Psychosomatics 49:218-224, May-June 2008
doi: 10.1176/appi.psy.49.3.218
© 2008 Academy of Psychosomatic Medicine
Depression After Diagnosis of Advanced Non-Small Cell Lung Cancer and Survival: A Pilot Study
William F. Pirl, M.D.,
Jennifer S. Temel, M.D.,
Andrew Billings, M.D.,
Constance Dahlin, APRN, B.C.,
Vicki Jackson,, M.D., M.P.H.,
Holly G. Prigerson, Ph.D.,
Joseph Greer, Ph.D., and
Thomas J. Lynch, M.D.
Received January 4, 2007; revised April 3, 2007; accepted April 18, 2007. From Massachusetts General Hospital Cancer Center, Boston, MA; Dana-Farber Cancer Institute, Boston, MA. Send correspondence and reprint requests to William F. Pirl, M.D., Massachusetts General Hospital, WACC 812, 15 Parkman St., Boston, MA 02114. e-mail: wpirl{at}partners.org
© 2008 The Academy of Psychosomatic Medicine
|
ABSTRACT
|
BACKGROUND: Major depressive disorder is estimated to occur in 10%–25% of people with cancer, and it has been inconsistently linked to increased mortality. OBJECTIVE: This pilot study investigates the association of depression and survival in advanced non-small cell lung cancer (NSCLC) patients. METHOD: Forty-three recently-diagnosed advanced NSCLC patients completed the Hospital Anxiety and Depression Scale and were followed prospectively. RESULTS: Patients with depression had poorer survival. Median survival was four times shorter than those without depression. Controlling for baseline performance status, depression predicted 6-month mortality, but was not significant for overall survival. CONCLUSION: Although depression after advanced-NSCLC diagnosis was associated with poorer survival at 6 months, this association was not present for overall survival; however, further research with larger samples should be pursued.
Key Words: Depression • Lung Cancer • Survival
|
INTRODUCTION
|
Major depressive disorder is estimated to occur in 10%–25% of people with cancer, over twice the rate of the general population.1,2 In cancer patients, major depression appears to be associated with poorer adherence to treatment, increased length of hospital stays, and greater utilization of healthcare services.3–5 Depression has also been associated with increased mortality in cancer patients. However, the literature is mixed, with both positive and negative studies, leaving the association between and mood and cancer mortality unclear.6 If such a relationship does exist, the presence of depression could provide prognostic information that could be used in decision-making as well as a potential target for increasing cancer survival.
Patients with advanced non-small cell lung cancer (NSCLC) represent a specific cancer population, with both elevated rates of mortality and depression. Median survival for advanced NSCLC is 10 months, and reported rates of depression in this population range from 16% to 29%.7–9 Four studies have been published investigating the association of depression and mortality specifically in lung cancer patients.10–13 In a study of 133 patients with heterogeneous bronchogenic cancers, Buccheri and colleagues10 found that patients with depression, assessed by the Zung Depression Rating Scale, had lower rates of survival than the nondepressed patients. However, the other three studies found no association between depression and mortality. With the exception of a study of early-stage NSCLC survivors, the negative studies are unfortunately complicated by inadequate assessments of depression and heterogeneous samples in terms of cancer cell type and stage.
Because prognoses differ according to type and stage of lung cancer, regardless of depression, the association between depression and survival in people with advanced NSCLC is still unknown. In order to investigate the association of depression and survival in this population, we prospectively followed newly-diagnosed patients with Stage III-B (with effusions) or IV NSCLC with good performance status. These patients received medical care in the ambulatory setting of a tertiary-care center and were participants in a feasibility trial of introducing early palliative care in addition to standard cancer treatment.
