Acute and Maintenance

Treatment of Bipolar
Depression

Terence A. Ketter, M.D.

 Teaching Points
Mood stabilizers are foundational agents and should be
considered first line treatments, with the strongest
evidence supporting the use of lithium and lamotrigine.

Emerging data suggest atypical antipsychotics provide

benefit in acute bipolar depression, with the strongest
evidence supporting the use of quetiapine monotherapy
and the olanzapine plus fluoxetine combination.

The utility of adjunctive antidepressants in bipolar

depression is controversial, as these agents can yield
switching into mania or hypomania in some patients.
 Pre-Lecture Exam
Question 1

1. The most pervasive symptoms in bipolar disorder are

those of: (choose one)

A. Mania, hypomania

B. Hypomania

C. Depression

D. Mixed States

E. None of the above

 Question 2

Which of the treatments below is the LEAST appropriate strategy in

bipolar depression: (choose one)

A. Mood stabilizer without antidepressant

B. Mood stabilizer with antidepressant

C. Atypical antipsychotic with antidepressant

D. Antidepressant with neither mood stabilizer nor atypical antipsychotic

 Question 3

Which antidepressant option carries the greatest risk of

hypomania/mania: (choose one)

A. Tricyclic antidepressants (TCAs)

B. Selective serotonin reuptake inhibitors (SSRIs)

C. Mirtazepine

D. Bupropion
 Question 4
Which of the following treatments do NOT have controlled data
suggesting utility in bipolar depression: (choose one)

A. Lithium

B. Lamotrigine

C. Olanzapine plus fluoxetine combination

D. Quetiapine

E. Citalopram

F. Pramipexole
 Question 5
Which of the following statements best describes the role of
maintenance adjunctive antidepressants in patients with
bipolar disorder: (choose one)

A. Long-term adjunctive antidepressants are always beneficial.

B. Long-term adjunctive antidepressants are never beneficial.

C. Long-term adjunctive antidepressants are beneficial in most

patients.

D. Long-term adjunctive antidepressants may be beneficial in

some patients.
 Overview

 Treatment options

– Mood stabilizers

– Atypical antipsychotics

– Adjunctive antidepressants

– Alternative treatments

 Treatment of acute bipolar depression

 Prevention of bipolar depression

 Bipolar disorders symptoms
are chronic and predominantly depressive
146 Bipolar I Patients 86 Bipolar II Patients
followed 12.8 yrs followed 13.4 yrs
1% 2%

6%
9%

53% 50% 46%

32%
% of Weeks
Asymptomatic
Depressed
Hypomanic
Judd et al 2002 Cycling / mixed Judd et al 2003
 Treatment Options in Bipolar Depression
Mood Stabilizers Adjunctive Alternative Treatments
Lithium Antidepressants Pramipexole
Lamotrigine Fluoxetine + Olanzapine Gabapentin
Carbamazepine Bupropion Omega-3 fatty acids
Divalproex SSRIs Phototherapy
ECT Venlafaxine Psychotherapy
Nefazodone Sleep deprivation
Atypical Antipsychotics Mirtazapine Thyroid hormones
Quetiapine MAOIs
Olanzapine TCAs

Jefferson JW, Greist JH. Textbook of Psychiatry, Washington, DC, American Psychiatric Press, 1994; Post RM, et al.
Neuropsychopharmacology 1998; Worthington JJ III, Pollack MH. Am J Psychiatry 1996; Amsterdam J. J Clin
Psychopharmacol 1998; Barbini B, et al. Psychiatry Res 1998; Wirz-Justice A, et al. Biol Psychiatry 1999; Stoll AL, et al. Arch
Gen Psychiatry 1999; Bowden CL. J Clin Psychiatry 1998; Tohen M, et al. Arch Gen Psychiatry 2003;60:1079-88; Calabrese
JR, et al. J Clin Psychiatry 1999;60:79-88; Goldberg JF, et al. Am J Psychiatry 2004;161:564-6.
Acute Treatment of Bipolar
Depression
 Lithium in Acute Bipolar Depression

 Li > placebo in 5/7 studies (N=158)1

– Pooled data
 19% little or no antidepressant effect
 81% significant antidepressant effect

 Li versus TCA studies1,2

– Some included unipolars

– TCA  Li in 3 studies (N=98)1,2

Mendels J. Am J Psychiatry 1976;133:373-8 1 Watanabe S, et al. Arch Gen Psychiatry 1975;32:659-668 2

 Lithium and Suicide Risk in
Major Affective Disorder

28 Reports* (16,800 Patients)

No. of Annual risk

reports of suicide
7 to 8-fold
}
With lithium 22 0.26 ± 0.4
difference
Without lithium 10 1.68 ± 1.5 p<0.0001

*19 of 28 reports (16,000 patients) recorded only actual suicides.

