Search

Background: ETP-ALL was included as a provisional identity in the 2016 WHO classification of ALL. This subtype was first identified by Coustan-Smith et al. in 2009. However, this immunophenotype-based classification does not fully enclose all ETP-ALL cases identified by gene expression profile (GEP). Although early studies in small series of ETP-ALL suggested a very poor outcome for ETP ALL more recent and larger series have showed improvement in outcome treating children with a contemporary protocol based on chemotherapy schedules, or after allogeneic stem cell transplantation (allo-SCT) in adults. Aim: To investigate the clinical and biological features of ETP-ALL cases in the Spanish cohort of adult T-cell ALL (T-ALL) patients and to assess the potential impact of high-risk therapy on their outcome. Method: One hundred eighty-five adults with T-ALL treated according to two consecutive MRD-oriented high-risk adult PETHEMA protocols -ALL-HR-2003 (NCT00853008) and ALL-HR-11 (NCT01540812; still ongoing)- were included in this study. The EGIL criteria were used to define the immunologic subtype of T-ALL after central review of immunophenotype reports, and the criteria proposed by Zuurbier et al. (Zuurbier L. et al. Haematologica 2014; 99:94-102) were used to define ETP-ALL. These later criteria consist of a combination of immunophenotypic markers (absence of CD1a-/CD4-/CD8-, cut-off 25%), that resemble those published by Coustan-Smith, with the advantage that include most ETP-ALL cases, as identified by GEP, avoiding the use of partial expression of CD5 marker to immuno-classify these patients. Results: Thirty-four out of 167 evaluable patients (20%) with T-ALL showed an ETP-ALL immunophenotype. Patients with ETP-ALL were older (mean 39 [SD 12] vs. 33 [12] yr; p=0.011), showed more frequently lymph node involvement (79% vs. 56%; p=0.021) and lower WBC counts at diagnosis (mean, 72 [155] vs. 97 [112] x109/L; p=0.004). At genetic level, ETP-ALL patients were associated with the absence of deletions in CDKN2A/B gene cluster (91% vs. 26%; p10% BM blasts on day+14) vs. 37% of non-ETP (p Conclusions: ETP-ALL accounts for 20% of adult T-ALL in the PETHEMA cohort and it is associated with a poorer initial response to treatment (lower CR rate, poorer MRD clearance) than the remaining T-ALL patients. Such poorer outcome is not overcome by allo-SCT in our series. Supported by grants from: Asociación Española Contra el Cáncer, AECC (GC16173697BIGA), Instituto Carlos III (PI14/01971 FI), 2017-SGR288 (GRC), CERCA Program from Generalitat de Catalunya, and "La Caixa" Foundation. Figure 1. Figure 1. Disclosures Montesinos: Daiichi Sankyo: Consultancy, Speakers Bureau; Novartis: Research Funding, Speakers Bureau.

Background & Objective: Acute Lymphoblastic Leukemia (ALL) is an aggressive neoplasia characterized by a high genetic heterogeneity both at diagnosis and at relapse. Due to the high incidence of relapse in adults and the dismal prognosis beyond recurrence, diagnosis and relapse samples of adult ALL patients were carefully analyzed in order to identify genetic alterations related with drug resistance and disease progression. Patients & Methods: Paired diagnosis-relapse bone marrow samples from 5 adult B-cell precursor ALL (B-ALL) patients were analyzed (Ph+ ALL [n=2], normal karyotype [n=1], t(1;19)(q23;p13) [n=1] and t(8;13)(p21-22;q12) [n=1]). Copy Number Alterations (CNA) were studied with Multiplex Ligation-dependent Probe Amplification (MLPA, kits P-335 and P-202 from MRC-Holland, Amsterdam, Netherlands) and Affymetrix CytoScan HD arrays (Affymetrix, Santa Clara, USA). In the array analyses, only the CNA that encompassed at least 25 markers were considered significant. Results: Regarding karyotype, 2 patients (1 Ph+ and 1 t(1;19) at diagnosis) showed the same chromosomal translocations within a complex karyotype at relapse. On the contrary, the other Ph+ patient showed a normal karyotype at relapse, while 2 patients did not experience any karyotypic change. Regarding immunophenotype, 3/5 patients showed changes on antigen expression from diagnosis to relapse such as expression of markers of immaturity (CD34, TdT positivity and CD38 negativity), loss of lymphoid markers (CD20 and CD22) and/or acquisition of myeloid markers (CD33 and CD66c). Concerning CNA, all relapse samples were genetically related to the diagnosis clone (common clonal origin). All relapsed populations lost CNA detected at diagnosis and/or acquired new CNA but retained some of the CNA showed at diagnosis revealing clonal evolution from ancestral clones. CNA in B-ALL key genes involved in lymphoid development (IKZF1, PAX5, EBF1,VPREB1 and BLNK), proliferation (CDKN2A/B, RB1, CRLF2, C-MYC and ERG), apoptosis (BTG1, TP53 and ATM), hematopoiesis transcription factors (ETV6 and MLL) and histone modifications (KDM6A) were detected, among others. Losses in 9p were the most recurrent event both at diagnosis and at relapse. CDKN2A/B deletions were observed in all relapse samples (3/5 in homozygosis) while PAX5 deletions were present in 4/5 relapsed cases. Interestingly, all relapse samples showed CNA favoring the activation and/or the transcription of proteins involved in the Akt/C-MYC signaling pathway. Another common feature (4/5 patients) were CNA affecting genes involved in drug transport such as several ABC transporter genes and genes related to drug resistance such as PRKDC and RUNX1T1 (in 3/4 of the cases, the CNA appeared exclusively at relapse or were already present at diagnosis and increased their frequency at relapse). CNA in genes that may confer stem cell characteristics (EGR1 and USP16) were another recurrent event at relapse (3/5 samples, 2 of them were not present at diagnosis). CNA affecting the X/Y PAR1 region (CRLF2, CSF2RA and IL3RA) or VPREB1 at 22q11.22 were detected in 3/5 relapse samples, respectively. An important apoptosis cluster at 11q21q24.2 (BIRC2/3, CASP1/4/5/12, hsa-miR-34b/c, ATM and BTG4) was lost in 2/5 relapse samples (one of them was not detected at diagnosis and the other increased its frequency at relapse). Finally, ETV6 deletion (12p13.2) and duplication of Xq26.2q28 (containing ABCD1, BCAP31 and genes coding for several cancer/testis antigens) were observed in 2 relapse samples. Conclusions: SNP arrays analysis of paired B-cell precursor ALL samples at diagnosis and at relapse allows the identification of genetic alterations potentially related with ALL progression. The systematic analysis of relapse samples could contribute to the identification of specific genetic targets with potential therapeutic impact for each patient (personalized medicine). Disclosures Martínez-López: Novartis: Honoraria, Speakers Bureau. Sole:Celgene: Membership on an entity's Board of Directors or advisory committees.

