Re-evaluating the role of antidepressants in cancer-related depression: a systematic review and meta-analysis Natalie Riblet M.D., M.P.H., Robin Larson Larson M.D., Bradley V. Watts M.D., M.P.H., Paul Holtzheimer M.D., M.S. PII: DOI: Reference:

S0163-8343(14)00116-9 doi: 10.1016/j.genhosppsych.2014.05.010 GHP 6861

To appear in:

General Hospital Psychiatry

Received date: Revised date: Accepted date:

10 March 2014 8 May 2014 13 May 2014

Please cite this article as: Riblet Natalie, Larson Robin, Watts Bradley V., Holtzheimer Paul, Re-evaluating the role of antidepressants in cancer-related depression: a systematic review and meta-analysis, General Hospital Psychiatry (2014), doi: 10.1016/j.genhosppsych.2014.05.010

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer Title: Re-evaluating the role of antidepressants in cancer-related depression: a

T

systematic review and meta-analysis

RI P

Authors: Natalie Riblet, MD MPH1,2; Robin Larson MD MPH1,3,4; Bradley V. Watts

SC

MD MPH3,4 Paul Holtzheimer, MD, MS2,3,4

Affiliations:

MA

Drive, Lebanon, NH 03766

NU

1) The Dartmouth Institute for Health Policy and Clinical Practice, 30 Lafayette

2) Dartmouth-Hitchcock Medical Center, 1 Medical Center Drive, Lebanon, NH 03766

ED

3) White River Junction VA Medical Center, 215 North Main Street, White River

PT

Junction, VT 05009

4) Geisel School of Medicine at Dartmouth, 1 Rope Ferry Drive, Hanover, NH

CE

03755

AC

Research Support: None.

Corresponding Author Information: Natalie Riblet, MD MPH, 1 Medical Center Drive, Lebanon, NH 03766; 603-650-5000; email: [email protected]

Disclaimers: The authors have no conflicts of interest to disclose.

Running Title: Review of antidepressants for depression in cancer

Key words: cancer, depression, antidepressive agents, review

-1-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer

NU

SC

RI P

T

Highlights  We reviewed the efficacy and tolerability of antidepressants in cancer patients.  Paroxetine, fluoxetine and mianserin improve cancer-related depression.  Paroxetine, fluoxetine and mianserin may vary in efficacy and tolerability.  High quality, randomized trials of newer antidepressants are needed.

ABSTRACT

MA

Objective: Prior reviews evaluating the role of antidepressants in cancer-related depression have drawn conflicting conclusions. These reviews have also not explored

ED

differences in efficacy and tolerability between antidepressants. We conducted a metaanalysis to address these limitations.

PT

Method: We searched Medline (1948-2013), the Cochrane Library (1800-2013),

CE

CINAHL (1986-2013), ClinicalTrials.gov (2013) and meeting abstracts. We included randomized trials comparing antidepressants to placebo or no treatment for cancer-

AC

related depression. We used random effects to calculate standardized mean differences (SMD).

Results: Of 5,178 potentially eligible citations, 9 trials (1,169 subjects) met inclusion criteria. Trials of mianserin found a robust reduction in depression scores at ≥ 4 weeks of treatment (SMD: 0.60,95%CI: 0.24 – 0.95). Similar, but less robust results were observed with paroxetine (SMD: 0.22, 95%CI: 0.01 – 0.42) and fluoxetine (SMD 0.34, 95%CI: 0.02 – 0.66). Conversely, there was no advantage with amitriptyline or desipramine. Compared to placebo, the odds of dropping out due to side effect were higher with fluoxetine and paroxetine and lower with mianserin. Methodological quality was moderate.

-2-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer Conclusions: Paroxetine, fluoxetine and mianserin improve cancer-related depression but may vary in efficacy and tolerability. High quality, randomized trials of newer

T

antidepressant agents are needed to identify optimal treatments for managing cancer-

SC

RI P

related depression.

NU

BACKGROUND:

MA

Depressive symptoms are common in cancer, affecting as many as half of patients and ranging from self-limited adjustment disorder to major depressive disorder [1,2]. A number of factors likely contribute to the risk of depression in cancer patients including

ED

difficulties with psychosocial adjustment and changes in neuroendocrine functioning

PT

[2]. Left untreated, depression can lead to a number of adverse outcomes including reduced quality of life and higher healthcare costs [1, 3-6]. A 2010 meta-analysis also

CE

found that co-morbid depression was associated with a 22% increased risk for cancerrelated death [5]. Similarly, cancer patients are twice as likely as the general population

AC

to commit suicide [6].

While antidepressants are widely prescribed and proven treatments for depression in non-cancer patient populations, their role in patients with cancer remains unclear. Although some studies suggest they reduce depressive symptoms [7-9], other studies find no advantage over placebo [10-12]. In addition, concerns have been raised that various antidepressants may be less well tolerated by cancer patients [13-14]. It is also unclear whether the benefits of antidepressants are limited to those with severe depression [9,12]. Alternatively, cancer patients may uniquely benefit from antidepressant medications because they also address other common cancer-related symptoms such as insomnia and fatigue [7, 15-16].

