The Role of Artificial Intelligence in Identifying Tumor-Reactive CD8 + TILs

Zhiwen Shi^1*Chao Yang¹Hongwei Sun²Teming Zhang³Yifen Shi^2*Yan Zhang^1*Yingying Sun^3*

• ¹School of Laboratory Medicine and Life Sciences, Wenzhou Medical University, Wenzhou, China
• ²Wenzhou Medical University, Wenzhou, China
• ³School of Basic Medical Sciences, Wenzhou Medical University, Wenzhou, China

Tumors remain one of the leading causes of mortality worldwide, with conventional therapeutic approaches facing significant challenges, such as limited efficacy and high recurrence rates. These obstacles underscore the urgent need for novel treatment strategies. Recent advances in immunotherapy have opened new avenues for cancer treatment, with tumor-reactive CD8+ tumor-infiltrating lymphocytes (TILs) emerging as key players in tumor immunotherapy. Understanding the biological characteristics of these cells is crucial for optimizing therapeutic outcomes. Traditionally, the identification of tumor-reactive CD8+ TILs has relied on the detection of known antigen epitopes. However, this approach presents several limitations, including restricted coverage of antigen epitopes, as well as high costs and extended timelines. These challenges hinder the broad application of this method in clinical practice. The rise of artificial intelligence (AI) has introduced innovative strategies for identifying tumor-reactive CD8+ TILs. By leveraging gene expression profiles, clonal expansion patterns, and the structures of T cell receptor-peptide major histocompatibility complex (TCR-pMHC), AI-driven methods have demonstrated high sensitivity and adaptability, offering a promising alternative to traditional approaches. Despite their potential, AI applications in this field still face significant hurdles, particularly in terms of clinical validation and large-scale implementation. This review explores the limitations of conventional CD8+ TILs identification methods and examines the role of AI in overcoming these challenges. It also discusses the current obstacles to the clinical adoption of AI-based strategies and highlights their future potential in advancing tumor immunotherapy.