A diagnosis of cancer is a significant life threat and psychosocial stressor, often leading to emotional distress in vulnerable individuals. Frequently, depressive syndromes are under-recognized in cancer patients; this is partly due to the difficulty of distinguishing depressive syndromes from adjustment disorders in response to cancer, and also because the neurovegetative and cognitive symptoms of depression may be considered normative in the context of cancer and its treatment. Conversely, dysphoric or apathetic cancer patients are sometimes misdiagnosed as depressed when they are in fact delirious.

DEPRESSION AND RELATED DISORDERS

Cancer increases prevalence of certain psychiatric symptoms and syndromes, including depressive disorders, above baseline population levels. Historically, highly variable rates of depressive disorders were noted in cancer patients, primarily due to lack of diagnostic and methodologic standardization. A recent high-quality meta-analysis found prevalence of 16% for major depression, 19% for minor depression (a mood disorder that does not meet full criteria for major depression, but features at least two depressive symptoms for 2 weeks), and 19% for adjustment disorder.¹ This prevalence of major depression appears to be three to four times higher than that in the general population, whereas rates of dysthymia are relatively similar (Fig. 6-1).² Review of specific studies suggests an association between year of publication and prevalence, with higher rates of depression reported in earlier studies. It is possible that prevalence of depression in cancer has in fact decreased in recent years; if true, this may be a result of improved oncologic outcomes, decreased cancer-associated stigma, availability of palliative care, and increased attention to screening.³ Alternatively, it may be that more recent studies employ more rigorous methods and therefore report more accurate rates; this hypothesis is supported by the finding that studies with more rigorous methods tend to report lower prevalence rates.¹

Young age and low levels of social support are consistent risk factors for depression in cancer patients (Box 6-1).⁴ Personal and family history of depression likely play roles as well.⁵ In contrast to findings in the general population, women with cancer are not at higher risk for depression than men with cancer.^6,7 Ethnicity (particularly Latino) and lower income may be risk factors for depression in cancer.⁸ Anemia⁹ and hypogonadism may also be risk factors for depression in cancer patients, although confirmatory studies are needed. Although there is no consistent association between cancer type and incidence of depression, one large study found that patients with Hodgkin’s lymphoma and lung, pancreatic, brain, and head and neck cancers had more distress than patients with other types of cancer.¹⁰ In particular, early reports suggest depression more often precedes the diagnosis of pancreatic cancer than other types of cancer. However, the specificity of this association is unclear, and may be confounded by associated pain disorders and the poor clinical and functional status of many patients with pancreatic cancer at the time of diagnosis. Alternatively, there may be a shared pathophysiological mechanism that has not yet been determined.

Thoughts of suicide may be present in up to 17% of cancer patients.¹¹ In instances of suicidal ideation, hopelessness is at least as strong a contributor as depression.¹² More common than active suicidal ideation is some form of desire for hastened death, found in between 10% and 30% of patients who are terminally ill with cancer. Of note, the desire for hastened death is a unifying construct which encompasses passive wishes to die sooner than would occur naturally, active plans to commit suicide, and requests for euthanasia or assisted suicide.¹³ Similarly, depression and hopelessness are strong yet independent contributors to the desire for hastened death.¹³ An unknown but substantial frequency of cancer patients demonstrate contingent suicidality, linking their suicidal intent to a future time point or particular life circumstance (e.g., severe pain or dyspnea as natural death approaches).¹⁴ There is a case report suggesting that some family members of cancer patients also struggle with severe anticipatory grief and are at risk for contingent suicide upon the death of their loved one.¹⁵

