original articles
Annals of Oncology Annals of Oncology 25: 378–384, 2014 doi:10.1093/annonc/mdt526 Published online 16 December 2013
Efficacy of short-term psychodynamic psychotherapy (STPP) with depressed breast cancer patients: results of a randomized controlled multicenter trial M. E. Beutel1,†, G. Weißflog2, †*, K. Leuteritz2, J. Wiltink1, A. Haselbacher1, C. Ruckes3, S. Kuhnt2, Y. Barthel2, B. H. Imruck1, R. Zwerenz1,‡ & E. Brähler2, ‡ 1 Department for Psychosomatic Medicine and Psychotherapy, University Medical Center Mainz, Mainz; 2Department for Medical Psychology and Medical Sociology, University Leipzig, Leipzig; 3Interdisciplinary Center for Clinical Trials, University Medical Center Mainz, Mainz, Germany
Background: There is a lack of trials of psychodynamic treatments of depression in breast cancer patients. The purpose of this trial was to determine the efficacy of short-term psychodynamic psychotherapy (STPP) in non-metastatic breast cancer patients diagnosed with depression, one of the most frequent mental comorbidities of breast cancer. Patients and methods: In a multicenter prospective trial, 157 breast cancer patients with comorbid depression were randomized to either individual STPP (intervention group, N = 78) or ‘treatment as usual’ (control group, TAU, N = 79). As our primary outcome measure, we hypothesized a higher rate of remission defined as no diagnosis of depression (Structured Clinical Interview for DSM-IV) and reduction in depression score by at least 2 points (Hospital Anxiety and Depression Scale, HADS-D) in STPP versus TAU at treatment termination. Secondary outcomes mainly refer to quality of life (QoL). Results: In the intention to treat (ITT) analysis, 44% of the STPP group achieved highly significantly more remission than TAU (23%). STPP treatment (OR = 7.64; P < 0.001) was the strongest predictor for remission post-treatment; time was also significant (OR = 0.96; P < 0.05). A high effect favoring STPP (d = 0.82) was observed for the HADS-D score posttreatment (secondary outcome). Regarding further secondary outcomes (QoL), analyses of covariance yielded main effects for group (favoring STPP with an effect size of at least d = 0.5) for global QoL, role, emotional and social functioning, pain, treatment side-effects, breast symptoms and upset by hair loss. Conclusions: STPP is an effective treatment of a broad range of depressive conditions in breast cancer patients improving depression and functional QoL. Findings are limited by the drop-out rate (∼1/3) and delayed post-treatment assessments. Future trials may consider stepped-care approaches, tailored to patients’ needs and requirements in the acute treatment phase. Key words: Short-term Psychodynamic Psychotherapy (STPP), depression, breast cancer, quality of life (QoL), efficacy
introduction As the leading cause of cancer death among females, breast cancer is associated with serious personal, social and health threats and losses. Depressive disorders are the most frequent mental comorbidities in ∼22% of cancer patients leading to substantial decrements of quality of life (QoL), compliance, work ability and prognosis [1]. Based on a meta-analysis of 23 randomized, controlled trials (RCT’s), cognitive behavioral therapy *Correspondence to: Dr G. Weißflog, Department of Medical Psychology and Medical Sociology, Psychosocial Oncology, University Leipzig, Philipp-Rosenthal-Straße 55, 04103 Leipzig, Germany. Tel: +49-341-9715415; Fax: +49-341-9715419; E-mail: gregor.weissfl[email protected] †
Joint 1st authorship. Joint last authorship.
‡
(CBT) had a positive, but small effect (d = 0.34) on depression in breast cancer patients [2]. Although there is evidence for the efficacy of psychodynamic group therapy in patients with advanced cancer [1], it is lacking in non-metastatic breast cancer. Combining supportive (e.g. alliance-building) and interpretative interventions, psychodynamic psychotherapy aims at gaining insight into and modifying intrapsychic and interpersonal conflicts underlying mental disorders. As supportive-expressive individual psychotherapy has been increasingly supported as an effective treatment of mental disorders such as depression [3], a short-term psychodynamic psychotherapy (STPP) manual was developed for depressed breast cancer patients [4]. The main purpose of the multicenter RCT was to determine the efficacy of manualized STPP in non-metastatic, depressed
© The Author 2013. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Received 13 June 2013; revised 26 August 2013 & 14 October 2013; accepted 21 October 2013
original articles
Annals of Oncology
breast cancer patients. Unlike previous trials, we assessed depression with a screening self-report measure (HADS) [5] and a diagnostic interview (SCID) [6] for those above cut-off. Patients who fulfilled both criteria were randomized either to STPP or to ‘treatment as usual’ (TAU). As our primary outcome measure, we hypothesized a higher remission rate in STPP versus TAU at follow-up. Secondary outcomes referred to an improved QoL.
