Search

Introduction: Chimeric antigen receptor (CAR) T-Cell therapy is approved for relapse/refractory diffuse large B cell lymphoma (DLBCL) patients after two lines of therapy. Although it is an effective treatment, approximately 20-30% of patients have an early relapse. In this context, biomarkers that helps to identify those patients who will be refractory to this therapy are relevant. Cell-free DNA (cfDNA) has emerged as a new tool for non-invasive monitoring of patients with lymphoma, therefore the aim of this study was to evaluate dynamic cfDNA concentration in addition to other biomarkers (LDH, CRP, ferritin) before CAR T-cell infusion to detect early progressors. Methods: We selected 44 r/r DLBCL patients treated with CAR T-cell in our centre. Plasma samples were collected pre-apheresis (PA) and pre-infusion (PI)(∆cfDNA). Other biomarkers (LDH, CRP, Ferritin) were studied PA, pre-lymphodepletion (PL) and PI, tumour volume metabolic (TMV) are also studied. CfDNA were obtained from plasma using QIAamp® Circulating Nucleic Acid (Qiagen) and quantified by QuantiFluor dsDNA System (Promega). The Mann-Whitney test was used to compare differences between two independent quantitative variables . ROC analysis was conducted to determine the cut-off value of biomarkers. Cumulative incidence of progression (1 and 3 months) was calculated from the date of CAR T-cell infusion. Data analysis was performed using Stata. Results: Cumulative incidence of relapse at 1 month and 3 months was 20% and 30% respectively. The correlation between the progression (1 month, 3 months) and the different biomarkers is shown in Table 1. The CRP PL, CRP PI, LDH PI and ∆cfDNA (PI-PA, median time 41.5 days [range:31-107]) was correlated with early progression (1 month). No differences were found with progression at 3 months. The different cut-off for the biomarkers selected was 9 ng/mL for ∆cfDNA, 225 U/L for LDH and 1.35 mg/dL for CRP. Cumulative incidence of progression at 1 month was calculated for these biomarkers (figure 1): ∆cfDNA, HR: 4.3 (1.05-17), p=0.042; CRP PL: HR: 6.9 (1.5-31.1), p=0.012; CRP PI, HR: 11.2 (1.5-83), p=0.019 and LDH PI, HR: 3.4 (1-17) p=0.11. It should be noted that 83.3% (10/12) of the patients who progressed during the first month had at least one of the above variables. Conclusions: Our results highlight that increase in cfDNA higher than 9 ng/mL, CRP PL and PI >1.35 mg/dL and LDH PI > 225 U/L before CAR T-cell therapy may detect early progressors within the first month after treatment. Therefore, ∆cfDNA, CRP and LDH could be useful to identify patients who highly probable will not benefit from the CAR T infusion, in which another prior strategy could be considered in an attempt to control the disease. These results need to be confirmed with a larger cohort. Figure 1 Figure 1. Disclosures Bailén: Pfizer: Honoraria; Kite-Gilead: Honoraria; Gilead: Honoraria. Kwon: Gilead: Honoraria.

