Applications of whole-genome sequencing in hematologic malignancies: Evaluating myeloid and lymphoid cancers in the Genome Era

Introduction

In 1973, Janet Rowley at the University of Chicago uncovered the cytogenetic basis of chronic myelogenous leukemia (CML)—in nearly every cell she studied, she found a translocation between chromosomes 9 and 22. This discovery built on the identification of the Philadelphia chromosome in patients with CML in 1959 by Peter Nowell and David Hungerford, marking the first direct genetic link to cancer. In the 1990s, molecular characterization revealed that this reciprocal translocation event results in a fusion gene—BCR-ABL1, from the B-cell receptor BCR and the kinase domain of ABL1. The identification of the BCR-ABL1 fusion gene paved the way for targeted therapies. Imatinib (brand name Gleevec) was the first tyrosine kinase inhibitor developed to target the BCR-ABL1 protein in CML. Approved by the Food and Drug Administration in 2001, it marked a new era in cancer treatment, achieving unprecedented success in CML patients with a five-year survival rate of 89%.

Most cases of CML—over 95%—are characterized by the typical t(9;22) translocation. Other myeloid malignancies, like acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS), and lymphoid malignancies such as multiple myeloma (MM) and chronic lymphocytic leukemia (CLL) are characterized by genetic heterogeneity with a wide variety of variants. For instance, in AML, key genetic findings include recurrent chromosomal translocations, such as t(8;21), inv(16), and t(15;17), and mutations in genes like FLT3, NPM1, CEBPA, IDH1/2, and DNMT3A, which define specific AML subtypes with distinct prognostic implications. These genetic alterations influence disease behavior, response to treatment, and risk stratification. This diversity of genetic findings is typical of other myeloid and lymphoid blood cancers.

The generalized workflow in leukemia management involves initial diagnosis through blood counts and bone marrow biopsy, classification, risk stratification using genomic tools, and treatment through chemotherapy, stem cell transplantation, and maintenance therapy. Appropriate classification and risk stratification are critical to patient management and essential to key clinical decision-making regarding stem cell transplantation and other interventions.

Current genetic and molecular testing in blood cancers follows a multimodal approach, and these lab findings are used to classify and risk-stratify disease according to World Health Organization and other clinical guidelines. Conventional cytogenetics (karyotyping) is used to detect large chromosomal changes, while fluorescence in-situ hybridization (FISH) helps identify specific rearrangements and cryptic translocations. Polymerase chain reaction (PCR) and reverse transcription PCR detect specific gene mutations (for example, FLT3, NPM1) and fusion transcripts, aiding in diagnosis and minimal residual disease (MRD) monitoring. Next-generation sequencing, including targeted panels, offers comprehensive mutation profiling for risk stratification. Chromosomal arrays and optical genome mapping methods are used in selected situations.

Whole-genome sequencing (WGS) has been shown to be an efficient alternative by providing comprehensive genomic profiling that detects all relevant genetic abnormalities, including chromosomal changes, structural variants, and cryptic mutations that can be missed by traditional methods like cytogenetics and FISH. This allows for better efficacy in classification and risk stratification, and one study in AML/MDS demonstrated that nearly 25% of patients had additional diagnostic findings and around 17% received a different risk stratification when WGS was used and compared with current standard testing modalities.¹ WGS can also enable emerging areas, such as discovery of targets for plasma-based MRD-enabling pathways, to move away from using bone marrow aspirates for longitudinal testing. WGS streamlines the diagnostic process by consolidating multiple tests into a single assay under a unified workflow, with a more efficient turnaround time of around five days. While historically WGS has been considered an expensive test, decreasing sequencing costs are making it more economically viable. Additionally, the methodology has been included in National Comprehensive Cancer Network guidelines for AML/MDS management and has reimbursement coverage under Medicare.

How it works

Existing Illumina products and workflows can be used to generate comprehensive and high-quality genomes from specimens obtained from subjects with hematological malignancies. Depending on the research context, DNA extracted from peripheral blood and/or bone marrow aspirate specimens is suitable as input. For detection of somatic variants, deeper average coverage is recommended, as critical events may be present in only a subset of cells. To achieve higher sequencing depths, pooling is typically performed with a relatively small number of libraries. Because of this, care should be taken when selecting sample indexes to avoid data loss due to adapter sequence similarity. Refer to our Index Adapters Pooling Guide or reach out to your local Illumina support team for more information.