patients and methods
procedure Patients were recruited after surgery, while still in oncological treatment (chemotherapy, radiation, hormonal treatment). Following detailed information and written informed consent, eligible, consecutive patients were entered into the study and filled out the screening questionnaire. Participants were required to fulfill the HADS-D criteria (depression score ≥8) and the diagnosis of a depressive disorder by the SCID-I interview. They were randomly assigned to STPP or TAU using computer-generated sets of numbers with random length (www.randomizer.org). Randomization was carried out in each study center by trial-independent research staff; group assignment was provided to the trial staff by closed, opaque envelopes. Quality assurance was carried out by the independent Interdisciplinary Center for Clinical Trials.
interventions We assumed that depression results from repetitive, maladaptive relationship patterns. The Central Core Conflict Relationship Theme [3] consists of unrealistic expectancies toward significant others (‘wish’), anticipated negative responses of the others to the wish and a negative response of the self (e.g. disappointed retreat). As in other psychodynamic treatments, therapists used supportive (alliance building) and interpretative interventions in order to modify these patterns. The treatment manual developed conjointly by the study groups of both sites [4] outlined specific strategies dealing with lifethreatening disease (e.g. hopelessness, suicidality, resource orientation). Up to 5 pre-treatment and 20 psychotherapy sessions were offered once weekly; treatment could be terminated earlier, when mutually agreed upon goals were achieved. A total of 20 psychodynamic psychotherapists in private practice (board certified or advanced post-graduate trainees) were trained (at least 2 × 2 h) in a group format by the respective local study center before entering the first patient. In order to assure treatment fidelity, psychotherapists presented each patient at three time points (focus formulation, middle, termination phase) in group supervision. Conceptualized as TAU, the ‘control group’ obtained written information on local cancer counseling centers. Additionally, they were offered written
Volume 25 | No. 2 | February 2014
study measures Demographic data were obtained by the baseline questionnaire; medical information was obtained from the medical records. At baseline, post-treatment (STPP: treatment termination; TAU: 6 months after baseline due to expected duration of STPP) and follow-up (6 months after second assessment), patients filled out the HADS, a reliable and valid measure of depression and anxiety in medically ill. Heightened depression was defined by a cut-off score of 8 (Cronbach Alpha 0.84). The SCID [7] was carried out by trained and supervised mental health professionals (at least bachelor degree), who were blinded to the intervention. Remission at follow-up as the primary outcome was defined as no diagnosis of depression (SCID) and a decrease in depression by at least two HADS-D points (based on reliable change index using means and standard deviations provided by meta-analysis). As main secondary outcomes, we assessed general and breast cancer-specific QoL by the EORTC Quality of Life Questionnaires C30 [8] and BR23 [9] supplemented by personality and helping alliance measures [7]. Medical, psychotherapeutic and psychopharmacological treatments were documented throughout the study.
statistical analysis For STPP of depression remission rates between 45% and 70% at treatment termination were published, while remission rates in primary-care settings (comparable with TAU) ranged between 20% and ∼50% [10]. For complex medical illness (cancer plus comorbid depression), we pragmatically expected a lower remission rate of 25% for TAU and 50% for STPP. Identifying this group difference of 25% by χ 2 test (two-tailed) at a power of 0.80 requires a total sample of N = 156. Analyses were done on an ITT basis. The primary end point (remission) was analyzed by a logistic regression with fixed effects for study center, HADS-D baseline score, treatment group, time until post-treatment and anti-depressant medication. Drop-outs were regarded as non-remitters. Remission was determined at P < 0.05; all secondary analyses were exploratory. The single components (HADS-D improvement ≥2 points, no SCID diagnosis) were analyzed separately by the same analysis model. Additional per protocol analyses were carried out for completers (for STPP defined by complete data and at least five sessions of psychotherapy; for TAU, complete data) by an analysis of covariance (ANCOVA) model with fixed effects for treatment and center and the baseline values (HADS-D and QoL scores) and time between randomization and end of treatment as covariates. Standard mean differences were calculated with Cohen’s d for completers for primary and secondary outcomes (Supplementary data, available at Annals of Oncology online). Data were analyzed by SPSS Version 20.0.
