Association of postoperative recurrence with radiological and clinicopathological features in patients with stage IA–IIA lung adenocarcinoma. Issue 141 (August 2021)

• Record Type: Journal Article
• Authors: Zhang, Yanyan; Zhao, Fengnian; Wu, Minghao; Zhao, Yunqing; Liu, Ying; Li, Qian; Zhou, Guiming; Ye, Zhaoxiang
• Is Part Of: European journal of radiology. Issue 141(2021)
• Publication Date: 2021-08
• Subjects: NSCLC non-small cell lung cancer -- OS overall survival -- CT computed tomography -- PACS picture archiving and communication system -- GGO ground-glass opacity -- CI confidence intervals -- OR odds ratios -- ROC receiver operating characteristic -- AUC area under the curve -- NCCN National Comprehensive Cancer Network; Computed tomography -- Clinicopathological features -- Recurrence -- Node-negative early-stage lung adenocarcinomas; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.109802
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: Six radiological and two clinicopathological features were associated with 2-year relapse. Five radiological and two clinicopathological features were associated with 5-year relapse. Stage, solid texture, pleural attachment and bronchovascular bundle thickening could help predict postoperative relapse. The combined model may better identify high risk groups of stage IA-IIA postoperative recurrence. Abstract: Objectives: To retrospectively investigate whether radiological and clinicopathological characteristics were associated with the presence of stage IA–IIA lung adenocarcinoma in patients at high risk for a postoperative recurrence. Materials and Methods: Three hundred twelve patients with biopsy-proven node-negative early-stage (IA–IIA) lung adenocarcinoma met the inclusion criteria for this study. Demographics data and histopathological findings were collected from medical records. Computed tomography (CT) performed approximately 1 month before surgery was manually scored using 23 CT descriptors. Univariate analyses were applied to demonstrate an association between clinicopathological and radiological features and 2-/5-year recurrences. Multivariate logistic regression was performed to assess the ability of radiological and clinicopathological features to discriminate low and high-risk factors for recurrence. A ROC curve was used to evaluate prediction performance. Results: Univariate analysis revealed that the 2-year recurrence was associated with six radiological features and two clinicopathological features, while 5-year recurrence was associated with five radiological features and two clinicopathological features. A multivariate logistic regression model of combined clinicopathological and radiological features showed that stage IIA (OR = 2.87), solid texture (solid part > 50 %: OR = 4.81; solid part = 100 %: OR = 3.61), pleural attachment (OR = 3.97) and bronchovascular bundle thickening (OR = 2.16) were associated with the independent predictors of 2-year recurrence, and stage IIA (OR = 3.52), solid texture (solid part > 50 %: OR = 3.56; solid part = 100 %: OR = 2.44) and pleural attachment (OR = 4.57) were associated with 5-year recurrence. Combined radiological and clinicopathological features could be significant indicators of 2- and 5-year recurrences (AUC = 0.784 and AUC = 0.815, respectively). Conclusions: The combination of radiological and clinicopathological features has the potential to help predict postoperative recurrence in patients with stage IA–IIA lung adenocarcinomas and guide oncologists and patients whether to undergo additional treatment after surgery.