Development of predictive models for assessing the progression and invasiveness of satellite lesions in patients with multiple pulmonary ground glass nodules.

ABSTRACT

Background: Currently, the appropriate treatment of satellite lesions is still controversial. With this study, we aimed to construct a set of nomograms to determine the characteristics of satellite lesions in patients with multiple pulmonary ground glass nodules (MPGGNs) and propose a reference for the management of satellite lesions. Methods: We retrospectively analyzed patients with MPGGNs who had undergone multiple rounds of surgical resection of primary and satellite lesions, including pathologic examinations after surgical resection. Results: A total of 125 lesions from 105 patients were included in the analysis; 85 lesions were advanced and 40 lesions were not advanced. Among them, 55 invasive pulmonary adenocarcinomas (IPA) and 70 noninvasive pulmonary adenocarcinomas were identified. After the final regression analysis, the patients' age, satellite lesion location, consolidation tumor ratio (CTR), lesion border clarity, and lesion diameter were used to predict satellite lesion progression. Patients' gender, satellite lesion location, lesion diameter, and computed tomography (CT) attenuation values were used to predict the invasiveness of the satellite lesion. The constructed nomograms showed strong discrimination with concordance indices (C indices) of 0.816 and 0.823, respectively. Conclusions: We developed a set of nomograms that can predict the risk of advanced or invasive satellite lesions in patients with MPGGNs. The area under the receiver operating characteristic (ROC) curve (AUC), the C-index, and the calibration curve suggest that the nomogram may be useful in the clinical setting. This model has the potential to help clinicians make treatment recommendations for the remaining lesions while treating the primary lesion in patients with MPGGNs.