results patient recruitment was carried out from December 2008 to May 2011 (Figure 1) Of a total of 5989 breast cancer patients from 16 cooperating hospitals, 3044 (50.8%) did not meet inclusion criteria, and 1639 (27.4%) declined to participate for the following reasons: 55% no return of questionnaires or no reason, 28% no need or time, 8% enough support, 4% overburdened by disease or treatment, 5% unwilling to participate in a clinical trial. Of those eligible patients screened (N = 1306), a depression score ≥8 was fulfilled by N = 256 (19.6%). Of those, 99 did not meet SCID criteria for a depressive disorder or declined further
doi:10.1093/annonc/mdt526 |
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Patients were recruited from oncological centers by the two study centers Mainz and Leipzig, Germany [7]. Eligible women fulfilled the following criteria: (1) breast cancer (stages T0-4, N0-1, M0), (2) curative treatment, (3) age 18–69 years, (4) German language competence, (5) depression score (HADS-D) ≥ 8, (6) depressive disorder according to SCID-I (ICD-10 diagnoses: depressive episode F32.-, recurrent depressive episode F33.-; dysthymia F34.1, adjustment disorder F43.2), (7) written consent with study participation. Exclusion criteria were: (1) severe additional medical disorder, (2) psychotic disorder, suicidality, acute substance related disorder, personality disorders except for cluster C, organic mental disorder, (3) concurrent psychotherapy. The study was approved by the Ethics Committees of the State of Rhineland-Palatinate [reference number 837.380.06 (5478)] and the University of Leipzig [reference number 218-2007].
psychodiagnostical information to their general practitioner who may then initiate psychotherapeutic referral or antidepressant treatment (‘as usual’).
original articles Enrollment
Annals of Oncology
Assessed for eligibility (N = 5989)
Excluded (N = 4683) Not meeting inclusion criteria (N = 3044) Declined to participate (N = 1639) Eligible & screened (N = 1306) Excluded (N = 1149) Screening negative (N = 1050) Declined to participate or SCID-I negative (N = 99)
Randomized (N = 157)
Allocated to intervention (STPP): N = 78 Received allocated intervention: N = 68 Did not receive allocated intervention: N = 10 . N = 3 no sufficient motivation . N = 3 severe somatic/mental comorbidity . N = 4 other reasons
Allocated to control group (TAU): N = 79 Received allocated intervention: N = 78 Did not receive allocated intervention: N = 1 . Withdrew consent after randomization to TAU
Follow-Up T2 Lost to follow-up: N = 20 . N = 18 HADS & SCID missing . N = 1 HADS missing . N = 1 SCID missing
Lost to follow-up: N = 25 . N = 20 HADS & SCID missing . N = 4 SCID missing . N = 1 HADS missing Discontinued: N = 0
Discontinued intervention: N = 12 . N = 5 no sufficient motivation . N = 4 somatic/mental comorbidity . N = 3 other reasons (incl. 1 ongoing) Completed intervention: N = 52
Analysis T2 Analysed N = 52 Excluded from analysis: N = 26 . N = 18 no follow-up & not participated or dropout of intervention . N = 6 follow-up complete but dropout of intervention . N = 2 follow-up incomplete but intervention complete
Analysed N = 54 Excluded from analysis: N = 25 . N = 24 follow-up incomplete . N = 1 withdrawal of consent & follow-up incomplete
Figure 1. CONSORT diagram showing patient flow through the trial.
participation. A total of 157 patients were randomized to STPP (N = 78) or TAU (N = 79). In STPP, 10 patients did not start the allocated intervention due to insufficient motivation, somatic or mental comorbidity or other reasons. In TAU, one randomized patient withdrew consent. Fifty-two patients received a minimum of five psychotherapy sessions and were defined as completers. Per protocol analyses were conducted with N = 52 (66.7%) in STPP and N = 54 patients (69.2%) in TAU.
| Beutel et al.
