Table 1.
Summary of included studies.
| Study | Population | Design | Results | Conclusions |
|---|---|---|---|---|
| 31 | 17 euthymic BD-I on Li+ × 35±37 months (7 ♂; 10 ♀) =37±10 vs. 21 HCs (7 ♂; 14 ♀) =30±8 (the latter were significantly younger, p<0.05) | Self-rated mood stability of pts. with BD and HCs with VAMS × 11 days | Patients on Li+ did not differ from HCs on VAMS ratings, but rated themselves as less swinging than HCs (6.9±4 vs. 12.5±6; p<0.01); two patients withdrawn from Li+ developed a manic episode; another did not, but developed mood instability | Li+ confers mood stabilization independently from age or gender |
| 32 | 120 BDI, aged >60 (=40.3±15.3; 52 ♂ and 68 ♀). Two groups: manic (N=50); mixed (N=70), further divided in those who have received Li+, AEs, or both | Naturalistic, retrospective (length of hospital stay). Outcome measures: between-group differences; Remission criteria: CGI-I score ≥2 | Serum VPA levels were higher in mixed (84.7 ± 13.9) than manic (64.6 ± 12.8) group. In mixed, pts. receiving AE+Li+ have a slower remission than those receiving Li+ alone or AE alone. Many pts. on Li++AE had the AE added after Li+ proved ineffective. After adding AE to Li+, they achieved remission in 2-3 weeks. In the whole pt. group, weeks needed to achieve adequate MS serum level was a predictor of remission | Effectiveness of Li+ and AEs in acute treatment of mixed or pure mania are similar. Time course to remission seems to be influenced by the speed with which patients achieve a therapeutic serum MS levels |
| 33 | 72 BD inpts. aged >60 (mean age 67.19 ± 5.34); 27 ♂ and 45 ♀ | Retrospective chart review; Pts. were maintained on MS monotherapy with Li+ (41), CBZ (11), or VPA (20) during hospitalization; Outcome measures: length of treatment and change in GAF scores | Length of hospital stay was 27.5±15.4 days for pts. on CBZ, 22.1±11.2 days for pts. on Li+, and 24.8±13.5 days for pts. on VPA. ↑GAF 28.8±11.8 for pts. on CBZ, 29.9±15.8 for pts. on Li+, and 35.2±10.9 for pts. on VPA | No significant differences in outcome measures of acutely ill geriatric pts. with BD who were treated with Li+, VPA, or CBZ alone |
| 34 | 10 BD-I, mixed, aged 37-72 (=50.4±2.8; 9 ♂; 1 ♀) | Open, ad- on gabapentin 900 2700 mg/day × 4 weeks; outcome measures: reduction of HDRS and BRMaS | Add on gabapentin ↓ HDRS (15 8) and BRMaS scores (8 1) at week 1. BRMaS scores stabilized through week 4; HDRS scores continued to ↓ (8 4) | Gabapentin potentiates the effect of the other mood stabilizers in subjects with BD and a mixed state |
| 35 | 27 BD, depressed, aged 37-72 (=41±12; 9 ♂; 18 ♀, 11 BD-I, 16 BD-II). Two consecutive HDRS>16 | DB, add-on paroxetine (36 mg/day) or additional MS (Li+, 1300 mg/day; VPA 1200 mg/day) Li+, in pts. already treated with an MS (Li+ or VPA, same dosages as above) × 6 weeks. Outcome measures: HDRS; YMRS; GAF scores | After 6 weeks, significant ↓ in HDRS scores in both groups, with no differences between groups. YMRS unchanged in both groups | Both adding another MS and adding an AD are effective in treating BD depression |
| 36 | 31 adults with DSM-IV BD, last episode mixed | Open add-on risperidone =4.2 mg/day on ongoing mood stabilizer × 6 months, weekly assessed with YMRS, CGI, PANSS, and HDRS. Response criteria: YMRS ≥50%↓from BL AND CGI ≥2↓from BL | 26 completers (84%); 74% responders at week 4; significant improvement on all scales, from week 1 onward on YMRS and HDRS, and from week 4 on CGI and PANSS; 73% asymptomatic at 6 months | Risperidone effective as add-on, but results need replication |
| 37 | 158 depressed BD-I inpts. ( age=52.6±15; 50♂; 108♀) | Retrospective, naturalistic, evaluation of the impact of AD treatment on switch incidence from depression to mania/hypomania | 25% switches; TCAs associated with higher risk of switching, reduced risk if ADs+MS | Better add AD only when MS is established |
| 38 | 11 BD, aged 19 – 46 (=29.4±10.7; 10♂; 1♀; 6 BD-I, 4 BD-II) | Open, add-on vitamin and mineral capsules (36 supplements) ×≥6 weeks (range: 6-21 weeks). Outcome measures: reduction in HDRS, YMRS, and BPRS scores; Response criteria (≥20% improvement on all scales) | As compared with the first assessment, at last assessment, HDRS decrement was 55% and YMRS decrement was 66%. Rates of response were 87.5%. Number of standard medications/pt ↓ from 2.7±2.0 to 1.0±1.1 | Vitamin and mineral capsules have beneficial psychotropic effects |
| 39 | 155, BDI, manic; 4 groups: MS + atypical AP (RISP or OLZ, N=69, =39.72 ± 14.50 11 BD, aged 19 – 46 (=29.4±10.7; 32♂; 37♀); MS + typical AP (N=69, =40.86 ± 16.11, 37♂; 32♀); MS + combination of typical and atypical (first typical than switch with atypical, N= 17, 41.06 ± 18.08, 8♂; 9♀) | Naturalistic, retrospective; outcome measures: length of stay, CGI improvement | No differences in length of stay were found. At discharge, subjects with MS + atypical AP and subjects with MS + combination between typical and atypical AP showed smaller CGI improvement scores (1.59±.58; 1.56±.63 respectively) than those with MS + typical AP (2.04±.73). Same results when subgroups showing psychotic features were selected. Subgroup of subjects treated with MS + RISP showed greater CGI score improvement than those treated with MS + typical AP | SGAs, in particular RISP, might be more effective than typical APs combined with MS, to treat manic episodes. If pts. require initial treatment with MS+FGA, they might have a better outcome if they switch to an SGA after the 1st week of treatment |
