Abstract 5742: Comparison of variants identified by next-generation sequencing panels: A 48-gene panel for myeloid neoplasms and a 35-gene panel for hematologic neoplasms

Abstract

Abstract Background: Identification of genomic variants by molecular testing plays an important role in classification, risk stratification, and management of myeloid neoplasms. Next-generation sequencing (NGS) assays that target frequently mutated genes are commonly used in clinical settings. Owing to the complexity of NGS workflow, the performance requirements of each NGS assay must be validated for the intended use. However, the scarcity of optimal validation samples challenges efficient validation of NGS assays. We previously reported development and validation of a 48-gene NGS panel for molecular profiling of myeloid neoplasms including acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), and myeloproliferative neoplasms (MPN). In this study, we compared the variant calls of a clinically validated 48-gene NGS panel for myeloid neoplasms to those of an independently developed and clinically validated 35-gene NGS panel for hematologic neoplasms. Method: The 48-gene NGS assay is a hybrid capture-based NGS assay using an Illumina NextSeq 500; it was developed to detect single nucleotide variants (SNVs), insertions/deletions (indels), and FLT3 internal tandem duplications (FLT3-ITD). The 35-gene NGS assay HemeSEQ (performed by med fusion, Lewisville, TX) uses Illumina TruSeq custom-amplicon library preparation chemistry, which is sequenced on an Illumina MiSeq. A total of 33,812 bp in 26 genes overlapped between the 2 assays; genes included ASXL1, CALR, FLT3, IDH1, IDH2, JAK2, NPM1, RUNX1, SF3B1, TET2, and TP53. For this study, 141 specimens harboring at least 1 pathogenic mutation within the overlapping genes were selected and sequenced by the 2 assays in CLIA-certified and CAP-accredited laboratories. Detected variants included those for which a targeted therapy or an experimental drug is available. Results: In total, 4.8 million individual base calls (33,812 bp X 141 specimens) were compared. A total of 1,094 variants (278 unique) were concordantly detected by both assays: 1,007 (219 unique) SNVs, and 87 (59 unique) indels; no calls were discordant. Thus, sensitivity was 100% (95% CI: 99.7-100) and specificity was 100% (95% CI: 99.9-100) for overlapping genes. In addition, agreement in variant allele frequency between the 2 assays was good (R2=0.986). The 48-gene assay detected an additional 36 pathogenic mutations from 26 specimens in 14 non-overlapping genes including PTPN11, ATM, BCOR, NF1, and STAG2. Conclusions: Of 4.8 million base calls compared, variant calls were 100% concordant between the 2 NGS assays, and variant allele frequency was in good agreement. The 48-gene assay detected an additional 36 pathogenic mutations in 14 myeloid neoplasm related genes. This study highlights the reliability of robustly validated NGS assays in detection of potentially actionable mutations in clinical settings. Citation Format: Sun Hee Rosenthal, Hansook Chong, Charles Ma, Daniel Sugganth, Renius Owen, Joseph Catanese, Christopher Elzinga, Andrew Grupe, Tracey Freitas, Jennifer Graham, Kristen Champion, Amanda Larsen, Frederick Racke, Felicitas Lacbawan. Comparison of variants identified by next-generation sequencing panels: A 48-gene panel for myeloid neoplasms and a 35-gene panel for hematologic neoplasms [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5742.