ALBERT

Description: Journal of the American Chemical Society DOI: 10.1021/ja408815k

Description: Abstract 536 Introduction: Mantle cell lymphoma (MCL) and chronic lymphocytic leukemia (CLL) are B-cell malignancies of different postulated origin, genetics, clinical presentation and prognosis. Several studies have reported that both MCL and CLL individually exhibit aberrant methylation in comparison to normal B-cells. However, a comprehensive comparison of the methylation profiles of these two B-cell disorders has not been performed yet. This strategy has the potential to identify cellular pathways and genes that are specifically targeted in each disease. Methods: We applied the genome-wide Illumina Infinium HumanMethylation27 BeadChip array (Illumina, San Diego, USA) which measures methylation levels at 27,578 CpG dinucleotides covering 14,495 genes, to compare the methylation profiles in: (i) 20 MCL cases; and, (ii) 30 CLL cases, 15 each with unmutated stereotyped subset #1 (IGHV1-5-7/IGKV1(D)-39) B cell receptors (BCRs) or mutated stereotyped subset #4 (IGHV4-34/IGKV2-30) BCRs, where these two subsets represent prototypes of unmutated and mutated CLL. The methylation status for each detected CpG site ranged between 0.1 (completely unmethylated) to 1 (completely methylated). Results: As expected, major differences in methylation patterns between MCL and CLL were observed. When the methylation profiles of the two entities were compared, 51 genes were identified as differentially methylated in all comparisons (MCL versus both CLL subsets combined and each subset separately). Among the 19 genes highly methylated in MCL were six (32%) homeobox or homeodomain-containing transcription factors (e.g. POU4F1, PITX3), whereas genes enhancing cell proliferation and tumor progression such as MERTK and CAMP were hypomethylated in MCL. Of the 32 genes hypermethylated in CLL were six pro-apoptotic genes, including DYRK2 and CYFIP2, the tumor suppressor PRDM2 and the cell cycle regulator CCND1. Conclusions: We report for the first time disease-biased methylation profiles for different functional classes of genes in MCL or CLL. Homeobox genes were highly methylated in MCL, whereas CLL was characterized by methylation of apoptosis-related genes. The identified differences in global methylation profiles between MCL and CLL may assist in unfolding distinct epigenetic silencing mechanisms involved in the pathogenesis of these B-cell malignancies. Disclosures: No relevant conflicts of interest to declare.

Description: According to WHO’s current criteria, follicular lymphoma is morphologically divided into grades 1, 2, and 3, and grade 3 is further subdivided into 3a and 3b, depending on the presence of centrocytes and solid sheets of centroblasts. The clinical relevance of this grading system is unclear. Although there is consensus that grades 1 and 2 share clinical characteristics of indolent lymphoma, it is debated whether grade 3 should be treated intensively with anthracycline-containing regimens. Furthermore, the 3a/3b subdivision is not universally implemented. We wanted to investigate the clinical relevance of the grading system with respect to clinical characteristics, treatment and survival. Totally, 186 cases were diagnosed between January 1994 and January 2004 in South Stockholm County with de novo follicular lymphoma and without concomitant transformation. The biopsies were re-graded according to the current WHO criteria by two experienced hematopathologists. 