Gene Analysis of Advanced Pancreatic Carcinoma Using EUS-Guided FNA with Focused DNA Array: Possible Predictor of the Chemotherapeutic Efficacy

Abstract

Background: Endoscopic ultrasound (EUS) has evolved from EUS imaging to EUS-guided fine-needle aspiration (FNA). EUS-guided FNA is now beyond cytology diagnosis and makes it possible to screen hundreds of genes simultaneously using cDNA microarray technique. The prediction of the effectiveness of anti-cancer agent is very important in the treatment of pancreatic carcinoma. The purpose of this study was to determine of the usefulness of Focused DNA Array (FDA) analysis using the pancreatic carcinoma tissue obtained by EUS-guided FNA for the prediction of chemotherapeutic effect. Methods: Pancreas tissues of 95 patients were obtained by EUS-guided FNA. mRNA was harvested from each sample and cDNA was composed by reverse transcription polymerase chain reaction (RT-PCR). Expression of mRNA was evaluated using FDA which was restricted to 133 genes including drug-resistance related genes. The expression of mRNA was measured as the relative value to house keeping gene, GAPDH. Results: 81 (85.3%) of 95 mRNA samples obtained by EUS-guided FNA had adequate quantities and qualities for analysis. The mean amount of 81 mRNA samples was 0.80 ± 0.18μg. Gemcitabine (GEM) monotherapy composed 1000 mg/m2 weekly administration for three weeks following one week rest was given for 49 patients. Thirty-five patients who had completed 2 courses or more were divided into the following two groups. Patients with partial response (PR) and those with stable disease (SD) whose tumor marker reduced by 50 % or more were classified as responders. The rest of the patients were classified as non-responders. The mRNA expression of deoxycytidine kinase (DCK), ENT1 and ENT2 which may act as GEM-sensitive factors, and dCMP deaminase (DCD), NT5, and RRM1 which may act as GEM-resistant factors were used for the following score analysis. The score for each gene was defined as relative values to the average expression of individual gene in 81 samples. The score for GEM-sensitivity was calculated as DCK+ENT1+ENT2-DCD-NT5-RRM1 gene scores. The score for GEM-sensitivity of 0.3 or more was seen in 8 (66.7%) of 12 responders. On the other hand, the gene score less than 0.3 was seen in 17 (73.9%) of 23 non-responder. (P = 0.002). When the cut-off value of the score for GEM-sensitivity was optimally defined as 0.3, the sensitivity, specificity and accuracy were 67%, 74% and, 71%, respectively. Conclusion: The gene analysis using FDA might be useful to predict the chemothrapeutic efficacy for patients with advanced pancreatic cancer.