PB2000 A comprehensive NGS panel approach for clinical genomic analysis in haematological malignancy

Mason, J., et al (2023) HemaSphere 7 (Suppl)


Delivery of comprehensive genomic analysis in hematological malignancies (HM) has until recently relied on multiple techniques for the detection of the range of aberrations of clinical relevance. Increasingly laboratories are replacing multiple workflows with next generation sequencing (NGS) based approaches; targeted panels are favored by many laboratories as the most cost-effective approach for the detection of clinically relevant CNVs and SVs in addition to SNVs and indels.

Our laboratory provides clinical testing services for a population of 6 million. The existing myeloid NGS panel reports on 42 genes, and detects SNVs and other clinically relevant mutations such as FLT3-ITDs and KMT2A-PTDs. All variants are interpreted by expert scientists according to recognized guidelines and captured in a curated database.


1)To review all clinically significant (CS) variants identified using our existing myeloid panel over the preceding 20 months and provide a ‘real world’ dataset on most frequently mutated genes based on expert curation.

2)To design a comprehensive hybridization-capture panel to detect all variant types and encompassing all regions identified as clinically relevant in the UK NHSE National Genomic Test Directory (TD; https://www.england.nhs.uk/publication/national-genomic-test-directories), plus additional genes likely to be included in future updates to the TD. This ‘Pan-Haem Panel’ (PHP) will be used to deliver testing across myeloid and lymphoid neoplasia and will replace multiple existing workflows.


PHP was co-designed by our laboratory and Nonacus, UK to the following specification:

1.Detect SNVs with a limit of detection (LoD) of 1% across 132 clinically relevant genes, including all those listed on the TD, plus additional genes expected to be included in future updates to the TD e.g. UBA1:VEXAS.

2.Detect CNVs with a LoD of 20% in target regions

3.Inclusion of specific intronic regions in the design of the panel and deployment of variant callers in the bioinformatic pipeline to enable the detection and characterization of FLT3-ITDs, b.KMT2A(MLL)-PTDs and rearrangements, BCR::ABL1 rearrangements

4.Backbone comprising 2213 SNPs, for gross CNV calling

A bespoke bioinformatics pipeline was created using a python script employing Illumina DRAGEN Bio-IT Platform v3.10 software, converting BCL files to ORA compressed FASTQs, before aligning reads to genome build GRCh38 with BWA-MEM. For variant calling, the DRAGEN variant caller tool is used for identifying SNVs while an extended version of Manta is implemented for calling SVs. Each sample concatenated VCF file is uploaded to Alissa Interpret for filtering and annotation.

Results: 1)CS variants across all myeloid referrals – most frequently mutated genes

An external file that holds a picture, illustration, etc.
Object name is hs9-7-e76738e4-g001.jpg

2)PHP coverage was >99% at >500X. A range of SNVs were detected at varying Variant Allele Fractions as low as 1%, plus indels in TP53 and CEBPA, FLT3-ITDs and KMT2A-PTDs.

More extensive data will be presented in the poster.


PHP provides comprehensive genomic analysis in a single workflow across a wide range of HM. In addition to reporting clinically relevant SNVs, indels, SVs, and CNVs, the panel will facilitate the characterization of key rearrangements relevant to bespoke measurable residual disease analysis in ALL and AML. Also included in the design are targets which will assist in a differential diagnosis from MDS, and genes implicated in familial haematological neoplasia.