IMPACT OF DEEP TARGETED NEXT GENERATION SEQUENCING ON THE WORK-UP OF PATIENTS WITH PANCREAS CYSTS OR DILATED DUCT - A PROSPECTIVE STUDY WITH EUS-GUIDED FNA

Aims Pancreas cysts or dilated pancreatic ducts were often detected by cross sectional imaging methods, but only mucinous neoplasia may develop pancreas cancer. Precise distinction between mucinous and non-mucinous cysts is crucial for clinical decision. We performed a prospective study (NCT03820531) on patients with pancreas cyst(s) and/or dilated main duct including Next Generation Sequencing (NGS) in the diagnostic work up of EUS-guided FNA.

Methods Pancreas cyst (≥ 10mm) or main duct fluid was analysed by CEA, cytology and deep targeted NGS using the Oncomine Tag Sequencing Barcode Set of 14 known gastrointestinal cancer genes with a limit of detection level down to 0.01 % mutant allele frequency. Results were compared with histopathology and clinical follow up.

Results 93 patients with incidental pancreas cyst(s) and/or dilated pancreas main duct ≥ 5mm were screened. 51 patients were excluded, mainly due to inoperability or small cyst size ≤ 10mm. 42 patients and 7 negative controls were enrolled in further analysis. KRAS/GNAS-mutations were identified most often and were detected in all histologically proven mucinous neoplasia but not in non-mucinous cysts, chronic pancreatitis or acute pseudocysts. Some cysts harboured multiple KRAS/GNAS-mutations. Deep targeted NGS, cytology and CEA had a sensitivity and specificity of 100%/100%; 50%/100% and 58.3%/85.7% on the basis of 24 patients with final diagnosis. KRAS/GNAS-analysis had an impact on clinical decision in 29/42 (69%) patients (operation versus follow up versus end of follow up).

Conclusions Deep targeted NGS in pancreas cyst/duct fluid specifies the diagnostic work up of patients with pancreas cysts ≥ 10mm and/or main duct dilation ≥5mm and may direct clinical decision. Results have to be proven by a greater cohort.