Search

Hydroxyurea is efficacious against sickle cell anaemia (SCA)-related complications in African children. Prior studies demonstrated conflicting results on the effect of hydroxyurea on risk of infection, the most common cause of morbidity and mortality in African children with SCA. We evaluated the incidence of infections before and after starting hydroxyurea in 117 children aged 1-5 years with SCA enrolled in the Zinc for Infection Prevention in Sickle cell anaemia (ZIPS) clinical trial that received zinc or placebo treatment for one year (Clinicaltrials.gov, NCT03528434). Children were enrolled between March 2018 and November 2019 and initiated on hydroxyurea (20 mg/kg/day) if they met Uganda SCA guideline criteria for hydroxyurea treatment at any time during the study. We compared the incidence of infections before and after hydroxyurea therapy, adjusting for zinc treatment. Overall, the mean duration on hydroxyurea was 223.8 (85.2) person days. The mean(SD) incidence of any severe/invasive infections (infections meeting strict clinical and laboratory or radiological diagnostic criteria) was 6.2(9.0) vs. 1.9(2.3) infections per child per year before and after hydroxyurea (incidence rate ratio [IRR]: 0.40, 95%CI: 0.29-0.54, p In Ugandan children with SCA, hydroxyurea therapy not only decreases the incidence of SCA-related complications, but also substantially reduces the incidence of infections. Research to understand the underlying mechanisms of protection from hydroxyurea against infection and exploration of its potential use for infection prevention is warranted. Disclosures Ware: Bristol Myers Squibb: Research Funding; Addmedica: Research Funding; Hemex Health: Research Funding; Nova Laboratories: Research Funding; Novartis: Other: DSMB Chair; Editas: Other: DSMB Chair.

COVID-19 and its prevention has put considerable strain on health care systems in low and middle-income countries (LMIC). In Uganda, a national lockdown was declared on March 18, 2020, in response to COVID-19 pandemic and concern of spread of cases without aggressive measures to prevent spread. The lockdown consisted of closure of all offices except essential ones, orders to stay at home unless an emergency occurred, school closure, a ban on all meetings of more than 10 people, a ban on public and private transport, closing down of all shops, malls, restaurants, places of worship and other facilities in which group meetings might occur, keeping a distance of at least 2 metres from other people in public places and a 7:00 p.m. to 6:30 a.m. curfew. Hospitals however remained open and operational. We describe the impact of the lockdown in Uganda in response to the COVID-19 pandemic on the morbidity and mortality in children with sickle cell anaemia (SCA) at a tertiary hospital in Uganda. The number of clinic visits for SCA related complications and death were compared in the pre-lockdown (November 2019 to February 2020) and during COVID-19 lockdown periods (March 2020 to June 2020) in children aged 1- 4.99 years enrolled in a SCA research study [Zinc for Infection Prevention in Sickle cell anaemia (NCT03528434)] at Jinja Hospital, Uganda. In the study, children with SCA are asked to return to the hospital for evaluation whenever they are unwell. Follow up phone calls are made to ascertain the wellbeing of the children and identify any who are unable to come to the hospital. During the lockdown, follow up calls continued and facilitation was provided for caregivers to bring any child who was unwell to the hospital for evaluation. A total of 238 children with a mean (standard deviation) age of 2.7(1.1) years were enrolled and were being followed up when the pandemic started. The incidence of hospital sick visits pre-lockdown and during the lockdown period was 7.7 vs 4.0 person-year, (p= Disclosures No relevant conflicts of interest to declare.

Introduction Hyperhemolysis has been well described in delayed hemolytic transfusion reaction (DHTR), which is one of the most serious complications of transfusion especially in patients with sickle cell disease (SCD). In the most severe cases, this life-threatening syndrome may reach 10% lysis of total red blood cells (RBCs) (Habibi A et. al. Am J Hematol 2016). DHTR is characterized by an acute anemia with increased plasma free Hb and free heme. Hyperhemolysis can also be found in other hemoglobinopathies and extracorporeal circulation procedures. Most in vitro studies on hyperhemolysis used only purified Hb or heme which did not allow to investigate all deleterious effects on the endothelium. We have developed a global and physiopathologic approach to assess the mechanism of hyperhemolysis-induced endothelial dysfunction. For this purpose, we created a microfluidic model reproducing hyperhemolysis observed in the most severe DHTR but it's also applicable for other pathogenesis of hyperhemolysis. Among the potential therapeutic approaches, carbon monoxide (CO) is beneficial on endothelial cells by its anti-inflammatory, antioxidant and vasodilator effect which reduce the toxicity of Hb and improve tissue perfusion (Motterlini R et.al. Nat. Rev. Drug Discov 2010). Here we investigated the therapeutic effects of a CO-releasing molecule, CORM-401, in hyperhemolysis-induced endothelial dysfunction. Materials and methods Flow culture Human Umbilical Vein Endothelial Cells (HUVECs) in fibronectin-coated µ-Slides were preconditioned for 4 hours under shear stress of 1dyn/cm² with sonicated RBCs of healthy donors (AA) reconstituted in serum of either AA or SCD patients (SS) at 7 g/L of free Hb. In negative control conditions, cells were pretreated similarly with either AA serum or culture medium. Cells were then collected for analyses of actin network, membrane markers and endothelial function. Treatment with CORM-401 or inactive CORM (iCORM) was performed at fixed concentration during the preconditioning and perfusion steps. ResultsHemolysis induces multiple endothelial damage and dysfunction After exposure to hemolysate, we noted a significant increase of subendothelial matrix exposition (fibronectin-coated surfaces) due to actin network reorganization and cells detachment. Immunofluorescence staining showed a moderate activation of HUVECs induced by hemolysate with membrane expression of adhesion molecules such as ICAM-1, E-Selectin and VCAM-1, but not VWF and P-Selectin. At a similar free Hb level, preconditioning with hemolysate without microparticles (MPs) had the lowest deleterious impact on HUVECs suggesting a role of MPs in hyperhemolysis. Perfusion of AA whole blood on HUVECs pretreated with hemolysate resulted in aggregation and activation of platelets in a GPIIbIIIa-dependent manner at injury sites. We also observed a significant increase of adhered RBCs (14-fold compared to control). Compared to whole blood RBC lysis, purified Hb induced similar subendothelial exposure and RBC adherence but did not lead to significant HUVECs activation. To study the DHTR in RBC transfusion of SCD patients, SS serum was used to perform hemolysis preconditioning of HUVECs. Adherence of RBCs was similar to condition using AA serum suggesting that endothelial damage in DHTR of SCD patients depend on hemolysate compositions and RBC MPs, rather than on patients' serum.Effects of CORM-401 on hemolysis-induced endothelial dysfunction Adding CORM-401 (50 to 100 µM) to the hemolysate produced 15% free COHb leading to a concentration-dependent decrease in hemolysis-induced HUVECs activation and RBC adherence (Figure 1 C, E). While no significant effects at the subendothelial exposure were observed, the aggregation of platelets at injury sites was decreased after CORM-401 treatment compared to controls (Figure 1 A, B, D), confirming the anti-platelet aggregation effect of this molecule (Chlopicki S et. al. Naunyn Schmiedebergs Arch Pharmacol 2012). Conclusion Our in vitro model enabled the reproduction of endothelial damage by hyperhemolysis and to determine the deleterious effects of each hemolysate component. The in vitro beneficial effects of CORM-401 were also demonstrated and an in vivo study on SCD mice is underway to explore the therapeutic effects of this molecule against hyperhemolysis-induced endothelial damage. Figure 1. Figure 1. Disclosures Bartolucci: Addmedica: Research Funding; GBT: Membership on an entity's Board of Directors or advisory committees; Fondation Fabre: Research Funding; Novartis US: Membership on an entity's Board of Directors or advisory committees.

