Pulmonary Surface Irregularity score: A Quantitative CT Biomarker for Idiopathic Pulmonary Fibrosis

项目概要： 特发性肺纤维化（IPF）是最常见的纤维化肺病，每年有50，000例新发病例。 在美国的一年。IPF是一种快速进展的致死性疾病，半数患者在2年内死亡 年这种疾病是异质性的，对现有抗纤维化药物的反应性存在很大差异。 治疗IPF和其他形式的肺纤维化的新治疗剂的开发受到以下因素的阻碍： 缺乏外部验证的定量生物标志物来对患者表型进行分层， 对治疗的反应，并作为替代终点。因此，迫切需要制定定量的 生物标志物，准确预测预后和评估疾病活动，以指导管理计划， 这些病人。目前的生物标志物包括肺功能测试、复合生物标志物和患者- 相关结果不足以指导临床试验和临床实践。高分辨率计算机 断层扫描（CT）常规用于所有IPF患者，并且是一种可应用的定量CT生物标志物 将避免额外的患者成本和辐射。目前的CT生物标志物是基于 肺混浊的主观视觉评估或量化，并且尚未成功地用于 评价对治疗的纵向反应或作为临床试验的替代终点。一种新的定量CT 需要生物标志物。在IPF和其他形式的肺纤维化患者中， 肺表面不规则性（PSI）进行性恶化。我们开发了一种定量CT生物标志物， 在高分辨率CT图像上测量PSI，以生成以十分之一毫米为单位的PSI分数。PSI评分是 独立于肺阴影，是无移植生存期的预测指标 问题研究我们建议外部验证PSI分数使用的数据和高分辨率CT从 肺纤维化基金会（PFF）患者登记处，一个大型前瞻性多中心数据库， 收集基线临床数据，纵向患者相关结局和生存数据， 美国IPF患者（N=1200）和其他形式的肺纤维化患者（N=603）。更具体地说，我们的目标是 验证基线高分辨率CT图像上PSI评分预测无移植的准确性 IPF患者的生存期和肺功能检查的纵向变化以及患者相关结局 和其他形式的肺纤维化。我们假设PSI评分可以预测 无移植存活率和肺功能参数的纵向变化以及患者相关 IPF和其他形式肺纤维化患者的结局。我们还致力于开发一种全自动的 PSI评分，并根据半自动PSI评分进行验证。我们假设全自动的 PSI评分将具有高度相关性，并且与半自动方法具有相当的准确性。本研究将 将PSI软件推向FDA许可和完全自动化，用于临床试验和临床实践， 从而加速了用于IPF和其他形式的肺纤维化的抗纤维化药物的开发。

PROJECT SUMMARY: Idiopathic pulmonary fibrosis (IPF) is the most common fibrotic lung disease, with 50,000 new cases per year in the United States. IPF if a rapidly progressive and fatal disease, with half of patients dying within 2 years. The disease is heterogeneous, and there is wide variability in responsiveness to existing antifibrotic therapies. Development of new therapeutic agents for IPF and other forms of pulmonary fibrosis is hindered by a lack of externally validated quantitative biomarkers to stratify patient phenotype, evaluate longitudinal response to therapy, and serve as a surrogate endpoint. Thus, there is an urgent need to develop quantitative biomarkers that accurately predict prognosis and assess disease activity to guide management plans among these patients. Current biomarkers including pulmonary function tests, composite biomarkers, and patient- related outcomes are insufficient for guiding clinical trials and clinical practice. High-resolution computed tomography (CT) is routinely used in all patients with IPF, and a quantitative CT biomarker that can be applied to existing images would avoid additional patient cost and radiation. Current CT biomarkers are based on subjective visual assessment or quantification of lung opacities and have not been successfully used to evaluate longitudinal response to therapy or as a surrogate endpoint for clinical trials. A new quantitative CT biomarker is needed. In patients with IPF and other forms of pulmonary fibrosis, subpleural fibrotic scars result in progressive worsening of pulmonary surface irregularity (PSI). We developed a quantitative CT biomarker to measure PSI on high-resolution CT images to generate a PSI score in tenths of a millimeter. The PSI score is independent of lung opacities and is prognostic of transplant-free survival in pilot single-institution retrospective studies. We propose to externally validate the PSI score using data and high-resolution CTs from the Pulmonary Fibrosis Foundation (PFF) patient registry, a large prospective multicenter database that has collected baseline clinical data, longitudinal patient-related outcomes, and survival data from patients in the United States with IPF (N=1200) and other forms of pulmonary fibrosis (N=603). More specifically, we aim to validate the accuracy of the PSI score on the baseline high-resolution CT images for predicting transplant-free survival and longitudinal changes in pulmonary function test and patient-related outcomes in patients with IPF and other forms of pulmonary fibrosis in the PFF patient registry. We hypothesize that the PSI score will predict transplant-free survival and longitudinal changes in pulmonary functional parameters and patient-related outcomes in patients with IPF and other forms of pulmonary fibrosis. We also aim to develop a fully automated PSI score and validate this against the semi-automated PSI score. We hypothesize that the fully automated PSI score will have high correlation and comparable accuracy to the semi-automated method. This study will move the PSI software towards FDA clearance and full automation for use in clinical trials and clinical practice, thereby speeding up anti-fibrotic drug development for IPF and other forms of pulmonary fibrosis.