Five latent patterns of immune checkpoint inhibitor-related adverse event reporting across cancer types: an integrated pharmacovigilance and transcriptomic analysis
Year: 2026
Type: article
Study design: Observational Study
Abstract: Immune checkpoint inhibitor (ICI)-related adverse event (irAE) profiles vary across cancer types, but whether they form shared cross-cancer patterns remains unclear. We analyzed 200,099 ICI-exposed cancer reports from the US Food and Drug Administration Adverse Event Reporting System and 65,309 from the Japanese Adverse Drug Event Report database (2015-2025). Cancer-system organ class (SOC) reporting profiles were quantified by disproportionality analysis, and non-negative matrix factorization identified latent patterns. Pattern-defined cancer groups were characterized by organ-system involvement, fatal outcome reporting, and time to onset; FAERS-derived profiles were evaluated in JADER, and exploratory transcriptomic annotation used TCGA and GTEx data. FAERS showed heterogeneous cancer-specific reporting profiles (representative reporting odds ratios, 2.11-6.42; all P < 0.001), with major profiles reproduced in JADER. Five latent patterns were identified: ocular, hematologic, neurologic, systemic-like, and visceral-like. The visceral-like group, dominated by gastrointestinal, hepatobiliary, renal and urinary, and respiratory involvement, had the highest proportion of irAE reports with recorded fatal outcomes in FAERS (19.0%) and JADER (4.8%); in FAERS, its median onset was 34 days. Median onset in FAERS was earliest in the systemic-like group (32 days) and latest in the hematologic group (43 days; P < 0.001). At the cancer-type level, exploratory transcriptomic annotation showed broad immune activation in cancers of the systemic-like group and macrophage, dendritic-cell, bile acid metabolism, and cytokine-signaling features in cancers of the visceral-like group. The five-pattern framework organizes cross-cancer heterogeneity in relative organ-system irAE reporting profiles and provides hypotheses for cancer-informed safety surveillance and prospective biological investigation.
Source: Frontiers in Immunology
Authors: Linghao Ni(Conceptualization, Formal analysis, Methodology, Writing – original draft), Xue Min(Data curation, Validation, Writing – review & editing), Dingqian Yong(Data curation, Validation, Writing – review & editing), Xiaoyu Shan(Data curation, Validation, Writing – review & editing), Guohan Song(Data curation, Validation, Writing – review & editing), Jiawei Zhou(Conceptualization, Methodology, Writing – review & editing), Bin Peng(Conceptualization, Supervision, Writing – review & editing)
Institution: Chongqing Medical University
Language: English
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,801
OpenAlex ID: W7167076733
