ALBERT

Description: Key Points Pomalidomide-cyclophosphamide-prednisone is an active combination in multiple myeloma patients who are relapsed/refractory to lenalidomide.

Description: Abstract 446 Background: The outcome of myeloma (MM) patients who are no longer responding to thalidomide, lenalidomide or bortezomib is poor, with a median event-free survival of 5 months and median overall survival (OS) of 9 months (Kumar SK et al, Leukemia 2012). The newer immunomodulatory drug pomalidomide, has shown significant activity in these clinical conditions. Aims: We assessed dosing, efficacy and safety of pomalidomide-cyclophosphamide-prednisone (PCP) in MM patients relapsed/refractory to lenalidomide. Methods: Pomalidomide was administered in doses ranging from 1 to 2.5 mg/day on days 1–28, cyclophosphamide at 50 mg every other day on days 1–28 and prednisone at 50 mg every other day on days 1–28 for 6 cycles, followed by maintenance therapy with pomalidomide-prednisone. Thromboprophylaxis with aspirin 100 mg/day or low-molecular weight heparin was recommended at physician's discretion. Results: The maximum tolerated dose (MTD) of pomalidomide was defined as 2.5 mg/day. Fifty-two patients were enrolled at the MTD and evaluated after completing at least 1 PCP cycle. Median age was 69 years (range 41–83). The median time from diagnosis to enrolment was 55 months (range 15–203). Best responses to PCP included 6% of complete response (CR), 19% of at least very good partial response (VGPR), 54% of at least partial response (PR) and 75% of at least minimal response (MR). Time to PR was rapid (median 1.8 months). After a median follow-up of 11 months (range 1–18), 1-year progression-free survival (PFS) and OS rates were 52% and 78%, respectively. PFS was not significantly different in patients with high-risk cytogenetic compared with patients with standard-risk disease and in patients younger or older than 75 years. Toxicities were primarily hematologic and included grade 4 neutropenia (13%) and thrombocytopenia (4%). At least grade 3 non-hematologic toxicities included infections (8%), rash (6%) and neurologic (6%). Thromboembolism occurred in 1 patient. Four patients discontinued treatment for toxicity (2 infections, 1 neurologic and 1 hepatic toxicity). Conclusions: PCP induced high response rates and prolonged PFS after prior exposure to lenalidomide and bortezomib (Table), without adding significant toxicity. PCP could be considered a valuable salvage option for pre-treated MM patients. Disclosures: Palumbo: Celgene: Consultancy, Honoraria. Larocca:Celgene: Honoraria. Sciacca:Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees. Guglielmelli:Celgene: Honoraria. Giuliani:Celgene: Research Funding. Boccadoro:Celgene: Consultancy, Honoraria.

Description: Introduction The standard treatment for elderly untreated diffuse large B-cell lymphoma (DLBCL) is RCHOP21, however up to 40% of patients experienced failures. Lenalidomide showed activity in heavily pretreated DLBCL and in vivo and in vitro data demonstrated a synergism with rituximab. In the phase I trial REAL07 (Chiappella et al, Haematol 2013), FIL demonstrated that the association of LRCHOP21 was feasible in elderly untreated DLBCL and identified 15 mg lenalidomide from day 1 to day 14 as the maximum tolerated dose in combination with RCHOP21. Patients and methods. The phase II trial REAL07 was designed based on Simon's two stage design to demonstrate an improvement of overall response rate (ORR) of 15% in LRCHOP21 compared to 70% of standard RCHOP21. Secondary endpoints were progression-free survival (PFS), overall survival (OS), event-free survival (EFS) and to correlate outcome with cell of origin (COO) profile. Response was evaluated according to 2007 Cheson criteria. Inclusion criteria were: age 60-80 FIT at the comprehensive geriatric assessment; untreated CD20+ DLBCL; Ann Arbor stage II/III/IV; international prognostic index (IPI) at low-intermediate/intermediate-high/high (LI/IH/H) risk. Treatment plan was: RCHOP21 plus 15 mg lenalidomide from day 1 to 14 for 6 courses. All cases were centrally reviewed by expert pathologist; COO profile analysis was conducted with immunohistochemistry according to Hans' algorithm and with gene expression profile (DASL assay). Results. From April 2010 to May 2011, 49 patients were enrolled. Clinical characteristics were: median age 69 years (range 61-80); stage III/IV 43 (88%), IPI IH/H 30 (61%). At the end of 6 LRCHOP21, ORR was 92%. Complete remissions (CR) were 42 (86%) and partial remission 3 (6%); 3 patients (6%) did not respond and one (2%) died for homicide. At a median follow-up of 28 months, 2-year OS was 92% (95% CI: 79-97), 2-year PFS was 80% (95% CI: 64-89) and 2-years EFS was 70% (95% CI: 55-81); 2-year PFS for IPI LI was 89% (95% CI: 62-97) and for IPI IH 