Table 1 shows social and medical ‘baseline’ characteristics for STPP and for TAU. The mean age of participants was 51.8 years. Previously, 32.7% had undergone psychotherapy, of those N = 22 (43.1%) for previous depression. The HADS-D score was significantly higher in STPP compared with TAU, but there was no difference regarding any other medical or social baseline characteristic. At baseline, over 70% suffered from a depressive episode (11.6% recurrent); 10.3% were on anti-depressant
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Allocation T1
original articles
Annals of Oncology
Table 1. Social and medical baseline characteristics: STPP versus TAU TAU (N = 78)a
Total (N = 156)
P-value
51.7 (9.9)
51.9 (8.3)
51.8 (9.1)
NS
12.8 (22.4)
13.4 (16.5)
13.1 (19.3)
NS
12.4 (3.2)
11.2 (2.8)
11.8 (3.1)
0.017
11 (14.1) 46 (59.0) 21 (26.9) 56 (71.8) 62 (79.5)
9 (11.5) 41 (52.6) 28 (35.9) 53 (67.9) 64 (82.1)
20 (12.8) 87 (55.8) 49 (31.4) 109 (69.8) 126 (80.8)
NS
24 (30.8) 14 (17.9) 8 (10.3) 13 (16.7) 17 (21.8)
34 (43.6) 16 (20.5) 7 (9.0) 11 (14.1) 9 (11.5)
58 (37.2) 30 (19.2) 15 (9.6) 24 (15.4) 26 (16.7)
NS
16 (20.5) 37 (47.4) 25 (32.1) 29 (37.2)
17 (21.8) 36 (46.2) 25 (32.1) 22 (28.2)
33 (21.2) 73 (46.8) 50 (32.1) 51 (32.7)
NS
13 (16.7) 39 (50.0) 6 (7.7) 6 (7.7) 14 (17.9) 8 (10.3) 69 (88.5)
13 (16.7) 34 (43.6) 5 (6.4) 8 (10.3) 18 (23.1) 8 (10.3) 71 (91.0)
26 (16.7) 73 (46.8) 11 (7.1) 14 (9.0) 32 (20.5) 16 (10.3) 140 (89.7)
5 (6.4) 33 (42.3) 29 (37.2) 6 (7.7) 54 (69.2) 50 (64.1) 17 (21.8)
9 (11.5) 32 (41.0) 23 (29.5) 8 (10.3) 54 (69.2) 45 (57.7) 19 (24.4)
14 (8.9) 65 (41.7) 52 (33.3) 14 (8.9) 108 (69.2) 95 (60.9) 36 (23.1)
NS NS
NS NS
NS NS NS
NS NS NS
SD, standard deviation; NS, no statistical significant difference in χ 2, respectively, t-tests for independent samples. N = 1 case not analysed due to withdrawal of consent after randomization. b NSTPP = 43, NTAU = 51 on sick leave. c To missing data per group: NTAU = 2. d NSTPP = 2, NTAU = 1. e NSTPP = 26, NTAU = 24. f NSTPP = 1. g NSTPP = 5, NTAU = 6. h NSTPP = 5, NTAU = 13. i NSTPP = 5, NTAU = 15. j NSTPP = 13, NTAU = 17. a
treatment. For over 90%, this was the first manifestation of breast cancer. All patients completed medical treatment or were in ongoing medical treatment (71.8% radiation, 59.0% hormone, 57.7% chemotherapy, 11.5% neoadjuvant chemotherapy, 8.3%
Volume 25 | No. 2 | February 2014
trastuzumab); 89.1% had undergone operation (60.3% breast conserving). The first follow-up assessment took place after a mean of 56 (SD = 12.64) weeks in STPP, later than in TAU with a mean of
doi:10.1093/annonc/mdt526 |
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Age Mean (SD) Work disability last year (weeks)b Mean (SD) HADS score (depression) Mean (SD) Marital status [n (%)] Single Married Separated, divorced, widowed Partnership (yes)c [n (%)] Children (yes) [n (%)] Employmentd [n (%)] Full time Part time Household Unemployed Pension Education [n (%)]
Annals of Oncology Annals of Oncology 25: 378–384, 2014 doi:10.1093/annonc/mdt526 Published online 16 December 2013
Efficacy of short-term psychodynamic psychotherapy (STPP) with depressed breast cancer patients: results of a randomized controlled multicenter trial M. E. Beutel1,†, G. Weißflog2, †*, K. Leuteritz2, J. Wiltink1, A. Haselbacher1, C. Ruckes3, S. Kuhnt2, Y. Barthel2, B. H. Imruck1, R. Zwerenz1,‡ & E. Brähler2, ‡ 1 Department for Psychosomatic Medicine and Psychotherapy, University Medical Center Mainz, Mainz; 2Department for Medical Psychology and Medical Sociology, University Leipzig, Leipzig; 3Interdisciplinary Center for Clinical Trials, University Medical Center Mainz, Mainz, Germany
