Plasma metabolism-related biomarkers enable non-invasive precision detection and individualized treatment in metastatic colorectal cancer

Abstract

Colorectal cancer (CRC) is one of the most deadly malignancies worldwide, especially metastatic CRC (mCRC), whose diagnosis and therapy are limited. Here, the targeted metabolomes of patients with CRC and mCRC were analyzed to explore new biomarkers and treatment strategies. First, plasma was collected from patients with CRC (n = 50) and mCRC (n = 50). Serum metabolites of amino acids, bile acids, and fatty acids were detected by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Subsequently, 60 CRC and mCRC were randomly selected as the training set. A total of 27 amino acids, nine bile acids, and 16 amino acids were detected, among which 13 metabolites were significantly different. KEGG functional analysis showed that these differential metabolites play an essential role in the metabolism of fatty and bile acids. ROC analysis showed that CRC and mCRC could be well discriminated by 13 different metabolic indices, among which the sensitivity and specificity of the single index C18:2 were 0.833 and 0.800. To identify these markers, 40 CRC and mCRC patients were used as validation sets. The identification results of GCA, C17:0, and C18:2 were consistent with previous results. The AUC of GCA, C17:0, combined with C18:2, can reach 0.86, much higher than the traditional index CEA (0.70) and CA19.9 (0.80).Furthermore, Pearson analysis showed a significant correlation between GCA and CA19.9. Twenty-five mCRC patients were selected, and the AUC of GCA, CEA, and CA19.9 were 0.74, 0.74, and 0.70, respectively. In addition, the AUC of GCA and CEA combined with CA19.9 was significantly increased to 0.87. Taken together, our study showed that GCA, C17:0, in combination with C18:2, is superior to the clinical classic in differentiating CRC and mCRC. At the same time, GCA, in combination with CEA and CA19.9, can significantly improve the diagnosis of mCRC with liver metastases. This study is expected to provide new indicators for the diagnosis of mCRC and new strategies for the treatment of mCRC.