Category	Name	Description
Library preparation	Illumina DNA PCR-Free Prep Illumina DNA Prep	Standard WGS library prep
Sequencers and flow cells	NovaSeq 6000 with S2/S4 NovaSeq X and NovaSeq X Plus with 10B/25B	High-capacity genome sequencing
Bioinformatics	DRAGEN v4.3 Somatic with Heme Recipe	Mapping and variant calling tuned for heme WGS research
Interpretation and reporting	Illumina Connected Insights	Tertiary analysis platform for oncology applications

Table 1: Available Illumina products empowering WGS research for hematologic malignancies

Full vision

Comprehensive genomic analysis of hematologic malignancies requires the evaluation of multiple variant types, including small variants (single-nucleotide variants and insertions-deletions smaller than 50 base pairs), copy number alterations (CNAs), structural variants (SVs) such as translocations, and loss of heterozygosity (LoH) calculation. Using a combination of variant calling, depth of coverage analysis, break-end detection, and other algorithms, the DRAGEN Somatic WGS Heme Tumor Only recipe provides accurate and comprehensive results for variants associated with hematologic malignancies.

The ability to detect somatic variants is highly dependent on the sequencing depth of the sample. In Figure 1, we performed titration experiments on DNA extracted from cancer reference cell lines (Seraseq Myeloid Mutation DNA Mix, NOMO-1, Kasumi-1, and HCC1187) to calculate the limit of detection (LoD) of variants across a range of coverage depths. A coverage of 140× provides a limit of detection of small variants with an allele frequency greater than or equal to 5%. Table 2 further details LoD for other variant classes when evaluated at 140× for consistency. Depending on the needs for a specific research application, the LoD can be shifted by increasing or decreasing sequencing coverage as shown in Figure 2.

Analytical sensitivity (limit of detection)
Average coverage	140×
Small variant (VAF)	0.05
Structural variant LoD (VAF)	0.07
0.5 Mb–5 Mb CNA deletion (fold change)	0.09
0.5 Mb–5 Mb CNA duplication (fold change)	1.09
5 Mb–10 Mb CNA duplication (fold change)	1.07
0.5 Mb–5 Mb LoH (tumor purity)	0.17
5 Mb–10 Mb LoH (tumor purity)	0.15

Table 2: Analytical sensitivity (limit of detection) for small variants, structural variants, CNA, and LoH established at 140×

In collaboration with researchers, we performed PCR-free WGS at an average of about 220× coverage on 23 clinical samples with hematological malignancies (Figure 3). Our team correctly called all clinically relevant small variants (n=73, including four FLT3-ITD), structural variants, and more than 96% of CNAs across this cohort. Specifically, two inversions approximately 45 Mb and 56 Mb in length, eight translocations, 22 CNAs ranging from 4.5 Mb to 100 Mb, and seven whole-chromosome events were evaluated in total.  

To further evaluate the performance of WGS, we evaluated an additional 30 peripheral blood or bone marrow aspirate specimens from research participants with AML with variant calls from WGS compared against TruSight Oncology 500 results. The results of this study are summarized in Table 3 and show high performance for relevant variant types.

Analytical performance of AML samples
Number of AML samples	53
Average coverage	~200×
Small variant (SNV/indel) accuracy	> 99%
Structural variant / copy number alteration accuracy	> 95%
Analytical specificity	> 99.9%

Table 3: WGS assay performance in a research cohort at about 200× coverage

The Genome Era

According to a paper published in the journal Blood, “In patients with myeloid neoplasms, whole-genome sequencing represents a potential replacement for both conventional cytogenetic and sequencing approaches, providing rapid and accurate comprehensive genomic profiling.”²

Illumina’s comprehensive whole-genome sequencing and informatics solutions for hematologic malignancies enables a unified workflow on a single platform, which improves laboratory efficiency and clinical efficacy compared to conventional methods. Every level of detail, from large chromosomal rearrangements to single-nucleotide variants, can be assessed accurately with this new technology platform. Contact your local account manager or sales representative for additional support.

Resources

illumina.com/products/by-type/sequencing-kits/library-prep-kits/illumina-dna-prep.html

illumina.com/products/by-type/sequencing-kits/library-prep-kits/dna-pcr-free-prep.html

help.dragen.illumina.com/product-guides/dragen-v4.3/dragen-recipes/somatic-wgs-heme-to

References

1. Duncavage EJ, Schroeder MC, O’Laughlin M, et al. Genome Sequencing as an Alternative to Cytogenetic Analysis in Myeloid Cancers. N Engl J Med. 2021;384(10):924-935. doi:10.1056/NEJMoa2024534
2. Duncavage EJ, Bagg A, Hasserjian RP, et al. Genomic profiling for clinical decision making in myeloid neoplasms and acute leukemia. Blood. 2022;140(21):2228-2247. doi:10.1182/blood.2022015853