ALBERT

Description: Abstract 3165 Introduction: Patients with cancer have a 7- to 10-fold overall increased risk of developing venous thromboembolism (VTE). Some tumors shed small membrane vesicles called microparticles (MPs) containing tissue factor (TF) and membrane phospholipids (PL) leading to procoagulant activity (PCA). TF, the most potent initiator of coagulation cascade, plays a critical role in hemostasis. The circulation of active TF associated with MPs has been considered as a risk factor for VTE in cancer. Aims: The primary objectives of this study were to characterize structure, size and PCA of tumor-cell derived MPs released by breast cancer cells MDA-MB231 (MDA) using thrombin generation (TGT), flow cytometry (FCM) and transmission electron microscopy (TEM). The secondary objectives were to determine the PCA of MDA and MPs derived from MDA in order to study the effect of stirring on MPs production and PCA. The study of the effects of the size of MP on PCA and the contribution of TF and PL to the PCA will also be performed. Methods: In vitro generation of MPs: Cultured MDA-MB-231 breast cancer cells were adjusted to the desired concentrations (600000 cells/ml) in PBS. Cells suspensions were incubated for 45 min at 37°C under stirring condition or without any stirring. Samples were then centrifuged at 4500g for 15 min and the cell supernatants (Sup) were used for EM, FCM and TGT. Alternatively, MP fractions were filtered through 0.1, 0.22, 0.45 or 0.65μm membranes (Ultrafree-MC) and subjected to activity assays. PCA: TGT was performed with Calibrated Automated Thrombogram (CAT). Cells suspensions or in vitro-generated MP fractions from 500000 cells were used as the source of TF and PL and added to normal pool plasma (NPP). Counting and expression of TF and MUC-1: Quantification of MPs and expression of TF (CD142) and MUC-1 (CD227) on MDA and MDA Sup were studied by flow FCM. The size of PMPs was defined using a blend of monodisperse fluorescent beads (Megamix). Tumor MPs sizing: A 5 μL sample of cells or MP fraction derived from 500000 cells diluted 100X was gently put onto a non-BSA precoated formvar grid and allowed to settle undisturbed for 48h before analysis under a TECNAI 10 TEM. Results: Effects of MDA-MB-231 and MDA-MB-231 Sup on PCA MDA and MDA Sup significantly increased the active thrombin in comparison to human NPP alone. Indeed, for example the lagtime was reduced by 11,7-fold and 7,2-fold, respectively. The difference between MDA and its Sup were highly significant (p

Description: Abstract 2942 Introduction: In 2004, a variant of del(20q) was described in MDS, i.e. an isochromosome of the long arm of chromosome 20 with interstitial loss of material [ider(20q)]. At the present time, 40 cases of ider(20q) have been reported in hematological malignancies, i.e. twenty-nine in MDS. The reported frequency of ider(20q) compared to del(20q) in MDS and AML, is 0.49% versus 2.49% and 0.26% versus 2.63%, respectively. No specific morphological anomaly was reported for MDS with del(20q) or ider(20q). Del(20q) preferentially involved the erythroid and megakaryocytic precursors and was associated with elliptocytosis. The prognostic significance of del(20q) is generally considered as good with some controversy. Until now, the prognostic significance of ider(20q) remains debated. Aims: The discovery of highly dysplastic neutrophils with neutrophil erythrophagocytosis in a patient with MDS and ider(20q) prompted us to analyze and compare the morphology of blood (PB) and bone marrow (BM) cells in MDS patients with del(20q) or ider(20q). The second aim was to compare clinical features of patients with MDS associated with del(20q) and ider(20q). Materials and methods: 34 patients with MDS displaying del(20q) and/or ider(20q) were included in this retrospective multicentric study. May-Grunwald-Giemsa stained PB and BM smears were analyzed for dysplasia in the erythroid, myeloid and megakaryocytic lineages. All cases were revised independently by at least 2 cytologists from 2 or 3 centers. Cytogenetic analysis was performed at diagnosis on short term BM cultures using conventional techniques, and karyotypes were described