Improved Genetic Characterization of Aggressive B-Cell Lymphoma at Diagnosis Using Next-Generation Sequencing

Introduction

Large B cell lymphoma (LBCL) is the most common aggressive lymphoma, including varieties such as diffuse large (DLBCL) and high-grade B cell lymphoma (HGBCL). Genomic rearrangements in the MYC locus can occur in all these histologies and are associated with inferior outcomes. Nowadays, in-situ hybridization (FISH) is universally used for genetically characterizing these patients. However, this strategy cannot identify the breakpoint location and partner, both dates that are important for prognosis. Likewise, other mutations are also necessary to properly classify LBCL. Next generation sequencing (NGS) increasingly seeks to find its place in the study of these patients. For this, we have designed a customized NGS panel (DX-NGS-LYMPHOMA), which includes the evaluation of fusion genes, point variants, and CNVs, with the intention of including it in clinical practice.

Materials and methods

We started with a group of 60 patients with DLBCL and HGBCL who had MYC, BCL6, or BCL2 rearranged by FISH. FISH was performed using break-apart probes (Vysis, Abbot Molecular). DNA was extracted from a paraffinized tissue sample using the Generead kit (Qiagen, USA). The DX-NGS-LYMPHOMA capture panel (Twist, USA) includes the following regions (Rearrangements: MYC, BCL2, BCL6, IGH, IGK, IGL, IRF4/DUSP22. Specific variants: EZH2, TNFRS14, NOTCH2, MYD88, NOTCH1, TP53, TCF3, ID3 and CCND3 CNVs: 11q, PDL1, REL). Library preparation was performed using the Kappa Hyperplus kit (Roche) and sequencing on a NextSeq or NovaSeq sequencer (Illumina). To confirm the results, we conducted a commercial panel EuroClonality-NDC assay (Univ8) Regarding analysis, reads were aligned to the human reference genome version GRCh38/hg38 using BWA and custom scripts. Low-quality reads (Q20) and PCR duplicates were removed using Picard tools. Variants and small indels were identified using the GATK algorithm. To identify structural variants, a combination of Manta software and custom Python scripts were used to identify breakpoints using soft clipping reads.

Results:

Point variants were identified in 100% of the patients; the most frequently altered genes were MYC (60%), TP53 (29%), EZH2 (27%), and BCL2 (15%).

NGS identified 85% of cases involving translocations, and we found the partner in all rearranged cases.

Thirty-six cases had rearranged MYC by FISH; NGS identified 28, but 6 were not identified by any of the two NGS panels (all in complete remission). Twelve breakpoints were in MYC locus, 1 upstream and 4 downstream. The partners identified for this group were 11 with IGH, 2 with DMD, 1 with ROCR3, 1 with RNF114, 1 with LPP and 1 with chromosome 5 (unknown gen). Non-genic rearrangements occurred all downstream of the MYC gene in PVT1 region, the partner identified were 1 with IGH, 3 with IGK, 1 with RNF11A, 1 with PRKDC, 1 with Chr7, 1 with chr8 and 1 with chr16 (unknown genes). In 11 patients, the FISH technique failed to identify MYC rearrangement; in two, the MYC rearrangement was identified in 2 by NGS. In no patient without MYC rearrangement by FISH, NGS identified rearrangement.

Among 18 cases with BCL2 rearranged by FISH, NGS identified all, and the partner was IgH, but in 3 FISH technique failures, one had BCL2 rearrangement, and three of the 46 patients in whom FISH did not identify rearrangement had it by NGS.

Regarding BCL6, of 15 patients with rearrangement by FISH, 11 had it by NGS. In 4 patients, BCL6 rearrangement could not be identified by the two NGS panels. In 1 of 3 patients who presented technical failure by FISH, the rearrangement was identified by NGS; in none, FISH did not show rearrangement. The partner identified were 8 IGH, 1 SETD3, 1 Chr6, 1 chr3 (unknow genes)

Conclusions.

NGS provides a comprehensive diagnostic characterization, utilizing only one methodology to identify various mutations that offer crucial information for patient diagnosis and prognosis. In cases where two different NGS panels have been unsuccessful in detecting the rearrangement, we may be dealing with a “false positive” result from the FISH test. Considering this, we are currently expanding our research to address this issue, and we are also working on CNV analysis. Additionally, we are working on improving the coverage of certain areas of our panel to optimize the results and correlate the clinical patient's outcome, with the ultimate goal of applying our panel in clinical practice.

Bastos-Oreiro:Genmab: Honoraria; Sobi: Honoraria; Astrazeneca: Honoraria; Janssen: Honoraria; Kite: Honoraria, Research Funding; Incyte: Honoraria; Lilly: Honoraria; Abbvie: Honoraria, Research Funding; Gilead: Honoraria, Research Funding; Takeda: Honoraria; BMS: Honoraria; Roche: Honoraria, Research Funding. Garcia-Sanz:Janssen: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Takeda Pharmaceutical: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; MSD: Honoraria; BMS/Celgene: Consultancy, Honoraria.