PF175 MULTICENTRE STANDARDIZATION OF MINIMAL RESIDUAL DISEASE DETECTION AND QUANTITATION USING THE EUROCLONALITY‐NGS ASSAY

Background: Amplicon‐based next‐generation sequencing (NGS) is increasingly used for identification of clonal immunoglobulin (IG) and T‐cell receptor (TR) gene rearrangements as markers for minimal residual disease (MRD) monitoring in acute lymphoblastic leukaemia (ALL). A standardized assay integrating amplicon library preparation as well as bioinformatic data analysis for marker identification has been established and validated in a multicentre study by the EuroClonality‐NGS working group, including laboratories from the EuroMRD consortium (Brüggemann et al., Leukemia, in press). The use of NGS for MRD detection represents a universal assay which offers better specificity compared to the currently used real‐time quantitative (q)PCR as well as information about background repertoire, however a reliable quantitation of MRD levels by this approach still represents a challenge. Aims: 14 laboratories from the EuroMRD consortium and EuroClonality‐NGS working group performed IG / TR marker identification and MRD quantitation in 3 follow‐up samples to assess the reproducibility of MRD detection and quantitation. Methods: All participating laboratories followed the standardized library preparation protocols developed by the EuroClonality‐NGS working group and performed 2 sequencing runs. The 1 st run aimed at marker identification of potential target IGH, IGK, TRB, TRG and TRD rearrangements present in the diagnostic sample. Based on the screening results, at least one gene locus was selected by each laboratory for a 2 nd run of all 3 follow‐up samples diluted into a central polyclonal buffy coat control with 500ng input DNA used (equivalent to 75,000 cells) for MRD detection and assessment of false positivity. MRD levels in these samples were assessed by qPCR independently. Libraries were prepared by 2 rounds of PCR in 12 laboratories, whereas 2 laboratories used a modified 1‐round approach. The EuroClonality‐NGS‐developed central In‐Tube Quality/Quantitation Control (cIT‐QC) ‐ spike‐ins of known DNA quantity, was used by 2 labs. Sequencing was performed on Illumina MiSeq (2x250 v2 kit) or IonTorrent S5 XL instruments. Results were analysed with ARResT/Interrogate. Results: The 6 rearrangements (3 IGH‐VJ, 1 IGH‐DJ, 1 TRD, 1 TRG) present in the diagnostic sample above the 5% threshold were identified as markers in 71/76 cases (93%). MRD detection was performed in the IGH‐VJ system by all 14 laboratories, IGH‐DJ by 4, TRD by 4 and, TRG by 2, with mean coverage of 992,129 reads per sample. In one case IGH‐VJ libraries did not meet the quality control standards for analysis and were therefore excluded. In 12 of the remaining 13 laboratories, all 3 IGH‐VJ targets were detected in all 3 follow‐up samples, as well as in all IGH‐DJ, TRD and TRG libraries. MRD levels (4.35E‐3; 1.8E‐3 and 1.72E‐4) correlated well with the expected dilution (8E‐3; 4E‐3 and 4E‐4) and qPCR results in TRG systems, but appeared over‐estimated in other systems in the absence of normalization to the cIT‐QC (selected IGH‐VJ target means vs. cIT‐QC normalized value: 6.34E‐2 vs 1.19E‐02; 3.18E‐2 vs. 8.40E‐03 and 3.98E‐3 vs. 2.31E‐04). Summary/Conclusion: We have shown that the EuroClonality‐NGS‐developed approach can be successfully used for MRD‐PCR marker identification. In addition, MRD detection using this assay showed very good inter‐laboratory concordance. For accurate quantitation of MRD the use of the cIT‐QC is essential and should be further validated in a large‐scale multicentre testing before the method can be used in a clinical setting.