A nomogram to predict invasiveness in lung adenocarcinoma presenting as ground glass nodule.

ABSTRACT

BACKGROUND: Adenocarcinoma in situ, minimally invasive adenocarcinoma, and invasive adenocarcinoma (IA) can all appear as a ground glass opacity (GGO) on chest computed tomography (CT). However, their respective prognoses are considerably different. This study aimed to predict IA in radiological examinations of patients with GGO lesions. METHODS: We retrieved the clinical records and high-resolution CT (HRCT) images of 124 patients with GGO lesions, who underwent various lung resections between 2016 and 2017. Correlations between the imaging features of preoperative HRCT and the postoperative pathology were analyzed. Receiver-operating characteristic (ROC) curve analysis, chi-square test, and one-way analysis of variance and multiple logistic regression were performed. A nomogram was developed and analyzed using a multiple logistic model. RESULTS: The maximum sensitivity and specificity were obtained at a cutoff value of -410 Hounsfield units (HU) for the mean CT value (m-CT), 10 mm for the maximum tumor dimension (MTD), and 0.25 for the consolidation tumor ratio (CTR). Further, there were significant differences in MTD, CTR, margin characteristics, air bronchogram, pleural indentation, and multiple GGOs (P<0.05). The independent predictive factors of IA included MTD [risk ratio (RR), 5.047; P=0.018], air bronchogram or vacuole sign (RR, 4.054; P=0.025), pleural retraction (RR, 4.742; P=0.008), and m-CT value (RR, 5.874; P =0.005). The scoring nomogram model was as follows -3.50744 + 1.26374 × (MTD>10 mm=1) + 2.41978 × (m-CT value≥-410 HU=1) + 1.77779 × (with air bronchogram or vacuole sign=1) + 1.60913 × (with pleural retraction=1). The area under the ROC curve was 0.9. The cutoff score was -0.5502 with a sensitivity of 86.8% and a specificity of 78.9%. CONCLUSIONS: IA in patients with GGO lesions can be predicted by evaluating the MTD, m-CT value, air bronchogram, and pleural retraction on HRCT by using a nomogram model.