| 40 | 64 depressive pts. (IDS≥16) while on mood stabilizers, aged 22.5-75,3 yrs (= 44.8±11.8; 38 ♂; 26 ♀; 43 BD-I, 19 BD-II; 1 BD NOS, 1 SABT) | 5-site DB; randomization to add-on bupropion 100 450 mg/day, sertraline 50 200 mg/day, or venlafaxine 75 375 mg/day, on ongoii8yhoy po60nk0long mood stabilizers × 10 weeks; nonresponders, rerandomized to other AD × 10 weeks, if still nonresponders, to the third; 1-year continuation on what works; Assessment with CGI-BP, response criterion, score 1 or 2 | 21 nonresponders rerandomized, 10 of them further rerandomized; a total of 95 acute-phase treatments were available for outcome assessment. 35 pts. exposed to acute phase treatment were responding (37%). During the 95 acute AD-exposure phases, there were 13 (14%) switches to mania/hypomania | Some depressions subsided and some switched to mania, but conclusive considerations could not be made due to the peculiarities of the design. Furthermore, final results were not available and 1 site could not administer bupropion |
| 41 | 19 depressed outpts. BD-II ( age=29; 13 ♂; 6 ♀) | 12-weeks open, divalproex sodium single dose 250 mg increased by 250 mg every 4 days until symptom relief in medication naïve vs. MS naïve; response criterion, ≥50% ↓ in HDRS scores | 63% responders; higher response in medication naïve group | Results support divalproex sodium monotherapy in BD-II depression |
| 42 | 156 BD-I, manic/mixed, aged 18-65 yrs. Three groups: RISP+MS (N=52; median=41 yrs; 26 ♂; 26 ♀); HAL+MS (N=53; median=44 yrs; 30 ♂; 23 ♀); placebo + MS (N=51; median=43 yrs; 24 ♂; 27 ♀) | 3-week DB, placebo-controlled trial of RISP (range 2-6 mg)+MS (VPA: 65.4±27.1 µg/ml; Li+: 0.7±0.3 meq/l) and HAL (4-12 mg)+MS (VPA: 76.2±25.6; Li+: 0.7±0.2 meq/l). Outcome measures: changes in YMRS scores, % of pts. scoring 1 on the CGI-I (“very much improved”) | RISP+MS and HAL+MS lowered YMRS scores more than placebo+MS (-14.3 vs. -13.4 vs. -8.2, respectively). RISP+MS and HAL+MS obtained higher rates of “very much improved” on the CGI-I than placebo+MS (50% vs. 53% vs. 30%, respectively). No between-group differences in psychotic vs. nonpsychotic and manic vs. mixed subpopulations. The HAL+MS group worsened more than placebo+MS scores on the Extrapyramidal Symptom Rating Scale from BL to endpoint (2 vs. -0.1, respectively, and maximum score 1.9 vs. 5.4) | Both HAL and RISP are effective adjunctive treatments for manic and mixed states. RISP has a safer profile |
| 43 | 115 BD-I, manic or mixed, aged 18-70 yrs (=39yrs; 60♂; 58♀) | DB, placebo-controlled trial of OLZ (5-20 mg/day)+MS (Li+, VPA). Each group further divided in MS nonresponders vs. other (responders or those who were not exposed to MSs before randomization (MS-naïve). Outcome measures: YMRS score ↓, percentage of remission | YMRS ↓ more in the OLZ+MS than in the OLZ+placebo group. Previous response to VPA or Li+ does not affect results (YMRS ↓ in OLZ+VPA responders vs. OLZ+VPA-others: -14.7 vs. -14.8; YMRS ↓ in placebo+VPA responders vs. placebo+VPA-others: -8.8 vs. -8.0; YMRS ↓ in OLZ+Li+ responders vs. OLZ+Li+-others: -15.9 vs. -13.9; YMRS ↓ in placebo+ Li+-responders vs. placebo+Li+-others: -6.5 vs. -8.9; proportion of pts. who remitted in OLZ+VPA responders vs. OLZ+VPA-others: 61.9% vs. 60.6%; proportion of pts. who remitted in placebo+VPA responders vs. placebo+VPA-others: 40.0% vs. 34.8%; proportion of pts. who remitted in OLZ+Li+ responders vs. OLZ+Li+-others: 66.7% vs. 56.7%; YMRS score ↓ in placebo+Li+-responders vs. placebo+Li+-others: 33.3% vs. 36.8% | OLZ was superior to placebo in treating mania. This secondary analysis suggests that OLZ monotherapy is similarly effective for pts. whether or not they have previously failed to respond to another MS for mania |
| 44 | 36 BD depressive, aged 18-70. Two groups: TPX+MS+SGAs (N=18; =39 yrs; 11 ♂; 7 ♀;10 BD-I, 8 BD-II); and bupropion+MS+SGAs (N=18; =43 yrs; 10 ♂; 8 ♀; 9 BD-I, 9 BD-II) | 8-week, single blind add-on therapy with TPX (50-100 mg/day), or bupropion (100-400 mg/day) to MS (Li+: 980.0±388.3 mg; VPA: 1106.25±400.36 mg) and SGAs. Outcome measures: response rates, changes in YMRS, HDRS, and CGI-I | After 8 weeks, TPX+MS+SGA and bupropion+MS+SGA showed similar response rates (56% vs. 59%, respectively). Time to response ranged from 2 to 4 weeks. Both groups showed similar rates of ↓ from BL on the HDRS (20.5 to 10; 20 to 9.5 respectively), CGI-I, and YMRS (7 to 2; 8 to 2 respectively) | Both adjunctive TPX and bupropion were associated with reductions in depressive symptoms |
| 45 | 60 BD, depressive; two groups: Paroxetine+MS (N=30; =47.1±15.2 yrs; 11 ♂; 19 ♀;23 BD-I, 7 BD-II); Venlafaxine+MS (N=30; =45.5±13.7 yrs; 9 ♂; 21 ♀;21 BD-I, 9 BD-II); HDRS>17 | 6-week RCT of add-on paroxetine (=32.3mg ± 11.2 or venlafaxine (=32.3mg ± 11.2) to MS (Li, 0.7 mg/L; VPA, 50 µg/ml; CBZ, 4µg/ml). Outcome measures: response, remission, switch rates. | After 6 weeks: Groups did not differ in HDRS ↓ (paroxetine+MS: -6.9; venlafaxine+ MS: -9.0); similar proportions of responders (paroxetine+MS: 50%; venlafaxine+MS: 59%); similar remission rates (paroxetine+MS: 37%; venlafaxine+MS: 41%); similar switch rates (paroxetine+MS: 3%; venlafaxine+MS: 13%) | Both venlafaxine and paroxetine are effective add-on treatments to MSs for bipolar depression. Switch rates, especially during treatment with venlafaxine, raise some concerns |
