ALBERT

Description: Abstract 2720 Introduction: Bendamustine is a promising agent that has been showed high efficacy in combination with Rituximab in indolent lymphomas; however experiences on its use in combination with other cytotoxic agents are scant, particularly in FL. A previous FIL trial showed that a brief R-FND induction chemoimmunotherapy with only 4 courses of R-chemotherapy followed by 4 rituximab doses as consolidation can achieve high CR and PFS rates in FL patients, supporting the feasibility of this regimen in the elderly (Vitolo et al, ASH 2011). These promising results prompted FIL to investigate safety and efficacy of a similar combined brief regimen substituting Bendamustine for Fludarabine aimed at reducing toxicity and maintaining efficacy. Patients and methods: From September 2009 to November 2011, 76 elderly patients (age 65–80) with advanced stage FL at diagnosis were enrolled. Inclusion criteria were: untreated grade I, II and IIIa FL; advanced disease requiring treatment (stage III/IV) or stage II including at least one of the following conditions: bulky disease (〉7 cm), active disease with rapid lymph node progression, lactate dehydrogenase or β2 microglobulin above normal, systemic symptoms and extranodal involvement. A comprehensive geriatric assessment (ADL, I-ADL, CIRS) was also performed at diagnosis and only “FIT” patients were enrolled into the study. Treatment plan was: 4 monthly courses of R-BM (375 mg/m2 Rituximab day 1, 90 mg/m2 Bendamustine days 1–2, 8 mg/m2 Mitoxantrone day 1), followed by consolidation with 4 weekly Rituximab infusions. Polymerase chain reaction (PCR) for BCL2/IgH rearrangement was performed on bone marrow samples at diagnosis, after R-BM and after consolidation. Results: At the time of analysis, 69/76 patients are evaluable for clinical characteristics, response to treatment and toxicity. Median age was 71 years (range 65–80); 26 males, 43 females; WHO grading was as follow: I 16%, II 55% and IIIa 29%; 19% had advanced stage II disease, 27% stage III and 54% stage IV; 49% had BM involvement, 19% B symptoms and 7% leukemic dissemination; 58% patients had no comorbidity, 20% one and 22% two or more comorbidities. According to FLIPI patients were: 12% at low, 30% at intermediate and 58% at high risk. PCR analysis was carried out in 64 patients at diagnosis: 58% were Bcl-2 positive. Sixty three patients completed the planned treatment and the whole program was completed according to the planned time in 60/63 patients. Six patients did not complete the treatment: 1 for progressive disease, 4 for adverse events (2 haematological toxicity with prolonged neutropenia; 1 CMV colitis and 1 for infection and concomitant worsening of pre-existing oral pemfigo) and 1 patient for worsening of performance status. Overall response after R-BM + Rituximab was observed in 64 (96%) patients with 52 (75%) patients in complete remission and 12 (18%) in partial remission (PR). Response to 4 cycles of R-BM was as follow: 27 (39%) in CR, 37 (54%) in PR and 5 (7%) patients in stable (SD) or progressive disease. Of the 40 patients in PR or SD after 4 R-BM, 26 (65%) converted to CR following further Rituximab consolidation. At the time of this analysis, of the 37 patients Bcl-2 rearranged at diagnosis, 19 (51%) were evaluable for molecular response during and after treatment. PCR negativity was achieved in 16/19 (84%) patients after R-BM and in 18/19 (95%) patients at the end of treatment. A total of 521 courses of R-BM and Rituximab were given. The most frequent severe toxicity (CTC grade 3–4) was neutropenia reported in 18% of the courses; nevertheless only four serious infections were recorded (all due to bacterial agents). Pulmonary and cardiac toxicities were 〈 1%. Two deaths were recorded: one due to lymphoma after second line treatment and one due to hepatic carcinoma occurred 3 months after completion of therapy. Conclusions: A brief course of chemo-immunotherapy R-BM followed by Rituximab consolidation is safe and effective with a high CR rate in elderly patients with untreated advanced stage FL. Planned future analysis of the entire study will give further information on late toxicity and outcome. Disclosures: Off Label Use: In Europe use of Bendamustine in first-line treatment of follicular lymphoma is off-label. Drug was provided free by Mundipharma. Vitolo:Roche: Membership on an entity's Board of Directors or advisory committees.

