Targeting TB transmission hotspots to find undiagnosed TB in South Africa: a genomic, geospatial and modeling study (TARGET- TB)

项目摘要 尽管重新作出了公共卫生努力，包括更有效的治疗，但结核病的发病率仍在上升， 只是逐渐减少，这是由于无法控制社区结核病传播造成的。早期 传染性个体的识别和治疗是打破传播链的关键， 高达40%的结核病病例仍未得到诊断。相关的持续时间延长 传染性和治疗开始的延误是结核病持续传播的重要原因。 未确诊的病例包括医疗保健系统遗漏的患病个体， 无症状（亚临床结核病），其传播结核病的能力未知。在我们的初步数据中，使用 活动病例发现和全血RNA生物标志物，我们确定了亚临床结核病的比例， 接近或超过有症状的活动性结核病。这些病例与存在存活的 痰液中的细菌，指向一个巨大的潜在传染性的个人池。在高传输设置中， 像家庭接触者调查这样目标明确的方法只能捕获一小部分结核病例， 然而，一般人群的方法效率太低，不切实际。需要新的病例发现方法， 通过在高流行率和高传播率的亚群中进行有针对性的筛查，提高诊断率。 在低发病率环境中，标准绘图工具已被用于确定目标人群，以加强 寻找病例。类似的方法在地方性环境中是否足够尚不清楚，这对于推进新的 个案调查方法。为了制定适当的战略，我们必须首先了解机制， 包括亚临床结核在内的社区结核病传播的空间模式。空间和基因组研究进展 统计建模与敏感诊断相结合，现在能够对空间靶向结核病筛查进行评估 在高负担的社区。我们假设，传播热点窝藏大量的个人， 未确诊和亚临床结核病，当有针对性时，可以提高结核病病例发现的效率。目标1： 确定在空间组织的热点内发生的TB传播的比例。在目标2中，我们测试 空间定位的病例发现是否比更广泛的方法更有效和更高效， 活动性和亚临床流行性结核病。为了实现我们的目标，我们将创新的空间统计 用贝叶斯随机动态方法建模，使用全基因组测序数据推断结核病传播， 并使用新的RNA生物标志物和Xpert Ultra结合胸部X光检查来检测社区中流行的结核病。 如果未被发现的结核病流行，包括亚临床形式，事实上，集中在传播地点， 将对有针对性的社区结核病筛查策略产生重要和实际的影响， 及早发现感染者，及早开始结核病治疗以阻断传播。

Project Summary Despite renewed public health efforts, including more effective treatment, tuberculosis (TB) incidence has reduced only incrementally, an effect driven by the inability to contain community TB transmission. Early identification and treatment of infectious individuals is central to breaking the chain of transmission and is limited by the fact that up to 40% of incident TB cases remain undiagnosed. The associated prolonged duration of infectiousness and delays in treatment initiation contributes significantly to ongoing TB transmission. Undiagnosed cases comprise diseased individuals who have been missed by the healthcare system and those without symptoms (subclinical TB) where the ability to transmit TB is unknown. In our preliminary data, using active case finding and whole blood RNA biomarker, we identified subclinical TB disease at proportions that approach or exceed that of symptomatic active TB. These cases were associated with the presence of viable bacilli in the sputum, pointing to a large potentially infectious pool of individuals. In high-transmission settings, highly targeted approaches like household contact investigation will capture only a small proportion of TB cases, yet general-population approaches are too inefficient to be practical. New case finding methods are needed that increase diagnostic yield through targeted screening in high-prevalence and high-transmission subpopulations. In low-incidence settings, standard mapping tools have been used to identify target populations for enhanced case-finding. Whether similar methods are sufficient in endemic settings is unknown and critical to advance new case-finding approaches. To develop appropriate strategies, we must first understand the mechanisms and spatial patterns of community-level TB transmission that include subclinical TB. Advances in spatial and genomic statistical modeling coupled with sensitive diagnostics now enable evaluation of spatially targeted TB screening in high-burden communities. We hypothesize that transmission hotspots harbor large number of individuals with undiagnosed and subclinical TB that when targeted can improve efficiency of TB case finding. In Aim 1, we determine the proportion of TB transmission that occur within spatially organized hotspots. In Aim 2, we test whether spatially targeted case-finding will be more effective and efficient than broader approaches for identifying active and subclinical prevalent TB. To accomplish our aims, we incorporate innovative spatial statistical modeling with Bayesian phylodynamic methods to infer TB transmission using whole genome sequencing data, and use novel RNA biomarker and Xpert Ultra with chest radiography to detect prevalent TB in the community. If undetected prevalent TB, including subclinical forms are, in fact, concentrated in locales of transmission, this would have important and practical implications for targeted community TB screening strategies as a means to identify infectious individuals early and interrupt transmission by early initiation of TB treatment.