key: cord-0004700-uq9y96lu authors: nan title: ACUTE LEUKEMIAS XV: Biology and Treatment Strategies Munich, Germany, February 22– 25, 2015 date: 2015-02-04 journal: Ann Hematol DOI: 10.1007/s00277-014-2297-4 sha: 585845b90b767c6f7faddafd3e841ea3c0dd7775 doc_id: 4700 cord_uid: uq9y96lu nan Pediatric Hematology and Oncology, Hannover Medical School, Germany; 2 Institiute of Human Genetics, Hannover Medical School, Germany; 3 It is now realized that T cell immune surveillance of cancer occurs in many cases. One approach that has now been developed to deal with tolerance to tumors has been checkpoint blockade, which can enable the natural immune response to respond to antigens encountered in the tumor microenvironment. In this seminar I will outline recent results in the area of synthetic biology, which uses engineered T cells to overcome immune tolerance. There are three types of T cell therapy that are being actively tested in the clinic (1) . In one case tumor infiltrating lymphocytes are obtained from surgical biopsies. This therapy is most advanced in melanoma, and trials for patients with metastatic melanoma are currently underway with phase 3 trials in Europe. Gene transfer technologies can also be used to produce engineered T cell therapies. One approach is a chimeric antigen receptor or CAR T cell approach. Another is the use of engineered T cells that express T cell receptors of known specificity and affinity. There are number of considerations for successful T cell therapy with adoptively transferred T cells. It is necessary to have an adequate number of T cells in order to eradicate large numbers of tumor cells. If a kilogram of tumor represents approximately 1x10 12 T cell tumor cells, then it is likely that a similar number of T cells will be required for an equivalent effector to target ratio in vivo. The failure to achieve a critical mass of T cells numerically may explain many previous trials with disappointing results following therapeutic vaccines. There are two potential solutions to overcome inadequate effector T cell function in cancer patients with adopted cell transfer therapy. On the one hand it is possible to infuse very large numbers of T cells, as has been done in tumor infiltrating lymphocyte therapy 2) . The other approach is to infuse smaller numbers of cells that are programmed to expand extensively in the patient 3, 4) . There are three essential factors to consider when developing engineered T cell therapies (Fig1). One issue under consideration is the nature of the lymphocyte subset to be infused. At present it is not known whether the optimal cell type is a memory cell or a naïve cell. Related to this is the question of the proportion of CD4 helper cells and CTLs. Secondly it is important to consider the aspects of genetic engineering that are used to reprogram the cells 5) . Finally it is necessary to have an optimized approach for the manufacturing of the T cells. The design of CAR T cells has become increasingly sophisticated (Fig2). The initial CAR T cells were so-called first-generation CARs and only had signaling from the T cell receptor zeta chain or related molecules. More recently second and third generation CAR T cells have been designed, incorporating progressively more complicated signaling domains. A number of strategies have been used to make CARs more potent or to enhance their proliferative capacity or their cytokine secretion. For example it has been shown that so-called Signal 1 from the Lafferty model 6) can be enhanced by encoding a ubiquitin-resistant linker of activated T cells (LAT) 7) . A number of costimulatory domains have been used to augment Signal 2. These include CD28, and the tumor necrosis family members 4-1BB, Ox40, and CD27, for example. Recent studies have shown that the ICOS signaling domain, a molecule related to CD28, provides CAR T cells with enhanced persistence in preclinical animal models. These studies were initiated by Crystal Paulos and showed that when T cells were activated through their natural ICOS receptor molecule with agonistic antibodies, that this promoted the proliferation of TH 17 cells and T follicular helper cells that had highly potent effector functions (Fig 3) . In contrast, if the T cells were costimulated with a CD28 dominant costimulatory signal, then the TH17 cells were restrained in their antitumor activity and inflammatory properties 8) . More recently, Sonia Guedan has demonstrated that when ICOS is expressed in CD4 cells by encoding the cytosolic domain in the CAR endodomain that this leads to enhanced persistence of the CD4 cells and that those cells promote the persistence of CD8 CAR T cells 9) . This latter effect is independent of the costimulatory domain it is expressed in the CD8 CAR T cell. Other studies in our laboratory have recently identified examples of CAR designs that have either a continuous and ligand-independent growth phenotype, or are quiescent in the absence of stimulation by surrogate antigen. Finally, studies by Yangbing Zhou have recently shown that it is possible to increase the therapeutic index of CARs by affinity-tuning of the single chain variable fragments. Ongoing research is defining the characteristics of CAR design that lead to a continuously active CAR. These T cells are capable of proliferating for months without stimulation or addition of exogenous cytokines. Factors that dictate ligand independent growth of the CAR T cells depend on the particular scFv used, and so far they have only been observed with CD28 endodomains. The growth is independent of exogenous ligands and there is evidence for constitutive signaling through CD28 involving AKT, NF-kappaB and the MAP kinase pathway. The continuous CAR T cells have constitutive secretion of Th1 and Th2 cytokines, and they tend to highly express T-bet, Eomes, GATA3, and Bcl-xL. Transformation not observed, and surprisingly, CAR T cells with the continuous growth phenotype are less effective in pre-clinical tumor models. One implication of this research is that it is important to screen newly designed CAR T cells for constitutive activity. The initial CAR T cell trials were conducted in patients with HIV, using a first-generation CAR (10) . At the University of Pennsylvania CAR T cell trials in cancer patients targeting CD19 were initiated in 2010 11) . There were three unique aspects of this the CAR, termed CART19, that were used in this trial that have not been tested previously. First a lentiviral vector was used to express the CAR. Secondly, a short duration T cell culture process using antibody-coated beads was used for T cell manufacturing. Thirdly, the CAR employed a TNF family member for co-stimulation, CD1 37, also referred to as 4-1BB. The initial patients that were treated had refractory chronic lymphocytic leukemia 12) . All three of the initial patients responded to the CART19 therapy. The patients had large bulky tumors and it was possible to conclude that each CART19 cell or its progeny through cell division killed more than 1000 tumor cells. As of December 2014 the patients continue to express the CARs and the remissions have been durable in two of the three patients having no evidence of leukemia for more than 4 years. The level of CAR T cell proliferation in vivo is a predictive biomarker of clinically beneficial antitumor responses. In more recent studies the CART19 T cell approach has been used to treat patients with acute lymphoblastic leukemia 9ALL). We recently reported the first 30 patients on this phase 1 trial and a striking 90% rate of complete remission was observed in both pediatric and adult cases 13) . Updated results as of December 1, 2014 are shown in Fig4. Durable B cell aplasia is associated with a low risk of relapse of ALL following CART19 T cell therapy. There have been a number of toxicities associated with CAR T cell therapy. B cell aplasia is observed in responding patients following therapy with CD19 or CD20 specific CARs. Tumor lysis syndrome has been observed in many patients, and this is managed as per standard oncology practices and the unique aspect of this is that it may be delayed for 20 to 50 days after infusion of the CAR T cells. Cytokine release syndrome characterized initially with fever occurs in many patients, and this is related to the tumor burden in the patient. Finally, a subset of patients develops macrophage activation syndrome which is characterized in the serum by very high levels of ferritin, C-reactive protein and evidence of coagulopathy. The hallmark of this syndrome is very high levels of IL-6 and interferon gamma in the serum. This syndrome is rapidly reversed with the IL-6 receptor antibody tocilizumab, indicating that this inflammatory disorder is dependent on IL-6 (Fig5). : Tocilizumab Anti-Cytokine Therapy for Cytokine Release Syndrome. A patient with leukemia was given tocilizumab on day 10 after CART19 therapy, with immediate resolution of fever In summary, engineered T cells are finally on the path towards widespread use. A number of trials are now underway that should lead to commercial approval in the United States and in Japan. The interactions between the bone marrow niche and AML stem cells may offer niche-specific protection against therapy and lead to relapse (1, 2) . Forthcoming therapies that target the molecular connections between leukemia cells and surrounding shelter offer promise for eradication of residual AML cells and cures (3) . The marrow is composed of at least two separate environmental compartments relevant to hematopoietic cells: the endosteal (osteoblastic) niche and the endothelial (vascular) niche. Research has supported the notion that both are involved in normal hematopoietic stem cell function. Both niches also appear to be instrumental in abnormal leukemia stem cell (LSC) maintenance, in a way that likely mimics their role in normal blood cell function (2, (4) (5) (6) (7) . Given that leukemia initiating cells are present in more than one hematopoietic progenitor phenotypic compartment, and that various leukemia stem cell phenotypes have been identified in AML, it is conceivable that each niche supports different leukemia stem/progenitor cell states (8, 9) . Interactions that occur between various niche cells and LSC encompass a complex interplay that relies on a multitude of cells, signals, conditions, and contacts (Fig1). : Complexity of complementary niches. Leukemia stem cells may reside in the endosteal niche and/or the endothelial niche. Both rely on microenvironment signals (such as CXCL12) and cell surface receptors (such as CXCR4 and E-selectin) to guide and maintain stem cell states. The proliferation of leukemia cells in and around the niche promotes a hypoxic environment that may metabolically favor LSC survival. Other cellular components of the marrow niche, such as mesenchymal stem cells (MSC) and osteoblasts, also communicate with and contact niche leukemic cells. A particular target for mobilization is not restricted to a singular niche. Emerging therapies target complementary elements of the bone marrow niche, and thus may also target different stem cell fractions in AML. Targeting the CXCR4/CXCL12 axis or E-selectin are both viable strategies for helping mobilize and eliminate leukemia cells. To further evaluate the microenvironment in vivo, we have incorporated a humanized bone marrow model in mice. The purpose of the current study was to determine whether human cancellous bone chips could serve as scaffolding in a model using subcutaneous implantation of mesenchymal stem cells (MSC), endothelial colony forming cells (ECFC), and matrigel, along with concomittant addition of bulk or FACS-sorted subpopulations of leukemia cells. Novel approaches to anti-CXCR4 therapy Identifying the elements linking leukemia cells and the microenvironment, and understanding those interactions mechanistically, is critical for disruption of leukemia resistance. One interaction that occurs between LSC and the bone marrow microenvironment is through C-X-C chemokine receptor type 4 (CXCR4, or CD184) on the leukemic cell, and its attractant ligand C-X-C chemokine receptor 12 (CXCL12, or stromal cell-derived factor-1α, SDF-1) in the marrow niche (10) (11) (12) . Interestingly, the level of CXCR4 expression on AML cells correlates with worse outcomes (13, 14) . CXCR4/CXCL12 inhibition is a known mobilizer of marrow stem cells, and the CXCR4 inhibitor plerixafor is approved by the US Food and Drug Administration for mobilization of CD34+ cells prior to autologous stem cell transplantation for non-Hodgkin's lymphoma and multiple myeloma (15) . Inducing mobilization of LSC is a method that is also being used to treat leukemia. In pre-clinical in vivo leukemia models, inhibition of CXCR4 resulted in mobilization of leukemia into peripheral blood and subsequent sensitization to cytotoxic chemotherapy (16) . A recent phase 1 study demonstrated this same approach was useful in humans, where AML patients treated with plerixafor prior to chemotherapy demonstrated modest (two-fold) leukemia mobilization and increased sensitivity to a standard ara-C plus anthracycline regimen (17) . A more recent study used plerixafor plus G-CSF in the pre-allogeneic stem cell transplant setting for patients with myeloid malignancies (18) . Although outcomes were not superior compared to historical patients, efficient mobilization of leukemic clones occurred (18) . Other CXCR4 inhibitors are under development including a monoclonal antibody (PF-06747143) (19) , peptide inhibitors (BL-8040) (20) , and other small molecule inhibitors (such as LY2510924) (21) . These agents by themselves exerted anti-leukemic effects, with complete remissions and reduction of bone marrow blasts in pre-clinical and clinical studies. The concept of CXCR4 inhibition was initially developed with the aim of releasing leukemic cells from the niche, and thus removing microenvironmental protection from LSC. Combinations of CXCR4-inhibitors with other agents are being investigated, but optimized synergistic lethality with other AML therapies has yet to be determined. This concept was complemented by the observation that CXCR4 inhibition can lower the apoptotic threshold of AML cells, independent from mobilization, through an axis consisting of the SDF-1-receptor CXCR4, the transcription factor Yin Yang 1, the microRNA let7a, and its targets BCL-XL and c-MYC (22) . Combinations with other mobilizing agents such as granulocyte-colony stimulating factor (G-CSF), are designed to potentiate the effects of CXCR4 inhibition. G-CSF itself works through several mechanisms to promote mobilization (23) . These agents may be combined also with other targeted therapies, such as FLT3 inhibitors. A trial that investigated the combination of plerixafor, G-CSF, and sorafenib in relapsed/ refractory FLT3-mutated AML resulted in 36% CR/p rates, an overall response rate of 62% and some durable responses (24) . Recent work has revealed the interactions between HSCs and the vascular niche by way of E-selectin binding and signaling, and in particular the effects of E-selectin on stem cell function (25) . Accumulating evidence suggests that inhibition of this pathway also induces AML stem cell mobilization and blocks protection from the vascular niche (26, 27) . The E-selectin inhibitor GMI-1271 is progressing into clinical trials for patients with AML. Targeting the vascular niche to eradicate LSC may become a complementary strategy to the targeting of other niche components. The interactions between AML stem cells and the endothelium is not fully understood (28) . Interestingly, hematopoietic and endothelial cells arise from the same primitive embryonic precursors (the hemogenic endothelium), and many genes important in the hematopoietic and endothelial differentiation process are known to be involved in AML (29) . A special relationship may be forged between leukemia stem cells and the endothelium that has yet to be understood and exploited for treatment and prevention of relapse. In order to better understand how the leukemia niche functions, it may be studied using a variety of experimental tools. For example, the leukemia niche has been shown to be hypoxic compared to normal bone marrow. Treatments that capitalize on the hypoxic microenvironment, such as hypoxia-activated pro-drugs, may be particularly useful as another niche-directed therapy (30) . New findings suggest that leukemic cells and LSC are generating energy not through the Warburg effect, but through oxidative phosphorylation (Oxphos). Thus LSC scavenge oxygen and adapt to thrive under hypoxic conditions (31). Hypoxic culture conditions are beneficial to the maintenance of a stem cell state (32), and for the growth and expansion of MSC. Recently, advances have been made by creating models to study the leukemia microenvironment that allows for the recreation of a human marrow niche in mice. One such model is carried out by adding MSC and endothelial colony forming cells (ECFC) to matrigel and injecting subcutaneously into the flanks of mice (33). After an extended period of time, humanized bone and bone marrow forms and engrafts AML cell lines as well as primary patient samples. The model allows for the microenvironment to be genetically modified prior to injection and bone formation (33). Inclusion of ECFC appears to be optional, and vascular formation may arise from both MSC and ECFC (34). Such models are aimed at creating a better representation of the human microenvironment in the context of leukemia and LSC xenografts. Knockdown of HIF-1α, CTGF, or NFkB in MSC all affected AML engraftment (33, 35, 36). A previous pilot study used the addition of cancellous bone chips implanted in the flanks of mice with MSC and ECFC (1 x 10 6 cells each) in matrigel and showed appropriate niche formation within 4 weeks (Figure 2A-B) . To further test whether a scaffolding nidus for bone formation could be used to improve bone and niche formation with concomitant addition of human leukemia, we used 2-3mm cubiform human cancellous bone chips (Medtronic Cat#P00035) combined with 200 μL matrigel containing MSC (2x10 5 cells) and ECFC (2x10 4 In addition, we added bulk leukemia cells or FACS-sorted subpopulations (including CD34+CD38-CD123+ LSC) from 2 primary AML patient samples. Five NOD-scid IL2rγ null (NSG) mice were implanted with single bone chips, along with 8x10 3 to 1.3x10 6 AML cells per mouse. Mice were monitored daily for signs of illness, and retro-orbital peripheral blood sampling was performed at regular intervals to detect engraftment of human leukemia by flow cytometry of human CD45 and CD33. All mice tolerated procedure and implantation. Peripheral blood evaluation showed no engraftment of human AML up to 16 weeks. Interim analysis for leukemia of one mouse at 16 The use of human cancellous bone chip implanted with MSC, ECFC, and leukemia cells as adapted from previously described humanized bone marrow model in mice did not so far result in improved or faster marrow niche formation or AML engraftment. Monitoring of xeno-transplanted mice is ongoing. The previously described and established model of humanized heterotopic bone marrow model in mice continues as the gold standard, and may be used to genetically modify the AML microenvironment (22, (33) (34) (35) (36) . As our understanding of the LSC evolves, so has our recognition that these cells function as a product of their surrounding microenvironment. Recent data suggests that clonal remnants in the form of pre-leukemic cells often remain after therapy, and may be responsible for detection of AML mutations by deep sequencing in remission (37). These preleukemic stem cells may occupy the bone marrow niche. Eradication of these cells could potentially take place with selective targeting of the niche where they reside. The optimal way to eliminate these cells within their niche is not yet clear, and niche-directed therapies, such as CXCR4 or E-selectin antagonists, may prove instrumental to prevent relapse. Enhanced testing of emerging therapies can be carried out using in vivo models that replicate structural and functional components of the human bone marrow microenvironment. The authors declare that they have no conflict of interest High-level Hox gene expression is a recurrent feature of several different malignancies including diverse subtypes of acute myeloid leukemia (AML). We have previously shown that that H3K79 dimethylation (H3K79me2) by the enzyme Dot1L is required to maintain Hox gene expression and proliferation of MLL-rearranged leukemias. This has prompted the development of small molecules that are now in clinical trails. More recently we have studied the regulation of the Dot1L complex and identified the role of Dot1L in the control of gene expression. We find that H3K79me2 is converted to mono-methylation (H3K79me1) at critical Hox loci as cells transition from hematopoietic stem cell-enriched populations to myeloid progenitor cells thus coinciding with a decrease in Hox gene expression. Conversion of H3K79me1 to H3K79me2 is critical for continued expression of Hox genes in leukemia cells and H3K79me2 but not H3K79me1 is dependent on Af10, a core constituent of the Dot1l complex. Af10 loss of function reverses aberrant epigenetic profiles found in leukemia cells and dramatically impairs the transforming ability of MLL-AF9, MLL-AF6 or NUP98-NSD1 fusionsdiverse, mechanistically distinct oncogenes that drive abnormal Hox gene expression. Furthermore, NUP98-NSD1 transformed cells, much like those transformed by MLL-fusion proteins, are highly sensitive to small-molecule inhibition of Dot1l. Also, we have performed a genome scale RNAi screen and found that suppression of the histone deacetylase Sirt1 and histone methyltransferase Suv39h1 rescued MLL-rearranged leukemia cells from their dependence on Dot1L. We find that Sirt1 and Suv39h1 are required for the acquisition of a repressive chromatin state consistent with facultative heterochromatin around MLLfusion target genes after Dot1L inhibition. Epigenomic profiling revealed that Dot1L-dependent genes demonstrate an H3K9ac high /me2 low signature that is induced at MLL-AF9 bound loci. This signature is also normally found on a distinct set of Dot1L dependent genes in hematopoietic stem and progenitor cells. Finally, pharmacological activation of Sirt1 enhances the efficacy of Dot1L inhibitors against MLL-rearranged leukemia via more rapid and complete silencing of leukemogenic gene expression. These data show that the Dot1L complex is highly regulated in hematopoiesis and in leukemia. A major function of Dot1L is to inhibit Sirt1-mediated acquisition of a heterochromatic state to maintain leukemogenic gene expression, and to maintain specific gene expression programs during normal hematopoietic development. These mechanisms are being exploited therapeutically in ongoing clinical trials. [1, 2] . It is expected that by the end of 2014 the WHO will have agreed upon which of the recommended changes from the CAC meeting will be incorporated into the 2015 WHO classification. These will be discussed. In adults, T-cell acute lymphoblastic leukemias (T-ALL) account for around 25% of all cases. In both pediatric and adult populations, T-ALLs are more frequently associated with a high whiteblood cell count (WBC), a mediastinal tumor mass and a cerebrospinal fluid involvement. Contrarily to what is observed in BCP-ALL, T-ALL are more frequently observed in adolescents and young adults, making patients with T-ALL younger in the whole adult population. Moreover, less heterogeneity in disease characteristics was reported between children and adults, even if unfavorable immature T-ALL seems to be more frequent in adults. 1 Even more than in BCP-ALL, patients with T-ALL benefit from pediatric-inspired approaches with complete remission (CR) rates >90%, relapse incidence around 30%, and approximately 55-60% event-free survival and 65% overall survival at 5 year. In most cooperative group, prognostication in T-ALL relies on initial WBC, phenotype, and response to induction therapy. A dismal prognosis has been reported for pro/pre-T (EGIL T-I and T-II) and mature T (EGIL T-IV) ALL compared to cortical T-ALL. The lack of CD1a and CD8 expression and the weak expression of CD5 define a subset of T-ALL, named "early T-cell precursor" ALL, that was associated with myeloidrelated molecular features and a higher risk of treatment failure. 2 In contrary to BCP-ALL, conventional cytogenetic analysis is uncommonly used for risk stratification in T-ALL. The presence of t(10;14)(q24;q11)/ TLX1 has been associated with a better prognosis, while the negative impact of t(10;11)(p12;q14)/PICALM-MLLT10 (formerly CALM-AF10) has been discussed. Complex karyotype (≥5 abnormalities) found in both BCP-and T-ALL has also been reported as an unfavorable feature. 3 In the last decade, molecular biology, has helped to refine the risk in T-ALL patients. Low expression of ERG and BAALC genes have been associated with a lower relapse risk and a better survival. 4 More recently, mutations of the NOTCH pathway involving NOTCH1 or FBXW7 genes have been shown to confer a favorable prognosis in T-ALL but also T-lymphoblastic lymphoma. 5 This feature may be modulated by the presence of RAS/ PTEN alterations, so that patients with NOTCH1/FBXW7 mutations w/o RAS/PTEN alterations have a particularly favorable outcome. 6 Finally, early response to therapy and quantitative minimal residual disease (MRD) monitoring during and after induction, either by clone-specific IgH/TCR gene rearrangement detection or by flow cytometry, may be used to summarize the risk at the individual patient level. As in many other diseases, these prognosis factors remain often dependent on the protocol used. Open questions concerning this rare disease diagnosed in relatively young patients remain to be addressed: 1) What is the optimal way to combine disease-related prognostic factors and response evaluation to assess the risk of treatment failure?, 7 2) Which patients are good candidates for allogeneic SCT, in other words which of these prognostic factors are good predictors of survival improvement after SCT? 3) Is there any subgroup of patients that may benefit from new therapeutic approaches? Among evaluated drugs, the nucleoside analogue nelarabine, already approved in relapsed patients, is currently being investigated front-line as postremission therapy. Targeting NOTCH1, IL7-R and especially PI3K/AKT/mTOR pathways might be promising therapeutic approaches in this disease. international collaborations validation and standardization of methods is performed: The EuroFlow network established fully standardized 8-color antibody panels as well as tools for bioinformatic assisted expert independent automated analysis of the data acquired. 2;3 Within the EuroMRD network guidelines for standardized real-time quantitative (RQ)-PCR of IG/TR gene rearrangements have been published already several years ago. 4 The network organizes periodic ring trials for external quality control of analysis. For BCR-ABL PCR standardization efforts were initially performed within the European Working Group for Adult Acute Lymphoblastic Leukemia (EWALL), and these efforts were pooled with the EuroMRD experience in 2008, when joined EWALL/EuroMRD meetings started. To date several issues regarding methodology and interpretation guidelines are addressed for generating reliable quantitative data for BCR-ABL transcripts in Ph+ ALL. With the availability of high throughput technologies also next generation sequencing (NGS) emerged as molecular tool for MRD quantification. IG/ TR NGS relies on amplification and high throughput sequencing of IG or TR gene rearrangements using multiplex PCR approaches. One major advantage of this workflow compared to ASO PCR is the applicability of a universal PCR assay without the need of patient-specific reagents. Recent developments offered the proof of principle that NGS-based MRD assessment is feasible and potentially even more sensitive than the standard options. [5] [6] [7] [8] [9] However, several aspects and potential pitfalls have to be adequately considered when MRD is analyzed and reported using NGS approaches. Therefore, also for IG/TR NGS MRD assessment a European consortium formed for setting standards in IG/TR NGS methodology and its applications in hemato-oncology. This EuroClonality-NGS consortium consists of several EuroClonality laboratories (formerly BIOMED-2; euroclonality.org) experienced in design of assays for detecting IG/TR rearrangements, supplemented by laboratories with an expertise in MRD measurement by IG/ TR gene analysis (EuroMRD; euromrd.org), IG/TR repertoire studies and immunoinformatics (European Research Initiative on CLL, ERIC; ericll.org). Within the German Multicenter ALL study Group (GMALL) we analyzed MRD in a total of 580 Ph-negative patients with standard risk (SR, n = 434) and high risk (HR, n = 146) features. 10;11 A complete MRD response after induction 2 and/or consolidation 1 was associated with a comparable clinical benefit irrespective of pretherapeutic risk factors. MRD was the only parameter with significant prognostic impact in multivariate analysis. The biologic differences between SR and HR patients were reflected by the significant different proportion of patients reaching a complete MRD response between SR and HR patients with about 20% point lower rates of MRD negativity in HR patients. MRD persistence after consolidation I identified patients with molecular failure as a new high-risk group. 11 Within the GMALL 07/03 trial, patients with persistent MRD >10 -4 after induction and/or first consolidation were therefore allocated to the MRD-HR group and qualified for allo-SCT. In 47% out of 120 MRD-HR patients, SCT was realized in first CR, with the SCT rate being significantly higher in HR compared with SR patients (71% vs 39%, P < .002). The probability of CCR after 5 years was significantly higher for patients receiving an MRD-directed SCT in first CR compared with those without SCT in first CR (66% vs 12% P < .0001). This also translated into a better OS at 5 years (54% vs 33%, P = .06). 11 Within a phase II trial Blinatumomab was administered to GMALL B-cell precursor ALL patients with MRD persistence or MRD relapse. Sixteen out of 20 evaluable patients (80%) became MRD negative after one cycle of treatment. This translated into a relapse free survival of 66% after a median follow up time of 33 months for those 9 patients who received an allogeneic stem cell transplantation (SCT) after blinatumomab treatment. 12;13 Of the 11 patients, who received no subsequent allo SCT RFS was 60% at a median follow-up time of 31 months. Currently, the confirmatory single-arm, phase 2 BLAST trial evaluates blinatumomab in patients with MRD+ ALL in a larger population. MRD has emerged as powerful and independent predictor of outcome in ALL and has surpassed other widely used and accepted prognostic indicators. Several international consortia aim for optimization and standardization of MRD methodology which is of utmost importance when MRD is used for treatment tailoring. Trials for AML have been continuously available to investigators for nearly 50 years, Since 1988 there has been a protocol for patients generally younger than 60 years, frequently including children, and a protocol for older patients. In general questions in relation to induction and post induction and by risk group have been set. Both tested intensive therapy approaches. Each protocol poses several question which is facilitated by factorial design. Since 1994 a new approach was made available for older patients who were not considered fit for intensive therapy. Similarly, apart from pilot trials all interventions are subject to randomisation. Mylotarg: Recent studies for younger patients were AML15 and 17. For induction AML15 assessed the value of the addition of gemtuzumab ozogamicin (GO: Mylotarg) to induction therapy. This delivered an improvement in overall survival. When assessed by risk group it was clear that at a single dose of 3mg/m 2 did not benefit high risk patients. Since that observation a single dose of 3mg/m 2 was compared to a single dose of 6mg/m 2 in AML17. There was no benefit for 6mg overall, or in the favourable or intermediate risk patients. However 6mg improved survival in high risk patients (n=133) from 18 to 32% at 3 years. Since closure of the 3mg vs 6mg randomisation all favourable risk patients have received GO with induction which has delivered a survival >85%. Where a single dose is better than a fractionated schedule will be tested in the AML19 trial. FLAG-Ida: Flag-Ida is a popular schedule for treatment of relapse, but has never been tested in a randomised trial as initial therapy. This was undertaken in AML15. Although there was no difference in in remission rate when compared with DA or DAE, 90% of the remission were achieved with one course. All patients received a second induction course. It was clear that FLAG-Ida resulted in more myelosuppression in later courses, and more deaths in CR. However the relapse risk was significantly reduced so the disease free survival was significantly improved but overall survival was not. Of interest was the retrospective, non-randomised, observation that 2 courses of FLAG-Ida was the same as 2 DA or DAE inductions plus 2 Ara-C consolidation. The value of the second FLAG-Ida course (which for 90% of patients was a consolidation course) is not clear, and will be a randomised in the AML19 trial. Daunorubicin Dose: A daunorubicin dose of 90mg/m 2 has been suggested as a standard of care, however it has never been compared with the widely used dose of 60mg/m 2 X 3 in the first induction course. In AML17 >1200 patients were randomised. At a median to 2 years follow up, there was no survival difference between the doses, overall, or in any subgroup. The day 60 mortality was higher in the 90mg arm, but thereafter the survival was identical. Consolidation Randomisation: The traditional consolidation in MRC trials was MACE (amsacrine/Ara-C/ etoposide) MidAC (Mitoxantrone/ Ara-C). When compared with conventional high dose Ara-C there was no overall difference. However for high risk patients MACE/MidAC was superior. Ara-C Dose in Consolidation: 3g/m 2 has been the traditional dose in consolidation. However we found that 1.5g/m 2 for 2 consolidation courses was as effective with reduced myelosuppression. Four vs Five Courses: Few studies has defined precisely how many sources are needed in total. We tested a total of 5 courses vs 4 courses. There was no survival difference. Additional of Targeted Treatments: In AML15 the effect of m-TOR inhibition was tested by the addition of everolimus or not. This produced no benefit. In AML15 and 17 patients with a FLT3 mutation were randomised to the FLT3 inhibitor, lestautinib, but this did not improve overall survival, although there was an observed improvement in patients receiving azole anti-fungal prophylaxis. Transplantation in CR1: We developed and validated a risk group which added age/secondary disease/response to course 1/ WBC to cytogenetics. This refined the traditional cytogenetic risk grouping by moving about 20% of intermediate risk to high risk. We found no benefit for the residual intermediate risk of a myeloablative transplant, except for the 8% who had the FLT3+/NPM1-genotypes/ High risk patients had benefit. For patients over 40 years with a matched sibling (but not a MUD) with intermediate risk had a survival benefit after a reduced intensity trandplant. Any benefit in those with high risk is uncertain. Mylotarg: In a similar approach to younger patients, the addition of mylotarg to induction in older patients was tested. A small but significant survival benefit was confirmed. The benefit was not limited to any recognised subgroup. Induction Therapy: In an effort to improve remission DA was compared with Daunorubicin/ Clofarabine (n=806), and the addition of ATRA to DA or ADE (n=616), without showing any benefit. Two Courses vs Three Courses: For patients who had at least a PR following course 1 and were in CR after the second induction course could be randomised to have a third course or not. No benefit was seen by the addition of course 3. We found the immunophenotypic assessment of MRD was the most important independent prognostic factor after entering remission. Maintenance with Azacitidine: Patients in remission after 2 or 3 chemotherapy were randomised to receive 5 days of azacitidine 6-weekly for 9 courses. This was well tolerated. There was some improvement in DFS, but overall survival was not improved. However in patients who had only 2 induction courses or were MRD-ve there was benefit in adding demethylation maintenance. These patients represent an increasing and important subgroup who collaborative group have, until recently, ignored. We developed our "Pick a Winner" program which is a rolling randomised phase 2 design with a futility analysis after approximately 87 events. Low dose Ara-C 20mg b.d. for 10 days is the standard of care control arm. To date the following treatments have been compared: 1)LDAC + tipifarnib/ 2)LDAC + Arsenic trioxide/ 3)low dose Clofarabine/ 4)LDAC + mylotarg/ 5)Sapactabine/ 6)Vosaroxin/ 7)LDAC+ Vosaroxin. /8) LDAC + ganetespib /9) LDAC + tosedostat /10) LDAC + Selinexor /11) LDAC + Lenolidomide /LDAC+ AC220. The aim was to double the remission rate as a surrogate for improving survival. Options 1,2,5,6,and 7 failed at the initial assessment. Options 3, 4and 8 qualified for the full randomisation, having improved CR in the first assessment. Option 3 and 4 however did not improve overall survival, and option 8 is on-going. Options 9, 10, 11 and 12 are recruiting to the first stage. This design is efficient and aims to screen for novel approaches in this difficult patient group. AML is a heterogeneous disease which is relatively uncommon. This presents considerable challenges in developing therapy. The will be no "one size fits all" solution. So factorial design with large recruitment is necessary. Since new treatments are likely to be expensive it is reasonable to assume that large differences will be require to justify use, particularly in older patients. This enable novel designs such as "pick a winner" to rapidly screen novel approaches which may offer benefit or indeed no benefit. Chromosome abnormalities are a common feature of cancer. It is generally believed that there are proto-oncogenes and tumor suppressors in chromosome amplifications and deletions, respectively. However, the role of these alternated regions in leukemiagenesis, is still unclear. Recurring deletions of chromosome 7 and 7q [-7/del(7q)] occur in myelodysplastic syndromes and acute myeloid leukemia (AML) and are associated with poor prognosis. Using RNAi and CRISPR/Cas9 approaches, we show that a~50% reduction in gene dosage of the mixed lineage leukemia 3 (MLL3) gene, located on 7q36.1, cooperates with other events occurring in -7/del(7q) AML to promote leukemogenesis. Mll3 suppression impairs the differentiation of HSPC. Interestingly, restoring the expression of Mll3 releases the differentiation block and impairs the progression of established AML. Further, shRNA library screening identifies additional tumor suppressors in -7/del(7q), whose loss might cooperate with Mll3 deficiency to promote AML. Thus, our in vivo functional analysis identifies tumor suppressors in -7/del(7q) in AML. Besides identifying the individual tumor suppressors, we also studied the functions of large chromosome deletions as a whole. Chromosome 17p12 (11B3 in mice), containing the well-known tumor suppressor TP53, is a common deleted region in human cancer, including lymphoma and leukemia. Here, by mouse genetic engineering, we report that 11B3 heterozygous knockout promotes lymphoma and leukemia genesis in mice. Tumors started with 11B3 one-allele-loss have their p53 mutated or silenced on the second allele, resulting in one allele with large deletion while the other allele with p53 mutation or silenced, which precisely mimics the alterations on Chromosome 17p in the majority of human cancer. However, tumors started with p53 one-allele-loss or mutations only lose the second allele of p53 by UPD. Moreover, 11B3 deficient tumors had significantly shorter latency than those with p53 loss only, indicating there are additional tumor suppression functions in this common deleted region beyond p53 only. Further, we show that another tumor suppressor eif5a in Chromosome 17p cooperates with p53 to suppress tumorigenesis, possibly through promoting apoptosis. Thus, large chromosome deletions promote tumorigenesis more than single tumor suppressor, suggesting synergy between multiple tumor suppressors from the same regions. Taken together, our studies conferred chromosome deletions as driver mutations in tumorigenesis. Our murine AML models recapitulate the pathology and drug response of human disease and therefore would provide a unique tool to understand the pathogenesis and therapeutic susceptibility of human AML with these chromosome alterations. Expression R. Delwel Department of Hematology, Erasmus University Medical Center, Rotterdam, the Netherlands Abstract Acute myeloid leukemia (AML) is the most common form of acute leukemia. Understanding of the molecular-and biological abnormalities is essential for the development of hypothesis driven therapy of AML. Based on large scale gene expression (1) and gene methylation profile studies (2) we identified AML subtypes with unique genetic and epigenetic aberrations and study the biological consequences of these abnormalities. In particular, we discovered a unique mechanism of transformation of AMLs with either an inv(3)(q21q26.2) or a translocation t(3;3)(q21;q26.2) [inv(3)/t(3;3], malignancies associated with a dismal clinical outcome. Applying functional genomics (4C-Sequencing and ChIP-Sequencing) and genomeengineering (TALENS and CRISPR/CAS9), we that a distal enhancer of GATA2 located at chromosome 3q21 had translocated to the leukemia disease gene EVI1, located at 3q26 (3). This unique translocation of the GATA2 enhancer causes ectopic activation of EVI1 in those leukemia cells. This novel mechanism of transactivation of EVI1 in myeloid disorders was also reported by Yamazaki and colleagues in an elegant humanized mouse model. Besides EVI1 overexpression, we also demonstrated that as the result of the elimination of the enhancer from the 3q21 site, GATA2 at the affected allele is not expressed. Functional haplo-insufficiency arising from inactivating mutations in GATA2 zinc finger DNA binding domains or in GATA2 promoter sequences has been well established as an underlying cause of syndromes characterized by predisposition to myeloid leukemia. We therefore hypothesize that mono-allelic expression of GATA2 is an additional oncogenic event in these AMLs with aberrant EVI1 expression. Thus, our data show that structural rearrangements involving single chromosomal repositioning of enhancers can cause deregulation of two unrelated distal genes, with cancer as the outcome. Importantly, we found that the rearranged enhancer near EVI1 had turned into a BET-inhibitor hypersensitive "super-enhancer". Inv(3)/t(3;3) AML cells appear to be highly sensitive to these BET-inhibitors, which cause strong downregulation of EVI1 expression accompanied by a cell cycle block and induction of differentiation and apoptosis. Our studies emphasize that molecular understanding of defective transcriptional control may provide leads for hypothesis driven targeting of cancer. Therefore, Although BET-inhibitors are promising new agents which are included in clinical trials in AML, we propose that understanding of the molecular defects that drive disease development may predict the success of BET-inhibitor response in AML. We recently demonstrated many more aberrations in myeloid malignancies with inv(3)/t(3;3) rearrangements (5) . We showed that 98% of those patients harbor mutations in genes activating RAS/receptor tyrosine kinase (RTK) signaling pathways. Moreover, in 15% of the patients mutations in GATA2 were found. These mutations always occurred in the gene located on nonrearranged chromosome 3q21. Thus, in these patients no wild-type GATA2 was expressed, since the GATA2 gene at the rearranged allele was not expressed as the result of the enhancer-rearrangement. We also found frequent mutations in the gene encoding the transcription factor RUNX1 (12%). In addition, we identified mutations in genes encoding splice factors as well as in epigenetic modifier genes in the inv(3)/t(3;3) cases. The high incidence of activating RAS/RTK signaling mutations may provide a target for a rational treatment strategy in this high-risk patient group. Despite the powerful aspects of immune reactions, most often tumor cells are able to evade immune recognition and destruction. Mechanisms exploited by tumor cells to escape T cell immunity include disruption of antigen presentation, down-regulation of HLA molecules, secretion of immune suppressive cytokines, as well as recruitment of regulatory T cells (T REG ) and myeloid-derived suppressor cells. In the last decade, another powerful immune suppressive mechanism gained much attention: the repressive action of co-inhibitory molecules (CIM) 1 . Activation of T cells is predominantly dependent on both co-stimulatory and co-inhibitory members, including members of the CD28:B7 and TNFR family. The balance between positive and negative co-signals determines the functionality of T cells during immunity and tolerance. In addition to the native role of cosignaling, tumor cells can evade immune control by down-regulating co-stimulatory molecules such as CD80 and CD86, and up-regulating various co-inhibitory ligands such as PD-L1, thereby limiting the therapeutic potential of current immunotherapy against cancer. Standard treatment for hematological cancers includes chemotherapy and radiotherapy, which reduce tumor burden and can induce long-term remission. Moreover, in the past years new therapeutics, including Dasatinib, Rituximab, Bortezomib, Lenalidomide and DNMT inhibitors, have been developed that target tumor cells. However, drug resistance and relapse remain major problems. In addition, cellular immunotherapy is an attractive treatment option to cure hematological malignancies. Such immune cell-based immunotherapies include allogeneic stem cell transplantation (alloSCT), T cell and NK cell adoptive transfer, and vaccination-based approaches using various antigen formulations or dendritic cells (DC). AlloSCT can be regarded as the most powerful immune cell-based immunotherapy. Donor-derived alloreactive T cell responses eradicate the malignant cells upon recognition of polymorphic HLA-presented peptides, known as minor histocompatibility antigens (MiHA). AlloSCT greatly enhanced the cure rate for aggressive hematological cancers, although many patients fail to launch productive immune responses and develop relapses. Moreover, a major drawback of alloSCT is the occurrence of graft-versus-host disease (GVHD), a potentially life-threatening side effect predominantly caused by alloreactive T cells recognizing healthy tissues, notably the skin, liver and gastrointestinal tract. Because hematorestricted MiHA are solely expressed by the redundant patient hematopoietic system and the hematological malignancy, they hold the key to separate GVT from GVHD. Studies by us and others demonstrated that the cellular immunotherapies described above are often hampered by the action of CIM that attenuate tumor-reactive T cell responses, resulting in sub-optimal clinical results. Here, we address the role of CIM in immune evasion by myeloid malignancies, and discuss options to circumvent T cell inhibition without severe adverse effects, which was reviewed in more detail by us previously 2 . Today a variety of CIM, including PD-1 and CTLA-4, have been implicated in suppressing anti-tumor immunity against myeloid malignancies. PD-1: Programmed death 1 (PD-1; CD279) is a member of the B7/CD28 family 3 . It is inducibly expressed on stimulated CD8 + T-cells, CD4 + Tcells, B cells and monocytes, and binds two ligands, PD-L1 (B7-H1; CD274) and PD-L2 (B7-DC; CD273). While PD-L1 is expressed on various non-lymphoid tissues, PD-L2 expression is mainly restricted to APC, like DC and macrophages. Furthermore, multiple tumor types express PD-L1 and its expression is elevated upon IFN-γ exposure. PD-L1 molecules on tumor cells can deliver negative signals towards PD-1-expressing tumor-reactive T cells, thereby inhibiting anti-tumor immunity. Indeed, PD-L1 expression has been associated with poor prognosis in solid tumors 4 . In addition to solid tumors, PD-1/PD-L interactions were shown to be of importance in myeloid malignancies 2 . For instance, long-term persistent murine leukemia cells were shown to sequentially up-regulate PD-L1 and CD80, thereby impairing immune destruction 5 . Upon PD-L1 or CTLA-4 blockade, CTL-mediated lysis of these persistent AML cells was improved. In addition to PD-L1 expression on APCs, PD-1 expression on CD8 + T cells and T REG was also important for tumor persistence of murine AML. Combining PD-L1 blockade with T REG depletion, showed superior efficacy in clearance of AML, due to alleviation of PD-1dependent T REG -mediated suppression 6 . The involvement of PD-1 in alloSCT has been investigated both in mouse models and in the human setting. In a study using a retrovirus-induced CML model, it was demonstrated that tumor-specific T cells can become exhausted 7 . In this model, consisting of PD-1 + tumor-reactive T-cells and PD-L1 + CML cells, exhaustion could be overcome by administration of either PD-L1 antagonistic antibody or PD-1 deficient T cells. In accordance, it was shown that the bulk T cell population from CML patients exhibited increased expression of PD-1 7 . Also, in the alloSCT setting, we observed high PD-1 expression on alloreactive CD8+ T cells that specifically recognize hemato-restricted MiHA in myeloid leukemia patients 8 . Furthermore, proliferation of these PD-1 + MiHA-specific T cells by stimulation with MiHA-loaded DC ex vivo was suboptimal, indicating their dysfunction due to PD-1 signaling. Importantly, upon treatment with anti-PD-1 or anti-PD-L1 blocking antibodies ex vivo proliferation of the MiHA-specific CD8+ T cells was reinvigorated. These and many more studies have led to clinical trials exploring the potency of PD-1 blocking antibodies, putting the PD-1 pathway in the forefront of antitumor therapy. Three antagonistic anti-PD-1 antibodies are currently in advanced clinical trials, i.e. pidilizumab (CT-011), nivolumab (BMS-936,558) and pembrolizumab (MK-3475). Furthermore, several anti-PD-L1 antibodies, BMS-936,559, MEDI4736, MPDL3280A and MSB0010718C, are being investigated in clinical trials. In 2012 exciting reports on the use of the anti-PD-1 nivolumab and the anti-PD-L1 blocking antibody BMS-936,559 in patients with advanced malignancies were published 9, 10 . Response rates upon administration of anti-PD-1 to patients with solid tumors ranged from 18-28%, depending on tumor type. Notably, the therapy was relatively well tolerated 11 . Interestingly, also blocking PD-L1, could induce durable tumor regression with an objective response rate of 6-17%, with prolonged responses of over a year. In 2013, the results were reported for the anti-PD-1 antibody pembroluzimab in melanoma 12 . In this more homogenous patient group, a response rate of 38%, and even 52% in the highest dose, were obtained. As in the previous studies, most responses were durable. These exciting results have encouraged registration of these PD-1 blockers. Currently, studies involving hematological cancers are being initiated. A phase I clinical trial performed with CT-011 in patients with varying hematological malignancies showed clinical responses in 6 out of 17 patients, with few adverse events 13 . Furthermore, several trials are underway either exploring PD-1 blockade as monotherapy or in combination with antitumor vaccines in AML and high-risk MDS. CTLA-4: Cytotoxic T lymphocyte associated antigen-4 (CTLA-4; CD152), was the first identified CIM, and is partly similar to the cosignaling molecule CD28. While CD28 is constitutively expressed on the membrane of naïve T cells, CTLA-4 is primarily localized in intracellular compartments and quickly translocates to the cell membrane upon T cell activation. The inhibitory function of CTLA-4 was revealed in knockout mice, which showed multi-organ T cell infiltration