Background. About 25-35% of adult patients with acute lymphoblastic leukemia (ALL) show the Philadelphia (Ph) chromosome. Their outcome has improved with the combination of tyrosine kinase inhibitors (TKI) and chemotherapy, generally followed by allogeneic hematopoietic stem cell transplantation (alloHSCT). However, few series have evaluated the prognostic value of the additional cytogenetic alterations (ACA) to the Ph chromosome, with contradictory results. The aim of this study was to analyse the frequency, type and prognostic significance of ACA to the Ph chromosome in young adults with Ph+ ALL treated with the ALL Ph-08 trial from the PETHEMA group. Patients and methods.Between 2008 and 2016, 95 patients with Ph+ ALL from 30 Spanish hospitals were included in the ALL Ph08 trial. This trial includes concurrent administration of imatinib (600 mg/d) and standard induction and consolidation chemotherapy, followed by alloHSCT (or autologous HSCT if alloHSCT not feasible) and maintenance chemotherapy with imatinib in cases of MRD persistence of reappearance after alloHSCT (Ribera et al, Br J Haematol 2012; 159: 78-81). The cytogenetic reports were centrally reviewed. Major molecular response (MMR) was defined as BCR-ABL/ABL ratio ²0.1% and complete molecular response (CMR) as BCR-ABL/ABL ratio ²0.01% or undetectable. Results. In 74 out of 95 patients the karyotype was evaluable after review (in the remaining 21 cases the diagnosis was carried out by FISH [n=9] or PCR [n=12]). The median age of the 74 patients was 39 years (limits 19-63) and 44 patients (59%) were males. The median WBC count was 17.6x109/L (limits 0.2-390) and 8 out of 66 evaluable patients had CNS involvement at diagnosis. The CR rate was achieved in 72/74 patients (97%) (1 patient did not achieve CR and 1 died during induction therapy). To date, 57 patients underwent to HSCT, of them 46 were in MMR, at least, at the time of HSCT. With a median follow-up of 1.89 years (limits 0.10-7.38) the probabilities of the CR duration, OS and EFS at 4 years were 69% (95%CI: 54%-84%), 39% (23%-55%) and 38% (23%-53%), respectively. Out of 74 patients, 52 (70%) showed at least one ACA and in the remaining 22 (30%) the t(9;22) was the only cytogenetic alteration. No clinical or biologic differences were observed on comparison of the two groups of patients. Twenty (27%) out of 73 evaluable patients showed trisomies and 19 (26%) monosomies. No differences in CR attainment were observed according the presence and type of ACA. The 4-yr. probability of CR duration in patients with t(9;22) and one or more monosomies (monosomal karyotype) vs. those without monosomies was 23% (95%CI: 0%;49%) vs. 88% (77%;99%) (p Conclusions. In young patients with Ph+ ALL treated within the ALL Ph08 trial, the frequency of ACA was high, being trisomies and monosomies shown in similar frequency. The monosomal karyotype was the most important unfavourable prognostic factor in this series of patients. Funded in part by grants PI10/01417 (FIS), RD12-0036-0029 from RTICC, Instituto Carlos III and RD14-SGR225(GRE), Generalitat de Catalunya. Figure Figure. Disclosures No relevant conflicts of interest to declare.

Background and objective. Asparaginase (ASP) is an essential drug for ALL treatment. Two types of E.coli ASP (native and pegylated) are used in clinical trials for treatment of children and adults with ALL, but to our knowledge direct comparison between these two types of ASP in the same protocol has not been performed. This study aimed to compare the efficacy and safety of native vs. PEG-ASP in adult patients with HR, Ph-negative ALL. Patients and methods. HR ALL included one or more of the following parameters at baseline: age 30-60 yr., WBC count >30x109/L for B-cell precursor ALL or >100x109/L for thymic T-ALL, pro-B, early or mature T-ALL, 11q23 or MLL rearrangements or complex karyotype. Induction therapy consisted of vincristine, prednisone, daunorubicin and ASP (native 10,000 IU/m2, i.v., days 16-20 and 23-27 or PEG 2,000 IU/m2, iv, day 15 depending on center decision) for 4 weeks (Induction-1). FLAG-Ida was administered as intensified induction (Induction-2) in patients not achieving CR or in those in CR with MRD≥0.1% at the end of induction, and those patients proceeded to allogeneic HSCT if CR was attained. Patients in CR and MRD Results. Ninety-onepatients received native ASP and 35 PEG-ASP in Induction-1. The two groups of patients were comparable for the main clinical and hematologic ALL parameters. No differences were observed in the CR rate (86% vs. 86%), in the frequency of MRD level Conclusions. In HR, Ph-negative adult ALL patients included in the PETHEMA ALL-HR 11 protocol the type of E.coli ASP (native vs. PEG) did not have impact on response and outcome. Allergic reactions were more frequently seen with native ASP and coagulation abnormalities and hepatic toxicity were most frequent with PEG-ASP. Most of the differences in toxicity can be explained by the schedule of ASP given in consolidation (single dose of native ASP vs. single dose of PEG-ASP in each cycle). Supported in part by grants RD12/0036/0029 (RTICC, FEDER), PI14/01971 FIS, Instituto Carlos III, and SGR225 (GRE), Spain Disclosures No relevant conflicts of interest to declare.

Risk associated to FLT3 internal tandem duplication (FLT3-ITD) in patients with acute myeloid leukemia (AML) may depend on mutational burden and its interaction with other mutations. We analyzed the effect of FLT3-ITD/FLT3 wild-type (FLT3wt) ratio depending on NPM1 mutation (NPM1mut) in 303 patients with intermediate-risk cytogenetics AML treated with intensive chemotherapy. Among NPM1mut patients, FLT3wt and low ratio (= 0.5) patients had a worse outcome. In NPM1wt AML, FLT3-ITD subgroups showed a comparable outcome, with higher risk of relapse and shortened overall survival than FLT3wt patients. Allogeneic stem cell transplantation in CR1 was associated with a reduced relapse risk in all molecular subgroups with the exception of NPM1mut AML with absent or low ratio FLT3-ITD. In conclusion, effect of FLT3 burden is modulated by NPM1 mutation, especially in patients with a low ratio. (Blood. 2013;121(14):2734-2738)