Background: There is a lack of trials of psychodynamic treatments of depression in breast cancer patients. The purpose of this trial was to determine the efficacy of short-term psychodynamic psychotherapy (STPP) in non-metastatic breast cancer patients diagnosed with depression, one of the most frequent mental comorbidities of breast cancer. Patients and methods: In a multicenter prospective trial, 157 breast cancer patients with comorbid depression were randomized to either individual STPP (intervention group, N = 78) or ‘treatment as usual’ (control group, TAU, N = 79). As our primary outcome measure, we hypothesized a higher rate of remission defined as no diagnosis of depression (Structured Clinical Interview for DSM-IV) and reduction in depression score by at least 2 points (Hospital Anxiety and Depression Scale, HADS-D) in STPP versus TAU at treatment termination. Secondary outcomes mainly refer to quality of life (QoL). Results: In the intention to treat (ITT) analysis, 44% of the STPP group achieved highly significantly more remission than TAU (23%). STPP treatment (OR = 7.64; P < 0.001) was the strongest predictor for remission post-treatment; time was also significant (OR = 0.96; P < 0.05). A high effect favoring STPP (d = 0.82) was observed for the HADS-D score posttreatment (secondary outcome). Regarding further secondary outcomes (QoL), analyses of covariance yielded main effects for group (favoring STPP with an effect size of at least d = 0.5) for global QoL, role, emotional and social functioning, pain, treatment side-effects, breast symptoms and upset by hair loss. Conclusions: STPP is an effective treatment of a broad range of depressive conditions in breast cancer patients improving depression and functional QoL. Findings are limited by the drop-out rate (∼1/3) and delayed post-treatment assessments. Future trials may consider stepped-care approaches, tailored to patients’ needs and requirements in the acute treatment phase. Key words: Short-term Psychodynamic Psychotherapy (STPP), depression, breast cancer, quality of life (QoL), efficacy
introduction As the leading cause of cancer death among females, breast cancer is associated with serious personal, social and health threats and losses. Depressive disorders are the most frequent mental comorbidities in ∼22% of cancer patients leading to substantial decrements of quality of life (QoL), compliance, work ability and prognosis [1]. Based on a meta-analysis of 23 randomized, controlled trials (RCT’s), cognitive behavioral therapy *Correspondence to: Dr G. Weißflog, Department of Medical Psychology and Medical Sociology, Psychosocial Oncology, University Leipzig, Philipp-Rosenthal-Straße 55, 04103 Leipzig, Germany. Tel: +49-341-9715415; Fax: +49-341-9715419; E-mail: gregor.weissfl[email protected] †
Joint 1st authorship. Joint last authorship.
‡
(CBT) had a positive, but small effect (d = 0.34) on depression in breast cancer patients [2]. Although there is evidence for the efficacy of psychodynamic group therapy in patients with advanced cancer [1], it is lacking in non-metastatic breast cancer. Combining supportive (e.g. alliance-building) and interpretative interventions, psychodynamic psychotherapy aims at gaining insight into and modifying intrapsychic and interpersonal conflicts underlying mental disorders. As supportive-expressive individual psychotherapy has been increasingly supported as an effective treatment of mental disorders such as depression [3], a short-term psychodynamic psychotherapy (STPP) manual was developed for depressed breast cancer patients [4]. The main purpose of the multicenter RCT was to determine the efficacy of manualized STPP in non-metastatic, depressed
© The Author 2013. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Received 13 June 2013; revised 26 August 2013 & 14 October 2013; accepted 21 October 2013
original articles
Annals of Oncology
breast cancer patients. Unlike previous trials, we assessed depression with a screening self-report measure (HADS) [5] and a diagnostic interview (SCID) [6] for those above cut-off. Patients who fulfilled both criteria were randomized either to STPP or to ‘treatment as usual’ (TAU). As our primary outcome measure, we hypothesized a higher remission rate in STPP versus TAU at follow-up. Secondary outcomes referred to an improved QoL.