129 cases were grade 1 or 2, 44 grade 3a, and 13 grade 3b. Clinical characteristics in the patients with follicular lymphoma grade 1–2 and 3a are shown in Table 1. The local policy has been to treat all patients with grade 3 disease aggressively. In grade 3a disease, front-line anthracyclines did not seem to improve overall survival (Figure 1). Grades 1–2 and 3a show similar overall survival curves of indolent and incurable lymphoma, while the 13 patients with grade 3b disease (of whom 12 received aggressive front-line therapy) seem to reach a plateau after four years (Figure 2). We conclude that follicular lymphoma grade 3a is not clinically different from grade 1 and 2 disease, and that patients with grade 3a do not seem to benefit from front-line anthracyclines. Although our findings are limited by the few cases of grade 3b, it seems that grade 3b is another disease entity and, when treated aggressively, it is maybe curable with chemotherapy. The histopathological distinction between grade 3a and 3b is clinically of great importance. Table 1 Grade 1–2 Grade 3a P Clinical characteristics in patients with follicular lymphoma grade 1-2 and 3a. Age 〉60 years 50% 61% NS Elevated LDH 32% 49% 0.046 Stage III or IV 68% 80% NS Front-line anthracyclines 41% 73% 0.0004 Subsequent transformation 21% 19% NS Figure Figure Figure Figure

Description: Abstract 4618 Background: In CLL the prognostic importance of cytogenetic abnormalities in bone marrow or peripheral blood is well-known. However data is lacking on the frequency and importance of the aberrations in other tissues such as lymph nodes and spleens. Aims: To develop a method to assess the chromosomal aberrations commonly seen in CLL (13q-, 11q-, 17p- and +12) on paraffin-embedded tissue sections from spleens and to relate the findings to data on abnormalities found in bone marrow or peripheral blood. Material and methods: 24 out of 61 CLL patients splenectomized between 1990–2009 have so far been analyzed. Paraffin-embedded sections were obtained from stored tissue material from spleen. Fluorescence in situ hybridization (FISH) was used to detect cytogenetic abnormalities and the results were compared with analyses made on blood or bone marrow. Monosomy 12, an abnormality rarely occurring in CLL was used as a surrogate marker to avoid false-positive results for deletions due to incomplete nuclei. In all but one case this was below 20%. Therefore the cut-off for the deletions (13q-, 11q- and 17p-) were set to 30%. For +12 we used 5%. Data on flow cytometry for quantification of the CLL clone in the spleen was available in 13 patients with a median of 71% clonal cells (range 11–86%). Indications for splenectomy were splenomegaly (n=13), AIHA (n=8) or ITP (n=3). All but five patients had received chemotherapy before splenectomy including purine analogs (n=7), alemtuzumab (n=1) or both (n=2) The remaining patients were treated with alkylators and/or anthracycline-containg regimens. Result: In all but one patient, FISH results from the splenic sections could be obtained with cytogenetic aberrations detected in 18 of 23 patients (78%). In 11 patients a single abnormality was detected, whereas multiple aberrations were present in 7 cases. The most common aberration was 13q- (70%), followed by 17p- (22%), 11q- and +12 (17% each) (Table 1). In 8 cases FISH samples from blood/bone marrow were available from time points both pre- and post- splenectomy and 6 additional cases had FISH data before splenectomy. All FISH abnormalities affecting 〉20% CLL cells in blood or bone marrow were also found in the spleen. In contrast an additional 17p clone was found in one patient and in two patients the 13q- and 17p- clones were significantly larger in the spleen. Remarkably one patient had both homo- and heterozygous 13q- clones in blood and in the spleen. In repeated blood samples nine years after splenectomy only a heterozygous clone remained. Conclusion: FISH on paraffin-embedded sections is a useful tool for the evaluation of genetic abnormalities. Clonal evolution does appear to occur in the spleen in CLL and with splenectomy selected clones might be obliterated. Disclosures: Kimby: Roche, Bayer-Schering, Mundipharma: Membership on an entity's Board of Directors or advisory committees, lecturer.