| 46 | 318 BD aged 24-89 (=53.3±15.1; 41% ♂; 59% ♀) | Retrospective, naturalistic; Evaluation of anxiety comorbidity and response to MS (Li+ or AEs); remission criterion, no mood episodes for 2 years | 24% with anxiety comorbidity; anxiety comorbidity associated with poorer response to AEs. No differences in response to Li+ | Anxiety comorbidity ↓response to AEs |
| 47 | 150 BDI, manic or mixed, aged 19 – 65; two groups (MS + RISP, N=75, median=37y, 32♂; 43♀); MS + Placebo, N=75, median 42y, 31♂; 44♀)); median:37y for MS + RISP group, 42y for MS + placebo group; 32♂; 43♀ for MS + RISP group, 31♂; 44♀ for MS + placebo group | Randomized, double blind, placebo controlled. MS + RISP (4mg/day) or MS + placebo x 3 weeks. Outcome measures: changes in YMRS at day 8 and endpoint (last available observation), % subjects showing 50% YMRS improvement, time to response (30% YMRS score reduction), CGI 1/2 | Compared to BL, ↓ YMRS was significantly greater in the MS + RISP group (-10.2±1.1) than MS + Placebo group (-6.7±1.0) at week 1. At endpoint, 59% of pts. in the MS+ RISP group showed ≥50% ↓YMRS scores, compared with 41% in the MS + placebo group. 48% at week 1 and 61% at endpoint of the MS + RISP group scored 1 or 2 on the CGI, compared with 31% at week 1 and 43% at endpoint in the placebo group. Compared to BL, at week 1 and endpoint, BPRS ↓ was significantly greater in the MS+RISP group (-7.5±.09 and -10.1±1.1, respectively) than in the MS+placebo group (-3.8± 0.8 and -4.8± 1.1 respectively). At endpoint, the MS+RISP group had significantly greater improvement in the hostility and thought disturbance subscales of the BPRS than did the MS+placebo group |
RISP, in association with MS, is more efficacious than placebo in the improvement in manic symptoms. Improvement in manic symptom is also rapid |
| 48 | BD aged 18-65 with at least one manic episode requiring hospitalization; DB study: 156 BD (80 ♂; 76 ♀); Open: 85 BD (39 ♂; 46 ♀) | 3-week DB: MS+RISP, MS+placebo or MS+haloperidol; then 10-week open-label, add-on RISP; remission criterion, YMRS≤12 | Greater remission in RISP or haloperidol group compared to placebo; 79% of pts. in remission after 10-week open-label treatment with RISP | RISP+MS combination is efficacious in manic episodes requiring hospitalization |
| 49 | 22 BD with TRD. Two groups: pramipexole+MS (N=12; =40.9;±8.2%; 4♂; 8♀; 9 BD-I, 3BD-II; placebo+MS (N=10; =43.3±6.2; 7♂; 3♀; 6BD-I, 4BD-II); YMRS<12; HDRS >18 | 6-week, double-blind, placebo-controlled, single center trial of add-on pramipexole (1.0-2.5 mg/day) to MS (Li+ 1137.5±381.5 mg/day; VPA 916.7±129.1 mg/day; CBZ 400.0 ± 282.8 mg/day; LAM 283.3 ± 144.3 mg/day; GPT 450±212.1 mg/day). Outcome measures: ≥↓ 50% in HDRS scores; changes in CGI-S. | After 6 weeks, 67% of pts. on pramipexole+MS showed ≥↓ 50% on HDRS compared to 20% in the placebo+MS group. Mean change from baseline in HDRS was greater in pts. taking pramipexole+MS (48.0%±33.1) than for those taking placebo+MS (21.4%±36.3). Pramipexole+MS showed lower CGI-S (2.7±1.4) than placebo+MS (4.4±1.3). After 6 weeks, pramipexole+MS showed greater CGI-S score ↓ (-2.4±1.8) than placebo (-0.30±1.3) | Pramipexole is effective for TRD in BD. |
| 50 | 45 BD, depressed, aged 18-75 yrs (=42.2±11.5; 30♂; 15♀; 30 BD-II; 15 BD-II); HDRS ≥10; YMRS≤12; GAF≤70 | Open-label, add-on citalopram to MS (Li+, VPA, or CBZ). 8-week acute phase and for those who responded, 16 weeks continuation phase. Outcome measures: changes in YMRS and HDRS, occurrence of anger attacks (using modified AAQ) at 8 weeks; survival analyses at 8 and 16 weeks | At BL 38.6% have anger attacks, which dropped to 14.6% after 8 weeks. 73.3% of those having anger attacks at baseline did not have them after 8 weeks, whereas 8.7% of those who did not have anger attacks at baseline reported anger attacks after 8 weeks. Treatment response and rates of response to depression were unrelated. Survival analyses were similar between groups. Trait anger predicts anger attacks at BL and at week 8 | Anger attacks in BD respond favorably to add-on citalopram and are better predicted by trait anger than hypomanic or depressive symptoms |
| 51 | 99 BD-I aged 18-70, in remission from a manic/mixed episode; two groups: OLZ+MS (N=51; =43.5; 52.9%♂; 46.1%♀); Placebo+MS (N=48; =39.0; 43.8% ♂; 55.2% ♀); DSM-IV A and B criteria severity for current manic episode ≤3, no more than two B criteria, DSM-IV A criteria severity for current depression ≤3, no more than three A criteria | Double-masked placebo-controlled trial of add-on OLZ (5-20 mg) to MS (Li+, 954.6-1174.7 mg/day or VPA, 1060.4-1512 mg/day). Outcome measures: a) syndromic relapse: occurrence of DSM-IV manic, depressive, or mixed episode; b) symptomatic relapse: HDRS or YMRS≥15 | Time to symptomatic relapse into either mania or depression was significantly longer for the combination group compared with the monotherapy group (163 days for OLZ+MS; 42 days for Placebo+MS). Women and white patients with OLZ+MS showed longer times to symptomatic relapse than pts with Placebo+MS (84 vs. 67 days, respectively) | OLZ+MS reduced relapse in BD episodes |
