- Top of page
Objective: To compare narrative therapy (NT) plus escitalopram versus escitalopram plus usual care on quality of life and depressive symptomatology of depressed patients with oncologic disease.
Methods: A total of 72 subjects (mean age 54.6 years), predominantly female with non-metastatic breast, lung and colon cancer and depressive disorder (DSM-IV-TR) were randomized to receive treatment with NT plus escitalopram (n=39) or escitalopram (10–20 mg QD) plus usual care (n=33). Main endpoints were improvement in dimensions of quality of life measured by the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire C-30 and reduction of depressive symptoms using the Hospital Anxiety and Depression Scale at weeks 12 and 24.
Results: The combined therapy group showed significantly greater improvement in all the functioning dimensions (p<0.01), pain scale (p=0.02), global health (p=0.02), and global quality of life (p=0.007) at weeks 12 and 24. There were no statistically significant differences in depressive symptomatology between the groups. From week 12 to week 24 study retention was higher in the combined treatment group (p=0.01).
Conclusions: Brief NT in combination with escitalopram was superior to usual care and escitalopram in improving functioning dimensions of quality life. Copyright © 2010 John Wiley & Sons, Ltd.
- Top of page
The search for effective treatments for anxiety and depression remains one of the main goals of Psycho-oncology 1, 2. Depression is the most prevalent psychological disorder among cancer patients 3–5 and has a negative impact on quality of life, treatment adherence and the level of emotional burden on the caregivers 6.
Data suggest that the most effective treatment strategies for major depression include the combination of psychotherapeutic and pharmacologic interventions 7–11 and based on this, combined approaches to the treatment of depression in cancer patients may be the most effective 12, 13. To our knowledge, two clinical trials have examined the efficacy of combined care (antidepressant plus psychotherpeutic intervention) among cancer patients with major depression 14, 15.
In Newell's meta-analysis, it is concluded that the methodological quality of most of the RCT in the study of depression in oncology have been suboptimal due to small sample sizes and short follow-up periods 16. Furthermore, some studies fail to report the clinical characteristics of the sample, the expertise of the therapist and/or the number of sessions required for the intervention 17. Finally, they recommend testing the efficacy of other psychotherapy interventions apart from cognitive–behaviour therapy. Narrative Therapy (NT), an individually tailored, multi-component, integrative intervention, could be an appropriate one in the field of oncology due to its flexibility and its focus in fitting a person's particular and unique needs 18–20. In contrast to cognitive therapy, NT does not claim that there is one universal encompassing metanarrative to give account of reality. Also, in contrast to other making meaning interventions, NT is not limited to existential issues 21. The lack of studies using NT 22 could be related to its recent emergence as an empirical paradigm and its emphasis on a subjective and qualitative understanding of the problems. However, there are not any clinical trials of NT in depression in cancer, though there is a clinical trial for depressive disorders 19 and an RCT focused on pain in cancer patients 23.
In the present study we consider health related quality of life (HQoL), and not only symptomatic scores, an important endpoint. Previous studies recommend including HQoL as a result variable in the design of any psychosocial intervention 24–26, since assessing HQoL could contribute to improve outcome and may indicate the directions needed for more efficient treatments. Studies have shown that psychological factors predict subsequent quality of life and the treatment of mental disorder influences the HQoL 27–29.
In order to improve the quality of the evidence available we have carried out a randomized controlled trial following previous recommendations 16. The main objective of the study was to evaluate the effects of adding a psychotherapeutic intervention, NT, to escitalopram (combined therapy) on measures of quality of life and depressive symptoms compared with usual care plus escitalopram.
- Top of page
From week 12 to week 24 significantly more patients (p = 0.01) were discontinued from the study, due to poor adherence and missing appointments in the usual care group (n = 7, 21.9%) than in the combined treatment group (n = 1, 2.8%). At week 24, six patients (15.4%) randomized to combined therapy were lost to follow up with 10 (30.3%) in the usual care group (p = 0.129). Of 12 possible sessions of NT, the mean number of sessions completed was 8.4 in the combined therapy.
