HIV感染孕妇抗病毒药物耐药预测模型研究

Mother-to-child transmission of HIV poses great harm to children’s health. HIV resistant strains are detected among HIV-infected pregnant women on anti-retroviral drugs. Studies in China reported that the prevalence rates of resistance to therapy were between 10% and 12% and the rates were even higher reported by studies conducted abroad. However, testing on resistance to therapy is not routinely provided for HIV-infected pregnant women on anti-retroviral drugs in China. Our study aims to follow-up a cohort of HIV-infected pregnant women and collection data from epidemiologic survey, clinical research, laboratory techniques, etc. Genome sequences techniques will be used to expand HIV pol locus and the results of sequences will be submitted to Stanford University HIV Drug Resistance Database(www.hivdb.stanford.edu/) for resistance assessment. The samples will be separated into training samples and testing samples randomly and Cox regression model will be conducted for assessment and statistical tests using resistance to therapy as outcomes. Independent factors include duration of HIV infection, duration of therapy, ARV regimens, withdrawal time, viral load, CD4+T counting , behavior characteristics,nutrition,clinical features etc. Practice value will be evaluated from professional perspective. The research results will be critical for promoting clinical therapy precision and improving control and prevention strategies on mother-to-child transmission of HIV in our country.

HIV母婴传播给儿童造成了极大的危害，在HIV感染孕产妇服用抗病毒药物人群中发现了HIV耐药株，国内有研究发现耐药率为10%-12%，但是目前我国在HIV孕妇抗病毒药物防治中尚没有开展HIV耐药常规检测。本研究拟建立HIV感染孕妇研究队列，通过流行病学调查、临床研究、实验室技术等收集资料。利用测序技术将HIV Pol区进行扩增，所得序列提交斯坦福大学耐药数据库(www.hivdb.stanford.edu/)进行耐药评价。选用耐药产生为结局事件，HIV感染时间、用药时长、用药方案、停药时长、病毒载量、CD4+T数、行为特征、营养状态和临床特征等影响因素为自变量，将样本随机分为训练样本和检验样本，利用Cox回归模型进行建模、评估和检验，并从专业角度评估实际意义和价值。研究结果对提高母婴阻断抗病毒药物临床用药的精准性和我国HIV母婴传播的防治水平有重要意义。

HIV耐药株的母婴传播给儿童造成了极大的危害，然而我国针对HIV感染孕妇尚没有开展HIV耐药常规检测。本研究通过建立HIV感染孕妇随访队列，进行现场流行病学调查和实验室检测，收集HIV感染孕妇一般情况、HIV感染相关行为、接受预防HIV母婴传播服务情况、抗病毒用药情况和社会支持等情况。通过自建耐药检测方法，利用测序技术将HIV Pol区进行扩增，探索HIV耐药分子进化规律，进行耐药评价。选用耐药产生为结局事件，影响因素为自变量，利用Cox比例风险模型进行建模。按照纳入标准，本研究研究对象207例，其中从145例样本中成功扩增HIV，检出耐药26例，经过影响因素分析，最终发现妊娠前应用药物、妊娠前用药时间、感染途径3个因素在耐药人群和不耐药人群分布存在差异。在本次妊娠前应用过抗病毒药物（23.10%）、妊娠前用药时间≥2年（28.07%）、血液或吸毒途径感染（28.57%）的孕妇中，HIV耐药检出率较高。在数据分析和专业分析的基础上，将妊娠前用药时间、本次妊娠用药时间、CD4+T淋巴细胞计数、HIV病毒载量等作为自变量构建Cox回归预测模型，经评估，所建模型评价预测能力的一致性指数为0.871，效度较高，具有应用价值。研究发表文章6篇，培养研究生8人，举办研究现场培训/会议/质控共17次，覆盖贵州省9个市州的88个地区，提高了研究地区预防HIV母婴传播相关科研、临床和检测能力。研究结果为提高我国预防HIV母婴传播防治水平、提高母婴抗病毒用药的精准性具有重要意义，研究结果为修订我国预防HIV母婴传播工作实施方案及相关文件提供了参考依据。