Internet-based Cognitive Behavioral Therapy for Breast Cancer Patients With Climacteric Symptoms

Background: Breast cancer is the most common cancer among women in the Netherlands, with approximately 14.000 new cases reported in 2011 (1). Nearly 30% of all women with breast cancer are premenopausal at the time of diagnosis (2). Ovarian damage is a major long-term sequela of chemotherapy in premenopausal women (3,4). Ovarian failure induced by chemotherapy and/or endocrine treatment leads to an earlier onset of menopause, with age and duration of treatment being its strongest predictors (5).

Premature menopause is a major concern of younger women undergoing (adjuvant/preventive) therapy for cancer (6). Primary menopausal symptoms include hot flushes, night sweats, vaginal dryness, decreased libido, dysuria and urinary incontinence. Secondary symptoms include insomnia due to night sweats, dyspareunia because of vaginal dryness, weight gain, and psychological distress (3,7,8). Previous studies have indicated that the severity of hot flushes is increased in women who experience transition into menopause during cancer treatment or after PBSO, compared to women going through natural menopause (9,10,11).

Among menopausal symptoms, hot flushes are considered to be the most disruptive, with prevalence rates between 63% and 80% in breast cancer patients (7,12-15). Moreover, vasomotor symptoms are an important reason why some women discontinue tamoxifen treatment (16,17). Hormonal replacement therapy (HRT) is highly effective in alleviating vasomotor symptoms associated with menopause (18). However, HRT is contraindicated in women with a history of breast cancer (19), even after prophylactic bilateral salpingo oophorectomy (PBSO) (20).

There is increasing evidence that behavioral interventions have a positive impact on symptoms in women with naturally occurring and treatment-induced menopause (21-27). A cognitive behavioral therapy (CBT) intervention including information about symptoms, monitoring and modifying precipitants, relaxation and stress management, cognitive restructuring of unhelpful assumptions and automatic thoughts, and encouraging helpful behavioral strategies has been found to be beneficial in the alleviation of menopausal symptoms in women with breast cancer (25,26), but compliance with face-to-face CBT programs can be problematic (26). A promising approach is to make this form of CBT more accessible and feasible for participants by having it available via the internet. Although the efficacy of CBT for this patient population has been demonstrated in the form of face-to-face group therapy, its efficacy has not been investigated when delivered via an internet platform in an individual setting.

Design. For this trial patients will be recruited from several hospitals in the Netherlands (predominantly Amsterdam region). Participants will be randomly allocated to either the guided intervention group, the self-management group or the control group (N = 83 per group). Upon completion of the study, the patients assigned to the control group will be given the opportunity to undergo the internet-based cognitive behavioral therapy program.

Women in the two intervention and control group will be asked to complete a battery of questionnaires prior to randomization (TO), at 10 weeks (T1) and at 6 months (T2).

Study measures. The primary outcomes are menopausal complaints, as assessed with a menopausal complaints questionnaire, the Functional Assessment of Cancer Therapy - Endocrine Symptoms (FACT-ES), and hot flushes and night sweats as assessed by the Hot Flushes Rating Scale (HFRS). The secondary study outcomes are sexuality problems (the Sexual Activity Questionnaire (SAQ)); sleep quality (The Groningen Sleep Quality Score (GSQS)); psychological distress (The Hospital Anxiety and Depression Scale (HADS)) and health related quality of life (The Medical Outcome Study 36-Item Short Form-36 Health Survey (SF-36)).

Patients' evaluation of the intervention program. The women in both intervention groups will be asked to provide their opinion of the internet-based CBT program at the 10 week evaluation point. All these women will be asked to complete a short questionnaire about the perceived efficacy of and satisfaction with the intervention program, whether they would suggest any changes to the program, and if they would recommend it to other women experiencing symptoms of premature menopause. In addition, telephone interviews (30 minutes) will be held with women who: (1) stated in the questionnaire that the intervention did not have the desired effect; and/or (2) gave the intervention a low rating (<6) and/or (3) would not recommend the program to others.

Cost effectiveness. The cost-effectiveness of the internet-based guided CBT versus the internet-based self-management CBT versus usual care will be expressed as: (1) cost per patient with a clinically relevant significant reduction on the HFRS and (2) cost per quality adjusted life year (QALY) gained.

Sample size and statistical power calculations: In total, 248 women will be recruited into the study, to allow for an attrition rate of approximately 20% (i.e., women who discontinue participation in the study entirely, including failure to complete all follow-up questionnaires; those women who discontinue participation in one of the groups but complete the follow-up assessments will be included in the analysis). Thus, a minimum of 130 women will be available for the primary intention-to-treat analysis. With a total sample of 130 women (65 per group), the study will have 80% power to detect an effect size of 0.5 with a p value of 0.05 (28).

Statistical analyses: Analyses will first be performed to evaluate the comparability of the intervention groups (guided versus self-management) and control group at baseline in terms of sociodemographic and clinical characteristics. ANOVA tests or appropriate non-parametric statistics will be used, depending on the level of measurement. If, despite the stratified randomization procedures, the groups are found not to be comparable on one more background variables, those variables will be employed routinely as covariates in subsequent analyses.

We will evaluate between-group differences over time in self-reported endocrine symptoms, sexual functioning, psychological distress, sleep quality and HRQOL. Scores for the FACT-ES, the HFRS, the SAQ, the HADS, the GSQS, and the SF-36 and the utility version of the SF-36, the SF-6D will be calculated according to published scoring algorithms.We will compare both intervention groups with the control group over time using multilevel procedures with repeated measures, using a restricted maximum likelihood (REML) solution to model specific contrasts between groups and follow-up assessment (29). Within each multilevel model the control group will be the reference category. Differences in mean change scores over time between the intervention groups and the control group will be accompanied by effect sizes (ES). As indicated previously, the FACT-ES and the HFRS will be used as the primary study endpoints, and the remaining measures will be considered as secondary endpoints. The p-value for overall model effects will be set at 0.05, and for specific contrasts at 0.01, lowering the risk of Type errors due to multiple testing.

All analyses will be conducted on an intention-to-treat (ITT) basis. In addition, per-protocol (PP) analyses will be performed on patients who met criteria for minimal compliance with the intervention(s). Supplementary analyses will be carried out in which data relating to compliance with the program elements is taken into account. Specifically, we will determine whether the level of compliance (based on completed homework assignments and self-report data) is associated significantly with the changes over time in symptom relief, sexuality, psychological distress, sleep quality, and cost-effectiveness. In addition the potential mediating effect of hot flush beliefs and behaviors will be examined through mediator analysis.

A cost-effectiveness analysis (CEA) will be performed using a validated health economic model as developed for use in the EVA-Trial (30). The analysis will consider the societal perspectives, the third-party payer perspective and the healthcare system perspective in comparing the effectiveness and costs of web-based CBT for alleviating treatment-induced menopausal symptoms versus waiting list control. Effectiveness data as gathered from the trial will be combined with the cost data from the cost-diary and hospital administration data.