Tissue cadherin 17 (CDH17): A favorable prognostic determinant of colorectal cancer using digital image analysis

e14651 Background: Colorectal cancer (CRC) is the third most diagnosed malignancy and the second leading cause of cancer-death globally. The identification of robust prognostic markers for risk stratification of CRC would facilitate the accurate selection of high-risk patients for adjuvant therapies, hence improving patients’ survival and quality of living. Cadherin 17 (CDH17) is a validated biomarker for CRC characterized by its overexpression in tumor tissues (Zheng et al., 2021). CDH17 expression in tissue samples can be reliably detected by immunohistochemistry (IHC), which could potentially provide useful information for malignancy monitoring and prognosis. Nevertheless, the implementation of IHC in clinical settings is hindered by the lack of standardized guidelines. In this regard, the present study employed a standardized automated IHC platform and a digitalized sample scoring system to investigate CDH17 tissue expression and examine its clinical prognostic value in CRC. Methods: IHC was performed to assess CDH17 expression on paired tumor and adjacent non-tumor tissues from a cohort of 198 retrospective CRC cases, using the Lic3 antibody and an automated IHC system integrated with AI-digital imaging analysis. The M-score was generated for CDH17 membrane staining according to the formula (1 × % of weakly stained cells + 2 × % moderately stained cells + 3 × % of strongly stained cells)/6. Clinical correlation analysis with TNM, differentiation, and metastasis was determined. The prognostic value of tissue CDH17 was evaluated by Kaplan-Meier curves and Cox regression analyses on patients’ OS and RFS. Comparison of M-score and pathologist visual scoring was made to assess the validity of software-derived IHC scoring. Results: A significantly higher tissue CDH17 level was observed in CRC patients with advanced tumour staging (Early vs. Late stage, p = 0.0493; Mann-Whitney U test) and distant metastasis (p = 0.0188; Mann-Whitney U test). CRC patients with high M-score (M-Score High ; cut-off = 30) were associated with poorer OS and shorter RFS (OS, p = 0.015; RFS, p = 0.018; Log-rank test). Subgroup analysis revealed that M-Score High in early-stage CRC predicts patients’ disease recurrence of less than 2 years after curative surgery (p = 0.025; Log rank test). Multivariate cox regression analysis showed that CDH17 was an independent prognostic factor for OS (HR = 1.918, 95% CI = 1.001-3.676, p = 0.05). A comparison of software-derived CDH17 M-score with pathologist scoring demonstrated good correlation of 0.7257 (p < 0.0001, n = 36; Pearson correlation). Conclusions: The present study suggests that CDH17 is a potential biomarker of poor prognosis for CRC. Quantification of tissue CDH17 level using a standardized automated IHC platform and digital image analysis could aid in patient stratification which allows efficient treatment strategies for better survival outcomes.