Predictors of Low-risk Phenotypes after Traumatic Brain Injury Incorporating Proteomic Biomarker Signatures

项目总结 在美国，创伤性脑损伤(TBI)是主要的死亡原因，治疗选择有限。 脑损伤的临床治疗试验结果令人失望，部分原因是会计上的困难 对于临床上重要的脑损伤内的异质性。颅脑损伤后早期给予治疗对减少 继发性脑损伤，但不加限制地治疗所有脑损伤可能是有害的。脑损伤刺激了 中枢和外周免疫反应的复杂级联反应。这些外周免疫 对脑损伤的反应可以作为早期风险表型的传感器，因为快速，容易测量 血液中的反应。提高入院时对患者进行风险分层的能力将简化患者选择 积极的干预措施--如侵入性神经监测--与选择那些有能力的患者 安全地被观察到减少潜在的危害。 霍莉·E·辛森医学博士是俄勒冈健康与科学大学的神经学家和神经强化专家。 她照顾严重急性脑损伤的病人。这个应用程序的目标是开发有监督的 脑损伤后可操作的短期和长期结果的学习模型，并在预先指定的情况下进行讯问 免疫调节蛋白仅在临床特征上就为模型增加了预测能力。她的中枢 假说是免疫调节蛋白质组特征提高了我们对低风险临床进行分类的能力 颅脑损伤后的表型。辛森博士的初步数据表明，外周细胞因子水平与 颅脑损伤后可操作的临床事件。该项目使用了一种高度敏感的单分子 免疫阵列(SiMoA)检测与无偏倚蛋白质组互补的免疫调节蛋白 利用全球发现质谱学的方法。她将开发和评估一系列模型 结合蛋白质组特征进行分类：急性进行性颅内出血(目标1A)，急性 神经恶化(目标1B)和6个月格拉斯哥结果量表测量的长期结果 (目标2)。她将在定义明确的临床试验人群(开发集)中开发这些模型，并进行测试 他们在OHSU(测试集)的一个独立的、预期登记的队列中正确分类结果的能力。 在一个由专家导师组成的多学科团队下，该项目将对低风险表型产生新的见解 急性颅脑损伤后的认知和结果分类。拟议的以患者为导向的研究项目将是 在预测建模和患者表型分型的原则上通过结构化的教学计划得到加强 (包括蛋白质组学)，这将为辛森博士提供她需要进行的关键技能 神经创伤领域的独立、创新的转化性临床研究。

PROJECT SUMMARY Traumatic brain injury (TBI) is a leading cause of death in the US, and treatment options are limited. Therapeutic clinical trials in TBI have yielded disappointing results owing in part to the difficulty in accounting for clinically important heterogeneity within TBI. Early delivery of therapy is essential after TBI to reduce secondary brain injury, but unrestricted treatment of all brain injuries could be harmful. TBI stimulates a complex cascade of immunologic responses, both centrally and peripherally. These peripheral immune responses to TBI could serve as an early sensor of risk phenotype given the rapid, readily measurable response in the blood. An improved ability to risk-stratify patients on admission will streamline patient selection for aggressive interventions—such as invasive neuromonitoring—versus selection of those patients who can safely be observed reducing potential harms. Holly E Hinson, MD MCR is a Neurologist and Neurointensivist at Oregon Health and Science University where she cares for patients with severe acute brain injury. The objective of this application is to develop supervised learning models of actionable short- and long-term outcomes post-TBI and to interrogate if pre-specified immunoregulatory proteins add predictive power to the models over clinical features alone. Her central hypothesis is that immunoregulatory proteomic signatures improve our ability to classify a low-risk clinical phenotype after TBI. Dr. Hinson’s preliminary data suggest peripheral cytokine levels are associated with actionable clinical events acutely after TBI. The project employs a highly-sensitive, single molecule immunoarray (SIMOA) to detect immunoregulatory proteins complemented with an unbiased proteomic approach utilizing global discovery mass spectrometry. She will develop and assess a series of models incorporating proteomic signatures to classify: acute progressive intracranial hemorrhage (Aim 1A), acute neurologic deterioration (Aim 1B), and long-term outcomes measured by the 6-month Glasgow Outcome Scale (Aim 2). She will develop these models in a well-defined, clinical trial population (development set), and test their ability to correctly classify outcome in an independent, prospectively enrolled cohort at OHSU (test set). Under a multidisciplinary team of expert mentors, the project will generate new insights into low-risk phenotype recognition and outcome classification after acute TBI. The proposed patient-oriented research project will be enhanced by a structured didactic program in the principles of predictive modeling and patient phenotyping (including proteomics), which will provide Dr. Hinson with the critical skills she will need to conduct independent, innovative translational clinical research in the field of neurotrauma.