| 52 | 909 BDI, manic or hypomanic, aged 16-60 (=35.1±13.7; 8♂; 10♀). YMRS>20 (for manics), YMRS>7 (for hypomanics) | Open label, add-on RISP to MS (Li+, VPA, CBZ, TPX), × 6 weeks; Outcome measures: reduction in YMRS, SARS, and CGI; response criterion (>50% reduction YMRS) | After 6 weeks, reduction of scores on the YMRS (from 32.9 ± 10.8 to 9.5 ±8.4) and CGI-severity (from 4.8 ± 1.1 to 2.1 ±.8) was significant. Response rates at week 6 were 70.7%. A higher reduction in the YMRS and CGI scores was found in the subgroup with psychotic features (24.2 ± 11.9; 2.9 ± 1.4 respectively) compared to the subgroup without psychotic features (22.6 ± 11.6; 2.5 ± 1.5 respectively) | RISP is effective for BD manic/hypomanic episode treatment |
| 53 | 18 BDI, manic or hypomanic (=35.1 ± 13.7; 8♂; 10♀) | Open label, add-on QTP (mean 267.9±105.4 mg/day) × 4 weeks. Outcome measures: ↓ in YMRS, HDRS, BPRS and CGI; response criterion (>50% ↓ on the YMRS) | After 4 weeks, YMRS ↓ from 28.2 ± 7.6 to 9.3 ± 5.7; HDRS ↓ from 2.7 ± 2.4 to .9 ± 1; BPRS ↓ from 32.8 ± 11.2 to 15.8 ± 11.6; CGI ↓ from 5±0.8 to 2.3 ± .7. Response rates at week 4 were 72.2% | Add-on QTP is an effective treatment for manic/hypomanic episode of BD |
| 54 | 59 UP-MDD, but scoring high on Angst's hypomania; good responder group: age=47; 17% ♂; 83% ♀; poor responder group: age=52; 19% ♂; 81% ♀ | Retrospective, naturalistic, comparison between good and poor responders to MS augmentation on ADs; remission criterion, clinical judgment of treating psychiatrists | 30% good responders, 70% poor responders; greater delay in prescribing MS augmentation and lower rate of MS in poor responders. No differences in hypomania and temperaments. Higher suicidal risk and agitation in poor responders | MS augmentation should be instituted without delay in patients with MDD meeting Angst's criteria for hypomania |
| 55 | 479 BD aged 18-65 (215 ♂; 264 ♀) | Retrospective, naturalistic; BD discharged on MS, MS+FGA, MS+SGA; relapse criterion, rate and time to rehospitalization | No differences between groups in rehospitalization time and rate (23% BD discharged on MS, 27% MS+FGA, 25% MS+SGA) | Augmentation with antipsychotics does not improve mood stability |
| 56 | 287 BD-I manic or mixed; aged 18-70 yrs; two groups: TPX+MS (=41.0±12.2; 58♂; 85♀); Placebo+MS (=39.0±11.9; 67♂; 77♀). YMRS ≥18 | 12-week, DB, placebo controlled, add-on TPX (50-400 mg/day) to MS (Li+, serum level 0.7 mEq/l, or VPA, serum level 70 µg/ml) and AP. Outcome measures: score change on YMRS, MADRS, CGI-S, BPRS, GAS | ↓ YMRS scores in both TPX (-10.1±8.7) and placebo (-9.6±8.2) groups, with no between-group differences. CGI, BPRS, MADRS and GAS improved, without between-group differences | Add-on treatment with TPX is not superior to placebo in ↓ manic/mixed episodes |
| 57 | 159 BD, depressive (=41.6±12.2; 83♂; 76♀; 115 BD-I, 42 BD-II). 228 acute AD trials, 111 AD continuation phase trials | Randomized, add-on bupropion (=286±132 mg/day), venlafaxine (=195±112 mg/day), and sertraline (286±132 mg/day) to MS (Li+, AE, AP). Acute phase: 10 weeks, those improved entered continuation phase (1 yr). If patients did not respond acutely to initial AD trial, they were randomly assigned to another AD. Outcome measures: response (CGI-BP=1 or 2; occurrence of 1) brief hypomania; 2) recurrent brief hypomania; 3) switch to full hypomania; 4) switch to mania | Acute phase: 48.7% of the trials reached response, that dropped to 32.5% after excluding those who had a switch. Switch rate to full (hypo)mania was 19.3%. 111 subjects reached sufficient response to enter the continuation phase. Continuation phase:67.8% trials showed AD response, but response rate in absence of a switch was 42.5%. 36.8% of trials switched to (hypo)mania. AD switch rates were not significantly different among the three ADs. In both acute and continuation phases, the threshold/subthreshold switch ratio was lowest with bupropion (acute: 0.85; continuation: 1.2), intermediate with sertraline (1.6 and 1.65, respectively) and higher with venlafaxine (3.6 and 3.75, respectively) | AD augmentation is not likely to yield a high rate of sustained AD response without a switch throughout both the acute and continuation treatment phases. Venlafaxine was associated with the highest relative risk of switch and bupropion with the lowest |
| 58 | 10 drug-naïve BD-II aged 18-65 with monthly mood episodes | Randomized, DB, escitalopram 10 mg vs placebo for 9 months | Reduction in depression severity, percentage of impaired days in the escitalopram group | Results support usefulness of SSRIs in BD-II treatment |
| 59 | 1127 BD-I after a recent manic or depressive episode (456 ♂; 671 ♀) | Open, LAM 100-200 mg/day+sedative/hypnotics vs LAM 100-200 mg/day+other psychotropics; stabilisation criterion, CGI≤3 for ≥4 weeks | Higher stabilization rates for LAM 100-200 mg/day+sedative/hypnotics vs other psychotropics | Adjunctive therapy with sedative/hypnotics may be useful in BD acute symptom control |
| 60 | 55 BD-I ( age=35±12.8, 34 ♂; 21 ♀) | Retrospective, naturalistic, SGA monotherapy vs. SGAs+MS for 6 months; relapse criterion, recurrence of mood episode | Clinical improvement in both groups, no differences in relapse | SGAs can be useful in the long-term management of BD-I |
| 61 | 10 BD-I outpts. (11–17 years) using a single MS and/or SGA, who had shown weight gain >5% of BL weight | Open. 11-week; medication switched to TPX during the first 4 weeks until 150 mg/day; YMRS main outcome to measure treatment response | Significant ↓in both YMRS score (F=10.21; p<0.01) and weight (F=8.04; p<0.01). weight loss=2.62 kg at endpoint. 6/7 completers (85%) did not show symptom worsening on the YMRS after 11 weeks. Significant BMI ↓ from BL to endpoint (p=0.017). No increase in adverse events | TPX seems to have antimanic effects during the treatment maintenance phase associated with weight reductions |
