1

New Approaches to Insomnia

and Depression

Andrew Krystal, MD, MSDirector, Sleep Research Laboratory and Insomnia Program

Director of Neurosciences Medicine, Duke Clinical Research Institute

Professor of Psychiatry and Behavioral Sciences

Duke University School of Medicine

Durham, North Carolina

2

Disclosures

• My content will include reference to commercial

products; however, generic and alternative

products will be discussed whenever possible.

• Consulting: Abbott, AstraZeneca, Attentiv, Teva,

Eisai,, Jazz, Janssen, Merck, Neurocrine,

Novartis, Otsuka, Lundbeck, Roche, Somnus,

Sunovion, Somaxon, Transcept, Vantia

• Grants/research support: NIH, Teva, Sunovion,

Astellas, Abbott, Neosync, Brainsway, Janssen,

ANS St. Jude, Novartis.

3

Overview

• The relationship of insomnia and

depression

• General treatment strategy

• Available data on the treatment of

insomnia in patients with insomnia and

depression

• Trying to make sense of the findings

• Conclusions

4

90% with MDD Have Sleep Problem:

DSM-IV Major Depression

• 2-week period of depressed mood or loss of interest

• Clinically significant distress or impairment of functioning

• Symptoms not due to substance abuse or medical condition

• 4 or more of the following symptoms:

– Insomnia/hypersomnia

– Weight loss/decreased appetite

– Psychomotor agitation/retardation

– Fatigue/loss of energy

– Worthlessness and guilt

– Diminished concentration/indecisiveness

– Thoughts of death and suicide

Armitage R, et al. Depress Anxiety. 1997;5:97-102; Reynolds CF III, Kupfer DJ. Sleep.

1987;10:199-215; APA, 2001.

5

Major Depressive Disorder

• Sleep alterations reported include:– Difficulty falling and staying asleep

– Increased light, stage 1 sleep

– Decreased SWS

– Decreased (<65 min) REM latency

– Prolonged first REM sleep

– Increased total REM sleep

• REM and SWS changes not currently believed relevant to insomnia diagnosis or outcome

Gillin et al., Arch Gen Psychiatry 1979; Kupfer et al., Arch Gen Psychiatry 1985; Walter et al., . Biol Psychiatry 1989;

Berger et al., Biol Psychiatry 1982; Tsuno et al., J Clin Psych. 2005

6

Changes in Sleep in MDD

• Decreased amount of sleep– Prolonged sleep latency

– Increased wake time in middle of night

– Early morning awakenings with inability to return to sleep

– Reduced sleep efficiency

– Decreased total sleep time

• Alterations in sleep stages –implications?

– Decreased slow-wave sleep (stages 3 and 4)

– Shortened REM latency (<65 minutes)

– Increased total amount of REM and REM%

Armitage R. Can J Psychiatry. 2000;45:803-809; Kupfer, et al. 1995; Benca, et al. 1992.

7

Normal

MDD

Sleep in MDD

Stg 4

Stg 3

Stg 2

Stg 1

REM

Waking

87654321

Time (Hours)

Time (Hours)

87654321

Stg 4

Stg 3

Stg 2

Stg 1

REM

Waking

8

Insomnia in MDD: Symptom or

Co-Morbid Conditions?• Are sleep problems best thought of as

symptoms or conditions that are co-morbid

with psychiatric disorders?

• Long considered symptoms: 1983 NIH

Consensus Conference:• Chronic insomnia is caused by medical and

psychiatric disorders

• Insomnia-specific treatment is not needed

• Treating the “underlying disorder” should address the

insomnia

9

Evidence for Bidirectionality, Insomnia

Independence - Insomnia:• Increases the risk of future depression

• Decreases antidepressant response

• Is independent risk factor for suicidality, attempts

and completed suicide in MDD pts

• Insomnia is a stronger predictor of near-lethal suicide

attempts than a specific suicide plan

• Relative risk of suicide death in studies up to 2.4

• Is the most frequent residual symptom in

antidepressant responders

• Residual insomnia increases relapse riskFawcett J. J Clin Psychiatry.1988;49;Suppl 7-8; Fawcett J, et al. Am J Psychiatry.1990;147(9):1189-1194; Reynolds CF, et al. Am J Psychiatry. 1997;154:958-962; Breslau N, et al. Biol Psychiatry.1996;39:411-418; Ford DE, Kamerow DB. JAMA.1989;262(11):1479-1484; Adam K, et al. J Psychiatr Res. 1986;20(4):301-306; Vgontzas AN, et al. J Clin Endocrinol Metab.2001;86(8):3787-3794; Ford DE, Cooper-Patrick L. Depress Anxiety.2001;14(1):3-6; Roberts RE, et al. Am J Psychiatry. 2000;157(1):81-88; Fava GA, et al. J Affect Disord. 1990;19:149-152; Nierenberg AA, et al. J Clin Psychiatry. 1999;60:221-225; Weissman, et al. Gen Hosp Psych. 1997. Hall et al., Psychosomatics. 1999; McCall et al., Sleep Medicine 2010

10

Sleep Disturbance Major Depression

Bidirectional Effects

11

Strategies for Treatment of

Insomnia in Depressed Patients • Monotherapy with sedating antidepressant

– Mirtazapine: 15–45 mg qhs; tricyclic antidepressant• Advantages: Single medication — good compliance?