according to the International System for Human Cytogenetic Nomenclature (ISCN 2009). When possible (n=20), metaphase and interphase FISH studies were performed on the same suspension as used for conventional cytogenetic analysis to confirm the presence of ider(20q) when cytogenetically suspected, to look for subclones with del(20q), and to establish the percentage of nuclei with ider(20q) and del(20q). All patients were followed from the date of clinical diagnosis to transformation in acute leukemia or death. The presence of autoimmune manifestations and the International Prognostic Scoring System (IPSS) were also registered. Results: For the first time, morphological abnormalities on PB and BM were found sensitive and specific for this recently described cytogenetic entity. Indeed, hypogranulated and vacuolized neutrophils and neutrophil erythrophagocytosis (Figure 1) were present on PB and BM in MDS with ider(20q), with a sensitivity ranging from 65% to 75%. Moreover, the sensitivity and specificity of a morphological score based on the assessment of 9 morphological criteria reached 70% and 90.5%, respectively, in MDS patients with isolated ider(20q). This score included hypogranular and vacuolized neutrophils (in association with non dysplastic neutrophils), hypogranular and vacuolized eosinophils (with sometimes difficult differentiation between neutrophils and eosinophils), neutrophil erythrophagocytosis, elliptocytosis and thrombocytophagocytosis on PB. Hypogranular and vacuolized neutrophils and eosinophils, erythrophagocytosis by neutrophils and deeply lobulated and hyperlobulated megakaryocytes (stag-horn-like) were observed in BM. In the present study, the outcome of patients with del(20q) and with ider(20q) was not statistically different. Indeed, the number of progressions and deaths was slightly higher in the ider(20q) group. However, the frequency of autoimmune manifestations (AIMs) in del(20q) and ider(20q) group were 9.5% (2/21) and 8% (1/13), respectively (p〉0.05). The mean IPSS was 0.5 in both groups. The OS (p=0.3744) and PFS (p=0.2497) did not differ statistically in MDS with del(20q) and ider(20q). All deaths were related to the MDS except for one patient with ider(20q) whose dead was from unknown origin. The median OS and PFS were 68 and 65 months, respectively, in the ider(20q) group whereas they were not reached in the del(20q) group. Conclusion: We proposed a morphological score based mainly on hypogranulated and vacuolized neutrophils and neutrophil erythrophagocytosis on PB and BM which can be used to predict ider(20q) in MDS patients. The prognosis seems not to differ between MDS with ider(20q) and del(20q). Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 1169 INTRODUCTION: New oral anticoagulants (NOACs) have been recently approved by the European Medicine Agency (EMA) and the Food and Drug Administration (FDA) for several indications. New oral anticoagulants include anti-IIa agent (dabigatran etexilate) and anti-Xa agents (rivaroxaban, apixaban and edoxaban). NOACs do not require monitoring nor frequent dose adjustment. However, searching for the optimal dose in the individual patient may be useful in some situations. Recent studies have shown that aPTT, HTI and ECT could be used to monitor dabigatran whereas PT and anti-Xa chromogenic assays could be used to monitor anti-Xa agents, while standardizing the time between the last intake of rivaroxaban and the sampling is mandatory. These tests only measure the initiation phase of the coagulation cascade. Thrombin generation assay (TGA) which measures the entire thrombin generation process could be used to better discriminate the inhibitory profile of the NOACs in patients. OBJECTIVES: To study the impact of dabigatran and rivaroxaban in patients by thrombin generation assay compared to other traditional coagulometric and chromogenic assays. MATERIALS AND METHODS: Five patients under dabigatran and 5 patients under rivaroxaban for atrial fibrillation were included in this study. Blood samples were taken at different intervals: at Ctrough, 2h and 3h after drug administration. The following tests were performed at each timepoint Rivaroxaban: Prothrombin