| 62 | 89 pregnant women ( age 32.7 years±5.4) BD-I (N=61) or BD-II (N=28) | Prospective observational study; Two groups based on MS status: 1) use of at least one MS at conception and continued ≥12 weeks of pregnancy; 2) MS discontinuation during 6 months before conception to 12 weeks of gestation. Follow up each trimester and at 6, 12, 24, and 52 weeks postpartum to ascertain recurrence of mania, hypomania (lasting ≥1 week), major depression, or a mixed state, and current treatments | During pregnancy, a total of 70.8% (63/89) of women experienced ≥1 episode of illness. Recurrence risk was 2.3 times greater after discontinuation of MS (53/62, 85.5%) than with continued treatment (10/27, 37.0%). Discontinuers spent >40% of pregnancy in an illness episode, vs. 8.8% of pregnancy of women continuing on MS. Median time to first recurrence was 9.0 (95% CI=8.0–13.0) weeks for discontinuers and >40 weeks (95% CI indeterminate) for continuers. Abrupt or rapid discontinuers (1–14 days; N=35) had 50% risk of recurrence within 2.0 (95% CI=1.0–6.0) weeks, gradual discontinuers (≥15 days, N=27) required 22.0 (95% CI=16.0–38.0) weeks to reach 50% recurrence risk (χ2= 25.9, df=1, p<0.0001). Excess of depressive-dysphoric polarity vs. manic-hypomanic episodes after discontinuation of MS (55/62 recurrences, 88.7%, versus 12/62, 19.3%, or 4.6-fold) compared to continued treatment (5/27, 18.5%, vs. 9/27, 33.3%, or 1.8-fold). Treatment-related risk factors, besides MS discontinuation, included: 1) polytherapy with two or more psychotropics (RR=2.3, p<0.001); 2) use of AD (RR=2.0, p<0.001); 3) primary MS other than Li+ (RR=1.6, p<0.001); 4) previous switch from depression to mania/hypomania during past AD treatment (RR=1.5, p<0.009); 5) abrupt MS discontinuation (RR=1.4, p=0.008). AD use and treatment discontinuation each operated independently as risk factors, even after adjusting for other indices of illness severity | Discontinuation during pregnancy of MS, particularly if abruptly, carries a high risk for new morbidity in women with BD, especially for early depressive and dysphoric states. However, this risk is reduced markedly by continued MS treatment |
| 63 | 232, BD, euthymic, aged 22-79 (=52.2 ± 9.7; 81 ♂; 158 ♀, BD-I N=91; BD-II N=141). 6 groups: QTP monotherapy (N=41); Li+ monotherapy (N=39); VPA monotherapy (N=73); LAM monotherapy (N=31); QTP + Li+ (N=25); QTP + VPA (N=23) | Naturalistic, 4-yr follow-up. Mean QTP doses: monotherapy, 214 mg/day; QTP+Li+, 223.5 mg/day; QTP+VPA, 237.4 mg/day; LAM 72.2 mg/day. Mean plasma levels for Li+ or VPA, 0.7 mEq/l ± 0.2 for Li+ monotherapy, 0.7 mEq/l ± 0.1 for Li++QTP, 52.1±17.2 ng/ml for VPA monotherapy, and 60.5 ng/ml ± 17.9 SD for VPA+QTP. Outcome measures: duration of euthymia/proportion of pts with no mood recurrences | After 4 years, pts. with Li++QTP and Li++VPA showed higher proportion of no mood episode (80% and 78.3%, respectively) than those with QTP alone (29.3%), Li+ alone (46.2%) and LAM alone (41.9%). Pts. with Li++QTP and Li++VPA did not relapse for longer times (41.4 and 39.2 months, respectively) than pts. on QTP (33.1 months) and VPA (30.1 months). Only pts. with Li++QTP did not relapse for longer times than Li+ alone (33.1 months). Pts. with Li+ were superior to those with QTP in proportion of subjects without relapses and time spent without relapse | QTP as either monotherapy or combination therapy (with Li+ or VPA) has been found to be effective in preventing both major and sub-threshold depressive episodes |
| 64 | 108 BD, euthymic, drug-free, (=52.2 ± 9.7; 43♂; 65 ♀, BD-I N=39; BD-II N=69). 3 groups: BD with early onset (<30y), middle onset (>30 and <45y), and late onset (>45y) | Naturalistic, 24-month follow-up. Drug free pts. received SGAs, Li+ or VPA. Outcome measure: relapse rates | After 24 months, depressive relapses were less in the early-onset group (=0.66) than middle- (=1.37) and late-onset (=1.26) | MS treatment seemed to be more effective in preventing depressive episodes in early-onset BD pts. compared to middle- and late-onset pts. Middle- and late-onset BD pts. were similar |
| 65 | 966 BD-I (open-label phase), with a recent depressive episode, aged ≥18 yrs.; 463 BD-I (randomization phase), aged ≥18 yrs. | First phase: open-label trial on LAM monotherapy (200-400 mg/day) for 16 weeks. Second phase: double-blind placebo- controlled trial on LAM monotherapy (200-400 mg/day). Outcome measures: occurrence of an intervention for manic/hypomanic/mixed symptoms, YMRS score ≥ 4, YMRS score ≥ 8, and YMRS score ≥ 14, survival analyses | Open-label phase: Compared to BL, YMRS ↑ by ≥14 points in 10% of pts., ↑ by ≥8 points in 20% of pts., by ≥4 points in 35% of pts. YMRS ↑ predicted by number of manic/hypomanic/mixed episodes in the preceding year. Randomized phase: no differences in % or HR of event occurrence between groups, LAM had consistently higher estimates of survival than placebo across all 4 thresholds of mania MRS scores at screening, and presence of ≥3 manic/hypomanic/mixed episodes in the preceding year significantly increased HR of reaching an event | LAM showed similar rates of manic relapse to placebo. During maintenance treatment, the likelihood of emergent manic or hypomanic features appears driven more by the pre-existing or historical burden of mania features, rather than the use of LAM |