Description: Background Follicular lymphoma (FL) is the second most common lymphoma in adults. Although responsive to therapies it is still considered incurable. The introduction of the CD20 antibody rituximab is well known to have improved outcome. Rituximab acts through complement-mediated cytotoxicity, antibody-dependent cellular cytotoxicity (ADCC) and direct induction of apoptosis. To enhance the efficacy of rituximab, different combination regimens have been used, mostly with chemotherapy but also with cytokines. Lenalidomide, an immunomodulatory agent commonly used in the treatment of multiple myeloma, has been shown to induce durable responses with manageable toxicity in indolent lymphomas and mantle cell lymphoma, especially in combination with rituximab. It acts on both the malignant cells and their microenvironment. The drug modulates signaling pathways, enhances the capacity of T cells and increases ADCC by natural killer (NK) cells, as well as suppresses angiogenesis. When combined with rituximab, the clinical effects seem to be synergistic (Fowler 2014). We aimed to investigate the dynamics of immune cell subsets in peripheral blood in patients given rituximab with or without lenalidomide. Patients and Methods FL patients included in a multicenter randomized phase II trial performed by the Swiss Group for Clinical Cancer Research (SAKK) in collaboration with the Nordic Lymphoma Group (NLG) were randomized 1:1 to treatment either with rituximab alone or rituximab and lenalidomide. Inclusion criteria were histologically confirmed CD20+ FL grade 1, 2 or 3A in disease stage Ann Arbor III-IV (or II not suitable for radiotherapy). Patients had to be in need of systemic therapy because of clinical symptoms, cytopenia, bulky disease or significant disease progression. In both treatment arms rituximab was administered as 4 single infusions of 375 mg/m2 weeks 1, 2, 3 and 4; in patients who showed at least a minor response 4 additional infusions were administered at weeks 12, 13, 14 and 15. In the combination arm, lenalidomide, 15 mg p.o. daily, was started 14 days before the first infusion and given continuously until 14 days after the last. Peripheral blood cells were sequentially sampled: at baseline, after 2 weeks' use of lenalidomide, 24 hours after first rituximab infusion and at follow-up at weeks 10 and 23. Analyses of CD3+, CD4+, CD8+ and CD56+CD3- (NK) cells were performed with flow cytometry. Results Immune cell activity was assessed on blood samples of 28 Norwegian and Swedish patients until July 2015. In all patients, irrespective of treatment arm, NK cell numbers markedly decreased at 24 hours after the first rituximab infusion compared to baseline counts (P=0.046), but returned to baseline levels by week 10 in most. However, patients in the combination arm exhibited a heterogeneous response with a diverse NK cell depletion/proliferation pattern, some showing a transient rise already after 14 days of lenalidomide use (Figure 1). CD3 levels were not affected at 24 hours after rituximab but increased over time in 15 of 18 patients (without differences between treatment arms). The increase at week 23 was statistically significant (P=0.004) with a median of 1.4 x 109 /L CD3+ cells compared to a baseline median of 0.88 x 109/L. In all patients, independent of treatment arm, the CD4/CD8 ratio increased compared to baseline already 24 hours after rituximab (P=0.011) and persisted throughout the study (week 10, P=0,005; week 23, P=0.019). The increased ratio was due to a large rise in CD4 counts (week 10, P=0.014; week 23, P=0.003), and a less pronounced rise in CD8 counts (week 10, P=0.094; week 23, P=0.007; Figure 2). Conclusion We found changes in the composition of immune cell subsets in peripheral blood in rituximab treated FL patients, with a larger interindividual variation when combined with lenalidomide. Ongoing analyses will reveal whether these patterns of immune cell response correlate with clinical outcome and long-term treatment effects. Figure 1. NK cell absolute counts (x 109/L) in (a) patients treated with rituximab and in (b) patients treated with rituximab plus lenalidomide. 1=baseline, 2=after 14 days of lenalidomide (b only), 3=24h after rituximab, 4=week 10, 5=week 23. Figure 1. NK cell absolute counts (x 109/L) in (a) patients treated with rituximab and in (b) patients treated with rituximab plus lenalidomide. 1=baseline, 2=after 14 days of lenalidomide (b only), 3=24h after rituximab, 4=week 10, 5=week 23. Figure 2. CD4/CD8 ratios in all 28 patients. The y scale is logarithmic. 1=baseline, 2=after 14 days of lenalidomide (14 patients only), 3=24h after rituximab, 4=week 10, 5=week 23. Figure 2. CD4/CD8 ratios in all 28 patients. The y scale is logarithmic. 1=baseline, 2=after 14 days of lenalidomide (14 patients only), 3=24h after rituximab, 4=week 10, 5=week 23. Figure 3. Figure 3. Disclosures Off Label Use: Lenalidomide was used together with rituximab in a randomized clinical trial.. Kimby:Gilead: Membership on an entity's Board of Directors or advisory committees; Jansen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees; Roche: Honoraria, Research Funding; Pharmacyclics: Membership on an entity's Board of Directors or advisory committees; Pfizer: Research Funding.

