Tumor-infiltrating Lymphocytes in Breast Cancer: Prognostic Significance, Spatial Heterogeneity, And Emerging AI-based Assessment Approaches

Abstract

Tumor-infiltrating lymphocytes (TILs) represent a crucial component of the tumor microenvironment in breast cancer (BC), playing a significant role in tumor progression, immune surveillance, and therapeutic response. This review provides a comprehensive synthesis of current evidence regarding the prognostic and predictive significance of TILs across breast cancer molecular subtypes, with particular emphasis on luminal breast cancer and recent advances in artificial intelligence (AI)-based assessment approaches. A narrative review with a structured literature search was conducted using major biomedical databases, including PubMed, Scopus, and Web of Science, to identify relevant studies published between 2014 and 2025. Current evidence demonstrates that high stromal TIL levels are strongly associated with improved overall survival, disease-free survival, and pathological complete response in triple-negative breast cancer (TNBC) and HER2-positive breast cancer. In contrast, the prognostic role of TILs in luminal breast cancer remains controversial, largely due to lower baseline immune infiltration, biological heterogeneity, and methodological variability across studies. Novel computational approaches, including deep learning-based TIL assessment systems, may provide more consistent quantification of TIL density and spatial distribution patterns, with potential to improve prognostic stratification. Emerging evidence suggests that not only TIL density but also spatial organization, including aggregated and diffuse immune infiltration patterns, significantly influences clinical outcomes. Compared with conventional histopathological evaluation, AI-driven models have shown potential to improve reproducibility, quantitative precision, and spatial characterization. This review highlights the need for standardized TIL assessment methodologies and supports the integration of digital pathology and AI-based tools into clinical practice to improve prognostic accuracy and facilitate personalized therapeutic decision-making in breast cancer.