76% (95% CI: 47-90) and for IPI H 72% (95% CI: 36-90). Hematological and extra-hematological toxicities were mild, with no grade IV extra-hematological events and no toxic deaths during treatment. Of the 294 planned courses of LRCHOP21, 277 (94%) were administered; median dose of lenalidomide delivered was 1185 mg (94% of the planned dose); at least 90% of the planned dose of each drug was administered in 91% of the RCHOP21 courses. Median interval time between RCHOP21 courses was 21 days (range 19-48). All 49 cases underwent central pathology review and diagnosis of DLBCL was confirmed. Regarding COO analysis, tissue block or stained slides were collected in 40/49 (82%), of which 32 were adequate for analysis. At the time of this abstract, COO analysis was reported according to immunohistochemistry data; DASL analysis is ongoing. Clinical characteristics between germinal center (GCB, 16 patients) and non-GCB (16 patients) were superimposable, excepted for a majority of H IPI risk in non-GCB group (p 0.067). ORR for GCB and non-GCB were 88% (CR 81%) and 88% (CR 88%), respectively. At a median follow-up of 28 months, 2-year PFS was 71% (95% CI: 40-88) in GCB-group and 2-years PFS was 81% (95% CI: 51-93) in non-GCB-group (Figure 1). Conclusions. In conclusion, LRCHOP21 is effective, also in poor risk patients, namely in non-GCB subgroup. These encouraging data warrant a future phase III randomized trial comparing LRCHOP21 vs. RCHOP21 in untreated non-GCB DLBCL. Disclosures: Off Label Use: lenalidomide in first line DLBCL is off lable. drug provided free by Celgene. Vitolo:Roche: Speakers Bureau; Celgene: Speakers Bureau; Takeda: Speakers Bureau.

Description: Introduction: Diffuse large B-cell lymphoma (DLBCL) of the testis (PTL) is a rare presentation of extranodal lymphoma at poor prognosis with a 5-yr overall survival of 40-55%. Contralateral testis, CNS and extranodal sites relapses are the main cause of failures. IELSG 10 study is a prospective phase II international trial for patients with stage I or II PTL and was designed to define a standard treatment for PTL. It aims to evaluate efficacy and toxicity of a combined treatment of R-CHOP, intrathecal methotrexate and prophylactic scrotal radiotherapy (RT) with the addition, in stage II, of loco-regional RT. The trial was conducted by IELSG and Intergruppo Italiano Linfomi. Patients and methods: From June 2001 to June 2006, 50 pts with stage I-II PTL were enrolled from 26 centres. Treatment plan was: R-CHOP21 (R 375mg/m2, Ctx750 mg/m2, Doxo50 mg/m2, Vcr1.4 mg/m2 day 1 and Pdn40 mg/m2 days 1–5) for 6 courses and 2 additional ones in stage II patients with slow response; intrathecal methotrexate (IT MTX) 15mg for 4 doses during courses 1 and 2; at the end of chemotherapy 30 Gy scrotal RT to contralateral testis was planned to all pts and 30–36 Gy nodal loco-regional RT for those with stage II disease. Results: To date 45 pts who completed the treatment are evaluable. Median age was 64 (range 22–80); 36 stage I and 9 stage II disease; 3 had bilateral testicular involvement; 4 LDH〉normal and 2 B symptoms. All received R-chemotherapy as planned. Forty-two (93%) pts received adequate CNS prophylaxis (at least 4 IT MTX); 3 less than 4 IT MTX because of poor tolerance or toxicity. Scrotal RT was given to 40 pts (89%) as planned; 5 did not performe it (3 refusal, 1 progressive disease and 1 bilateral orchiectomy). Forty-four patients (98%) achieved a CR and one progressed after 4 R-CHOP courses. With a median follow-up of 28 months, 3-yr OS, 3-yr PFS and 3-yr EFS were: 88% (95% CI 69–96%), 82% (95% CI 63–92%) and 78% (95% CI 58–89%). Seven patients relapsed or progressed: 2 in nodal sites, 3 in extranodal ± nodal sites and 2 in CNS (1 isolated meningeal and 1 eningeal + nodal relapse). The actuarial risk of CNS relapse at 3 years is 2.5% (95% CI 0–7%). No contralateral testis relapses were observed. Five patients died: three because of DLBCL; one of peripheral T-cell lymphoma and one of ANLL 21 months off therapy while in CR. No toxic death occurred during treatment. Main grade 3–4 toxicities were: hematological 27% and neurological 15%. Conclusions: Although the follow-up is short, these results compare favorably with those previously reported in IELSG retrospective study in PTL (Zucca et al J Clin Oncol 2003). Contralateral testis relapses has not been observed and the incidence of CNS relapse seems to be reduced. If these results are confirmed with a longer follow-up, a combined treatment of R-CHOP, IT MTX and prophylactic scrotal RT ± nodal RT in stage II patients would improve the prognosis of these pts and should be regarded as the standard treatment in localized stage PTL. However, if further relapses are observed there will be a need for innovative strategies to address the issues of systemic and CNS relapse.

