ALBERT

Description: Introduction T-cell prolymphocytic leukemia (T-PLL) is a rare and aggressive T-lymphoid malignancy with poor response to current treatment strategies. We recently demonstrated single agent activity of venetoclax in relapsed/refractory (r/r) T-PLL, however resistance developed on mono-therapy. We here set out to identify partners for an effective combinatorial treatment concept. Methods To overcome bcl-2 inhibitor resistance we utilized primary T-PLL patient samples and applied a combinatorial next-generation functional drug screen for venetoclax and 25 additional therapeutic agents (Fig.1a). Molecular mechanisms of drug combinations were evaluated by BH3-family member profiling and mass spectrometry. Protein expression was assessed by Western Blot and viability by AnnexinV/Hoechst staining. The best scoring combination was evaluated in two late stage r/r T-PLL patients. Results Pairwise combinations screen of venetoclax with candidate small molecule inhibitors and chemotherapeutic drugs on primary T-PLL cells revealed synergistic action of venetoclax with ibrutinib, idelalisib, and 5-azacytidine, and to lower extents Olaparib, Temsirolimus, Ruxolitinib and Belinostat whereas cisplatin antagonized the effect of venetoclax across all patient samples tested (Fig 1b). The combination of venetoclax and ibrutinib resulted in substantial reduction of viability in primary T-PLL cells (Fig 1c). BH3-profiling on primary T-PLL samples with and without ibrutinib treatment demonstrated enhanced overall priming with predominant increase of bcl-2 dependency upon ibrutinib treatment (Fig 1d). Addition of ibrutinib to venetoclax led to decreased phosphorylation of ITK in vivo (Fig 1e). Two patients suffering from r/r T-PLL after failing at least two treatment lines including alemtuzumab were treated with the combination of venetoclax and ibrutinib resulting in significant clinical responses with substantial drops in leukocytosis and LDH as well as substantial clinical improvement (Fig 2a, b). The dynamic BH3 profiling in samples taken from these patients confirmed that the addition of ibrutinib is indeed increasing overall priming and Bcl-2 dependency (Fig 2c, d). Conclusion Our findings suggest efficacy of combinatorial treatment of venetoclax with ibrutinib in T-PLL. Mechanistically, ibrutinib dephosphorylated ITK in T-PLL cells and, furthermore, enhanced BCL2 dependency, both, in-vivo and in-vitro. Patients treated with the combination venetoclax and ibrutinib experienced profound clinical responses which needs further evaluation in an prospective clinical study on a larger cohort of r/r T-PLL patients. Disclosures Herbaux: Abbvie: Honoraria; Janssen: Honoraria; Takeda: Honoraria; BMS: Honoraria; Gilead: Honoraria. Mayerhoefer:Siemens: Research Funding, Speakers Bureau; BMS: Speakers Bureau. Jaeger:Novartis, Roche, Sandoz: Consultancy; AbbVie, Celgene, Gilead, Novartis, Roche, Takeda Millennium: Research Funding; Amgen, AbbVie, Celgene, Eisai, Gilead, Janssen, Novartis, Roche, Takeda Millennium, MSD, BMS, Sanofi: Honoraria; Celgene, Roche, Janssen, Gilead, Novartis, MSD, AbbVie, Sanofi: Membership on an entity's Board of Directors or advisory committees. Davids:AbbVie, Acerta Pharma, Adaptive, Biotechnologies, Astra-Zeneca, Genentech, Gilead Sciences, Janssen, Pharmacyclics, TG therapeutics: Membership on an entity's Board of Directors or advisory committees; AbbVie, Astra-Zeneca, Genentech, Janssen, MEI, Pharmacyclics, Syros Pharmaceuticals, Verastem: Consultancy; Acerta Pharma, Ascentage Pharma, Genentech, MEI pharma, Pharmacyclics, Surface Oncology, TG Therapeutics, Verastem: Research Funding; Research to Practice: Honoraria. Staber:Takeda-Millenium: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; MSD: Honoraria, Speakers Bureau; Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees; AbbVie: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Roche: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Janssen: Honoraria, Speakers Bureau; Gilead: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. OffLabel Disclosure: Ibrutinib - BTK inhibitor Venetoclax - bcl2 inhibitor

