External Quality Assessment on Molecular Tumor Profiling with Circulating Tumor DNA-Based Methodologies Routinely Used in Clinical Pathology within the COIN Consortium

Paul van der Leest, Pim Rozendal, John Hinrichs, Carel J. M. van Noesel, Karen Zwaenepoel, Birgit Deiman, Cornelis J. J. Huijsmans, Ronald van Eijk, Ernst Jan M. Speel, Rick J. van Haastert, Marjolijn J. L. Ligtenberg, Ron H. N. van Schaik, Maurice P. H. M. Jansen, Hendrikus J. Dubbink, Wendy W. de Leng, Mathie P. G. Leers, Menno Tamminga, Daan van den Broek, L. on C. van Kempen, Ed Schuuring

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

BACKGROUND: Identification of tumor-derived variants in circulating tumor DNA (ctDNA) has potential as a sensitive and reliable surrogate for tumor tissue-based routine diagnostic testing. However, variations in pre(analytical) procedures affect the efficiency of ctDNA recovery. Here, an external quality assessment (EQA) was performed to determine the performance of ctDNA mutation detection work flows that are used in current diagnostic settings across laboratories within the Dutch COIN consortium (ctDNA on the road to implementation in The Netherlands). METHODS: Aliquots of 3 high-volume diagnostic leukapheresis (DLA) plasma samples and 3 artificial reference plasma samples with predetermined mutations were distributed among 16 Dutch laboratories. Participating laboratories were requested to perform ctDNA analysis for BRAF exon 15, EGFR exon 18-21, and KRAS exon 2-3 using their regular circulating cell-free DNA (ccfDNA) analysis work flow. Laboratories were assessed based on adherence to the study protocol, overall detection rate, and overall genotyping performance. RESULTS: A broad range of preanalytical conditions (e.g., plasma volume, elution volume, and extraction methods) and analytical methodologies (e.g., droplet digital PCR [ddPCR], small-panel PCR assays, and next-generation sequencing [NGS]) were used. Six laboratories (38%) had a performance score of >0.90; all other laboratories scored between 0.26 and 0.80. Although 13 laboratories (81%) reached a 100% overall detection rate, the therapeutically relevant EGFR p.(S752_I759del) (69%), EGFR p.(N771_H773dup) (50%), and KRAS p.(G12C) (48%) mutations were frequently not genotyped accurately. CONCLUSIONS: Divergent (pre)analytical protocols could lead to discrepant clinical outcomes when using the same plasma samples. Standardization of (pre)analytical work flows can facilitate the implementation of reproducible liquid biopsy testing in the clinical routine.
Original languageEnglish
Pages (from-to)759-767
Number of pages9
JournalClinical Chemistry
Volume70
Issue number5
DOIs
Publication statusPublished - 1 May 2024

Cite this