|
METHOD
|
Patient Selection
For this study of depression and survival among NSCLC patients, we studied a subset of the patients who were accrued to a larger, parent study, "Impact of Early Intervention with Palliative Care on Quality of Life in Patients with Advanced NSCLC (EIPC)." Broadly described, the EIPC study sought to assess the feasibility and expected adherence of a clinical approach to patients with incurable NSCLC that explicitly integrates subspecialty palliative and medical oncologic care.14 Specifically, the EIPC participants who completed the baseline depression measure, the Hospital Anxiety and Depression Scale (HADS), served as the analytic sample for this study. Participants were recruited to the EIPC from a convenience sample of patients receiving care in the ambulatory thoracic oncology clinic at the Massachusetts General Hospital Cancer Center. Patients with histopathologically- or cytologically-confirmed Stage III-B with pleural or pericardial effusions or Stage IV disease (advanced NSCLC) were assessed for eligibility. Patients were required to be within 8 weeks of diagnosis of advanced NSCLC to participate. Other eligibility requirements included age over 18 years, and being capable of reading and responding to questions in English. Also, participants needed to have an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, which correspond respectively to being fully active without any restrictions from disease or being able, ambulatory, and able to perform light activities, but restricted in terms of strenuous activities. The protocol was approved by the Massachusetts General Hospital (MGH) Institutional Review Board, and all patients provided written informed consent before participation.
The EIPC study was a pilot feasibility trial for introducing advanced NSCLC patients to the palliative care service within 8 weeks of diagnosis, in contrast to the more common practice of referring to palliative care closer to the end of life. Participants were required to meet with the palliative care service at least once every 2 months and to complete quality-of-life assessments. All participants continued to receive standard oncological care by the thoracic oncology team. Demographics (age, sex, race, marital status) and disease factors (stage and initial therapy received for advanced disease) were collected prospectively from patient surveys and the medical record. Also, a retrospective medical review was performed after completion of the study to determine other clinical variables, including 1) time from patient’s report of first noticing symptoms to initial oncology evaluation; 2) presence of brain metastases at the time of diagnosis; 3) chemotherapy dose delays and reductions; 4) number of hospitalizations; 5) number of infections; and 6) hospice referral within the first 6 months of diagnosis.
The HADS was used to assess depressive symptoms. This is a well-validated self-report instrument that has been used to measure the severity of depressive and anxiety symptoms in cancer patients.15 Designed for medically ill populations, it is a 14-item self-report questionnaire that minimizes the inclusion of physical symptoms in assessing mood and anxiety. It contains two 7-item scales to measure depression and anxiety over the previous week. Although the HADS is not a diagnostic instrument, a range of cutoff scores on the subscales have been established to identify probable cases of anxiety and depressive disorders. In an effort to capture all possible cases of anxiety and depressive disorders in the sample, the lowest cutoff scores based on the psychometrics of the instrument,  8 on each subscale, were used to define anxiety and depression.15
Statistical Analysis
The primary objective of this prospective observational study within an intervention study was to evaluate the association between the presence of depressive symptoms after diagnosis of advanced NSCLC and survival. The secondary objective was to determine the rates of possible cases of depressive and anxiety in this group of patients. Associations between depression and covariates that may influence mortality were evaluated with a chi-square test for categorical variables and a two-sample t-test for continuous variables. Associations between depression and survival were evaluated in three ways. We first used the Wilcoxon test of equality of survival functions, given the relatively long follow-up period, with the likely difference in survival being observed earlier. Variables that were significant in the bivariate analyses at the 0.1 significance level were then entered into a logistic-regression model for 6-month mortality and a Cox regression model for overall mortality. All analyses were performed in STATA 8.0 SE (College Station, Tex.).
|
RESULTS
|
From October 2003 through June 2005, 382 newly-diagnosed cases of Stage III-B (with and without effusions) or Stage IV NSCLC were seen in the ambulatory thoracic oncology clinic. A convenience sample of 46 patients who were ECOG PS 0 or 1 was accrued to the EIPC study. One participant was determined to be ineligible for the study because of his lower performance status at the time of enrollment. Two participants in the EIPC study did not have baseline HADS scores. Forty-three patients (95.6%) who completed the baseline HADS as part of the trial are the analytic sample for these analyses. The median time from diagnosis to baseline assessment was 15 days, with an interquartile range of 10 to 22 days. No participants were lost to follow-up, and, at the time of the analyses, the maximum follow-up of survivors was 30 months. The baseline characteristics are shown in Table 1. Median age of the participants was 65.6 years (standard deviation [SD]: 8.32), and 60.5% (26/43) were women. The majority of patients had Stage IV disease.
