Notable Patterns in the Genomic Landscape of Adult T-Cell Leukemia/Lymphoma Encountered in HTLV-1 Endemic Western World Regions

Background: Adult T-cell leukemia/lymphoma (ATLL) is an aggressive malignancy with dismal prognosis and associated with clonal T-cell expansion driven by Human T-Lymphotropic Virus 1 (HTLV-1) infection. Comprehensive genomic studies in Japan have identified recurrent alterations affecting TCR-NF-kB signaling (i.e. PRKCB, PLCG1, CARD11, VAV1, and IRF4), T-cell trafficking pathways (i.e. CCR4 and CCR7), and the tumor suppressor genes CDKN2A and TP53. HTLV-1 endemic regions include Africa, the Caribbean, and South America in addition to Japan. Retrospective studies from the Western population have reported distinctive features from the Japanese cohort, e.g. younger age, more common lymphomatous presentation, and worse outcomes. Our group sought to evaluate the unique molecular features of ATLL in a large cohort of patients from the Caribbean and South America.

Methods: We performed a multimodal genomic study on specimens from 169 patients encountered in the United States (Miami), Peru, Brazil, France, and Spain. Data types included Oncoscan/Copy Number Variation (CNV) data for 129 patients, RNA-seq data for 97 patients, and whole exome sequencing (WES) data for 125 patients. Patients were ethnically classified based upon single nucleotide polymorphisms, under 3 main groups: African (n=80), native American (n= 32), or South Asian/Islander (n=12). 46 specimens without WES data could not be ethnically classified. Somatic variants were called using Mutect. Putative driver mutations were identified by frequency-based criteria. CNV significance was determined using GISTIC2.0. Data were compared to whole exome and targeted sequencing data published by Kataoka et al.

Results: Our cohort replicated trends reported in Japanese datasets but included several distinctive findings. South American and Caribbean patients had fewer mutations in CCR4 and CD58. Three putative tumor suppressors not previously implicated in ATLL were identified based on recurrent damaging mutations. These included ANKRD11 (n=3), DGKZ (n=3), and PTPN6 (n=3). Both ANKRD11 and PTPN6 were only mutated in Afro-Caribbean patients with aggressive (acute and lymphomatous) cases. CNV analysis revealed ANKRD11 deletions in a significant portion of cases (n=16).

As previously reported, STAT3 mutations were more common in indolent subtypes. IRF4 mutations (n=14) or amplifications (n=19) were only observed in aggressive ATLL subtypes. L70V was the most common IRF4 variant (n=5). Among Japanese samples, K59R mutations were seen twice as often as L70V mutations. Both K59R and L70V are located within the IRF4 DNA binding domain. In samples from patients with disease relapse (N = 10), IRF4 was the only gene mutated significantly more often (p = 0.03). 3 of 10 patients who were previously treated with interferon (IFN)-based therapy relapsed with new IRF4 mutations (L70V n=2, and K59R n=1), suggesting IRF4 may be associated with IFN resistance.

A total of 11 patients in the Western cohort had FOXO3 mutations including R177W (n=7) and D199N (n=3) variants. These only occurred in Afro-Caribbean patients with aggressive subtypes. R177W is located within the FOXO3DNA binding domain, suggesting that dysregulation of its transcriptional targets may contribute to disease. In the Japanese cohort, only one patient had a FOXO3 mutation, (D199N). The majority of primary ATLL samples analyzed by Western Blot showed significantly reduced or no expression of FOXO3, or ANKRD11.

Conclusion: The genomic landscape of ATLL encountered in patients from South America and the Caribbean resembles that of ATLL in Japanese patients. However, our study identified novel variants and tumor suppressor genes not previously implicated in ATLL that differ from the Japanese population. Furthermore, we correlate our genomic analysis with clinical findings to implicate ATLL driver genes in IFN resistance and disease prognosis. Functional Studies to determine the prognostic and functional roles of the gene alterations we identified in ATLL are ongoing.