Table 1 presents baseline demographic variables and the mean scores for each scale of QLQ C-30 and HADS-D. Demographic variables did not differ significantly between the study groups. The mean age was 54.6 ± 10.4 years (range 35–75 years). The sample is unbalanced by the distribution of gender and age according to cancer type and the different prevalence of the cancer type. The distribution between genders is one man for every four women (19.4% compared with 80.6%). Mean HADS-D scores in both groups was 12.48 (:95% CI: 11.4–13.5) in usual care group, and 12.46 (95% CI: 11.4–13.4) in combined therapy group.
Table 1. Baseline characteristics of study participants
| ||Usual care N = 33||Combined therapy N = 39||Total N = 72||p|
|Age-yr. Mean (SD)||56 (10.8)||53.2 (9.5)||54.6 (10.2)||0.252|
|Female sex n(%)||24 (72.7)||34 (87.2)||58 (80.6)||0.145|
| Breast||19 (57.6)||29 (74.4)||48 (66.7)||0.263*|
| Colo rectal||8 (24.2)||7 (17.9)||15 (20.8)|| |
| Lung||6 (18.2)||3 (7.7)||9 (12.5)|| |
| No||6 (18.2)||4 (10.3)||10 (13.9)||0.417|
| Yes||27 (81.8)||35 (89.7)||62 (86.1)|| |
|Prior psychiatric history n(%)|
| No||27 (81.8)||29 (74.4)||56 (77.8)||0.573|
| Yes||6 (18.2)||10 (25.6)||16 (22.2)|| |
|History of cancer in the family n(%)|
| No||24 (72.7)||19 (48.7)||43 (59.7)||0.581|
| Yes||9 (27.3)||20 (51.3)||29 (40.3)|| |
|HADS- D||12.5 ± 3||12.5 ± 3.1||12.5 ± 3.03||0.974|
|QLQ C-30 functioning dimensions|
| Physical||31.7 ± 15.2||30.8 ± 14.2||31.2 ± 14.6||0.786|
| Social||8 ± 12||1.7 ± 6.4||4.6 ± 9.8||0.005|
| Emotional||8.8 ± 10.2||4 ± 5.7||6.2 ± 8.4||0.015|
| Cognitive||13.6 ± 21||6.8 ± 9.9||9.9 ± 16.2||0.076|
| Role||14.6 ± 21.5||7.7 ± 15.2||10.9 ± 18.6||0.114|
| Fatigue||85.0 ± 15||86 ± 18.7||85.8 ± 17||0.899|
| Nausea||64.1 ± 35.9||71.8 ± 36.5||68.3 ± 36.1||0.375|
| Pain||71.7 ± 33.7||75.6 ± 30.5||73.8 ± 31.9||0.606|
| Dyspnea||36.4 ± 35.7||35.9 ± 31.9||36.1 ± 33.4||0.953|
| Insomnia||78.8 ± 20.1||79.5 ± 24.9||79.2 ± 22.7||0.897|
| Appetite loss||66.7 ± 25||56.4 ± 30.7||61.1 ± 28.5||0.129|
| Constipation||35.3 ± 31.1||24.8 ± 29.3||29.6 ± 30.4||0.143|
| Diarrhea||34.3 ± 30.7||26.5 ± 33.5||30.0 ± 32.2||0.306|
| Financial difficulties||9 ± 17.2||12.8 ± 23.7||11.1 ± 20.9||0.455|
| Global health||14.1 ± 17.2||13.7 ± 13.7||13.9 ± 15.3||0.899|
| Global quality of life||16.2 ± 15.3||19.7 ± 12.01||18.1 ± 13.6||0.281|
Depression scores measured by HADS-D were reduced over time in both treatment groups, but there were not statistically significant differences, (Figure 2). Depression scores for patients in the combined therapy group remained lower than scores in patients in the usual care group. Figure 3 shows functioning dimensions and pain scale scores of QLQ-C30 at each timepoint.
