Antidepressants for the treatment of depression in palliative care: systematic review & meta-analysis Lauren Rayner, Dr Annabel Price, Dr Alison Evans, Dr Koravangattu Valsraj, Professor Irene Higginson, Professor Matthew Hotopf Department of Palliative Care, Policy & Rehabilitation, CSI Department of Psychological Medicine, IOP King’s College London

Depression in palliative care • Prevalence: - median 15% advanced disease (Hotopf 2002) • Associations: - poorer functioning (Booth 2001) - reduced adherence to treatment (DiMatteo 2000) - poor prognosis (Hata 2006) - higher mortality (Lloyd-Williams 2008) - increased costs (Boulanger 2009) • Barriers to detection: - understandable distress vs. depressive disorder (Horwitz 2007) - overlap of somatic symptoms (Endicott 1984)

www.kcl.ac.uk/palliative

Antidepressants in palliative care • Effective in physically healthy people but doubts about whether they are appropriate in palliative care • Drug contraindications and interactions • Need to take account of prognosis and time required to take effect (Kendrick 2009; Lloyd-Williams 1999) • Evidence of efficacy? - cochrane review of antidepressants for depression in people with a physical illness (Rayner 2010) - effective and acceptable

www.kcl.ac.uk/palliative

Our systematic review • Palliative care defined broadly as “the prevention and relief of suffering in patients with a life-threatening illness” (WHO 2002) • There is no consensus about which diseases are lifethreatening • We included cancer, renal failure, COPD, chronic heart failure, Parkinson’s disease, multiple sclerosis, HIV/AIDS • These are all progressive illnesses which benefit from palliative care

www.kcl.ac.uk/palliative

Method • Selection criteria: RCTs comparing antidepressants and placebo for depression in ‘life-threatening illness’ • Search strategy: Cochrane trials registers, EMBASE, MEDLINE, PSYCINFO, reference lists… • Primary outcome: efficacy 6-8 weeks • Secondary outcomes: efficacy at 4-5 & 9-18, acceptability (drop-outs), tolerability (adverse events) • Data analysis: odds ratios for dichotomous data (response to treatment); standardised mean differences for continuous data (mean depression score & SD). Random effects meta-analysis. www.kcl.ac.uk/palliative

Results: study selection Records identified through database searching (n=2215)

Additional records identified through other sources (n=15)

Records after duplicates removed (n=2018) Records screened (n=141)

Records excluded (n=73)

Full-text articles assessed for eligibility (n=68)

Full-text articles excluded (n=43)

Studies included in systematic review (n=25)

Reasons: •Illness not life-threatening (n=26) •Depression not primary outcome (n=11) •Outcome measure not validated (n=3) •Prevention study (n=2) •Trial duration < 4 weeks (n=1)

Studies included in meta-analysis (n=21)

www.kcl.ac.uk/palliative

Antidepressants used in the trials Selective serotonin reuptake inhibitors (SSRIs)

Tricyclic antidepressants (TCAs)

Other antidepressants

Fluoxetine (6 trials)

Desipramine (3 trials)

Mianserin (2 trials)

Paroxetine (5 trials)

Nortriptyline (2 trials)

Mirtazapine (1 trial)

Citalopram (3 trials)

Imipramine (2 trials)

Atomoxetine (1 trial)

Sertraline (1 trial)

Doxepin (1 trial)

Fluvoxemine (1 trial) Escitalopram (1 trial)

www.kcl.ac.uk/palliative

Results: response to treatment (6-8 weeks)

Odds ratio 2.25 favouring antidepressants www.kcl.ac.uk/palliative

Results: mean depression score (6-8 weeks)

Standardised mean difference -0.54 favouring antidepressants

www.kcl.ac.uk/palliative

Results: response to treatment over time Response to treatment Time since Number of randomisation trials (patients)

Odds ratio (CI) OR>1 favours antidepressant

Number needed to treat (NNT)

4-5 weeks

5 (292)

1.93 (1.15-3.42) p=0.001

9

6-8 weeks

12 (685)

2.25 (1.38-3.67) p=0.001

6

9-18 weeks

7 (231)

2.71 (1.50-4.91) p=0.001

5

www.kcl.ac.uk/palliative

Results: Acceptability over time Number of drop-outs Time since randomisation

Number of trials (patients)

Odds ratio (CI)

4-5 weeks

3 (212)

1.00 (0.25-4.07) p=1.00

6-8 weeks

10 (597)

1.08 (0.65-1.81) p=0.76

9-18 weeks

6 (228)

2.09 (1.02-4.31) p=0.05

www.kcl.ac.uk/palliative

OR>1 favours placebo

Results: tolerability Number of adverse Number of trials events (patients)

Odds ratio (95% CI)

Dry mouth

12 (546)

2.05 (1.29-3.26) p=0.003

Nausea

11 (504)

1.51 (0.63-3.63) p=0.36)

Headache

10 (447)

1.50 (0.58-3.86) p=0.40

Dizziness

9 (595)

1.23 (0.62-2.44) p=0.55

Sexual dysfunction

5 (242)

1.92 (0.80-4.65) p=0.15

Insomnia

5 (240)

0.58 (0.10-3.29) p=0.54

www.kcl.ac.uk/palliative

OR>1 favours placebo

Results: Selective serotonin reuptake inhibitors vs. Tricyclic antidepressants Response to treatment

Acceptability

OR>1 favours antidepressant

OR