Imaging to predict therapeutic outcomes.
Imaging to predict therapeutic outcomes.
复制标题
影像学预测治疗结果。
DOI：
10.1164/rccm.201502-0285ed
复制
发表时间：
2015
影响因子：
24.7
通讯作者：
Washko,GeorgeR
中科院分区：
文献类型：
--
作者：
Washko,GeorgeR
肺减容手术提高了严重肺气肿患者的生存率(1)，但术中发病率和死亡率已转向微创手术，如“瓣膜”(2)、“密封剂”(3)和“蒸汽”(4)，以实现非切除术的内镜下肺减容（ELVR）。虽然这些研究都没有达到国家肺气肿治疗试验(1)的范围，但它们继续支持计算机断层扫描（CT）扫描可能是术前评估的关键组成部分的论点。在这一期的期刊中，Schuhmann和他的同事（第767-774页）(5)报道了通过利用自20世纪80年代引入密度测定法以来在信号处理方面取得的许多进步来扩展程序前评估的努力（6,7）。他们通过使用新的、临床相关的、基于图像的特征以及适当的假设来做到这一点，假设这些特征可以区分慢性阻塞性肺疾病的亚型，并改进我们预测治疗反应的能力。因此，研究人员进行了回顾性分析，定量评估了146名接受基于瓣膜的ELVR的患者的CT扫描所确定的多种特征。分析的目的是利用影像学来预测对干预的反应（反应的定义是在治疗后3个月靶肺叶体积减少至少350毫升），并在一部分受试者中比较定量CT与Chartis支气管内系统的预后表现（2,8）。预测基于瓣膜的治疗反应的一个主要挑战是肺叶内和叶间侧支通气的存在。虽然后者可以通过不完全的叶间裂隙检测到，但前者的评估需要有创设备，如Chartis支气管内系统（2,8）。为了从影像学上确定瓣内侧支通气，作者采用了一种由Mishima及其同事首次描述的技术(9)。三岛认识到，肺气肿的进展可能以实质中肺气肿簇的发展和合并为特征。那些患有轻度疾病的受试者有许多小簇，而那些患有严重疾病的受试者有较少的簇，但每个簇的大小都大得多。由于肺叶内侧支通气可能与组织破坏程度成正比，低衰减聚类（LAC）分析在生物学上是合理的，是一种利用密度测定法测定肺气肿百分比以外的新方法。另一类肺实质内特征越来越多地被探索为疾病亚型是肺血管形态。早期血管造影研究报告肺气肿患者的小肺血管变窄和缩小（10）。当我们将这一过程表示为肺或肺叶远端与总肺实质内血容量的比值时，结果证明，在预测临床相关结果方面，该分数与密度测定法具有高度的互补作用（11）。在目前的研究中，Schuhmann及其同事定义并应用了一种类似的方法，称为小血管容积相对于总血管容积，其中比例越高意味着疾病越严重(5)。
Lung volume reduction surgery improves survival in select patients with severe emphysema (1), but the periprocedure morbidity and mortality have shifted focus to minimally invasive procedures such as “valves”(2),“sealants”(3), and “vapor”(4) to achieve nonresectional endoscopic lung volume reduction (ELVR). Although none of these studies have approached the scope of the National Emphysema Treatment Trial (1), they continue to support the contention that computed tomography (CT) scan may be a key component to preprocedure evaluation.In this issue of the Journal, Schuhmann and colleagues (pp. 767–774)(5) report on efforts to expand on that preprocedure evaluation by leveraging the many advances made in signal processing since densitometry was introduced in the 1980s (6, 7). They do so by using new, clinically relevant, image-based features with the appropriate hypothesis that they may discriminate subtypes of chronic obstructive pulmonary disease and refine our ability to predict therapeutic response. The researchers therefore performed a retrospective analysis to quantitatively assess multiple features ascertained from the CT scans of 146 subjects who underwent valve-based ELVR. The goals of the analysis were to use imaging to predict response to intervention (where response was defined as a decrease in target lobe volume by at least 350 ml at 3 months after treatment) and in a subset of subjects to compare the prognostic performance of quantitative CT to the Chartis endobronchial system (2, 8). A major challenge in predicting response to valve-based therapies is the presence of both the intralobar and interlobar collateral ventilation in the lung. Although the latter may be detected by incomplete interlobar fissures, assessment of the former requires invasive devices such as the Chartis endobronchial system (2, 8). In an effort to radiologically ascertain intralobar collateral ventilation, the authors applied a technique first described by Mishima and colleagues (9). Mishima recognized that the progression of emphysema may be characterized by the development and coalescence of clusters of emphysema in the parenchyma. Those subjects with mild disease have many small clusters, whereas those with severe disease have fewer clusters, but each is much larger in size. Because intralobar collateral ventilation is likely proportional to the degree of tissue destruction, low attenuation cluster (LAC) analysis is biologically plausible and is a novel method to leverage densitometry beyond the percentage emphysema. Another category of intraparenchymal features increasingly being explored for disease subtyping is pulmonary vascular morphology. Early angiographic studies reported narrowing and reduction of the small pulmonary vessels in patients with emphysema (10). When we expressed this process as a ratio of distal to total intraparenchymal blood volume in the lung or lobe of interest, the resulting fraction proved to be highly complementary to densitometry in the prediction of clinically relevant outcomes (11). In this current work, Schuhmann and colleagues define and apply a similar method termed the small vessel volume relative to total vessel volume, whereby a higher ratio signifies more severe disease (5).