-3-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer Multiple prior reviews have evaluated antidepressant efficacy in cancer or physically ill populations but these reviews have drawn conflicting conclusions [1, 3-4,

T

17-23]. Among the more recent meta-analyses, three concluded that antidepressants are

RI P

superior to placebo in physically ill and palliative care populations but these reviews were not focused on patients with cancer [21-23]. Two reviews of cancer-related

SC

depression also suggested that antidepressants are favored over placebo [19, 24]. These prior reviews and meta-analysis leave unresolved two important clinical issues: which

NU

antidepressant agents are most effective in cancer-related depression and which

MA

antidepressant is most tolerable in cancer patients. Resolving these questions would help identify the specific antidepressant agents that are most useful in cancer-related depression.

ED

To address these questions, we conducted a meta-analysis to examine the relative

PT

efficacy and tolerability of antidepressants used for cancer-related depression. Our aim is to provide clinicians with specific and actionable guidance when treating cancer-

METHODS:

CE

related depression.

AC

Study Eligibility Criteria We defined a priori inclusion criteria for this review. Included studies required (1) randomized or quasi-randomized design of (2) patients with active cancer of any type and stage who had co-morbid depression of any severity (3) treated with antidepressant medications [25] (4) for a duration of at least 4 weeks (5) compared to placebo or no treatment and (6) reporting depression outcomes. We excluded populations of patients whose cancer was in remission because the current controversy centers around patients with active disease [3, 18]. Outcome measures Primary Outcome-Depressive Symptoms:

-4-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer We specified improvement in depressive symptoms as our primary outcome of interest and defined this as a mean change from baseline in depression scores on a

T

validated depression rating scale. In the event that a trial used multiple rating scales [7,

RI P

26-27], we selected scales that were used across included studies and have been well studied in the cancer population. In addition to our main analysis of depression

SC

symptoms as a continuous measure, we also evaluated depression as a dichotomous outcome by looking at the proportion of subjects with symptom response. We accepted

NU

response thresholds as defined by the individual studies. We intended to evaluate rates

MA

of depression remission but too few studies reported on this outcome. Secondary Outcome-Treatment Tolerability:

Our secondary outcome was treatment tolerability, which we defined as the

ED

proportion of subjects who dropped out due to side effect. While we planned to further

PT

subgroup these results by type of side effect, we were unable to perform this analysis due to limitations in reporting tolerability across included studies.

CE

Search Methods

Databases, search terms, limits and special strategies

AC

We searched Medline (1948-December 31, 2013), the Cochrane Library (1800December 31, 2013), PsycINFO (inception – December 31, 2013), and Cumulative Index to Nursing and Allied Health Literature (CINAHL) (1986 – December 31, 2013) to identify published articles addressing our research question. We used exploded MeSH terms and key words to create sets for our two main themes: cancer and antidepressants. We then used the Boolean term “AND” to determine their intersection and to locate pertinent studies. We applied no limits or language restrictions. Additional Search Methods: We attempted to locate additional published and unpublished studies by reviewing references of included articles, searching ClinicalTrials.gov (as of December 31, 2013)

-5-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer and reviewing abstracts of the American Psychosocial Oncology Society Meetings (2010-2013).

T

Study Selection

RI P

Based on a priori inclusion criteria, one non-blinded reviewer screened the titles and abstracts of all potentially relevant studies, excluding those that were clearly ineligible

SC

for the review. The same reviewer then assessed the full text of remaining studies to make a final determination regarding eligibility for review.

NU

Data Collection

MA

Two non-blinded independent reviewers extracted data in duplicate from included studies. We translated two foreign articles into English through a translator [28-29]. We

Analysis

PT

Depression Outcomes

ED

extracted data related to demographics, methods, and risk of bias.

For consideration of a continuous measure of depressive symptoms, we pooled mean

CE

changes in depression scores using standardized mean difference (SMD) (i.e Hedges’s adjusted g) and 95% confidence intervals (CI) because studies used different validated

AC

instruments to assess depression. For examination of dichotomous outcomes, we pooled studies using relative risk (RR) and 95% CI. Tolerability Due to low event rates, we used Peto odds ratio (OR) and 95%CI to pool drop outs due to side effects [30]. Data synthesis We used Cochrane’s Review Manager software Revman 5.1[31] to pool our quantitative results. We analyzed the results for our primary outcome by comparing improvements in depressive symptoms from baseline to 4 weeks or more of treatment. Assessment of heterogeneity

-6-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer We used Revman 5.1[31] to assess for heterogeneity and to determine the appropriateness of combining quantitative results. We relied on a conventional

T

threshold of a p 50% to indicate significant heterogeneity was present.

RI P

We first pooled the results of our primary and secondary outcomes using a fixed effects model. Due to significant heterogeneity among some of the studies reporting on

SC

depression, we presented our results using a random effects model. This is a more conservative approach for identifying statistically significant effects.

NU

Assessment of reporting biases

MA

We created a funnel plot for our primary outcome of interest of depression in order to determine whether publication bias may have affected reported results. We evaluated whether a correlation existed between study size and effect size to assess for publication

ED

bias.

