Abstract
INTRODUCTION: Gastrointestinal malignancies account for 25% of all cancer cases and 35% of cancer-related mortality. Next-generation sequencing (NGS) can elucidate the genomic landscape of gastrointestinal cancers; tissue-based genotyping has traditionally been used, but liquid biopsy-based genotyping is a noninvasive alternative. Moreover, geographical variations in the genomic landscape of gastrointestinal cancers have not been fully elucidated. This retrospective study aimed to gain insight into the genomic landscape of patients with gastrointestinal cancers from the Asia and Middle East (AME) region using plasma-derived circulating tumor DNA (ctDNA). METHODS: From routine clinical practice, 2,601 plasma samples were collected from 2,062 patients with gastrointestinal cancers in the AME region. NGS profiling was conducted using the Guardant360® assay. The frequency of biomarkers that can aid decision-making in cancer patients was investigated. RESULTS: Single-nucleotide variants affected most commonly TP53 (70.4%), KRAS (44.0%), APC (25.7%), ATM (15.1%), and PIK3CA (12.3%). Copy number alterations were most often observed in EGFR (13.7%), CCNE1 (5.9%), PIK3CA (5.0%), MYC (4.7%), and FGFR1 (4.6%); fusions were detected in 1.6% of patients and most frequently affected FGFR2, RET, ALK, FGFR3, and NTRK1/3. In patients with pancreatic adenocarcinoma, the most frequently observed clinically informative genomic biomarkers occurred in KRAS (G12C, 1.6%; all others, 67.1%), BRCA1/2 (4.1%), BRAF (V600X, 1.5%), and microsatellite instability-high (MSI-H) (1.0%). In patients with colorectal cancer, the most common clinically relevant alterations were KRAS (49.0%), BRAF (V600E, 7.6%), and NRAS (5.7%) mutations; ERBB2 amplifications (2.5%); and MSI-H (1.8%). In patients with biliary tract cancers, actionable alterations included IDH1 mutations (11.1%), ERBB2 amplifications (4.6%), FGFR2 fusions (2.0%), MSI-H (2.0%), and BRAF V600E (1.5%). In patients with gastric or gastroesophageal junction adenocarcinomas, actionable alterations included ERBB2 amplifications (10.1%) and MSI-H (3.6%). CONCLUSION: Our data provide insight into the genomic landscape of patients with gastrointestinal cancers from the AME region using ctDNA analysis. These findings highlight the potential utility of liquid biopsy as a noninvasive tool for characterizing tumor genomic profiles and support its role in clinical practice.