Of note, all patients studied received a first-line oncologic therapy; 65% of patients (28/43) received chemotherapy; 23.3% (10/43) received radiation therapy; and 9.3% (4/43) received an oral epidermal growth factor tyrosine kinase inhibitor (EGFR-TKI). One patient received combined treatment with both chemotherapy and radiation; 37 patients (86.0%) were diagnosed at advanced stage, and 6 (14.0%) had known NSCLC that progressed to advanced stage. There was no difference in the mean number of visits with the palliative care team between the two groups; patients meeting study criteria for depression had a mean of 5.78 visits (SD: 4.91), compared with a mean of 5.68 visits (SD: 2.46) for those without depression.
Depression and Anxiety
Using a cutoff score of 8 for each subscale of the HADS, 23.3% of patients (10/43) met study criteria for depression, and 32.6% (14/43), for anxiety. Eleven patients in total (24.4%) were taking antidepressants at baseline, including two of the patients meeting study criteria for depression. Four patients (8.9%) had a past history of depression, but none of the patients who met study criteria for depression had a past history of depression.
Bivariate analyses of patient characteristics by baseline HADS Depression subscale scores are summarized in Table 2. Of note, there were no differences in patient characteristics between patients who had baseline HADS scores meeting criteria for depression and those with baseline HADS scores not suggestive of depression. Additional factors such as marital status, smoking status, and initial cancer treatment did not differ between the two groups. Only one of the patients who met study criteria for depression had progressed to advanced NSCLC, and there was not a significant difference in progression between patients who had baseline HADS scores meeting criteria for depression and those with baseline HADS scores not suggestive of depression. Exploratory analyses of the data from the retrospective chart review revealed no differences between the two groups with respect to time from symptoms until diagnosis of lung cancer or the presence of brain metastases at diagnosis, chemotherapy dose reductions/delays, hospitalizations, documented infections, or early referral to hospice.
Survival
The median survival for the sample was 9.9 months, with an interquartile range of 6.2 months to 18.4 months. Over two-thirds of the patients (30/43) were alive at 6 months, and over one-third of the patients (17/43) were alive at 12 months.
Unadjusted predictors of 6-month mortality are shown in Table 3. Patients with depression had greater mortality at 6 months (p=0.02). That is, 60.0% of patients with depression (6/10) had died by 6 months, as compared with 21.2% (7/33) of those without depression. There was no association between anxiety and survival. There was a trend for increased mortality with higher baseline performance status. No other patient characteristics or disease factors in Table 3 were significantly associated with lower survival. Additional factors, such as marital status, smoking status, and initial cancer treatment did not differ between those who survived or died within 6 months.
The Kaplan-Meier survival curve for patients with and without depression is shown in Figure 1. Patients with depression had poorer survival (Wilcoxon p=0.03). The median survival for patients with depression was 2.5 months, as compared with 10.4 months for patients without depression.
View larger version (46K):
[in this window]
[in a new window]
|
FIGURE 1. Survival Curves by HADS Scores
HADS: Hospital Anxiety and Depression Scale. Median survival: HADS <8, 2.5 months; HADS 8, 10.4 months.
|
Depression and performance status were entered into both logistic and Cox regression models. The adjusted associations with 6-month mortality are shown in Table 4. Depression remained a significant predictor of 6-month mortality after controlling for performance status (p=0.04), with those who were depressed at baseline experiencing over five times the odds of death within 6 months, as compared with those patients who were not determined to be depressed at baseline (odds ratio [OR]: 5.30; 95% confidence interval [CI]: 1.04–26.88). This two-covariate model was very predictive of mortality, with a Hosmer and Lemeshow goodness-of-fit test value of 0.87. Additional factors instead of performance status, such as the time from symptoms to initial oncology evaluation, presence of brain metastases, chemotherapy dose delays or reductions, and hospice referrals, were substituted in this model, and the effect of depression at baseline remained significant. The Cox regression that adjusted for performance status is shown in Table 5. Depression was not a significant predictor of overall mortality after controlling for performance status (p=0.10). Those with depression at baseline had almost twice the risk of death, as compared with those without depression (hazard ratio [HR]: 1.89; 95% CI: 0.88–4.06).