Introduction: Non-Hodgkin B lymphomas (NHL-B) are a large group of heterogeneous diseases, with well-defined, histological and clinical features. Currently, the lymph node biopsy is essential for diagnosis, often obtained from hard to reach areas. Our aims with this study was to analyze genetics variants at diagnosis trying to discriminate between different NHL-B histologic subtypes using formalin fixed paraffin embedded (FFPE) tissue sections and circulating tumor free DNA (ctDNA) samples. Material and Methods: This is a retrospective, cross-sectional, single-center study. Sixty patients were selected with a diagnosis of different NHL-B subtypes: Diffuse large B cell lymphoma (DLBCL), follicular lymphoma (FL), marginal lymphoma (ML), mantle cell lymphoma (MCL) and plasmablastic lymphoma (PL). Sixty FFPE tissue samples and 23 ctDNA specimens obtained at diagnosis were selected. We performed an enrichment panel of 54 genes recurrently mutated in lymphomas (Lymphoma Solution; Sophia Genetics) by next generation sequencing (NGS; NextSeq; Illumina). The depth of 90% of the readings was greater than 2600x. Quantitative variables were expressed as median and range. Categorical variables were expressed as frequency and percentage. The Fisher exact test was used to compare the distribution of categorical variables. The Mann-Whitney test was used to compare differences between quantitative variables. Statistical significance was set at p < 0.05. Results: Our study shows that 97% (58/60) of patients presented any variant. The most frequently mutated genes were KMT2D, EP300 and SOCS1. Exclusive variants were detected in FL in the genes CCND1, PAX5, CREBBP, BCL2 and REL genes. In the same way, in DLCBL the variants detected in the BRAF, TCF3, XPO1, PIM1, CHD2, ID3 and BCL6 genes were exclusive of this subtype. Furthermore, the genes BCL2 (p=0.0021), CREBBP (p=0.0004), NOTCH2 (p=0.03), PAX5 (p=0.01) and TNFRSF14 (p=0.04) are more frequently altered in patients with FL, ATM (p=0.02) gene in MCL; NFKBIE (p=0.03), CIITA (p=0.02), MYC (p=0.009) and PIM1 (p=0.04) in DLCBL, among which it was found that high-grade lymphomas are more frequently associated with mutations in CHD2 (p=0.02), MYC (p=0.03) and MYD88 (p=0.02; data not shown). In the subgroup of 23 patients with paired samples (FFPE/ctDNA), 95% (22/23) of patients had mutations in any of the genes on the panel in FFPE tissue samples and 82% (19/23) in ctDNA specimens. The number of variants detected was different depending on the type of sample analysed, 52% of the variants were detected in both specimens, 39 only in FFPE tissue and 9% in ctDNA; Figure 1). In variants detected in both samples, the average value of the VAF was higher that when is detected only in one of the samples (FFPE: 28.1 vs 5.8 (p We did not find significant differences in the number of variants depending on the stage, classification, tumour burden or bulky mass (Table 1). Conclusion: In our study, we were able to identify genetic variables that could characterize different histological groups of NHL-B in FFPE tissue and ctDNA samples, by using a commercial gene panel of NGS. The mutational profile obtained from ctDNA and FFPE tissue is comparable in all histological subtypes, regardless tumour burden, aggressiveness or stage, throwing each one of them complemental information. These results support the hypothesis that could be possible to distinguish different histologies through non-invasive strategies, specially oriented to lesions where obtaining a tissue sample is difficult. We are working on the development of clinical-genetic algorithms that allow us to achieve our goal. Disclosures No relevant conflicts of interest to declare.

Around 20% to 30% of patients with Hodgkin lymphoma (HL) do not benefit from standard therapies and finally succumb to their disease. The factors that influence the outcome of HL have not been elucidated, underscoring the demand for the identification of biologic risk factors and new therapeutic targets. We analyzed the gene expression profiles of samples from 29 patients with advanced classic HL treated with standard therapy and compared the expression profiles of patients with favorable and unfavorable clinical outcome. Using supervised methods, we identified 145 genes associated with outcome, which were grouped into 4 signatures representing genes expressed by either the tumoral cells (genes involved in the regulation of mitosis and cell growth/apoptosis) or the tumor microenvironment. The relationship between the expression of 8 representative genes and survival was successfully validated in an independent series of 235 patients by quantification of protein expression levels on tissue microarrays. Analysis of centrosomes and mitotic checkpoint confirmed the existence of an abnormal transition through mitosis in HL cells. Therefore, genes related to tumor microenvironment, cell growth/apoptosis, and regulation of mitosis are associated with treatment response and outcome of patients with HL.

Splenic marginal zone lymphoma (SMZL) is a newly recognized lymphoma type whose precise molecular pathogenesis is still essentially unknown. This hampers differential diagnosis with other small B-cell malignancies. With the aim of characterizing this tumor more comprehensively, and of identifying new diagnostic and prognostic markers, we performed cDNA microarray expression profiling and tissue microarray (TMA) immunohistochemical studies in a relatively large series of 44 SMZLs. The results were related to immunoglobulin heavy chain variable region (IgVH) mutational status and clinical outcome. SMZLs display a largely homogenous signature, implying the existence of a single molecular entity. Of the genes deregulated in SMZLs, special mention may be made of the genes involved in B-cell receptor (BCR) signaling, tumor necrosis factor (TNF) signaling and nuclear factor-κB (NF-κB) activation, such as SYK, BTK, BIRC3, TRAF3, and LTB. Other genes observed were SELL and LPXN, which were highly expressed in spleen, and lymphoma oncogenes, such as ARHH and TCL1. In contrast, the genes CAV1, CAV2, and GNG11 located in 7q31, a commonly deleted area, were down-regulated in the entire series. A comparison with the genes comprising the signature of other small B-cell lymphomas identified 3 genes whose expression distinguishes SMZL, namely ILF1, SENATAXIN, and CD40. Shorter survival was associated with CD38 expression, naive IgVH genes, and the expression of a set of NF-κB pathway genes, including TRAF5, REL, and PKCA. (Blood. 2005;106:1831-1838)

