Clinicopathological characteristics and prognosis of colon cancer with lung metastasis without liver metastasis: A large population-based analysis.

Distant metastasis explains the high mortality rate of colon cancer, in which lung metastasis without liver metastasis (LuM) is a rare subtype. This study is aimed to identify risk factors of LuM and LLM (lung metastasis with liver metastasis) from colon cancer, and to analyze the prognosis of patients with LuM by creating a nomogram. Patients' information were obtained from the Surveillance, Epidemiology, and End Results (SEER) database. Multivariable logistic regression analysis was used to determine the risk factors for LuM and LLM. Prognostic factors for cancer-specific survival (CSS) and overall survival (OS) were identified by multivariate Cox proportional hazards regression and nomogram models were established to predict CSS and OS. Multivariate logistic regression analysis showed that blacks, splenic flexure of colon tumor, tumor size >5 cm, T4, N3, and higher lymph node positive rate were associated with the occurrence of LuM. Meanwhile, age >65 years old, female, splenic flexure of colon, higher lymph node positive rate, and brain metastasis were independent risk factors for CSS. The C-index of the prediction model for CSS was 0.719 (95% CI: 0.691-0.747). In addition, age, primary site, tumor size, differentiation grade, N stage, and bone metastasis were significantly different between LuM and LLM. The nomograms we created were effective in predicting the survival of individuals. Furthermore, patients with LuM and LLM from colon cancer might require different follow-up intervals and examinations.