Description: Introduction: Accordingto National Comprehensive Cancer Network (NCCN) and ESMO guidelines on Waldenstrom’s Macroglobulinemia (WM) bendamustine may be considered as a therapeutic option in first line treatment or in relapsed refractory disease. Even though there are only two clinical trials including a limited number of patients addressing the role of bendamustine and rituximab (BR) treatment in WM. Patients and Methods: To define the efficacy and tolerability of BR combination as salvage regimen in WM patients, we retrospectively analyzed the outcome of symptomatic refractory relapsed patients treated with BR in 14 Italian centres. All patients receiving at least one day of treatment were included in the study. Treatment consisted of: R 375 mg/sqm iv day 1 and B iv days 1, 2. Therapy was administered every 4 weeks up to 6 courses. Results: Seventy-one patients are included in the study. As regards B dosage; 45 patients (63%) received the highest dose of 90 mg/sqm while 22 (31%) were treated with 70 mg/sqm. The 4 patients (6%) with a cumulative illness rating scale ≥ 6, received the lowest dose of 50 mg/sqm. At treatment, median age was 72 years (49-88), sex ratio M/F 46/25. Mediannumber of prior regimens was 2 (range 1-6). Twenty-four patients (34%) presented with refractory disease. The majority (90%) of patients had been previously treated with alkylating agents, 30% had also received purine analogues based treatments. Previous R was administered in the 77% of cases. The main reason (62%) for starting treatment was anemia followed by adenopathy and/or splenomegaly (35%). Median IgM level at treatment was 3815 mg/dL.Overall 361 courses of BR treatment were administered, median number 6 (range 1-6) with 47 (66%) of patients completing the 6 planned courses. Toxicity was discontinuation cause in 10 patients (14%): 4 infection, 1 fatal, 6 myelosuppression. In the remaining 14 treatment was discontinued for clinical clinical decision after disease reassessment. No difference in terms of treatment discontinuation was observed according to B dosage and age. Overall response rate (ORR) was 80.3% including: 7% complete remissions (CR), 15.5 % very good partial remissions (VGPR), 52.2% partial remissions (PR) and 5.6% of minor responses. A stable disease was observed in 16.9% of patients. One (1.4%) disease progression and one death were recorded. A progressive decrease of IgM level was observed during follow-up leading to an amelioration of response in 4 cases leading to a final ORR of 84.5%. None of the clinical and biological characteristics considered (age, sex, disease status, previous lines of treatment, previous fludarabine, bulky disease, Hb and IgM level, beta 2 microglobulin, B dosage) had an impact on ORR achievement. A better quality of response (CR plus VGPR) was observed in patients with an IgM level 〈 3000 mg/dL and in those treated with the higher dosage of B (90 mg/sqm). After a median follow-up of 19 months (3-54) 11 of the 57 responding patients met the criteria for disease progression. No difference was observed when patients were stratified according to the quality of response. B dosage did not impact disease progression. Considering that most of the patients received prophylactic growth factors, grade 3-4 neutropenia developed in only 13% of courses, 36% of patients. Dose modification or delayed treatment administration was necessary in 4 and 10% of courses respectively. During treatment we recorded 14 episodes of FUO and 5 major infections, leading death in one case. After a median follow up of 19 months none of the patients developed secondary myelodisplastic syndrome, acute leukemia or diffuse large B-cell lymphoma. In 3 cases a solid cancer was observed. Conclusion: BR combination showed to be as effective as more intensive salvage regimens in pretreated WM patients. Treatment showed to be well tolerated even in elderly patients with limited episodes of myelosuppression and infections when compared to purine analogues including regimens. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 1796 Poster Board I-822 The rationale of the study stems from three considerations: i) to date, few clinical predictors of asymptomatic multiple myeloma (MM) progression are available; ii) no study has assessed the role of beta-2-microglobulin (B2M) on risk of asymptomatic MM progression; iii) B2M is a serum marker of tumor burden and represents a key variable of the ISS staging system for symptomatic MM. The study aims at assessing the impact of B2M on risk of asymptomatic MM progression. The study was based on a consecutive series of 148 asymptomatic MM diagnosed according to the IMWG. Cumulative probability of progression to symptomatic MM was calculated from diagnosis of asymptomatic MM to progression to symptomatic disease according to IMWG. Survival analysis was performed by Kaplan-Meier method using log-rank to test for associations. Cox proportional hazard regression was used to build a multivariate model. Harrell's c-statistics was used to evaluate the discriminatory value of the prognostic models. Clinical features at asymptomatic MM diagnosis were as follows: median age 67 years, male:female ratio 1.02, previous MGUS 32/148 (21.6%), IgG monoclonal component (MC) 113/148 (76.4%), IgA MC 32/148 (21.6%), light chain MC 3/148 (2.0%), median serum monoclonal component (sMC) 1.1 g/dl, positive urinary immunofixation 27/148 (18.2%), median urinary monoclonal component (uMC) 98 mg/24h, median bone marrow plasma cell percentage (BMPC%) 15%, median B2M 2.0 mg/l, median albumin 4.3 g/dl, median C reactive protein 0.3 mg/l. After a median follow-up of 48 months, 29/148 asymptomatic MM progressed to symptomatic disease, accounting for a 29.5% 5-year probability of progression. According to Youden's index, best cut-off values for prediction of progression were 2.5 mg/l for B2M, 1.5 g/dl for sMC, 500 mg/24h for uMC, and 20% for BMPC%. Univariate analysis identified B2M 〉2.5 mg/l (5-year risk: 64.5%; HR=3.85; p1.5 g/dl (5-year risk: 49.1%; HR=5.76; p500 mg/24h (5-year risk: 68.7%; HR=6.60; p1.5 g/dl (HR=4.07; p=.003), uMC 〉500 mg/24h (HR=3.66; p=.015) and BMPC% ≥20% (HR=3.20; p=.007). Based on multivariate analysis, B2M, sMC, uMC, and BMPC% were combined into a model for predicting progression to symptomatic MM. This model stratified patients into four risk groups: very low risk (0 risk factors: 5-year risk 0%), low-intermediate risk (1 risk factor: 5-year risk 19.6%), high-intermediate risk (2 risk factors 5-year risk: 60.7%), and high risk (3 or 4 risk factors 5-year risk 80.7%) (Fig. 1). Three lines of evidence suggest that B2M may improve the conventional risk stratification model for prediction of asymptomatic MM progression defined by sMC, uMC, and BMPC%. First, B2M 〉2.5 mg/l adds prognostic information within the low risk group characterized by sMC '1.5 g/dl, uMC '500 mg/24h and BMPC 1.5 g/dl and/or uMC 〉500 mg/24h and/or BMPC ≥20% (p=.034) by segregating a high risk group of patients with B2M 〉2.5 mg/l who are virtually all projected to progress (5-year risk: 89.0%). Third, based on c-statistics, the prognostic model including B2M along with sMC, uMC, and BMPC% (c-statistics=.760; SE=.045) allowed to predict asymptomatic MM progression better than a model based only on sMC, uMC, and BMPC% (c-statistics=.726; SE=.049). The implications of our results are twofold: i) B2M predicts progression of asymptomatic MM to symptomatic disease independent of conventional risk factors (sMC, BMPC%, light chain burden); ii) B2M refines the conventional model for predicting progression to symptomatic MM by allowing the identification of a very low risk group of patients who never progress and a high risk group of patients who are virtually all projected to progress. Disclosures No relevant conflicts of interest to declare.

