ALBERT

Description: Peripheral T-cell lymphomas (PTCLs) are a heterogeneous group of non-Hodgkin lymphomas frequently associated with poor prognosis and for which genetic mechanisms of transformation remain incompletely understood. Using RNA sequencing and targeted sequencing, here we identify a recurrent in-frame deletion (VAV1 Δ778–786) generated by a focal deletion-driven alternative splicing mechanism as well as novel VAV1 gene fusions (VAV1-THAP4, VAV1-MYO1F, and VAV1-S100A7) in PTCL. Mechanistically these genetic lesions result in increased activation of VAV1 catalytic-dependent (MAPK, JNK) and non–catalytic-dependent (nuclear factor of activated T cells, NFAT) VAV1 effector pathways. These results support a driver oncogenic role for VAV1 signaling in the pathogenesis of PTCL.

Description: Introduction Extranodal NK/T-cell lymphoma (ENKTL) is a rare disease; in Western countries it represents less than 1% of all Non-Hodgkin lymphoma. When localized, ENKTL is associated with a good prognosis. In contrast, patients (pts) with a disseminated disease still have a dismal outcome despite the use of asparaginase (ASPA) containing regimens that have significantly improved the prognosis of this lymphoma. Production of neutralizing anti-ASPA antibodies leading to inactivation of the enzyme can reduce its activity. Inactivation of ASPA activity is correlated with a worse prognosis in acute lymphoblastic leukemia (ALL). Monitoring of ASPA activity is therefore recommended in ALL pts with 100 IU/L as threshold value of serum enzymatic activity considered to be sufficient for complete depletion of l-asparagine in serum. However, ASPA monitoring is not routinely performed in pts with ENKTL, despite the frequent use of ASPA-containing regimens. The main objective of this study was to determine the proportion of pts with an insufficient ASPA activity corresponding to production of neutralizing anti-ASPA antibodies, risk of allergic reaction and inefficacy of the drug. Methods Adult pts with histologically confirmed ENKTL who received an ASPA-containing regimen between 2014 and 2018 and had a monitoring of ASPA activity were included. ASPA activity measurement was usually performed with a quantitive enzyme assay 48 hours after the last ASPA injection of each cycle for native forms of ASPA and 14 days after injection of the pegylated form. Activity below 100 UI/L was considered as insufficient. ASPA activity was correlated with pts outcome. The choice of initial form of ASPA and reasons to switch between two forms of ASPA were also analyzed. Results From 2014 to 2018, a total of 21 pts received an ASPA-containing regimen and were monitored for ASPA activity. Median age was 53 years with 14 men and 7 women. More than half of these pts had a stage IV disease (n=11/21), 6 were in stage I and 4 in stage II. Fifteen pts were in first line and 6 pts were in relapse. Pts have received either native e-coli L-asparaginase (Kidrolase®: KID) (n= 13 in the treatment-naïve group, n=2 in the relapsing group) or a pegylated form of e-coli- L-asparaginase (Oncaspar®: ONC) (n=2 in the treatment-naïve group, n=4 in the relapsing group). Most of the pts received ASPA (8 injections at each cycle for native forms and 1 injection for the pegylated form) associated with gemcitabine, methotrexate and dexamethasone, plus oxaliplatine for disseminated diseases. Six pts had optimal ASPA activity, 3/6 had localized disease: 2 had persistent RC and 1 relapsed, 3/6 had a disseminated disease: 2 progressed during treatment and 1 relapsed. Fifteen pts displayed low ASPA activity, 12/15 pts with a KID containing regimen after 1 cycle (n=10) or 3 cycles (n=2) and 3/6 pts with an ONC containing regimen after the first cycle (n=2) or the second (n=1). Among these 3 pts, 2 have been previously treated with KID that could induce an immunization against ONC. Ten pts received a second form of asparaginase: Erwinia asparaginase (Erwiniase®: ERW) in 9 pts and ONC in 1 pt. In 9/10 cases, this switch was justified by detection of low ASPA activity. Measurement of enzymatic activity in these 9 pts showed satisfactory levels in 2/5 pts treated with ERW and monitored for ASPA activity and in the pt receiving ONC, 4/6 pts with a localized form and 1/3 with a disseminated form are in persistent RC. Six pts with low ASPA activity including one case with a localized form and 5 cases with a disseminated disease, did not receive another form of ASPA, they all progressed or relapsed. Conclusion More than 2/3 of pts had sub-optimal ASPA activity after 1 to 3 cycles of ASPA containing regimens. This finding is probably due to the development of anti-ASPA inhibitory antibodies and may explain the poor outcome of pts with a disseminated disease despite the remarkable efficacy of ASPA in this disease. Although the series reported here is small and heterogeneous, our results still suggest a better outcome in pts with good ASPA activity or in case of switch between ASPA molecules in the context of low activity. ASPA activity monitoring should be recommended in pts with ENKTL, to avoid allergic reaction and ineffective treatment by switching ASPA molecules. Pegylated forms of ASPA, or encapsulated in red cells may, be less immunogenic, should be used in the first line setting instead of native forms. Table. Table. Disclosures Bachy: Celgene: Consultancy; Janssen: Honoraria; Gilead Sciences: Honoraria; Takeda: Research Funding; Sandoz: Consultancy; Amgen: Honoraria; Roche: Research Funding.