Tumoral cells in Hodgkin lymphoma (HL) display an increased growth fraction and diminished apoptosis, implying a profound disturbance of the cell cycle and apoptosis regulation. However, limitations of molecular techniques have prevented the analysis of the tumor suppressor pathways and cell-cycle checkpoints. Tissue microarray (TMA) is a powerful tool for analyzing a large number of molecular variables in a large series of tumors, although the feasibility of this technique has not yet been demonstrated in heterogeneous tumors. The expression of 29 genes regulating the cell cycle and apoptosis were analyzed by immunohistochemistry and in situ hybridization in 288 HL biopsies using TMA. The sensitivity of the technique was validated by comparing the results with those obtained in standard tissue sections. The results revealed multiple alterations in different pathways and checkpoints, including G1/S and G2/M transition and apoptosis. Striking findings were the overexpression of cyclin E, CDK2, CDK6, STAT3, Hdm2, Bcl2, Bcl-X(L), survivin, and NF-kappaB proteins. A multiparametric analysis identified proteins associated with increased growth fraction (Hdm2, p53, p21, Rb, cyclins A, B1, D3, and E, CDK2, CDK6, SKP2, Bcl-X(L), survivin, STAT1, and STAT3), and proteins associated with apoptosis (NF-kappaB, STAT1, and RB). The analysis also demonstrated that Epstein-Barr virus (EBV)-positive cases displayed a characteristic profile, confirming the pathogenic role of EBV in HL. Survival probability depends on multiple biologic factors, including overexpression of Bcl2, p53, Bax, Bcl-X(L), MIB1, and apoptotic index. In conclusion, Hodgkin and Reed-Sternberg cells harbor concurrent and overlapping alterations in the major tumor suppressor pathways and cell-cycle checkpoints. This appears to determine the viability of the tumoral cells and the clinical outcome.

Classical Hodgkin´s lymphoma (cHL) is a clonal proliferation of the characteristic Hodgkin and Reed Sternberg cells (HRS), diluted in a reactive inflammatory background, and characterized by a defective B cell immunophenotype. Few studies have been aimed to the identification of the driver mutations in this disease, and only occasional gene mutations in members of the NF-kappaB and JAK/STAT pathways, and, more recently B2M, have been previously described. We analyzed 57 cHL tumor samples (FFPE) using massive parallel sequencing (Ion Torrent™ semiconductor technology). To increase coverage a previous tumor cell enrichment process by punch tissue cores from selected tumor-rich areas was implemented. In addition, we sequenced 7 cHL cell lines using the same technical approach. Bioinformatics analysis, including alignment with germline sequences and SNPs filtering, were done using the Torrent Suite Software and Variant Caller. Pathogenic prediction of single nucleotide variants (SNVs) and mutation interpretation were performed using the Alamut and Provean softwares. All experiments were done in duplicate to minimize false positive rates. Overall, the results show very high genomic heterogeneity with a range of 10-400 SNVs per sample, most of them (~60%) missense type. We found a relatively large number of genes recurrently mutated at low frequency and only a few genes mutated in up to 15-20% of the patients, reflecting a high level of genomic instability in the neoplasm. Specific mutations in genes previously described in cHL (NFKBIA, TNFRSF14) and in diffuse large B-cell lymphomas (CARD11, STAT6, CREBBP, CMYB) were consistently found, as well as new SNVs in genes not previously described (BTK, NFKB2). Mutations affecting selected genes (B2M, CARD11, NFKBIA, CSF2RB, and STAT3) in cell lines were additionally validated by Sanger sequencing. Of note, we found high prevalence of mutations affecting BTK and the BCR pathway, suggesting some dependencies of active BCR signaling albeit the absence of BCR expression by HRS cells. Consistent with this interpretation, incubationof a panel of cHL cellular models with either Ibrutininb or AVL-292, selective-BTK inhibitors, in vitro constrains cell proliferation and causes cell death. Interestingly, the concentration to induce cell apoptosis in cHL was comparable to other cell models of diffuse large B-cell and mantle cell lymphoma sustaining the requirement of an active BCR signaling for cHL cell viability. In conclusion, cHL is characterized by high genomic instability, including numerous mutations in genes related with B-cell function and specific signaling pathways. Pathogenic predictions of the different SNVs identify potential driver mutations that can be associated with the pathogenesis of the disease and might represent new therapeutic targets. Disclosures No relevant conflicts of interest to declare.

