Diagnostic value of spectral CT multiparametric quantitative imaging in assessing pathological grading of invasive lung adenocarcinoma

ZHU Zirui; SHEN Leilei; LIU Mengqi; XUE Zhiqiang; CHEN Zhiye

doi:10.12122/j.issn.1674-4500.2024.04.14

Abstract：Objective To investigate the value of spectral CT multi-parametric quantitative imaging for evaluating the pathological grading of invasive lung adenocarcinoma(ILA).Methods A total of 61 patients with pathologically confirmed pulmonary invasive adenocarcinoma at Hainan Hospital of the PLA General Hospital from July 2019 to July 2022 were retrospectively enrolled,which preoperatively scanned by spectral enhanced CT.Patients were divided into two groups according to the pathological grading system of ILA proposed by the International Association for the Study of Lung Cancer in 2020,with grades G1 and G2 as the low-grade group(n=31)and grade G3 as the high-grade group(n=30).The clinical characteristics(gender,age,smoking history,stage)and morphological parameters(tumor size,lobulation sign,spicule sign,pleural depression sign,vacuole sign)were analyzed.Spectral parameters[iodine concentration(IC),normalised iodine concentration(NIC),effective atomic number(Zeff),single energy CT values(CT40keV,CT70keV),slope of the curve K values(40-70 keV)]in both the arterial phase and venous phase were measured.Multiphase software was used to construct extracellular volume fraction(ECV)maps to measure the ECV values of the lesions.Logistic regression analysis was used to construct the combined models of arterial phase parameters and venous phase parameters,which expressed in a nomogram.AUC was used to assess the performance of each parameter and model to identify the pathological grade of ILA.Decision curves analysis was used to assess the clinical net benefit of the models.Results The morphological parameters of ILA in the low-grade and high-grade groups were not substantially different(P>0.05).The IC,NIC,Zeff and ECV values of the low-grade group were higher than those of the high-grade group in both arterial and venous phases(P<0.05).ROC curve analysis revealed that the IC,NIC,Zeff,ECV in the arterial phase had better diagnostic efficacy than the venous phase,AUC of arterial phase:0.734-0.831,AUC of venous phase:0.702-0.749.The combined arterial phase parameter in the arterial phase had better diagnostic efficacy than the parameters alone,with AUCs of 0.855,83.3%,83.9%,respectively.Decision curves analysis demonstrated that the model of combined arterial phase parameters had higher clinical net benefit than the venous combined parameters model.Conclusion Spectral CT arterial phase combined parameter can be used as an imaging marker for pathology grading diagnosis of ILA,and spectral CT multiparametric quantitative imaging can be used as a simple,effective and non-invasive diagnostic tool for ILA diagnosis.