Time (PT) using the following reagents: Triniclot PT Excel S.. and Innovin.., Thrombin Generation Assay (TGA) using PPP-Reagent and PPP-Reagent High, Biophen Direct Factor-Xa Inhibitor.. (DiXaI). Dabigatran: Activated Partial Thromboplastin Time (aPTT) using CK-Prest.. and Synthasil.., Hemoclot Thrombin Inhibitor.. (HTI) and Thrombin Generation Assay (TGA) using PPP-Reagent and PPP-Reagent Low. All the tests were calibrated by spiking rivaroxaban or dabigatran at increasing concentrations in pooled citrated normal human platelet poor plasma (PPP). RESULTS AND DISCUSSIONS: In vitro Rivaroxaban The Peak and mVRI were the most sensitive CAT parameters with a high sensitivity (Peak IC50 was 3ng/mL with PPP-Reagent Low and PPP-Reagent and 14ng/mL with PPP-Reagent High; mVRI IC50 was 1ng/mL with PPP-Reagent Low and PPP-Reagent and 3ng/mL with PPP-Reagent High) and both reagent showed a low variability (CV

Description: Abstract 1442 Introduction: Type II heparin-induced thrombocytopenia (HIT) is a serious immune-mediated adverse effect of heparin treatment. The release of procoagulant platelet microparticles (PMPs) is considered as major component of this thrombogenic disease. Early diagnosis of HIT is essential in order to improve clinical outcomes. However, such a diagnosis is challenging. The current diagnostic tests consist of the combination of the clinical scoring system (“4T's rule”) with serological or functional tests such as the gold standard 14C-serotonin release assay (14C-SRA). Serological tests are acceptable for ruling out type II HIT but lack still specificity and are also not standardised. 14C-SRA is time-consuming and is only available in highly specialized laboratories and not in all countries. Aims: The aim of our study is to develop a PMPs generation assay (PMPGA) in whole blood that could be routinely used to diagnose HIT. The ultimate goal is to provide a more specific functional test than the existing serological assays. In addition, this test should be easy to use,faster and as accurate as 14C-SRA. Method: The platelet-poor plasma of HIT-suspected patients (n=71) was first incubated 20 minutes at 37°C with citrated 109 mM whole blood from a healthy donor containing 0, 1 or 500 IU heparin/ml. Those three concentrations aimed at determining a basal PMP concentration, the PMP concentration generated by the HIT antibodies and checking the specificity, respectively. In the second step, PMPs positive or negative for annexin-V FITC (phosphatidylserine (PS)) and anti-CD62P-PE (P-selectin) were quantified on a BDIS FACS Aria® flow cytometer (BD Biosciences).The size of the PMPs was defined using a blend of monodisperse fluorescent beads (Megamix, BioCytex, Marseille, France) of three diameters (0.5, 0.9 and 3 μm) according to a previously described protocol. ROC Curves were performed to determine the optimal cut-offs for MPs concentrations and ratios. 4T's rule and clinical follow-up were registered. Results for ELISA, Light Transmission Aggregometry (LTA), PMPGA, and gold standard14C-Serotonin Release Assay (SRA) were compared by Chi-Square test. Results: During the incubation at low heparin concentration, PMPs expressing PS are generated due to the formation of immune complexes whereas PMPs rate decreases in presence of higher heparin concentration, allowing the definition of positive HIT. The determination of an optimal cut-off for MP concentrations requires standardization of many preanalytical and analytical conditions. However, ratios are much more convenient and the optimal cut-off for the ratio of PMP at low and high heparin concentration was calculated as 2. The relation between PMPGA and SRA was much more pronounced (p