A case–control study of the suicide risk associated with medical illness among older Americans found that cancer was the only nonpsychiatric medical condition independently associated with completed suicide (OR 2.3), and that metastatic disease further increased the suicide risk.¹⁶ Adjusted rates of suicide are 31.4/100,000 person-years among those with cancer, compared with 16.7 in the general population.¹⁷ Nonetheless, completed suicide is rare, particularly in the setting of adequate supportive treatment. In a study of over 17,000 patients cared for at home by palliative care teams, only 0.03% of deaths occurred due to suicide.¹⁸ Another population-based study found that only 0.2% of all deaths in cancer patients resulted from suicide, with the highest risk in the first 6 months after diagnosis.¹⁹ More recent work has confirmed that the relative risk of suicide as compared to cancer-free controls is very high in the acute phase following a cancer diagnosis (RR 12.6 in the first week after diagnosis, RR 3.1 in the first year after diagnosis).²⁰ This finding suggests the need for early psychiatric assessment and intervention for at-risk patients, as well as communication training for primary care providers, emergency room physicians, oncologists and other providers who may be most likely to see patients during this time period. Risk factors for suicide among patients with cancer include male sex, white race, and unmarried status; lung cancer carries the highest rates, followed by stomach and oropharyngeal/laryngeal cancers.¹⁷ Notably, these types of cancer are often associated with underlying alcohol or substance abuse, which are also risk factors for suicide. The risk is highest in the first 5 years after the diagnosis of cancer, but remains elevated for at least 15 years.¹⁷ Mood disorders are present in 80% of patients with cancer who commit suicide.²¹

Cancer activates many inflammatory pathways that may potentiate depression (Fig. 6-2), and the tumor burden may extend to the CNS (e.g., primary brain tumors, brain metastases, and leptomeningeal disease) where it may directly disrupt mood (Box 6-2). Surgery, chemotherapy, and radiation, which are often administered sequentially and repetitively, particularly in patients with advanced disease, may result in tissue damage and destruction and thus activate an innate immune response. Several chemotherapeutic agents (cyclophosphamide, 5-fluorouracil) are known to cross the blood–brain barrier. In animal models, these as well as other chemotherapeutic agents that do not cross the blood-brain barrier decrease cell proliferation in the hippocampal dentate gyrus, an area of the brain implicated in the pathogenesis of mood disorders.²² In addition, chemotherapy and radiation may directly induce nuclear factor κ B (NFκB) and its downstream inflammatory gene products.²³ Circulating IL-6 levels are increased in patients with cancer and major depression as compared to healthy comparison subjects and cancer patients without depression.^24,25 There are other neuroendocrine alterations present that might increase the inflammatory response, including dexamethasone nonsuppression^24,26 and flattening of the diurnal cortisol curve.²⁵ In patients receiving injectable IFN-alpha for treatment of melanoma, exaggerated secretion of ACTH and cortisol in response to the first injection is associated with subsequent development of depression.²⁷

These physiologic factors coexist with profound psychological stress from the diagnosis of cancer, which is often followed by ongoing suffering, worry about recurrence and mortality, regardless of cancer stage. Given the known contribution of stress to inflammatory responses, it is probable that both physical and psychosocial challenges contribute to the biological mechanisms underlying depression in cancer patients. Still, there is a relative lack of data on biomarkers in cancer patients with major depression, as compared to a more ample literature on the biological parameters associated with cancer fatigue.²⁸

There is little evidence of unique phenomenological distinctions among major depression, minor depression, and adjustment disorder with depressed mood in patients with cancer. However, approximately 30% of advanced cancer patients may ultimately cross over between major depressive disorder (MDD) and adjustment disorders (generally from adjustment disorder to MDD), reinforcing the importance of early intervention and longitudinal follow up.²⁹ Notably, several diagnostic criteria for MDD are confounded by symptoms resulting from cancer or its treatment, specifically low energy, poor appetite, altered psychomotor activity, and impaired concentration. In assessing major depression in patients with cancer, criteria of particular assistance in the differential diagnosis include anhedonia, early morning awakening, disability out of proportion to condition, guilt, suicidal ideation, and somatic symptoms out of proportion to known organic disease processes. Further, fatigue that is already present early in the morning, as opposed to fatigue that sets in gradually over the course of the day, may suggest depression as opposed to the expected effects of cancer and cancer treatment. Although controversy continues over how to weigh somatic versus psychological symptoms in patients with cancer and comorbid depression, most clinicians favor an inclusive approach (in which somatic symptoms are counted toward a diagnosis of MDD), especially given that proinflammatory cytokines may contribute both to depression and somatic symptoms, such as pain and fatigue. Other non-DSM criteria may support a diagnosis of MDD (Box 6-3).