PT

RESULTS Description of Studies

CE

Results of Search

We identified 5,178 potentially eligible records, of which 4,700 remained after

AC

removing duplicates and ongoing studies. As outlined in Figure 1, 183 records remained after title and abstract screening, and 9 trials met inclusion criteria based on full text review [7-11, 26-29] . Included Studies Table 1 describes the characteristics of 9 randomized trials included in the review [711, 26-29]. All were published between 1985 and 2011. Studies represented a total of 1,169 patients from various countries and with varying severity of depression. Eight trials compared active drug to placebo and one trial compared active drug to no treatment [28]. Most studies included multiple cancer types but 3 were restricted to breast cancer [8, 11, 27]. Overall, 83% of subjects were female and the mean age was

-7-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer 54.3 years. Treatment duration ranged from 4 to 12 weeks. Attrition was high with only 2 trials reporting < 20 % loss to follow-up [26, 28].

T

Excluded Studies

RI P

Two additional trials would have met inclusion criteria but were excluded because we were unable to obtain continuous outcomes measures for depression despite attempts to

SC

contact these authors [12, 32]. Effects of Interventions

NU

Primary Outcome

MA

Depressive Symptoms-Continuous

As shown in Figure 2, two trials of mianserin found robust and statistically significant improvements in depressive symptoms when compared to placebo at ≥ 4

ED

weeks of treatment (SMD: 0.60,95%CI: 0.24 – 0.95) [7, 8].

PT

Three trials of fluoxetine [9, 10, 28] suggested that treatment was favored over placebo (SMD: 0.71, 95%CI: 0.00 -1.42). There was significant heterogeneity,

CE

however, among included studies (I2=84%, p < 0.05). After removing one trial due to poorer methodological quality including the absence of a placebo arm [28], the two

AC

homogeneous trials (I2=0%) continued to favor fluoxetine, although the results were less robust (SMD 0.34, 95%CI: 0.02 – 0.66) [9-10]. Three heterogeneous studies of paroxetine (I2=57%) suggested no difference in depressive symptom improvement with treatment versus placebo (SMD: 0.14, 95%CI: – 0.22 - 0.50) [11, 26-27]. However, after removing the small, lower quality study [11], the heterogeneity resolved (I2=18%) and paroxetine was favored over placebo (SMD: 0.22, 95%CI: 0.01 – 0.42) [26-27]. The two small studies of amitriptyline [29] and desipramine [11] found no significant differences in improvements in depressive symptoms between treatment and placebo. Depressive Symptoms-Dichotomous

-8-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer Depression response rates were significantly greater with mianserin versus placebo at

≥ 4 weeks of treatment (RR: 1.67, 95%CI: 1.22; 2.28) [7-8]. A similar trend was seen in

T

the two fluoxetine trials reporting on depression response rates, although the findings

RI P

were not significant (RR: 1.22, 95%CI: 0.85; 1.74) [9-10]. A trial of paroxetine and a trial of desipramine both suggested that there was no significant benefit in depression

SC

rates when comparing antidepressant treatment and placebo (RR: 0.49, 95%CI: 0.13; 1.92 and RR: 0.83, 95%CI: 0.36; 1.94, respectively) [11].

NU

Secondary Outcomes

MA

Tolerability

As shown in Figure 3, mianserin was associated with a lower, although non-

ED

significant, risk of dropping out due to side effect versus placebo (OR 0.57, 95% CI: 0.14, 2.14) [7-8]. Two trials of fluoxetine, however, showed an increased risk of study

PT

drop-out with fluoxetine compared to placebo (OR: 5.06, 95%CI: 1.64, 15.64) [9-10]. After removing the smaller study [10] due to heterogeneity (I2 =60%), placebo

CE

continued to be favored (OR: 1.92, 95%CI: 0.38, 9.74). Three trials of paroxetine also

AC

showed an increased risk of drop outs with paroxetine compared to placebo (OR: 1.54, 95%CI: 0.57, 4.15) [11, 26-27]. A trial of desipramine suggested that treatment was associated with lower risk of dropping out due to side effect (OR: 0.40, 95% CI: 0.04, 4.30) [11]. Publication Bias Based on review of a funnel plot for included studies reporting changes in mean depression scores at ≥ 4 weeks, we did not find evidence to suggest publication bias. The results are fairly symmetrical around the mean, arguing against the existence of a correlation between study size and effect size (see Figure 4) [7-11, 26-29]. The

-9-

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer interpretation, however, is limited by the paucity of overall studies contributing to the primary outcome.

T

DISCUSSION

RI P

Summary of main results

The alpha-2-adrenergic receptor antagonist, mianserin, and the selective serotonin

SC

reuptake inhibitors (SSRIs), fluoxetine and paroxetine, appear to be effective in treating

NU

symptoms of cancer-related depression. Mianserin also showed a higher depression response rate compared to placebo whereas paroxetine and fluoxetine did not. The

MA

finding of low response rates, even among agents that separated from placebo, suggests that only modest changes in depressive symptoms may be typical with antidepressant

ED

treatment in this population.

