A large multicenter study combining clinical, epidemiological, and molecular data reports that esophageal adenocarcinoma (EAC) appears to arise through a shared biological pathway linked to Barrett’s esophagus (BE), even when BE is not identified at diagnosis.
The analysis included 3,100 patients with EAC from 25 UK centers, with detailed clinical annotation and genomic data available for a substantial subset. The investigators aimed to clarify whether EAC can develop independently of BE or whether BE-related changes are present in all cases.
Although BE is considered the main precursor lesion, it is not detected in approximately half of patients with EAC at presentation. In this cohort, patients with and without detectable BE showed similar demographic and risk factor profiles, including age, sex, smoking status, and obesity.
Genomic analyses showed substantial overlap between BE-positive and BE-negative tumors. Key driver mutations, including alterations in CDKN2A, TP53, and ARID1A, were present in both groups. Overall mutational burden, mutational signatures, and large-scale genomic changes were also comparable, indicating similar molecular processes underlying tumor development.
As shown in the genomic comparisons, BE-associated molecular features were retained in both tumor types, with few meaningful differences after statistical correction.
Further analysis using multiregional sequencing demonstrated similar evolutionary patterns in both groups. Early driver mutations – particularly in TP53 – were observed across tumors regardless of BE status, and measures of tumor heterogeneity were comparable.
Importantly, spatial transcriptomic and protein analyses identified markers of intestinal metaplasia within tumor tissue in both groups. Markers such as TFF3 and REG4, typically associated with BE, were detected even in tumors without histologically identifiable BE. The spatial data show overlap between BE-related gene expression and tumor regions, suggesting persistence of precursor biology.
The only consistent clinical difference was tumor stage: cases without detectable BE were more likely to present at a more advanced stage.
Overall, the findings indicate that EAC may develop through a common pathway involving intestinal metaplasia, even when BE is not visible on endoscopy or histopathology. This study highlights that absence of detectable BE does not exclude underlying precursor changes. Molecular and biomarker-based approaches may therefore support earlier detection and risk stratification where morphology alone is insufficient.
