The clinical value of predicting lymphovascular invasion in patients with invasive lung adenocarcinoma based on the intratumoral and peritumoral CT radiomics models.
Authors
Affiliations (3)
Affiliations (3)
- Department of Radiology, The First Affiliated Hospital of Guangxi Medical University, No. 06 Shuangyong Rd, Nanning, 530021, People's Republic of China.
- Department of Radiology, The People's Hospital of Guangxi Zhuang Autonomous Region, Guangxi Academy of Medical Sciences, No. 6 Taoyuan Road, Nanning, Guangxi, 530021, People's Republic of China.
- Department of Radiology, The First Affiliated Hospital of Guangxi Medical University, No. 06 Shuangyong Rd, Nanning, 530021, People's Republic of China. [email protected].
Abstract
Lung cancer remains the leading cause of cancer-related deaths worldwide, and lymphovascular invasion (LVI) is an important pathological indicator affecting the prognosis of lung cancer patients. Traditional imaging techniques face challenges in effectively and accurately predicting vascular invasion, but integrating clinical indicators with radiomics features is expected to improve the non-invasive preoperative prediction of LVI, providing valuable reference for clinical treatment decisions. This study aimed to investigate the clinical value of predicting LVI in patients with invasive lung adenocarcinoma (LUAD) based on the intratumoral and peritumoral CT radiomics models. The 384 patients with invasive LUAD from Institution 1 were randomly divided into training (<i>n</i> = 268) and internal validation (<i>n</i> = 116) sets with a ratio of 7:3, and 251 patients from Institution 2 were used as the external validation set. Altogether, 1226 features were extracted from the tumor gross (GT), gross tumor and peritumor (GPT), and peritumor(PT), respectively. Clinical independent predictors for LVI in patients with invasive LUAD were screened using univariate and multivariate logistic regression analysis, a combined model that included clinical predictors and optimal Radscore was constructed, and a nomogram was drawn. All cases were diagnosed using histopathological examination results as the gold standard. The GPT model showed better predictive efficacy than the GT and PT models, with the area under the curve (AUC) of 0.83, 0.79, and 0.75 in the training, internal validation, and external validation sets, respectively. In the clinical model, the preoperative carcinoembryonic antigen (CEA) level, tumor diameter, and spiculation were the independent predictors. The combined model containing the independent predictors and the GPT-Radscore significantly predicted LVI in patients with invasive LUAD, with AUCs of 0.84, 0.82, and 0.77 in the three cohorts, respectively. The CT scan-based radiomics model which including intratumoral and peritumoral radiomics features could effectively predict LVI in LUAD patients, and the predictive efficacy were further improved by combining clinically independent predictors. This study holded significant clinical importance, as it provided a non-invasive biomarker for the preoperative prediction of LVI status in lung cancer patients, thereby identifying subgroups with poor prognosis. It demonstrated great potential for risk stratification and guiding personalized treatment strategies in clinical practice.