The value and application of a poorly differentiated cluster-based Ezrin marginal score system in the prognostic assessment of colorectal cancer.

Aim: This study aimed to explore the distribution of Ezrin-positive cells in poorly differentiated cluster (PDC) of colorectal cancer (CRC), and its correlation with both traditional and auxiliary pathological indicators, including E-cadherin and tumor-stroma ratio (TSR), as well as patient prognosis.

Methods: We selected 59 patients with stage I-III who underwent radical surgery to observe the expression of CRC under hematoxylin and eosin (H&E) and immunohistochemistry (IHC) staining. This study establishes the poorly differentiated cluster-based Ezrin marginal score (PDC-EMS), which quantifies the tendency of Ezrin-positive cells to accumulate at the margins of PDC, to evaluate the validity and superiority of this assessment method.

Results: The results indicate that PDC grading is significantly correlated with the TNM stage, pN stage, macroscopic configuration, and tumor budding (TB) grading. The PDC-EMS shows significant correlations with PDC and TB grading and demonstrates good consistency with traditional pathological indicators. E-cadherin expression significantly negatively correlates with TNM staging and markedly associates with pN stage, perineural invasion (PNI) and TSR. Univariate analysis suggests that the PDC-EMS system, PDC grading (three-tiered system) and pT stage are risk factors that affect prognosis. Multivariate regression analysis identifies WHO grading, E-cadherin, and tumor location to be independent influencing factors. When considering only PDC classification, pT stage, and lympho-vascular invasion (LVI), PDC emerges an independent influencing factor affecting postoperative survival.

Conclusion: The innovative PDC-EMS scoring system developed in this study demonstrates functional consistency with both PDC grading and TB grading. Furthermore, this scoring system offers significant advantages in predicting prognosis and shows promising potential for clinical applications.