on mcrovsculr nd microvsculr outcomes in ptients with type 2 dibetes mellitus (the ADVANCE tril): rndomised controlled tril. Lncet 2007; 370:829 840 10. Mourd JJ, Weber B, Znnd F, Lville M, Duru G, Andréjk M, Investigtors of the STRATHE tril: Comprison of different therpeutic strtegies in hypertension: low-dose combintion of perindopril/indpmide versus sequentil monotherpy or stepped-cre pproch. J Hypertens 2004; 22:2379 2386; correction, 2007; 25:258 11. Rush AJ, Trivedi MH, W isniewski SR, Nierenberg AA, Stewrt JW, Wrden D, Niederehe G, Thse ME, Lvori PW, Lebowitz BD, Mc- Grth PJ, Rosenbum JF, Sckeim HA, Kupfer DJ, Luther J, Fv M: Acute nd longer-term outcomes in depressed outptients requiring one or severl tretment steps: STAR*D report. Am J Psychitry 2006; 163:1905 1917 12. Rush AJ, Trivedi MH, W isniewski SR, Stewrt JW, Nierenberg AA, Thse ME, Ritz L, Biggs MM, Wrden D, Luther JF, Shores-Wilson K, Niederehe G, Fv M: Bupropion-SR, sertrline, or venlfxine-xr fter filure of SSRIs for depression. N Engl J Med 2006; 354:1231 1242 13. Blier P, Wrd HE, Trembly P, Lberge L, Hébert C, Bergeron R: Combintion of ntidepressnt medictions from tretment initition for mjor depressive disorder: double-blind rndomized study. Am J Psychitry 2010; 167:281 288 14. Kli M: Neurobiologicl bsis of depression: n updte. Metbolism 2005; 54(5 suppl 2):24 27 15. Fv M, Rush AJ: Current sttus of ugmenttion nd combintion tretments for mjor depressive disorder: literture review nd proposl for novel pproch to improve prctice. Psychother Psychosom 2006; 75:139 153 16. Trivedi MH, Fv M, Wisniewski SR, Thse ME, Quitkin F, Wrden D, Ritz L, Nierenberg AA, Lebowitz BD, Biggs MM, Luther JF, Shores-Wilson K, Rush AJ: Mediction ugmenttion fter the filure of SSRIs for depression. N Engl J Med 2006; 354:1243 1252 17. Nelson JC, Mzure CM, Jtlow PI, Bowers MB Jr, Price LH: Combining norepinephrine nd serotonin reuptke inhibition mechnisms for tretment of depression: double-blind, rndomized study. Biol Psychitry 2004; 55:296 300 18. Blier P, Gobbi G, Turcotte JE, de Montigny C, Boucher N, Hebert C, Debonnel G: Mirtzpine nd proxetine in mjor depression: comprison of monotherpy versus their combintion from tretment initition. Eur Neuropsychophrmcol 2009; 19:457 465 19. Kroenke K, Spitzer RL, Willims JB: The Ptient Helth Questionnire-2: vlidity of two-item depression screener. Med Cre 2003; 41:1284 1292 20. Kroenke K, Spitzer RL, Willims JBW: The PHQ-9: vlidity of brief depression severity mesure. J Gen Intern Med 2001; 16:606 613 21. Americn Psychitric Assocition: Dignostic nd Sttisticl Mnul of Mentl Disorders, 4th Edition, Text Revision (DSM- IV-TR). Wshington, DC, APA, 2000 22. Hmilton M: A rting scle for depression. J Neurol Neurosurg Psychitry 1960; 23:56 61 23. Fv M, Rush AJ, Alpert JE, Blsubrmni GK, Wisniewski SR, Crmin CN, Biggs MM, Zisook S, Leuchter A, Howlnd R, Wrden D, Trivedi MH: Difference in tretment outcome in outp- 12 jp.psychitryonline.org AJP in Advnce
Rush, Trivedi, Stewrt, Et Al. tients with nxious versus nonnxious depression: STAR*D report. Am J Psychitry 2008; 165:342 351 24. Clery P, Guy W: Fctor nlysis of the Hmilton Depression Scle. Drugs Exp Clin Res 1977; 1:115 120 25. Rush AJ, Zimmermn M, W isniewski SR, Fv M, Hollon SD, Wrden D, Biggs MM, Shores-Wilson K, Shelton RC, Luther JF, Thoms B, Trivedi MH: Comorbid psychitric disorders in depressed outptients: demogrphic nd clinicl fetures. J Affect Disord 2005; 87:43 55 26. Sngh O, Stucki G, Ling MH, Fossel AH, Ktz JN: The Self- Administered Comorbidity Questionnire: new method to ssess comorbidity for clinicl nd helth services reserch. Arthritis Rheum 2003; 49:156 163 27. Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Wrden D, Ritz L, Norquist G, Howlnd RH, Lebowitz B, McGrth PJ, Shores- Wilson K, Biggs MM, Blsubrmni GK, Fv M, STAR*D Study Tem: Evlution of outcomes with citloprm for depression using mesurement-bsed cre in STAR*D: implictions for clinicl prctice. Am J Psychitry 2006; 163:28 40 28. Trivedi MH, Rush AJ, Gynes BN, Stewrt JW, W isniewski SR, Wrden D, Ritz L, Luther JF, Stegmn D, DeVeugh-Geiss J, Howlnd R: Mximizing the dequcy of mediction tretment in controlled trils nd clinicl prctice: STAR*D mesurementbsed cre. Neuropsychophrmcology 2007; 32:2479 2489 29. Rush AJ, Bernstein IH, Trivedi MH, Crmody TJ, Wisniewski S, Mundt JC, Shores-Wilson K, Biggs MM, Woo A, Nierenberg AA, Fv M: An evlution of the Quick Inventory of Depressive Symptomtology nd the Hmilton Rting Scle for Depression: A Sequenced Tretment Alterntives to Relieve Depression tril report. Biol Psychitry 2006; 59:493 501 30. Rush AJ, Gullion CM, Bsco MR, Jrrett RB, Trivedi MH: The Inventory of Depressive Symptomtology (IDS): psychometric properties. Psychol Med 1996; 26:477 486 31. Wisniewski SR, Rush AJ, Blsubrmni GK, Trivedi MH, Nierenberg AA: Self-rted globl mesure of the frequency, intensity, nd burden of side effects. J Psychitr Prct 2006; 12:71 79 32. Wisniewski SR, Eng H, Meloro L, Gtt R, Ritz L, Stegmn D, Trivedi M, Biggs MM, Friedmn E, Shores-W ilson K, Wrden D, Brtolo wits D, Mrtin JP, Rush AJ: Web-bsed communictions nd mngement of multi-center clinicl tril: the Sequenced Tretment Alterntives to Relieve Depression (STAR*D) project. Clin Trils 2004; 1:387 398 33. Fv M: Augmenttion nd combintion strtegies in tretmentresistnt depression. J Clin Psychitry 2001; 62(suppl 18):4 11 34. Leuchter AF, Lesser IM, Trivedi MH, Rush AJ, Morris DW, Wrden D, Fv M, Wisniewski SR, Luther JF, Perles M, Gynes BN, Stewrt JW: An open pilot study of the combintion of escitloprm nd bupropion-sr for outptients with mjor depressive disorder. J Psychitr Prct 2008; 14:271 280 35. McGrth PJ, Stewrt JW, Fv M, Trivedi MH, Wisniewski SR, Nierenberg AA, Thse ME, Dvis L, Biggs MM, Shores-Wilson K, Luther JF, Niederehe G, Wrden D, Rush AJ, STAR*D Study Tem: Trnylcypromine versus venlfxine plus mirtzpine following three filed ntidepressnt mediction trils for depression: STAR*D report. Am J Psychitry 2006; 163:1531 1541 36. Mundt JC, Mrks IM, Sher MK, Greist JH: The Work nd Socil Adjustment Scle: simple mesure of impirment in functioning. Br J Psychitry 2002; 180:461 464 37. Frisch MB, Clrk MP, Rouse SV, Rudd MD, Pweleck JK, Greenstone A, Kopplin DA: Predictive nd tretment vlidity of life stisfction nd the Qulity of Life Inventory. Assessment 2005; 12:66 78 38. Levine J, Schooler NR: Generl versus specific inquiry with SAFTEE (letter). J Clin Psychophrmcol 1992; 12:448 39. Altmn EG, Hedeker D, Peterson JL, Dvis JM: The Altmn Self- Rting Mni Scle. Biol Psychitry 1997; 42:948 955 40. Fv M, Iosifescu DV, Pedrelli P, Ber L: Relibility nd vlidity of the Msschusetts Generl Hospitl Cognitive nd Physicl Functioning Questionnire. Psychother Psychosom 2009; 78:91 97 41. Stewrt JW, McGrth PJ, Deliynnides RA, Quitkin FM: Does dul ntidepressnt therpy s initil tretment hsten nd increse remission from depression? J Psychitr Prct 2009; 15:337 345 42. Quitkin FM, Petkov E, McGrth PJ, Tylor B, Besley C, Stewrt J, Amsterdm J, Fv M, Rosenbum J, Reimherr F, Fwcett J, Chen Y, Klein D: When should tril of fluoxetine for mjor depression be declred filed? Am J Psychitry 2003; 160:734 740 43. Poirier MF, Boyer P: Venlfxine nd proxetine in tretmentresistnt depression: double-blind, rndomised comprison. Br J Psychitry 1999; 175:12 16 44. Perry PJ: Phrmcotherpy for mjor depression with melncholic fetures: reltive efficcy of tricyclic versus selective serotonin reuptke inhibitor ntidepressnts. J Affect Disord 1996; 39:1 6 45. Khn A, Dger SR, Cohen S, Avery DH, Scherzo B, Dunner DL: Chronicity of depressive episode in reltion to ntidepressntplcebo response. Neuropsychophrmcology 1991; 4:125 130 46. Crpenter LL, Ysmin S, Price LH: A double-blind, plcebo-controlled study of ntidepressnt ugmenttion with mirtzpine. Biol Psychitry 2002; 51:183 188 47. Mes M, Libbrecht I, vn Hunsel F, Cmpens D, Meltzer HY: Pindolol nd minserin ugment the ntidepressnt ctivity of fluoxetine in hospitlized mjor depressed ptients, including those with tretment resistnce. J Clin Psychophrmcol 1999; 19:177 182 AJP in Advnce jp.psychitryonline.org 13