|
DISCUSSION
|
There is a growing literature on the impact of depression on mortality. In coronary artery disease, depression appears to be an independent risk factor for mortality in the 5 years after a myocardial infarction.16–18 This negative association between depression and survival seems to also be present in other medical illnesses, and possibly even in the general population.19–21 However, previous studies of the association between depression and cancer survival have yielded mixed results.6,10–13
The data from this study support the hypothesis that depression in recently-diagnosed advanced NSCLC patients is associated with poorer survival at 6 months, but not associated with overall survival. Although the sample size in this pilot study may not allow for a significant finding for overall survival, further research with larger samples should be pursued because of the increased mortality risk suggested in these data.
Although the majority of previous studies in lung cancer patients did not find an association, these studies have methodological challenges related to heterogeneous samples in terms of cancer-cell type and stage and differing time-points for the assessment of depression. The results presented here are drawn from the most homogeneous cohort of advanced NSCLC patients, in which the association between depression and survival has been studied. In the same way that response to cancer treatments differs by cancer cell type and stage, associations between depression and mortality may also differ by these variables. Even after controlling for type of cancer and stage, a true positive finding might be lost in a heterogeneous sample that may not have sufficient power to detect an association. Even though depression did not appear to be associated with survival in the study by Nakaya and colleagues,13 a homogeneous cohort of patients who received curative treatment for early-stage NSCLC, this finding may not generalize to those patients with advanced NSCLC. One major difference between these populations is that patients with advanced NSCLC will die from their cancer and/or its complications, whereas lung cancer survivors are more likely to die of other causes.
Also, previous studies of survival in lung cancer patients used different time-points to assess depressive symptoms. Because depressive symptoms may be part of a normal, time-limited reaction to being diagnosed with cancer, the diagnosis of a major depressive episode is usually reserved until this reaction period is over, typically 3 to 4 weeks after diagnosis.22 Studies that assess depressed mood in patients within 3 days of diagnosis may be investigating the effect of a person’s reaction to their diagnosis, rather than an actual depressive episode. Although there was a range in the amount of time since diagnosis in this sample, depressive symptoms were assessed further out from diagnosis, with a median of 15 days, and 97.7% (42/43) had more than 3 days since diagnosis.
Although an association between depression within 8 weeks of diagnosis and survival was found, this pilot study is not able to identify any underlying mechanisms for the relationship. Our retrospective chart review did not find any significant differences between the people with and without probable cases of depression in this small sample in terms of adherence to chemotherapy regimens, hospitalizations, infections, brain metastases, and early referral to hospice. However, several possible mechanisms can be postulated. Depression may result in behaviors that decrease survival. Patients with depression are less likely to be adherent to cancer treatment, which may limit the benefits of their therapy.3 Depression is also associated with greater desire for death in advanced cancer patients, and depressed patients may be more likely to stop treatment or choose hospice sooner.23
Depression may also be associated with biological changes. Previous research has found that cancer patients with depression have significantly lower lymphocyte counts, which could be a possible mechanism for tumor progression.24,25 Enhanced platelet activation and aggregation have been reported in post-myocardial infarction patients with depression.26 In cancer patients, this might increase the risk for pulmonary emboli, a common and often fatal complication of cancer. However, there were no significant differences in rates of pulmonary emboli between our two groups. Finally, depression may not be associated with a cause of mortality, but rather be a byproduct of another process, such as the patient’s underlying malignancy. Current theories on the origin of depression in patients with cancer postulate that depression may be the result of cytokines that are caused by aggressive malignant disease.27,28 The fact that the biggest effect on mortality was seen early may suggest that the depressive symptoms could have been a psychological manifestation of cytokines from more aggressive cancers.