| 66 | 109 BD-II, depressive, aged 18-65 yrs ( =40.3±11.5; 27♂; 82♀). CGI-BP-S depression score≥3 for >12 weeks on maintenance with Li+ or VPA+other medications (ADs and APs allowed) | Naturalistic, follow up (52 weeks) of add-on LAM (145.5±113.2 mg/day) to MS (Li+ or VPA) and APs (QTP, OLZ, ZIPR). Outcome measures: change in CGI-BP-S depression score | CGI-BP-S depression score ↓after 4 weeks of LAM add-on. Scores on the CGI-BP-S ↓ by about 1.8 points in the first 12 weeks and then remained stable. 49% completers. Completers and drop-outs differed for number of psychiatric hospitalizations (0.7±0.9 vs. 1.4±2.2) and history of suicide attempts (11.4% vs. 30.8%, respectively) | LAM is an effective add-on treatment for bipolar depression. Number of prior hospitalizations for depression and history of attempted suicide may be associated with poor response to adjunctive LAM treatment |
| 67 | 23 adult BD inpts. (12 ♂; 11 ♀; 17 BD-I, 6 BD-II), HDRS ≥20 | Randomized, DB, MS+citalopram 40 mg/day +ARP 10-30mg/day vs. MS+citalopram 40 mg/day+placebo for 6 weeks; remission criterion, HDRS ≤9 | No differences between the two groups | Augmentation with ARP does not improve mood stability |
| 68 | 50 BD, depressive pts., aged 18-70 (=439.7±10.3; 23 ♂; 31 ♀; 20 BD-I, 34 BD-II); MADRS>20 and YMRS<12 | 6 week – double-blind, placebo-controlled trial of VPA monotherapy (mean 1606±44 mg/day, range 1000–2000 mg/day). Outcome measure: improvement on MADRS, YMRS, CGI-BP, HARS. % pts. achieving response (≥50% ↓ in MADRS from BL) | VPA group improved more on the MADRS compared with placebo group at weeks 3, 4, 5, and 6 (mean change of MADRS total score for VPA over placebo=4.32). This is mainly driven by differences in the BDI subgroup. VPA group responded for 38.1% and remitted for 23.1%, placebo group improved and remitted by 10.7% | VPA is effective in treating bipolar depression in the BD-I subset |
| 69 | 139 BD-I with manic or mixed episode after 6-week olanzapine vs. haloperidol vs. placebo DB trial. Monotherapy group (N=100): (=41.8; 41% ♂; 59% ♀); Combination group (N=39): (=43.2; 46% ♂; 54% ♀) | Open, 56-site (Japan) study; olanzapine 5-20 mg/day monotherapy switch from previous trial × 18 weeks, if lack of efficacy: olanzapine + MS (Li+, CBZ or VPA); safety assed by treatment-emergent adverse events; Remission criterion, YMRS≤12 | Monotherapy group: 59% treatment-emergent adverse events, remission 93%; Combination group: treatment-emergent adverse events 79.5%, remission 61.5% | Results support efficacy of combination therapy of olanzapine + MS if olanzapine monotherapy lacks of efficacy |
| 70 | 40 BD aged 24-84 yrs (=49±16; 10 ♂; 30 ♀; 21 BD-I, 19 BD-II) CGI-BP≥5 | Open, add-on memantine 10-30 mg/day × 12 months; response/remission criterion, CGI-BP 1 or 2 | After 6 months, 47.5% scored 1 and 25% 2; after 12 months, 52.5% 1 and 20% 2 | Memantine helps overcoming resistance if added on ongoing treatment |
| 71 | 13 BD, depressive, aged ≥18 yrs (6 ♂; 7 ♀; 12 BD-I, 1 BD-II). HDRS>15, YMRS>12. | 8-week, open-label, add-on nefazodone 300-600 mg/day+MS or AP (Li+ 450- 600 mg/day; LAM 400 mg/day; VPA, 750-1250 mg/day; CBZ 400-600 mg/day, CLZ, 325 mg/day). Outcome measures: changes in HDRS and CGI-BP, remission, response | 69% responded after 8 weeks, 31% remitted. HRDS ↓ from 26.1 ± 5.1 at BL to 18.5 ± 10.1 at week 8; CGI-BP ↓ from 24.2± 0.6 at BL to 3.4 ± 1.3 at week 8, both significant | The effectiveness of add-on nefazodone therapy is moderate |
| 72 | 83 BD-I aged 18-65 (31 ♂; 52 ♀) with sleep disturbances, ≥5 Pittsburgh Sleep Quality Index |
Randomized, DB, add-on ramelteon 8 mg × 24 weeks vs. placebo; relapse criterion, MADRS score ≥16 and/or YMRS≥15 or need of drug treatment | 48.2% relapse. Ramelteon group less likely to relapse | Ramelteon has a potential utility in maintain mood stability |
| 73 | 7423 PBD (N 3131 aged 6–12; N 4292 aged 13–18). Age 12.73±3.38 on either MS (2479) or SGA (4944) | Retrospective cohort study. The outcome measures were psychiatric hospital admission, all cause medication discontinuation and treatment augmentation | Pts. who initiated on MS and SGA had comparable risk of psychiatric hospital admission (HR = 1.172, 95%CI: 0.827–1.660). Compared with those who initiated on MS, pts. who initiated on SGA were less likely to discontinue treatment (HR=0.634, 95% CI: 0.419–0.961) and less likely to receive treatment augmentation (HR=0.223, 95%CI: 0.103–0.484) | SGAs might be more effective and better tolerated than traditional MSs in BD maintenance treatment |
| 74 | 23 euthymic BD (Age =48.13 yrs.±14.73; 16 ♂; 7 ♀) | Naturalistic, follow up. Outcome measures: mean time to recurrence after MS discontinuation | Median time of recurrence (all manic relapses) is 10 months. Total number of episodes and number of manic episodes negatively correlate with time to recurrence | Rates of relapse in the Indian population are similar to those present in western countries |
| 75 | 3240 BD (1270 ♂; 1970 ♀) with no AD treatment during the previous year | Retrospective, naturalistic; AD monotherapy vs. AD+MS; switch to mania criterion, rate of mania 0-3 months, 3-9 months | ↑ risk of switch was confined to pts. on AD monotherapy | AD monotherapy is associated with ↑ risk of mania |