Description: Abstract 4392 Introduction: Patients with cancer have a 7- to 10- fold increased risk of developing venous thromboembolism. Circulating microvesicles (MVs) could be a predictive biomarker for venous thromboembolism in cancer. Thrombin generation assay is a useful technique to detect procoagulant activity of MVs. However, thrombin generation assay suffers from a lack of sensitivity due to the presence of Tissue Factor Pathway Inhibitor (TFPI) in plasma. Aims: To improve the sensitivity of thrombin generation assay to tissue factor (TF) by limiting the interference of TFPI. Methods: Serial dilutions of MDA-MB231 cells were incubated for 45 min at 37°C to generate MVs. Samples were then centrifuged and supernatants which contain MVs were used for thrombin generation assay. Normal pooled plasma was incubated with inhibitor of TFPI or was diluted twice to decrease plasma level of TFPI. Lagtime was used as a surrogate marker of thrombin generation assay to detect procoagulant activity of MVs. Results: i) Inhibition of TFPI decreased twice the cell concentration needed for a significant reduction of lagtime and decreased 2.4-fold the intra-assay variability. ii) Plasma dilution had no impact on the thrombin generation assay sensitivity when thrombin generation assay was triggered by MVs derived from MDA-MB-231. Conclusions: Thrombin generation is a very sensitive method to study the procoagulant activity of TF-MVs. The sensitivity can be increased by inhibition of TFPI with specific monoclonal antibody against its Kunitz Domain I. A twice plasma dilution is an interesting alternative to study the procoagulant activity of MVs by thrombin generation assay with a good sensitivity, especially when low plasma quantities are available. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 2275 Introduction: Apixaban is direct factor-Xa inhibitor that reached the market for the prevention of venous thromboembolism in patients undergoing major orthopaedic surgery. It is also being evaluated in the reduction of recurrent ischemic events when added to antiplatelet therapy after an acute coronary syndrome and in the prevention of stroke in patients with non-valvular atrial fibrillation. Thanks to its predictable pharmacokinetic profile, biological monitoring is not required. Nevertheless, evaluation of plasma drug concentration may be valuable in specific situations such as recurrent thrombosis, bleedings, before urgent surgery, in case of bridging and in case of at least two risk factors among the following ones: drug interactions with caution, moderate renal impairment and moderate hepatic impairment; Monitoring may also be useful in infants, pregnant women or in extreme body weights, although no relevant data on drug levels associated with approximate therapeutic and harmful ranges are currently available. Material and Methods: Apixaban was spiked at increasing concentrations (0, 5, 10, 20, 50, 100, 200 and 500 ng/mL) in pooled citrated normal human platelet poor plasma (PPP) to measure Prothrombin Time (PT) and dilute PT with different thromboplastin, Thrombin Generation Assay (TGA) with different inducers and activity on different anti-Xa chromogenic assays. Activated Partial Thromboplastin Time with different reagents, Thrombin Time (TT), Ecarin Clotting Time (ECT) and Reptilase Time (RT), measurement of fibrinogen (Clauss method and PT-derived method) and antithrombin (anti-IIa and anti-Xa based chromogenic assays) were also tested. We also evaluated the impact of apixaban on assays used for the determination of lupus anticoagulant such as the DRVV-T.. (Screen and Confirm) as well as the PTT-LA.. and the Staclot-LA.. . Results and Discussion: As mentioned in previous studies, PT showed a weak sensitivity towards apixaban in comparison with the plasma range obtained in short pharmacokinetic studies. Indeed, the concentration needed to double the clotting time was 154 ng/mL with the most sensitive reagent while the mean Cmax obtained in a short PK study after one oral intake of 5 mg apixaban (dose given in atrial fibrillation) was 96 ng/mL. Therefore, the sensitivity of PT is not strong enough to allow accurate quantitative measurement of the plasma drug concentration (Table 1). Activated Partial Thromboplastin Time presented a better sensitivity but showed a plateau after 100 ng/mL reflecting the uselessness of this test for the quantification of apixaban. Thrombin Time, ECT and RT were logically not affected while DRVV-T.. showed a sensitivity of 205 ng/mL (Screen), which is once again not enough sensitive. On the opposite, chromogenic anti-Xa assays seemed to be very sensitive (Figure 2 and Table 1). Nevertheless, the