The myelodysplastic syndromes (MDS) are a heterogeneous group of clonal hematopoietic stem cell disorders characterized by ineffective hematopoiesis, peripheral blood cytopenias, dysplasia and a propensity for transformation to acute myeloid leukemia (AML). MDS is frequently associated with epigenetic gene silencing via methylation of cytosine residues in gene regulatory regions, and DNA methyl-transferase 1 (DNMT1) inhibitors, such as 5'azacytidine and 5-aza-2'-deoxycytidine (decitabine, DAC), are two of the three agents that are FDA approved for treatment of MDS. Although these drugs are not curative, they induce hematological improvement or improved survival in a significant fraction of MDS patients. Two novel, thiol-substituted 2'-deoxycytidine (dCyd) analogs designated T-dCyd (4'-thio-2'-deoxycytidine) and Aza-T-dCyd (5-aza-4'-thio-2'-deoxycytidine) were synthesized and shown to be potent DNMT1 inhibitors in vitro. We evaluated these drugs in vivo using the NUP98-HOXD13 (NHD13) mouse model for MDS. To mimic human MDS hematopoiesis, in which a portion of the hematopoietic output is provided by the MDS clone, and a portion provided by normal, non-MDS cells, we transplanted wild-type (WT) mice with a mixture of WT murine hematopoietic cells and NHD13 (MDS) hematopoietic cells. This bone marrow transplant (BMT) produces chimaeric recipients with bone marrow comprised of hematopoietic cells derived from both the MDS clone as well as normal hematopoietic precursors. WT and MDS cells in the mice can be distinguished by differential CD45 alleles (CD45.1 and CD45.2, respectively), which enables analysis and purification of the MDS and WT cells; this feat is not easily achieved with human MDS patient samples, which lack cell surface antigens specific for the MDS clone. At 8 weeks post-transplant; engraftment of MDS cells was documented by the presence of CD45.2+ cells in the peripheral blood, and the starting CBCs showed signs consistent with MDS including peripheral blood cytopenia and macrocytosis. Mice were randomly assigned to one of the three groups. 1) PBS, 2) T-dCyd, 3) Aza-T-dCyd. T-Cyd was dosed at 4 mg/kg/d intraperitoneally (IP) on weekdays for 2 weeks (10 doses), followed by three weeks rest; this constituted one cycle of therapy. Aza-T-dCyd was administered on the same schedule at 4 mg/kg/d IP. Flow cytometry and CBC were assessed on day 21 of each cycle, and treatment continued for up to one year, or until mice were humanely euthanized due to tachypnea, lethargy, or other signs of AML. Between four and six mice were treated per group, and the entire experiment was repeated three times and results pooled for T-dCyd, once for Aza-T-dCyd. The T-dCyd treated chimaeric mice showed significantly enhanced overall survival associated with hematological improvement including hemoglobin concentration, platelet and absolute neutrophil count compared to PBS treated mice (median survival 45.4 vs 28 weeks, p=0.0187). In addition to a survival advantage, AML onset was significantly delayed in the T-dCyd treated mice (median time to AML transformation 35 weeks for PBS vs unreached for T-dCyd, p=0.0111), although there was no significant change in MDS (CD45.2) engraftment between the T-dCyd and PBS treated mice. For Aza-T-dCyd group, we did not detect a survival benefit nor hematologic improvement, although we suspect this may have been secondary to unexpected toxicity at the selected dose. In sum, these results demonstrate the utility of chimaeric WT/MDS mice as a pre-clinical model for human MDS, and show that treatment with T-dCyd, a new DNMT1 inhibitor, leads to a survival advantage, hematologic improvement, and delayed transformation to AML. Disclosures Aplan: NIH Office of Technolgy Transfer: Employment, Patents & Royalties: NUP98-HOXD13 mice.

The vascular adhesion molecule E-selectin is a key component of the Bone Marrow (BM) Haematopoietic Stem Cell (HSC) niche prompting HSC to proliferate at the expense of self-renewal (Winkler, Nat Med 2012).Only 3 - 5% of BM endothelial cells express E-selectin in steady state however E-selectin is greatly upregulated (5 - 10 fold) in BM of mice with acute myeloid leukemia (AML) raising the question; how do AML stem cells (LSC) respond to E-selectin at the vascular niche & does E-selectin signalling in AML & HSC differ? Using models of murine AML generated by retroviral transduction of MLL-AF9 or AML1-ETO oncogenes, we found leukemic blasts rapidly upregulate E-selectin-binding upon oncogenic transformation. Furthermore E-selectin adhesion promoted LSC survival to cytarabine in vitro as well as in vivo. LSC survival to chemotherapy in wildtype compared to E-selectin knockout (E-/-) mice quantified by rigorous limiting-dilution transplantation assay of 1%, 0.1%, 0.01% femur BM demonstrated that E-selectin deletion increased sensitivity of LSC to high-dose cytarabine therapy ~11-fold (900mg/kg cytarabine n=6 donors &15 recipients/gp p=0.0037). Thus E-selectin is a critical vascular niche component mediating LSC chemo resistance. Importantly these findings could be replicated by administration of a potent small molecule glycomimetic E-selectin antagonist (GMI-1271) to wt mice. Furthermore treatment with GMI-1271 (40mg/kg bidaily) for 10 days in combination with standard mouse version of 7+3 induction chemotherapy (5 days cytarabine 100mg/kg; 3 days doxorubicin 1mg/kg) was able to significantly double mouse survival over chemotherapy alone (p=0.0054; no chemotherapy median survival 25 d, AraC/Dox alone 32 d, AraC/Dox plus GMI-1271 survival 41 d; n=8 mice/gp). To understand mechanisms of this chemo-sensitisation, mice with MLL-AF9 monomyelocytic (11q23 translocation) or AML1-ETO granulocytic t(8;21) -induced AML were administered GMI-1271 or vehicle control for 5 days before sorting BM AML blasts for RNA sequencing. Analysis of differentially expressed transcripts by CuffDiff / DSeq2 revealed 170 RNAs differed following in vivo E-selectin blockade. KEGG pathway analysis indicated a pathway potentially dampened in AML blasts following GMI-1271 administration was PI3K - NF-kB signalling - raising the hypothesis that adhesion to E-selectin activates pro-survival NF-kB signalling in AML cells leading to enhanced chemoresistance. Using two in vitro assays, we confirmed E-selectin to be unique among vascular adhesion molecules tested in being able to directly activate NF-kB. Activation of NF-kB was only observed upon E-selectin mediated adhesion & was not observed following adhesion to P-selectin, PECAM-1 or VCAM-1 using either myeloid NF-kB GFP reporter cell lines or by induction of p65 NF-kB (Ser 536). Importantly E-selectin mediated NF-kB activation was completely inhibited when E-selectin antagonist GMI-1271 added. Assays repeated in presence of a specific NF-kB activation antagonist (BMS-345541 10uM 24hrs) demonstrated that NF-kB blockade alone reversed E-selectin-mediated chemoresistance in vitro. Upstream blockade of E-selectin by GMI-1271 not only inhibits NF-kB activation but also mobilizes LSC out of the protective BM niche & prevents re-entry thereby breaking the chemo resistance observed with these cells. A Phase I/II Clinical trial to study efficacy of GMI-1271 in combination with chemotherapy in AML patients (NCT02306291) is currently in progress Disclosures Winkler: GlycoMimetics: Research Funding. Magnani:GlycoMimetics: Employment, Equity Ownership, Membership on an entity's Board of Directors or advisory committees. Levesque:GlycoMimetics: Equity Ownership.