To the editor: Peripheral T-cell lymphomas (PTCLs) are a group with poor outcome and nonspecific therapeutic regimens. Recently, Palomero et al[1][1] and Sakata-Yanagimoto et al[2][2] identified a recurrent heterozygous mutation in the RHOA small GTPase gene encoding a p.Gly17Val alteration (G17V)

Abstract 3494 The search of an efficient therapy for Peripheral T-cell lymphomas (PTCL) patients is still a challenge, in part due to the very little knowledge about the PTCL pathogenesis, and the absence of appropriate models. This heterogeneous group of very aggressive malignancies can not be cured with conventional therapies; therefore, new therapeutic strategies are needed to improve the poor outcome in these patients. PIM family is composed of 3 kinases (PIM1, PIM2 and PIM3) which play an essential role in cell proliferation and survival. They are mainly activated through JAK/STAT signaling pathway, and are frequently altered in human malignancies by amplification, mutation and overexpression. The aim of this study is to determine the efficiency and the mechanism of action of PIM inhibition in PTCL. Gene expression profiling of twenty two PTCL cases and seven reactive lymph nodes was performed. We observed a strong overexpression of the three PIM family genes in PTCL cases, especially PIM2. In addition, Gene Set Enrichment Analysis identified an overexpression of STAT3 and IL-2 pathways in PTCL cases, probably responsible for the strong expression of PIMs we found. Furthermore, PIM genes expression was confirmed by quantitative RT-PCR in 6 PTCL-derived cell lines compared to normal T cells from healthy donors, highlighting again the relevance of PIM2. Genetic inhibition was carried out using small interference RNA to specifically abolish the expression of each PIM1, PIM2 and PIM3 in a panel of 6 PTCL cell lines belonging to different PTCL subgroups. Additionally, pharmacological inhibition with one PIM inhibitor (ETP-39010) was performed. Surprisingly, genetic inhibition of each of the PIM gene alone did not show any cellular effect, neither cell cycle arrest nor apoptosis. But interestingly, we found that specific inhibition of each of the PIM genes caused an increased expression of the other PIM family members, probably leading to a compensatory mechanism among these kinases balancing the lack of one of them, avoiding pro-apoptotic effects and allowing cell survival. Accordingly, a simultaneous inhibition of PIM1, PIM2 and PIM3 using the pharmacological pan-PIM inhibitor produced a decrease in cell viability and a strong induction of apoptosis in all cell lines, without cell cycle arrest. Several PIM inhibitor biomarkers have been identified at the mRNA level, involving the DNA damage response signaling. In conclusion, our results indicate that PIM kinases inhibition could be an effective therapeutic approach for PTCL. Disclosures: No relevant conflicts of interest to declare.

Abstract 4859 BACKGROUND: Most patients with classical Hodgkin's Lymphoma (CHL) are cured with primary treatment. However, a proportion of them fail to first line treatment needing to be rescued with subsequent lines of chemotherapy and/or autologous or allogeneic stem cell transplantation (auto-SCT and allo-SCT, respectively). The identification of clinical and biological characteristics of these patients at diagnosis is still a challenge and most prognostic systems fail to identify a proportion of patients with worse prognosis. In this context, different groups are currently analyzing several biological markers as determinants of clinical outcome. It has been reported that Bcl-2 immunohistochemical expression in Hodgkinxs Reed Sternberg cells (HRSC) might confer a worse prognosis. OBJETIVE: To analyze clinical outcomes following 1st line chemotherapy according to Bcl-2 expression at diagnosis of CHL. PATIENTS AND METHODS: CHL patients, older than 16 years old, receiving at least 1 line of treatment, were retrospectively studied for Bcl-2 expression in diagnostic samples. For this purpose, tissue sections were immunostained and semiquantitatively assessed for this marker. Cumulative incidence (CI) of treatment failure, treatment failure free survival (TfFS) and overall survival (OS) were defined as primary outcomes. Treatment failure was considered when a different treatment regimen was set up due to relapse after CR or failing to achieve CR following 1st line. RESULTS: 103 patients (55 Bcl-2 positive patients and 48 Bcl-2 negative patients) were analyzed. Main patient and clinical features are shown in Table 1. Both cohorts were well balanced for the main prognostic factors. At a median follow up of 36m (2-221), 34m (2-140) for Bcl-2 negative patients and 38m (4.5-221) for Bcl-2 positive patients, CI of 3 years treatment failure was 19% and 50% for the negative and positive cohorts, respectively (p=0.02). 3 years TfFS after diagnosis was 75% for Bcl-2 negative patients vs 47% for Bcl-2 positive patients (p=0.1). Within the cohort of Bcl-2 negative patients, 9/48 (19%) underwent auto-SCT as part of rescue treatment while 18/55 (33%) of Bcl-2 positive patients received an SCT (13 auto-SCT and 5 allo-SCT) and 3 of them, a second SCT (allo) for the treatment of post-auto-SCT relapse. 3 years OS was 84.5% for negative and 86% for positive patients (p=NS). CONCLUSIONS: According to these preliminary results, Bcl-2 expression in HRSC at diagnosis may constitute an independent biological prognostic marker in CHL patients, seemingly associated with worse outcome and need of second line chemotherapy. More studies and a longer follow up is needed in order to confirm that aggressive treatment strategies such as SCT may overcome the negative impact in survival of Bcl-2 expression in this population. Disclosures: No relevant conflicts of interest to declare.