Background Cytogenetics at diagnosis is the most important prognostic factor in acute myeloid leukemia (AML). Of note, intermediate cytogenetic risk group (IR-AML) is a very heterogeneous subset including normal karyotypes and all the cytogenetic abnormalities not included in the favorable or the adverse groups. Molecular alterations affecting NPM1, FLT3 and CEBPA show a prognostic impact in IR-AML. MLL partial tandem duplications (MLL-PTD) have also been described in this group of AML, but their prognostic impact have not been well established. Aim To analyze the prognostic relevance of MLL-PTD in the subset of patients diagnosed with IR-AML since 2003, and included in the CETLAM protocols LMA-2003 and LMA-2012. Methods Between 2003 and 2004 MLL-PTD were analyzed by Southern Blot (enzymes employed BglII, EcoRI, BamHI). Since 2004, a long PCR strategy was used to identify this abnormality. Results NPM1 mutations (NPM1mut), FLT3 internal tandem duplications (FLT3-ITD) and MLL-PTD were available for 893 patients. No MLL-PTD was found among 111 and 161 patients of the good and poor cytogenetic risk groups, respectively. The IR-AML group included 621 patients, and 37 carried a MLL-PTD (6%), thus only this cytogenetic group of patients was analyzed. NPM1mut were found in a 41% of patients and none of them had a concomitant MLL-PTD (p Conclusions MLL-PTD is a genetic alteration found in a 6% of IR-AML. Patients with this abnormality have a worse LFS and OS than the rest of patients of the IR-AML group. Based on these results, patients with MLL-PTD should be considered as patients with poor cytogenetic risk AML for treatment allocation. Disclosures No relevant conflicts of interest to declare.

Introduction: Recent studies have shown that young to middle-aged adults who receive a pediatric-inspired chemotherapy regimen for treatment of Ph-neg ALL do not appear to require an alloHSCT if they achieve good response on MRD testing after induction therapy. Patients (pts) who are not good MRD responders achieve better outcomes with alloHSCT than their counterparts who do not receive alloHSCT. However, it is not clear if this approach can be translated to adult ALL pts with HR features at baseline. The aim of the prospective ALL-HR-11 trial from the Spanish PETHEMA Group was to evaluate the response to a differentiated post-induction therapy (chemotherapy or alloHSCT) according to MRD levels (assessed by 8-color, centrally-performed flow cytometry at the end of induction-week 5- and consolidation therapy-week 17-) in HR Ph-neg adult ALL patients. Patients and methods: HR ALL included one or more of the following parameters at baseline: age 30-60 yr, WBC count >30x109/L for B-cell precursor ALL or >100x109/L for thymic T-ALL, pro-B, early or mature T-ALL, 11q23 or MLL rearrangements or complex karyotype. Induction therapy included vincristine, prednisone, daunorubicin and asparaginase (E coli native or PEG according to center availability) for 4 weeks (Induction-1). FLAG-Ida was administered as intensified induction (Induction-2) in pts not achieving CR or those in CR with MRD≥0.1% at the end of induction. For pts in CR and MRD Disclosures No relevant conflicts of interest to declare.

INTRODUCTION: The myelodysplastic syndromes (MDS) are a group of hematologic disorders characterized by ineffective hematopoiesis and increased risk of transformation to acute myeloid leukemia (AML). Aberrant DNA methylation is the dominant and most well-studied epigenetic alteration in MDS. Various genes, including cell cycle regulators, apoptotic genes, and DNA repair genes, are epigenetically silenced and play a role in pathogenesis and transformation to leukemia. The clinical response of MDS and AML to drugs that revert the aberrant hypermethylation, such as 5-aza-2x-deoxicitidine and 5-azacitidine (AZA), suggests that this hypermethylation could have an important role in the disease and it is not a secondary effect to other mechanisms. The aim of this study was to define the methylation pattern of DNA at diagnosis in patients with high-risk MDS and secondary AML to determine if there are some pattern predictive for response to AZA treatment or relapse. MATERIAL AND METHODS: Genomic DNA was obtained from bone marrow (n=100): 39 patients at diagnosis prior AZA treatment (24 MDS, 10 AML and 5 CMML), 10 control samples (peripheral blood of donors of bone marrow stimulated with hematologic growth factors) and 51 samples at different moments of follow-up (3 months, 6 months, 12 months and/or 18 months). Genome-wide DNA methylation profiling was performed using the Illumina Infinium 450K methylation array. We have used the IWG-2006 criteria for classifying the responders versus non responders. RESULTS: We have focused on MDS patients treated with AZA comparing methylation profiles at diagnosis and at the time of first detected response to treatment (mostly at 3 months). We performed different experiments of DNA methylation assays using RnBeads tool. The global methylation dendogram (Fig.1) clearly demonstrate that the subgroup of responders (7 patients) show strong demethylation after the treatment, while in non-responders methylation profiles has not changed that dramatically even after 6 months from the beginning of the treatment. We also have found several hundreds differentially methylated regions (DMR) at gene promoters, gene bodies and other genomic locations, which are significantly hypomethylated after the treatment in responders, while remain almost unchanged after the treatment in non responders. CONCLUSIONS: Illumina Infinium 450K methylation array serves as an informative methodology to study the methylation changes in bone marrow samples. A more detailed investigation of obtained DMR might shed light on the mechanisms of the response to AZA treatment and as result to allow detection of methylation markers of response to treatment or relapse. Acknowledgments: Instituto de Salud Carlos III, Ministerio de Sanidad y Consumo, Spain (PI 11/02519); 2014 SGR225 (GRE) Generalitat de Catalunya; Fundació Josep Carreras, Obra Social "La Caixa" and Celgene Spain. Diana Domínguez for her excellence technical assistance. Figure 1. Bi-clustering of the methylation levels of gene promoters in responders (green: diagnosis and orange: response) Figure 1. Bi-clustering of the methylation levels of gene promoters in responders (green: diagnosis and orange: response) Disclosures Valcárcel: Celgene Corporation: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees; GlaxoSmithKline: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau.

Current prognostic factors for acute myeloblastic leukemia (AML) are not sufficient to accurately predict the group of patients in the intermediate-risk category who will successfully respond to treatment. Distinct patterns of inherited functional genomic polymorphisms might explain part of these heterogeneous prognoses. We used the allelic discrimination method to identify polymorphisms in GSTT1, SULT1C2, CDA, SXR (drug metabolic pathways), XPD, XPA, XPG, ERCC1, TOP2A (DNA repair), VEGF (angiogenesis), and MDR1 (multidrug resistance) genes in 110 adult patients with intermediate-risk AML, enrolled in the CETLAM-99 prospective trial. A multivariate prognostic model adjusted for age, white blood cell (WBC) count, French-American-British group, cytogenetics, MLL rearrangement, internal tandem duplication of FLT3 (FLT3-ITD), induction courses to achieve complete remission, and germline polymorphisms, was used to detect independent risk factors associated with clinical outcome. This analysis showed an increased risk of refractoriness to chemotherapy in the group of patients with XPA variant alleles (RR = 14; P = .02). In the same model, increased relapse risk was associated with SULT1C2 heterozygosity (RR = 4.1; P = .004), FLT3-ITD (RR 3.3; P = .003), and MDR1 variant alleles (RR = 2.4; P = .02). Adverse prognostic variables for overall survival were XPA (RR = 3.4; P = .02) and MDR1 (RR = 2.1; P = .02) variant alleles, and WBC count (RR = 2.1; P = .02). These findings might be useful in selecting risk-adapted treatment strategies in intermediate-risk AML.