Description: Abstract 2871 Introduction. The addition of Rituximab to standard CHOP chemotherapy has improved outcome of DLBCL; however, 30–40% of patients do not response or relapse after induction treatment. The use of Lenalidomide, as single-agent in relapsed or refractory aggressive NHL has been encouraging, with an overall response rate (ORR) of 19% for DLBCL. Preclinical evidence suggests that Lenalidomide may be synergistic with Rituximab. However the safety and efficacy of Lenalidomide combined with R-CHOP is not yet fully evaluated. Aims. On these basis, the IIL is running a prospective multicenter dose finding phase I-II pilot trial to evaluate efficacy and safety of Lenalidomide treatment plus R-CHOP21 (LR-CHOP21) for elderly patients with untreated DLBCL (registered at http://www.clinicaltrials.gov, NCT00907348). The primary endpoint for the phase I part of the study was: definition of Dose Limiting Toxicity (DLT) considered as the maximum dose inducing any grade ≥3 non-hematologic toxicity, or a 〉15 days delay of planned cycle date. The primary endpoint for the phase II part of the study was: complete response (CR) and ORR. Herein we report the results of phase I. Patients and Methods. Inclusion Criteria were: age 60–80; histologically proven CD20+ DLBCL; no prior chemotherapy and no prior malignancies in the last 3 years; Ann Arbor stage II, III, IV; International Prognostic Index (IPI) at low-intermediate, intermediate-high or high risk score. Treatment consisted of 6 courses of R-CHOP21 in association with Lenalidomide for days 1–14 at the established dose level (5, 10, 15, 20 mg), with Peg-Filgrastim support and prophylactic sc. low-molecular weight heparin. Phase I of the study was planned to define the Maximum Tolerated Dose (MTD) that is the dose that achieves a DLT in 33% or less patients evaluated after the first three courses of LR-CHOP21. The study was designed with the Continual Reassessment Method, a Bayesian “memory design” that begins with a subjective prediction of the dose-response relationship and uses, as dose allocation rule of the sequentially incoming patients, the re-estimated probability of toxicity based on the results obtained for the patients already observed. Four doses of Lenalidomide were tested: 5, 10, 15 and 20 mg. By decision of the steering committee of the study, a dose of 10 mg was administered to the first cohort of 3 patients; in the light of the observed toxicity the probability model assigned the next cohort to the dose with the higher probability to met the required MTD. At the end of each cohort, the dose level associated with an updated DLT probability closest to 33% was recommended to be administered to the next patient cohort. Results. In phase I, from May 2008 to February 2010, 21 patients were enrolled. Median age was 69 years (61-78), stage II/III/IV were 4/4/13 respectively, B symptoms in 11, PS 2 in 8, bone marrow involvement in 6, abnormal LDH in 8, intermediate-high or high IPI score in 15. Patient allocation by Lenalidomide dose was: none patients received 5 mg/day, nine patients received 10 mg/day, nine patients 15 mg/day and three patients 20 mg/day on days 1–14 of each LR-CHOP21 courses. The flow of the dose allocation and the observed DLTs are shown in figure 1. DLTs in the first three courses were recorded in seven patients (see figure 1 for details). These data, according to the continual reassessment method, determined Lenalidomide 15 mg as MTD in association to R-CHOP21. One-hundred-fifteen courses of LR-CHOP21 were performed in the whole series of 21 patients. Overall hematological toxicity was moderate: grade III or IV thrombocytopenia occurred in 10%, anemia in 4% and neutropenia in 28% of the courses. Extra-hematological toxicities were mild: only one patient with grade IV (increase of CPK), two patients with grade III cardiac, three with grade III neurological toxicities, and three patients with grade III infections (two pneumonias and one febrile neutropenia with diarrhea). At the end of six LR-CHOP21 courses, 15/21 patients achieved CR, 1/21 partial remission and 5/21 were not responsive. Conclusions. Lenalidomide 15 mg on days 1–14 in association with R-CHOP21 is the MTD for LR-CHOP. This schedule is safe and well tolerated in a population of elderly DLBCL patients. Preliminary efficacy results are promising. The ongoing phase II part of the LR-CHOP21 trial aims at evaluating the efficacy of 15 mg of Lenalidomide in association with R-CHOP21. Disclosure: Vitolo: Roche Italy: advisory committee; Celgene Italy: advisory committee; Janssen-Cilag: lecture-fee. Off Label Use: The use of Lenalidomide is off-label in untreated DLBCL.