relation was not always linear and some methodologies needed to be adapted to ensure a broader range of application. TGA (Figure 1) may be useful to assess the pharmacodynamics effects of apixaban on the coagulation process. Nevertheless, the turn around time and the lack of standardisation are currently limitations that restrict the use of this method. In the case of the exploration of an haemorrhagic event, specific tests such as RT, fibrinogen (Clauss and PT-derived method (dFib)), TT and clotting factor activity may be used. Apixaban did not interfere with these tests. Antithrombin determination if also of importance and chromogenic anti-IIa based assays should be used in face of patients treated with apixaban to avoid misdiagnosis since an overvaluation of 12% by 100 ng/mL was shown using one chromogenic anti-Xa based assay. Conclusion: PT may not be used as screening test to assess the risk of bleedings. A more specific and sensitive assay such as chromogenic anti-Xa assays using calibrators should be used to correctly assess the concentration of apixaban. Determination of lupus anticoagulant using DRVV-T.. and PTT-LA.. or Staclot LA.. as well as the determination of antithrombin using factor-Xa based chromogenic assays, were influenced by apixaban. Finally, standardization of the time between the last intake of apixaban and the sampling is mandatory. Figures: Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 3332 Introduction: Recently, 2 oral anticoagulants have been marketed: dabigatran, a direct thrombin inhibitor, and rivaroxaban, a direct FXa inhibitor. Thanks to their safety and efficacy profiles, routine coagulation monitoring is generally not required. However, the most commonly reported adverse reactions with these drugs are bleedings. A sensitive laboratory assay might be valuable to measure the pharmacodynamics of these 2 drugs in different situations including risks of drug interaction (i.e. strong P-gp inhibitors), overdose or to measure patient compliance. Due to their different mechanism of action, the measurement of their anticoagulant level in plasma may require different types of clotting assays. Aims: The aim of the present study was to determine which coagulation assay(s) could be used to measure the impact of rivaroxaban and dabigatran among a range of routine or more specific tests. Methods: Both drugs were spiked at increasing concentrations in pooled citrated normal human platelet-poor plasma (PPP). The final concentrations were ranging from 10 nM to 2 μM for dabigatran and from 25 nM to 5 μM for rivaroxaban. Such concentrations covered the plasma range in patients during orthopedic surgery. The following routine clotting assays were performed with the spiked plasma according to the manufacturer's protocols: Prothrombin Time (PT) with Neoplastin CI+ and Neoplastin R on STA-R (Roche Diagnostics), with Innovin on BCS (Siemens) and with Recombiplastin on ACL-TOP (Instrumentation Laboratories); activated Partial Thromboplastin Time (aPTT) with CK-Prest, PTT-A and Cephascreen on STA-R, with Actin FS on BCS and with Synthasil on ACL TOP and Thrombin Time (TT) on STA-R. More specific tests included Prothrombinase-induced Clotting Time (PiCT, Pefakit®, Pentapharm) and Ecarin Clotting Time (ECT) on STA-R; dilute PT (dPT) and activated Clotting time (ACT) on KC10. Thrombin Generation Tests (TGT) were also performed with Calibrated Automated Thrombogram (CAT) using PPP or PPP LOW reagents (Thrombinoscope). Sensitivity of a coagulation assay was defined as the concentration required to double clotting time (2XCT). Results: Dabigatran. A concentration-dependent prolongation of PT, dPT and aPTT was observed. However, at high concentrations, sensitivity of aPTT decreased whereas sensitivity of PT increased. The results varied depending on the clotting reagent used. TT was too sensitive leading to high variability for concentrations higher than 250nM. PiCT showed a linear concentration coagulation time (CT) relationship until a plateau at 750nM. ECT showed a high sensitivity (2X CT ≈ 93 nM), a linear relationship in the whole concentration range and a very low variability (CV≤ 1,8%). ACT gave a similar profile to