Introduction Recurrent Copy Number Alterations (CNA) in genes potentially involved in the pathogenesis of ALL have been identified in genes involved in B-cell development, cell cycle regulation, proliferation, apoptosis and drug resistance. Their independent prognostic significance in adult ALL patients is controversial. The aim of this study was to analyze the prognostic significance of CNA in a series of 96 high-risk, Ph-negative, B-precursor adult ALL patients treated according to risk-adapted protocols from the Spanish PETHEMA Group. Methods MLPA assays (MRC-Holland) were performed for the following genes: IKZF1, IKZF2, IKZF3, EBF1, CDKN2A/B, PAX5, ETV6, BTG1, RB1, hsa-miR-31, X/Y PAR1 region genes (CRLF2, CSF2RA, IL3RA) and 14q32.33 region genes (IGH D, MTA1, KIAA0284) . Fragment analysis was made by Genescan in an ABI-3130 sequencer (Applied Biosystems). Data normalization provided a value indicative of the presence or absence of CNA: 0-0.20 homozygous deletion, 0.21-0.70 heterozygous deletion, 0.71-1.30 normal, 1.31-1.70 heterozygous duplication and 1.71-2.20 homozygous duplication. Univariable and multivariable analyses including the most relevant clinical parameters (age, WBC count, phenotype, cytogenetics, CNS involvement) were performed for CR attainment, CR duration and OS. Results The median age [range] of the 96 patients was 39 [15-72] years, 50 (52%) patients were males, with a median WBC count 14.3 x109/L [0.4-388]. Phenotype: early pre-B 19 (20%), common 51 (54%), pre-B 22 (24%), unknown 2 (2%). Cytogenetics: normal 18 (19%), hyperdiploid 5 (5%), hypodiploid 2 (2%), near haploid 6 (6%), t(1;19) 7 (8%), 11q23/ MLL 11 (12%), complex 1 (1%), other 27 (29%), no growth 17 (18%). The most frequent CNA deletions involved CDKN2A /B (43/96, 45%), PAX5 (34/94, 36%), IKZF1 (32/95, 34%), hsa-miR-31 (25/96, 26%), 14q32.33 region (18/96, 19%), RB1 (17/96, 18%), EBF1 (12/91, 13%) and X/Y PAR (10/96, 10%). The most frequent duplications involved X/Y PAR (11/96, 12%) and 14q32.33 region (7/96, 7%). The CR rate was 83% (80/96), the median (95%CI) of CR duration was 2.7 years (0-5.9) and the median (95%CI) of OS was 2.1 (1.0-3.2), being the median (range) follow-up of the series of 3.8 (0.6-8.0) years. Table 1 shows the results of univariable and multivariable analyses. By multivariable analyses advanced age and EBF1 deletions were significantly associated with less CR rate, WBC count and X/Y PAR duplication were associated with shorter CR duration, and advanced age and CDKN2A /B deletion were associated with shorter OS. Conclusions The CNA of EBF1 , X/Y PAR1 genes and CDKN2A/B have independent prognostic significance in adult patients with high-risk, Ph-negative, B-precursor ALL. This study suggests that these genetic studies should be added to the initial work-up of these patients for more accurate prognostic assessment Supported by grants PI10/01417, RD12-0036-0029 from Instituto Carlos III, 2014 SGR225 (GRE) from Generalitat de Catalunya and a grant from the Spanish Society of Hematology and Hemotherapy (2012). | Variable | CR rate | CR duration | OS | | ---------- | -------- | ------------------ | -------- | ------------------- | | | P (univ) | OR (95%CI) | P (univ) | HR (95%CI) | P (univ) | HR (95%CI) | | Age | 0.011 | 0.93 (0.89 - 0.98) | NS | - | 0.005 | 1.03 (1.01 - 1.05) | | WBC | NS | -

Introduction The improvements in genetic characterization of AML allowed the ELN to propose a prognostic classification into 4 categories: Favorable, Intermediate I, Intermediate II and Adverse. Several groups have tested the outcome of AML patients based on these genetic subgroups in registries and databases from prospective trials. Since ALLOSCT recommendations are being established based on the ELN risk categories, we considered of interest to investigate whether these also have an impact on the results of allogeneic transplants. This could be helpful to identify the patients that may benefit the most from ALLOHCT in the first complete remission (CR1). Objectives To investigate whether the ELN genetic groups of AML have any impact on the results of ALLOHCT performed in CR1. To identify the patients who are best candidates for transplantation, based on the ELN categories as well as in other characteristics. Methods Patients were transplanted between 1990 and 2012. In all cases, treatment prior to transplant had consisted of anthracycline based induction and intermediate-dose cytarabine consolidation according to the CETLAM AML-88, AML-94, AML-99 and AML-03 trials. Upper age limit for transplantation was 60 years until 2002 and 70 years thereafter. Most patients above 50 years received either CD34 selected grafts or reduced intensity conditioning. GVHD prophylaxis after transplant was usually based on cyclosporine and prednisone or methotrexate. The main characteristics of patients and transplants were included in the exploratory analysis of variables influencing survival; those with a p- value up to 0.1 were incorporated as covariates to the multivariable model. Results One hundred ninety two patients in CR1 received an ALLOHCT from HLA-identical siblings (n=140), adult unrelated donors (n=26) or umbilical cord blood (n= 26). Age distribution of the patients was as follows: 16-35 years-old (y-o) n=65 (34%), 36-50 y-o n= 55 (29%), 51-60 y-o n= 54 (28%), >60 y-o n=18 (9%). Twenty-three patients (12%) were classified as in the favorable ELN category, 54 (28%) in the intermediate I, 41 (21%) in the intermediate II and 74 (39%) in the adverse. One-hundred fifty-seven patients (82%) had achieved CR1 after a single course of chemotherapy. Conditioning was myeloablative (MA) in 139 (72%) patients and reduced-intensity (RIC) in 53 (28%). In 76 cases of the MA regimens total body irradiation was included. Thirty-eight patients (20%) died due to transplant complications other than relapse and 47 (24%) experienced a leukemia recurrence. Overall survival (OS) and leukemia-free survivals at 8-years were 55±4% and 53±4%, respectively. Multivariate analysis of factors influencing OS included as covariates age at diagnosis of AML, ELN categories, courses to CR1 and conditioning regimen (MA versus RIC), since they had a p-value 60, p=0.01) and courses to CR (1 vs more, p=0.03). Conclusions The ELN genetic categories have impact on OS of AML patients who receive an ALLOHCT in CR1. The results were very good in the favorable category and, most important, in intermediate II patients. Even patients in the adverse category had a substantial probability of long-term survival. This is remarkable, since the outcome of patients with adverse genetic features when treated with CT only is very poor. Disclosures: No relevant conflicts of interest to declare.