patients and methods
procedure Patients were recruited after surgery, while still in oncological treatment (chemotherapy, radiation, hormonal treatment). Following detailed information and written informed consent, eligible, consecutive patients were entered into the study and filled out the screening questionnaire. Participants were required to fulfill the HADS-D criteria (depression score ≥8) and the diagnosis of a depressive disorder by the SCID-I interview. They were randomly assigned to STPP or TAU using computer-generated sets of numbers with random length (www.randomizer.org). Randomization was carried out in each study center by trial-independent research staff; group assignment was provided to the trial staff by closed, opaque envelopes. Quality assurance was carried out by the independent Interdisciplinary Center for Clinical Trials.
interventions We assumed that depression results from repetitive, maladaptive relationship patterns. The Central Core Conflict Relationship Theme [3] consists of unrealistic expectancies toward significant others (‘wish’), anticipated negative responses of the others to the wish and a negative response of the self (e.g. disappointed retreat). As in other psychodynamic treatments, therapists used supportive (alliance building) and interpretative interventions in order to modify these patterns. The treatment manual developed conjointly by the study groups of both sites [4] outlined specific strategies dealing with lifethreatening disease (e.g. hopelessness, suicidality, resource orientation). Up to 5 pre-treatment and 20 psychotherapy sessions were offered once weekly; treatment could be terminated earlier, when mutually agreed upon goals were achieved. A total of 20 psychodynamic psychotherapists in private practice (board certified or advanced post-graduate trainees) were trained (at least 2 × 2 h) in a group format by the respective local study center before entering the first patient. In order to assure treatment fidelity, psychotherapists presented each patient at three time points (focus formulation, middle, termination phase) in group supervision. Conceptualized as TAU, the ‘control group’ obtained written information on local cancer counseling centers. Additionally, they were offered written
Volume 25 | No. 2 | February 2014
study measures Demographic data were obtained by the baseline questionnaire; medical information was obtained from the medical records. At baseline, post-treatment (STPP: treatment termination; TAU: 6 months after baseline due to expected duration of STPP) and follow-up (6 months after second assessment), patients filled out the HADS, a reliable and valid measure of depression and anxiety in medically ill. Heightened depression was defined by a cut-off score of 8 (Cronbach Alpha 0.84). The SCID [7] was carried out by trained and supervised mental health professionals (at least bachelor degree), who were blinded to the intervention. Remission at follow-up as the primary outcome was defined as no diagnosis of depression (SCID) and a decrease in depression by at least two HADS-D points (based on reliable change index using means and standard deviations provided by meta-analysis). As main secondary outcomes, we assessed general and breast cancer-specific QoL by the EORTC Quality of Life Questionnaires C30 [8] and BR23 [9] supplemented by personality and helping alliance measures [7]. Medical, psychotherapeutic and psychopharmacological treatments were documented throughout the study.
statistical analysis For STPP of depression remission rates between 45% and 70% at treatment termination were published, while remission rates in primary-care settings (comparable with TAU) ranged between 20% and ∼50% [10]. For complex medical illness (cancer plus comorbid depression), we pragmatically expected a lower remission rate of 25% for TAU and 50% for STPP. Identifying this group difference of 25% by χ 2 test (two-tailed) at a power of 0.80 requires a total sample of N = 156. Analyses were done on an ITT basis. The primary end point (remission) was analyzed by a logistic regression with fixed effects for study center, HADS-D baseline score, treatment group, time until post-treatment and anti-depressant medication. Drop-outs were regarded as non-remitters. Remission was determined at P < 0.05; all secondary analyses were exploratory. The single components (HADS-D improvement ≥2 points, no SCID diagnosis) were analyzed separately by the same analysis model. Additional per protocol analyses were carried out for completers (for STPP defined by complete data and at least five sessions of psychotherapy; for TAU, complete data) by an analysis of covariance (ANCOVA) model with fixed effects for treatment and center and the baseline values (HADS-D and QoL scores) and time between randomization and end of treatment as covariates. Standard mean differences were calculated with Cohen’s d for completers for primary and secondary outcomes (Supplementary data, available at Annals of Oncology online). Data were analyzed by SPSS Version 20.0.