Abstract 3493 Peripheral T-cell lymphomas (PTCL) are a heterogeneous group of very aggressive malignancies lacking efficient therapy. Unfortunately, there are neither animal models nor representative cell lines for most PTCL types, making functional and pharmacodynamic studies even more difficult. PI3K signaling is essential for cell proliferation and survival, is frequently altered in human cancer and seems to play a critical role in T-cell development and activation. The aim of this work is to determine the efficiency of PI3K inhibition in PTCL, looking for pharmacodynamic biomarkers, and to identify markers that could distinguish responders from non-responders. Twenty two PTCL cases and seven reactive lymph nodes were studied using gene expression profiling. We performed an in silico analysis using the Connectivity Map program to identify drugs that could potentially reverse the PTCL gene expression signature. Among them, several PI3K/mTOR inhibitors were found. Moreover, genomic studies using Gene Set Enrichment Analysis identified PIK3CD gene (encoding for the delta isoform of PI3K) to be the only one significantly correlated to the activation of CD40, NF-kB and TCR pathways. Quantitative RT-PCR confirmed the strong overexpression of PIK3CD in 6 PTCL-derived cell lines compared to normal T cells from healthy donors. Sequence analyses for the coding region of the PIK3CD gene identified a point mutation in one of these cell lines, described as activating in solid tumors. A panel of 6 PTCL cell lines belonging to different PTCL subgroups was treated with 3 PI3K inhibitors (LY294002, ETP-45658, GDC-0941). Moreover, genetic inhibition was also carried out using small interference RNA to specifically abolish the expression of alpha and delta isoforms of PI3K (PIK3CA and PIK3CD genes, respectively). In vitro studies showed very similar results with the three pharmacological PI3K inhibitors we used: they induced G1 cell cycle arrest in all cell lines, and apoptosis in some of them, in a time/dose-dependent manner. We also observed a decrease in the levels of pAKT(S473) in all cell lines, while pGSK3B(S9) and p-p70S6K(T389) were reduced after treatment only in sensitive cell lines. Our results indicate that genetic inhibition of PI3K delta isoform could induce apoptosis in those PTCL cell lines that were sensitive to PI3K inhibitors, but not in the resistant cell lines; while genetic inhibition of PI3K alpha isoform did not display such effects. Taken together these results could highlight the relevance of PI3K delta isoform in at least a subset of PTCL, indicating that PI3K inhibition, especially delta isoform, could be an effective therapeutic approach for PTCL and identifying potential markers for patients' stratification and pharmacodynamic assessment. Disclosures: No relevant conflicts of interest to declare.

