In:
Journal of Clinical Oncology, American Society of Clinical Oncology (ASCO), Vol. 37, No. 15_suppl ( 2019-05-20), p. e18557-e18557
Kurzfassung:
e18557 Background: The advent of NGS technologies revolutionized patient management. Compiling a comprehensive list of genomic alterations and summarizing the clinical implications is complex and time-consuming. One main difficulty lies in detecting relevant genomic alterations with actionable, diagnostic and prognostic value. The development of bioinformatic tools is essential to interpret genomic data. Here we describe the validation of an end-to-end genomic NGS workflow to diagnose myeloid disorders. Methods: Performed on 81 patients previously diagnosed using non-NGS diagnostic tests, this study aims at finding out if an NGS diagnostic approach can accelerate the diagnosis process. DNA was extracted from purified mononuclear cells (n = 54), polymorphonuclear cells (n = 50) and/or total leucocytes (n = 58) from peripheral blood and bone marrow (n = 9). Genomic analysis was performed using a capture-based solution (MYS by SOPHiA GENETICS) covering 30 relevant genes associated with MDS, MPN and Leukemia. Sequencing was performed on Illumina MiSeq and results were analyzed using the SOPHiA platform. Results: We performed 2 validation runs with high raw reads quality (Phred-score 〉 35). Over 90% of reads were on-target and coverage was larger 〉 1000x with a mean of 10 regions falling below this threshold for each sample. Repeatability and reproducibility were evaluated with 6 samples and showed a good correlation between intra (R = 0.993) and inter run (R = 0.97) replicates. Average number of driver mutation was 1.3 for patients with MPN (n = 38), 3 for patients with CMML (n = 9), 2.3 for patients with MDS (n = 10), 2.5 for patients with AML (n = 13) and 1.2 for other neoplasms (n = 11). Obtained NGS results supported previous diagnoses. However for patients where peripheral blood and bone marrow samples were used to perform the analysis, one patient with mastocytosis has a discrepancy of variants found in each sample type. We detected 4 variants in peripheral blood and one additional variant in DNA extracted from bone marrow. The mutation occurred in the RUNX1 gene and was associated to an unfavourable prognosis with higher risk of relapse. Conclusions: We described here the successful validation of a single NGS panel to diagnose multiple myeloid neoplasms. The workflow shows excellent reliability in terms of reproducibility and repeatability which is essential for clinical implementation. Each method has its own benefits and limitations and interpretation of results should be made taking into account the type of the sample used.
Materialart:
Online-Ressource
ISSN:
0732-183X
,
1527-7755
DOI:
10.1200/JCO.2019.37.15_suppl.e18557
Sprache:
Englisch
Verlag:
American Society of Clinical Oncology (ASCO)
Publikationsdatum:
2019
ZDB Id:
2005181-5