Women with anticardiolipin antibodies (ACLA) are at increased risk for fetal loss. One potential explanation for this outcome is that sera from these individuals contain antibodies reactive with trophoblast cells, which are involved in the establishment of the uteroplacental vasculature and maintenance of placental blood fluidity. To examine this hypothesis, we compared the incidence of trophoblast-reactive antibodies in 27 patients with ACLA and a history of fetal loss with that in 29 normal pregnant women. Sera from 20 patients, but only one control, contained trophoblast-reactive antibodies (P < .001). These antibodies were not directed against major histocompatibility class I antigens, and reacted with both term and first-trimester trophoblast cells. In most cases, sera from which ACLA were adsorbed by cardiolipin- containing liposomes maintained reactivity against cells. In addition, patient Ig fractions immunoprecipitated an approximately 62-kD protein from the trophoblast cell surface, stimulated the release of arachidonic acid and thromboxane A2 by trophoblasts, and inhibited the binding of prourokinase to trophoblast urokinase receptors. These observations show that sera from women with ACLA and a history of fetal loss contain antitrophoblast antibodies. These antibodies may be serologically distinct from ACLA, and may contribute to the pathogenesis of fetal demise.

INTRODUCTION: The Bruton's tyrosine kinase (BTK) inhibitor ibrutinib is a covalent inhibitor of BTK, a member of the B-cell receptor (BCR) signaling pathway and induces objective clinical responses in the majority of CLL patients (Byrd et al., NEJM 2013). Interestingly this drug also inhibits L2-inducible T cell kinase (ITK), an essential enzyme for the development and effector function of Th2 and Th17 cells, and has been shown to shift the balance towards a Th1 response. The purpose of the study was to determine how ibrutinib influences the Th1/Th17/T regulatory cell response, expression of immune checkpoint blockade molecules on T cells and the functional pathogen-specific T cell recovery. METHODS: Here we present data from a clinical trial of ibrutinib versus ibrutinib + rituximab in previously treated patients (NCT02007044). Peripheral blood and serum were collected at baseline, 3 months and 6 months during therapy. Multicolor flow cytometry was used to characterize B cell subsets, T-cell subsets, expression of PD-1, PD-L1 and CTLA-4 and T-cell effector function. For statistical analysis of pre-treatment to on-treatment measurements the paired Student t-test was used. RESULTS: Here we report on the phenotypic and functional recovery of immune subsets in 41 CLL patients treated with ibrutinib (n=17) or ibrutinib + rituximab (n=25). Both PD-1 and PD-L1 were expressed at high levels on CLL cells. Interestingly, by 3 and 6 months, there was a significant decrease in PD-1 expression from a pre-treatment median of 15% to 4% (at 3 months) and 3% (at 6 months; P CONCLUSION: Based on these data we propose that ibrutinib therapy can modulate the T cell response through multiple mechanisms which include (i) direct inhibition of ITK and skewing of the T cell response toward a Th1 profile; (ii) reduction in PD1/PDL1 expression on B and T cells and (iii). suppression of Tregs. It is unclear how ibrutinib influences the PD1/PDL1 interaction and whether this is a direct effect of the drug on essential components of B and T cell-receptor signaling or whether it is an indirect effect related to a reduction in the leukemia burden. Mechanistic studies to understand how ibrutinib modulates the PD1/PL1 axis are currently underway. Figure 1. Ibrutinib therapy is associated with a reduction in PDL1 expression on the surface of CD19+ B cells. Figure 1. Ibrutinib therapy is associated with a reduction in PDL1 expression on the surface of CD19+ B cells. Figure 2. Ibrutinib therapy is associated with a reduction in PD1 expression on the surface of CD3+ T cells. Figure 2. Ibrutinib therapy is associated with a reduction in PD1 expression on the surface of CD3+ T cells. Disclosures Burger: Pharmacyclics LLC, an AbbVie Company: Research Funding. Wierda:Glaxo-Smith-Kline Inc.: Research Funding; Celgene Corp.: Consultancy. Rezvani:Pharmacyclics: Research Funding.

Background: Regulatory CD4 T cells (Treg) are potent to suppress the responses of conventional CD4+ (Tcon) and CD8+ T cells to alloantigens and prevent graft-vs.-host diseases (GVHD). Since the mechanisms for thymic selection of Treg and Tcon are distinct, we hypothesized that the two cells types use distinct TCR repertoires in the response to same alloantigen. We have used high throughput deep sequencing to study the T cell receptor (TCR) Vbeta CDR3 repertoire of CD4 Treg and CD4 Tcon at baseline and again after expansion with alloantigen. Methods: CD4+CD25+CD127- Treg and CD4+CD25-CD127- Tcon were FACS-sorted from healthy donor PBMCs or umblical cord blood mononuclear cells. Aliquots of sorted Treg and Tcon were expanded separately by DC from the same Major Histocompatibility (MHC)-mismatched donor. Treg were cultured with DCs, IL-2, IL-15 and rapamycin, while Tcon were cultured with DCs and IL-2. Genomic DNA from baseline or expanded purified Treg and Tcon was sequenced, at least 10^6 deep, by Adaptive Biotechnologies multiplex kit to detect unique T cell receptor (TCR) Vbeta CDR3 sequences. Data was analyzed by the algorithm established for the Adaptive Biotechnologies software and characterized according to the IMGT (International ImmunoGeneTics information system) nomenclature. Results: The TCR Vbeta CDR3 sequences of baseline natural Treg and Tcon T cell have minimal overlap, indicating that the thymus shapes distinct TCR repertoires in these two cell types. However, the CDR3 length and the frequency of nucleotide deletion or insertion at the Vbeta-Dbeta and Dbeta-Jbeta junction were similar. By employing the "robust regression model", we identified expanded TCR Vbeta CDR3 sequences among both Treg and Tcon after in vitro culture with the same alloantigens. These expanded Treg and Tcon used unique TCR Vbeta CDR3 sequences that are not shared by the other cell type. Expanded Treg and Tcon displayed fewer nucleotide deletions and insertions at the Vbeta-Dbeta and Dbeta-Jbeta junction than at baseline. The frequency of the expanded sequences, insertions and deletions were of the same magnitude in Treg and Tcon suggesting that both undergo similar processes of antigen-driven TCR selection and magnitude of cell expansion in vitro. Conclusion: CD4 Treg and Tcon utilize largely distinct TCR Vbeta CDR3 repertoires at baseline and after expansion against the same alloantigens. Questions remain whether Treg and Tcon recognize the same or distinct peptides from the same antigen, and whether they bind peptide with different avidity. TCR Vbeta CDR3 deep sequencing ought to be used to track single Treg and Tcon after adoptive cell therapy. Figure 1. CDR3 nucleotide sequences of natural Treg and Tcon do not overlap before (Baseline) and after (End) expansion against alloantigen Figure 1. CDR3 nucleotide sequences of natural Treg and Tcon do not overlap before (Baseline) and after (End) expansion against alloantigen Disclosures No relevant conflicts of interest to declare.