There were observed differences in tolerability between the three effective agents:

PT

mianserin showed a non-significant trend toward lower rate of drop out than placebo,

CE

paroxetine showed a non-significant trend toward higher drop-out than placebo, and fluoxetine had significantly higher-drop out than placebo.

AC

Overall completeness and applicability of evidence There is a scarcity of evidence to address the role of different antidepressants in cancer-related depression. Studies of alpha-2-adrenergic receptor antagonists, TCAs and monoamine-oxidase inhibitors (MAOIs) are generally underrepresented [7, 8, 11, 29, 33-34]. The paucity of studies of TCAs and MAOIs may reflect their reputation for having more significant side-effect profiles, which would be particularly problematic in patients with cancer [25]. The only alpha-2-adrenergic receptor antagonist included in this review was mianserin which is not available for sale in the United States. Although a related compound, mirtazapine, is available [35], we are not aware of any placebocontrolled RCTs that have been conducted in patients with cancer-related depression.

- 10 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer While included studies are generalizable to various settings, countries and cancers, they provide limited information about the relationship between antidepressants and

T

variables such as cancer type and prior psychiatric history. Similarly, the evidence of

RI P

antidepressants may be limited to patients with severe cancer-related depression [3, 12, 21]. There is insufficient data available, however, to examine other severities of

SC

depression. Several studies also excluded patients with a projected survival less than 3 months [9-10]. Males were generally underrepresented in the studies [7-8, 10-11, 26-27,

NU

29]. Few studies explored the role of antidepressants in addressing other cancer-related

MA

outcomes such as quality of life [9-10] and survival [7, 9]. Thus, any conclusions about these sub-populations of cancer patients or other important outcomes are not possible. Limitations of the evidence

ED

Our conclusions are limited by the quality and amount of evidence available

PT

regarding the treatment of cancer related depression. Few antidepressants had been studied in even a single randomized clinical trial. Those agents that have been studied

CE

were examined in a limited number of small studies. Thus, the overall amount of evidence is limited. There were also a number of limitations related to the design of the

AC

studies we included. Antidepressants may not have always been optimally dosed for therapeutic effect [36]. This concern was compounded by the fact that trials were generally relatively short in their duration. Many studies experienced high attrition rates and none of the included studies controlled for co-administration of behavioral interventions. Thus, it is possible that some unknown behavioral intervention was responsible for the improved depression. One study also compared active drug to no treatment [28]. While each of these factors could threaten the ability to draw valid conclusion, most would have resulted in an under-estimation of the true effect of antidepressants on cancer-related depression. Most authors also described intention-to-treat analysis [7-8,

- 11 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer 11, 26-28] and only two studies reported significant differences between study completers and non-completers [7-8]. In both cases, differences were driven by higher

T

rates of drop out in the placebo arm due to a perceived lack of efficacy [7-8]. Although

RI P

reported side effects did not consistently differ between antidepressants and placebo for any of the drug classes, dropouts due to side effects appeared to be more common

SC

among patients receiving certain types of SSRIs.

Adverse drug-drug interactions between chemotherapeutic agents and antidepressants

NU

may also influence the tolerability of antidepressants in patients with cancer [14, 37-

MA

39]. Notably, tamoxifen, which plays an important therapeutic role in the treatment of breast cancer [40], is converted to its active metabolite through the cytochrome P450 2D6 pathway [41]. Since SSRIs inhibit the CYP2D6 pathway to varying degrees [42],

ED

concurrent treatment with tamoxifen may lead to an increased risk for cancer recurrence

PT

and mortality [37]. The potency of CYP2D6 inhibition may also contribute to risk with paroxetine, a strong inhibitor, carrying a much greater risk than citalopram, a weak

CE

inhibitor [37-38]. For example, a population based cohort study found an increased risk for mortality in patients treated concurrently with tamoxifen and paroxetine [14]. A

AC

recent review by Callari et al, however, concluded that there is no evidence to suggest that concurrent treatment with tamoxifen and SSRIs contributes to a heightened risk for breast cancer recurrence [37]. We were unable in our review to explore to what degree (if any) drug-drug interactions may have impacted treatment adherence, disease prognosis or survival. There have been three prior meta-analyses of antidepressant efficacy in cancerrelated depression [17, 19, 24]. While these reviews favored antidepressants over placebo, they differed in their opinion about whether the findings were statistically significant. Given that these reviews were limited to a few select trials and did not explore in depth the role of type of antidepressant or tolerability on outcomes, it is

- 12 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer difficult to compare them directly to our findings. Our review, however, does support antidepressants as an efficacious treatment for cancer related depression.

T

There have also been several meta-analysis of antidepressant efficacy and tolerability

RI P

in physically ill populations (ie. cardiovascular disease, pulmonary disease, stroke, etc) with co-morbid depression randomized to antidepressant or placebo. Consistent with

SC

our review, two of these meta-analyses suggested that SSRIs are more effective than placebo in treating depression, however, have delayed treatment onset and are not as

NU

well-tolerated [21, 23]. Furthermore, one of the reviews found that mianserin and

MA

mirtazapine appeared to be more effective than SSRIs and better tolerated [21]. Contrary to our findings, however, TCAs were found to be associated with a robust and significant improvement in depressive symptoms [21, 23]. This discrepancy may reflect

ED

the limited number of TCA trials included in our review and/or differences in the

PT

populations studied.