results patient recruitment was carried out from December 2008 to May 2011 (Figure 1) Of a total of 5989 breast cancer patients from 16 cooperating hospitals, 3044 (50.8%) did not meet inclusion criteria, and 1639 (27.4%) declined to participate for the following reasons: 55% no return of questionnaires or no reason, 28% no need or time, 8% enough support, 4% overburdened by disease or treatment, 5% unwilling to participate in a clinical trial. Of those eligible patients screened (N = 1306), a depression score ≥8 was fulfilled by N = 256 (19.6%). Of those, 99 did not meet SCID criteria for a depressive disorder or declined further
doi:10.1093/annonc/mdt526 |
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Patients were recruited from oncological centers by the two study centers Mainz and Leipzig, Germany [7]. Eligible women fulfilled the following criteria: (1) breast cancer (stages T0-4, N0-1, M0), (2) curative treatment, (3) age 18–69 years, (4) German language competence, (5) depression score (HADS-D) ≥ 8, (6) depressive disorder according to SCID-I (ICD-10 diagnoses: depressive episode F32.-, recurrent depressive episode F33.-; dysthymia F34.1, adjustment disorder F43.2), (7) written consent with study participation. Exclusion criteria were: (1) severe additional medical disorder, (2) psychotic disorder, suicidality, acute substance related disorder, personality disorders except for cluster C, organic mental disorder, (3) concurrent psychotherapy. The study was approved by the Ethics Committees of the State of Rhineland-Palatinate [reference number 837.380.06 (5478)] and the University of Leipzig [reference number 218-2007].
psychodiagnostical information to their general practitioner who may then initiate psychotherapeutic referral or antidepressant treatment (‘as usual’).
original articles Enrollment
Annals of Oncology
Assessed for eligibility (N = 5989)
Excluded (N = 4683) Not meeting inclusion criteria (N = 3044) Declined to participate (N = 1639) Eligible & screened (N = 1306) Excluded (N = 1149) Screening negative (N = 1050) Declined to participate or SCID-I negative (N = 99)
Randomized (N = 157)
Allocated to intervention (STPP): N = 78 Received allocated intervention: N = 68 Did not receive allocated intervention: N = 10 . N = 3 no sufficient motivation . N = 3 severe somatic/mental comorbidity . N = 4 other reasons
Allocated to control group (TAU): N = 79 Received allocated intervention: N = 78 Did not receive allocated intervention: N = 1 . Withdrew consent after randomization to TAU
Follow-Up T2 Lost to follow-up: N = 20 . N = 18 HADS & SCID missing . N = 1 HADS missing . N = 1 SCID missing
Lost to follow-up: N = 25 . N = 20 HADS & SCID missing . N = 4 SCID missing . N = 1 HADS missing Discontinued: N = 0
Discontinued intervention: N = 12 . N = 5 no sufficient motivation . N = 4 somatic/mental comorbidity . N = 3 other reasons (incl. 1 ongoing) Completed intervention: N = 52
Analysis T2 Analysed N = 52 Excluded from analysis: N = 26 . N = 18 no follow-up & not participated or dropout of intervention . N = 6 follow-up complete but dropout of intervention . N = 2 follow-up incomplete but intervention complete
Analysed N = 54 Excluded from analysis: N = 25 . N = 24 follow-up incomplete . N = 1 withdrawal of consent & follow-up incomplete
Figure 1. CONSORT diagram showing patient flow through the trial.
participation. A total of 157 patients were randomized to STPP (N = 78) or TAU (N = 79). In STPP, 10 patients did not start the allocated intervention due to insufficient motivation, somatic or mental comorbidity or other reasons. In TAU, one randomized patient withdrew consent. Fifty-two patients received a minimum of five psychotherapy sessions and were defined as completers. Per protocol analyses were conducted with N = 52 (66.7%) in STPP and N = 54 patients (69.2%) in TAU.
| Beutel et al.
