Novel Genomic Alterations in MCL1 and ARID1A Identified in Pediatric Burkitt Lymphoma Using Targeted High-Throughput Sequencing

Abstract 899

While the majority of children with Burkitt lymphoma (BL) are cured with conventional chemotherapy, outcome for patients with relapsed disease is poor (overall survival <30%). Cure rates are lower in developing countries, which bear the brunt of this disease, including the endemic form in sub-Saharan Africa. Targeted therapy would be of benefit to all patients with BL to improve outcomes and decrease the reliance on conventional toxic chemotherapy. In order to develop such therapies a comprehensive understanding of the BL cancer genome is needed. The hallmark of BL is the translocation of MYC to either the immunoglobulin heavy or light chains. In contrast to adult cases, pediatric tumors often have multiple additional cytogenetic abnormalities, the consequences of which are less well understood. In order to characterize the genomic landscape of pediatric BL and to investigate potential therapeutic targets, we performed targeted massively parallel sequencing on a cohort of primary pediatric BL samples.

Large genomic studies in rare tumors such as pediatric BL have been limited by the need for frozen tissue. In the current study we utilized a platform that has been adapted for genomic analysis of formalin-fixed, paraffin-embedded (FFPE) tissue, which allows the use of archival FFPE cases (Lipson, et al Nat Med. 2012 Feb 12;18(3):382-4). Targeted genomic sequencing was performed on 29 pediatric BL cases evaluating 3,230 exons of 182 cancer related genes and 37 introns from 14 genes often rearranged in cancer. Tumor samples were collected at diagnosis from patients in the US, Brazil, and Kenya and included the sporadic (n=24), endemic (n=3) and HIV associated (n=2) forms of disease.

At an average coverage of 653-fold, 89% of cases were found to have at least one genetic alteration. The number of observed alterations correlated with EBV status. EBV negative cases demonstrated significantly more genomic alterations than EBV-positive cases. Among the EBV(–) cases, 13/15 had >1 alteration compared with 2/13 EBV(+) cases with >1 alteration (p<0.001). This is consistent with a tumorigenic role of EBV in BL. The most common mutations observed were in MYC (58.6%) and p53 (41.4%). MYC mutations spanned the coding region and included hot spots previously documented in lymphomas. Cases with MYC mutations also demonstrated MYC translocations, confirming that mutations may functionally cooperate with translocation to promote MYC-mediated oncogenesis.

Novel recurrent alterations were identified in the chromatin remodeling gene, ARID1A as well as the anti-apoptotic gene, MCL1. ARID1A is a member of the SWI/SNF family of complexes that regulate chromatin structure and has been implicated as a tumor suppressor in a variety of solid tumors, but has not yet been described in BL. We identified truncating mutations in 5/29 cases (17.2%), including one case that had a secondary mutation in SWF5, also a member of the SWI/SNF family. Immunohistochemistry for ARID1A demonstrated decreased expression in cases where the truncating mutation was upstream of the antibody epitope.

MCL1 is a member of the Bcl2 family of apoptotic proteins and has been implicated as an oncogene in lymphomas. Of the 29 cases sequenced, 6 (20.7%) had alterations in the MCL1 pathway including amplification of MCL1 (n=5) and point mutation in FBWX7, a ubiquitin ligase that targets MCL1 for degradation (n=1). Amplification of MCL1 was confirmed by FISH. In an independent cohort, 5/17 (29.4%) cases demonstrated MCL1 amplification by FISH. MCL1 protein expression, as evaluated by immunohistochemistry, was increased in 19/40 cases including the cases known to have MCL1 amplification.

This study is the first genomic analysis of BL using FFPE tissue and demonstrates the feasibility of next generation sequencing of rare lymphomas using archival FFPE tissue. Our findings highlight the potential role of anti-apoptotic and chromatin remodeling genes in BL pathogenesis.