Although this pilot study is provocative, it does not allow for inferences about causality, and it also has limitations. Although an association was found between elevated depressive symptoms within 8 weeks of diagnosis and mortality at 6 months, this study did not investigate the effects of the formal diagnosis of major depressive disorder. Patients were assessed for probable cases of depression by use of a self-report measure of depressive symptoms, and not by formal diagnostic criteria. Using the HADS may have increased the number of false positives for depression, but even the diagnosis of major depression is problematic because of the confounding of physical symptoms shared with cancer. Although the small sample size in this pilot study limited our ability to control for multiple variables in the regression models, the restrictive eligibility criteria provided some controlling of disease variables by assembling a homogeneous cohort. Another limitation of this group is that it was a convenience sample and was only a small portion of the potentially eligible patients. Although the median survival for the sample is in the expected range, and the rate of depression is similar to the rate in the thoracic oncology clinic, there may be something about participants in this sample who chose to participate in the EIPC study that could bias the results or limit generalizability to all patients with advanced NSCLC. Also, the homogeneous racial composition of the sample limits generalizability, especially in light of observed racial differences in NSCLC mortality.29
To address these limitations, the association between depression and mortality in people with advanced NSCLC needs to be investigated by use of standardized diagnostic criteria for major depressive disorder in larger samples that are adequately powered to evaluate potential mediators. Longitudinal assessments would allow the use of time-varying covariates to further examine causal mediators. Also, this study raises the question of whether depression at the time of diagnosis can be used as a clinical prognostic indicator among similar good performance-status patients. If this relationship is confirmed, a logical question will be whether the treatment of their depression has an effect on survival. The answer to this question will ultimately depend on the underlying mechanism. However, if an intervention can eliminate the discrepancy in survival and increase lifespan by another 7 months, it may present more benefit to advanced-NSCLC patients with depression than any current cancer treatments in this population.
|
ACKNOWLEDGMENTS
|
Drs. Pirl and Temel contributed equally to this article and should both be considered first-authors.
This paper was presented at the Academy of Psychosomatic Medicine Annual Meeting in Tuscon, AZ on 11/18/06.
This study was partially funded by NIH/NCI K23 CA115908 (Pirl) and an investigator-initiated research grant from Amgen (Temel).
|
REFERENCES
|
- Kessler RC, Berglund P, Demler O, et al: The epidemiology of major depressive disorder: results from The National Comorbidity Survey replication (NCS-R). JAMA 2003; 289:3095–3105[Abstract/Free Full Text]
- Pirl WF: Evidence report on the occurrence, assessment, and treatment of depression in cancer. JNCI Monogr 2004; 32:32–39
- Richardson J, Sheldon D, Krailo M, et al: The effect of compliance with treatment on survival among patients with hematologic malignancies. J Clin Oncol 1990; 8:356–364[Abstract]
- Herrmann C, Brand-Driehorst S, Kaminsky B, et al: Diagnostic groups and depressed mood as predictors of 22-month mortality in medical inpatients. Psychosom Med 1998; 60:570–577[Abstract/Free Full Text]
- Ingram D, Browne G, Reyno L, et al: Prevalence, correlates, and cost of anxiety and affective disorder in men with prostate cancer one year after initial diagnosis (abstract). Psycho-oncology 2003; 12(suppl):S229
- Spiegel D, Kato PM: Psychosocial influences on cancer incidence and progression. Harv Rev Psychiatry 1996; 4:10–26[Medline]
