ALBERT

Description: Purpose: Recurrently mutated genes have been identified in patients with myelodysplastic syndromes (MDS) and, more recently, in patients with unexplained cytopenia. (Kwok et al. Blood 2015, Hansen et al. American Journal of Hematology 2016 and Malcovati et al. Blood 2017). In this study, we investigated the prognostic impact of these mutated genes in patients with idiopathic cytopenia and compared them to a control cohort of patients with low risk MDS. Methods: We included patients with idiopathic cytopenia after routine assessment, without cytogenetic abnormalities. For comparison, a group of low risk MDS patients without cytogenetic abnormalities, excess of blasts or ring sideroblasts were included. All samples were sequenced covering at least the 20 most recurrently mutated genes in MDS, and a subset of cases underwent a blinded morphology review by two hematopathologists. Results: Two hundred and forty nine patients, 171 with idiopathic cytopenia and 78 with low risk MDS, were included in this study. Of these, 80 (47%) and 53 (68%), respectively, had one or more detectable mutations. There was no difference in survival between the groups, however a predefined subset of "adverse mutations" (ASXL1, NRAS, SRSF2, U2AF1, TP53, RUNX1, EZH2, IDH2 and GATA2, adopted from Bejar et al. Current Opinion in Hematology 2017) was associated with inferior survival in the MDS group (p= 0.035), but not in the group with idiopathic cytopenia and at least one mutation (p= 0.43) (Figure 1). However, if an adverse mutation was present in the idiopathic cytopenia group the risk of progression to MDS or AML increased significantly (HR [CI:95%] 12.01 [1.47; 98.23], p= 0.02), after adjusting for age and sex. Thus mutational screening identified the patients with unexplained cytopenia at risk of progressing to an overt myeloid neoplasm (Figure 2). A total of 18 patients (23%) progressed to a myeloid neoplasm during follow up, of those 12 had material available at time of progression. All patients who progressed to AML (n=4) acquired a new driver mutation at time of progression, in contrast to the patients who progressed to MDS or CMML (n=8) without excess of blasts, who showed a clonal expansion or a steady variant allele frequency at the time of progression. TET2 and DNMT3A mutations were more frequent in patients with idiopathic cytopenia, and were associated with less dysplasia of bone marrow cells. A total of 109 cases with idiopathic cytopenia underwent a blinded morphology review by two independent reviewers; ten cases were concordantly reclassified to fulfill the criteria for MDS, and all of these had at least two mutations. None of these have progressed to higher risk MDS and these ten are not included in the 18 patients mentioned above, who progressed to MDS, CMML or AML during follow up. Conclusion: We here show that mutational profiling can identify patients with idiopathic cytopenia who are at risk of progression, but in contrast to low-risk MDS, the presence of adverse mutations in patients with idiopathic cytopenia do not predict inferior survival. Disclosures Hansen: Otsuka Pharma: Membership on an entity's Board of Directors or advisory committees. Grønbæk:Janssen Pharma: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees; Otsuka Pharma: Membership on an entity's Board of Directors or advisory committees.

Description: Key Points The intensified standard-of-care regimens for younger patients with MCL do not overcome the deleterious effects of TP53 mutations. MCLs with TP53 mutations should be considered for alternative frontline treatment.

Description: INTRODUCTION: During the past decades, the outcome of Mantle Cell Lymphoma (MCL) treatment has improved substantially in younger patients. In a recent update of the Nordic MCL2 trial we show very long response durations after a median follow-up of 11.4 years, but we also observe a continuous pattern of relapses even after 10 years of remission.(Eskelund et al, 2016) The course of this disease remains very heterogeneous, and with the current prognostic indexes we are unable to identify patients who might be cured by the current standard-of-care and others who would possibly benefit from alternative frontline approaches. Recently, next-generation sequencing (NGS) studies have explored the mutational landscape of MCL, however, in inhomogeneous and diversely treated cohorts or with short follow-up. Still, TP53 and NOTCH1/2 mutations have been shown to be prognostic markers. In our current study we examine the prognostic impact of aberrations in the most frequently mutated genes in MCL in a homogenously and optimally treated patient cohort, with a long-term follow-up. MATERIAL AND METHODS: Freshly frozen DNA from diagnostic bone marrow samples from patients included in two prospective Nordic trials, MCL2 and MCL3 were analysed. In both trials patients received intensified first line induction therapy with alternating courses of R-CHOP and R-HD-Cytarabine and consolidation with high-dose therapy and ASCT. All patients signed an informed consent.(Geisler et al, 2008; Kolstad et al, 2014). NGS was performed using the Ion Torrent Technology. A targeted panel of 8 genes frequently mutated in MCL was constructed on the basis of previous NGS studies.(Bea et al, 2013; Zhang et al, 2014) The panel included all coding regions of the following genes: ATM, CCND1, TP53, KMT2D, NOTCH1, NOTCH2, WHSC1 and BIRC3. Cut-off for calling a mutation was set to a variant allele frequency 〉3%. Mean depth was 〉1500X in all patients. RESULTS: So far, we have mutational data from 72 patients. Patients were previously untreated and 1 mutation (2-4). Mutations were distributed as follows: ATM 15 (21%), KMT2D 11 (15%), WHSC1 7 (10%), TP53 6 (8%), CCND1 5 (7%), NOTCH2 4 (6%), NOTCH1 3 (4%), BIRC3 2 (3%). In univariate analyses, mutations in TP53 were highly predictive of an inferior outcome (median OS and PFS were 14 and 10 months, respectively; p1 mutations. In multivariate analyses, the only prognostic significant mutations were TP53 (p150 patients from the combined MCL2 and MCL3 cohort. Disclosures Jerkeman: Amgen: Research Funding; Gilead: Research Funding; Janssen: Research Funding; Celgene: Research Funding; Mundipharma: Research Funding. Geisler:Roche: Consultancy; Janssen: Consultancy; Celgene: Consultancy; Sanofi: Consultancy.

