MicroRNA expression profiling signatures to separate chronic inflammation from Pancreatic Cancer. Preliminary results

Background and Aims: Chronic pancreatitis (CP) is a persistent inflammation of the pancreas that results in permanent structural damage with fibrosis and ductal strictures, followed by a decline in exocrine and endocrine function, and debilitating pain as the main symptom. Patients are also at increased risk for pancreatic cancer (PC), a disease well known for its poor prognosis. However, the early diagnosis of chronic pancreatitis and the separation from PC can be quite challenging, particularly in the background of chronic pancreatitis. It has been reported that microRNAs (miRNAs) are increasingly found and applied as targets for the diagnosis and treatment of various diseases. The aim of the present study was to determine a set of miRNAs that can help in the differentiation between normal, CP and PC tissue specimens. Methods: Pancreatic tissue samples (n = 20) were collected during operations at the 1st Department of Surgery (Semmelweis University), snap-frozen and stored long term at -80 °C. CP patients (n = 10) were grouped based on the M- ANNHEIM clinical classification (grades B-D, stages Ib-IVb), PC patients (n = 5) were all of stage III. Normal tissue (n = 5) was also obtained. A senior pathologist checked for tumor content and graded chronic pancreatitis samples using the classification adapted by Klöppel. Total RNA including the small RNA fraction was extracted from the tissue slides using a modified RNeasy Mini Kit (Qiagen), and after RNA quality check, the samples were loaded onto Affymetrix GeneChip miRNA 3.0 microarrays that contained 1733 human mature microRNA sequences. Statistics included the linear model, SAM and PAM methods. Results: Statistical analyses demonstrated very high sensitivity of distinction between healthy people and patients with either CP or PC. A set of only 7 miRNAs differentiated between controls and CP patients (overexpressed: miR-548ai, miR-3197, miR-3613 – 3 p, miR- 4668 – 5 p; underexpressed: miR-17*, miR-29c*, miR-378 d) with an accuracy = 0.93, p-value = 0.02, specificity = 100% and sensitivity = 80%. In addition discrimination between cancer and chronic pancreatitis was also achieved with a set of 9 markers including miR-21*, miR-181b, miR-221 and miR-222. Conclusion: Expression patterns of miRNAs are significantly altered in pancreatic diseases. Microarray analyses identified statistically unique profiles which have the potential to separate between the different tissue types and could serve as molecular signatures that differ according to pathologic features.