Description: Mantle cell lymphoma is a non-Hodgkin lymphoma with, in general, a poor prognosis. A minor subset of patients with an indolent disease course has however been recognized (1,2). The various growth patterns of MCL, i.e. mantle zone (MZ), nodular (N) or diffuse (D) is assumed to correlate to stage and to disease course. The genetic aberrations underlying the pathogenesis are well defined and correlate to high tumour cell proliferation and poor prognosis. However, the effect of the lymphoma microenvironment in disease development and sustainability is largely unknown. We have used flow cytometry to investigate the non-malignant cell composition of the lymph node microenvironment in a population-based cohort of 154 MCL cases diagnosed from January 1, 1998 to December 31, 2012. Flow cytometry analyses of lymph nodes, performed as part of the diagnostic process, were used to evaluate percentages of tumour cells, remaining non-malignant B-cells and T-cell subsets (CD3+, CD3+CD4+, CD3+CD8+). As lymph node T-cell numbers reflect a high tumor load in the lymph node we also investigated the CD4/CD8 ratio, which is not dependent on T-cell percentage. Data from 26 non-malignant lymph nodes were used for comparison. T-cell percentages are shown in Table 1. Clinical and other pathological parameters of the MCL cases, including MIPI, cell morphology, tumor growth pattern and cell proliferation were also evaluated. Indolent disease (n=15), defined here as requirement of treatment 〉 two years from diagnosis, was associated with higher amount of CD3, CD3+CD4+ and higher CD4/CD8 ratio (p=0.0429, p=0.0211 and p= 0.0032 respectively). Higher tumor cell proliferation correlated negatively with the CD4/CD8 ratio (p= 0.0007). There was a significant difference in CD3 percentages between reactive lymph nodes and MCL irrespective of growth pattern (all p

Description: Background: Mantle cell lymphoma (MCL) is an aggressive B-cell lymphoma with a high rate of relapses after therapy. Chronic lymphocytic leukemia (CLL) is a heterogeneous disease with varied outcome. For both diseases there is a need for new therapies. Cannabinoid receptors (CBs), which are overexpressed in most cases of MCL and CLL compared to normal B cells (Islam et al., 2003; Gustafsson et al., 2008; Freund et al., 2016) are promising novel therapeutic targets. CBs are membrane-bound receptors that convey signals from the microenvironment to the cells. There are two types of CBs: CB1 and CB2. CB1 is suggested to be involved in retention and/or egress of MCL cells from the tissue to the blood circulation (Wasik et al., 2014). CB2 is expressed by normal B-cells where it regulates positioning and retention of cells in tissue (Pereira et al., 2009; Basu et al., 2011; Muppidi et al., 2011) and in pre-B-cell acute lymphoblastic leukemia, involved in the energy metabolism (Chan et al., 2017). The retention/egress of the B-cell lymphoma cells is mainly regulated by chemokine receptors and adhesion molecules. The chemokine receptor CXCR4 is one of the most highly expressed chemokine receptors in MCL and CLL. 2-arachidonoylglycerol (2-AG, CB1/CB2 endogenous ligand) and CXCL12 (CXCR4 ligand) are synthetized and secreted by stromal cells in the bone marrow (Kose et al., 2018; Burger and Gribben, 2014). The endocannabinoids levels in cancer are suggested to have a role in cancer progression (Sailler et al., 2014) while CXCL12 is already a candidate target for therapy using a CXCR4 inhibitor AMD3100. Aim: To investigate a possible crosstalk between CBs and CXCR4 in MCL and CLL cells. Methods: Patients with newly diagnosed MCL (n=8) or CLL (n=25) gave informed consent to participate in the study. Lymphoma cells were enriched by negative selection. Fifteen primary lymphoma samples and the JeKo MCL cell line were subjected to chemotaxis towards CXCL12 and/or 2-AG. CXCR4 membrane expression was assessed by flow cytometry. Selective CB1 and CB2 antagonists were used to investigate the underlying mechanisms. CB1, CB2 and CXCR4 encoding genes levels were measured by qPCR and normalized to B cells from tonsil. Results and Conclusion: 