Suicidal ideation and/or requests for physician-assisted suicide often support a diagnosis of MDD, and require careful psychiatric evaluation. After assessment of any suicide risk factors other than medical illness, followed by decisive action to assure safety, patients may benefit from gentle exploration of the meaning behind their request. Contributing psychodynamic factors may include a need for control, ambivalence (i.e., communicating desires to the physician reflects an underlying wish for rescue), or a split in the experience of the self, in which the medically ill self is experienced as an intrusive entity that needs to be eradicated for the good self to survive. Other possibilities include suicide as a means of revenge against loved ones, atonement for guilt (especially if the cancer was initially attributed to bad deeds, for instance smoking of tobacco, and unacceptable emotions), and actualizing a “felt experience” of already being dead.³⁰

Patients with cancer experience many losses, including vitality, bodily integrity, independence, mental clarity, fertility, virility, and others in the course of their illness that may predispose them to depression. Specific stressors vary according to cancer type (Table 6-1).

Depression is associated with various psychiatric comorbidities in cancer patients, including anxiety disorders which may include symptoms of posttraumatic stress, though full syndromal PTSD is uncommon (Box 6-4).³¹

Depressive symptoms appear to be most prominent at specific points in the course of the illness: at diagnosis, completion of active treatment when patients are confronted with existential angst and fears of recurrence in the setting of less active follow-up with medical providers, and at diagnosis of recurrence or metastasis. In patients with breast cancer, there is a nonlinear decline in the point prevalence of depression and anxiety in the years following diagnosis (50% in the first year; 25% in years 2–4; 15% in year 5), with an increase to 45% within 3 months of diagnosis of a recurrence.³² In general, however, longitudinal studies using DSM-based criteria for caseness (as opposed to scores on depression scales) are lacking. Progressive disease may not cause substantially greater rates of depression than less symptomatic disease, a finding particularly robust in breast cancer.³³ However, metastasis is associated with elevated rates of depression, although the direction of causality remains to be determined.^34,35 Depression in patients with metastasis increases with increased physical and psychosocial symptom burden, though not with closeness to death per se.³⁶ Long-term cancer survivors (>5 years from diagnosis) who are in remission do not appear to have higher rates of major depression than the general population.³⁷

Cancer patients with depression often exhibit maladaptive thoughts and behaviors, such as irritability, amplification of somatic symptoms, blaming themselves for their illness, withdrawing from family and professional caregivers, and limiting their travel and activities outside the home. These issues often result in reduced satisfaction with medical care, less effective communication with caregivers and physicians, reduced adherence to medical treatment plans, and longer hospitalizations.^38,39 They also demonstrate higher rates of complementary and alternative medication utilization.⁴⁰

Importantly, even after adjusting for clinical factors affecting prognosis, depression has been shown to adversely affect quality of life and overall survival (i.e., all-cause mortality) in advanced cancer patients.^41–43 It does not, however, affect cancer progression.⁴³

The primary differential diagnosis for depression in patients with cancer includes grief and adjustment disorders, substance-induced mood disorders, and hypoactive delirium (Box 6-5). Assessing the quality of tearfulness may be clinically useful in distinguishing adjustment disorders and grief from that of major depression: cathartic crying is more characteristic of adjustment disorders and grief, while draining or unrelieving crying may accompany a major depression. Grief reactions may include intense sadness, appetite loss, and sleep disturbance, but usually do not include the pervasive guilt and loss of self-esteem that accompanies major depression. In addition, grief-related emotions tend to come in waves interspersed with periods of time when the patient is able to enjoy activities; by contrast, depression is usually characterized by persistent anhedonia. Adjustment disorders also tend to be less severe and less functionally impairing than major depression. Obtaining collateral history from family members can help clarify the extent of impairment in day-to-day activities. It is also important to obtain a careful medication history, including any recent changes that would suggest a substance-induced depressive disorder.