Figure 2. HADS-Depression subscale at each timepoint. Boxes show the interquartile range (IQR) with continuous lines for mean scores. Whiskers indicate range. Circles and dots beyond the whiskers are outliers or extreme define as any points outside the IQR by anymore than 1.5 times the IQR
Download figure to PowerPoint
Figure 3. Functioning dimensions and pain scale of QLQ-C30 at each timepoint. Panel A: Physical. Panel B: Cognitive. Panel C: Emotional. Panel D: Social. Panel E: Role. Panel F: Pain. Data are presented as box-plot that shows the mean and interquartile range
Download figure to PowerPoint
Table 2 shows adjusted data by generalized linear model. Comparisons of adjusted mean scores between groups over time found that, at 12 weeks, combined treatment group had significantly higher scores on physical (p<0.001), social (p<0.001), emotional (p<0.001), cognitive (p<0.001), and role (p<0.001) in functioning dimensions on the QLQ-C30 questionnaire. The significant differences were maintained in favor of the combined therapy group at 24 weeks. The combined therapy intervention also showed significant pain reduction (p = 0.02) and better global health (p = 0.02), and global quality of life (p = 0.007) at 12 and 24 weeks.
Table 2. Adjusted data by generalized linear model
| ||Usual care||Combined therapy|| |
| ||Mean (SE)||Change from baseline||95% CI||Mean (SE)||Change from baseline||95% CI||p|
| Baseline||12.48 (0.51)|| || ||12.46 (0.47)|| || ||0.463|
| 12 week||7.42 (0.51)||5.06||3.64 to 6.48||6.52 (0.47)||5.95||4.64 to 7.25|| |
| 24 week||6.72 (0.51)||5.7||4.33 to 7.18||5.54 (0.47)||6.9||5.61 to 8.23|| |
|QLQC-30 functioning dimensions|
| Baseline||31.49 (2.81)|| || ||30.96 (2.58)|| || ||<0.001|
| 12 week||41.79 (2.81)||−10.33||−18.12 to −2.47||56.60 (2.58)||−25.64||32.83 to −18.44|| |
| 24 week||39.77 (2.81)||−8.28||−16.1 to 0.4||66.52 (2.58)||−35.55||−35.55 to −28.3|| |
| Baseline||6.18 (3.62)|| || ||3.31 (3.32)|| || ||<0.001|
| 12 week||29.42 (3.62)||−23.23||−33.22 to 13.24||49.04 (3.32)||−45.73||54.92 to −36.54|| |
| 24 week||34.47 (3.62)||−28.28||−38.3 to 18.3||61.86 (3.32)||−58.55||−67.8 to −49.3|| |
| Baseline||7.21 (2.94)|| || ||5.44 (2.70)|| || ||<0.001|
| 12 week||24.38 (2.94)||−17.17||−25.3 to −9.03||46.04 (2.70)||−40.59||48.08 to −33.11|| |
| 24 week||29.68 (2.94)||−22.47||−30.6 to 14.3||56.72 (2.70)||−51.28||−58.8 to −43.8|| |
| Baseline||10.70 (3.19)|| || ||9.32|| || ||<0.001|
| 12 week||28.38 (3.19)||−17.67||−26.53 to −8.82||53.34||−44.01||52.16 to −35.58|| |
| 24 week||33.93 (3.19)||−23.23||−32.1 to −8.8||68.29||−58.97||−67.1 to −6.8|| |
| Baseline||12.44 (3.12)|| || ||9.56 (2.87)|| || ||<0.001|
| 12 week||27.59 (3.12)||−15.15||−23.83 to −6.47||54.86 (2.87)||−45.3||53.28 to −37.31|| |
| 24 week||34.16 (3.12)||−21.71||−30.4 to −13.03||69.81 (2.87)||−60.26||−68.2 to −52.3|| |
| Baseline||73.23 (3.87)|| || ||74.36 (3.56)|| || ||0.002|
| 12 week||57.07 (3.87)||16.16||5.38 to 26.94||46.58 (3.56)||27.77||17.86 to 37.69|| |
| 24 week||48.99 (3.87)||24.24||13.5 to 35||29.49 (3.56)||44.87||34.9 to 54.8|| |
| Baseline||79.03 (3.96)|| || ||79.29 (3.64)|| || ||NS|
| 12 week||43.67 (3.96)||35.35||24.32 to 46.39||41.68 (3.64)||37.61||27.46 to 47.76|| |
| 24 week||32.56 (3.96)||46.47||35.46 to 57.5||23.73 (3.64)||55.56||45.41 to 65.7|| |
| Baseline||63.54 (4.49)|| || ||59.05 (4.13)|| || ||NS|
| 12 week||38.29 (4.49)||25.25||12.77 to 37.73||31.7 (4.13)||27.35||15.87 to 38.83|| |
| 24 week||28.19 (4.49)||35.35||22.88 to 47.83||22.3 (4.13)||36.75||25.27 to 48.23|| |
| Baseline||66.83 (4.4)|| || ||69.52 (4.05)|| || ||NS|
| 12 week||61.38 (4.4)||15.15||2.89 to 27.41||51.57 (4.05)||17.95||6.61 to 29.23|| |
| 24 week||43.6 (4.4)||23.23||10.97 to 35.49||34.9 (4.05)||34.61||23.34 to 45.89|| |
| Baseline||36.24 (4.06)|| || ||36 (3.73)|| || ||NS|
