Systematic review highlights a lack of evidence about using antidepressants to treat cancer patients with depression


Having a life threatening disease such as cancer and undergoing gruelling treatment can have detrimental psychological effects. According to a recent review, for instance, the prevalence of depression among cancer patients is 10.8%, when assessed by a standardised clinical assessment (Ng et al. 2011).

The authors of this meta-analysis make the argument that established criteria for depression may not be suited to detect a depressive episode in cancer patients as these diagnostic criteria include symptoms that overlap with physical symptoms present in cancer patients such as loss of appetite or sleep disturbance. Therefore, they claim that ‘depressiveness’ – symptoms associated with depression without a diagnosis of depression:

may have adverse effects on prognosis and quality of life in cancer patients (ps. 2-3).

The aim of this systematic review and meta-analysis is to determine whether antidepressants are effective in the treatment of depression and depressive symptoms in cancer patients.


Inclusion criteria included presence of depression or depressive symptoms assessed by the application of clinical criteria or depression rating scales. The primary outcome of the studies needed to be severity of depression and depressive symptoms and secondary outcomes were number of dropouts and adverse effects and quality of life (studies included were published before end of 2010). The authors selected double-blind RCTs and for the meta-analysis placebo-controlled trials.

The authors used a random effects model and risk ratio to summarise the findings and also computed I2 to assess the overall variability attributed to the variability among studies. They used Cochrane Collaboration’s tool to detect potential risk of bias and funnel plot for publication bias.


Out of 7,000 articles, 9 studies were selected: 3 were head-to-head trials (antidepressants were compared to each other) and 6 were placebo-controlled trials and therefore included in the meta-analysis. Half of the studies included in the meta-analysis had breast cancer patients only and the sample sizes were between 35 to 193 patients.


Measuring heterogeneity helps us decide whether the trials included in the review are similar enough to be pooled in a meta-analysis

Initial analysis revealed I2 of 71% that indicated high levels of heterogeneity among studies. After the exclusion of the study by Navari et al which had a higher risk ratio than other studies, the results showed that overall relative risk (RR) was reduced from 1.56 (95% CI: 1.07-2.28, p <.05) to RR = 1.39 (95% CI: 1.09-1.77, p = .008) and I2 was 10% (acceptable level of heterogeneity). In other words, the meta-analysis of placebo-controlled trials indicated that antidepressants were more effective than placebo in creating a therapeutic response.

Subgroup analyses indicated that the antidepressant mianserin (RR = 1.67) was more effective than SSRIs (R = 1.16). Meta-analysis on adverse effects was not performed due to large amounts of missing data. They found no significant difference between dropouts in the drug condition versus placebo groups.

The overall risk of bias was determined to be ‘unclear’ which means that:

the meta-analysis cannot provide a high degree of level of evidence (p.8).

There was no evidence of publication bias.


The authors conducted a rigorous meta-analysis and followed guidelines set out by the Cochrane Collaboration in determining whether antidepressants are effective at reducing depression and depressive symptoms in cancer patients.

Overall, they found that:

antidepressants are effective in the treatment of depression/depressive symptoms in cancer patients (p. 9).

However, they issue caution: the number of studies included is very small and their quality means that we can’t obtain a high degree of evidence on the basis of this meta-analysis. They also appropriately alert the readers to the need to require a minimum of symptom severity to benefit from antidepressants so that this meta-analysis is not used as an excuse to rush to prescribe antidepressants for cancer patients.

This review

This review does not help us develop a clear set of recommendations for using antidepressants in depressed cancer patients

In reading the article, one should be cautious in terms of evidence cited throughout. For instance, the authors use Andersen et al’s study on the impact of psychological intervention on survival among breast cancer patients as an evidence for decrease in severity of depressive symptoms being associated with increased survival in cancer patients (without mentioning the intervention in question). What the authors fail to acknowledge is the well-cited analysis by Coyne et al (2007) that showed unequivocally that no study that was adequately powered to show whether a psychological intervention increases survival (as a primary endpoint) has shown this effect (this includes Andersen et al’s study: See Stefanek et al’s critique of the paper).

In addition, the authors cite Satin et al’s meta-analysis as a support for depression influencing mortality from cancer as well as cancer progression. However, Satin et al’s meta-analysis only found a predictive role for depression on mortality and not cancer progression (raising questions about the direct effect of depression on mortality), unlike how it is presented by the authors in the introduction. In addition, Satin et al also add that the effect, albeit significant, is a small one which is not apparent in the current meta-analysis.

Overall, the evidence that antidepressants are helpful for cancer patients with depression and depressive symptoms is not clear and we need to exercise caution in our assumptions about the needs of cancer patients and how we can effectively address them.


Laoutidis ZG et al Antidepressants in the treatment of depression/depressive symptoms in cancer patients: a systematic review and meta-analysis. BMC Psychiatry 2013, 13:140

Andersen BL, Yang HC, Farrar WB, Golden-Kreutz DM, Emery CF, Thornton LM, Young DC, Carson WE 3rd (2008). Psychologic intervention improves survival for breast cancer patients: a randomized clinical trial. Cancer 113: 3450–3458

Coyne JC, Hanisch LJ, et al. (2007). “Psychotherapy does not promote survival (Kissane et al., 2007): now what?” Psychooncology 16(11): 1050-1052. [PubMed abstract]

Navari RM, Brenner MC, Wilson MN (2008) Treatment of depressive symptoms in patients with early stage breast cancer undergoing adjuvant therapy. Breast Cancer Res Treat 112: 197–201 [PubMed abstract]

Ng CG, Boks MP, Zainal NZ, de Wit NJ (2011) The prevalence and pharmacotherapy of depression in cancer patients. J Affect Disord 131: 1–7. [PubMed abstract]

Satin JR, Linden W, Phillips MJ (2009). Depression as a predictor of disease progression and mortality in cancer patients: a meta-analysis. Cancer 115: 5349–5361 [PubMed abstract]

Stefanek ME, Palmer SC, Thombs BD, Coyne JC (2009). Finding what is not there: unwarranted claims of an effect of psychosocial intervention on recurrence and survival. Cancer. 115(24):5612-6. doi: 10.1002/cncr.24671. [PubMed abstract]

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+