Trial of Diet, Physical Activity and Breast Cancer Recurrences: the DIANA-5 Study (DIANA-5)

Aims The investigators have been proposing a randomized intervention trial of diet and physical activity to reduce BC relapse (local, distant), and second ipsilateral or contralateral BC risk, in BC patients at high metabolic-endocrine of recurrence risk.

Secondary aims:

• to evaluate the effect of the combined dietary and physical activity change on the prevalence of Metabolic Syndrome
• to evaluate the effect of the combined dietary and physical activity change on the prevalence on risk biomarkers (BMI > 24.9 kg/m^2)
• to evaluate the effect of the combined dietary and physical activity change on the prevalence on risk biomarkers (testosterone >= 0.4 ng/ml)
• to evaluate the effect of the combined dietary and physical activity change on the prevalence on risk biomarkers (insulin > pmol/l)
• to evaluate the effect of the intervention on the development of other life-style related health conditions (other cancers, diabetes, hypertension, dyslipidemia, total mortality)

Recruitment:

Potential participants have been recruited at the time of diagnosis through follow-up clinics, patients' associations, or the media. Patients have been requested to sign an informed consent, including authorization for getting blood samples, for storing samples for future studies. The investigators collected demographic information, fasting measure body weight, height, waist circumference, and blood pressure, blood glucose, LDL and HDL cholesterol, triglycerides, testosterone and insulin. Eligible patients have filled in a baseline questionnaire on medical history, medication use, reproductive and body weight history, usual physical activity, and a validated semi-quantitative food frequency questionnaire.

Biological bank: the investigators collected and stored at -80°C one aliquot of whole blood, three 2-ml serum aliquot and two 2-ml buffy coat aliquot.

Randomization:

The investigators recruited 1542 patients' high risk and randomized the participants in two groups: 773 (control group) have received general standard written recommendation for healthy lifestyle without, however, any active support; 770 (intervention group) have received a combination of individual and group contacts over the course of one year, including kitchen courses, gym and dance classes, common meals and reinforcing meetings, with emphasis on a comprehensive dietary and life-style change.

Randomization has been carried out within strata of age, treatment (no adjuvant therapy, chemotherapy only, hormonal therapy only, both hormonal and chemotherapy) and axillary nodal status at diagnosis.

Patients who did not satisfy the high-risk criteria have been given the same general recommendations as the control group, and have been followed up as an external low risk group.

Intervention:

The intervention wanted aim to increase physical activity, controlling weight, and promoting healthy diet.

1. Physical activity: achieve and maintain regular participation in a moderate intensity physical activity program of 210 minutes/week (30 min on average per day) over at least 3 days /week; decrease sedentary behaviors by 30 minutes/day on at least 5 days/week. During the first 12 months one group physical activity session per month offered to enhance program adoption. Women who wanted to take up vigorous sports encouraged to do so. For those who do not progressed to more vigorous activity, the focus was on maintaining moderate activities, such as walking. For self-monitoring and compliance enhancement, study participants had use logs, fill-in questionnaires and armband to monitor physical activity.
2. Weight control: reducing energy intake relative to expenditure was the primary dietary focus for promoting weight loss in overweight or obese participants, while maintaining a healthy energy balance was the primary focus for normal weight participants. Participants were encouraged to include whole grains, pulses and high-fiber vegetables.
3. Healthy diet: reducing calorie intake, through the preferred consumptions of highly satiating foods, such as unrefined cereals, legumes and vegetables, reducing high glycemic index food, high insulinemic foods, saturated fat, preferring instead unrefined vegetable fats, such as extra virgin olive oil, nuts and oleaginous seeds, reducing protein intake, mainly animal protein (except fish).

Baseline and yearly measurements:

Height and weight: electronic scale with women in light clothes and without shoes. In a sample of cases body fat and lean mass will be measured with bioelectric impedance (BIA).

Blood pressure: electronic device. Serum glucose, triglycerides and cholesterol: standard quality-controlled laboratory techniques.

Serum Testosterone and Insulin: Radioimmunoassay (RIA). For testosterone the investigators have used a direct RIA kit which has been validated by comparison with indirect assay after organic extraction of serum samples and celite purification.

Compliance:

Compliance has been monitored through lifestyle questionnaires, dietary questionnaires, weight change and plasma glucose, triglycerides, cholesterol, insulin, testosterone changes after one year of intervention. A compliance score has been computed, based on the direction of change in all these biomarkers.

Follow-up and outcome events:

The main outcome is the new BC events including:

• new primary breast cancer
• local/regional recurrence
• distant/metastatic recurrence The follow up have been based on the routine clinical follow-up at the collaborating hospitals, the periodic questionnaires to study participants, the regional cancer registry and hospital discharge diagnosis system, and death certificates.

Statistical power:

The survival of BC patients in Italy is dramatically increasing: 5-year relative survival increased from 80.6% for patients diagnosed in the early 1990s to 85.6% for patients diagnosed in the early 2000s. For these patients the estimated relapse free survival is 81.9. Such a survival improvement is accompanied by a postponement of the incidence of relapse.

The original aim was to recruit 1,200 high risk patients in order to have 600 patients per arm, which shall guarantee 80% chance of getting a significant difference also with 25% reduction, and allowing for 90% compliance in the intervention group and 10% contamination of the control group.

Statistical analysis:

Nowadays the investigators are assessing the baseline association between androgens, insulin, and several measures of adiposity including body mass index (BMI), waist circumference and percentage body fat, through Spearman correlation coefficients. Furthermore, the investigators are comparing the change in food consumption, body weight, and geometric means of hormones end points from baseline to 12 months in the intervention and control group.

The main analysis of the intervention effect on the incidence of recurrences is by intention to treat, i.e. based on assigned treatment at the time of randomization, regardless of adherence. Later, as a secondary analysis the investigators shall assess the effect by change in body weight and biomarkers. The investigators shall compute total survival and disease-free survival.

Hazard ratios and confidence intervals will be computed by the Cox proportional hazard model, with standard clinic-pathological prognostic covariates as potential confounders. Separate analyses will be carried out by Estrogen Receptors, Progesterone Receptors and erbB2 status. The investigators decided to carry out another follow-up in 2022 as the relapse time has lengthened to 10 years, due to new therapies, and to evaluate the effects on lifestyle due to closures during the pandemia severe acute respiratory syndrome-CoV-2.

Feasibility:

The principal investigator (at first dr. Franco Berrino and later dr. Anna Villarini) has long term experience in carrying on and coordinating epidemiological studies, including small-scale dietary intervention studies. The Milano National Cancer Institute and the others collaborating center is fully equipped with facilities for kitchen courses and gym facilities.

Control of potential biases:

Performance bias-patients have been recruited after the main treatments have been planned or delivered. In no case the allocation to the intervention or control group or the compliance will affect treatment.

Measurement bias-patients and researchers cannot be blinded, but the ascertainment of outcome was carried out by clinicians that were not involved in the study.

Attrition bias-Based on previous studies the investigators expect a fairly high compliance in the intervention group (>90%) but also some modification in the control group. After the main analysis by intention to treat, therefore, secondary analyses will be carried out by compliance score.