Long‐term efficacy of meaning‐centered group psychotherapy for cancer survivors: 2‐Year follow‐up results of a randomized controlled trial

Abstract Objective Meaning‐centered group psychotherapy for cancer survivors (MCGP‐CS) is an effective intervention to improve personal meaning, psychological well‐being, and depressive symptoms until 6 months after the intervention. In this study, the long‐term effects of MCGP‐CS (i.e., at 1‐ and 2‐year follow‐up) on meaning, psychological well‐being and posttraumatic growth were assessed, in comparison to supportive group psychotherapy (SGP) and care as usual (CAU). Methods Cancer survivors (n = 170) were randomized into MCGP‐CS, SGP, or CAU. Assessments were scheduled at baseline, 1 week, 3 months, 6 months, 1 year, and 2 years postintervention. Outcome measures were the Personal Meaning Profile, Ryff's Scales of Psychological Well‐Being (SPWB), the Posttraumatic Growth Inventory, and their subscales. Linear mixed models (LMM) were used and results were both reported on an intention‐to‐treat (ITT) basis, as well as for intervention completers only. Results LMM and post hoc analyses with Bonferroni correction revealed that MCGP‐CS participants reported more improvement on positive relations (subscale of SPWB) than CAU participants of 2‐year postintervention (ITT analysis, Cohen's d = .82). Completers also reported more personal growth (subscale of SPWB) after MCGP‐CS than after SGP 1‐year postintervention (Cohen's d = .94). No long‐term effects were found on the other outcome measures. Conclusions In the 2 years after MCGP‐CS, the short‐term significant effects on personal meaning and most positive effects related to psychological well‐being faded. However, MCGP‐CS had a long‐term positive effect on positive relations with others and on survivors' sense of personal growth. Trial registration Netherlands Trial Register: NTR3571


| BACKGROUND
Many cancer survivors encounter physical hindrances and are confronted with psychosocial and existential problems, also years after curative treatment is completed. 1,2 There is growing evidence that meaning-focused coping is a viable way to successfully adjust to the aftermath of cancer, 2 especially if meaning can be made from the cancer experience. 3,4 Breitbart and colleagues developed meaningcentered group psychotherapy (MCGP) to improve psychological wellbeing in patients with advanced cancer. 5,6 This intervention is grounded in the work of the psychiatrist Viktor Frankl, 7 founder of logotherapy (ie, meaning therapy). MCGP was adapted for cancer survivors (MCGP-CS) by Van der Spek et al. 8  There is evidence that MCGP and MCGP-CS are effective in enhancing a sense of meaning, psychological well-being, and reducing depressive symptoms. 5,[8][9][10] MCGP-CS is likely to be cost-effective as well. 11 In a randomized controlled trial (RCT) among advanced cancer patients, MCGP was more effective than supportive group psychotherapy (SGP) in improving quality of life, spiritual well-being, and reducing depression and hopelessness. These improvements were sustained during the 2-month follow-up period. 5 Van der Spek et al. 9  Several other types of existential interventions have been developed, [12][13][14][15][16] and a few studies reported on the long-term effects of these interventions. In four RCT's on experiential-existential, 17 cognitiveexistential, 13 or supportive expressive group psychotherapy, 18 participants improved over the 1-year follow-up period, but not more than after a non-meaning-focused type of group psychotherapy 16,17 or the care as usual condition. 12,18 In a RCT on cognitive existential couple therapy, couples did better after the existential therapy compared to care as usual, and these results were maintained during the 9-month follow-up period. 19 The aim of this study is to investigate the long-term follow-up results of the RCT on the efficacy of MCGP-CS by Van der Spek et al. 9 Survivors' sense of personal meaning as well as psychological well-being and posttraumatic growth were compared for MCGP-CS, SGP, and CAU until 2 years after the intervention, both for all participants (intention-to-treat, ITT), and only for those who completed the intervention. Because many other things can also influence one's sense of meaning over the course of 2 years, in additional sensitivity analyses, psychological treatment and cancer recurrence during these 2 years were taken into account. Insight into the long-term MCGP-CS effects reveals whether this intervention supports survivors enduringly to experience a sense of meaning, well-being, and growth, despite the limitations of having had cancer.

| Study design and population
This study is an extension of a multicenter RCT on the efficacy of MCGP-CS compared to SGP and CAU with three follow-up assessments: postintervention and at 3-and 6-month follow-ups. In the present study, assessments were added at 1-and 2-year follow-ups. To limit participant burden, only personal meaning, psychological wellbeing, and posttraumatic growth were assessed at the long-term follow-ups. The study protocol and extension were approved by the Medical Ethics Committee of the Leiden University Medical Center and the trial was registered in the Netherlands Trial Register (NTR3571).
Details of the study procedure can be found in the previous report on the efficacy of MCGP-CS. 8 In brief, eligible patients were adult cancer survivors who were diagnosed in the last 5 years and who had completed treatment with curative intent. Participants had to have an expressed need for psychological care and at least one psychosocial complaint. Exclusion criteria were severe cognitive impairment, current psychiatric or psychological treatment elsewhere, and insufficient mastery of the Dutch language.
Informed consent was obtained from all the individual participants included in the study.

| Randomization and blinding
An independent researcher prepared a computer-generated randomization table with random block sizes and made a list of sequentially numbered allocations. Participants were placed in a group, and when a consecutive group had 7 to 10 participants, the independent researcher allocated the group to a study arm.

| Meaning-centered group psychotherapy for cancer survivors
MCGP-CS is a manualized intervention consisting of eight weekly 2-hour sessions. 20 The following themes were addressed: sources of meaning, meaning before and after cancer, past and future life stories as sources of meaning, participants' attitude toward life's limitations, creative sources of meaning, and experiential sources of meaning. In addition, important existential concepts played a role in MCGP-CS, such as identity, existential guilt, isolation, and freedom.

