ALBERT

Description: ASCT is considered the standard of care for patients up to 65 years of age with MM. The optimal induction treatment prior to ASCT is still unknown. VAD is a commonly used regimen but is associated with a number of toxicities, administrative inconvenience, and a complete remission (CR) rate of typically less than 10%. Recently the combination thalidomide/Dex was reported to induce a higher overall response rate than Dex alone in newly diagnosed patients, but with significantly more toxicities, especially deep vein thrombosis, and a 4% CR rate (Rajkumar et al, J Clin Oncol 2006;24:431–6). In a phase II trial we have evaluated the combination of Vel 1.3mg/m2 D1, 4, 8, 11 with Dex 40mg/D D1-4, D9-12, for 4 consecutive 21-D cycles (Dex only on D1-4 during cycles 3–4). This combination was well tolerated and yielded a 21% CR/nCR rate with successful stem cell harvest and engraftment (Harousseau et al, Haematologica 2006). In 2005, the IFM initiated a multicenter randomized phase III trial comparing Vel/Dex with VAD as induction treatment of patients with newly diagnosed MM up to the age of 65. The primary objective of the study is the CR (negative immunofixation) plus nCR (CR but positive immunofixation) rate after 4 cycles. A secondary objective is to evaluate the role of consolidation with 2 courses of DCEP (Dex, cyclophosphamide, etoposide, cisplatin). Patients were randomized at diagnosis among 4 arms (A1: 4 cycles of VAD; A2: 4 cycles of VAD + 2 cycles of DCEP; B1: 4 cycles of Vel/Dex; B2: 4 cycles of Vel/Dex + 2 cycles of DCEP); stratification was by cytogenetics and β2 microglobulin. Stem cell collection was performed between cycle 3 and cycle 4 after priming with G-CSF (10μg/kg/D × 6 D). At the end of induction (± DCEP consolidation) treatment, patients received melphalan 200mg/m2 plus ASCT. Enrollment of 480 patients is planned, to detect a 10% higher CR/nCR rate with Vel/Dex (20% vs 10%) with 80% power (two-sided test, significance level 0.05). An interim analysis is planned when 222 patients have been evaluated for response at the end of induction. As of July 2006, 304 patients have been enrolled. Randomization of the 222 patients to be considered for interim analysis was completed on May 19th. Baseline characteristics include: median age 55y; 140M, 82F; β2 microglobulin 〉3mg/L in 54%; del13 by FISH in 45%. The occurrence of serious adverse events (SAEs) was similar between the VAD and Vel/Dex arms (31% with VAD, 33% with Vel/Dex). The most frequent SAEs were pneumopathy (5.3% VAD, 4.0% Vel/Dex), thrombosis (2.7% VAD, 3.3% Vel/Dex), and peripheral neuropathy (1.3% VAD, 3.3% Vel/Dex). We intend to present results from the interim analysis.

