A 220-Gene Targeted Next-Generation Sequencing Panel for the Detection of Variants in Diffuse Large B-Cell Lymphoma, Follicular Lymphoma, and Mantle Cell Lymphoma: Application to a Cohort of 85 Formalin-Fixed Paraffin-Embedded Diffuse Large B-Cell Lymphoma Biopsies

Background:Recent advancements in comprehensive sequencing technologies has enabled researchers to uncover numerous somatic variants within many disease types, but their respective contributions to disease pathogenesis and potential roles as outcome biomarkers are less well-defined.  Even fewer studies have integrated somatic mutation events with genomic gain and loss events of respective genes, mostly due to the need to utilize a second platform to robustly assess genomic copy number. To further understand the roles of these genomic aberrations in the disease biology of mature B-cell neoplasms, we developed a next-generation sequencing (NGS) panel of 220 genes for profiling of Diffuse Large B-cell Lymphoma (DLBCL), Follicular Lymphoma (FL), Mantle Cell Lymphoma (MCL), and secondarily for Chronic Lymphocytic Leukemia (CLL).

Methods: The 220 genes represented in the panel included genes based on reported frequency of mutations, targets of therapeutic agents, involvement in disease-enriched pathways, prognostic value, and associated with targeted therapy resistance. Also included were genes that mapped to sites of genomic gain and loss commonly observed in these neoplasms. Of all the genes, 163 were enriched for DLBCL, 144 for FL, 44 for MCL, and 107 for CLL. Probes were designed to cover the coding exons of the 220 genes using a hybrid-capture approach (Nimbledesign, Roche, Inc.), with a total capture size of 0.92 Mb comprising 4086 target regions.  Optimization of DNA fragmentation and library preparation was performed for DNA extracted from formalin-fixed paraffin-embedded (FFPE) biopsy material, being the most common intended tissue type and also for which tumor burden could be assessed.  Sequencing was performed using a Miseq sequencer (Illumina, Inc.) with the goal to enable detection of somatic variants down to an allele variant frequency (AVF) of 5%. CLC Genomics Workbench (Qiagen, Inc.) was used for alignment and variant calling.  Filtering was performed to exclude synonymous, intronic, and untranslated region variants, variants within homopolymer regions (>6), and variants detected below 5% AVF.

Results: The average depth of coverage achieved across an initial sample set of over 50 FFPE DLBCL specimens was approximately 350X, with only one sample having more than 10% of the target regions not achieving a coverage of 60X or greater (within 95% of the target region).  As part of a proof-of-concept study, we sequenced a cohort of 85 DLBCL FFPE samples from a single institution from patients undergoing biopsy as part of their routine clinical care. Clinical outcome, cell-of-origin subtype (by the Hans method) and genomic gain/loss data by array-CGH were available.  All studies were performed with IRB approval. Of the 85 cases, 80 had also been submitted to bi-directional Sanger’s sequencing for Exons 5-8 of TP53 to establish accuracy of variant detection. Of cases completed to date, additional TP53 mutations have been detected by NGS due to an AVF below that detectable by Sangers or only detected in one direction. In addition, of the 14 cases completely sequenced and filtered to date, EZH2 variants in four samples have been detected in the GCB subtype while four CD79B variants were in samples of the non-GCB subtype, consistent with previously reported associations. Furthermore, six of the eight samples which harbored an EZH2 or CD79Bvariants had mutations within the catalytic domain at Y646 (EZH2) or the ITAM residue Y196 (CD79B), both of which have clinical implications.  Initial assessment of the ability of the panel to infer copy number showed concordance with aCGH in detecting large chromosomal aberrations. The sensitivity and specificity of this approach is currently being optimized.

Conclusion: Overall then, a hybrid-capture NGS panel has been developed, targeting genes enriched for involvement in the most frequent subtypes of mature B-cell neoplasms and optimized for application to FFPE biopsy specimens. While emerging pathogenomic patterns observed in the cases sequenced to date have confirmed previously observed patterns, completion of the NGS analysis of the 85 DLBCL cohort will permit more complex integrated analysis of somatic variants and genomic gain/loss with pathologic and clinical outcome data.