Description: Background. Measurement of serum M-spike is used to determine response to therapy and treatment free survival in Waldenstrom Macroglobulinemia (WM). However, there are many limitations to the use of IgM M-spike, and new sensitive markers are needed to determine early response or progression in these patients. We have previously shown that involved serum free light chain (sFLC) accurately diagnosed patients with WM, and correlated with markers of poor prognosis, specifically beta- 2 microglobulin (B2M). In this study, we sought to determine the role of levels of involved in the response to therapy in patients with WM treated on clinical trials, and compare it to the response observed using the traditional M-spike measurement. Methods. We prospectively studied involved sFLC in 61 WM patients enrolled on 2 clinical trials, at diagnosis (N=8) and with relapse/refractory disease (N=53). Patients were treated on one of two clinical trials: either perifosine (N=30; given 150mg oral daily) or combination of bortezomib and rituximab (N=30; given IV bortezomib 1.6mg/m2 at days 1, 8, 15 q 28 days × 6 cycles and rituxan 375 mg/m2 at days 1, 8, 15, 22 on cycles 1 and 4). Response was assessed after cycle 2, confirmed on 2 consecutive measurements, and included minor response (MR), partial response (PR) or complete remission (CR). Results. The baseline characteristics of the patients were as follows: the median age was 65 years (44–83), male/female ratio 2.38, serum B2M 3.0mg/L (1.00–7.30), hemoglobin 11.0g/dL (7.00–14.90), platelet count 216 ×109/mm3 (46–563), serum M-spike 2.24g/L (0.41–4.62), and involved sFLC 95.2mg/L (4.92–868). The WM International staging system (WM-ISS) showed that 28% of patients had low ISS, 31% intermediate, and 41% high ISS scores. The overall response rate for the entire cohort was 62.3%, assessed by M-spike measurement (MR=23, PR=13, nCR=2). The overall response rate using involved sFLC level was 72.1% (MR=14, PR=29, and CR=1), (p

Description: Abstract 1307 Background. Mutation of MYD88 gene has recently been identified in activated B-cell like diffuse B-cell lymphoma, and enhanced JAK STAT and NF-kB signalling pathways. Whole exome sequencing study in Waldenstrom macroglobulinemia (WM) suggested a high frequency of MYD88 L265P mutation in WM. Although the genetic background is not fully deciphered in WM, the role of NF-kB and JAK STAT pathways has been demonstrated in WM; which underlying mechanisms of deregulation remain to be elucidated. We aimed to analyze MYD88 mutation in exon 5 and to characterize the clinical significance of this genetic alteration in 67 WM. Method. 67 patients (42 males, 25 females) diagnosed with WM were included in this study, along with 9 patients with chronic lymphocytic leukemia (CLL), 4 multiple myeloma (MM) and 9 marginal zone lymphoma (MZL) were also studied. Patients were untreated at time of BM collection and gave informed consent prior to research sampling. Clinical features, immunophenotypic markers using flow cytometry (Matutes score panel, CD38, CD138, CD27, CD80), conventional cytogenetic, FISH and SNP array data (n = 46) were analysed. B cells from bone marrow and T cells from blood were isolated respectively using B cell isolation kit and Pan T isolation kit (Myltenyi Biotech). For DNA sequencing of exon 5 of MYD88, the exon 5 of MYD88 gene was amplified from genomic DNA by PCR. The purified PCR products were directly sequenced in both directions using BigDye® Terminator Cycle Sequencing Kit (Applied Biosystems, CA, USA) and analyzed on the Applied Biosystems 3130xl Genetic Analyzer. Data were analyzed with SeqScape software version 2.5 (Applied Biosystems). Results. MYD88 L265P mutation (MYDmut) was observed in 79% of patients, including homozygous mutation in two patients (3%). MYD88 mutation was not identified in T lymphocytes isolated from 4 WM patients that confirmed MYD88 mutation was acquired in the tumoral cells. We haven't observed any other mutation on exon 5. We then sought for other mechanisms of MYD88 gene alteration, such as copy number alteration (CNA) and copy neutral –loss of heterozygosity (CN-LOH) also considered as an acquired UPD (uniparental disomy) at MYD88 locus. We found an UPD at MYD88 locus in solely one patient (2%), and haven't identified any deletion at 3p22. On the contrary, we observed a gain on chromosome 3 at 3p22 locus (including MYD88 gene) in 7/57 (12%) patients. Taking together, we identified alteration of the MYD88 