- Hughes JE: Depressive illness and lung cancer, II: follow-up of inoperable patients Eur J Surg Oncol 1985; 11:21–24
- Hopwood P, Stephens RJ: Depression in patients with lung cancer: prevalence and risk factors derived from quality-of-life data. J Clin Oncol 2000; 18:893–903[Abstract/Free Full Text]
- Walker MS, Zona DM, Fisher EB: Depressive symptoms after lung cancer surgery: their relation to coping style and social support. Psychooncology 2006; 15:684–693[CrossRef][Medline]
- Buccheri G: Depressive reactions to lung cancer are common and often followed by poor outcome. Eur Respir J 1998; 11:173–178[Abstract/Free Full Text]
- Faller H, Bulzebruck H, Drings P, at al: Coping, distress, and survival among patients with lung cancer. Arch Gen Psychiatry 1999; 56:756–762[Abstract/Free Full Text]
- Faller H, Schmidt M: Prognostic value of depressive coping and depression in survival of lung cancer patients. Psychooncology 2004; 13:359–363[CrossRef][Medline]
- Nakaya N, Saito-Nakaya K, Akizuki N, et al: Depression and survival in patients with non-small cell lung cancer after curative resection: a preliminary study. Cancer Sci 2006; 97:199–205[CrossRef][Medline]
- Temel JS, Jackson VA, Billings JA, et al: Phase II study: integrated, palliative care in newly-diagnosed advanced non-small-cell lung cancer. J Clin Oncol 2007; 25:2377–2382[Abstract/Free Full Text]
- Zigmund AS, Snaith RP: The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 1983; 67:361–370[Medline]
- Grace SL, Abbey SE, Kapral MK, et al: Effect of depression on five-year mortality after an acute coronary syndrome. Am J Cardiol 2005; 96:1179–1185[CrossRef][Medline]
- Wulsin LR, Evans JC, Vasan RS, et al: Depressive symptoms, coronary heart disease, and overall mortality in The Framingham Heart Study. Psychosom Med 2005; 67:697–702[Abstract/Free Full Text]
- Rumsfeld JS, Jones PG, Whooley MA, et al: Depression predicts mortality and hospitalization in patients with myocardial infarction complicated by heart failure. Am Heart J 2005; 150:961–967[CrossRef][Medline]
- Kouzis A, Eaton WW, Leaf PJ: Psychopathology and mortality in the general population. Soc Psychiatry Psychiatr Epidemiol 1995; 30:165–170[CrossRef][Medline]
- Ganzini L, Smith DM, Fenn DS, et al: Depression and mortality in medically ill older adults. J Am Geriatr Soc 1997; 45:307–312[Medline]
- Katon WJ, Rutter C, Simon G, et al: The association of comorbid depression with mortality in patients with type 2 diabetes. Diabetes Care 2005; 28:2668–2672[Abstract/Free Full Text]
- Pirl WF, Roth AJ: The diagnosis and treatment of depression in cancer patients. Oncology 1999; 13:1293–1301[Medline]
- Breitbart W, Rosenfeld B, Pessin H, et al: Depression, hopelessness, and desire for hastened death in terminally ill patients with cancer. JAMA 2000; 284:2907–2911[Abstract/Free Full Text]
- Reiche EM, Nunes SO, Morimoto HK: Stress, depression, the immune system, and cancer. Lancet Oncol 2004; 5:617–625[CrossRef][Medline]
- Reiche EM, Morimoto HK, Nunes SM: Stress and depression-induced immune dysfunction: implications for the development and progression of cancer. Int Rev Psychiatry 2005; 17:515–527[CrossRef][Medline]
- Serebruary VL, Glassman AH, Malinin AI, et al: Enhanced platelet/endothelial activation in depressed patients with acute coronary syndromes: evidence from recent clinical trials. Blood Coagul Fibrinolysis 2003; 14:563–567[CrossRef][Medline]
- Cleeland CS, Bennett GJ, Dantzer R, et al: Are the symptoms of cancer and cancer treatment due to a shared biologic mechanism? a cytokine-immunologic model of cancer symptoms. Cancer 2003; 97:2919–2925[CrossRef][Medline]
- Raison CL, Miller AH: Depression in cancer: new developments regarding diagnosis and treatment. Biol Psychiatry 2003; 54:283–294[CrossRef][Medline]
- Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2006. CA Cancer J Clin 2006; 56:106–130[Abstract/Free Full Text]
Get information about faster international access.
Privacy Policy
Copyright © 2008
Academy of Psychosomatic Medicine.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