Description: Nucleoside analogs (NA) are used alone or in combination in low grade B-cell lymphoproliferative disorders for 20 years. The incidence of disease transformation (Richter syndrome: RS) and development of MDS/AML among patients (pts) with indolent B-cell malignancies receiving NA are still controversial. We have reviewed the crude incidence of RS and MDS/AML in 165 WM patients treated with 3 F-based regimens. Results: Retrospective Study 1 (Leblond J Clin Oncol 1998): F (25mg/m2 iv D1–5) in 71 WM pts in relapse/refractory treated with alkylator, median time from diagnosis (dg) to tt: 5.9yrs, median prior tt: 2 (1–4). Randomized prospective study 2 (Leblond Blood 2001) in 92pts in first relapse after alkylator: Arm 1: F (25mg/m2 ivD1–D5) (45pts) vs Arm 2: CAP (doxorubicine 25mg/m2 IV D1, cyclophosphamide 750mg/m2 IV D1, Prednisone 40mg/m2 D1–5) (45pts), median time from dg to tt: 3.9yrs. Retrospective study N°3 (Tamburini Leukemia 2005) in first-line WM (14pts) or relapse WM (35pts): F 30mg/m2 D1–3 by oral route + Cyclophosphamide (C) 300mg/m2 D1–3 by oral route, median time from dg to tt: 2.1 yrs. The incidences of MDS/AML and RS are shown in Table 1 Discussion: the crude incidence of RS varied from 6,6% to 8% and was not statistically different in the randomized prospective study N°2 between the two groups. However, the incidence reported in pts treated with only alkylating-based regimens was lower: 3/167(1.8%) (Facon J Clin Oncol1993) and 3/217 (1.4%) (Garcia Sanz Br J Hematol 2001) and could be explained by a F use as salvage treatment in the CAP arm. The impact of fludarabine on RS needs to be confirmed in larger prospective studies. The crude incidence of MDS/AML varied from 1.4 to 8.9% in pts treated with fludarabine alone and was 6% in pts treated with F+C. The incidence reported in pts treated with alkylating based regimens varied from 1% (2/167, Facon 1993, 3/217 Garcia-Sanz 2001) to 6% (3/46, Kyle Br J Hematol 2000,) and was not different. In other low grade B-cell malignancies, a higher incidence of MDS/AML has been reported in pts treated with analog-alkylator combination compared to F or alkylator alone (Morrison J Clin Oncol 2002, Bowcock Br J Hematol 2006, Tam Hematologica 2006). In our study, the rate was the same in the two populations. The impact of late effects, particularly second malignancies, need to be better evaluated in prospective studies, especially in young patients and clinicians should discuss the risk when recommending such therapy to pts. MDS/AML and RS incidences in the three studies Study N°, pts Median survival time from study Median follow-up time from study MDS/AML (N,%) RS (N,%) N°1 71 pts 23 mo 34 mo 1/71 (1.4%) 5/71 (7%) N°2- 92 pts 41mo( arm1), 45 mo (arm2) 34mo 4/45 (8.9%, arm1), 2/45( 4.5%, arm2) 3/45 (6.6% arm1), 2/45 (4.5%, arm2) N°3-49pts 64% at 60 mo 42 mo 3/49 (6%) 4/49 (8%)

Description: The Intergroupe Francophone du Myélome (IFM) initiated 2 trials in 1999 to study patients with high-risk (β2-microglobulin level greater than 3 mg/L and chromosome 13 deletion at diagnosis) de novo multiple myeloma. In both protocols, the induction regimen consisted of vincristine, doxorubicin, and dexamethasone (VAD) followed by first autologous stem cell transplantation (ASCT) prepared by melphalan 200 mg/m2. Patients with an HLA-identical sibling donor were subsequently treated with dose-reduced allogeneic stem cell transplantation (IFM99-03 trial), and patients without an HLA-identical sibling donor were randomly assigned to undergo second ASCT prepared by melphalan 220 mg/m2 and 160 mg dexamethasone with or without anti–IL-6 monoclonal antibody (IFM99-04 protocol). Two hundred eighty-four patients—65 in the IFM99-03 trial and 219 in the IFM99-04 trial—were prospectively treated and received at least one course of VAD. On an intent-to-treat basis, overall survival (OS) and event-free survival (EFS) did not differ significantly in the studies (medians 35 and 25 months in the IFM99-03 trial vs 41 and 30 months in the IFM99-04 trial, respectively). With a median follow-up time of 24 months, the EFS of the 166 patients randomly assigned in the tandem ASCT protocol was similar to the EFS of the 46 patients who underwent the entire IFM99-03 program (median, 35 vs 31.7 months), with a trend for a better OS in patients treated with tandem ASCT (median, 47.2 vs 35 months; P = .07). In patients with high-risk de novo MM, the combination of ASCT followed by dose-reduced allogeneic transplantation was not superior to tandem dose–intensified, melphalan-based ASCT.