Description: Peripheral T-cell lymphomas (PTCL) are malignant and highly aggressive hematologic tumors arising from mature post thymic T-cells. The diagnosis of PTCL includes diverse lymphoma subgroups, altogether accounting for about 15% of all non-Hodgkin lymphomas. Despite much effort in developing reliable diagnostic markers, the diagnosis of PTCLs is challenging and 20-30% of cases are diagnosed as PTCL-NOS (not otherwise specified). This heterogeneous and poorly defined group of lymphomas is frequently characterized by chemotherapy resistance and a very poor prognosis. Here we report the presence of recurrent driver activating genetic alterations in the VAV1 gene in PTCL, NOS. RNA-seq analysis of a comprehensive series of 154 PTCLs and targeted sequencing identified VAV1 gene fusions with different partners including VAV1-THAP4, VAV1-MYO1F and VAV1-S100A7. In all cases the resulting oncoproteins lack the C-terminal SH3 domain of VAV1, a motif implicated in the negative regulation of VAV1 signaling, leading to increased activation of VAV1 catalytic-dependent (MAPK, JNK) and non-catalytic-dependent (NFAT) VAV1 effector pathways. In addition, and most notably, we also identified focal microdeletions at the VAV1 intron 25-exon 26 boundary, which result in the activation of an alternative intraexonic splice acceptor site and the consequent expression of mis-splicing-driven mutant transcripts harboring a recurrent VAV1 Δ778-786 in-frame deletion. Mechanistically, the VAV1 Δ778-786 mutation removes 9 amino acids proximal to the C-terminal VAV1 SH3 domain and induces in increased VAV1 activation and signaling in biochemical assays. In all, these results support a driver role for oncogenic VAV1 signaling in T-cell transformation of major importance for the design of targeted therapies for the treatment of PTCL, NOS. Disclosures No relevant conflicts of interest to declare.

Description: We recently reported a truncating deletion in the NFKBIE gene, which encodes IκBε, a negative feedback regulator of NF-κB, in clinically aggressive chronic lymphocytic leukemia (CLL). Because preliminary data indicate enrichment of NFKBIE aberrations in other lymphoid malignancies, we screened a large patient cohort (n = 1460) diagnosed with different lymphoid neoplasms. While NFKBIE deletions were infrequent in follicular lymphoma, splenic marginal zone lymphoma, and T-cell acute lymphoblastic leukemia (

Description: Deregulated NF-κB signaling is a hallmark of most, if not all, lymphoid malignancies, and recurrent gene mutations in both the canonical and non-canonical NF-κB pathway are known to lead to NF-κB activation. However, the full compendium of NF-κB gene mutations in lymphoid malignancies remains to be elucidated. Recently, we reported a 4-bp truncating mutation in the NFKBIE gene, which encodes IκBε, a negative regulator of NF-κB, in patients with chronic lymphocytic leukemia (CLL). The NFKBIE deletion was enriched in clinically aggressive CLL patients (7-8%) and associated with a worse clinical outcome. At the functional level, NFKBIE-deleted CLL showed reduced IκBε levels and decreased p65 inhibition, along with increased phosphorylation and nuclear translocation of p65, compared to wildtype patients. Preliminary data has indicated an increased frequency of NFKBIE aberrations in other lymphoid malignancies as well. To explore this further, we screened for NFKBIE deletions in a large cohort of patients diagnosed with a range of different lymphoid neoplasms. Overall, NFKBIE deletions were identified in 76 of 1414 patients (5.4%). While NFKBIE deletions were relatively infrequent in patients diagnosed with follicular lymphoma (3/225, 1.3%), splenic marginal zone lymphoma (3/175, 1.7%), and T-cell acute lymphoblastic leukemia (1/94, 1.1%), moderate frequencies were observed among diffuse large B-cell lymphoma (18/521, 3.5%), mantle cell lymphoma (8/189, 4.2%), and primary CNS lymphoma (1/34, 2.9%) patients. In contrast, a remarkably high frequency of NFKBIE deletions (41/176 cases, 23%) was observed among primary mediastinal B-cell lymphoma (PMBL) patients. Noteworthy, the prevalence of NFKBIE-deleted PMBL cases was similar in the different contributing centers. All PMBL patients in the present series received a CHOP based treatment regime; in ~75% of cases rituximab was added and ~25% were treated with dose intensified schemes. For the latter, the vast majority of patients received CHOEP, while individual cases were treated with MegaCHOEP, DA-EPOCH or ACVBP. Regarding clinicobiological associations, there were no significant differences between NFKBIE-deleted and wildtype PMBL patients with respect to age, sex, Ann Arbor stage, IPI risk-groups, extranodal or bone marrow involvement, bulky disease, and LDH elevation. However, NFKBIE-deleted patients were more likely to be refractory to primary chemotherapy (31% vs. 3%, P=.001) and had a shorter overall survival compared to wildtype patients (5-year overall survival: 63% vs 84%, P=.013). In multivariate analysis (including age, gender, Ann Arbor stage, IPI, and NFKBIE mutation status), NFKBIE mutation status (95% CI: 1.23-10.61; HR: 3.61; P=0.020) remained an independent factor for poor prognosis. In summary, we document NFKBIE deletions as a common genetic event across B-cell malignancies, albeit at varying frequencies. The high frequency of NFKBIE deletions in PMBL alludes to the critical role of this aberration in the pathophysiology of the disease. NFKBIE deletions were associated witha worse clinical outcome, hence potentially representing a novel poor-prognostic marker in PMBL. *Contributed equally as first authors. **Contributed equally as senior authors. Disclosures Stamatopoulos: Gilead: Consultancy, Honoraria, Research Funding; Abbvie: Honoraria, Other: Travel expenses; Novartis: Honoraria, Research Funding; Janssen: Honoraria, Other: Travel expenses, Research Funding.

