In immuno-oncology research, tumor mutation load is a promising marker with the potential to guide immunotherapeutic decisions (1). NGS-based approaches are the preferred methods to interrogate this marker in DNA extracted from tumor biopsies.
The QIAseq Tumor Mutation Burden Panel has been designed to target variants in 486 genes related to tumor and immune biology (Figure 1), covering a total 1.3 Mbp of DNA with Single Primer Extension (SPE) technology. The panel makes use of Unique Molecular Indices (UMIs), supporting sequencing error correction and reduction of PCR amplification bias during bioinformatic analysis. After DNA extraction, library preparation and sequencing, the NGS reads can be analyzed using the QIAseq Tumor Mutational Burden (TMB) ready-to-use bioinformatics workflows delivered with the Biomedical Genomics Analysis plugin for CLC Genomics Workbench.
Figure 1. Coverage of the QIAseq Tumor Mutational Burden Targeted TMB Panel.
Through a series of tools and filters, the TMB workflow reports genetic variants detected, computes the TMB score and provides a genome viewer for convenient investigation of variants and read mappings.
The TMB scores computed by the workflow correlate closely to whole exome sequencing (WES) results reported in two recent quality assurance initiatives focusing on TMB score harmonization across academic centers, pharma companies and diagnostic service providers (2, 3) (Figure 2).
The microsatellite instability (MSI) status of the sample can also be assessed if the MSI booster panel (cat no. SDHS-10101-11981Z-48) is used during sample preparation. The MSI algorithm compares the length profiles of selected microsatellite markers in the sample to reference profiles to assess the stability status at each MSI locus and in aggregate for all loci for the sample. A reference profile comes with the software; however, it is recommended that users construct their own reference profiles using tools provided. This is encouraged, as sequencing chemistries and laboratory procedures used may introduce experimental variation not captured in the provided reference profiles. For MSI calling, the nine quasimonomorphic mononucleotide repeats of the panel give the most consistent results, as also recommended by Buhard et al (4).
The workflows allow for calling copy number variants (CNVs) if appropriate normal reference samples are available and specified in the workflow.
Controlling for quality in the analysis workflow
UMI consensus reads are generated from the data. Using the mapping of these reads against the reference sequence forms the basis of the subsequent analysis, substantially reducing false-positive variant calls. In addition, filtering steps applied at several levels ensure that only quality results of interest are reported. These steps include removing reads originating from homologous genes and pseudogenes from the read mapping, and, after low variant frequency detection, removing variants likely to be artifacts, such as low-quality variants and germline variants (homozygous variants and variants considered common in the population as found in dbSNP). Only variants called at sites of 100X coverage or more are included in the final number, and the TMB score is then calculated by dividing with the sequence length of the panel covered at minimum 100X in mega base pairs (MBP).
The variants reported by the workflow can then be submitted to QIAGEN Clinical Insight Interpret (QCI Interpret) to identify pathogenic variants.
Figure 2. Variability in TMB estimates for each tumor cell line across all 15 participating laboratories (5). The QIAseq Tumor Mutation Burden Panel is LAB 12.
Data quality and TMB analysis
It should be noted that TMB can only be reliably scored on high-quality samples. The quality of the sample can be evaluated prior to library construction by checking the amount and purity of DNA obtained after extraction, as well as ensuring that the DNA fragment size peak after fragmentation is 300–350 bp. High-quality samples will have more than 1 MB of target region with greater than 100X coverage. This measure is reported in the QC for Targeted Sequencing report and the TMB report. Another indicator of high-quality samples is the average number of reads per UMI, reported in the UMI Group Report. This value should be between 2 and 4. Lower values (e.g., 1) or higher values (e.g., >5) are indicative of poor samples (too little DNA input, fragments are too short, or both). For other recommendations relating to QC values, please refer to the manual.
Availability and customization
Ready-to-use QIAseq TMB analysis workflows tailored for Illumina and for Ion Torrent are delivered with the Biomedical Genomics Analysis plugin for CLC Genomics Workbench and via a CLC Genomics Server plugin, supporting those with many samples to analyze, such as a hospital or other enterprise setting. By running analyses on CLC Genomics Server nodes, either dedicated job nodes or those of an existing HPC cluster, multiple samples can be analyzed in parallel.
As for all ready-to-use workflows, these workflows can be easily customized, as described in this tutorial. The TMB and MSI calling parts of the workflows are generic, and can be adapted to any other DNA panel workflow. For the development of TMB analysis workflows for Illumina’s TSO500 or panels from other vendors, please contact our Custom Solutions team.
1. Samstein R. et al. (2019) Tumor mutational load predicts survival after immunotherapy across multiple cancer types. Nature Genetics 51, 202–206.
2. Stenzinger A. et al. (2019) Tumor mutational burden standardization initiatives: Recommendations for consistent tumor mutational burden assessment in clinical samples to guide immunotherapy treatment decisions. Genes Chromosomes Cancer 58, 578–588. doi:10.1002/gcc.22733
3. Friends of Cancer Research [press release]. Friends of Cancer Research Announces Launch of Phase II TMB Harmonization Project. September 18, 2018. https://www.focr.org/news/friends-cancer-research-announces-launch-phase-ii-tmb-harmonization-project
4. Buhard O. et al. (2004) Quasimonomorphic Mononucleotide Repeats for High-Level Microsatellite Instability Analysis. Disease markers 20, 251–257. doi:10.1155/2004/159347.
5. Merino DM. et al. (2019) TMB Standardization by Alignment to Reference Standards Phase 2 of the Friends of Cancer Research TMB Harmonization Project. ASCO Meeting Library; Presented Saturday, June 1, 2019.https://meetinglibrary.asco.org/record/172797/abstract