置信区间
危险系数
优势比
肺
腺癌
内科学
医学
表皮生长因子受体
病理
胃肠病学
肿瘤科
癌症
作者
Yang Zhang,Shengping Wang,Hui Zhu,Yuan Li,Yang Zhang
标识
DOI:10.1093/ejcts/ezac297
摘要
Abstract OBJECTIVES We comprehensively investigated the morphology patterns of lung cancers associated with cystic airspaces. Our goal was to determine the predictive value of imaging features in a clinical environment. METHODS We collected information about patients with resected lung cancers associated with cystic airspaces from January 2010 to December 2019. Radiological features, clinicopathological characteristics, gene mutations and survival data were analysed comprehensively. RESULTS A total of 384 resected lung cancers associated with cystic airspaces were identified and categorized as 4 types: I, thin-wall type (n = 31, 8.1%); II, thick-wall type (n = 113, 29.4%); III, cystic airspace with a nodule type (n = 162, 42.1%) and IV, mixed type (n = 78, 20.3%). There were 27 (7.0%) adenocarcinomas in situ/minimally invasive adenocarcinomas; 237 (61.7%) lung adenocarcinomas; 115 (29.9%) squamous cell carcinomas; and 5 (1.3%) other tumours. The epidermal growth factor receptor mutation rate for type III was the highest (68.4%, P = 0.004). Pre-/minimally invasive adenocarcinomas were commonly featured as thin, pure ground-glass wall-surrounded cystic airspaces with smooth inner surfaces and margins. For patients with lung adenocarcinomas associated with cystic airspaces, type III (odds ratio 2.10; 95% confidence interval 0.55–8.06; P = 0.028) was an independent factor associated with a worse differentiation level. Type I was associated with excellent survival and type II, with the worst prognosis (P < 0.001). Type II (hazard ratio 2.29; 95% confidence interval 1.30–4.04; P = 0.004) was an independent prognostic factor for overall survival. CONCLUSIONS Morphological patterns could be predictors for gene mutations, invasive status, pathological differentiation and postoperative prognosis for lung adenocarcinomas associated with cystic airspaces.
科研通智能强力驱动
Strongly Powered by AbleSci AI