Description: Peripheral T cell lymphomas (PTCL) are a heterogeneous group of lymphoid tumors resulting from the malignant transformation of mature T-cells. Among these, angioimmunoblastic T-cell lymphoma (AITL), which accounts for almost 20% of PTCL cases, represents one of the most aggressive forms of non-Hodgkin lymphoma with a dismal survival rate of 30% at 5 years. AITL tumors are characterized by loss of function mutations in the Ten-Eleven Translocation 2 (TET2) epigenetic tumor suppressor and a highly recurrent mutation (p.Gly17Val, G17V) in the RHOA small GTPase gene, present in almost 70% of AITL cases. Immunohistochemical analyses and gene expression profiling studies have established follicular T helper cells (Tfh), a as the cell of origin of AITL. Tfh cells are a CD4+ T-cell population characterized by high levels of expression of the chemokine receptor CXCR5, the inhibitory receptor PD-1 and the transcription factor BCL6. To investigate the specific role and mechanisms of RHOA G17V in AITL we first generated a conditional inducible Rhoa G17V knock-in mouse by introducing a floxed Rhoa wild type minigene upstream of the Rhoa G17V mutant exon in the endogenous RhoA locus by homologous recombination. RhoA G17V conditional knockin mice express wild type RhoA but switch to express the mutant G17V allele upon expression of Cre recombinase. Crossing this line with a CD4CreERT2 strain, which expresses a tamoxifen-inducible form of Cre, results in tamoxifen inducible selective activation of Rhoa G17V specifically in CD4 T-cells. Given the close association of the RHOA G17V mutation with AITL we hypothesized that expression of mutant Rhoa G17V could promote Tfh cell polarization and transformation, explaining the high prevalence of this mutation in AITL cases. Notably, tamoxifen-induced expression of Rhoa G17V in CD4CreERT2 RhoA G17V knockin mice resulted in spontaneous accumulation of Tfh cells expressing Cxcr5, Pd-1 and Bcl6 in absence of immunization. Mechanistically, RhoA G17V Tfh cells expressed increased levels of the ICOS receptor, a major factor in Tfh development, and showed enhanced ICOS signaling. Moreover, suppression of ICOS signaling with an anti-ICOSL blocking antibody impaired Rhoa G17V-induced Tfh polarization. Next, we analyzed the specific interaction of RHOA G17V and loss of Tet2 in AITL. Expression of mutant RHOA G17V in hematopoietic progenitors from Tet2 knockout mice resulted in the specific development of AITL with a median latency of 14 months. Tet2-/- RHOA G17V tumors are clonal and transplantable and present typical AITL Tfh-like features including expression of CXCR5, PD-1, BCL6 and ICOS, and are critically dependent on ICOS signaling for growth and proliferation in vivo. In all, our results demonstrate an instructive role for the RHOA G17Vmutation in Tfh cell fate specification and in AITL transformation and support an oncogenic role for ICOS signaling in the pathogenesis of AITL. The AITL animal model reported here may facilitate the preclinical testing of new therapeutic approaches for the treatment of this disease. Disclosures No relevant conflicts of interest to declare.