Our findings suggest that there is a need for high quality randomized clinical trials

CE

that explore the role of antidepressants in treating cancer-related depression. Alpha-2adrenergic receptor antagonists show particular promise in cancer patients, possibly due

AC

to their unique pharmacological profile. These agents promote norepinephrine and serotonin transmission in the central nervous system by blocking presynaptic alpha-2adrenergic receptors and postsynaptic serotonin type 2 (5-HT2) and type 3 (5-HT3) receptors [43]. They may be better tolerated than TCAs or SSRIs due to their lower affinity for muscarinic and dopaminergic receptors coupled with a higher affinity for histamine-H1 receptors and 5-HT2AC and 5-HT3 receptors [43]. Thus, alpha-2adrenergic receptor antagonists act as effective anti-depressant, anti-emetic and anxiolytic agents, while avoiding other common serotonin-related side effects such as headache, agitation, jitteriness or sexual dysfunction [43]. Although their high affinity for histamine H1 receptors contributes to sedation, this effect may be mitigated by an

- 13 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer increase in norepinephrine neurotransmission [44]. Several studies of mirtazapine in cancer patients have also found improvements in cancer-related cachexia, nausea,

T

insomnia, and even hot-flashes [16, 45-47].

RI P

Given that current randomized trials of alpha-2-adrenergic receptor antagonists in the cancer population are limited to mianserin, it is unclear whether similar benefits

SC

would be observed with other agents in this drug class. While mirtazapine is the pharmacological analogue of mianserin and may have a similar mechanism of action,

NU

there may be important differences between these two medications especially in their

MA

treatment of cancer related depression [48]. CONCLUSIONS

Our review suggests that several antidepressants including one alpha-2-adrenergic

ED

antagonist, mianserin, and two SSRIs, fluoxetine and paroxetine, are generally effective

PT

in treating symptoms of cancer-related depression. When selecting antidepressants providers should consider several characteristics (in addition to efficacy of the

CE

antidepressant agent) such as side effects, time to therapeutic onset and cost of treatment. It is possible that alpha-2-adrenergic receptor antagonists may be more

AC

effective and better tolerated than SSRIs in the cancer population. Available evidence, however, rests entirely on RCTs of a single agent, mianserin, which is not available for sale in the United States. Evidence for the efficacy of other antidepressant agents including TCAs in cancerrelated depression is lacking. Similarly, the role of antidepressants such as selective norepinephrine reuptake inhibitors (SNRIs) and norepinephrine reuptake inhibitors (NRIs) remains largely unknown. It is also unclear whether depression severity, gender, cancer type or stage may impact antidepressant efficacy and tolerability.

- 14 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer REFERENCES: 1. Reich M. Depression and cancer: recent data on clinical issues, research

T

challenges and treatment approaches. Curr Opin Oncol 2008; 20(4): 353-9.

RI P

2. Pasquini M, Biondi M. Depression in cancer patients: a critical review. Clinical Practice and Epidemiology in Mental Health 2007; 3:2.

SC

3. Rodin G, Lloyd N, Katz M, Green E, Mackay JA, Wong RKS. The treatment of depression in cancer patients: a systematic review. Support Care Cancer 2007;

NU

15(2): 123-36.

2010; 14(5): 395-9.

MA

4. Irving G, Lloyd-Williams M. Depression in advanced cancer. Eur J Oncol Nurs

5. Pinquart M, Duberstein PR. Depression and cancer mortality: a meta-analysis.

ED

Psychol Med 2010; 40(11): 1797-810.

PT

6. Misono S, Weiss NS, Fann JR, Redman M, Yueh B. Incidence of suicide in persons with cancer. J Clin Oncol 2008; 26(9): 4731-8.

CE

7. Costa D, Mogos I, Toma T. Efficacy and safety of mianserin in the treatment of depression in women with cancer. Acta Psychiatr Scand 1985; 72(suppl 320):

AC

85-92.

8. Van Heeringen K, Zivkov M. Pharmacological treatment of depression in cancer patients: a placebo-controlled study of Mianserin. Brit J Psychiat 1996; 169(4): 440-3. 9. Fisch MJ, Loehrer PJ, Kristeller J, Passik S, Jung SH, Shen J, et al.Fluoxetine versus placebo in advanced cancer outpatients: a double-blinded trial of the Hoosier oncology group. J Clin Oncol 2003; 21(10): 1937-43. 10. Razavi D, Allilaire J-F, Smith M, Salimpour A, Verra M, Desclaux B, et al.. The effect of fluoxetine on anxiety and depression symptoms in cancer patients. Acta Psychiatr Scand 1996; 94(3): 205-10.

- 15 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer 11. Musselman DL, Somerset WI, Guo Y, Manatunga AK, Porter M, Penna S, et al.. A double-blind, multicenter, parallel-group study of paroxetine,

T

desipramine, or placebo in breast cancer patients (stage I, II, III, and IV) with

RI P

major depression. J Clin Psychiatry 2006; 67(2): 288-96.