Abstract 5218 INTRODUCTION: Diffuse large B-cell lymphoma (DLBCL) represents more than 80% of high grade lymphomas and is considered a heterogeneous entity regarding clinical presentation, morphology, molecular and cytogenetic features. The Revised International Prognostic Index (R-IPI) is a clinically useful prognostic index that may help guide treatment. In this context, BCL2 and MYC gene rearrangements, that have been associated to specific variants of lymphoma, seem to have a prognostic implication. Consideration of these genetic markers together with classical parameters could contribute to a better stratification of patients. OBJETIVE: To analyze the incidence and possible clinical relevance of BCL2 and MYC gene alterations in the outcome of patients with DLBCL. PATIENTS/METHODS: This retrospective analysis included 20 patients with DLBCL (January 2008-April 2011), 74% (20/27) males with a median age of 70 years (range 17–82). Median follow-up was 300 days (range 17–1134). Fluorescence in situ hybridization (FISH) was performed on paraffin-embedded (n=12) or fresh (“touch-down” preparation) samples (n=8) from different tissues (7 lymphadenopathies, 3 central nervous system, 3 biologic fluids and 7 others). Each sample was hybridized with break-apart probes for BCL2 (18q21) and MYC (8q24) genes (Vysis Inc). 200 nuclei were scored per slide to reach 7% sensitivity. RESULTS: Seventy percent (14/20) of patients analyzed showed BCL2 and/or MYC alterations at diagnosis (rearrangement and/or extra signals). BCL2 was present as unique alteration in 30% (6/20) of the samples, MYC was present as unique alteration in 30% (6/20) and 10% (2/20) showed both BCL2 and MYC rearrangement (Table 1). Two out of six patients with BCL2 as unique alteration showed rearragement, were treated with standard chemotherapy (R-CHOP or high dose of Citarabine/Methotrexate in central nervous system) and both patients relapsed. The other 4 patients presented extra signals as BCL2 alteration and also received standard chemotherapy. Fifty percent of these patients relapsed (Table 1). On the other hand, 3 out of 6 patients with MYC as unique alteration presented gene rearrangement, they were treated with standard chemotherapy, and all 3 relapsed. The other 3 patients with MYC alterations presented extra signals, received standard chemotherapy and 33.3% (1/3) relapsed. Finally, the 2 patients with simultaneous BCL2 and MYC rearrangement were treated with intensive chemotherapy and they reached and maintained complete remission. CONCLUSIONS: According to the results obtained from this study, FISH analysis on paraffin-embedded or fresh samples from lymphoma tissues is a feasible and useful technique at diagnosis in DLBCL patients. Although genetic markers are not included in the R-IPI score, the identification of BCL2 and/or MYC alterations at diagnosis, associated to other biological markers, could help to identify a subgroup of high risk patients who could benefit from more aggressive therapies. Further studies of larger series and genes are needed to confirm this observation. Disclosures: No relevant conflicts of interest to declare.

Abstract 4825 Introduction: The majority of patients with classical Hodgkinxs Lymphoma (cHL) are cured with primary treatment. However, a significant proportion of patients will not achieve a complete response (CR) or relapse after completion of initial therapy and need to be rescued with high-dose chemotherapy and stem cell transplantation (SCT): autologous (aSCT) and/or alogeneic (Alo-SCT). The identification of clinical and biological characteristics of these patients at diagnosis is still a challenge and the most prognostic systems used to date fail to identify a proportion of patients with worse prognosis. The best understanding of the biology of cHL could help to identify these patients and different groups works in the use of biological marker as determinants of clinical outcome. Bcl2 over-expression has been described in cHL and seems to be an independent marker associated to bad prognosis in probably relation with alterations in apoptosis regulatory molecules. Objetive: To retrospectively analyze the frequency of Bcl2 protein over-expression in tissue biopsies of patients diagnosed with cHL, which required intensive treatment in our centre. Patients and Methods: We revised clinical data and samples at diagnosis of patients with cHL who received SCT (aSCT or/and Allo-SCT) due to partial remission, relapse or progression and we determined Bcl2 expression protein by immunohistochemistry. All patients received at least 2 lines of treatment previous to intensive treatment with SCT. We analyzed the expression of Bcl2 according to each patient histology: Nodular Sclerosis (NE), Mixed cellularity (MC), Lymphocyte-rich (LR), and Lymphocyte-depleted (LD). Results: Between September 1997 and May 2010, 39 patients with cHL required intensive treatment in our center: 32 aSCT and 7 Allo-SCT due to partial remission, refractory, relapse or progression after first line of treatment. Average age was 32 years (range: 18–64), males: 25/females 14. The characteristics of patients and remission status at transplant are described in table 1. We had available samples at diagnosis from 31 out of 39 patients (80%): 20 NE, 5 MC, 6 LR and we found over-expression of Bcl2 in 17/20 (85%) NE; 5/5(100%) MC and 2/6 LR (33%). Conclusions: Our results suggest that over-expression of Bcl2 is frequent in patients with cHL which needed intensive treatment after failure of the first line of therapy (particularly with NE and MC). Further studies are needed to confirm the worse prognosis of Bcl2 expression in cHL and if may be useful at diagnosis in association with other clinical parameters to identify patients with poor prognosis. Table 1 . CLASSICAL HODGKIN LYMPHOMA (c HL) NODULAR SCLEROSIS cHL MIXED CELLULARITY cHL LYMPHOCYTE-RICH cHL LYMPHOCYTE-DEPLETED cHL N (39) 23 8 7 1 Sex (F/M) 11/12 8/0 5/2 1/0 Age, Years: Md(R) 30(18-64) 37(31-62) 36(19-64) 54 Stage: I, II/III, IV 11/12 3/5 5/2 0/1 Bulky disease 14YES/9NO 1YES/7NO 2YES/5NO 1NO B Symptoms: (N) YES/NO (13) YES/(10) NO (3) YES/(5) NO (6) YES/(1) NO (1) YES Intensive treatment 5Alo-SCT/18aSCT 2AloSCT/6aSCT 7aSCT 1aSCT N:type of trasplant (status at transplant) 18: aSCT(11CR/5PR/2RF); 5:AloSCT(2PR with previous aSCT/1PR/1CR/1RF) 6: aSCT (4CR/2PR) 7:aSCT (6CR/1PR) 1:aSCT (in CR) 2: Alo-SCT(1CR/1PR with previus aSCT) Sample revised: N 20 (23 total NE) 5 (8 total MC) 6 (7 total LR) 0/1 Bcl2 expression: P/Nv (%) 17P/3Nv (85%) 5P/0Nv (100%) 2P/4Nv (33%) Not aplicable Alo-SCT: Alogeneic stem cell transplant; aSCT: autologous stem cell transplant; CR: Complete Remission; F: female; LD: Lymphocyte-deplete; LR: Lymphocyte rich; M: male; MC: mixed cellularity Md: median; n: number; NE: nodular sclerosis; Nv: negative; P: positive; PR: Parcial remision R: range; RF: refractory Disclosures: No relevant conflicts of interest to declare.

