An exploratory study in potential biomarkers discovery for LNM risk assessment in Chinese patients with esophageal squamous cell carcinoma at pT1a (mucosal) stage

Background Esophageal cancer (EC) is a major global malignant neoplasm, with China accounting for over 40% of cases. Esophageal squamous cell carcinoma (ESCC), the predominant subtype, continues to show poor long-term survival. Endoscopic resection has become a valuable, minimally invasive strategy for managing pT1a (mucosal) ESCC. However, accurate assessment of lymph node metastasis (LNM), a key determinant of treatment suitability, remains challenging because the current clinical imaging techniques have limited sensitivity for detecting micrometastatic nodal disease. Therefore, no reliable consensus criteria are currently available. This study is hypothesis-generating rather than confirmatory, given the limited sample size (n = 9). Objectives To identify potential biomarkers for LNM risk assessment in pT1a (mucosal) ESCC. Methods Clinical specimens from Chinese patients with pT1a (mucosal) ESCC (n = 9, LNM+ n = 5, LNM- n = 4) were subjected to transcriptomic profiling. All samples were processed in a single batch without batch-effect normalization. RNA integrity was assessed by RIN (RNA Integrity Number) with a threshold of ≥7.0, and each sample generated at least 10 Gb of data (≥32 million reads/sample; mean reads per sample and range are reported in Supplemental Table 2). Molecular features associated with LNM were systematically characterized through comprehensive transcriptomic analysis. Key findings were validated using quantitative PCR on the same set of clinical specimens. Publicly available data from The Cancer Genome Atlas (TCGA), which predominantly included T1/T2 stage Caucasian patients and did not fully match the Chinese training cohort, were incorporated to evaluate predictive performance. Results Two significantly dysregulated signaling pathways were identified, and five major expression changes (TPTEP1, ERAP2, ADAMTS16, CST4, and PRSS8) associated with LNM were discovered in the transcriptomic analysis and confirmed by qPCR in the same samples. Based on these biomarkers, a preliminary predictive model was developed to estimate the risk of metastasis in early-stage ESCC. Although the model demonstrated promising predictive capacity, further optimization is still needed with larger, independent datasets. Conclusions This study enhances the current understanding of the molecular mechanisms underlying LNM in early-stage ESCC and identifies candidate biomarkers with potential utility for metastatic risk assessment. These findings are exploratory and require validation in larger, independent cohorts, particularly within the Chinese population. These findings may support the development of more precise and personalized therapeutic strategies for patients with early-stage ESCC.