Background: Esophageal adenocarcinoma (EA) arises from Barrett’s epithelium (BE), and chronic gastroesophageal reflux disease is considered the strongest risk factor for disease progression. All BE patients undergo acid suppressive therapy, surveillance, and BE removal by surgery or endoscopic ablation, yet the incidence of EAC continues to increase. Despite the known side effects and mortality, the one-size-fits-all approach is the standard clinical management as there are no reliable methods for risk stratification.

Methods: Paired-end Illumina NextSeq500 RNA sequencing was performed on total RNA extracted from 20-week intervals (0, 20, 40, and 60 W) of an in vitro BE carcinogenesis (BEC) model to construct time series global gene expression patterns (GEPs). The cells from two strategic time points (20 and 40 W) based on the GEPs were grown for another 20 weeks, with and without further acid and bile salt (ABS) stimulation, and the recurrent neoplastic cell phenotypes were evaluated.

Results: Hierarchical clustering of 866 genes with ≥ twofold change in transcript levels across the four time points revealed maximum variation between the BEC20W and BEC40W cells. Enrichment analysis confirmed that the genes altered ≥ twofold during this window period associated with carcinogenesis and malignancy. Intriguingly, the BEC20W cells required further ABS exposure to gain neoplastic changes, but the BEC40W cells progressed to malignant transformation after 20 weeks even in the absence of additional ABS.

Discussion: The transcriptomic gene expression patterns in the BEC model demonstrate evidence of a clear threshold in the progression of BE to malignancy. Catastrophic transcriptomic changes during a window period culminate in the commitment of the BE cells to a “point of no return,” and removal of ABS is not effective in preventing their malignant transformation. Discerning this “point of no return” during BE surveillance by tracking the GEPs has the potential to evaluate risk of BE progression and enable personalized clinical management.

背景：食管腺癌 (EA) 起源于巴雷特上皮 (BE)，慢性胃食管反流病被认为是疾病进展的最强风险因素。所有 BE 患者均接受抑酸治疗、监测和通过手术或内镜消融去除 BE，但 EAC 的发生率继续增加。尽管存在已知的副作用和死亡率，但一刀切的方法是标准的临床管理方法，因为没有可靠的风险分层方法。

方法： 对从 20 周间隔（0、20、40 和 60 W）提取的总 RNA 进行双端 Illumina NextSeq500 RNA 测序。 体外用于构建时间序列全局基因表达模式 (GEP) 的 BE 致癌 (BEC) 模型。来自基于 GEP 的两个战略时间点（20 和 40 W）的细胞再生长 20 周，有或没有进一步的酸和胆汁盐 (ABS) 刺激，并评估复发性肿瘤细胞表型。

结果：在四个时间点转录水平变化≥两倍的 866 个基因的分层聚类显示 BEC20W 和 BEC40W 细胞之间的差异最大。富集分析证实，在与致癌和恶性肿瘤相关的这个窗口期内，基因发生了≥两倍的改变。有趣的是，BEC20W 细胞需要进一步暴露于 ABS 以获得肿瘤性变化，但即使在没有额外 ABS 的情况下，BEC40W 细胞也会在 20 周后进展为恶性转化。

讨论：BEC 模型中的转录组基因表达模式证明了 BE 进展为恶性肿瘤的明确阈值。窗口期的灾难性转录组学变化最终导致 BE 细胞进入“不归路”，并且去除 ABS 不能有效防止它们的恶性转化。通过跟踪 GEP 在 BE 监测过程中发现这个“不归路”，有可能评估 BE 进展的风险并实现个性化的临床管理。