Abstract 4785 Background The karyotype is an important predictor of outcome in adults with acute lymphoblastic leukemia (ALL). Some groups have reported the negative prognostic value of complex karyotype (CK, defined as ≥5 unrelated chromosomal abnormalities) in adult ALL (Moorman et al, Blood. 2007:109;3189-97). On the other hand, monosomal karyotype (MK, defined as ≥2 distinct autosomal chromosome monosomies or 1 single monosomy in the presence of structural abnormalities) has been associated with a worse outcome in patients with acute myeloid leukemia. We aimed to assess the prognostic value of cytogenetic abnormalities, especially CK and MK, in adults with ALL treated with protocols of the Spanish PETHEMA Group. Patients and Methods The karyotypes of 783 adult ALL patients from 63 Spanish centers treated according to the protocols of the PETHEMA Group between 1993 and 2011 were reviewed. The several PETHEMA protocols were risk-adapted (standard-risk –SR–, high-risk –HR–) or subtype-oriented (Philadelphia chromosome [Ph+] ALL -with or without imatinib-, and Burkitt's ALL [BL]). The impact of the main cytogenetic abnormalities as well as of the CK and MK on complete remission (CR) rate, CR duration, overall survival (OS) and event-free survival (EFS) was analyzed. Results The median age of the series was 33 years (range 15–82) and 448 patients (57.2%) were male. The karyotypes of 560 out of 783 patients were evaluable after review: normal karyotype 153 patients, t(9;22) 120, t(v;11q23) 30, t(8;14), t(8;22) or t(2;8) 47, high hyperdiploidy (>50 chromosomes) 53, low hyperdiploidy (47–50 chromosomes) 52, hypodiploidy (45–39 chromosomes) 32 and extreme hypodiploidy ( Conclusions Our study confirms that cytogenetics is a very important tool for risk assessment in adult ALL. Patients with t(9;22) and t(v;11q23) had the worst prognosis and the t(1;19) did not have prognostic significance. The introduction of imatinib in patients with t(9:22) ALL significantly improved their outcome. The CK and the MK were not associated with a worse prognosis in patients treated with risk-adapted or subtype-oriented protocols of the PETHEMA group. Disclosures: No relevant conflicts of interest to declare.

Abstract 1452 Introduction: Acute myeloid leukemia (AML) is a heterogeneous group of neoplastic disorders characterized by the presence of acquired genetic alterations in hematopoietic progenitor and stem cells. The prognostic impact of immunophenotypic markers has long been controversial. Despite this, only a very limited number of studies have investigated the prognostic impact of the levels of cell surface membrane (Sm) proteins commonly expressed on AML blasts. In this study we analyzed the prognostic impact of the levels of expression of several markers associated with AML blasts and normal hematopoietic precursors, as asessed by Multiparameter Flow Cytometry (MFC) in de novo AML patients (pts). Methods: Overall, 122 adult de novo AML pts diagnosed between 12/2003 and 08/2011 were studied, after excluding acute promyelocytic leukemia. All cases were diagnosed and classified according to the WHO criteria and they were immunophenotyped using an extensive 4-color panel of monoclonal antibodies by MFC, as previously described. Patients were treated according to the multicenter CETLAM AML-03 trial. Blast cells were gated after excluding all other cell populations, as well-delineated clusters of SSC low/int/CD45dimevents (“blast gate”) using the merge function of the Infinicyt software (Cytognos SL, Salamanca, Spain). For each case, the reactivity for the following markers was evaluated in terms of mean fluorescence intensity (MFI; arbitrary units) in bone marrow samples: CD123, CD117, CD34 and CD7. Normal residual bone marrow lymphocytes were used as reference for internal quality control purposes. Cytogenetic and molecular risk stratifications were based on the European LeukemiaNet (ELN) recommendations and the molecular lesions were investigated using well established protocols. Overall survival (OS), disease-free survival (DFS) and relapse rate (RR) were measured by the Kaplan–Meier method and curves were compared with the log-rank test. Multivariate analysis was performed using the Cox regression model. P-values ≤0.05 were considered to be statistically significant. Results: Median AML patient age was of 54 years (range: 20–70 y) with a M/F ratio of 63/59. Twelve pts (10%) had good cytogenetic/molecular risk, 83 (68%) intermediate, 17 (14%) high risk and 10 (8%) pts were unknown. Fifteen cases (12%) had NPM1mut, 33 (27%) showed FLT3-ITDmut, 5 (4%) pts carried CEBPAmut, 41 (33.6%) pts with no mutations and 28 (23%) pts were unknown. The median follow-up of pts alive was 19 months (range 0–97 months) and the 5-year OS of all pts was 42±5%. Univariate analysis of prognostic factors showed an association between higher CD45 (MFI >232; p=.01), CD117 (MFI>259; p=.008), CD34 (MFI >242; p=.007) and CD7 (MFI> 19; p=.03) expression and both a shorter OS and DFS. In addition, high CD123 expression (MFI >248; p=.04) was associated with a shorter DFS. In multivariate analysis only a high CD45 (p=.03) and a high CD34 (p=.03) expression were identified as independent predictors for a shorter OS. In turn, higher levels of CD123 (p=.03) and of CD34 (p=.02) were independent predictors for DFS and relapse, even when age, gender and WBC were taken in account. Conclusion: In this study, we show that higher levels of expression of immunophenotypic markers associate with immature myeloid precursors (CD45, CD117, CD34 and CD7) as assessed by MFC, have a significantly adverse prognostic impact in AML, both as regards OS and DFS. Accordingly, higher levels of CD45 and CD34 were independent predictors for both a shorter OS whereas, greater CD123 and CD34 values were associated with an increased risk of relapse and a shorter DFS, supporting the adverse prognostic impact of a more immature blast cell phenotype in de novo AML. Disclosures: No relevant conflicts of interest to declare.

