Radiomics Signature to Predict Prognosis in Early-Stage Lung Adenocarcinoma (≤3 cm) Patients with No Lymph Node Metastasis

Objectives: To investigate the predictive ability of radiomics signature to predict the prognosis of early-stage primary lung adenocarcinoma (<= 3 cm) with no lymph node metastasis (pathological stage I). Materials and Methods: This study included consecutive patients with lung adenocarcinoma (<= 3 cm) with no lymph node metastasis (pathological stage I) and divided them into two groups: good prognosis group and poor prognosis group. The association between the radiomics signature and prognosis was explored. An integrative radiomics model was constructed to demonstrate the value of the radiomics signature for individualized prognostic prediction. Results: Six radiomics features were significantly different between the two prognosis groups and were used to construct a radiomics model. On the training and test sets, the area under the receiver operating characteristic curve value of the radiomics model in discriminating between the two groups were 0.946 and 0.888, respectively, and those of the pathological model were 0.761 and 0.798, respectively. A radiomics nomogram combining sex, tumor size and rad-score was built. Conclusion: The radiomics signature has potential utility in estimating the prognosis of patients with pathological stage I lung adenocarcinoma (<= 3 cm), potentially enabling a step forward in precision medicine.

Automated summary

This study investigated the predictive ability of radiomics signature for the prognosis of early-stage primary lung adenocarcinoma (<= 3 cm) with no lymph node metastasis (pathological stage I). The radiomics signature was found to be significantly associated with prognosis, and the radiomics model showed good performance in predicting prognosis. A radiomics nomogram incorporating other factors was built to facilitate individualized prognostic prediction.