Background: Bevacizumab, the first approved anti-angiogenic agent for metastatic colon cancer, is associated with an increased risk of arterial thromboembolism (ATE) and potentially venous thromboembolism (VTE). The molecular mechanisms underlying this association are incompletely understood. The transcription factor Kruppel-like factor 2 (KLF2) is a master regulator of cellular inflammatory responses, and decreased expression of KLF2 is associated with a prothrombotic phenotype. KLF2 is particularly important in regulating inflammatory and prothrombotic genes in endothelial cells and monocytes. Objective: To determine whether bevacizumab therapy was associated with alterations in the expression of KLF2 or other genes potentially associated with thrombosis in peripheral blood mononuclear cells. Methods: We conducted a prospective observational cohort study of consecutive patients initiating bevacizumab therapy for metastatic colorectal, lung, or brain cancer. Patients with prior history of VTE or on anticoagulation were excluded. Blood was collected at baseline (prior to initiation of treatment) and every 4 weeks (± 1 week) for up to 16 weeks while on Bevacizumab therapy. Total RNA was isolated and levels of KLF2 mRNA as well as mRNA encoding 18 other genes potentially associated with thrombosis were measured using a customized qPCR microarray (Qiagen RT2 Profiler). Changes in gene expression levels were standardized through inclusion of housekeeping genes, RT, and PCR controls on each array. Associations with clinical events were assessed by stepwise linear regression. Results: The study population comprised 25 patients who received from one to four courses of bevacizumab. Primary sites of cancer included colorectal adenocarcinoma (N=18), non-small cell lung cancer (N=2), primary brain tumor (N=4) and small bowel adenocarcinoma (N=1). The average patient age was 58 years. Average length of study participation was 71.8 days. PCR analyses demonstrated a significant decrease in KLF2 mRNA after the first cycle of bevacizumab, and levels of KLF2 remained suppressed in patients who remained on study through at least three cycles of therapy ( p = 0.0015, 0.0139, 0.011, respectively). No significant change in the expression of other prothrombotic genes was demonstrated. Levels of KLF2 correlated inversely with levels of D-dimer (p = 0.008). Conclusions: Therapy with Bevacizumab was associated with significant and durable decreases in the expression of KLF2 in peripheral blood mononuclear cells. These findings suggest a potentially new mechanism which may contribute to the prothrombotic phenotype of these patients through loss of thromboprotective gene expression in monocytes. Enhancement of KLF2 expression may provide a potential strategy to explore for circumventing thrombosis in patients treated with bevacizumab. Acknowledgment: This work was supported by funds from the Scott Hamilton Cares Initiative, the Stephen Hardis Chair in Oncology Research, and the V Foundation. Disclosures No relevant conflicts of interest to declare.

BACKGROUND: Glioblastoma (GBM) is most common primary malignant brain tumor, and has a median survival of 15-18 months. Dovitinib, an oral multi-tyrosine kinase inhibitor of vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF), and platelet derived growth factor (PDGF) is currently under study in a phase II trial for GBM at the Cleveland Clinic. Dovitinib is administrated 5 days on and 2 days off every 4 weeks until progressive disease (PD) or intolerable toxicity are observed. Extracellular vesicles (EV) are submicron particles that express or contain cellular proteins and nucleic acids and are released from a variety of non-malignant cells (e.g. endothelial cells, platelets, leucocytes) and malignant cells. In some settings, EV may serve as biomarkers of inflammation, thrombosis and tumor spread/burden. OBJECTIVE: The aim of our study was to characterize levels of circulating EV and their relation to disease course in patients with GBM enrolled in the Dovitinib study (with or without prior treatments with anti-angiogenic agents). We also examined the association between EV levels and the development of venous thromboembolism (VTE). METHODS: Patients previously treated with anti-angiogenic therapy (Group 1, n=14) or without prior anti-angiogenic treatment (Group 2, n=14) were examined separately. EV were measured at study enrollment (pre-treatment), at the end of cycle 1 (day 28), and at PD. EV were isolated from citrated whole blood by differential centrifugation and incubated with fluorochrome-conjugated monoclonal antibodies to CD144-PE (endothelial cells), CD41-PECy4 (platelets), CD14-PE (monocytes) and CD142 (tissue factor, Alexa Fuor 647), then analyzed by flow cytometry. Depending on sample size, the Student t-test or Wilcoxon test was used to compare EV levels (due to the small sample size and skewed distribution of EV levels). P RESULTS: Three patients from group 1 and 6 patients from group 2 were not included in the analysis secondary to lack of an EV sample, withdrawal of consent or complications leading to early drug discontinuation. Of theremaining 11 patients in Group 1, 3 had PD and 8 had stable disease (SD) at the end of cycle 1. Of the 8 patients in group 2 available for analyses after cycle 1, 2 had PD and 6 had SD (one of these developed VTE but continued on the study). In the pretreatment sample of patients from group 1, patients who developed PD had significantly higher levels of CD14+ EV (89977±12121 vs. 42237±27651, p =0.048) and CD142+ EV (68701±9010 vs. 9695±12462, p=0.048) compared to those with SD. However, there was no statistically significant difference in EV levels (all sub-populations) from pre-treatment to the end of cycle 1 in patents with either PD or SD. EV levels did not correlate with peripheral blood counts. Due to the small number of patients in group 2 with progressive disease, we were unable to assess the correlation with EV. Six (2 in group 1, 4 in group 2) of the 27 patients for which pre-treatment EV were available developed VTE during the study. The EV levels were not significantly different between patients who developed VTE compared to those who did not both at pretreatment and at the day 28 evaluation. However, most patients who developed VTE demonstrated profound increases in EV before or in association with their thrombotic event. CONCLUSIONS: In patients with GBM receiving Dovitinib without prior exposure to anti-angiogenic therapy, elevated pre-treatment levels of CD14+ and CD142+ EV were associated with progressive disease, suggesting their potential role as a predictor of poor response to Dovitinib. Due to the relatively small sample size, no significant differences were observed between patients that developed VTE and those that did not, either pretreatment or at the Day 28 evaluation; however, these studies are ongoing. In the majority of patients with VTE, EV levels increased substantially before or in association with VTE development. Acknowledgment: This work was supported by a grant from the Scott Hamilton Cares Initiative Disclosures No relevant conflicts of interest to declare.