Description: Abstract 1898 Background: Multiple myeloma (MM) cells growth is sustained by several stimuli derived from surrounding cells of bone marrow (BM) microenvironment. Besides the increase in growth factors production, a constitutive activation of several pathways determining protection from apoptosis and growth advantage have been reported too. Aberrant activation of Met/HGF (Hepatocyte Growth Factor) pathway has been described in several tumors causing cell proliferation, cell migration and neoplastic angiogenesis. A qualitative analysis demonstrated that Met transcription is present in MM plasma cells and increase plasmatic levels of HGF has been related to a bad prognosis subgroup of patients. However a comprehensive investigation of quantitative analysis on Met expression, plasmatic HGF values and correlation with clinical outcome on a large amount of MM patients treated with novel agents is still missing. Aim: : to investigate the role of Met mRNA expression and HGF levels in a large panel of myeloma patients treated with novel drugs. Patients and Methods: one hundred and five samples of purified plasma cells derived from newly diagnosed myeloma patients have been included in this study. Fifty two patients received 9 courses of Velcade-Melphalan-Prednisone (VMP) as part of the VMP-VMPT trial (Palumbo A, 2009 ASH Meeting, abs 128) while fifty three patients have been included in the PAD-MEL100-LP-L trial (Palumbo A, JCO 2010). Met and HGF mRNA quantitative expression have been investigated on both purified plasma cells (CD138+ bone marrow fraction) and on bone marrow CD138 negative fraction. mRNA expression has been evaluated using a quantitative Real-Time PCR (qRT-PCR) on Abi Prism 7900 (Applied Biosystems) with a relative quantification based on ΔΔCt approach. JUM2 cell line was used as calibrator and Gus as housekeeping gene. HGF serum level has been evaluated on 76 of those patients too. ELISA assay has been employed to determine HGF serum value. On 51 samples with higher levels of mRNA Met expression, a FISH analysis reaching for Met amplification has been performed. Purified plasma cells were treated with a dual-color FISH assay using a MET/CEP7 probe cocktail (Cytocell, Cambridge, UK). Results: Met mRNA expression was higher in CD138+ cells than in CD138- fractions (median 76,90 range 0,81-916,51 vs 11,03 range 0–243,88 respectively; p=0,0001). Similarly HGF mRNA expression was higher in CD138+ cells than in CD138- population (median 2,07 range 0–65,34 vs 0,49 range 0,11-3,22 respectively; p=0,03). Patients with high and low Met levels were divided using the median value of Met expression as cutoff. At a median follow up of 30 months, patients with low Met mRNA expression displayed a higher progression free survival (PFS) and overall survival (OS) compared with those with high Met levels (PFS 73% vs 42% respectively, p

Description: High-dose (200 mg/m2, MEL200) and intermediate-dose melphalan (100 mg/m2, MEL100) showed significant activity in myeloma. In a phase 3 study, 298 patients were randomly assigned to receive 2 autologous transplantations after conditioning with MEL200 or MEL100. Ninety-six of 149 (64%) completed MEL200 and 103 of 149 (69%) MEL100. Best response to MEL200 was: complete remission 22 of 149 (15%); partial remission 95 of 149 (64%), for an overall response rate of 79%. Best response to MEL100 was: complete remission 12 of 149 (8%); partial remission 95 of 149 (64%), for an overall response rate of 72%. Overall survival did not differ (P = .13); median progression-free survival (31.4 vs 26.2 months, P = .01), median time to progression (34.4 vs 27.0 months, P = .014) were longer in the MEL200. Treatment-related mortality was 3.1% in the MEL200 and 2.9% in the MEL100 group. Severe neutropenia and infections were marginally superior, whereas severe thrombocytopenia, mucositis, gastrointestinal adverse events, and the overall occurrence of at least 1 nonhematologic grade 3 or 4 adverse event were significantly higher in the MEL200 cohort. We conclude that MEL200 leads to longer remission duration and should be considered the standard conditioning regimen for autologous transplantation. This study was registered at www.clinicaltrials.gov as #NCT00950768.

