70P - Diagnostic performance of liquid biopsy for pancreatic solid lesion as alternative to endoscopic ultrasound-guided fine needle aspiration (EUS-FNA)

Date 10 October 2016
Event ESMO 2016 Congress
Session Poster display
Topics Staging Procedures (clinical staging)
Translational Research
Basic Principles in the Management and Treatment (of cancer)
Presenter David Sefrioui
Citation Annals of Oncology (2016) 27 (6): 15-42. 10.1093/annonc/mdw363
Authors D. Sefrioui1, F. Blanchard2, P. Basile1, E. Toure2, C. Dolfus2, L. Beaussire3, N. Vasseur3, A. Perdrix4, A. Gangloff1, L. Schwarz5, F. Clatot4, J. Tuech5, J. Sabourin2, T. Frebourg6, P. Michel1, F. Di Fiore4
  • 1Department Of Of Hepato-gastroenterology, IRON (equipe de recherche onco normande) within INSERM unit U1079., 76031 - Rouen/FR
  • 2Department Of Pathology, Rouen University Hospital, Rouen/FR
  • 3Institute For Biomedical Research And Innovation, Inserm U1079, University of Rouen, Rouen/FR
  • 4Medical Oncology, Centre Henri Becquerel, Rouen/FR
  • 5Surgery, Rouen university Hospital, Rouen/FR
  • 6Department Of Genetics, Rouen University Hospital, Rouen/FR



EUS-FNA is considered as the reference procedure for the diagnosis of solid pancreatic tumor. Liquid biopsy (CA19.9, circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA)) is an attractive alternative approach. We previously reported that the detection of CTCs had a diagnostic accuracy of 70% for pancreatic adenocarcinoma (PA) (Am J Gastroenterol 2013;108:152-155). The aim was to evaluate the diagnostic performance of each of these biomarkers (alone or combined) in an extended series of patients with pancreatic solid tumor.


From 01/2011 to 03/2014, all patients with pancreatic solid tumors diagnosed on CT-scan and referred for a EUS-FNA were included. For each patient, 1 EDTA tube (CTC) and 1 heparinized plasma tube (CA19.9) were systematically collected before EUS-FNA. CTCs isolation was performed using the Screencell® Cyto filtration. CTCs were characterized by an experienced cytopathologist blinded to the histological diagnosis. CtDNA extraction was performed on the remaining heparinised plasma sample if available. CtDNA was analysed with QX200TM Droplet DigitalTM PCR System (ddPCR) and a multiplex assay allowing screening in a same sample for multiple (n = 7) KRAS mutations (c.34G > A, c.34G > C, c.34G > T, c.35G > A, c.35G > C, c.35G > T and c.38G > A).


A total of 68 patients were included (58 with a malignancy tumor (52 PA and 6 other malignant tumors) and 10 with benign lesion. The stage at diagnosis for PA was localized, locally advanced and metastatic in 13, 17 and 22 patients respectively. Sensitivity (Se) of EUS-FNA performed at inclusion for the PA diagnosis was 65%. Se of CTCs, ctDNA and CA19.9 was 64%, 63% and 79% respectively. Se of each marker increased proportionally with stage (i.e. Se (metastatic PA) = 77%, 85% and 80% for CTCs, ctDNA and CA19.9). Specificity (Spe) of these biomarkers was 81%, 75% and 93% respectively. All 3 biomarkers were available for 51 patients. Positivity of at least 2 on 3 was associated with a Se of 77% and a Spe of 91%.


Our results confirm that CA19.9 alone or in combination with CTCs and/or ctDNA represents a non-invasive and effective method as an alternative to EUS-FNA for PA diagnosis.

Clinical trial identification

Legal entity responsible for the study

CHU Rouen


Association de Cancerologie Digestive de h^Haute Normandie


All authors have declared no conflicts of interest.