12. Stockler MR, O’Connell R, Nowak AK, Goldstein D, Turner J, Wilcken NR, et

SC

al. Effect of sertraline on symptoms and survival in patients with advanced cancer, but without major depression: a placebo-controlled double-blind

NU

randomized trial. Lancet Oncol 2007; 8(7): 603-12.

MA

13. Pae CU, Kim YJ, Won WY, Kim HJ, Lee S, Lee CU, et al. Paroxetine in the treatment of depressed patients with haematological malignancy: an open-label study. Hum Psychopharmacol Clin Exp 2004; 19(1): 25-9.

ED

14. Kelly CM, Juurlink DN, Gomes T, Duong-Hua M, Pritchard KI, Austin PC, et

PT

al. Selective serotonin reuptake inhibitors and breast cancer mortality in women receiving tamoxifen: a population based cohort study. BMJ 2010; 340: c693.

CE

15. Moss EL, Simpson SA, Pelletier G, Forsyth P. An open-label study of the effects of bupropion SR on fatigue, depression, and quality of life of mixed-site

AC

cancer patients and their partners. Psycho-Oncol 2006; 15(3): 259-67.

16. Cankurturan ES, Ozalp E, Soygur H, Akbiyik DI, Turhan L, Alkis N. Mirtazapine improves sleep and lowers anxiety and depression in cancer patients: superiority over Imipramine. Support Care Cancer 2008; 16(11): 1291-8. 17. Iovienio N, Tedeschini E, Ameral VE, Rigatelli M, Papakostas GI. Antidepressants for major depressive disorder in patients with a co-morbid axis-III disorder: a meta-analysis of patient characteristics and placebo response rates in randomized controlled trials. Int Clin Psychopharm 2011; 26(2): 69-74.

- 16 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer 18. Ng C, Boks MPM, Zainal NZ, de Wit N. The prevalence and pharmacotherapy of depression in cancer patients. J Affect Disorders 2011; 131(1-3): 1-7.

T

19. Hart SL, Hoyt MA, Diefenbach M, Anderson DR, Kilbourn KM, Craft LL, et

RI P

al.Meta-analysis of efficacy of interventions for elevated depressive symptoms in adults diagnosed with cancer. J Natl Cancer Inst 2012; 104(13): 990-1004.

SC

20. Ujeyl M, Muller-Oerlinghausen B. Antidepressants for treatment of depression in palliative patients: a systematic literature review. Der Schmerz 2012; 26(5):

NU

523-36.

MA

21. Rayner L, Price A, Evans A, Valsraj K, Higginson IJ, Hotopf M. Antidepressants for depression in physically ill people (review). Cochrane DB Syst Rev 2010; (3). doi: 10.1002/14651858.CD007503.

ED

22. Rayner L, Price A, Evans A, Koravangattu V, Hotopf M, Higginon J.

PT

Antidepressants for the treatment of depression in palliative care: a systematic review and meta-analysis. Palliat Med 2011; 25(1): 36-51.

CE

23. Taylor D, Meader N, Bird V, Pilling S, Creed F, Goldberg D. Pharmacological interventions for people with depression and chronic physical health problems:

AC

systematic review and meta-analysis of safety and efficacy. Br J Psychiatry 2011; 198(3): 179-88.

24. Laoutidis ZG, Mathiak K. Antidepressants in the treatment of depression/depressive symptoms in cancer patients: a systematic review and meta-analysis. BMC Psychiatry 2013; 13: 140-60. 25. Kim HF, Fisch MJ. Antidepressant use in ambulatory cancer patients. Current Oncol Rep 2006; 8(4):275-81. 26. Morrow GR, Hickok JT, Roscoe JA, Raubertas RF, Andrews PL, Flynn PJ, et al. Differential effects of paroxetine on fatigue and depression: a randomized,

- 17 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer double-blind trial from the University of Rochester Cancer Center Community Clinical Oncology Program. J Clin Oncol 2003; 21(24): 4635-41.

T

27. Roscoe JA, Morrow GR, Hickok JT, Mustian KM, Griggs JJ, Matteson SE, et

RI P

al. Effect of paroxetine hydrochloride (paxil) on fatigue and depression in breast cancer patients receiving chemotherapy. Breast Cancer Res Tr 2005;

SC

89(3): 243-9

28. Wang D-S, Li G-L, Chen J-H, Liu X-M. Effects of psychological interventions

NU

in cancer patients undergoing radiotherapy. Journal of Chinese Psychology

MA

2011; 19(4): 561-3.

29. del Carmen LM, Plancarte R, de la Fuente JRU. La amitriptilina comocoanalgesico en pacientes con cancer. Salud Mental Volume 1990: 13(4):

ED

1-6.

PT

30. Higgins JPT, Deeks JJ, Altman DG (editors). Chapter 16: Special topics in statistics. In Higgins JPT, Green S (editors). Cochrane Handbook for

CE

Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. www.cochrane-handbook.org. Accessed

AC

October 1, 2012.

