癌症免疫治疗中的预测性生物标志物：跨选定实体瘤的叙述性综述

Cancer immunotherapy has transformed the therapeutic landscape for several malignancies, offering durable responses in select patient populations.

However, response rates remain variable, underscoring the need for robust predictive and prognostic biomarkers. This narrative review synthesizes current evidence on biomarkers guiding immunotherapy in non-small cell lung cancer (NSCLC), melanoma, breast cancer, and colorectal cancer. Literature was identified through manual screening of PubMed and Scopus using key terms related to immunotherapy biomarkers, with no date restrictions applied; studies published up to May 2025 were included. Clinically validated markers such as programmed death-ligand 1 (PD-L1) and microsatellite instability-high (MSI-H) demonstrate utility but face limitations including assay variability and tumor heterogeneity.

Emerging candidates, such as tumor-infiltrating lymphocytes (TILs), relative eosinophil count (REC), lactate dehydrogenase (LDH), and S100 calcium-binding protein B (S100B), offer additional prognostic or predictive value but require further validation. This review adds value by integrating and comparing both validated and emerging biomarkers across tumor types and by emphasizing practical considerations for real-world application.

Given the narrative format and focus on selected cancers, the scope is inherently limited, and not all emerging biomarkers or real-world implementation data are fully addressed. Methodological limitations and current gaps in biomarker validation are also discussed.