| 12 week||27.14 (4.06)||9.09||−2.22 to 20.4||18.05 (3.73)||17.95||7.55 to 28.35|| |
| 24 week||19.07 (4.06)||17.17||5.86 to 28.48||15.49 (3.73)||20.51||10.11 to 30.92|| |
| Baseline||85.71 (3.64)|| || ||85.97 (3.35)|| || ||NS|
| 12 week||70.9 (3.64)||14.81||4.66-24.96||61.37 (3.35)||24.5||15.17 to 33.84|| |
| 24 week||61.81 (3.64)||32.32||22.57 to 42.07||53.39 (3.35)||32.48||23.14 to 41.81|| |
| Baseline||13.96 (2.77)|| || ||13.83 (2.55)|| || ||0.02|
| 12 week||24.06 (2.77)||−11.61||−19.96 to −3.27||33.92 (2.55)||−19.32||−26.9 to −11.55|| |
| 24 week||27.09 (2.77)||−15.15||−23.5 to −6.8||43.74 (2.55)||−30.77||−38.4 to −23.1|| |
|Quality of life|
| Baseline||17.68 (2.99)|| || ||18.37 (2.76)|| || ||0.007|
| 12 week||29.30 (2.99)||−10.1||−17.82 to −2.37||37.6 (2.76)||−20.08||27.19 to −12.9|| |
| 24 week||32.84 (2.99)||−13.13||−20.8 to −5.4||49.14 (2.76)||−29.91||−37 to −22.8|| |
Post-hoc analysis of all the functioning dimensions of QLQ C-30 revealed a greater improvement in combined therapy group at 12 weeks. Pain scale was reduced in combined therapy while global quality of life and global health were improved at 12 weeks. The treatment effect for all the mentioned variables was sustained at 24 weeks, though the differences were lesser.
With respect to HADS-D, mean change from baseline to 12 weeks was 5.06 (95% CI: 3.64–6.48) and 5.7 (95% CI: 4.33–7.18) at 24 weeks for usual care group. For combined therapy group, mean change from baseline at 12 weeks was 5.95 (95% CI: 4.64–7.25), and 6.9 (95% CI: 5.61–8.23) at 24 weeks. Adjusted mean differences between both groups was 0.89 (95% CI: −0.99–2.77).
- Top of page
Our study suggests that using combined therapy for major depression in oncologic patients results in significant improvements in functioning dimensions, global health and global quality of life as well as reduction of pain at 12 weeks. These gains are sustained at the 6-month follow-up. We have also found a non-significant trend towards a reduction in depressive symptoms in the combined therapy group compared with the usual care group. Globally these results support the notion that combined treatment compared with usual care improves significantly overall patient's quality of life but does not reduce significantly depressive symptoms in oncologic patients. An explanation for the negative result in depressive symptomatology could be that HADS may not be a sensitive measure to evaluate therapeutic efficacy 47, 48.
HQoL seems really useful as a result variable when two treatments give the same level of symptoms reduction, since we can expect differences respect to HqoL 49. It is possible that HQoL scores are more sensitive to inherent changes to the adaptation process to illness or ‘response change phenomenon’ 50, than the measures of depressive symptoms 51. Combined intervention, which adds psychotherapy, could be more useful to impulse the adaptation process to the multiple challenges that the patient must face.
To our knowledge, this study represents the first randomized, controlled clinical trial to assess the effectiveness of NT plus escitalopram compared with escitalopram delivered by oncologists on measures of quality of life and depressive symptomatology among Spanish patients with non-metastatic cancer who met depression diagnostic criteria according to DSM-IV-TR. In addition, this study represents the first one in which an adaptation of NT for depression has been used. NT is an integrative intervention designed to address components that are of critical importance in depressed patients. In the oncological field, we have found that using somatosensorial interventions that modulate emotion arousal and mind–body interactions, as well as interventions directed to family and other important caregivers whose attitudes might influence psychological distress is crucial 18, 19.