Tondo, et al. 1997.
 Suicide and Suicide Attempts
with Randomized Lithium or Carbamazepine
30-month prospective study
in 285 recently hospitalized patients
(175 bipolar, 110 schizoaffective)

Suicide Suicide Total Suicidal

Attempts Behavior

Lithium 0 0 0

Carbamazepine 5 4 9

Thies-Flechtner et al. Pharmacopsychiatry 1994;29:103-7.

 Mood Stabilizer Choice and Suicide Events in
Bipolar Disorder Patients in Two Large HMOs

Events per 1,000 pt-years

Medication # of Outpatient Inpatient Completed
PtÕs Attempts Attempts Suicides
Lithium 11,308 9.5 4.3 0.7

Divalproex 12,358 26.8* 10.65* 1.75*

Lithium + 3067 25.8* 11.8* 1.60

Divalproexa
a
Treatment-resistant patients; *Sig. Diff from Lithium alone (p<.05)

Goodwin et al. JAMA 2003;290:1467-73

 Mood Stabilizer Choice and Suicide Events in
Bipolar Disorder Patients in Two Large HMOs
Risk ratios of events relative to patients on lithium
(Adjusted for age, sex, year of treatment, comedications, comorbidity)
Medication Outpatient Inpatient Completed
attempts attempts Suicides
Lithium 1.0 1.0 1.0

Divalproex 1.7* 1.6* 2.6**

Divalproex 2.1* 2.1* 2.6

a
+ Lithium
a
Treatment-resistant patients; Sig. Diff from Lithium alone (*p<.001; **p<.004)
Goodwin et al. JAMA 2003;290:1467-73
 8-Week Randomized Double-Blind Divalproex
Monotherapy in Acute Bipolar Depression
Mean Change from Baseline

-2

-4
Placebo (N=22)
-6 ±
±
-8

-10
Baseline - PBO 20.5, DVPX 22.6 Divalproex 62 ug/mL (N=21)
-12
0 1 2 3 4 5 6 8
Week ±
p < 0.1 vs PBO, LOCF

Sachs G, et al. 40th Ann ACNP Mtg, December 9-13, 2001, Waikaloa, HI.
 8-Week Randomized Double-Blind Divalproex
Monotherapy in Acute Bipolar Depression
0
From Baseline (LOCF)
Mean HAM-D Change

-2

-4
Placebo (N = 12)
-6

-8
P = 0.0002

-10
Divalproex 82 ug/mL (N = 13)
-12
0 1 2 3 4 5 6 7 8
Week
Baseline HAM-D: Placebo, 19.9; Divalproex 22.0. Last observation carried forward.
Davis LL, et al. J Affective Disord 2005;85:259-66.
Summary of 4 Acute Bipolar Depression Studies
Response Rates

Active-Placebo Response Rate Difference Placebo Response Rate

60%
(≥ 50% decrease in depression rating)

50%
Response Rate

40%

30% 36 36 29 25 29 35 35 25

20%

10%

0% 22 22 25 24 19 11 4 8
QTP Q TP L TG OFC L TG L i Pax L i IM I Olz
600mg 300mg 200mg 50mg
Sachs GS. In Ketter TA (ed). Advances in the Treatment of Bipolar Disorders. Am Psychiatric Press, Inc. 2005.
 7-Week Randomized Double-Blind Lamotrigine
Monotherapy in Acute Bipolar I Depression
Last Observation Carried Forward Observed Cases
0 0
Change From Baseline

-5 -5
Placebo (n = 65)
MADRS

Placebo
-10 LTG 50 mg/d (n = 64) -10
±
*
Switch Rates ± † * LTG 50 mg/d
* * *
* *
-15 LTG 200 mg/d (n = *63) -15
PBO 5%
* *
LTG 50 3% * * *
LTG 200 8% LTG 200 mg/d
-20 -20
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Week Week
±
P<0.1; † * P<0.05.
Calabrese et al. J Clin Psychiatry. 1999;60:79-88.
 16-Week Randomized Open Adjunctive Therapy of
Treatment Resistant Bipolar Depression a

23.8%
[5.8-41.8]

17.4%
[2.4-32.4]

4.6%
Switch Rates [0-14.6]

Lamotrigine 19% 138 9429 1.5

mg/d mg/d mg/d
Inositol 13%

Risperidone 13%

Nierenberg AA, et al. Am J Psychiatry 2006;163;210-6. a

54% BPI, 46% BPII.
 8-Week Randomized Double-Blind Olanzapine ±
Fluoxetine in Acute Bipolar I Depression
0
Mean Change in MADRS Scores

Fluoxetine monotherapy arm excluded

due to risk of mania induction. Switch Rates

PBO 7%
-5 *
OLZ 6%
* OFC 6%
*
-10 * *
* PBO (N = 355)