Description: Imatinib (IM) at 400 mg daily is the first line therapy for newly diagnosed CML patients (pts); however, less than 50% of major molecular responses (MMR) are obtained at 12 months. To improve these results, we designed a phase III, multicenter, open-label, prospective randomized trial. The reference arm was IM 400 mg daily (n=159). The 3 experimental arms were IM 600 mg daily (n=160), IM 400mg daily in combination with Ara-C, (20 mg/m2/day, days 15–28 of 28-day cycles)(n=158) and IM 400mg in combination with Peg-IFN alfa-2a (Peg-IFN2a, 90 μg weekly) (n=159). Treatment was delivered at least 12 months or until treatment failure (disease progression) or major toxicity. The primary endpoint is the overall survival. Other endpoints are: rate and duration of hematologic and cytogenetic responses, major (MCyR) and complete (CCyR), molecular response (major molecular response ie MMR) and the tolerability. Using treatment allocation ratio 1.1.1.1, randomization was stratified according to Sokal risk groups. The current interim analysis of the first 636 patients (α=0.85%, β=10%) at 1 year from randomization was planned in order to select the best experimental arm for further comparison with IM 400. The increased dose of IM or a combination regimen would be considered as promising if it increased the 4 log reduction response rate by at least 20 percentage points, e.g. from 15% to 35%, with an acceptable tolerability. Evaluation of molecular response up to 12 months was centralized, blinded and calculated according to International score (IS). Pts were recruited between 9/2003 and 10/2007.[median age 51 yrs (18–82), 62% of pts were male; Sokal distribution was low risk 33%, intermediate risk 41% and 27% high risk]. Median follow-up is 36 months (range 8–57) at the time of analysis. Overall, at 3 months 86 % of pts achieved complete hematologic response. The MCyR, CCyR and MMR rates at 6 and 12 months are: IM-400 IM-600 IM-Ara-c IM-PegIFN *p〈 10−2 (overall); ** p

Description: Most CML patients are sensitive to Imatinib Mesylate (IM), however, a small fraction develop resistance, mostly through the onset of BCR-ABL mutations. More than 30 mutations have been described, mostly in advanced phase, and confer different levels of clinical resistance. In this setting, clinical trials with 2nd generation tyrosine kinase inhibitors (TKI) provide encouraging results, however neither in vitro nor clinical activity has been demonstrated when the most frequent BCR-ABL mutation, T315I, has been identified. In this retrospective study from 5 French centers, we analysed the features and clinical outcomes of 27 CML patients treated with IM and presenting either clinical, cytological, cytogenetic resistance or molecular progression, and harboring a BCR-ABLT315I mutation detected by direct sequencing (same method in the 5 different laboratories, quality control exchanges). The 27 patients were in chronic phase (CP) at diagnosis, with 17 M and 10 F with a median age at diagnosis of 52 (25–70). Sokal scores were high for 8 pts, intermediate for 6 pts, low for 3 pts and unknown for 8, 2 were in blast crisis (BC) at CML diagnosis. Transcripts were M-BCR for 22 patients and m-BCR for 2 patients and unknown for 3 pts. At diagnosis 4 pts had additional chromosomal abnormalities as a variant Ph1 chromosome, a -7, a -Y, and an additional t(3;7;12) to the Ph1. Progression has been defined as a 2-fold rise in BCR-ABL transcripts levels, loss of any previous response to IM, and a transition towards a more advanced phase of the disease. At T315I discovery, 11 pts were in CP, 4 in accelerated phase, and a majority in BC [7 in myeloid and 5 in lymphoid]. The median interval between diagnosis and IM was 20 Mo. (0–145.2). The median initial dose of IM was 464 mg/day. Most of the patients were poor responders to IM of which 12 pts that obtained no more than a CHR, 2 pts a PCyR, 7 pts a CcyR, 2 pts a MMR, 1 pt no response, and 3 unknown. Twenty-one pts harboured a T315I alone, 1 a T315I with a Y253H, 2 with a M351T, 1 with a E255K+E255V, 1 with a L324Q, and 1 with a F311L, with no impact on survival. The median time for progression from day 1 IM was short [13 Mo. (0–49.6)] regardless of the phase of the disease at T315I identification, and the median interval IM start-T315I identification was 20 Mo. (0–57.6). Median overall survival from D1 of IM was 17.5 Mo. for advanced phases and 42.5 Mo. for CP (p=0.08). All patients progressed with no difference for time to progression between phases (p〉0.05). Figure Figure Neither additional mutations (p=0.91), nor additional chromosomal abnormalities (p=0.11), nor Dasatinib treatment (p=0.15) modified overall survival. In conclusion, the onset of BCR-ABLT315I mutations occurs preferentially in high-risk CML, seems more frequent in advanced phases, and is always responsible for progression and poor survival, underlying the need for alternative treatments, in patients lacking a histocompatible donor.