Similarly, patients may appear withdrawn or dysphoric if they are using alcohol or illicit substances or misusing prescribed psychoactive medications, such as benzodiazepines or opioids for “chemical coping.” Although chronic daily marijuana use may contribute to depressive symptoms, patients utilizing cannabis for medicinal purposes may report improvements in mood and sleep as well as pain relief.⁴⁴ Other patients with alcohol or substance abuse histories opt to turn over a new leaf after a cancer diagnosis, which occasionally results in full withdrawal.

Patients with occult hypoactive delirium may demonstrate flat affect, psychomotor slowing, decreased engagement, and suicidal ideation; this necessitates careful cognitive examination when evaluating the patient with cancer for depression. Differential diagnosis of suicidal ideation in this population must always include encephalopathic conditions; the presence of hallucinations or delusions is a risk factor for suicide attempts.⁴⁵ Finally, a complete differential diagnosis may include several other constructs (demoralization and conservation-withdrawal) that have been proposed to describe psychological symptoms of low mood, interest, or rumination. Demoralization syndrome includes existential distress, hopelessness, helplessness, isolation, and potential suicidality, but is differentiated from major depression by the lack of generalized anhedonia.⁴⁶ Conservation–withdrawal is a construct proposed to explain behavioral disengagement and inactivity as a means of conserving energy while fighting a serious illness; patients may display diminished speech and movement and withdrawal from usual activities, but usually lack depressed mood per se.⁴⁷

There is an enhanced emphasis on depression screening in cancer settings as a result of the National Cancer Institute’s recent mandate that all comprehensive cancer centers should screen oncology patients for emotional distress. Clinicians have historically underestimated the severity of depressive illness in the cancer population. One large study found a concordance of only 13% between oncologist and patient ratings in patients with severe depressive symptoms; nursing staff ratings had no greater concordance^48,49 with patient scales. The six-item Brief Edinburgh Depression Scale, initially developed for post-partum settings, may be useful as a screening tool for oncology practices; other options include the Patient Health Questionnaire-Depression Module (PHQ-9) and the Beck Depression Inventory (BDI). Although the Hospital Anxiety and Depression Scale (HADS) is the most extensively validated screening tool for emotional distress in cancer patients, thresholds for clinically significant distress are widely variable across studies.⁵⁰ Although there is no clear evidence that depression screening alone improves outcomes compared to usual care,⁵¹ screening that is linked to collaborative care interventions in cancer patients shows promise in early trials.⁵²

Opioids, corticosteroids, interferon, adjuvant hormonal treatments, and several chemotherapeutic agents, including tyrosine kinase inhibitors,⁵³ are medications used in treating cancer that are also implicated in substance-induced depression (Box 6-6). Recognition of depression as a potential side effect can help clinicians tailor treatment when appropriate, for example, by minimizing steroid use, or to identify and treat any resulting depression in a timely fashion if it occurs.

INTERFERON

After high-risk early-stage or metastatic melanoma a year-long course of interferon alpha may be given. Among its many constitutional and neuropsychiatric side-effects, this recombinant cytokine can gradually trigger a depressive syndrome that, in instances, necessitates dose reduction or even discontinuation. Thus, mental health management for patients receiving interferon is especially important.

A small body of research supports use of antidepressants in patients receiving interferon alpha, both in resolving interferon-related depressive symptoms and allowing patients to successfully complete the interferon course.⁵⁴ Pretreatment with a serotoninergic antidepressant (in one small RCT, paroxetine) may delay or prevent onset of interferon-related depression in melanoma patients.⁵⁵ However, pretreatment is best reserved for patients who, at time of interferon initiation, are either already on an antidepressant regimen or report prominent depressive symptoms.⁵⁶ Interferon occasionally precipitates mania, hypomania, or a mixed state,^56,57 and this ill-effect may be fostered by antidepressants.