leading to lethal lymphoproliferative disease. Like CD28, CTLA-4 has an extracellular domain containing the MYPPPY binding motif, enabling both receptors to interact with CD80 and CD86 expressed by APC. However, the binding affinity of CTLA-4 for these ligands is higher by a factor 10-100, thus outcompeting CD28 and thereby promoting immune inhibition 14 . In addition, CTLA-4 signaling in immunosuppressive T REG mediates the control of auto-reactive T cells 15 . Therefore, the effect of CTLA-4 interference could either be due to depletion and/or inhibition of T REG . CTLA-4 signaling can attenuate adaptive immune responses in chronic viral infections and cancer. CTLA-4, like PD-1, as such is not a marker of exhausted cells, but elevated levels on tumor-reactive T cells correlated with their dysfunction in patients with metastatic melanoma 16 . Moreover, in various CD80 and CD86-positive tumor models, monotherapy with CTLA-4 blocking antibody resulted in elimination of established tumors and long-lasting antitumor immunity. Interestingly, CTLA-4 also has an influence on the motility of T cells. After addition of a CTLA-4 antagonist in a mouse model, it was shown that T cells exhibited increased motility, indicating that CIM blockade does not only restore cytolytic activity, cytokine secretion and proliferation, but could also enhance T cell migration 17 . Although anti-CTLA-4 treatment works in vivo, CTLA-4 blockade in vitro has not been successful in reversing T cell dysfunction. This can be due to limitations of the in vitro models, as CTLA-4 blockade may exert it's in vivo action via multiple immune mediators (e.g. effector T cells, antibody responses, T REG ). All these preclinical findings have stimulated clinical exploration of anti-CTLA-4 blocking antibodies. At the moment, two blocking antibodies exist, ipilimumab (MDX-010) and tremelimumab (CP-675,206). Most studies have been performed with ipilimumab in melanoma, and in these patients the median overall survival almost doubled 18 . In follow-up studies, the effects of CTLA-4 blockade were consistent 19 and in 2011, the FDA and EMA approved ipilimumab treatment for advanced melanoma, thereby paving the way for further exploration of therapies targeting CIM in cancer. For both blocking antibodies, not all patients gained clinical benefit, and the occurrence of severe adverse toxic effects upon CTLA-4 blockade remains an issue 20 . Experimental and clinical studies have demonstrated that CTLA-4 can also hamper T cell immunity against hematologic cancers in both the autologous and allogeneic setting. This might be due to native expression of CD80 and/or CD86 on hematologic tumor cells, interacting with the CTLA-4 expressing T cells. For instance, in multiple myeloma (MM) patients, CD86 but not CD80 was expressed by tumor cells, while CTLA-4 was up-regulated on T cells, which led to anergy of the tumor-reactive T cells 21 . Similar to these results, T cells from chronic lymphocytic leukemia (CLL) patients responded to anti-CD3 activation by a decrease in CD28 and an increase in CTLA-4 expression, resulting in an inhibitory phenotype 22 . Also AML cells heterogeneously express CD86, but CD80 levels were generally low or absent 8, 23 . In addition to autologous antitumor T cell responses, alloreactive T cell function after alloSCT is also strongly influenced by CIM 2 . The importance of CTLA-4 in modulating allogeneic immune responses has been observed by association of certain CTLA-4 genotypes with overall survival and the incidence of leukemia relapse after alloSCT. Furthermore, it was demonstrated in mice that CTLA-4 blockade shortly after alloSCT increased GVHD 25 . However, when anti-CTLA-4 was administered at later time-points after alloSCT, the GVT effect was boosted without signs of GVHD. In patients, ipilimumab administration at late time-points after alloSCT has been explored in one phase I trial 26 . Following a single infusion of ipilimumab in 29 alloSCT patients with a recurrent or progressive hematological malignancy, 3 clinical responses were observed. Importantly, no induction or exacerbation of clinical GVHD was reported, although similar to other CTLA-4 blockade trials 14% of the patients showed organ-specific immune-related adverse events. The lack of GVHD induction is likely attributed to the median interval of one year between last DLI and ipilimumab administration. This provides a window for antitumor immunotherapy in the post-alloSCT setting and emphasizes the importance of appropriate timing of CIM blockade. It has been recognized that CTLA-4 and PD-1 exert their role in attenuating T cell activation at different physiological locations and moments of the immune response. CTLA-4 is mostly involved in the inhibition of CTL priming in the lymph node, while PD-1 seems to limit T-cell proliferation and function in lymphoid tissues as well as in the periphery, i.e. at the tumor site. Therefore, the effects of concurrent PD-1 and CTLA-4 blockade are of great interest. In a mouse tumor model, it was demonstrated that double-positive CD8 + TIL were more dysfunctional than either single PD-1 or CTLA-4 positive CD8 + T-cells. In addition, double PD-1/CTLA-4 blockade led to reversal of TIL dysfunction and subsequent tumor rejection in the majority of mice 27 . These distinct roles of PD-1 and CTLA-4 warranted combined clinical trials to investigate whether administration of blocking both CIM would have an additive or synergistic clinical effect. Recently, very impressive clinical responses have been reported with combined PD-1 and CTLA-4 blockade in advanced melanoma 28 . These striking clinical responses were associated with grade 3/4 adverse events in 53% of the patients, which generally were reversible and were not more severe than observed with monotherapy. In addition to PD-1 and CTLA-4, other co-inhibitory players were shown to contribute to T cell inhibition and/or exhaustion in hematologic cancers. These include BTLA, TIM-3, LAG-3, CD200R, and PD-1H/VIS-TA, which are reviewed by us in reference 5. Several pre-clinical studies indicate that these CIM are co-expressed on tumor-reactive T cells, but can have non-redundant functions from PD-1. For instance, coexpression of PD-1, BTLA and TIM-3 rendered melanoma-specific CD8 + T cells highly dysfunctional, which could be reversed by combined blockade of all three CIM 29 . Therefore, combinatorial CIM blockade holds great promise in T cell-based cancer immunotherapies. Several therapeutic strategies are being developed to dampen the inhibitory signaling by CIM in order to enhance antitumor T cell immunity. The challenge of interference with immune checkpoints is to boost antitumor reactivity while avoiding systemic toxicity. This can potentially be achieved by combining the alleviation of co-inhibition with other therapeutic options or optimal dosage and timing of antibody administration. Appealing combinations are the simultaneous targeting of multiple CIM, use of co-stimulatory agonists in parallel with CIM antagonist, or incorporation of CIM interference in existing cellular therapies. For example, DC vaccination may be applied together with blocking antibodies against CIM to boost T cell-mediated antitumor immunity. In addition, siRNA-mediated silencing of PD-1 ligands on tumor antigen-loaded DC vaccines or in vivo in the tumor micro environment may be attractive way to exploit CIM interference to boost antitumor immunity 30 . Altogether, CIM play a pivotal role in natural and therapeutic T cell-mediated immunity against hematological cancers. With increasing knowledge of a growing number of CIM, novel mono-and combinatorial treatment options are becoming available. In the end, this can lead to optimized immunotherapy against cancer. Targeting of Ubiquitin Ligases for the Treatment of Hematologic Malignancies B. Levine Ebert Harvard Medical School, Brigham and Women's Hospital, Boston, USA Lenalidomide is a highly effective drug for the treatment of myelodysplastic syndrome (MDS) with del(5q), multiple myeloma, and some additional B cell lymphomas. Lenalidomide binds directly to the CRBN-DDB1 E3 ubiquitin ligase. We have used a combination of quantitative proteomic approaches to demonstrate that lenalidomide acts by a novel mechanism of action for a therapeutic agent: in multiple myeloma cells, lenalidomide increases the binding of two substrates, IKZF1 and IKZF3, to the CRBN substrate adapter; increases the ubiquitination of these substrates; and causes the targeted degradation of IKZF1 and IKZF3, transcription factors that are essential for the differentiation and survival of plasma cells including multiple myeloma cells. In del(5q) MDS, lenalidomide targets CSNK1A1 for destruction. Haploinsufficiency for CSNK1A1, a gene located within the common deleted region on chromosome 5q, results in selective targeting of the del(5q) clone. Using deletion mutants of IKZF3 we identified a 75 amino acid degron in the N-terminal zinc finger domain that is sufficient for lenalidomide-induced degradation. Based on sequence alignment of that region between lenalidomide responding Ikaros proteins IKZF1 and IZKF3 vs. non-responding IKZF2, IKZF4 and IKZF5 we substituted a single amino acid (IKZF3 Q147H ) that prevented binding of IKZF3 to CRBN and conferred resistance to lenalidomide induced degradation. IKZF1 and IKZF3 are essential transcription factors for terminal B cell differentiation, and we found that genetic inactivation of IKZF1 or IKZF3 in lenalidomide-sensitive multiple myeloma cells causes severe growth inhibition. Lung Infections Affecting the Relevance of Minimal Residual Disease in Acute Myeloid Leukaemia D. Grimwade Cancer Genetics Lab, Guy's Hospital, London, United Kingdom Abstract Acute myeloid leukaemia (AML) is highly heterogeneous at the molecular level. Nevertheless, risk stratification used to guide transplant practice in younger adults remains largely based on pretreatment assessment of cytogenetics, an extremely limited panel of molecular genetic markers and morphological assessment of bone marrow blast percentage following induction therapy. However, application of more objective methodology such as multiparameter flow cytometry (MFC) has highlighted the limitations of morphology for reliable determination of remission status. Moreover, there is now overwhelming evidence that detection of submicroscopic levels of leukaemia (i.e. Minimal Residual Disease, MRD) using MFC or molecular-based approaches provides powerful independent prognostic information. Implementation of MRD assessment into clinical practice remains a major challenge, hampered by differences in the assays and preferred analytical methods employed by different laboratories. While this should be addressed through adoption of standardized assays with external quality control, it is clear that the molecular heterogeneity of AML coupled with increasing understanding of its clonal architecture dictates that a "one size fits all" approach to MRD detection in this disease is not feasible. However, with the range of platforms now available there is considerable scope to realistically track treatment response in every patient. The seminal TCGA consortium study that involved sequencing 200 AML genomes, has highlighted the molecular complexity of the disease, with the average case harbouring over 10 mutations 1 . Amongst the 200 AMLs analysed,~200 genes were identified as recurrent mutation targets and over 1000 genes found to be mutated in at least 1 case 1 . The availability of high throughput sequencing technologies has fostered interest in the development of targeted sequencing panels to inform patient management 2,3 ; however, the degree of complexity presents a major challenge to define the individual genetic abnormalities or combinations of markers that provide significant independent prognostic information. Indeed, in children and younger adults decisions regarding allogeneic stem cell transplantation (SCT) still remain largely based on cytogenetic risk group and very few molecular markers (FLT3-ITD, CEBPA and NPM1). There is however increasing evidence that assessment of minimal residual disease (MRD) status using flow cytometry or molecular approaches can provide powerful independent prognostic information and is increasingly being incorporated into algorithms to guide consolidation therapy. Acute promyelocytic leukaemia (APL) has provided a paradigm for MRDdirected therapy (reviewed 4 ), with achievement of molecular remission (CRm) in the bone marrow (BM) incorporated into the standard response criteria in this subset of AML 5 . MRD monitoring used to guide early intervention with arsenic trioxide (ATO) has been reported to improve outcome 6 , particularly in high risk patients and is now considered as a standard of care, as recognised in international disease guidelines 7 . The role of MRD monitoring in other subsets of AML is less well established. Having provided proof of principle in APL, we have investigated the prognostic impact of MRD in the commonest molecular subtype of AML, i.e. characterised by mutations involving the nucleophosmin (NPM1) gene. The mutations involve carboxy-terminal tryptophan residues required for nucleolar localisation and generate a nuclear export signal inducing delocalisation of the protein to the cytoplasm (designated NPM1c mutations) 8 . NPM1c leukaemias are molecularly heterogeneous, with outcome influenced by the pattern of cooperating mutations. This has generated interest in molecular profiling to guide treatment approach, particularly concerning allogeneic SCT in first complete remission (CR1). Patients with NPM1c AML with FLT3-ITD and/or DNMT3A mutations have been associated with poorer outcome and are widely considered candidates for SCT. Conversely, those with the NPM1c/FLT3-ITDneg genotype are no longer routinely transplanted in CR1, due to their overall relatively favourable outcome [9] [10] [11] . Given the molecular heterogeneity, we hypothesised that detection of MRD using RNA-based real time quantitative PCR (RT-qPCR) assays could improve outcome prediction. However, its value has been questioned by recent studies providing evidence that relapse can arise from pre-leukaemic clones 12, 13 . Serial MRD samples were analysed in a cohort of 346 NPM1c patients (median age 50 years) treated on the UK National Cancer Research Institute (NCRI) AML17 trial, with a median follow-up of 35 months 14 . Patients were recruited between April 2009 and June 2012, during which time clinicians were blinded to the PCR results so that their prognostic significance could be reliably established. Diagnostic samples were subject to PCR and sequence analysis to define mutation type. Samples were analysed by RT-qPCR assays using the common forward primer and probe published by Gorello et al 15 , in conjunction with mutationspecific reverse primers, allowing MRD detection in all patients, covering 27 different NPM1 mutations. Assays were confirmed to be mutant specific, with a median sensitivity of 1 in 10 5 . Overall 2,569 follow-up samples (902 BM, 1667 PB) were analysed (median 6 samples/patient). To determine if MRD assessment provided independent prognostic information, targeted sequencing (Haloplex, Agilent Technologies) of 51 genes was undertaken in 223 cases with available diagnostic DNA. The panel included established recurrent mutation targets in NPM1c AML and 14 genes found to be mutated in whole exome sequencing of 27 cases with differing kinetics of relapse or sustained molecular remission. For patients in documented morphological complete remission (CR), early MRD assessment distinguished those at markedly differing risk of relapse and overall survival (OS). For patients with NPM1 mutant transcripts still detectable in peripheral blood (PB) following chemotherapy course 2, cumulative incidence of relapse was 77% at 3 years, compared to 28% in those testing PCR negative (p<0.0001), associated with poorer OS (25% vs 77%, p<0.0001). In multivariable analysis MRD provided the most powerful predictor of disease outcome (p<0.0001), far stronger than the mutational profile as determined by the 51 gene panel, including FLT3-ITD and DNMT3A mutation status. Serial monitoring of paired BM and PB samples showed that analysis of marrow increases MRD detection rate, affording a median 1-log increment in sensitivity, associated with a longer time from molecular conversion to relapse. Detection of MRD beyond consolidation predicted progression in 53 of 64 patients, preceded by a median 0.7-log rise in transcripts per month; 8 patients remain in CR following pre-emptive therapy including SCT or epigenetic therapy and only 3 have not yet relapsed at 2-10 months follow-up. Targeted sequencing analysis at molecular or clinical relapse, showed mutant allele frequencies of genes including DNMT3A and TET2 exceeding the NPM1 mutation in~40% of cases, consistent with presence of pre-leukaemic clones. Moreover, using digital PCR we have shown that the DNMT3A mutation persists following chemotherapy, based on analysis of long-term follow-up samples from patients in ongoing molecular remission of NPM1c AML. Irrespective of the clonal architecture, the NPM1 mutation was found to be a stable marker of AML status, detected at relapse in 69 of 70 cases analysed. In this remaining patient the diagnosis of "relapse" was based purely on morphological grounds with~9% blasts in the bone marrow. However, retrospective targeted sequencing of this sample using the 51 gene panel did not show any disease-associated mutations calling into question the clinical diagnosis of relapse. In conclusion, while there is evidence that NPM1 mutant AML frequently arises on a background of pre-leukaemia, NPM1c RT-qPCR assays track the leukaemic clone with high fidelity, providing the most powerful predictor of relapse risk and therefore merit adoption into routine practice to assess response and inform patient management. The purpose of our work is to functionally characterize tumor cells from patients using suitable preclinical in vivo models, focusing on subpopulations within individual patient's samples and on challenging subpopulations requiring novel therapies. Towards this aim, we used the individualized xenograft mouse model of acute leukemia. Primary patients' acute leukemia cells were transplanted into severely immuno-compromised mice to generate patient-derived xenograft cells (PDX cells). Genetic engineering was performed in PDX cells by lentiviral transduction. Stable expression of transgenes allowed molecular marking of tumor cells for isolation of phenotypic and functional distinct cell subpopulations as well as bioluminescence in vivo imaging. Using these techniques, we could show that tumor cells within a single sample consist of functionally diverse cells in vivo. All samples consisted of a mixture of rapidly and slowly growing cells as well as of drug resistant and drug sensitive tumor cells. Slowly growing tumor cells tended to be more drug resistant in vivo than rapidly growing tumor cells. We conclude that acute leukemia cells from a single patient inherit functionally diverse subclones. Favorable subclones are eliminated by standard polychemotherapy, while challenging subclones require research effort to develop effective subclonedirected treatment options in the future in order to prevent disease relapse. Patients with acute leukemia (AL) show highly different rates of tumor cell clearance upon first line polychemotherapy and AL appear as functionally heterogeneous diseases. Within cells obtained from a single patient, most tumor cells might respond to anti-tumor therapy, while a minority might persist in residual disease and might display major treatment resistance. Functional heterogeneity is a hallmark of AL and the subpopulation with the most dismal characteristics ultimately determines patient's prognosis. The purpose of the present study was to functionally characterize tumor cells from patients with AL in the xenograft mouse model in vivo, focusing on subpopulations within individual patients' samples and on challenging subpopulations requiring novel therapies. Towards this aim, primary tumor cells from children and adults suffering AL, either acute lymphoblastic leukemia or acute myeloid leukemia, were transplanted into severely immunocompromised NSG mice (lacking B-, T-and functional NK-cells) according to published protocols (1). As primary AL cells are mainly reluctant towards in vitro growth, mice were used for cell amplification to generate patient-derived xenograft cells (PDX cells). In both acute lymphoblastic leukemia and acute myeloid leukemia, PDX cells were transduced with lentiviruses to stably introduce transgenes (2, 3) . Positively transduced cells were enriched by flow cytometry using a recombinant fluorochrome as marker (Fig1). Fig1 Construct for stable overexpression of transgenes Expression of fluorochromes and surface tags enabled re-isolation of rare numbers of leukemia cells from bone marrow. Expression of luciferase enabled bioluminescence in vivo imaging for highly sensitive and reliable monitoring of disease development and treatment effects (Fig2) (2) . Bioluminescence in vivo imaging in a mouse harboring transgenic cells from an adult with acute myeloid leukemia We studied the two functional characteristics, spontaneous proliferation and drug sensitivity in vivo. In line with clinical experiences, in vitro data and published in vivo data (4, 5) , AL samples from different patients varied widely regarding the rate of spontaneous proliferation in mice in vivo. In addition, we found that within a sample of individual patients, not all cells equally participated in leukemic growth and a mixture was present consisting in rapidly and slowly growing cells. Major differences were also detectable regarding in vivo drug sensitivity. Differences were present not only between different samples, but also within single samples which displayed major heterogeneity regarding drug sensitivity. In addition, we found that within a single sample, slowly proliferating cells tended to be more resistant compared to rapidly proliferating cells. Taken together, genetic engineering of PDX cells in the individualized xenograft mouse model of AL enabled their detailed functional characterization in vivo. Our preclinical in vivo studies revealed a marked functional heterogeneity not only between samples, but also within each sample. Thus, within the same sample of a single patient, functionally heterogeneous cells exist regarding both spontaneous proliferation in vivo and drug sensitivity in vivo. We conclude that novel therapeutic approaches are required targeting important subpopulations present within samples from patients with AL. Our in vivo techniques enable isolating, characterizing and targeting challenging subclones from patients' AL samples for preclinical studies aiming to prevent disease relapse and to increase the prognosis of patients suffering AL. The relevance of minimal residual disease (MRD) assessment in acute myeloid leukemia (AML) is currently being studied in a number of clinical trials. The most widespread techniques used are multiparameter flow cytometry (MFC) as well as molecular-based approaches. While MFC has limited sensitivity, it is applicable in most patients, whereas molecular MRD assessment is confined to those cases with detectable molecular aberrations. Essentially, the assessment of MRD is thought to result in optimizing treatment strategies and ultimately, improving the outcome of patients. To achieve this, it is important to optimize the current assays and maybe more importantly, to gain further insight into the biology of those cells detected by MRD assays. Here, we highlight some of the current developments in the usage and improvement of MRD detection by flow cytometry. The assessment of minimal residual disease (MRD) has transformed the risk assessment and treatment of acute lymphoblastic leukemia (ALL). In acute myeloid leukemia (AML), there also is a large body of evidence correlating the presence of MRD with outcome [1] . This is true for MRD detection by molecular assays as well as by multiparameter flow cytometry (MFC). We and others have demonstrated that the detection of residual leukemic cells by MFC below the threshold of cytomorphology has adverse prognostic implications on relapse free as well as overall survival beyond current pre-treatment risk classifications, including cytogenetic and molecular aberrations [2, 3] . However, more than other techniques of MRD assessment, MFC relies on manual gating and is therefor prone to variations between laboratories. The most widespread gating strategy is based on the definition of leukemia-associated immunophenotypes (LAIP) at primary diagnosis. During follow-up, cells expressing this phenotype are defined as MRD. However, there is still variation on whether the final MRD value refers to the percentage of CD45+ cells, total nucleated cells or mononuclear cells. Furthermore, some studies include corrections for the degree of dilution by peripheral blood [4] . Additionally, from a more clinical perspective, it is still subject of investigation whether (and even more importantly which) treatment of MRD, i.e. achieving a MRD-conversion from positive into negative will impact patients' prognosis [5] . Studies investigating the prognostic impact of MRD show relapse rates in the MRD negative group of around 40% for younger patients and up to 70% for older patients, highlighting that additional factors play a role [3, 6] . Clinical trials prospectively assessing the value of MRD detection as well as treatments based on the detection of MRD are underway. Bone marrow samples from patients with AML (excluding patients with acute promyelocytic leukemia) were analyzed by flow cytometry after Ficoll density-gradient centrifugation. From 2000 to 2011, MFC analysis was performed using a FACS Calibur flow cytometer (BD Biosciences, San Jose, CA, USA) and from 2011 onwards using a NAVIOS flow cytometer (Beckman Coulter, Brea, CA, USA). To be eligible for further analysis, patients needed to receive intensive induction chemotherapy within the German AMLCG trials AMLCG99, AMLCG2004 and AMLCG2008 or analogous to these study protocols. RFS was calculated from the time patients achieved complete remission (CR) to the time of hematological or extramedullary relapse or death from any cause. OS was calculated from primary diagnosis of AML to the time of death from any cause. SPADE as well as viSNE algorithms were performed as previously described [7, 8] . The Kaplan-Meier estimator and log-rank test were used to analyze survival data. In our first cohort (data obtained on the FACS Calibur flow cytometer), 178 bone marrow samples from patients subsequently achieving CR were available for MRD analysis by MFC during aplasia (typically on day 16-18 during induction therapy). 68 patients (38%) were MRD positive and 110 patients (62%) MRD negative. The 5-year relapse free survival (RFS) rate was 16% for MRD-positive patients and 43% for MRD-negative patients (log-rank test p=0.001). Five-year overall survival (OS) rates were 32% and 55%, respectively (p=0.019). Of note, these analyses were performed by defining the cutoff for MRD positivity at 0.15%. Lowering the cutoff value to 0.1% did not change the observed shorter RFS rate (p=0.005). Additionally, if analyzed as a continuous variable, MRD levels were associated with a significant influence on RFS (p=.004) as well as OS (p=0.028). As expected, age (<60 vs. ≥60 years) and cytogenetic risk group showed a high degree of correlation with survival. In multivariate analysis, MRD status during aplasia remained an independent determinant of RFS (p=0.009). Concerning OS, flow MRD status during aplasia was not an independent predictor of OS in this cohort (p=0.3). Concerning a lowered cutoff of 0.1%, multivariate analysis showed a significant association of MRD positivity during aplasia with RFS (p=0.03). We are in the process of analyzing data obtained from a more comprehensive antigen panel obtained on the NAVIOS flow cytometer. This second cohort was analyzed not only utilizing "traditional" gating strategies, but also by determining the malignant fraction of the total primitive (stem) cell compartment as proposed by Terwijn et al. [9] as well as automated analysis utilizing SPADE [7] and viSNE [8] algorithms. First results of these efforts will be presented at the symposium. Unlike ALL, MRD assays in AML are only slowly making their way into clinical trials and consequently only have little impact on clinical practice to date. This is most likely due to two reasons: (1) AML already is a highly heterogenous disease and there is no clear consensus on the optimal application of thealbeit limitedtreatment options. This is especially true for patients considered to have intermediate risk, where there is considerable debate over the optimal management (allogeneic transplant vs. chemotherapy). Despite numerous studies showing the additional prognostic impact of MRD detection in AML, it is very unlikely that this issue will be resolved easily. More likely, MRD status is going to be an additional variable to be considered when determining post-remission treatment regiments. (2) False-negative and false-positive rates for MRD detection are higher than those seen in ALL. In standardrisk adult ALL, early response with continued MRD-negativity starting on day 11 of induction has been associated with extremely low relapse rates approaching 0%, whereas those patients with persistent MRD almost all experience relapse [10] . For AML, even MRD negative patients show relapse rates that are unsatisfactory and some patients deemed MRD positive will not relapse despite receiving no additional therapy [1, 5] . Some of this could be due to limited sensitivity of the assays used. Indeed, increasing sensitivity of improved assays including nextgeneration sequencing has the inherent appeal of detecting residual disease well below the threshold of current assays. For ALL, high sensitivity of the MRD assay seems to be of clinical benefit, since it enables the treating physician to clearly identify patients at high risk of relapse. For AML, this could be true as well, however formal proof of this concept is lacking to date. Rather it is at least questionable whether there is any clinical benefit in lowering the threshold of MRD positivity to those used in ALL [11, 12] . Therefor, when considering the clinical implications of MRD assessment in acute leukemia, it has become increasingly clear that there are fundamental differences between ALL and AML [11] . In conclusion, since the main obstacle in AML is finding improved treatment approaches as well as optimizing selection of existing therapies, we believe MRD detection can offer significant insight in disease biology (as a starting point to evaluate novel approaches as well as an assay on in vivo chemosensitivity). However, the results from MRD-guided clinical trials are awaited urgently to define the role of MRD in clinical practice. The authors declare that they have no conflict of interest In analogy to normal haematopoietic stem cells leukaemic stem cells (LSC), the origin of disease progression and relapse, are thought to reside in specific niches in the bone marrow microenvironment (BMM), where they are protected from therapy. Targeting the BMM is a novel approach for eradicating LSC. Using a retroviral transduction/transplantation model of BCR-ABL1-induced chronic myelogenous leukaemia (CML) we showed that increased bone remodeling due to parathyroid hormone (PTH), the most potent regulator of bone, led to a 15fold reduction of LSC in CML via release of transforming growth factor (TGF)-β1 from the bone, but the same did not apply to acute myeloid leukaemia (AML). PTH treatment also led to reduced engraftment of human CML cells in immunedeficient mice. In summary, these data suggest that the BMM differs between CML and AML and that targeting of the BMM can lead to reduction of LSC, a requirement for cure of CML.(1) Haematopoietic stem cells, the normal counterparts to LSC, are situated in an HSC niche in the BM which consists of osteolineage cells, sinusoidal endothelial cells, mesenchymal stem cells, adipocytes and sympathetic neurons. Osteoblastic cells support haematopoiesis, have been shown to regulate HSC quiescence (2) and to influence the HSC pool size (3). Specifically, activation of osteoblastic cells in the bone marrow by PTH led to an increased number of HSC (3). Consequently, we tested whether PTH-induced bone remodelling andsignaling could, similarly, affect leukaemo-genesis. We used the well-described retroviral transduction/ transplantation model of BCR-ABL1 + CML(4) and MLL-AF9 + AML (5) to induce leukaemia in mice with osteoblastic cell-specific activation of the receptor for PTH and PTH-related peptide (Col1-caPPR mice, promoter: 2.3 kb fragment of the collagen type I α1 promoter) (6) . Secondary transplantation assays were used to assess LSC frequency and Southern blots were employed to test engraftment of transplanted cells. We used short hairpin-mediated knockdown (in the CML model) or overexpression (in the AML model) of the TGF-β receptor (TGFBR1), as well as immunofluoresence of LSC in AML and CML, in order to assess TGFBR1 signaling via pSMAD2/3. Transplantation of BCR-ABL1 + bone marrow from 5-fluorouracilpretreated syngeneic donor mice into Col1-caPPR mice recipients led to a significant delay in the latency and incidence of CML-like myeloproliferative neoplasia (MPN) compared to wildtype (WT) controls. This was not due to decreased engraftment in Col1-caPPR, as assessed by Southern blotting for retrovirally transduced clones. In fact, the prolonged latency of CML-like MPN in Col1-caPPR mice was due to a reduced number of LSC and a reduced frequency of cycling LSC compared to WT recipients, as tested by flow cytometry and secondary transplantation assays. In contrast, transplantation of bone marrow transduced with the oncogene MLL-AF9 + , known to be associated with AML in humans and mice, into Col1-caPPR recipients led to accelerated AML-like disease and increased disease clonality compared to WT recipients. Administration of an Fc fusion protein of osteoprotegerin, a decoy receptor for the receptor activator of nuclear factor kappa B ligand (RANKL) and, thereby, a potent inhibitor of the bone phenotype in Col1-caPPR mice, led to restoration of the CML-like MPN in Col1-caPPR mice. We hypothesized that increased levels of TGF-β1, known to be released by PTH-stimulated and rapidly turning over bone, may suppress CML cells, which express TGFBR1 at greater levels than MLL-AF9 + AML cells. Indeed, TGF-β1 protein and biologically active TGF-β1 was increased in the rapidly remodeling bones of Col1-caPPR mice. Further, in-vitro treatment of the BCR-ABL1 + cell line K562, but not the MLL-AF9 + cell line THP-1 with TGF-β1 led to suppression of the CML, but not the AML cell line. In order to test, if modulation of TGF-β1 signaling could restore or 'rescue' CML and decelerate AML, we knocked down TGFBR1 by shRNA-expressing lentivirus in the CML model and overexpressed TGFBR1 in AML on the donor bone marrow cotransduced with BCR-ABL1 or MLL-AF9, respectively. Knockdown of TGFBR1 on the CML cells restored the CML like disease and overexpression of TGFBR1 decelerated AML in Col1-caPPR mice. Consistently, in-vitro treatment with TGF-β1 led to a significant increase of pSMAD2/3 (downstream of TGFBR1) in CML LSC, but to no difference in staining intensity in AML LSC. This suggested that MLL-AF9-induced AML is insensitive to TGF-β1 derived from the BMM due to lack of TGFBR1 expression or higher constitutive pSMAD2/3 signaling, whereas TGF-β1 suppresses CML-like MPN. In order to enumerate LSC in a PTH-treated BMM, we transplanted the bone marrow from saline-or PTH-treated WT mice with CML into irradiated, untreated recipients in limiting dilution. This revealed that modification of the BMM by PTH led to a 15fold reduction of LSC in CML. Furthermore, cotreatment of WT mice with CML with imatinib, considered standard of care in CML, plus PTH led to longterm survival in 33% of mice compared to 6% in mice treated with imatinib alone and significantly reduced the number of cycling LSC in CML. Finally, in experiments to extend these studies to clinical samples, we demonstrated that PTH-treatment of immunedeficient NOD-SCID interleukin-2 receptor γc-deficient (NSG) recipients led to significant reduction of human CML cells without affecting the engraftment of normal human cells. These data suggest that modulation of the niche may also be deleterious to human CML progenitors. This study suggests that & a myeloid leukaemia is influenced by the BMM & there is differential regulation of CML versus AML by the BMM & modification of the BMM leads to reduction of LSC. Therapeutic targeting of the BMM may be a complementary approach to conventional therapies. Research support from GlycoMimetics Inc. The introduction of ATRA (all-trans retinoic acid; tretinoin) and ATO (arsenic trioxide) into routine remission induction and post-remission consolidation therapy has dramatically improved the outcomes for patients with acute promyelocytic leukemia (APL) with typical t(15;17) and PML/RARA rearrangements. Following treatment, most patients achieve complete molecular remissions (CRm) with undetectable disease at levels of 10 -5 . Whether there is any clinical benefit from prolonged maintenance therapy for these patients, and if so, whether a high-risk subset can be identified, remains controversial. Patients who fail to achieve a CRm due to drug resistance are not appropriate candidates for maintenance or observation, due to the high likelihood of relapse; these patients require additional rescue chemotherapy and/or stem cell transplantation. The benefit of maintenance therapy is inversely proportional to the antileukemia effect of the preceding induction/consolidation therapy. As initial treatment with ATRA plus anthracycline (with or without cytarabine) or ATO has become more effective, the need for maintenance therapy appears to have diminished. Thus, access to using ATO for frontline treatment regimens may in large part determine the need for maintenance. Large randomized trials evaluating the role of maintenance therapy have been confounded by marked variability in induction and consolidation regimens, as well as by the pretreatment risk characteristics of the cohorts under study. 1 Several smaller studies have included maintenance with an observation only arm. Early randomized trials showed that maintenance therapy with ATRA +/chemotherapy decreased the incidence of relapse and improved the disease-free survival (DFS) when only ATRA or ATRA plus chemotherapy had been used for remission induction and consolidation. 2, 3 The European APL93 trial evaluated 4 different maintenance arms after induction therapy containing ATRA. 3 Compared to observation, ATRA maintenance provided improved DFS at 5 years (74% vs 55%) and lower rates of relapse at 10 years (33% vs 43%). 2, 3 However, other randomized trials failed to show benefit. In the APL97 study by the Japan Adult Leukemia Study Group (JALSG), the addition of 6 courses of intensive maintenance chemotherapy compared to observation only gave significantly worse 6-year DFS (63% vs 80%) and overall survival (OS; 86 vs 99%). 4 The AIDA 0493 trial also did not show a benefit for maintenance after achieving CRm. 5 These studies have generally used ATRA (45 mg/m2 daily) on an intermittent schedule (15 days every 3 months, or 7 days every 2 weeks) with or without mercaptopurine (60 mg/m2 daily) and methotrexate (20 mg/m2 weekly) or lowdose chemotherapy alone. More recent studies have called into question the need for maintenance therapy when ATO is used early in frontline treatment of APL. In a European cooperative group trial, 77 patients with low-intermediate risk APL were randomized to receive 6 cycles of ATRA and ATO, followed by observation only. 6 Their 2-year event-free survival(EFS) was 97% compared to 88% for a control group randomized to induction and consolidation with ATRA plus chemotherapy, followed by ATRA plus low-dose chemotherapy maintenance. There was less hematologic toxicity and fewer infections but more hepatic toxicity observed in the ATRA/ATO arm, and their OS was superior (p=0.02).. In the North American Intergroup noninferiority study (SWOG 0521), 105 patients with low-intermediate risk APL received ATRA plus chemotherapy induction followed by consolidation with 2 courses of ATO and 2 courses of ATRA plus anthracycline. 7 Among 68 CRm patients who agreed to be randomized to observation only or to maintenance with ATRA plus mercaptopurine and methotrexate, no relapses were observed on either arm after a median follow up of 36 months. An earlier North American Intergroup trial (CALGB 9710) that used the same remission induction and consolidation regimen had randomized an initial cohort of 50 patients, including all risk groups, to either observation alone or to maintenance with ATRA alone. 8 There were no significant differences in DFS or OS between these two arms although the power to detect a difference was low because of small numbers of patients. The CALGB 9710 trial had enrolled 481 evaluable adults (15-79 years old) and showed a significant advantage for APL patients who received 2 cycles of ATO as their initial post-remission consolidation after induction with ATRA plus daunorubicin (50 mg/m2 for 4 days) and cytarabine (200 mg/m2 for 7 days). 9 After consolidation, 331 patients still in hematologic CR agreed to be randomized to maintenance therapy for one year either with ATRA alone (45 mg/m2 for 7 days every 2 weeks) or to the same dose and schedule of ATRA plus mercaptopurine (60mg/m2 daily) and methotrexate (20 mg/m2 weekly). After a median follow up time of >6 years, there were no significant differences in DFS or OS between the two maintenance arms. There was a suggestion of improvement in DFS with the 3-drug combination in patients who had not previously received ATO, but there was no improvement in OS. Stratified analyses by risk group showed no significant differences in DFS between the 2 maintenance regimens, either for the 258 patients with low-intermediate risk APL or for the 69 high-risk APL patients. Age, gender, CD56 expression and FLT3-ITD or TKD mutations at diagnosis did not have an impact on outcome by maintenance arm. Although no treatment-related deaths were reported during maintenance therapy in the CALGB 9710 study, hematologic adverse events were more common in the combination ATRA plus chemotherapy arm (maximum grade 3 or 4, 18% vs 4%; p<0.0001), as were non-hematologic adverse events (maximum grade 3 or 4, 36% vs 25%; p=0.03). In the Japanese APL97 study, toxicity was significantly greater during the intensive maintenance chemotherapy, and 2 cases of therapy-related leukemia were reported compared with none on the observation arm. Patients with high-risk APL, defined by initial white blood cell count >10, 000/ul, have a higher incidence of both early death and relapse. Maintenance therapy with tamibarotene, a synthetic retinoid, showed significantly improved relapse-free survival compared with ATRA (both given orally for 14 days every 3 months) in a subset analysis of high-risk APL patients although no benefit was observed overall in a randomized trial by the JALSG. 10 However, no ATRA had been given during consolidation courses in this trial. The role of maintenance therapy for APL patients who achieve CRm after initial therapy that includes both ATRA and ATO is controversial. Relapse is uncommon in patients who received ATO during induction or consolidation. Prospective studies with uniform induction and consolidation in which patients in CRm are randomly assigned to maintenance (with ATRA and/or chemotherapy) vs observation only would be required. Routine maintenance for 1-2 years involves additional toxicities and costs. Patients who received ATO as part of frontline treatment probably do not need maintenance, whereas those who do not have access to ATO for initial treatment may benefit from maintenance with a lower incidence of relapse. The author declares that he has no conflict of interest During the last decades, acute promyelocytic leukemia (APL) has developed from a highly fatal to the best curable type of acute myeloid leukemia. Approximately 80% of patients can currently be cured [1] . However important issues remain: (1) There is still need of further improvement, in particular in patients at high risk for early death (ED) or relapse. (2) The commonly used relapse risk score based on initial white blood cell (WBC) and platelet counts (Sanz score) was developed on the basis of treatment protocols consisting of all-trans retinoic acid (ATRA) and anthracyclines [2] . The applicability of this score to patients treated with other regimens including high dose ara-C or arsenic trioxide (ATO) has not been proven so far. (3) It is of high interest, whether the profile of molecular changes detected in APL blasts provides prognostic parameters, which may reliably identify patients with a high risk of relapse and which are independent of the administered therapy. (4) Up to now, the therapeutic management of patients with relapsed APL has not been standardized. ATO is presently regarded as the treatment of choice for APL relapses after frontline therapy with ATRA and chemotherapy. However, the optimal therapy to sustain ATO-induced 2 nd remission is still unknown. Results of international studies for newly diagnosed APL and registry data may contribute to answer these open questions. As shown by registry results and by data of the German AML Cooperative Group (AMLCG), patients with initially high white blood cell (WBC) counts, poor performance status or older age have an increased risk to die early, either before the start of treatment or during the early phase of the therapy. Most frequent causes of ED were bleeding complications, mainly in the CNS [3] [4] [5] [6] . The ED rate of patients older than 60 years registered by the AMLCG was 48% in non-eligible and 19% in study patients, whereas the ED rate of the study patients below 60 years was only 8% [6, 7] . In both age groups, the early death rate was significantly higher in patients with pretreatment WBC counts over 10,000/μl. Bleeding complications were the cause of ED in 64% of the younger and in 30% of the elderly patients, respectively [6, 7] . Concerning the widely used Sanz relapse risk score, the results of the German AMLCG in newly diagnosed APL could not reproduce the score for relapse free survival (RFS) and cumulative incidence of relapse, when an intensified therapy with double induction including high dose ara-C was administered [7] . Similarly, in a study from the USA reported by Powell et al, additional treatment with ATO consolidation therapy reduced the difference in RFS between patients with initially high or low WBC counts, in comparison to patients treated with conventional ATRA and chemotherapy without ATO [8] . Results from our center showed that the expression profile of several genes including ERG (Ets' Related Gene), BAALC (Brain And Acute Leukemia Cytoplasmatic) and WT1 (Wilms' Tumor 1), which were associated with prognostic importance in other types of leukemia, had prognostic impact in APL patients, who received uniform treatment according to the protocols of the German AMLCG. In particular, low BAALC expression had a favorable prognostic impact, whereas high ERG expression had an independently negative influence on overall survival (OS), RFS and remission duration [9, 10] . In patients, who achieved a CR, low or high WT1 expression was significantly associated with an inferior OS compared to intermediate WT1 expression [11] . Remarkably, an optimized relapse risk stratification model combining the different expression levels of these three markers identified patients with an exceptionally good prognosis regarding the cumulative incidence of relapse [12] . Furthermore, SNP array analysis of APL blasts showed that patients with a higher number of copy number alterations (CNA) had an inferior prognosis [13] . Deletions of the chromosomal subband 1q31.3 were associated with a higher number of CNAs and were a strong independent prognostic factor for an increased risk of relapse [13] . In 2008, a European registry of relapsed APL was established and coordinated at the University Hospital of Mannheim. Of 237 patients registered in eight European countries, documented outcomes of 155 patients in first relapse of APL treated with arsenic trioxide (ATO) are available. Ninety-one percent of patients in hematological relapse (n=104) achieved complete hematological remission (CR), 7% had induction death and 2% resistance; 27% developed differentiation syndrome and 39% leukocytosis, whereas no death or side effects occurred in patients with molecular relapse (n=40). All patients with extramedullary relapse (n=11) entered clinical and molecular CR. Ninety-five patients underwent transplantation (60 autologous, 35 allogeneic). The remaining patients received heterogeneous postconsolidation therapy, most frequently ATO or chemotherapy of variable intensities. The overall survival of the 155 patients at three years was 70% and the event free survival 54% [14] . Guidelines and continuous education to raise the clinical awareness of the immediate actions needed for diagnosis and therapy of APL and rigorous supportive care are important to reduce the rate of ED. Data of the German AMLCG and of other clinical trials indicate that biological factors and the modification of APL therapy have influence on the relapse risk and that scores solely based on blood counts cannot be applied without taking into account the treatment regimen. Therefore, molecular markers may contribute to an improved risk stratification of APL. The results in relapsed APL suggest that at least 50% of the patients in first relapse have a chance to be cured by salvage therapy with ATO followed by individually adapted consolidation therapy. These results indicate the need of further improvement of relapse therapy. Most recently, a series of candidate gene and whole genome studies have identified recurrent somatic mutations in AML patients including TET2, ASXL1, DNMT3A, and EZH2 mutations. We and others have shown these mutations are of prognostic relevance, and can be used to improve risk stratification in AML. We identified genetic predictors of outcome that improved risk stratification in AML independent of age, WBC count, induction dose, and postremission therapy and validated their significance in an independent cohort. Importantly, these mutational predictors involved complex genotypes, suggesting combinations of mutations mark prognostically relevant groups and segregate AML into distinct, biologically significant subsets. Integrating mutational data with dose-intensity revealed that high-dose daunorubicin improved survival in patients with DNMT3A/NPM1 mutations or MLL translocations relative to treatment with standard dose daunorubicin, but not in patients wild-type for these alterations. These data provide important clinical implications of genetic alterations in AML by delineating mutation combination genotypes that predict outcome in AML and improve AML risk stratification. Of biologic importance, the TET family of proteins have been shown to place a hydroxyl mark on methylated DNA and lead to DNA demethylation. We and others have found that TET2 mutations leads to loss of DNA hydroxymethylation and a hypermethylation phenotype in leukemia patients. In addition, in vitro and in vivo studies show that TET2 loss leads to impaired hematopoietic differentiation, increased stem cell self-renewal, and myeloid transformation in vivo. These data demonstrate that novel mutations coopt the epigenetic state of hematopoietic stem/progenitor cells in order to contribute to transformation and that these mutations have biologic and prognostic relevance. Novel FLT3 Inhibitors and Targeted Therapies in AML M. Levis Johns Hopkins University, Baltimore, USA Acute myeloid leukemia (AML) patients who harbor an internal tandem duplication of the FLT3 gene (FLT3/ITD mutation) have a generally worse outcome than their counterparts with wild type FLT3, although this difference can be mitigated somewhat if they undergo allogeneic transplant in first remission [1] . Given that the ITD mutation leads to constitutive activation of the