| 76 | 180 BD-I aged 18-65 (59♂; 121 ♀) in acute manic episode | Randomized, DB, add-on allopurinol 300 mg/die vs placebo × 6 weeks; response criterion, ≥50% ↓ in YMRS | No differences between groups in response | Results do not support add-on allopurinol as a treatment for acute mania |
| 77 | 59 BD-II > 18 years old ( =4±12.5; 27 ♂; 32 ♀;) responded to treatment, HDRS ≤ 16 | Randomized, DB, venlafaxine or Li+ monotherapy × 12 weeks, 6 additional months to evaluate relapse; response criterion, ≥50% ↓ in HDRS; relapse criterion, HDRS≥14+CGI≥4 for ≥14 days | 67.7% venlafaxine versus 34.4% lithium subjects responded; no difference in relapse between treatment conditions during continuation monotherapy | Continuation venlafaxine and Li+ monotherapies provide similar prophylactic effectiveness |
| 78 | 201 adult BD-I hospitalized for a manic episode (113 ♂; 88 ♀) | Retrospective, naturalistic; MS monotherapy, MS+SGA, MS+FGA; response criterion, 1-year rehospitalization | 1-year rehospitalization rates lower in MS+SGA group (6.3%) compared to MS monotherapy group (24.3%) and to MS+FGA group (20.6%) | Results support efficacy of atypical antipsychotic adjunctive therapy to MS |
| 79 | Drug naïve BD aged 18-65 ( age= 38.8 yrs.; 56 ♂; 12 ♀) with mixed depression | Open, CBZ, Li+ or VPA monotherapy × 8 weeks; response criterion, ↓ ≥50% in HDRS+one mania scale (YMRS, CARS-M, BRMaS) | High agreement between the three mania rating scales; response on HDRS+YMRS=22.1%, on HDRS+BRMaS=20.6%, and on HDRS+CARS-M=23.5% | Results support the use of any scale to assess the efficacy of MSs in mixed depression; overall response is about 20% |
| 80 | 344 BDI, depressive, aged 17-70 (=45.2±12.6; 78 ♂; 122 ♀). HDRS >18 and YMRS>8 | 8 week-DB, 67-cent3r (15 countries) RCT of add-on agomelatine (25-50 mg) on VPA or Li+. Additional 10-month continuation phase. Outcome measures: MADRS improvement; response criteria (MADRS improvement ≥50%); changes in HDRS, HARS, CGI, LSEQ, QLESQ | No differences. Results became significant (greater improvement in the agomelatine group vs. placebo) after excluding sites with pts. showing high placebo response | Agomelatine added on Li+ or VPA MS is ineffective for bipolar depression. Concerns for patient recruitment in some centers |
| 81 | 159 BD-I aged ≥17 yrs. (=37.9±13.49; 79 ♂; 79 ♀). Patients remitted from recent manic episode on Li+ or VPA and RISP or OLZ add-on | Patients randomized to placebo substitution of RISP or OLZ at week 0 (N=52), after 25 weeks (N=54), and no substitution for 52 weeks (N=53) (endpoint). Outcome measures: Event rates (occurrence of any mood episode), time of any mood episode | After 52 weeks, event rate was higher in the 0-week group than in the 24- and 52-week groups. Time to any mood episode was longer in the 24- and 52-week groups compared to the 0-week group. No differences were found between 24- and 52-week groups. Time to mood episode was unchanged after considering subgroups taking OLZ. Instead, in those taking RISP, time to any mood episode in the 52-week group was similar to that in the 0-week group and shorter than in the 24-week group | Use of adjunctive SGAs to MS is beneficial for 24 weeks. However, these benefits are not apparent over 24 weeks. Relapse prevention by OLZ lasts longer than the one provided by RISP |
| 82 | 80 BD aged 18-55 (23 ♂; 57♀; 65 BD-I, 15 BD-II) | Retrospective, naturalistic, treatment with Li+ and/or VPA for more than 2 years; response criterion, total Alda Scale score>5 | 34% good responders; no differences between Li+ and VPA groups. In the Li++VPA group, psychotic, mixed, and atypical features associated with poorer response | Li+ and VPA show similar efficacy. Polypharmacy is associated with poor response |
| 83 | 32 BD with subthreshold symptoms, aged 18-65 (=43.75±10.1; 12 ♂; 20 ♀; 21 BD-I, 11 BD-II). YMRS< 14 and/or MADRS>8 and <14 | 12-week, double blind-placebo controlled trial of QTP-XR, 300-600 mg, in addition to MS (Li+, VPA, LAM). Outcome measures: MADRS, HDRS-5, YMRS, CGI-BD score changes. Rates of remission, and early response: % of patients with HDRS<8 and YMRS<8, level of functioning (as assessed by FAST, GAF, EQ-5D, TOOL) and functional remission | The mean changes in MADRS total score from BL to week 6 were -2.44 in the QTP-XR and +2.50 in the placebo group. Changes in HDRS-5 at week 6 were -1.44 in the QTP-XR and +0.28 in the placebo group. At week 12, the QTP-XR group scored higher on the FAST-autonomy subdomain | QTP-XR is effective in treating subthreshold depressive symptoms |
| 84 | 243 BDI, manic, aged 18-83 (=49.1±13.7; 107 ♂; 137 ♀) | Naturalistic, 34-center (Italy) 12-week follow-up study. Pts. starting/switching to AP and/or MS. Outcome measures: predictive factors of remission (YMRS ↓ ≥50%) and changes from baseline on YMRS, MADRS, FAST, CGI-BP | After 12 weeks, remission rate was 82.3%. No variables found to associate with remission. After 12 weeks, YMRS change was -22.0 ± 10.7. BL CGI-BP depression weakly predicted YMRS change. MADRS change was -6.1±8.2. BL YMRS weakly predicted MADRS change. CGI-mania and CGI-total score change was -6.1±8.2 and -2.7±1.6, respectively. BL YMRS scores weakly predicted CGI-mania and CGI-total score change. Mean FAST change was -17.4±17.3. BL YMRS scores predicted FAST score change | The initiated/changed pharmacological treatment for mania was associated with rapid improvement in manic symptoms and functioning. In contrast, the study has not clearly shown the association of any of the examined intrinsic and extrinsic factors with remission and clinical improvement |
