AUTHOR=Li Shengshu , Geng Ziying , Hong Shuang , Zhang Jianxin , Yang Yanli , Wei Qin , Zhang Xinxin , Zhuang Xiaofei , Huo Rujie , Han Songyan , Wang Jie 

TITLE=Advances in the mechanisms, imaging characteristics and management strategies for immune checkpoint inhibitor-related pneumonitis

JOURNAL=Frontiers in Immunology

VOLUME=Volume 16 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2025.1656063

DOI=10.3389/fimmu.2025.1656063

ISSN=1664-3224

ABSTRACT=In recent years, the introduction of immune checkpoint inhibitors (ICIs) has revolutionized the treatment landscape for malignant tumors, markedly improving survival outcomes across various cancers, such as lung cancer, esophageal cancer, and melanoma. Consequently, ICIs have become a cornerstone of first-line therapy for numerous malignancies. However, while ICIs effectively modulate immune responses to combat tumor cells, they may also trigger excessive immune activation and T-cell dysfunction, thereby leading to a spectrum of immune-related adverse events (irAEs). The organs most frequently affected by these irAEs include the skin, gastrointestinal tract, endocrine system, and lungs. Among these adverse events, the development of severe immune checkpoint inhibitor-related pneumonitis (CIP) may result in significant disability, permanent discontinuation of ICIs, and even death, with real-world incidence rates exceeding those reported in clinical trials. Early detection, precise diagnosis, and timely intervention are critical for optimizing patient outcomes. However, diagnosing CIP remains challenging because it relies heavily on high-resolution chest CT imaging and a detailed treatment history. The radiological features of CIP are often nonspecific, complicating its identification. This complexity is further exacerbated in patients receiving consolidative immunotherapy following concurrent or sequential chemoradiotherapy for stage III unresectable non-small cell lung cancer, where distinguishing between radiation pneumonitis and CIP becomes particularly difficult. To address these challenges, an increasing number of imaging experts are investigating the potential of radiomics and machine learning techniques in predicting the occurrence and assessing the prognosis of CIP. This article comprehensively reviews the pathogenesis of CIP, the predictive value of radiomics in identifying this condition and recent advancements in treatment strategies, with the aim of providing novel insights for future research and clinical management of CIP.