Description: Abstract 3514 Isocitrate dehydrogenase (IDH) is a metabolic enzyme that catalyzes a reaction in the tricarboxylic acid cycle. Gain of function mutations in the IDH1/2 genes have been reported in different malignancies and are observed in 15–30% of de novo AML with association to a normal karyotype and to NPM1 mutations. The exact role of IDH1/2 mutations in leukemogenesis remains to be determined. IDH mutations have not previously been studied in a cohort of therapy-related myelodysplasia (t-MDS) and therapy-related acute myeloid leukemia (t-AML). To evaluate the frequency of IDH1/2 mutations in t-MDS and t-AML, and their possible association to type of previous therapy and to other genetic abnormalities, DNA from 140 well-characterized patients with t-MDS (n=89) and t-AML (n=51) were analyzed with high-resolution melting followed by sequencing. All patients have previously been examined cytogenetically and investigated for mutations in 12 other genes: FLT3(ITD, TKD), KIT, JAK2, KRAS, NRAS, BRAF, PTPN11, RUNX1, MLL(ITD), CEBPA, NPM1, and TP53. In total, IDH mutations were detected in 12 of 140 patients (9%). 3 patients had a mutation in IDH1 and 9 patients had a mutation in IDH2 (Table 1), all mutations previously reported in de novo AML. No patients had concurrent IDH1 and IDH2 mutations. IDH mutations were not related to previous therapy with alkylating agents, topoisomerase II inhibitors or radiotherapy, but were significantly associated with other types of therapy not firmly established to be leukemogenic (p=0.004). The latency period to development of t-MDS/t-AML was not different between IDH1/2 positive (+) cases and cases with IDH (wt) (64 and 48 months, respectively, p=0.118). 4/5 cases with t-MDS and IDH+ progressed to AML compared to 27/84 t-MDS cases with IDHwt (p=0.048).Table 1:Characteristics of 12 patients with t-MDS/t-AML and mutations in IDH1/2CaseAge/sext-AML/t-MDSPrevious therapyKaryotypeOther mutationsIDH Mutation1974/FAMLAlk45,XX,-7/48,XX,der(1;7)(q10;p10),+11, +13/46,XX–IDH1 R132G2963/FAMLRT46, XXNPM1 FLT3-ITDIDH1 R132G3663/FAMLAlk46,XX,+2,+8/47,XX,der(6)t(1;6) (q?25;p21),+8N-RASIDH2 R172K4472/MMDSAlk46,XY,+1,der(1;7)(q10;p10)/46,XY–IDH2 R140Q5562/FMDS→AMLRT46, XXRUNX1IDH2 R140L7272/FMDS→AMLAlk, T II, RT46,XX,+1,der(1;7)(q10;p10)/50,XX,idem, +8,+9,14+21RUNX1IDH2 R140Q8178/MMDS→AMLAlk46,XY,der(17)t(11;17)(q13;p13),i(13) (q10)/47,idem,+der(13)t(11;13) (q13;p11)IDH2 R172K10443/FMDS→AMLAlk47,XX,+1,der(1;7)(q10;p10),+8RUNX1IDH1 R132C10944/FAMLMtx, Aza46, XXIDH2 R140Q11952/FAMLAlk, T II, RT46, XXNPM1IDH2 R140Q13325/MAMLVCR, MTX, Asp,6-MP46, XXIDH2 R140Q18060/MMDS→AMLMtx46, XXMLL-ITDIDH2 R140Q6-MP, 6 mercaptopurine; Alk, alkylating agent; Asp, l-asparaginase; Aza, azathioprine; Mtx, methotrexate; RT, radiotherapy, T II, topoisomerase inhibitor, VCR, vincristine. IDH mutations were significantly associated with a normal karyotype (6/12 cases with IDH+ vs. 18/128 with IDHwt, p=0.006) and der(1;7)(q10;p10) resulting in trisomi 1q and loss of 7q (4/12 cases with IDH+ vs. 7/128 with IDHwt, p=0.008), but was inversely correlated to other chromosome 7 abnormalities (1/12 cases with IDH+ vs. 54/128 with IDHwt, p=0.03). No patient with mutated IDH had chromosome 5 abnormalities, TP53 mutations or recurrent balanced translocations. 7/12 patients with mutated IDH1/2 had other gene mutations characteristic of AML (Table 1). The frequency of each of these other mutations were not different from patients with wildtype IDH1/2 (RUNX1, p=0.4; NPM1, p=0.2; FLT3, p=1.0; MLL, p=0.165; N-RAS, p=1.0). In conclusion, mutations of IDH1/2 were observed in 9% of patients with t-MDS/t-AML. They were not related to any specific type of therapy but perhaps associated with transformation from MDS to AML. IDH mutations clustered in the genetic pathway characterized by a normal karyotype and mutations of NPM1, and the pathway characterized by 7q−/−7 and RUNX1 point mutations. The significant association observed between IDH1/2 mutations and der(1;7)(q10;p10) may indicate that this cytogenetic aberration represents a specific entity, biologically distinct from other chromosome 7 abnormalities. This is also supported by the different clinical outcome between cases with der(1;7) and other cases with -7/7q- (Sanada et al, Leukemia 2007). Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 3859 The MPL gene is located on chromosome 1p34 and encodes the thrombopoietin receptor. It includes 12 exons and is a key factor for growth and survival of megacaryocytes. Acquired mutations in this gene activate the thrombopoietin receptor constitutively. MPL515 somatic mutations are stem cell-derived events that involve both myeloid and lymphoid progenitors. Two distinct exon 10 mutations are found in 15% of JAK2-V617F negative myeloproliferative neoplasms (MPN),), i.e. 3% of essential thrombocythemia (ET) and 10% of primary myelofibrosis (PMF).: W515L, W515K and the rare W515A variant. A hereditary mutation, S505N, is associated with familial thrombocytosis. MPL mutation detection is a helpful new tool to detect clonality in JAK2 V617F negative MPN and to establish the diagnosis of MPN. As many laboratories use very different methods and interpretations, standardization is highly warranted. Particularly the methodology is diverse and the results need to be comparable, requiring comprehensive testing. This quality control project was established within workgroup 2 of the MPN&MPNr-EuroNet network, (www.mpneuronet.eu, supported by the European program COST (CoOperation in Science and Technology)). The lab from the ‘Hopital H. Mondor AP-HP Paris' provided 29 samples containing randomized concentrations (between 100% and 1%) of the four mutations MPL W515L, W515K, W515A and S505N. The plasmids used for this quality control experiment spanned exons 9, intron 9 and exon10 of the MPL with S505N, W515L, W515K and W515A mutants diluted first with wild-type plasmid gene and then diluted in human genomic DNA. Thirteen European laboratories tested these 29 samples, each using their own chosen methods (14 altogether). The following methods were used: Mutascreen W515L/K Kit (Ipsogen, France):(n=4), allelic discrimination real-time PCR (n=2), high resolution melting (HRM) (n=7) and sequencing (n=2, 1 Sanger, 1 pyrosequencing)). There were no false positive results in any of the labs. All labs using the Mutascreen W515L/K Kit detected all W515L and W515K mutations, from 100% mutated down to 1% mutated plasmids. The allelic discrimination assays which were also designed for W515L and W515K only, detected the mutations down to 2%. The HRM methods were all designed differently. All except one (which did not recognize S505N) detected all 4 mutations with a sensitivity, down to 5% mutated plasmids, with few exceptions detecting either lower or higher amounts. The Sanger sequencing and pyrosequencing assays had a detection limit of 5–10%, with the pyrosequencing assay not being designed for the S505N mutation. All participating labs detected the most frequent MPL mutations in MPN W515L and W515K, with many designs not including W515A and S505N. Achieved sensitivities differed between methods with cutoffs of 1% to 10% (1.5% for the Ipsogen kit). Most laboratories reported the results as either positive or negative. However, the percentages of mutated alleles reported by a few labs differed greatly from each other and from the original dilutions (range 2–50 times) In conclusion, these results show that the diverse methods for MPL mutation detection used by different European labs yielded comparable specificity with varying sensitivity. Smaller clones might be missed by the less sensitive methods, and quantification of mutated alleles should be interpreted very carefully until standardised reference material for MPL mutation testing will be available. More extensive interlaboratory testing including patient samples is needed to identify the most robust assays suitable for diagnostic mutation detection and particularly for quantification of mutated alleles. Disclosures: No relevant conflicts of interest to declare.