| 85 | 273 BD aged >18 (=40.8±11.07; 81 ♂; 192 ♀; 173 BD-I, 100 BD-II). Two trajectory-based groups: adherent (N=210) and nonadherent (N=63) | Naturalistic, 12-week follow-up. Outcome measures: factors differentiating adherent vs nonadherent group | The nonadherent group spent less time in euthymia (47%) than the adherent group (67.8%). Women more represented in the nonadherent (82.5%) than in the adherent group (66.7%) | Characteristics associated with belonging to the less adherent class were more time with symptoms (i.e., not euthymic), and female gender |
| 86 | 413 youth BD aged 7-17.11; N=886 Li+, N= 1.752 MS | Naturalistic, longitudinal study. Data from the Course and Outcome of Bipolar Youth (COBY) study. Follow-up every 6 months over a mean follow-up of 10 years. “Lithium blocks”: Li+ for more than 75% of the follow-up weeks, regardless of other medications. “MS blocks”: MS but not Li+, for more than 75% of the time, i.e., antimanic AEs, FGAs and SGAs, and/or LAM. Clinical outcomes: suicide attempts and suicidal ideation; threshold and subthreshold depression; threshold and subthreshold (hypo)mania; psychosocial functioning; hospitalization; aggression; and SUD | During Li+ (vs. MS) follow-up periods, pts. were older, less likely to have lifetime anxiety, and less likely to be on AD (p<0.005). After covariate adjustment, the Li+ group (vs. MS) had half as many suicide attempts (p=0.03), fewer depressive symptoms (p=0.004), less psychosocial impairment (p=0.003), and less aggression (p=0.0004) | Li+ is associated with decreased suicidality, less depression, and better psychosocial functioning than MSs in a population of youths with BD |
| 87 | 91 BD aged 18-70 ( age=41.6±14.2; 47 ♂; 44♀; 63 BD-I, 28 BD-II) | Retrospective, naturalistic, 2-site Campanian study; impaired glucose metabolism (N=49) vs. normal glucose metabolism (N=42); response to MS criterion, Alda scale score ≥7 | Impaired glucose metabolism predicted low treatment response to MSs; in the impaired glucose metabolism sample, good response 4.4%, moderate response 23.1%, and poor response 22%; in the normal glucose metabolism sample, good response 17.6% (p=0.012), moderate response 23.1% (n.s.), and poor response 9.9% (p=0.001) | BD with impaired glucose metabolism are at risk for a poor response to MSs; poor metabolism predicts significantly less good response |
| 88 | 16224 BD aged <65. Three treatment strategies: 6775 pts. (41.8%) MS (Li+, VPA, valpromide, CBZ, LAM) without SGA. 7268 pts. (44,8%) SGA (ARP, OLZ, RISP, QTP) without MS; 2181 pts. (13,4%) combination of MS+SGA | Historic cohort study using French national healthcare databases. Outcomes: treatment discontinuation, switch or addition, psychiatric hospitalization, suicide attempt, and death | The 1-year adjusted cumulative incidence of treatment failure was 75.7% (95%CI 74.9;76.3) in pts. using MS, 75.3% (74.6;76.0) in pts. using SGA, and 60.5% (58.3;62.6) in pts. with the combination. The adjusted difference in incidence of treatment failure for SGA compared with MS was -0.40% (-1.4;0.6 p=0.4) in the whole population, -2.2% (-3.3; -1.2 p<0.002) in pts. <65 years. Early discontinuation was the most often first occurring event over follow-up for all three treatment strategies. Treatment addition and discontinuation were slightly less frequent with SGA than with MS, while psychiatric hospitalization occurred more often in the SGA group. The 1-year adjusted cumulative incidence of partial discontinuation was 59.6% (95%CI 57.3;61.7) in pts. with combined treatment. The incidence of mortality was quite high as first event, particularly in the group receiving combinations. Treatment failure occurred in 64.8% (95%CI 64.0–65.6) of pts. with MS, 67.4% (95%CI 66.7;68.2) of pts. with SGAs, and 50.8% (95%CI 48.6;53.0) of pts. with a combination. The adjusted cumulative incidence of early treatment discontinuation was 37.0% (95%CI 36.1;37.8) with MS, 38% (95%CI 37.3;38) with SGAs, and 28.4% (95%CI 26.4;30.2) with a combination. Discontinuation was significantly more frequent with SGAs than with MSs with an adjusted difference in cumulative incidence=2.6% (95%CI 1.5;3.7; p<0.002) | The rate of treatment failure is very high in all age groups and for all treatment strategies. SGAs did not perform better than MSs in the whole study population and were even worse in the sensitivity analyses. SGAs are slightly more effective than MSs in younger pts., but less effective in older ones |
| 88 | 3862 BD aged 65 and over. Three treatment strategies: MS: N 1450 (37.5%) SGA: N 2074 (53.7%) MS+SGA: N 338 (8.8%) | Historic cohort using French national healthcare databases. Outcomes: treatment discontinuation, switch or addition, psychiatric hospitalization, suicide attempt, and death | The adjusted difference in incidence of treatment failure for SGA compared with MS was +6.7% (4.1;9.1 p<0.002). Treatment addition was less frequent with SGA than with MS, while early discontinuation, psychiatric hospitalizations, and death were more frequent with SGA. When considering each type of outcome separately without stopping the follow-up when another type of outcome occurred, pts. with SGA (with or without MS) were at higher risk of death | SGAs are less effective in older pts. and fail more often than MSs in older pts. Mortality was particularly high in older pts. treated with SGA or a combination |