Abstract 1712 The study of chromosome alterations helps to classify acute myeloid leukemia (AML) into three prognostic groups, favorable, intermediate and adverse. The prognostic value of favorable risk and adverse risk abnormalities is well defined; in contrast, the outcome of intermediate-risk group is heterogeneous. Molecular characterization of patients with intermediate-risk AML has identified subcategories with diverse prognosis. All this knowledge has translated into a recent ELN proposal for the genetic classification of AML. Additionally, the intermediate and particularly the cytogenetically adverse groups have been refined by the HOVON group by introducing the concept of “monosomal karyotype” (MK), consisting of at least two autosomal monosomies or one monosomy plus a structural abnormality. Objective: To validate the recent ELN classification in a large series of AML patients and to investigate if the inclusion of MK improved the prognostic stratification. Patients and methods: 804 consecutive patients treated under the CETLAM AML-99 (n=353) and CETLAM-03 (n=451) trials were analyzed. The two protocols included idarubicin, intermediate-dose cytarabine and etoposide as induction and mitoxantrone and intermediate-dose cytarabine as consolidation. G-CSF priming was given in the CETLAM-03 trial. Following, risk-adapted treatment with chemotherapy or hematopoietic transplantation was administered. Parameters analyzed were relapse rate (REL), leukemia-free survival (LFS) and overall survival (OS). Results: Median age of the series was 46 years (range 16–60). Median follow-up of patients alive was 15 months. The ELN classification resulted in different prognostic risk groups. At 5 years, ELN favorable risk category had an OS of 60±4%, intermediate-I of 32±%, intermediate-II of 46±% and adverse of 17±3% (p Conclusion: In the present study, the ELN genetic classification did not discriminate the prognosis of patients in the intermediate-I and intermediate-II categories. In contrast, genetic grouping that considered CBF AML, favorable mutations in the intermediate cytogenetics category and that separated adverse karyotype with or without MK translated into significantly different OS. This classification, together with the study of mutations recently described and the expression of certain genes may contribute to a more precise prognostic stratification and risk-adapted therapy. Disclosures: No relevant conflicts of interest to declare.

Abstract 322 Introduction: Current therapeutic protocols for adult ALL consider MRD together with the baseline risk factors (age, WBC count, immunophenotype, cytogenetics) and speed in response to therapy for treatment decisions. On the other hand, the systematic use of allogeneic SCT for all adult patients (pts) with Ph- HR-ALL is still a matter of debate. The aim of the prospective study ALL-AR-03 from the Spanish PETHEMA Group was to evaluate the response to a differentiated therapy (chemotherapy or allogeneic SCT) according to early bone marrow blast clearance ( Patients and methods: HR ALL included one or more of the following baseline parameters: age 30–60 yr, WBC count >25×109/L and 11q23 or MLL rearrangements. Induction therapy included vincristine, prednisone and daunorubicin for 4 weeks. In pts with slow cytologic response to therapy intensified induction with high dose ARA-C and mitoxantrone was administered. Early consolidation therapy included 3 cycles with rotating cytotoxic drugs including high-dose methotrexate, high-dose ARA-C and high-dose asparaginase. Pts. with slow cytologic response on d14 or MRD level >0.05% after consolidation were assigned to allogeneic SCT (related or unrelated) and those with standard cytologic response on d14 and MRD level Results: On June 2009, 235 HR ALL pts were evaluable [mean (SD) age 37(14) yr, 129 males, 155 precursor B-ALL, 80 T-ALL, WBC count 64(96) ×109/L]. Induction death: 20(8%), resistance: 11 (5%), CR: 202 (87%). Slow cytologic response on d14 was observed in 137/186 (74%) pts, MRD Conclusions: These results suggest that in HR Ph- adult ALL pts with adequate response to induction and adequate clearance of MDR after consolidation the results of therapy are not hampered by avoiding allogeneic SCT. MRD is the main prognostic factor for CR, DFS and OS. Supported by grants P-EF/07 from FIJC and RD 06/0020/10556 from RETICS, and PI051490 from FIS, Instituto Carlos III. Disclosures: No relevant conflicts of interest to declare.

Abstract 4711 Aim To analyze the prognostic impact of cytogenetics and main molecular abnormalities on the achievement of complete remission (CR), in patients with primary (de novo) AML (M3 excluded). Patients and Methods Between december-2003 and july-2009, 605 patients up to 70 years-old were included in the CETLAM AML-03 protocol. Induction therapy consisted in one or two courses of idarubicin, intermediate dose ara-C and etoposide, in addition to G-CSF priming from day 0. Cytogenetics classification was as in the MRC studies: Favorable prognosis (FP), intermediate (IP) and adverse (AP). In the IP group, molecular analysis of NPM1 (NPM1+) mutations, CEBPA mutations and internal tandem duplication of FLT3 gene (ITD/FLT3) was performed. In the AP group, the absence (MK-) or presence of monosomal karyotype (MK+) was studied; MK+ was defined as 2 or more autosomal monosomies, or one monosomy and ≥1 structural alteration. Results Median age of the series was 53 years (range, 16-73). In 538 (89%) patients cytogenetic data were available; of them, 64 (10,5%) had FP, 375 (61,8) IP (included 255 with normal cytogenetics) and 99 (16,3%) AP. In the FP group, 33 (5,4%) had the AML1/ETO fusion and 31 (5,1%) the CBF/MYH11. In the IP group, 121 (31%) of 279 patients studied were NPM1+, 100 (26,7%) of 346 were DIT/FLT3+ and 22 (6%) of 235 analyzed were CEBPA+. In patients with AP, 47 were MK- and 33 MK+. Overall, 447 (73,8%) of 588 patients evaluable at the moment of the analysis achieved a CR. CR rates according cytogenetics were: FP 92,2%, IP 76,5% and AP 69,4%, p=0.01. In AML, with AML1-ETO+ and CBF/MYH11+, CR rates were 91% and 93.5%, respectively; of note no chemorefractory cases were observed. In the IP group, CR rates according to mutational status were: NPM1+/FLT3- 91.2%, CEBPA+/FLT3- 94.4%, no mutations 79.1% and DIT/FLT3+ 69.7% (p=0.003). Again, no refractoriness was observed in patients in IP patients belonging to the two groups with favourable molecular profile (p=0.003). In patients with AP, CR rate was 76.1% if MK- and 65.6% if MK+ (p=0.46). The prognostic impact of the herein described cytogenetic and molecular findings was confirmed in multivariate analyses. Comments Genetic characterization is nowadays mandatory in AML. CR rate is high and refractoriness exceptional in the presence of AML1-ETO or CBF/MYH11 rearrangements, or NPM1 or CEBPA mutations without DIT/FLT3. In contrast, CR probability is lower in AP group patients, mainly in those with MK+. These patients require new strategies within clinical trials. Supported in part by grants: GR1-01075, ECO07/90065, PI080672 and RD06/0020/0101. Disclosures: No relevant conflicts of interest to declare.