PT (Innovin) with a slightly higher sensitivity (2X CT ≈ 716 nM vs 842nM). Dabigatran induced a concentration-dependent delay and inhibition of the tissue factor-induced thrombin generation with 5 pM TF or 1 pM TF with 4 μM PL and 16.7 mM CaCl2. The drug strongly increased the lag time and Tmax whereas it slightly decreased the Cmax and ETP. The lag time was the most sensitive CAT parameter with a high sensitivity (2x lag time ≈ 145 nM) and a low variability (CV≤6%). Rivaroxaban. A concentration-dependent prolongation of PT, dPT and aPTT was observed. However, at high concentrations, sensitivity of aPTT increased. Sensitivity of dPT (2X CT ≈ 109 to 550 nM) and PT (2X CT ≈ 138 to 764 nM) were similar and superior to aPTT (2X CT ≈ 897 to 2050 nM). The results varied depending on the clotting reagent used. TT was insensitive to rivaroxaban until 2500nM. Both PiCT and ECT showed a low sensitivity (2X CT ≈ 2536 nM and 5750 nM, respectively). A concentration-dependent prolongation of ACT was observed until 2500nM (2X CT ≈ 2275 nM). Rivaroxaban induced a concentration-dependent reduction and delay of the TF-induced thrombin generation. The Cmax was more strongly decreased than the ETP whereas the Tmax was more prolonged than the lag time, showing a major influence on the amplification phase. The Cmax was the most sensitive CAT parameter with a high sensitivity (Cmax EC50 ≈ 50 nM) and a low variability (CV

Description: Abstract 2041 INTRODUCTION: Hereditary Spherocytosis (HS) is characterized by weakened vertical linkages between the membrane skeleton and the red blood cells (RBC) lipid bilayer, leading to the release of microparticles (MPs). SDS-PAGE and ektacytometry are the two reference tests to establish the diagnosis. However, SDS-PAGE lacks sensitivity to very mild or asymptomatic HS carriers, and a high reticulocyte count might mask a reduction in ankyrin-1 in SDS-PAGE whereas an ankyrin-1 defect represents 40–65% of HS in the USA and Europe. Ektacytometry is only available in specialised laboratories. Consequently, diagnosis of HS is currently based on a combination of clinical and family histories, physical examination and laboratory data, especially RBC indices and sometimes a hypertonic cryohaemolysis test and eosin-5-maleimide binding in flow cytometry (EMA). Besides the release of MPs due to destabilisation of the lipid bilayer, one previous study showed that in HS, but not in AIHA, the surface area loss is already present at the circulating reticulocyte stage. AIMS: The aim of this study is to use the release of MPs at the reticulocyte stage to develop an algorithm on 2 haematological analysers: XE-2100 and XE-5000 (Sysmex) as simple, fast, accurate, sensitive, and specific diagnostic laboratory test for HS. METHODS: HgB, MCV, MCHC, Reticulocytes (Ret), Immature Reticulocytes Fraction (IRF), Hypochromic RBC (Hypo-He), Microcytic RBC (MicroR) and indices were determined using XE-2100 (n= 15) and XE-5000 (n= 30). %HYPO−He and %HYPER−He are specific to the XE−5000 (Figure 1). We first studied the efficiency of discordance between Ret and IRF on the XE-2100 and XE-5000, the MicroR/Hypo-He ratio on XE-5000 and the combination of those 2 rules forming an algorithm, to screen 45 confirmed HS (22 women, 23 men, mean age 13.1, from 1 day to 76 years old). They were classified as severe (n= 6, Hg 12g/dl). We then studied the efficiency of the algorithm to differentiate HS from other haemolytic disorders (n=108), microcytic RBC disorders (n=93), healthy subjects (n=61) and a routine database (n=1230). RESULTS: HS is characterized by a high reticulocyte count (Ret) without an equally elevated Immature Reticulocytes Fraction (IRF) (Figure 1). The algorithm includes a precondition to screen all cases of HS, and a second rule taking into account the degree of anaemia. All 45 HS have reticulocytes 〉 80 G/l and a Ret (G/l)/IRF (%) ratio higher than 7.7. The second rule is different according to HgB level. All trait and mild cases of HS have a Ret/IRF ratio greater than 19. The efficiency of the algorithm to screen trait and mild HS is very interesting, as the diagnosis of trait and mild HS is currently difficult even by SDS-PAGE. The severity