Zhongyu Mou, Patra Volarath, Rebecca Racz, Kevin P Cross, Mounika Girireddy, Suman Chakravarti, Lidiya Stavitskaya
{"title":"预测心脏不良反应的定量结构-活性关系模型","authors":"Zhongyu Mou, Patra Volarath, Rebecca Racz, Kevin P Cross, Mounika Girireddy, Suman Chakravarti, Lidiya Stavitskaya","doi":"10.1021/acs.chemrestox.4c00186","DOIUrl":null,"url":null,"abstract":"<p><p>Drug-induced cardiotoxicity represents one of the most common causes of attrition of drug candidates in preclinical and clinical development. For this reason, the evaluation of cardiac toxicity is essential during drug development and regulatory review. In the present study, drug-induced postmarket adverse event combinations from the FDA Adverse Event Reporting System were extracted for 2002 drugs using 243 cardiac toxicity-related preferred terms (PTs). These PTs were combined into 12 groups based on their clinical relevance to serve as training sets. The optimal classification scheme was determined using a combination of data sources that included drug labeling information, published literature, clinical study data, and postmarket surveillance data. Two commercial QSAR platforms were used to construct 12 models, including general cardiac toxicity, cardiac ischemia, heart failure, cardiac valve disease, myocardial disease, pericardial disease, structural heart disease, cardiac arrhythmia, Torsades de Pointes, long QT syndrome, atrial fibrillation and ventricular arrhythmia, and cardiac arrest. The cross-validated performance for the new models reached a sensitivity of up to 80% and negative predictivity of up to 80%. These new models covering a wide range of cardiac endpoints will provide fast, reliable, and comprehensive predictions of potential cardiotoxic compounds in drug discovery and regulatory safety assessment.</p>","PeriodicalId":31,"journal":{"name":"Chemical Research in Toxicology","volume":" ","pages":"1924-1933"},"PeriodicalIF":3.7000,"publicationDate":"2024-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Quantitative Structure-Activity Relationship Models to Predict Cardiac Adverse Effects.\",\"authors\":\"Zhongyu Mou, Patra Volarath, Rebecca Racz, Kevin P Cross, Mounika Girireddy, Suman Chakravarti, Lidiya Stavitskaya\",\"doi\":\"10.1021/acs.chemrestox.4c00186\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>Drug-induced cardiotoxicity represents one of the most common causes of attrition of drug candidates in preclinical and clinical development. For this reason, the evaluation of cardiac toxicity is essential during drug development and regulatory review. In the present study, drug-induced postmarket adverse event combinations from the FDA Adverse Event Reporting System were extracted for 2002 drugs using 243 cardiac toxicity-related preferred terms (PTs). These PTs were combined into 12 groups based on their clinical relevance to serve as training sets. The optimal classification scheme was determined using a combination of data sources that included drug labeling information, published literature, clinical study data, and postmarket surveillance data. Two commercial QSAR platforms were used to construct 12 models, including general cardiac toxicity, cardiac ischemia, heart failure, cardiac valve disease, myocardial disease, pericardial disease, structural heart disease, cardiac arrhythmia, Torsades de Pointes, long QT syndrome, atrial fibrillation and ventricular arrhythmia, and cardiac arrest. The cross-validated performance for the new models reached a sensitivity of up to 80% and negative predictivity of up to 80%. These new models covering a wide range of cardiac endpoints will provide fast, reliable, and comprehensive predictions of potential cardiotoxic compounds in drug discovery and regulatory safety assessment.</p>\",\"PeriodicalId\":31,\"journal\":{\"name\":\"Chemical Research in Toxicology\",\"volume\":\" \",\"pages\":\"1924-1933\"},\"PeriodicalIF\":3.7000,\"publicationDate\":\"2024-12-16\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Chemical Research in Toxicology\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1021/acs.chemrestox.4c00186\",\"RegionNum\":3,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2024/11/13 0:00:00\",\"PubModel\":\"Epub\",\"JCR\":\"Q2\",\"JCRName\":\"CHEMISTRY, MEDICINAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Chemical Research in Toxicology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1021/acs.chemrestox.4c00186","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/11/13 0:00:00","PubModel":"Epub","JCR":"Q2","JCRName":"CHEMISTRY, MEDICINAL","Score":null,"Total":0}
引用次数: 0
摘要
药物引起的心脏毒性是候选药物在临床前和临床开发过程中最常见的损耗原因之一。因此,在药物开发和监管审查过程中,对心脏毒性的评估至关重要。本研究使用 243 个与心脏毒性相关的首选术语(PTs),从 FDA 不良事件报告系统中提取了 2002 种药物的上市后不良事件组合。根据这些术语的临床相关性将其分为 12 组,作为训练集。最佳分类方案是综合使用各种数据源确定的,这些数据源包括药物标签信息、公开发表的文献、临床研究数据和上市后监测数据。使用两个商业 QSAR 平台构建了 12 个模型,包括一般心脏毒性、心脏缺血、心力衰竭、心脏瓣膜疾病、心肌疾病、心包疾病、结构性心脏病、心律失常、Torsades de Pointes、长 QT 综合征、心房颤动和室性心律失常以及心脏骤停。经交叉验证,新模型的灵敏度高达 80%,负预测率高达 80%。这些涵盖广泛心脏终点的新模型将为药物发现和监管安全评估提供快速、可靠和全面的潜在心脏毒性化合物预测。
Quantitative Structure-Activity Relationship Models to Predict Cardiac Adverse Effects.
Drug-induced cardiotoxicity represents one of the most common causes of attrition of drug candidates in preclinical and clinical development. For this reason, the evaluation of cardiac toxicity is essential during drug development and regulatory review. In the present study, drug-induced postmarket adverse event combinations from the FDA Adverse Event Reporting System were extracted for 2002 drugs using 243 cardiac toxicity-related preferred terms (PTs). These PTs were combined into 12 groups based on their clinical relevance to serve as training sets. The optimal classification scheme was determined using a combination of data sources that included drug labeling information, published literature, clinical study data, and postmarket surveillance data. Two commercial QSAR platforms were used to construct 12 models, including general cardiac toxicity, cardiac ischemia, heart failure, cardiac valve disease, myocardial disease, pericardial disease, structural heart disease, cardiac arrhythmia, Torsades de Pointes, long QT syndrome, atrial fibrillation and ventricular arrhythmia, and cardiac arrest. The cross-validated performance for the new models reached a sensitivity of up to 80% and negative predictivity of up to 80%. These new models covering a wide range of cardiac endpoints will provide fast, reliable, and comprehensive predictions of potential cardiotoxic compounds in drug discovery and regulatory safety assessment.
期刊介绍:
Chemical Research in Toxicology publishes Articles, Rapid Reports, Chemical Profiles, Reviews, Perspectives, Letters to the Editor, and ToxWatch on a wide range of topics in Toxicology that inform a chemical and molecular understanding and capacity to predict biological outcomes on the basis of structures and processes. The overarching goal of activities reported in the Journal are to provide knowledge and innovative approaches needed to promote intelligent solutions for human safety and ecosystem preservation. The journal emphasizes insight concerning mechanisms of toxicity over phenomenological observations. It upholds rigorous chemical, physical and mathematical standards for characterization and application of modern techniques.