Description: Introduction Patients with relapsed or refractory aggressive non Hodgkin B-cell lymphomas (NHL) not eligible to high dose chemotherapy and transplantation had a dismal prognosis. Lenalidomide showed activity in this setting as single agent or in combination. On this basis, we conducted a single center retrospective study to investigate efficacy and safety of lenalidomide alone or in association with rituximab or steroids in patients with heavily pretreated aggressive B-cell NHL. Methods Primary end points of the study were response rate (RR), defined as complete response (CR), partial response (PR) and stable disease (SD), and duration of response (DOR); secondary end points were feasibility and safety. Inclusion criteria for the analysis were: patients with relapsed/refractory aggressive B-cell NHL, aged 〉18 years, treated with lenalidomide between August 2007 to June 2012. Treatment scheme were: standard dose of oral lenalidomide 25 mg/day for 21 days every 28 days as single agent; standard dose of lenalidomide with the same schedule in association to weekly dexamethasone (20 mg bolus); lenalidomide 20 mg/day for 21 days every 28 days in combination  with rituximab (375 mg/sqm) every 28 days. Patients were treated until disease progression or unacceptable toxicities. Results A total of 53 patients were analyzed. Different histotypes of NHL were included in the study: 34 diffuse large B-cell lymphomas (DLBCL), 11 mantle cell, 5 follicular, 2 primitive mediastinal B-cell and one Burkitt lymphoma. At relapsed before lenalidomide treatment, the majority of the patients presented an advanced disease: 40 (75%) stage 3-4; intermediate high/high risk  international prognostic index (IPI) 23 (43%). Bone marrow was involved in 20 (38%) patients and 20 (38%) cases presented a bulky disease. Prior treatment lines were as follows: 8 (15%) patients received lenalidomide at first relapse while 24 (45%) underwent more than 3 previous lines of therapy; 14 (26%) patients received high dose chemotherapy and autologous stems cell transplant, one (2%) patient was allogenic transplanted and  4 (8%) did both before lenalidomide. All patients analyzed received lenalidomide: 31 patients (58%) underwent lenalidomide as single agent, 11 (21%) received a combination scheme of lenalidomide plus rituximab and 11 cases (21%) were treated with lenalidomide plus steroids. Median time from diagnosis and the beginning of lenalidomide was 25,3 months (3,7- 145,9), while median time from last previous therapy and lenalidomide treatment was 3,2 months (0,4- 38). At the time of this analysis response assessment was done in 51 patients: Response rate was 35% (18 patients), with CR 8 (15%), PR 5 (10%), SD 5 (10%). All patients who obtained CR underwent more than 3 courses of therapy, while among 31 patients who did not respond to treatment, 21 failures occurred during the first three cycles. Concerning different schemes of therapy: in the arm treated with lenalidomide as single agent RR was 24%, among patients underwent lenalidomide plus rituximab was 55% and in the group receiving lenalidomide plus steroid 45%. Among 34 DLBCL patients, RR was 41% (n= 14: CR 5, PR 4, SD 5), while in 11 patients affected by mantle cell lymphoma RR was 27% (n= 3: CR2, PR 1). Median DOR for all 18 responding patients was 12 months (0,2-24). At a median follow-up of 20 months, 5 patients were in stable CR, 7 continued lenalidomide, 11 relapsed and 28 died. Patients received a total of 257 cycles of lenalidomide, of which 25 were earlier interrupted and 48 were reduced in dose or duration; 50% of patients had at least one interruption in the planned treatment, however globally 91% of the expected dose was given. One patient died due to heart failure during the treatment. Toxicity was globally mild: most common grade 3-4 adverse events were neutropenia (19%), anemia (17%) and thrombocytopenia (17%). Five patients had grade 3-4 infections and 3 patients had thromboembolic  events (only one grade 3).  Two cases of neuropathies, both grade ≤ 2 were observed. Conclusions Lenalidomide single agent or in association with rituximab or steroids is effective and safe in patients with relapsed or refractory aggressive B-cell NHL, showing a promising response rate also in patients with heavily pretreated disease and with a mild toxicity profile. Disclosures: Vitolo: Roche: Speakers Bureau; Celgene: Speakers Bureau; Takeda: Speakers Bureau.