-15 OLZ 9.7 mg

(N = 351)
OLZ 7.4 mg
†
† † + FLX 39.3 mg
-20
(N = 82)
0 1 2 3 4 6 8
Week
Baseline MADRS 31.3 PBO, 32.6 OLZ, 30.8 OLZ+FLX. Tohen M, et al. Arch Gen Psychiatry 2003;60:1079-88.
* P < 0.05 vs OLN, OLN+FLX. † P < 0.05 vs OLN.
 8-Week Randomized Double-Blind Olanzapine ±
Fluoxetine in Acute Bipolar I Depression
Responders Remitters
60 54% †

50 47% †
Percentage of Patients

40 37%
31%
29% *
30
23% *
Switch Rates
20
PBO 7%
10 OLZ 6%
OFC 6%
0
OFC OLN PBO OFC OLN PBO

* P < 0.05 vs OLN, OLN+FLX. † P < 0.05 vs OLN. ITT-LOCF Tohen M, et al. Arch Gen Psychiatry 2003;60:1079-88.
 7-Week Randomized Double-Blind Lamotrigine vs
Olanzapine + Fluoxetine in Acute Bipolar I Depression
Week
0 1 2 3 4 5 6 7 Switch Rates
0 LTG 5%
Mean Change in MADRS Scores

OFC 4%
-5
Weight Change (kg)

LTG -0.3***
-10
OFC +3.1
*
-15 LTG 106 mg
* (N = 205)
-20 * OLZ 10.7 mg
* * + FLX 38.3 mg
(N = 205)
-25
Baseline MADRS 30.9 OFC, 31.4 LTG. *P < 0.05, ***P < 0.001 OFC vs LTG. Trade-off: 3 lbs/MADRS point.
Brown EB, et al. J Clin Psychiatry 2006;66:1025-33.
 7-Week Randomized Double-Blind Lamotrigine vs
Olanzapine + Fluoxetine in Acute Bipolar I Depression
Responders ≥ 7% Weight Gain
70 69%
± 60%
60
Percentage of Patients

50

40

30 23%

20 Switch Rates

LTG 5%
10
OFC 4%
***
0%
0
OFC LTG OFC LTG
±
P < 0.08, *** P < 0.001 OFC vs LTG. Trade-off: 9% response vs 23% weight gain.
Brown EB, et al. J Clin Psychiatry 2006;66:1025-33.
 8-Week Randomized Double-Blind Quetiapine
Monotherapy in Acute Bipolar Depression
Study Week
0 1 2 3 4 5 6 7 8
0
Change From Baseline

-5
(LS Means)

† Placebo (N = 169)
-10 †
†
† † Quetiapine 300 mg (N = 172)
Switch Rates †
†
-15 PBO 4% † † † †
QTP 300 4% †
† † † †
QTP 600 2%
-20
Quetiapine 600 mg (N = 170)
Baseline MADRS 30.3 PBO, 30.4 QTP 300, 30.6 QTP 600.
†P<0.001 (quetiapine vs placebo) ITT, LOCF
Calabrese JR, et al. Am J Psychiatry 2005;162:1351-60.
8-Week Randomized Double-Blind Quetiapine
Monotherapy in Acute Bipolar Depression
Depression Rating Scale
Montgomery-Asberg

Implrovement

Switch Rates
PBO 7%
QTP 300 2%
QTP 600 4%

Baseline MADRS 29.6 PBO, 31.1 QTP 300, 29.9 QTP 600.
*P<0.01, †P<0.001 (quetiapine vs placebo). ITT, LOCF
Thase ME, et al. J Clin Psychopharmacol 2006;26:600-9.
 8-Week Randomized Double-Blind Quetiapine
Monotherapy in Acute Bipolar Depression
Response Rates
BOLDER I BOLDER II
Switch Rates
*** ***
Percentage of Patients Responding

60 58% 58%
PBO 4%
Switch Rates
(≥ 50% MADRS Decrease)

QTP 300 4%
50
QTP 600 2% ** PBO 7%

40
40% *
37% QTP 300 2%
36%
QTP 600 4%
30
24%
20

10

0
QTP QTP PBO QTP QTP PBO
300 600 300 600
Calabrese JR, et al. Am J Psychiatry 2005;162:1351-60. Thase ME, et al. J Clin Psychopharmacol 2006;26:600-9.
*p < 0.05, **p< 0.01, *** p < 0.001 vs placebo.
 BOLDER I and II: MADRS Total Score
Bipolar I vs. II Disorder
Bipolar Disorder I Bipolar Disorder II
(N=657) (N=321)
Change From Baseline
0
MADRS LS Mean

-4
Improvement

-8

-12

-16 Quetiapine 300

† Quetiapine 600
†
-20 Placebo
‡ ‡

p<0.01; ‡p<0.001 vs. placebo (N at baseline); ITT = intent to treat; AstraZeneca (data on file); Thase
†

ME (2006), Presented at the 159th Annual Meeting of the APA. Toronto, Canada; May 20-25;
Calabrese JE et al. (2005), Am J Psychiatry 162(7):1351-1360
 Magnitudes of Effects in Controlled Trials in
Acute Bipolar I Depression
Effect Size Response Rate (%)a
0.9 70