locus in 85% of patients with WM, by either gain-of-function mutation (79%) or CNA (12%). Interestingly, we found gain on chromosome 3 more frequently in the MYDwildgroup than in the MYDmutgroup (p=0.02). Twenty one percent of the patients with WM had no mutation of MYD (MYDwild), and were characterized with a female predominance, a splenomegaly, gain of chromosome 3 and CD27 expression. We did not observed difference in terms of survival according to the MYD88 mutation status. MYD88 mutation was not related to deletion 6q, gain of 4, deletion 11q, deletion 17p, deletion 13q14 in our study. Interestingly, deletion 7q, a frequent cytogenetic aberration in marginal zone lymphoma, was rare in our series (4/57; 7%) and was independent of MYD88 mutation status (2 in the MYDwild and 2 in the MYDmut) (p=ns). No MYD88 L265P mutation was observed in CLL and MM. In MZL, 1/9 patient without M monoclonal component had a MYDL265p mutation. Conclusion. These results confirm a high frequency of MYD88 L265P mutation in WM that may become a useful biomarker for diagnostic in WM and may help better understand the physiopathogeny of WM. Disclosures: No relevant conflicts of interest to declare.

Description: Background Quantification of monoclonal immunoglobulins in serum is the foundation of IMWG multiple myeloma (MM) response criteria for patients with intact immunoglobulin disease. However, electrophoretic methods for quantifying M-Ig in serum are subject to well documented limitations including inaccuracy at high concentrations (due to dye saturation) and poor sensitivity at low concentrations. Moreover, the methods require skilled interpretation and can be time consuming; however their clinical utility is well established. Recently heavy/light chain (HLC) assays, which quantify kappa and lambda isotypes of intact immunoglobulins have become available. Here we compare responses assigned by the immunoassays to IMWG criteria and evaluate the clinical impact of discordance. Methods Sequential sera, from enrolment to progression, were available for 107 MM patients (59 IgGκ, 29 IgGλ, 12 IgAκ, 7 IgAλ) enrolled onto either the IFM 2009-02 end-stage relapsed or refractory MM or the IFM 2010-02 in del17p and t(4;14) relapsed or refractory MM trials. The inclusion criterion was a measurable intact immunoglobulin MM according to IMWG criteria (M spike ≥ 10 g/L), using serum and/or urine protein electrophoresis. IgA HLC (IgAκ and IgAλ) and IgG HLC (IgGκ and IgGλ) analysis was compared to historic SPEP, IFE, UPEP, uIFE and serum free light chain (sFLC) results (measured using polyclonal antisera based assays). Responses were assigned at approximately 90 days (median 90, range 61-107 days) and at maximum response (if different) according to IMWG criteria using changes in monoclonal protein concentrations measured either by SPEP or dHLC (clonal - non clonal). Complete response was assigned either by the absence of monoclonal protein on IFE or by a normal HLC ratio. Results At the time of enrolment, 87/88 IgG and 17/17 IgA patients had abnormal HLC ratios which were concordant with IFE results, and quantification of the monoclonal protein by dHLC was similar (median (range): 29.2 (2.5-77.5) g/L) and SPEP (30.7 (1.8-66.9) g/L). After 3 months of treatment, responses assigned by HLC/FLC assays showed near-perfect agreement with responses assigned using SPEP and IFE (Weighted Kappa 〉0.81, p

Description: Abstract 5076 Background. Waldenstrom macroglobulinemia (WM) is a low grade B cell lymphoma characterized by bone marrow infiltration of lymphoplasmacytic tumor cells that secrete monoclonal IgM (M-protein) into the serum. Measurement of the serum M-protein [using serum protein electrophoresis (SPEP)] and measurement of total IgM (using nephelometry) are used to diagnose and monitor WM. There are, however, many limitations with these techniques and new markers are needed. IgG and IgA Hevylite® immunoassays have been reported to be more sensitive than SPEP and nephelometry for identifying monoclonal immunoglobulins in multiple myeloma and unlike immunofixation, provide quantitative information. We hypothesized that serum IgM Hevylite assays (specifically measuring IgMkappa and IgMlambda, separately) would accurately identify serum IgM M-proteins. We also evaluated the association between known