Development of a Precision Medicine-based Diagnostic Tool for Membranous Nephropathy

项目摘要 该计划的目标是启动基于质谱（MS）的工作流程的商业化， 膜性肾病（MN）是肾病综合征的主要原因， 导致三分之一的患者出现肾衰竭自20世纪50年代以来，大多数形式的MN被认为是特发性的， 起源，然而，基于近年来的大量出版物，现在广泛接受MN是由 通过自身抗体对抗大约20种不同的内源性人类蛋白质。最常见的MN 自身抗原是磷脂酶A2受体（PLA2R）， 这种蛋白已被证明是监测疾病进展、严重程度和 缓解（1 - 3）。Arkana和其他人已经在MN病例中鉴定了另外的新的自身抗原（4 - 9），并给出了 PLA2R血清学检测用于个性化治疗的成功，表征和临床验证 剩余的MN自身抗原是迫切需要的。此外，对阳性患者的随访调查 因为这些新发现的自身抗原与危及生命的合并症有关， 特异性自身免疫性疾病因为根据MN自身抗原对患者进行分类可能会发现严重的 基础疾病，PLA2R血清学只能识别70%的原发病例，Arkana已经与 蛋白质组学专家在阿肯色州大学医学科学（UAMS）开发一个公正的，MS- 的方法对所有抗原类型的MN进行分类。I期研究评估了近300例组织活检， 先前确定自身抗原包括PLA2R、血小板反应蛋白1型结构域的MN患者 含有7A（THSD 7A）（10）和exostosin 1/2（EXT 1/2）（11）的细胞，还包括三阴性病例。后 评估一系列计算，统计和人工智能数据分析模式，排名Z- 分数被发现是自身抗原分类的有效度量，正确分类超过95%的 样品在拟议的第二阶段计划中，Arkana将继续与UAMS合作， 临床实践和商业可用性的有力方法。质量控制标准将为 纳入/排除进样，以确保MS分析和支持的可靠性和可重复性 自身抗原分类的验证。分析管道也将扩大， 另外的自身抗原靶点，其中许多是通过Arkana最近对数百个 生物库MN样品。最后，该方法将被转移到Arkana的CLIA实验室，并在 与UAMS的比较研究，包括约300个盲态样本，代表了 MN自身抗原分类。将此工作流程作为商业服务推出，不仅可以为临床医生提供 对于MN患者的诊断和治疗，前所未有的细节，但也揭示了潜在的危及生命的 前代策略无法检测到的合并症。

Project Summary The goal of this program is to initiate commercialization of a mass spectrometry (MS)-based workflow for the diagnosis and classification of membranous nephropathy (MN), a leading cause of nephrotic syndrome, leading to kidney failure in a third of patients. Since the 1950’s, most forms of MN were considered idiopathic in origin, however, based on numerous publications in recent years, it is now broadly accepted that MN is caused by autoantibodies against approximately 20 different endogenous human proteins. The most common MN autoantigen is the phospholipase A2 receptor (PLA2R), and serological analysis of autoantibody titers against this protein has proven to be an indispensable serological tool for monitoring disease progression, severity, and remission (1-3). Arkana and others have identified additional novel autoantigens in MN cases (4-9), and given the success of PLA2R serological testing for personalized treatment, the characterization and clinical validation of the remaining MN autoantigens is urgently required. Furthermore, follow-up investigation of patients positive for these newly identified autoantigens have been linked to life-threatening comorbidities including cancer and specific autoimmune diseases. Because classifying patients based on MN autoantigen may reveal severe underlying conditions, and PLA2R serology will identify only 70% of primary cases, Arkana has partnered with proteomics experts at the University of Arkansas for Medical Sciences (UAMS) to develop an unbiased, MS- based approach to classifying MN of all antigen types. Phase I studies evaluated nearly 300 tissue biopsies from MN patients with previously determined autoantigens including PLA2R, thrombospondin type-1 domain containing 7A (THSD7A) (10), and exostosin1/2 (EXT1/2) (11), with triple negative cases also included. After evaluating a range of computational, statistical, and artificial intelligence data analysis modalities, ranked Z- scores were found to be an effective metric for autoantigen classification, correctly classifying over 95% of samples. During the proposed Phase II program, Arkana will continue collaborating with UAMS to translate this powerful method to clinical practice and commercial availability. Quality control metrics will be developed for the inclusion/exclusion of incoming samples, to ensure the reliability and repeatability of MS analysis and support verification of autoantigen classifications. The analytical pipeline will also be broadened to accommodate additional autoantigen targets, many of which were identified by Arkana’s recent MS analysis of hundreds of biobanked MN samples. Finally, the method will be transferred to Arkana’s CLIA laboratory and validated in a comparative study with UAMS comprising approximately 300 blinded samples representing the full breadth of MN autoantigen classes. Launching this workflow as a commercially available service will not only offer clinicians unprecedented detail for the diagnosis and treatment of MN patients, but also reveal potentially life-threatening comorbidities undetectable by prior generation strategies.