receptor and contribute to the transformed phenotype [2] , clinically useful small molecule tyrosine kinase inhibitors (TKIs) of FLT3 have been a goal of numerous researchers around the world for more than a decade [3] . Initial efforts to develop a FLT3 inhibitor began with re-purposing multitargeted kinase inhibitors, lestaurtinib (CEP-701) and midostaurin (PKC-412), as FLT3 inhibitors [4, 5] . These agents were clinically disappointing, largely because the plasma levels of these drugs required to achieve complete FLT3 inhibition resulted in unacceptable off-target effects [6, 7] . Another re-purposed TKI, sorafenib, is somewhat more selective than previous agents, has in vitro and in vivo activity against FLT3 [8] [9] [10] , and is the agent that has thus far been used to demonstrate a clear clinical benefit of FLT3 inhibition for FLT3/ITD AML patients, particularly in the post-transplant setting [11] [12] [13] . However, because of its multi-targeted nature, sorafenib use in patients is clearly associated with hand-foot syndrome, gastro-intestinal disturbance, and vascular events, rendering it somewhat difficult to use even in the best of circumstances. Quizartinib (AC220) is the first FLT3 TKI to be developed specifically for FLT3-ITD AML. The drug is more potent in vitro and in vivo than any previously developed FLT3 inhibitor [14] [15] [16] , and it's in vivo potency is probably directly attributable to its high degree of selectivity-cKit is the only other major target receptor affected. As monotherapy in patients with relapsed, refractory FLT3-ITD AML, quizartinib induced clearance of peripheral blasts and terminal differentiation of marrow blasts in roughly half of treated patients [17] . Importantly, most of these responses were classified as "CRi" (complete remission with incomplete count recovery) because of the persistent red cell and platelet transfusion dependence in these patients, despite their having a marrow blast percentage of well below 5% [16] . The failure to recover normal hematopoiesis in most patients could have been in part due to inhibition of c-Kit [18] , but more likely was due to the marrow suppressive effects of minimal residual disease. The clinical benefit of these responses, however, was manifest by a large number of patients being successfully bridged to an allogeneic transplant [19] . The importance of proceeding relatively quickly to a transplant was underscored by the emergence of resistance-conferring point mutations within the FLT3 coding sequence in many patients while on quizartinib monotherapy [20] . Quizartinib has now advanced to a pivotal trial in which patients with relapsed/refractory FLT3-ITD AML are randomized to receive either salvage chemotherapy or monotherapy with quizartinib (NCT02039726). These clinical studies have provided additional proof of the benefit of FLT3 inhibition, and the development of resistance mutations offer the best evidence yet of the relative importance of this oncogene in leukemogenesis. However, they highlighted two areas in which the approach could possibly be improved, namely by addressing the resistance mutations, and by avoiding inhibition of c-Kit. Two inhibitors have recently emerged into clinical development with the potential to address these issues. Crenolanib is a TKI originally developed as a PDGFR inhibitor [21] . It is a potent FLT3 inhibitor both in vitro and in vivo, but with relatively less activity against the other type III RTK, c-Kit [22] . In addition, it has activity against the most common type of resistance-conferring mutation, namely that found at residue D835 in the kinase domain. Early clinical studies have confirmed a meaningful response rate induced by crenolanib as monotherapy [23] , and the drug is now being studied in combination with salvage chemotherapy in a randomized trial (Vinay Jain, Arog Pharmaceuticals, personal communication). ASP2215 is a novel small molecule TKI originally developed as an inhibitor of Axl. It has potent activity in vivo against FLT3, including both the kinase domain mutations as well as the so-called gatekeeper mutation at residue F691 [24] . A number of trials with this agent are underway (NCT02014558) or are in active planning, including those in combination with chemotherapy. It seems inevitable that one or more of these agents will eventually make it through the drug approval process. The challenge going forward will be to find the best way to incorporate these TKIs into current AML treatment regimens to maximize the survival for these patients. While using these agents to improve remission rates and to maintain remission after allogeneic transplant seems likely to be the initial approach, combinations with additional targeted agents, specifically those that can overcome the protective effect of the bone marrow microenvironment, may ultimately be the optimal approach. The clinical heterogeneity of myelodysplastic syndromes (MDS) and secondary acute myelogenous leukemia (AML) is a reflection of molecular diversity at multiple levels, including karyotypic defects and combinations of diverse mutations. While only a few of the newly discovered somatic mutations showed impact on clinical outcomes, most of them did not improve diagnostic or prognostic resolution. Apart of discovery of new somatic mutations, and germ line alterations, introduction of next generation sequencing (NGS) has led to recognition of intra-tumor variability and complex clonal architecture of myeloid neoplasms. As a result, ancestral events initiating malignant evolution can be identified. Founding lesions may shape the biology, subsequent clinical course such as succession of secondary events, and be more impactful in terms of the molecular definition of disease subtypes and prognostic utility. In addition, subgrouping of cases according to the ancestral mutations greatly reduces the molecular complexity and allows for clinical comparison between individual defects. While some somatic mutations are exclusively of secondary nature, for others specialized analysis of clonal hierarchy is needed as they can occur as primary or secondary lesions. Using various methods, including serial exome sequencing, analysis of allelic frequencies and averaging of clonal sizes, we have identified the ancestral events and performed comprehensive genotype/phenotype association studies and analyzed clinical correlations. Acute myeloid leukemia (AML) is an aggressive cancer of myeloid precursors in the bone marrow that leads to the accumulation of immature blasts that disrupt normal hematopoiesis [1] . The development of massively parallel next generation DNA sequencing has been extensively applied to the investigation of AML genomes leading to the identification of recurrently mutated genes in AML, most notably by The Cancer Genome Atlas (TCGA) AML group, who reported the analysis of 200 cases of human AML [2] . This study, along with prior molecular genotyping studies of AML, demonstrated that an individual case of AML contains an average of 5 mutations in recurrently mutated genes. This observation raises the key question of how do these multiple mutations accumulate in a single lineage of cells, particularly in the myeloid lineage where the mature cells and their precursors have short lifespans measured in days to weeks. We hypothesized that since most cells in the myeloid development hierarchy are short-lived, then leukemic mutations must be serially acquired in clones of self-renewing, long-lived HSC [3] . Based on this model, the residual HSC population at the time of diagnosis should contain some cells retaining normal function, but harboring mutations in some, but not all, of the genes mutated in the patient-matched AML cells. We would term these cells pre-leukemic HSC that harbor preleukemic mutations. In this study, we investigated this hypothesis through the analysis of residual HSC in cases of de novo AML. We took advantage of cell surface markers (TIM3 and CD99) that we identified to be differentially expressed on leukemic cells compared to normal HSC [4, 5] . Our general approach was to: (1) isolate residual HSC from AML samples taken at diagnosis using fluorescence-activated cell sorting (FACS); (2) conduct exome sequencing of each individual AML case using purified patient-matched T cells as the reference; (3) conduct targeted deep re-sequencing of each mutation in the residual HSC to determine which are pre-leukemic mutations; and (4) genotype single clones of HSC using SNP-specific assays to determine the order of mutation acquisition. This approach allowed us to determine the frequency and genotype of pre-leukemic HSC in de novo AML. We analyzed 19 cases of de novo AML from which we could isolate residual HSC based on differential expression of TIM3 and CD99 [6, 7] . Exome sequencing of these 19 cases identified mutations in many of the common recurrently mutated genes in AML including a mix from most of the major categories including: activated signal transduction mediators, DNA methylation regulators, chromatin modifiers, the cohesin complex, and transcription factors. Targeted deep re-sequencing of these leukemia mutations in the patientmatched residual HSC identified many, but not all, of these mutations in the residual HSC. Specifically, we identified pre-leukemic mutations in TET2, IDH1, IDH2, DNMT3A, RAD21, SMC1A, CTCF, ASXL1, and NPM1. Thus, consistent with our hypothesis, the residual HSC population consisted of some pre-leukemic HSC with some pre-leukemic mutations. From these cases, we explored whether there was enrichment or common patterns of pre-leukemic mutations. We found that preleukemic mutations were more commonly identified in genes regulating the epigenome through DNA methylation, chromatin modifications, or the cohesin complex. Notably, genes involved in activated signal transduction, such as FLT3 and RAS, were absent among the pre-leukemic mutations. Thus, it appears that the development of AML is initiated by mutations in genes that regulate the epigenome in HSC, followed by proliferative mutations that likely occur in downstream progenitors. These observations suggest that the development of AML occurs from pre-leukemic HSC that persist in patients at the time of their diagnosis, a finding that has clinical implications. In particular, clinical experience with AML has demonstrated that most patients are able to enter remission after treatment with induction chemotherapy. However, the majority of such patients will eventually relapse and die from this relapsed disease. Thus, key questions are how do pre-leukemic HSC respond to induction chemotherapy, and is it possible that they can contribute to relapsed disease through the acquisition of novel mutations? To explore these questions, we next investigated the persistence of preleukemic HSC in clinical remission after high dose induction chemotherapy. Out of 4 patients with detectable pre-leukemic HSC at diagnosis, all 4 showed persistence of these pre-leukemic HSC in remission where they gave rise to both lymphoid and myeloid cells, consistent with a normal function. Lastly, we investigated several cases of paired diagnosis and relapsed disease to determine the clonal basis of relapse. From these 3 cases, we detected 3 patterns. In the first, the relapsed disease was clonally identical to the disease at diagnosis indicating a failure to eradicate this dominant diagnostic clone. In the second, a minor subclone from the time of diagnosis was present as the dominant clone at relapse. In the third, a new mutation occurred at relapse suggesting further clonal evolution of a leukemic, or possibly, pre-leukemic clone. Thus, there may be multiple clonal paths to relapse in AML, and even potentially multiple paths in an individual patient. In summary, we have shown that AML develops from the serial acquisition of mutations in self-renewing HSC that we term preleukemic HSC. Moreover, mutation acquisition is not random with patterns indicating that mutations in genes affecting the epigenome occur first, followed by mutations in genes that affect activated signal transduction and other pathways. It remains to be seen if this order is recapitulated universally in AML, or if there are clinical and biological differences if the order is reversed. In addition, we showed that pre-leukemic HSC persist in clinical remission, and that relapsed disease may arise from multiple mechanisms, including the possibility of new mutations occurring in pre-leukemic HSC. Thus, pre-leukemic HSC may constitute a cellular reservoir that must be eradicated in order to cure AML. The author declares that he has no conflict of interest MDS are a group of heterogeneous clonal disorders that mainly affect the elderly (45/100,000 in >70 years old). They are characterized by ineffective production of mature blood cells and have a propensity to evolve to acute myeloid leukemia or bone marrow failure. Patients are classified according to risk-score systems, such as the World Health Organization (WHO) or the International Prognostic Scoring Systems (IPSS and IPSS-R), which are used to adapt therapeutic options to individual patients. It is important to note that, so far, no treatment has been able to alter the natural course of this disease, except for hematopoietic stem cell transplantation, which is not an option for over 90% of MDS patients (due to advanced age and/or lack of a suitable donor). Therefore, there is an urgent need to better understand MDS biology and develop relevant model systems by which we can evaluate the effect of new therapeutic strategies. Growing evidence in mice indicates that a large spectrum of hematological malignancies, including pre-malignant diseases such as myeloproliferative neoplasms (MPN) [1, 7] and myelodysplastic syndromes (MDS) [5] , can be triggered by abnormal functional properties of the bone marrow niche cells. However, the contribution of the niche to human MDS is much less well established. In this report we investigated the complex interplay between hematopoietic cells and their mesenchymal stromal niche cells in lower risk human MDS. Our data revealed a striking dependency of MDS-propagating stem cells on disease-associated MSCs, which is likely mediated by a number of deregulated niche factors involved in intercellular cross-talk. Finally, we could show that MDS hematopoietic cells appear to "educate" their niche environment into a self-supporting one, thus highlighting the existence of a bidirectional cross talk that likely contributes to MDS pathogenesis. Our study is based on xenotransplantation of material from 24 MDS patients classified as follows: Xenotransplantation procedure: MDS CD34+ cells were co-injected with patient-derived bone marrow stromal cells (MSCs) directly into the bone marrow cavity (i.f) of NOD.Cg-PrkdscidIl2rgtm1Wjl/Szj (NSG) or NSGS mice (NSG mice expressing human SCF, IL3 and GM-CSF). The mice were analyzed between 14 and 28 weeks post transplantation. Molecular tracking of primary MDS bone marrow was carried out by whole exome sequencing or targeted next generation sequencing interrogating up to 17 recurrently mutated genes. Mesenchymal stromal cells (MSCs) were used as a germline control. Mutational allelic burdens in xenografted cells were quantified with ultra-deep sequencing or custom pyrosequencing. Large scale lesions were quantified using interphase-FISH or a newly developed PCR assay that quantifies allelic loss at heterozygous short tandem repeat (STR) loci within deleted regions (JC. Jann et al., manuscript in preparation). Using a large number of primary lower risk MDS cases, we demonstrated that co-injection of MDS CD34+ cells with their corresponding MSCs in NSG mice leads to significant and long-term engraftment in over 70% of the cases analyzed (14/20 patients, range hCD45+= 1-22%). Most importantly, this effect was much less prominent when using MSCs derived from healthy age-matched donors. Likewise, and in line with previous reports [2, 3, 4, 6] , injection of CD34+ cells alone only gave rise to very limited engraftment (1/7 patients; hCD45+=3%). Transplanted samples exhibited a clear myeloid bias and significant engraftment of cells with progenitor (CD34+CD38+) and stem cell phenotype (CD34+CD38-) that could be serially transplanted. When compared to NSG recipients, NSGS mice exhibited an augmented human chimerism with readily detectable morphologically dysplastic cells, one of the main clinical features of MDS. Furthermore, molecular analysis of the engrafted cells (in NSG and NSGS strains) confirmed their "diseased" origin as they carried identical lesions to the ones present in the primary bone marrow, and recapitulated the clonal heterogeneity seen in the patients. Using this PDX model, we were further able to demonstrate that diseasepropagating stem cells in lower risk MDS exclusively reside within the lin-CD34+CD38-stem cell fraction. In order to investigate the molecular mechanisms underlying the enhanced ability of MDS MSCs to support MDS propagation, we performed RNA sequencing comparing healthy age-matched MSCs with MDS MSCs. Our data revealed that MDS MSCs displayed an altered differentiation program and deregulated expression of a number of niche factors involved in intercellular communication (secreted factors, extracellular matrix proteins, adhesion molecules, etc). More strikingly, in co-culture experiments using primary patient material, MDS hematopoietic cells, but not their healthy counterparts, were able to instruct healthy MSCs to adopt "MDS-like" features, thus highlighting the existence of a bi-directional cross-talk between hematopoietic MDS cells and their mesenchymal niche counterparts. This may well contribute to the establishment of the "MDS stem cell-niche unit" that promotes disease propagation in vivo. In this study we have identified patient-derived MSCs as a critical functional component of lower risk MDS. Together with MDS stem cells, these patient MSCs form a functional stem cell-niche unit that allows for the propagation of the disease in a xenograft recipient. The striking changed expression in diseased MSCs of genes involved in processes like cytokine-cytokine receptor interaction, cellular adhesion, and ECM remodeling, further suggests that diseased MDS cells alter the function of the normal HSC niche into one that can support the requirement of MDS cells. This unique niche-dependent PDX model of lower risk MDS will now allow us to study the biology of human MDS in vivo and provide a platform for unraveling the molecular basis and putative role for niche cells in the establishment of the clonal dominance observed in this disease. More importantly, our model constitutes a robust platform for preclinical drug testing. The authors declare that they have no conflict of interest We studied 275 adult patients with acute myeloid leukemia (AML) enrolled on the AMLCG-2008 multicenter trial by targeted amplicon sequencing, focusing on a panel of 70 leukemia driver genes. We found a median of 4 mutations per patient and identified patterns of co-occurring and mutually exclusive alterations. Based on the observed variant allele frequencies, we were able to discriminate clonal founder mutations and subclonal variants. Analyses of clinical outcomes are currently ongoing. Recurrent mutations in more than 100 different genes have been identified in adult AML patients by large-scale exome and genome sequencing efforts, 1 but the clinical relevance of most of these alterations has not been defined. Moreover, "deep sequencing" studies revealed that AML patients frequently harbor multiple, genetically related disease subclones, and that the neoplastic cell population may undergo clonal evolution during treatment. 2 It is unclear whether clonal heterogeneity at diagnosis associates with clinical characteristics or outcomes. In order to address these questions, we set out to characterize a relatively large, uniformly treated patient cohort for mutations in known and putative AML driver genes. We studied adult AML patients who received high-dose cytarabine-based induction chemotherapy within the German multicenter AMLCG-2008 trial. The AMLCG-2008 protocol was a randomized phase III trial comparing two different induction chemotherapy strategies: a high-dose cytarabine-containing double induction regimen (standard arm) versus the dose-dense, high-dose cytarabine-containing sequential HAM (sHAM) regimen. From 2009 unteil 2012, 396 patients were enrolled, and 387 patients are evaluable for outcomes. Genomic DNA was isolated from pretreatment blood or bone marrow specimens. Sequence alterations in 70 genes known to be recurrently mutated in AML or other hematologic neoplasms were analyzed by multiplexed amplicon sequencing (Agilent Haloplex; target region, approx. 321 kilobases). Sequencing was performed on an Illumina MiSeq instrument using 2x250bp paired-end reads. Single nucleotide variants and insertions/deletions (InDels) up to approx. 150bp were identified using an in-house custom analysis pipeline that included the VarScan and Pindel tools as variant callers. A variant allele frequency (VAF) threshold of 2% was set for single-nucleotide variant detection, corresponding to heterozygous mutations present in 4% of cells in a specimen. Variants were classified as known/putative driver mutations, variants of unknown significance, or known germline polymorphisms based on published data (including dbSNP, the Catalogue of Somatic Mutations In Cancer [COSMIC] and The Cancer Genome Atlas [TCGA]). In patients with more than one single nucleotide variant, the chi square test was used assess if the observed VAFs, adjusted for ploidy, were compatible with the presence of a single clone. Material for genetic analyses was available for 275 of the 387 participants (71%) enrolled on the AMLC-2008 trial (140 male, 135 female; median age, 56 years; range 18-86 years). Mean sequencing coverage of target regions was approximately 600-fold, and on average, 98.2% of target bases were covered >30-fold. We detected a total of 1043 mutations in 47 genes, including 39 genes mutated in >1 patient. Nine genes (NPM1, FLT3, DNMT3A, NRAS, WT1, IDH2, RUNX1, TET2 and ASXL1) were mutated in >10% of patients (Fig1, dashed line). We found a median of 4 mutations per patient (range: 0-10). Of note, only 1 patient had no detectable mutation and no abnormality on cytogenetic analysis. Patients with Intermediate-risk cytogenetics according to the MRC classification 3 harbored a higher number of driver gene mutations (median, 4) compared to patients with MRC Favorable (median, 2 mutations) or Unfavorable (median, 3 mutations) cytogenetics (P<.001). When analyzing patterns of co-occurring and mutually exclusive mutations, we confirmed well-known associations (e.g., between CEBPA and GATA2 mutations) and identified novel pairs of mutations that frequently occur in combination and, to our knowledge, have not yet been reported in AML (e.g., ASXL1/STAG2, SRSF2/STAG2). These findings may guide functional studies on the molecular mechanisms of leukemogenesis. We found evidence for clonal heterogeneity in 129 (52%) of 248 pts, based on the presence of mutations with significantly (P<.001) different VAFs within the same sample. Our analyses reveal differences in allele frequencies between different AML driver genes. Mutations can be grouped into "early" events that often are present in the founding clone, and "late" events that frequently appear to be restricted to subclones (Fig3). In the entire patient cohort including all cytogenetic risk groups, mutations in three genes were associated with event-free survival (EFS): NPM1-mutated patients had favorable EFS compared to NPM1-wildtype patients, while RUNX1 and TP53 mutations associated with shorter EFS (P=.002 and P<.001, respectively). Targeted sequencing allowed detection of mutations affecting a panel of known and putative AML driver genes in clinical specimens with high sensitivity. Our data from the AMLCG-2008 patient cohort reveal patterns of cooperating gene mutations, and show that the presence of subclonal driver mutations is a frequent event in AML patients. Differentiating between "founding clone" mutations and subclonal mutations that typically occur later in the disease process has implications for choosing targeted therapies aimed at disease eradication. Introduction ALL is the commonest childhood malignancy and a leading cause of cancer related death in young people 1 . While less common in adults, the outcome is poor and improved treatment approaches required in this age group 2 . ALL is comprised of multiple subtypes that historically have been defined by genetic alterations identified by cytogenetic analysis, including aneuploidy (hyperdiploidy and hypodiploidy) and translocations that disrupt key hematopoietic transcription factors, tyrosine kinases and epigenetic regulators (e.g. ETV6-RUNX1, TCF3-PBX1, BCR-ABL1 and rearrangement of MLL) 3 . Such alterations are important determinants of treatment outcomes, but do not explain the full spectrum of genetic alterations or disease behavior. Moreover, with the exception of BCR-ABL1 and MLL-rearranged ALL, these subtypes are less common in adults, and the genetic basis of ALL in this age group has until recently been poorly understood. 2 Genetic subtypes of ALL Genomic profiling, initially using microarray analysis of gene expression and DNA copy number alterations, and more recently, next generation sequencing, have identified multiple new subtypes of ALL of clinical importance, and have shown that in the majority of patients, ALL is a polygenic disease, with mutations in at least 4 key cellular pathways observed in the majority of cases 4 . While the genes involved, and the mechanism of alteration (deletion/amplification, mutation, or translocation) varies between subtypes, these are important common themes. These include lymphoid development; cytokine receptor, kinase and Ras signaling; cell cycle regulation and tumor suppression; and epigenetic regulation. Ph-like ALL is characterized by a gene expression profile similar to BCR-ABL1 (Ph+) ALL, absence of the BCR-ABL1 oncogene, genetic alteration of lymphoid transcription factors such as IKZF1 (IKAROS), and poor outcome [5] [6] [7] . The frequency of Ph-like ALL rises with increasing age, from 10% in children to over 25% in young adults, with the prevalence in older adults to be defined 8 . Genome sequencing has shown that Ph-like ALL cases harbor alterations activating several cytokine receptor and tyrosine kinase signaling pathways. These include rearrangements of "ABL1-class" kinases ABL1, ABL2, CSF1R and PDGFRB; rearrangements of CRLF2 (often with mutations of JAK1/JAK2) 9,10 , rearrangements of JAK2/EPOR, mutations of JAK-STAT signaling genes (IL7R, SH2B3, and others) and less common alterations targeting Ras signaling and tyrosine kinases 8, 11 . Importantly, the majority of alterations result in activation of signaling pathways sensitive to ABL1 inhibitors (imatinib, dasatinib et al) and JAK2 inhibitors (e.g. ruxolitinib), and anecdotal data has shown efficacy of tyrosine kinase inhibitors (TKI) in otherwise treatment resistant Ph-like ALL 8, 12, 13 . The activity of TKIs will soon be tested in prospective clinical trials of Ph-like ALL in children, and is a priority in adults where this entity is more common. CRLF2-rearranged ALL. CRLF2 encodes cytokine receptor-like factor 2, which with the interleukin 7 receptor alpha chain (IL7RA) forms the receptor for the cytokine thymic stromal lymphopoietin (TSLP). CRLF2 is nor normally expressed in B lymphoid cells, but is deregulated by rearrangement to IGH or by a deletion upstream of the CRLF2 locus that results in P2RY8-CRLF2 fusion. Both result in overexpression of CRLF2 detectable by flow cytometry. CRLF2 rearrangement is present in approximately half of Ph-like ALL cases but is also present in non-Ph-like B-ALL, particularly Down syndrome-associated B-ALL. 9, 10 The majority of CRLF2-rearranged cases have concomitant genetic changes activating JAK-STAT signaling, particularly JAK1/2 mutations, but also JAK3, IL7R and FLT3 mutations. CRLF2-rearranged leukemic cells typically exhibit activated JAK-STAT signaling that is inhibited by JAK inhibitors 14, 15 , but the efficacy of these agents in vivo is presently unknown. ERG deregulated ALL comprises 5-7% of childhood and adult ALL and is characterized by genomic alterations perturbing expression of ERG, a transcription factor. Identification of this subtype is important as such cases commonly have alteration of IKZF1 (IKAROS), which in the context of Ph+ or Ph-like ALL is a poor prognostic feature, but here is not associated with adverse outcome 16, 17 . Hypodiploid ALL comprises up to 5% of ALL and is characterize by marked aneuploidy. Genome sequencing has defined the genomic basis of this entity with several clinical implications 18 . Near haploid ALL cases have 24-31 chromosomes, an exceptionally high frequency of activating Ras pathway mutations, particularly NF1 (otherwise rare in ALL) and mutations of IKZF3 (AIOLOS). Low hypodiploid cases have less severe aneuploidy (32-39 chromosomes), alteration of IKZF2 (HELIOS), and near universal alteration of TP53 which is inherited in over half of low hypodiploid cases. Thus, TP53 mutational testing is warranted in hypodiploid ALL, particularly low hypodiploid cases in children. The majority of cases show activation of PI3K and Ras-Raf-MEK-ERK signaling pathways, and sensitivity to PI3K inhibitors and BCL-2 inhibitors 18 . A key goal of genomic profiling studies is the identification of genomic features that predict treatment success or failure in ALL. These can be grouped into three categories: (1) subtypes, (2) genomic alterations identified at diagnosis; and (3) acquired alterations that facilitate resistance to specific therapeutic agents and promote relapse. Key subtypes associated with poor outcome in ALL include Ph+, MLL-rearranged and hypodiploid ALL have long been recognized to be associated with poor outcome. In addition, Ph-like ALL and ALL with intrachromosomal amplification of chromosome 21 (iAMP21) are also associated with poor outcome. Detection of these entities is thus important at diagnosis, although the biologic basis of poor treatment outcome remains poorly understood (with the exception of Ph+ ALL, which is in part due to alteration of IKZF1. Specific genetic alterations transcend these key subtypes and influence outcome in ALL. The most important of these is IKZF1 alterations. IKZF1 encodes IKAROS, a transcription factor required for the development of all lymphoid lineages. Genetic alterations of IKZF1 are present in up to 25% of B-ALL cases, and include deletions that result in loss of IKAROS activity, and focal deletions and sequence mutations that result in the expression of dominant negative alleles. These alterations result in the acquisition of a primitive, stem cell like phenotype and abnormal bone marrow stromal adhesion and resistance to therapy. Genetic alterations and relapse in ALL Several classes of genetic alteration are acquired and promote resistance to therapy in ALL. These include deletions or mutations of CREBBP (encoding CREB-binding protein), NT5C2 and epigenetic regulators 19 . CREBBP is a histone and non-histone acetyl transferase that in part mediates the transcriptional response to glucocorticoids 20 , a mainstay of ALL therapy. CREBBP mutations result in resistance in glucocorticoids, thus strategies to inhibit histone deacetylation are being explored to reverse this phenotype. NT5C2 encodes a 5' nucleotidase that regulates metabolism of nucleosides and mutations in this gene confer resistance to thiopurines used in ALL therapy. NT5C2 mutations are rarely, if ever observed at diagnosis, but in contrast are acquired during the course of therapy 21, 22 . Thus, the detection of these mutations may be a harbinger of relapse and may prompt implementation of alternative treatment strategies. Additional pathways targeted by genetic alteration at relapse include Ras and JAK-STAT signaling and epigenetic regulation. Mutation of epigenetic regulators is increasingly recognized in ALL. Such mutations include a diverse range of histone writers, readers and erasers 23, 24 . Although there are currently limited data regarding the manner in which such mutations contribute to leukemogenesis and treatment resistance, this is an active area of research and a key area of potential therapeutic intervention. An important observation is that there is substantial clonal evolution and mutation turnover during therapy in ALL, and that relapse commonly arises from a minor clone at diagnosis that shares only partial identity with the predominant clone at diagnosis, but that acquires mutations that facilitate resistance to therapy and relapse. Identification of these low level clones, and the mutations that facilitate relapse, represents an important challenge and opportunity to identify patients at high risk of relapse who may benefit from alternative therapeutic approaches. The author declares that he has no conflict of interest The Alliance for Clinical Trials in Oncology has an extensive portfolio of clinical and translational trials in acute myeloid leukemia. In the last few years, Alliance clinical trials have been focused on novel induction strategies based on cytogenetic or molecular risk groups, induction strategies for the elderly, acute promyelocytic leukemia and post-remission maintenance. The leukemia correlative sciences team has utilized nextgeneration sequencing and other novel technologies to molecularly characterize thousands of bone marrow and peripheral blood specimens obtained in conjunction with Alliance trials. These efforts have lead to the development of highly enriched prognostic stratification systems and provide significant insights into the biology of AML. The objective of this abstract is to briefly review the most important clinical and basic science publications during the last year based on Alliance trials, as well as to summarize recently completed and ongoing trials for all phases of AML treatment. The Alliance has been a world leader in the molecular characterization and sub-classification of cytogenetically normal AML, the largest and most heterogeneous subgroup of the disease. Marcucci et al. have identified novel, prognostically relevant AML subgroups using a seven-gene score based on epigenetic and genetic information. [1] Using nextgeneration sequencing analysis of methylated DNA, differentially methylated regions (DMRs) associated with prognostic mutations were identified in 134 older (≥60 years) adults with cytogenetically normal AML. Seven genes were identified which had promoter DMRs and expression level associated with overall survival, including CD34, RHOC, SCRn1, F2RL1, FAM92A1, MIR155HG, and VWA8. For all of these genes, lower expression and higher DNA methylation were associated with longer overall survival. The authors determined that patients who had high expression of only zero or one of these genes had excellent rates of complete remission (CR, 91% and 100%, respectively) and overall survival (82% and 76%), compared to very poor outcomes in patients with six or seven highly expressed genes (CR 71%, 53%; OS 24%, 5%; for 6 and 7 genes, respectively). This seven-gene score performed very favorably compared to other published models and it is hoped that a practical quantitative assay can be developed to facilitate rapid integration into clinical practice. [1] Data from Alliance AML trials were used as a validation cohort for a combined molecular and clinical prognostic model for relapse and survival in cytogenetically normal AML patients of all ages. [2] Pastore et al. used data from 572 cytogenetically normal AML patients treated on the German AML Cooperative Group 99 (AMLCG99) study comparing thioguanine, cytarabine, and daunorubicin plus high-dose cytarabine and mitoxantrone (HAM) versus HAMHAM as induction therapy were used for. Validated prognostic indices for relapse and survival were developed based on age, performance status, white blood cell count and mutation status of NPM1, CEBPA and FLT3-internal tandem duplication (ITD). These data are readily available to clinicians and application of this schema should be straightforward. The authors conclude that theirs is the first prognostic system to combine clinical and molecular features specifically for cytogenetically normal AML and irrespective of age. [2] Metzeler et al. studied 364 patients (ages 18-83) with cytogenetically normal AML treated on Alliance trials in an effort to determine whether a previously identified stem cell-like gene expression signature is independently associated with inferior outcomes. [3] It is believed that AML is derived from a distinct, quiescent, chemotherapy-resistant population of leukemia stem cells (LSCs). Eppert et al. compared gene expression profiles between functionally defined LSCs and leukemic cell populations without stem cell activity and identified a "core enriched (CE)" signature with 44 stem-cell associated genes that are highly expressed in LSCs. In this work, patients with cytogenetically normal AML who had an increased CE signature had worse overall outcomes. [4, 3] These data were confirmed using the Alliance cohort. Furthermore, it was shown that a high CE score was an independent predictor of decreased rates of complete remission, disease-free survival and overall survival even after adjusting for standardly accepted clinical and molecular prognostic indicators. Future studies to identify the functional role of LSC signature genes and to investigate the relationships between CE signature and other molecular abnormalities in AML are required. [3] Niederwieser et al. investigated the clinical significance of the DNA methyltransferase DNMT3B using samples from 210 older cytogenetically normal AML patients (age 60-83) treated with intensive chemotherapy on Alliance trials. [5] Patients with high expression of DNMT3B had significantly inferior rates of complete remission, disease-free survival and overall survival. Also, DNMT3B expression was an independent predictor or poor outcomes in multivariate analyses. Further studies are ongoing to investigate the clinical and therapeutic relevance of this finding. [5] The prognostic significance of GAS6 expression was also investigated in 270 older AML patients with cytogenetically normal AML treated with intensive chemotherapy on Alliance clinical trials. [6] GAS6 is the common ligand for all three TAM receptor tyrosine kinases (TYRO3, AXL and MERTK) and binds AXL with high affinity. AXL expression has been reported as a poor prognostic indicator in cytogenetically normal AML and it is possible that aberrant autocrine expression of GAS6 may result in activation of AXL. [7] Whitman et al. demonstrated that GAS6 expression is an adverse prognostic indicator in cytogenetically normal AML and suggest development of targeted treatments against GAS6/AXL signaling. [6] Although patients with normal cytogenetics comprise the largest and best characterized subgroup in AML, chromosomal abnormalities are common and extremely important for prognosis. [9] Outcomes were poor, with 64% CR and 15% overall survival at 5 years. There were no significant differences in CR, DFS or OS between younger and older patients with sole +8. The abnormality was found to be molecularly heterogeneous, with frequent mutations in RUNX1, ASXL1, IDH2, DNMT3A, NPM1 and FLT3-ITD. High BAALC expression and FLT3-ITD were associated with inferior outcomes in younger patients and TET2 mutations and wild-type RAS resulted in worse outcomes for older patients with isolated +8. [9] Patients with acute promyelocytic leukemia (APL) enjoy excellent outcomes, made even better by the use of arsenic trioxide as part of up-front treatment of newly diagnosed patients. In previously untreated patients with low or intermediate risk APL, it is possible that treatment with arsenic trioxide during consolidation (CALGB 9710) may eliminate the need for maintenance chemotherapy. [10] A non-inferiority trial, terminated due to poor accrual, demonstrated no relapses in either group of 105 APL patients randomized to maintenance or observation. In another follow-up study of Alliance 9710, 48% of 245 newly diagnosed adult patients with APL were found to have mutations in FLT3, 31% FLT3-ITD and 14% with a point mutation (FLT3-D835). [11] Neither type of mutation affected rates of CR, induction death, DFS or overall survival and treatment with arsenic trioxide during consolidation improved outcomes regardless of mutational status. However, even after treatment with arsenic trioxide, APL patients with complex karyotype had significantly decreased overall survival. [11] Treatment outcomes for older patients with AML remain dismal worldwide and novel strategies are clearly needed. Alliance study 10502 investigated the addition of the proteasome-inhibitor bortezomib to standard daunorubicin/cytarabine induction and to intermediate-dose cytarabine consolidation in 95 newly diagnosed patients with AML age 60-75 years. [12] Bortezomib inhibits the master transcription factor nuclear factor kappa-B (NFκB), which is increased in AML cells and LSCs. [13] Although bortezomib has virtually no activity as a single agent in AML, it showed preliminary efficacy when combined with chemotherapy in patients with relapsed disease. [14] In newly diagnosed patients, the combination of bortezomib and standard chemotherapy was feasible, had an acceptable toxicity profile and resulted in an encouraging rate of complete remission (65%). Further investigation of this regimen in older adults with AML is warranted. The role of bortezomib in AML induction was explored in Alliance trial 11002, in which newly diagnosed older AML patients were randomized between ten days of decitabine with and without the addition of bortezomib. The trial was designed based on favorable clinical outcomes of the ten-day induction schedule from three single-center studies and previously demonstrated clinical and pharmacodynamics activity for the combination. [15] [16] [17] [18] CALGB 11002 also included comprehensive geriatric assessments and extensive correlative scientific studies. Accrual has been completed and data are anticipated imminently. CALGB 11002 ran concurrently with trial 11001, a phase II trial of conventional daunorubicin and cytarabine combined with sorafenib for newly diagnosed AML patients ≥60 years with FLT3 mutations. Preliminary results from this single-arm, multicenter phase II trial were presented at the 2013 American Society of Hematology Annual Meeting and showed CR or complete remission with incomplete platelet recovery (CRi) in 69% of 52 evaluable patients. There was no excess toxicity attributed to sorafenib. [19] Interestingly, a similar trial conducted in Europe failed to show a benefit for this regimen in older AML patients. [20] Conventional intensive chemotherapy induction is often withheld from older patients due to well-justified morbidity and mortality concerns. However, although the median age at diagnosis of AML is over 70 years, it is well-known that clinical trials include few patients with very advanced age. Investigators from the Alliance, German AML Cooperative Group (AMLCG), German-Austrian AML Study Group (AMLSG), and the Acute Leukemia French Association (ALFA) pooled data from several large studies of older patients with AML treated with intensive induction and identified 138 octagenarians. [21] The majority of the patients had a performance status of zero or one (69%) and normal cytogenetics (47.1%). Complex karyotype was identified in 14.5% of patients and isolated trisomy 8 in 9.4%. The complete response rate for patients surviving beyond 60 days was 39%. Overall survival at 1 year was 24%, with only 8% surviving for 3 years. FLT-3 internal tandem duplication was not associated with overall survival, while NPM1 mutations were associated with significantly longer survival. [21] AML is a highly molecularly heterogeneous disease and efforts are underway to adapt induction strategies to specific molecular subtypes. CALGB 10801 was a phase II study of induction (daunorubicin/ cytarabine) and consolidation (high-dose cytarabine) chemotherapy plus the tyrosine kinase inhibitor dasatinib, followed by continuation therapy with dasatinib alone in newly diagnosed patients with core binding factor AML. Diagnosis of core binding factor AML was defined by the presence of t(8;21)(q22;q22) or the molecular equivalent RUNX1/RUNX1T1 fusion transcript or inv(16)(p13q22) or t(16;16)(p13;q22) or the molecular equivalent CBFB/MYH11 fusion transcript. Preliminary data are highly anticipated and will be presented at the 2014 ASH meeting. Data are also expected from CALGB 10603, an international, randomized phase III study of induction (daunorubicin/cytarabine) and consolidation (highdose cytarabine) chemotherapy combined with midostaurin or placebo in treatment-naive patients with FLT3 mutated AML. The role of maintenance therapy in AML remains unclear. Although it is well-known that substantial residual disease remains after AML therapy and it is intuitive that ongoing treatment could eliminate or suppress this population of residual cells over time, clinical trials of chemotherapy and immunotherapy have failed, to date, to demonstrate an improvement in overall survival. Several trials have shown improvements in relapse-free survival, but the regimens have not been adopted into standard practice. [22] [23] [24] CALGB 10503 was a phase II study of maintenance therapy with decitabine following standard induction and cytogenetic risk-adapted intensification in previously untreated patients with AML < 60 years. The goals were to determine the efficacy, feasibility, and toxicities associated with one year of maintenance therapy with decitabine administered to patients (ages ≥ 15 and < 60 years) with untreated AML and to assess the 1year disease free survival rate after decitabine maintenance. The trial has been difficult to evaluate due to patient refusals and noncompliance. Similarly, the efficacy of immunotherapy with IL-2 as a post-remission maintenance therapy for younger patients with newly diagnosed AML (CALGB 19808) could not be assessed due to patient and physician refusals to proceed with the planned protocol-directed therapy. [25] Conclusion Despite the enormous efforts of AML patients and investigators, it is painfully clear that therapeutic strategies in AML still lag far behind the substantial breakthroughs in understanding disease pathobiology. Major goals for the next decade should include creative incentives to encourage patients and physicians to participate in AML clinical trials; faster incorporation of novel agents into clinical trials of newly diagnosed AML patients; streamlining arcane and bureaucratic processes in academic medicine, industry and government that impede new trials; and the rapid development of real-time, feasible assays to allow direct application of novel biology-based prognostic scoring systems into clinical practice. Perhaps most importantly, we should strive for increased cooperation among and collaboration between national and international cooperative groups to limit clinical trials in unselected AML patient populations and facilitate rapid accrual of distinct, biologically defined subgroups of AML patients onto rationally designed protocols. Background Sorafenib is a multi-kinase inhibitor with activity against several oncogenic kinases that may play a role in the pathogenesis of acute myeloid leukemia (AML). We present the results of the randomized placebocontrolled SORAML trial testing sorafenib versus placebo as add-on to standard induction and consolidation treatment in AML patients ≤60 years. 276 patients from 25 centers were enrolled in the SORAML trial (NCT00893373). The main eligibility criteria were newly diagnosed AML, age from 18 to 60 years and suitability for intensive therapy. The treatment plan for all patients included two cycles of induction with DA, followed by three cycles of high-dose cytarabine. At study inclusion, patients were randomized to receive either sorafenib (800 mg/day) or placebo as add-on to standard treatment in a double blinded fashion. Study medication was given on days 10-19 of DA I+II or HAM, from day 8 of each consolidation until 3 days before the start of the next consolidation and as maintenance for 12 months after the end of consolidation. The primary endpoint of the trial was event-free survival (EFS). 