Description: Background: ATRA combined to anthracycline based chemotherapy (CT) is the reference treatment of newly diagnosed APL. ATRA has also demonstrated a role during maintenance treatment in randomized studies (Fenaux, Blood 99; Talmann, NEJM 1997), and may even be useful during consolidation courses (Sanz, Blood, 2008, although this was not a randomized study). During induction treatment, the optimal duration of ATRA treatment, however, is unknown. We tried to assess this point using long-term results of APL 93 and APL 2000 trials, conducted by the European APL group in newly diagnosed APL between 1993 and 2004. Methods: APL 93 and 2000 trials combined ATRA plus 3 courses of daunorubicin (DNR) based CT and included a total of 902 pts with different randomizations (for date of introduction of CT, use or not of AraC and of maintenance treatment). Only treatment arms that proved “optimal” (ie with early addition of the first course of CT to ATRA, AraC in combination to DNR for the 3 CT courses, and combined maintenance with intermittent ATRA and low dose continuous CT) (Fenaux, Leukemia 2000;Ades JCO 2006) and adult pts who had achieved CR in those arms, ie 414 pts, were considered for this analysis of the influence of duration of ATRA during induction treatment on the cumulative incidence of relapse (CIR) and survival. Per protocol, ATRA (45 mg/m2/d, rounded to a daily dose of 8 pills of 10 mg per day, ie 80 mg/day in most pts) was to be administered “until CR”, and to be transiently discontinued only in case of severe ATRA syndrome. No dose reduction was allowed except in children, who were therefore excluded from this analysis. Results: In the 414 pts, who Included 263 (64%) pts with WBC 〈 10G/L and 151(36%) pts with WBC 〉10G/L, the median cumulative dose of ATRA administered during Induction treatment was 2160 mg (corresponding to 27 days at 80 mg/d) Early ATRA Interruptions, Ie before recovery from aplasia and patient discharge, were made almost exclusively for ATRA syndrome or unexplained fever, and were followed or not by restart of the drug. Final discontinuation of ATRA for induction treatment was made a median of 27 days after onset of ATRA treatment. Although both trials stated that ATRA should be continued “until CR”, ATRA discontinuation was more often made at the end of the period of aplasia and patient discharge, than at the time of documented CR on bone marrow examination, especially as marrow abnormal promyelocytes are known to disappear slowly in APL. No difference in 5 y CIR or survival were found in pts who had received more or less than 2160 mg of ATRA during induction in the whole population and in pts with baseline WBC less or greater than 10 G/L. 89/414 (21%) pts had received less than 1500 mg of ATRA during Induction (corresponding to less than 19 days at 80 mg/day), their 5 y CIR and survival were similar to those of pts who had received more than 1500 mg of ATRA during induction in the whole population, and in pts with baseline WBC 〉 10 G/L. However, in pts with baseline WBC

