Value of morphological characteristics on CT scans in predicting lymphovascular invasion in patients with early-stage lung adenocarcinomas

ZHANG Lan; WANG Lixia; LI Yanpeng

doi:10.3969/j.issn.1002-7386.2025.11.009

Abstract：Objective To investigate the value of preoperative CT features in predicting lymphovascular invasion(LVI)in patients with early-stage lung adenocarcinomas.Methods Clinical data,preoperative CT images,and pathological data of 413 patients with stage Ⅰ-Ⅱ lung adenocarcinomas who underwent radical surgical resection at the First Affiliated Hospital of Hebei North University from January 2019 to December 2022 were retrospectively analyzed.There were 143 males and 270 females,with an average age of 59.81±9.79(31-82)years.According to the postoperative pathological determination of LVI,patients were divided into the LVI positive group(24 cases)and LVI negative group(389 cases).Using independent sample t-test,Mann Whitney U test,Chi-square test or Fisher's exact test,clinical data,CT imaging morphology,and pathological features between LVI positive and LVI negative groups were compared.Independent risk factors for LVI were identified by binary multivariate logistic regression analysis,and then,a prediction model was created and validated for its performance using the receiver operating characteristic(ROC)curve and calculating the area under the curve(AUC).DeLong test was performed to compare the performance in prediction models,and Hosmer-Lemeshow test was conducted to evaluate the model calibration.Results There were no significant differences between LVI positive and LVI negative groups in the age,body mass index(BMI),family history of tumors,smoking status,alcohol consumption,tumor location,tumor-lung interface,lobulation sign,vascular bundle sign,vacuole sign,air bronchogram sign,and occupation(P＞0.05).There were significant differences between the LVI positive group and the LVI negative group in the gender,maximum tumor diameter,tumor CT imaging type,hair prick sign,pleural indentation sign(P＜0.05).The statistically significant indicators from the above univariate analysis were included in the binary multivariate logistic regression analysis,and the results showed that gender,maximum tumor diameter,tumor CT imaging type were independent risk factors for LVI.Based on the above indicators,a maximum diameter model,an imaging model(tumor maximum diameter+tumor CT imaging type),and a combined model(gender+tumor maximum diameter+tumor CT imaging type)were constructed.The ROC results showed that the combined model had good predictive ability,with an AUC of 0.844,sensitivity of 79.2％,and specificity of 79.9％.The Hosmer Lemeshow test showed that the prediction model had a good calibration ability(P=0.822).Conclusion A prediction model based on preoperative CT imaging of tumor maximum diameter,patient gender,and tumor CT imaging type is helpful for predicting LVI status in early-stage lung adenocarcinoma patients.