267 patients received study treatment, 134 in the sorafenib arm and 133 in the placebo arm. Demographic and disease characteristics were equally distributed between the two arms; the incidence of FLT3-ITD was 17%. After a median observation time of 36 months, the median EFS was 9.2 months in the placebo arm and 20.5 months in the sorafenib arm, corresponding to a 3-year EFS of 22% versus 40% (p=0.013). Median RFS after standard treatment plus placebo was 23 months and not yet reached after sorafenib treatment, corresponding to a 3-year RFS of 38% and 56%, respectively (p=0.017). The median OS had not been reached in either arm; the 3-year OS was 56% with placebo versus 63% with sorafenib (p=0.382). in favor of sorafenib was observed. The most common reported AEs Grade ≥3 were fever (40%), infections (22%) and bleeding events (2%). The risk for fever, bleeding events and handfoot syndrome was significantly higher in the sorafenib arm while the incidence of all other AEs showed no significant differences. In younger AML patients, the addition of sorafenib to standard chemotherapy in a sequential manner is feasible and associated with antileukemic efficacy. We observed a higher incidence of infections and bleeding events under sorafenib. Whereas OS in both treatment arms was similar, sorafenib resulted in a significantly prolonged EFS and RFS. The authors declare that they have no conflict of interest The detection of minimal residual disease (MRD) has become part of the state-of-the-art diagnostics to guide treatment in pediatric acute lymphoblastic leukemia (ALL). This applies to the treatment of de novo and recurrent ALL. In high risk ALL, MRD detection is considered an important tool to adjust therapy, in some patients to adjust the optimal time for hematopoetic stem cell transplantation. Precise quantification and quality control is instrumental to avoid false treatment assignment. MRD detection may be used to modulate treatment intensity once it has been demonstrated at well-defined informative checkpoints that certain levels of MRD can reliably predict the risk of relapse. In addition, MRD is used as endpoint to determine the activity of a given agent or treatment protocol. If activity translates into antileukemic efficacy, MRD may be considered a surrogate clinical endpoint. More recently, the prognostic power of MRD detection at given time points was demonstrated in novel genetic subtypes of ALL which may comprise a significant proportion of disease recurrences. This observation illustrates again that so far no genetic subtype with a uniform response pattern to contemporary treatment schedules has been described. In contrast, all subtypes display heterogeneity in treatment response and outcome. In vivo sensitivity of acute lymphoblastic leukemia (ALL) as measured by the early blast cell reduction in peripheral blood (PB) or bone marrow (BM) after exposure to one or several antileukemic agents is used to risk-stratify patients with ALL because response is of high prognostic relevance. [24, 35] Lack of adequate response, in particular at the end of remission induction, indicates poor prognosis even though this may vary significantly according to individual patient characteristics. [34] The choice of technique for MRD detection mainly depends on the aims of the clinical trial and on the availability of resources. [6, 15, 39] [3, 5, 11] It is now widely acknowledged that MRD detection is part of state-of-theart diagnostics and needed in the management of ALL. More importantly, MRD detection may even replace other prognostic factors. [4, 8, 18, 33] Prognostic information based on response to treatment In the largest study for de novo ALL published so far, the BFM (Germany, Austria, Switzerland) and AIEOP (Italy) study groups utilized MRD (measured by RQ-PCR) for risk stratification in a total of 3184 pcB-ALL and 464 T-ALL patients. All patients were treated by identical chemotherapy in the first 9 weeks of therapy. Large differences in pEFS between MRD defined subgroups were found. [8, 33] MRD was analyzed in the BM at end of induction (day 33), and at end of induction consolidation (day 78). Importantly, within pcB-ALL, the prognostic impact of MRD was maintained even in the two large subgroups of TEL/AML1 positive and hyperdiploid ALL. [8] The AIEOP-BFM ALL 2000 trial confirmed the strong prognostic impact of MRD at end of induction when combined with a subsequent timepoint. This trial also demonstrated that in pcB-ALL the MRD-intermediate risk group (MRD-IR) comprises the largest group of patients (51.7%), but also the majority of relapses: 69% of all relapses were found in this group which showed an overall relapse rate of 21% at five years. [8] The situation in T-cell ALL was similar: 62.9% of all patients were defined as IR, comprising 55% of all relapses (cumulative incidence of relapse 17.6% at 5 years). [33] This observation illustrates the need for more refined approaches to identify the patient at high risk to relapse. Recent work of several study groups has identified intriguing properties in subsets of pcB-ALL which in part appear to show prognostic impact independent of MRD. This may be clinically quite relevant as the distribution of relapses as shown before requires further refinement in risk assessment and treatment adaptation. [7, 9, 10, 23, 25, 26, 28, 31, 40] Differential response in ALL subgroups Two major differences between pcB-and T-ALL can be found: i) MRD at end of induction (day 33) is more informative in pcB-ALL, while MRD at end of consolidation (day 78) is more informative in T-ALL. ii) MRD levels in pcB-ALL correlate with risk of systemic relapse whereas MRD in T-ALL is predictive of both systemic and extramedullary relapse. The results obtained here in combination with the results obtained by FCM in the same trial has changed the risk group definition for HR patients as used by AIEOP-BFM: Any patient who has more than 10% leukemic blasts by FCM on day 15 is enrolled into the HR group; any patient with pcB-ALL who has MRD ≥ 10 -3 (0.1%) at day 33, and is still MRD positive at day 78, is stratified into the HR group; any pcB-ALL patient with MRD <0.1% at day 15 is eligible for SR if PCR results (with at least one highly sensitive molecular target, quantitative range up to 10 -4 ) at day 33 and day 78 are also negative. [3, 17] Post-remission MRD surveillance: Should MRD before and after hSCT be monitored? Additional post-remission MRD assessment is performed in several clinical trials. In AIEOP-BFM ALL 2000, all patients with MRD at a level of ≥10 -3 at day 78 were stratified into the HR group, and then monitored after each reconsolidation element. This strategy is currently used to adjust further chemotherapy and prepare for allogeneic stem cell transplantation (SCT). While post-induction MRD was also found to have significant prognostic impact in relapsed ALL [12] , MRD monitoring of pre-SCT response demonstrated the necessity to optimize the quality of remission before transplant to prevent post-SCT relapse. [1, 2, 22] Philadelphia-chromosome positive ALL The long-term follow-up of Ph+ patients treated in COG with imatinib in addition to intensive chemotherapy provides interesting insights into the efficacy of a tyrosine-kinase inhibitor when combined with chemotherapy. [20, 36] In those patients treated most intensively with imatinib, the prognostic value of MRD was nearly abrogated. At the same time, the authors were able to demonstrate that additional genetic abnormalities had a significant impact on event-free survival. [36] In adult Ph+ ALL, the picture was similar, demonstrating a limited prognostic impact of persisting MRD when patients were treated with a tyrosine kinase inhibitor (TKI) in addition to chemotherapy. [30] Very intriguingly, Foa et al. demonstrated that a TKI (dasatinib) combined only with steroids and intrathecal chemotherapy can induce complete remission in patients with Ph+ ALL, even though a significant number relapsed later on. [16] This indicates a potentially significant progress as such approaches may contribute to future treatment regimen that carry less toxicity due to the replacement of (some) chemotherapy by more targeted agents. MRD monitoring in relapsed patients carries some potential pitfalls mainly due to clonal evolution. [13] Detailed analysis of all molecular markers at first diagnosis and at time of relapse may reveal a different origin of the predominant clone. [37] Regular monitoring of MRD by those markers defined at first diagnosis may fail if clonal evolution occurs, also when occurring after the first relapse. [19] Ongoing research will determine if next generation sequencing (NGS) is the appropriate tool to prevent such diagnostic "failures". [14, 21] In adult patients, a striking difference was found between de novo and relapsed Ph+ ALL with regard to the levels of bcr-abl kinase domain mutations. This may not only explain the lack of treatment efficacy when treating with TKI but also impact the utility of RT-PCR monitoring of relapsed Ph+ ALL patients. [29] Endpoint in clinical trials To define antileukemic activity, MRD response can also be utilized as primary endpoint in clinical trials. It needs to be taken into account, however, that MRD response is only a more sensitive tool to assess response when compared to the cytological response. It may completely depend on the scenario, if activity translates into improved efficacy of a given intervention, thus improving the relevant clinical endpoints such as event-free survival and overall survival. This said, it is noteworthy that recent novel immunotherapeutic strategies e.g. with an anti-CD3, anti-CD19 bispecific monoclonal antibody showing strong antileukemic activity (as measured by MRD response) may contribute to long-term remission as such agents may improve the quality of remission needed to perform subsequent allogeneic SCT successfully. [38] [18, 32] By contrast, a British study comparing reinduction therapy with mitoxantrone versus idarubicin in relapsed pediatric ALL demonstrated that MRD response (activity) was not predictive of treatment efficacy. [27] This observation indicates that MRD response may be misleading in drug evaluation as long as activity is considered to be equal to efficacy. Essentially all subtypes of ALL display heterogeneity in response. MRD detection has evolved as one of the most powerful diagnostic tools in the prognostic assessment and also for the clinical management of ALL. In its prognostic power, MRD cannot be replaced by upfront diagnostic markers. New techniques such as whole genome sequencing may overcome some of the shortcomings of current MRD technologies if validation proves successful. Clinical intervention based on MRD results must rely on robust prospective data indicating the precise prognostic impact of a given MRD level at predefined timepoints. In well-defined circumstances, MRD detection can be a very useful guide for modulating therapy. The author declares that he has no conflict of interest Minimal residual disease (MRD) has been found to offer an independent post-remission prognostic marker in acute myeloid leukaemia (AML). We have confirmed this in a large blinded study in which MRD was assessed in parallel to a HOVON therapy protocol. Subsequently, MRD is now included as part of a new risk classification in a HOVON/SAKK treatment protocol where intermediate risk patients, previously all scheduled for allogeneic stem cell transplantation, are scheduled for autologous transplantation in case of MRD-negativity. In parallel, leukaemia stem cell load at clinical follow up was identified as a factor that adds significantly to the prognostic value of MRD load. Especially relapse in MRD-negative patients can now be better predicted based on leukaemia stem cell (LSC) frequency. A single tube assay was therefore developed which facilitates stem cell load assessment and allows this in the majority of patients. Lastly, false-negative MRD and false-negative LSC cases may be identified by monitoring upcoming aberrant populations that, due to low frequencies, were not properly identified at diagnosis. Treatment of AML patients usually is performed according to classification in risk groups defined by molecular, cytogenetic and clinical parameters, which thereby identify patients with favourable, intermediate and poor/very poor risk. 1 Although very useful, this classification is still far from the ideal situation in which patients should have individualized risk assessment, potentially allowing individualized therapy: all risk groups still harbour patients with good and poor outcome, even the most favourable patient group, while the poor risk patient group still harbours good performing patients. Parameters that would integrate all known and unknown factors contributing to individual patient's clinical outcome were therefore thought to be urgently needed. A post-remission parameter like MRD has been proposed to represent such a factor and is meanwhile extensively studied. 2 Our recent study in the HOVON/SAKK consortium (http://www. HOVON.nl) has shown for young patients (HO42A, for details see website) that MRD assessed by flowcytometry in an unbiased way (prospective sampling and MRD quantitation in a multiinstitutional way, with complete lack of knowledge on clinical outcome until final analysis) is an independent prognostic factor in all individual risk groups. However, despite these improvements, still a considerable portion of MRD-negative group (defined as having MRD below a threshold level of 0.1% of WBC) relapse. 3 Also, the terms MRDnegative and MRD-positive (above the threshold level of 0.1%) have a different meaning in the different risk groups: MRDpositive patients in the favourable risk group still have relatively good prognosis, while MRD-negative patients in the poor risk group still have relatively poor prognosis. Intrinsic properties of MRD cells may underlie this phenomenon and one of the factors may be the LSC load. LSC are hypothesized to initiate leukaemia, to be therapy resistant, and at the basis of re-growth of the leukaemia after treatment. 4 As a single independent parameter it has been shown that stem cell load at diagnosis is a prognostic marker 5,6,7,8 and, more recently in the HO42A study, we also showed prognostic impact after therapy. 7 The stem cell load, when taken into account in a risk assessment that also includes MRD, may further improve MRD based risk assessment. Flowcytometric MRD was performed using the approach in which leukaemia associated immunophenotypes (LAIP) were defined at diagnosis. 3 In the HOVON42A study risk was based on the combination of cytogenetics, molecular aberrations and clinical parameters. LSC can be defined as CD34+CD38-. Strategies to define such population at diagnosis and at follow up have been described elsewhere. 7 LSC specific markers (not expressed on CD34+ CD38-normal hematopoietic stem cells, HSC), included lineage markers like CD7, CD19, CD56 9 and, in addition, CLL-1 (CLEC12A) 10 , were originally included in the LSC panel. 7 In a new protocol, tested in more than 500 AML samples, additional markers have been included (e.g CD123, CD33, CD96, TIM-3, CD44). The combination of MRD and LSC frequencies after induction therapy was tested for relapse free survival in a small (n=91) patient group. Threshold levels used for LSC were much lower than for MRD: 1 in 100,000 cells to one in a million cells. 7 MRD: Based on the prognostic impact of MRD in the HO42A clinical study, the present HO132 protocol contains a risk group previously designated as intermediate risk, but now, based on MRD positivity (MRD % >0.1) after induction courses, is part of the poor risk group. These patients, whenever possible, are scheduled for allogeneic transplantation (Table 1) . New risk group Since MRD had the strongest prognostic impact after the second induction course, for logistical and financial reasons it was decided to restrict the protocol to MRD assessment after the second induction course. MRD status after consolidation, although informative 3,11 is not considered in decision making, since it does not comply with the timely finding of an allogeneic donor for part of the patients. Based on previous validation study 12 , it was decided to perform MRD assessment in a largely centralized way. Accrual of patients has started (target number 800; 83 patients already included). LSC ± MRD: When combining MRD load and stem cell load ( Figure 1 ) it was found that: 1. The MRD negative group consists of two different groups with different prognosis based on LSC load; 2. The LSC negative group can be divided in a MRD-positive and a MRD-negative group with different prognosis; 3. Double negativity goes along with good prognosis, while double positivity goes along with very poor prognosis. 7 LSC could be accurately determined in part of the patients only. 7 Therefore, additional markers were studied with the aim to increase that part. Using the markers indicated in Methods, it is now possible to assess LSC in >90% of the patients. Based on extensive study of the redundancy of markers, we are now able to assess total LSC load in a given AML both at diagnosis and after therapy in a one tube assay that includes 13 different markers. This approach is included as a side study in the HO132 protocol, where the possible contribution of LSC load assessment to prognosis will be studied. Immunophenotype shifts in MRD and LSC assessment: Although LSC markedly improves risk stratification based on well-known risk factors and now also including MRD load, still the double negative patient group (MRD-LSC-) contains relapsing patients while the two single positive groups still contain both relapsing and non-relapsing patients (Figure 1 ). We and others found immunophenotype shifts in around 60% (median in different studies ,13 ) of patients, and these contribute to MRD-and LSC false negatives. With the knowledge of immunophenotype and molecular shifts in mind, the present HO132 protocol, apart from the inclusion of LSC load assessment, also assesses the contribution of immunophenotype shifts to MRD (and LSC) negativity. We conclude that, based on many studies, and especially on our recent unbiased MRD study, MRD contributes significantly to the existing risk group stratification that is based on cytogenetic and molecular factors as well as on clinical factors. The decision-making process whether a patient is allocated to an allogeneic stem cell transplantation can now be guided by MRD assessment. Moreover, the inclusion of other factors like stem cell load and immunophenotype changes, as seen for MRD and LSC, will contribute to the final goal of individualized therapy. The evolution of acute myeloid leukemia (AML) has been previously described either in studies of large patient cohorts with focus on only a restricted number of AML-associated genes or in smaller series of relapsed patients studied by genome-wide techniques [1] . We set out to comprehensively characterize the genetic evolution in a large AML cohort in order to understand molecular mechanisms of relapse and therapy-resistance. We performed exome-sequencing of matched bone marrow or peripheral blood samples taken at diagnosis, complete remission and relapse from 47 patients with cytogenetically normal AML (CN-AML). Samples were collected within the German Cancer Consortium (DKTK) at the partner sites in Berlin and Munich. The median age at diagnosis was 65y (range: 21-89y). FLT3 internal tandem duplication (ITD) and NPM1 mutation status at diagnosis was available for all but one patient (FLT3-ITD -/NPM1 -, n=5; FLT3-ITD + /NPM1 -, n=9; FLT3-ITD -/NPM1 + , n=16; FLT3-ITD + /NPM1 + , n=16). On average, 96% of the target sequence was covered at least 10-fold (minimum coverage defined for variant calling). The following criteria were applied for identification of somatic mutations: Variant allele frequency (VAF) ≥20% either at diagnosis or at relapse and VAF<5% at remission. We filtered for mutations with translational consequences, excluded known error-prone genes and dismissed common germline polymorphisms (dbSNP 138; MAF≥1%). We identified a total of 777 genes to be somatically mutated, of which 104 were recurrently affected. Mutation frequencies of 12 genes found mutated both in our cohort and in 86 CN-AML patients reported by The Cancer Genome Atlas [2] are shown in Figure 1 A. Seven genes were recurrently altered only at diagnosis (e.g. CBL) and 16 genes were recurrently altered only at relapse in our cohort (e.g. KDM6A, SF3B1 and SRSF2). At diagnosis, the number of somatic mutations per patient varied between 5 and 34 (median: 17). At relapse, the number of mutations ranged from 2 to 57 (median: 17). Mutations in several AML-associated genes (e.g. DNMT3A, RUNX1, IDH1 and IDH2) showed similar VAFs at diagnosis and relapse in the vast majority of cases. In contrast, WT1 mutations were gained at relapse in 4/6 (67%) patients and FLT3 point mutations were below 5% VAF at relapse in 7/12 (58%) patients initially positive for these variants. In total, 92 mutations present at diagnosis were lost at relapse (VAF <5%) while 116 mutations were acquired during disease progression. Clonal evolution, i.e. gain or loss of mutations at the time point of relapse, was observed in 42/47 (89%) of patients. We evaluated clonal heterogeneity based on data from targeted re-sequencing. So far, analysis is complete for 32 patients. At diagnosis, 31/32 samples showed evidence for the presence of multiple subclones. In 19 patients (59%), the major clone at diagnosis remained predominant at relapse whereas expansion of a subclone during disease progression was observed in 13 patients (41%). Based on cytogenetics and copy number alteration (CNA) analysis of exome data, we detected partial or complete gain/loss of chromosomes. Five patients (11%) acquired chromosomal alterations during disease progression. Trisomy 8 was the only recurrent chromosomal abnormality gained in 3 patients (6%) at relapse. To detect pre-leukemic lesions [3, 4] , we evaluated our exome data for the persistence of mutations in 40 AML-associated driver genes during remission. We limited our analysis to mutations previously reported as confirmed somatic (COSMIC annotation) to avoid confounding with private germline variants. Strikingly, 22/47 (47%) of patients carried non-silent mutations in these genes with VAF>5% (median: 31%, range: 9-75%) at remission (27 mutations in total). In contrast, other mutations (e.g. in FLT3 or NRAS) found in these patients could not be detected at remission, consistent with therapy response. Based on VAF, 22/27 81%) persistent mutations showed a dynamic pattern over the course of disease with a relative change of >20%, likely due to partial eradication/expansion of leukemic or pre-leukemic clones. Persistent mutations in DNMT3A, TET2, SRSF2 and IDH2 were observed in 28%, 11%, 4% and 4% of patients in our cohort, respectively (Figure 1 B) . Among patients with DNMT3A mutation at diagnosis, those with persistent mutations tended to relapse earlier (n=13; median time to relapse 270 days; range: 81-586) than patients without detectable DNMT3A mutations at remission (n=7; median time to relapse 508 days; range: 235-1697; p=0.111). Our findings provide insights into the genetic evolution during the course of disease in a large cohort of relapsed CN-AML. Information about the dynamics of genetic lesions (e.g. persistent or relapse-specific mutations) may have prognostic significance and allow for tailored approaches to treat or to prevent relapse of AML. Acute myeloid leukemia (AML) is characterized by disruption of hematopoietic stem and progenitor cell differentiation. Disruption of function of stem cell and differentiation stage-specific transcription factors is frequently observed in AML patients, and recent work has shown that AML is associated with mutations in genes encoding epigenetic modifiers. Furthermore, newer data suggest that aberrations at the transcriptional and epigenetic level are founding events in the multistep pathogenesis of AML. We studied DNA cytosine methylation during physiological hematopoietic stem cell (HSC) differentiation hypothesized, reasoning that analysis of alterations in DNA methylation patterns during healthy HSC commitment and differentiation would yield epigenetic signatures that could be used to identify stage-specific prognostic subgroups of AML. We performed a nano HpaII-tiny-fragment-enrichment-by-ligation-mediated-PCR (nanoHELP) assay to compare genome-wide cytosine methylation profiles between highly purified human long-term HSC, short-term HSC, common myeloid progenitors, and megakaryocyte-erythrocyte progenitors. We observed that the most striking epigenetic changes occurred during the commitment of short-term HSC to common myeloid progenitors and these alterations were predominantly characterized by loss of methylation. Analysis of changes at the level of individual loci showed that between the progression of LTHSC to STHSC, 40% of the 509 significantly differentially methylated loci were demethylated and 60% were more methylated in the STHSC cells. Similarly, transition from CMP to MEP identified 127 loci (52%) with decreased methylation levels and 116 (48%) with increased methylation levels, which indicates predominantly balanced changes in the numbers of hypomethylated and hypermethylated loci. However, the transition between STHSC to CMP showed that, of the 793 differentially methylated loci, 95% were more methylated in STHSCs than in CMPs. We developed a metric of this HSC commitment-associated methylation pattern, and then applied this epigenetic signature to analyze published gene expression data and clinical outcome data from patients with AML. The HSC commitment-associated signature proved to be highly prognostic for overall survival in 3 independent large AML patient cohorts, regardless of patient treatment and epigenetic mutations. Interestingly, mutations in genes that are known to regulate DNA cytosine methylation, such as isocitrate dehydrogenase 1 (IDH1), IDH2 and methylcytosine dioxygenase 2 (TET2), were not enriched in either the 'signature score high' or the low-scoring patients. Application of the epigenetic signature metric for AML prognosis was superior to evaluation of previously published commitment-based gene expression signatures. Together, our study has delineated a human stem cell commitmentassociated methylome, and shows that that assessing the status of HSC commitment using an epigenetic signature is an independent tool for prognostication of overall survival in patients with AML. A provisional patent application for the diagnostic use of the described stem cell commitment-associated methylation signature in AML has been filed by the Albert-Einstein-College of Medicine. Postremission therapy of patients with acute myeloid leukemia (AML) is critical for the elimination of minimal residual disease (MRD) and a prerequisite for achieving cure. Cellular immunotherapy is a highly effective treatment option as demonstrated by the low relapse rate after allogeneic stem cell transplantation (SCT). However, many patients are not eligible for this treatment. Therapeutic vaccination with autologous dendritic cells (DCs) is a promising strategy to induce anti-cancer immune responses. We have developed a GMP-compliant protocol for the generation of next-generation DCs. A short 3-day differentiation period is combined with a novel maturation cocktail including a TLR7/8 agonist, resulting in DCs characterized by a positive costimulatory profile, high production of IL-12p70, polarization of T helper cells into Th1 and efficient stimulation of cytotoxic T lymphocytes and NK cells. In a current proof-of-concept phase I/II clinical trial we are evaluating next-generation DCs as postremission therapy for AML patients with a non-favorable risk profile (NCT01734304). Standard exclusion criteria apply, and patients have to be ineligible for allogeneic SCT. DCs are generated from patients´monocytes and loaded with RNA encoding the leukemia-associated antigens WT1, PRAME or CMVpp65 as an adjuvant and surrogate antigen. Patients are vaccinated intradermally with 5x10 6 DCs of each of the three different batches up to 10 times within 26 weeks. Primary endpoints are feasibility and safety, and secondary endpoints include immune responses and disease control with a particular focus on MRD conversion. Phase I will include 6 patients, and phase II another 14 patients. So far, six patients have been enrolled and three of them have been vaccinated for at least six times each. DCs fulfilled all quality criteria (phenotype, viability, sterility, cell count, purity), and after thawing maintained their positive costimulatory profile as well as their capacity to secrete high amounts of IL-12p70. DCs expressed all three antigens and were able to induce a selective T cell response in vitro, suggesting proper antigen processing and presentation. In vaccinated patients a delayed type hypersensitivity reaction developed. Apart from erythema and itching at the injection site, no higher grade adverse events have been observed. As of yet, all vaccinated patients are relapse-free. The authors declare that they have no conflict of interest IDH is an enzyme that interconverts isocitrate and a-ketoglutarate (aKG) as part of the citric acid cycle. Mardis and colleagues first identified an IDH mutation in AML after performing massively parallel sequencing on 188 AML samples, noting the presence of IDH1 mutations in 8% of patients 1 . Cancer-associated point mutations in IDH1 lead to the acquisition of aneomorphic activity that allows for the NADPH-dependent reduction of aKG to the oncometabolite (R)-2 hydroxyglutarate (2HG) which is present in vast excess in patients with IDH1 mutations 2 . The demonstration of elevated 2HG levels in IDH1-wildtype AML patients ultimately led to the discovery of IDH2-mutant AML, which occurs in 8-12% cases [3] [4] [5] . IDH mutations affect one of three active-site arginine residues of the enzyme, IDH1/R132, IDH2/R140, or IDH2/R172. The expression of either the IDH1 or IDH2 mutant in AML induces an increase in global DNA hypermethylation, and leads to a block in cell differentiation 6, 7 . IDH1 and IDH2 mutations are mutually exclusive in the great majority of cases, and the IDH2/R172 mutation appears to be mutually exclusive with mutations in NPM1 and FLT3-ITD 4,8-11 . There have been inconsistent findings regarding the prognostic impact of IDH mutations in AML. One study showed a lack of a prognostic association with IDH1 or IDH2 mutations 9 . Two studies reported that IDH2 mutations are associated with a favorable prognosis when IDH-2 mutations co-occur with NPM1 mutations, although the favorable effect was restricted to patients with the IDH2/R140 mutations 10, 12 . One study reported a negative prognostic association to IDH1 mutation alone 8 , yet others report both IDH1 and IDH2 mutations as associated with an unfavorable prognosis 4, 11 . Needless to say, this is an area that warrants further investigation prior to drawing final conclusions regarding the prognostic impact of these mutations. Wang and colleagues developed a small molecule IDH2 inhibitor, AGI-6780, which was demonstrated to potently and selectively inhibit the IDH2/R140Q mutant. The investigators treated ex vivo human AML cells (IDH2/R140Q mutant cells and IDH2/wild-type cells) with AGI-6780. The IDH2 mutant cells demonstrated a dose-dependent reduction in the amount of both intracellular and extracellular 2HG. Further, a burst of cell proliferation was observed in the IDH2 mutant cells treated with AGI-6780, with an increase in the number of viable, more mature CD45-positive cells. After 6 days of treatment, there was a dose-dependent decrease in the percentage of blast cells. These effects did not occur in the IDH2/wild-type cells. The authors concluded that AGI-6780 is able to promote outgrowth and differentiation of the cells harboring the IDH2 mutation, and should be investigated clinically 13 . AG-221 is a potent, selective inhibitor of the IDH2 mutant enzyme, and it has been demonstrated to reduce 2HG levels by >90%, to reverse histone and DNA hypermethylation in vitro, and to induce differentiation in leukemia cell models. Yen et al. evaluated the efficacy of AG-221 in an IDH2/R140Qmutant AML xenograft model and found that AG-221 reduced 2HG in bone marrow, plasma, and urine of the engrafted mice. This conferred a survival benefit to treated mice, and survival correlated with the onset of cellular differentiation 14 . The pre-clinical data for IDH2 inhibition led to the development of a phase I, multi-center, open-label, dose-escalation study, examining AG-221, a first-in-class, potent, reversible, selective inhibitor of the mutant IDH2 enzyme. The data from the phase I trial (NCT01915498) were presented at the 56 th American Society of Hematology (ASH) meeting in December 2014. Eligibility criteria included presence of an IDH2 mutation in advanced hematologic malignancies. The medication was continuously administered orally as a single agent either once daily or twice daily in 28-day cycles. The first cohort received a dose of 30 mg twice daily, and sequentially higher doses are ongoing (currently at 150 mg twice daily or 200 mg once daily) 15 . To date, 48 patients have been dosed and 27 remain on study drug as of July 2014. AG-221 has been well tolerated with the majority of adverse effects being grade 1 and 2, and maximum tolerated dose has not yet been reached. Of the nine deaths, eight occurred within the first cycle of therapy, with only one death being reported as "possibly" related to study drug in a patient with severe pneumonia 15 . At time of the ASH presentation, 32 patients had undergone day 28 bone marrow biopsies and were evaluable for efficacy. Investigator-assessed objective responses have been seen in 20 patients, with 8 complete responses (CR), 1 CR with incomplete platelet recovery, 3 CR with incomplete blood count recovery, and 8 partial responses. Five patients have stable disease and continue on study drug, and seven patients have had progressive disease ( Table 1) . Responses seem durable with a CR lasting up to 4.5 months and ongoing among current subjects. Five patients who achieved CR proceeded to allogeneic stem cell transplantation 15 . Conclusion IDH2 inhibition has shown unprecedented activity in relapsed/refractory AML patients whose leukemia cells express IDH2 mutations in this phase I study. These findings warrant further investigation in the phase II setting. Additional considerations for the future include whether IDH2 inhibition can be moved into the front-line setting in combination with conventional induction therapy and whether IDH2 inhibition should be continued post-transplant for maintenance. IDH1 inhibitors are now in clinical trials for patients with IDH1 mutant AML. Ultimately, IDH2 inhibitors have served as a powerful model for the translation of genomic findings into the clinic. CIK cells were transduced with SFG-retroviral vectors encoding anti-CD33.CAR and anti-CD123.CAR. For the in vitro studies first generation CAR have been used, while for the in vivo studies third generation CARs have been designed, carrying CD28 and 0X40 costimulatory molecules within the endodomain of the CAR. The efficacy and safety profiles of transduced cells were characterized in vitro by short and long-term cytotoxicity assays. Their potential reactivity against HSPCs was evaluated by colony-forming unit assay after 4-hours co-incubation. In vivo tests were performed using a xenograft model of immunocompetent mice (NSG), that were treated with CAR-redirected CIK cells, upon injection of primary AML cells to test the efficacy profile, or CD34+ cord-blood-derived cells to the safety profile against HSPCs. Anti-CD123.CAR + CIK cells were able to strongly kill CD123 + THP-1 cell line (60%±5.4%, Effector:Target -E:Tratio of 5:1, n=3), as well as primary AML blasts (59%±5.4%, E:T ratio of 3:1, n=4) in 4-hours citotoxicity assays. The same results were observed with anti-CD33.CAR + CIK cells, as expected according to the overexpression of both target antigens. In long-term cytotoxicity assay we used as targets the , n=4) . Furthermore, a limited killing of normal CD123 + monocytes and CD123low expressing endothelial cells was measured. Once injected into low-level AML engrafted NSG mice (median of hCD45 + CD33 + 0.6% before treatment), genetically modified T cells had a potent antitumor effect. Indeed, the bone marrow of control untreated animals or mice treated with un-manipulated CIK cells, was infiltrated by leukemic cells (86% and 81% leukemic engraftment), while in 7/8 anti-CD33-CD28-OX40-ζ and in 8/10 anti-CD123-CD28-OX40-ζ treated mice AML cells were not detectable. Similar results were obtained when T cells were injected in mice with an established high AML burden (median of hCD45 + CD33 + 70% before treatment). One week after the last CIK injection the level of AML engraftment was 96%, 87%, 0.35% and 0.34% for untreated mice, mice treated with un-manipulated CIK cells and with anti-CD33-CD28-OX40-ζ and anti-CD123-CD28-OX40-ζ transduced CIK-cells, respectively. We performed secondary transplantation on residual AML cells and mice were treated again with transduced CIK cells. Residual AML cells were still sensitive to CARs approach, leading once again to an almost complete eradication of the disease (median level of hCD45 + CD33 + engraftment was 98%, 0.02% and 0.04% respectively for untreated mice, anti-CD33-CD28-OX40-ζ and anti-CD123-CD28-OX40-ζ transduced CIK-cells). Furthermore, a fundamental issue was to determine the safety profile of such approach against normal haematopoietic precursors. In untreated mice injected with primary cord blood-derived CD34 + cells the level of engraftment of hCD45 compartment was 42%, whilst in mice treated with unmanipulated, anti-CD33-CD28-OX40-ζ or anti-CD123-CD28-OX40-ζ transduced CIK-cells the levels of human compartment was 40%, 11.7% and 26.3% respectively. Moreover, when we considered specifically the CD34 + CD38compartment, enriched in HSC, the level of engraftment was 1.92%, 1.02%, 0.55% and 0.83%. Secondary transplantations using remaining HSC repopulating cells confirmed these findings. Taken together, our results indicate that both anti-AML CARs strongly enhances anti-leukemic CIK functions towards AML, having anti.-CD123 CAR with an improved safety profile on normal HSPCs compared to anti-CD33.CAR, paving the way for the development of novel immunotherapy approaches for the treatment of resistant forms of AML. The authors declare that they have no conflict of interest Ch. Thiede Medical Dept. 1, University Hospital Carl Gustav Carus, Dresden, Germany The detection of minimal residual disease (MRD) has become one of the most important aspects in the treatment of patients with leukemia, especially in CML, ALL and AML. Several studies have shown that detection of MRD is able to identify patients at risk for early disease recurrence. While translocations, like in CML and small subgroups of patients with AML or the IgV H or TCRB rearrangements are very suitable targets for sensitive and quantitative detection of MRD, the predominant genetic lesion in AML and MDS are single base pair point mutations or small insertion or deletions, which can be much more difficult to detect at a comparable level of sensitivity. Next generation sequencing (NGS) is a novel technology, which is able to generate massive amounts of sequencing data (up to 10 12 bases). Several groups have shown that this technology can be used not only to resequence genetic material but also to identify variant alleles at very low frequency (1, 2) . We aimed to use NGS for the sensitive detection of MRD, especially in AML. However, pilot experiments using established protocols clearly showed insufficient sensitivity due to an increased background, predominantly induced by DNA-polymerase base misincorporation. In order to overcome these inherent limitations and artifacts induced during data analysis due to subsampling and barcode allocation, we performed extensive optimization of the reaction parameters and developed novel algorithms for data analysis. Using the optimized conditions, we were able to detect variant alleles in genes like DNMT3A, JAK2, IDH1, IDH2, BRAF, KIT, KRAS, NRAS with a sensitivity down to 1/50.000 cells, making the method applicable to detect MRD. A major advantage of this procedure compared to other currently available methods, e.g. digital-PCR, is the quantitative, scalable and flexible mode of detection allowing for MRD assessment based on genetic alterations present in DNA-stretches up to 100 bp. The author declares that he has no conflict of interest HLA-haploidentical hematopoietic stem cell transplantation (haplo-HSCT) is a valuable treatment option for patients with various hematologic disorders who lack an HLA-matched donor. The high and ready availability of a one HLA-haplotype-mismatched family donor is an important advantage of this alternative strategy, especially in patients who do not have an HLA-sibling donor, but need an urgent transplant due to aggressive, active disease. Historically, the main risks limiting the benefit of haplo-HSCT were graft rejection, severe graft-versus-host disease (GvHD) and early death due to toxicity and infections. To allow for crossing of the HLA-barrier, HLA-haploidentical transplantation strategies were developed involving intensive preparative regimens and in-vivo and ex-vivo T cell depletion (TCD). Unwanted consequences of these approaches were reduction of cellular immunity, and delayed immune reconstitution resulting in high rates of mortality due to infections, toxicity and relapse. The use of reduced intensity conditioning [1, 2] , infusion of mega doses of CD34+ cells [3] , and graft manipulations such as selective T cell depletion [3] were helpful to achieve engraftment with lower rates of GvHD, but relapse and infectious complications remained a problem, in particular in patients with high-risk, nonremission diseases. Thus, the development and employment of strategies that allow a safe engraftment with low rates of GvHD, toxicity and infection while providing a sufficient disease control have been an important task in haplo-HSCT. Recently, novel strategies using T-cell-replete grafts (TCR) and high-dose cyclophosphamide (HD CY) post-transplantation have made haplo-HSCT markedly safer (TCR/HD CY approach; Figure 1) . Encouraging results inducing remarkable tolerance were achieved showing low rates of graft failure (13%), GvHD (acute grade II-IV: 6%; chronic: 5%) and non-relapse mortality (NRM at one-year: 15%) [4] , as well as improved immune reconstitution and less opportunistic infections in comparison to a TCD strategy using CD 34+ selected grafts and in-vivo T cell depletion with anti-thymocyteglobulin (ATG) [5] . So far, over a number of years the Baltimore/Seattle group was able to successfully translate the experimental hypothesis of a bi-directional immunologic tolerance which is induced by the use of HD CYapplied in a narrow time window post-transplantation into a clinical protocol. They hypothesized, that HD CY applied after transplantation results in a selective depletion of proliferating, allo-reactive T cells which are responsible for graft rejection and GvHD, while resting cells such as memory T cells which provide post-transplant immunity to infections were spared. It was been shown that resistance of hematopoietic cells in the graft to the toxic effects of CY can be mediated through the enzyme aldehyde dehydrogenase (ALDH) which is contained in those cells on a high expression level [6] . In addition, regulatory T cells have also the ability to up-regulate the cytoplasmatic enzyme ALDH which is responsible for the detoxification of activated CY in an inactivated form [7] . Thus, acute and chronic GvHD was effective prevented by this TCR/HDCYapproach with no impact of the higher degree of the HLA-disparity on the rate of GvHD or NRM while a trend to a lower relapse risk and improved event-free survival was observed [8] . However, disease control achieved by those non-myeloablative TCR/HD CY haplo-HSCT protocols was disappointing with more than one half of the patients relapsing after one year, and relapse being the predominant cause of treatment failure. One attempt to improve disease control in patients with high-risk disease (refractory, relapse) is to itensify conditioning, the other is to use peripheral blood stem cells (PBSCs) instead of bone marrow. Accordingly, the Atlanta group treated 20 patients with hematologic malignancies (high-risk relapsed/refractory n=11) using the TCR/HD CY platform for myeloablative conditioning with fludarabine and busulfan [9] . All patients engrafted. While the incidence of acute GvHD grade II-IV was 46% in patients with relapsed and refractory diseases, and although they observed a rather high rate of CMV reactivation (81%) and BK-virus associated hemorrhagic cystitis (75%), NRM (10% at one year) was unexpected low. Control of disease was favorable for the 20 patients of the entire cohort (one-year DFS 50%), but survival data for the high-risk group with relapsed and refractory disease were not shown in particular. They conclude that myeloablative T-cell-replete haplo-HSCT is a valid option in patients with high-risk disease who lack the access to a conventional donor in a reasonable timeframe. The Genua group used the TCR/HD CY platform and performed a myeloablative conditioning consisting of thiotepa, busulfan and fludarbine or fludarabine and total body irradiation (TBI) in 50 patients with hematologic malignancies [10] . In contrast to the Baltimore protocol, the calcineurin inhibitor and MMF were started before the initiation of HD CY to better control GvHD, and the second dose of HDCy was applied on day +5 rather than of day +4 to reduce toxicity. While the impact of these alterations in immunosuppression remains unclear, this protocol allowed for a successful engraftment (90%) with low rates of GvHD (acute grade II-IV: 12%; chronic 10%) and acceptable NRM for relapsed patients (26%). However, with a short follow up (in median 8 months) DFS at 22 months was 68% for patients in remission, while it was only 37% in patients with active disease. They observedconsistent with the findings of the Atlanta group -a rather high rate of hemorrhagic cystitis (40%) with predominance after the busulfan-based conditioning. In contrast to the Baltimore group, the TCR haplo-approach of the Peking group utilized an intensive preparative regimen including in-vivo T cell depletion with ATG and a combined graft (peripheral blood stem cells and bone marrow, both G-CSF mobilized), while post-grafting immunosuppression consisted of cyclosporine A, methotrexate and MMF [11] . The leukemic control for the patients with highrisk disease was acceptable (3-year DFS 59% for AML 59%, and 24 % for ALL), but NRM was markedly high in high-risk patients, reaching 30% in AML and exceeding 50% in ALL patients mostly due to severe infections. Rates of acute (46%) and chronic (54%) were also markedly higher as observed after the TCR approach of the Baltimore group using HD CY posttransplantation. Results of sequential T-cell-replete HLA-haploidentical transplantation in the treatment of patients with high-risk acute leukemia and relapse after a first allogeneic transplantation To enhance the disease control in patients with advanced, non-remission acute leukemia we adopted the TCR/HD CY haplo-HSCT approach of the Baltimore group and introduced sequential therapy utilizing clofarabine as cytoreduction [12] . Clofarabine has shown anti-leukemic acitivity, both in myeloid and lymphoid blasts. Over the past years the concept of a sequential therapy consisting of a cytoreductive chemotherapy followed by RIC in preparation for allogeneic hematopoietic transplantation (allo-HSCT) has been demonstrated to be an efficient and tolerable treatment strategy for patients with relapsed, refractory or high-risk acute myeloid leukemia undergoing HLA-matched transplantation at our center and elsewhere [13] . Sequential therapy addresses one major concern in non-remission patients: the reduction of the leukemic burden and the enhancement of the disease control by administration of a highly effective anti-leukemic chemotherapy before allo-HSCT in order to provide more time for the presumed graft-versus-leukemia (GvL) effect after allo-HSCT. We treated 18 patients (median age 39 years; nine male) with advanced, acute leukemia (AML n=15; ALL=3) who lacked a suitable HLA-matched donor. No patient was in remission prior to transplant. All patients received cytoreduction with clofarabine (5 x 30 mg/m2 IV over 5 days), followed after three days of rest by conditioning for T-cell-replete haplo-HSCT. HD CY (50 mg IVon day +3 and 4), tacrolimus (targeted levels 8-12 μg/l) and MMF (4 x 500 mg IV) were used for post-grafting immunosuppression. In absence of GvHD, MMF was stopped on day +35 and tacrolimus was tapered and discontinued until day +180. In contrast to the Baltimore protocol disease-specific and individual modifications were undertaken in the last part of conditioning in most patients due to the high variability given among patients with relapsed and refractory non-remission acute leukemia. Hence, conditioning consisted of fludarabine, cyclophosphamide plus either melphalan, treosulfan/etoposide or TBI. The treatment schedule in detail is shown in Figure 2 . Bi-directional tolerance was induced: while allowing successful engraftment, rates of acute GvHD II-IV (22%) und chronic GvHD (28%; no severe) were low. Cumulative incidence of NRM was 6% at day + 30, but increased to23 % at one-year with infectious complications being the main cause of NRM. Most commonly seen severe toxicities (grade III-IV) were transient elevation of liver enzymes (44%) and mucositis (40%), while the incidence of hemorrhagic cystitis was low (11%). Given the high-risk profile of our patients results were encouraging. This is reflected by a CR rate of 78% on day + 30, while OS was 56% and RFS was 39% one-year after sequential haplo-HSCT. In particular, outcome was favourable for relapsed-ALL patients with 2 out of 3 patient still alive two-years after haplo-HSCT. We conclude that sequential therapy is feasible in the context of the TCR-haplo HSCT approach and provides an acceptable toxicity profile and anti-leukemic activity in high-risk acute leukemia patients with active disease at time of transplant, while rates of GvHD and NRM were comparable to results of HLA-matched transplantation. Using HLA-haploidentical donors and the TCR/HD CY haplo-HSCT platform proved also to be feasible and showed encouraging results in patients with acute leukemia relapse after a first allogeneic transplantation [14] . We retrospectively analyzed the courses of 20 consecutive adults (median age 37 years; 12 male) with AML (n=14), ALL (n=5) and acute bi-phenotypic leukemia (n=1). Conditioning consisted of fludarabine, CY and either melphalan or TBI or tresosulfan+/-etoposide. While engraftment was achieved in 85%, the rate of grade II-IV acute GvHD was 35% and chronic GvHD occurred in five patients. Toxicity was moderate. Most commonly observed severe toxicities (grade III-IV) were mucositis (30%), hyperbilirubinemia (20%), elevation of transaminases (20%) and creatinine (20%), while invasive fungal infection affected 30%. One-year NRM was 36%. At a median follow-up of 17 months, one-year OS was 45% and RFS was 33%. We conclude that early outcome and NRM are at least comparable with results after a second allo-HSCT from HLAmatched donors without donor change at HSCT2. In addition, we observed that this TCR/HD CY approach may allow a fast immune reconstitution, even after a second allogeneic transplantation, which is comparable to that of HLA-matched transplantation. Accordingly, CD4+ T cells already reached median counts >200/μl on day +100 and achieved an almost normal level (median: 430/μl) on day +360 after a second haplo-HSCT. Recently retrospective studies have shown that the TCR/HD CY haplo-HSCT approach is as safe and effective as transplantation from an HLAmatched related and unrelated donor [15] , and is comparable to use of two umbilical cord blood transplantation using two units in adults [16] . In a single center study the Genua group has compared outcomes of patients undergoing their TCR/HD CY haplo-approach (as above described) with results after HLA-matched sibling donor, matched unrelated donor, mismatched unrelated donor and umbilical cord blood transplantation. They suggest that TCR/HD CY haplo-HSCT is comparable to that of sibling transplantation, while UCB transplantation provides an inferior survival [17] . Introduction of sequential therapy in the TCR/HD CY haplo-HSCT concept by our group proved to be feasible, safe and effective in treatment of patients with advanced, high-risk malignancies. Moreover, the TCR/ HD Cy approach can provide a safe platform for a second allogeneic HSCT in patients with relapse of acute leukemia after a first HLA-matched transplantation resulting in encouraging outcomes. Since the almost universal and ready availability of an HLA-haploidentical family donor and since the TCR/HD CY approach is un-expensive and simple requiring no expertise in graft manipulations and avoiding complicated, cost and time intensive GMP procedures, this approach can challenge strategies of donor choice in patients lacking an HLA-matched donor. In patients with active and aggressive hematological disease lacking an HLA-matched sibling donor sequential therapy utilizing HLAhaploidentical family donors and TCR/HD Cy platform might be our first choice in future. Further, in patients with advanced, non-remission disease at time of transplant the utility of this TCR/HD CY haplo-HSCT in particular regarding myeloablative conditioning regimens remains to be determined; comparative and prospective trials are needed. Approximately one-third of B-cell acute lymphoblastic leukemias (B-ALLs) among adults harbor BCR-ABL rearrangements [1] . These cases have a particularly poor prognosis with conventional cytotoxic chemotherapy. However, the introduction of the ABL tyrosine kinase inhibitors (TKIs) imatinib, nilotinib and dasatinib into up-front regimens against BCR-ABL ALL has markedly improved rates of complete remission and disease-free survival [2] . Single-agent treatment of BCR-ABL ALL with a TKI results in complete hematologic remission (equivalent to >99.9% reduction in leukemia burden) in a significant fraction but is commonly followed by relapse [3] . The relapsed clones typically harbor mutations that affect TKI binding, most notably ABL T315I. The TKI ponatinib has activity against clones harboring T315I, but leads to hematologic remission in only one-third of patients after relapse [4] . Allosteric inhibitors of ABL kinase activity that bind the myristoylation site in ABL have recently been developed. We aimed to determine the therapeutic potential of a newergeneration myristate mimetic (NVP-ABL001), both alone and in combination with type II ABL inhibitors, in human BCR-ABL ALL. At the same time, we utilized these compounds to define the curative potential of prolonged and sustained inhibition of BCR-ABL in cases of B-ALL with this rearrangement. We have established a panel of primary murine BCR-ABL B-ALLs that propagate in secondary transplant recipients. In addition, we established 7 BCR-ABL B-ALL patient-derived xenografts (PDXs) that propagate in NOD.Scid.IL2R -/-(NSG) mice. The latter include two PDXs with ABL T315I mutations, one PDX with a GNB1 K89E mutation [5] and 3 PDXs with BCR-ABL p210. Upon engraftment, mice were randomized to treatment with vehicle, single-agent nilotinib, single-agent ABL001, single-agent ponatinib, or combinations of ABL001 with nilotinib or ponatinib (n=8 per group/line). Sentinel mice (n=2 per group/line) were sacrificed on treatment day 7 for pharmacokinetic and pharmacodynamic evaluation. This included peripheral blood flow cytometry for hCD19+/hCD45+ cells, BCR-ABL qRT-PCR, and spleen and bone marrow immunohistochemistry for leukemic burden, immunoblotting for phosphorylated BCR-ABL and downstream signal transduction molecules for on-target drug effect. Mice were monitored on treatment by weekly peripheral blood flow cytometry until relapse (defined as >5% peripheral blood human CD45/human CD19-positive cells), at which point circulating