Description: Abstract 1784 Introduction. The outcome of MCL is unfavourable, with continuous relapses. The MIPI, a clinical score, defined performance status, age, LDH and leucocyte counts as predictors of MCL outcome. Ki-67 as cell proliferation index was evaluated in biological-MIPI (MIPI-b). Aim of the study was to tested MIPI on a retrospective group of MCL patients treated with Rituximab-chemotherapy with or without High Dose Chemotherapy and autologous stem cell transplantation (R-HDC); secondary endpoints were: to evaluate the feasibility of MIPI-b on a retrospective population and to quantify the predictive discrimination of IPI, MIPI, MIPI-b on the outcome of MCL in the Rituximab era. Methods. Between 1999 and 2009, 206 MCL 〉18 years at diagnosis consecutively treated in seven Italian institutions were included into the study. Histology was centrally reviewed and, if possible, Ki-67 evaluation was performed. Overall survival (OS) and failure-free survival (FFS) curves were estimated both overall and stratified by MIPI, MIPI-b and IPI score. Differences between curves were tested using the 2-tailed log-rank test. In order to quantify the predictive discrimination of MIPI, MIPI-b and IPI scores, a Cox's model analysis and univariate logistic models (with death and failure event as binary outcomes) were fitted and the area under the receiver operating characteristic (ROC) curves (c-index) was estimated in a subgroup of 120 patients that fulfilled MIPI, MIPI-b and IPI scores. Results. Clinical characteristics were: median age 61 (34-85) years, 78% stage IV, 73% with bone marrow involvement, 16% with blastoid variant; median leucocyte counts at diagnosis was 7.53×103 (2.38-175). First-line treatments were: R-HDC in 51%, Rituximab-Fludarabine based chemotherapy in 12%, Rituximab-CHOP in 32% and other Rituximab containing regimens in 5%. Ki-67 evaluation was performed in 135 patients; median Ki-67 value was 30%. Patients at high-risk (HR) were 29% according to MIPI, 18% according to MIPI-b and 33% to IPI. With a median follow-up of 48 months, 4-year OS was 72% (95% CI:65-78) and 4-year FFS was 49% (95%CI: 41–56). Four-year OS according to MIPI by risk groups was: LR 94% (95%CI: 85–97), IR 66% (95%CI: 47–80), HR 41% (95%CI: 26–55) and according to IPI: LR 87% (95%CI: 73–94), IR 88% (95%CI: 74–95), HR 41% (95%CI: 26–55) (Figure 1). In the subgroup of 120 patients that fulfilled MIPI, MIPI-b and IP scores an univariate logistic model and a Cox's model analysis were fitted. The c-index and Cox-index for death event were 76% and 75% for MIPI, 69% and 69% for MIPIb, 76% and 71% for IPI respectively; the c-index and Cox-index for failure event were 61% and 68% for MIPI, 59% and 64% for MIPIb, 68% and 66% for IPI respectively. A sub-analysis was conducted to validate the role of MIPI in R-HDC group; an univariate logistic model and a Cox's model analysis were fitted and the c-index and Cox-index for death event and for failure event were calculated (Table 1). Conclusions. MIPI score was confirmed as a good predictor of death event in MCL retrospective patients treated with Rituximab-chemotherapy regimens. MIPI score should be a good predictor of death event also in patients treated with R-HDC. The impact of MIPI-b score should be tested in prospective trials, with central histology review. New therapeutic strategies are warranted to improve the outcome of MCL namely in MIPI-HR group. Disclosures: No relevant conflicts of interest to declare.