tumor burden markers and prognostic factors with IgM Hevylite results in patients with WM. Method. We retrospectively measured IgMkappa and IgMlambda in sera from 59 WM patients: 44 patients were at diagnosis and 15 had relapsed disease. The diagnosis of WM was made according to the current guidelines. All serum samples were kept frozen in the Lille serum bank since collection. All patients gave informed consent prior to the collection and none were treated at time of collection of the serum. Approval of this protocol was obtained from the CHRU Lille and was in accordance with the Declaration of Helsinki. Hevylite measurements were made at The Binding Site Ltd, Birmingham, UK. A normal range was produced from normal (blood donor) sera (n=120), median (and 95%ile ranges) were; IgMkappa 0.634g/L (0.29-1.82), IgMlambda 0.42g/L (0.17-0.94), IgMkappa/IgMlambda ratio 1.6 (0.95-2.3). For ease of comparison IgM hevylite ratios were expressed as the involved monoclonal immunoglobulin/uninvolved polyclonal immunoglobulin (IgMi). To describe the distribution of IgMi Hevylite levels in patients with WM, the median and range (min-max) were reported. Median values were compared using the Wilcoxon rank-sum test and ANOVA. Fisher's exact test was used to compare proportions. All statistical tests were two-sided. All analyses were conducted using SPSSv12 software. Results. The baseline characteristics of the patients were as follows: the median (range) age was 68 years (41-86), male/female 38/21, serum b2M 3.0mg/L (1.2-9.0), hemoglobin 11.8g/dL (7.6-15.4), platelet count 267 ×109/mm3 (55-741), serum M-spike 19g/L (3.0-52.7). In our series, 18 (31%), 22 (37%) and 19 (32%) patients had low, intermediate and high risk disease respectively, in the WM-IPSS scoring system. The median (min-max) IgMkappa ratio was 134 (8.7-2850) and IgMlambda ratio was 0.03 (0.0007-0.39). IgMi Hevylite ratio was 98.07 (2.59-2850). The IgMi Hevylite correlated well with the M-spike measured using SPEP (r=0.601, p

Description: Abstract 3628 Background. Waldenstrom's macroglobulinemia (WM) is a rare lymphoproliferative disorder characterized by bone marrow (BM) infiltration of lymphoplasmacytic cells and monoclonal IgM gammopathy. The deletion 6q is the most frequent cytogenetic aberration in WM and it has been suspected that this region harbours a tumour suppressor gene of pathogenic significance for WM. Comparative genomic hybridization (CGH) array delineated several minimal deleted regions (MDR) on 6q and pointed out key regulator genes of the NFKB pathway involved in these deletions, including A20 gene. The zinc finger protein A20 has been characterized as dual inhibitor of NFKB activation and was recently described as a tumor suppressor gene in lymphoma. However, the mechanisms of A20 deregulation are not fully understood in WM. We aimed to study the mechanisms of A20 deregulation in WM using gene expression profiling (GEP) and single nucleotide polymorphism (SNP) based arrays, a powerful high resolution method allowing both the detection of loss of heterozygosity (LOH) and copy number alteration (CNA) analysis in the same experiment. Methods. We have studied BM samples from 42 patients (pts) with WM (31 males, mean age: 67 years). Conventional karyotype study was analyzed following stimulation with IL2 and DSP30 of BM cells. FISH analysis was performed to detect deletion 6q21 (MYB probe). DNA and RNA were extracted following BM CD19 B cell negative selection. Quantification of copy number of A20 gene was performed using real time PCR, RNAse P gene was used as reference gene. Genome-Wide Human SNP Array 6.0 (Affymetrix chips) was used to detect both LOH and CNA in 31 pts. Paired samples (B cells/normal T lymphocytes) were used as an intra-individual reference to distinguish germline polymorphisms from somatic abnormalities in 29 patients. Size, position and location of genes were referenced using UCSC HG18 assembly, LOH and CNA were identified using genotyping console 3.02 software (Affymetrix) and Partek genomic suite. Gene expression profiling (GEP) was performed in 11 pts using Affymetrix U133A arrays. Gene-expression intensity values were log transformed, normalized and analyzed using GeneSpring GX software. Results. SNP array identified deletion 6q in 9/31 (29%) patients, confirmed by FISH