0.8
60
0.7
50
0.6

0.5 40

0.4 30
0.3
20
0.2 OLZ OLZ
7.5 mg 7.5 mg
+ 10 +
0.1 OLZ FLX QTP QTP LTG LTG OLZ FLX QTP QTP LTG LTG
10 mg 40 mg 300 mg 600 mg 50 mg 200 mg 10 mg 40 mg 300 mg 600 mg 50 mg 200 mg
0 0
Tohen 2003.
1
Calabrese 2004.
2 3
Calabrese 1999. Tohen 2003.
1 2
Calabrese 2004. 3
Calabrese 1999.

Effect Size (ES) = (improvement over PBO) / (pooled SD)(small <0.4; mod 0.5-0.9; large >1.0). a
>50% MADRS decrease

Tohen M, et al. Arch Gen Psychiatry 2003;60:1079-1088; 2Calabrese JR, et al. 157th APA Annual Meeting, May 1-6, 2004,
1

New York, NY. Abstract NR756. Page 284; 3Calabrese JR, et al. J Clin Psychiatry 1999;60:79-88.
6-week Randomized Double-Blind Adjunctive
Pramipexole in Acute Bipolar Depression
67% Response Rates
(8/12) 60%
70 (6/10)
(≥ 50% HDRS/MADRS decrease)

60 15 BPI, 7 BPII on 21 BPII on

Li (N=6, 0.7 mEq/L), Li (N=12, 0.8 mEq/L),
50 DVPX (N = 9, 81 ug/mL), DVPX (N=9, 73 ug/mL)
Percent responders

45 LTG (N=6),
GBP (N=3),
40 CBZ (N=2)
P<0.02
35
30 P<0.04
20%
25 (2/10)
20
9%
15
(1/11)
10
1.7 1.7
5 mg/d mg/d
0
Pramipexole Placebo Pramipexole Placebo

Goldberg JF, et al. Zarate CA, et al.

Am J Psychiatry 2004; 161:564-6 Biol Psychiatry 2004; 56:54-60.
6-week Randomized Double-Blind Adjunctive
Pramipexole in Acute Bipolar Depression
Switch Rates
70
Percent with Hypomania or Mania

60 15 BPI, 7 BPII on 21 BPII on

Li (N=6, 0.7 mEq/L), Li (N=12, 0.8 mEq/L),
50 DVPX (N = 9, 81 ug/mL), DVPX (N=9, 73 ug/mL)
45 LTG (N=6),
GBP (N=3),
40 CBZ (N=2)
35
30
18%
25
(2/11 hypomania)
20
8% 10%
15 (1/10 hypomania)
(1/12 mania)
10 0%
1.7 1.7
5 mg/d (0/10) mg/d
0
Pramipexole Placebo Pramipexole Placebo

Goldberg JF, et al. Zarate CA, et al.

Am J Psychiatry 2004; 161:564-6 Biol Psychiatry 2004; 56:54-60.
 6-week Randomized Double-Blind Adjunctive
Modafinil in Acute Bipolar Depression
60 Response Rates

50 *
44%
(≥ 50% IDS decrease)
Percent Responders

40

30
22%
20
Switch Rates
10 Placebo 11.4%
177 mg/d Modafinil 4.9%
N = 41 N = 44
0
Modafinil Placebo
*p < 0.05 vs placebo.
Frye M, et al. Am J Psychiatry 2007;164:1242-9.
Response in Randomized Controlled Trials
of Antidepressants vs. Placebo in Bipolar Depression

Gijsman et al, American Journal of Psychiatry. 2004;161:1537-1547.

 Paroxetine, Imipramine, Placebo Added to
Lithium in Bipolar Depression
Li + PBO (n=43) Li + PXT 33 mg/d (n=35) Li + IMI 167 mg/d (n=39)
60
*
50

40
*
% Responders

30

20

10
Switch Rates
0
Li + PBO 2%
All Subjects Lithium < 0.8
Li + PXT 0%
Nemeroff CB, et al. Am J Psychiatry. 2001;158:906-912. *p < 0.05
Li + IMI 8%
 Adjunctive Paroxetine vs Second Mood
Stabilizer in Bipolar Depression
Similar Efficacy* Better Tolerability
Hamilton Depression Scale (17 item)

25
40

20 •• 38%
(6/16)

Dropout Rate (%)

• Double-Blind Adjunctive
Response

• • •
15 2nd Mood Stabilizer
Rates

10 • • •• •• • 64% 20 p < 0.05

• Paroxetine
50%

5
0%
(0/11)
0 0
2nd Mood
Baseline 1 2 3 4 5 6 Paroxetine
Stabilizer
Treatment Duration (wks)
*Last Observation Carried Forward Analysis Young, et al. Am J Psychiatry 2000;157:124-6.
26-Week Double-Blind Adjunctive Antidepressant
vs Placebo in Acute Bipolar Depression
Recovery Rates Switch Rates
30 P=0.40
NNT = 26

