Trajectories and Predictors in the Clinical High Risk for Psychosis Population: Prediction Scientific Global Consortium (PRESCIENT)

项目摘要/摘要 精神病通常首先出现在年轻人身上，导致广泛的痛苦、长期的残疾、过早死亡和巨大的经济负担。及早干预是减轻这一负担的重要战略。精神障碍之前会有一段痛苦、功能受损和阈值以下症状的前驱期。我们最初的研究从操作性上定义了临床高风险(CHR)状态，该状态预测早期精神病的风险大幅增加。在慢性阻塞性肺疾病人群中，临床轨迹有很大的异质性。目前，该领域无法在早期可靠地识别这些轨迹，特别是在单个患者层面上。到目前为止，这些模型(使用临床、神经认知、神经成像、神经生物学和遗传数据)对转化为精神障碍和其他结果只产生了适度的预测价值。这对有针对性的干预开发和开发强大的病因学模型提出了挑战。目前的项目旨在为慢性再生障碍性疾病人群中的一系列结果(转化为精神病障碍、持续性和偶发性非精神病障碍、慢性再生障碍性贫血状态不缓解、持续性阴性症状、完全恢复、功能结果)开发更可靠的预测模型，并引入经过验证的工具用于临床实践。这些预测模型和相关的临床工具将使用由生物标志物(神经成像、神经认知、神经生理学、生物谱学)、临床数据和数字瞬时评估组成的多模式数据来开发。这些预测模型将有助于选择慢性阻塞性肺疾病患者参加临床试验，作为早期治疗效果的衡量标准，并监测疾病进展以及临床和功能结果。该项目以四大支柱为基础： 1.现有的全国性临床基础设施(网络)，以支持在短时间内(2年招募，2年随访)招募和跟踪大量CHR青年队列(n=1000)，以及临床对照组(n=300)。 2.使用该数据集：验证现有和即将建立的预测模型，并利用最新的方法学进步和探索性生物标志物开发新的、更精细的预测模型。 3.在各国际中心招募一个独立的慢性循环样本，对澳大利亚网络中产生的模型进行外部验证，以确保结论的普遍性。或者，可以将这些站点用作网络中的额外分支，并将替代数据集用于外部验证目的(见2.5.4)。这一网站和研究专业化网络将为根据当前工作计划的结果在这一临床人群中进行未来的治疗试验提供临床研究基础设施。 4.作为气候变化研究领域先驱的独特记录和在最先进的预测建模方面的专门知识，包括开创新的预测方法(动态预测、多模式概率预测、网络理论)，以及使用数字技术支持必要数据的收集。我们在管理临床人群中的多站点研究网络以及在当地、全国和国际上快速招聘方面也拥有无与伦比的记录。

PROJECT SUMMARY/ABSTRACT Psychotic illnesses usually first emerge in young people and result in widespread suffering, protracted disability, premature death, and a huge economic burden. Early intervention represents a vital strategy to reduce this burden. Psychotic disorders are preceded by a prodromal period of distress, impaired functioning and subthreshold symptomatology. Our original research operationally defined the Clinical High Risk (CHR) state, which predicts a substantially increased risk of incipient psychosis. There is substantial heterogeneity in clinical trajectories in the CHR population. The field is currently unable to reliably identify these trajectories early on, particularly on an individual patient level. The models to date (using clinical, neurocognitive, neuroimaging, neurobiological and genetic data) have yielded only modest predictive value for conversion to psychotic disorder and other outcomes. This presents a challenge for targeted intervention development and developing robust aetiological models. The current project seeks to develop more robust prediction models for a range of outcomes in the CHR population (conversion to psychotic disorder, persistent and incident non-psychotic disorder, non-remission of CHR status, persistent negative symptoms, full recovery, functional outcome) and introduce validated tools for use in clinical practice. These prediction models and associated clinical tools will be developed using multimodal data consisting of biomarkers (neuroimaging, neurocognition, neurophysiology, biospecimens), clinical data, and digital momentary assessments. The prediction models will facilitate selection of CHR patients for enrolment in clinical trials, serve as measures of early treatment effects, and monitor disease progression and clinical and functional outcomes. The project is based on four pillars: 1. An existing nationwide clinical infrastructure (network) to support recruitment and follow up of a large cohort of CHR young people (n=1000) over a short timeframe (2 year recruitment period, 2 year follow up), as well as a clinical comparison group (n=300). 2. Use of this dataset to: validate existing and forthcoming prediction models and develop new, more refined prediction models using recent methodological advances and exploratory biomarkers. 3. Recruitment of an independent CHR sample across international centres for external validation of models generated in the Australian network to ensure generalizability of findings. Alternatively, these sites could be used as additional spokes in the network, with alternative data sets used for external validation purposes (see 2.5.4). This network of sites and research specialization will provide the clinical research infrastructure for future treatment trials in this clinical population informed by findings of the current program of work. 4. Unique track record as pioneers of the CHR field and expertise in state-of-the-art predictive modelling, including pioneering new approaches to prediction (dynamic prediction, multimodal probabilistic prediction, network theory), and use of digital technologies to support collection of requisite data. We also have unrivalled track record in management of multisite research networks in this clinical population and rapid recruitment locally, nationally, and internationally.