31. Review Manager (RevMan) [Computer program]. Version 5.1. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011. 32. Navari RM, Brenner MC, Wilson MN. Treatment of depressive symptoms in patients with early stage breast cancer undergoing adjuvant therapy. Breast Cancer Res Tr 2008; 112(1): 197-201. 33. Uzer Y, Shnider BI, Gold GL. A double blind study with iproniazid in patients with far-advancd cancer. Antibiotic Medicine & Clinical Therapy 1960; 7: 77781.

- 18 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer 34. Purohit DR, Navlakha PL, Modi RS, Eshpumiyani R. The role of antidepressants in hospitalized cancer patients. Jr Asso Phys Ind 1978; 26: 245-

T

8.

RI P

35. Croom KF, Perry CM, Plosker GL. Mirtazapine: a review of its use in major depression and other psychiatric disorders. CNS Drugs 2009; 23(5): 427-52.

SC

36. Solvason HB, DeBattista C. Antidepressant dosing for the acute treatment of unipolar depression. Primary Psychiatry 2009; 16(10): 30-6.

NU

37. Callari A, Mauri M, Miniati M, Mancino M, Bracci G, Dell’Osso L, et al.

Tumori 2013; 99: 623-33.

MA

Treatment of depression in patients with breast cancer: a critical review.

38. Jin Y, Desta Z, Stearns V, Ward B, Ho H, Lee K-H, et al. CYP2D6 genotype,

ED

antidepressant use, and tamoxifen metabolism during adjuvant breast cancer

PT

treatment. J Natl Cancer Inst 2005; 97(1): 30-9. 39. Lash TL, Cronin-Fenton D, Ahern TP, Rosenberg CL, Lunetta KL, Silliman

CE

RA, et al. Breast cancer recurrence risk related to concurrent use of SSRI antidepressants and tamoxifen. Acta Oncol 2010; 49(3): 305-12.

AC

40. den Hollander P, Savage MI, Brown PH. Targeted therapy for breast cancer prevention. Frontiers in Oncology 2013; 250(3): 1-15.

41. Crewe HK, Notley LM, Wunsch RM, Lennard MS, Gillam EM. Metabolism of tamoxifen by recombinant human cytochrome P450 enzymes: formation of the 4-hydroxy, 4’-hydroxy and n-desmethyl metabolites and isomerization of trans4-hydroxytamoxifen. Drug Metab Dispos 2002; 30(8): 869-74. 42. Crewe HK, Lennard MS, Tucker GT, Woods FR, Haddock RE. The effect of selective serotonin re-uptake inhibitors on cytochrome P4502D6 (CYP2D6) activity in human liver microsomes. Br J Clin Pharmacol 1992;34:262–5.

- 19 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer 43. Fawcett J, Barkin RL. Review of the results from clinical studies on the efficacy, safety and tolerability of mirtazapine for the treatment of patients with

T

major depression. J Affect Disorders 1998;51(3): 267-85.

RI P

44. Puzantian T. Mirtazapine, an antidepressant. Am J Health-Syst Pharm 1998; 55(1): 44-9.

SC

45. Kim SW, Shin IS, Kim JM, Kim YC, Kim KS, Kim KM, et al. Effectiveness of mirtazapine for nausea and insomnia in cancer patients with depression.

NU

Psychiat Clin Neuros 2008; 62(1): 75-83.

MA

46. Riechelmann RP, Burman D, Tannock IF, Rodin G, Zimmermann C. Phase II trial of mirtazapine for cancer-related cachexia and anorexia. Am J Hosp Palliat Care 2010;27(2): 106-10.

ED

47. Biglia N, Kubatzki F, Sgandurra P, et al. Mirtazapine for the treatment of hot

PT

flushes in breast cancer survivors: a prospective pilot trial. Breast J 2007;13(5):490-5.

CE

48. De Boer T, Nefkens F, Van Helvoirt A, Van Delft AM. Differences in modulation of noradrenergic and serotonergic transmission by the alpha-2

AC

adrenoreceptor antagonists, mirtazapine, mianserin and idazoxan. J Pharmacol Exp Ther 1996; 277(2):852-60.

- 20 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer Legends/Titles for included figures:

T

Figure 1 PRISMA Flow Diagram

RI P

APOS: American Psychosocial Oncology Society; CINAHL = Cumulative Index to

SC

Nursing and Allied Health Literature

Figure 2: Efficacy of antidepressant therapy versus placebo or no treatment on mean

NU

change in depressive scores at ≥ 4 weeks of treatment.*

MA

Black square = individual effect size of study Diamond = mean weighted effect size

ED

*Summary estimate for fluoxetine and paroxetine are presented after the outliers (Wang

PT

and Musselman) have been removed

placebo.*

CE

Figure 3: Odds of dropping out due to side effect with antidepressant therapy versus

AC

Black square = individual effect size of study Diamond = mean weighted effect size *Summary estimate for fluoxetine is presented after the outlier (Razavi) has been removed

Figure 4: Funnel plot assessing for publication bias in studies reporting mean changes in depressive symptoms for antidepressant therapy versus placebo or no treatment at