analysis. All cases were monoallelic deletion. In 90% of cases, deletion 6q was associated with others genetic abnormalities. In one case, we observed a loss of 6q combined with a gain of 6p. Deletion 6q was associated with gain of chromosome 4 in 2 cases, and with deletion of 13q in 3 cases. The MDR was located on chromosome 6q21-6q25. This MDR contained several candidate genes included always deletion of A20 gene, located on 6q23, and was confirmed by real time DNA PCR quantification in all cases. Overall, the frequency of A20 gene deletion in our entire WM cohort (N=42 pts) was 29.2% (12/42), determined by real time DNA PCR of the copy number of A20 gene. In patient without deletion 6q, we looked at potential deregulated mechanisms of A20 by LOH with no CNA, e.g. UPD (uniparental disomy). We have not seen any UPD targeting the A20 gene. We then looked at A20 gene expression by GEP, and found no significant variation of A20 gene expression related to A20 monoallelic deletion as compared to pts with two copies, reflecting probably the existence of other mechanisms of A20 gene deregulation. This result was consistent with the absence of clear difference in genome-wide expression profiling in pts with 6q deletion and those without the deletion. As A20 is a key player in the negative feedback loop regulation of NFKB, we also looked at the gene expression of a set of genes involved in NFKB pathway. The presence or absence of A20 deletion did not influence the gene expression profiling of this NFKB pathway gene set. Conclusion. We described a high frequency of deletion of A20 gene. In most cases, deletion of A20 was associated with other genetic abnormalities. A20 deletion was not associated with a significant signature by GEP. Additional studies are needed to understand the cellular consequences of A20 deletions in the pathogenesis of WM. Disclosures: Leleu: Celgene: Consultancy, Research Funding; Janssen Cilag: Consultancy, Research Funding; Leo Pharma: Consultancy; Amgen: Consultancy; Chugai: Research Funding; Roche: Consultancy, Research Funding; Novartis: Consultancy, Research Funding.

Description: Background. Waldenstrom macroglobulinemia (WM) is a B-cell malignancy characterized by bone marrow (BM) infiltration of clonal lymphoplasmacytic cells, which produce a monoclonal immunoglobulin M. MYD88L265P mutation may be considered as a founder event because of it high frequency in WM. WM cells may acquire additional genetic hits that may potentially promote disease progression: CXCR4 or CD79B mutations, copy number variation,…TP53 is a tumor suppressor gene that functions as regulator influencing cellular responses to DNA damage. Little is known regarding TP53 alteration in WM. Our aim was to screen TP53 mutation in a large cohort of WM at diagnosis to analyze the genomic landscape of WM using targeted next generation sequencing (NGS) and genome wide single nucleotide polymorphism array (SNPa) and to identify clinical and biological characteristics. Method. BM samples of 125 WM (mean age: 67 years) were analyzed at diagnosis. Tumoral DNA was extracted following CD19 B cell selection. TP53 mutations were analyzed by targeted NGS to scan the coding exons of TP53. MYD88L265P, CD79A, CD79B, and CXCR4mutations were analyzed by sanger sequencing and/or NGS. Genome-Wide Human SNP Array 6.0 (Affymetrix chips) was performed in 62 cases. CN-LOH (copy neutral- loss of heterozygosity) and CNA (copy number aberration) were mapped using console 3.02 software (Affymetrix). Flow cytometry was performed to assess P53 and p21 expression after nutlin3a exposition to characterize functional mutant of TP53. Viability and cell growth of treated cells were determined using the MTS assay. Results. We have identified TP53 mutations using NGS in 7.3 % of WM (6 non-sense, 3 frameshift mutations located in the DNA binding domain) (TP53mut WM). The mutation load of TP53 varied from 13% to 98.9% (mean: 62.0%) using the variant allele frequency in NGS. We next examined the effects of nutlin-3a which is an mdm2 inhibitor on WM patients CD19+ cells genotyped for TP53 mutation. Nutlin-3a increased the expression of p53 and p21 in TP53Wild WM patients using flow cytometry (n=6). In contrast, in TP53MutWM cells, no significant variation of p53, p 21 and viability using MTS assay was observed suggesting the presence of functional mutation of TP53. The minimal deleted region of 17p in 17p deleted (TP53Del) samples was mapped using SNP array and contained 79 genes, among which was systematically comprised the loss of TP53. A high correlation between TP53 mutation and deletion 17p (p

