A total of 116 patients with persistent malignant pulmonary nodules (54 in the growing group and 62 in the stable group) were included in this study. The clinical characteristics of the patients in the 2 groups are listed in Table 1. Compared with the stable group, patients in the growing group include a higher fraction of males (53.7% in growing vs 35.5% in stable group, p = 0.061) and smokers (37.0% vs 19.4%, p = 0.038), and experience of surgical history (42.6% vs 22.6%, p = 0.028), and higher patient age (63.4 ± 9.0 in growing group vs 53.8 ± 10.5 in the stable group, p < 0.001). The tumor history (p = 0.753), tumor initial size (p = 0.428), and follow-up duration (p = 0.121) were similarly observed in the two groups. After several months of follow-up, all GGNs were confirmed by surgery experts as lung adenocarcinoma. The neoplasm stage (p < 0.001) and invasive carcinoma (p < 0.001) were significantly different as well between the two groups.

In 116 patients, lobulation sign, spiculation sign, pleural indentation, air bronchogram, vessel convergence, vacuole sign were observed in 45 (38.8%), 22 (18.9%), 24 (20.7%), 10 (8.6%), 59 (50.8%), and 41 (35.3%), respectively (Table 2). Compared with the stable group, air bronchogram (16.7% vs 1.6%, p = 0.006), vessel convergence sign (70.4% vs 33.9%, p < 0.001), vacuole sign (48.1% vs 24.2%, p = 0.011), and the median vessel diameter of the GGO nodules (1.25 vs 0, p < 0.001); were significant different. However, similar fractions of lobulation sign, spiculation sign, and pleural indentation were observed in the two groups. The diameter of microvascular and the solid component of GGN were significantly different (p < 0.001).

Effect of univariate and multivariate variables in patients with growing GGO in (Table 3).A univariable analysis demonstrated that growth of malignant pulmonary nodules was closely associated with patient age (odds ratio [OR], 1.108; 95% confidence interval [CI], 1.057–1.160; p < 0.001), gender (OR, 2.109; 95% CI, 1.000–4.448; p = 0.050), Smoking history (OR, 2.451; 95% CI, 11.060–5.665; p = 0.036), surgical history (OR, 2.544; 95% CI, 1.139–5.680; p = 0.023). According to the CT images, the vacuole sign (OR, 2.910; 95% CI, 1.322–6.405; p = 0.008), the tumor blood vessel diameter through pulmonary lesions (OR, 4.288; 95% CI, 2.307–7.971; p < 0.001), air bronchus sign (OR, 12.200; 95% CI, 1.492–99.785; p = 0.020), vascular convergence sign (OR, 4.673; 95% CI, 2.113–10.176; p < 0.001), and solid component (OR, 12.347; 95% CI, 4.737–32.185; p < 0.001) were also significantly different in the patients of two groups, indicating these are the risk factors. In a multivariable analysis, the patient age (OR, 1.065; 95% CI, 1.004–1.131; p = 0.037), the solid component of GGN (OR, 11.38; 95% CI, 3.004–43.107; p < 0.001), the vacuole sign (OR, 6.542; 95% CI, 1.943–22.025; p = 0.002), and the tumor blood vessel diameter through pulmonary nodules (OR, 2.933; 95% CI, 1.260–6.825; p = 0.013) were closely related to the growth of persistent malignant pulmonary nodules.

A ROC analysis was carried out to determine the dividing point of the patients’ age and the size of pulmonary nodules vessel (Table 4). A positive correlation was observed between the patient age and vessel diameter of both groups. Specifically, patient age (AUC 0.750, p < 0.001) and vessel diameter (AUC 0.785, p < 0.001) were significantly effective. Results showed that the age of 54-year old is crucial for the nodule growth The cut-off value between Growing group and Stable group patients of patient age score was found to be equal to 54.0 with a sensitivity of 0.889 and specificity of 0.548 (Fig. 3) and the vessel diameter cut-off point of 0.90 with sensitivity of 0.667 and specificity of 0.823 (Fig. 4). According to Kaplan–Meier analysis, patient age (p < 0.001; Fig. 5), solid component of GGN (p < 0.001; Fig. 6) and the diameter of vessel penetration in GGN (p < 0.001; Fig. 7) were significantly different as well between two groups.

The ROC analysis of patient age showing that the cut-off value was 54.0 with sensitivity of 0.889 and 1-specificity of 0.548

The ROC analysis of patient age showing that the vessel diameter cut-off point was 0.9 mm

The variation in the cumulative percentages of growing nodules of two patient age groups with the increasing follow-up time

Kaplan–Meier plot for time to nodule growth according to CTR (the maximum diameter of consolidation in tumor (C) and the maximum diameter of the whole tumor including GGO (T). The CTR means C/T ratio. Part-solid GGNs (CTR > 0.5) show significantly higher cumulative percentages of growth than GGNs with CTR < 0.5

Kaplan–Meier plot for time to nodule growth according to the vessel diameter entering the GGN. those with Vd ≥ 0.9 mm growth much faster than those with Vd < 0.9 mm

Because the solid component was the main risk factor for growth, the analysis of the subgroup was necessary for the GGO with a solid component ratio < 0.5 (Table 5). A univariable analysis demonstrated that growth of GGO was closely associated with patient age (odds ratio [OR], 1.082; 95% confidence interval [CI], 1.020–1.148; p = 0.009), For the CT images, the vacuole sign (OR, 4.333; 95% CI, 1.498–12.532; p = 0.007), the main tumor blood vessel diameter through GGO lesions (OR, 3.570; 95% CI, 1.571–8.108; p = 0.002). Multivariable analysis showed the vacuole sign (OR, 4.272; 95% CI, 1.266–14.420; p = 0.019), and the vessel diameter through pulmonary nodules (OR, 2.561; 95% CI, 1.116–5.879; p = 0.027) were the risk factors for the growth.