| Supportive group psychotherapy
SGP is a manualized intervention that aims to help survivors cope better with the cancer-related difficulties. 21 Like MCGP-CS, this intervention consists of eight weekly 2-hour sessions. The themes addressed were as follows: need for support, communicating with health care providers, coping with medical tests, with family and friends, with vocational issues, body image, limitations in physical functioning, and coping with the future.
Fidelity to both treatment protocols was ensured in several ways. 22

| Outcome measures
The primary outcome measure was personal meaning, measured as the total score of the Personal Meaning Profile (PMP). 23 The 49-item Dutch version of the Ryff's Scales of Psychological Well-Being (SPWB) was used to measure psychological well-being. 25,26 This measure has no total score. The original scale consists of six subscales: positive relations (α = .83), autonomy (α = .84), environmental mastery (α = .77), personal growth (α = .71), purpose in life (α = .79), and self-acceptance (α = .81). In the Dutch version, two subscales of spiritual well-being were added: inner strength (α = .75) and higher power (α = .91). 26 Items were scored on a six-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree). Subscale scores were calculated as the mean item score and higher scores indicated greater well-being.
Posttraumatic growth was measured using the total score of the Posttraumatic Growth Inventory. 27,28 This 21-item measure (α = .91) has five subscales: relating to others (α = .85), new possibilities (α = .80), personal strength (α = .80), spiritual change (α = .70), and appreciation of life (α = .75). Items were rated from 0 (not at all) to 5 (very great degree). The total score was calculated as the sum of all items and a higher score represented more posttraumatic growth.
A study-specific questionnaire was used to obtain sociodemographic characteristics. Clinical characteristics were retrieved from medical records. Uptake of psychological treatment was measured at baseline and all follow-up assessments, except postintervention, using the items about psychiatric and psychological treatment of the Treatment Inventory of Costs in Patients with psychiatric disorders. 29

| Statistical methods
Linear mixed models (LMM) with fixed effects for study arm, time, and their two-way interaction, as well as a random intercept for subjects, were used to investigate the differences in the course of the outcome measures over time in the three study arms. A correction was made for patient charactersitics if there were significant baseline differences across study arms. Also, analyses were corrected for the baseline scores of outcome measures in the case of significant differences between study arms at baseline. Results were reported on an ITT basis and for participants who attended six, seven, or all therapy sessions (completers).
If the course of an outcome measure differed significantly over time between the study arms, post hoc analyses were performed to assess which two groups differed significantly, using LMM, and between which points in time, using independent-samples t tests. Post hoc analyses were corrected for multiple testing by Bonferroni's correction. Cohen's d effect sizes were calculated by dividing the difference in change since baseline between the study arms by the pooled SD, calculated at all separate follow-up time points. Effect sizes of 0.2 were categorized as small, 0.5 as medium, and 0.8 as large.
As sensitivity analyses, all analyses were repeated without participants (a) who received psychological treatment during follow-up and (b) who faced cancer recurrence during follow-up. Analyses were performed in SPSS 24 and a two-sided P value < .05 was considered to indicate statistical significance.

| Study population
Basic information about the participant flow during the recruitment period and drop-out in various phases of the study can be found in  Table 2 and Table S1).
Post hoc LMM analyses with Bonferroni correction did not show a significant difference between two of the study arms in the course of personal meaning (PMP total score) and purpose in life (SPWB) from baseline to 2-year follow-up (Table 3)

| Completers
For completers, the results were largely comparable ( Table 2)   It is striking that none of the sources of meaning investigated in this study (eg, goal-orientedness) were significantly affected by MCGP-CS in the long-term. The few long-term effects that were identified all occurred on the measure of psychological well-being (SPWB). It is possible that the SPWB is more sensitive for change than the measure that was used for personal meaning (PMP). An alternative explanation may be that the long-term improvements in the area of psychological well-being were not strong enough to be translated into an enhanced sense of meaning. MCGP-CS's few long-term effects on psychological well-being were slightly stronger when analyzing completers only and without participants who received other psychological treatment during the follow-up period. No long-term effects were found when participants who faced cancer recurrence during follow-up were left out. Further research is not only needed to validate these long-term findings, but should also address the question how these long-term intervention effects interact with other major events in life.

| Sensitivity analyses
The long-term results of MCGP-CS seem to be in line with the results of previous studies on long-term effects of existential interventions. Overall, these effects seem to be quite modest. However, while some other studies did not find significant differences between the long-term effects of an existential intervention and a non-meaningfocused intervention 16,17 or care as usual, 13,30 the present study did find stronger improvement on some outcomes at long-term follow-up.

| Clinical implications
MCGP-CS is a useful addition to the current mental health care available in the oncology field. It is a brief intervention that is effective and cost-effective. 8,11 Some of its effects linger on for 1 or even 2 years.
Relatively easy adaptations could be made to stimulate stronger longterm improvements of psychological well-being and personal meaning.
Meta-analyses show that long-term effects could be stimulated by more contact hours, 31 possibly in the form of booster sessions. 32 MCGP-CS could also be extended with an online component, which can facilitate participants to remind and practice the skills they have learned. 33

| Study limitations
A strength of this study is its conservative ITT analyses with Bonferroni correction. The statistical methods decrease the chance of false positive findings, lending more credibility to the effects that were found. Yet, the possibility of chance findings could not be ruled out and the significant long-term effects should be interpreted tentatively; especially, because some results appeared to be inconsistent.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. LMM analyses are corrected for baseline score. *P < .05; **P < .005.