Abstract 2936 Poster Board II-912 Mantle cell lymphoma (MCL) pathogenesis is still partially unexplained. Although the overexpression of CyclinD1 dependent of t(11;14) is a distinctive molecular hallmark of this neoplasm, this event alone cannot account for the increased survival signaling that characterizes this lymphoma type. Here we investigate whether microRNA (miRNA) expression profile may help to explain the changes in the expression of gene pathways that are characteristic of MCL. Twenty-three frozen MCL samples, 11 frozen control tissues (7 lymph nodes and 4 tonsils), 8 MCL cell lines and 3 samples of CD19+IgD+CD27- cells obtained from tonsils, were studied for miRNAs and gene expression. MiRNA one color microarray data for 470 human miRNA were analyzed using SAM (Significance Analysis of Microarrays) algorithm. MiRNA targets were predicted by miRanda and TargetScan methods. Pathways identification and analysis was carried out by GSEA (Gene Set Enrichment Analysis) online resource. The analysis of 23 MCL cases compared to 11 control tissues showed a miRNA signature that included 117 miRNAs with FDR MiRNAs were used also for the identification of survival prognostic markers; using different analysis (22 frozen specimens) and validation (54 paraffin embedded cases) series. A single miRNA distinguished a group of MCL cases with a 72.2% survival at 60 m. This study identifies a set of miRNAS involved in MCL pathogenesis, which could be used in MCL recognition and clinical prognostication. Disclosures: No relevant conflicts of interest to declare.

WHO criteria defines platelet counts above 600×109/L as the threshold for essential thrombocythemia (ET) diagnosis. It has been argued that such threshold excludes a number of patients with ET with platelet counts below 600×109/L. Recently, a proposal for revision of the World Health Organization (WHO) diagnostic criteria for ET has been published, which includes the combination of histological bone marrow study and testing of JAK2 mutation. Design and methods : Retrospective analysis of 92 patients with ET diagnosis between 1989 and February 2008, isolating the subgroup of patients with platelet counts below 600×109/L. The aim of this study was to analyze the applicability of the 2008 WHO criteria in this subgroup. Results : Of the 92 patients, 30 patients did not fulfill the WHO criteria due to platelet counts 3.5 | 30/88 (34%) | 8/28 (29%) | | Fibrosis grade I | 2/90 (2%) | | | Compatible histology | 79/89 (86%) | 21/28 (75%) | | Abnormal Cytogenetics | 2/26 (8%) | | | Symptoms | 13 (14%) | 4 (13%) | | Thrombotic event | 16 (18%) | 5 (17%) | | Haemorrhagic event | 3 (4%) |

