A nomogram predicting the risk of extrathoracic metastasis at initial diagnosis of T≤3cmN0 lung cancer.

BACKGROUND

The risk and risk factors of extrathoracic metastasis at initial diagnosis in T≤3cmN0 lung cancer patients are not fully understood. We aimed to develop a model to predict the risk of extrathoracic metastasis in those patients.

METHODS

Clinicopathological data of patients were collected from Surveillance, Epidemiology, and End Results (SEER) database. Univariable and multivariable analyses using logistic regression were conducted to identify risk factors. A predictive model and corresponding nomogram were developed based on the risk factors. The model was assessed using the area under the receiver operating characteristic curve (AUC), Hosmer-Lemeshow test, and decision curve.

RESULTS

A total of 20,057 T≤3cmN0 patients were enrolled, of whom 251 (1.25%) were diagnosed with extrathoracic metastasis at the initial diagnosis. Aged ≤50 [odds ratio (OR): 2.05, 95% confidence interval (CI): 1.19-3.53, P=0.01] and aged ≥81 [1.65 (1.05-2.58), P=0.03], Hispanic [1.81 (1.20-2.71), P=0.004], location of bronchus [3.18 (1.08-9.35), P=0.04], larger tumor size, pleural invasion, and a history of colorectal cancer [2.01 (1.01-4.00), P=0.046] were independent risk factors. In the training cohort and validation cohort, the AUCs of the developed model were 0.727, 0.728 respectively, and the results of Hosmer-Lemeshow test were P=0.47, P=0.61 respectively. The decision curve showed good clinical meaning of the model.

CONCLUSIONS

Extrathoracic metastasis at initial diagnosis in T≤3cmN0 lung cancer patients was not rare. The model based on the risk factors showed good performance in predicting the risk of extrathoracic metastasis.