Table 1 shows social and medical ‘baseline’ characteristics for STPP and for TAU. The mean age of participants was 51.8 years. Previously, 32.7% had undergone psychotherapy, of those N = 22 (43.1%) for previous depression. The HADS-D score was significantly higher in STPP compared with TAU, but there was no difference regarding any other medical or social baseline characteristic. At baseline, over 70% suffered from a depressive episode (11.6% recurrent); 10.3% were on anti-depressant
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Allocation T1
original articles
Annals of Oncology
Table 1. Social and medical baseline characteristics: STPP versus TAU TAU (N = 78)a
Total (N = 156)
P-value
51.7 (9.9)
51.9 (8.3)
51.8 (9.1)
NS
12.8 (22.4)
13.4 (16.5)
13.1 (19.3)
NS
12.4 (3.2)
11.2 (2.8)
11.8 (3.1)
0.017
11 (14.1) 46 (59.0) 21 (26.9) 56 (71.8) 62 (79.5)
9 (11.5) 41 (52.6) 28 (35.9) 53 (67.9) 64 (82.1)
20 (12.8) 87 (55.8) 49 (31.4) 109 (69.8) 126 (80.8)
NS
24 (30.8) 14 (17.9) 8 (10.3) 13 (16.7) 17 (21.8)
34 (43.6) 16 (20.5) 7 (9.0) 11 (14.1) 9 (11.5)
58 (37.2) 30 (19.2) 15 (9.6) 24 (15.4) 26 (16.7)
NS
16 (20.5) 37 (47.4) 25 (32.1) 29 (37.2)
17 (21.8) 36 (46.2) 25 (32.1) 22 (28.2)
33 (21.2) 73 (46.8) 50 (32.1) 51 (32.7)
NS
13 (16.7) 39 (50.0) 6 (7.7) 6 (7.7) 14 (17.9) 8 (10.3) 69 (88.5)
13 (16.7) 34 (43.6) 5 (6.4) 8 (10.3) 18 (23.1) 8 (10.3) 71 (91.0)
26 (16.7) 73 (46.8) 11 (7.1) 14 (9.0) 32 (20.5) 16 (10.3) 140 (89.7)
5 (6.4) 33 (42.3) 29 (37.2) 6 (7.7) 54 (69.2) 50 (64.1) 17 (21.8)
9 (11.5) 32 (41.0) 23 (29.5) 8 (10.3) 54 (69.2) 45 (57.7) 19 (24.4)
14 (8.9) 65 (41.7) 52 (33.3) 14 (8.9) 108 (69.2) 95 (60.9) 36 (23.1)
NS NS
NS NS
NS NS NS
NS NS NS
SD, standard deviation; NS, no statistical significant difference in χ 2, respectively, t-tests for independent samples. N = 1 case not analysed due to withdrawal of consent after randomization. b NSTPP = 43, NTAU = 51 on sick leave. c To missing data per group: NTAU = 2. d NSTPP = 2, NTAU = 1. e NSTPP = 26, NTAU = 24. f NSTPP = 1. g NSTPP = 5, NTAU = 6. h NSTPP = 5, NTAU = 13. i NSTPP = 5, NTAU = 15. j NSTPP = 13, NTAU = 17. a
treatment. For over 90%, this was the first manifestation of breast cancer. All patients completed medical treatment or were in ongoing medical treatment (71.8% radiation, 59.0% hormone, 57.7% chemotherapy, 11.5% neoadjuvant chemotherapy, 8.3%
Volume 25 | No. 2 | February 2014
trastuzumab); 89.1% had undergone operation (60.3% breast conserving). The first follow-up assessment took place after a mean of 56 (SD = 12.64) weeks in STPP, later than in TAU with a mean of
doi:10.1093/annonc/mdt526 |
Downloaded from http://annonc.oxfordjournals.org/ at New York University on October 9, 2014
Age Mean (SD) Work disability last year (weeks)b Mean (SD) HADS score (depression) Mean (SD) Marital status [n (%)] Single Married Separated, divorced, widowed Partnership (yes)c [n (%)] Children (yes) [n (%)] Employmentd [n (%)] Full time Part time Household Unemployed Pension Education [n (%)]
