. 2019 Sep 17;9(9):e030502. doi: 10.1136/bmjopen-2019-030502

Table 2.

Variables, measures and methods of analysis

Variable/Outcome	Outcome measure	Method	Statistical analysis
Primary
Efficacy of supplemental DHA provided with standard neoadjuvant chemotherapy as measured by change in Ki67.	Ki67 labelling index.	Immunohistochemistry.	95% t-CI for mean per cent change in Ki67. Independent t-test to compare change between the study groups.³⁶
Secondary
1. DHA incorporation into plasma phospholipids.	Fatty acid composition of plasma phospholipids.	Gas chromatography.	Paired t-test will be used to compare the mean per cent change in the DHA level of patients after each cycle with their baseline values. If the data are not normally distributed, the Wilcoxon signed-rank test will be employed for this comparison. A 95% t-CI for the mean per cent change in the DHA from baseline will be compared with patients receiving placebo.
2. Systemic immune function.	Immune cell subset identification. Plasma cytokines. Ex vivo stimulated immune cell response.	Flow cytometry. ELISA. Meso Scale.	Repeated-measures analysis of variance with post-hoc analysis.
3. Identify factors that may affect DHA incorporation into tumour tissue and plasma phospholipids.	Factors assessed after calculating high and low DHA incorporators: Weight (body mass index-BMI). Age. Usual diet estimated from the FFQ. Composition of dietary fat estimated from the FFQ. Histology of the tumour (provided from the biopsy). Amount of DHA consumed (adherence to the supplement). % incorporation of other fatty acids.		Independent t-test will be conducted to compare the mean values between the two study groups. χ² test will be conducted to determine correlation between two categorical variables for outcome measures listed.
4. Examine changes in markers for apoptosis.	Caspase-3.	Immunohistochemistry.	Within-subject and between-subject variability between the two groups will be tested using GEE method.
5. Examine changes in markers for tumour infiltrating lymphocytes.	CD4+/CD8+.	Immunohistochemistry.	Within-subject and between-subject variability between the two groups will be tested using GEE method.
6. Describe the rate of pCR in breast and in axillary nodes.	Absence of invasive cancer on H&E evaluation.	Immunohistochemistry.	pCR=ypT0/is ypN0. 95% t-CI using independent t-test for mean per cent change between treatment groups.
7. Describe the rate of grade 3 and 4 chemotherapy-associated toxicities.	Rate of grade 3/4 toxicities and chemotherapy-associated hospitalisations.	Chart review.	95% t-CI using independent t-test for mean per cent change in events between treatment groups.
Exploratory outcomes
1. FFQ.	Diet History Questionnaire II	Questionnaire.	Independent t-test of macronutrient and fat content/composition between groups.
2. Quality of life.	Baseline and endpoint questionnaires.	Questionnaire.	Paired t-test for continuous variables and McNemar’s for categorical variables for mean per cent change in events between treatment groups.
3. Exercise.	Godin Leisure-Time Exercise Questionnaire.	Questionnaire.	Paired t-test for continuous variables and McNemar’s for categorical variables for mean per cent change in events between treatment groups.
4. Assess the rate of breast conservation.	Rate of lumpectomy and mastectomy.	Chart review.	χ² tests.
5. Assess the volume of surgical blood loss.	Review surgical reports for quantitative/qualitative loss of blood.	Chart review.	Independent t-test.
6. Analyse local control, relapse-free survival and overall survival.	Electronic medical record and/or paper medical chart review at 3, 5 and 10 years to explore possible effects on long-term outcome.	Chart review.	Kaplan-Meier estimates along with the survival curves, log-rank test will be used for statistical comparison between groups.

DHA, docosahexaenoic acid; FFQ, Food Frequency Questionnaire; GEE, generalised estimating equation; pCR, pathological complete response.