Despite the major advances in the treatment of classical Hodgkin Lymphoma (cHL) patients, around 30% to 40% of cases in advanced stages may relapse or die as result of the disease, and current markers to predict prognosis are rather unreliable. The identification of molecular events and biological processes associated with treatment failure are essential to develop new predictive tools. We used gene expression data from 29 samples of advanced cHL patients and HL-derived cell lines in order to identify transcriptional patterns from both tumoral cells and cell microenvironment. Student t-test was used to detect genes differentially overexpressed in cell lines and in tumor samples, thus creating two databases that report for genes expressed by the tumor HRS cells and genes expressed by the microenvironment. Using Gene Set Enrichment analysis (GSEA) we identified specific gene sets enriched in both databases in patients with favorable and unfavorable outcome, respectively. To validate these pathways we designed a novel Taqman low-density array (LDA) to examine the expression of the most relevant genes in 60 formalin-fixed, paraffin embedded (FFPE) tissue samples, and correlated the results with treatment outcome. Functional pathways related to unfavorable outcome significantly enriched in the HRS cells included the regulation of the G2/M checkpoint of the cell cycle, S phase and G1/S transition, chaperons, histone modification and other signaling pathways with an important representation of the MAPK pathway. On the other hand, genes reporting for specific T-cell populations (T-cytotoxic and T-regulatory cells) and macrophage activation were found to be overexpressed in the microenvironment. The final model presents a balanced representation of these genes, including also genes encoding factors implicated in drug resistance (RRM2, TYMS and TOP2A). RNA extracted from FFPE sections yielded analyzable data for 80% of samples. LDA analysis of the genes included in the model confirmed the feasibility of this approach, and the capacity for identifying cases with increased risk of failure.LDA provides an effective technique for analyzing gene expression in FFPE tissues, and it can be used for clinical prediction in diagnostic samples, using a selection of genes identified after GSEA analysis of the initial molecular signatures. This novel Taqman LDA will be used to develop a new molecular predictor of the outcome of patients with advanced cHL.

Detection of specific chromosomal abnormalities is essential in the diagnosis of several lympholiferative disorders. However, conventional cytogenetic studies are not frequently carried out from biopsies as it is in bone marrow (BM) specimens. On the contrary, FISH is usually performed on paraffin embedded tissue, an alternative with potential technical nuances both in its application and its interpretation. In our experience, FISH on tissue imprints is the ideal alternative to overcome these problems. In the present study, 46 tissue imprints and 17 BM smears from 43 patients with lymphomas were selected to investigate the presence of t(14;18)(q32;q21), t(11;14)(q13;q32), t(8;14)(q24;q32) and t(3;var)(q27;var). Representativity of the samples was assured prior to FISH by rapid May-Grümwald staining (Diff-Quick, QCA, Spain). Twenty imprints from reactive palatine tonsils and adenoids were used as negative controls. FISH was performed with specific dual-color dual-fusion FISH (D-FISH) probes for the first 3 translocations and a dual-color break-apart FISH probe for t(3;var)(q27;var) following the instructions of the probes supplier (Vysis, Inc). All except one sample (a BM smear stored at room temperature over 9 years), rendered satisfactory FISH hybridizations. In any case, all 43 patients could be successfully studied by FISH (the case referred above in which FISH failed in a first attempt was successfully analyzed using a different sample). The results supported the suspected diagnosis of follicular lymphoma in 22 patients, mantle cell lymphoma in 12 patients, large B-cell lymphoma in 5 patients, marginal zone lymphoma in 2 patients and B chronic lymphocytic leukemia in 2 patients. In one case, the observation of t(11;14) by FISH allowed to reclassify the case from follicular to mantle cell lymphoma and also to demonstrate clonal evolution by studying sequential tissue imprints stored for more than 6 years. Negative results also aided in the assignment of patients to proper diagnostic categories. FISH performed on tissue imprints and BM smears constitutes an optimal strategy for retrospective and prospective investigation of chromosomal abnormalities in lymphomas. Tissue imprints and BM smears conventionally stored at room temperature even for long periods of time can be safely used and sent by ordinary mail without special considerations for this purpose. Based on our experience we recommend to perform and routinely store tissue imprints whenever fresh tissue is available.