≥ 4 weeks of treatment. SE = Standard Error; SMD = Standardized mean differences

- 21 -

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

Review of antidepressants for depression in cancer

Figure 1

- 22 -

ACCEPTED MANUSCRIPT

ED

MA

NU

SC

RI P

T

Review of antidepressants for depression in cancer

AC

CE

PT

Figure 2

- 23 -

ACCEPTED MANUSCRIPT

ED

MA

NU

SC

RI P

T

Review of antidepressants for depression in cancer

AC

CE

PT

Figure 3

- 24 -

ACCEPTED MANUSCRIPT

ED

MA

NU

SC

RI P

T

Review of antidepressants for depression in cancer

AC

CE

PT

Figure 4

- 25 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer

Alpha-2-Adrenergic Receptor Antagonists Rom ania/ 51. Costa Hosp 6 1 Mi Stage et al, ital 7 yrs 0 xed 2-4 1985 & 3 (10. 0 Clini 8) c

52 yrs (8)

1 0 0

Bre ast

4 week s

30

None

XRT (100)

Mianse rin NR (3060mg/ day)

Plac ebo

6 week s

38

NV Organ on

MAD RS 26.6 (7.4)

NR

Fluoxe tine 20mg/ day

Plac ebo

5 week s

24

Lilly Franc e/ Benel ux

Adva nced canc er

BZS DS 23.8 (6.3)

Chemo (64) XRT (15)

Fluoxe tine 20mg/ day

Plac ebo

12 week s

20

Eli Lilly

NR

CESD 15.3 (10)

Chemo (100)

Paroxe tine 20mg/ day

Plac ebo

8 week s

13

Glaxo Smith Kline

NR

CESD≥ 19 (28% )

Chemo (100)

Paroxe tine 20mg/ day

Plac ebo

8 week s

23

Glaxo Smith Kline

Stage 1 &2

AC

CE

Selective Serotonin Reuptake Inhibitors Belg 52. ium/ 91% 9 Razav Fran 9 8 Mi with yrs i et al, ce/ 1 0 xed no (11. 1996 Clini mets 3) c Fisch et al, 2003

USA / Clini c

1 6 3

60. 0 yrs (N R)

6 1

Mi xed

Morr ow et al, 2003

USA / Clini c

5 4 9

56. 4yr s (12. 4)

7 6

Mi xed

USA / Hosp ital & Clini c

1 2 2

52. 2yr s (9.7 )

1 0 0

Rosco e et al, 2005

Bre ast

Indust ry Spons or

Plac ebo

HDR S 20.7 (3.7)

MA

5 5

Con trol

Attri tion (%)

Chemo &/or XRT (97)

ED

Neth erlands / Hosp ital

Current Tx (%)b

Treat ment Leng th

Mianse rin 44.5m g/day (3060mg/ day)

PT

Van Heeri ngen et al, 1996

Depre ssion mean (SD)a

RI P

F )

Canc er Stage (s)

Interve ntion Mean Dose/F req (Range )

SC

N

Ca nce r Ty pe

NU

Autho r, year

Cou ntry/ Setti ng

Ag e me an (SD )

S e x ( %

T

Table 1: Baseline characteristics of randomized trials evaluating antidepressants in cancer-related depression

HDR S 21.3 (4.5)

- 26 -

ACCEPTED MANUSCRIPT Review of antidepressants for depression in cancer

Chin a/ Hosp ital

6 2

60. 2 yrs (N R)

3 9

Mi xed

NR

HDR S 18.4 (6.2)

XRT (100)

Fluoxe tine 20mg/ day

8 week s

6

None

RI P

Selective Serotonin Reuptake Inhibitors and Tricyclic Antidepressants

No Dru gd

T

Wang et al, 2011c

3 5

54. 1yr s (12. 2)

1 0 0

Bre ast

Stage 1-4

HDR S 22.5 (5.6)

Chemo (17) XRT (3)

HDR S 19.6 (6.3)

Palliativ e Treatme nt(100)

MA

USA / Clini c

Mexi co/ Clini c

1 9

49. 1 yrs (14. 8)

AC

del Carm en et al, 1990

CE

Tricyclic Antidepressants

PT

ED

Muss elman et al, 2005

NU

SC

Arm 1: Desipr amine 113mg /day (50175mg /day)

9 5

Mi xed

NR

Arm 2: Paroxe tine 31mg/ day (2040mg/ day)

Amitri ptyline NR (25100mg /day)

Plac ebo

6 week s

40

Glaxo Smith Kline

Plac ebo

4 week s

58

None

BZSDS = Brief Zung Self-Rating Rating Depression Scale; CES-D: Center for Epidemiologic Studies Depression Scale; Chemo = chemotherapy; F = female; freq: frequency; HDRS: Hamilton Depression Rating Scale; MADRS: Montgomery-Asberg Depression Rating Scale; mets = metastasis; mg = milligram; mixed = various cancer diagnoses; N = number; NR = not reported; % = percent; SD = standard deviation; Tx = treatment; XRT = radiation; yrs = years; a Severity of depression at baseline as measured by clinical scales or diagnostic exam; higher scores on scales mean more severe depression b Type of cancer treatment received in past 30 days c Trial used quasi-randomization d Additional arm received music therapy only and was excluded from this meta-analysis

- 27 -