Description: Abstract 297 Background. Current models of cancer progression are based on evolution and clonal selection i.e. evolution of tumour cells inducing expansion of cells that acquire genetic lesions over time, related to ongoing mechanisms of genomic instability. An initiating event may be followed by gene mutation, copy number alteration (CNA) or copy neutral loss of heterozygosity (CN-LOH) that drive tumor progression and emergence of mechanisms of resistance to drugs. WM is a lymphoproliferative disorder characterized by bone marrow (BM) infiltration of lymphoplasmacytic cells that secrete monoclonal IgM antibody. The clinical course is characterized either by an indolent or smoldering status or a symptomatic profile that needs chemotherapy to control the progression of tumour cells. The majority of patients (pts)evolved from indolent to symptomatic, and the mechanisms of progression of WM are not fully understood to date. We hypothesized that we could gain insights into clonal evolution underlying disease progression of WM on a paired serial analysis of WM samples using genome wide SNPa, that allow both the detection of LOH and CNA. Method. BM samples of 19 untreated pts with WM (12 males, mean age: 67 years, 11 symptomatic pts) were analysed. All patients had Genome-Wide Human SNP Array 6.0 (Affymetrix chips) on at least two sequential tumor samples 〉 6 months apart (42 samples from 19 patients with two to three points). Tumoral DNA was extracted following CD19 B cells selection. Paired samples (tumor/normal T lymphocytes) were used as an intra-individual reference to identify germline polymorphisms. Size, position and location of genes were identified with UCSC Genome Browser HG18 assembly, LOH and CNA using genotyping console 3.02 software (Affymetrix). FISH analysis was performed to detect deletion 6q; 13q14, 11q22, TP53, trisomy 4 and 12 chromosomal aberrations using Vysis probes. P53 and MYD88 mutation were analyzed by sanger sequencing. Results. At initial sampling, SNPa detected a total of 76 CNA genetic aberrations (range 0 – 24 per genome) including 22 gains and 54 losses; 85% of patients had MYD88 L265P mutation. During the follow-up of all indolent WM that remained indolent, we haven't observed any new genetic aberrations, gain or loss (either CNA or CN-LOH) (n=8, mean follow-up: 63 months, range: 16–107 months). Among the 11 remaining patients (mean follow-up: 44.2 months, range: 7–92 months), a clonal evolution was observed in 6 cases. Three of them were symptomatic and acquired CNA or CN-LOH during clinical evolution: one case with acquisition of several CN-LOH, including one at 3p22 involving MYD88 locus associated to a mutation, one case with an additional deletion 7q31 which was observed in a patient with a chromothripsis at diagnosis, one case with emergence of subclonal del17p. The two remaining patients evolved from smoldering to symptomatic with a major variation in CNA number. Finally, loss of CNAs displayed at diagnosis was observed in one patient after chemotherapy and during indolent follow-up. In 5 cases, no detectable CNA changes were identified between the initial and subsequent sample at relapse. Conclusion. Our results using high resolution SNP array support the hypothesis that a symptomatic WM disease will favour genetic clonal evolution of tumor cells. In our study, 3 distinct genotypic patterns were observed: (i) absence of genomic variation in stable smoldering WM disease or in symptomatic relapsed patient. (ii) unstable genotype in symptomatic or asymptomatic patients switching to symptomatic suggesting high risk tumors that are less stable and more prone to change with time. (iii) loss of abnormalities suggesting that chemotherapy eradicated the dominant clone or that a genetically distinct relapsed clone has emerged. This study refines our understanding of the dynamic genetic changes in progressive WM. Further confirmation of the role of some candidate genes is underway. Disclosures: No relevant conflicts of interest to declare.