leukemia cells were collected for evaluation of resistance mechanisms. Relapsed mice from single-agent nilotinib or ponatinib arms then crossed-over from monotherapy to combination therapy with ABL001. When mice progressed on combinations, they were sacrificed and leukemia cells were isolated from bone marrow and spleen. We tracked the clonal architecture of BCR-ABL resistance mutations in the PDX mice by PCR amplification of the ABL kinase domain followed by next-generation sequencing (NGS). By sequencing to depth~500,000, we can easily identify variants present at 10 -3 -10 -4 and distinguish these from background sequencing errors [6] . Mice in which acquired BCR-ABL mutations did not explain relapse (i.e., clonal composition of the relapsed leukemia was primarily wild-type) were further subjected to unbiased genetic evaluation to identify BCR-ABL-independent mechanisms of resistance. Continuous variables were compared using the t-test while the Fisher's exact test was performed to compare categorical variables. Overall survival was analyzed by the method of Kaplan and Meier and compared using the log-rank test. Results from these studies will be outlined at the Acute Leukemias XV meeting. Therapy-related acute myeloid leukemia (t-AML) and therapy-related myelodysplastic syndrome (t-MDS) are defined as AML or MDS arising after a previous exposure to cytotoxic chemotherapy or radiotherapy [1] . Compared to their de novo counterparts, these disorders are associated with a higher incidence of cytogenetic abnormalities including deletions in chromosomes 5 and 7, a poorer response to chemotherapy, and a worsened overall survival [2] . The key difference between the pathogenesis of t-AML and de novo AML is the exposure of the former to cytotoxic therapy during its evolution. However, the exact mechanism by which this exposure to cytotoxic therapy contributes to leukemogenesis remains unclear. In this presentation, we will discuss how recent findings by our research group on the genetic heterogeneity of AML and normal hematopoietic cells may shed light on the pathogenesis of t-AML and t-MDS. Recent genomic studies performed by our research group have revealed AML to be a clonally heterogeneous disease [3] . Although somatic mutations arising early in the evolution of AML are present in every leukemic cell (i.e. in the founding clone), we are also often able to identify multiple subclonal leukemic populations at the time of diagnosis. Each subclone harbors its own set of unique somatic mutations in addition to the mutations present in the founding clone. These additional mutations have presumably provided the tumor cell acquiring them with a competitive advantage, allowing it to clonally expand. The selective pressure of cytotoxic therapy can dramatically alter the clonal architecture of AML, with certain (often minor or even undetectable) subclones at diagnosis becoming the dominant leukemic population at the time of relapse after induction chemotherapy [4] . The clonal evolution of tumor cells under the selective pressure of cytotoxic therapy is increasingly being recognized as a common phenomenon in a variety of cancers [5] . This is important clinically as curative therapy requires the elimination of both the founding clone and all of its subsequently arising subclones. Our research group has revealed the hematopoietic progenitors from healthy individuals to be also genetically heterogeneous [6] . By performing exome sequencing on the progeny of sorted hematopoietic stem and progenitor cells (HSPCs) obtained from healthy donors, we discovered that these cells acquire somatic mutations as a function of age. HSPCs obtained from cord blood samples possessed a low number of validated somatic mutations. On the other hand, HSPCs from adult donors possessed an increased number of mutations, with the total number of mutations strongly correlating with the age of the donor. Thus, each individual possesses a genetically heterogeneous HPSC population with each HSPC having its own unique set of somatic mutations and with the total number of mutations increasing as the individual ages. The majority of these age-related mutations are likely biologically and clinically insignificant. However, multiple studies have shown that, in elderly individuals without a known hematologic disorder, oligoclonal hematopoiesis often is identifiable [7] [8] [9] [10] [11] [12] [13] . This suggests that, in these individuals, a genetically distinct hematopoietic cell has gained a competitive fitness advantage over its counterparts and clonally expanded. Of particular note, our research group recently analyzed blood leukocyte sequence data from 2,728 patients enrolled in The Cancer Genome Atlas without a known hematologic malignancy [14] . Somatic mutations were identified in over 2% of the subjects and in over 5% of the subjects greater than 70 years of age. Many of these mutations were in genes with welldescribed roles in hematologic malignancies including DNMT3A, TET2, ASXL1, and TP53. This data suggests that mutations in these classic AML and MDS "driver" genes may provide the HSPCs acquiring them with a competitive advantage. Whether the presence of a clonally expanded HSPC population carrying a mutation in these genes predisposes an individual to the future development of AML or MDS remains to be determined. In the case of t-AML and t-MDS, this selection is occurring under the influence of cytotoxic chemotherapy and/or radiotherapy. As with relapsed AML, cytotoxic therapy may potentially alter the hematopoietic clonal architecture of an individual with "normal" hematopoiesis, providing certain hematopoietic progenitors with a competitive fitness advantage. In this presentation, we will discuss how the influence of cytotoxic therapy on a genetically heterogeneous hematopoietic progenitor population may contribute to the pathogenesis of t-AML and t-MDS, resulting in diseases that are both clinically and biologically distinct from AML and MDS arising in the absence of cytotoxic therapy exposure. Diagnosing the infectious causes of pneumonia and lung infiltrates that affect patients with leukaemia and recipients of a stem cell transplant presents several challenges to the clinician and to the laboratory. Firstly, there is no such thing as the optimal sample other than obtaining tissue directly from the lung which is not only hazardous but is also seldom possible given the patient population. Hence the aim is to obtain the best possible specimen that fits the purpose best. Sputum is seldom, if ever, useful. not least because the lack of neutrophils means it is not produced during neutropenia. Rather the respiratory secretions that are expectorated are just that but often very mucoid and viscous and may only represent the slime frequently associated with mucositis. Moreover bacteria recovered from culture typically represent the oral microbial flora unless, of course, a professional pathogen such as Streptococcus pneumoniae is recovered. Most often pulmonary infectious diseases are caused by opportunistic pathogens making any distinction between colonization and infection all but impossible. Many years ago it was assumed that aetiology of pulmonary infectious diseases could be deduced from blood cultures when bacteraemia occurred at the same time. Hence the aetiology was ascribed to bacteria and certain yeasts such as Candida albicans on that basis alone. However this notion has been rightly discarded as being wrong. With the advent of better radiology and fibre optic bronchoscopy it has become more the norm to sample the lung by instilling fluid typically saline into the lungs, washing the site and attempting to recover most of it for microbiological investigation. The technique is widely known as bronchoalveolar lavage and requires a skilled and experienced respiratory physician to perform it reliably and safely but only if indicated by radiography. It is also true to say that not all bronchoscopies are the same and neither are all BAL fluids. None the less the specimen obtained at least has the merit of giving a more representative sample of the affected area of the lung. Hence it is recommended in almost every guideline for diagnosing the cause of pulmonary infection of immunocompromised patients The real challenge appears once the specimen has been obtained. What should be done with it and who should get what? Clearly there has to be an agreement if the right people are to get the right specimen to perform the right tests in a timely fashion. Before exploring specific microbiological investigations it is worth considering the question that is being asked in terms of the likelihood of any given infectious disease. For instance, if the prevalence of a given infection is very low then the test should lean towards excluding the diagnosis rather than confirming it. By contrast, if the prevalence is relatively high the test would be better employed to confirm the diagnosis reliably. Detecting galactomannan to diagnosis aspergillosis provides a useful example. Firstly, a bronchoscopy is only done when certain abnormalities are seen and there is least a real possibility of an infectious aetiology. Moreover nodules with or without a halo sign already meet the definition of possible disease thereby increasing the likelihood of aspergillosis markedly to perhaps as high as 30%. It is often forgotten that positive and negative predictive values of a given test are the product of test sensitivity or specificity and prevalence, and that one has usually to sacrifice sensitivity to obtain the highest specificity and vice versa. In the example of galactomannan it can be shown that although the analytical threshold is set at an optical index of 0.5 the clinical threshold should be set at a higher value least 1.0 to obtain the highest positive predictive value so that a positive test result is credible and confirms the diagnosis of aspergillosis. In most cases, a so-called typical radiographic appearance as described by the EORTC/MSG definitions during neutropenia is strongly associated with Aspergillus infection rather than one due to any other mould. Should a range of 20 plus tests be conducted "just in case"? If so not only is this costly but the value of the information will be difficult to ascertain as the result may be wholly inconsistent with the clinical signs and radiographic appearance. Moreover, the information may only add confusion rather than any real value. Hence the choice of tests needs to be guided by the clinical context and the population epidemiology to provide at least an estimate of likelihood. Far from solving the problem, the adoption of an increasing number nucleic acid detection assays, typically PCR tests, may have aggravated it since detection of nucleic acid from a potentially pathogenic pathogen does not of itself help distinguish colonization from infection. There is also the additional problem of contamination to contend with and the lack of approved standards. The European Aspergillus PCR initiative (www. eapcri.eu) is making progress towards establishing a standard for Aspergillus PCR but, apart from CMV detection [http://www.nibsc.org/ documents/ifu/09-162.pdf], no other PCR as been formally standardised and validated for this purpose. Quantitative techniques may help but there have been few, if any, studies done that help define the clinical as opposed to the analytical threshold. Hence, two things must drive the selection of tests. Firstly, the likelihood of a given opportunistic pathogen causing infection in a particular setting and secondly, whether or not the test chosen confirms or excludes the aetiology reliable. This aspect will be explored in full with the aim to proposing a core set of diagnostic tests and guidance on when to consider supplementary tests. Combination of pattern recognition with knowledge of the clinical setting is the best approach to pulmonary infectious processes. A specific pattern of involvement can help suggest a likely diagnosis in many instances. Different combinations of parenchymal and pleural abnormalities may be suggestive for additional diagnoses. When pulmonary infection is suspected, knowledge of the varied radiographic manifestations will narrow the differential diagnosis, helping to direct additional diagnostic measures, and serving as an ideal tool for follow-up examinations. The aim of this presentation is to summarize the radiologic and CT findings of some of the most common infectious processes affecting the immunocompromised hosts. Despite advances in diagnosis and treatment, pulmonary infections are a major cause of morbidity and mortality in adult patients. Pneumonia is the sixth most common cause of death in the United States and more than 6 millions cases of bacterial pneumonia occurs each year in the immunocompetent population (1). Pulmonary infection is also a major cause of morbidity and mortality in patients with impaired immune function. In the last several decades, AIDS epidemic, advances in the treatment of cancer, organ transplantation, and immunossupresive therapy has resulted in large numbers of patients who develop abnormalities in their immune system (2) (3) . Mildy impaired host immunity as it occurs in chronic debilitating illness, diabetes mellitus, malnutrition, alcoholism, advanced age, prolonged corticosteroid administration and chronic obstructive lung disease have also been regarded as predisposing factors of pulmonary infections (4). Diagnosis of pneumonia calls for a combination of clinical awareness, appropriate microbiologic tests, and radiographic studies. Plain chest x-ray is an inexpensive test that can rapidly demonstrate the presence of pulmonary abnormalities. According to American Thoracic Society guidelines, posteroanterior (PA) (and lateral when possible) chest radiography should be obtained whenever pneumonia is suspected in adults (5). In most cases the plain film findings may be diagnostic for pneumonia and may eliminate the need for additional radiographic procedures. The more common radiographic findings include segmental or lobar consolidations and interstitial lung disease. Other less common radiographic findings include mediastinal lymphadenopathy, pleural effusion, cavitation, and chest wall invasion. Despite that, the nonspecificity of radiographic findings as well as the wide range of potential causes often lead to frustration when evaluating the imaging findings of a patient with suspected pneumonia. Computed tomography (CT) is a useful adjunct to conventional radiography in selected cases (6) (7) (8) (9) . There is a large literature indicating that CT is a sensistive method capable of imaging the lung with excellent spatial resolution providing anatomical detail similar to that seen by gross pathological examination. Differences in tissue attenuation and parenchymal changes caused by an acute inflammatory process can be seen readily by CT (8, 9) . Unlike chest radiography, CT provides cross sectional images and the pattern and distribution of pulmonary processes are therefore much more readily appreciated than on conventional examinations. High-resolution CT can be helpful in the detection, differential diagnosis, and management of immunocompromised patients with pulmonary complications (9) (10) (11) . Although accurate clinical information is essential to narrow the differential diagnosis, it is often still imposible to determine the cause of parenchymal abnormalities in this group of patients. Combination of pattern recognition with knowledge of the clinical setting is the best approach to pulmonary infectious processes. A specific pattern of involvement can help suggest a likely diagnosis in many instances. The most common patterns seen at HRCT in acute pulmonary infections include nodules, tree-in-bud appearance, ground-glass attenuation, consolidation and airway disease. Angioinvasive aspergillosis occurs almost exclusively in immunocompromised patients with a severe neutropenia (12) . The characteristic CT findings consist of nodules surrounded by a halo of ground-glass attenuation ("halo sign") or pleural based wedge-shaped areas of consolidation. In severely neutropenic patients the halo sign is highly suggestive of angioinvasive aspergillosis (Fig.1) . The ground-glass halo reflects the presence of hemorrhage surrounding the nodule. A similar appearance has been described in a number of other conditions including infection by herpes simplex and cytomegalovirus. Close-up view of a CT scan shows the classical CT appearance of an angioinvasive pulmonary aspergillosis. A nodular opacity representing a lung infarction is surrounded by a ground-glass pattern of attenuation ("halo"). Although non-specific, this particular CT finding, in the appropriate clinical setting is highly diagnostic of angioinvasive aspergillosis. Nontuberculous mycobacteria infection may present as nodular lesions on the chest radiograph. The nodules may be single or multiple and are frequently associated with centrilobular nodules and branching opacities resulting in a "tree-in-bud" appearance (13) . "Tree-in-bud" pattern The "tree-in-bud" pattern, first described in diffuse panbronchiolitis and endobronchial spread of tuberculosis, represent bronchioles filled with mucus or inflammatory material resulting in centrilobular tubular, branching, or nodular structures (13) . A variety of bacterial, mycobacterial, fungal, and viral pathogens may cause bronchogenic dissemination and bronchiolar impaction by mucus or pus, resulting in typical "tree-in-bud" appearance. The centrilobular nodules have a patchy distribution in the lung and are similar to those seen in a number of different infectious conditions, including endobronchial spread of pulmonary tuberculosis, M. Aviumintracellulare, viral and mycoplasma pneumonia. Ground-glass opacity is defined as a localized increase in lung attenuation that allows visualization of vascular structures coursing through the affected region. Ground glass attenuation is a common but nonspecific HRCT finding in the non-HIV immunocompromised patients. This finding may represent either alveolar or interstitial disease and may also be associated with different entities. In immunocompromised patients the differential diagnosis includes infections such as PCP (Fig3), CMV (Fig2) and Mycoplasma, drug induced lung disease, pulmonary hemorrhage and cryptogenic organizing pneumonia (10) (11) 14) . Thin-section CT scan shows bilateral and multiple patchy areas of ground-glass opacities in a 28-year-old woman 48 days after allogeneic BMT. Concurrent small nodules surrounded with a "halo" of groundglass attenuation are also visible throughout the lungs. Transbronchial biopsy was diagnostic for CMV infection. Pneumocystis jirovecii pneumonia in a 46-year-old inmunossupressed man. High-resolution CT image at the level of lung bases shows extensive ground-glass attenuation in both lungs, in a geographical appearance, with only a few focal areas of sparing Focal or multifocal areas of consolidation may be seen with various infections. Bacterial infections are more frequent than fungal infections in non-HIV immunocompromised patients. Infections caused by fungi are most commonly seen in neutopenic patients with hematological diseases. A distinct form of aspergillus infection called semi-invasive or chronic necrotizing aspergillosis, may be seen in patients with chronic debilitating illness, diabetes mellitus, malnutrition, alcoholism, advanced age, prolonged corticosteroid administration, and chronic obstructive lung disease. Radiologically, unilateral or bilateral segmental areas of consolidation with or without cavitation and/or adjacent pleural thickening, and multiple nodular opacities are demonstrated (5) . Mycobacterial infection may also cause focal areas of consolidation, with or without cavitation. Infectious causes for airway disease include bacterial bronchitis, tuberculous involvement of the airway, and necrotizing tracheobronchial aspergillosis. Pyogenic infectious airways diseases have been increasingly recognized in recent years. Necrotizing bronchial aspergillosis is a rare form of invasive aspergillosis that may be seen on CT as an endobronchial mass, an obstructive pneumonitis and/or collapse, or as a hilar mass.. The diagnosis of this form of aspergillus infection is usually based on the presence of abnormal chest radiograph and bronchoscopic biopsy specimen consistent with tissue invasion (15) . The author declares that he has no conflict of interest Cancer patients with high-risk neutropenia developing lung infiltrates carry a high risk of treatment failure. Microbial etiology frequently remains undetermined, and differential diagnoses include lung involvement by the underlying malignancy, capillary leak syndrome, cryptogenic organizing pneumonia, immune reconstitution syndrome and lesions caused by chemotherapy or radiation 1-3 . It has not yet been demonstrated that diagnostic procedures such as bronchoscopy and bronchoalveolar lavage improve clinical outcome in these patients 4 . It has been reported that response to broad-spectrum antibacterial treatment is less than 30%, whereas mold-active systemic antifungals added to first-line antimicrobial treatment increases the response rate to up to 78% 5 . However, inadequate overuse of broad-spectrum antifungal agents must be avoided. In patients currently receiving mold-active antifungal prophylaxis, invasive diagnostic intervention is considered in order to detect rare pathogens such as Nocardia and non-microbial causes of LI. In contrast, patients who have been treated with nucleoside analogs such as fludarabine or cladribine or with long-term T-cell depleting agents such as alemtuzumab or antithymocyte globulin may be affected by microorganisms typical for severe T-cell immunosuppression such as Cytomegalovirus, P.jirovecii, mycobacteria, yeasts, pneumococci, Staphylococcus aureus, Pseudomonas aeruginosa or Aspergillus spp. Respiratory viruses such as Influenza, Parainfluenza, Coronavirus, Rhinovirus, Human Metapneumovirus or Respiratory Syncytial Virus may be relevant pathogens for respiratory tract infections 6 , particularly during the winter seasons, however, there are almost no effective antiviral drugs for the treatment of these infections available. An updated guideline 7 was prepared by a multi-disciplinary group of experts in a stepwise consensus process. Systematic literature search was conducted by all co-authors, data were extracted and tabulated, preliminary recommendations were proposed for discussion and sent to all co-authors. Tables were revised after email-based discussion and presented for final discussion at the AGIHO guideline conference. If no clear consensus was achieved, majority voting of the conference was adopted. The 2001 criteria for strength of recommendation and quality of evidence published by the Infectious Diseases Society of America were used. For microbiologic diagnostics, only criteria for the strength of recommendation were used 8 . These patients with should be treated with a broad-spectrum beta-lactam with anti-pseudomonal activity, as used for empirical treatment of unexplained fever PLUS (if lung infiltrates are not typical for PcP or lobar bacterial pneumonia) mold-active systemic antifungal therapy with voriconazole or liposomal amphotericin B. There is a significant benefit from prompt as compared to delayed mold-active antifungal therapy. In patients pretreated with voriconazole or posaconazole for systemic antifungal prophylaxis and in whom a breakthrough filamentous fungal pneumonia is suspected, measurement of antifungal drug levels and invasive diagnostic procedures should be discussed, and treatment should be switched to liposomal amphotericin B. Systemic antifungal treatment should be continued until hematopoietic recovery and regression of clinical and radiological signs of infection. In patients who had not received cotrimoxazole for prophylaxis and have a CT scan suggesting PcP, and who have a rapid rise of serum lactate dehydrogenase, prompt start of high-dose trimethoprim-sulfamethoxazole therapy should be started, even before bronchoscopy is performed. Antiviral agents such as ganciclovir are not recommended for early pre-emptive therapy, and glycopeptides, fluoroquinolones or macrolide antibiotics without a specific pathogen documented from clinically significant samples should not be used as well. In individual patients undergoing high-dose chemotherapy and autologous hematopoietic stem cell transplantation with febrile neutropenia and LI of unknown origin, whose conditioning regimen included total body irradiation or who have been treated with alemtuzumab, anti-thymocyte globulin or fludarabine, cytomegalovirus should be considered. If a rapid culture or "immediate early antigen" is positive, prompt ganciclovir treatment is recommended, while antifungal therapy should not be given pre-emptively in this patient cohort. The interpretation of microbiological findings, typically originating from blood cultures or bronchial lavage samples, may be difficult. If etiologically significant pathogens are detected, particularly multidrug-resistant bacteria, critical revision of antimicrobial treatment therapy is recommended. For Pseudomonas aeruginosa pneumonia, primary combination antibacterial therapy including an antipseudomonal beta-lactam plus an aminoglycoside or ciprofloxacin is preferred, however, adequate beta-lactam monotherapy may also be appropriate 9 . Depending on their in-vitro susceptibility pattern, multi-resistant gram-negative aerobes such as Extended-Spectrum-Beta-Lactamase-producing E.coli, Enterobacter spp. or Klebsiella spp. as well as Acinetobacter spp. or P.aeruginosa require antimicrobial treatment selected appropriately according to this pattern. In individual patients, aerosolized colistin has been successfully used as a part of the antimicrobial strategy. For Stenotrophomonas maltophilia pneumonia, high-dose trimethoprim-sulfamethoxazole (15 to 20 mg/kg/day of trimethoprim) is recommended. In individual patients, tigecycline-based treatment may be an appropriate alternative 10 . Pneumonia caused by methicillin-susceptible Staphylococcus aureus should be treated with oxacillin or flucloxacillin, while methicillinresistant Staphylococcus aureus (MRSA) should preferably be treated with vancomycin. Linezolid is a possible alternative for first-line treatment. First-choice antiviral treatments for cytomegalovirus pneumonia are foscarnet or ganciclovir. For treatment of possible, probable or proven pulmonary aspergillosis, intravenous voriconazole or liposomal amphotericin B are recommended, while for mucormycosis (zygomycosis), liposomal amphotericin B is preferred. Combination antifungal first-line treatment in patients with invasive mould infections is controversial. If Pneumocystis pneumonia is suspected, treatment with trimethoprimsulfamethoxazole (TMP/SMX; co-trimoxazole) at a dosage of TMP 15-20 mg/kg plus SMX 75-100 mg/kg daily should be initiated immediately after asservation of representive samples (induced sputum or BAL). Delay of treatment onset may increase the mortality rate. Therapy should be continued for at least 2 weeks. In patients with respiratory failure, systemic corticosteroids are controversially discussed. Recent studies could not show a clinical benefit 11 . This article highlights the importance of the acute respiratory failure (ARF) in patients with hematologic malignancies (HM) and stem cell transplant (SCT) recipients in terms of prevalence, prognostic impact, indications and limitations for intensive care unit (ICU) therapies, early recognition of patients at risk, as well as therapeutic management with focus on ICU specific issues. The ARF depicts the most common reason for ICU admissions in patients with HM and SCT recipients. [1] [2] [3] In fact, the development of critical illness without respiratory involvement is rather uncommon in these patients. 4 Most cases of ARF are caused by pulmonary infections or extrapulmonary sepsis. 5 However, in newly diagnosed acute leukemia, ARF is predominantly caused by disease-related mechanisms, such as hyperleukocytosis, pulmonary leukemic infiltration, or acute lysis pneumopathy. 6 Patients with undetermined causes of ARF have worse outcomes. Thus, identifying the underlying aetiology in order to establish the proper causative treatment is of upmost interest. Today, several prospectively evaluated algorithms are available, which assist physicians to choose the most appropriate diagnostic strategy in individual cancer patients with ARF. 7, 8 Prognostic impact The ARF is one of the most important prognostic factors in patients with HM and SCT recipients and independently associated with adverse outcome. 1, 3 Trivial clinical signs like cough, sputum, rales, or requirement of only 1 L/min oxygen already correlate with the development of ARF and, therefore, mortality. 4, 9 Over the last decades ARF mortality rates have markedly dropped in patients with HM, but stay above 50% if mechanical ventilation (MV) becomes necessary. 1,5 However, mortality rates of SCT recipients with ARF and MV remain unacceptably high, especially in the presence of uncontrolled or refractory graft-versus-host disease (GVHD). 2,3,10 In principal, ICU treatment should be considered if 1) the critical condition may be reversible from the perspective of an intensivist, 2) the hematologic or long-term prognosis and underlying comorbidities allow improvement of survival or quality of life, e.g. by intensive therapies, and 3) the patient does not refuse ICU treatment. Accordingly, an evidence-based expert consensus recommends offering a "full code" management with unrestricted allocation of ICU resources including MV to critically ill cancer patients in case of: complete remission of the malignant disease, availability of curative therapeutic options (e.g. patients with induction or consolidation chemotherapy), newly diagnosed malignancy with life expectancy >1 year, complications of autologous SCT, myeloma with partial remission, and selected patients with low-grade hematologic malignancies (e.g. chronic lymphatic or myeloid leukemia and low-grade lymphoma). In patients after allogeneic SCT, full code management including MV is recommended in 1) the earliest phase of SCT, 2) in patients with no or controlled GVHD, and 3) in patients with status epilepticus due to posterior reversible encephalopathy syndrome. 11 No ICU ICU admission cannot be recommended on a routinely basis if no lifespan-extending therapies exist for the malignant disease, lifeexpectancy with ongoing anti-cancer therapy is <1 year (which, of course, can be adapted according to the individual situation of the respective patient), the patients has been mainly bedridden within the last 3 months, or in case of a patients refusal. In SCT patients, no aggressive ICU treatment such as MV can be recommended in patients with uncontrolled or refractory GVHD. 11 In patients with intermediate hematologic prognosis (not fitting in any of the above-described categories), unrestricted ICU therapy can be offered for a limited period of time with a successive reevaluation of the goals of therapy. Importantly, the development of the clinical courses of these patients cannot be estimated within the first 3 to 5 days after the ICU admission. Those, who progress to multi-organ failure after the 3 rd ICU day have markedly reduced chances of survival. However, end-of-life decisions should not be taken before the 3 rd ICU day. 12 Patients with dyspnoea or any respiratory events should be evaluated thoroughly concerning the presence or onset of ARF. Especially patients with neutropenia should be managed rigorously concerning diagnostic procedures and (empiric) antimicrobial treatment. 13 Early recognition of patients at risk and prompt transfer to the ICU at an early phase of ARF is recommended. Although no definite cut-off parameters have ever been evaluated prospectively, PaO 2 /FiO 2 ratios < 200, respiratory rates >30/min, or any combination with ≥1 organ dysfunction seem to be agreeable triggers for an ICU referral. However, decisions for an even earlier ICU admission should be evaluated regularly on an individual basis. Besides the need for aggressive diagnostic measures and specific treatment of the underlying ARF trigger, choosing the appropriate mode of MV deserves special considerations. According to the landmark-trial of Hilbert et al, immunocompromised, mainly HM patients with pulmonary infiltrates and ARF (PaO 2 /FiO 2 ratio < 200) profit from non-invasive ventilation (NIV) via a face mask in the ICU setting in terms of reduced intubation and mortality rates. 14 However, a trial in SCT recipients could not reproduce these findings in the setting of a standard care hematology ward. 15 Furthermore, several case control studies and observational data have shown conflicting results on the impact of NIV in HM and SCT patients. [16] [17] [18] [19] However, patients with secondary intubation after NIV failure have consistently been shown to have markedly reduced survival rates. Taking the existing evidence in HM and SCT patients into account, NIV can be recommended early in the course of mild to moderate ARF without additional extra-pulmonary organ failure and without risk factors for NIV failure. 20 Further detailed evidence on the impact of NIV in immunosuppressed patients shall be derived from the currently ongoing large prospective IVNIctus-trial. 21 A series of investigations documented higher mortality rates in patients with delayed admissions to the ICU. 1, 22, 23 Although not proven in prospective trials, early admissions are thought to be capable of preventing evolution to multi-organ failure. In fact direct admission to the ICU in patients with high-risk acute myeloid leukemia and physiologic disturbances but no organ dysfunction seems to be associated with improved outcome regarding rate of intubation and survival. 24 As of today there is no compelling evidence to perform ICU measures in HM or SCT patients in a standard hematology ward. However, early consultation of an intensivist may be useful not only with regard to developing an appropriate strategy concerning MV, but also concerning general ICU management issues, such as fluid balance. 25 Interestingly, the use of NIV with curative intention concerning the ARF in cancer patients who have declined tracheal intubation is associated with considerable 90-day survival rates. In addition the ICU burden of survivors and their relatives is in favour of offering curative NIV as ceiling therapy to selected patients. 26 The use of extracorporeal membrane oxygenation (ECMO) has been described in a very limited number of HM patients and SCT recipients in therapy-resistant acute respiratory distress syndrome or as "awake ECMO" to avoid intubation. 27, 28 Although this approach may be a therapeutic alternative for selected patients in the future, it has to be regarded experimental at this point. The acute respiratory failure is a frequent and serious complication of the underlying disease or its therapy in patients with HM and SCT recipients. Due to its prognostic impact and tendency for rapid progression, the ARF can be considered as a medical emergency. Awareness, early recognition, rigorous assessment of the underlying causes and initiation of the appropriate treatment must be instituted promptly. In patients eligible for a full code management or an ICU trial, early referral to the ICU and close collaboration between hematologists and intensivists should be warranted on short notice. In the absence of additional extra-pulmonary organ dysfunctions, an early NIV trial may be performed. Optimal induction pre-supposes that achieving complete remission is an essential requirement for durable survival. This is generally accepted, but perhaps may be different in older more frail patients. Improving the rate of CR is desirable and measureable by microscopy, but optimising induction, even if it does not change the CR rate is still important because it can reduce subsequent risk of relapse. For nearly 3 decades "3 + 7" comprising an anthracycline and Cytarabine has been the standard of care, with a strange reluctance to try alternatives. Other anthracyclines than daunorubicin or idarubicin have failed to make a consistent difference to survival based on an extensive meta-analysis. Dose intensification of daunorubicin to 90mg/m 2 X 3 has been shown to be superior to a 45mg/m 2 dose, but 60mg/m 2 is widely used and has until recently never been compared with 90mg/m 2 . In a recent large randomised study with 1200 patients randomised there was no evidence that 90mg was better overall or in any subgroup. Escalation of the Ara-C daily dose has similarly not resulted in a consistent survival benefit, and there seems to be little difference between a continuous infusion or twice daily dosing. Adding a third drug such as thioguanine or etoposide is unconvincing. There has been recent interest in newer nucleoside analogues. Substitution of cytarabine with clofarabine in combination with daunorubicin did not improve outcome in older patients, but the addition of cladrabine to DA has recently delivered interesting results, particularly in patients with adverse cytogenetics where it showed a survival benefit. An alternative mechanism of delivering the D and A combination is in a fixed ration, stabilised in a liposome (CPX571). A large randomised trial of this approach, based on a successful randomised phase 2 experience is currently underway. FLAG-Ida (fludarabine/Ara-C/ G-CSF and Idarubicin) in various schedules has been widely used for relapsed disease but seldom compared as frontline therapy. The MRC AML 15 trial compared it with DA or ADE. It did not improve the overall remission rate, but 90% of those who achieved CR, did so with one course. It is associated with more myelosuppression, so fewer patients progressed to consolidation, but it was of interest that a non-randomised retrospective comparison of patients only receiving two induction (FLAG-Ida) courses with four courses of DA or ADE + HD Ara-C showed no difference. For those FLAG-Ida patients who did receive FLAG-Ida and consolidation the OS was 66%, which provides an incentive to learn how to deliver this schedule. Flag-Ida had a significantly reduced relapse rate. There has been much interest for more than 10 years in the immunoconjugate, gemtuzumab ozogamicin (GO). One strategy that has been addressed has been to augment remission induction by adding it to conventional induction chemotherapy. Dose is crucial. A single or fractionated dose of 3mg/m 2 is well tolerated. The alternative of 6mg/m 2 is generally well tolerated, but some degree of hepatotoxicity is expected. A recent meta-analysis of the 5 completed trials in adult using individual patient data, showed and overall survival benefit which was present in favourable and intermediate risk patients but not those with adverse risk. The UK group randomised over 800 patients to a comparison of a single dose of 3mg/m 2 vs 6mg/m 2 and found no benefit using the higher dose, with the possible caveat that there may be a benefit for adverse risk using the 6mg dose level. Similar data as seen in the 5 adult trials has emerged in children. It remains to be established whether a repeated 3mg dose is superior to a single 3mg dose schedule. Such studies are planned in both adults and children. What is remission anyway? The accepted conventional definition requires marrow blasts to be <5% as well as recovery of neutrophils and platelets to 1.0 x 10 6 /L and 10 x 10 9 /L respectively. During the early experience with GO, the observation was that several patients achieved the marrow criteria without full peripheral blood count recovery. This appeared to have some survival benefit, but usually less than that of a true CR. So the definition of CRi and CRp was born. There is a certain imprecision in this definition, depending on when the marrow and count recovery were assessed. For example if a marrow is done and shows <5% blasts but on that day the counts are not fully recovered, but a week later without more therapy they have, is that CR or CRi? There is a trend to define CR based on minimal residual disease status in the response marrow. A fair proportion (perhaps 50%) of "remission" remission marrows are MRD+ve, and this has important implications for survival. Recent studies have suggested that patients with CRi are more likely to be MRD +ve. So far MRD has not been accepted as a surrogate for survival benefit, and it is unclear that the +ve patients are the ones who will benefit from intensification treatment. It maybe that the negative patients are the most curable with intensification. This issue needs to be carefully resolved by randomised comparisons. There are several alternative induction therapies to the traditional "3+7" approach. Some of these require prospective confirmation in further trials, and some options will be delayed due to due unavailability (e.g.GO). Although in young patients the remission rate is already high, and therefore difficult to improve upon, a better "quality" of remission may be achieved, which manifests itself, not in a changed remission rate, but in a reduction of relapse risk. It has always been accepted that the more sensitive patients will get into CR. There is now the opportunity to corroborate this with objective MRD techniques. But it should not be assumed that the negative patients do not need intensification. Trials to test this will need to be done. The advent of reduced intensity conditioning regimens has permitted the extension of a potentially curative graft-versus-leukaemia (GVL) effect to older patients with haematological malignancies in whom allogeneic transplant was previously precluded because of excess toxicity. Nowhere has this advance been of more clinical significance than in the management of AML where allografting is now a central component of the treatment algorithm in selected fit older patients. A number of substantial issues remain to be resolved however if allogeneic transplantation is to be deployed to maximal effect in this population. Chief amongst these is the identification of patients most likely to benefit both on grounds of tolerability of the attendant transplant related toxicities. Beyond this the optimal conditioning regimen for older patients with AML remains to be resolved both with regard to anti-tumour activity and transplant toxicity, particularly graft-versus-host disease (GVHD). Separately the development of post-transplant pharmacological or immunotherapeutic interventions raise the possibility of uncoupling GVHD from GVL with the aim of improving relapse free survival. Abundant evidence attests to the potency of the graft-versus-leukemia (GVL) effect in adults allografted for acute myeloid leukemia (AML)(1). The decreased toxicity of reduced intensity conditioning (RIC) regimens now permits the exploitation of an alloimmune anti-tumor response in the large number of older patients with AML whose outcome with myelosuppressive chemotherapy can be predicted to be poor (2) . Recent cooperative group and registry studies confirm the promise of RIC allografts in delivering long term survival older patients with high risk AML but organ toxicity, graft-versus-host disease (GVHD) and disease relapse remain major causes of treatment failure. Importantly all of these transplant complications demonstrate increased frequency and different characteristics in older patients and therefore require specific attention if transplant outcomes are to be optimized in this particular population. At the same time challenges remain in both deciding optimal treatment options in patients with primary refractory or relapsed AML and the identification of older patients fit enough to safely proceed to allograft. Considerable progress has been made in the identification of older patients in whom allogeneic SCT can be predicted to be associated with an acceptable toxicity. Central to this process has been the development of the Hematopoietic Cell Transplant Comorbidity Index (HCT-CI)(3). This has recently been shown to accurately predict both early transplant related mortality and overall survival in patients over the age of 60 years transplanted using either a T replete stem cell inoculum or in vivo T cell depletion (4, 5) . Additional precision is however required in terms of appropriately weighting individual co-morbidities aswell as prospectively validating this scoring system in the setting of individual conditioning regimens with particular reference to both stem cell source and the degree of patient: donor HLA disparity. Alternative pre-transplant scoring systems such as the pretransplantation assessment of mortality (PAM) will provide valuable data in this regard (6) . Acute and chronic GVHD remain major challenges in older patients transplanted using T replete stem cell inoculums making the use of in vivo T cell depletion an attractive form of GVHD prophylaxis in this population. Both the use of ATG and alemtuzumab have been shown to reduce the incidence and severity of GVHD in RIC allografts without necessarily increasing the risk of disease relapse. There have been no prospective comparisons of ATG and alemtuzumab in RIC allografts but one registry study has suggested a higher risk of relapse in patients with AML allografted using an alemtuzumab based regimen. Alternative strategies aimed at reducing the risk of GVHD such as the infusion of regulatory T cells or more effective treatment of acute and chronic extensive GVHD using extracorporeal photapharesis are of particular interest in the elderly given their increased risk of GVHD (7) . Disease relapse remains the major cause of treatment failure in patients allografted for AML. The increased frequency of adverse risk cytogenetics in older patients with AML accordingly increases the risk of disease relapse in this patient population. The great majority of patients destined to relapse after an allograft for AML do so within the first year posttransplant and consequently strategies with the potential to reduce the risk of disease relapse focus around either intensification of the conditioning regimen, without a concomitant increase in transplant toxicity, or optimization of a GVL effect early post-transplant. The development of sequential RIC regimens incorporating a prior course of myelosuppressive chemotherapy followed by a fludarabine/ low does TBI (FLAMSA) based RIC allograft appear to have the capacity to reduce the risk of disease relapse in patients allografted for high risk AML (8) . There remain questions concerning their toxicity in older patients and this has led to the substitution of intravenous busulphan for low dose TBI. The activity and tolerability of this modified FLAMSA-Bu regimen compared with standard RIC regimens is currently being examined in a prospective randomized trial in the U.K. An alternative approach towards intensification of the conditioning regimen without increasing toxicity is the use of BCNU combined with fludarabine and melphalan which also shows encouraging Phase II data in older patients (9) . A number of strategies aimed at optimization of a GVL effect in patients allografted for AML are currently under investigation. Prosaically it is important to remember that the intensity of post-transplant cyclosporine (CsA) exposure, particularly in the first 21 days post-transplant, is an important determinant of disease relapse, particularly in recipients of T cell depleted allografts (10) . Consequently meticulous attention should be paid to the degree and duration of immunosuppression delivered posttransplant with adjustments made according to donor stem cell source, degree of donor:patient HLA disparity, the use of T cell depletion and crucially the predicted risk of disease relapse. The administration of donor lymphocyte infusions (DLI) either prophylactically in patients deemed to be at high risk of relapse or pre-emptively in the setting of molecular or immunophenotypic evidence consistent with relapse is of considerable interest. The main challenges associated with the use of DLI in this setting include the attendant risk of GVHD, particularly in recipients of alternative donor transplants or when DLI is administered in the first six post post-transplant. An additional important limitation of this approach is that it perforce can only be implemented a number of months post-transplant when a significant number of patients have already relapsed. The posttransplant administration of targeted therapies with the capacity to manipulate either the kinetics of disease relapse or agents with the potential to manipulate the alloreactive response therefore has the potential to overcome this limitation. One area of active interest is the posttransplant administration of epigenetic therapies, such as the DNA methyltransferase inhibitor azacitidine, which both up-regulates the expression of epigenetically silenced putative tumour antigens and hastens T regulatory cell reconstitution post allograft (11) . RIC allografts represent one of the most important recent therapeutic advances in the management of AML in older adults. It is becoming increasingly apparent that the complications of transplant have particular features in this patient population requiring the development of specific peri-and post transplant strategies. At the same time the challenge of how best to identify older patients with AML in whom the toxicity of allogeneic transplantation is acceptable will become even more acute as we move towards the prospect of universal donor availability. The author declares that he has no conflict of interest Immuno-oncology is a rapidly evolving field, with recent therapeutic advances seen in the development of medicines that enhance the ability of the immune system to eradicate tumors. Stimulating the anti-tumor activity of T cells has emerged as an effective therapeutic strategy, using methodologies such as chimeric antigen receptor-expressing T cells, vaccines, immune checkpoint inhibitors, and BiTE® antibody constructs. BiTE® molecules are single-chain Fv antibody constructs designed to bind with one arm to a tumor-specific cell-surface antigen and with the second arm to CD3, which is part of the T-cell receptor complex. Redirecting T cells to tumor cells results in T-cell activation, proliferation, and serial tumor-cell lysis. 1 BiTE® antibody constructs bring the T cell and tumor cell in close proximity to one another, thereby inducing the formation of a cytolytic synapse between the effector and target cell and triggering the T-cell receptor signaling cascade. 2 Granules containing granzyme and the pore-forming protein perforin fuse with the T-cell membrane and ultimately induce apoptosis of the tumor cell. 3 Expression of MHC class I receptors is not required for BiTE®-mediated tumor cell lysis, indicating that BiTE® activity does not depend on a specific T-cell receptor or peptide antigen presentation. The BiTE® mechanism of action is therefore therapeutically advantageous in that it may circumvent escape mechanisms frequently employed by tumors to evade T-cell-mediated immune responses. This article focuses on two BiTE® antibody constructs that have been developed for the treatment of hematologic malignancies: blinatumomab (bispecific for CD3 and CD19) and AMG 330 (bispecific for CD3 and CD33). CD19 is a highly specific B-cell marker that is expressed throughout Bcell development and on >90% of B-cell lineage cancers. 4 Blinatumomab is designed to redirect CD3+ cytotoxic T cells to CD19+ normal and malignant B cells. 5 Blinatumomab was first tested in clinical studies in patients with NHL and subsequently in patients with ALL. The serum half-life is 2-3 hours, and so it is administered by continuous IV infusion via a portable mini-pump to maintain constant steady-state serum concentrations. Continuous IV infusion of blinatumomab induces predictable pharmacodynamics according to its mode of action. Complete depletion of peripheral B cells occurs rapidly after initial administration, resulting in the sustained absence of circulating B cells throughout treatment. T-cell activation and redistribution also occur rapidly, followed by T-cell expansion due to the proliferation of memory T-cell subsets (CD8+ T EM and T EMRA cells, and CD4+ T CM and T EM cells). A transient release of cytokines occurs within 48 hours after initial administration (mainly IL-6, IL-10, and IFN-γ), predominantly in the first cycle of treatment. 