Description: The development of nonmyeloablative conditionings has recently reduced the transplant-related mortality (TRM) and extended the eligible age for transplantation up to 65–70 years. From January 2000 to June 2005, 106 newly diagnosed patients younger than 65 years were enrolled in a prospective phase II study at 15 Italian Centers. Fifty-eight were also previously described in a comparison of autografting with allografting based on a genetic randomisation (Bruno et al. N Engl J Med 2007). Here we report on a larger GITMO experience with a longer follow-up. Induction chemotherapy consisted of VAD-based regimens, followed by a cytoreductive autograft with melphalan 200 mg/m2, and by a non-myeloablative 2 Gy TBI-based allograft from an HLA-identical sibling. Graft-vs-host disease (GVHD) prophylaxis included cyclosporin and mycophenolate mofetil. Primary endpoints were overall (OS) and event-free (EFS) survivals. Secondary endpoint was TRM. One-hundred-two (96%) patients, median age 54 (30–65), completed the tandem program whereas 4 withdrew their consent. After a median follow-up of 54 (21–94) months, OS was not reached and median EFS was 35 (31–56) months post-transplant. Incidences of acute grade II-IV GHVD and extensive chronic GVHD were 40% and 50% respectively. Fourteen (13%) patients died from TRM, 14 (13%) from disease progression, 2 from lung cancer (2%) and 1 from lymphoma (1%). Overall response, defined as complete (CR) and partial remission, was 91% (93/102), with 53 patients achieving CR. Overall 39/102 patients relapsed, however only 8/53 of those who reached CR post-transplant. By multivariate-analysis disease response prior to allografting was significantly associated with longer OS (HR 0.27, CI 0.09–0.80, p

Description: Abstract 2928 Poster Board II-904 Introduction: A new prognostic clinical index (MIPI) and a biological one with cell proliferation (Ki-67) evaluation (MIPIb), were defined specifically for MCL to give a more reliable estimation of outcome (Hoster 2008). Aim of our analysis was to test MIPI and MIPIb on a retrospective series of MCL patients treated with R-chemotherapy. Patients and methods: Between 1999 and 2008, 136 MCL at diagnosis consecutively treated in five institutions entered into the study. Histology was centrally reviewed. Clinical characteristics were: median age 62 (37-84) years, 78% stage IV, 73% with bone marrow involvement and 15% with blastoid variant. First-line treatments were: R-high-dose chemotherapy with Autologous Stem Cell Transplantation (R-HDC) in 48 patients, R-Fludarabine based chemotherapy in 22, R-CHOP-like in 50 and other R containing regimens in 16. Ki-67 evaluation was performed in 93 patients; 43 were not, due to inadequate pathological materials. Overall Survival (OS) and failure-free survival (FFS) curves were estimated both overall and stratified by MIPI, MIPIb and IPI score. Differences between curves were tested using the 2-tailed log-rank test. In order to quantify the predictive discrimination of MIPI, MIPIb and IPI scores, in a subgroup of 84 patients fulfilled MIPI, MIPIb and IPI scores, a Cox's model analysis and univariate logistic models (with death and failure event as binary outcomes) were fitted and the area under the receiver operating characteristic (ROC) curves (c-index) was estimated. Results: Prognostic index stratification was as follows: according to MIPI 45 patients (33%) were at low-risk (LR, 0-3), 36 (26%) at intermediate-risk (IR, 4-5), 43 (32%) at high-risk (HR, 〉5) and 12 missing; according to MIPIb 70 patients (51%) were at LR (0-5.699), 7 (5%) at IR (5.7-6.499), 16 (12%) at HR (〉6.5) and 43 missing; according to IPI 38 patients (28%) were at LR, 41 (30%) at LIR, 47 (35%) at IH-HR and 10 missing. Responses were as follows: complete 74, partial 29, no response 22, not yet evaluable 11. With a median follow-up of 28 months, 2-year OS was 80% (95% CI:71%-86%) and 2-year FFS was 60% (95%CI: 51%-69%). 2-year OS and 2-year FFS rates according to MIPI, MIPIb and IPI were shown in table 1. Eighty-four patients had all the factors to accurately calculate MIPI, MIPIb and IPI; in this subgroup, an univariate logistic model and a Cox's model including the time at the event were performed. The c-index and Cox-index for death event were 73% and 77% for MIPI, 72% and 73% for MIPIb, 67% and 65% for IPI respectively; the c-index and Cox-index for failure event were 66% and 72% for MIPI, 66% and 65% for MIPIb, 67% and 64% for IPI respectively. A further analysis for death event was performed to adjust the effect of MIPI for other known risk factors (Ann-Arbor stage, Bone Marrow involvement, blastoid variant, number of extranodal sites). In a Cox model, MIPI score and number of extranodal sites were confirmed as independent predictors of death event: adjusted hazard ratio was 8.75 (95%CI: 3.14-24.4, p=