Description: Waldenström Macroglobulinemia (WM) is a disorder of lymphoplasmacytoid cells that inhabit lymph nodes and the bone marrow. WM cells are characterized by secretion of monoclonal pentameric IgM. These cells are CD19+, CD20+, CD22+, CD38+, CD138+/- and phenotypically resemble IgM plasmablasts or plasma cells. In addition, 91% of WM cases carry an activating mutation of MyD88 (L265P). Mature resting B cells can be driven to differentiate to IgM secreting plasmablasts and plasma cells with similar phenotypes using the TLR4 ligand lipopolysaccharide (LPS). We have demonstrated that LPS (+ cytokine)-differentiated cells become Bcl-xL dependent during this process, rendering them sensitive to the Bcl-xL/Bcl-2 inhibitor ABT-737. For this reason, we hypothesized that activation of MyD88 in WM cells could drive Bcl-xL dependence in a similar manner conferring ABT-737 sensitivity. We treated three WM cell lines, BCWM.1, MWCL-1 and RPCI-WM1 which all harbor the MyD88 (L265P) mutation with ABT-737. We found varying levels of resistance to ABT-737 with an IC50 〉 2 μM for all three lines as compared with the ABT-737 sensitive multiple myeloma cell line MM.1s which has an IC50 of 0.4 μM. The RPCI-WM1 cell line was the most insensitive to ABT-737-induced apoptosis with no apoptosis above baseline up to 1.6 μM of drug. Since the WM cell lines were not sensitive to direct inhibition of intrinsic survival regulators, we then examined the sensitivity of these cell lines to other activators of the intrinsic apoptosis pathway. Two of the three cell lines were moderately sensitive to bortezomib with IC50 ≈ 5 nM as compared with the sensitive multiple myeloma cell line MM.1s with an IC50 of 2 nM. The RPCI-WM1 cell line was insensitive to bortezomib with no apoptosis above baseline up to 20 nM bortezomib. Similarly, we found that two of the cell lines were moderately sensitive to arsenic trioxide with an IC50 ≈ 6 μM as compared with the multiple myeloma cell line MM1.s (IC50 ≈ 4 μM). The RPCI-WM1 cell line was insensitive to ATO as well with an IC50 〉 20 μM. Given the lack of sensitivity of the three WM cell lines we tested to Bcl-xL/Bcl-2 inhibition with ABT-737 treatment, and that RPCI-WM1 appears insensitive to multiple inducers of intrinsic apoptosis, we examined the expression levels of Bcl-2 family members in these cells. Both BCWM.1 cells and MWCL-1 cells expressed Bim mRNA at very low levels with MWCL-1 expressing no detectable Bim at the protein level. Surprisingly, more moderate levels of Bim were detected in RPCI-WM1 cells. These findings were confirmed at the mRNA level by qRT-PCR. Bcl-xL and Mcl-1 were detectable in all three lines at moderate levels while Bcl-2 which was only expressed at significant levels in MWCL-1 cells and undetectable in BCWM.1 cells. We examined the expression levels of the Bax and Bak in these cells and remarkably there was no detectable Bax and very small amounts of Bak protein in RPCI-WM1 cells. Consistent with a defect in gene expression, Bax mRNA was also low in RPCI-WM1. This was not due to copy number variation, as determined by array-CGH in both the initial patient isolate and the established cell line. Additionally, no loss of Bax, Bak or Bim (Bcl2l11) was observed in SNP array analysis of 46 patients with WM. Interestingly, Bak mRNA levels in RPCI-WM1 were similar to the other WM lines, suggesting a defect in translation or post-translational regulation is responsible for the low protein expression. These results lead us to conclude that these WM cell lines are not sensitive to Bcl-xL/ Bcl-2 inhibition despite activation of MyD88. We have further shown that there are multiple and distinct differences in Bcl-2 family protein expression that lead to this insensitivity. While low levels of Bim combined with expression of Mcl-1 confer resistance to ABT-737 in MWCL-1 and BCWM.1, the lack of Bax and Bak confers resistance to intrinsic apoptotic stimuli in general in RPCI-WM1. Moreover, the loss of Bax and Bak protein expression occur through distinct mechanisms. These WM cell lines demonstrate that sensitivity to agents that kill through the intrinsic apoptotic pathway may vary within a disease that is characterized by a single activating mutation and suggests that additional heterogeneous events regulate the expression of Bcl-2 family proteins in WM. Disclosures: Leleu: CELGENE: Honoraria; JANSSEN: Honoraria. Lonial:Millennium: Consultancy; Celgene: Consultancy; Novartis: Consultancy; BMS: Consultancy; Sanofi: Consultancy; Onyx: Consultancy. Boise:Onyx Pharmaceuticals: Consultancy.