6 Cytokine release appears to correlate with dose of blinatumomab and with B-cell counts. Cytokine release syndrome (CRS) as a clinically significant adverse event has been observed in patients with ALL, 7 but can be mitigated by step-wise dosing of blinatumomab in the first cycle (i.e., 9 μg/day on days 1-7 and 28 μg/day on days 8-29) and premedication with dexamethasone. No differences have been detected in the timeframe and hierarchy of cytokine release between clinical responders and nonresponders, but slightly lower mean cytokine concentrations were observed in nonresponders. Blinatumomab has been shown to induce rapid and long-term responses in patients with relapsed or refractory NHL and ALL, both representing difficult-to-treat populations with historically poor rates of response and overall survival. In patients with relapsed/refractory diffuse large B-cell lymphoma, the most recent analysis of 21 evaluable patients from a phase 2 study demonstrated an overall response rate of 43%, with a median response duration of 11.6 months. 8 The activity of blinatumomab was examined in a pivotal single-arm phase 2 study in a population of adult patients with relapsed/refractory ALL with particularly high risk of poor outcome. 9 Of 189 patients treated, 43% achieved a CR/CRh (CRh = complete response with partial recovery of peripheral blood counts). The median relapse-free survival was 5.9 months and the median overall survival was 6.1 months. A molecular response (MRD negativity) was achieved by 82% of the evaluable patients with a complete response. Three patients had Grade 3 CRS. Neurologic events, which were typically reversible, occurred in 11% (Grade 3) and 2% (Grade 4) of patients. Based on these data, blinatumomab was recently approved by the FDA for the treatment of Philadelphia chromosome-negative relapsed or refractory B-cell precursor ALL. Overall survival of patients with AML is poor, and standard treatment with intensive chemotherapy regimens followed by stem cell transplant is associated with significant morbidity and mortality. The CD33 antigen is expressed at varying levels on >99% of AML patient samples and correlates with known cytogenetic and phenotypic abnormalities, providing a rationale for its potential as a therapeutic target. AMG 330 is a novel investigational BiTE® antibody construct directed at CD33 and CD3, developed for the treatment of AML. 10 In long-term co-cultures of primary AML samples, incubation with AMG 330 led to expansion of T cells from within the patient sample and caused significant AMG 330mediated T-cell lysis of leukemic blasts, even at low effector to target ratios. Evaluation of AMG 330 in patients with AML is warranted. Other BiTE® antibody constructs are also under development targeting different antigens with specificity for various types of solid tumors ( Table 1) . According to the FAB classification APL is designated as "M3 leukemia" and assigned to the WHO defined type of AML with recurrent cytogenetic abnormalities, "acute promyelocytic leukemia with t(15;17)(q22;q12), (PML/RARα) and variants". Detection of PML-RARα fusion is carried out by conventional cytogenetics including fluorescence in situ hybridization (FISH) and/or RT-PCR. Alternative fusion partners are the zinc finger (PLZF), the nucleophosmin gene (NPM), the nuclear mitotic apparatus (NUMA) or the STAT5b gene. These fusion partners may be of therapeutic relevance, e.g. for the sensitivity to all-trans-retinoic-acid (ATRA), which does not exist with the involvement of the PZLF gene [t(11;17)(q23;q21)]. In addition to age, the most important prognostic factor is the leucocyte count at diagnosis which divides patients into a high risk (>10 Gpt / l) and a non-high-risk group ( 10 Gpt / l). Through use of a risk-guided therapeutic approach the outcome of patients within both groups including risk of relapse has become almost comparable 1-3 . After a common initial induction treatment (AIDA), subsequent riskguided (high-risk: WBC ≥ 10, non high-risk: WBC <10) consolidation cycles have been shown to equalize the risk of relapse between both APL risk groups. Current efforts focus on the reduction of toxicity associated with standard therapy including the implementation of less toxic and highly effective agents such as ATO. ATO is the most effective single agent in APL 4;5 . ATO acts via complex mechanism in APL that is not yet fully explained. At a high concentration (0.5 to 2.0 μmol/L) ATO induces apoptosis in vitro, while at lower concentrations (0.1 to 0.5 μmol/l) it induces partial differentiation of leukemic promyelocytes through PML/ RARα degradation and inhibits angiogenesis [6] [7] [8] [9] . ATO is licensed for the treatment of relapsed and refractory APL in the USA and Europe [10] [11] [12] [13] [14] [15] [16] [17] [18] and can achieve remission rates in up to 90 % of patients. Concerning its toxicity profile, ATO is usually well tolerated and its use is associated with a series of manageable adverse events (hyperleucocytosis, increase of liver enzymes, APL differentiation syndrome, prolongation of the QT interval 19 ). Most of the adverse events mentioned above are usually mild and manageable 10;18 . The antileukemic efficiency of ATO is increased when combined with ATRA. ATO as single agent is able to induce durable molecular remission after two cycles in the majority of patients treated for disease recurrence 17 . Results of various studies conducted with ATO as single agent or combined with ATRA for newly diagnosed APL patients reported CR rates of 86-95 %, molecular remission rates after two cycles of 76-100 % and survival rates of 86-88 %, with significantly better responses being obtained in patients with non-high risk disease as compared to high-risk patients [20] [21] [22] [23] [24] . Recent data of the APL0406 -Intergroup Study (GIMEMA/ AML-SG/SAL) showed that ATO plus ATRA is at least as effective as AIDA-based therapy as first-line treatment in non-high risk APL patients 1;24 . In particular, early mortality was almost absent in the experimental treatment arm combining ATO and ATRA. In addition, the results of a recent published randomized trial evaluating ATO in first-line therapy during consolidation demonstrated that ATO further reduced the risk of recurrence and improved survival 25 . Thus available data strongly suggest that at least non-high risk APL patients may be cured without chemotherapy (with ATO/ATRA only), which is a pivotal step in the treatment of APL 26 . Based on this APL0406 study in non-high-risk a randomized study (APOLLO) in high-risk APL is planned which will compare standard treatment with an ATO/ATRA combination supplemented with low-doses of conventional chemotherapy only during induction. The author declares that he has no conflict of interest Aim: In this study we aim to dissect the clonality of the leukemiainitiating cell (L-IC) population in MLL+ infant ALL, a very aggressive form of leukemia occurring within the first year of life, with a poor outcome. Method: To achieve this goal we used a xenotrasplantation assay into immune-deficient mice, and performed Ig/TCR rearrangement analysis and genome-wide SNP array screening of diagnostic samples from patients and xenograft leukemias from transplanted mice Results:We have observed that MLL-AF4+ infant ALL is composed of a branching subclonal architecture already at diagnosis, frequently driven by an Ig/TCR-rearranged founder clone. Each clone is potentially able to contribute to the diverse phenotypic pool of MLL-AF4+ L-ICs. Interestingly, some clones which were minor at diagnosis can reactivate and become dominant upon serially transplantation in mice, whereas other major dominant clones at diagnosis can become quiescent and disappear. Investigation of paired diagnostic and relapse samples suggested that relapses often occur from subclones already present (but quiescent) at diagnosis. Finally, Copy-number alterations arising at relapse might contribute to the activation and expansion of previously quiescent subclones. , to characterize acute leukemias. In the model we primarily focused on the dependence of the duration from the first mutation to the diagnosis of the leukemic disease (time-to-diagnosis) on leukemic stem cell's (LSC) selfrenewal and proliferation rates. We simulated a healthy hematopoietic system and added a small amount of mutated stem cells with predefined proliferation (range: 0-2) and selfrenewal (range: 0-1) to the system. The development of a leukemia is observed for 2000 days and the time-to-diagnosis (days) is calculated for each parameter combination. Results: The model shows that even slowly proliferating LSC can become a leukemia when selfrenewal is high enough. The analysis of the time-to-diagnosis showed that LSC with high proliferation and high selfrenewal are responsible for "fast" leukemias, i.e., lead to a short time between initial mutation and diagnosis, while decreased proliferation or decreased selfrenewal leads to "slower" leukemias. We could also show that selfrenewal is the more sensitive parameter with regards to "leukemic pace". In particular, there seems to exist a lower threshold for selfrenewal. Below this threshold no leukemia occurs any more, at least in the model. Conclusion: Mechanistic mathematical models are useful tools to analyze hematological diseases. In particular, we analyzed the emergence of leukemia before therapy starts to generate an individualized model based risk classification, i.e. the mechanistic model can be used as basis for a statistical model to fit longitudinal patient data to determine a patient's "leukemic pace type". The next step will be to incorporate chemotherapy into the model and generate hypotheses about optimal therapy schemata. Conflict of interest: None Chromosome 17p deletions are among the most frequent genetic abnormalities in various cancers and associated with a dismal prognosis. However, they have long been simply considered as the inactivation of tumor suppressor gene TP53. Previous work in our laboratory identified additional tumor suppressor genes next to TP53, indicating that large deletions on 17p may contribute to cancer biology beyond their impact on eliminating wildtype TP53 allele. To explore this further, we generated a novel geneticallyengineered mouse model that incorporates conditional deletion of chromosome 11B3, which is syntenic to the common deletion region on human 17p13. We found that besides p53, heterozygous deletion of linked genes on 11B3 not only promotes Myc-driven lymphomagenesis or Nf1; Mll3-defective acute myeloid leukemogenesis as seen by shorter tumor latency and overall survival than controls with only p53 loss, but also contributes to the poor outcome of chemotherapy treatments as shown by additional resistance to cyclophosphamide, vincristine and methotrexate. Furthermore, most tumor cells generated from heterozygous deletion of 11B3 have spontaneously missense or frame-shift mutations on the other wildtype p53 allele during the loss of p53 heterozygosity, which represents the major p53 configurations in human cancers. We further identified key players in 17p13 region that is capable of promoting Eμ-Myc lymphoma development by its own or cooperating with p53 suppression in tumorigenesis. To Introduction: The nature of leukemia initiating cells (LIC) in acute lymphoblastic leukemia (ALL) remains elusive, although being of key clinical relevance since LICs need to be eradicated in order to cure the disease. The hierarchical concept with few LIC being able to propagate leukemia upon transplantation onto immunodeficient mice has been challenged by data showing LIC-activity also in committed cells supporting a stochastic model. However, as a common strategy, these studies are based on defining sub-populations by distinct surface marker profiles. Material and methods: Aberrant energy metabolism was reported in cancer including altered redox states in cells with stem cell properties. In this study, we analyzed redox states of primary ALL cells and analyzed whether cellular energy metabolism determines LIC activity. Patient derived xenograft ALL samples were investigated for reactive oxygen species (ROS) in combination with cell cycle analysis. Cellular sub-fractions sorted according to high or low ROS levels were analyzed for repopulating activities in our NOD/SCID/huALL xenograft model. Interestingly, we identified a low oxidative state in cells of early G0/G1 cell cycle phases, which also show increased NOD/SCID repopulation. In contrast, ROS high cells originate from later cell cycle phases (G2/M) suggesting that the ALL cells' oxidative state is indicative for its leukemia initiating activity. To functionally address this hypothesis, sorted ROS high and ROS low patient derived leukemia cells were transplanted onto NOD/SCID mice. Both sub-fractions led to leukemia engraftment with ROS low cells showing higher repopulating activity compared to ROS high ALL cells. Conclusion: In conclusion, our data indicate that all cells in ALL show LIC activity, with cells of low energy metabolism representing the driving leukemia initiating cell compartment, thus pointing to redox modulation as a potential therapeutic target in ALL. Conflict of Interest: None Purpose: The purpose of this study was to identify genes associated with leukemic stem cell (LSC) activity in primary human acute myeloid leukemia (AML) specimens to gain novel insights into the molecular networks controlling LSC self-renewal. Methods: We combined next-generation sequencing technologies with functional in vivo assays to identify LSC associated genes and used combinatorial cell sorting strategies to determine the enrichment for LSCs in sorted fractions. Results: We identified the G-protein coupled receptor 56 (GPR56) as a novel LSC marker in AML. Sorting experiments revealed a hierarchical architecture, in which GPR56 identified more primitive cells compared to GPR56 negative cells, both in vitro and in vivo. GPR56 + CD34 + double positive cells were most highly enriched for engrafting LSCs not only in fresh samples, but also after in vitro culture in optimized conditions that support LSC activity ex vivo. AML samples with high GPR56 gene expression were significantly enriched for adverse genetic aberrations such as FLT3-ITD, EVI1 rearrangements, and mutations in TP53 and RUNX1. Furthermore, high GPR56 gene expression on its own was associated with poor prognosis in independent patient cohorts. Conclusion: Here we established GPR56 as a novel LSC marker that reveals a leukemic hierarchy and as a factor of poor prognosis in AML. Our data provide evidence that LSC phenotypes are heterogeneous and remain to be defined in GPR56 negative samples. The hematopoietic stem cell differentiates in progenitors and then in mature blood cells. Mechanisms of leukemogenesis block this differentiation. We defined the Cell of Arrest of Cancer (CAC) as the stage of hematopoietic differenciation in which leukemic cells are blocked. We thought that the cell of arrest of acute myeloid leukemias (AML) dictated functionnal properties of AML and was induced by specific genetic anomalies. Patients and methods: We extensively phenotyped normal hematopoietic progenitors to create signatures of CAC of AML. These signatures allowed us to classify 932 AML patients seen in our Leukemia Unit between 2000 and 2012 as well as 142 AML patients published by the Cancer Genome Atlas Research Network project (NEJM, 2013) . This study thus relies on our own data and databases publicly available. Results: Our phenotypes of HSC, MPP, CMP, GMP and GP/MP allowed us to classify AML as HSC-like, MPP-like, CMP-like, GMP-like or GP/MP-like. We validate the phenotypic classification of AML with a transcriptomic signature of normal stem cells and hematopoietic progenitors (HSCP) and gene set enrichment analysis. Based on normal HSCP properties, we show that CAC influences functions of AML as proliferation, chemoresistance and metabolism. Earlier the blockage occurs in the differentiation process and poorer the prognosis of patients is. Identification of the CAC allows to highlight the role of CEBPA in stage -specific blockage. This method gives insight of leukemogenesis mechanisms and defines unexpected stage of blockage. Discussion: We propose a new concept in cancer that we called Cell of Arrest of Cancer and used AML to characterise its validity. The identification of the CAC will permit to study in details leukemic blockage mechanisms in a heterogenous disease as AML. CAC classification will allow to consider new therapeutic approaches including terminal differentiation induction in specific subtypes of AML. Conflict of interest: None Introduction: The detection of several mutations in a blast population of pediatric acute myeloid leukemia (pAML) is supposed to be caused by a clonal evolution from a leukemic stem cell (LSC) to leukemic blasts. Since LSC and potential leukemic subclones are only present in small subpopulations, it has been a major technical challenge to particular analyse only the specific population. To acquire a better understanding of the underlying mechanisms of mutagenesis, clonal evolution and leukemogenesis, the aim of this study was to establish methods that allow the analysis and detection of mutations in single cells of a subpopulation known to contain HSC as well as LSC (CD34+CD38-). Methods and results: The mutational status of sorted single cells from three patients suffering from pAML was analysed and compared to mutations detected at initial diagnosis in DNA from a bulk of bone marrow (BM) cells. Single CD34+CD38-cells were sorted, applied to Whole Genome Amplification (WGA) and whole exome sequencing. The mutational status was compared to mutations initially detected. WGA from single cells resulted in an amount of 29 to 31.7μg DNA from each of five single cells. A 4bp insertion in exon 12 of NPM1, (MutA) initially detected from a bulk of cells, was identified in amplified DNA from single cells in 2/3 patients. Internal tandem duplications in Flt3, indicated by mismatches in the alignment, could be detected in amplified DNA from single cells of two patients. The detected ITD resemble those initially detected in DNA from a bulk of BM cells. Conclusion: Single cell sequencing provides a useful tool to amend the detection of genetic aberrations from a bulk of cells and to confirm the presence of specific mutations in single cells from small subpopulations. It therefore helps to get further insights into the clonal evolution in pAML. Background: Acute myeloid leukemia (AML) is characterized by the accumulation of malignant blasts with impaired differentiation due to recurrent mutations, amongst which IDH mutations occur in 15% of AML patients. Methods and results: We showed both in vitro and in a xenografted mice model, that clinically achievable doses of ATRA are sufficient to achieve a terminal granulocytic differentiation, on primary AML samples and AML cell lines harboring IDH1-R132H mutation. There is no effect at this concentration on the WT controls. This is associated with reduction of both proliferation and colony formation, and further lead to apoptosis, improving then overall survival of mutant xenografted mice. We further showed that ATRA sensitivity is due to a constitutive reduced Lyn activation in the presence of the mutation. This will also sensitizes to clinically achievable doses of dasatinib, a Lyn inhibitor. Moreover, as ATRA induces a brief Lyn activation that transiently reduces ATRA activity, its association with dasatinib synergistically increases differentiation. This association might also be considered for others IDH mutations producing 2-hydroxyglutarate, since treatment with the mutantspecific oncometabolite (eg. 2-hydroxyglutarate) sensitizes AML cells to ATRA-induced differentiation. Summary: Constitutive reduction of Lyn in IDH1-R132H blasts sensitizes this subgroup of AML to both dasatinib and ATRA and results in a highly synergistic effect between these two drugs. Conclusion: As ATRA and dasatinib are already used in the clinic, and because IDH mutations are systematically conserved at relapse, combining ATRA to dasatinib should be promising to achieve a long-term remission for this AML patient subgroup. Conflict of interest: None at the time of enrollment into the trial. In 68% of patients, BMI was ≥ 25, and this cut-off level was chosen for further analysis. Both groups were well balanced concerning genetic risk groups but more male patients were overweight. There was no significant difference in CR rates between both groups (60% vs. 58%). However, in younger patients (<60 years), BMI ≥ 25 was significantly associated with reduced OS and RFS as compared to patients with BMI < 25 (OS: 10.0 vs. 4.9 years, p<0.02, RFS: 2.7 vs. 1.8 years, p<0.02). This result was independent from the type of AML (de novo or secondary), karyotype and FLT3 mutation. Also, BMI conferred a worse prognosis in AML with intermediate and unfavorable cytogenetics and in patients with FLT3-ITD. Multivariate analysis confirmed the increased risk for younger overweight patients of 28% for OS. Interestingly, RNA expression analysis in 474 patients showed not one single differently expressed gene in AML blasts from overweight patients, making overweight a truly independent prognostic factor. Discussion: This analysis establishes overweight as an independent negative prognostic factor in younger patients with AML. This is of particular importance since overweight is the only factor that could be altered in during the course of treatment. Further research is needed to clarify if overweight correction could improve prognosis. Conflict of interest: None Background: Besides genetic alterations, epigenetic changes are recognized as an additional mechanism contributing to AML. To identify marker genes that may provide a basis for improved patient outcome prediction, diagnosis and prognosis we performed a systematically genome-wide screening of DNA methylation profiles across 23.000 CpG islands of 28 AML samples using the Methyl-CpG-Immunoprecipitation assay combined to CpG island microarrays. These profiles were then used to select a subset of aberrantly methylated CpG islands (380 regions covering 15.000 CpGs) for quantitative DNA methylation profiling in a larger cohort of 196 AML patients using MALDI-TOF analysis of bisulfite-treated DNA. Meta-analysis clearly separated a subgroup of CpG island regions showing highly correlated DNA methylation changes that were also marked by H3K27me3 in normal hematopoietic progenitor cells (HPC). The residual group of CGIs not targeted by polycomb group repressors in HPC displayed heterogeneous methylation patterns across patients that clustered with genetic markers. A fraction of AML patients (5/196) displayed aberrant hypermethylation at almost all studied loci, representing a rare CpG island methylator phenotype (CIMP) in AML. These patients present immature leukemia with various chromosomal aberrations but very few mutations and show high resemblance with a recently reported CEBPA methylated subgroup displaying a similar strong methylation signature (Figueroa et al, 2009 ). DNA methylation analysis revealed that most of these CEBPA-silenced cases fall into our CIMP category. Method: To explore the whole range of epigenetic alterations in this rare group of hypermethylated patients we performed in-depth global DNA methylation analyses. Interestingly, hyper-rmethylated gene promoters represent many transcriptional regulators that are involved in the differentiation of myeloid lineages (e.g. CEBPA, IRF8, GATA2). In addition, these patients frequently show hypermethylation of the TET2 promoter which could result in a loss of maintenance DNA demethylation and therefore successive hypermethylation at CGIs. Conclusion: It is likely that the aberrant silencing of key lineage regulators stabilizes the differentiation arrest in these cells, and that these patients may particularly benefit from therapies that revert DNA methylation. Conflict of interest: None In our study, we have described 7 pts. among clean-up workers with various cytological variants of AML (AML with minimal differentiation, acute myelomonocytic leukemia and acute erythroid leukemia) in whom the overt leukemia was preceded by MDS (15.2% of all AML cases diagnosed in clean-up workers). At the same time, only six (1.5%) cases of preceding MDS were found upon examination of AML patients in general population of Ukraine in our study. Conclusion: Study of the secondary MDS and its relation to AML in post-Chernobyl period seems to be relevant taking into account the data suggesting that MDS risk among A-bomb survivors remains to be significantly increased even in 55-60 years after radiation exposure. Conflict of interest: None The Histone Methyltransferase EZH2 Controls Drug Resistance in Acute Myeloid Leukemia (AML) Leukemia (AML). However, the underlying mechanisms still remain incompletely understood. Here we report that loss of the histone methyltransferase EZH2 and subsequent reduction of H3K27 trimethylation contribute to chemoresistance in AML. Method: In Myelodysplastic Syndrome (MDS) and Myeloproliferative Neoplasms (MPN) EZH2 is often inactivated due to mutations which is associated with poor prognosis. By use of quantitative PCR and immunohistochemistry we show that a decrease of EZH2 mRNA and protein also correlated with a poor prognosis of AML patients indicating a tumor suppressor role of EZH2 in AML. EZH2 is located on chromosome 7q36.1 and it is not yet fully clear whether EZH2 expression is affected in MDS and AML patients with del(7)/del(7q) who are largely refractory to chemotherapy and have a poor prognosis. We found EZH2 levels to be reduced in del(7)/del(7q) AML patients as determined by Western Blot. Notably, the reduction of EZH2 protein levels via treatment with H3K27 Methyltransferase inhibitors or lentiviral knockdown was sufficient to induce chemoresistance of Normal Karyotype (NK)-AML cells in vitro and in a xenograft mouse model. Furthermore, we observed that EZH2 loss occurred during the acquisition of drug resistance in a Tyrosine Kinase Inhibitor-and Cytarabine-resistant AML cell line. Functionally, the loss of EZH2 directly induced upregulation of HOX genes, which suggests a stem-cell-like signature to be associated with the resistance phenotype. Pharmacological inhibition of Cyclin-dependent kinase 1 (CDK1) and treatment with the proteasome inhibitor Bortezomib, respectively, increased EZH2 protein levels and restored drug sensitivity of the resistant cell line. To evaluate the effect of Bortezomib on EZH2 levels in patients`blasts we treated NK-AML blasts ex vivo with Bortezomib. Importantly, in those patient samples in which EZH2 protein levels were increased by Bortezomib we observed a significantly increased cell death. Conclusion: Our data strongly indicate that restoration of EZH2 protein levels e.g. via Proteasome inhibitors and thereby restoration of EZH2 function might be a novel promising approach to increase therapy response in AML. Background: The cytogenetically cryptic t(5;11)(q35;p15.4) is a frequently observed chromosomal aberration in AML, which results in the translocation of the NUP98 and NSD1 genes and expression of a NUP98-NSD1 fusion protein. Due to its association with poor prognosis and dismal clinical outcome, there has been increasing interest to improve the identification, therapy and minimal residual disease monitoring of patients with NUP98-NSD1. Method: In the present study, we applied next generation RNA-sequencing, aCGH and molecular cloning techniques together with capillary sequencing to study NUP98-NSD1 fusion transcripts in adult de novo AML patients with t(5;11)(q35;p15.4). By sequencing full-length NUP98-NSD1 cDNA transcripts cloned from the leukemic patient cells, we were able to identify two in-frame transcripts and one out-of-frame transcript. One of the in-frame transcripts joined NUP98 exon 11 to NSD1 exon 6, while the other joined NUP98 exon 12 to NSD1 exon 6. The latter had intra-exonic splice variation at the end of NUP98 exon 7. The out-offrame transcript joining NUP98 exon 12 to NSD1 exon 6 lacked NSD1 exon 7. The relative expression levels of all transcripts were assessed from the RNA-seq data based on aligned read counts. The RNA-seq data indicated that the in-frame transcripts are expressed at a higher level than the out-of-frame transcript, while also indicating that the transcript joining NUP98 exon 12 to NSD1 exon 6 is the predominant transcript. Background: BCL2 family proteins are key regulators of apoptosis. BCL2 and BCLXL promote survival by preventing mitochondrial proteasome inhibitors than wild type samples (n=13) (mean IC50 42nM vs 170nM) and this sensitivity is strongly correlated with a higher FLT3-ITD allelic burden (p=0.0084, r=-0.719). We uncovered by immunoblotting, electron microscopy and confocal analyses that proteasome inhibitors induce autophagy in AML cells which is responsible for the early degradation of FLT3-ITD leading to the inhibition of FLT3-ITD signaling and to cell death. Moreover, proteasome inhibitors overcome resistance to quizartinib (a potent second generation tyrosine kinase inhibitor) induced by mutations in the kinase domain of FLT3, suggesting that these compounds may prevent the emergence of mutant clones arising from tyrosine kinase inhibitor treatments. Finally, in vivo studies using a mouse xenograft model, confirmed that bortezomib was able to decrease FLT3-ITD protein level and to inhibit leukemic cell growth. Conclusion: Therefore, selecting patients according to FLT3-ITD mutations could be a new way to detect a significant clinical activity of proteasome inhibitors in AML patients Conflict of interest: None Background: Recent data suggest that TRK (tropomyosin-related kinase) signalling might contribute to the pathogenesis of mastocytosis by autocrine and paracrine loops. We have recently demonstrated a potential role of neurotrophins including nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF) and TRK signalling in leukemia development. Material and methods: In the present study, 57 C57BL/6J mice were transplanted with retrovially gene-modified primary hematopoietic stem/progenitor cells (TRKB/BDNF=17, TRKB=8, BDNF=5, tCD34=7 in 3 independent experiments, and TRKA/NGF=8, TRKA=6, NGF=6 in 2 independent experiments). Interestingly, lymphoblastic leukemia was diagnosed in 5 animals with co-expression of TRKB/BDNF, whereas the other 12 animals unexpectedly developed systemic mastocytosis (SM), affecting mainly spleen, liver, and bone marrow with multifocal compact mast cell infiltrates. Mast cells demonstrated mainly features of mature hyper-granular mast cells. Most SM animals followed an indolent course, while leukemic mice survived <6 months after transplantation. In contrast, no animals with TRKB alone, BDNF alone or tCD34 showed SM or other hematological malignancies. These data strongly suggest that activation of TRKB by BDNF is important to promote mastocytosis. Interestingly, the frequency of SM disease in Kit D816V transgenic mice was lower than in our model (8/28=29% vs. 12/17=71%), whereas retroviralmediated expression of Kit D816V even failed to induce SM in transplanted animals. Five out of 8 animals co-expression of TRKA/NGF developed acute leukemia (mainly mixed phenotype acute leukemia and erythroleukemia). It is interesting to know whether the remaining 3 animals from the TRKA/NGF group will also develop mastocytosis. Conclusion: In summary, we provide the first direct evidence for induction of mastocytosis by activation of TRKB in hematopoietic stem/progenitor cells in vivo. Our data indicate an important role of TRK receptors in the pathogenesis of acute leukemia and mastocytosis. with significantly longer latency. Recently, SRC was identified as a crucial interaction partner of FLT3-ITD for STAT5 activation. Method: Here, we have investigated the role of STAT5 for the latency and phenotype of the FLT3-mutant induced diseases using a conditional Stat5 knockout approach. Our studies proved that STAT5 is necessary for the induction of a MPN by FLT3-ITD, while FLT3-TKD is not dependent on activation of STAT5. Mice with deletion of the Stat5 allele displayed expansion of lymphoid cells with an increased latency similar to the FLT3-TKD disease. In additional experiments, we were able to show differential activation of STAT5 in myeloid and lymphoid progenitors. Furthermore, Oncostatin M, a STAT5 target gene, is exclusively expressed in Flt3-ITD + myeloid progenitor cells and in Flt3-ITD + mice with intact Stat5. Strikingly, we could show that FLT3-TKD expression in combination with Oncostatin M is sufficient to induce a MPN, pointing to a pivotal role of STAT5 for induction of FLT3-induced myeloid neoplasms. Furthermore, significantly elevated Src expression levels could be detected in proliferating myeloid progenitor cells compared to lymphoid progenitors. Background: Leukemia-associated-immuno-phenotypes (LAIP) defined by multiparameter flow cytometry (MPFC) are present at diagnosis iñ 90% of Acute Myeloid Leukemias (AML) and are used to identify and measure MRD. We assessed in parallel LAIP profiles and gene mutational status in AML to better identify patients at high risk of relapse. Screening for FLT3 mutations is relevant for patient risk stratification and recent studies suggest that relapse of FLT3-ITD AML may originate from very small subclones which are undetectable at diagnosis using conventional techniques. Method: We have previously developed a patient-specific real-time quantitative PCR (ps-RQ-PCR) to implement FLT3-ITD detection, however this approach is not applicable at diagnosis to detect minor subclones. To improve detection of small populations harboring FLT3-ITD and potentially associated to chemoresistance, 9-color MPFC and molecular analyses were used in parallel. Within the CD34+ve cell fraction, we identified a discrete population expressing high-levels of the IL-3 receptor alpha-chain (CD123) and MIC-2 (CD99) in combination with the IL-2 receptor alpha-chain (CD25). The presence of this population positively correlated with FLT3 mutations. Receiver operating characteristics (ROC) showed that within the CD34+ve cell fraction, a percentage of CD123/CD99/CD25+ve cells≥11.7% predicted FLT3 mutations with a specificity and sensitivity of >90%. CD34/CD123/CD99/ CD25+ve clones were retrospectively evaluated and detected at presentation in 3 patients with FLT3wt/NPM1+ AML, who relapsed with FLT3-ITD/NPM1+ AML. The ps-RQ-PCR designed for each FLT3-ITD confirmed the presence of low copy numbers of the mutation in diagnostic samples in all three cases. Finally, MPFC revealed that the CD34/CD123/CD99/CD25+ve population correlated with absent or low CD38 expression, suggesting that this LAIP may identify leukemia initiating cells. Results: Our data show that patients with FLT3-wild type/NPM1+ AML by routine PCR who carry at diagnosis a CD34/CD123/CD99/CD25+ subpopulation likely harbor small FLT3-ITD subclones potentially associated with chemoresistance and disease relapse. We propose that such patients are kept under "special surveillance" and stringently monitored for possible emergence of FLT3-ITD after remission induction and during follow-up. Purpose: Cancer cells emit microvesicles to their microenvironment to represent an important mode of intercellular communication by serving as vehicles for transfer cytoplasmic proteins and RNAs. Since, microRNA21 is overexpressed in leukemia, this study considered the effect of microvesicles from leukemic microenvironment on microRNA21 expression in umbilical cord blood hematopoietic stem cells. Methods: Umbilical cord blood hematopoietic stem cells sorted by MACS (CD34 antibody), were treated with 30 ug microvesicles isolated from bone marrow samples of patients with acute myeloid leukemia in diagnosis phase (written consent form was obtained) and were divided to three groups: 1) control group without microvesicles, 2) normal group treated with microvesicles from normal bone marrow samples and 3) leukemia group treated with microvesicles from leukemic bone marrow samples. Cell count, qPCR for microRNA21 expression and flowcytometry analysis for CD34 and viability test were performed after 7days. Results: Number of cells in leukemia group showed a significant increase in comparison with other both groups (P<0.05). Also, 2.7 fold change of microRNA21 expression was observed in leukemia group (P<0.01). Surface CD34 results did not show any significant different among all studied groups in comparison with 97% CD34 in day0 sorted hematopoietic stem cells (P>0.05). Cell viability test done by 7AAD was more than 90% in all groups. Conclusion: This study showed that microvesicles from leukemic microenvironment are able to deregulate microRNA21 expression in cord blood hematopoietic stem cells. Recent reports indicate this might be due to different subcellular localization and downstream signaling of the various FLT3 mutants. However, it remains unclear whether FLT3-ITD receptors are consistently retained within intracellular compartments or whether they are rapidly internalized after ligand stimulation. Therefore, we set out to study the differential localization of FLT3 mutants. Methods: FLT3 mutants, namely N676K, D835Y, K644R and two different ITDs, were expressed in U2OS or BaF3 cells and analyzed by confocal and high resolution microscopy in combination with immunofluorescence (IF). To study the subcellular localization of FLT3 and to determine qualitative and quantitative differences upon ligand stimulation we used subdiffraction multicolor 3D structured illumination microscopy (3D-SIM). The FLT3 trafficking/activation kinetics are analyzed by live cell microscopy utilizing a spinning disc microscope. Imaging conditions are optimized for fluorescently tagged expression constructs (FLT3-GFP/ FLT3-RFP), as well as tetracysteine motif tagged expression constructs allowing successive pulse labelling with two different colors. FLT3 inhibition is performed using PKC412 and AC220. Results: Our IF staining confirmed a predominant ER-localization of FLT3-ITD, whereas FLT3 wildtype and point mutations localize to the cell membrane. While ITD activates STAT5 signaling, N676K and D835Y activate the MAPK pathway. Summary: Disease relapse remains a major cause of mortality for patients with acute myeloid leukaemia (AML). FLT3 is a receptor tyrosine kinase expressed by normal and leukemic myeloid progenitors and 25% of AML cases harbor a FLT3-ITD mutation, associated with an even inferior outcome due to increased relapse rate. AML relapse might be correlated to interactions between leukemic blasts and the bone marrow microenvironment (e.g. hypoxia), promoting leukaemia survival e.g. chemo-resistance. Hence, the effects of hypoxia on FLT3 expression could provide novel insight into relapse biology. Material and methods: Patient samples and the murine pro-B-cell line Ba/f3, stably transfected with FLT3-WT and with different FLT3-ITD mutants (W51, W78, NPOS) were used. Experiments were carried out in parallel under standard laboratory conditions and reduced oxygen environment of 1% O 2 . Applied methods comprise cell culture, flow cytometry (cell cycle analysis, apoptosis), real-time PCR and western blot. Survival analyses were done according to the method of Kaplan-Meier. Results: AML patients samples were cultured under hypoxic conditions for 48 hours and 40% (n=6) of those expressing FLT3, showed a down-regulation of total FLT3 protein level independent on the mutational state. This finding could not be observed in any human AML cell line screened (n=14), however both Ba/f3 WT and all three ITD mutants checked showed FLT3 down-regulation after 72 hours at 1% O 2 . This was accompanied by a significant decrease in cell growth as compared to 21% O 2, without significant increase in apoptosis. Real-Time PCR analysis showed no difference in the transcriptional regulation of FLT3 under 1% O 2 compared to 21% O 2. Susceptibility against FLT3-TKI was unaffected despite decreased FLT3 expression. Interestingly, half-life of FLT3 protein was shorter at 1% O 2 (1 hr) as compared to 21% O 2 (2 hrs). Degradation of FLT3-ITD is mediated by proteasome, which could reliably be inhibited under normoxic and hypoxic conditions. Re-oxygenation of hypoxic AML blasts with 21% O 2 revealed that down-regulation of FLT-3 is reversible within 24 hours. The mechanisms of down-regulation for FLT3 seem to be complex: degradation is certainly mediated by proteasome and seems not to be transcriptionally dependent. Clinically, FLT3 regulation status might be of prognostic importance (n=8, relapse free survival (RFS) for non-regulators 2.2 months versus 10.0 months for regulators, p=0.07). Conclusion: This data shows that hypoxia influences FLT3 expression in primary patient samples and the Ba/f3 model. The functional consequences are still unknown, but patients that fail to down-regulate FLT3 tend to have a shorter RFS. The mechanisms involved in down-regulation comprise a proteasome-dependent degradation. Ongoing projects focus on molecular mechanisms responsible for oxygen-dependent regulation of FLT3 and on biological consequences. Conflict of interest: none Purpose: This study comprehensively records the kinetics of global gene expression changes that accompany the transition from a normal hematopoietic precursor to a transformed pre-leukemic cell. Employing MLL-fusion transformed cells as a paradigma of hematopoietic transformation the hierarchy and wiring of control circuits governing the decision between self-renewal and differentiation are visualized. Methods and results: With the help of a knock-in model introducing a conditional MLL-fusion oncogene into the mouse germ-line we are presenting a "reverse-engineering"strategy that allows kinetic sampling of the "transformed to normal" transition. By determination of global transcription rates through nascent RNA sequencing coupled with genome-wide ChIP-Seq probing for oncogene presence and function we identify the crucial downstream mediators of MLL-fusion induced transformation. Combining kinetics with occupancy data allows to discriminate drivers from bystanders. Thus novel players, pathways and principles determining the transforming activity of MLL fusion proteins could be identified. Biological verification experiments involving a select set of hitherto unknown MLL-fusion targets confirmed their role in MLL-induced transformation and revealed the importance of pathways including stable RNA-synthesis, splicing and control of metabolism that had not yet been associated with transformation by MLL-fusion proteins. This comprehensive knowledge of the MLL-triggered transforming network will allow for the rational design of novel inhibitors as potential therapeutic strategies. Conclusion: MLL-fusions are multifunctional oncoproteins that control a discrete and highly redundant network of control circuits that cooperate to keep hematopoietic precursors in an undifferentiated state. The analysis strategy applied in this work can be universally adapted for other oncogenes facilitating the elucidation of different transforming principles. 6 Dept. of Pediatrics, Univ. of Regensburg; 7 Dept. of Pediatric Hematology/Oncology, Univ. of Tübingen, Germany; 8 Dept. of Pediatric Intensive Care Medicine, UKBB Basel; Switzerland, 9 Dept. of Oncology, Dr von Haunersches Kinderspital, LMU München, Munich, Germany Purpose: In the individualized xenograft mouse model of AML, primary patients' AML cells are transplanted into immunocompromised mice. We aimed at improving this model to enable reproducible, sensitive and reliable disease monitoring for preclinical treatment trials. Methods: Primary AML patient cells were engrafted in immunocompromized NSG mice and serially retransplanted. Patientderived xenografted (PDX) cells were lentivirally transduced in vitro to express a fluorochrome and enhanced firefly luciferase. Transgenic t-PDX cells were enriched by flow cytometry, reinjected into mice, and monitored over time using bioluminescence in vivo imaging (BLI). Mice were treated in preclinical treatment trials; effect of drugs on leukemia burden was monitored by BLI. Results: About 30% of primary AML patient samples allow serial transplant-ation in NSG mice and thereby generation of stable PDX cells. PDX cells were derived from genetically distinct patient samples, mimicking the known heterogeneity of AML. Targeted resequencing of 43 genes important for AML leukemogenesis revealed identical mutations in primary and PDX cells after initial or serial transplantation, besides two losses of minor suclones. Until now, five PDX samples were genetically engineering by lentiviral transduction. Recombinant expression of luciferase enabled BLI for reliable follow up of leukemia growth in mice. Imaging was highly sensitive and detected t-PDX cells on MRD level (10 -4 leukemic cells within mouse BM). Furthermore, imaging facilitated reliable analysis of preclinical treatment trials, visualizing drug effects in single mice over time. Introduction: Hematological disorders such as leukemia are associated with disruptions of the microenvironment within the bone marrow (BM) niche. The aim of this study was to establish a 3D perfusion culture system for leukemic cells and human mesenchymal stromal cells (MSC), which constitute an important supportive cellular component of the bone marrow niche. Methods: We used a modular perfusion culture system (Will W. Minuth, Regensburg, Germany). MSC were cultured on porous membrane filters consisting of mixed cellulose which were permanently perfused with fresh culture medium. The viability and induction of apoptosis of the MSCs and leukemic cells in different culture conditions was tested. Results: A perfusion rate of 20.8 μl per minute and a concentration of 50 mM HEPES buffer were observed to be optimal. We could show that the amount of viable MSCs cultured in perfusion culture was considerably higher (OD 450 : 0.37 ± 0.03 vs. 0.2 ± 0.01) than in conventional cell culture. The basal level of apoptotic cells in a co-cultivation of MSC and a leukemic cell line is 1.7x increased in conventional cell culture compared to perfusion culture. MSC showed no differences in the expression of exemplary chosen genes relevant for the BM niche (CXCL12, Jag1) due to the different culture conditions. Additionally, MSC cultured on membranes formed 3D-like networks. No differences in morphology were observed with respect to the different culture conditions. Discussion and conclusion: The 3D perfusion culture system is sufficient to increase the viability and proliferation of MSC without changing the gene expression profile of exemplary chosen genes. The system can be used for further experiments including co-culturing experiments with MSCs and leukemic blasts and might be an important option to mimic the hematopoietic stem cell niche in vitro. The DNA-Repair Machinery as a Novel Target in a Humanized T(8;21)+ Pre-Leukemia Model with Activated C-KIT C. Wichmann 1,2 , L. Chen-Wichmann 1,2 , and R. Henschler 2 1 Georg-Speyer-Haus -Frankfurt, Germany; 2 Ludwig Maximilian University Hospital, AöR, Department of Transfusion Medicine, Cell Therapeutics and Hemostaseology -Munich, Germany Introduction: The RUNX1/ETO fusion protein, which originates from the t(8;21) chromosomal rearrangement, is one of the most frequent translocation products found in de novo acute myeloid leukemia (AML). In RUNX1/ETO leukemias, activated forms of the c-KIT tyrosine kinase receptor are frequently found, thereby suggesting oncogenic cooperativity between these oncoproteins in the development and maintenance of t(8;21) malignancies. Method and discussion: We show that activated c-KIT cooperates with RUNX1/ETO to expand human CD34+ hematopoietic progenitor/stem cells ex vivo. Morphologically and cytometrically, CD34+ cells expressing both oncogenes precisely resemble the AML-M2 myeloblastic leukemia cell phenotype. In contrast, RUNX1/ETO-expressing cells largely undergo granulocytic differentiation. Oncogenic c-KIT amplifies RUNX1/ETO-depended clonogenic growth and protects cells from exhaustion. Of note, activated c-KIT reverts the RUNX1/ETO-induced DNA-damage followed by apoptosis, a recently described RUNX1/ETOdependent phenomenon. In the presence of activated c-KIT, RUNX1/ ETO-downregulated DNA-repair genes including OGG1 and BRAC2 are re-expressed leading to an enhancement of DNA-repair efficiency via homologous recombination. Through this mechanism activated c-KIT protects from RUNX1/ETO-induced cellular stress. This points to a direct link between the high frequency of mutated c-KIT and the occurrence of t(8;21) and herewith provides a rational mechanism of oncogenic cooperativity and drug resistance. Results: Together, our results provide new mechanistic insight into RUNX1/ETO and c-KIT oncogenic cooperativity in human primary CD34+ progenitors and suggest that augmented DNA-repair might account for the increased chemoresistance and decreased overall survival rates observed in t(8;21)-positive AML patients with activated c-KIT mutations. This cell protective mechanism represents a new therapeutic target, as RUNX1/ETO cells with activated c-KIT are highly sensitive to pharmacological inhibitors of DNA-repair. Introduction: The discovery of new drugs and drug targets can help to improve the cure rates of Acute Myeloid Leukemia (AML). DNA methyltransferases (DNMTs) inhibitors represent a group of promising compounds for AML treatment because leukemic cells have altered genome methylation, as hypermethylation of tumor suppressor genes promoters. Method: The present study aims the in vitro evaluation of two DNMTs inhibitors, called B and Q that are nanaomycin A analogs. Cytotoxic activity was carried out in K562 AML cell line by XTT reduction assay, performed 48 hours after addition of compounds increasing concentrations. Enzymatic activity inhibition of DNA methyltransferase DNMT1 and DNMT3B were accessed by Western Blot technique followed by quantitative RT-PCR. In addition, we investigated the methylation of CpG islands located in promoter regions of six tumor suppressor genes, by pyrosequencing, before and after treatment with both compounds. Discussion: K562 showed an IC50 of 10,4uM and 6,6uM for compounds B and Q respectively. At protein level, the treatment with B compound resulted in decrease of both DNMT1 and DNMT3B levels while Q treatment resulted in decreased DNMT1 and increased DNMT3B isoforms, when compared to control sample (DMSO). At mRNA level, the treatment with B showed a decreased DNMT1 with no significant alteration on DNMT3B while the Q treatment showed a decreased DNMT1. We also verified an increased DNMT3B mRNA level when cells were treated with Q compound at 0,5μM. Regarding methylation of tumor suppressor genes promoters, we observed that no ESR1 and RASSF1 genes promoters were methylated in K562 cells since methylation levels of these promoters in control cells (DMSO) were also very low. CADM1, Mir141, MGMT and THBS1 promoter regions showed methylated patterns, however no difference in methylation of these regions was observed between treated and control cells. Conclusion: We concluded that compounds B and Q were both cytotoxic for K562 cell line and decreased DNMT1 protein levels. Besides, compound B also decreased DNMT3b protein levels. Currently, we are working on DNMT1 and DNMT3B in vitro inhibition assay to elucidate the compounds mechanisms of action and on K562 gene expression profiles to understand some of the genes pathways affected by these compounds. Background: MLL-translocations are frequent in infant acute myeloid leukemia (AML). Their leukemogenic potential is widely accepted, but unfortunately MLL-fusions are not amenable to therapeutic intervention. We thus set out to identify novel therapeutic targets downstream of MLL-AF9 with the prospect to improve the still unsatisfactory prognosis of MLL+ AML. Methods: Gene expression profiles were generated and bioinformatically interpreted after efficient and specific silencing of endogenous MLL-AF9 in THP1 (AML) cells. Results: A strategy to prioritize genes which mediate MLL-AF9 leukemogenic effects resulted in a set of 41 out of hundreds of differentially regulated genes. Seven of these qualified as druggable and thus could offer novel therapeutic targets for MLL+ AML. We show that targeting one of the encoded proteins, the dopamine receptor