Examination of Genetic Aberrations in Chlamydophila Psittaci negative MALT Lymphomas of the Ocular Adnexa

Abstract 1569

Ocular adnexal mucosa associated lymphoid tissue lymphomas (OAMALTL) are the most common lymphomas of the eye. The etiology and pathogenesis of ocular adnexa MALT lymphomas (OAMALTL) are still unknown and the association with Chlamydophila psittaci (C. psittaci) has been shown in only some geographic regions. Only few small studies specifically examined for the presence of t(11;18)(q21;q21)/API2-MALT1, t(1;14)(p22;q32)/BCL10-IGH, t(14;18)(q32;q21)/IGH-MALT1, t(3;14)(p13;q32)/FOXP1-IGH translocations or for TNFAIP3 (A20) mutations/deletions, which may contribute to the activation of canonical nuclear factor (NF)-kB pathway in OAMALTL. Herein we sought to comprehensively examine the frequency of these translocations as well as CARD11 and MYD88 (L265P) mutations, in addition to A20 mutations /deletions in a large cohort of C. psittaci negative OAMALTL. A total of 47 OAMALTL originating in the orbit (22), conjunctiva (19) and lacrimal gland (6) were used for analysis. Dual color fusion probes for t(14;18)(q32;q21) IGH-MALT1, t(11;18)(q21;q21) BIRC3-MALT1 and dual color break apart probes for MALT1, BCL6, and CEP18 (chromosome 18 centromere) were used for FISH analysis in all the analyzed tumors. Dual color break apart probe for IGH was used for selected tumors. Extracted DNA was used for PCR amplification and sequencing of the coiled-coil domain of CARD11, exon 5 of MYD88, in which the L265P mutation was previously reported in diffuse large B-cell and MALT lymphomas, and all the coding exons of the A20 gene. A20 gene copy number variation was analyzed by TaqMan Copy Number Assay (Applied Biosystems). DNA extracted from peripheral blood lymphocytes of 3 healthy volunteers and OCI-LY8 and Jeko-1 cell lines served as calibrator and positive controls for A20 deletion. CARD11 mutations were not found in all the analyzed tumors. The MYD88 L265P mutation was detected in 3 (6.4%) tumors. A total of 9 A20 mutations were identified in 7 (14.9%) tumors. One tumor harbored 3 distinct mutations (F149C; 1bp deletion in exon 3 and 2bp deletion in exon 5). Among the 9 detected A20 mutations, the majority (89%) would produce truncated proteins due to out-frame insertion/deletion (7), while one of these deletions was located in a known splicing site. Only 2 missense mutations were observed, including one in a tumor in which 2 concomitant deletions were also present (described above). A20 heterozygous deletions were detected in 7 (14.9%) tumors. There was no association between A20 mutations and heterozygous deletion in any of the analyzed tumors. None of the tumors harbored a concomitant A20 mutation/deletion and MYD88 L265P mutation.

A total of 5 tumors harbored chromosomal alterations: additional copy of BCL6, most probably due to trisomy 3, in 2 tumors, an additional copy of IGH in 1, extra copies of both IGH and MALT1 in 1, and extra copies of IGH, MALT1, BIRC3 together with IGH rearrangement to unidentified partner in 1. t(14;18)(q32;q21) IGH-MALT1 and t(11;18)(q21;q21) BIRC3-MALT1 were not detected in any of the analyzed tumors. A20 mutations were detected in a tumor with extra signals of both IGH and MALT1 genes as well as in one of the tumors with an additional copy of BCL6. The tumor with complex chromosomal aberrations including the IGH rearrangement to an unidentified partner also harbored A20 deletion. 26 of the analyzed tumors previously were shown to exhibit evidence of antigen selection based on the analysis of IGHV mutation pattern (PLoS One. 2011;6(12):e29114). There was no association between antigen selection and A20 mutations/deletions, MYD88 (L265P) mutation and chromosomal alterations. Overall, our data suggest that translocations characteristic for MALT lymphomas are rarely observed in OAMALTL. The MYD88 (L265P) mutation is also uncommon, while most of the A20 mutations/deletions affect only one allele and thus most probably do not lead to NF-kB activation. This raises a question on what are the mechanisms for canonical NF-kB signaling pathway activation in OAMALTL and if it is indeed activated in these tumors. It is possible that NF-kB signaling pathway activation maybe due to B cell receptor signaling, as may be reflected by association with presence of antigen selection that is observed in major fraction of these tumors. Immunohistochemical studies addressing these questions are underway and will be presented at the meeting.