Abstract 1003 Poster Board I-25 Introduction: Recently, the cooperative group HOVON-SAKK has refined the prognostic impact of cytogenetic abnormalities in acute myeloid leukemia (AML) by introducing the concept of monosomal karyotype (MK). This consists of ≥ 2 autosomal monosomies or one autosomal monosomy in addition to a structural alteration. In their experience, MK would explain the poor prognosis of AML with a complex karyotype. Objective: To investigate the prognostic impact of MK in patients with primary (de novo) AML enrolled in the Spanish CETLAM group protocols (AML 94/99/03). Also, to determine whether considering MK added predictive value to the cytogenetic classification of the Medical Research Council (MRC). Methods: Retrospective analysis of data from 1149 AML patients. Chromosomal formula was centrally reviewed with karyotypes being classified by the presence of MK and allocated into the MRC risk categories. Complete remission (CR) rate, disease-free survival (DFS) and overall survival (OS) were calculated. Results: The karyotype was assessable in 904 (79%) of the 1149 cases. In 145 of the 904 cases (16%), abnormalities involving CBF gene were detected and in 437 (48%) the karyotype was normal (NK). In 253 (28%) additional patients the karyotype was not monosomal; of them, 61 (24%) belonged to the unfavorable MRC with 17 cases harboring a complex karyotype ≥ 5 abnormalities, 7 cases with rearrangements 3q, 13 cases with -7, 9 cases with 5q abnormalities and 16 cases with t(6;9)). The remaining 69 (7.7%) patients had a MK; of them, 59 (85.5%) were from the unfavorable MRC category and included 43 cases with complex karyotype ≥ 5 abnormalities, 6 cases with rearrangements 3q, 5 cases with -7, 5 cases with alterations of 5q). The following table summarizes the results in terms of CR rate, DFS and OS: Conclusions: The addition of MK to the MRC cytogenetic classification refines the prognostic prediction. In our series, the dismal outcome of patients with MK is confirmed; these patients had worse prognosis than those with adverse cytogenetics without MK. Alternative treatment strategies are mandatory for MK+ patients. Supported in part by grants: GR1-01075, ECO07/90065, PI080672 and RD06/0020/0101. Disclosures: No relevant conflicts of interest to declare.

Abstract 1062 Poster Board I-84 Background: Arsenic trioxide (ATO) is currently regarded as the best treatment option in relapsed acute promyelocytic leukemia (APL). The efficacy and safety of salvage therapy using an ATO-based approach compared with chemotherapy-based regimens is not well established. We analyze the clinical outcome of 110 APL patients relapsing after front-line therapy with ATRA and anthracycline, who received second-line therapy with chemotherapy- or ATO-based regimens. Methods: From June 1997 to May 2009, 110 patients (69 M/41 F; median age: 40 years, 2-81) relapsed after front-line therapy with PETHEMA trials (LPA96, n=30, LPA99, n=60; and LPA2005, n=20). Patients presented with either molecular relapse (n=37) or hematological relapse (n=73, 14 of them involving central nervous system). Sixty-seven patients (61%) followed salvage therapy with chemotherapy-based regimens. Therapy consisted of induction with mitoxantrone plus cytarabine plus ATRA (n=46), EMA (n=7), or other chemotherapy regimens (n=14). Thirty-four patients (51%) received one consolidation cycle followed by stem-cell transplantation (SCT) (autologous, 20; allogeneic, 14). Patients not eligible for SCT received consolidation therapy with or without maintenance therapy. From October 2003, 43 patients (49%) received salvage therapy with ATO-based regimens, comprising induction with ATO (0.15 mg/kg intravenously until CR) followed by one consolidation cycle with ATO plus ATRA. Twenty-three patients (53%) received intensification therapy with SCT (autologous, 19; allogeneic, 4). Patients not eligible for SCT received maintenance therapy with ATO plus ATRA with or without Mylotarg or other chemotherapy. Results: Baseline characteristics, including age at relapse, were similar in both cohorts of patients, but patients treated with ATO-based regimens were more frequently late relapses (>18 months after initial APL diagnosis) (67% vs. 40%, P=0.005). The median follow-up in the chemotherapy-based group was 62 months (range, 6-134), and 18 months (range, 2-53) in the ATO-based group. CR rates were 84% in the chemotherapy-based group (8 deaths and 3 resistances) and 88% in the ATO-based group (3 deaths and 2 resistances) (P=0.48). Molecular remission was achieved after consolidation in 79% and 91%, respectively (P=0.13). Twenty-two patients of the chemotherapy-based group that achieved CR were not transplanted because of ineligibility for SCT (n=4), early relapse before planned SCT (n=7), mobilization failure (n=4), clinical decision/toxicity (n=6), and short follow-up (n=1). Reasons for not SCT in the ATO-based group were ineligibility for SCT (n=6), early relapse before planned SCT (n=5), and short follow-up (n=4). The 2-year overall survival (OS), disease-free (DFS), and relapse-free survival (RFS) in the chemotherapy-based group and in the ATO-based group were 44% vs. 63% (P=0.05), 34% vs. 33% (P=0.51), and 37% vs. 34% (P=0.61), respectively. For patients not receiving SCT, the 2-year OS, DFS, and RFS in the chemotherapy-based group and in the ATO-based group were 38% vs. 26% (P=0.98), 34% vs. 23% (P=0.57), and 37% vs. 23% (P=0.46), respectively. For patients receiving autologous SCT, the 2-year OS, DFS, and RFS in the chemotherapy-based group and in the ATO-based group were 60% vs. 87% (P=0.03), 40% vs. 38% (P=0.38), and 42% vs. 41% (P=0.37), respectively. For patients receiving allogeneic SCT, the 2-year OS, DFS, and RFS in the chemotherapy-based group and in the ATO-based group were 56% vs. 100% (P=0.15), 26% vs. 37% (P=0.46), and 28% vs. 37% (P=0.53), respectively. Conclusions: this study performed in a large series of APL patients relapsing after upfront therapy with ATRA and anthracycline shows high rates of CR either with ATO (88%) or chemotherapy regimens (84%). The 2-year DFS and RFS were similar in patients who received second-line therapy with chemotherapy- or ATO-based regimens. However, an apparent benefit in terms of OS was observed in the ATO-based group (P=0.05), suggesting a potential role in minimizing toxicity after administering ATO as salvage therapy. Disclosures: No relevant conflicts of interest to declare.

Objective. To study the incidence, type and prognostic significance of clonal chromosomal abnormalities following SCT in patients with AML. Patients and methods. From February 2000 to March 2008, 74 patients were studied. Karyotypes were analysed by G-banded chromosomes obtained from 24 hours bone marrow cultures, and were described according to ISCN 2005. Results. Clonal abnormalities were observed in 17/74 patients (23%). Median follow-up was 11 months (range 5–47). Initial diagnosis: AML2 (6 cases), AML5a (3), AML6 (3), AML with inv(16)(p13;q22) (1), AML with multilineage dysplasia (1), AML5b (1), AML1 (1) and AML0 (1). Treatment before SCT: idarubicin, cytarabine and etoposide (16), FLAG-ida (1). All patients presented a normal karyotype at the time of SCT. SCT type: autologous (12 patients), allogeneic (3), allogeneic of reduced intensity (2). Conditioning regimen: TBI and cyclophosphamide (14 cases), busulfan and cyclophosphamide (1), fludarabine and busulfan (2). The median time between SCT and the appearance of clonal abnormalities was 8 months (range 3–36). At the time of clonal abnormality detection, 16 patients were in cytological and/or clinical relapse and 1 in complete remission. In 4 cases the initial clone reappeared, 1 showed the initial abnormalities with an acquired abnormality and 12 presented de novo clonal abnormalities. The median survival from the appearance of clonal abnormalities was 2 months (range 1–40). At the time of the analysis all patients had died. Conclusions. The appearance of clonal abnormalities following SCT in patients with AML is frequent. Time between SCT and the appearance of clonal abnormalities is short. The majority of patients presented de novo clonal abnormalities with cytological and/or clinical relapse and poor prognosis.