D5, was able to reduce leukemic cell characteristics of THP1 cells. Department of Hematology, Città della Salute e della Scienza, Turin, Italy, 2 Bone Marrow Transplant Center, Città della Salute e della Scienza, Turin, Italy, 3 General Surgery 2U and Liver Transplantation Center, University of Turin, Città della Salute e della Scienza, Turin, Italy Background: Myeloid sarcoma (MS) is an extramedullary localization of immature granulocytic precursor cells. Gastrointestinal tract involvement is a rare phenomenon. A 62 year old man was admitted to our department with fever, abdominal pain and jaundice due to a common bile duct mass constricting the choledochus duct. Method: After endoscopic stent implantation, an explorative laparotomy with cholecystectomy lead to diagnosis of myeloid sarcoma with involvement of cholecystis and choledochus. Further tumor manifestations, in particular bone marrow infiltration, were excluded. Due to present febrile cholestasis subcutaneous low dose Cytarabine (LDAC, 20mg/bid) together with broad spectrum antibiotics were initiated. Staging after one cycle of LDAC documented a significant tumor mass reduction, thus two intensive chemotherapy cycles with Idarubicin, Cytarabine, and Etoposide were performed. During both aplasia periods a biliary drainage change and broad antimicrobical therapy were required because of febrile cholestasis. Consequent re-staging documented a complete remission (CR), therefore a consolidation radiotherapy with 24 Gy was added. After four months of follow-up (FU) fever and petechial skin due to piastrinopenia and leucocytosis with 74% of myeloid blast in peripheral blood appeared. Intensive chemotherapy induced a CR of the AML, thus a second consolidation cycle was performed, both complicated by febrile cholestasis requiring again biliary drainage replacement. Given the fact that a CR of AML and MS was obtained, a semi-intensive Daunorubicin/ Citarabin cycle was given and after myeloablative chemotherapy hematopietic stem cell transplantation (HCT) with peripheral blood stem cells from a 10/10 matched unrelated donor realized. Graft versus host disease prophylaxis contained Ciclosporin, Methotrexate and antithymoglobulin. The aplasia period was complicated by fever without cholestasis. During the recovery the drainage was not opened and the patient could be discharged after 21 days. The patient is currently at 4 month FU maintaining a CR of AML with 100% chimerism and no MS or infection signs. Purpose: Conventional cytogenetic analysis is a mandatory component of AML workup. In older pts (>60 y), presence of adverse cytogenetic changes is associated with poor response to standard induction chemotherapy and short overall survival. Conventional cytogenetic studies are time-consuming, and therapeutic decisions are often delayed until the karyotyping result is available. Thus, we investigated whether a targeted FISH screening would allow the rapid identification of the most common unfavorable chromosomal changes in AML. Patients and methods: We studied 318 newly diagnosed AML pts (263 pts>60 y). Conventional cytogenetic analyses were performed using standard techniques. For FISH analyses probes directed against 7 genomic loci were used: 5p15.2 (D5S721, D5S23), 5q31 (EGR1), 7cen (D7Z1), 7q31 (D7S486), 17p13 (TP53), 17cen (D17Z1) and 3q26.2 (EVI1). The result was considered positive if the signal pattern was consistent with one of the following abnormalities: del(5)(q31) or -5, -7, abn(17p13) or -17, or inv(3)/t(3;3). Results: FISH screening was positive in 75 pts, and 73/75 belonged to the ELN adverse genetic group. Thus, a positive FISH result indicated the adverse karyotype in 97% of cases (pos. predictive value). Contrary, 21 out of 243 FISH neg pts belonged to the adverse group according to their karyotype. These pts mostly carried complex structural changes without involvement of chromosome 5, 7, or 17. Conclusion: The median time from receipt of the specimen until the FISH screening result became available was two days. Our targeted screening assay using a limited number of FISH probes could rapidly identify a large majority of older AML pts with unfavorable cytogenetics and represents a valuable addition to the standard genetic workup of AML. As an additional warranty to minimize the probability of relapse accepted the need to intensify therapy in patients with the expression of the interleukin 2 receptor alpha (CD25) clonal blast cells. Algorithm for risk-adapted induction therapy is as follows. When the content of CD25 + cells before treatment and at least 10% in the absence (0%) of cells with LAF after the first course of remission induction therapy protocol "7 + 3" with escalated doses of anthracyclines (in 14 days from start of treatment) were limited to one course of induction remission, the patient is transferred to consolidation remission therapy with high-dose cytarabine protocol HiDAC. When the content of CD25 + cells before treatment and at least 10% of cells with LAF on day 14 of 0.01% or more was conducted two courses of remission induction protocol "7 + 3", and the first course with escalated, and the second with standard doses of anthracyclines. When the content of CD25 + cells prior to treatment is 10% or more and no (0%) cells at day 14 LAF performed two cycles of remission induction protocol "7 + 3", a first course with escalated, and a second standard doses of anthracycline . When the content of CD25 + cells before treatment to 10% or more, and cells LAF on day 14 of 0.01% or more, conducted two courses of induction of remission, the first course protocol "7 + 3" with escalated doses of anthracyclines, and the second course FLAG-Ida. Result: As a result of applying this algorithm frequency cytomorphological complete remission was 89.5% complete remissions cytomorphological and immunophenotypic -84%. Conflict of interest: None Purpose: The Hedgehog pathway plays an important role in stem-cell biology and malignant transformation. Therefore we investigated expression and prognostic impact of Hedgehog pathway members in acute myeloid leukemia (AML). Methods: Pre-treatment samples from 104 newly diagnosed AML patients (AMLSG 07-04 trial) were analyzed by qPCR and expression of Hedgehog family members was correlated to clinical outcome. Inhibition of GLI by GANT61 or shRNA was investigated in AML cells in vitro and in vivo. Results: Expression of receptors SMO and PTCH1 and their downstream mediators GLI1, GLI2 and GLI3 was detected in AML patients while no expression of Hedgehog ligands was found.GLI2 expression had a significant negative influence on event-free (EFS), relapse-free (RFS) and overall survival (OS; p=0.037, p=0.026 and p=0.013, respectively) and was correlated to FLT3 mutational status (p<0.001). Analysis of a second, independent patient cohort confirmed the negative impact of GLI2 on EFS and OS (p=0.007 and p=0.003, respectively; n=290). Within this cohort, also GLI1 had a negative prognostic impact (p<0.001 for both EFS and OS). Although AML cells did not express Hedgehog ligands, AML patients had significantly increased Desert Hedgehog (DHH) plasma levels compared to healthy subjects (p=0.002). Provision of DHH could be ascribed to BM niche cells which showed elevated DHH expression levels compared to the whole cell BM compartment. Moreover, the GLI inhibitor GANT61 or knockdown of GLI1/2 by shRNA caused anti-leukemic effects including induction of apoptosis, reduced proliferation and colony formation in AML cells and a survival benefit in a mouse model. Results: Feeder cells only slightly prevented spontaneous cells death of patient-derived ALL cells in vitro. While several cytotoxic drugs induced marked apoptosis when used in clinically relevant concentrations in vitro in the absence of feeder cells, the presence of feeder cells markedly reduced apoptosis induction. In certain drugs, reduction of apoptosis by feeders was up to 50 %. Conclusion: From our data we conclude that the environment inherits major inhibitory effects on drug sensitivity of ALL cells and should be considered in novel treatment protocols for ALL. Here, we aimed to functionally discriminate clonal subpopulations in order to link adverse characteristics. Towards this aim, we characterized several single cell clones regarding their functional properties in vitro and in vivo using tumor cells from a patient with ALL. Method: We amplified the patient's ALL cells in immuno-compromised mice within the individualized xenograft mouse model. RGB marking enabled multicolor staining of patient-derived xenograft ALL cells based on lentiviral genetic engineering; RGB marking allowed isolation of individually colored viable single cell clones from the ALL patient. Results: Single cell clones were characterized for their functional properties in vitro and in vivo as well as for phenotypic markers. When 2 differently RGB marked clones were mixed within the same recipient mice, one clone overgrew the other indicating different proliferation rates of both clones. In vitro, chemosensitivity clearly differed between several clones tested, especially regarding stimulation with glucocorticoids, indicating that resistant subclones existed before start of treatment. Introduction: Treatment resistance and relapse in acute lymphoblastic leukemia (ALL) are considered to be due to apoptosis defects. Since inhibitor of apoptosis proteins (IAPs') negatively regulating apoptosis signaling are highly expressed in leukemias, small molecule inhibitor antagonists (SMAC mimetic, SM) provide a promising strategy to overcome apoptosis deficiency and effectively treat high risk ALL. In this study, we addressed and characterized effects and effectivity of SM in ALL. Method: Exposure of ALL cells to SM at nanomolar concentrations led to clear cell death induction in ALL cell lines and 70% of primograft ALL samples (n=42). Interestingly, apoptosis induction could be inhibited by the soluble TNF-a receptor etanercept, thus indicating TNF-a dependency. Previously, we described that rapid NOD/SCID/huALL engraftment is associated with deficient apoptosis signaling and indicative for early patient relapse. Importantly, SM activated constitutive deficient apoptosis signaling in high risk/early relapse ALL. Next, we investigated the effectivity of SM on high risk ALL in a preclinical in vivo setting and treated leukemia bearing recipients either with vehicle, SM, multidrug chemotherapy, or a combination of both. Results: Importantly, a profound reduction of tumor load and prolonged survival of animals was observed upon SM treatment alone, and even more pronounced in combination with multi-agent chemotherapy. Interestingly, concomitant in vivo application of etanercept revoked the SM effect, indicating that SM induced apoptosis sensitization involves signaling via TNF-a also in vivo, thereby pointing to a potential biomarker for the identification of patients who would benefit from SM treatment. Conclusion: In conclusion, the small molecule SM induce cell death ex vivo and in vivo in primary patient-derived ALL overcoming apoptosis deficiency, pointing to a novel strategy for targeted therapy of high-risk ALL. Background: Expression of meningioma 1 (MN1) gene can be associated with ATRA resistance. Purpose of the study was to investigate MN1 expression levels in patients (pts) with ALL treated by all-trans retinoic acid (ATRA) on the protocol ALL-MB-2008. 1 that in relapsed APL patients (treated upfront with ATO) mutations in the B2 domain of PML in PML-RARA gene are involved in resistance and poor clinical outcome (NEJM 2014). We had previously reported that Bortezomib (Bo) was able to synergize with ATO by inducing apoptosis (Blood. 2012;120 ,3552). We further evaluated the mechanism of degradation of PML-RARA when ATO was combined with Bo. Method and results: We generated in-house ATO resistant NB4 cell lines (NB4EV-ASR1 (with A216V mutation), ASR2 and ASR3). The combination of ATO and Bo induced a significant apoptosis in the resistant cells similar to naïve NB4 cells. We observed a time dependent accumulation and degradation of ubiquitinated proteins and ubiquitin binding protein p62 in ATO+Bo treated naïve NB4 cells, correlating with induction of autophagy and degradation of PML-RARA. Similar results were seen in the resistant cells. Blocking autophagy by 3-methyl adenine or knock down of p62 showed an accumulation of PML-RARA in the combination treated cells. We also observed an interaction between P62,LC3II and PML-RARA suggesting that PML-RARA was cleared by autophagy via p62 protein. In a transplantable APL mice model, combination prolonged the life span of the mice. A reduction in the LIC was demonstrated by secondary transplantation experiments. A phase II clinical study combining Bo with ATO and chemotherapy has been initiated for relapsed APL (NCT01950611). 11 patients have been enrolled, all achieved complete molecular remission. The combination was well tolerated. Beyond induction the combination does not require in-patient care and there is no significant cytopenia. Background: Mass cytometry allows for multi-parametric analysis of single cells to distinguish phenotypic and functional features of subpopulations from leukemia samples. Mass cytometric analysis of pediatric Ph+ BCP ALL constructs a novel model of ALL organized along the normal B cell developmental trajectory that may be used as a lens through which to examine signaling features, drug response, and clinical data. Methods: Mononuclear cells from diagnostic bone marrow samples were obtained from pediatric patients with BCP-ALL (n=52) or healthy controls (n=6). Mass cytometry analysis of 40 proteins was performed at baseline state and perturbed state (cytokines and tyrosine kinase inhibitors). Healthy bone marrow samples were gated as previously described along the trajectory of developing B cells (Davis et al., 2014) . These populations were used as the foundation for a classifier in which each leukemia cell was assigned to its nearest healthy population based on Mahalanobis distance in nine dimensions. Results: ALL samples demonstrated phenotypic heterogeneity but maintained features consistent with stages of healthy B cell development. The single-cell classifier assigned each cell to its most related healthy B cell population. Across all samples, leukemic cells preferentially straddled the preproB cell to preB1 transition with expansion of these compartments at the expense of progenitor and immature B cell compartments. Each leukemia had a distinct occupancy of the developmental trajectory. Within the developmental compartments, blast cells retained functional features of their healthy counterparts including developmental enzymes TdT and RAG as well as transcription factors (Ikaros, PAX5). Activation of relevant signaling pathways including JAK-STAT and PI3K signaling was assessed in each compartment. Response to TKI's varied based on developmental compartment of the blasts. Finally, the association between state of developmental arrest and clinical features (age, MRD, relapse, genetics) will be presented. Conclusions: Deep proteomic profiling of BCP ALL establishes a singlecell classification linking phenotype with functional attributes of leukemic cells. This data demonstrates that heterogeneity within leukemia may be related to developmental state of the leukemic cell. Conflict of interest: None 59 Primary Induction Failure and Early Relapses in Children with AML Sorafenib and sunitinib, putative VEGF receptor inhibitors approved for use in metastatic renal cell carcinoma are relatively good FLT3 inhibitors. Sorafenib in particular has been shown to have both single agent activity 16 and has been combined with both standard induction chemotherapy 17 and hypomethylating agent 18 therapy. A phase III trial of chemotherapy plus or minus sorafenib in newly diagnosed AML patients (not restricted to mutant FLT3) suggests that sorafenib may lead to a superior event free survival with a trend toward improved overall survival. 19 The frequent and appropriate use of stem cell transplant to consolidate first remissions in AML patients with FLT3 mutations may cloud the ability to note an overall survival advantage with any new agent in AML, and FLT3 agents in particular, when used early in the course of the disease. Moreover, FLT3 inhibitors such as sorafenib have shown impressive responses in patients with FLT3 AML relapsing after a stem cell transplant. 20 Why this setting might yield a high number of responses with FLT3 inhibitors is unclear, but may have to do with changes in the immunologic millieu engendered by these compounds or perhaps by the different disease biology present in the post-transplant relapse setting. In any event, most subsequent trials will likely employ FLT3 inhibitors after transplant alone and/or in addition to early use In summary, the use of FLT3 inhibitors as single agents or in combination with chemotherapy in AML has not yet led to an approved agent. However, the story is far from over. Both sorafenib and midostaurin as adjuncts to standard chemotherapy in FLT3 mutant AML (and sorafenib in all AML), quizartinib as a single agent in AML, and crenolinib with its relatively unusual ability to inhibit both FLT3 ITD and TKD, may each have a role in AML. While FLT3 inhibitors are unlikely to have major long term disease mediating activity as single agents, drugs which can 'bridge' patients to transplant and/or potentiate the effects of chemotherapy may be useful. Thus, while FLT3 inhibitors have not been the straightforward hoped-for panacea in AML, it is much too early to 'close the book.' Conflict of interest: The author declares that he has no conflict of interest Adoptive Immunotherapy for Cancer or Viruses Exploiting the curative potential of adoptive T-cell therapy for cancer Principles of adoptive T cell cancer therapy Adoptive T cell therapy for cancer in the clinic Adoptive T cell transfer for cancer immunotherapy in the era of synthetic biology A new analysis of allogeneic interactions Enhanced function of redirected human T cells expressing linker for activation of T-cells that is resistant to ubiquitylation The inducible costimulator ICOS is critical for the development of human TH17 cells ICOS-based chimeric antigen receptors program bipolar TH17/TH1 cells Long-term in vivo survival of receptor-modified syngeneic T cells in patients with human immunodeficiency virus infection Epigenetic Signatures Identify New Clinically Relevant Subtypes in Acute Myeloid Leukemia Wouter de Laat and Ruud Delwel An oncogenic enhancerrearrangement causes concomitant deregulation of EVI1 and GATA2 in leukemia A remote GATA2 hematopoietic enhancer drives leukemogenesis in inv(3)(q21;q26) by activating EVI1 expression Mutational spectrum of myeloid malignancies with inv(3)/t(3;3) reveals a predominant involvement of RAS/RTK signaling pathways Co-inhibitory molecules of the B7-CD28 family in the control of T-cell immunity Coinhibitory molecules in hematologic malignancies: targets for therapeutic intervention Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads to negative regulation of lymphocyte activation Overexpression of PD-L1 significantly associates with tumor aggressiveness and postoperative recurrence in human hepatocellular carcinoma In a model of tumor dormancy, long-term persistent leukemic cells have increased B7-H1 and B7.1 expression and resist CTL-mediated lysis Program death-1 signaling and regulatory T cells collaborate to resist the function of adoptively transferred cytotoxic T lymphocytes in advanced acute myeloid leukemia Programmed death 1 signaling on chronic myeloid leukemia-specific T cells results in T-cell exhaustion and disease progression PD-1/PD-L1 interactions contribute to functional Tcell impairment in patients who relapse with cancer after allogeneic stem cell transplantation Activity, and Immune Correlates of Anti-PD Antibody in Cancer Safety and Activity of Anti-PD-L1 Antibody in Patients with Advanced Cancer Phase I study of single-agent anti-programmed death-1 (MDX-1106) in refractory solid tumors: safety, clinical activity, pharmacodynamics, and immunologic correlates Safety and tumor responses with lambrolizumab (anti-PD-1) in melanoma Phase I safety and pharmacokinetic study of CT-011, a humanized antibody interacting with PD-1, in patients with advanced hematologic malignancies Complementarity determining region 1 (CDR1)-and CDR3-analogous regions in CTLA-4 and CD28 determine the binding to B7-1 Immunologic self-tolerance maintained by CD25(+ )CD4(+) regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4 Exhaustion of tumor-specific CD8 T cells in metastases from melanoma patients Cytotoxic T lymphocyte antigen-4 blockade enhances antitumor immunity by stimulating melanoma-specific Tcell motility Anti-CTLA-4 antibody therapy: immune monitoring during clinical development of a novel immunotherapy Ipilimumab: an anti-CTLA-4 antibody for metastatic melanoma Phase III randomized clinical trial comparing tremelimumab with standard-of-care chemotherapy in patients with advanced melanoma The expression of T cell related costimulatory molecules in multiple myeloma Alterations of the expression of T-cell-related costimulatory CD28 and downregulatory CD152 (CTLA-4) molecules in patients with B-cell chronic lymphocytic leukaemia Regulation of CD80/B7-1 and CD86/B7-2 molecule expression in human primary acute myeloid leukemia and their role in allogenic immune recognition CTLA-4 genotype and relapse incidence in patients with acute myeloid leukemia in first complete remission after induction chemotherapy Opposing roles of CD28:B7 and CTLA-4:B7 pathways in regulating in vivo alloresponses in murine recipients of MHC disparate T cells CTLA4 blockade with ipilimumab to treat relapse of malignancy after allogeneic hematopoietic cell transplantation Dual blockade of PD-1 and CTLA-4 combined with tumor vaccine effectively restores T-cell rejection function in tumors Nivolumab plus ipilimumab in advanced melanoma CD8(+) T cells specific for tumor antigens can be rendered dysfunctional by the tumor microenvironment through upregulation of the inhibitory receptors BTLA and PD-1 siRNA silencing of PD-L1 and PD-L2 on dendritic cells augments expansion and function of minor histocompatibility antigen-specific CD8+ T cells Cancer Genome Atlas Research Network. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia Prognostic relevance of integrated genetic profiling in acute myeloid leukemia A novel hierarchical prognostic model of AML solely based on molecular mutations Can we say farewell to monitoring minimal residual disease in acute promyelocytic leukaemia? Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia Prospective minimal residual disease monitoring to predict relapse of acute promyelocytic leukemia and to direct pre-emptive arsenic trioxide therapy Management of acute promyelocytic leukemia: recommendations from an expert panel on behalf of the European LeukemiaNet Acute myeloid leukemia with mutated NPM1: diagnosis, prognosis and therapeutic perspectives Mutations and treatment outcome in cytogenetically normal acute myeloid leukemia Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet The European LeukemiaNet AML Working Party consensus statement on allogeneic HSCT for patients with AML in remission: an integrated-risk adapted approach Preleukemic mutations in human acute myeloid leukemia affect epigenetic regulators and persist in remission Identification of preleukaemic haematopoietic stem cells in acute leukaemia Molecular detection of minimal residual disease provides the most powerful independent prognostic factor irrespective of clonal architecture in Nucleophosmin (NPM1) mutant acute myeloid leukemia Quantitative assessment of minimal residual disease in acute myeloid leukemia carrying nucleophosmin (NPM1) gene mutations A model of human acute lymphoblastic leukemia in immune-deficient SCID mice Defining minimal residual disease in acute myeloid leukemia: which platforms are ready for "Prime Time Prognostic relevance of treatment response measured by flow cytometric residual disease detection in older patients with acute myeloid leukemia Normalization of bone marrow aspirates for hemodilution in flow cytometric analyses Should persons with acute myeloid leukemia have a transplant in first remission? High Prognostic Impact of Flow Cytometric Minimal Residual Disease Detection in Acute Myeloid Leukemia: Data From the HOVON/SAKK AML 42A Study Extracting a cellular hierarchy from high-dimensional cytometry data with SPADE Differential regulation of myeloid leukemias by the bone marrow microenvironment Tie2/Angiopoietin-1 signaling regulates hematopoietic stem cell quiescence in the bone marrow niche Osteoblastic cells regulate the haematopoietic stem cell niche The P190, P210, and P230 forms of the BCR/ABL oncogene induce a similar chronic myeloid leukemia-like syndrome in mice but have different lymphoid leukemogenic activity Transformation from committed progenitor to leukaemia stem cell initiated by MLL-AF9 Activated parathyroid hormone/parathyroid hormonerelated protein receptor in osteoblastic cells differentially affects cortical and trabecular bone Very long-term outcome of acute promyelocytic leukemia after treatment with all-trans retinoic acid and chemotherapy: the European APL Group experience All-trans retinoic acid in acute promyelocytic leukemia: long-term outcome and prognostic factor analysis from the North American Intergroup protocol A randomized study with or without intensified maintenance chemotherapy in patients with acute promyelocytic leukemia who have become negative for PML-RARalpha transcript after consolidation therapy: the Japan Adult Leukemia Study Group (JALSG) APL97 study AIDA 0493 protocol for newly diagnosed acute promyelocytic leukemia: very long-term results and role of maintenance Retinoic acid and arsenic trioxide for acute promyelocytic leukemia Arsenic trioxide during consolidation for patients with previously untreated low/intermediate risk acute References Acute promyelocytic leukemia: from highly fatal to highly curable Definition of relapse risk and role of non-anthracycline drugs for consolidation in patients with acute promyelocytic leukemia: a joint study of the PETHEMA and GIMEMA cooperative groups Continuing high early death rate in acute promyelocytic FLT3 mutations in acute myeloid leukemia: what is the best approach in 2013? Internal tandem duplication of the FLT3 gene is a novel modality of elongation mutation which causes constitutive activation of the product Will FLT3 inhibitors fulfill their promise in acute meyloid leukemia? A FLT3-targeted tyrosine kinase inhibitor is cytotoxic to leukemia cells in vitro and in vivo Inhibition of mutant FLT3 receptors in leukemia cells by the small molecule tyrosine kinase inhibitor PKC412 Phase IIB Trial of Oral Midostaurin (PKC412), the FMS-Like Tyrosine Kinase 3 Receptor (FLT3) and Multi-Targeted Kinase Inhibitor, in Patients With Acute Myeloid Leukemia and High-Risk Myelodysplastic Syndrome With Either Wild-Type or Mutated FLT3 Results from a randomized trial of salvage chemotherapy followed by lestaurtinib for patients with FLT3 mutant AML in first relapse Antitumor activity of sorafenib in FLT3-driven leukemic cells Mutant FLT3: a direct target of sorafenib in acute myelogenous leukemia A pharmacodynamic study of sorafenib in patients with relapsed and refractory acute leukemias Phase I study of sorafenib in patients with refractory or High activity of sorafenib in FLT3-ITD-positive acute myeloid leukemia synergizes with allo-immune effects to induce sustained responses Phase I Trial of Maintenance Sorafenib after Allogeneic Hematopoietic Stem Cell Transplantation for Fms-like Tyrosine Kinase 3 Internal Tandem Duplication Acute Myeloid Leukemia FLT3-mutant allelic burden and clinical status are predictive of response to FLT3 inhibitors in AML Phase I Study of Quizartinib Administered Daily to Patients With Relapsed or Refractory Acute Myeloid Leukemia Irrespective of FMS-Like Tyrosine Kinase 3-Internal Tandem Duplication Status Terminal myeloid differentiation in vivo is induced by FLT3 inhibition in FLT3/ITD AML Inhibition of c-Kit by tyrosine kinase inhibitors References 1. The Cancer Genome Atlas Research Network. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia Clonal evolution in relapsed acute myeloid leukaemia revealed by whole-genome sequencing Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic Acute lymphoblastic leukaemia Adolescents and young adults with acute lymphoblastic leukemia. Hematology / the Education Program of the Cytogenetics of paediatric and adolescent acute lymphoblastic leukaemia Genomic characterization of childhood acute lymphoblastic leukemia Deletion of IKZF1 and prognosis in acute lymphoblastic leukemia A subtype of childhood acute lymphoblastic leukaemia with poor treatment outcome: a genome-wide classification study Independent prognostic value of BCR-ABL1-like signature and IKZF1 deletion, but not high CRLF2 expression, in children with B-cell precursor ALL Targetable kinase-activating lesions in Ph-like acute lymphoblastic leukemia Deregulated expression of cytokine receptor gene, CRLF2, is involved in lymphoid transformation in B-cell precursor acute lymphoblastic leukemia Rearrangement of CRLF2 in B-progenitor-and Down syndrome-associated acute lymphoblastic leukemia Genetic alterations activating kinase and cytokine receptor signaling in high-risk acute lymphoblastic leukemia Tyrosine kinase inhibitor therapy induces remission in a patient with refractory EBF1 PDGFRB-positive acute lymphoblastic leukemia Successful tyrosine kinase inhibitor therapy in a refractory B-cell precursor acute lymphoblastic leukemia with EBF1-PDGFRB fusion Targeting JAK1/2 and mTOR in murine xenograft models of Ph-like acute lymphoblastic leukemia Aberrant STAT5 and PI3K/mTOR pathway signaling occurs in human CRLF2-rearranged B-precursor acute lymphoblastic leukemia ERG Deletions Define a Novel Subtype of B-Progenitor Acute Lymphoblastic Leukemia An intragenic ERG deletion is a marker of an oncogenic subtype of B-cell precursor acute lymphoblastic leukemia with a favorable outcome despite frequent IKZF1 deletions The genomic landscape of hypodiploid acute lymphoblastic leukemia CREBBP mutations in relapsed acute lymphoblastic leukaemia Conditional modulation of glucocorticoid receptor activities by CREB-binding protein (CBP) and Relapse-specific mutations in NT5C2 in childhood acute lymphoblastic leukemia Activating mutations in the NT5C2 nucleotidase gene drive chemotherapy resistance in relapsed ALL Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel Review of the relevance of aberrant antigen expression by flow cytometry in myeloid neoplasms High prognostic impact of flow cytometric minimal residual disease detection in acute myeloid leukemia: data from the HOVON/SAKK AML 42A study High stem cell frequency in acute myeloid leukemia at diagnosis predicts high minimal residual disease and poor survival High levels of CD34+CD38low/-CD123+ blasts are predictive of an adverse outcome in acute myeloid leukemia: a GroupeOuest-Est des LeucemiesAigues et Maladies du Sang (GOELAMS) study Leukemic stem cell frequency: a strong biomarker for clinical outcome in acute myeloid leukemia Heterogeneity of leukemia stem cell candidates at diagnosis of acute myeloid leukemia and their clinical relevance The novel AML stem cell associated antigen CLL-1 aids in discrimination between normal and leukemic stem cells Toward optimization of postremission therapy for residual disease-positive patients with acute myeloid leukemia Defining consensus leukemia-associated immunophenotypes for detection of minimal residual disease in acute myeloid leukemia in a multicenter setting Tumor Heterogeneity Makes AML a "Moving Target" for Detection of Residual Disease Clonal evolution in relapsed acute myeloid leukaemia revealed by whole-genome sequencing Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia Preleukemic mutations in human acute myeloid leukemia affect epigenetic regulators and persist in remission Identification of pre-leukaemic haematopoietic stem cells in acute leukaemia Recurring mutations found by sequencing an acute myeloid leukemia genome Cancer-associated IDH1 mutations produce 2-hydroxyglutarate The common feature of leukemiaassociated IDH1 and IDH2 mutations is a neomorphic enzyme activity converting alpha-ketoglutarate to 2-hydroxyglutarate IDH1 and IDH2 gene mutations identify novel molecular subsets within de novo cytogenetically normal acute myeloid leukemia: a Cancer and Leukemia Group B study Cancer-associated metabolite 2-hydroxyglutarate accumulates in acute myelogenous leukemia with isocitrate dehydrogenase 1 and 2 mutations Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation IDH mutation impairs histone demethylation and results in a block to cell differentiation Acquired mutations in the genes encoding IDH1 and IDH2 both are recurrent aberrations in acute myeloid leukemia: prevalence and prognostic value Molecular alterations of isocitrate dehydrogenase 1 and 2 (IDH1 and IDH2) metabolic genes and additional genetic mutations in newly diagnosed acute myeloid leukemia patients The prognostic impact and stability of Isocitrate dehydrogenase 2 mutation in adult patients with acute myeloid leukemia IDH1 and IDH2 duplication Prognostic relevance of integrated genetic profiling in acute myeloid leukemia Targeted inhibition of mutant IDH2 in leukemia cells induces cellular differentiation AG-221 Offers a Survival Advantage In a Primary Human IDH2 Mutant AML Xenograft Model AG-221, an Oral, Selective, First-in-Class, Potent Inhibitor of the IDH2 Mutant Metabolic Enzyme Acute myeloid leukaemia Monoclonal antibody therapy directed against human acute myeloid leukemia stem cells Chimeric antigen receptors for the adoptive T cell therapy of hematologic malignancies Expression of immunoglobulin-T-cell receptor chimeric molecules as functional receptors with antibody-type specificity Expansion of cytolytic CD8(+) natural killer T cells with limited capacity for graftversus-host disease induction due to interferon gamma production In vivo trafficking and survival of cytokine-induced killer cells resulting in minimal GVHD with retention of antitumor activity Repeated infusions of donorderived cytokine-induced killer cells in patients relapsing after allogeneic stem cell transplantation: a phase I study Cytokine induced killer cells as promising immunotherapy for solid tumors High-throughput sequencing detects minimal residual disease in acute T lymphoblastic leukemia Nextgeneration sequencing for minimal residual disease monitoring in acute myeloid leukemia patients with FLT3-ITD or NPM1 mutations Department of Hematology and Oncology Partially matched, nonmyeloablative allogeneic transplantation: clinical outcomes and immune reconstitution Unmanipulated HLA 2-3 antigen-mismatched (haploidentical) stem cell transplantation using nonmyeloablative conditioning Full haplotype-mismatched hematopoietic stem-cell transplantation: a phase II study in patients with acute leukemia at high risk of relapse HLA-haploidentical bone marrow transplantation for hematologic malignancies using nonmyeloablative conditioning and high-dose, posttransplantation cyclophosphamide Direct demonstration of elevated aldehyde dehydrogenase in human hematopoietic progenitor cells Aldehyde dehydrogenase expression drives human regulatory T cell resistance to posttransplantation cyclophosphamide Nonmyeloablative HLA-haploidentical bone marrow transplantation with high-dose posttransplantation cyclophosphamide: effect of HLA disparity on outcome Haploidentical transplantation using T cell replete peripheral blood stem cells and myeloablative conditioning in patients with high-risk hematologic malignancies who lack conventional donors is well tolerated and produces excellent relapse-free survival: results of a prospective phase II trial Treatment of acute leukemia with unmanipulated HLA-mismatched/haploidentical blood and bone marrow transplantation Feasibility of clofarabine cytoreduction followed by haploidentical hematopoietic stem cell transplantation in patients with relapsed or refractory advanced acute leukemia Sequential regimen of chemotherapy, reduced-intensity conditioning for allogeneic stem-cell transplantation, and prophylactic donor lymphocyte transfusion in high-risk acute myeloid leukemia and myelodysplastic syndrome T-cell-replete HLA-haploidentical hematopoietic transplantation for hematologic malignancies using posttransplantation cyclophosphamide results in outcomes equivalent to those of contemporaneous HLA-matched related and unrelated donor transplantation Acute lymphoblastic leukemia First-line treatment of Philadelphia chromosome-positive acute lymphoblastic leukaemia in adults New developments in the treatment of chronic myeloid leukemia and Philadelphiapositive acute lymphoblastic leukemia Ponatinib in refractory Philadelphia chromosome-positive leukemias Aggressive Langerhans cell histiocytosis following T-ALL: clonally related neoplasms with persistent expression of constitutively active NOTCH1 Molecular ontogeny of donor-derived follicular lymphomas occurring after hematopoietic cell transplantation Therapy-related myeloid leukemia The impact of therapy-related acute myeloid leukemia (AML) on outcome in 2853 adult patients with newly diagnosed AML Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia Clonal evolution in relapsed acute myeloid leukaemia revealed by wholegenome sequencing Clonal evolution in cancer The origin and evolution of mutations in acute myeloid leukemia Nonrandom X-inactivation patterns in normal females: lyonization ratios vary with age Recurrent somatic TET2 mutations in normal elderly individuals with clonal hematopoiesis Age-related somatic structural changes in the nuclear genome of human blood cells Acquired skewing of X-chromosome inactivation patterns in myeloid cells of the elderly suggests stochastic clonal loss with age Somatic mutations found in the healthy blood compartment of a 115-yrold woman demonstrate oligoclonal hematopoiesis Detectable clonal mosaicism and its relationship to aging and cancer Detectable clonal mosaicism from birth to old age and its relationship to cancer Diagnosis and antimicrobial therapy of lung infiltrates in febrile neutropenic patients (allogeneic SCT excluded): updated guidelines of the Infectious Diseases Working Party (AGIHO) of the German Society of Hematology and Medical Oncology (DGHO)dagger The prevalence of nosocomial infection in intensive care units in Europe Epidemiology of community-acquired respiratory tract infections in adults: incidence, etiology, and impact Lung disease data 1994 Semiinvasive pulmonary aspergillosis in chronic obstructive pulmonary disease:radiologic and pathologic findings in nine patients The changing face of Pneumocystis carinii pneumonia in AIDS patients HRCT in acute diffuse lung disease in the immunocompromised patient Acute pulmonary complications in immunocompromised non-AIDS patients: Comparison of diagnostic accuracy of CT and chest radiography Acute lung disease in the immunocompromised host: CT and pathologic findings Acute parenchymal lung disease in immunocompetent patients: diagnostic accuracy of highresolution CT HRCT in acute diffuse lung disease in the immunocompromised patient Spectrum of pulmonary aspergillosis : histologic, clinical, and radiologic findings Tree-in-bud pattern: frequency and significance on thin section CT Cytomegalovirus pneumonitis: spectrum of parenchymal CT findings with pathologic correlation in 21 AIDS patients Necrotizing aspergillosis of large airways : CT findings in 8 patients Acute pulmonary failure during remission induction chemotherapy in adults with acute myeloid leukemia or high-risk myelodysplastic syndrome Pulmonary infiltrates in neutropenic patients with acute leukemia during chemotherapy Pulmonary infiltrations in febrile neutropenic patients. Risk factors and outcome under empirical antimicrobial therapy in a randomized multicenter trial Diagnostic strategy for hematology and oncology patients with acute respiratory failure: randomized controlled trial A prospective, randomized multicenter trial of the empirical addition of antifungal therapy for febrile neutropenic cancer patients Fourth European Conference on Infections in Leukaemia (ECIL-4): Guidelines for diagnosis and treatment of human Respiratory Syncytial Virus, Parainfluenza Virus, Metapneumovirus, Rhinovirus, and Coronavirus Diagnosis and antimicrobial therapy of lung infiltrates in febrile neutropenic patients (allogeneic SCT excluded): updated guidelines of the Infectious Diseases Working Party (AGIHO) of the German Society of Hematology and Medical Oncology (DGHO) Infectious Diseases Society of America. Guide to development of practice guidelines Falagas ME. β-Lactam plus aminoglycoside or fluoroquinolone combination versus β-lactam monotherapy for Pseudomonas aeruginosa infections: a meta-analysis Tigecycline as a therapeutic option in Stenotrophomonas maltophilia infections Adjunctive steroid in HIV-negative patients with severe Pneumocystis pneumonia Outcomes of critically ill patients with hematologic malignancies: prospective multicenter data from France and Belgium-a groupe de recherche respiratoire en reanimation oncohematologique study Outcome and prognostic factors of hematopoietic stem cell transplantation recipients admitted to a medical ICU Critical care of the hematopoietic stem cell transplant recipient Incidence and prognostic value of respiratory events in acute leukemia Acute respiratory distress syndrome in patients with malignancies Respiratory events at the earliest phase of acute myeloid leukemia Diagnostic strategy for hematology and oncology patients with acute respiratory failure: randomized controlled trial Clinical assessment for identifying causes of acute respiratory failure in cancer patients Predictive factors of intensive care unit admission in patients with haematological malignancies and pneumonia Outcome of critically ill allogeneic hematopoietic stem-cell transplantation recipients: a reappraisal of indications for organ failure supports Intensive care of the cancer patient: recent achievements and remaining challenges The ICU trial: a new admission policy for cancer patients requiring mechanical ventilation How to treat severe infections in critically ill neutropenic patients? Noninvasive ventilation in immunosuppressed patients with pulmonary infiltrates, fever, and acute respiratory failure Respiratory failure in patients undergoing allogeneic hematopoietic SCT-a randomized trial on early non-invasive ventilation based on standard care hematology wards Improved survival in cancer patients requiring mechanical ventilatory support: impact of noninvasive mechanical ventilatory support Outcome in noninvasively and invasively ventilated hematologic patients with acute respiratory failure Ventilatory support in critically ill hematology patients with respiratory failure Noninvasive versus invasive ventilation for acute respiratory failure in patients with hematologic malignancies: a 5-year multicenter observational survey Mechanical ventilation in cancer patients: clinical characteristics and outcomes Early non-invasive ventilation for acute respiratory failure in immunocompromised patients (IVNIctus): study protocol for a multicenter randomized controlled trial Delayed intensive care unit admission is associated with increased mortality in patients with cancer with acute respiratory failure Early intervention on the outcomes in critically ill cancer patients admitted to intensive care units Intensive care unit management of patients with newly diagnosed acute myeloid leukemia with no organ failure Early CPAP prevents evolution of acute lung injury in patients with hematologic malignancy Noninvasive mechanical ventilation in patients having declined tracheal intubation Extracorporeal membrane oxygenation in adult patients with hematologic malignancies and severe acute respiratory failure Extracorporeal membrane oxygenation instead of invasive mechanical ventilation in patients with acute respiratory distress syndrome UK References 1. Appelbaum FR. Optimising the conditioning regimen for acute myeloid leukaemia. Best practice & research Clinical haematology Long-term outcomes among older patients following nonmyeloablative conditioning and allogeneic hematopoietic cell transplantation for advanced hematologic malignancies Hematopoietic cell transplantation (HCT)-specific comorbidity index: a new tool for risk assessment before allogeneic HCT Risk assessment before allogeneic hematopoietic cell transplantation for older adults with acute myeloid leukemia Impact of pre-transplant co-morbidities on outcome after alemtuzumabbased reduced intensity conditioning allo-SCT in elderly patients: A British Society of Blood and Marrow Transplantation study. Bone Marrow Transplant A risk score for mortality after allogeneic hematopoietic cell transplantation Role of regulatory T cell populations in controlling graft vs host disease. Best practice & research Clinical haematology Sequential regimen of chemotherapy, reduced-intensity conditioning for allogeneic stem-cell transplantation, and prophylactic donor lymphocyte transfusion in high-risk acute myeloid leukemia and myelodysplastic syndrome Reduced-toxicity conditioning with fludarabine, BCNU, and melphalan in allogeneic hematopoietic cell transplantation: particular activity against advanced hematologic malignancies Factors predicting long-term survival after T-cell depleted reduced intensity allogeneic stem cell transplantation for acute myeloid leukemia Azacitidine augments expansion of regulatory T cells after allogeneic stem cell transplantation in patients with acute myeloid leukemia (AML) Serial killing of tumor cells by cytotoxic T cells redirected with a CD19-/CD3-bispecific singlechain antibody construct Induction of regular cytolytic T cell synapses by bispecific single-chain antibody constructs on MHC class I-negative tumor cells Immunomodulatory therapy of cancer with T cell-engaging BiTE antibody blinatumomab Flow cytometric study of potential target antigens (CD19, CD20, CD22, CD33) for antibody-based immunotherapy in acute lymphoblastic leukemia: analysis of 552 cases Tumor regression in cancer patients by very low doses of a T cell-engaging antibody Immunopharmacologic response of patients with B-lineage acute lymphoblastic leukemia to continuous infusion of T cell-engaging CD19/CD3-bispecific BiTE antibody blinatumomab Phase II trial of the anti-CD19 bispecific T cell-engager blinatumomab shows hematologic and molecular remissions in patients with relapsed or refractory B-precursor acute lymphoblastic leukemia Treatment of relapsed/refractory diffuse large B-cell lymphoma with the bispecific T-cell engager (BiTE®) antibody construct blinatumomab: primary analysis results from an openlabel, phase 2 study Phase II trial of the anti-CD19 bispecific T cell-engager blinatumomab shows hematologic and molecular remissions in patients with relapsed or refractory B-precursor acute lymphoblastic leukemia CD33 target validation and sustained depletion of AML blasts in long-term cultures by the bispecific T-cell-engaging antibody AMG 330 Retinoic Acid and Arsenic Trioxide for Acute Promyelocytic Leukemia Definition of relapse risk and role of nonanthracycline drugs for consolidation in patients with acute promyelocytic leukemia: a joint study of the PETHEMA and GIMEMA cooperative groups Modern Approaches to Treating Acute Promyelocytic Leukemia All-trans-retinoic acid in acute promyelocytic leukemia Impact of arsenic trioxide in the treatment of acute promyelocytic leukemia Acute promyelocytic leukemia: recent advances in therapy and molecular basis of response to arsenic therapies Mechanisms of action of arsenic trioxide Arsenic trioxide induces dose-and time-dependent apoptosis of endothelium and may exert an antileukemic effect via inhibition of angiogenesis Arsenic trioxide as an inducer of apoptosis and loss of PML/RAR alpha protein in acute promyelocytic leukemia cells Use of arsenic trioxide (As2O3) in the treatment of acute promyelocytic leukemia (APE) .2. Clinical efficacy and pharmacokinetics in relapsed patients Complete remission after treatment of acute promyelocytic leukemia with arsenic trioxide Arsenic trioxide in comparison with chemotherapy and bone marrow transplantation for the treatment of relapsed acute promyelocytic leukaemia Use of arsenic trioxide (AS(2)O(3)) in the treatment of patients with acute promyelocytic leukemia -The M. D. Anderson experience Studies on treatment of acute promyelocytic leukemia with arsenic trioxide: Remission induction, follow-up, and molecular monitoring in 11 newly diagnosed and 47 relapsed acute promyelocytic leukemia patients Combined treatment with arsenic trioxide and all-trans-retinoic acid in patients with relapsed acute promyelocytic leukemia Arsenic trioxide therapy in relapsed or refractory Japanese patients with acute promyelocytic leukemia: Updated outcomes of the phase II study and postremission therapies United States multicenter study of arsenic trioxide in relapsed acute promyelocytic leukemia Arsenic trioxide treated 72 cases of acute promyelocytic leukemia Use of all-trans retinoic acid plus arsenic trioxide as an alternative to chemotherapy in untreated acute promyelocytic leukemia Treatment of children with newly diagnosed acute promyelocytic leukemia with arsenic trioxide: a single center experience Treatment of acute promyelocytic leukemia with arsenic trioxide without ATRA and/or chemotherapy Single-agent arsenic trioxide in the treatment of newly diagnosed acute promyelocytic leukemia: durable remissions with minimal toxicity All-trans retinoic acid/AS(2)O(3) combination yields a high quality remission and survival in newly diagnosed acute promyelocytic leukemia Arsenic trioxide improves eventfree and overall survival for adults with acute promyelocytic leukemia: North American Leukemia Intergroup Study C9710 Stability of mutations was tested in 7 patients (13 BM samples), initially carrying R882 mutation, at the time of complete remission (CR) and/or disease relapse by Sanger sequencing. Results: Four different mutations were identified in 11/25 patients: 2/4 mutations are novel and not described yet (L347P and Y347S), other two mutations are common (R882C and R882H). 