Abstract 3290 Background: Patients with high risk AML, defined as those with age > 60 years or multiple medical co-morbidities determined by Charleston comorbidity index (CCI) carry a poor prognosis and inferior outcomes after 7+3 induction chemotherapy. CR rates tend to range from 6–51% and induction death rates between 9–48%. We present here a single institution experience of high risk AML patients treated with an induction regimen consisting of high dose mitoxantrone and cytarabine (HiDAC/MITO). Methods: We performed a retrospective analysis of all patients with AML who received HiDAC/MITO induction from January 2009- January 2010 at our institution. Patients with age ≥60 or age Results: 20 AML had received HiDAC/MITO for remission induction. The median age was 66.5 years (range 47 to 78), those with age ≥ 70 was 8 (40%). CCI was ≥ 5 in 18 (90%) patients. Other high risk features included high risk cytogenetics in 8 (40%) and non-denovo AML (AML with AHD, t-AML or relapsed AML) in 11 (55%). Overall CR rate was 17 (85%, CI: 61%-96%) and 3 (15%) patients had refractory disease. There was no treatment related mortality. Median time to neutrophil recovery (>1000/ul) was 27 (range 19 to 37) days and median time to platelet recovery (>100,000/ul) was 28 days (range 23 to 44) days. Patients with non–denovo AML were more likely to be refractory to treatment or relapse after day 30. Median follow up of the entire cohort is 288 (range 29 to 530) days. 3 month and 6 month overall survival (OS) was 94.7% and 73.3% and progression free survival (PFS) 93.8% and 87.5%, respectively. The median OS was 410 days (CI: 243-*); (denovo 410 vs. others 381 days). Median PFS is 524 days (CI: 381-*); (denovo *not reached vs. others 381 days). 11(55%) patients were able to proceed to autologous (4) or allogeneic (7) stem cell transplantation (SCT) after receiving HiDAC/MITO. The time to transplant ranged from 44 to 195 days. Median OS of the patients who underwent SCT is 524 days versus 269 days for the non transplant group (p =0.0038). The HiDAC/MITO induction regimen was well tolerated. Cardiac toxicity defined by symptomatic CHF was noted in 6/20 patients. Of the six patients 2 had prior cardiac history and 1 had prior anthracycline exposure and 1 had both anthracycline exposure and cardiac history. Cardiac toxicity was delayed and identified by echo at a median of 90 range (42 to139) days after induction chemotherapy. None of these patients died from cardiac toxicity. Conclusions: In this high risk AML population, HiDAC/MITO induction was well tolerated and demonstrated an overall response rate of 85% and no induction deaths, allowing a substantial number (55%) of patients to proceed to SCT. Contrary to our expectations advanced age or multiple medical co-morbidities did not affect CR rate or survival, thus high lighting the utility of this regimen for high risk newly diagnosed elderly patients with AML. Disclosures: No relevant conflicts of interest to declare.

The Hedgehog (Hh) signaling pathway is critical for normal development and dictates the self-renewal, proliferation and differentiation of normal stem cells and progenitors. Aberrant reactivation of Hh signaling has been described in a wide variety of human cancers and its role in normal stem cells suggest that pathway dysregulation contributes to oncogenesis and influences the cell fate decisions in cancer stem cells (CSC). Like their normal counterparts, CSC appear to undergo self-renewal as well as give rise to differentiated progeny, and these properties implicate that CSC are responsible for continual tumor cell production that underlies the initiation, maintenance and progression of clinical disease. Myeloid leukemias have long served as the model system for human CSC, but the cellular processes responsible for regulating these rare biologically distinct cell populations have remained unclear. We hypothesized that Hh pathway activation contributes to the pathogenesis of acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) and studied Hh signaling in these settings. Using both RT-PCR for pathway components and a Gli1 reporter assay, we have found that Hh signaling is active in several human AML derived cell lines (Kasumi-1, KG1, KG1a) and primary AML and MDS samples. Approximately 80% (19/24) of primary AML samples tested express the downstream effectors GLI1 or GLI2 indicative of active Hh signaling. Furthermore, inhibition of Hh signaling with the naturally derived SMOOTHENED antagonist cyclopamine reduces the clonogenic growth of KG1 cells implicating the pathway in self-renewal. In contrast, cyclopamine failed to affect colony growth in the HL-60 cell line that lacks expression of Hh pathway signaling components, confirming that the effect of Hh inhibition is specific. In addition, the ectopic expression of Gli1 in KG1 cells partially rescued the effect of cyclopamine on colony formation further demonstrating the specific nature of this compound. We also studied normal CD34+ bone marrow cells and found that they expressed components of Hh pathway by RT-PCR. However, in contrast to KG1 cells, cyclopamine had little effect on the recovery of either normal hematopoietic progenitors or stem cells in an in vitro long-term culture assay. Therefore, it appears that Hh inhibition may preferentially inhibit myeloid leukemias. We further studied the role of Hh pathway activation on normal hematopoiesis and developed a transgenic mouse model in which SMOOTHENED is conditionally over-expressed in the myeloid lineage via Cre recombinase activity regulated by the Lysozyme promoter. Analysis of these mice demonstrated only subtle changes in peripheral blood counts, but further analysis of cells expressing the transgene revealed a significant reduction in the number of mature myeloid cells. This was confirmed by analyzing blood cells for the granulocyte marker Gr1 and pan-myeloid marker Mac1, both of which were significantly reduced in the SMOOTHENED over-expressing cells. These defects are reminiscent of MDS and further suggest that the Hh signaling pathway plays a role in normal hematopoiesis. Therefore, aberrant Hh pathway activation is a feature of myeloid leukemias and inhibitors such as cyclopamine may have a therapeutic role in the treatment of AML and MDS.