of the disease shown by HgB level is due to the intensity of release of MPs, which is reflected at best by MicroR. Thus, for moderate and severe cases, MicroR and MicroR/Hypo-He were combined. The optimal cut-offs for MicroR and MicroR/Hypo-He in cases of HgB between 8 and 12 g/dl, were 3.5% and 2.5 respectively and if HgB was lower than 8 g/dl, were 3.5% and 2.0 respectively. Interestingly, the algorithm was also valid for the 4 neonates and the 4 cases of ABO incompatibility patients. We also showed that this algorithm is much more performant than single parameters (Ret/IRF index) or the existing rules (reticulocytosis, % microcytes, combination of increased Mean Corpuscular Haemoglobin Concentration (MCHC) with increased red distribution width (RDW), combination of increased RDW with hyperdense cells (Hyperchromic cells), combination of increased MCHC with increased hyperdense cells). The Area Under the Curve (AUC), sensitivity, specificity, predictive positive value (PPV) and negative predictive value (NPV) were respectively 0.997, 100%, 99.3%, 75% and 100% for the HS algorithm. Finally, preliminary data on a small group of patients indicate that HS induced the release of significantly more RBC MPs than healthy subjects, paroxysmal nocturn hemoglobinuria and heparin-induced thrombocytopenia. CONCLUSIONS: We have developed a simple and inexpensive algorithm that could be used as an excellent screening device for the diagnosis of HS. This rapid algorithm also works on mild HS, ABO incompatibilities, neonates and overcomes the lack of sensitivity of SDS-PAGE in ankyrin deficiencies. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 4788 Introduction: Thrombosis is a common complication of patients with malignancies. Patients with hematological malignancy have a 28 fold increase risk to develop venous thromboembolism (VTE). A population-based cohort was used to determine the incidence and risk factors associated with development of venous thromboembolism (VTE) among Californians diagnosed with acute leukemia between 1993 to 1999. Principal outcomes were deep vein thrombosis in the lower and upper extremities, pulmonary embolism, and mortality. Among 5,394 cases with acute myelogenous leukemia (AML), the 2-year cumulative incidence of VTE was 281 (5.2%). Sixty-four% of the VTE events occurred within 3 months of AML diagnosis. The induction of hypercoagulable state mechanisms is not fully understood to date. Multifactorial aspects such as physic immobility, chemotherapy adverse effects or the overexpression of several procoagulant substances (cytokines, cysteine protease and tissue factor) by cancer cells are often provoked. Several studies strongly suggest that microvesicles (MVs) harboring tissue factor activity may have a primary role in VTE. MVs are small membrane vesicles shed from normal and/or tumor cells following activation or apoptosis. MVs may present TF and negatively charged phospholipids (PL) such as phosphatidylserine on their membrane. These elements are thought to be implicated in the procoagulant activity (PCA). Objectives: The aim of this study was to assess the capacity of untreated acute promyelocytic leukemia cells to shed procoagulant MVs. Methods: Acute promyelocytic leukemia (APL) cells lines (NB4 and HL-60) were cultured 48h in medium at 600,000 cells/mL. Cells and MVs were separated by filtrations (0.1–0.22–0.45–0.65μm). The PCA was assessed by thrombin generation assay. Alternatively, MVs were incubated with anti-TF antibodies (10μg/mL) or with annexin V (0,5μM to assess the contribution of TF and phospholipids to the PCA. Moreover HL-60 cells were incubated with HgCl2 which promote di-S bond formation (activation of TF). Results: NB4 cells and HL-60 cells can stimulate thrombin generation. HL60 cells reduced the lagtime 3.9-fold and increased the peak 1.6-fold in comparison to CTL and NB4 cells decreased the lagtime 10.9-fold and increased the peak 6.7-fold in comparison to CTL. No PCA was observed in HL-60 filtered with 0.65 μm membrane (no statistical difference in lagtime peak and ETP). By contrast, NB4 cells can support thrombin generation activity when filtered at different sizes. MVs of sizes