Description: Abstract 1288 Background. Approximately 30% of the patients who fulfil the criteria of Waldenström macroglobulinemia (WM) are diagnosed while asymptomatic, and will not require immediate therapy; these cases are called indolent WM (IWM). However, patients with a disease-related event will be considered for therapy, these cases are called symptomatic or aggressive WM (AWM). The physiopathology of these 2 groups remains unclear, and the mechanisms of progression have not been fully understood so far. We hypothesized that a gene signature that differentiates these two categories could be identified to better understand the underlying mechanisms of progression of WM. Methods. Seventeen patients diagnosed with WM (8 IWM and 9 AWM) were included in this study. We selected tumour cells from the bone marrow (BM) using mononuclear cell isolation, then B cell enrichment (B cell isolation kit, Myltenyi-Biotec, USA). The purity was confirmed by flow cytometry. Total RNA was extracted using the Trizol method. Gene expression profiling was performed using U133A arrays (Affymetrix, USA). Gene expression was normalized using the RMA algorithm. We ranked genes by fold-change of expression levels on a first series of 11 patients (5 IWM and 6 AWM) calculated with the ‘limma’ package in R. Next, we used a supervised classification to establish a gene expression profile to distinguish IWM from AWM. Therewith, we validated this profile on an independent set of 6 patients (3 IWM and 3 AWM). We then performed a pathway analysis using Ingenuity® analysis software. We confirmed gene expression deregulation with qRT-PCR on 3 candidate genes in the first series of patients. Genome-wide detection of copy number alteration and loss of heterozygosity were performed on 13 of the 17 WM cases, using the Genome-Wide Human SNP Array 6.0 (Affymetrix, USA). Finally, we investigated the functional consequences of the deregulation of these candidate genes in BCWM1 and MWCL1, both B cell lines originated from WM. Survival was studied using a colorimetric method with MTS (Promega, USA). Proliferation was analyzed using incorporation of a nucleoside analog (EdU) into DNA during active DNA synthesis (Invitrogen, USA). Results. The differential analysis has identified 82 probes, corresponding to 48 genes, significantly deregulated and capable of differentiating samples from IWM and AWM in an unsupervised classification. Moreover, with a supervised classification, this gene expression profile accurately classified 94% of the 17 WM samples, including the 6 WM of the independent validation set. The two molecular networks that appeared to play a major role in the physiopathology of IWM versus AWM were the plasma cell differentiation pathway and the AKT pathway. We have then identified 3 key genes in those 2 pathways, BACH2 and CIITA on the one hand and PTEN, respectively. We have then confirmed the deregulation of these gene expression levels by qRT-PCR in 3 IWM and 4 AWM; these 3 genes were over-expressed in IMW relatively to AMW. BACH2 is a B-cell-specific transcription factor known to be a tumour suppressor gene. It was shown that BACH2 reduces proliferation and induces cell death when over-expressed in B lymphoma tumour cells. We have thus pharmacologically over-expressed BACH2 in BCWM1 and MWCL1 and significantly reduced the proliferation and the survival of the two cell-lines. Further studies using BACH2 specific overexpression with lentiviral infection are underway, in vitro. The data will be presented at ASH. In order to further study the mechanisms of deregulation of BACH2 in IWM and AWM, we have conducted a genome wide SNP array study of 13 patients. Among those, 7 patients (4 IWM and 3 AWM) demonstrate a deletion of long arm of chromosome 6 (del6q), the most frequent chromosomal abnormality in WM. BACH2 gene is located on the 6q15 locus. Interestingly, we found that 3 out of the 3 AWM had a del6q that took in the 6q15 region, whereas 3 out of 4 of the IWM had a del6q preserving the 6q15 region. Therefore, haploinsufficiency could participate in the under-expression of BACH2 in aggressive WM; this hypothesis will be verified by using DNA qRT-PCR of BACH2. Conclusion. To the best of our knowledge, we have identified for the first time a specific gene expression signature that differentiates IWM and AWM. We have exposed several genes from this dataset, including BACH2, which is a candidate to better understand the underlying mechanisms of progression of WM. Disclosures: No relevant conflicts of interest to declare.