Automated Quantification of sTIL Density with H&E-Based Digital Image Analysis Has Prognostic Potential in...

Area of Science:

• Oncology
• Digital Pathology
• Biomarker Discovery

Background:

• Triple-negative breast cancer (TNBC) is an aggressive subtype with limited treatment options.
• Stromal tumor-infiltrating lymphocytes (sTIL) are emerging as a key prognostic biomarker for TNBC.
• Manual assessment of sTIL faces limitations hindering widespread clinical adoption.

Purpose of the Study:

• To develop and validate a fully automated computational pipeline for assessing sTIL in TNBC.
• To ensure the computational method adheres to the International Immuno-Oncology Biomarker Working Group (TIL-WG) guidelines.
• To evaluate the prognostic significance of the computational sTIL score for overall survival (OS) in TNBC patients.

Main Methods:

• Development of a hematoxylin and eosin (H&E)-based digital image analysis pipeline for automated sTIL quantification.
• Correlation of the computational sTIL score with manual pathologist-derived sTIL status.
• Prognostic analysis of the computational sTIL score in a retrospective TNBC cohort, adjusting for clinical variables.

Main Results:

• The automated pipeline generated a quantitative and interpretable sTIL score.
• The computational score demonstrated significant correlation with manual sTIL assessment.
• The sTIL score was prognostic for OS (HR: 0.81, p=0.001), independent of age, tumor size, nodal status, and tumor type.
• The pipeline successfully stratified patients into distinct prognostic groups.

Conclusions:

• The developed H&E-based computational pipeline offers a robust and automated method for sTIL assessment in TNBC.
• This approach aligns with TIL-WG guidelines, addressing limitations of manual assessment.
• The computational sTIL score holds significant prognostic value for OS in TNBC, supporting its potential clinical utility.