Discovery of the Biomarker Signature for Neuropathic Corneal Pain

项目摘要 神经性角膜疼痛(NCP)是眼部神经性疼痛的一种。它会导致患者有严重的 不适和严重损害的生活质量(QOL)。缺乏标准观察到的迹象 检查导致误诊为干眼病(DED)，导致对 医疗保健基金。NCP诊断标志物的鉴定及检测方法的建立 这种方法可以使患者得到充分和及时的治疗，提高患者的生活质量，减少医疗保健 系统的财政负担。可以使用激光活体共聚焦显微镜进行光学活组织检查 (IVCM)，它允许在准组织学水平上显示角膜基底部下神经。初步 数据显示，IVCM发现了微神经瘤(由积聚引起的切断神经末端的球状突起 分子成分)在NCP患者中存在，但在DED患者中不存在。我们建议验证微神经瘤 作为一种新的NCP生物标志物。在目标1中，我们将使用超过2,000个DED/NCP受试者和超过 500,000个IVCM图像，以确认微神经瘤的存在是NCP的合适生物标志物 微神经瘤诊断NCP患者的敏感性和特异性与其他IVCM的比较 参数。三名观察员将分别对这一确认的生物标志物的图像进行两次评分，以评估相互间和 观察员内部的精确度和IVCM数据集的描述性统计将有助于确定 微神经瘤检测的高精度所需的最小图像数量。AIM 2将提供 微神经瘤的生物学验证。眼部疼痛的强度和对生活质量的影响 由眼痛评估调查(OPAS)评估的眼痛将在 有微小黑色素瘤的和没有微黑素瘤的。此外，眼部疼痛/不适的变化对 在患有微神经瘤的患者和患有微神经瘤的患者之间，向眼睛内滴注高渗盐水将进行比较 那些没有的。在目标3中，我们将为自动化开发一个经过验证的人工智能(AI)程序 识别微神经瘤，以使临床医生能够快速和广泛地采用。程序的准确性 将通过评估AI计划对IVCM图像的评估与 评估2名观察员。类似的准确性评估将使用从以下位置获得的图像进行评估 一个独立的站点，以便可以评估站点间的精度。人工智能项目还将接受评估，以满足其 NCP诊断的特异性和敏感性。目标4将建立微神经瘤的临床应用 在一项前瞻性的多中心研究中，IVCM观察到NCP的生物标志物。生物标志物的精确度， 参考间隔、站点间性能的协调以及灵敏度和特异度 NCP诊断的概率将使用这一前瞻性队列来确定。下一步，微神经瘤的发现将是 与OPAS和高渗功能试验相关，以进行生物学验证。最后，人工智能程序的 将使用这项研究的IVCM图像来测试提供NCP诊断的能力。

PROJECT ABSTRACT Neuropathic corneal pain (NCP) is an ocular type of neuropathic pain. It causes patients to have severe discomfort and a severely compromised quality of life (QoL). The lack of signs observed by standard examination has resulted in misdiagnosis as dry eye disease (DED) resulting in an inefficient use of healthcare funds. The identification of a diagnostic biomarker for NCP and development of a detection method would allow adequate and timely treatment, improve patients’ QoL, and decrease the health care system’s financial burden. An optical biopsy can be performed using laser in vivo confocal microscopy (IVCM), which allows for visualization of subbasal corneal nerves at a quasi-histological level. Preliminary data has shown that IVCM identified microneuromas (a bulb at the end of a severed nerve caused by build-up of molecular constituents) are present in NCP, but not DED, patients. We propose to validate microneuromas as a novel biomarker for NCP. In Aim 1 we will use our database of over 2,000 DED/NCP subjects and over 500,000 IVCM images to confirm that the presence of microneuromas is an appropriate biomarker for NCP by comparing the sensitivity and specificity of identification of NCP patients via microneuromas to other IVCM parameters. Three observers will each grade images twice for this confirmed biomarker to assess inter- and intra-observer precision, and descriptive statistics of the IVCM datasets will allow for determination of the minimum number of images necessary for high precision of microneuroma detection. Aim 2 will provide biological validation of microneuromas. Both the intensity of ocular pain and the compromise to QoL caused by ocular pain as assessed by the Ocular Pain Assessment Survey (OPAS) will be compared between those with microneuromas and those without. Further, the change in ocular pain/discomfort in response to instillation of hyperosmolar saline into the eyes will be compared between those with microneuromas and those without. In Aim 3 we will develop a validated artificial intelligence (AI) program for automated identification of microneuromas to allow rapid and wide-scale adoption by clinicians. Accuracy of the program will be determined by evaluating the agreement of the AI program’s assessment of IVCM images with the assessment of 2 observers. A similar assessment of accuracy will be assessed using images obtained from an independent site so that inter-site precision can be evaluated. The AI program will also be assessed for its specificity and sensitivity in NCP identification. Aim 4 will establish the clinical utility of microneuromas observed by IVCM as a biomarker for NCP in a prospective, multi-center study. The biomarker’s precision, reference intervals, and harmonization of performance between sites as well as the sensitivity and specificity of NCP diagnosis will be determined using this prospective cohort. Next, the microneuroma findings will be correlated with the OPAS and hyperosmolar functional tests for biological validation. Finally, the AI program’s ability to provide a diagnosis of NCP will be tested using the IVCM images from this study.