Central Nervous System (CNS) relapse of aggressive non-Hodgkin’s lymphoma is a devastating clinic event. Significant controversy exists however, as to whether CNS prophylaxis affects the risk of CNS relapse and if so, which patients should be offered CNS prophylaxis. To address these questions we analyzed the twenty year follow up data of SWOG 8516, a prospective randomized trial comparing CHOP (n=225 patients), MACOP-B (n= 218), ProMACE-cytaBOM (n=233), and m-BACOD (n=233) for patients with newly diagnosed intermediate or high grade non-Hodgkin’s lymphoma. Patients that received ProMACE-cytaBOM who were bone marrow (BM) positive at baseline and BM negative after 4 cycles of treatment, received prophylactic cranial irradiation (XRT). Patients that received MACOP-B who were BM positive at baseline received intrathecal methotrexate and Ara-C (IT chemo) twice weekly for six doses. In contrast, no patient that received CHOP or m-BACOD received any CNS prophylaxis. Patient clinical characteristics were well balanced between these four arms. Results: Of the 515 patients who relapsed with documented lymphoma, 348 (68%) relapsed only in nodal areas; 167 patients (32%) had extranodal relapse, the most common site of which was the CNS (n=34, 3.8%). Among the 34 patients with CNS relapse, 31 (91%) presented with isolated CNS relapse, 2 (6%) presented with CNS and nodal relapse and 1 (3%) presented with CNS, nodal and other extranodal relapse. 97% of all CNS relapses occurred by year 2, compared to only 73% of non-CNS relapses (p= .002). Survival after CNS relapse was significantly worse compared to that of patients with non-CNS relapse (2 year estimate 9% vs. 30%, respectively; p= .002). Using logistic regression analysis, significant differences in the incidence of CNS relapse were evident in patients with > 1 extranodal sites (5.7% vs. 2.7%, p=.03) and in patients with higher International Prognostic Index (IPI) scores (6.3% in high IPI, 4.8% in high intermediate, 3.4% in low-intermediate and 1.6% in low IPI, p=.02). There was no evidence that patients receiving any type of CNS prophylaxis (i.e., either cranial XRT or IT chemo) had different rates of CNS relapse compared to patients who did not receive any CNS prophylaxis (5.2% vs. 3.6%, p=.40). In separate analyses, there was no evidence that patients receiving cranial XRT had different rates of CNS relapse than other patients (p=.65) and no evidence that patients receiving IT chemo had different rates of CNS relapse than other patients (p=.48). Finally, if the analysis is restricted to the 238 patients that were BM positive at diagnosis, 96 patients had CNS prophylaxis and 142 did not. The rate of CNS relapse in those patients that received CNS prophylaxis was no different than that seen in the patients that did not receive CNS prophylaxis (5.2% vs. 4.2%, p=.72). Conclusion: CNS relapse almost always occurs within 2 years of initial treatment, is usually isolated, and portends a poor prognosis. Patients with > 1 extranodal sites or those having a high IPI score have a higher incidence of CNS relapse. There is no evidence to suggest however, that CNS prophylaxis with either cranial radiation or intrathecal MTX/Ara-C decreases the risk of CNS relapse.

Limbic encephalitis is a rare complication reported in patients receiving an allogeneic hematopoietic stem cell transplant (HCT). It is characterized by a syndrome of short-term memory deficit, altered mental status, seizures and coma accompanied by characteristic MR imaging findings of abnormal high signal intensity on flair and T2 weighted images in the mesial temporal lobes. The incidence and risk factors for the development of of limbic encephalitis are not well studied in patients undergoing HCT. In this study we retrospectively reviewed the medical records of 399 allogeneic HCT recipients transplanted at our institution from June 1995 to June 2006 to determine the incidence of limbic encephalitis among these patients. Forty-nine of these patients received umbilical cord blood transplants (UCBT). We identified 66/399 patients who underwent MRI of the brain to evaluate neurologic symptoms; 28 patients had abnormal MRIs. The MRIs were then examined to determine if the findings were consistent with limbic encephalitis. Seven of the twenty-eight MRIs showed changes compatible with limbic encephalitis. Five of the seven patients with limbic encephalitis had undergone UCBT and 2 had received matched unrelated donor HCT. The median age of the patients was 44 years (range 32–54), 5 were female. Diseases treated were AML (3), Hodgkins disease (2), Burkitt’s lymphoma(1) and Multiple Myeloma (1). Four patients had failed prior autologous transplant. Six patients had persistent disease at the time of transplant. Total body irradiation based conditioning was given in 4 of 7 patients. All patients had received anti-thymocyte globulin (ATG) as a part of their conditioning. GVHD prophylaxis was with methylprednisolone + tacrolimus (5), mycophenolate mofetil + tacrolimus(1) and methotrexate + tacrolimus(1). The 2 patients that did not have methylprednisolone as part of GVHD prophylaxis received it as treatment for acute GVHD. All of the patients received prophylactic high dose acyclovir or valacyclovir. The median time from transplant to diagnosis was 39 days, (range 28–148). Symptoms included: short term memory deficit/confusion (6), seizures (5) and coma(4). SIADH was seen in 4 patients with a mean serum sodium of 125 mmol/L. Four patients had elevated CSF protein (median for the entire cohort 48 mg/dL; range 20–105) and glucose (82 mg/dL; 68–97). Three patients had an elevated CSF WBC count (4 cells/μL; 0–34) with lymphocyte pleocytosis (81%; 0–100). CSF bacterial and viral cultures were negative on all of the patients. Five patients had HHV-6 PCR performed; 4 of these had HHV-6 DNA detected in the CSF. Six of the seven patients recieved second line antiviral therapy with either ganciclovir or foscarnet. Three of the patients who had therapy started at the onset of symptoms had clinical improvement. Short term memory deficit persisted in all the patients. Six of the seven patients have died with the median survival of 39 days (range 4–110) from diagnosis of limbic encephalitis. One patient is alive 11 months after diagnosis with severe neurologic dysfunction. We describe a rare, but devastating complication of allogeneic transplantation affecting approximately 2% of our allografts and 10% of UCBT recipients. Heavily pretreated patients, and those receiving ATG and corticosteroids appear to have a higher risk of limbic encephalitis.

