Identification of an individualized immune-related miRNA pair signature for survival prediction of neoadjuvant chemoradiotherapy in esophageal squamous cell cancer.

Background: Immune-related microRNAs (ImiRNAs) have emerged as potential biomarkers owing to their universality and tissue-specificity. Although neoadjuvant chemoradiotherapy (NCRT) has been incorporated into standardized guidelines for esophageal squamous cell carcinoma (ESCC), its efficacy is hindered by tumor and individual heterogeneity. This study aimed to develop a specific ImiRNA pair signature to predict the long-term survival benefits of NCRT for ESCCs.

Methods: Endoscopic biopsies of 215 ESCCs were collected from three centers. ImiRNAs were identified via miRNA sequencing and immune gene annotations. A pairwise-comparison method and LASSO-Cox regression constructed prognostic models, validated in two cohorts (n = 113). Immunohistochemistry was employed to gain insights into the real tumor immune microenvironment after treatment.

Results: The ESCC NCRT immune-related miRNA pair signature (EN-ImiRPS), comprising six miRNA pairs, demonstrated robust prognostic accuracy for both overall survival (OS; HR: 9.76; P < 0.001) and recurrence-free survival (RFS; HR: 5.31; P < 0.001) in the training cohort. This performance was consistently validated in both internal and external cohorts. Furthermore, the EN-ImiRPS score outperformed pathological complete response in predicting OS and RFS. Notably, an individual miRNA expression-based model was significantly less predictive than the pair-based approach. Low-risk patients, as defined by EN-ImiRPS, exhibited significantly increased CD8 T cell infiltration (P < 0.05), indicative of enhanced anti-tumor immunity within the tumor microenvironment.

Conclusions: We established a multicenter, ImiRNA pair-based signature that accurately predicts long-term survival in ESCC patients receiving NCRT. The model's stability across cohorts and its association with immune microenvironment features highlight its potential for guiding personalized therapy. This study also identifies key ImiRNAs for further mechanistic exploration of tumor-immune interactions.