Background: Therapeutic plasma exchange (PE) is the mainstay of treatment of thrombotic thrombocytopenic purpura (TTP). In many patients, prolonged PE is required to achieve remission and between 30% and 60% of patients relapse after a variable period of months to years, with a further episode of TTP or a chronic picture of recurrent relapses. Various immunosuppressive agents have been used in acute refractory and chronic relapsing cases. Despite advances in treatment, mortality remains at 15–20%. Limited clinical data from small groups of patients suggest that the administration of anti-CD20 antibody (Rituximab) may also be useful in acute refractory and chronic relapsing TTP. However, whether rituximab could be and effective strategy for adults with relapsed or refractory TTP remains unresolved. Patients and Methods: We have studied the clinical response to rituximab in 22 adult patients (16 F/6 M; median age, 43 years, range 24–72) from 11 Spanish centers who presented with acute refractory (13 cases) or an acute relapse (9 cases) episode of idiopathic TTP. Second-line therapies previously administered included vincristine (11 cases) and splenectomy (2 cases). At the time of the study, five (28%) patients presented neurological symptoms, three (17%) renal failure, and two (17%) fever. On admission, every patient received daily PE and corticosteroids during the acute episode. Results: Rituximab was administered a median of 16 days (range, 7–53) after starting PE. It was administered immediately after PE at a dose of 375 mg/m2 each week for a median (range) of 4 (1–8) doses. No severe acute or delayed toxicity associated with rituximab was observed. Two (9%) patients died due to TTP-related causes 24 and 48 hours after relapse and were not evaluable for response. The remaining 20 (91%) patients achieved complete remission in a median of 21 days after initiating rituximab (range, 2–35). After a median follow-up of 13 months (range, 1.5–33), 17 patients are in remission and 3 patients have relapsed at 7, 29, and 29 months after rituximab administration. Conclusion: Based on these results, rituximab appears a safe, effective, targeted therapy associated with a high response rate in adult patients with acute refractory/relapsing idiopathic TTP.

We analyzed the complete remission (CR) rate and toxicity in patients (pts) with “de novo” AML enrolled in an intensive CT protocol including induction and intensification CT concurrently with G-CSF. Between December-2003 and April-2005 167 adults 60 years n=43). Cytogenetical analysis was available in 96% of pts. All pts with 30x109/L (n=3), bone pain (n=1) or infectious complications (n=6). In 21 of the 53 pts (39%) with WBC ≥ 30x109/L at diagnosis G-CSF was administered after the counts were reduced (median period 4 days). After induction therapy CR was obtained in 69% of pts (n=115) (≤ 60 years: 69% and >60 years: 67%), 85% of them with a single CT course and 14% of pts (n=25) were considered as refractory. Treatment related mortality (TRM) was 16% (n=27) (≤ 60 years: 14% and >60 years: 21%) due to infection (n=17), hemorrhage (n=4) or multiorganic failure (n=6). The median duration of neutropenia (60 years n=28), 96% with concurrent G-CSF (in 8 cases was suppressed because of WBC >30x109/L). Median neutropenia was 19 days (6–47) and thrombocytopenia had a median duration of 22 days (2–78). The intensification TRM was 7% (n=8): 5 pts >60 years (18%) vs 3 pts ≤ 60 years (3%) (p=0.02): being the infections (87% of cases, n=7) the main TRM cause. No significant differences were observed in the induction results between the present protocol and the previous one (LMA-99); post-intensification neutrophil and platelet recovery time were significantly longer for the current protocol: 21 and 22 days vs 17 and 16 days respectively (p60 years had a high TRM, mainly post-intensification CT. Longer follow-up is needed to evaluate the G-CSF effect on relapse and survival.

In this study we tested whether distinct patterns of functional genomic polymorphisms in genes involved in drug metabolic pathways (GSTT1, SULT1C2, CD-EX, TOP2A, SXR-1, SXR-2), DNA repair (XPD, XPA, XPG, ERCC1), angiogenesis (VEGF-1, VEGF-2), and multidrug resistance (MDR1-1, MDR1-2) predict clinical outcome in patients with AML. Two hundred adult patients ≤60 year-old with primary AML (M3 excluded), enrolled in the AML-99 protocol of the CETLAM group between November-98 and July-03, have been studied. Induction therapy consisted of idarubicin, intermediate dose ara-C (IDAC) and VP16. Intensification included mitoxantrone and IDAC. Patients were following treated with high-dose ara-C or transplantation depending on the risk category. According to the MRC cytogenetic classification, 32 patients had a favourable karyotype, 23 had abnormalities indicating poor prognosis, and 110 patients were included in the intermediate prognosis (IP) group. We focused on IP group patients with the aim of improving its prognostic stratification. The characteristics considered were: age (20x109/l), FAB classification, MLL rearrangement, internal tandem duplication of FLT3 (ITD-FLT3), induction courses to achieve complete remission (CR) (1 vs 2), and germline polymorphisms of the above-mentioned genes. Prognostic variables in the univariate analysis (Kaplan-Meier method) with a p value ≤0.2 (log-rank test) were included for the multivariate analysis (proportional hazard method). Of the 110 patients included in the IP group, 86 (78%) achieved a CR, 11 (10%) were chemoresistant and 13 (12%) died during induction. After a median follow-up of alive patients of 24 months (range 5–64), overall survival was of 31% at 5 years. In the multivariate analysis, adverse prognostic variables for survival were polymorphism of XPA (RR=3.4; p=0.02) and of MDR1-1 (RR=2.1; p=0.02), and WBC >20x109/l (RR=2.1; p=0.02). Increased risk of relapse was associated with polymorphism of SULT1C2 (RR 4.1; p=0.004), ITD-FLT3 (RR 3.3; p=0.003), VEGF2 (RR 2.8; p=0.04) and MDR1-1 (RR 2.4; p=0.02). Finally, in the multivariate analysis for refractoriness to chemotherapy, XPA polymorphism was the only independent factor increasing the risk (RR=14; p=0.02). In conclusion, germline polymorphisms, which can easily be analyzed from DNA of peripheral blood, have independent prognostic value in intermediate risk AML.