创新链/学科链/研发链/产业链

新药研发前沿动态 / 医药领域趋势进展

基于FAERS数据库的替尔泊肽不良事件影响因素及风险预测模型构建

Influencing Factors and Risk Prediction Models for Tirzepatide-Associated Adverse Events Based on FAERS Database

  • 摘要: 目的 探讨基于美国食品药品监督管理局不良事件报告系统(Food and Drug Administration Adverse Event Reporting System,FAERS)数据库的替尔泊肽不良事件影响因素及风险预测模型构建。方法 收集FAERS数据库中2023年3月至2025年3月的数据,提取与替尔泊肽有关的药品不良事件(adverse drug event,ADE)报告。运用报告比值比(reporting odds ratio,ROR)分析技术、英国药品和保健品管理局(Medicines and Healthcare Products Regulatory Agency,MHRA)综合标准法对数据进行深度挖掘并对风险信号进行检测,以系统性识别药物不良事件信号。采用国际医学用语词典中的系统器官分类及首选术语框架,对不良事件进行标准化术语映射与层级化分类统计。采用受试者工作特征曲线(receiver operating characteristic curve,ROC)分析目标药物发生目标不良事件的预测价值。结果 本研究最终纳入14 929 854例ADE报告,其中目标药物ADE为53 145例报告,其他药物ADE为14 876 709例报告。采用ROR、MHRA法对ADE报告相关信号进行挖掘,发现有效信号132个。其中,报告数前5位分别为给药剂量错误(14 370例)、注射部位疼痛(7 711例)、超说明书使用(6 119例)、恶心(5 638例)及额外剂量给药(3 624例),信号强度前5位分别为注射部位发冷(ROR=103.59)、意外漏注(ROR=73.26)、注射部位损伤(ROR=71.26)、产品损坏(ROR=50.61)及额外剂量给药(ROR=50.17)。ROC曲线构建预测模型,替尔泊肽发生目标ADE的曲线下面积为0.958,具有较高的预测价值。结论 通过ROR与MHRA方法对ADE报告进行系统分析,能够揭示替尔泊肽上市后监测中的关键风险信号及其潜在机制。

     

    Abstract: Objective To explore the influencing factors of tirzepatide-associated adverse events and to construct risk prediction models based on the Food and Drug Administration Adverse Event Reporting System (FAERS) database. Methods Data from the FAERS database from March 2023 to March 2025 were collected, and adverse drug events (ADE) reports related to tirzepatide were extracted. The reporting odds ratio (ROR) analysis technique and the United Kingdom's Medicines and Healthcare Products Regulatory Agency (MHRA) standardized method were applied for in-depth data mining and detection of risk signals to systematically identify ADE signals. ADEs were standardized and hierarchically classified using the system organ class (SOC) and preferred term (PT) framework of the Medical Dictionary for Regulatory Activities (MedDRA). The predictive value of the target drug for specific ADEs was evaluated using receiver operating characteristic (ROC) curve analysis. Results A total of 14 929 854 ADE reports were included, comprising 53 145 reports for the target drug and 14 876 709 for other drugs. Mining with ROR and MHRA methods yielded 132 valid signals. The top five ADEs by reported count were dosing errors (14 370 cases), injection site pain (7 711 cases), off-label use (6 119 cases), nausea (5 638 cases), and extra dose administration (3 624 cases). The top five signals by strength were injection site coldness (ROR = 103.59), missed dose (ROR = 73.26), injection site injury (ROR = 71.26), product damage (ROR = 50.61), and extra dose administration (ROR = 50.17). The predictive model based on the ROC curve demonstrated the area under the curve (AUC) of 0.958 for tirzepatide-related target ADEs, indicating high predictive value. Conclusion Systematic analysis of ADE reports using ROR and MHRA methods can reveal key risk signals and underlying mechanisms in post-marketing surveillance of tirzepatide.

     

/

返回文章
返回