This study aims to establish and validate a new nomogram for predicting brain metastasis from lung cancer by integrating data.
266 patients diagnosed as lung cancer between 2016 and 2018 were collected from Guangdong Academy of Medical Sciences. The first 70% of patients were designated as the primary cohort and the remaining patients were identified as the internal validation cohort. Univariate and multivariable logistics regression were applied to analyze the risk factors. Independent risk factors were used to construct nomogram. C-index was used to evaluate the prediction effect of nomogram.100 patients diagnosed as lung cancer between 2018 and 2019 were collected for external validation cohorts. The evaluation of nomogram was carried out through the distinction and calibration in the internal validation cohort and external validation cohort.
166 patients were diagnosed with brain metastasis among the 266 patients. The gender, pathological type (PAT), leukocyte count (LCC) and Fibrinogen stage (FibS) were independent risk factors of brain metastasis. A novel nomogram has been developed in this study showed an effective discriminative ability to predict the probability of lung cancer patients with brain metastasis, the C-index was 0.811.
Our research provides a novel model that can be used for predicting brain metastasis of lung cancer patients, thus providing more credible evidence for clinical decision-making.