25 27.3%
Percentage of Patients

(51/187) 68% BPI, 32% BPII

23.5% Bupropion - 300 mg/d (median, N = 86)
Paroxetine - 30 mg/d (median, N = 93)
20 (42/179)

15 P=0.84
NNH = 167

10 10.7%
10.1% (20/187)
(18/179)
5

0
Mood Stabilizer Mood Stabilizer Mood Stabilizer Mood Stabilizer
+ Antidepressant + Placebo + Antidepressant + Placebo

Sachs GS, et al. N Engl J Med 2007;356:1711-22.

Adding antidepressant no better or worse than adding placebo to mood stabilizer(s).

52-Week Adjunctive Intensive Psychosocial Intervention
vs Collaborative Care in Acute Bipolar Depression

N = 130
Intensive Collaborative p = 0.01
Median time to 50% recovery (days) 169 [138-230] 279 [-] N = 163
Median time to 25% recovery (days) 98 [88-112] 125 [105-168]
Recovered at 1 year (%) 64.4 51.5
3 Up to 30
sessions sessions

Miklowitz DJ, et al. Arch Gen Psychiatry 2007;64:419-27.

52-Week Adjunctive Intensive Psychosocial Intervention
vs Collaborative Care in Acute Bipolar Depression

p < 0.05
N = 130

N = 163

3 Up to 30
sessions sessions

Miklowitz DJ, et al. Arch Gen Psychiatry 2007;64:419-27.

52-Week Adjunctive Intensive Psychosocial Intervention
vs Collaborative Care in Acute Bipolar Depression

N = 163

N = 130

Odds of Being Well in Any Month

Intensive 1.58 x Collaborative
p = 0.003

3 Up to 30
sessions sessions

Miklowitz DJ, et al. Arch Gen Psychiatry 2007;64:419-27.

 Do Antidepressants Induce Mania?

 41% Natural switch rate depression to mania

(on no antidepressants) 1

 Switch rate on medications 2

– 53% Imipramine
– 28% Lithium plus imipramine
– 26% Lithium

Lewis JL, Winokur G. Arch Gen Psychiatry 1982 1; Prien RF, et al. Arch Gen Psychiatry 1984.
Switch Rate From Index Depression Into Mania

By Era and Prevailing Treatment

10

Tricyclics
8 (N=509)
Switch Rate (%)

6 ECT
(N=100)

4 Nano-
leptics
(N=100)
Spontaneous
2
(N=200)

0
1920 1930 1940 1950 1960 1970 1980
Year
Angst J. Psychopathology 1985.
Increased Mania Switch Rates with Tricyclics
**
11.2%
12 (14/125)

10
% With Manic Switch

6 4.2%
3.7%
(2/48)
4 (9/242)

0
PBO Sertraline / Paroxetine TCAs
** p < 0.01 vs PBO
Peet M. Br J Psychiatry. 1994;164:549-550.
Switch Rates With Tricyclic vs.
Other Antidepressants

Gijsman et al, American Journal of Psychiatry. 2004; 161: 1537-1547

Manic Switch Rates in Randomized Controlled Trials
of Antidepressants vs. Placebo

Gijsman et al, American Journal of Psychiatry. 2004;161:1537-1547

 10-Week Randomized Adjunctive
Antidepressants in Acute Bipolar Depressiona
60
Response Rates
53%
51%
49%
50
Either ≥ 50% IDS Decrease or
Percentage of Patients with

≥ 2 point CGI Improvement

40

30

20

10
286 mg/d 192 mg/d 195 mg/d
N = 51 N = 58 N = 65
0
Bupropion Sertraline Venlafaxine
a
73% Bipolar I, 26% Bipolar II, 1% Bipolar NOS; 82% double-blind, 12% open.
P = NS.
Post RM, et al. Br J Psychiatry 2006;189:124-31.
 10-Week Randomized Adjunctive
Antidepressants in Acute Bipolar Depressiona
Switch Rates
40 YMRS >13 CGI-M YMRS >13 or
Increase ≥ 2 CGI-M ≥ 3
P = 0.05. P < 0.01. P = 0.03. 31%
Percentage of Patients

29%
30

20
15% 16%
14%
10% 9%
10 7%
4%
N = 51 N = 58 N = 65 N = 51 N = 58 N = 65 N = 51 N = 58 N = 65
0
BUP SERT VEN BUP SERT VEN BUP SERT VEN
a
73% Bipolar I, 26% Bipolar II, 1% Bipolar NOS; 82% double-blind, 12% open.
Post RM, et al. Br J Psychiatry 2006;189:124-31.
Do Antidepressants Induce Rapid Cycling?