Description: SHP-1 and SOCS-1 are members of a family of negative regulators of signaling that act downstream of cytokine receptors, receptor tyrosine kinases and receptor complexes of the immune and hemopoietic system. SHP-1 and SOCS-1 act through the inhibition of the JAK2/STAT pathway and their inactivation is considered to be a relevant mechanism in JAK2/STAT pathway induction. Several observations point to induction of the JAK/STAT pathway as an important mechanism of lymphomagenesis. Since DNA methylation is a mechanism of gene silencing involved in the pathogenesis of human cancers, including lymphoma, here we tested the involvement of SHP-1 and SOCS-1 methylation in immunodeficiency-related lymphoma. Tumor samples from 34 HIV-related non-Hodgkin lymphoma (HIV-NHL) and 25 post-transplant lymphoproliferative disorders (PTLD) were analyzed for SOCS-1 and SHP-1 methylation by methylation-specific polymerase chain reaction. The tumor panel included 15 HIV-diffuse large B cell lymphoma (HIV-DLBCL), 10 HIV-Burkitt lymphoma (HIV-BL), 9 HIV-primary effusion lymphoma (HIV-PEL), 10 PTLD-immunoblastic (PTLD-IB), 7 PTLD-centroblastic (PTLD-CB), 4 polymorphic PTLD (P-PTLD), 3 PTLD-BL and 1 PTLD multiple myeloma (MM). Among HIV-NHL, SHP-1 methylation occurred in 23/34 (67%) cases, including 12/15 (80%) HIV-DLBCL, 4/10 (40%) HIV-BL, and 7/9 (78%) HIV-PEL. SOCS-1 methylation occurred in 6/34 (17%) HIV-NHL, including 5/9 (55%) HIV-PEL and 1/15 (7%) HIV-DLBCL. When considering EBV status, SHP-1 was methylated in 10/11 (91%) EBV-negative HIV-NHL and in 12/22 (54%) EBV-positive HIV-NHL. Among PTLD, SHP-1 methylation was detected in 20/25 (80%) cases, including 7/10 (70%) PTLD-IB, 6/7 (86%) PTLD-CB, 3/4 (75%) P-PTLD, 2/3 (66%) PTLD-BL and 1 PTLD MM. SOCS-1 methylation was detected in 3/25 (12%) PTLD, including 1/10 (10%) PTLD-IB and 2/7 (28%) PTLD-CB. When considering EBV status, SHP-1 was methylated in 12/13 (92%) EBV-negative PTLD and in 6/11 (54%) EBV-positive PTLD. The implications of our data are threefold. First, SHP-1 methylation is involved in the majority of HIV-NHL and PTLD. Second, it is remarkable that virtually all EBV-negative HIV-NHL and PTLD carry SHP-1 methylation. In EBV-positive B-cells, EBV infection activates the STAT pathway. It is conceivable that, in EBV-negative HIV-NHL and in EBV-negative PTLD, SHP-1 inactivation through aberrant methylation may surrogate EBV infection for STAT activation. Third, similar to observations in NHL of the immunocompetent host, SOCS-1 methylation is rarely implicated in the pathogenesis of immunodeficiency-related NHL. This notion is supported by the phenotype of SOCS1-deficient mice, that die of a myeloproliferative disease but do not develop a lymphoproliferative disease.

Description: Introduction : MYC, BCL2 and BCL6 overexpression, assessed by IHC, with the latter conferring a better prognosis, have been reported to be a prognostic factor in DLBCL, but data are not consistent and sometimes contradictory. The aim of the present study was to assess the prognostic impact of overexpression of MYC, BCL2, and BCL6 in a retrospective cohort of de-novo DLBCL, selected for an high proliferation index (MIB1 ≥70%), treated consecutively with R-CHOP regimen. Methods: Patients with de-novo DLBCL diagnosed between January 2010 and December 2013 were included into the study. Inclusion criteria were: high proliferation index MIB1 ≥ 70% and a full course of R-CHOP regimen. Paraffin-embedded tumor samples were collected and investigated using immunohistochemistry (IHC) for MYC, BCL2 and BCL6. Fluorescence in situ hybridization (FISH) is ongoing. MYC/BCL2+ or MYC/BCL6+ double expression cases were identified if they had rearrangements of MYC and BCL2 or BCL6. MYC immunochemistry was done on TMA sections using the antibody clone Y69. BCL2 and BCL6 staining had been evaluated previously at diagnosis. Tumor cells were defined positive for MYC and BCL2 or BCL6 protein expression by immunostaining if 〉40%, 〉40% and 〉25% of cells showed positive expression, respectively. Progression free survival curves (PFS) were estimated using the Kaplan-Meier method and compared between groups using the log-rank test and Cox models. Results : One hundred and sixty seven patients are evaluable for clinical characteristics and 69/167 had paraffin embedded tumor samples available for immunohistochemistry at the time of present analysis. Clinical characteristics of the 69 cases were: median age 66 years (IQR 57;73), 45 (65%) male, 47 (68%) stage III-IV, 35 (54%) with elevated LDH levels and 46 (67%) at International Prognostic Index (IPI) high intermediate or high risk. Overexpression of MYC was detected in 28 cases (41%), 50 (72%) and 38 (55%) showed BCL2 and BCL6 overexpression respectively. Nineteen (28%) cases showed MYC/BCL2+ and 17 (25%) MYC/BCL6+ double expression. With a median follow up of 26 months, the median 2-years PFS was 59%. Overexpression of MYC and BCL2 proteins and low expression of BCL6 