Constructing genetic-immune prognostic subtypes of primary duodenal adenocarcinoma through whole exome sequencing and AI-assisted immune microenvironment analysis.

Abstract

4197 Background: Primary duodenal adenocarcinoma (DA) is a rare gastrointestinal tumor with a poor prognosis. This study aimed to investigate the tumor immune microenvironment (TIME) and genetic landscape of DA to depict the unique DA subtypes, distinguishing its prognosis and identifying potential therapeutic targets. Methods: To reveal the heterogeneity of DA’s genetic landscape and TIME, 88 treatment-naïve DA tumor samples were analyzed via multiplex immunofluorescence (mIF) staining, whole-exome sequencing, and RNA sequencing. The quantity of infiltrating cells, tertiary lymphoid structure, and spatial analysis was conducted using an automated platform (APTIME), an artificial intelligence-based analysis tool for analyzing pathology images from formalin-fixed paraffin-embedded slides stained with mIF. Results: Significant heterogeneity was observed in the genetic and immune landscapes of DA. A genetic-immune classification was established, identifying four distinct subtypes: MSI (microsatellite instable), InA(inflamed active), ML (macrophage-low), and MH (macrophage-high). The InA subtype exhibited high levels of infiltrating immune cells, while the MH subtype, characterized by enriched tumor-associated macrophages, was associated with the worst overall survival. The MH subtype frequently harbored TGF-β pathway mutations, particularly in SMAD4, while the ML subtype showed alterations that predominantly in the SWI/SNF pathway, specifically in ARID2. Spatial analysis indicated that closer proximity between macrophages and both tumor cells and T cells correlated with worse prognosis in DA patients. Closer interactions between PD-L1 and PD1+ T cells in the MH subtype suggested that PD-1/PD-L1 interactions contributed to an immunosuppressive tumor microenvironment. Conclusions: This study enhances the understanding of DA's molecular characteristics, particularly through the identification of a novel genetic-immune subtype, and provides a foundation for developing precision treatment strategies for this malignancy.