118P - Nationwide External Quality Assessment (EQA) of EGFR testing in circulating tumor DNA : the French experience

Date 11 September 2017
Event ESMO 2017 Congress
Session Poster display session
Topics Cancers in Adolescents and Young Adults (AYA)
Personalised/Precision medicine
Translational Research
Basic Principles in the Management and Treatment (of cancer)
Presenter Marc Denis
Citation Annals of Oncology (2017) 28 (suppl_5): v22-v42. 10.1093/annonc/mdx363
Authors M. Denis1, A. Vallée1, S. Charpentier1, I. Nanni-Métellus2, L. Lacroix3, C. Jovelet3, M. Francilette3, D. Fetique4, J. Bellocq4, E. Rouleau5
  • 1Molecular Biology, CHU Nantes, 44093 - Nantes/FR
  • 2Laboratoire De Biologie Médicale, Faculté de médecine, Marseille/FR
  • 3Bmo, Institut Gustave Roussy, 94800 - Villejuif/FR
  • 4Afaqap, Hôpital Hautepierre, Strasbourg/FR
  • 5Département De Biologie Et Pathologie Médicales, Gustave ROUSSY, 94805 - VILLEJUIF/FR



Detection of EGFR mutation in circulating tumor DNA (ctDNA), a powerful blood-based biomarker with multiple clinical applications for lung cancer patient, is technically challenging because it requires sensitivity and specificity. In order to evaluate the performance of the french laboratories performing this assay, we set up a national EQA scheme for circulating tumor DNA.


Artificial samples were prepared by spiking DNA extracted from control FFPE sections containing specific mutations (Horizon Diagnostics, from 25 to 350 copies/mL of plasma, as determined by digital PCR) in normal plasma (Clinisciences). Aliquots (2 ml) of 10 different samples were sent in dry ice to 43 laboratories. DNA extraction and EGFR testing were performed according to local practice. Laboratories were requested to search for exon 19 deletions, p.L858R, p.G719X and p.T790M mutations. Data were collected on a web questionnaire within one month and compared to the expected results.


We collected 30 complete sets of data. DNA was extracted using the QIAmp® circulating DNA kit (Qiagen; n = 13), the Cobas® cfDNA sample preparation kit (Roche; n = 9) or the Maxwell® system (Promega; n = 7). The most widely used methods were the Cobas® EGFR Mutation Test v2 (Roche; n = 10), digital PCR (n = 8) and NGS (n = 6). Among the 10 labs using the Cobas®, 9 obtained the 10 expected genotypes. This number dropped to 3 (out of 6 labs) for NGS, and 2 (out of 8 labs) for dPCR, because of false negative results, false positive results, and not contributive tests.


Digital PCR and NGS are known to be highly sensitive techniques. However, the results of this EQA suggest that in routine clinical practice, ctDNA analysis requires technical skill or/and validated bioinformatic pipeline to reach high sensitivy and specificity. Under the specific conditions of this scheme, the Cobas® method appeared to be the most robust approach. This external control will allow the laboratories to evaluate their practice and improve their process. A similar approach targeting other genes (BRAF, KRAS and NRAS) is being developed, and additional EQA schemes will be set up at the nationwide level. Supported by a grant from AstraZeneca.

Clinical trial identification

Legal entity responsible for the study





All authors have declared no conflicts of interest.