 Increased rapid cycling since TCAs introduced 1

 Mania rates over 2 years 2

– 67% Imipramine
– 33% Placebo
– 18% Lithium
 Antidepressants induce reversible rapid cycling in

double-blind placebo-controlled studies.3

Angst J. Psychopathology 19851; Prien RF, et al. Arch Gen Psychiatry 1973 2;
Wehr TA, Goodwin FK. Psychopharmacol Bull 1987 3
 Tricyclics Shorten Cycle Length

10 Bipolar Disorder Patients

300 TCAs TCAs

Cycle Length (days)

250

200

150

100

50

0
Off On Off On Off
Tricyclic Treatment Status
Wehr TA, Goodwin FK. Am J Psychiatry 1987.
 Acute Bipolar I Depression Algorithm

 Optimize current mood stabilizer (if applicable) before

initiating additional treatment for depression
 Patients on Li - optimize (serum Li level ≥ 0.8 mEq/L) to
determine whether adjunctive intervention necessary
 Patients with recent and/or severe history of mania - receive
or add an effective antimanic agent

 Stage 1
 Adjunctive LTG if depression persists after mood stabilizer
optimization

Number of iterations at each level and adjunctive treatment(s) to be determined by clinician judgment
Suppes T, et al. J Clin Psychiatry 2005;66:870-86.
 Acute Bipolar I Depression Algorithm

 Stage 2: If Stage 1 ineffective or not tolerated*

 QTP monotherapy or OFC
 Although onset of action faster than LTG, overall efficacy and
long-term tolerability evidence favors LTG (at Stage 1)

 Stage 3: If Stages 1 and 2 ineffective or not tolerated*

 Combination of two agents already introduced in algorithm
 Li, LTG, QTP, and OFC combination

 OFC a two-drug combination, so adding another agent yields

three-drug combination

Number of iterations at each level and adjunctive treatment(s) to be determined by clinician judgment
Suppes T, et al. J Clin Psychiatry 2005;66:870-86.
 Acute Bipolar I Depression Algorithm

 Stage 4: If Stages 1, 2, and 3 ineffective or not tolerated*

 ECT and combination therapy ( Li, LTG, QTP, OFC
combination, VPA or CBZ in combined with SSRI, bupropion,
or venlafaxine)
 Minority opinion that Stage 4 should precede Stages 2 and 3

 Stage 5: If Stages 1, 2, 3, and 4 ineffective or not tolerated*

 MAO-I, other atypical antipsychotics not included,
pramipexole, new combinations of drugs included in the
algorithm, inositol, stimulants, and thyroid supplementation

Number of iterations at each level and adjunctive treatment(s) to be determined by clinician judgment
Suppes T, et al. J Clin Psychiatry 2005;66:870-86.
Maintenance Treatment of
Bipolar Depression
Summary of Double-Blind Lithium Monotherapy
vs Placebo Maintenance Trials in 1970s
Lithium Compared to Placebo, Primarily After Manic/Mixed Episodes
Superior Superior Superior
Episode Depression Mania
100 Prevention Prevention Prevention
Percentage of Patients

81%
80 Lithium (n=251)
Placebo (n=263) 56%
60

37%
40 34%
21% 23%
20

0
Overall Relapse Relapse Into Depression Relapse Into Mania
or Hypomania
Goodwin FK, Jamison KR: Manic-Depressive Illness, Oxford University Press, New York 1990:688-9.
Lithium Prevention of Any Relapse
in Bipolar Disorder

Areas of blue boxes reflect weights of studies in meta-analysis.

Lower confidence interval extends beyond graph (0.08).
b

Geddes JR et al. Am J Psychiatry 2004;161:217-222.

Lithium Prevention of Depressive Relapse
in Bipolar Disorder

Areas of blue boxes reflect weights of studies in meta-analysis.

Lower confidence interval extends beyond graph (0.10).
b

Geddes JR et al. Am J Psychiatry 2004;161:217-222.

Lithium Prevention of Manic Relapse
in Bipolar Disorder

Areas of blue boxes reflect weights of studies in meta-analysis.

Geddes JR et al. Am J Psychiatry 2004;161:217-222.
 12-Month Double-Blind Divalproex, Lithium
Monotherapy vs Placebo Maintenance
Fewer Dropouts Due to Depression with Divalproex vs Placebo After Manic/Mixed
Episodes

Overall Patients Receiving SSRI Rescue

Discontinuing Due to Depression

Discontinuing Due to Depression

P = 0.03
Percentage of Patients

20% 50%
P = 0.017
45%

Percentage of Patients
40%
15%
35%
30%
10% 25%
20%
15%
5%
10%
5%
0% 0%
DVP (n=187) PBO (n=94) DVP + SSRI PBO + SSRI
(n=41) (n=20)