2-AG induced chemotaxis in 11/15 MCL and CLL samples. In JeKo, 2-AG-induced migration was blocked by a CB2 antagonist, suggesting that signaling via CB2 is involved. When the primary cells were subjected to migration towards CXCL12, two patterns of chemotaxis were observed. The first pattern was seen in 7/15 samples that migrated towards CXCL12. In these samples, the migration was inhibited when 2-AG was combined with CXCL12. The second type of response was observed in 8/15 samples, those samples did not migrate towards CXCL12 but chemotaxis was enhanced by combining 2-AG and CXCL12. MCL and CLL samples expressed variable mRNA levels of CB1 (RFI range: 0.0-204) and CB2 (RFI range: 0.8-14.3) and all expressed CXCR4 at mRNA (RFI range: 0.1-215.8) and protein (MFI range: 1278-19301) levels that did not differ neither between the two diseases nor between the two migratory groups. When all 15 samples were combined, CB1 mRNA levels, but not CB2 mRNA, correlated to the chemotaxis towards CXCL12 (Spearman correlation coefficient = 0.626; p=0.01). In contrast, CB2 mRNA levels, but not CB1, correlated to chemotaxis towards 2-AG (Spearman correlation coefficient = 0.532; p=0.04), which is in agreement with the effects observed in JeKo. Furthermore, CB1 and CB2 mRNA levels correlated to chemotaxis towards the combination of CXCL12 and 2-AG both (for CB1 mRNA: Spearman correlation coefficient= 0.588; p=0.02 and for CB2 mRNA: 0.589; p=0.02). Neither CXCL12-induced CXCR4 receptor internalization, nor recycling was influenced by 2-AG incubation. Our findings indicate a novel pathway regulating chemotaxis of MCL and CLL implicating a cross-talk between CBs and CXCR4. The fact that the capacity to internalize CXCR4 remained intact after incubation with 2-AG suggests that the reduced CXCL12-mediated migration when 2-AG was combined could be due to an impaired downstream signaling in lymphoma cells. Importance: Lymphoma cells residing in the tissue receive pro-survival stimuli and are protected from chemotherapy by signals from the microenvironment. A better understanding of how lymphoma cell migration and tissue retention are regulated can be a step towards more efficient therapies. Disclosures Wahlin: Gilead: Consultancy, Honoraria, Research Funding; Roche: Research Funding.

Description: Introduction: The SAM domain and HD domain 1 (SAMHD1) protein is a deoxynucleoside triphosphate (dNTP) triphosphohydrolase, which has been initially described to restrict human immunodeficiency virus type 1 (HIV-1) in certain cell types through depletion of intracellular dNTP substrates required for HIV-1 reverse transcription. Mutations of SAMHD1 gene have been linked to Aicardi-Goutières syndrome (AGS) and have been identified as putative drivers of chronic lymphocytic leukemia resulting in decreased SAMHD1 mRNA and protein levels. More recently, SAMHD1 mutations have been reported in T-prolymphocytic leukemia (T-PLL). Based on these findings and the fact that SAMHD1 limits the dNTP pool in the cell, it may play a role in oncogenesis as a tumor suppressor. In addition, SAMHD1 may confer resistance to nucleoside-based chemotherapies by hydrolysing their active triphosphate metabolites, with cytarabine in acute myeloid leukemia being an example (Herold et al, Nat Med 2017; 23(2):256-263). The expression patterns and the potential role of SAMHD1 in the pathogenesis of peripheral T-cell lymphomas (PTCL) are not yet known. Methods: The patient cohort included 64 PTCLs of various histologic types which were diagnosed and treated at Karolinska University Hospital (Sweden). A control group of 4 reactive lymph nodes and 2 reactive tonsils was included in the study for comparison. All tissue samples were obtained prior to therapy. SAMHD1 expression was assessed by immunohistochemistry performed on a PTCL tissue microarray (TMA) with duplicate tumor cores from each case or full tissue sections using dual immunostaining (SAMHD1 / CD68) and a monoclonal antibody against SAMHD1 (Bethyl Laboratories, San Antonio, TX). At least 500 lymphoma cells were counted to calculate the percentage of SAMHD1-positive tumor