Dados do Resumo

Título

Transcriptomic Analysis of Lauren's Histological Subtypes in Gastric Cancer

Introdução

Gastric cancer (GC) is a major global cause of cancer-related deaths. Lauren’s classification divides GC into intestinal and diffuse types. The intestinal type has a known progression model, aiding early detection and prevention. In contrast, the diffuse type lacks defined precursor lesions and is less comprehensively understood.Advances in Next Generation Sequencing offer the potential to uncover gene expression profiles that could serve as biomarkers for GC management.

Objetivo

The aim was to analyze transcriptomic profiles of gastric cancer based on Lauren classification to characterize molecular differences between intestinal and diffuse subtypes.

Métodos

A cohort of 43 patients with intestinal subtype gastric cancer and 21 with diffuse subtype was included, with paired tumor and non-tumor samples collected for RNA sequencing (RNA-seq). The study was approved by the Research Ethics Committee (CAAE 47580121.9.0000.5634). RNA-seq was performed on the NextSeq® platform using the NextSeq® 500 MID Output V2 kit. In the dry-lab pipeline, initial processing involved demultiplexing of raw sequencing reads and quality-filtered with Fastp (v0.23.2). Taxonomic classification was conducted with Kraken 2 (v2.1.3) against a comprehensive database. Alignment was performed using Salmon with Gencode version 43. Differential expression analyses were conducted with DESeq2, using |Log2(Fold-Change)| > 1 and adjusted p-value < 0.05. Receiver operating characteristic (ROC) analysis and discriminant analysis of principal components (DAPC) were performed. Gene Set Enrichment Analysis (GSEA) with Reactome sets was conducted using the Fgsea package.

Resultados

Differential expression analysis identified 4,732 mRNAs in intestinal tumors and only 376 lncRNAs in diffuse tumors, suggesting diffuse tumors resemble adjacent non-tumoral tissue. Comparing intestinal and diffuse tumors, 693 genes showed significant differences, while only 2 genes differed in adjacent non-tumoral samples. DAPC showed distinct molecular profiles, with adjacent non-tumoral samples showing greater homogeneity. To refine transcriptomic profiling, 33 high-sensitivity and specificity genes (AUC > 0.75) were selected for subtype classification, with clustering supporting clear separation. GSEA highlighted enrichment in cell cycle, checkpoint, apoptosis, and viral infection pathways, which are crucial for gastric carcinogenesis. Additionally, microbiota analysis found higher abundance (p-adj <0.01) of Faecalibacterium and Flavonifractor in diffuse tumors, suggesting potential microbiota-cancer interactions.

Conclusões

Despite the need for further validation, specific expression profiles were identified between gastric cancer subtypes. The presence of bacteria like Faecalibacterium and Flavonifractor in the diffuse type suggests a potential role in disease progression. This study highlights molecular features that could help to guide future investigations aiming to establish personalized management of GC.

Financiador do resumo

FAPESPA - Fundação Amazônia de Amparo a Estudos e Pesquisas (convênio 004/2021)
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Palavras Chave

gastric cancer; Transcriptomics; biomarkers