Evaluation and development of cardiovascular risk prediction algorithms in HIV

心血管疾病（CVD）是HIV感染患者的一个重要问题，但最新的CVD风险预测工具准确预测HIV患者风险的程度尚不清楚。在这份协议中，我们将 评估2013年11月发布的新的美国心脏病学会（ACC）/美国心脏协会（AHA）CVD风险预测算法，以评估其在HIV患者中的表现。然后，我们将开发一种新的风险预测算法，结合艾滋病毒和艾滋病毒相关因素，试图提高风险预测。进行这项研究的基本原理是HIV相关CVD的独特病理生理学，这被认为是不完全由传统的风险因素解释，并在很大程度上由炎症和免疫失调驱动。虽然已建立的CVD风险预测工具已应用于HIV人群，但没有证据表明它们适合使用，因为它们不能反映这些潜在的免疫和炎症变化。准确预测CVD风险尤为重要，因为它是新发布的2013年胆固醇指南的关键组成部分，该指南指导临床医生识别需要CVD风险调整治疗的患者。通过拟议研究的目标1，我们将评估新的ACC/AHA CVD风险预测算法，评估其准确预测HIV患者CVD风险的程度，并假设其将预测风险。我们将对长期存在的FRADIOR风险评分进行平行分析。在目标2中，我们将开发一种新的CVD风险预测算法，专门用于HIV人群，首次将HIV状态作为CVD风险因素纳入预测函数，并假设其纳入将改善风险预测，超越传统的CVD风险因素。我们将通过评估纳入HIV相关变量是否能进一步改善风险预测，而不仅仅是纳入HIV状态来完善这一分析。为了实现这些目标，我们将利用一个独特适合进行研究的既定队列，由一个具有相关领域特定专业知识的跨学科团队进行研究。我们将与Ralph D 'Agostino博士及其团队合作，他们在风险预测统计分析方面拥有数十年的经验，应用复杂的风险预测方法，以严格比较现有和新的风险预测功能。心血管风险预测是HIV相关心脏病的一个关键方面， 因为它是临床医生识别需要风险调整治疗的高风险个体的能力的基础。在不同环境中验证新的ACC/AHA风险预测算法并评估新风险标志物的获益是新的2013年ACC/AHA风险评估指南中确定的特定优先事项，这两项工作都将通过本研究进行。这项拟议的研究提供了一个机会，可以及时回答一个具有重大临床和公共卫生影响的问题，优化HIV CVD风险预测方法，从而改善这一高危人群的CVD预防策略。

DESCRIPTION (provided by applicant): Project Summary/Abstract Cardiovascular disease (CVD) is a significant problem for HIV-infected patients, yet the extent to which the newest CVD risk prediction tools accurately predict risk for HIV patients is not known. In this grant, we will evaluate the new American College of Cardiology (ACC)/ American Heart Association (AHA) CVD risk prediction algorithm, released in November 2013, to assess its performance in HIV patients. We will then develop a new risk prediction algorithm incorporating HIV and HIV-related factors to attempt to improve risk prediction. The rationale for performing this study is the uniqu pathophysiology underlying HIV-associated CVD, which is thought to be incompletely explained by traditional risk factors and driven in large part by inflammation and immune dysregulation. While established CVD risk prediction tools have been applied to HIV groups, there is not evidence that they are appropriate for use as they do not reflect these underlying immunologic and inflammatory changes. Accurate prediction of CVD risk is particularly important as it is a key component of the newly released 2013 cholesterol guidelines, which guide clinicians in identifying patients in need of CVD risk modifying treatment. Through Aim 1 of the proposed study, we will evaluate the new ACC/AHA CVD risk prediction algorithm, assessing the degree to which it accurately predicts CVD risk for HIV patients and hypothesizing that it will under predict risk. We will conduct a parallel analysis of the longstanding Framingham Risk Score. In Aim 2, we will develop a new CVD risk prediction algorithm tailored for use in HIV populations, for the first time incorporating HIV status as a CVD risk factor within a prediction function and hypothesizing that its inclusion will improve risk prediction beyond that by traditional CVD risk factors alone. We will refine this analysis by assessing whether inclusion of HIV-related variables indicating disease and treatment status further improve risk prediction beyond inclusion of HIV status alone. To achieve these aims, we will leverage an established cohort uniquely suited to perform the study to be conducted by a cross- disciplinary team assembled with specific expertise in the relevant fields. We will collaborate with Dr. Ralph D'Agostino and his team, who have decades of experience in risk prediction statistical analysis, to apply sophisticated risk prediction methodology in order to rigorously compare established and new risk prediction functions. Cardiovascular risk prediction is a critical aspect of HIV-related heart disease as it underlies a clinician's ability to identify high-risk individuals in need of risk-modifying therapy. Validating the new ACC/AHA risk prediction algorithm in diverse settings and assessing the benefit of novel risk markers were specific priorities identified in the new 2013 ACC/AHA risk assessment guidelines, both of which will be performed through this study. The proposed study presents an opportunity to answer a timely question with a significant clinical and public health impact, optimizing methods for CVD risk prediction in HIV and thereby improving CVD preventative strategies for this at-risk group.