Integrated metabolomics and lipidomics reveal plasma markers of non-small cell lung cancer with brain metastasis.

Brain metastasis (BM) is a fatal complication of non-small cell lung cancer (NSCLC). The lack of non-invasion methods for early diagnosis and risk stratification is the major cause for the poor prognosis of NSCLC with BM. The metabolic state of BM has a significant change characterized with high reactive oxygen species accumulation and the subsequent antioxidant metabolic response. We sought to screen plasma markers based on metabolomics and lipidomics for the early diagnosis and prognostic evaluation of NSCLC patients with BM. Plasma samples collected from 48 NSCLC patients with BM and 49 gender- and age- matched primary lung cancer (PLC) patients were randomly divided into train set and test set, then analyzations of metabolomics and lipidomics were performed using liquid chromatography-tandem mass spectrometry (LC-MS). Differential metabolites and lipids were discovered from the train set. A diagnostic biomarker panel was constructed using logistic regression. The test set were utilized to validate the accuracy of the diagnostic model. Furthermore, a risk score was established to stratify risk for NSCLC patients. There were 34 differential metabolites and 35 differential lipids annotated in the train set. The diagnostic biomarker panel consisting of homocysteine, ascorbic acid, LPC (22:0) and LPC (20:0) is able to discriminate BM from PLC with an excellent performance. The risk score based on protocatechuic acid and LPC (20:0) could efficiently stratify risk for NSCLC patients with BM. Our study reports plasma biomarkers and predictive models for the early diagnosis and prognostic evaluation of NSCLC patients with BM.