High-Resolution Architecture and Partner Genes of MYC Rearrangements in Lymphoma with DLBCL Morphology

Background

Genomic structural rearrangements of the MYC gene locus are reported in ~12% of lymphomas with diffuse large B-cell morphology (DLBCL) and are associated with inferior outcomes. Concurrent rearrangement of MYC and either the BCL2 or BCL6 locusis seen in ~8% of DLBCL cases. These lymphomas are included in the revised WHO disease entity high grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangements (HGBL-DH), and their observed poor prognosis is considered sufficient for treatment guidelines to recommend dose intensive chemotherapy regimens. The majority of studies that have examined MYC rearrangements in DLBCL have utilized FISH assays, which can inform on the presence or absence of rearrangements but lack the resolution to identify breakpoint locations and characterize breakpoint architecture.

Methodology

We identified 112 patients with tumors of DLBCL morphology (95 de novo DLBCL and 17 transformed from follicular lymphoma) that were reported to harbor rearrangements in the MYC region based on FISH break-apart assays. Cases were selected if they were break-apart positive using VYSIS and/or DAKO probes. Seventy-five cases were HGBL-DH by FISH (49 with BCL2 rearrangement; 11 with BCL6 rearrangement; 15 with both). We designed a hybrid capture assay to perform targeted sequencing in regions of interest including the MYC, BCL2, BCL6, and immunoglobulin ( IGH, IGL, IGK ) loci using DNA derived from formalin-fixed paraffin-embedded biopsies. We used multiple structural variant detection tools in an ensemble approach to identify predicted genomic rearrangements in MYC and the other loci of interest. These predictions describe the exact location of the breakpoint along with the identity of the partner locus.

Results & Discussion

Predicted MYC rearrangements matching the observed FISH status were identified in 93 cases (83%). Possible reasons for the remaining rearrangements not being identified include low quality input DNA, insufficient sequencing depth, or breakpoints located in hard-to-align repetitive regions. Two types of breakpoints were seen in the MYC region, those clustering in close proximity to the MYC coding sequence (the "genic cluster" spanning 1kb upstream of the transcription start site to the end of intron 1), and those outside of this cluster. The vast majority (76%) of the breaks outside the genic cluster were located telomeric of MYC up to 0.51Mb downstream. Translocations in the genic cluster were highly enriched for IGH partners (32/38; 84%), while numerous partners were identified in the rearrangements outside this region, mostly non- IGH (48/55; 87%). The most frequent non- IG partners were the BCL6 and PAX5 / ZCCHC7 loci - known recurrent MYC rearrangement partners. One novel recurrent partner was identified, the RFTN1 gene (chr3p24.3; n= 4). In cases where MYC was the sole translocation, the majority were MYC-IGH rearrangements (17/28; 61%) and most of these were in the genic cluster (14/17; 82%). Conversely, HGBL-DH cases more frequently harbored MYC -non- IGH translocations (43/63; 68%).

Many FISH break-apart probe assays (e.g. the DAKO assay) were designed for Burkitt lymphoma, where most breakpoints occur within the genic cluster. Of note, many of the MYC rearrangements with non- IGH partners have breakpoints located telomeric to the binding region of these probes, making the rearrangements undetectable using these assays. The DAKO assay failed to detect the MYC rearrangement in 32% (19/59) of the HGBL-DH cases evaluated in this study.

Breakpoints in the IGH locus clustered according to genomic features, with BCL2 translocations occurring in the VDJ segment regions and MYC rearrangements occurring upstream of constant regions. The location of the IGH breakpoints with respect to enhancer elements suggests that MYC over-expression is likely driven by 3'RR elements.

Conclusions

To our knowledge, this study is the largest to date assessing MYC rearrangements at base-pair resolution in lymphomas with DLBCL morphology. We characterized breakpoint architecture and identified rearrangement partners in 93 cases, and described the relationship between breakpoint location and partner locus. Our work reinforces the importance of FISH probe design in identifying MYC rearrangements in patient samples.