Modeling of Viral Load Trajectories for HIV Cure Research

世界迫切需要推进艾滋病毒治愈研究议程，以解决全球艾滋病毒持续高企的问题。 患病率和相关死亡率。尽管联合抗逆转录病毒疗法(ART)成功地实现了 持续控制病毒复制，对副作用、药物-药物相互作用、耐药性和 费用要求需要确定实现艾滋病毒根除或无抗逆转录病毒药物缓解的战略。遵循艺术 停药后，患者的病毒载量水平通常会迅速上升到一个峰值，然后下降，然后稳定在一个 病毒载量设定点。描述病毒反弹轨迹的特征(例如，病毒反弹时间和病毒 设定值)在分析抗逆转录病毒治疗中断(ATI)并确定宿主、病毒学和 预测这些特征的免疫学因素是艾滋病毒治愈研究的核心。但这样做需要 应对各种分析挑战，包括非线性病毒反弹轨迹、粗糙数据 由于检测的量化限制，病毒载量的间歇性测量，小样本来自 个人研究和高维候选预测指标。受到我们与艾滋病毒的持续合作的激励 治愈研究人员，并在我们之前工作的基础上，我们的目标是通过以下方式解决关键的方法差距 利用艾滋病临床试验小组进行的多项随机研究和 苏黎世主要HIV感染队列。目标1建议开发一套新的方法来预测 基于全面的病史资料的病毒反弹，例如ART开始后的病毒衰减率， 扩展拟合算法和为区间删失结果开发的变量选择技术。目标2 建议使用平滑的模拟伪最大似然方法来拟合病毒反弹模型 最大化基于蒙特卡罗近似构造的平滑模拟目标函数 平滑反应的前两个时刻，并开发方法来评估时间之间的关联 以反弹和病毒设定点，并同时选择影响不同细微特征的生物标志物 病毒反弹的轨迹。目标3建议开发以最佳方式集成来自多个 队列和不同阶段的病毒载量轨迹，同时适当地考虑同质性和 不同研究之间协变量效应的异质性。创新在于新技术的开发和应用 解决现有数据分析中各种固有挑战的病毒反弹建模方法。 意义在于这些方法在更好地表征病毒反弹轨迹、识别前病毒 ATI预测指标，并评估新型治疗剂的效果和机制。评选结果 建议的研究可以为未来ATI研究的优化设计提供信息，并提供新的工具，可以提取更多 来自已完成和正在进行的ATI研究中收集的数据的信息。这些新的见解在 ATI前对ATI后病毒血症控制效果较好的预测因子的发现和针对性干预措施的评估 病毒反弹过程的不同组成部分，最终提高了我们找到治愈艾滋病毒的能力。

The world urgently needs to advance the HIV cure research agenda to address the persistently high global HIV prevalence and associated mortality. Despite the success of combined antiretroviral therapy (ART) in achieving sustained control of viral replication, the concerns about side-effects, drug-drug interactions, drug resistance and cost call for a need to identify strategies for achieving HIV eradication or an ART-free remission. Following ART withdrawal, patients' viral load levels usually increase rapidly to a peak followed by a dip, and then stabilize at a viral load set point. Characterizing features of the viral rebound trajectories (e.g., time to viral rebound and viral set points) after analytic antiretroviral treatment interruption (ATI) and identifying host, virological, and immunological factors that are predictive of these features are central to HIV cure research. But doing so requires addressing a variety of analytical challenges, including the non-linear viral rebound trajectories, coarsened data due to the assay's limit of quantification, intermittent measurements of viral load values, small sample sizes from individual studies, and high-dimensional candidate predictors. Motivated by our ongoing collaborations with HIV cure research investigators and built on our previous work, we aim to address key methodological gaps by leveraging data from multiple randomized studies conducted by the AIDS Clinical Trials Group and from the Zurich Primary HIV Infection Cohort. Aim 1 proposes to develop a new set of methods for prediction of time to viral rebound based on comprehensive history profiles, such as the rate of viral decay after ART initiation, extending fitting algorithms and variable selection techniques developed for interval-censored outcomes. Aim 2 proposes to fit the viral rebound model using a Smoothed Simulated Pseudo Maximum likelihood method which maximizes a smoothed simulated objective function constructed based on a Monte Carlo approximation of the first two moments of the smoothed responses, and to develop methods to assess the association between time to rebound and the viral set point and to simultaneously select biomarkers that affect different finer features of the viral rebound trajectory. Aim 3 proposes to develop methods that optimally integrate data from multiple cohorts and different phases of viral load trajectories while properly accounting for the homogeneity and heterogeneity in covariate effects across studies. Innovation lies in the development and application of new methods for modeling viral rebound that address various inherent challenges in analyses of available data. Significance lies in the role of these methods in better characterizing viral rebound trajectories, identifying pre- ATI predictors, and assessing the effects and mechanisms of novel therapeutic agents. The results of the proposed research can inform optimal design of future ATI studies and provide new tools that can extract more information from data collected in completed and ongoing ATI studies. These new insights are useful in the discovery of pre-ATI predictors of better viremia control post ATI and evaluation of interventions that target different components of viral rebound process, ultimately improving our capacity to find a cure for HIV.