Description: Imatinib (IM) at 400 mg daily has emerged as the preferred therapy for newly diagnosed CML pts. Despite impressive results, only a minority of pts treated with IM achieved a molecular remission. To improve upon these results, the CML French Group designed a phase III, multicentre, open-label, prospective randomized trial. The experimental arms are IM 400mg daily in combination with Peg-IFN-α2a (Peg-IFNα2a, 90 μg weekly) or IM 400mg daily in combination with Ara-C, (20 mg/m2/day, days 15–28 of 28-day cycles) or IM 600mg daily. The reference arm is IM 400mg daily. All pts (over 18 years of age with Bcr-Abl positive CML in chronic phase within 3 months of diagnosis) receive IM 400 mg/day as monotherapy days 1–14 and then start the assigned randomized regimen. Treatment continues at least 12 months or until treatment failure (disease progression by hematologic criteria) or major toxicity. The primary endpoint will be the overall survival. Other endpoints will be: rate and duration of hematologic and cytogenetic responses, major (MCyR) and complete (CCyR), molecular response and the tolerability. Using treatment allocation ratio 1.1.1.1, randomization is stratified according to Sokal risk groups. An interim analysis of the first 636 patients (α=0.85%, β=10%) at 1 year from randomization will allow evaluation of molecular response rates, one of the experimental arm being selected for further comparison with IM 400. The increased dose of IM or a combination regimen would be considered as promising if it increased the 4 log reduction response rate by at least 20 percentage points, e.g. from 15% to 35%, with an acceptable tolerability. Evaluation of molecular response up to 12 months is centralized and blinded. This evaluation is based on a cohort of 315 pts with a median time of observation of 12 months, recruited between 9/2003 and 6/2005.[median age 53 yrs (18–78), 60% of pts were male; Sokal risk distribution: 39% of pts low risk, 38% intermediate risk, and 23% high risk]. At 3 months 82% of pts achieved complete hematologic response. Cytogenetic data are available from 154 pts. At 6 months, 135 pts (87%) achieved a MCyR, being complete in 105 pts (68%). Grade 3/4 neutropenia and thrombocytopenia occurred in 5% and 0% of IM400 pts, in 5% and 1% of IM600 pts, in 30% and 4% of IM+IFN pts and in 24% and 13% of IM+Ara-c pts respectively. Dose of Peg IFN was reduced in 16% of pts, 45 μg per week being well tolerated. Grade 3/4 non hematological toxicity occurred in 6% of IM400 pts (mainly skin rash and muscle cramps), in 10% of IM600 pts, in 5% of IM+IFN pts (maily skin rash) and in 13% of IM+Ara-c pts (mainly diarrhea). Discontinuation of experimental treatment occurred in 15% of IM600 pts, 28% of IM+IFN pts and in 13% of IM+Ara-c pts. This first analysis of a randomized trial has proven feasibility of IM combinations in addition to high response rates. However a substantial hematological toxicity was recorded with IFN or Ara-c combination, which requires a careful assessment during the first 6 months of treatment. Long-term observation will demonstrate whether these promising results will have the potential to improve survival of CML pts.

