EDITORIAL article

Front. Med.

Sec. Translational Medicine

Exploring adverse drug reactions: monitoring, mechanism, intervention, and resolution

  • 1. Yangzhou University, Yangzhou, China

  • 2. National Cancer Institute Center for Cancer Research, Bethesda, United States

  • 3. Shandong First Medical University, Jinan, China

  • 4. China Pharmaceutical University, Nanjing, China

Article metrics

View details

217

Views

The final, formatted version of the article will be published soon.

Abstract

has shifted toward the comprehensive capture and interpretation of complex clinical events, embracing a full continuum of patient-specific responses across various biological contexts. This Research Topic highlights the importance of systematically characterizing delayed or atypical ADRs. For example, clinically significant immune-related adverse events with multisystem involvement emerge long after treatment initiation or even following drug discontinuation, underscoring how delayed onset may lead to repeated exposure and life-threatening outcomes if early causality attribution fails. As illustrated by Hao et al., sintilimab-induced colitis progresses to intestinal obstruction and hemorrhage despite initial corticosteroid therapy. Early endoscopic evaluation, timely escalation to biologic agents, and multidisciplinary management are crucial to mitigate severe complications. Vigilant monitoring and individualized treatment plans are essential for improving outcomes in high-risk patients (1). Osilodrostat is a novel steroidogenesis inhibitor for the treatment of Cushing's syndrome. However, prolonged endocrine suppression may persist even after the discontinuation of osilodrostat, potentially due to extensive off-target effects (2). In addition, investigations into postacute COVID-19 vaccination syndrome also highlight the need to establish systematic efforts for long-term and delayed adverse events, challenging the conventional temporal bounds of pharmacovigilance frameworks (3).Machine learning-based prediction models have a considerable positive impact on proactive risk identification and stratification. Yue et al. recognized gastrointestinal comorbidities, alcohol consumption, and concomitant medication use as key multifactorial, pathology-driven risk factors for semaglutide-associated adverse events. A logistic regression-based nomogram scoring tool was subsequently developed to estimate the risk of semaglutide-induced gastrointestinal adverse events (4). Similarly, machine learning algorithms such as XGBoost were employed to stratify key determinants, such as IL-6, distinguishing outcomes between mono-and multimodal treatment strategies in pancreatic cancer patients (5). Wang et al. evaluated the association between KRT8 expression and the development of immunerelated pneumonitis (IRP) via statistical models, including multivariate logistic regression, bootstrap validation, and Bayesian analysis, suggesting that elevated KRT8 expression in tumor tissue may serve as a potential biomarker for predicting IRP in lung adenocarcinoma patients receiving immunotherapy (6). At the core of biologically relevant heterogeneity lies the molecular interaction between drugs and biological systems that drives diverse adverse events. Several studies in this collection provide a mechanistic dissection of adverse events, unravelling the complex causal chains between drug exposure and clinical manifestations. Zhang et al. revealed that ibrutinib induces hair apoptosis and hearing loss by inhibiting GPR83-AKT signalling. GPR83 overexpression activates AKT to protect hair cells, while the caspase inhibitor Z-VAD-FMK mitigates ototoxicity.These findings identify GPR83-AKT as a potential target for preventing chemotherapy-induced hearing loss (7).Rare events such as paroxetine-induced pancreatitis have been confirmed through involuntary rechallenge in the absence of conventional risk factors.Paroxetine may induce pancreatitis through multiple nonexclusive mechanisms.Elevated serotonin levels caused by SERT inhibition can impair β-cell function and disrupt insulin and exocrine secretion. Additionally, reactive catechol metabolites and individual genetic polymorphisms in CYP2D6/CYP2C19 may trigger T-cell-mediated hypersensitivity reactions, further contributing to pancreatic injury (8). Tranexamic acid (TXA) is a synthetic lysine analogue that inhibits fibrinolysis to reduce perioperative bleeding during posterior lumbar interbody fusion (PLIF). TXA reduces perioperative blood loss in PLIF primarily by inhibiting the fibrinolytic system, stabilizing fibrin clots, protecting coagulation components, and exerting beneficial anti-inflammatory actions without significantly increasing systemic thrombotic risk when used at appropriate doses (9). Several studies have demonstrated that natural compounds can simultaneously increase therapeutic efficacy and mitigate toxicity by modulating convergent signalling pathways involved in the immune response. Tao et al. demonstrated that cycloastragenol (CAG) inhibits the JAK/STAT5 signalling pathway, reducing the levels of the neutrophil chemotactic cytokines CXCL3 and CCL5, thereby limiting neutrophil infiltration into brain metastases and enhancing the efficacy of radiotherapy. Concurrently, CAG suppresses the IKK/NF-κB pathway, promotes microglia/macrophage polarization toward an anti-inflammatory phenotype, mitigates neuroinflammation, and protects brain function from radiation-induced injury (10). In addition, homoplantaginin alleviates DSS-induced ulcerative colitis in mice by modulating the MMP9-RLN2 signalling axis, reducing proinflammatory cytokines and immune cell infiltration while protecting the intestinal barrier (11). Similarly, Folium isatidis and its active ingredients mitigate COPD-related lung lesions by correcting aberrant immune cell infiltration, restoring immune homeostasis, and improving pulmonary function (12). Evidence-based approaches provide the foundation for managing ADRs by identifying risk factors, quantifying incidence, and evaluating intervention efficacy.

Summary

Keywords

Adverse drug reactions (ADRs), Causality attribution, machine learning, Precision safety management, proactive pharmacovigilance

Received

29 January 2026

Accepted

10 February 2026

Copyright

© 2026 Tao, Li, Wu and Shan. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Li Tao; Xin Li; Yunhao Wu; Yunlong Shan

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Share article

Article metrics