专家评估在杂质遗传毒性(定量)构效关系评价中的应用

Application of Expert Review in the Evaluation of(Quantitative)Structure-Activity Relationship of the Genotoxicity of Impurities

目的通过专家评估获得或变更(定量)构效关系[(Q)SAR]评价分类结果实例,提供一种基于人用药品技术要求国际协调理事会(ICH)M7(R1)指导原则进一步利用专家进行评估的杂质遗传毒性评价程序。方法采用Lhasa公司的Nexus 2.5.0版软件平台(整合了Derek Nexus和Sarah Nexus)对杂质进行遗传毒性评价,任意选择3个预测得到阳性、阴性、相互矛盾或未分类结果的实例,进行专家回顾评估,获得最终分类结果。结果实例1,Derek预测结果为阴性,Sarah预测结果为阳性,专家评估后分类结果由3类变更为5类;实例2,Derek预测结果为阴性,Sarah预测结果为模棱两可,专家评估后界定为5类;实例3,Derek和Sarah预测结果均为阴性,专家评估后分类结果由5类变更为1类。结论通过对实例化合物遗传毒性分类结果论证过程的描述,为专家评估在杂质遗传毒性(Q)SAR评价中的应用提供了基本的参考程序。

Objective To provide an evaluation procedure for impurity genotoxicity with expert review based on the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use(ICH)M7(R1)guidelines by examples of(quantitative)structure-activity relationship[(Q)SAR]classification results,which were obtained or modified by using expert evaluation.Methods Nexus 2.5.0 software platform including Derek Nexus and Sarah Nexus provided by Lasha Limited were used to evaluate the genotoxicity of impurities.Three cases of compounds with positive,negative,conflicting or inconclusive prediction results were randomly selected,and experts'review and evaluation were conducted to obtain the final classification results.Results For case 1,Derek prediction result was negative while Sarah prediction result was positive,and the compound should be assigned class 5 instead of class 3 after implementing the expert review.For case 2,Derek prediction result was negative and Sarah prediction result was ambiguous,and the compound was classified into class 5 after implementing the expert review.For case 3,Derek and Sarah prediction results were both negative,and the compound should be assigned class 1 instead of class 5 after implementing the expert review.Conclusion The description of the genotoxicity classification results of example compounds can provide a reference of the basic expert review program in the evaluation of the genotoxicity of impurity by using(Q)SAR.