• Disadvantages: Limited antidepressant options; risks residual daytime sedation, weight gain, etc.

• Nonsedating antidepressant plus:

– Sedating antidepressant: trazodone: 50–200 mg qhs; mirtazapine 15–30 mg qhs; low-dose tricyclic antidepressant

– Hypnotic agent• Advantages: greater flexibility in antidepressant selection; More predictable rapid sleep

improvement; can d/c sedating agent and continue antidepressant

• Disadvantages: Compliance? Cost

– Cognitive behavioral therapy for insomnia (CBTI)

Nowell PD, Buysse DJ. Depress Anxiety. 2001;14:7-18; Nierenberg AA, et al. Am J Psychiatry.

1994;15:1069-1072; Dording CM, Mischoulon D, Petersen TJ, et al. Ann Clin Psychiatry. 2002;14:143-147;

Saletu-Zyhlarz GM, Abu-Bakr MH, Anderer P, et al. Prog Neuropsychopharmacol Biol Psychiatry.

2002;26:249-260.

12

Cognitive Behavioral Insomnia

Therapy

• Multiple components frequently

administered in combination

– Sleep Hygiene

– Stimulus Control

– Sleep Restriction

– Cognitive Therapy

– Relaxation Techniques

13

Sleep Hygiene Education

• Caffeine - sources & effects

• Nicotine

• Role of exercise

• Light bedtime snack (milk, peanut butter)

• Alcohol, tobacco & other substances

• Environment: light, noise, temperature

17

CBTI+Antidepressant Meds in

MDD• MDD patients receiving CBTI along with

escitalopram had a greater depression

remission rate than those administered a

control behavioral intervention along with

escitalopram (62% vs 33%)

• Two trials recently completed of anti-

depressant meds plus CBTI

– Improvement in insomnia mediates MDD

improvement; Detailed results pending

Manber et al., Sleep. 2008

18

Available Data on Insomnia

Pharmacotherapy• Studies of sedating antidepressants don’t

specifically assess sleep effects

– Don’t indicate utility of treating sleep

• No studies of Sedating antidepressant +

Non-sedating antidepressant

• Studies of treatment of residual insomnia

with drug for sleep + antidepressant

– Zolpidem, Trazodone

• “Hypnotic” + Non-sedating antidepressant

– clonazepam, eszopiclone, zolpidem CR

19

Improvement in Residual Insomnia

in SSRI-Treated Patients with MDD

Greater improvement in sleep compared

with placebo found with adjunctive:

• Trazodone

• Zolpidem

– Improved reported function and QOL

Asnis GM, et al. J Clin Psychiatry. 1999;60:668-676.

Nierenberg AA, et al. Am J Psychiatry. 1994;151:1069-1072.

20

Trials of Antidepressant/Sleep

Agent Co-Therapy

• Clonazepam

• Eszopiclone

• Zolpidem CR

2121

Screening

21

Trial design:

Clonazepam in patients with insomnia and MDD

Open-label fluoxetine 20 mg/d

Double-blind treatment(21 days)

Week 0 6 8

Clonazepam 0.5–10 mg/d (n=40)

Placebo (n=40)

3

20–40 mg/d

Taper(12 days)

Smith WT, et al. Am J Psychiatry. 1998;155:1339-45

2222

22

Clonazepam:

Depression outcomes in patients with insomnia and MDD

FLU (20–40 mg/d) + PBO (n=40)

FLU (20–40 mg/d) + CLO 0.5–1.0 mg (n=40)

Smith WT, et al. Am J Psychiatry. 1998;155:1339-1345

HAM-D17 Total Score

Week 1 Week 3 Week 40

5

10

15

20All patients received fluoxetine

Improvement18

14 13.5

11.5

15

11

**

***

Mean

HA

M-D

17 t

ota

l sco

re

**p<0.01, ***p<0.001 vs fluoxetine + placebo

23

Screening

Trial design:

Eszopiclone in patients with insomnia and MDD

Open-label fluoxetine 20–40 mg/d

Eszopiclone 3 mg/d (n=269)

Placebo (n=274)

Single-blind

placebo

Double-blind treatment Run-out

Week –2 0 4 8 10

Fava M, et al. Biol Psychiatry. 2006; Krystal et al., J Clin Sleep Med. 2007

24

****

*

*

0

30

60

90

120

150

-1 0 1 2 3 4 5 6 7 8

Placebo+Fluoxetine

Eszopiclone+Fluoxetine

Min

ute

s (

media

n)

Week of Study

*p<0.0002 vs placebo

p values reflect results from change from baseline analyses using ANCOVA.