Description: CML-BP has a poor prognosis and is refractory to most therapies. Dasatinib (SPRYCEL®), the most potent BCR-ABL inhibitor, is 325-fold more potent than imatinib in vitro against unmutated BCR-ABL. Dasatinib also inhibits important tyrosine kinases (eg, SRC family kinases) that may play a role in imatinib resistance and disease progression. Previous studies have demonstrated the efficacy and safety of dasatinib 70 mg BID for patients with imatinib-resistant or -intolerant CML-BP, leading to its approval in this patient group. CA180-035 is an open-label phase III study comparing dasatinib dosing at 70 mg BID vs 140 mg QD in patients with advanced CML (accelerated phase or BP) or Ph+ ALL who are resistant or intolerant to imatinib. The primary trial objective was to compare major hematologic response (MaHR) rates with QD or BID dosing. The main secondary objectives were to compare major cytogenetic response (MCyR) rates, time to and duration of response, progression-free survival (PFS), overall survival (OS), and safety profiles between the two schedules. Preliminary analyses from this study have demonstrated that QD treatment is associated with equivalent efficacy and less frequent key adverse events (AEs) compared with BID treatment. Here, 2-year results from 149 patients with myeloid BP (MBP) CML (75 QD and 74 BID) and 61 patients with lymphoid BP (LBP) CML (33 QD and 28 BID) are reported. Following dasatinib treatment, durable MaHRs and MCyRs were achieved by patients in both the MBP and LBP groups, and extended PFS and OS were observed in a proportion of patients (Table). Excluding patients that were BCR-ABL positive, Ph negative (n=10), rates of MCyR were comparable. Dasatinib was generally well tolerated, and with both schedules, cytopenias were the most commonly reported AEs. Grade 3/4 treatment-related nonhematologic AEs were infrequent with both schedules, with diarrhea, nausea, headache most commonly reported and each occurring in

Description: Abstract 3293 Poster Board III-1 Introduction: Nilotinib, a potent and highly selective BCR-ABL kinase inhibitor, is approved for the treatment of Ph+ CML patients (pts) in chronic phase (CML-CP) and accelerated phase (CML-AP) who are resistant or intolerant to prior therapy, including imatinib. The ENACT study, which is a Phase IIIb, open label, multicenter study, was initiated to obtain additional efficacy and safety information in pts with imatinib-resistant or -intolerant CML in CML-CP, CML-AP, or blast crisis (BC) in a clinical practice setting outside of a registration program. Methods: Pts received nilotinib 400 mg twice daily (BID). Dose escalation was not permitted. Cytogenetic and hematological responses were provided by investigator assessment and monitored according to the current European Leukemia Network (ELN) recommendations. Safety assessments included monitoring for all adverse events (AEs), serious AEs and cardiac procedures. A centralized approach to molecular testing in France facilitated matching of the PCR results to ENACT pts who were monitored for molecular response as a standard of care. BCR-ABL/ABL ratio (in %) were reported on the international scale (IS), French physicians followed the current ELN recommendations and monitored pts by PCR at baseline and every 3 months thereafter. Major molecular response (MMR) was defined as IS ≤ 0.1%. In addition, pts with ≤ 0.003% IS are reported. Confirmed loss of MMR was defined as IS 〉 0.1% on two consecutive PCR tests in pts previously in MMR Results: A total of 168 French CML-CP pts enrolled in the ENACT study between Jan. 2006 and Oct. 2008 with molecular response data available for 77% of pts. The median age of pts was 57 years, 56% were imatinib-resistant and 43% were imatinib-intolerant as determined by physician assessment. At study completion, 108 (64%) pts were continuing on nilotinib. The main reasons for treatment discontinuation were unsatisfactory therapeutic effect and AEs (16% for both). Median (range) duration of nilotinib exposure was 363 (4–765) days; median (range) average dose intensity was 793 (222–913) mg/day. Dose interruptions and reductions lasting longer than 5 days occurred in 36.9% and 11.9% of pts, respectively, with AEs being the main reason. The majority of nilotinib-related grade 3/4 AEs were hematologic, with thrombocytopenia (22.6%) and neutropenia (14.9%) being the most common. Non-hematologic AEs were mostly grade 1/2, not study drug-related and included fatigue, headache and rash. No incidence