Search

5-Fluorouracil (5-FU) and its pro-drug Capecitabine have been widely used in treating colorectal cancer. However, not all patients will respond to the drug, hence there is a need to develop reliable early predictive biomarkers for 5-FU response. Here, we report a novel potentially functional Single Nucleotide Polymorphism (pfSNP) approach to identify SNPs that may serve as predictive biomarkers of response to 5-FU in Chinese metastatic colorectal cancer (CRC) patients. 1547 pfSNPs and one variable number tandem repeat (VNTR) in 139 genes in 5-FU drug (both PK and PD pathway) and colorectal cancer disease pathways were examined in 2 groups of CRC patients. Shrinkage of liver metastasis measured by RECIST criteria was used as the clinical end point. Four non-responder-specific pfSNPs were found to account for 37.5% of all non-responders (P<0.0003). Five additional pfSNPs were identified from a multivariate model (AUC under ROC = 0.875) that was applied for all other pfSNPs, excluding the non-responder-specific pfSNPs. These pfSNPs, which can differentiate the other non-responders from responders, mainly reside in tumor suppressor genes or genes implicated in colorectal cancer risk. Hence, a total of 9 novel SNPs with potential functional significance may be able to distinguish non-responders from responders to 5-FU. These pfSNPs may be useful biomarkers for predicting response to 5-FU. [ABSTRACT FROM AUTHOR]

Early diagnosis of hepatocellullar carcinoma (HCC) remains a challenge. The current practice of serum alpha-fetoprotein (AFP) measurement is inadequate. Here we utilized a proteomic approach to identify novel serum biomarkers for distinguishing HCC patients from non-cancer controls. We profiled the serum proteins in a group of 58 resectable HCC patients and 11 non-HCC chronic hepatitis B (HBV) carrier samples from the Singapore General Hospital (SGH) using the RayBio® L-Series 507 Antibody Array and found 113 serum markers that were significantly modulated between HCC and control groups. Selected potential biomarkers from this list were quantified using a multiplex sandwich enzyme-linked immunosorbent assay (ELISA) array in an expanded SGH cohort (126 resectable HCC patients and 115 non-HCC chronic HBV carriers (NC group)), confirming that serum prolactin and monocyte chemoattractant protein-1 (MCP-1) were significantly upregulated in HCC patients. This finding of serum MCP-1 elevation in HCC patients was validated in a separate cohort of serum samples from the Mochtar Riady Institute for Nanotechnology, Indonesia (98 resectable HCC, 101 chronic hepatitis B patients and 100 asymptomatic HBV/HCV carriers) by sandwich ELISA. MCP-1 and prolactin levels were found to correlate with AFP, while MCP-1 also correlated with disease stage. Subsequent receiver operating characteristic (ROC) analysis of AFP, prolactin and MCP-1 in the SGH cohort and comparing their area under the ROC curve (AUC) indicated that neither prolactin nor MCP-1 on their own performed better than AFP. However, the combination of AFP+MCP-1 (AUC, 0.974) had significantly superior discriminative ability than AFP alone (AUC, 0.942; p<0.001). In conclusion, prolactin and MCP-1 are overexpressed in HCC and are conveniently quantifiable in patients’ sera by ELISA. MCP-1 appears to be a promising complementary biomarker for HCC diagnosis and this MCP-1+AFP model should be further evaluated as potential biomarker on a larger scale in patients at-risk of HCC. [ABSTRACT FROM AUTHOR]