This is a randomized, prospective study with a small group of 30 patients (pilot study). Once participants are enrolled, they will be randomized in a 2:1 ratio of either the nutritional intervention (personalized counseling on increasing fiber intake and maintaining adequate caloric intake during breast cancer treatment) or an educational handout on increasing fiber intake. Randomization will be concealed using a random number generator. Investigators and healthcare providers will not be blinded due to the nature of the intervention itself, but all documentation related to the personalized nutritional intervention will not be part of the participant's electronic medical record. Co-investigators completing the data analysis will be blinded to participants' group assignments.
Consent will be obtained upon enrollment. Variables to be collected as part of the baseline demographics questionnaire include: age, gender, marital status, education level, employment status, annual income, residence (ZIP code), BMI, medications and dose, TNBC stage at diagnosis, and treatment type(s).
Participants will complete: interim medication updates including antibiotic use (Interim Survey), gastrointestinal symptoms using the PROMIS GI symptom assessment, dietary intake (NCI-DSQ) with two 24-hour food logs, performance status (FACT-G Scale), anxiety (GAD-7), and depression (PHQ-9), and physical activity (RAPA). If participants report moderate-severe anxiety and/or depression through the GAD-7 and/or PHQ-9 surveys, respectively, they will be notified by telephone or email to contact their primary care provider or oncologist to seek prompt evaluation. Stool samples will be collected at baseline (pre-treatment), 6-week, 12-week, 18-week, and 24-week timepoints. The 6-week interval period was selected based on expected treatment cycles with dose-dense doxorubicin/cyclophosphamide and Taxol +/- pembrolizumab. This interval was also selected based on the minimum amount of time expected for potential changes in dietary intake and symptoms during treatment. Survey reminders will be sent via the app every 6 weeks and data will be exported to RedCap.
The nutritional intervention for the treatment group -- personalized counseling on increasing fiber intake and maintaining adequate caloric intake during treatment -- will be administered as (1) a 60-minute initial telehealth consultation within the first week of study enrollment, and (2) up to two 30-minute follow-ups throughout the study, ideally the first follow-up within 6 weeks of study enrollment. These sessions will be led by a registered dietitian using cultural awareness and symptom assessment. Session notes will be included in the participants' medical records. Control participants' intervention will be an educational handout on increasing fiber intake. San Diego State University (SDSU) masters students in Exercise and Nutritional Science will assist all participants with completing two 24-hour ASA24 food logs at each timepoint along with other surveys. Nutritional composition reported in 24-hour logs will be analyzed at SDSU using ASA24.
Participants will collect stool samples using the Zymo Research kit, which has previously been shown to be equally efficacious as a larger stool sample (scoop),10 mailed to their home. This kit includes a Fecal Collection Tube, a Feces Catcher, a Biohazard Bag, Gloves, and multi-language instructions. To encourage participation and retention, participants will be given a $40 gift certificate per stool sample with up to $200 total compensation. Participants will mail stool samples to the Scripps Biorepository and processed for storage at -80 degrees C. Once all samples are collected, they will be sent to the Scripps Genomics Core for DNA extraction and 16S sequencing. Sequencing data analysis by the CCBB will include taxonomic classification, abundance tables, diversity analysis (alpha and beta indices), and principal component analysis (PCA) plots. Subset analyses will compare participants with high/low fiber consumption, high/low caloric intake, overweight/not overweight by BMI, and cancer-directed treatment (chemotherapy alone or with immunotherapy).
Survey data will be imported into GraphPad Prism for visualization and analyses. For each survey item collected at all 5 time points, a mixed effects model will be used where the survey item result is the dependent variable, group (control or nutritional intervention) and time point are fixed effects, and participant is treated as a random effect. The mixed effects model uses restricted maximum likelihood estimation to account for missing values. Residuals will be examined to ensure approximation of normality is met. P-values will be reported for each fixed effect (group and time) and the interaction between these effects, reported as time x group.
For TME analysis, up to 3 participants with high-fiber dietary intake who experience pCR and 3 participants with residual disease (i.e. up to 6 total) will be selected based on retrospective chart review, and biopsy samples pre- and post-treatment will be identified through the Department of Pathology. Samples will be further processed and mounted onto slides, each with 1 participant with matched pre- and post-neoadjuvant treatment samples per slide for GeoMx spatial transcriptional profiling of 96 regions of interest (ROIs) through the Scripps Genomics Core. The CCBB will perform statistical analysis to identify signaling pathways and immune cell composition in TNBC.
At the end of the study, participants will be offered participation in a one-on-one session via video conferencing platform (such as Zoom) to provide an opportunity for participants to share in their own words their experiences in the study and in using the survey app. Specifically, participants will be asked to discuss the utility of educational materials provided, ease of navigating the study app and participating in the study, and other feedback that will help medical oncologists and dietitians continuously improve patients' experience during neoadjuvant treatment. Participants who uninstall the CareEvolution app or have not engaged with the app for over 60 days will be disenrolled from the study.