5/11 patients carried mutation R882C, 4/11 mutation R882H. In 4/7 patients, initially carrying mutation in R882 codon, mutations were detected also at the time of CR. Conclusions: Our results confirm that mutations in DNMT3A gene are a frequent event in NK-AML patients. In some patients, DNMT3A mutations persist during CR. These results may indicate the presence of preleukemic AML stem cells with mutation in DNMT3A gene, which are resistant to chemotherapy Heckman 1 1 Institute for Molecular Medicine Finland The samples with complex cytogenetics were largely resistant. Sensitivity correlated to expression level of BCL2 and resistance to expression of BCLXL and MCL1. Method: In this study, we assessed the ex vivo sensitivity of fresh BM MNCs from 11 diagnosed and 34 relapsed/refractory AML patients to venetoclax (22 samples) and navitoclax (45 samples) We observed responses across all mutational groups (FLT3-ITD, NPM1, TP53, NRAS, IDH) and also in samples with complex or monosomal karyotype. Although BCL2 transcript levels tended to correlate to navitoclax sensitivity, a better predictor for sensitivity was the gene expression of specific proapoptotic BCL2 family members. Conclusion: In conclusion Acute leukemia (AL) is the most widespread malignancy in childhood and frequent in adults. Detection of molecular genetic abnormaities in AL is of great significance as they define biological properties of malignant cells, associated with treatment response and can be markers for MRD monitoring and targets for therapy. Number of clinically relevant mutations is constantly increasing and even with introduction of new and effective high-throughput approaches for detection of molecular genetic abnormalities, development of simple and quick method for the simultaneous detection of multiple clinically significant molecular-genetic alterations (chimeric genes) can be useful. Method: During the project execution multiplex nested PCR was developed for the detection of expression of main described splice-variants of clinically relevant chimeric genes Proteasome Inhibitors Induce FLT3-ITD Degradation through Introduction: Internal tandem duplication of the Fms-like tyrosine kinase-3 receptor (FLT3-ITD) is found in 30% of acute myeloid leukemia (AML) and is associated with a poor outcome To document the prevalence of PML/RARα (bcr1 and bcr3 Methods and results: Blood/bone marrow samples were collected from CD 13, CD 33, CD 45, CD117, CD 34 and CD 64. Summary: Acute myeloid leukaemia (AML) is a type of cancer that affects the blood and bone marrow and one of the leading haematological malignancies in Assam CD 45, CD117, CD 34 and CD 64 are most common in these patients The Mutatome of CBFB/MYH11 Rearranged Acute Myeloid Leukemia (AML) Another patient gained a CBL mutation at relapse. Conclusion: The mutatome of CBFB/MYH11 rearranged AML is genetically complex with frequent secondary mutations in kinase genes such as NRAS, KIT, FLT3. These kinase mutations seem to be unstable during disease progression and the actual driver of relapse remains elusive in many cases Methods: mFISH analysis of the MOTN-1 cell line was performed as well as paired-end next generation sequencing (NGS; Illumina HiSeq2000) of MOTN-1 and one T-LGL sample. In addition, a distinct 6q region was characterized in three T-LGL samples using high resolution comparative genomic hybridization (FT-CGH) and LM-PCR. Gene expression was studied by RNA sequencing (RNAseq; SOLID5500). Results: Rearrangements were detected within 1p and 2q in MOTN-1 affecting expression of FGR, ZEB2 and CASP8, and within 6q in MOTN-1 and one T-LGL sample affecting MAP3K5 and IFNGR1. Nineteen genes, among them FON3, RIN3, AKT1, PPP2R5C, were overexpressed as a result of an amplification in 14q in one T-LGL patient Acute lymphoblastic leukemia (ALL) is a frequent disease in children and adults which remains difficult to treat in many patients XIAP, the X-linked inhibitor of apoptosis, is a potent kinase inhibitor, highly expressed in hematological malignancies and contributing to substrates, Coelenterazine and Luciferin. Thus, two different cell populations were comparatively monitored in real time in a single mouse Roeder 2 , I. Glauche 2 , B. Fehse 1 1 Department of Stem Cell Transplantation Affymetrix gene expression profiles of 207 adult B-cell precursor ALL patients were classified as Ph-like ALL according to published algorithms. Patients and methods: Clinical and outcome analysis were restricted to patients treated within GMALL trials 06/99 and 07/03 (n=107). The molecular characterization included copy number variation analysis and targeted amplicon sequencing. We identified 26 patients (13%) with Phlike gene expression profile. Among 105 B-cell precursor ALL patients that were negative for Ph-and MLL-translocations (B-other) the incidence of the Ph-like subtype was 25% Myelodysplasia in Adults with Near-Tetraploidy and Lymphoid (ALL) or Ambiguous Lineage (ALAL) Acute Leukemias P. Lemez ) pts -in CCR. Results: Level of MN1 expression in pts with ALL was correlated with resistance to ATRA treatment. Patients with low MN1 expression (5/9, 55,6%) didn′t have relapse. 3/4 (75%) relapsed pts had high MN1 expression. Patients with low MN1 expression had prolonged relapsefree survival (P=0.06) comparing with pts with high level of MN1. Conclusion: Our results suggest that the increase of relapses in high risk MLL-AF4-positive ALL patients treated according to ALL-MB-2008 protocol was associated with high MN1 expression. MN1 expression independently predicts outcome in patients with ALL. Conflict of interest: None 50 Incidence of IKZF1 Deletions In Russian Cohort of Pediatric B-Precursor ALL The Netherlands) to identify whole or intragenic IKZF1 deletions. Results: In total, 22 IKZF1 deletions were identified in 171 cases According to ALL-MB2008 treatment protocol 14 pts were stratified to ImRG, 5 to SRG and 3 to HRG. Conclusions: Our data on genetic IKZF1 alteration is compatible with data published by others groups. However, the analyzed patient group is not sufficiently large and has a prominent imbalance in genetic variants (66 cases with t(12;21)). Moreover, the follow-up is too short to analyze survival and risk of relapse Pfizer Worldwide Research & Development Asparaginases are a cornerstone in the treatment of ALL, but their utility is limited by toxicities including hypersensitivity. ERYASP improves pharmacokinetics and tolerability, especially hypersensitivity reactions, while maintaining circulating asparaginase (ASPA) activity due to the protective barrier of the erythrocyte membrane. Patients and methods: This phase III study enrolled patients with primary relapsed or refractory Ph-ALL. The endpoints for efficacy were comparison of the duration of ASPA activity >100IU/l and the incidence of ASPA hypersensitivity during induction. Patients, aged 1-55 years IU/kg) or L-ASP(10.000IU/m 2 ) in combination with COOPRALL. The patients with prior ASPA hypersensitivity received ERYASP. Results: ERYASP significantly reduced the incidence of ASPA hyper In the treatment of acute lymphoblastic leukemia (ALL) response to initial remission induction (RI) therapy is a strong prognostic factor used for patient risk stratification. As 20% of ALL patients relapse, new anti-leukemic agents are essential, which most likely will be applied to already existing multi-agent chemotherapy regimens and will undergo in vivo preclinical testing. Study aim: In this study we aimed to (i) establish an in vivo treatment strategy, which resembles RI therapy in our NOD/SCID/huALL mouse model and to (ii) analyse the in vivo response to RI treatment of distinct patient-derived xenograft (pdx) ALL samples in the context of pretreatment leukemia features and patient outcome. Patients and methods: We have treated 7 pdx samples with RI therapy combining vincristine TTL short ) being associated with an early relapse in contrast to superior patient outcome of TTL long leukemias. Interestingly, rapid TTR after VDA therapy is significantly associated with TTL short phenotypes and TTL long leukemias display a longer TTR indicating increased in vivo sensitivity for VDA treatment. Conclusion: In summary, we have established an effective in vivo RI protocol allowing evaluating novel anti-leukemic agents in a preclinical setting for ALL. In addition, short TTR is associated with inferior relapse free survival and NOD/SCID engraftment suggesting that pre-treatment high risk features reflect leukemia-intrinsic characteristics such as posttreatment leukemia reoccurrence and relapse Method: Flowcytometric analyses of CD70 surface expression in 28 patient-derived xenograft samples (TTL short n=8, TTL long n=20) revealed in BCP-ALL, the efficacy of monoclonal antibody based CD70-directed immunotherapy was evaluated in NOD/SCID mice. Result: A marked reduction of leukemia load in peripheral blood, bone marrow and spleens of the animals was detected in anti-CD70 treated CD70 hi primograft ALL cells. This effect could be abrogated both by NK-cell depletion in the NOD/SCID recipients and by using NK-cell deprived NSG mice as recipients, indicating that decreased in vivo leukemia growth upon anti-CD70 treatment is mediated by NK-cell induced cytotoxicity of CD70 positive ALL cells. Consistently, CD70 antibodydependent cell Proteasome Activity is Dispensible for the Degradation of PML-RARA: Efficacy of Bortezomib along with Arsenic Trioxide in the Treatment of ATO Sensitive and Resistant Acute Promyelocytic Leukemia Degradation of PML-RARA upon treatment with ATO is predominantly mediated by the proteasome complex In group PIF (n=26) pOS was 22% in HSCT and 0% in CHEMO (p=0,0003) Conclusions: Cure of patients with refractory AML and early Leukemia Background: Children with trisomy 21 (Down syndrome, DS) are predisposed to develop acute megakaryoblastic leukemia (DS-AMKL) as well as the antecedent transient leukemia (DS-TL). Mutations in the transcription factor GATA1 (GATA1s) are present in nearly all children with DS-AMKL and DS-TL. GATA1s is both essential and sufficient to cause DS-TL in synergy with trisomy 21. Study aim and methods: To elucidate cell viability, apoptosis or differentiation in DS-AMKL cell lines (CMK and CMY) All 12 candidate genes acted synergistically to enhance the self-renewal efficiency of murine fetal liver cells in vitro. Results: Using an in vivo biotinylation approach, we showed that bio GATA1 is associated with protein-complexes of 10 different hsa21-oncogenes, involved in splicing, deubiquitination and transcriptional regulation. Direct interactions as well as transcriptional regulation of several of these factors are perturbed in N-terminal truncated GATA1s. Conclusion: Thus, we deciphered a complex interactive oncogenic network on hsa21 centered on GATA1, positively regulating megakaryopoiesis. Deregulation of this network results in synergistic effects on hematopoietic differentiation ETP) Phenotype Impact on Outcome of Pediatric Acute Lymphoblastic Leukemia (ALL) Portuguese Institute of Oncology; 2 Prof. Dr. Fernando Fonseca Hospital study including all patients (pts) younger than 16 years diagnosed with T-ALL between Mar2007 and Jun2014 to determine differences in presenting features, response to remission-induction therapy, relapse rate (RR), overall survival (OS) and event-free survival (EFS), between ETP ALL, defined by immunophenotyping (CD5weak/-CD1a-CD8-and ≥1 myeloid or stemcell related antigen), and non-ETP pediatric T-ALL. Method and results: We included 35 pts, 9 (25.7%) had an ETP phenotype and 22 were male Conclusions: Our series showed a higher frequency of ETP-ALL than reported, and a similar outcome for ETP and non-ETP ALL. This could be, at least in part, explained by the inclusion of CD5-cases, not mentioned in the first reports. However, ETP-ALL showed a higher induction-failure rate: there may be some benefit in introducing new drugs in their induction regimen. To clarify these observations, we will extend the follow up time and the sample. The outcome for relapsed T-ALL, ETP and non-ETP, remains extremely poor: there is an urgent need for new therapeutic approaches Boos 1 1 Department of Pediatric Hematology and Oncology, University Childrens' Hospital of Münster, Germany; 2 Department of Medical Biotechnologies and Translational Medicine Methods: Between 02/2011 and 10/2012 1,135 CSF samples from 523 patients were analyzed for ASN content by HPLC with a limit of quantification of 0.2 μmol/L. Results: Administration of 2,500 U/m 2 PEG-ASE significantly reduced ASN levels in CSF. In protocol I after administration of 2 doses at 2 weeks interval ASN reduction in CSF lasted for 40 days and treated with a BFM AML-98 protocol. Results: There were 83 newly diagnosed patients in this period. Median age was 9.8years (range: 1-15) and there were 52(62.7%) males. Cytogenetic data was available in 71 Leukemia is the most common childhood malignancy, translocation of the mixed-lineage leukemia (MLL) gene on chromosome (chr.) 11 1,2 is frequent. Method and discussion: We present an infant New mFISH analysis showed a complex karyotype including t(5;9;11) (q21;p21;q21). Multiplex-PCR and sequencing (gDNA) confirmed a MLL-AF9 fusion gene. Retrospective measurement proved that the MLL-AF9 fusion gene was the same at initial diagnosis and in relapse Conflict of interest: None 65 Therapy Adaption in the Genetically High Risk Group in Childhood Acute Myeloid Leukemia: A Report University Children's Hospital Complete Remission (CR) was achieved by 31 of these patients. Out of this group the patients which were transplanted in 1.CR (n=11, pEFS 73%, SE 13%) had a significantly better prognosis than those without HSCT in 1.CR and median survival of more then 4,3 months (the median time to allogeneic transplantation, HSCT) (n=19, pEFS 20%, SE 10%, p(Mantel-Byar)=0.04). Conclusion: Our results show that allogeneic HSCT in 1.CR seems to be the therapeutic option in patients with genetically high risk factors. It is desirable to improve the CR rate in this group for example by introduction of new agents upfront in therapy as the achievement of complete remission seems to give these HR patients a realistic chance of cure. Conflict of Interest: None 66 Expression of MECOM Transcripts in Childhood Acute Leukemia In AML expression of сEVI1 was higher in MLL-AF9-positive patients (p<0,05), including those with M7. Higher levels of MDS1/ EVI1 and cEVI1 were found in AML with complex karyotype (p<0,01). In B-ALL with MLL rearrangements expression of MDS1/EVI1 and сEVI1 was lower comparing with group without rearrangements NG2-positivity represented the highest overall correct prediction (OCP) rate for presence of MLLrearrangements (90.6%). Diagnostic accuracy of CD20-negativity and CD45-positivity was lower (81.2% and 81.9% respectively) while OCP for other markers was not sufficient enough Mutations included 7 deletions, 3 insertions and 1 indel distributed in the gene functional domain in the following way: 2 mutations in TAD1, 4 -in TAD2 and 5 -in bZIP region. wt DNA, in one case there was nonspecific amplification of wt DNA 10 cycles later. Additionally, for patients with GCG deletion in TAD2 domain we have found useful detection of mutation using HRM. Conclusion: Detection of CEBPA mutations can be useful for MRD detection of NK AML patients, being at the same time challenging task for RQ-PCR Fechina 1 Regional Children's Hospital To estimate prognostic significance of minimal residual disease (MRD) in bone marrow (BM) and peripheral blood (PB) by qualitative detection of different MLL fusion gene transcripts (FGt) in infants with ALL treated by MLL-Baby protocol Median of follow-up period in the observed group was 5.2 years. TPs for MRD assessment were as follows: day 15 of remission induction (time point (TP) 1), at the end of remission induction (TP2) This was most prominent within the subpopulation of effector memory T-cells. Furthermore, in response to AMG 330-mediated T-cell activation, we observed an upregulation of PD-L1 on primary AML cells (16/19). This phenomenon was cytokine-mediated as the sole addition of IFN□ and TNF□ also induced PD-L1 expression. Interestingly, we also observed a PD-L1 upregulation on T-cells upon activation with AMG 330, but to a much lower extent compared to primary AML cells (n=17; mean MFI ratio: Tcells: 4.7; AML cells: 12.1). Blockade of the PD-1/PD-L1 interaction increased T-cell proliferation resulting in enhanced cytotoxicity against primary AML cells (with/without PD-1 blocking: 75 vs 44%) and a significant increase in IFN□ production. This effect was most prominent in co-cultures with a low effector : target cell ratio are employees of AMGEN Research Autologous Hematopoietic Stem Cell Transplantation for Adults with Acute Myeloid Leukemia: A Single-Centre Experience Poland Introduction: Hematopoietic stem cell transplantation (HSCT) remains the most efficacious therapy in patients (pts) with acute leukemias. For older pts with high risk or relapsed disease lacking a related HLAcompatible donors autologous transplantation (AutoHSCT) is a valid alternative therapeutic option. Material: From 1997 till 2014 in Department of Haematooncology and Bone Marrow Transplantation Medical University of Lublin there were performed 29 AutoHSCT in pts with acute myeloid leukemia (AML: 15 M and 14 F; median age: 52.2 years). There were diagnosed following FAB types of AML: M0 -3, M1 -4, M2 -6, M4 -10 and M5 -6. The pts with AML were classified into 3 cytogenetic prognostic groups: high risk -9, intermediate -16 and low risk -4. Twenty five pts with AML were in first complete remission (CR) and 4 in second CR The 3-year disease-free survival (DFS) and overall survival (OS) was 60% and 68%, respectively. Conclusion: Our data confirm that AutoHSCT is a very valuable therapeutic option for pts with AML, especially older and lacking a related HLA-compatible donors Haploidentical 2 ND ALLO-HSCT For Treatment of Acute Leukemia (AL) Relapse after 1 ST ALLO-HSCT: Retrospective Registry Analysis of 63 PTS Relapse was initially treated by cytoreductive chemotherapy in all cases; stage at start of conditioning for haploidentical second allo-HSCT was CR in 27%, active disease in 66% and not evaluated in 8%. Conditioning for 2 nd HSCT was myeloablative/reduced in 14%/86% To overcome the HLA barrier, 23 patients (36%) received ex vivo T-cell depletion (TCD), following either CD3/CD19 negative or CD34 positive selection. 4 pts received in vivo TCD only, 2 received no TCD, 35 pts (55%) received high-dose cyclophosphamide post-transplant according to the Baltimore protocol. Results: Neutrophil engraftment was achieved after a median of 12 days (range, 8-26). 50 pts (78%) achieved CR after 2 nd haploidentical allo-HSCT, out of which 23 (46%) relapsed again. After a median follow-up of 425 days, 47 pts had died, 22 from leukemia, and 25 from treatmentrelated causes. Kaplan-Meier estimated overall survival at 1 and 2 years from haploidentical 2 nd HSCT was 41+/-6% and 19+/-6% Allogeneic Transplantation in Elderly Patients with AML and MDS Comparing two Reduced-Intensity Regimes S Germany Study aim and method: Different reduced-intensity conditioning (RIC) n=24 ≥66 years) or FBM (n=48; n=35 ≥66 years) followed by serotherapy with ATG in all pts. Median patient age was 66 years (68 years FBM; 64 years FLAMSA-Bu). Results: The hematopoietic cell transplantation comorbidity index (HCT-CI) was higher for FBM patients than for the FLAMSA-BU with 26 the FBM vs 58% and 56.7% for the FLAMSA-Bu group, respectively. Both conditioning regimens are feasible and provide similar rates of OS, NRM and GvHD. Conflict of interest Diagnosis and antimicrobial therapy of lung infiltrates in febrile neutropenic patients (allogeneic SCT excluded): updated guidelines of the Infectious Diseases Working Party (AGIHO) of the German Society of Hematology and Medical Oncology (DGHO) How I manage pulmonary nodular lesions and nodular infiltrates in patients with hematologic malignancies or undergoing hematopoietic cell transplantation Pulmonary infiltrates in non-HIV immunocompromised patients: a diagnostic approach using non-invasive and bronchoscopic procederes Etiologies, diagnostic strategies, and outcomes of diffuse pulmonary infiltrates cuasing acute respiratory failure in cancer patients: a retrospective observational study Pneumonia Turing remission induction chemotherapy in patients with acute leucemia Incidence and outcome of pneumonia in patients with acute leucemia receiving first induction therapy with anthracycline-containing regimens Pulmonary complications following hematopoietic stem cell transplantation: diagnostic approaches Rational approach to pulmonary infiltrates in leukemia and transplantation Bacteremic pneumonia in neutropenic patients with cancer. Causes, empirical antibiotic therapy, and outcome Surveillance of nosocomial sepsis and pneumonia in patients with a bone marrow or Peripherals blood stem cell transplant. A multicenter project Pneumonia in allogeneic stem cell transplantation recipients: a multicenter prospective study Incidence, risk factors, and mortality from pneumonia developing late after hematopoietic stem cell transplantation Parainfluenza virus lower respiratory tract disease alter hematopoietic cell transplant: viral detection in the lung predict outcome Fatal human metapneumovirus infection following allogeneic hematopoietic stem cell transplantation Respiratory syncytial virus infection in patients with hematological diseases: singlecenter study and review of the literature Outcome of pandemic H1N1 infections in hematopoiteic stem cell transplant recipients Invasive aspergillosis complicating pandemic influenza A (H1N1) infection in severely immunocompromised patients Neuraminidase inhibitors improbé outcome of patients with leucemia and influenza: an observational study The characteristics and outcomes of parainfluenza virus infection in 200 patients with leucemia or recipients of hematopoietic stem cell transplantation Internal tandem duplication of the flt3 gene found in acute myeloid leukemia The presence of a FLT3 internal tandem duplication in patients with acute myeloid leukemia (AML) adds important prognostic information to cytogenetic risk group and response to the first cycle of chemotherapy: analysis of 854 patients from the United Kingdom Medical Research Council AML 10 and 12 trials Inhibition of mutant FLT3 receptors in leukemia cells by the small molecule tyrosine kinase inhibitor PKC412 FLT3 internal tandem duplication mutations associated with human acute myeloid leukemias induce myeloproliferative disease in a murine bone marrow transplant model Phase 1 clinical results with tandutinib (MLN518), a novel FLT3 antagonist, in patients with acute myelogenous leukemia or high-risk myelodysplastic syndrome: safety, pharmacokinetics, and pharmacodynamics Single-agent CEP-701, a novel FLT3 inhibitor, shows biologic and clinical activity in patients with relapsed or refractory acute myeloid leukemia Results from a randomized trial of salvage chemotherapy followed by lestaurtinib for FLT3 mutant AML patients in first relapse Phase IIB Trial of Oral Midostaurin (PKC412), the FMS-Like Tyrosine Kinase 3 Receptor (FLT3) and Multi-Targeted Kinase Inhibitor, in Patients With Acute Myeloid Leukemia and High-Risk Myelodysplastic Syndrome With Either Wild-Type or Mutated FLT3 Phase IB study of the FLT3 kinase inhibitor midostaurin with chemotherapy in younger newly diagnosed adult patients with acute myeloid leukemia Phase I study of quizartinib administered daily to patients with relapsed or refractory acute myeloid leukemia irrespective of FMS-like tyrosine kinase 3-internal tandem duplication status Results of a phase 2 randomized, open-label, study of lower doses of Quizartinib (AC220;ASP2689) in subjects with FLT3-ITD positive relasped or refractory acute myeloid leukemai Validation of ITD mutations in FLT3 as a therapeutic target in human acute myeloid leukemia Results of a phase I study of quizartinib (AC220, ASP2689) in combination with induction and consolidation chemotherapy in younger patients with newly diagnosed acute myeloid leukemia Crenolanib is a potent inhibitor of FLT3 with activity against resistance-conferring point mutants Mutant FLT3: a direct target of sorafenib in acute myelogenous leukemia Phase I/II Study of Combination Therapy With Sorafenib, Idarubicin, and Cytarabine in Younger Patients With Acute Myeloid Leukemia Phase 2 study of azacytidine plus sorafenib in patients with acute myeloid leukemia and FLT-3 internal tandem duplication mutation Sorafenib Versus Placebo in Addition to Standard Therapy in Adult Patients >=60 Years with Newly Diagnosed Acute Myeloid Leukemia: Results From the Randomized-Controlled Soraml Trial Compassionate use of sorafenib in FLT3-ITD-positive acute myeloid leukemia: sustained regression before and after allogeneic stem cell transplantation Conflict of interest: The author declares that she has no conflict of Up to 25% of patients with profound neutropenia lasting for >10 days develop lung infiltrates, but pathogens isolated from blood cultures, bronchoalveolar lavage or respiratory secretions may not be etiologically relevant. Testing for Aspergillus galactomannan, beta-D-glucan or DNA in blood, bronchoalveolar lavage or tissue samples may facilitate the diagnosis. Pulmonary toxicity from cytotoxic drugs or radiotherapy as well as pulmonary involvement by the underlying malignancy should be included into differential diagnosis and eventually give reason for invasive diagnostic procedures. In high-risk patients, pre-emptive treatment with moldactive systemic antifungal agents improves clinical outcome, while other microorganisms are preferably treated only when documented. High-dose trimethoprim-sulfamethoxazole is first choice for treatment of Pneumocystis pneumonia, while cytomegalovirus pneumonia is treated primarily with ganciclovir or foscarnet. 4 Hematology Unit, Medical School of Athens, University General Hospital "Attikon", Athens, Greece Purpose: Mutations in the additional sex comb-like 1 (ASXL1) gene were recently shown in various myeloid malignancies including AML. The vast majority of these mutations which affect the twelfth exon of the gene, are frameshift and nonsense mutations and they are generally associated with an aggressive phenotype and poor clinical outcome. The encoded protein is involved in the epigenetic regulation of gene expression. This study aimed to identify the ASXL1 mutations in de novo and secondary AML patients and associate them with the cytogenetic findings. Methods: Our study included 200 AML patients (148 de novo AML and 52 s-AML) and 10 healthy donors. Conventional cytogenetic analysis was performed on unstimulated bone marrow cells. In order to detect ASXL1 exon 12 mutations molecular analysis was performed by PCR and subsequent direct Sanger sequencing in all patients' and controls' samples. Results: A successful karyotypic analysis was performed in all AML samples at diagnosis. Among them, 128 (64%) exhibited clonal karyotypic abnormalities while 72 (36%) showed a normal karyotype. ASXL1 mutations were detected in 38 out of 200 patients (19%). A significantly higher frequency of ASXL1 mutations was observed in s-AML compared to de novo AML patients (28.8% vs 15.5%, respectively; p=0.035). None ASXL1 mutation was observed in the control group. The most frequent ASXL1 mutation was the c.1934dupG (31/38, 81.5%). ASXL1 mutations were more frequently found in males than in females (20.4% vs 18.8%, respectively), while stratification of patients according to age revealed a higher frequency of ASXL1 mutations in patients ≥60years than patients <60years (p=0.004). Among the 38 AML patients with ASXL1 mutations, 9 patients had trisomy 8 (23.7%), 12 had a normal karyotype (31.5%), 2 had t(8;21)(5.3%), 2 had trisomy 13 (5.3%) and 13 patients other chromosomal abnormalities (34.2%). Conclusions: These results suggest that ASXL1 mutations are frequent in AML, especially in s-AML. ASXL1 mutations are highly correlated with older age, male sex and specific karyotypic aberrations. The most common chromosomal abnormality associated with ASXL1 mutations is trisomy 8. 6 Institute for Medical Informatics, Biometry and Epidemiology, LMU Munich, Germany Introduction: Whole exome sequencing (WES) or custom gene panel sequencing (GPS) allows to detect gain or loss of genomic material in tumor cells (i.e. copy number alteration; CNA) and thereby whole chromosome aneuploidies and submicroscopic amplifications or deletions of genomic regions. We performed WES at diagnosis and complete remission of AML patients with partial deletions of chromosome 9 (AML del(9q), n=5) as well as GPS of known mutational targets in AML and candidate genes located on 9q of 26 diagnostic AML del(9q) samples and 21 AML patients without any detectable cytogenetic aberration on chromosome 9. Method: CNA profiling based on WES showed somatically acquired stretches of significantly reduced read counts for genes located on 9q in 2 of the 5 patients, which is in line with corresponding SNP array results. CNA profiling based on GPS defined a common deleted region with a size of at least 8.4 Mb, detected in 18 out of 26 (70%) patients. Patient samples without detected 9q deletion in our CNA analysis have significantly fewer metaphases with 9q deletion (median 27%, range 8-77%) compared to samples tested positive for 9q deletion by our CNA analysis (median 95%, range 24-100%; p=0.001), as reported by routine cytogenetics. Furthermore, the varying clone size harboring the 9q deletion in the diagnostic samples is limiting for CNA detection. Result: Beyond the identification of somatic mutations with single nucleotide resolution, sequencing data enables the efficient detection of recurring and/or somatic CNAs in AML. Purpose: Deregulated DNA repair is one of the hallmarks of cancers including Acute Myeloid Leukemia (AML), as it results in genomic instability. ATM gene functions as a sensor, activates cascade of events leading to stimulation of multiple DNA damage-responsive signaling pathways. Principal DNA repair mechanism activated in the hematopoietic stem cells is the Non Homologous End Joining (NHEJ) pathway. However, this pathway was shown to be error prone. Functional SNPs in the genes involved in DNA repair might influence the gene expression leading to altered DNA repair which might confer the risk to AML. Methods: This hospital-based case-control study included 225 AML patients and 326 cancer-free controls from South Indian population. Six polymorphisms of XRCC5, XRCC6, XRCC7 and ATM were genotyped using polymerase chain reaction (PCR)-Restriction Fragment Length Polymorphism (PCR-RFLP) method. Statistical analyses were performed by using SPSS (version 20.v) and SNPSTAT online tool. Protein-Protein Interaction (PPI) analysis was also done to see the relationship between these genes. Results: We found that there was an elevated risk of AML associated with the XRCC5 VNTR 0R repeat and A allele of 2408G>A polymorphism (p-0.04 and p<0.0001 respectively). The frequencies of G allele (p-<0.0001) of XRCC6 -1310C>G and T allele (p-0.003) of ATM -5144A>T polymorphisms were also significantly increased in AML cases. Further, analyses of the variant genotypes with epidemiological and clinical variables revealed a significant association of the risk genotypes with development and progression of AML. However, XRCC6 -61C>G and XRCC7 6721G>>T polymorphisms did not show any association with AML Introduction: Chronic lymphocytic leukemia (CLL) is a high-risk disease for developing secondary neoplasms due to immunodeficiency associated with the disease and/or with the treatment. Acute myeloid leukemia(AML)or myelodysplastic syndrome (MDS) was a rarely entity, these were observed in 3.5% to 6.1% of patients treated with fludarabine alone or in combination. AML or MDS coexisting with untreated CLL were also reported. Case report: We report here a case of 60-year female with the diagnosis of CLL Rai stage 1 with no chromosomal changes, who developed AML preceded by t-MDS, 3 years after completion of CLL therapy consisted with 6 courses with FC-R regimen. She achieved CR with incomplete recovery of platelet and WBC count. In August 2009 she was admitted to hospital with symptoms of pleuro-pneumonia and deep pancytopenia with 5% of blasts in peripheral blood (PB)and 15% in bone marrow aspirate-RAEB was diagnosed. A week later performed BM revealed 42% of blasts which displayed a myelomonocytic phenotype.Cytogenetic analysis showed complex chromosomal abnormalities; del 5 (13q33), trisomy 8, trisomy 20, monosomy 21. No abnormalities concerning chromosomes; 11, 12, 13 and 17 were seen. The diagnosis of AML/RAEB-t was made. No CLL cell clone was found in BM, no organomegaly was seen in CT examination. The patient was still in CR of CLL. The patient received induction chemotherapy with daunorubicine plus cytarabine(3+5)and obtained CR. Relapse occurred three months later. The patient died 8 months later due to leukemia with its complications (infections, haemorrhages from GI tract). Conclusions: In presented case we can consider drug-related origin of AML in patient with complete remission of CLL or another myeloid neoplasm from unknown origin. Purpose: Genetic barcodes (BCs) are short sequences that are introduced into the genome as an expression-independent method for stable and unique marking of cells and their progeny. We used a BcrAbl-BCtransduced cell line to investigate clonal competition in vitro as well as in the development and progression of BcrAbl-induced leukaemia in the mouse model. Methods: We constructed a γ-retroviral vector encoding BcrAbl in conjunction with GFP. In this vector, our BC32-sequence was introduced to establish a γ-GFP-BC32-BcrAbl-vector plasmid library. γ-retroviral vectors were generated and used to transduce the murine, IL-3-dependent cell line Ba/F3. After IL-3 withdrawal, cells were cultured for 52 days; samples were taken every 2-3 days to follow the clonal dynamics in vitro. Background: The success of anti-leukemia treatment depends on the response of leukemia cells towards drug treatment; treatment resistance leads to persistence of tumor cells, post-therapeutic residual disease and potentially disease relapse. Patients and methods: Here, we studied the influence of the feeder cells on drug sensitivity of patient-derived ALL cells in vitro. Towards this aim, primary ALL cells from two children with high-risk or relapsed BCP-ALL were amplified in immunocompromised mice and cells freshly isolated from mice were subjected to in vitro experiments. Several cytotoxic drugs of routine ALL treatment were used to stimulate patient-derived ALL cells in vitro. Stimulations were performed in the presence Introduction: The T315I gatekeeper mutation confers resistance to majority of approved ABL1 inhibitors, with only ponatinib demonstrating efficacy in BCR-ABL1(T315I)-driven disease. However, vascular adverse events and selection of resistant compound mutations limit its clinical utility. Hence, there is an unmet need for novel treatment strategies for patients with gatekeeper-mutated Ph+ leukemia. Method: In this study we integrated comprehensive drug sensitivity testing with structural analysis to characterize the tyrosine kinase inhibitor axitinib as a potential novel therapy for BCR-ABL1(T315I)-driven leukemias. Ex vivo drug sensitivity testing of primary cells derived from a Ph+ ALL patient revealed a cancer-selective response to the VEGF inhibitor axitinib (currently approved for renal cell carcinoma). Strikingly, axitinib exhibited higher sensitivity in T315I positive Ph+ patient samples in comparison to T315I negative patient samples. In accordance, axitinib inhibited the kinase activity of ABL1(T315I) with similar potency as its intended primary target VEGFR2, while showing >30-fold lower potency towards non-mutated ABL1. Analogously, in engineered Ba/F3 cells, axitinib showed a 10-fold higher potency to inhibit T315I than nonmutated ABL1-driven autophosphorylation and cell growth. The axitinib:ABL1(T315I) crystal structure revealed that axitinib bound to a T315I-mutant induced active conformation of ABL1(T315I), different than the binding mode in non-mutated ABL1, likely explaining the increased potency towards the gatekeeper mutated protein.Result: Compassionate two week treatment of a CML patient harboring the T315I mutation with axitinib, resulted in a rapid 5-fold reduction of T315I transcript levels in the bone marrow, further suggesting that axitinib can produce specific and effective responses in patients with BCR-ABL1(T315I)-driven leukemia. Conclusion: We demonstrate that axitinib potently inhibits BCR-ABL1(T315I) via a gatekeeper mutant-selective mechanism. Since axitinib is in clinical use for treatment of refractory renal cell carcinoma with a manageable safety profile, our data provide a sound basis for readily repurposing axitinib for BCR-ABL1(T315I)-driven leukemia. Finally, the distinct mechanism of inhibition by axitinib serves as an exemplar for development of even more effective gatekeeper-mutant selective inhibitors targeting ABL1 as well as other clinically important kinases, such as EGFR, FLT3 and KIT. ). NALM6 cALL line was retrovirally transduced with MLXIP short hairpin RNA, which reduced the MLXIP expression by 80%. Upon MLXIP knock down (kd), kd and control lines were injected iv; CD10+ blasts in blood, spleen and bone marrow were assessed after 30 days. Importantly, in vivo MLXIP maintained 90-99% of CD10+ leukemic blasts. Spleen size and weight became normal by MLXIP kd: Results: While signs of leukemia engraftment were observed in all mice, the differences of CD10+ blasts in blood, bone marrow and spleens in the control vs. the MLXIP kd group were highly significant (p=0.008). The decrease of leukemic splenomegaly after MLXIP kd was most impressive: Median spleen weight was 0.22g vs. 0.08g in the control vs. the kd group. In conclusion, these findings demonstrate that MLXIP maintains leukemic burden and malignancy of cALL in vivo. Introduction: Personalized molecular medicine is gaining traction as more doctors are insisting on reliable biomarker testing to diagnose and stratify their patients for treatment or enrolment in clinical trials. Pharmaceutical companies seek to shorten drug development by being more selective (recruiting "the right patient for the right drug") before trial enrolment. Wider recruitment to access sufficient patient populations to power development of new therapeutic agents requires more laboratories with harmonized testing for reliable stratification and selection for trials. Standardized and harmonized international clinical testing is now required. Analytical metrics were used to harmonize and validate a next generation sequencing (NGS) assay manufactured under ISO 13485 for clinical use in our US and EU clinical labs. Methods: All reagents were manufactured under cGMP and QC tested under a QSR compliant regulatory system for use in validation studies. Site to site validation tested Limit of Detection (LOD), Blank (LOB), Linearity, Precision and Reproducibility (Repro) to pre-specified standards. The LymphoTrack® IGH PGM assay was used for this validation. LOD and LOB validation consisted of 1 operator using 1 instrument to run a dilution series of a cell line DNAs in tonsil DNAs in each lab. Samples were run in duplicate with 1 lot of master mix and a 1 lot of tonsil DNA control. The Repro study included 2 runs by each of 2 operators in each lab (8 runs total). JMP software assessed the LOD, LOB and Repro runs of the assay based on 96 data points. Results: Concordance between the labs was excellent with %CVs <20% and both labs showing inter and intra repeatability and reproducibility between labs, demonstrating harmonization among labs that use cGMP reagents and assays. (see table) . Furthermore, patients with trisomy 8 exclusively had a poor outcome. Interestingly, monosomies (excluding -7) were detected in 19 patients, out of whom 12 belonged to the genetically high risk group due to other aberrations. However, the remaining 7 patients showed a dismal outcome, too. Background: Allogeneic hematopoietic stem cell transplantation (HSCT) has become a cornerstone in the treatment of patients at high risk for relapse of their acute leukemia. Initially monitoring of chimerism after HSCT was applied to confirm donor cell engraftment, but, since relapse is host derived, the increase of host chimerism might indicate imminent relapse. However, commonly used STR assays are limited by low sensitivity, whereas XY-FISH, CD34+ cell chimerism and MRD approaches are limited by applicability. Study aim: In this prospective, non-interventional study we evaluated the accuracy, reliability and feasibility of a quantitative PCR based commercially available assay (Allele SEQR® Chimerism Assay, Abbott) and its potential for early relapse detection. Between 05/11 and 01/13 95 consecutive patients that received allo-HSCT for AML/MDS, were enrolled. Method and results: According to local standards, bone marrow samples were tested at day 30, 90 and 180 after allo-HSCT and patients were followed up until 03/14. Informative markers were available in all screened patients (n=68). High sensitivity (<0,05% at 100ng DNA) translated in a remarkable relapse prediction/ exclusion rate in 61 repetitively tested patients (relapse rate 86% vs. 15%, p=<0,001, Mantel-Byar) with a median time from positive testing to relapse of 68 days resp. no relapse within >60days after negative testing. Conclusion: Hence, this approach might facilitate personalized monitoring and chimerism-triggered therapy. Late Submissions Abstract Lung infiltrates are frequent in patients treated for acute leukemia and in hematopoietic stem cell transplant (HSCT) recipients. The development of pulmonary infiltrates is associated with significant morbidity and mortality in these severely immunocompromised patients. The differential diagnosis of lung infiltrates is wide and should include infectious and non-infectious causes. Pneumonia remains a major cause of mortality among leukemia and HSCT patients, espeially when complicated with bacteremia. Among non-infectious causes of pulmonary infiltrates, diffuse alveolar hemorrhage, and pulmonary edema are particularly common. This short review specifically focuses on the current epidemiology of lung infiltrates in patients with acute leukemia and HSCT recipients. The development of lung infiltrates in patients treated for acute leukemia and in patients undergoing stem cell transplantation (HSCT) is a frequent life threatening complication, requiring early diagnosis and treatment. [1] [2] [3] [4] In particular, it has been estimated that patients with high-risk neutropenia, develop pulmonary infiltrates with a likelihood of 20-25%. 1, 5, 6 Pulmonary infiltrates in leukemia and HSCT patients pose a major challenge for clinicians because of the wide differential diagnosis, including infectious and non-infectious causes. 7, 8 In a recent study, noninvasive and bronchoscopic diagnostic techniques were applied to determine the etiology of the pulmonary infiltrates in 200 consecutive patients with different non-HIV conditions, of whom 121 (60%) had hematologic malignancies or were HSCT recipients. 3 The etiology of lung infiltrates was infectious in 23% of cases, non-infectious in 23%, and in 19% remained undetermined. The main infectious etiologies were bacterial (24%), fungal (17%), and viral (10%). The most frequent pathogens were Aspergillus fumigatus, Staphylococcus aureus, and Pseudomonas aeruginosa. Among the noninfectious etiologies, diffuse alveolar hemorrhage, pulmonary edema, and bronchiolitis obliterans organizing pneumonia were the most common causes. 3 Other non-infectious causes that should be carefully considered in the differential diagnosis of pulmonary infiltrates include lung involvement by the underlying malignancy, immune reconstitution syndrome and lesions caused by chemotherapy or radiation. Importantly, pneumonia is a common cause of mortality among leukemia and HSCT patients. In a retrospective study of 801 patients with acute leukemia who underwent induction chemotherapy, pneumonia occurred in 233 (29%) of them. 5 One or more pathogens were identified in 79 (34%) cases; gram-positive bacteria were present in 67% of patients, gram-negatives in 30%, fungi in 23%, and viruses in 10%. In a multivariate analysis, age greater than or equal to 60 years, acute myeloid leukemia, low platelet count, low albumin level, neutropenia, and neutrophil count greater than 7,300 were risk factors. The case fatality rate of pneumonia was 17% (40 of 233 patients). 5 In a retrospective study of 458 patients affected by acute leukemia receiving an anthracycline-containing induction regimen, 109 (23.8%) developed pneumonia. 6 Advanced age and total blast count were the only factors independently associated with a higher risk of developing pneumonia. The case fatality rate of pneumonia was 38.5% (42 of 109 patients). 6 Attributable mortality of pneumonia was 55% and overall mortality reached 62.5% in a prospective series of 40 consecutive cases of bacteremic pneumonia in adult neutropenic patients with cancer. 9 In that study, P. aeruginosa and Streptococcus pneumoniae caused 72.5% of all episodes of bacteremic pneumonia. According to data from the Hospital Infection Surveillance System for Patients with Hmatologic/Oncologic Malignancies (ONKO-KISS), initiated in Germany in 2000, the pooled mean site-specific incidence density per 1000 neutropenic days was 5.9 for pneumonia (6.1 and 5.6 in patients with allogeneic and autologous transplantation, respectively The development of pneumonia and chronic graft-versus-host disease (GVHD) was associated was associated with increased mortality after allogeneic HSCT, and the probability of survival was significantly lower in patients that had at least one pneumonia episode. Pneumonia development in the first 100 days after transplantation, fungal etiology, GVHD, acute respiratory failure, and septic shock were associated with increased mortality after pneumonia. 11 The incidence, etiology, outcome, and risk factors for developing pneumonia late after HSCT were investigated in 1359 patients who underwent transplantation in Seattle. 12 A total of 341 patients (25%) developed one or more episodes of pneumonia. Etiological diagnoses was established in 144 patients, including virus (10%), bacteria (9%), P. jirovecii (5%), and other fungi (6%). The overall cumulative incidence of first pneumonia at 4 years after discharge home was 31%. Factors independently associated with development of late pneumonia after allografting were increasing patient age (RR 0.5 for <20 years, 1.2 for >40 years, P=0.009), donor HLA-mismatch (RR 1.6 for unrelated/mismatched related, P=0.01), and chronic GVHD (RR 1.5, P=0.007). 12 Finally, it should be pointed out, that respiratory viruses such as Influenza, Parainfluenza, Coronavirus, Rhinovirus, Human Metapneumovirus or Respiratory Syncytial Virus are an important yet underestimated cause of respiratory tract infection and lung infiltrates in immunocompromised patients, particularly among allogeneic HSCT recipients. [13] [14] [15] [16] In individual patients, respiratory viruses such as Influenza A (H1N1)pdm09 may also facilitate the development of invasive aspergillosis. 17 In immunosuppressed patients affected by these viruses, neutropenia, poor APACHE II score, age over 65 years and severe lymphocytopenia have been reported prognostic factors resulting in a higher risk of fatal outcome. In summary, the development of lung infiltrates occur frequently in leukemia and HSCT patients and is associated with significant morbidity and mortality. The differential diagnosis of pulmonary infiltrates is wide and includes infectious and non-infectious complications. In particular, pneumonia remains one of the most significant causes of mortality among these severely immunocompromised patients. Two decades ago when AML blasts were found to harbor FLT3 mutations in approximately 35% of patients 1 there was hope that this gain-of function mutation in a tyrosine kinase inhibitor might be an important therapeutic target in this relatively resistant disease. Moreover, recognition that the most common type of mutation, namely a length or internal tandem duplication in the juxtamembrane region of the FLT3 receptor, was associated with relatively poor prognosis 2 further spurred therapeutic development. A number of FLT3 inhibitors were shown to have important in vitro activity in murine models. Such agents inhibit the growth of leukemic cell lines whose growth factor independence was based on a FLT3 ITD transduced gene 3 or improve outcomes of a fatal myeloproliferative disorder from a FLT3 ITD stem cell transplant. 4 However, the lack of an approved FLT3 inhibitor in 2015 suggests that this potential therapeutic avenue was paved with rocks and stones rather than gold. However, there is still significant hope that one or more FLT3inhibitors might yet find their way into the therapeutic armamentarium of AML drugs either as a single agent or in combination with chemotherapy. Some of the pitfalls in the therapeutic development of FLT3 inhibitors in AML have been inconsistent potency, protein binding leading to the lack of sufficient drug levels for prolonged enzyme inhibition, off-target effects, non-specificity, high FLT3 ligand levels, and probably most importantly the issue that a FLT3 mutation is a relatively late 'hit' in the pathophysiology of AML. The FLT3 inhibitors which have undergone clinical development in AML to greater or lesser degrees to fall into three categories: 1) "early" agents (tandutinib, lestaurtinib, and midostaurin), "later" agents (quizartinib, PLX3397, and crenolinib) and 3) sorafenib. Tandutinib emerged out of solid pre-clinical testing, had activity in early phase trials 5 , but a business decision led to its abandonment as a therapeutic agent. Lestaurtinib displayed single agent activity in a 14 patient trial conducted at Johns Hopkins. 6 However, a randomized trial in which chemotherapy (the choice of which dependent on the disease free interval) plus or minus lestaurtinib was given to patients with relapsed FLT3 ITD mutant AML showed no benefit for the addition of lestaurtinib to chemotherapy. 7 Ancillary studies revealed that the failure of this trial may have been due in part to the fact that insufficient lestaurtinib levels in plasma were present in many patients. Those who achieved adequate lestaurtinib levels for long enough period of time did experience a benefit. Nonetheless, because of the negative results lestaurtinib will likely not be developed further in AML. The third agent which has been in development for over 10 years, midostaurin, is a multi-targeted kinase inhibitor that inhibits both tyrosine kinases and serine-theonine kinases. Single agent trials with this drug at various doses showed that it had the ability to reduce the blast counts in many patients with FTL3 mutant AML (and in a few patients with FLT3 wild-type AML) 8 but important clinical responses were rare. Nonetheless, the biological activity of this drug was encouraging enough so that it was combined with chemotherapy in a phase IB trial. While higher doses of midostaurin were not well tolerated in combination with standard induction chemotherapy, a tolerable dose of 50 miligrams twice daily in combination with chemotherapy was retrospectively shown to improve the disease free and overall survival of FLT3 ITD patients to a range seen in FLT3 wild type patients. 9 These relatively encouraging results prompted a large phase III trial (CALGB 10603; RATIFTY) in patients aged 18-60 with FLT3 mutant AML in which standard induction chemotherapy and high dose ara-C consolidation was combined with either placebo or midostaurin for 14 days per cycle followed by continuous maintenance placebo or midostaurin for a year. Midostaurin, unlike most of the other FLT3 inhibitors (except for crenolinib to be described below) can inhibit both the FLT3 ITD and FLT3 tyrosine kinase domain mutants, allowing accrual of patients with both FLT3 mutant subtypes (n=714 over 4 years). The primary endpoint of the trial was overall survival (not censored for stem cell transplant) to be analyzed after 504 deaths; however, perhaps because of a high rate of stem cell transplantation (55% overall, 24% in CR1), these events have not yet occurred. Event-free survival data will hopefully soon be available; discussions with regulatory agencies are ongoing to allow release of this data. Thus, it is possible that midostaurin might be approved as a drug to be combined with chemotherapy for mutant FLT3 AML. One of the newer agents, PLX3397 10 , has undergone phase I/II testing in patients with mutant FLT3 AML. This drug is more specific and perhaps better tolerated than the first generation agents. Initial results are encouraging; it is unclear if and when this drug will be developed further in AML. Probably the most promising agent for potential use as a single agent in AML is quizartinib which has the advantage of a high degree of potency and specificity for the FLT3 ITD mutation. In phase I/II trials the drug showed an impressive level of activity in terms of reducing the blast count in blood and marrow, 11, 12 although true complete remissions were noted in less than 10% of treated patients. Interestingly for this specific drug, there were also some responses in FLT3 wild-type patients; however, the overall response rate of 60% in FLT3 mutant patients is impressive. One of the potential drawbacks of this agent is that it selects for secondary tyrosine kinase domain mutations, providing mechanism for early resistance. 13 These results have led to registration ongoing trial of this drug vs physicians'-choice chemotherapy for patients with relapsed mutant FLT3 AML. The drug has also been tolerably combined with chemotherapy 14 ; the U.S. Intergroup has plans to perform a randomized trial of chemotherapy plus or minus quizartinib during induction, post-remission therapy and specifically after stem cell transplant.Crenolinib have the potential advantage of inhibiting both FLT3 ITD and FLT3 TKD mutations. 15 Phase I evaluation of this agent has suggested that there is a significant level of activity. Potentially this drug will also be combined with chemotherapy in preparation for a phase III chemo plus or minus crenolinib trial.