were associated with an inferior 2-years PFS in univariate analysis: MYC- vs MYC+ 64% vs 55%; BCL2- vs BCL2+ 71% vs 56%; BCL6+ vs BCL6- 61% vs 54%. In a Cox multivariate regression model adjusted for IPI and age, MYC overexpression, BCL2 positivity and BCL6 negativity showed prognostic relevance as significant independent indicators with different risk (Hazard ratio 2.53 for MYC+, 2.08 for BCL2+ and 1.62 for BCL6-). Established that the three variable contributed with different risk in the multivariate analysis, an IHC sum additive score of 0-5 was calculated proportionally to the coefficient estimated (coefficient [Log hazard ratio] 0.92 for MYC+, 0.73 for BCL2+ and 0.48 for BCL6-), assigning an individual risk of 2 points for MYC or BCL2 positivity and 1 point for BCL6 negativity. Two years-PFS was significantly different between all separate groups (Hazard ratio for unit increase 1.57 95% CI 1.11-2.22, p=0.01). After pooling scores 0-1 (with or without BCL6), 2 (presence of MYC or BCL2 only), and 3-4-5 (MYC+/BCL6-, BCL2+/BCL6-, MYC+/BCL2+, MYC+/BCL2+/BCL6-) 2-yrs PFS rates were different across the three groups: 100% vs 64% vs 50% (log rank p= 0.04) (figure 1). Conclusion: Our data showed, with the limits of a small sample size, that MYC overexpression alone or with high expression of BCL2 and/or low expression of BCL6 correlates with a worse prognosis independently by IPI score in a cohort of DLBCL selected for high proliferation index and treated with R-CHOP. Assessment of MYC, BCL2 and BCL6 expression by IHC represents a rapid, inexpensive, and reproducible technique. These results need to be confirmed in our complete series of 167 patients (analysis ongoing) and validated prospectively in a larger cohort, using standardized staining and scoring methodologies. Thus, MYC and BCL2 represent relevant biomarkers that should be tested in future clinical trials using novel effective and targeted agents in order to improve the prognosis of DLBCL. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2660 Inconsistent information concerning the pattern of CNS relapse have been reported in MCL pts. We retrospectively analyzed the clinical variables at diagnosis and outcome, with special reference to CNS relapse, in a population of consecutive pts with confirmed diagnosis of MCL from two hematology centers. Among 2426 non Hodgkin's lymphoma pts treated from 1979 to 2011, 142 cases (44 female, 98 male) of MCL were selected. Median age at diagnosis was 68 years (17–94 years); 116 pts (82%) had stage III-IV, 89 (67% of the 132 cases in whom the data was available) had intermediate-high/high International Prognostic Index (IPI) risk. Extranodal disease was reported in 127 pts (89%), serum LDH was elevated in 45 pts (40% of 113 tested pts). Information concerning first line treatment was available in 139 pts. Fourteen pts (10%) did not receive active treatment at diagnosis, in 7 (5%) of these, systemic treatment followed the initial expectant strategy. Four pts (3%) only received radiotherapy and/or surgery. One hundred twenty two pts (88%) were treated with chemotherapy, 46 pts (33%) had rituximab, alone (6 pts, 4%) or in combination with chemotherapy. Eighteen pts (13%) received chemotherapy regimens including drugs crossing the blood-brain barrier or prophylactic intrathecal chemotherapy, 10 pts (7%) had autologous stem cell transplantation. After median follow-up of 8 years, CNS relapse was observed in 11 cases (7.7%; 95%CI:4–13%). CNS disease occurred at a median of 13.8 months (range:3.7–95 months) from diagnosis. Cumulative risk of CNS relapse raises until 10 years, being 7.2% (95%CI:4–14%) at 3 years, 10.6% (95%CI:6–19%) at 5 years and 13.6% (95%CI:7–25%) at 10 years. Elevated serum LDH at diagnosis was significantly associated with higher risk of CNS relapse at univariate analysis (P=0.006). Actuarial risk of CNS relapse was significantly shorter in pts with higher risk according to IPI (P=0.018). Median survival after CNS relapse was 6.3 months (range: 1.5–77 months). CNS relapse had a dismal impact on survival (P=0.04). No specific treatment approach at diagnosis, including autologous stem cell transplantation, intrathecal chemotherapy, high dose cytarabine or rituximab, alone or in combination with chemotherapy, significantly reduced the risk of CNS relapse. CNS relapse is one of the most challenging events in the management of MCL. Our data confirmed the adverse clinical outcome of MCL after CNS relapse. Better definition of clinico-pathological profile at diagnosis suggesting higher risk of CNS relapse could select pts candidate to prophylactic approach addressed to prevent CNS disease. Disclosures: No relevant conflicts of interest to declare.