Introduction: Pts with Chronic Lymphocytic Leukemia (CLL) are reported to have quantitative and qualitative T and NK cell dysfunction. While NK cells act through non-specific killing, T-cells are more specific. The 2 types of T-lymphocytes, CD4+ (Th; helper) and CD8+ (Ts; cytotolytic/suppressor) are subcategorized based on cytokine secretion profile upon activation. Release of different cytokines from these immune cells modulates the host response. T1 cells (Th1, Ts1) secrete IL-2 and interferon-g which initiate the Th1 response- mainly CD4+ activation along with B and T cells, leading to proliferation and differentiation of these cells. T2 (Th2, Ts2) cells initiate the Th2 response (release of TNF-a, IL-10) resulting in direct lysis of the target cell by production of cytokines such as IL-4, IL-5 and IL-10. Hypothesis: To decipher this antitumor mechanism of L in CLL pts we investigated its effect on the efferent arm of immune response by evaluating the T cell population and the afferent response by change in expression of co-stimulatory molecules on B-CLL cells and cytokine profile in these pts treated on a phase II clinical study. Methods: CLL pts treated with L were evaluated for absolute number of T (CD4+, CD8+) and NK (CD56+) cells by flow cytometry on day before (day0) start and on Day 8 of treatment with L. Peripheral blood was collected and ficolled to obtain enriched mononuclear cells. The serum was used to study the cytokines. Activation status was determined by co-expression of CD45+. Serum cytokine profile was measured by Flow cytometry using the Luminex system. B-CLL surface co-stimulatory molecules were detected by flow cytometry and analyzed by FACS. These responses were correlated with the tumor flare (TF) reaction that the patients developed during the first week of treatment with L. Results: Eighteen out of 45 pts have so far been evaluated for immunomodulatory activity of L. There were 2 complete responders (CRs) and 6 partial responders (PRs); while 4 had stable disease (SD), 4 were clinically unevaluable and 2 were too early for response in this group. Mean baseline (bl) NK cell count pretreatment was 251 (range 31–1510) vs. post treatment was 193 (range 6–13,482). Six out of 18 patients showed an increase, ranging from 20 −199% in the absolute NK (CD16+/CD56+/CD45+). While there was no appreciable change in CD4+ numbers there was a general trend in increase of CD8+ cells. No change in monocyte population was noted. Concurrent increase in the expression of co-stimulatory molecules such as CD95 and CD80 was noted. This response in co-stimulation was confirmed by in vitro experiments done on isolated B-CLL cells (n=4)treated with L. An increase in Th-2 cytokines such as IL-4, IL-5, IL-6 and IL-10 was noted in all eight responders, while VEGF levels were decreased in 6/18 patients. 99% of patients had a TF and the grade of TF correlated with the changes in T cells and cytokine profile. Conclusion: It appears that in vivo L is able to orchestrate an anti-tumor response in CLL by modulating the NK cells, changing the cytokine profile and up-regulating co-stimulatory molecules. This change in the immune effector cell repertoire and the Th2 skewing may explain the initial flare reaction noted in these L treated pts. Data from these correlative studies is being evaluated in the context of the phase II clinical trial to be reported at the 48th ASH annual meeting.

The interaction between T-cell costimulator, ICOS, and its ligand (ICOSL) expressed on B cells plays an important role in intercellular cognate interactions leading to lymphocyte activation. We found that a subset of chronic lymphocytic leukemia (CLL) B cells express ICOS and that ligation of this receptor induced leukemia-cell activation of the PI3K/AKT survival pathway. Moreover, expression of ICOS was associated with lower-level expression of ICOSL, apparently because of receptor-ligand down-modulation. We hypothesized that co-expression of ICOS and ICOSL on CLL B cells may enhance leukemia-cell stimulation and be conducive to more aggressive clinical disease. We evaluated the CLL cells from 208 patients for expression of ICOS and ICOSL. The expression of ICOS and ICOSL were characterized by the percentage of CD19+/CD3- cells that expressed ICOS and/or ICOSL using fluorescence thresholds established via parallel analyses on the same cell populations stained with isotype control mAbs. In addition, we evaluated these CLL cells for IgVH somatic mutations and for expression of ZAP-70. We then examined the relationship between expression of ICOS and ICOSL, the IgVH mutational status, ZAP-70, and the time from diagnosis to initial therapy, as per NCI working group criteria. The median proportion of cells that expressed ICOS was 3.5% (ranging from 0.1% to 99.6%), and median proportion of cells that expressed ICOSL was 62% (ranging from 0.9% to 97.4%). Ninety cases (44%) were found to use mutated IgVH genes, whereas 118 (56%) were found to use unmutated IgVH. Unmutated cases had significantly higher expression of ICOS than mutated cases (p=0.0014, Wilcoxon test), although there was no significant difference in the level of ICOSL (p=0.47). 95 cases expressed ZAP-70, and 113 cases did not. There was a significant association between ZAP-70 expression and ICOS, p=0.0048 by the Wilcoxon test, with the cases expressing ZAP-70 having higher than expected levels of ICOS. The association with ICOSL was also significant, p=0.04, with higher than expected levels of ICOSL found in the cases that did not express ZAP-70. Next we used recursive partitioning to identify the optimal threshold for distinguishing levels of ICOS and ICOSL expression that best could discriminate the time from diagnosis to initial therapy into two groups. This revealed that 35% for ICOS and 47.7% for ICOSL were the best-cut points. Median time to treatment was 5.1 years among the patients with high ICOS expression (n=25), and 6.7 years among those with low expression. The log rank p-value associated with this difference was 0.03. Median time to treatment was 4.2 years with low ICOSL (n=76), and 7.8 years among those with high expression. The log rank p-value associated with this difference was 0.0004. The stepwise model identified as significant risk factors associated with the need for early treatment: (1) expression of ZAP-70 (hazard ratio 4.53, Wald p-value < 0.0001); (2) expression of unmutated IgVH with >=98% germline sequence homology (hazard ratio 2.56, p = 0.0014), and (3) low-level expression of ICOSL (hazard ratio 1.62, p = 0.02). Because the analysis for ICOS and ICOSL by flow cytometry is relatively straightforward, low-level expression of ICOSL may be a useful surrogate marker for high-risk disease.

Patients (pts) with AML undergoing curative intent chemotherapy (CTX) are highly susceptible for septicemia due to profound and prolonged neutropenia, in conjunction with damage to the gastrointestinal mucosa, and the use of vascular access devices. Despite aggressive empiric antimicrobial therapy, and the use of prophylactic antibiotics, septicemia remains a major cause of morbidity in this group. Over the past decade the spectrum of pathogens that cause infections in neutropenic pts has changed, with gram+ve microorganisms replacing gram-ve bacilli. In order to identify the incidence, and cause of septicemia and the resistance pattern of bacteria in AML pts treated with curative intent CTX in the Leukemia/BMT program of BC, all positive blood cultures collected between Feb, 1999 and Feb, 2004 were evaluated retrospectively. 623 separate CTX cycles administered to 295 patients were reviewed (m=157, f=138). Median age at diagnosis was 52y (17–76). CTX regimen were classified as 7+3 or similar (conventional dose cytarabine + anthracycline) (group 1), or as high dose cytarabine containing regimen (with or without anthracycline) (group 2). 328 cycles were given as induction intend-CTX, while 295 where given as consolidation. 426 cycles were spend entirely as inpatients (IP) (from administration of CTX to ANC>0.5); 40 cycles were spend as early discharge (ED) (from CTX administration to discharge as outpatient prior to d+15 or ANC>0.5); 157 cycles were spend entirely as outpatients (OP). 126 episodes of septicemia were recorded in 623 cycles of CTX (20%), of which 21 episodes yielded more than 1 identified organism. Percent occurrence of septicemia stratified by CTX protocol indicated 17% (57/332) for group 1, and 24% (69/291) for group 2 (p Septicemia remains a frequent complication in high-risk neutropenic AML pts. Supported by our results, we have constructed an infectious risk stratification scheme pertaining to the CTX treatment of AML in IP, OP and ED settings. Furthermore, we report that while infections in this population constitutes a significant source of morbidity, mortality directly attibutable to septicemia is rare with prompt intervention.

