Shi Jinghan, Li Fei, Yang Fujun, Dong Zhengwei, Jiang Yan, Nachira Dania, Chalubinska-Fendler Justyna, Sio Terence T, Kawaguchi Yo, Takizawa Hiromitsu, Song Xiao, Hu Yang, Duan Liang
Department of Endoscopy, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, China.
Department of Radiology, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, China.
Transl Lung Cancer Res. 2021 Nov;10(11):4266-4280. doi: 10.21037/tlcr-21-896.
Visceral pleural invasion (VPI) is a clinical manifestation associated with a poor prognosis, and diagnosing it preoperatively is highly imperative for successful sublobar resection of these peripheral tumors. We evaluated the roles of computed tomography (CT) features and circulating tumor cells (CTCs) for improving VPI detection in patients with clinical T1N0M0 invasive lung adenocarcinoma.
Three hundred and ninety-one patients were reviewed retrospectively in this study, of which 234 presented with a pleural tag or pleural contact on CT images. CTCs positive for the foliate receptors were enriched and analyzed prior to surgery. Logistic regression analyses were performed to assess the association of CT features and CTCs with VPI, and the receiver operating characteristic (ROC) curve was generated to compare the predictive power of these variables.
Patients mostly underwent either segmentectomies (18.9%) or lobectomies (79.0%). Only 49 of the 234 patients with pleural involvement on CT showed pathologically confirmed VPI. Multivariate logistic regression analysis revealed that CTC level ≥10.42 FU/3 mL was a significant VPI risk factor for invasive adenocarcinoma cases ≤30 mm [adjusted odds ratio (OR) =4.62, 95% confidence interval (CI): 2.05-10.44, P<0.001]. Based on CT features, subgroup analyses showed that the solid portion size was a statistically significant independent predictor of VPI for these peripheral nodules with pleural tag, while the solid portion length of the interface was an independent predictor of pleural contact. The receiver operating curve analyses showed that the combination of CTC and CT features were highly predictive of VPI [area under the curve (AUC) =0.921 for pleural contact and 0.862 for the pleural tag, respectively].
CTC, combined with CT features of pleural tag or pleural contact, could significantly improve VPI detection in invasive lung adenocarcinomas at clinical T1N0M0 stage prior to the patient's surgery.
脏层胸膜侵犯(VPI)是一种与预后不良相关的临床表现,术前诊断对于这些周围型肿瘤成功进行亚肺叶切除至关重要。我们评估了计算机断层扫描(CT)特征和循环肿瘤细胞(CTC)在改善临床T1N0M0期侵袭性肺腺癌患者VPI检测中的作用。
本研究回顾性分析了391例患者,其中234例在CT图像上表现为胸膜结节或胸膜接触。术前对叶酸受体阳性的CTC进行富集和分析。进行逻辑回归分析以评估CT特征和CTC与VPI的相关性,并生成受试者工作特征(ROC)曲线以比较这些变量的预测能力。
患者大多接受了肺段切除术(18.9%)或肺叶切除术(79.0%)。在234例CT显示胸膜受累的患者中,只有49例经病理证实为VPI。多因素逻辑回归分析显示,对于直径≤30 mm的侵袭性腺癌病例,CTC水平≥10.42 FU/3 mL是显著的VPI危险因素[调整优势比(OR)=4.62,95%置信区间(CI):2.05-10.44,P<0.001]。基于CT特征的亚组分析表明,实性部分大小是这些有胸膜结节的周围型结节VPI的统计学显著独立预测因素,而界面的实性部分长度是胸膜接触的独立预测因素。受试者工作曲线分析表明,CTC与CT特征的组合对VPI具有高度预测性[胸膜接触的曲线下面积(AUC)=0.921,胸膜结节的AUC=0.862]。
CTC结合胸膜结节或胸膜接触的CT特征,可显著提高临床T1N0M0期侵袭性肺腺癌患者术前VPI的检测率。