Description: A fraction of Ph- CMPD is characterized by JAK2F617F mutation leading to constitutive JAK-STAT activation. The negative regulators of cytokine signaling SHP-1, SOCS-1 and SOCS-3 have a crucial function in the negative regulation of JAK2 activation/phosphorylation in response to EPO, G-CSF and a subset of cytokines. SHP-1, SOCS-1 and SOCS-3 may be silenced by aberrant DNA methylation in human malignancies. Here we tested chronic phase Ph- CMPD and acute myeloid leukaemia (AML) from Ph- CMPD for aberrant DNA methylation of SHP-1, SOCS-1 and SOCS-3. The study was based on: i) 85 Ph- CMPD, including 35 essential thrombocythemia (ET), 20 polycythemia vera (PV), 15 idiopathic myelofibrosis (IMF), 6 chronic myelomonocytic leukemia (CMML) and 9 Ph- chronic myeloid leukemia (Ph- CML); and on ii) 19 AML from Ph- CMPD, including 4 AML from PV, 10 AML from ET and 5 AML from IMF. Cases were analysed for SHP-1, SOCS-1 and SOCS-3 aberrant methylation by methylation-specific PCR and for JAK2V617F mutation by allele specific PCR. For comparison, 10 samples of normal bone marrow hematopoietic cells were also investigated. Among Ph- CMPD, methylation of SHP-1 occurred in 4/20 (20%) PV and 4/35 (11%) ET, while it was absent in IMF (0/15), CMML (0/6) and Ph- CML (0/9). Methylation of SOCS-1 occurred in 5/20 (25%) PV, 5/35 (14%) ET, 2/15 (13%) IMF and in 1/9 (11%) Ph- CML while it was absent in CMML (0/6). Methylation of SOCS-3 occurred in 11/20 (55%) PV, 13/35 (37%) ET, 4/15 (26%) IMF, 3/6 (50%) CMML and 3/9 (33%) Ph- CML. JAK2V617F mutation was detected in 47/85 (55%) Ph-CMPD, including 17/20 (85%) PV, 18/35 (51%) ET, 12/15 (80%) IMF, 0/6 CMML and 0/9 Ph- CML. SHP-1, SOCS-1 and SOCS-3 methylation was analysed according to JAK2 mutation status in PV, ET and IMF. In this group of patients, SHP-1, SOCS-1 and SOCS-3 methylation occurred in both JAK2 mutated cases (6/47, 13% for SHP-1; 10/47, 21% for SOCS-1 and 18/47, 38% for SOCS-3) and germline cases (2/38, 5% for SHP-1; 2/38, 5% for SOCS-1 and 10/38, 26% for SOCS-3). By combining the results of SHP-1, SOCS-1 and SOCS-3 methylation status, 21/47 (45%) JAK2 mutated cases carried SHP-1 and/or SOCS-1 and/or SOCS-3 methylation as opposed to 12/38 (31%) germline cases. This pattern of SHP-1 and SOCS-1 methylation was conserved also when the analysis was restricted to PV, ET and IMF each as a single group and after stratification for JAK2V617F mutation. Among AML from Ph- CMPD, methylation of SHP-1 occurred in 1/10 (10%) AML from ET, while it was absent in AML from PV and AML from IMF. Methylation of SOCS-1 occurred in 1/4 (25%) AML from PV and 1/10 (10%) AML from ET. One ET patient acquired SHP-1 methylation at transformation to AML. All normal bone marrow samples (n=10) scored negative for SHP-1, SOCS-1 and SOCS-3 methylation. The implications of these results are threefold. First, inactivation by aberrant methylation of SHP-1, SOCS-1 and SOCS-3 is involved in the pathogenesis of Ph- CMPD and is selectively associated with neoplastic hemopiesis. Second, among PV, ET and IMF, methylation of SHP-1, SOCS-1 and SOCS-3 may occur in both JAK2V617F positive and negative cases. Third, the low prevalence of SHP-1 and SOCS-1 methylation in AML from Ph- CMPD suggests that SHP-1 and SOCS-1 silencing is not involved in leukemic transformation.