Gut microbiome and serum cytokines as biomarkers for pathological complete response (pCR) and recurrence free survival in early locally advanced triple negative breast cancer (LA TNBC) receiving neoadjuvant chemotherapy (NA) and immune checkpoint blocker (ICI): Updated analysis from a pilot study.

Abstract

e12630 Background: Microbial metabolic pathways are known to influence immunotherapy resistance and these may similarly contribute to resistance in LA TNBC treated with NA ICIs. We previously reported differential microbial response predictors of treatment in this population. In this analysis, we are presenting updated data demonstrating differences in the expanded profile of microbiome, metabolites and cytokines. Methods: 35 patients who received NA pembrolizumab and chemotherapy for LA TNBC were identified at Moffitt Cancer Center and Tampa General Hospital between February 2022 to October 2023. Blood and stool samples were collected at initiation and completion of NA therapy and/or when the patients developed grade 3 or higher toxicity to therapy. The goal was to study the gut microbiome, composition and metabolites associated with pCR We conducted shotgun metagenomic sequencing on 59 stool samples, GC-MS sequencing in 59 stool and 56 plasma for metabolites and 25 serum cytokines. Patient characteristics were as follows: 26 Caucasian, 4 African-American, 2 Hispanic, 3 unknown, median age 52 years. 23 patients had stage II, while 12 patients had stage III, 17 patients were node negative while 18 were node positive disease. Results: Response was assessed in 34 patients, of which 12 had pCR (ypT0/is ypN0) while 22 had residual disease (RD), 5 (14%) patients had recurrence. At median follow up of 20.38 months, the percentage survival was 87.3% (95% CI: 69.3-95.1). One patient experienced disease progression while on NA therapy, and 6 developed grade 3-4 immune-related adverse events, resulting in treatment discontinuation, with one patient succumbing to these complications. Butyric acid is a known microbial metabolite associated with immunosuppressive effects. However, its role in TNBC role has not been explored. We report that lower butyric acid is associated with a significantly higher likelihood of pCR in LA TNBC,HR 0.9, p < 0.05. Meanwhile amino acid metabolites tyrosine and phenylalanine are associated with better pCR with HR 3.99 and HR 1.2, p < 0.05 respectively. No difference in alpha diversity at baseline or follow-up comparing patient pCR vs RD. Beta diversity at follow-up was statistically different between patients with pCR vs RD while no difference noted at baseline. Higher levels of CXCL5, CCL2 were noted in patients with RD. Conclusions: To our knowledge this is the first analysis in patients with LA TNBC treated with NA ICB showing differential immunosuppressive microbial metabolites, particularly Short Chain Fatty Acids which are associated with pCR. Biomarkers play a vital role in identifying patients who are most likely to respond to immune checkpoint blockade (ICB). The gut microbiome may act as a predictive biomarker for neoadjuvant ICB response in TNBC, necessitating validation in larger studies.