of grade 3/4 QT prolongation (QTcF 〉 500 msec) was observed and no pts experienced study drug-related grade 3/4 pancreatitis or pleural effusion on study. Overall major cytogenetic response (MCyR) and complete hematologic response (CHR) rates were both 54%, which are higher than those observed in the overall ENACT CML-CP cohort at 45% and 43%, respectively. Among the pts with molecular response data, 10 (7.8%) came to the study with MMR, one had a confirmed loss of MMR but subsequently regained MMR. The proportion of pts with post-baseline MMR (regardless of baseline response) was 42% (55 pts); 38% of pts achieved an MMR on study. The proportion of pts with a post-baseline ratio ≤ 0.003% IS was 16% (21 pts). The achieved molecular responses were durable with only 4 pts with confirmed loss of MMR, while another 3 pts met the criteria for loss of response but subsequently regained MMR. Of the 61 pts with CCyR and molecular data available, 72% also achieved MMR (see Table). Conclusions: This analysis of the French CML-CP subset of a large expanded access study further demonstrates that nilotinib is well tolerated and effective in heavily pretreated pts with CML-CP. A majority of pts (72%) with CCyR also had MMR. This data supports the use of nilotinib at 400 mg bid as the recommended dose in French pts. Disclosures: Nicolini: Bristol Myers Squibb: Consultancy, Honoraria, Research Funding, Speakers Bureau; Novartis Pharma: Consultancy, Honoraria, Research Funding, Speakers Bureau; Chemgenex: Honoraria, Speakers Bureau. Bordessoule:Novartis Pharmaceuticals: Honoraria. Belanger:Novartis Pharmaceuticals: Employment. Lamy:Novartis Pharmaceuticals: Honoraria. Gardembas-Pain:Novartis Pharmaceuticals: Honoraria. Maloisel:Novartis: Consultancy, Honoraria; Sanofi: Research Funding; Calgene: Research Funding. Rea:Novartis Pharmaceuticals: Honoraria. Johnson-Ansah:Novartis: Consultancy. Lenain:Novartis Pharmaceuticals: Honoraria. Szczudlo:Novartis: Employment. Wang:Novartis Pharmaceuticals: Employment. Duh:Novartis Pharmaceuticals: Research Funding.

Description: Abstract 878 Introduction: We investigated, in a prospective randomized study, the place of ASCT in the frontline treatment of CLL. Patients and methods: From March 2001 to December 2007, 241 patients were included in the trial. Eligibility criteria were previously untreated stage B and C CLL patients under 66 years, characterized by Matutes score 4-5, absence of cyclin D1 expression, baseline flow assessment of ZAP 70 and CD38 expression, karyotype and FISH analysis, IgHv mutational status (centralized). Preceding randomization, initial chemotherapy consisted of 3 monthly courses of mini CHOP regimen as previously described, followed by 3 monthly courses of fludarabine, IV or oral. Patients achieving CR (NCI 1996 criteria plus normal CT scan) were randomized between observation and ASCT. Non CR patients were offered cisplatin/cytosine-arabinoside/dexamethasone (DHAP) rescue and randomized whatever the response between ASCT or 3 subsequent monthly IV courses of Fludarabine-Cyclophosphamide (FC). Conditioning regimen for ASCT consisted of cyclophosphamide IV (60 mg/sqm d-5-4) and fractionated total body irradiation (10 Gy). The primary end-point of the study was event-free survival at 3 years. Responses after initial treatment (i.e.before randomization) and after completion of therapy, overall survival, side effects, prognostic significance of clinical and biological characteristics at baseline were other endpoints. Results: Baseline characteristics were: gender (M/F: 3), age (median 56.4 years, range 33.3-66), stage B (185 patients) or C (56 patients). All enrolled cases but five (236 patients) started the treatment. Among them, 206 completed the six planned courses of initial chemotherapy. For the 236 patients, CR rate was 43.6%, and overall response was 89.8%. Forty two patients were not randomized because of treatment failure (19) or patient/physician decisions (23). After an observed median follow-up of 40.2 months (Q1-Q3, 17.9-47.9) at the reference date (1/1/2009), the overall survival for the 241 patients was 87.8% (95% CI, 83.3-92.6) at 3 years and 75.4% (95% CI, 66.2-88.6) at 5 years. For the 199 randomized patients, the overall survival was 90.9% (95% CI, 86.7-95.3) at 3 years. Among them 105 patients were in CR after initial treament and were allocated to ASCT (53 patients) or observation (52 patients) with a 3 years EFS of respectively 78.7% (95% CI, 67.7-91.4) and 31.3% (95% CI, 20.1-48.8) (p