Participants and study design:
This study was designed in randomized, double blinded, placebo-controlled clinical trial. The patients’ recruitment was carried out from March 2020 to March 2021 among outpatients in hematology-oncology clinic of Omid hospital, Isfahan, Iran. Omid hospital is a 200-beds, referral, and university-affiliated hospital specialized for the treatment of hematology-oncology patients. The protocol of study has been approved in Isfahan University of Medical Sciences ethic committee (ID number: IR.MUI.RESEARCH.REC.1399.362) and also trial was submitted in Iranian clinical trial registry center (IRCT20180722040556N6).
We included adult patients (18 years and older) suffering from hormone-positive breast cancer of any degree (I to III) who had completed all course of chemotherapy and radiotherapy treatment while they were under the treatment by hormone therapy.
Hormone therapy was included by administration of selective estrogen receptor modulators (SERM) family drugs (including tamoxifen) or aromatase inhibitors (including steroidal and non-steroidal agents). Before starting recruitment, all patients who had signed the consent form firstly were screened by hospital anxiety and depression scale (HADS), translated and validated by Montazeri et al [13]. The HADS is a fourteen-item scale widely used for measuring psychological morbidity in cancer patients. Scoring for each item ranges from zero to three.
If the patients’ score was more than 11 in the early evaluation, the patient was asked to consult with a psychiatrist, if not, he or she were invited to participate in our study if they were complaining of mood or sleep changes during one month ago.
The exclusion criteria were included, patients with a history of any other malignancies, patients receiving anticonvulsant medications due to their active seizures, patients who had any known psychiatric problems (depression, bipolar, psychosis and etc.) or those who were receiving medications that affected the central nervous system, such as antipsychotics, hypnotic or sedative (such as benzodiazepines, non-benzodiazepines agents or any drugs with antihistamine properties) or antidepressants and anti-anxiety. We also excluded patients who were receiving drugs such as propranolol (due to the relative effect on the central nervous system) and warfarin (due to the high risk of interactions), any supplements such as soy supplements or vitamin E to reduce menopausal symptoms. Patients with any degree of liver and kidney failure and autoimmune diseases and those who had a history of smoking, chronic alcohol consumption were also excluded.
Seventy-eight patients were considered for patients sampling, among them, 72 patients fulfilled the inclusion criteria and signed the consent form. The intervention group received 1 gram of omega-3 (containing 180 mg EPA and 120 mg DHA) twice a day orally and the placebo group received identical soft gelatin capsule with the same shape and packaging containing liquid paraffin). Both omega-3 supplement and identical placebo were produced by Zahravi® Pharmaceutical Company, Tabriz, Iran.
For baseline data gathering, we recruited the patients and asked them to fill all questionnaires at the end of first week. Then, in the second week of follow-up period, we administered the omega-3 or placebo randomly twice daily for 4 weeks. The randomization of samples was carried out based on the blocked randomization method. Information such as the number of treatment groups (2 main intervention groups for example A and placebo for example B), block size (a multiple of the number of groups that were selected in this study) and total patients (the sample size of 60 people) were entered into the internet-based software machines to assigned specific codes to each patients in order to receive omega-3 or placebo soft gelatin capsule. Block randomization was commonly used to balance the number of samples assigned to each of the studied groups when there was sample size limitation. After completing the sample size allocation, the code of each patient was opened and matched with the software output. For keeping the study blinded, the data collector, the investigators and the statist were not informed of allocated drug code before data analyses.
Questionnaires:
At the end of 4-week follow-up, all questionnaires including center for epidemiological studies-depression scale (CES-D) [14], profile of mood states (POMS) [15], Pittsburgh sleep quality index (PSQI) [16] questionnaires were applied again and the results were compared with the baseline.
Center for epidemiological studies-depression scale is a brief, 20-question, self-report scale designed to measure self-reported symptoms associated with depression in six different scales of depressed mood, feelings of guilt and worthlessness, feelings of helplessness and hopelessness, psychomotor retardation, loss of appetite, and sleep disturbance which were experienced in the past week.
The POMS is a clinical psychological rating scale used to assess transient, distinct mood states in 40-item questionnaire. Options of this tool are rated from zero (I do not feel that way at all) to 4 (I have that feeling too much).
The last assessed questionnaire was PSQI which has 10 different questions about sleep quality to be asked from the patient in the past month. It is an effective self-administrated instrument includes 4 open-ended questions and 14 questions to be answered using event-frequency and semantic scales. The tool looks at seven areas: subjective sleep quality, habitual sleep efficiency, sleep disturbances, the use of sleep-promoting medication and daytime dysfunction.
To assess the validity of the Persian version of the questionnaire, it was evaluated by a group of specialists, including two psychologists, one literature expert, two oncologists, and 10 general physicians. Then, a forward-backward translation method was employed for translating the questionnaires, and the same questionnaires were given to 30 matched patients with 2 weeks apart. And finally, the reliability of the questionnaires with an intraclass correlation coefficient (ICC) were assessed. Those patients who were not consumed at least 80% of total cumulative doses of omega-3 or identical placebo were considered as non-adherent patients and were excluded from the study. In addition, we evaluated every reported adverse drug reaction based on common terminology criteria for adverse events (CTCAE) version 4 during 4-weeks’ follow-up [17, 18]. For the simple review of the proposed method, the graphical abstract was shown in Figure 1.
Statistical Approach:
Sample size calculation:
The following equation has been used to determine the number of experimental samples in this research and to investigate the changes of a specific parameter that is used in two different modes [19].
In this equation, N indicates the number of patients, is the reliability coefficient (95%), is the power factor (80%), and is the probability of beta power in the first and second groups (50%), and d is called the study error.
Due to difficulties in reaching optimal sample size, and based on the inclusion criteria and subjects who dropped out in addition to missing data, the authors had to terminate the study by fewer samples, adopting a pilot study design (See flow diagram).
Statistical analysis:
Per-protocol analysis was considered for comparison the data of those patients who completed the treatment originally allocated. In the data analysis stage, Kolmogorov-Smirnov test was used to evaluate the normality of data distribution in both intervention and placebo groups (patients’ demographic and clinical characteristics). Also, in order to evaluate the homogeneity of the distribution parameters in two groups chi-squared were applied. The differences between two groups were assessed by paired t-test, analysis of variance with repeated observations and Fisher's exact test. Paired-sample t-test was applied for analyzing the mean scores of CES-D, POMS, and PSQI questionnaires within a group post and pre-intervention. In this study, the Statistical Package for Social Sciences (SPSS) version 20 software was used and P-value less than 0.05 was considered significant.