Abstract:Abstract Background: The incidence of early onset (EO) gastrointestinal (GI) cancers prior to age 50 is rising, and the etiology is unknown. The gut microbiome may contribute to GI cancer pathogenesis, though how bacteria drive metastasis and treatment resistance is not known. Purpose: To define the microbiome contribution to EO-GI cancers and metastasis by analyzing longitudinal samples from previously untreated patients with GI cancers. Methods: We designed a prospective biospecimen collection platform. We selected patients with newly diagnosed, previously untreated colorectal cancer (CRC) and esophagogastric cancer (EGC). We collected stool, biopsy or surgical tissue, and peripheral blood mononuclear cells (PBMCs) at baseline prior to treatment and at each restaging scan for patients with metastatic disease. For those with locally advanced disease, we collected samples at baseline and after each treatment phase (surgery, radiation, chemotherapy). Stool samples were analyzed using shotgun sequencing. Tissue samples were banked for further analysis. Clinical data was manually curated. Relative abundances of bacteria at the species level were compared between groups. Alpha diversity was calculated using the inverse Simpson index and compared between groups using the Wilcoxon signed-rank test. Beta diversity was analyzed using the Bray-Curtis dissimilarity matrix and compared using PERMANOVA. Multivariate association between species abundance and clinical covariates was performed using MaAsLin2 R package. Results: We analyzed a total of 150 stool samples from 76 patients (colorectal n=53, esophagogastric n=23), including up to 6 samples from a single patient over time. Mean alpha diversity did not differ significantly by primary site or age at diagnosis in patients with CRC but was higher in stage IV compared with stage I disease (P=0.054). Beta diversity was significantly different between samples from patients with right- compared with left-sided CRC (P=0.005) but did not differ significantly by diagnosis age (<50 vs. > 50). Among those with EGC, mean alpha diversity was not significantly different in samples from patients with esophagus or gastroesophageal junction (proximal) tumors compared with gastric (distal) tumors and did not differ significantly by age group. Beta diversity was significantly different between patients with EGC over 50 compared with those under 50 (P=0.002). When CRC baseline samples were examined, several bacterial species were associated with age <50 at diagnosis, including Streptococcus anginosus group (P=0.001), Solobacterium moorei (P=0.013), and Firmicutes bacterium CAG83 (P=0.016). Conclusions/Future Directions: Microbiome composition may cluster by primary tumor site and age at diagnosis in patients with previously untreated GI cancers. Functional analysis and is ongoing and will be presented. Citation Format: Melissa A. Lumish, Asha R. Saxena, Nicholas Waters, Anqi Dai, Saskia Hartner, Teng Fei, Matthew Drescher, Jonathan Bermeo, Dorina Ismailgeci, Maggie Fox, Yelena Y. Janjigian, Luis A. Diaz, Martin R. Weiser, Jonathan Peled, Marcel van den Brink, Karuna Ganesh. Prospective platform to define microbiome correlates of metastasis and therapy resistance in early onset and average onset gastrointestinal cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5910.

Abstract 5910: Prospective Platform to Define Microbiome Correlates of Metastasis and Therapy Resistance in Early Onset and Average Onset Gastrointestinal Cancers