Sleep Latency (LOCF)

25

WASO (LOCF)

*

**

†

* *-15

0

15

30

45

60

75

90

-1 0 1 2 3 4 5 6 7 8

Placebo+Fluoxetine

Eszopiclone+Fluoxetine

Min

ute

s (

media

n)

Week of Study

*p<0.007 vs placebo

†p<0.05 vs placebo

p values reflect results from change from baseline analyses using ANCOVA.

26

Total Sleep Time (LOCF)

*****

*

240

270

300

330

360

390

420

-1 0 1 2 3 4 5 6 7 8

Placebo+Fluoxetine

Eszopiclone+Fluoxetine

Min

ute

s (

media

n)

Week of Study

p values reflect results from change from baseline analyses using ANCOVA.

*P<0.0001 vs placebo

27

Change from Baseline in HAM-D17

-10.9

-12.2-12.9

-14.8

-16

-14

-12

-10

-8

-6

-4

-2

0

Week 8 Week 10

-9.2 -9.5-10.4

-11.1

-16

-14

-12

-10

-8

-6

-4

-2

0

Week 8 Week 10

Placebo+Fluoxetine

Eszopiclone+Fluoxetine

All Items Excluding Insomnia Items

LS

Mean C

hange

p values reflect results from change from baseline analyses using ANCOVA

p=0.002

p<0.0001

p=0.001p=0.01

28

Antidepressant Response and Remission

(LOCF)

48

59

30

36

0

10

20

30

40

50

60

70

Week 4 Week 8

19

33

22

42

0

10

20

30

40

50

60

70

Week 4 Week 8

Placebo+Fluoxetine

Eszopiclone+Fluoxetine

Response

(50% HAM-D17 decrease)Remission

(HAM-D177)

Perc

ent

of P

atients

p=0.14

p=0.03

p=0.29

p=0.009

29

Improvement in MDD Not Seen

with Zolpidem CR• Identical study carried out with zolpidem

CR and sleep was improved but no

improvement in depression vs placebo

Sheehan et al., 2009

3030

30

Study design:

Zolpidem CR in patients with insomnia and MDD

Double-blind treatment Run-out

Zolpidem controlled-release 12.5 mg/d (n=190)

Placebo (n=190)

Open-label escitalopram 10 mg/d

Week –1 0 4 8 10

Screening

12 16 20 2624

Responder-randomized treatment

Responders ≥50%

improvement in HAM-D17

Fava M, et al. Poster presented at APA Annual Meeting, Washington, DC; 2008

3131

Sleep Latency

0 2 4 6 8-150

-100

-50

0

All patients received escitalopram

Improvement

**** ** **

Number of weeks of treatment

Mean

ch

an

ge f

rom

baseli

ne

in

sle

ep

late

ncy (

min

ute

s)

Wake After Sleep Onset

0 2 4 6 8-80

-60

-40

-20

0

All patients received escitalopram

Improvement

*** ****** ***

Number of weeks of treatment

Mean

ch

an

ge f

rom

baseli

ne

in

WA

SO

(m

inu

tes)

Total Sleep Time

0 2 4 6 80

50

100

150

All patients received escitalopram

Improvement

*** ****** ***

Number of weeks of treatment

Mean

ch

an

ge f

rom

baseli

ne

in

TS

T (

min

ute

s)

31

Zolpidem CR:

Sleep outcomes in patients with insomnia and MDD

**p<0.001, ***p<0.0001vs escitalopram + placebo

ESC 10 mg/d + PBO (n=190)

ESC 10 mg/d + ZOL CR 12.5 mg (n=190)

Fava M, et al. Poster presented at APA Annual Meeting, Washington, DC; 2008

3232

32

Zolpidem CR:

Depression outcomes in patients with insomnia and MDD

ESC 10 mg/d + PBO (n=190)

ESC 10 mg/d + ZOL CR 12.5 mg (n=190)

Fava M, et al. Poster presented at the 22nd Annual APSS Meeting (SLEEP); June 2008

Week 4 Week 8

-15

-10

-5

0

All patients received escitalopram

Improvement

-8.1-9.1

-10.9

-11.3

Me

an

ch

an

ge

fro

m b

ase

line

HA

M-D

17

to

tal sco

re

33

How Do We Explain Eszopiclone

vs Zolpidem CR Difference?