Following the recommendations of Sellick, Newell and the Revised Consort Statement 5, 16, 52 about clinical trials, we have clearly detailed randomization procedure, have made blind outcome assessments and have accounted for all patient's disposition and clinical characteristics of the treatment population. In addition, we have included up to 30 patients per treatment arm, with a longer follow-up period than in previous studies. The type of intervention and the number of sessions have been described. The intervention was carried out by well-trained clinicians (psychiatrists and psychologists) with ample experience in the field of psychotherapy. Another advantage of our clinical trial is that the population of patients is homogeneous in terms of their illness status (non-metastatic), and the study includes data in cancer types that are less well studied than breast cancer, such as colon and lung cancer. Moreover, all the subjects met depression diagnostic criteria using clinical interviews conducted by trained clinicians.
The proportion of patients who discontinued from the trial was different between the two interventions with an overall better compliance for patients randomized to the combined therapy group. The reasons given by patients to stop medication were mainly the belief that they did not need antidepressants and the fear of a possible pharmacological interaction with cancer medications. NT with its emphasis in working with patient's narrative and with an acceptation stance might have a positive impact in patient's fears and worries about medication interactions and side effects.
Our study has several limitations. First, the sample size in each arm is less than 50, a number advised in Newell's review. However, most of previous clinical trials comparing psychosocial interventions had less than 30 per treatment group 18. This fact reflects the difficulties in recruiting patients with strict inclusion criteria and high refusal rates to participate. In our study, before randomization, approximately half of the sample (78 subjects) refused to participate in the study though they met all inclusion criteria. A substantial number of patients communicated their fear of worsening or collapsing if they talked about their feelings with a therapist. This refusal rate number is similar to other trials of depressed patients 14, 15, 21. Second, the combined therapy group had a higher percentage of patients with breast cancer and lower percentage of lung cancer patiens compared with the usual care group. Third, we have used an interim analysis to estimate statistical power and sample size. However, the analysis was done by an external unconnected study analyst, blind to the randomization. Finally, the administration of the drug was done by a different specialist. We have tried to be closer to natural conditions and in our setting, oncologists may use antidepressants in some patients without derivation to psychiatry service. In addition, due to the design of the study and with the objective of reducing comparison bias, combined therapy has been compared with a highly intensified usual care regimen that is not commonly used in our setting. The fact that the study has been developed in a real clinical context with patients that present well-diagnosed depression adds effectiveness value to this study. The applicability of our results might be compromised by the feasibility in the ‘real world’ and its cost-effectiveness, since intervention was delivered by well-trained therapists. The implementation of our approach might require further adaptation that could imply group interventions. One thing in favour of our intervention is that it proved to be acceptable to patients, shows good potential for dissemination, is relatively easy to implement, and improved the compliance.
As a conclusion we could say that NT used in combination with antidepressants appears to be a potentially beneficial intervention in depression in oncologic non-metastatic illness. Combined therapy does not seem superior to pharmacological approach to reduce depressive symptoms, but it improves much more quality of life. Further research, with larger samples, is clearly needed to better understand the psychological mechanisms provided by NT.
- Top of page
The authors thank the contribution of patients, oncologists and the Statistics and Epidemiology department of La Paz University Hospital for his valuable help.
Special thanks to Jose R Arribas, M.D for his helpful suggestions.
The work was supported with unrestricted grants from Lundbeck, FIS PI 050737 and FIS PI 052062. The funding source had no role in the study design, data collection, analysis and interpretation of the data, preparation of the manuscript, or the decision to submit the manuscript for publication.
Author contributions: Conception and design: Beatríz Rodríguez Vega, Carmen Bayón, Ángela Palao.
Provision of patients and application of interventions: Angela Palao, Ana Hospital, María Dieguez, Beatríz Castelo.
Collection and assembly of data: Guadalupe Torres, Guillermo Benito.
Data Analysis and interpretations: Carmen Bayón, Beatriz Rodríguez Vega, Angela Palao, Guadalupe Torres, Elia Pérez.
Manuscript writing: Beatríz Rodriguez Vega, Carmen Bayón.