Aberrant expression of microRNAs in serum may identify individuals with pancreatic cancer

Abstract

Pancreatic cancer (PC) has the poorest survival rate among all types of human cancer due to the lack of sensitive and non-invasive diagnostic screen methods for PC screening. Our aim was to identify novel serum microRNA (miRNA) biomarkers for the early detection of PC. We used microarray to screen differential expression of miRNAs in two pooled serum samples (6 PC patients and 6 healthy controls). A panel of miRNAs (22 over-expression and 23 decreased) were deregulated in serum of PC patients in comparison to controls. The expressions of 8 selected miRNAs were further evaluated in sera from 49 PC patients and 27 controls using quantitative reverse transcription-polymerase chain reaction. The levels of serum miR-492 and miR-663a were significantly decreased in PC patients compared with controls (P < 0.05). ROC curve analysis showed that serum miR-492 and miR-663a yield an AUC of 0.787 with 75.5% sensitivity and 70.0% specificity and 0.870 with 85.7% sensitivity and 80.0% specificity, respectively, for discriminating between PC patients and healthy controls. In addition, the level of miR-663a was significantly and inversely associated with TNM stage (P = 0.027). These results suggested that serum miR-492 and miR-663a could have strong potential as novel non-invasive biomarkers for the early detection of PC.

Keywords: MicroRNA; miR-492; miR-663a; pancreatic cancer; serum.

Figure 1

Serum levels of miR-492 (A) and miR-663a (B) in 49 PCs and 27control subjects in the second-stage validation. The relative levels of miRNAs (Y-axis) were normalized to the spiked-in cel-miR-39. Data indicated significantly decreased expression of miR-492 and miR-663a in PC group with more than 2-fold change compared with the normal control group.

Figure 2

Receiver operating characteristics (ROC) curve analysis using miR-492 (A) and miR-663a (B) for discriminating PC from controls. MiR-492 yield an AUC of 0.787 (95% CI: 0.689-0.885) with 75.5% sensitivity and 70.0% specificity, and miR-663a yield an AUC of 0.870 (95% CI: 0.793-0.948) with 85.7% sensitivity and 80.0% specificity for discriminating PC from normal controls. Combined analysis of miR-492and miR-663a (C) revealed an AUC of 0.869 (95% CI: 0.791-0.947), with 85.7% sensitivity and 80.0% specificity.

Figure 3

Correlation of serum levels of miR-492 and miR-663a and patients’ different TNM stages. Box plots of serum levels of miR-492 (A) (P = 0.279) and miR-663a (B) (P = 0.027) in PC patients (n = 49) with different TNM stage.