• Sleep effect sizes are comparable but

significant difference in associated effect

on depression.

• Improvement in depression may not be

mediated by improvement in sleep; and

Either:

– Eszopiclone is an antidepressant and

Zolpidem CR is not

– Zolpidem CR is an anti-antidepressant and

Eszopiclone is not

34

BENEFIT SUSTAINED FOR AT LEAST 2 WEEKS POST

DISCONTINUATION OF INSOMNIA THERAPY

35

Discontinuation Effects

WASO

0

10

20

30

40

50

60

70

80

90

-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

FLX + PBO FLX + ESZ

Days after eszopiclone discontinuation

Min

ute

s (

media

n)

Week 8 = last week of DB treatment

BSL Wk 8

36

Discontinuation Effects

TST

200

250

300

350

400

450

-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

FLX + PBO FLX + ESZ

Days after eszopiclone discontinuation

Min

ute

s (

media

n)

Week 8 = last week of DB treatment

BSL Wk 8

37

Mechanism of ESZ vs Zolpidem

Difference?• Benzodiazepines Temazepam, Flurazepam, Triazolam etc.

– A group of compounds with related chemical structure

– Mechanism of action:• GABA receptor comprised of 5 peptides that form channel

which controls the flow of chloride ions in and out of the neuron.

• Generally, Cl concentration greater outside than inside the neuron. GABA binding opens the channel and resulting inward flux of CL hyperpolarizes neuronal membrane causing inhibition

• Benzodiazepines bind to a binding site on α subunit of GABA receptor complex and enhance this GABA-mediated inhibition

• “Non-Benzodiazepines” Zolpidem, Zaleplon, Eszopiclone, Indiplon

– A group of compounds unrelated to selves or benzos

– Mechanism of action:• Same as benzos, relatively greater α subunit binding specificity

38

The GABA Receptor Complex

39

GABA-A Subunit-Specific Effects

• The effects of binding to α subunits differ

because of location of GABA receptors

containing them– Greater binding to α subunits of GABA receptors in

the cerebellum will result in greater effect on balance.

– Greater binding to α subunits of GABA receptors in

the amygdala will result in greater effect on anxiety.

– Evidence of differential binding in animals

• Limited human data on differential α subunit binding

40

GABA Alpha Subunit Subtypes

41

Alpha

subunit

Agents with

Significant

Effects

Possible

Adjunctive

Therapeutic

Effects

Potential

Adverse Effects

a1

Triazolam, Temazepam,

Flurazepam, Estazolam,

Quazepam, Zaleplon,

Zolpidem, Zolpidem CR,

Eszopiclone

Anti-ConvulsantCognitive Impariment,

Ataxia

a2

Triazolam, Temazepam,

Flurazepam, Estazolam,

Quazepam, Eszopiclone

Anxiolytic,

Myorelaxant,

Antidepressant?

a3

Triazolam, Temazepam,

Flurazepam, Estazolam,

Quazepam, Eszopiclone

Anxiolytic,

Myorelaxant,

Antidepressant?

a4Analgesia Ataxia, Amnesia

a5

Triazolam, Temazepam,

Flurazepam, Estazolam,

QuazepamMyorelaxation

Cognitive Impairment

Tolerance

Animal Data on Effects of GABA Alpha Subunit Binding

Mohler et al., J Pharmacol Exp Ther, 2002; Kopp et al., Proc Natl Acad Sci. 2004; Van Rijnsgever et al., J

Neurosci. 2004; Chandra et al., Proc Natl Acad Sci. 2006; Crestani et al., Neuropharmacology, 2002;

Griebel et al., J Psychopharmacol. 1998

42

Conclusions

• Targeting pharmacologic treatment

specifically to insomnia can significantly

improve sleep and reported daytime

function and QOL

– Further studies needed to determine

• Optimal duration of treatment

• If treating insomnia decreases relapse rate

43

Conclusions

• Effects of treating insomnia on

antidepressant response are variable

– Esz but not Zolp appears to augment

antidepressant effect; Clonazepam?

• Further studies needed:

– To confirm antidepressant effect of ESZ and mechanism

(a2, a3 ?)

» To determine if antidepressant benefit is sustained and if there is

decreased risk of relapse

» To determine if ESZ alone is antidepressant

» With other (a2, a3 ?) agents

– Relationship between antidepressant and sleep effects

» Do drugs with sleep benefit and antidepressant

effects have greater augmentation effect?