DVP = divalproex PBO = placebo LI = lithium

Gyulai et al. Neuropsychopharmacol 2003;28:1374-82.
SSRI = selective serotonin reuptake inhibitor
 Lamotrigine and Lithium
Effective in Bipolar I Prophylaxis
Time to Intervention for Any Episode (pooled recently manic/dep pts)
Patients stabilized on lamotrigine prior to randomization.
1.0
0.9 LTG v. PBO, p < 0.001 Placebo (n=188)
Li v. PBO, p < 0.001 Lamotrigine 245 mg/d (n=223)
0.8
Survival Estimate

LTG v. Li, p = 0.629 Lithium 0.7 mEq/L (n=164)

0.7
0.6
0.5
50
0.4 37%
42%
40
0.3 30 22%
20
0.2 10

0.1 0
LTG Li PBO
18 Months
0.0
0 10 20 30 40 50 60 70
Week
Goodwin GM, et al. J Clin Psychiatry 2004;65:432-41.
 Lamotrigine Effective in
Bipolar I Depression Prophylaxis
Time to Intervention for Depression (pooled recently manic/dep pts)
Patients stabilized on lamotrigine prior to randomization.
1
LTG v. PBO, p = 0.009 Placebo (n=188)
0.9 Li v. PBO, p = 0.120 Lamotrigine 245 mg/d (n=223)
0.8 LTG v. Li, p = 0.325 Lithium 0.7 mEq/L (n=164)
Survival Estimate

0.7
0.6
0.5
57%
60 53%
0.4 50 41%
40
0.3 30

0.2 20
10

0.1 0
LTG Li PBO
18 Months
0
0 10 20 30 40 50 60 70
Week
Some patients considered intervention-free for depression could have had intervention for mania.
Goodwin GM, et al. J Clin Psychiatry 2004;65:432-41.
 Lamotrigine and Lithium Effective in
Bipolar I Mania Prophylaxis
Time to Intervention for Mania (pooled recently manic/dep pts)
1 Patients stabilized on lamotrigine prior to randomization.

0.9
0.8
Survival Estimate

0.7
0.6
0.5
80%
80
0.4 65%
60 53% Placebo (n=188)
LTG v. PBO, p = 0.034
0.3 Li v. PBO, p < 0.001 Lamotrigine 245 mg/d (n=223)
40
0.2 20
LTG v. Li, p = 0.030 Lithium 0.7 mEq/L (n=164)

0.1 0
LTG Li PBO
18 Months
0
0 10 20 30 40 50 60 70
Week
Some patients considered intervention-free for mania could have had intervention for depression.
Goodwin GM, et al. J Clin Psychiatry 2004;65:432-41.
 Incidence of Mania/Hypomania/Mixed
Episodes Reported as Adverse Events

Combined Analysis
Patients stabilized on lamotrigine prior to randomization.

25
Percent of patients

20

15

10 7%
5% 4%
5

0
Lamotrigine Lithium Placebo
(n=227) (n=166) (n=190)

In all bipolar controlled trials, adverse events of mania were reported as 5% lamotrigine, 3% lithium, and 4% placebo.

Goodwin GM, et al. J Clin Psychiatry 2004;65:432-41.

 12-Month Double-Blind Olanzapine
Monotherapy vs Placebo Maintenance
Olanzapine Compared to Placebo After Manic/Mixed Episodes
Superior Superior Superior
Episode Depression Mania
100 Prevention Prevention Prevention
Percentage of Patients

80.1%
80 p<.001 Olanzapine 12.5 mg/d (n=225)

Placebo (n=136)
60
46.7% p=.015 47.8%
41.2%
40 34.7% p<.001

20 16.4%

0
Overall Relapse Relapse Into Depression Relapse Into Mania
Stabilized on OLZ before randomization. Relapse criteria - hospitalized or YMRS or HAMD-21 >= 15.
Tohen MF, et al. Am J Psychiatry 2006;163:247-56.
 12-Month Double-Blind Olanzapine vs
Lithium Maintenance Monotherapy
Olanzapine Compared to Lithium After Manic/Mixed Episodes
Equivalent Equivalent Superior
50
Episode Depression Mania
Prevention Prevention Prevention
Percentage of Patients

40 38.8% Olanzapine 11.9 mg/d (n=217)

p=.055
Lithium 1103 mg/d (0.77 mEq/L) (n=214)
30.0%
30 28.0%
p<.001
p=.895
20 16.1% 15.4% 14.3%
10

0
Overall Relapse Relapse Into Depression Relapse Into Mania

Stabilized on OLZ+Li before randomization. Relapse criteria - YMRS or HAMD-21 >= 15.
Tohen MF, et al. Am J Psychiatry 2005;162:1281-90.
 26-Week Double-Blind Aripiprazole vs Placebo
Continuation/Maintenance Monotherapy
Aripiprazole Compared to Placebo After Manic/Mixed Episodes
Superior Equivalent Equivalent Superior
Episode Depression Mixed Mania
50 Prevention Prevention Prevention Prevention
43%
Percent of Patients

