Quantitative Feature Change of Pulmonary Subsolid Nodule on CT Image to Determine Its Invasiveness

Aim: To quantitatively evaluate the changes of the non-solid components in pulmonary subsolid nodules (SSN) between baseline and follow-up CT scan in order to identity SSN with invasive manifestations, i.e., minimally invasive adenocarcinoma (MIA) and invasive adenocarcinoma (IAC). Materials and methods: Patients with baseline and follow-up CT scans, who had one or more SSNs and histological results, were retrospectively included. The measurement parameters were the maximum diameter of the whole SSN, the mean, the maximum and the minimum CT value, and the standard deviation of CT value of the non-solid component. Time dependent linear regression were used to evaluate the progression of nodule measurements from the baseline to follow-up scan. The diagnostic performance for SSN with invasive manifestations was evaluated by using the receiver operating characteristic curve and the area under the curve (AUC). Results: 71 patients (32 males and 39 females) with 78 SSNs were included. Nineteen SSNs were benign, ten adenocarcinoma in situ (AIS), 36 MIA and 13 IAC. The diameter of all SSNs positively correlated with time (all P<0.05). The standard deviation of CT values of the non-solid components with invasive manifestations positively correlated with time (P=0.37), but the other three CT values did not correlate with time (all P>0.05). In benign and AIS, all four CT value parameters had no correlation with time (all P>0.05). Annual changes of the standard deviation of CT value significantly determined invasive SSN (P=0.014). The sensitivity, specificity and AUC were 68.7%, 77.8% and 0.747, respectively. Conclusion: The diameter of benign, AIS, MIA and IAC increases during follow-up. The changes of the standard deviation of CT value may be used to differentiate and diagnose SSN with invasive manifestations (MIA and IAC).