cells. Overall survival (OS) was defined as time from diagnosis to death or last follow-up. Event-free survival (EFS) was defined as time from diagnosis to relapse, death, or last follow-up. Survival analyses were performed using the Kaplan-Meier method (log-rank test) and Cox regression models. Two T-cell lymphomas cell lines (Mac1, Mac2A) were used as an in vitro system. As our preliminary findings from in silico analysis revealed potential binding sites for MYC on the SAMHD1 gene promoter, we hypothesized that MYC might regulate SAMHD1 expression. Therefore, the T-cell lymphoma cell lines were treated with the selective BET / MYC inhibitor JQ-1 or transiently transfected with a MYC-overexpressing plasmid or MYC gene-specific siRNA constructs, respectively. Western blot analysis was used to assess the protein levels. Results: SAMHD1 protein was expressed in reactive T-cells and histiocytes (CD68+) in all reactive lymphoid tissues (lymph nodes and tonsils) with strong staining intensity. SAMHD was differentially expressed among PTCL subtypes generally with weaker staining intensity as compared to normal T-cells and histiocytes, thus being positive in all (100%) angioimmunoblastic T-cell lymphomas (AILT), 67% PTCL-NOS, 45% ALK+ ALCL, 20% of ALK+ ALCL, and none (0%) of T-lymphoblastic lymphomas (p=0.0017, chi-square test). Among the SAMHD1- positive cases, the percentage of positive lymphoma cells ranged from 0 to 100% and its highest median was observed in AILT. SAMHD1 expression inversely correlated with CD30 expression (% CD30+ positive lymphoma cells) (p=0.0025, Mann-Whitney test). No significant associations between SAMHD1 levels and other clinicopathologic parameters or clinical outcome (EFS or OS) were found, however, the number of patients analyzed in each histologic subtype was limited. Inhibition of MYC activity by JQ-1 or MYC gene silencing with specific siRNA resulted in a substantial increase in the SAMHD1 protein level in T-cell lymphoma cell lines. Inversely, transient transfection of the cell lines with a MYC overexpressing plasmid resulted in decreased levels of SAMHD1. Taken together, the in vitro data suggest a possible MYC-associated regulation (repression) of SAMHD1 gene expression in T-cell lymphoma. Conclusions: SAMHD1 is shown for the first time to be differentially expressed among PTCL types and its regulation may involve MYC. Preliminary survival analysis shows no significant associations of SAMHD1 expression with EFS and OS in this cohort of PTCL, however, analysis of a larger PTCL study group is underway to draw definite conclusions. Disclosures Österborg: Gilead: Consultancy, Research Funding; Beigene: Research Funding; Pharmacyclics: Research Funding; Janssen: Research Funding; Abbvie: Research Funding.

Description: The prognostic role of the transcription factor SOX11 in mantle cell lymphoma (MCL) is controversial. We investigated prognostic markers in a population-based cohort of 186 MCL cases. Seventeen patients (9%) did not require any therapy within the first 2 years after diagnosis and were retrospectively defined as having an indolent disease. As expected, indolent MCL had less frequent B symptoms and extensive nodal involvement and 88% of these cases expressed SOX11. In our cohort 13 cases (7.5%) lacked nuclear SOX11 at diagnosis. SOX11− MCL had a higher frequency of lymphocytosis, elevated level of lactate dehydrogenase (LDH), and p53 positivity. The overall survival in the whole cohort, excluding 37 patients receiving autologous stem cell transplantation, was 3.1 year and in patients with indolent or nonindolent disease, 5.9 and 2.8 years, respectively (P = .004). SOX11− cases had a shorter overall survival, compared with SOX11+ cases, 1.5 and 3.2 years, respectively (P = .014). In multivariate analysis of overall survival, age 〉 65 (P = .001), Eastern Cooperative Oncology Group score ≥ 2 (P = .022), elevated LDH level (P = .001), and p53 expression (P = .001) remained significant, and SOX11 lost significance. We conclude that most indolent MCLs are SOX11+ and that SOX11 cannot be used for predicting an indolent disease course.