40
p=.013

30 25%
Aripiprazole 24.3 mg/d (n=77)
Placebo (n=83) 23%
20 p=.009
12% 13%
10 6% 8%
5%
0
Overall Relapse into Relapse into Relapse into
Relapse Depression Mixed Mania
Stabilized on ARI before randomization.
Keck PE, et al. 157th APA Annual Meeting; May 1-6, 2004; New York, NY. Abstract NR746.
Antidepressants After Depression Resolution

Disorder / Episode Pattern Begin Taper Comments

Unipolar 6–12 months Maintenance if

≥ 3 episodes

Bipolar
Monophasic 6–12 weeks Repeat if relapse
Biphasic - MDE Maintenance if
repeated relapses
Bipolar
Biphasic - DME 6–12 days Start taper after
Polyphasic first euthymic visit
Hx rapid cycling
Hx iatrogenic mania

Sachs G, 2000. Personal communication.

 Controlled Maintenance Studies of
Antidepressants for Bipolar Depression

Study N, Duration Efficacy Switch

Prien et al ’73 N=44, 24 mo Li > IMI = PBO
Wehr & Goodwin N=5, 27 mo Li = Li + DMI Li + DMI >> Li
‘79
Quitkin et al ‘81 N=75, 19 mo Li = Li + IMI Li + IMI > Li
Kane et al ‘82 N=22, 11 mo Li > PBO = IMI
Prien et al ‘84 N=117, 30 mo Li = Li + IMI> IMI IMI > Li + IMI =
Li
Sachs et al ‘94 N=15, 12 mo Li + BUP= Li + DMI >
Li + DMI Li + BUP

Kane et al Arch Gen Psychiatry 1982;39:1065-9; Prien et al Arch Gen Psychiatry 1984;41:1096-1104; Prien et al Arch Gen Psychiatry
1973;29:420-5; Quitkin et al Arch Gen Psychiatry 1981;38:902-7; Sachs et al J Clin Psychiatry 1994;55:391-3; Wehr & Goodwin Arch
Gen Psychiatry 1979;36:555-9.
 Bipolar Versus Unipolar
Maintenance Treatment Dissociation
IMI+LI Li IMI PBO
100 Bipolar 100 Unipolar
Cumulative Probability of Remaining Well

Cumulative Probability of Remaining Well

90 90
80 80
70 70
60 60
50 50
40 40
30 30
20 20
10 10

0 0
3 6 9 12 15 18 21 23 25 27 30 3 6 9 12 15 18 21 24 27
Months Months
Li 0.8 mEq/L; IMI 125 mg/d Adapted from Prien et al Arch Gen Psychiatry 1984;41:1096:1104.
Antidepressant Continuation Beneficial
in Some (15%?) Patients

n = 41
Prospective 1-year follow-up
(36% relapsed) Remission of MDE with AD
added to mood stabilizer

Tolerated AD ≥ 2 months

Continuation: AD > 6 months

Discontinuation: AD < 6 months
n = 43
(70% relapsed)

Altshuler et al. Am J Psychiatry. 2003;160:1252-62.

Treatment of Bipolar Depression
 Acute treatment
– Lithium, lamotrigine
– Olanzapine plus fluoxetine, quetiapine
– Adjunctive antidepressants
– Alternative treatments


– Lithium, lamotrigine
– Divalproex
– Adjunctive antidepressants (controversial)
– Alternative treatments


 Post-Lecture Exam
Question 1

1. The most pervasive symptoms in bipolar disorder are

those of: (choose one)

A. Mania, hypomania

B. Hypomania

C. Depression

D. Mixed States

E. None of the above

 Question 2

Which of the treatments below is the LEAST appropriate

strategy in bipolar depression: (choose one)

A. Mood stabilizer without antidepressant

B. Mood stabilizer with antidepressant

C. Atypical antipsychotic with antidepressant

D. Antidepressant with neither mood stabilizer nor atypical

antipsychotic
 Question 3

Which antidepressant option carries the greatest risk of

hypomania/mania: (choose one)

A. Tricyclic antidepressants (TCAs)

B. Selective serotonin reuptake inhibitors (SSRIs)

C. Mirtazepine

D. Bupropion
 Question 4
Which of the following treatments do NOT have controlled
data suggesting utility in bipolar depression: (choose
one)

A. Lithium

B. Lamotrigine

C. Olanzapine plus fluoxetine combination

D. Quetiapine

E. Citalopram

F. Pramipexole
 Question 5
Which of the following statements best describes the
role of maintenance adjunctive antidepressants in
patients with bipolar disorder: (choose one)

A. Long-term adjunctive antidepressants are always

beneficial.

B. Long-term adjunctive antidepressants are never

beneficial.

C. Long-term adjunctive antidepressants are beneficial

in most patients.

D. Long-term adjunctive antidepressants may be

beneficial in some patients.
Answers to Pre & Post Competency Exam

1. C
2. D
3. A
4. E
5. D