Pro-inflammatory monocyte-derived cytokines (monokines) such as interleukin (IL)-12, IL-15 and IL-18 work in concert to quickly induce IFN- γ in natural killer (NK) cells, while the anti-inflammatory monokine transforming growth factor beta 1 (TGF-β1) can suppress NK cell IFN- γ production. Here, we provide new molecular evidence as to how these monokines carry out their opposing functions. In the human NK-92 cell line we show that Smad proteins, mediators of TGF- β1 signaling, directly inhibit IFN- γ production. Using chromatin immunoprecipitation (ChIP), we found that both Smad2 and Smad4 associate with the IFN- γ promoter in the proximal region but not in other regions, such as exon 4. Luciferase reporter assays in 293T cells indicated that Smad2 alone, which does not possess a DNA binding domain, had little effect on the inhibition IFN-γ of promoter activity. In contrast, Smad3 and Smad4, both of which have DNA binding activity, dramatically suppressed IFN-γ promoter activity. Co-transfection of Smad2 and Smad4, or Smad3 and Smad4, in 293T cells resulted in synergistic repression. Interestingly, the Smad proteins also inhibited the transactivation of the IFN-γ promoter by T-BET, a recently discovered transcription factor controlling IFN- γ production in mouse T cells and NK cells. This was consistent with our data showing that TGF-β1 inhibited T-BET expression in human NK cells. Using IFN-γ promoter deletion constructs, we mapped the region responsible for Smad-mediated inhibition to the -204 proximal promoter, where Smad binding elements are enriched. In parallel, using micro-array analysis, western-blotting and semi-quantitative PCR in human primary NK cells and the NK-92 cell line, we demonstrated that the combination of the pro-inflammatory monkines IL-12 with IL-15 or IL-18 was able to antagonize multiple components of TGF- β1 signaling, including: 1) inhibition of the expression and the phosphorylation of Smad2; 2) suppression of Smad3 expression; and 3) down-regulation of the expression of the TGF- β receptor II subunit. Moreover, although these monokines had no effect on Smad4 expression, using ChIP assay we observed that Smad4 dissociated from the proximal IFN-γ promoter upon co-stimulation of NK cells with IL-12 and IL-18. In summary, our data show that TGF- β1 suppresses IFN- γ production via the direct binding of Smad proteins to the proximal IFN-γ promoter and via inhibition of T-BET, while the pro-inflammatory cytokines serve to inhibit or down modulate TGF- μ signaling components and reverse Smad binding to the proximal IFN-γ promoter. Thus, we suggest that NK cell production of IFN-γ is controlled by the integration of anti-inflammatory and pro-inflammatory monokine signals, leading to reciprocal antagonism.

Objectives: Acute myeloid leukemia (AML) with monosomy 7 or deletion of the long arm of chromosome 7 (-7/7q-) is a leukemia with a poor outcome. Microarray analysis resulted in a gene expression profile specific for this subset of AML and showed SKI, a nuclear Co-repressor protein, being strongly upregulated. SKI is described to be a transforming gene, which inhibits TGF-beta-, SMAD- and retinoic acid (RA)-signaling. Materials, methods and patients: Gene expression profiling was performed using cDNA microarrays (detecting 4608 genes). Gene expression pattern of 20 AML (9 pts. with -7/7q- and 11 pts. with normal karyotype) patients was compared to 23 healthy bone marrow donors. Expression data of interestings genes were confirmed by real-time PCR in samples of 111 AML patients. In cell culture experiments HL60 and U937 leukemic cell lines were used to investigate the influence of the identified oncogene Ski on differentiating properties. In a reporter gene assay retinoic acid response element (RARE) was transfected in QT6 cells. U937 were retrovirally transfected with Ski wild type and Ski mutant. Results: Microarray analysis in AML (-7/7q-), revealed SKI being upregulated in AML. The group of patients with a high SKI expression level had a poorer overall survival when compared to the low expressers (p = 0.0279). Real-time PCR analysis comparing 111 AML samples to healthy CD34+ haematopoietic stem cells revealed, that SKI levels were highest in AML with -7/7q-, moderately elevated in most other karyotypes except AML with translocations involving retinoic acid receptor alpha (RARalpha). This suggests that Ski might interfere with RARalpha function in leukemia. Using reporter gene assays we could show that expression of wild-type, but not mutated SKI blocked RA induced differentiation. SKI was down regulated in HL60 leukemia cell line when differentiation was induced by ATRA and or valproic acid. Conclusion: Upregulation of SKI has prognostic significance in AML. Inhibition of RA signaling may be more often implicated in pathogenesis of AML than anticipated so far. The Co-repressor of histone deacetylases (HDACs) Ski might therefore become a molecular target for the induction of differentiation in the treatment of AML.

Conventional amphotericin B (C-AMB) is one of the most frequently used antifungal first line agents in patients with hematological malignancies diagnosed to have invasive fungal infections or persistent neutropenic fever despite of its well documented adverse effect profile. To evaluate the influence of pretreatment with C-AMB on the outcome of allogeneic HSCT, a retrospective analysis including 270 consecutive allogeneic HSCT between 1/2002 and 12/2003 at out institution was performed. Median patient age was 42 years, female to male ratio 46/51 and CMV positivity 83 %. Diagnosis was AML (n=90), ALL (n=72), CML (n=45), MDS (n=21), NHL (n=14), MM (n=8) or micellaneous (n=20). Sixty percent (164/270) of patients were in early, the remainders in advanced disease stages. Conditioning therapy was TBI or chemotherapy in 237 and 33 patients, respectively. Donors were identical siblings (n=110), non-identical family members (n=3), matched unrelated (n=116) or mismatched unrelated (n=41). Pretreatment with antimycotics was reported in 74 cases (C-AMB, n=73; voriconazole, n=1) for FUO (n=23), not specified fungal pneumonia (n=23), pneumonia due to aspergillus sp.(n=18), hepatosplenic candidiasis (n=6), fungemia (n=3), and IFI by scedosporium prolificans (n=1). A Cox regression model was used to test C-AMB pretreatment as an independent risk factor on TRM. Age in decades, disease stage, preceding neutropenia, renal impairment (creatinine > 1mg/dl) was also included. Only previous treatment with C-AMB (p=0.001; RR 3), and advanced disease stage (p=0.008; RR2,5) remaind statistically significant. Subsequently, to identify patients at increased risk, we defined following strata: 1. patients without previous neutropenia; 2. patients with previous neutropenia; 3. patients with C-AMB pretreatment, creatinin < 1,0 mg%; 4. patients with C-AMB pretreatment and creatinin > 1,0mg%; 5. patients with renal impairment for other causes; 6. patients with renal impairment or C-AMB pretreatment and antimycotic therapy during transplant procedure. TRM was 3.2%, 5.43%, 19.57%, 26.67%, 29.63% and 24.14%, respectively. We conclude that patients being pretreated with conventional C-AMB have an increased independent risk for TRM (RR 4; p